An Imperial Possession, page 60
The second approach is to examine the contextual data for the distribution of specific artefact classes across Britain in terms of broad categories of sites: military, major towns, other towns, villas, other rural settlements, and so on. A pioneer analysis of coins from 140 British sites, spread across the range of types, revealed broad differences in coin-use profiles between towns and rural sites and between east and west Britain. One interesting point to emerge is that the suburbs of major towns have coin profiles that are more similar to the countryside as a whole than to the pattern within the walled areas of these centres.
An analysis of the distribution of oil lamps has shown an overwhelming predominance in military contexts and at the two major centres of early Roman government, Colchester and London. Rural sites (including villas) represent a tiny fraction of the distribution. Olive oil was available in the civil zone in the first century (when most lamps were imported), but evidently it was not much used for lighting by native Britons. On the other hand, analysis of a range of small metal toilet implements, often produced in sets of ‘tweezers, probes and nail cleaners’, reveals a very different pattern. These items were introduced into Britain from the Roman empire in the late Iron Age and remained popular personal grooming items throughout the Roman period, though they became much more rare on the Continent. As such, they represent an insular cultural trait. They are rare at military sites and major towns (doubly so when one considers the volume of excavation at those categories of site), but are abundant at small towns, villas and other rural sites (cumulatively 66 per cent of all examples). Villas and rural sites together represent 38 per cent of all sites on which they have been located, with small towns another 32 per cent. These were status indicators in the late Iron Age, but appear to have achieved a broader distribution in the Roman period, plausibly representing the long-term maintenance of indigenous traditions about personal grooming.
The rest of this chapter will concentrate on three particularly important areas where alternative rural identities manifested themselves: diet, funerary practice and religion.
Consumption and diet
Food is a basic requirement of life, but it also reflects social beliefs within society. The differences implicit in the truism ‘we are what we eat’ should be demonstrable in archaeological terms. Basic variability in foodways ought to link with notions of identity. In many societies, diet is a conservative area of experience, linked to established subsistence economies and to a tendency for cooking traditions within rural households to be passed down from generation to generation. Major departures from traditional foodways thus require explanation. The elite in any society may occupy a distinctive niche in that they often employ cooks from outside their close agnate group, and these servants or slaves bring not only independent ideas to bear but are also less bound by tradition. In a colonial contact situation, there is a high potential for experimentation with food, both to denote status and to define new forms of social behaviour. Conversely, passive resistance to social change can also find an outlet in the maintenance of traditional foodways as a means of reinforcing a sense of identity.
In Britain under Roman rule there were certainly major innovations in diet and in the way that food was prepared and served. The army and the urban and rural elites were the people at the forefront of new practices and a distinctive material culture of food consumption. Imported commodities such as wine, olive oil and fish sauce are proxies for broader experimentation with radically changed styles of cuisine and taste. The late Iron Age aristocracies of the eastern and southern kingdoms had already experimented extensively in this area. Underneath the palatial villa of Fishbourne there was a high-status late Iron Age site, from which an assemblage of pottery and food remains closely correlates with Roman styles of feasting. This process was taken further under Roman rule at some high-status sites as part of the elite demonstration of knowledge and understanding of Continental Roman culture. As we shall see, the process was neither uniform nor comprehensive. The occasional use of olive oil is something rather different from a permanent adoption of oil as the edible fat of choice.
Unsurprisingly, military sites and the major towns record the widest range of markers of changed dietary habits: higher ratios of cattle and pig bones and lower numbers of sheep and goats, the presence of numerous domestic fowl, a rise in consumption of fish and shellfish even with distance from major rivers and the sea, the occurrence of wine, oil and fish-sauce amphoras, botanical remains of ‘luxury’ and imported foodstuffs, and large numbers of specialized vessel forms in pottery and metal for the preparation, cooking and serving of food. Rural sites show less dramatic change in diet, though there appears to be both social and regional differentiation in the data. For instance, from non-villa sites, rural pottery assemblages tended to be dominated by jars used in cookery, with relatively smaller numbers of specialized vessels for food preparation and serving. Mortaria, flagons and amphoras – vessels that specifically link to foreign notions of food preparation and tastes – were notably rare, as were samian or other fineware serving dishes. Drinking vessels were generally larger-size beakers (suitable for beer) rather than the smaller cups normally used for wine. Animal-bone assemblages from rural settlements were dominated by sheep/goat, partly because these were the less sought-after animals in urban and military markets and thus more likely to be consumed in the country, partly because of an early development of wool production. Broadly speaking, there is a divide in British archaeology between the more ‘Roman’ sites where cattle and pigs predominate and those, primarily rural and ‘native’ in orientation, where sheep were the mainstay. Nonetheless, some rural sites show evidence of a shift towards beef production or pig-rearing in the Roman period and there is specific evidence from metrical analysis of cattle and sheep bones of increases in the size of stock, suggesting both introduction of new stock from the Continent and improved husbandry techniques. Regional differences were quite marked; for instance, cattle bones predominate in assemblages in the upper Thames Valley.
The role of meat in everyday diet has been much debated and it is generally agreed that in both Roman and British Iron Age societies meat was an occasional food of special status for most people. An association of meat with feasting and with religious sacrifice appears to have been a feature of both societies. Sheep were the preferred sacrificial animal at a number of British temple sites, such as Uley, Harlow and Great Chesterford, with a particular preference for young lambs. This suggests a seasonal cycle to cult activity at these religious sites. Otherwise, meat consumption was closely related to social status, and the financial wherewithal to eat it more frequently was advertised by social customs of inviting others to witness and participate.
In parts of Britain, the material culture of many rural settlements is best characterized as impoverished, with few ostensibly Roman artefacts present. Cornwall and west Devon, Shropshire, Wales, the northern Midlands and northern Britain all fall into this broad category and the exceptional sites immediately excite curiosity. Perhaps it is better to see this as innate conservatism rather than resistance, but it may also reflect the reaction of people in areas of Britain that were offered less chance of wealth accumulation or social advancement than others.
By and large, higher-status rural sites, primarily villas but also including large villages such as Dragonby in Lincolnshire, do tend to show more experimentation with cultivated foodstuffs, animal products, ceramics (large numbers of vessels from a wider range of forms and functional types). The differences between elite and non-elite foodways in Britain evidently widened in the Roman period. In the case of villas we can add in the architectural evidence of ornately decorated dining-rooms, serving as a suitable stage for the rich to play with their food. On the other hand, numbers of amphoras are low on all rural sites and this might suggest that despite a certain flirtation with wine and oil, perhaps for specific social functions, day-to-day cuisine remained fixed on the traditional mainstays of British diet, beer and butter. Nevertheless, even on lower-status sites, there are some indications of changes and experimentation of diet across the centuries of the Roman presence in Britain; as more evidence is accumulated it should be possible to delineate this picture more precisely.
Funerary rituals and commemoration
Funerary practices provide important evidence for the behaviour of the non-elite elements in rural society, though as ever the high-status burials tend to dominate discussion. It is apparent that the period of Roman rule brought about extraordinary changes in the treatment of the body in death, but this was neither a rapid nor a straightforward process. In the late Iron Age there were already great regional differences in funerary practices, complicated by the widespread practice of excarnation (exposure of the body), human sacrifice and head cults. Human body parts turn up with some regularity in unusual contexts on Iron Age sites; sometimes associated with foundation deposits for houses, and in boundary features. They are not uncommon finds in Romano-British contexts either, though on sites with Iron Age antecedents the tendency is to assume that they were residual. Human sacrifice is perhaps most famously marked by the Lindow Man bog body from Cheshire, probably dating to the late first century and the time of Roman annexation of the region. Head cults (the curation of the skulls taken as trophies) are attested by the Roman literary sources and the archaeological data for some of the Iron Age peoples of Gaul. Archaeological finds of skulls from settlement sites in Britain suggest that the practice was also probably significant in some regions of the islands. Tell-tale signs are knife marks indicating de-fleshing of the skull and evidence for the polishing of the exposed parietal bone. A cluster of finds has been made in association with late Iron Age and Roman-period rural settlements in the southern Fenland, while a de-fleshed skull was located in a Roman context at the Folly Lane cemetery by St Albans, and a large number of skulls have been recovered from the Walbrook in London. All of this suggests that the influence of regional British practices with regard to death and the body may have had a long afterlife in the Roman period.
Elite burials dominate the funerary record, whether because of rare epigraphic commemoration or because of the quality of finds. A recent discovery of a pair of exceptional cremations from Harpenden near St Albans yielded a total of over 150 separate items, including 13 bronze vessels, 14 samian vessels, 9 of glass, 2 silver brooches and fragments of ivory. The date of the assemblage appears to be the early second century and the graves lay close to a probable early villa on the site of an earlier Iron Age settlement. The presence of some local pottery and a bronze strainer of Iron Age type may hint at a local origin of those buried here, but the overall assemblage reflects not only an adoption of a wide range of Roman elite culture, but, more importantly, access to it. On the other hand, the average numbers of grave goods in cremations from this part of Britain at this date is less than two, so we are dealing with exceptional individuals, whatever their precise origin.
In general, the areas of south-east Britain that had adopted the practice of furnished cremation burials in the late Iron Age continued to do so under Roman rule, with a gradual expansion in the range of artefacts deposited, to include more characteristically Roman grave goods such as coins and lamps. The difficulty is to differentiate between native Britons and incomers, as at all times native Britons probably dominate the archaeological record and at the local level practices may quickly have converged. Richly furnished cremation burials were a distinctive feature of the south-eastern counties but were rare beyond this zone in parts of Britain where inhumation or other methods of disposal of the dead had predominated in the Iron Age. The exceptions are primarily military sites. Within the south-east zone, regional differences stand out. For instance, larger numbers of vessels are commonly found in burials south of the Thames than in those north of it. This is another indicator that the Harpenden couple could have been immigrants.
Status was conveyed in various ways, with grave furnishings and funeral ritual being supplemented by choice of container, use of surface markers and treatment of the body. Elaborate above-ground structures in the south-east of England include large barrows, but mausolea and other built stone structures are also occasionally attested. Villas and military sites show a higher proportion of elite burials than towns, and contribute more to the overall total. For example, the Bartlow Hills cluster of seven barrows (Essex) probably related to a nearby villa. Villas associated with mausolea include Lullingstone and Bancroft.
The shift from cremation to inhumation was generally slower in rural areas than urban ones. However, by the end of the third century, most of southern Britain favoured the inhumation rite, with limited grave furniture and sometimes with coffins. In the later Roman period the influence of the major towns as cultural trendsetters is more evident, and a number of other innovations introduced from other parts of the empire, such as plaster packing in some elite burials, were primarily focused on the towns.
Nonetheless, there are aspects of the rural burial rite that appear to have remained distinctive and non-Roman. Burials were still commonly placed within boundary features and stray body parts at rural sites suggest that excarnation continued to be practised to a degree. Decapitation of the dead, with the head sometimes placed between the legs of the deceased in the grave, is occasionally encountered in urban cemeteries, but it is much more prevalent in rural contexts, especially in a broad band running from Dorset to the Wash. There are notable concentrations of decapitated burials in Dorset, Somerset, the upper Thames Valley, the Cotswolds, the Nene Valley, Essex and Cambridgeshire, and striking absences in east Kent, Norfolk, Sussex, Lincolnshire, the northern West Midlands. The significance of these burials is much debated and it is clear that more than one procedure was followed. A small minority appear to have been beheaded alive; in other cases the decapitation was performed post mortem; in yet others the skull appears to have been separated from the body after primary decomposition. Men, women and even infants are all represented. Punishment of criminals or deviants is inadequate as an overall explanation, though it is conceivable that the individuals were perceived in some sense as outsiders or on the margins of society. There may be links to pre-Roman practices related to head cults. At any rate, it is difficult to account for the regular appearance of this bizarre rite in terms of normative Roman burial practices. Although attested at several villa sites, it is associated with a wider range of rural settlements. Prone (face-down) burials also occur more commonly at rural sites. Both prone burials and decapitations when encountered in the managed urban cemeteries tend to be located on the fringes, suggesting that those interred in these ways were to some extent seen as outcasts or undesirables.
Infants were commonly buried in a different way from older children and adults and this is particularly clear-cut at rural sites. They were often interred below the floors of buildings, or incorporated in foundation deposits below the walls. Rather than interpret such discoveries as evidence of infanticide, some scholars argue that infant mortality was very much bound up with notions of fertility. The frequent association of infant burials with agricultural buildings suggests that some inherent link was being made between the death of a child and the productive potential of the land. In a similar way, the fact that a high proportion of decapitated burials involved women may also be linked to ideas about fertility and death. At any rate, such unusual customs were particularly prevalent in rural areas, suggesting a significant degree of cultural conservatism and non-Roman behaviour.
Religion
The practice of religion in the countryside differed in important respects from that of the military community and the urban centres, though there was of course a degree of overlap. We need to take account of the movement of the elite between town house and rural estate and of soldiers from fort to farm or simply visiting rural shrines. Our evidence for rural religion is almost entirely limited to the villa/civil zone of the province (Fig. 17). The religious practice of a large part of rural Britain, from Cornwall, through Wales and northern Britain, evidently remained unmonumentalized, non-epigraphic and aniconic.
The overwhelming majority of rural shrines were of Romano-Celtic type, though a few had a more basilical layout, as at Lydney Park. Some temples were sited close to roads or within village-type settlements (Nettleton, Heybridge, Westhawk Farm), but the positioning of others in significant topographic or cultural landmarks is also evident – as in the shrines established in the hillforts of Maiden Castle, Chanctonbury or South Cadbury, or offshore (Hayling Island). A number of the best-embellished rural temples, especially those producing epigraphic material, were located close to major villas or suspected imperial or other major estates: in the case of Uley one can cite Kingscote, Woodchester and Frocester within a few miles and Pagan’s Hill lay close by Keynsham. Many of the more isolated rural temples were set within sacred enclosures and surrounded by subsidiary buildings. The fourth-century temple to Mars Nodens at Lydney Park was erected within an abandoned promontory fort and the complex included accommodation and baths as well as the main shrine. Some at least of these sites had pre-Roman origins – notably Uley, Hayling Island, Wanborough.
