The last general review of the evidence for Roman Warwickshire, published in 1996 (Booth 1996), took the form of a survey of the work of the previous 20 years or so since an earlier brief summary by Webster (1974). Recent years have seen some significant publications of sites old and new, and a steady stream of PPG16 driven fieldwork. The latter has generated a quantity of ‘grey’ literature, but much of this has yet to be converted into more widely accessible publication. There remains a backlog of very important excavations dating back to the early 1980s and beyond, the publication of which would still be of great value for the archaeology of the region, not just of Warwickshire itself (see further below). The most significant publications of the last few years relate largely to the small town of Alcester and importantly, to its immediate hinterland in the lower Arrow valley, discussed below.
This review, like its predecessor, follows a fairly traditional framework, but it is hoped that the potential of Warwickshire evidence to address a wide range of new questions will be apparent. The review outlines the character of the information base for Warwickshire and the state of knowledge in a series of chronological and subject based themes, and then summarises the main strengths and weaknesses of the data, indicating two types of areas for future work – those where our knowledge is deficient and requires active enhancement, and those where Warwickshire possesses datasets of high quality which by further promotion can be used as models for the wider region.
The character of the resource
Despite an antiquarian tradition extending as far back as references in Leland and with major local exponents such as William Dugdale, a tradition reflected in Haverfield’s 1904 survey in VCH1, there was no meaningful field examination of Warwickshire’s Roman archaeology until the 1920s. Even then, and allowing for Thomas May’s involvement at Tiddington (Fieldhouse, May and Wellstood 1931), this work was amateur and small scale and continued to be so after the Second World War. A more systematic approach to the problems of specific sites developed from the 1950s, in work still largely carried out by amateur groups, but with active encouragement and support from Graham Webster in the Department of Extra-Mural Studies of Birmingham University (eg Hughes 1960, 10). This pattern of activity was maintained for some considerable time, but began to change in the 1960s with the advent of Ministry of Public Buildings and Works funding for a variety of excavations including those of specifically threatened sites. The most notable of these early rescue excavations was that directed by Christine Mahany at the small town of Alcester – it is particularly unfortunate for the study of such sites generally, and in the west Midlands in particular, that these excavations were not finally published until 30 years after their commencement, by which time their potential impact was lost, though the data presented remain extremely valuable.
The scale of the 1960s work at Alcester was not matched again until the end of the 1970s when Warwickshire Museum initiated a series of major rescue excavations at Coleshill, Tiddington and Wasperton. Excavation in a development-led context, from small-scale housing development to gravel quarries, pipelines and infrastructure projects such as roads, has continued ever since. Research excavation has tended to be carried out by local groups and on a small scale – though these groups have also been involved in sites threatened by development. The most significant excavation not driven by a specific development threat was that at The Lunt, Baginton, resulting in one of the most complete fort plans known from lowland Britain (Hobley). Unusually, work at Broad Close, Mancetter, examining a part of the major pottery industry, was occasioned principally by the threat of ploughing to pottery kilns.
The results of excavations from the 1960s onwards have formed the backbone of our knowledge of the Roman period in Warwickshire. There has been no significant tradition of widespread fieldwalking, though a number of sites, for example in the Dene Valley, have been located in this way (McKay*). There is also a large body of data from aerial photography, concentrated in, but not confined to, the Avon Valley. This was the subject of a pioneering study (Webster and Hobley 1965), and more detailed plotting was carried out in the 1980s (subsequent work?), but the area remains less well known in this respect than, for example, the Upper Thames valley or parts of the Severn Valley. Geophysical survey has generally remained a site specific tool and recent attempts to use magnetometry on a wider scale, for example in the context of M6 Toll in north-west Warwickshire, were disappointing, probably principally because of general unresponsive soil types or conditions in this part of the county.
The combination of excavation allied to development and the preponderance of aerial evidence on the Avon valley, has tended to produce a somewhat skewed data set. Two major nucleated settlements, Alcester and Tiddington, are relatively well known through detailed excavation. The latter lies by the Avon and the former is situated at the confluence of two of its tributaries, so the relationship between nucleated and rural settlement, insofar as it is understood at all, is best considered in the context of these two key sites.
The late Iron Age
The transition from the late Iron Age into the early Roman period is also most clearly demonstrated in the Avon valley where large-scale excavation in advance of gravel quarrying has permitted the identification of settlement sequences that span this transition, most particularly at Wasperton. Unusually for the West Midlands, Warwickshire produces evidence for a distinctive late Iron Age defined by ceramics. Material of this type, with (inter alia) grog tempered fabrics and wheel thrown forms characteristic of the south-eastern ‘Belgic’ tradition (sensu Thompson 1982, 4), occurs at a number of sites, particularly in the Avon Valley. This material raises a number of important questions, of which its absolute dating remains one of the most crucial. There is at present no evidence that the appearance of this pottery can be dated much more than a generation before the Roman conquest, and it might lie rather closer to that event, but in the absence of other forms of dating evidence this remains a subjective judgement. The distribution of this material on present evidence seems to be based principally on the Avon Valley, with the most notable assemblages at Tiddington and Wasperton. The valley seems effectively to mark the frontier of the distribution of this pottery – in the north-west of the county it is conspicuous by its virtual absence, with tiny numbers of sherds from Coleshill and recently examined sites on M6 Toll at Wishaw and (straying just over the county boundary) at Sutton Coldfield. It is not clear that every site occupied in the late Iron Age, even in the Avon Valley, necessarily had access to this distinctive ceramic material. In these cases, therefore, it must be presumed that hand made pottery of middle Iron Age type potentially continued in use up to the Conquest period, if not beyond (see further below for the specific case of Coleshill). In the absence of clear evidence for chronologically diagnostic morphological change in settlement types the problem of distinguishing middle from ‘late’ Iron Age sites therefore remains, and for those sites which do not have clear evidence of Roman activity it is impossible to demonstrate if their decline is related to the Roman conquest, to social and economic developments in the LPRIA or should be assigned to other factors and potentially earlier dates.
There are no certain nucleated settlements of late Iron Age date. None of the known hillforts demonstrably continues in occupation at this time, though this judgement may simply reflect the lack of detailed evidence for these sites in the county. The occurrence of ‘Belgic type’ pottery associated with the large low lying defended site at Wappenbury is of interest, but this pottery does not date the probable secondary phase of the ‘hill fort’ rampart, which appears to have a terminus post quem of at least the later 1st century AD and may have been much later (Booth 1991a; contra Stanley and Stanley 1960, 9) – despite its lowland setting which might suggest a late Iron Age date this site is most likely to have originated earlier. Suggestions of a possible defended site to the south of Alcester can now be discounted (Jones et al 1997, 8). Where its context is known the distinctive late Iron Age pottery occurs in rural settlement enclosures of types seen both earlier and later.
There is still insufficient evidence to allow generalisations to be made about continuity of settlement from the Iron Age into the Roman period – there are remarkably few sites with a clear sequence from middle Iron Age to early Roman and beyond. Sites such as Park Farm, Barford, for example, seem not to have been occupied into the Roman period in any significant way (Cracknell and Hingley 1994, 28-29) but it is impossible to say if the end of occupation falls a little before, at, or a little after the time of the conquest. At Wasperton continuity of settlement seems likely, but at Tiddington, where middle and ‘late’ Iron Age and early Roman material are all present there seems to be a spatial distinction between middle and late Iron Age features, which indicates at least the reconfiguration of settlement, if not a hiatus. Another important site with a potential sequence is Coleshill, but here the chronology remains problematic. An unenclosed roundhouse settlement, associated with hand made pottery of middle Iron Age character, directly underlay the later temple complex, the earliest phase of which may perhaps be assigned to the later 1st century AD. The major problem is to determine the period of occupation of the roundhouses in the absence of absolute dating evidence. It is possible, for example, that the latest of them was post-conquest in date since in an area with no significant distinctive late Iron Age ceramic tradition (see above) the middle Iron Age style presumably remained in use until supplanted by ‘Romanised’ coarse wares in the second half of the 1st century AD. These questions are also of considerable significance for other aspects of the interpretation of the site, in particular the change of use from domestic settlement to temple complex.
Finally it may be noted that the distribution of Iron Age coins falls broadly into two, with concentrations of Corieltauvian coinage in the north of the county, and Dobunnic coinage in the south. The distributions overlap at sites such as Alcester and Tiddington and at the former, at least, there is the possibility that some of these coins are early post-conquest introductions in a military context. These distributions not only shed some light on the possible extent of tribal territories (cf Booth 1996, 27, 52; Van Arsdell 1994) but also tell us a little about the character of late Iron Age society in the region in contrast, for example, with the coinless Cornovii to the north-west.
There is some evidence for a Roman military phase at a number of sites in the county, three of which have to date demonstrated particularly significant concentrations of military activity of the conquest period and later, at Mancetter, Baginton and Alcester. At the first of these, where a ‘vexillation fortress’ was succeeded by smaller installations (Scott 1984; 2000), primary occupation may be dated to the early AD 50s, but apparently not earlier (Seaby and Perry 2000, 52). It is possible that a similar sequence was followed at Baginton, where early military activity might have been focussed at Home Farm (Frere 1984, 295), some 250 m to the south of the well known sequence of forts at The Lunt. The date of Home Farm is uncertain, but coins from the Lunt itself suggested that this site might not have been established until after the rebellion of Boudicca in AD 60 (Reece 1975, 24). It should be noted that phase I features at the Lunt appear to have extended beyond the line of the eastern defences of the later fort and their character remains unclear (Hobley 1975, 17; cf Booth 1996, 28). None of the work carried out subsequently to Brian Hobley’s excavations has been published. At neither Baginton nor Mancetter, however, is activity demonstrable beyond the early Flavian period at the latest, while at Alcester, where the earliest fort (south of the town) might perhaps have been of Claudian date (Booth 1996, 28), there are indications, though the evidence is less clear than one could wish, that military occupation with the area of the later town may have extended into the early 2nd century (Booth and Evans 2001, 301-3), to parallel the situation indicated at some other sites in the region (such as ?Wall, Dodderhill and possibly Metchley)*. The reason for this sequence of occupation, in contrast with the pattern seen at Baginton and Mancetter, is unclear. Development of Mancetter as a non-military settlement seems to have centred on Watling Street, some distance from the complex of military sites, while the character of undoubted later settlement at Baginton, and its relationship to the fort, remains obscure.
The most visibly obvious legacy of the conquest period is a network of major roads. The routes of these are tolerably well understood, but minor roads remain largely unknown.
Alcester, of course, developed into the principal nucleated settlement of the Roman period in the county, and is arguably, with the publications of recent years (Booth 1989a; 1989b; Booth and Evans 2001; Cracknell 1989; 1996; Cracknell and Mahany 1994; Mahany 1994; Mudd and Booth 2001), one of the better known small towns of Roman Britain, albeit that many important questions remain to be answered. Unusually for a site of this type there remains a significant lack of knowledge of the defended area compared with extramural settlement, and this area not only lay at some distance from the two major roads which ran through the settlement but was also probably not a primary focus of settlement. With a substantial occupied area, ?2nd century and 4th century phases of defences and considerable structural diversity, including probable or possible ‘public’ buildings - a major storehouse (Booth 1989b) and a potential mansio (Mahany 1994, 157) - Alcester is clearly amongst the upper order ‘small towns’ (Burnham 1995, 10) and may have been a centre for local administration in the northern part of Dobunnic territory.
For various reasons – though principally lack of development pressure and therefore of invasive fieldwork (not necessarily a bad thing) - knowledge of most other major nucleated settlements in the county has lagged behind. The sites in question are Tripontium, High Cross and Mancetter on Watling Street, Princethorpe and Chesterton on the Fosse Way and Tiddington, probably on a minor road running up the Avon valley from the salt way crossing at Stratford. Publication of the important early 1980s work at Tiddington would significantly enhance our knowledge of the range of site types within the ‘major nucleated settlement’ bracket.
The Watling Street sites all had defended enclosures. In the case of Mancetter and Tripontium these were probably of later 3rd century date, broadly comparable with those of Bannaventa to the south-east and Wall to the north-west, while the defences of High Cross are known only from the air. Tripontium, one of the more extensively examined nucleated sites in the county, saw salvage recording work in advance of gravel extraction principally in the 1960s (Cameron and Lucas 1969; 1973). Since that time, work has continued on the site of a probable mansio outside the defended enclosure (Lucas 1984; Black 1995, 56, 72 and 81). Many aspects of the site remain very poorly understood, however, and apart from the mansio complex meaningful structural evidence is largely lacking. At Mancetter the military sites and major foci of pottery production, both at some distance from the line of Watling Street, are both better known than the defended enclosure which bestrides the road itself. Recent work, however, has revealed a substantial structure.
The Fosse Way settlements also had defensive enclosures of which that at Chesterton may, like Alcester, have had two main phases (the 1960s excavations, never published, are conveniently summarised by Burnham and Wacher (1990, 249-252), cf Booth 1996, 35-37). This may indicate a difference of character between this and some of the ?smaller settlements along the major Roman roads in the county, but there is still insufficient evidence to provide detailed definition of these differences. Princethorpe, for example, has only seen very small-scale examination in recent years (Cuttler and Evans 2000). A recent reassessment of Dorn, just over the county boundary in Gloucestershire, provides useful comparanda, however (Timby 1998, 390-399).
Tiddington, although sited just away from the major road network, also had a defensive enclosure of late Roman date, but its extent is uncertain. In possible contrast to some of the other nucleated sites Tiddington seems likely, on the basis of the excavated sample, to have been principally agricultural in character, though ‘ancillary’ features included two pottery kilns (Palmer 1982; 1983; Burnham and Wacher 1990, 310-313).
Further minor nucleated settlements probably include Brailes (away from the main road network), Bidford on Avon on Ryknild Street, a small site at Billesley on the salt way between Stratford and Alcester, and perhaps also Coleshill, though the settlement aspect of this site is little known. The principal importance of Coleshill lies of course in its temple complex, which is the only one of its kind currently known in the county and is at least of major regional significance. Again publication of this crucial site is urgently needed. It is the location of this complex that suggests that the associated settlement may have been of nucleated character. In general terms the site may have been positioned in relation to a crossing point of a minor road over the river Cole just above its confluence with the Tame (cf Magilton 1980, 27-29), though there is still no concrete evidence for the existence of such a road. More importantly perhaps, it is increasingly clear that substantial buildings of this kind were routinely located in major nucleated settlements and for the present, at least, it seems that rural shrines in our region for the most part did not take the form of substantial structures.
In the context of Coleshill, however, the chicken and egg question of primacy, whether of temple or, as I would argue, settlement, remains to be resolved.
Rural settlement presents a different range of problems. The data set is very uneven: there are still relatively few well-excavated sites or well-understood landscapes recovered in other ways, although this situation has improved with recent work. There is still, understandably, a marked bias in our evidence towards what might be termed the broader Avon valley. A relative absence of villas has always been a characteristic of the rural settlement pattern – the 1996 estimate was 12-15, a figure based largely on the simple criterion of the presence of flue tile, in many cases known only from surface collection. There is now evidence for up to another 6 or 7 from recent work of various kinds, all in the south of the county, including an example at Long Compton located from the air, one at Burton Dassett indicated by finds in a trench in the churchyard and a recently exposed mosaic pavement of Corinian type at *?Lower Tysoe. There are very slight hints of clustering of villas in the vicinity of Alcester and Chesterton, but the sample size is probably too small for such conclusions to be very meaningful. Perhaps a third of all these sites have seen any excavation, the only extensive work being that carried out at Salford Priors during work on what is now the A46 south of Alcester (Palmer 2000). This site contained a number of well-scattered buildings, none of any great architectural pretension, and its relatively modest character may be typical of such sites in the county, but it is likely that the principal domestic building lay elsewhere, therefore giving a misleading impression of the character of the site from the structures examined.
Lower status rural settlements must have been numerous, but remain relatively little known in relation to their likely original numbers. Many sites on the Avon valley gravels known from the air could on morphological grounds be either Iron Age or Roman in date, and some are of course likely to be both (cf Booth 1996, fig 11). Rectangular or sub-rectangular enclosures seem generally to be characteristic of such settlements from the middle Iron Age onwards. The sequence is well illustrated at Salford Priors, where a rectilinear enclosure with a central roundhouse, of Iron Age date, had a smaller enclosure added to its east side in the Roman period. A few hundred metres to the south 1st-century features included another rectilinear enclosure while to the north contemporary settlement activity, including subrounded structures, may have been unenclosed. In the early 2nd century these areas were cut through by a trackway upon which later ditch alignments were based. In the 3rd century the southern part of this area became part of the villa complex already mentioned.
At the opposite end of the Avon valley on Dunsmore in north-east Warwickshire we see further evidence of continuity from the Iron Age into the Roman period, here in the form of redefinition of existing ‘land unit’ boundaries. For example at Ling Hall Quarry, Church Lawford, an early Roman farmstead was based on one of the earlier boundaries. An interesting aspect of this site is the suggestion that barley was a favoured cereal crop because it would grow relatively easily on the acid rich Dunsmore soils.
Another characteristic of the Dunsmore area, and again one that seems to be reflected in some other sites in the county, is the fact that none of the known Roman activity extends later than the 2nd century AD. It is difficult to assess the significance of such a date range when sites are only partly excavated and later activity could have occurred in adjacent areas, but elsewhere, for example in the Upper Thames Valley, there is evidence for significant dislocation of settlement patterns in the first half of the 2nd century AD, and it is possible that a related phenomenon may be occurring in the Avon valley. There are of course sites which have a longer sequence of rural settlement; so for example at Wasperton relatively low status occupation continued through the Roman period and at Wishaw, in the north west of the county, a Roman enclosure that may originally have related to a sequence of middle and late Iron Age boundaries was apparently occupied into the 4th century (but not up to the end of that century). Many if not most of the early Roman rural enclosures produce evidence for roundhouses if they contain any structural remains at all. Structural evidence on later Roman low status settlements appears to be scarce. Again this is a pattern seen in the Upper Thames Valley. There are, however, rural settlement sites in which a progression from circular to rectilinear structures can be recognised, for example contained within curvilinear enclosures at Bidford Grange (Hart 1991; Hart, Osgood and Palmer 1992, 18) and within a rectilinear enclosure at Crewe Farm, Kenilworth (Ford 1971a). Both sites may have seen a peak of activity in the 2nd century but the Bidford Grange farmstead was occupied through most of the Roman period.
The agricultural economy is being increasingly elucidated through the study of plant and animal remains, though until recently useful data for these have been confined to a small number of sites, particularly Alcester, Tiddington and Wasperton (cf Booth 1996, 45-46). At least the first of these can be characterised more as a consumer than a producer site so the evidence for plant and animal husbandry will be skewed to reflect this. Significant new data are now available from Salford Priors. Here it is interesting that in terms of volume the principal samples, particularly of charred plant remains, are all of later Roman date. It is also striking that the range of species represented seems very reminiscent of that encountered at sites within Alcester, the small town in whose hinterland Salford Priors lay. The major cereal at Salford Priors was spelt wheat (Moffett and Ciaraldi 2000), as it was also in Alcester at sites such as Coulters Garage, where large deposits of burnt chaff and grain were associated with a possible granary building (Colledge 1989). While spelt wheat was probably the principal cereal crop produced at many Warwickshire sites the suggestion that barley was favoured on the more acidic soils of the Dunsmore gravels is particularly interesting and may be a pointer to identification of other aspects of local variation in the character of rural settlement.
Despite these hints of local variation in agricultural production there is little indication that the Warwickshire evidence, important as it may be, is intrinsically remarkable. In another area of economic activity, however, that of ceramic production, the situation is very different, and Warwickshire has wide ranging evidence for pottery and tile manufacture at a variety of sites of differing scales (for a summary of some aspects, see Booth 1986). The most important of these is the Mancetter-Hartshill complex, an industry of national importance for its specialist production of mortaria. Excavations over a period from 1960 to 1984, of some 61 kilns with associated features and pottery assemblages, remain unpublished. Unfortunately none of the smaller kiln sites examined in the 1980s and 90s have been published either, so understanding of the range of industries present is based largely on interim data with, in some cases, no detailed characterisation of the products involved. With regard to other evidence for ‘industrial’ activity there have been no significant new discoveries since the mid 1990s.
Summary: strengths and weaknesses of the data for Roman Warwickshire
Warwickshire has important evidence for the early military phase in the region and, with a ceramic tradition that permits identification of a distinctive ‘late Iron Age’ on the basis of part of its material culture, the potential to identify non-military sites of this period and compare and contrast these with the settlement evidence of earlier and later periods. There can be little doubt, however, that routine application of radiocarbon dating to ‘middle Iron Age’ sequences is of great importance, not only for understanding these sites in their own right and but also for establishing the character of base line rural settlement as a background to possible political, social and economic changes in both the LPRIA and in the post-conquest period.
In terms of major nucleated settlement Warwickshire has good evidence, particularly from Alcester, which may have followed a well-established model of town development around an early military nucleus. Like Alcester, other major settlements developed mainly along the principal Roman roads but the importance, if any of a military background in their establishment is generally unclear. Equally, however, there is generally no evidence for the existence of significant pre-Roman settlement at these locations – Tiddington, not on a major road, is clearly an exception to this – and a primarily economic driving force may provide the best model to explain the growth of some of these sites, whether this was an organic process or one stimulated by the imposition of official establishments such as mansiones. The broad distribution pattern of upper echelon settlements is tolerably well understood, though detailed evidence for some is largely lacking and there may be a serious danger in assuming that Alcester is representative of all the sites in this group. Some minor nucleated settlements may yet remain to be located. The interrelation ship of the more important of these sites, Alcester and probably also Chesterton, with the others, and the degree of integration of all the larger sites into the wider rural settlement pattern is less clearly understood.
Other particularly strong aspects of the data for Roman Warwickshire include the evidence for ceramic production, but much of the potential of this evidence remains to be realised through detailed publication of the sites in question.
Relative weaknesses in our data relate most importantly to rural settlement. Evidence for this is good in places, but our understanding of the rural settlement pattern and the associated agricultural economy is patchy and still excessively dependent upon data from gravel sites in the Avon valley. A further major weakness concerns the archaeology of the dead. Cemeteries and more scattered burials are known from a number of sites, principally Alcester (eg Mahany 1994, 144-147), Tripontium (Cameron and Lucas 1969, 139 and 141-144; 1973 passim) and Tiddington (Palmer 1982, 5, 14-16) amongst the major settlements and Stretton on Fosse (Gardner et al 1982, 9 and 30) and Wasperton (Crawford 1983, 25-6) amongst rural ones. The last of these is of particular importance in relation to the question of rural population survival in the post-Roman period, and some of the Stretton on Fosse burials may also be relevant to this discussion (Ford 1971b). Overall, however, this is a very small sample and few even of the assemblages within it have been studied systematically. Such evidence as there is suggests a later Roman date for most if not all of these burials. In a loosely related sphere, that of religious practice, evidence, whether for communal or individual activities, is again scarce, with the singular exception of Coleshill.
Some questions and priorities for further work
It is perhaps inevitable that the preceding discussion and what follows bear strong similarities the summary and conclusions of the last review (Booth 1996, 54-55), though it is hoped that some of the outstanding questions are a little more nuanced than was the case previously. It is still true, however, that there is no single subject area relating to Roman Warwickshire for which we can claim to know all that we may reasonably want to. Future work should therefore allow for progress on a broad front, and most new projects arising through the planning process without regard to archaeological agendas can probably contribute significantly to our overall picture. Site specific issues remain important, both for large and small settlements, well-known or poorly-understood. Even for sites such as Alcester, for example, a simple research agenda might seek to address basic topics such as 1) definition and refinement of chronology of probable and possible military features, 2) location and definition of (particularly early) cemeteries, 3) characterisation of occupation within the defended area, 4) examination of the possible mansio to determine its true function, 5) further definition of the range and scale of economic activities. For most of the major settlements fundamental questions relating to their origins, chronology of development and range of economic and other functions remain to be addressed. Related to these issues is the fundamental one of how these sites were integrated and interacted with their rural hinterlands – itself an issue potentially of relevance for the very nature, density and economic character of rural settlements.
The data to characterise rural settlement in these terms still hardly exist, though they are more numerous in the Avon valley than elsewhere. There are hints that there may have been differences in the character of rural settlement between the three broad north-east to south-west trending zones of the county; Arden, Avon valley and Feldon. One such hint is the distribution of villa sites – particularly as expanded by recent discoveries. On present evidence only Salford Priors and Sambourne, both in the Arrow valley and in the hinterland of Alcester, lie north of the Avon. There is no suggestion that the Feldon and Cotswold fringe of the county should be seen as characterised or dominated by villa type settlement in a way that might have applied a little further south, in north Oxfordshire. Nevertheless the implication of recent discoveries is that this may be a landscape in which villas played a significant, rather than just an occasional, role. In the absence of detailed examination of any of these sites, however, we have no evidence for the interrelationship, if any, between villa and non-villa settlement. Examination of this relationship, both in this area and elsewhere in the county, will require much more evidence than is currently to hand – accumulation of such basic evidence must be a priority. Precisely the same situation prevails north of the Avon. The differences in settlement character between Feldon and Arden in the medieval period are well known (eg Roberts 1978) and while this need not bear any direct relationship to the settlement patterns of the Roman period it would be surprising if there were no variation to reflect the broad differences in landscape character which seem to have been established over a very long period. It is emphasised, however, that this is a model to be tested, not a statement of fact.
The chronological aspect of rural settlement also requires extensive and detailed consideration. A number of sites appear not to have outlasted the 2nd century AD, and it is increasingly possible that this is related to a broad chronological trend in rural settlement in the county, particularly in view of the rather clearer evidence for widespread dislocation in rural settlement patterns to the south in the Upper Thames Valley in the 2nd century (Henig and Booth 2000, 106-110). The Warwickshire rural settlement record needs to be considered in the light of such possible comparanda, but the evidence for possible major chronological (and other) trends has to be satisfactorily mapped before they can be interpreted.
Understanding of wider questions of society/regional identity may also be relevant to these as well as other issues. The Avon valley may have been a frontier zone for ‘Belgic type’ ceramics shortly before the Roman conquest, but does this reflect anything more than the adoption by the local populace of a new technology? Is there anything to be made of a correlation between this distribution and the tribal territory of the Dobunni? At a slightly later period the distribution of Severn Valley ware has been seen as potentially significant in these terms (Evans 1994, 148-9). This may be so, but I would place the border between the Dobunni and Corieltauvi a little further east than Evans implies. The potential significance of Coleshill as an indicator of tribal boundaries, in this case most likely between the Corieltauvi and the Cornovii, has been remarked on before (eg Magilton 1980, 29). With parts of the county lying potentially within at least three tribal territories, there is scope to develop further the question of the relationship between aspects of material culture – perhaps most clearly seen in terms of pottery distributions - and such territories. A notable characteristic of recent fieldwork on the line of M6 Toll through Staffordshire, the margins of Birmingham and Warwickshire, was the relative paucity of Roman finds (as well as those of several other periods) even from quite substantial sites. Arguments from such negative evidence can be difficult to sustain, but a superficial judgement is that Cornovian rural settlements were ‘poor’ in terms of material culture. Careful comparative studies could first examine the validity of this observation and then seek to explain it.
Questions of identity of communities and societies are also important at the end of the Roman period. Here we have particularly important evidence, with the suggestion that cemeteries at Wasperton and Stretton on Fosse show potential continuity of at least some rural populations into the Anglo-Saxon period. Any future discoveries of sites of this type should deploy DNA analysis and other techniques to the question of determining the degree of similarity or difference between successive or contemporary populations. This analysis should also embrace material from ‘early Anglo-Saxon’ cemeteries. Such an approach carries implications for standards of recovery and sampling of appropriate material in order to avoid contamination which will impact on field methodologies.
Repeated reference has been made to a small number of key outstanding publications, principally of large-scale excavations carried out in the 1980s. The most important of these sites, in no particular order, are Tiddington, for its contribution to understanding of nucleated settlement, Coleshill, for its contribution to the study of religion in the region, Wasperton, for its contribution to knowledge of the long term development of rural landscapes and, importantly, to burial archaeology and the question of population continuity in the post-Roman period, and Mancetter-Hartshill, for its contribution to pottery studies. All of these sites can be considered to be of major regional, if not of national importance and in all cases significant post-excavation work has already been carried out. The detailed publication of these key excavations, despite their vintage, will contribute very considerably to our growing understanding of the period, complementing the data being derived from current work. Completion of these reports should be prioritised where possible.
There are a number of lesser, but still locally/regionally important sites also deserving of publication. The pottery kilns at Lapworth were excavated in 1988 in advance of the construction of the M40, with no provision for further work. The status of 1980s excavations at Home Farm, Baginton is unclear. Of a much earlier vintage, the 1971 excavation at Crewe Farm still has an important contribution to make to understanding of rural settlement, but it is uncertain how much material survives from this site.
Note and acknowledgements
This survey is considerably shorter than that produced by the present writer in 1996. An attempt has been made here to reference all significant new work carried out since that time, but the reader is referred to the earlier survey for references up to that time, where not given here. I am deeply indebted to Nicholas Palmer and Stuart Palmer of Warwickshire Museum for helpful comments and suggestions and for providing much new information.
Black, E W, 1995 Cursus Publicus, the infrastructure of government in Roman Britain, Brit Archaeol Rep (Brit Series) 241, Oxford.
Booth, P, 1986 Roman pottery in Warwickshire: production and demand, J Roman Pottery Stud 1, 22-41.
Booth, P, 1989a Excavations at 64 Bleachfield St, Alcester, 1981, Trans Birmingham Warwickshire Archaeol Soc 93 (for 1983-4), 9-32.
Booth, P, 1989b Roman Store Buildings at Alcester, Trans Birmingham Warwickshire Archaeol Soc 94 (for 1985-6), 63-106.
Booth, P, 1991 Late Iron Age pottery and the date of Wappenbury Camp, Trans Birmingham Warwickshire Archaeol Soc 96 (for 1989-90), 1-3.
Booth, P, 1996 Warwickshire in the Roman period: a review of recent work, Trans Birmingham Warwickshire Archaeol Soc 100, 25-57.
Booth, P, and Evans, J, 2001 Roman Alcester: Northern extramural area excavations 1969-1988, Roman Alcester Ser Vol 3, Counc Brit Archaeol Res Rep 127, York.
Burnham, B C, 1995 Small towns: the British perspective, in A E Brown (ed), Roman Small Towns in Eastern England and Beyond, Oxbow Mono 52, 7-17.
Burnham, B C, and Wacher, J, 1990 The ‘small towns’ of Roman Britain, London.
Cameron, H, and Lucas, J, 1969 Tripontium: first interim report..., Trans Birmingham Warwickshire Archaeol Soc 83 (for 1966 and 1967), 130-179.
Cameron, H, and Lucas, J, 1973 Tripontium: second interim report..., Trans Birmingham Warwickshire Archaeol Soc 85 (for 1971-1973), 93-159.
Cracknell, S, 1989 Roman Alcester: recent archaeological excavations, Trans Birmingham Warwickshire Archaeol Soc 94 (for 1985-6), 1-62.
Cracknell, S, 1996 (ed) Roman Alcester: defences and defended area, Counc Brit Archaeol Res Rep 106.
Cracknell, S, and Hingley, R, 1994 Park Farm, Barford: excavation of a prehistoric settlement site, 1988, Trans Birmingham Warwickshire Archaeol Soc 98 (for 1993-94), 1-30.
Cracknell, S, and Mahany, C, 1994 (eds) Roman Alcester: Southern extramural area, 1964-1966 excavations part 2: finds and discussion, Counc Brit Archaeol Res Rep 97.
Cuttler, R, and Evans, J, 2000 A section through the Fosse Way and the excavation of Romano-British features at Princethorpe, Warwickshire, 1994, Trans Birmingham Warwickshire Archaeol Soc 102 (for 1998), 57-72.
Evans, J, 1994 Discussion of the pottery in the context of Roman Alcester, in Cracknell and Mahany (eds), 144-149.
Fieldhouse, W J, May, T, and Wellstood, F C, 1931 A Romano-British industrial settlement near Tiddington, Stratford-upon-Avon, Birmingham.
Ford, W J, 1971a Crewe Farm, Kenilworth, Warws, West Midlands Archaeol News Sheet 14, 14.
Ford, W J, 1971b Stretton-on-Fosse, Warwicks, West Midlands Archaeol News Sheet 14, 22.
Frere, S S, 1984 Roman Britain in 1983, sites explored, Britannia 15, 266-332.
Gardner, P J, Haldon, R and Malam, J, 1982 Prehistoric, Roman and medieval settlement at Stretton-on-Fosse: excavations and salvage 1971-76, Trans Birmingham Warwickshire Archaeol Soc 90 (for 1980), 1-35.
Henig, M, and Booth, P, 2000 Roman Oxfordshire, Alan Sutton, Stroud.
Hughes, H V, 1960 Recent work at Roman Alcester, Trans Birmingham Warwickshire Archaeol Soc 76 (for 1958), 10-18.
Jones, C, Eyre-Morgan, G, Palmer, S, and Palmer, N, 1997 Excavations in the outer enclosure of Boteler’s Castle, Oversley, Alcester, 1992-3, Trans Birmingham Warwickshire Archaeol Soc 101 (for 1997), 1-99.
Lucas, J, 1984 Tripontium, third interim report, Trans Birmingham Warwickshire Archaeol Soc 91 (for 1981), 25-54.
Magilton, J, 1980 The Coleshill Romano-Celtic temple: some reflections and new discoveries, West Midlands Archaeol 23, 27-39.
Mahany, C, 1994 (ed) Roman Alcester: southern extramural area, 1964-1966 excavations part 1: stratigraphy and structures, Counc Brit Archaeol Res Rep 96.
Moffett, L, and Ciaraldi, M, 2000 Charred plant remains, in Palmer 2000, 162-171.
Mudd, A, and Booth, P, 2001 Site of the former Hockley Chemical Works, Stratford Road, Alcester: excavations 1994, Trans Birmingham Warwickshire Archaeol Soc 104 (for 2000), 1-74.
Palmer, N, 1982 Roman Stratford: excavations at Tiddington 1980-81, Warwickshire Museum.
Palmer, N, 1983 Tiddington Roman settlement, West Midlands Archaeol 26, 36-46.
Palmer, S, 2000 Archaeological excavations in the Arrow Valley, Warwickshire, Trans Birmingham and Warwickshire Archaeol Soc 103 (for 1999).
Roberts, B K, 1978 The historical geography of moated homesteads: the Forset of Arden, Warwickshire, Trans Birmingham Warwickshire Archaeol Soc 88 (for 1976-77), 61-70.
Scott, K, 1984 Mancetter village: a first century fort, Trans Birmingham Warwickshire Archaeol Soc 91 (for 1981), 1-24.
Scott, K, 2000 Mancetter village: a 1st-century fortress continued, Trans Birmingham Warwickshire Archaeol Soc 102 (for 1998), 1-55.
Seaby, W, and Perry, G, 2000 Coins, in Scott 2000, 50-52.
Stanley, M, and Stanley, B, 1960 The defences of the Iron Age camp at Wappenbury, Warwickshire, Trans Birmingham Warwickshire Archaeol Soc 76 (for 1958), 1-9.
Thompson, I, 1982 Grog-tempered `Belgic' pottery of South-eastern England, Brit Archaeol Rep (Brit Ser) 108, Oxford.
Timby, J R, 1998 Excavations at Kingscote and Wycomb, Gloucestershire: A Roman Estate Centre and Small Town in the Gloucestershire Cotswolds with Notes on Related Settlements, Cirencester.
Van Arsdell, R D, 1994 The coinage of the Dobunni: money supply and coin circulation in Dobunnic territory, Oxford Univ Comm for Archaeol Monograph 38, Oxford.
Webster, G, 1974 The West Midlands in the Roman period; a brief survey, Trans Birmingham Warwickshire Archaeol Soc 86, 49-58.
Webster, G, and Hobley, B, 1965 Aerial reconnaissance over the Warwickshire Avon, Archaeol J 121 (for 1964), 1-22.