West Midlands Regional Research Framework for Archaeology, Seminar 3: Hodder1
From forts and roads to farms and kilns: the Roman period in Birmingham and the Black Country
Mike Hodder
Birmingham City Council
Introduction
Much of the evidence for the Roman period in Birmingham and the Black Country has been obtained from recent work including extensive excavations at Metchley fort and vicus in Edgbaston and post-excavation analysis and publication of previous work on this site, excavations on the line of the M6 Toll motorway in Sutton Coldfield, a newly-discovered farmstead in Kings Norton, the Birmingham Roman roads survey project, and fieldwalking in Sutton Coldfield and Lutley. The title reflects the types of sites investigated and the range of evidence now available.
Existing evidence: Forts and roads
Excavations at Metchley fort as part of new development have demonstrated that the fort was larger, had more phases of development and was occupied for a longer period than previously thought. The first fort on the site, established in the late 40s AD, was soon enlarged with annexes on three sides and had a short-lived vicus outside the fourth side. Subsequent modifications included a phase of structures including fenced enclosures inside the fort and funnelling ditches outside the west gate in the by then abandoned vicus which look more agricultural than military, suggesting its use as a supply base, followed by a further, smaller fort and at least one more phase of defences. Occupation of the site continued well into the second century.
The network of known Roman roads was of course a military network imposed on the existing landscape. Some of the Roman road lines in the area are followed by modern roads and others are visible as disused lines, most notably in Sutton Park, but it is difficult to trace the original lines of roads where they are in present built-up areas. Possible changes in alignment during the Roman period have been suggested, and at least those stretches of road still in use must have been used after their function as military communication lines had passed. There would have been many other trackways in use, often on more logical natural routes such as along the sandstone ridge running north from Birmingham to Sutton Coldfield and Lichfield. These may have been preferred by the local population to the better but less convenient roads, which might nonetheless have seen official traffic. How much did the imposed network influence the distribution and character of civilian settlements, and to what extent can road lines be used to predict the sites of these settlements?
Existing evidence: Farms and kilns
Much new evidence for civilian settlements and agricultural and industrial activity has emerged from recent excavations and fieldwalking projects. The vicus at the Metchley fort, a nucleated military-related settlement consisting of scattered timber buildings based around a road leading out of the fort’s west gate, is almost certainly atypical of other settlements in the area in terms of its form and location. Small excavations at Parsons Hill in Kings Norton, Birmingham, in the 1940s revealed gravel surfaces, possibly including part of Icknield Street, daub and charcoal and pottery dated to the 1st to 3rd centuries, but no recognisable structures. Only one kilometre away, and again close to the line of the Icknield Street, recent excavations at Longdales Road in Kings Norton have revealed a large triple-ditched enclosure over 80m across which contained pebble surfaces and a single circular house but no other recognisable structures. The bulk of the pottery from the site is 3rd to 4th century and predominantly Severn Valley ware, with a few earlier sherds. The size of the enclosure suggests that it may have served as a collection point for livestock from surrounding farms. The acid soil was unfortunately not conducive to bone survival. The excavation has revealed a ditched field system around it, probably at least one more enclosure, and pebble surfaces and ditches adjoining Icknield Street. The enclosure is in a prominent location and next to a farm which is known from documentary evidence to have been there in the 14th century.
There is a much greater quantity of pottery from this site, and from the Parsons Hill site considering the small area excavated, compared to known sites on the north of the conurbation. Did the proximity to a road make pottery, particularly Severn Valley ware, more easily obtainable or did it mean that the settlement was generally more Romanised because of the influence of traffic, including official traffic, to and from places like Alcester? Fieldwalking at Lutley, near Halesowen, has revealed three sites just over 1km apart, each indicated by large quantities of pottery(800, nearly 400 and 100+ respectively), predominantly Severn Valley ware. One site has a greater range of pottery than the others, potentially reflecting a higher status.
At Langley Mill Farm on the M6 Toll, adjacent to the Iron Age enclosure, there were four large irregularly-shaped ditched enclosures(40 to 70m across) of 2nd to 3rd century date, presumably for livestock and possibly for collection of stock from several farms, like the Longdales Road site. Some of the enclosures were conjoined and there were also probable ditched droveways. The only identifiable Roman structure was a small timber building, possibly a stockman’s or shepherd’s hut. The enclosures postdate a group of rectilinear ditched fields or paddocks. The Roman phase of this site was not identified in fieldwalking but there is a very thin scatter of Roman pottery in adjoining fields.
This site almost certainly saw continuous occupation from the Iron Age into the Roman period: the middle Iron Age enclosure followed by the field system which is potentially late Iron Age/early Roman with the actual farmhouse off-site, followed by the large enclosures, themselves representing successive phases of Roman activity. Curiously, the site lies on what was marginal land, heathland used as rough pasture, in the medieval and post-medieval periods, up to its enclosure in the 1820s.
Pottery from these sites shows that they were part of local market networks. The ditched enclosures and paddocks at Longdales Road and Langley Mill Farm and the funnelling ditches and fenced enclosures in one phase of the Metchley fort suggest a specialisation in livestock, possibly to meet the demands of places like Droitwich, Alcester and Wall for meat, hides, or fleeces.
Late 1st and early 2nd century pottery at the 2nd century kiln site in Perry Barr indicates earlier occupation of that site. Roman pottery and coins from Castle Old Fort indicate visits and at least temporary occupation of the hill fort in the Roman period. Small quantities of Roman pottery were found in excavations at Sandwell Priory, and there are antiquarian accounts of Roman pottery from Wolverhampton.
Fieldwalking in the Sutton Coldfield area has located concentrations of Roman pottery in small quantities which would be interpreted elsewhere as evidence for manuring with domestic debris, but because they can be seen in the wider context through extensive field survey they can be recognised as significant local concentrations. As was found in the Wroxeter Hinterland study, there seems to be relatively low use of pottery in these rural settlements with a rapid fall-off in pottery quantities with distance from the nearest town, in this case Wall. The two main concentrations, Over Green and Wigginshill, are just over 1km apart and each is close to a medieval farm in a medieval hamlet. It is of course possible that the Roman farm lies directly under the medieval site. If this is so, then the small quantity of pottery may be because the fieldwalking is in an area which is peripheral to the main settlement. The five sherds of Roman pottery from medieval contexts in two adjacent excavations in the Bull Ring in Birmingham city centre are, like the fieldwalking finds from Sutton Coldfield, a small but locally significant quantity and hint at a farmstead here.
The settlement site indicated by the surface pottery concentration at Over Green lies at one corner of a rectilinear field system at the Burrells in Sutton Coldfield, much of which survives as actual field boundaries and may be Roman or earlier in origin. It is similar to demonstrably Roman or pre-Roman field systems elsewhere in the country. Assarting documented on this land in the 1240s is unlikely to have resulted in such a regular field system. It is more likely that the medieval assarters reused existing boundaries in land that had been abandoned after the Roman period. The alignment of the field system continues as earthworks in adjoining fields is followed by the alignment of an open field system further south.
Chance finds of coins in the conurbation are numerically significant and therefore demand consideration. In the Birmingham SMR, 92 out of 150 Roman records are coins, mainly single coins, and there are four hoards, not all verifiable. The proportion is similar in the Black Country SMR. How do these finds relate to the location or pattern of settlements? They are not all ancient losses. Were the hoards buried near settlements?
Another chance find, a metalwork hoard from Pedmore Hall near Stourbridge, consisted of jewellery including finger rings, toe rings, armlets, necklaces and coins of 2nd century date and suggests a high-status owner.
Pottery kilns have been found at Perry Barr and Sutton Coldfield, both of them 2nd century and both chance discoveries. Both sites made tankards of Severn Valley Ware form. At Sutton Coldfield, carinated bowls were also made. The Sutton Coldfield kiln was not near an obvious clay source but proximity to fuel was probably more important. It is near Hill Wood, an extensive medieval and post-medieval coppiced woodland. The products met local markets: Sutton products have been identified at Coleshill. Neither site is on a known road line. The Perry Barr site is just under 1 km from Icknield Street and the Sutton site is over 2km from Icknield Street, but the latter is near a medieval road line which follows a natural ridge and could have been in use much earlier. The Sutton Coldfield kiln site, excavated in 1987, is as yet unpublished.
Limestone cobbles and lime used for mortar at Wall is Wenlock limestone, which outcrops at Walsall. Whilst Roman exploitation of this resource is not unlikely, there is as yet little firm evidence.
Environmental evidence
The environmental evidence is direct and indirect, ie what can be inferred from the archaeological evidence. Pollen from the fort ditches at Metchley shows woodland regeneration in the area in the 2nd century. The buried soil under the Roman road in Sutton Park has a podzol profile, suggesting heathland or possibly woodland in the mid 1st century. Large quantities of timber would have been required for building the various phases of the Roman fort and it was presumably obtained by clearing substantial areas of local woodland. Wood for wattlework in the fort’s buildings could have come from coppice and the fuel for the pottery kilns is similarly likely to have been from managed woodland. The ditched enclosures and paddocks at Longdales Road and the M6 Toll suggest livestock and therefore grazing land. Similarly the funnelling ditches in the vicus and fencing in the fort itself in one of the Metchley phases are also potentially related to livestock. If cattle were particularly important in this area in the Roman period then there may have been extensive areas of grassland.
Sub-Roman: 5th century and beyond:
There is little Anglo-Saxon archaeological evidence from the area, but the proximity of known Roman sites, either excavated or indicated by pottery concentrations, to medieval sites still in occupation is particularly noticeable. This may indicate the selection of the same site at different periods for the same reasons, or continuous occupation of the same site. This might mean that the medieval farm adjoins the Roman farm site, like a post-medieval house outside but adjoining a medieval moat, or the later farm is on exactly the same site as the Roman farm. If the latter were the case, it would of course mean that the Roman farm is directly under existing farm buildings and therefore difficult to locate. If there were continuity it means that these Roman sites are the successful ones, on good locations, in contrast to those site now visible as cropmarks which have been abandoned, were therefore failures and are unrepresentative. At the Burrells field system, abandonment or change of use at the end of the Roman period is indicated by medieval assarting of the area.
Curatorial issues
The existing evidence indicates a potentially large number of settlement sites, as little as 1km apart from each other. In the north of the area these may have used relatively little pottery, making them difficult to locate. Some sites are close to known Roman roads but others are some distance away from them. The proximity of Roman settlement sites to medieval sites might make the former difficult to detect, but also highlights the need to be aware that sites known to have been in existence in the medieval period may be on the same site as their Roman predecessors. Open spaces within the conurbation are important particularly important for the survival of sites and landscapes of the Roman period, as for other periods. The significance of chance finds of coins and other metalwork is unclear and needs to be assessed.
Future research
As noted for the prehistoric periods, assessment of existing air photos is required, and more flying of hitherto unflown areas should be undertaken. Open spaces within the conurbation are particularly important for survival of archaeological remains of this period and appropriate evaluation should always be required as part of consideration of development proposals affecting them.
Fieldwalking should be undertaken of all available areas. Roman sites should be detectable in fieldwalking(although the Langley Mill Farm site wasn’t). Fieldwalking methods need to take account of the small quantity of pottery indicating Roman sites in the north of the area; widely spaced traverses are not appropriate. There is an opportunity for local groups to get undertake this, and those involved need to be able to identify Roman pottery.
The existing evidence demonstrates the potentially large number of Roman settlement sites in the area. Where they lie within the present built-up part of the conurbation, they obviously cannot be located by fieldwalking but may be found by increasing public awareness. Much Roman pottery has probably been found in suburban gardens but has not been recognised as old by its finders because of the large quantities of modern debris and the expectation that everything is recent. Both of the known pottery kiln sites in the area were chance discoveries in gardens.
Many medieval sites in the area could be on or adjacent to their Roman predecessors. This needs to be borne in mind when work is undertaken on the medieval sites.
More work could be done on the interpretation of chance finds of coins. This would include examining the circumstances of finding; assessing the likelihood that particular finds are recent losses, from the circumstances of discovery or the coin type; and seeing whether the loss pattern by coin date across the region as a whole is consistent with that in the rest of the country.
Full analysis and publication of the existing data is needed. Post-excavation work on the recent excavations at Metchley fort and on the M6 Toll motorway is in progress, but the Sutton Coldfield pottery kiln remains a major post-excavation backlog with no immediate prospect of funding to undertake it. It is also important to ensure analysis and publication of assemblages found in fieldwalking.