Romanisation: a Kentish Perspective

ROMANISATION: A KENTISH PERSPECTIVE COUN ANDREWS This paper explores some aspects of romanisation in Kent. Firstly it touches on the patchy nature of villa distribution in Kent, and suggests that romanisation within the civitas of the Cantiaci may have been similarly uneven. It then examines romanisation more generally and poses questions about its nature and extent. The concept of romanisation has become increasingly important to an understanding of the Roman period. The basic premise is relatively simple; the elite elements of the native population wanted to adopt a Roman lifestyle as fully as possible. They wanted to continue exerting influence within their tribes (albeit in a more circumscribed way) but had to do so in a new Roman form. Not a great deal is known about how towns were governed but it is thought that the tribal elite became the decuriones or town councillors in a new system and that collectively they would form the ordo or town council and rule their civitas on behalf of the Romans. They would be responsible for tax collection, the recruitment of auxiliaries etc. The main manifestations of this process archaeologically are Roman style towns and villas. Towns required the appropriate facilities according to their status. A civitas capital such as Durovernum Cantiacorum needed a forum and basilica complex, but would also be expected to have a temple, baths etc. The members of this local elite would also require a landed estate, as in the Roman system land ownership was a prerequisite of political power. They would also probably require, or at least desire, a Roman style villa decorated with mosaics and wall paintings. It was through these Roman trappings and symbols, such as villas and romanised towns, that native aristocrats competitively displayed their status rather than through the forms, such as prestige metalwork and weaponry, of the late Iron Age. One of the clearest statements we have suggesting that romanisation was official policy is found in Tacitus's biography of his fatherin- law Agricola, governor of Britain from 78 AD. In a very famous passage, Tacitus tells us that Agricola gave private encouragement and official assistance to the building of temples, town squares and 25 COLIN ANDREWS private houses [templa fora domos]. 1 What form this assistance or official encouragement took is not considered here. For present purposes the passage in question does seem to make clear that certain important aspects of romanisation, notably the establishment of towns containing amenities appropriate to their status, were officially encouraged. The reason for this is that the Romans required towns in their provinces, for it is through towns that they ruled the empire. They were essential, so it should come as no surprise that they were officially encouraged. But in most areas little encouragement was probably required; the old tribal aristocracies wanted a Roman style town for their own reasons. One of them as mentioned above was this competitive display of status, what Tacitus calls honoris aemulatio or rivalry for honour, which he says, proved more effective than compulsion. At the end of this passage Tacitus, with delicious cynicism, refers to the Roman lifestyle that these tribal elites were so keen to embrace in the following terms: 'The unsuspecting Britons spoke of such novelties as 'civilization', when in fact they were only a feature of their enslavement'. What Tacitus is alluding to is that Roman provincial government was only made possible by elements of the native population colluding with them, and that this collusion was expressed by an enthusiasm for a Roman lifestyle. It is obvious that across the province as a whole the degree of romanisation was regional; its quality varied between different parts of Britannia. We might expect the frontier region of the north and the troublesome west to be less integrated into the Roman system than the southeast. But surely Kent must have been romanised? Caesar famously describes the people of Kent thus ' the most civilised of all these nations are those who inhabit Kent [ Cantium]' .2 What he meant presumably is that we were the most romanised, for surely that would be his criterion for civilisation. Ours was the first area to produce its own coinage, which Caesar mentions, we had wealthy burials from the first century BC which contain items imported from Italy, and even had proto-urban centres; surely we would have been in the vanguard of romanisation? Most discussions of the subject tend to make the same assumption and treat 'the south-east' as a single highly romanised block. Villas One way of assessing regional patterns of romanisation is to look at the distribution of villas (Fig. 1). These, along with towns whose distribution pattern is very similar, are good indicators because their existence suggests a Roman style of local government integrated by a 26 • • Fig. I The Distribution of Kentish Roman villas, definite and implied, showing their apparent paucity in mainland east Kent. (Various sources.) COLIN ANDREWS road network with a villa economy. A high density of both villas and towns has long been thought to demonstrate a high degree of romanisation. 3 So when we look at a distribution map of the known (and unconfirmed) villas in South East England it comes as something of a surprise to see the large gaps over much of inhabited Kent. It has long been noted that in the eastern half of the civitas, and especially around Canterbury, villas are comparatively rare. Several explanations have been advanced. One adopts the well-known archaeological maxim that 'absence of evidence is not evidence of absence' and suggests that archaeology may simply have just failed to find a large number of Kentish villas.4 Although a Roman villa seems the most unlikely class of find to go unrecorded, being comparatively large, solidly constructed and of continuing interest to laymen and archaeologists alike. If we assume that the current picture is reasonably accurate and that there were never many villas in parts of Kent how do we explain it? There have been some interesting theories. In the 1960s Rivet noted that it was not unprecedented for villas to cluster around the second town of a civitas rather than the civitas capital and went on to suggest that perhaps the farmers and landowners in the Canterbury area operated from within the town itself.5 Blagg developed this idea and proposed that as yet undiscovered large town houses in Canterbury may have had associated barns, corn-drying ovens, etc., as have been found in other towns such as Silchester and Cirencester. This would mean that these landowners did not need to build villas in the adjacent countryside.6 This may be an aspect of the situation, but does it explain the comparative lack of villas over the whole of east Kent in every period? It also implies that villas were fundamentally units of production, whereas it seems clear that they had other equally important socio-political functions. Villas were to some extent a display of their owner's status, taste and sophistication. Native villa owners were also making a statement about their relationship to the new order imposed by Rome. Frere makes the point that perhaps some of the large gaps in villa distribution may be due to those areas being imperial estates, although he does not apply this to Kent. 7 Black, in his survey of villas in south-eastern England, seems to turn this argument on its head by suggesting that the initial impetus for villa construction came from Gallic immigrants in the late first and early second centuries. These immigrants he argues would have settled primarily in confiscated lands; this is the mechanism he uses to explain why villas are common in west but not east Kent. He feels that after the battle of the Medway large appropriations of land will have taken place, and that 28 ROMANISATION: A KENTISH PERSPECTIVE Gallic immigrants settled on this land, built Gallic style villas and gave the idea to native farmers, who in a spirit of emulation began to build their own villas.8 This is an interesting notion, and there is clearly similarity of design between early villas in the southeast and those in Gaul. It is also fairly obvious that native Britons would have lacked the necessary skills to build or decorate buildings in the Roman style as early as the first century AD. But to suggest that large numbers of tradesmen and entrepreneurs emigrated from Gaul and then built villas here for their own use on confiscated land seems unnecessarily complicated. Might not this similarity of design be more simply explained by itinerant Gallic villa builders and other craftsmen crossing the channel to exploit a lucrative new market? If confiscated land was the key, wouldn't we then expect to find the lands of the Iceni positively stuffed with villas? Significance of the Military Presence Perhaps the answer has something to do with the degree of romanisation. If the correlation between villa density and the adoption of a fully integrated Roman system is correct then perhaps it is in these terms that we should seek to explain eastern Kent's lack of villas. If we Jook at the other parts of Britain that are similarly impoverished, we may be able to establish a link. The areas in question are the northern frontier region, the west generally including Wales, the south west peninsula, parts of Essex and Norfolk, and of course large parts of Kent. One factor that these areas have in common is the large-scale presence of the Roman armed forces or other officials. In relation to the northern frontier region and the west, Millett argues that the presence of these elements undermined the normal system of local government by a local tribal elite. 9 In these areas there was no need for native aristocrats to construct Roman style villas and try to emulate the lifestyle of a landed Roman noble because they were not governing their local areas in the same way that their equivalents in other areas were. Although the Roman army has traditionally been seen as a primary agent of romanisation, by intermarrying with locals, spreading the Latin language etc. So there is something of an ambiguity here. Another associated factor is Roman willingness to adopt and support the elite of any given tribe. For example, it is perhaps no surprise that there is a dearth of villas in the area inhabited by the Iceni. Their tribal aristocracy would not have been adopted by Rome after their revolt of AD 60-61, so the form of circumscribed autocracy established elsewhere was not established in the land of the Iceni. But 29 COLIN ANDREWS what about Kent? We don't appear to have been particularly troublesome, the battle of the Medway (if we can still assert there was such a battle) would be more likely to have retarded romanisation in the west of our region rather than the east. Might we not therefore extend Millett's idea and suggest that the answer lies with the presence of the Roman military? We know that the Class is Britannica had bases at Dover and Lympne from the first quarter of the second century, Reculver and Richborough both saw military activity from at least the early/mid third century. Then towards the end of the third century we have the construction of the so-called Saxon Shore forts. The eastern and coastal regions of Kent may initially have been too strategically sensitive to allow them to be ruled by natives in the same way as more secure areas. After the midthird century, when barbarian raids around the coast seem to have begun, it might also have been considered too dangerous. It is important to stress at this point that the differing character of eastern Kent does not imply that it was in any sense under populated. Stray finds of pottery, coins, etc., suggests that population density was high, just as it was in the Iron Age. But the land was being exploited from native farmsteads rather than Roman style villas. For whatever reason, the native landowners in the eastern part of Kent did not feel the need to display their status through Roman style villas. Iron Age antecedents Another approach is to look at the late pre-Roman Iron Age and see if it provides any clues about Kent's post-conquest development. It has consistently been argued in recent studies, that one of the keys to how successfully romanisation took root is to be found in the period between Caesar and Claudius. 10 Because of course this process did not start in AD 43; there was a long softening up period, in which native elites began to enjoy imported luxuries, such as wine, paraphernalia associated with drinking and feasting, fine pottery and a host of other things. Rich burials from this period show how these imported goods became symbols of high status. Caesar's achievements in Britain are often underrated, but he did bring the Roman Empire to the Channel, and this clearly altered the context in which the late Iron Age people lived, especially in the southeast. The first point to make about Kent, or Cantium, is that it was not a unified territory inhabited by a unified tribe called the Cantiaci. Caesar famously mentions that in his day, four kings ruled Cantium, and he names them.11 It seems clear that the tribe of the Cantiaci was a Roman confection, created out of several smaller units for the sake 30 ROMANISATION: A KENTISH PERSPECTIVE of administrative convenience. Caesar does not unfortunately tell us how Cantium was divided up, assuming that the four kings ruled separate areas, or even what its extent was. But one division seems clear, and has been noted by several scholars including Detsicas, Cunliffe and Burnham & Wacher; 12 all have observed that pottery styles and other evidence suggest that in the late Iron Age the west of the region was quite distinct from the east . Coin evidence further supports the idea of some sort of tribal and political division of Kent in the later Iron Age. In his review of Iron Age coinage in Kent, Holman interestingly attempts to reconstruct a fourfold territorial division based upon the numismatic evidence. 13 The writer supports Cunliffe's view that the Medway itself was not a border and that both banks of the Medway lay firmly within a western pagus. 14 Thus you had one territory centred on the Quarry Wood and Rochester, the other on Bigbury and Canterbury. We know that the Romans saw Britain as a patchwork of tribes and tribal territories, and that they dealt with the natives on a tribal basis. Rivet has suggested that Rochester and Canterbury may have shared power in the Roman period. 15 So might it not be that there was some factor in the late Iron Age political situation which encouraged the growth of villas in the west of the county but inhibited this development in the east? Could it just be coincidence that the sort of east west division we see in villa distribution seems to mirror a di vision that existed in the late Iron Age? The most recent studies of late Iron Age coinage are much less inclined to draw specific historical conclusions from the evidence of coins. Creighton for example, writes 'any comments about expansion, military conquest, contenders for the throne fighting it out etc., are all conjecture spun around very limited evidence' . 16 This is in part due to the work done by Hazelgrove and his contention that late Iron Age coins can only be dated within a generation or so, hence his division of the pre-Roman coinage into nine chronological periods.17 Even the tribal nomenclatures we have become used to are falling out of favour to be replaced by neutral geographical labels, such as southern, south-eastern, etc. Ultimately of course, finding 'foreign' coins in Kent such as those of Cunobelin, only tells us that they were accepted and used there. In ten years we might all be using Euro coins, will this mean we are ruled by Europe? You can see the difficulty. But we can still perhaps draw some limited historical conclusions from the coin evidence. After Caesar's incursions two main powers or dynasties emerged in southern England. The southern dynasty in Hampshire and Sussex traditionally associated with the Atrebates/ 31 COLIN ANDREWS Regni, and the eastern dynasty in Hertfordshire and Essex associated with the Catuvellauni/Trinovantes. Coins issued by rulers from both of these groupings are found in Kent in different periods. For example, coins of Tasciovanus ascribed to Hazelgrove' s period 7 (20 BC-AD 10) are found in the core Catuvellauni/Trinovantes territory but also in Kent, primarily in the west. A little later in period 8 (AD 10- 40 ) we find coins of Eppillus of the southern (Atrebates) dynasty turning up, though this time mainly in the east of the county. But then, also in period 8, presumably following a fairly short period of southern dynasty influence, we find the whole of Kent blanketed with the coins of Cunobelin. However one interprets this evidence it seems reasonable to suggest that in the first four decades of the first century AD her powerful neighbours heavily influenced Kent. We did not in this period produce coinage on our own behalf. So we either had our own rulers who were subject to, or clients of Cunobelin. Or we were actually absorbed into the very powerful eastern dynasty and were effectively ruled from Camulodunum. In a recent study of the imagery on late Iron Age coinage, Creighton has argued that Cunobelin strongly associated himself with the Augustan regime and was a 'friendly' king to Rome. 18 The rich burials discovered within Cunobelin' s core area suggest that he was able to almost monopolise trade with the Continent. His control over prestigious imported goods, would have enhanced his power, prestige and status among other rulers. This was probably to the detriment of Kent, which does not seem to have enjoyed the same degree of wealth in this period. For example, we have nothing to compare with the very rich burials of Hertfordshire and Essex. This is important because, as already noted, there is a correlation between how 'romanised' a particular tribe was before 43, and how romanised they subsequently became. The last coins issued in Kent are specific to that county and are inscribed with the name Amminus. He has been associated with the Adminius mentioned by Suetonius as fleeing from Britain to the continent in 39/40 where he surrendered to Gaius Caligula at Mainz. Suetonius also tells us that his father Cunobelin exiled him. 19 These events suggest that by the time of the Claudian invasion Kent might well have been in a state of disarray. Its indigenous ruling elite must surely have been undermined in consequence. They certainly would not have had the opportunity to become as familiar with imported luxury goods as the elite in Essex and Hertfordshire for example. They also appear to have been subservient to more powerful rulers from the beginning of the first century. Might these circumstances have somehow affected the native elite of eastern Kent's response to 32 ROMANISATION: A KENTISH PERSPECTIVE Rome; made them less susceptible to the seduction of luxury goods and a Roman lifestyle? Canterbury One criticism of the notion that eastern Kent was, for some reason, less romanised than western Kent is that in other respects evidence for the adoption of a Roman lifestyle seems clear. The most obvious factor is that east Kent was reasonably well urbanised with a respectable civitas capital - Durovernum Cantiacorum. Before looking at the Roman town let's briefly explore its origins. The archaeology suggests that the Iron Age oppidum was first settled sometime after Caesar's incursions, perhaps around 15 Bc.20 This foundation has traditionally been associated with the abandonment of the nearby Iron Age site of Bigbury at around the same period, although the date of each event does not seem to coincide as precisely as was once thought. The reason usually asserted for the establishment of a sprawling, undefended oppidum on a flat open site on the river Stour is trade and communications. It would appear that the late Iron Age people wanted to take advantage of the new trading conditions and new markets that developed in the wake of Caesar's conquest of Gaul. A fordable point on the Stour that was also on the overland route from the Channel was ideal. Excavation of the pre-Roman levels has shown some evidence for trade; amphora and Continental pottery sherds provide evidence of imports. Most of the early amphora sherds were found in post conquest levels but Arthur in his study of the Canterbury amphorae is reasonably confident that some foreign goods found their way to the oppidum.21 A comparatively high number of potin coins were also found, which has been interpreted as evidence for trade and commerce. The settlement seems to have been large, spreading over both sides of the Stour. Pellet moulds, possibly for producing coin blanks have also been found hinting at some quite important activities. But the new oppidum was presumably also a significant cultural centre or tribal focus for the people of east Kent. It would probably have been a meeting place and perhaps a centre of religious significance. So unlike many of the towns of Roman Britain Durovernum had a preRoman past, it did not grow out of a Roman fort. A key part of the romanisation theory concerns how a native elite would embellish their own town. In Roman Italy inscriptions attest to public buildings, repairs, renovations and specific facilities, being paid for by wealthy locals. This was an integral part of the urban Roman scene. The wealthy wanted to be seen paying for facilities 33 COLIN ANDREWS within their communities, they were competing with one another for power and influence. Blagg's study of the inscriptional evidence for architectural munificence in Britain throws up some interesting findings.22 He notes that in frontier provinces such as Britannia and the two Germanys the proportion of named individual benefactors who are members of the military is far higher that in other provinces. He also notes that 'there are remarkably few records of munificence by other prominent individuals' .23 Where Britain scores highly compared with other provinces is in munificence by corporate bodies. Such bodies may be the province as a whole, tribal civitates, collegia or military units. When we look at the beneficiaries we find that in southern Britain out of 41 inscriptions recording munificence twenty four of the beneficiaries were deities and fifteen of these were GraecoRoman or oriental. The next most popular beneficiary after deities is the emperor and his family, there are fifteen inscriptions of this type. Not a single inscription has yet been found in southern Britain recording a civil magistrate making an act of munificence. Obviously this is to some extent an accident of survival, but if local RomanoBritish notables were routinely paying for the facilities of their town and local area you would expect some inscriptions to have been discovered. Those that have been found tend to suggest that munificence was primarily the sphere of those most associated with the Roman regime, either as benefactors or beneficiaries. Excavations in Canterbury suggest that up until about AD 70 the Iron Age pattern largely remained unaltered. But then a programme of clearance was begun resulting in the disappearance of the circular post-built huts and emergence of some Roman style structures.24 This was presumably a result of the Flavian encouragement for urbanisation alluded to by Tacitus. It is unfortunate that we do not know how this building programme was funded. There are two distinct views; Millett argues that that the main impetus behind them was native. 25 The local elite would have hired Gallic architects, craftsmen, etc., but that this work may have been subsidised by tax concessions from the Roman authorities. Frere, on the other hand, suggests that Roman involvement was much more 'hands on', with the army actually carrying out much of the construction work in the early towns.26 Both views are at present speculative; we simply do not have the necessary evidence to prove either. Though it is a very important discussion in the context of romanisation. Perhaps the answer lies somewhere between these two views as suggested by Blagg who also describes the complex nature of the relationship between the Roman army, and civilian architects and builders from Gaul.27 Also dating to the early formative period was the first theatre. With 34 ROMANISATION: A KENTISH PERSPECTIVE seating carried on a curved gravel bank this structure appears to date to around AD 90/100.28 At such an early date it is difficult to see the native inhabitants clamouring for a production of Plautus or Terence. Though Canterbury's two successive theatres are often assumed to be clear evidence for romanisation. What could be more characteristic of a desire to adopt classical culture than the early erection of a large theatre? The second theatre, which was built c. 210 in an entirely Roman manner, probably seated at least 3,000 people, perhaps more. So, drawing on Tacitus, we can conjure up a picture in our minds of toga clad natives, chatting together in Latin as they wandered to the theatre to see a classical play. One might also argue that theatre must have been very popular indeed for it to be reconstructed so lavishly in the early third century. But is this idea of the theatre as a powerful symbol of romanisation accurate? Canterbury's theatre is intimately connected with an adjacent temenos or temple precinct. The temple precinct at Canterbury has not been fully excavated but it was clearly a large space surrounded by a portico. Evidence has been found for a Romano- Celtic style shrine or temple in one corner and a fountain in line with the central axis of the theatre. Fragments of decorative stonework and a column capital suggest that there may have been a purely Roman style temple within the sacred enclosure as well, but this is not yet proven.29 This association between temple and theatre is not unusual in a Roman context. But in Britain there are close parallels to the specific arrangement at Canterbury; Verulamium for example, and Gosbecks Farm outside Colchester. Gosbecks Farm is particularly interesting, for here too we find a sacred enclosure, containing a Romano-Celtic style temple in one corner, surrounded by a portico. Furthermore this enclosure was associated with a large capacity Roman style theatre with a date of construction similar to Canterbury's first theatre. The seating of the Gosbecks Farm theatre was also carried on earth banks, and it too was later rebuilt in masonry.30 Blagg has pointed out other architectural similarities between the two such as an apparent lack of substantial stage buildings and parados entrances.31 We might also note that Gosbecks, like Durovernum grew out of an Iron Age oppidum. Significantly Gosbecks Farm, located about 2 miles outside the colonia of Camulodunum, is widely accepted as a primarily native religious sanctuary. Excavations have demonstrated that the sacred enclosure was of pre-Roman origin. So the theatre's primary purpose was presumably religious (especially since according to Tacitus there was another theatre at Camulodunum) and specifically relevant to native or Romano-Celtic religious practice. So even though the 35 COLIN ANDREWS Gosbecks Farm sanctuary may have appeared Roman, with a Roman style theatre, porticos, etc., it was actually not a sign of romanisation but rather the opposite. While not arguing that Canterbury's theatre was never used for entertainment purposes, its primary function, its raison d'etre, should be sought within the context of Romano-Celtic religious practice. Arguably then, the large classical style theatre does not have to be seen as a cogent symbol of romanisation. It may have been built to fulfil some of the functions of the pre-Roman oppidum, perhaps preserving and monumentalizing a site of religious and cultural significance to the native inhabitants. The whole nature of Romano-British religion is very interesting in the context of romanisation because it is here that we see how uncoercive the Romans actually were. Fully-fledged Roman temples are rare, Romano-Celtic style temples comparatively common. The latter structure clearly met the needs of native ritual in a way that a Roman temple did not.32 The point is that the Romans were generally quite happy to allow the native population to worship in their traditional way; taking exception to only the most offensive rites. There were of course changes; the introduction of writing enabled the names of native deities to be written down, they began to be anthropomorphised in stone, temples were rebuilt in stone. So the outward form was altered, romanised if you like, but the substance, it is suggested, remained very much the same. This is one of the difficulties with the concept of romanisation; these Romano/Celtic temples are mainly confined to central and southern Britain, they are built using Roman techniques. Are they therefore evidence for romanisation, or are they evidence for the strength and deep-rootedness of pre-Roman ritual and belief ? Roman Coin Usage Another point often made to suggest how deeply native Britons were affected by the coming of Rome is the introduction of coinage and the development of a monetised economy. When the evidence is analysed however, some interesting points emerge. Roman coins are a fairly common find, long lists are found attached to most excavation reports. The data presented usually concentrates on the sheer quantity of coins in each period. They are primarily used as dating evidence and to suggest the amount of coin usage at various periods on different types of site; urban and rural for example. But what do they tell us about the everyday economy of Roman Britain? To explore this it is necessary to concentrate on denominations not just numbers of coins. The most likely coins to be lost are those that change hands 36 ROMANISATION: A KENTISH PERSPECTIVE most often, those which are used most. The most likely coins to stay lost and be recovered archaeologically are those of lower value for which less time will be spent searching immediately following their loss. For many years Richard Reece has been doing some very interesting analyses of site finds.33 For example a study of coins lost at Richborough in the second half of the first century AD shows that 952 dupondii and asses were lost against only 9 low value semisses and quadrans. It is clear from this ratio that dupondii and asses were used most and lost most. But when we try to work out roughly how much money they might have represented the answer is something of a surprise. We know that under Domitian, at the end of the first century AD, a private soldier was paid 300 denarii per year, not quite one a day.34 Out of this about one third went on paying for food and possibly as much as another third on clothing and equipment. It is surely difficult therefore to see a denarius as having a purchasing power of much less than twenty pounds in today's terms. If this is the case, and the whole calculation is fraught with difficulties, then that would make an as, at sixteen to the denarius, worth over a pound. Even if we accept Reece's conservative figure of a denarius being equal to five pounds (in 1987) that would still make an as worth thirty pence or so. Finds of semisses and quadrans on British archaeological sites are uncommon. They therefore must also have been uncommon as loose change in Roman Britain. This Augustan coinage system lasted until the economic crisis of the mid-third century swept it away. A general lack of coinage of all denominations is demonstrated from every type of site up until 259/60, when devaluation seems to have brought the value of coinage down to a level where they could be used as small change, although this situation is itself temporary. 35 So this phenomenon of a total lack of coins worth much less than a pound is not short term, it lasted for at least two hundred years. It is impossible to imagine a fully monetised economy operating under these circumstances. As Reece writes, referring to the period before 260 AD: 'the commonly occurring transactions in the markets and the streets [were] below the value of the commonly occurring coins'. 36 We surely have to assume that the huge bulk of the native population hardly used coinage at all in their everyday lives, even in the most highly romanised areas. The Divide between Roman and Native One newly discovered site, which seems to reflect the ambiguity of the relationship between native and Roman is Westhawk Farm, near 37 COLIN ANDREWS Ashford. This small town was investigated under the aegis of the Oxford Archaeological Unit.37 A geophysical survey was carried out over a wide area, which beautifully exposed the main features. The site is basically situated on a crossroads; the main road which runs roughly east-west through the whole site connected Canterbury with the iron-working sites of the Weald, another road coming from Lympne joins it at right angles from the south. The point at which these roads met seems to have been the centre of the settlement. The geophysical survey shows how much of the land flanking the northern side of the main east-west road had been divided into a series of rectangular plots delineated by ditches. (See article on Westhawk Farm, pages 1-23.) This can be seen particularly clearly in the north-western corner of the excavated area. What struck the excavators was that most of the structures built within these plots are circular and would appear to be fairly typical Iron Age type habitation structures. In fact the quantity of circular structures and the general lack of stone buildings across the whole site is surprising. Only a small scatter of roof tiles was found for example, suggesting that most structures were thatched. Little evidence was found for late Roman occupation and it looks as if the site was abandoned sometime around the middle of the third century having begun perhaps around 100. The circular structures were still being constructed in the mid/ late second century and appear to have been used throughout the settlement's life. A lot of evidence was found to suggest that iron-working was an important activity for the inhabitants of the Westhawk Farm settlement. This included smelting furnaces and slag for the first part of the iron-making process, and smithing furnaces and hammer scale for the purification of the resulting iron. If, as seems likely, the production of iron was the main, perhaps even sole, industry this may explain why it was abandoned in the third century. We know from other Wealden sites that the iron industry went into terminal decline during this period. It also broadly coincides with the departure of the Dover squadron of the Classis Britannica, which seems to have been involved in the iron industry. But this settlement was more than just a small iron-working town. There are open spaces and what looks like a religious shrine (see pages 1-23); there may have been a bathhouse. The people were not rich, most of the pottery found on site is locally produced. What this discovery gives us is evidence for the people who are usually hidden in the archaeological record. People whose lives and artefacts were ephemeral; they represent the hidden majority of Roman Britain and of Roman Kent. Archaeology has inevitably created a skewed picture 38 ROMANISATION: A KENTISH PERSPECTIVE of the Roman period. There has in the past been a tendency to concentrate on high status remains/artefacts, such things as towns, villas, coins, inscriptions, sculptures, mosaics, roads, baths, etc., generally serve to illustrate the lives of the elite and therefore most romanised members of Romano-British society. It has been calculated that the population of Roman Britain probably approached 4 million. This, it has been argued, means that only ten percent of the population lived in towns and villas. 38 It is this ten percent that the material culture represents, the other ninety percent of the population, living, as mentioned earlier, ephemeral lives, with few artefacts, have until fairly recently been hidden and largely ignored. Over the last twenty years or so this problem has been recognized by archaeologists and historians and an attempt made to redress the balance. The writer's interest in these questions, was provoked by the rapid and utter transformation of Kent in the fifth century from the civitas of the Cantiaci to the Anglo-Saxon kingdom of Kent. Within a couple of generations it is as if the Romans had never been here. The apparent lack of any sign of continuity must surely make one question the extent to which Britain was romanised. While thinking about this problem the following words, written by A. L. F. Rivet in 1963 sprang to mind; 'We have all read, and been irritated by Histories of England ... which either dismiss the Roman period too briefly or, even if they give it longer treatment present it as an episode which was quite without any effect on what followed after. Neither of these attitudes can be logically sustained'. 39 This quotation refers to Britain as a whole, but in Kent it is difficult to pin down any specific area in which the Roman period influenced the early Anglo-Saxon. Everyone is aware of the main areas in which a lack of continuity is relatively clear. Towns such as Canterbury were totally deserted during the early fifth century, villas were similarly abandoned, wheel-made pottery disappears along with Latin and literacy, coins are no longer used, building in masonry ceases, Christianity is swept away. All these things are demonstrable archaeologically. The debate about continuity has been raging for many years and attention has recently become focused on the countryside. It is thought that significant continuity might be found at the level of the basic, comparatively unromanised farmstead. 40 So we are back to that large group of people about whom we know so very little. The people who did not adopt Roman ways, who did not use many Roman artefacts, who continued to live much as they had done in the late Iron Age right through the Roman period. The argument is that they might also have carried on well into the Anglo-Saxon period. 39 COLIN ANDREWS We have good historical evidence for mass migration to the continent in the fifth century.4 1 Much of this was presumably from Kent and the South-East, and those who left probably represented the most romanised element of the population. The people who remained, largely our hidden ninety percent would not have had the same psychological or emotional investment in the Roman way of life. What we call romanisation was primarily a veneer, albeit a pretty and seductive one. During the early fifth century this veneer was stripped off, or more probably simply peeled away of its own accord in the vacuum left by our separation from the Roman Empire in 407. It is quite remarkable how complete and comparatively sudden was our secession from the Latin world. If we compare Britain with modern Romania the difference is striking. Trajan conquered Romania in 106, half a century after Britain's fall. In the mid third century it was invaded by the Goths and ceased to be part of the Roman Empire, yet Romanian, along with Italian, is the closest modern language to Latin; the name means 'land of the Romans'. We could look at Gaul where even after the collapse of Roman rule in the west an attempt was made to perpetuate Roman institutions and ways of life. They were conscious of their Roman past, in a way that Britons, especially in our area, apparently were not. In the mid fifth century the landscape in the South-East must have been littered with fragments of Rome: roads, large crumbling buildings like Canterbury's theatre, villas etc. The earliest AngloSaxon settlers seem to have been very conscious of the people who inhabited the land before them, but without knowing who they were. Fifth-century female graves often contain a collection of small Roman artefacts; coins, buckles, pieces of pottery.42 These seem to have been charms, or were perhaps thought to have apotropaic properties. The Anglo-Saxons seem in a sense to have been haunted by the remains of Roman Britain that they saw all around them. This sense of wonder, of mystery is beautifully evoked in one the earliest English poems probably written in the late seventh century. It is entitled The Ruin, and here are the first few lines:43 Well-wrought this wall: Wierds broke it, The stronghold burst .... Snapped rooftrees, towers fallen, The work of Giants, the stonesmiths, Mouldereth. Rime scoureth gatetowers Rime on mortar Shattered the showershields, roofs ruined age underate them. 40 ROMANISATION: A KENTISH PERSPECTIVE And the wielders and wrights? Earthgrip holds them - gone, long gone, Fast in gravegrasp ..... . Parts of this Roman infrastructure were still capable of resurrection when Christianity returned in the late sixth century. It was in places associated with the Roman past that the new Romans led by Augustine chose to site the first Kentish churches for two hundred years. Durovernum Cantiacorum became repopulated, but was now known as Cantwaraburg. Buildings in stone were once again erected, people wrote again. But all this took place within a firmly AngloSaxon context. However one looks at the evidence it is clear that Anglo-Saxonisation or Germanisation was entirely more successful than romanisation. ACKNOWLEDGEMENTS The writer would like to thank Paul Booth of the Oxford Archaeological Unit for his help in providing information about Westhawk Farm prior to publication. My thanks also to the editors of Current Archaeology for their help. NOTES 1 Tacitus, Agricola, 21. 2 Caesar 5.14. 3 Millett, M., 1990, The Romanization of Britain, CUP, Cambridge, pp. 64, 94, 119. 4 Detsicas, A., 1983, The Cantiaci, Gloucester, p. 84. s Rivet, A. L. F., 1966, The Civitas Capitals of Roman Britain, Leicester, p. I 05. 6 Blagg, T. F. C., 1982, 'Roman Kent', in Archaeology in Kent to AD I 500, P. E. Leach (ed.), London, CBA Research Report No. 48, pp. 51-61 (p. 56). 7 Frere, S. S., 1987, Britannia, 3rd ed., Routledge, London, pp. 266ff. 8 Black, E. W., 1987, The Roman Villas of South-East England, BAR British Series, 171, Oxford, pp. 25, 82. 9 Millett l 990, op. cit. (see note 3), pp. lO0ff. 10 Burnham, B. C., 1995, 'Celts and Romans', in The Celtic World, (ed.) M. Green. Routledge, London, pp. 121-141 (p. 129). 11 Caesar 5, 22. 12 Detsicas 1983, op. cit. (see note 4), p. 10; Cunliffe, B. W., 1982, 'Social & economic development in Kent in the pre-Roman Iron Age', in Archaeology in Kent 10 AD 1500, (ed.) P. E. Leach, London, CBA Research Report No. 48, pp. 40-51 (p. 47f0: Burnham, B. C. and Wacher, J., 1990, The Small Towns of Roman Britain, Batsford, London, p. 39. 41 COLIN ANDREWS 13 Holman, D., 2000, 'Iron Age Coinage in Kent: a Review of Current Knowledge', Archaeologia Cantiana, cxx, 205-233. 14 Cunliffe 1982, op. cir. (see note 12), 48. 15 Rivet 1966, op. cit. (see note 5), p. 109. 16 Creighton, J ., 2000, Coins and Power in Late Iron Age Britain, CUP, Cambridge, p. 78. See also Braund, D., 1996, Ruling Roman Britain, Routledge, London, pp. 67ff. 17 Hazelgrove, C. C., 1987, Iron Age Coinage in SE Britain, Oxford, BAR Report No. 174, pp. 99-122. 18 Creighton 2000, op. cit. (see note 16). 19 Suetonius, Caligula, 44. 20 Blockley, K., Blockley, M., Blockley, P., Frere S. S. and Stow, S., 1995, Excavations in the Marlowe car park and surrounding areas, Archaeology of Canterbury V, Part l, Canterbury, p. 48. 21 Arthur, P., 1986, 'Roman Amphorae from Canterbury', Britannia, 239-259. 22 Blagg, T. F. C., 1990, 'Architectural Munificence in Britain', Britannia, 13-27. 23 ibid., p. 20. 24 Blockley et al., op. cir. (see note 20), p. 53. 25 Millett I 990, op. cir. (see note 3), pp. 69ff. 26 Frere 1987, op. cit. (see note 7), pp. 230ff. 27 Blagg I 990, op. cit. (see note 22), p. 21. 28 Frere, S. S., 1970, 'The Roman Theatre at Canterbury', Britannia, 83-114 (p. 110). 29 Blagg 1982, op. cit. (see note 6), pp. 52ff. 3° Crummy, P ., 1986, ln Search of Colchester's Past, Colchester Archaeological Trust. 31 Blagg, T. F. C., 1989, 'Roman Art & Architecture', in Research on Roman Britain, (ed.) M. Todd, Britannia Monograph No. 11, pp. 203-219 (p. 211 ). 32 Muckelroy, K. W., 1976, 'Enclosed Ambulatories in Romano-Celtic Temples', Brilannia, 173-192. 33 Reece, R., 1987, Coinage in Roman Britain, London; Reece, R., 1995, 'Site-finds in Roman Britain', Brilannia, 179-207. 34 Suetonius, Domitian, 7; Webster, G., I 956, The Roman Army, Chester, pp. 30ff. 35 Reece 1995, op. cit. (see note 33). 36 Reece 1987, op. cir. (see note 33), pp. 37ff. 37 Booth, P. and Lawrence, S., 2000, 'Westhawk Farm Ashford', Current Archaeology, No. 168, May, 478ff. 38 Millett 1990, op. cit. (see note 3), pp. 181ff. 39 Rivet I 966, op. cit. (see note 5), p. I 10. 40 Esmonde Cleary, 1989, The Ending of Roman Britain, Batsford, London, pp. 200ft. 41 Hawkes, S. C., 1982, 'Anglo-Saxon Kent c. 425-725', in Archaeology in Kent to AD 1500, P. E. Leach (ed.), London, CBA Research Report No. 48, pp. 64-79 (pp. 66ff). 42 Drewett, P., Rudling, D. and Gardiner, M., 1988, The South-East to AD 1000, Longman, London, p. 272. 43 The Earliest English Poems, Penguin Books, 1964, p. 30 (Tr. Michael Alexander). 42