Encountering Romanitas: Characterising the Role of Agricultural Communities in Roman Britain

Abstract

There has been a hesitancy in academic discussion of Roman Britain to address the potential significance of the identity and agency of rural communities in shaping the provincial landscape. This article seeks to address the reasons for this before delineating some avenues by which we might better investigate this issue. Through two case studies the importance of kinship, agricultural peers and occupational identity (being farmers) are recognised as potential drivers for the course of rural life in Roman Britain. In so doing the extent to which ‘being Roman’ was really a central concern of many agricultural communities is questioned.

This article explores the role of rural identity and agency and the extent to which these can be applied to the writing of new narratives of agricultural social development in the first centuries of Roman rule in Britain. Arguably, emphasis in Roman archaeology has been placed on the impact of urban and military impositions and related population responses, rather than a consideration of the diversity of the rural landscapes encountered and their role in shaping provincial identities. Thus the first section includes a short critique of some current approaches to Roman-period rural landscapes and communities, and highlights the continuing absence of attention to agency in literature and perspectives considering the mass of the rural populace in the Western Roman Empire. The second part seeks to delineate routes by which we might investigate issues of local agency, and the importance of kinship or peer group in reinforcing the identity of rural communities and how they can contribute to our picture of the emerging character of rural landscapes in the provinces. Two contrasting case studies will help demonstrate the validity of this approach and its potential contribution to wider debates about cultural change in the Roman Empire. The importance of the approach outlined here lies in its willingness to recognise and engage with the diversity apparent in the archaeological evidence from agricultural landscapes and in its opening of new avenues for debate regarding how and why the Western provinces developed in the ways they did.

CURRENT APPROACHES TO IDENTITY

The last decade has seen substantial advances in the understanding of what it meant to be Roman in provinces such as Britain and Gaul. ¹ Since the late 1990s earlier perspectives on Roman identity have been criticised for their normative approach to the complex social interactions that were attendant upon the incorporation of the Western provinces within the Roman Empire. ² It was recognised that these developments, conventionally termed ‘Romanisation’, required us to assume that to be Roman was to be part of a unified experiential whole — something which the literature on discrepant experiences seriously called into question. ³

Several recent studies have done much to recognise the role of agency and broader communal identities in shaping the varied development of the Western Roman provinces and their populations. In Britannia: The Creation of a Roman Province, John Creighton notes that the creation of towns in Britain and their varied trajectories reflected the many different aspirations and interpretations of what it was to be Roman of their ‘stakeholders’. ⁴ In doing so, he stresses how the topography and meaning of these urban places were influenced by the differing balance in each between indigenous and Roman immigrant and military communities. David Mattingly in particular has done much to recognise the diverse identities and means of expression of different communities within the province, in particular emphasising the contextual nature of ‘being Roman’ and the need to recognise discrepant experiences of imperial discourses. ⁵ In a sophisticated account of late Roman military communities Andrew Gardner has focused on the importance of practice in the creation and reworking of identities through time and space. ⁶ Whilst the theoretical emphasis has changed, though, it would be nonetheless true to say that the academic ‘fields of battle’ remain predominantly those long familiar to earlier works, namely the military, ⁷ urbanism, ⁸ and élite political or economic life ⁹ in acting as driving forces for change. Creighton's Britannia, ¹⁰ for example, primarily focused on social contexts for the early development of urban landscapes in southern Britain, largely against the background of élite discourse between indigenous leaders and the Roman state.

RECOGNISING THE VALUE OF RURAL DATA

Whilst the complexity and sheer scale of our evidence for the contemporary rural landscapes of the period are recognised in these works, they still rarely feature in anything other than a passive role, often simply to be described or illustrated as examples of the scale of transformation under Roman rule (or its absence), or seen as an ‘other’ in response to or against which changes in other spheres of provincial life are compared. Indeed, agricultural communities (unless they are part of largely undefined élites) are on the whole denied any effective agency of their own in influencing the emerging identity of society in the Western provinces. Arguably these issues have only increased in the face of the generation of a huge amount of data through development-led archaeology over the last twenty years. This is partly because of our inability to keep pace with this output and to synthesise it, but also because in other regions paradoxically, where the pace of modern development is slower, a continuing emphasis on military and urban archaeology in research projects means that there are still significant gaps in our knowledge. In Britain, France and the Low Countries especially the rural data lie in drifts, so that in parts of the Thames Valley, for example, there are entire blocks of landscape that have been subject to sustained and continuous excavation and survey, whereas much of rural Devon offers very little excavated evidence of Roman period rural settlement. ¹¹ The varied impacts of suburban expansion, road and rail construction, and mineral extraction have equally rapidly created zones within the wider landscapes of northern France and the Netherlands around which new or revised histories of settlement can be contemplated and written. ¹² While the data expand and the potential for more detailed analysis increases nonetheless, access to and publication of these findings are underdeveloped. ¹³ As soon as we aspire to write regional syntheses in which we might wish to compare and contrast patterns of rural settlement in different areas we can see the scale of the problem. ¹⁴

These limitations aside, it is evident that the rural population of Britain, at the time of the Roman conquest, was large and that few habitable landscapes were not settled on some level and often by significant permanent rural communities. Equally, few scholars would now suggest that these rural communities represented less than 80–90 per cent of the population of the subsequent Western Roman provincial landscapes. Individually weak many of these farming communities may have been, but their sheer numbers mean they may have had a greater ability to influence provincial culture than is sometimes imagined. Given this, it can be suggested that their continued low profile in many accounts and debates about the provinces stems from the frequent inability to develop or conceive of models in which the great mass of the rural population might have had an active role in shaping the particular character, identity and consumption patterns of the emerging provincial landscapes. It is important to observe that this problem is not particular to the North-Western provinces: in a recent review of rural surveys in Roman Italy, Rob Witcher noted a similar theoretical stagnation in which interpretation has remained largely text-driven and processual; ¹⁵ focusing on issues such as agriculture, population, environment and economy rather than ideology, perception or community. In both regions there has been a tendency for archaeologists ‘to focus on processes rather than people and land rather than landscape’. ¹⁶

An emphasis on agricultural and craft economy and status remains true also for Roman Britain where, with some notable exceptions, ¹⁷ these issues still feature prominently in many regional and national overviews of the evidence. ¹⁸ In many studies, agricultural evidence in particular is often seen only in economic terms, not as an important means of social expression in itself. ¹⁹ Even where our data are good, the results are frequently constrained within normative interpretations of the evidence. The presence of architectural expressions that recognisably draw upon aspects of the classical architectural lexicon, for example, however transformed in the process, might still be seen unproblematically as evidence of the transformation of a rural landscape into the familiar pattern of hierarchical ‘villa dominated’ landscapes built around largely hypothetical estates. ²⁰ An underlying theme here, therefore, is that simply continuing to amass data only gets us so far. We have to first acknowledge and then be more confident in addressing the potential role of the wider provincial rural populace and its traditions, structures and concerns within debates about the emerging character of the provinces.

With some honourable exceptions, the archaeology of Roman rural landscapes has rather left theoretical approaches to ossify as piles of new data have accumulated. There have been few radically new models for understanding the development of Roman rural landscapes, perhaps because the transformation of Late Iron Age to Roman settlement in southern Britain and northern Gaul is often seen as familiar and fairly secure. ²¹ Exceptions are increasingly apparent as our evidence of, for example, large swathes of northern and western Britain and parts of the Low Countries and northern France improves, ²² demonstrating very different traditions of settlement as well as societies still seemingly more strongly associated with indigenous Iron Age beliefs, practices and structures.

DIVERSITY AND IDENTITIES IN A PROVINCIAL CONTEXT

In a series of important contributions Mattingly has signalled the need to recognise the multivocality of discourses that took place within an imperial system like Rome's. ²³ In particular, he cogently points out that in Britain under the Roman Empire ‘what was being constructed was not simply different variants of Roman identity, but also different ideas of what it meant to be non-Roman’. ²⁴ Whilst diversity, discrepancy and even the absence of material evidence are critically discussed, however, the debate is still usually framed within the concept of ‘becoming’ or ‘being Roman’. Consequently, overviews of the evidence from rural landscapes focus on things Roman and on the introduction and operation of imperial power. Gardner too has noted how most critical attention has been directed on to the categories of ‘Roman’ and ‘native’ rather than other features of identity, such as gender, age or occupation. ²⁵ To this category, a continuing focus on hierarchical status can be added. ²⁶ Thus, in one recent discussion of rural communities in Roman Britain we see variability in the uptake and use of epigraphy, mosaics, wall-painting, statuary, silver plate and stone architecture in the ‘fashioning of élite behaviour’, ²⁷ and the use of pottery, coinage and even toilet instruments in ‘non-élite behaviour’. Whilst not wishing to deny the fundamental transformations attendant on the adoption of this material or to deny that some (probably many) people in parts of the Western provinces shared aspirations and ideas about what constituted Romanitas, it is important to highlight that those that did not are still written out of the story here and in other Western provinces of the Roman Empire. There is a tendency to see diversity in the evidence in terms of ‘discrepancy’ (itself a normative concept) that is interpreted as a rather inert ‘resistance’, ²⁸ or ‘cultural conservatism’, ²⁹ rather than having a social logic of its own that was itself dynamic. This pacification of agency in imperial or colonial contexts is not unfamiliar to anthropologists, and in other contexts such a process might be seen as a dimension of hegemony. ³⁰ In Egypt Nick Thomas, ³¹ for example, has commented on how even where the colonised express resistance through an adherence to old or traditional forms of architecture, the new terms in which such resistance is expressed are themselves part of a ‘transformative endeavour’. Resistance or conservatism in this context is not inert, but actively involved in the creation of new and different identities for the colonised.

Related to this conceptual bias is the tendency in provincial studies to over-emphasise the effects of Roman culture and structures and under-emphasise or simply ignore the impact of indigenous societies on the development of Roman provinces. Peter Wells has stressed that in part this stems from our academic grounding in which most students of the period have been trained to look for material evidence of the former and not the latter. ³² Thus, in many accounts of the rural landscape we encounter much discussion of the adoption of particular forms of stone-built architecture and their decorative elaboration or the complex and highly visible material culture of food consumption and display. Rarely do we hear anything of the frequently more numerous and widespread examples of timber architecture or the presence of unfamiliar traditions of human and animal burial. ³³ The predilection of many Roman archaeologists for normative and hierarchical interpretations of the evidence is understandable — if regrettable — given the presence of forms of material culture (e.g. coinage, silver plate and domestic architecture) to which we continue to ascribe such status today. Many highly visible forms of consumption and display did occur in rural contexts in the Western Roman provinces, but one of the challenges for archaeology is to be wary of necessarily ascribing them to a single familiar cause. Equally, by focusing overly on plotting aspects of novel material consumption familiar from urban contexts (e.g. bathing, dining, domestic architectural elaboration), we are tending to ignore those spheres of material and working life that may have been crucial to agricultural communities themselves. In effect we may be exporting our own urban, metropolitan biases about what is important to a study of rural societies in a Roman context. ³⁴

There is scope to comprehend rural settings better through greater consideration of other sources and approaches to the subject. Thus, for example, some recent work in anthropology argues that ideas of kinship, identity and community in rural societies derive from the ways in which a group organises itself in the process of agriculture, craft or exchange; in other words the business of everyday life. ³⁵ A practice-based approach to identity and community group should be attractive to archaeologists who might add that patterns of everyday life are equally channelled, structured and influenced by physical circumstances such as buildings, field-systems, trackways, and tools. ³⁶ An important observation of the anthropological work is that for many agricultural communities, agricultural practice itself and the continuing maintenance of aspects of the rural landscape, such as field boundaries or traditional routes of access to grazing, that structure relations with neighbouring communities and the wider world are crucial to their sense of identity. Gerd Baumann, ³⁷ for example, goes so far as to point out that among the Miri of Sudan ‘a person who does not farm may be liked … but cannot be a villager’ and thus cannot take part in important decision-making processes in the community.

Equally, agricultural practice (how you farm) and land use (the maintenance or renegotiation of rights to cultivate or to rear livestock) are very important at an intra- and inter-communal level for agricultural communities. These concerns often operate at the scale of one's immediate peers or the wider agricultural community rather than in relation to the concerns and interests of urban communities or political authority. How this might have been expressed in the often very diverse rural landscapes of the Roman provinces is potentially a very important avenue of study. We should not be surprised at the idea that within farming communities both individual and group identities, and prestige could be intimately tied into the ownership and well-being of land or livestock. John Gray, ³⁸ for example, in a telling analysis of sheep farmers in the Borders of Scotland has shown that the obvious economic drivers behind their work are embedded within a complex web of social relations with their peers that are just as important to them. Much effort, for example, goes into the presentation of rams at annual auction, the farmers readily admitting that the aesthetics of the animals is as important as their economic value because their own social standing and identity is represented and enhanced by the rams. ³⁹ At particular issue here is their social status in relation to their own peer group of neighbouring sheep farmers rather than concerns of how they appear to the wider world. This internalised social discourse can be seen to affect how they farm (and indeed what is sometimes perceived to be their conservatism in relation to those from outside this community), but also the ways in which they choose to reinvest their wealth. This dimension to social discourse is too often missing from accounts of rural communities in the Roman period despite the fact that they might have had profound consequences for the development of the agricultural landscape under Roman rule.

Much recent work in anthropological contexts has also reminded us of the importance of the house in both its material and ideological sense in the creation and maintenance of particular discourses about identity in agricultural communities. ⁴⁰ Prominent in a number of these studies is the role houses play in mediating relationships between kin and the wider community through reception. Thus Janet Carsten, ⁴¹ for example, argues that it is the qualitative density of experiences within a house such as domestic labour, child rearing and the economy of the household that makes kinship, and ensures its importance in binding its inhabitants together. Furthermore, she suggests that the location and demarcation of where these often routine acts take place reproduce important values and meanings. Thus ‘… apparently neutral and insignificant acts such as washing clothes or eating a meal, and how apparently random placings within a house — where different household articles are stored, or who sits where at the table — are not just imbued with social meaning, they are crucially involved in the reproduction of meaning.’ ⁴² These rules learnt about both appropriate behaviour and status are then taken beyond the house through the reception of others within the house or in visits to the houses of others alongside the day to day occupational activities carried out in the surrounding landscape. In this way houses act as important points of mediation between the identity of the resident household and the wider world. Such practices can be very important in defining communal identity as in the case of the Nuer in Southern Sudan where Gudeman and Rivera have shown how communal identity is defined in terms of locality and shared household residence and not as used to be thought through descent or ethnicity. ⁴³

These examples highlight the importance of context and practice in the expression and reworking of rural identity. By looking at practice in rural dwellings and settlements as a whole, where particular emphasis may have been placed on kinship or occupational identity (rather than ethnicity or cultural affiliation), a very different dimension to provincial life may be revealed than that seen when the same people visited urban centres to fulfil wider economic, social or religious roles. Similar important discussions of identity in relation to ethnicity and the Roman world by Siân Jones ⁴⁴ and Simon James, ⁴⁵ and subsequently extended to a broader consideration of identity by Gardner ⁴⁶ have all stressed how individuals are typically able to have multiple identities or none at all in different contexts of social interaction. People in the Western Roman provinces are, therefore, likely to have had a number of ‘nested affiliations’ ⁴⁷ dependent upon circumstances. ⁴⁸ This more pluralistic approach allows us the opportunity to explore the role different human social identities played in a range of situational contexts (places or events) and at a series of scales, from the individual farmstead or community to broad regions. One such under-explored aspect of identity in the Roman provinces is occupational role. As Gardner has suggested occupational role is an important dimension of identity firmly based upon habitual activities. In a military context both James ⁴⁹ and Gardner ⁵⁰ have stressed that discipline, especially in relation to the body, alongside skills specialisms was a major aspect of the reinforcement of such identities, but there is no reason to think that similar concerns did not influence other communities.

To illustrate how these points might work in practice, two short case studies drawn from Roman Britain are offered. The first will show that even in those rural landscapes where we see the development of villa architecture, the reception of Romanitas and the creation of provincial Roman cultures are more complex and varied than is assumed; in addition we will see that rural communities could and did impact upon the shape of the resultant reworked provincial landscape. In the second case study away from these more ‘familiar’ regions, it will be suggested that we are only starting to get to grips with the potentially complex relationships between an often large rural populace and the wider Roman world with its urban and military communities and the complex social landscapes they created in the process.

CHALLENGING THE FAMILIAR: THE RURAL LANDSCAPES OF THE EAST MIDLANDS

The landscapes of the East Midlands covered by the modern counties of Leicestershire, Lincolnshire and Northamptonshire in some respects present a not unfamiliar picture of rural life in the Western Roman provinces. A region already undergoing significant social change in the Late Iron Age, ⁵¹ it experienced a comparatively short period of direct military intervention before the establishment of a colony in its north-east and the development of a civitas-capital in its south. But there was no immediate and wholesale restructuring of the rural landscape (at least outside the floodplains of the major rivers and some of the former wetlands of the Fens); rather the archaeology indicates a piecemeal process over a period from the mid-first to mid-second century a.d. that appears to re-gear much of the rural landscape partly to enhance the storage, processing and mobilisation of agricultural produce. ⁵² Thus, by the middle of the second century a.d. we see the development of a significant and growing network of modest villas with evidence for investment in corn-driers, barns and even mills to process and transport agricultural surpluses to the towns of the region and probably beyond. What stands out though is that these emerging villas lay within landscapes inhabited by a far larger number of more materially modest settlements, ⁵³ some single farms, others rather larger agglomerations, but all of which came to use a wide range of material culture associated with the Roman world.

Such a simple regional summary of course hides more complex patterns to be drawn from the excavated and survey evidence in the foundation, restructuring and abandonment of rural settlements that reflect different local histories of engagement with the Roman world and its attendant impact on the organisation of agricultural life in the East Midlands. Within this, what we might term ‘familiar’ landscape, however, we need to observe some aspects of the archaeology that suggest a rather more complex situation. A traditional focus on rather superficial regional and national maps of villas naturally conceals significant regional and local variations in architectural expression. One such example is the aisled hall, a tradition of architecture found widely on rural settlements in Britain from the Humber Estuary to the Isle of Wight, and thought to have begun in the first century a.d., but which becomes more widespread predominantly from the early second century a.d. Intriguingly, aisled halls were a form of architecture that was both a new departure from pre-existing LPRIA indigenous traditions and simultaneously distinct from more recognisably classically-inspired traditions, such as the strip-houses or winged-corridor buildings which were often their contemporaries. The aisled hall was an architectural form that could be adapted to a wide range of purposes and many are likely to have been flexible multi-purpose buildings. Straightforward equations of architectural type with function or status rarely work here, as comparable studies carried out in the East Midlands ⁵⁴ and Hampshire ⁵⁵ have shown. Critical to understanding better the role of these buildings in rural Roman Britain is to look at patterns of usage through time and across space.

For much of their lives these halls were largely open timber-built spaces that incorporated both living space and craft and agricultural processing facilities. However, unlike many superficially similar houses in the Lower Rhineland, there is little evidence that they ever incorporated byres. Closer inspection of a sample of 22 excavated halls in the East Midlands ⁵⁶ suggests that the location of many of these activities was carefully structured: hearths and evidence for cooking and eating were largely located centrally within one end of the hall (usually to the west), whilst the other (usually eastern) end was used for agricultural processing activities such as corn-drying or malting or craft activities such as iron smithing. The central part of the hall was frequently left clear and may have been used as a place of assembly and/or as a working floor (for threshing or malting) and storage area. During the third and fourth centuries a.d. many of these halls saw architectural elaboration or transformation into stone, reflecting the importance of the halls as a focus of material consumption and display over a long period of time. The particular enhancements (e.g. mortared stone, tessellated paving, and hypocausts) draw upon a wider classical architectural lexicon that, in some cases, introduces new modes of social behaviours such as bathing. Where this is done, however, it is still constrained within the existing spatial framework of activities across the aisled hall. The creation of separate rooms through the insertion of partition walls, the addition of small baths and the use of tessellated paving are all limited to the same ‘domestic’ quarter of the former open hall. The central hall space is still maintained and is conspicuously not embellished. The location of known entrances into these dwellings shows that these halls would still have been the first point of reception within the building. The hall part of these houses also still shows a focus on agricultural and craft work, as well as possibly being used for the display of stored agricultural surpluses.

Strikingly, in some notable cases, such as at Mansfield Woodhouse (Notts.) ⁵⁷ and Drayton (Leics.), ⁵⁸ the external façade of the hall is altered to create the appearance of a portico or pavilions. In doing so, at least superficially these buildings looked more like the winged-corridor houses that formed the core of many third- and fourth-century villas in Britain, but they contained a very different interior. Traditionally such transformations have been seen as forms of emulation of Roman norms, but it can instead be suggested that in their later form we see precisely the kind of plurality of identity discussed above: the hall's exterior face presented a ‘familiar’ form of prosperous architectural expression to the outside world, whilst the interior maintained important traditional spatial distinctions in the operation of the house. Thus established concerns of agricultural and craft practice within the house were retained, as were ideas about the ‘proper’ place to do this. In the grander scale of imperial villas these houses were fairly modest, but in a British context they are often no less impressive than contemporary corridor houses (the East Midlands halls range from 16 to 48 m long). They are not the residences of the poorest echelons of rural society, but can be viewed as important local foci within seemingly prosperous agricultural landscapes. This hybrid development, in which a central open reception and working space is combined with a gradually segregated and architecturally elaborated dwelling under one roof, demands that we recognise very different readings of what it was to be Roman in the rural landscapes of this part of the province.

VENTURING INTO THE UNFAMILIAR: THE WROXETER HINTERLANDS

What of those parts of the North-Western provinces which offer much less evidence for material transformation under Roman rule? Do Roman archaeologists use or even recognise the evidence from these rural landscapes adequately? These are what might be called the ‘unfamiliar’ landscapes of the Roman West. Comparatively materially impoverished (at least to our eyes) and often under-studied they may be, but there is nevertheless much to be gained from modelling the forms of social and economic expression used by the vast majority of inhabitants of these often large parts of the North-Western provinces.

A useful example to illustrate this problem relates to the area of modern Shropshire and the Welsh Marches and to the territory traditionally ascribed to the Cornovii and its principal town Wroxeter. Recent publication of the Wroxeter Hinterland Project ⁵⁹ and earlier excavations of the town ⁶⁰ provide the opportunity to contrast the urban and rural social contexts of this region. Wroxeter itself was a substantial and significant town whose forms and levels of architectural display, range of domestic occupation and patterns of material consumption are akin to other major towns in Roman Britain. ⁶¹ Similarly in terms of ceramics, published excavation reports for the town show high quantities of coarse and fine wares and of imported and locally produced wares similar to levels seen in other major towns. ⁶² The same is true for the use and consumption of glass, coinage, and metalwork objects of personal adornment. ⁶³

A striking contrast, however, is seen almost as soon as we step outside the walls of Wroxeter and look at the archaeology of the rural communities that lived in its wider region. In material terms one of the simplest places to start is through evidence for ceramic production and consumption. Few scholars would argue that pottery was ever a particularly luxurious or expensive product in the Roman world; its ubiquity, repertoire and availability seemingly supported by the excavations within Wroxeter. Its absence, therefore, is harder to argue on economic rather than cultural grounds. Data from the Wroxeter Hinterlands survey, however, show a striking contrast between the town and its own rural landscape. Within 1–5 km from the town walls the quantities of Roman ceramics recovered from probable settlements drop rapidly. Venture beyond 5 km from the city and known rural sites of the Roman period are almost aceramic. Indeed, information from excavations in and around Wroxeter that were part of the project largely confirms this unfamiliar impression of minimal engagement in the consumption of ceramic vessels on rural settlements ( fig. 1). As Vince Gaffney and Roger White note pottery use should be seen as the exception rather than the rule in Cornovian territory. ⁶⁴

FIG. 1. The density of recorded pottery by sherd count per 100 m² of excavated area from Roman settlement excavations in Shropshire and Leicestershire. ‘City’ refers to the combined density figure from several excavations in Wroxeter, Shropshire and Leicester.

Furthermore, where it does occur, the pottery is overwhelmingly sourced from rural Severn Valley kilns, not from those urban industries in Wroxeter, or through regional or provincial imports. In other words, whilst there is abundant evidence that pottery was both made at and imported into Wroxeter on a large scale, very little of it was exported into its hinterland. Jerry Evans assessing the so-called potters’ shops in towns across Britain has recognised their focus on fine wares, mortaria and Black Burnished Ware 1. ⁶⁵ With the exception of Black Burnished wares these products are largely absent from Wroxeter's rural sites as are the coarse wares produced in the city's immediate vicinity. Instead, rural consumption is both limited in scale and focused on just a few sources, notably the Severn Valley wares. If we consider the forms in which these different wares were made, it is possible to suggest that these patterns were linked to an absence of demand for the complex material paraphernalia of Roman-style dining in these rural communities. Whilst cups, flagons, bowls and platters were a common part of the urban repertoire of Wroxeter, jars and cooking-pots dominate the much more restricted rural suite and point to a quite different form of household expression. But there are good reasons to think that this material disjuncture between urban and rural pottery use is not unique to this part of Roman Britain. Initial comparison of pottery assemblages from other excavated sites, such as Exeter and York, ⁶⁶ similarly reveals urban and rural coarse ware supplies that appear largely mutually exclusive until the third century a.d.

Examining other aspects of the material correlates of economic and cultural life we see further striking contrasts between the town and the countryside. In a region without indigenous coin production — or even coin use — in the Late Iron Age it is perhaps unsurprising that early coin loss is largely limited to military installations or settlements closely associated with them. ⁶⁷ It is the army that introduces coinage to the region and its supply that determines its presence in the early years. Nevertheless, a contextual analysis of site losses from excavations and a qualitative assessment of coins reported via the Portable Antiquities Scheme suggest that coin loss levels on rural sites are generally very low in the Wroxeter region compared with regions to the south and east ( fig. 2). This disparity becomes particularly apparent in the late third and fourth centuries, at a time when in many regions to the east of Shropshire we see significant volumes of coinage being recorded on rural settlements, while coin loss levels around Wroxeter remain very low. Excavation of an area of 1,870 m² at the late Roman villa at Whitley Grange, for example, produced only seven later third- to fourth-century coins (a density of 0.37 per 100 m²). ⁶⁸ A meagre number when compared with the figures for the same period from typical sites in the East Midlands for example, such as the villa at Great Casterton (210 coins from an area of 2,600 m², 8.1 per 100 m²) ⁶⁹ or the small aisled villa at Empingham 2 (30 coins from 375 m², 8 per 100 m²), ⁷⁰ or even its neighbouring farm at Empingham 1 (30 coins from 3,600 m², 0.83 per 100 m²). ⁷¹ Whatever the detailed interpretation of this evidence (a subject that cannot be dealt with adequately here), it would seem that the circulation of coin in rural contexts in this region was limited and may simply confirm that the rural population was tied into the system of wider Roman taxation. Accordingly, the scarcity of later third- to fourth-century coins, in particular, may suggest the socially restricted use of coinage and indicate that rural trade continued to operate largely through barter or other socially embedded means of exchange.

FIG. 2. The number of brooches as a proportion of all recorded Roman metalwork from three Local Authority areas recorded on the Portable Antiquities Scheme database up to June 2011. Whilst the total for Shropshire is low (probably a reflection of modern patterns of land use and the level of metal-detecting), the proportion of all metalwork represented by coins is noticeably low compared with two nearby regions to the south and east.

That the reception and use of new material cultures in the countryside were not only restricted but also selective, however, is seen in the rather different patterns of the adoption and use of dress items, especially brooches. Brooches are a surprisingly common part of otherwise sometimes materially impoverished excavated assemblages on rural sites in the region, ⁷² and dolphin and enamelled bow brooches are frequently found in rural contexts south of the Mersey. It has even been argued by Rob Philpott that one brooch form with a marked rural distribution, namely the enamelled Wirral type, was manufactured on rural sites. ⁷³ More extensive recent data from the Portable Antiquities Scheme show that brooches form a very high proportion of all recorded Roman metalwork in Shropshire and Staffordshire ( fig. 3) and may support the idea that personal dress and display were of particular social significance to rural communities in the area in the Roman period.

FIG. 3. The proportion of brooches to all Roman metalwork from Local Authority areas in the West Midlands and eastern Wales as recorded on the Portable Antiquities Scheme database up to June 2010. The size of each roundel reflects total size of the recorded assemblage. A = Shropshire, B = Staffordshire.

If common material correlates such as pottery and coinage are comparatively scarce in rural contexts, we seem to have a greater problem in addressing the possibility that other spheres of material or social life were key to Cornovian communities both before and after the Roman conquest. Cattle are often cited as a principal product of this region on the grounds that the soils and environment are very well suited to a pastoral economy and pollen evidence indeed suggests that cleared extensive grasslands become fairly widespread in this area in the Late Iron Age. ⁷⁴ Here, archaeological data are in short supply owing to taphonomic problems of poor bone survival, but what little rural archaeozoological evidence there is suggests an emphasis on cattle and in Wroxeter (where conditions of preservation are better) the consumption of cattle and their probable processing for hides and other secondary products is clearly seen. ⁷⁵ Interestingly, however, there is little sign that this demand for cattle led to any ‘improvement’ in the animals themselves: Hammon's morphological analysis of the bones suggests no increase in the size of cattle in the Roman period, an apparent contrast with the pattern further south and east in the province around towns such as Colchester. ⁷⁶ This could indicate an absence of investment in improving rural agricultural practice, something certainly supported by other aspects of the archaeological record from rural sites in the region where there is precious little evidence for the construction of new facilities for the increased processing and storage of agricultural products in the Roman period, such as corn-driers, barns or mills. There may, however, have been no need for the development of bigger breeds of cattle if their use as draught animals for cultivation or large-scale transport was not locally important. The continued reliance on smaller breeds may equally have been a cultural preference, indeed in modern historical agronomies small breeds are often considered hardier, less susceptible to infection and requiring less management.

A second potentially important facet of indigenous economic life in the wider region was the exploitation and exchange of salt. An important resource for consumption, the processing of leather and textiles and for the preservation of meat and dairy products (such as cheese) the region incorporated three of four inland brine sources in Roman Britain. There is extensive evidence for trade in this commodity in the LPRIA through surviving briquetage, not just regionally but throughout eastern Wales and along the Welsh coast line. ⁷⁷ The Roman conquest appears to see the sequestration or at least direct military supervision of these sites until the later second century a.d. when civilian contractors probably took up the reins. ⁷⁸ Such acts of appropriation probably had a significant impact in the early Roman period on one key resource used in social differentiation among Cornovian communities, contributing to a distrust or wariness of the perceived benefits of engaging with agents of the wider Roman state that can only have contributed to greater social insularity.

When we look at the architecture of the rural settlements, villas are a rare and generally late development in the Wroxeter hinterland zone. Instead, small enclosed farms of less than 1 ha using traditional earth and timber architectures appear the dominant form. ⁷⁹ Part of a trend beginning in the Middle Iron Age and continuing into at least the second century a.d., this is a common feature of many of the Late Iron Age and Early Roman rural landscapes of western and northern Britain. ⁸⁰ The Shropshire settlements are often small and frequently have only a single dwelling, but many are endowed with substantial enclosing earthworks. Such conspicuous enclosure has been seen as symbolising the relative independence of groups living in such places. ⁸¹ Evidence for the monumentalisation of enclosure boundaries through acts of redigging and multivallation are thought to be as much about display as defence or boundary definition, and involved the use of considerable labour. The scale of these earthworks may even suggest that they were intended to convey permanence and longevity to contemporary observers.

Andy Wigley has suggested that enclosure ditches on settlements in Shropshire were dug via gang work and that their construction and periodic reworking were communal projects. ⁸² Even at relatively inconspicuous sites like the rectilinear enclosure at Bromfield, Stan Stanford has suggested clear evidence for gang work by as many as eleven separate groups. ⁸³ At multivallate enclosures such as Collfryn digging the boundary ditches would have entailed an effort ‘disproportionate to the resources of manpower available within the settlement’. ⁸⁴ Whilst encountering a landscape of ostensibly independent farming households, this evidence may point to a degree of communal interdependence in their construction and maintenance. In the Iron Age, Niall Sharples has suggested that among these overwhelmingly agricultural communities labour itself was seen as a gift, probably in exchange for agricultural produce or opportunities for feasting. ⁸⁵ The degree of success (and thus status) of individual households, therefore, may be reflected in the level of labour implied in long, sometimes monumental and multiple phases of boundary aggrandisement around otherwise small settlement sites. The deliberate location of some of these settlements on major Earlier Iron Age land boundaries, as seen at Sharpstones Hill, ⁸⁶ may mean that their construction was also a reaffirmation of claims to land access — always very important in the sometimes land hungry practices of a pastoral economy. ⁸⁷ The maintenance of earthwork enclosures appears fundamental to many rural settlements until the later second century a.d., but even in the late Roman period we may see the potential significance of these long-lived acts in the way that most of the small number of villas in the region were located within or on top of such enclosures (as at Ashford Carbonell, Berwick Alkmund Park, and Acton Trussell) when walls eventually replaced ditches and banks in enclosing settlements in the third and fourth centuries a.d.

What then are we to make of this rather unfamiliar pattern of evidence for the rural landscapes in the Wroxeter region? In the past the absence of many of the more readily identifiable material indices of ‘Romanisation’ might have been interpreted as a testament to the poverty of rural Cornovians or else an index of their resistance to Roman hegemony. ⁸⁸ Instead, we might perhaps think of it as a world in which for much of the time the region witnessed separate yet parallel social lives within the urban communities of Wroxeter and its immediate hinterland on the one hand, and among the majority of the rural populace on the other. To many in this latter agricultural society, the control of communal labour, livestock, and the resources for their processing and preservation, as well as possibly access to the appropriate means for personal adornment, may have been the traditional means of stressing identity and social status. In this context many of the forms of material consumption associated with say dining, bathing and the elaborate forms of domestic architectural display that we commonly associate with ‘being Roman’ were simply not necessary or relevant to the majority of the agricultural populace, at least in the first few generations after the Roman conquest. The comparative material poverty of these communities — at least in the sense of visible disposable wealth — may well have been a factor in the lack of uptake of some of the more elaborate means of display seen in the wider Roman world. Equally important, however, may have been the way in which rural identity and status were not rooted in money (transitory symbols of wealth), but largely tied to an ability to command or at least benefit from investment in periodic communal acts of labour. In such a situation prosperity would not be apparent through individual acts of material consumption or display but communal ones. Indeed identity, to the majority of the farming population, was defined more collectively and can be said to have been inscribed in the agricultural architecture of the landscape itself — namely its fields, enclosures and settlements — and in the ownership and display of its livestock.

DISCUSSION

Pulling this disparate array of evidence together is of course problematic, but combined it suggests a rather different perspective on ‘being Roman’ (or not) in the rural landscapes of the North-Western provinces of the Empire. In the East Midlands case study, a region which saw much change towards the greater mobilisation of agricultural resources and rural industry, the development of local markets and the adoption of many shared ritual and burial practices between rural and urban contexts, we also nevertheless saw the development of distinctive domestic architecture on many rural sites that emphasised places of communal reception and working in central halls. In this context the public face of many of these houses may have stressed understandings of Romanitas that would have been recognisable to members of Roman urban and élite society. In their internal working though, they stressed rather different expressions of identity that may have stressed the importance of ties of kinship, agricultural occupation or locality over those of ‘being Roman’. In the Shropshire case we saw a rather different landscape in which many rural communities may have looked inward to kin and agricultural peers for their sense of identity, only engaging with the demands and ideologies of the wider Roman province on those occasions when necessity demanded. The means by which social relations and identity were formed and maintained were through traditions of agricultural practice, labour and dress rather than modes of practice and material consumption more familiar in other Roman contexts. If so it is questionable whether some of these communities ever thought of themselves as ‘becoming’ or ‘being Roman’.

The two case studies sketched above could and should be supplemented, complemented and contrasted with others areas in the Western provinces. Nonetheless, hopefully the areas explored go some way to illustrating how a greater awareness of the interests and enduring agency of much of the rural populace could influence the character and development of Western provincial landscapes. The implications of these choices are not just cosmetic. Insularity or a lack of engagement in the economic and social life of the wider Roman world by a significant minority or even majority of the rural population of a region would have profound implications on, for example, a lack of development of the monetary economy, the relative social and economic isolation of urban communities from their own rural hinterlands, ⁸⁹ or the development of distinctive local understandings of what it meant to be Roman that created different traditions of domestic architecture or ritual and religious practice. It is perhaps of no small significance to reflect on the fact that the Empire could accommodate such a response, seemingly unproblematically.

Exploring such issues allows us to readdress the diversity apparent in the rural landscapes of many parts of the Western provinces and how variations in agricultural communities’ traditions and their participation in different facets or means of material expression influenced their development. ⁹⁰ In doing so we should be willing to investigate means of material and social expression that are sometimes rather unfamiliar to Roman scholars. We should not be surprised that among agricultural communities agricultural concerns, namely the supply of labour and control of agricultural wealth or the physical reaffirmation of traditional rights to land, might feature prominently in their social lives and sense of identity. We also need to pay more attention to conceptual issues and be prepared to question habitual assumptions about the nature of Roman rural society. Even when we encounter ‘familiar’ Roman material culture, we should remember that the context of its adoption and use may have transformed its meaning. Steve Willis for example has shown that what constitutes ‘normal patterns’ of samian use in Roman Britain cannot be applied to the south-west peninsula and north Wales which have their own distinctive traditions of consumption and use that are qualitatively different. ⁹¹ All too often archaeology has treated the transformation of Late Iron Age rural landscapes into Roman provincial ones as unproblematic (almost cast as a natural reified unfolding), ⁹² but it is increasingly clear that culturally and ideologically many in indigenous Late Iron Age societies behaved very differently from those in the Roman world; to them the transition to landscapes of villas and towns was neither necessarily obvious nor easy to make.