Introduction
Trial-trench evaluation and subsequent excavation were undertaken by Archaeology South-East in 2016 at the ‘Land East of Kings Warren’ site, Red Lodge, Suffolk (TL 7073 7034), in advance of residential development (Fig. 1). Prior to this, the 8.93 ha site was under arable cultivation with recent residential development encroaching to the west, occupying gently sloping land falling south to north from c. 23 m to c. 21 m A.O.D. The bedrock geology comprised Holywell Nodular Chalk and New Pit Chalk Pit formation, with no superficial deposits. Whilst encountered remains comprised material from the Early Neolithic through to the medieval period, only the Bronze Age landscape and its subsequent reuse in the Roman period are considered here; an account of other remains is provided in the final archive report.Footnote 1
Site location.
Fig. 1 Long description
Map detailing the archaeological investigation areas at Red Lodge, Suffolk. The main site is highlighted, with a specific excavation area marked in red. The map includes a scale bar indicating distances in meters and a north arrow for orientation. A smaller inset map shows the location of Red Lodge within a broader region of England, with major roads like the A11 labeled. The site is divided into sections, with areas designated for archaeological investigation clearly outlined.
A barrow was recorded within the site area (SHER: FRK008); although largely reduced by ploughing, it was noted in 2007 to survive to a height of c. 0.5 m. A geophysical survey conducted in 2013 had detected a corresponding ring-ditch and a larger rectangular ditched enclosure around it (Fig. 2).Footnote 2
Geophysical survey plot (Cranfield University 2013).
Fig. 2 Long description
Geophysical survey map depicting a Roman boundary ditch and various features within the surveyed area. The map includes a prominent ring-ditch labeled ′Ring-ditch′ at the center. Surrounding features are marked as ′Well′, ′Shrine′, ′Pond′ and ′Pit′. The survey area is outlined with a boundary and the unit of excavation is indicated. A scale bar shows a distance of 20 meters and a north directional arrow is present. The map provides detailed spatial information about archaeological features detected through geophysical methods.
The site lies in a landscape widely represented during the Bronze Age period, with several barrows, likely of this date, known to the east and south-east, in addition to that within the site. Bronze Age pits have also been encountered in the previously evaluated areas outside of the development area, and to the north-east of the site. There is little evidence for Roman or Early Medieval (Anglo-Saxon) remains being present in the vicinity, though there are oblique references to illicit metal-detecting of metal objects of reputed Iron Age, Roman and Anglo-Saxon date — including within the site itself. Prior trial-trench evaluation of a c. 300 ha development area to the immediate north and west found no significant archaeological remains predating the post-medieval period (Fig. 1).Footnote 3 The barrow is identified by the Suffolk HER as potentially having been later utilised as a gallows during the thirteenth century, being mentioned in a convention of 1238 between the Bishop of Rochester and the Abbot of St Edmunds regarding the location of the structure on the boundary of their Freckenham and Herringswell lands (HER: FRK146); Red Lodge was formerly part of Freckenham parish and the barrow now lies c. 100 m from the parish boundary.
The site
The earliest recorded land use remains were pits and a possible utilised tree-throw or hollow of earlier Neolithic date. No continued occupation into the Bronze Age is evident and this earlier prehistoric activity is not considered further in this article. The following describes the origins of the religious significance of this location in the Bronze Age funerary monument constructed here, focuses on its later incorporation into the Roman rural shrine complex and briefly considers its final use as a possible gallows site — all of which demonstrate the perpetuation of the specialness of this isolated or liminal place, albeit under changing circumstances.
The Bronze Age barrow
The Bronze Age land use is perceived to have been entirely of a sacred/funerary nature, with the vicinity dominated by a large round barrow, as denoted by the remains of a ring-ditch with two small pits in close proximity to its west (Fig. 3). A general lack of other contemporary features, both within the site and in the adjacent investigation areas, suggests the surrounding landscape was kept deliberately clear in order to enhance the prominence, and respect the sanctity, of the monument. The 45 m diameter ring-ditch (35–38 m internally) is presumed to have had a central mound, removed by later agricultural cultivation, though no trace of this survived. Where excavated, the ditch was 2.90–6.20 m wide and up to 1.80 m deep, with lower fills of compacted chalk and yellow silty sand suggestive of natural silting and slumping events. At least some of these may have derived from the posited earthwork mound, though their truncation due to later re-cutting of the ring-ditch made this difficult to ascertain with any confidence. Few diagnostic finds were recovered from these deposits, Optically Stimulated Luminescence (OSL) dating of the lower ring-ditch fill instead producing a date range of 2200–1600 b.c. (GL16085–6)Footnote 4 and confirming the monument’s Early Bronze Age origin. The ring-ditch occupied an area of slightly higher ground in the local topography, which was likely a deliberate action to accentuate the prominence of the barrow monument in its surrounding, relatively flat landscape.
Site plan.
Fig. 3 Long description
Site plan depicting features from the Early Roman Age and Roman period. The map includes a large ring ditch centrally located, marked in yellow, indicating Early Bronze Age. Roman features are shown with purple hatching, including a shrine labeled G41, a pond labeled G34/35 and a well labeled G14. Additional labeled points include BA pits, G15, G16, G17, G18, G19, G20, G21, G22, G23 and G24. The map also shows a medieval grave and the limit of excavation. Roman boundary ditches and geophysical survey results are indicated with dashed lines. A scale bar and north arrow are present for orientation.
Similar ring-ditch examples have been recorded elsewhere in Suffolk at Sicklesmere Road, Bury St Edmunds,Footnote 5 Upthorpe Road, Stanton,Footnote 6 Flixton Park QuarryFootnote 7 and two at Liberty Village, Eriswell.Footnote 8 Though these are generally smaller than the Red Lodge ring-ditch, measuring c. 23–30 m, larger barrows measuring c. 32 m and c. 41 m have also been excavated at Aldham Mill Hill, Hadleigh,Footnote 9 and one of the barrows at Flixton Quarry was projected as having a c. 39 m diameter.Footnote 10 While there was no evidence of any associated burial at Red Lodge, it would likely have been placed either at ground level or within the perceived mound and has therefore been subsequently removed by later agricultural activity. While the size of the ring-ditch could be suggestive of an enclosure, with such ring-ditches at Mucking having a c. 40 m diameter and Springfield Lyons measuring c. 53 m,Footnote 11 the lack of any entranceway or internal features imply that it more likely functioned as a barrow.
Other than the ring-ditch remains, Bronze Age features were limited to two pits to the west of the barrow. These may have acted as markers or formed a small adjacent structure. The paucity of other features is thought to be deliberate, to accentuate the presence of the barrow in the landscape.
During the Iron Age it is likely that the barrow was retained, or at least recognised as a conspicuous earthwork, with a quarter of an Iron Age jar being recovered from the upper fill of the ring-ditch. The perceived later re-cutting of the ring-ditch during the Roman period means it is difficult to know if this was residual in the wider landscape or is evidence of Iron Age deposition. The lack of other Iron Age features in the surrounding vicinity of the ring-ditch suggests that it was still a prominent feature on the landscape, if not utilised as such, then at least respected as an area of historic importance.
The Roman religious complex
The main period of site use was during the second to later fourth century, when a religious focus was founded around the barrow remains (Fig. 3). Very broadly, the development of this rural ritual complex divides into two phases of activity, with the creation of a temenos enclosure in the earlier Roman period and the addition of a shrine building in the later. In addition to these principal developments, a range of depositional events demonstrate a continuum of ritual activity throughout this period, up to a cessation in the later fourth century a.d.
Earlier Roman
The Roman use of this location seemingly began with the construction of a rectilinear enclosure or temenos around the prehistoric barrow monument, probably in the earlier second century a.d. (Ditches G1, G2, G4 and G5). The enclosed area is perceived to have measured 200 m east–west and 160 m north–south and, while the northern and eastern boundaries were not within the excavation area, the northern boundary was defined by the preceding geophysical survey, with the eastern edge postulated to lie just outside the eastern extent of the excavation area, on the route of the existing access road (Fig. 3). The enclosure did not contain any obviously domestic settlement features, though minor internal boundaries and isolated pits and post-holes, including a possible well and structured animal bone deposits, are thought to be part of the ritual use of the temenos interior. This paucity of remains suggests that the remains of the barrow monument was the focus of this early Roman activity.
The western boundary was formed of two ditches (G2 and G4) aligned NNW–SSE. The boundary originated with Ditch G4, which was fairly shallow, measuring 0.90 m wide and 0.60 m in depth and with a single fill of mid greyish-brown silty sand, with no dating material. The ditch was on the same alignment, and cut by, Roman Ditch G2, which was larger, measuring 1.40–3.50 m wide and 0.56–0.97 m deep. It contained four sherds of Roman pottery and flecks of ceramic building material (referred to as CBM hereafter). The increased size over its precursor perhaps indicated an increase in size or prominence of the religious precinct, with the recut ditch postulated to link with the northern boundary as detected by the geophysical survey (Fig. 2). Internal boundary Ditch G1 lies to the east of the G2 and G4 ditches and is thought to be indicative of an entranceway to the centre of the temenos, with a c. 57 m wide gap between the ditches (see below). The southern boundary of the enclosure (Ditch G5) was aligned ENE–WSW and was only partially visible, continuing off the southern edge of the excavation area, though postulated to have joined with the G2 ditch forming the south-western corner of the enclosure. The ditch was 2.75 m at its widest point, and its maximum depth was 0.60 m. It contained a single orange-brown silty sand fill with sparse finds including two sherds of a Roman jar (c. a.d. 100–325). The ditch was truncated by a modern irrigation pipe which was on the same alignment, and, along with its location on the edge of the excavation area, prevented the full width being excavated.
Within the enclosure were four smaller ditches or gullies (Ditches G6–9) interpreted to represent internal boundaries. Gully G6 ran parallel with the southern temenos boundary ditch G5, just to its north. The gully ran for 35.17 m and was 0.62 m wide and 0.37 m deep, with a single sherd of second-century samian pottery recovered from the fill. While the gully could indicate an internal boundary, it may instead represent the earliest manifestation of the enclosure, later replaced by larger Ditch G5. Both terminal ends were within the excavation area suggesting the full extent of the ditch was exposed, though its shallow nature means it may have formed a larger segmented boundary across the area; however, no evidence of this was forthcoming during excavation. At the western end of the gully was rounded pit G45, which measured 3.0 m by 2.20 m and 0.47 m deep, and contained three fills of mid to dark grey-brown silty sand. The near-complete skeleton of a cockerel was found in the upper fills, along with a cattle scapula with hook-hanging damage, shed deer antler, three iron nails and a small quantity of Roman pottery. The cockerel remains have been radiocarbon dated to 126–322 cal. a.d. (95.4% probability, SUERC-82635), which correlates with the establishment of the temenos enclosure and the subsequent shrine building. Cockerel remains were frequently used for funerary rites and sacrificial offerings during the Roman period,Footnote 12 with domestic fowl remains recovered in significant quantities from the temple sites such as Uley, Brigstock and Folly Lane.Footnote 13 The pit location, at the terminus of the ditch, suggests that it was a structured deposit placed at time of the enclosure inception.
North–south ditches (G7 and G8) lay within the enclosed area and are thought to form an interrupted internal subdivision of its eastern part, some 21 m long. Its southern segment (Ditch G7) was slightly narrower, measuring 1.1 m wide and 0.43 m deep, while the northern segment (Ditch G8) was up to 2.1 m wide and 0.82 m deep. Roman pottery was recovered from both, with the material generally suggestive of an early Roman date; a later Roman coin (a.d. 268–293) recovered from Ditch G7, is probably intrusive and derived from the more intense later activity in this part of the temenos. A first- to third-century coin was recovered from Ditch G8, with a polished Neolithic flint axe head retrieved from its northern terminus, the latter constituting the deliberate deposition of a curated object.
While the perceived northern boundary of the temenos is thought to lie north of the excavation area, as shown by the geophysical survey, a smaller east–west aligned gully lay along its northern edge. Gully G9 was only partly exposed and had shallow sloping sides, measuring 0.90 m wide and 0.29 m deep. A small number of pottery sherds of broad Roman date were recovered, though this boundary is thought to have originated during the early Roman period. The ditch is postulated to have formed a lesser (earlier?) boundary to the north of the ring-ditch, perhaps indicating the initial northern edge of the temenos that later was enlarged in the Late Roman period, but more likely denoting a contemporary subdivision. A post-hole was cut into its eastern end, which may be indicative of a fence structure along it, though the similarity of fills made it difficult to establish if further posts were present along its exposed length.
To the west of the G9 gully was a small north-east–south-west ditch (G1) which terminated within the site area. The ditch contained a single fragment of wall plaster along with undiagnostic Roman pottery sherds and bone fragments. A pit (G22) was located immediately adjacent to the ditch terminus and is likely contemporary, as it was indistinguishable in plan from the ditch and contained comparable fills. A glass bead of Late Iron Age or early Roman type was recovered from the pit, along with pottery of broad third-century date that defines its disuse. The bead may indicate a deliberate deposition, potentially marking the end of use of the ditch. The ditch location forms a c. 57 m wide gap from the western boundary of the enclosure. It could therefore have functioned as a boundary forming part of the approach into the centre of the temenos, defined by the barrow at its centre. A line of nine circular undated post-holes (G13) was located adjacent to the ditch terminus. They were evenly spaced and likely formed a structure such as a fence line, extending the G1 ditch alignment as part of an internal boundary or entrance.
Features within the enclosure were few in number (Fig. 3) and it is posited that this sacred space was kept deliberately clear, with the focus on the Bronze Age barrow. While no diagnostically early/mid-Roman pottery was recovered from its ring-ditch, it is likely to have been recut and maintained on its incorporation into the religious complex; while accumulated fills of this date are likely to have been removed by its final later Roman re-cutting.
The earlier Roman features occupying the enclosure were primarily pits and post-holes, only a few of which had a clear function; these are thought to have had a use related to that of the religious sanctuary. Pit G18, located immediately adjacent to the western boundary of the temenos and to the east of the enclosure ditch G4 terminus, contained a complete articulated dog skeleton accompanied by a horse scapula at its base. The dog remains have been radiocarbon dated to 88–251 cal. a.d. (95.4% probability, SUERC-82636). Dog burials were often deposited in auspicious places during the Roman period, as at Castle Hill, Cambridge, where they were recovered from under the floor surface of a possible shrine and within several later Roman shafts;Footnote 14 the complexity of these deposits is noteworthy.Footnote 15 Dog remains at Greyhound Yard, Dorchester, comprised multiple skeletons associated with joints of meat, as perhaps the Red Lodge example may have been. The Dorchester deposits have been posited to be the product of rituals relating to the foundation of the new town.Footnote 16 The placing of dog burials in significant locations is also apparent in the Late Iron Age, with two dog burials at Hallaton in an oblique slot which regulated access to the inner part of the shrine.Footnote 17
A probable well (G14) was located to the west of the ring-ditch, between the temenos entrance and the ring-ditch remains. The large shaft-like feature measured 3.80 m by 3.10 m and was excavated by hand and machine to a depth of 2 m without its base being reached. Pronounced slumping of lower unexcavated fills was/is suggestive of a significant further depth and of voids below this. It had steep, near-vertical sides, with four fills identified. A small quantity of mid-second- to third-century pottery, animal bone and CBM was recovered, none of which is distinctive of structured or votive deposition. If such deposits had been placed within the shaft, they were present at a greater depth, with deposits within wells being a well-established phenomenon. A Roman well at Exning was 16 m deepFootnote 18 and another at Rothwell Haigh, Leeds, 12.3 m deep,Footnote 19 both containing significant animal bone remains, while a 4.5 m deep shaft at Springhead contained pottery, dog and cow remains, and a human skull.Footnote 20 The Rothwell Haigh well remains comprised 19 individual dog remains, as well as up to 71 sheep and 38 goats, while the Exning well included sheep and pig remains at a depth in excess of 2 m. The location of well-like feature (G14), midway between the temenos entrance and reused barrow, suggests that it had a significant role in the functioning and rituals of the religious sanctuary, perhaps including that of cleansing.
The other discrete features within the enclosure generally had no clear use or contained conspicuous concentrations of artefacts, though their locations in relation to the enclosing ditch and the barrow hint that they were part of its ritual activities. While some were poorly dated within the Roman period, two adjacent pits (G15), were purposefully positioned near the entrance; notably, one contained a significant quantity of oyster shell (c. 2 kg) along with some pig canines. While this is not necessarily demonstrable of a feasting event, the general lack of oyster shell from elsewhere within the sanctuary suggests deliberate ritual deposition such as at Chanctonbury Ring.Footnote 21 The association of pigs with the later shrine may imply that the pig canines are a precursor to the head-and-hoof deposits, indicating a prolonged association of the site with pig remains.
Later Roman
The religious sanctuary continued in use into the later Roman period, with the focus retained on the Bronze Age barrow. Its ring-ditch was probably recut during this period, if not in the earlier Roman phase of occupation, and certainly became infilled at this time. The recut had a broadly V-shaped profile and measured 2.40–6.45 m wide and 0.69–1.55 m deep, being shallowest at its north-east. The upper fill was a distinct dark grey-brown silty sand that contained exclusively later Roman pottery (a.d. 250–325) and CBM, with the lower fill of yellow- and orange-brown silty sand generally containing no finds, though a small number of residual flints and pottery sherds were recovered. Eleven coins, a copper-alloy brooch (Fig. 4.1), two bracelets, a spoon and tweezers (Fig. 4.2) were also recovered from the upper fill of the ring-ditch. One of the bracelets is considered to show signs of deliberate breakage and all are likely to be part of the wider depositional activity within the religious complex. Animal bone and antler were also retrieved from three of the excavated segments, with a single fragment of human bone recovered from a segment in the east of the ring-ditch. The majority of finds from the ring-ditch were recovered from its eastern side, suggesting a focus of activity in this area, conceivably associated with the adjacent shrine building (see below).
Illustrations of the pertinent finds. (1) White metal coated copper-alloy Colchester Derivative brooch; (2) copper-alloy tweezers; (3) complete iron ox goad; (4) copper-alloy plaque fragment; (5) copper-alloy votive leaf plaque fragment.
The western part of the temenos ditch was again recut and extended to the north-west (G3), and the smaller internal divisions appear to have gone out of use, perhaps indicating an enlargement and/or simplification of the enclosure during this period. The recut of the western boundary was up to 3.2 m wide and decreased in depth to 0.82 m towards its northern end. Finds were sparse along the ditch length with undiagnostic Roman pottery (31 sherds), a coin dated a.d. 275–285, CBM and a single pig canine recovered from the excavated segments. A stone quern fragment and single pig canine were recovered from two segments which may indicate potential deliberate deposition, though the limited quantity makes it difficult to ascertain if these are deliberate depositions or incidental inclusions within the fill. While there is no evidence that the southern boundary of the enclosure was remodelled, it is thought to have continued in use, and it is assumed that the entire temenos remained in use into the Late Roman period.
Other than the addition of the shrine (see below), the sacred space surrounding the barrow appears to have continued to be kept clear into the later Roman period; the few features of identified later date are all interpreted to be associated with its religious functioning. To the west of the ring-ditch was a roughly rectangular pit (G19), measuring 1.0 m by 0.42 m and 0.15 m deep in which the remains of a near-complete cow skeleton were recorded. The animal was c. 12–30 months old and showed no evidence of butchery or consumption. The skeleton has been radiocarbon dated to 238–384 cal. a.d. (95.4% probability, SUERC-82633). The location of this animal deposit, on the opposite side of the ring-ditch to the shrine, is thought to be deliberate. It is probable that it represents a single event, with the later Roman date showing that it was deposited at some point during the shrine’s use.
In the south of the excavation area was a large irregular-shaped feature (G35), either a utilised natural hollow or pond, which measured 15.9 m by 12.04 m and up to 0.80 m deep. A roughly circular pit (G34) was later dug into the centre of the infilled pond that contained diagnostic pottery sherds of Late Roman date. The upper fills of the two features were indistinguishable in plan, with five coins recovered from the surface, four of which date between a.d. 260 and 296, with the other having a later date of a.d. 330–340. Further finds recovered included a votive plaque (Fig. 4.5), brooch fragment and ox goad (Fig. 4.3). The copper-alloy leaf plaque had been folded along its centre, showing a purposeful alteration of the object prior to deposition. This act of decommissioning an object by folding, prior to deposition, is seen at the Great Chesterford temple, where at least eleven votive leaves had similarly been folded, though, similarly, no deities could be identified.Footnote 22 Iron nails recovered from the fills and surface of the feature may indicate that wooden objects had also been deposited here.
To the north of the ring-ditch was a flue-like structure (G43), its surviving remains measuring 2.85 m long by 0.56 m wide and 0.22 m deep. The flue was aligned east–west and had evidently been tile-lined, with near-complete tegulae remaining in situ along its southern side. Five post-holes were located around the flue, with three consistent in size and shape; while no clear structure is discernible, these were likely associated with the flue use. The flue may be the below-ground remains of an oven associated with food production, for the ritual burning of foodstuffs as part of the deposition, though no charred seed remains or environmental evidence was recovered from the environmental samples. The tegulae lining the flue are contemporary with those found associated with the shrine building and likely suggest that it was constructed at the same time.
The shrine
The principal focus of activity was seemingly located to the immediate east of the barrow ring-ditch, in the form of a post-built structure interpreted as a shrine. The building was constructed on several localised deposits of mid-yellow and mid-brown silty sand redeposited natural which contained Roman pottery and CBM that are identified as levelling deposits laid prior to its construction. Dating of these deposits varies, with the upper layers producing pottery assemblages of a.d. 250–300, while the lower deposits contained pottery of a.d. 120–250 and four coins, the latest of which dates to a.d. 141. While the artefacts recovered from the lower layers might simply be residual here, and/or the coins curated and deliberately deposited as a foundation offering in the second half of the third century, it is possible that they hint instead at a precursor of the later Roman shrine structure.
The shrine structure itself (Figs 5 and 11, below) comprised ten post-holes defining a roughly rectangular outline, measuring c. 5 m north–south by c. 4.5 m east–west. This likely constitutes a six- or eight-post structure, with additional and intercutting post-holes being replacements or repairs. The post-holes varied in size, though all had steep/vertical sides and flat bases. All contained mid-orange-brown silty sand fills, with two of the corner post-holes also containing surviving post-pipes. A bulk soil sample from post-hole [1484] contained well-preserved oak charcoal and charred cereals, which have been radiocarbon dated to a.d. 19–176 (95.4% probability, SUERC-84345 and 84346). This may indicate that the origins of the building lie in the earlier Roman phase, with the later Roman material recovered from the underlying levelling deposits a result of the later disuse of the structure. Roman pottery fragments, CBM and iron nails were recovered from five of the post-holes, with wall plaster likely from the disuse of the building also retrieved from one.
Schematic plan of shrine.
Fig. 5 Long description
Schematic plan of a shrine with a rectangular outline defined by post holes. The outline measures approximately 5 meters north-south by 4.5 meters east-west. The post holes are marked around the perimeter, indicating a possible structure with a column in the corner. Votive deposits are located to the left, marked with stars indicating head and hoof deposits. A screen is positioned on the left side and a postulated altar or lararium is noted on the right. The area within the outline is labeled as a gravel surface. Surrounding the shrine is a chalk path, indicated by diagonal hatching. A scale bar at the bottom right shows measurements up to 2 meters and a north arrow is present at the top right for orientation.
The post-holes were potentially cut into a roughly rectangular, 0.10–0.30 m thick chalk deposit, which measured 7.0 m north–south and 8.20 m east–west and therefore extended beyond the walls of the shrine structure. This chalk layer was recorded in the field to be cut by the post-holes and to represent a floor-surface preparation, though the sporadic, truncated nature of the chalk deposit means that it could alternatively have been deposited after the wall posts were in situ. The chalk clearly defined an internal gravel surface, which may indicate that it functioned as an external walkway used to accentuate the building location. The absence of any material overlying the chalk implies it had been formed solely of redeposited chalk, and not acted as a base level for further overlying building deposits — constituting an external walkway or surface as well as an interior floor. As an external feature, it would make sense to have been placed after the construction of the shrine and placed around the post-holes forming the structure. In addition to its gravel floor, the shrine contained two settings for possible columns, in its south-east corner. As the only internal features, these are speculated to mark the position of an altar. The columns were c. 1.15 m apart and located on a north-east–south-west diagonal in the south-east corner of the building, speculated to have had a small shelf or ledge adjoining the two and being the interior focus of activity. Due to the size of the shrine building and possible altar it is likely that it would have been similar to a household lararium in form.
The footprint of the shrine suggests that it was single-storied and constituted a single room or cella, a gravel layer forming an internal floor surface. There was no clearly defined entranceway to the building, with it speculated that it could be to the north of the structure, due to the smaller post-holes here, perhaps indicating that they were less weight-bearing and provided an access point. The location to the north would mean anyone entering the building would have had the altar directly in front of them on entry. The lack of building stone on the site implies that the walls were wholly of wattle-and-daub construction. A significant quantity of painted wall plaster (c. 48 kg) was recovered from features and deposits overlying and around the shrine and are highly likely to have derived from its fabric (Fig. 6), potentially embellishing both the inside and outside of the structure, with some fragments bearing linear lath impressions on their back. Collectively, they attest to decorative schemes that included large areas or panels of colour, primarily white and red, with elements in pink and green also recovered. Similarly, the quantity of un-abraded roof tile fragments present in the immediate vicinity of the shrine indicates that the structure had a tile roof, with 111 kg (representing 76% of the total CBM from across the site) including near-complete tegulae collected during the excavation. Though some of this could perhaps have derived from the nearby oven/flue, the concentration around the shrine implies a majority originated from the shrine roof.
Painted wall plaster.
To the immediate west of the shrine was a series of five irregular pits forming a north–south line (Fig. 5). The two southernmost pits contained structured animal deposits in the form of pig skulls accompanied by articulated hindlimbs and forelimbs — so-termed ‘head-and-hoof’ deposits (Figs 7 and 8). All the skulls had been deliberately placed facing east, towards the shrine. The southernmost pit contained two head-and-hoof deposits, both of which showed signs of being disturbed or compressed (Figs 7 and 8). Skull [1425], from a juvenile pig aged 2–7 months that was slaughtered in the winter of its first year, returned a radiocarbon date of 142–337 cal. a.d. (95.4% probability, SUERC-82646). The pit to the immediate north contained a further four pig head-and-hoof deposits placed within a homogeneous mid-red-brown sandy silt fill. Three were placed towards the top of the fill, with two including in situ coins placed on the cranial vault, a sestertius of Antoninus Pius (a.d. 156/7) and a dupondius of Faustina (a.d. 142). A further skull underlay these and, while it did not have a placed coin, it displayed green copper staining on its frontal bone indicative of its former presence. This lower skull deposit [1521] has been radiocarbon dated to 84–229 cal. a.d. (95.4% probability, SUERC-82644), while skull [1429] from the upper part of the fill produced a 252–400 cal. a.d. date range (95.4% probability, SUERC-82645). This upper skull included the placed coin of Antoninus Pius which was heavily worn, suggesting it had been in circulation for some time before deposition, with such coins known to have remained in circulation in the early third century. Four further coins were recovered from the fill surrounding these pig deposits, the latest being dated a.d. 345–50. The layered nature of deposition, irregular shape of the pits and varying dates of the skull deposits and coins appear to indicate that the pit was revisited/reopened multiple times in order to place additional head-and-hoof deposits. The pig remains were predominantly of young adult males that were culled before reaching skeletal maturity and were most similar in size to Iron Age pigs.Footnote 23 The condition of the bones, as well as the presence of dentition, indicates that deposition occurred at or soon after the time of death, with a single example containing butchery marks indicating that the animal was decapitated.
Head-and-hoof deposit.
Head-and-hoof deposit (detail).
Directly to the north of the head-and-hoof deposits were three further pits positioned on the same north–south alignment. These contained noticeably different fills of very dark blackish-brown silty sand with common CBM and mortar fragments, the latter including backing mortar for wall plaster. The fills appear to incorporate debris material deriving from the adjacent shrine, and it is speculated that the pits were either cut after the building had passed out of use, or, more probably, represent the robbing/removal of some of the head-and-hoof deposits.
Enclosing the five pits and their structured deposits, and presumably segregating them from the barrow ring-ditch, was an L-shaped ditch-like feature (see Fig. 5: the ‘screen’), which lay parallel to the shrine building and measured c. 7.5 m in length. Its fill contained significant quantities of painted wall plaster and CBM fragments, posited to derive from the disuse and demolition of the adjacent shrine. A possible in situ tile fragment with attached wall plaster was located c. 5 m to the west of the shrine building and within the L-shaped feature. To the east of the tile fragment was a linear, heavily degraded piece of timber and associated iron nails. Combined with the tile, this could have formed a further screen or barrier to the west of the shrine, between the building and associated head-and-hoof deposits and the barrow.
Beyond the shrine itself and the structured animal deposits both alongside and within the wider sacred precinct, clear instances of the votive deposition of objects are less conspicuous and less easily identified — those in well or shaft G14 perhaps being a notable exception. However, the general lack of material culture present within many of the excavated features suggests that much of the material found in debris-rich layers both surrounding and overlying the shrine constitute objects that had been deliberately deposited within the temenos, perhaps after having been first utilised or placed within the shrine building itself. Twenty-seven of the fifty-two coins recovered were in its immediate vicinity, with a further eleven coming from the upper fill of the adjacent ring-ditch; collectively, these had a date range of c. a.d. 250–350+. Two copper-alloy pins, two brooches and a pair of tweezers were also found in the eastern side of the ring-ditch, adjacent to the shrine. A fragment of a second votive plaque (Fig. 4.4) was retrieved from a debris deposit to the immediate north of the shrine that is similar to those recovered from Kelvedon,Footnote 24 where they were thought to be remains of a border to an inscription. Although relatively modest in quantity, the distribution of these coins and items of personal adornment in demolition deposits overlying the shrine building remains, and in its near vicinity, shows a clear concentration. Items of personal adornment deposited as part of a suite of ritual offerings is not unusual, with deposited jewellery recovered along with small tools and sheep mandibles from within the temple precinct at Heybridge,Footnote 25 for example. Deposition of coins are also often dominant at religious sites, as is the case at the large temple site at Uley, Gloucestershire.Footnote 26 A Neolithic polished flint axe head was recovered from immediately adjacent to the shrine. While it may have been brought to the site in the earlier Bronze Age in conjunction with the construction and use of the barrow, it is more likely a Roman acquisition and deposition, along with the other Neolithic axe head found to the immediate east in internal subdivision ditch G8. Polished axes are known to have been valued during the Roman period, with objects often treated as ritual objects, curios or lucky charms.Footnote 27 A Palaeolithic hand axe had been deposited at the posited temple building at Kelvedon,Footnote 28 and more than 30 were placed at the Ivy Chimneys religious complex at Witham, where they were interpreted as being linked with a cult of Jupiter.Footnote 29 That said, it is unknown if the Red Lodge examples were found elsewhere and deliberately brought to this sanctuary site for deposition, or were found locally and opportunistically repurposed.
With the exception of the face fragment from a Hadham Ware face flagon (Fig. 9) found in the deposit to the immediate north of the shrine, the recovered pottery assemblage is of a primarily ‘domestic’ nature, with a notable absence of diagnostic ‘cult’ type vessels such as tazzas and lamps and a low incidence of burial of whole or modified vessels. This supports the suggestion that perhaps this was a local rural shrine, with attendants using locally made containers, potentially used to bring food or other goods to site as offerings. Two sherds are inscribed with graffiti, with the base of a grey-ware vessel incised with an ‘X’, and ‘VAR’ scratched into the exterior wall of a rim fragment. Examples of incised pottery sherds are widespread, if infrequent, across the east of England, with the ‘X’ graffiti considered to have links with the wheel and magical or superstitious connotations with Fortuna.Footnote 30 Alternative interpretations, such as at Heybridge, Essex,Footnote 31 where ‘X’ incised examples were more prolific, has suggested that it may have been used to mark out vessels for a specific purpose during dining or cooking. The sole example from the Red Lodge assemblage makes it difficult to ascertain its function, with the generally domestic nature of the pottery implying it was refunctioned in some way prior to deposition. A small number of other vessel fragments show signs of having been re-shaped, with one having been deliberately halved. The deliberate halving or re-shaping of domestic pottery is shown in several sanctuary sites,Footnote 32 showing the repurposing of domestic objects for a further purpose.
Pottery.
The exact sequence of abandonment/demolition of the shrine and wider site area was not clear from the archaeological evidence. The dispersed nature of the shrine demolition material, likely exacerbated by modern agriculture, restricted the understanding of the overlying deposits, meaning a clear sequence of disuse could not be discerned. The material overlying the shrine was formed of multiple deposits, which had unclear relationships in plan, and formed an irregular 27 m by 16 m spread overlying the structure, and primarily to its north and west (Fig. 3). The latest coin associated with the shrine head-and-hoof deposits dates to a.d. 348–50 and is suggested as being deposited towards the end of use of the shrine, with material from the overlying demolition/disuse deposits containing similarly dated late third- to fourth-century coins and pottery. Disuse of other areas of the site were less apparent, with many features having seemingly infilled naturally over time as opposed to being deliberately backfilled at the end of use. A shallow deposit of light-yellow brown silty sand overlay the flue, with material including two coins, dated a.d. 250–296, recovered. This material may have been indicative of its end of use, though due to its shallow nature and location immediately outside the ring-ditch may be potential mound material which has spread over the area after it went out of use. The lack of definitive evidence of demolition or deliberate backfilling of features across the religious precinct are suggestive that it fell into disuse as opposed to being deliberately demolished, with dating suggesting its level of use deteriorated towards the mid-fourth century. The rural nature of the site means that it likely retained a sense of importance and reverence within the local area, but its larger-scale use as a religious site was no longer utilised.
Post-Roman land use
Archaeological evidence of post-Roman activity was limited to a single inhumation burial of medieval date, found within the upper fill of the temenos ditch (Fig. 3). This prone interment was initially regarded as a late Roman casual insertion into the top of the ditch; however, radiocarbon dating subsequently returned a date of 1169–1266 cal. a.d. (95.4% probability, SUERC-83258). While the site area is identified in documentary sources as being utilised as gallows in the thirteenth century, no archaeological evidence of a gallows, or further burials or deposits which may indicate one nearby, were encountered. The barrow survived as a positive earthwork into the nineteenth century and, as recorded by the Suffolk HER, was even observed in 2007 to be c. 0.5 m higher than the surrounding field surface. It is entirely conceivable, therefore, that the prominence of this location, on or close to the parish boundary, was selected as the location of a gallows,Footnote 33 with its structural remains presumably having been lost to agricultural truncation. The adult male skeleton, aged 25–35 years, showed no sign of trauma consistent with hanging.
Discussion
While prehistoric barrows have long been appreciated often to have been reused in the Roman period, this Red Lodge site provides an example of extended and progressively more complex use of the monument over time, peaking with a developed religious complex in the Late Roman period. The Early Bronze Age monument is deliberately incorporated into the Roman precinct, though the reasons for this and the precise nature of the role it played in the rituals performed can only be speculated. Comparative sites in the East of England region cited in this discussion are located on Fig. 10.
Comparative shrine plans.
Fig. 10 Long description
Map of East England detailing locations of barrows and shrines with specific symbols. Barrows are marked with yellow circles, Roman re-use of a barrow with blue circles, shrines with green circles, both barrow and shrine with purple circles and other sites with red squares. Red Lodge is highlighted with a red square. Key locations include Aldham Mill Hill, Hadleigh (1), Ardeigh (2), Bixley (3), Bredonham Loop (4), Sixpaddle (5), Castle Hill, Cambridge (6), Exning (7), Fritton (8), Grimston End (9), Haddenham (10), Hadleigh (11), Heybridge (12), Kelvedon (13), Liberty Village, Lakenheath (14), Little Paxton (15), Monkton (16), Red Lodge (17), Scole (18), Sicklesmere Road, Bury St Edmunds (19), Stanwick (20), Stowmarket (21), Walbrook, London (22) and Witham (23). The map includes a scale bar indicating 25 kilometers for distance reference.
Pre-Roman
This location at Red Lodge evidently has a long history of special or ceremonial use that tangibly begins in the Early Bronze Age with the construction of a funerary monument — in the form of a relatively large and locally prominent round barrow. Although no associated burials or other structured deposits survive, it can reasonably be assumed that this earthwork monument had an enduring significance to the local population — primarily to those who built it and placed their honoured deceased within its mound, but also to subsequent generations who are likely to have renegotiated their relationship with it and perhaps reinterpreted its significance over time. Although it is possible only to speculate as to the nature of the relationship between the monument and the local community of the day, some sort of link endured into, or perhaps had been reinvented by, the Iron Age. The large part of a jar of Early Iron Age form in the barrow ring-ditch, though ostensibly found in the Roman-period recut and alongside Roman pottery sherds, might be interpreted as a displaced offering that had been inserted into the mound. Although residual, and with the lack of other Iron Age pottery from across the excavated features, this is thought likely to be indicative of the Iron Age use of the barrow. Given the intervening 1500 or so years, it is highly unlikely that the monument retained much more than a vague semblance of its original significance, meaning and connection, perhaps having been renegotiated and appropriated by different peoples along the way. Beyond the single instance of a placed offering, very little evidence for what activities this continued veneration and use entailed survives.
Roman
Perhaps long mythologised and revered as the resting place of a heroic ancestor, and by this time presumably imagined or created by the local population to suit their own social, political or religious needs,Footnote 34 in the Roman period the barrow site became the location of a formal constructed sacred space. Evidently, the barrow had sufficient perceived significance, perhaps even a religio-magical power, that a cult centre was developed around it from the second century a.d. onwards, the ritual reuse of prehistoric monuments being a well-established phenomenonFootnote 35 — as part of the wider use of a range of constructed and natural locations in the landscape.
The reuse of barrow monuments during the Roman period is not uncommon, though the nature and complexity of reuse varies. The most simplistic examples of reuse consist of the ring ditch or mound being used for the placement of structured depositions. Excavations at Liberty Village, Eriswell, recovered two pig head-and-hoof deposits within a barrow ring-ditch, with an associated scatter of second- to fourth-century a.d. coins in the interior of the ring-ditch.Footnote 36 Similar deposition is apparent at Grimstone End, Pakenham, with cremation vessels and food urns places on the inner lip of the barrow during the beginning of the second century a.d., probably at the base of any surviving mound.Footnote 37 The presence of temenos enclosures around barrows at Slonk Hill, West Sussex,Footnote 38 and Stanwick, Northamptonshire,Footnote 39 show that the barrow remains themselves were the foci, with the surrounding area deliberately kept devoid of features, as no further structures or features were present, apart from large coin and oyster-shell deposits at Stanwick indicative of votive offerings.Footnote 40 At Slonk Hill during the fourth century both of the two round barrow ditches were filled in and a square structure comprising timber posts constructed around the eastern mound. Associated votive offerings probably include a deposit of leg bones from lambs and piglets cut into infill of the western mound, and 12 fourth-century coins found in a nearby post-hole.Footnote 41 This form of deposition is perhaps similar to the Red Lodge barrow use during the Early Roman period as, while ‘special’ deposits could not be associated with the barrow itself, they were present within the establishment of the temenos, indicating that it was revered in some capacity. The surviving evidence from the above-mentioned sites, as at Red Lodge, contain no evidence of the perceived barrow mound, which likely acted as a focal point of activity. Remains of intact barrow mounds are sparse, though evidence of two phases of temple structure, including a sub-square post-built structure, were on the crown of a barrow mound at Snow’s Hill, Haddenham, Cambridgeshire.Footnote 42 It is quite plausible that this was a more common occurrence, with the regular removal of barrow mounds across East Anglia by modern agriculture meaning that they are rarely seen in the archaeological record.
Form of the sacred space
The establishment of a relatively extensive rectangular temenos, c. 3.2 ha in area (c. 200 m × 160 m), with the barrow located at its approximate centre, established the essence of what may be regarded as a religious sanctuary in an unequivocally Roman manner. The overall size/shape of this sacred space is difficult to fully compare, as limited size of excavation area on sites, such as Liberty VillageFootnote 43 and Orton’s Pasture,Footnote 44 means the full extent of the enclosure space is not exposed. Larger temple complexes incorporated other structures such as hostels and bath houses.Footnote 45 The fact that the full extent of the religious precinct is not always exposed means that comparison to the Red Lodge temenos is difficult.
It seems that the barrow was initially the single primary focus for the activity carried out within the sacred precinct, though no indication of its earlier Roman use can be discerned — any such activity probably having been removed either by the later re-cutting of its enclosing ring-ditch or the reduction of its mound by more recent agricultural processes. It is the fact that the ring-ditch was recut and received fills in the later Roman period, containing items identified as votive objects, that is taken as evidence that the barrow continued to have an active role in the sacred function of this place. The locations of 700 coins from across excavations of the barrow at Walkington Wold are indicative of the focus of existing barrows, with some objects having been deliberately damaged before deposition.Footnote 46 The construction of the shrine appears to be a later enhancement of the religious complex. Located immediately east of the barrow, it was clearly positioned in relation to it and perhaps purposefully aligned on the west entrance into the temenos that lay beyond. This said, it is observed that enclosed temple/shrine complexes elsewhere often instead had eastern entrances (e.g. Thetford, Springhead, Hayling Island, Great Chesterford, Lancing) and it must be conceded that nothing of the north and east, and little of the south, boundaries of the Red Lodge sanctuary were investigated to determine whether any other points of access were present. The introduction of the shrine is posited to have accompanied the remodelling of the western boundary and its entrance. The latter enlarged the area of the religious precinct, perhaps implying a perceived need for a greater capacity or a desire to reflect its enhanced status as a sacred place. As previously noted, the barrow ring-ditch was recut — perhaps as part of this enhancement — suggesting that its role was not diminished or replaced by the shrine. The mound may even have been enhanced or restored using the upcast from the recut. This said, the nature of its use is unclear; whether it served as a monument or cenotaph, or as a platform on which certain devotional acts were carried out, can only be speculated upon. The internal subdivision of the sacred space is defined by a number of minor boundaries positioned to the north and east of the barrow and shrine building. While the full extent of the subdivisions within the temenos remains unknown, they imply an internal special division of the temenos. The small number of internal features means that no distinct uses of the divided areas could be established, though the inner enclosure was obviously the focus of the religious activity. While the intermittent nature of the internal ditches means that they may not have formed a complete boundary, they may have been used to limit physical or visual access to the shrine and barrow. The use of boundaries to define space and create specific routes of access to sacred sites is commonly seen within wider temple sites and within the shrines themselves, the inner sanctum or cella of temples often being enclosed by an ambulatory, such as at larger temple sites such as Chanctonbury RingFootnote 47 and Brean Down.Footnote 48 Control of the wider space surrounding the shrine can be seen at Heybridge, with access to the temple cella becoming more restricted on approach. Four zones were identified, with the level of inclusion and area decreasing, suggesting selective access to the temple itself.Footnote 49 While the use of zones could not be clearly understood, the Red Lodge shrine lies within a wider area incorporating at least three zones leading to the shrine. As the northern third of the enclosure was not excavated, its role within the zones can only be speculated, but it is perceived that it would have formed the outer space, with entrants to the shrine then led through the access to the west of the barrow with final, likely limited, access granted to the shrine itself. The fact that the eastern end of the temenos is speculated to be outside the excavation area may mean that a further approach was available from the east, though again the internal boundaries may have provided a deliberate routeway through to the shrine and barrow at the centre.
As demonstrated by the extensive trenching investigations across the wider Red Lodge area, all Roman land-use activity is confined to the temenos; no associated ritual activity, or indeed any other activity, is present in its surrounding vicinity. As explored further below, the sacred complex sat in apparent isolation in its landscape.
The shrine
Although its site was perhaps already used as a location for deposition of votive offerings, the shrine building is considered to be a later introduction to the religious precinct, likely being constructed around the mid- to later third century a.d. As a small, rectangular (c. 5.0 by 4.5 m), single-roomed building, lacking embellishments such as an ambulatory or portico, it is regarded to have been a shrine rather than a more developed temple facility. However, this distinction may be fairly semantic, and its function may not have been very different other than in terms of scale and capacity. While of relatively simple timber-framed, single-storey construction with walls of wattle infill, its decorated plaster walls and probable tiled roof would have created a relatively imposing and distinctively Roman edifice. While the decorative scheme cannot easily be reconstructed, it evidently included large blocks of single colour (red and white), as well as areas of dot decoration and foliate design (red and green), and five fragments which joined together to form a purple scroll (Fig. 6). A group of fragments displayed a series of lines in shades of red and purples, which are thought to be an attempt to illustrate architectural features, such as pilasters and pediments, and potentially form an upper frieze portion of fresco. Fragments containing pink and yellow/orange and were also recovered. Detailed analysisFootnote 50 indicates multiple layers of redecoration, with repainting of the shrine walls being undertaken at pertinent points, perhaps within the religious calendar.
In contrast to its colourful and ‘high-spec’ walls, the internal floor of the shrine appears to have been a simple gravel surface, laid over a chalk bedding layer. No indication of an overlying surfacing, such as tesserae, was observed and no such items found elsewhere on site in secondary contexts. A shrine at nearby Lackford, c. 8 km east of Red Lodge, seems to have been of broadly similar date and construction, comprising an 8 m square post-and-timber plank-built structure with a chalk and clay floor and painted wall plaster wall decoration.Footnote 51 Although seemingly circular, the ‘temple-related’ structure at Hockwold cum Wilton was perhaps not dissimilar, being formed of a compacted chalk surface, with ridges of harder chalk with flint inclusions possibly indicating a wall, and with wall plaster and roof tiles in association.Footnote 52
Few internal features were present in the shrine, presumably reflecting the simple layout within such a small building. As previously mentioned, in situ fragments of two possible ?moulded mortar/plaster column bases were present in its south-east corner that are posited to constitute supports for a small altar — perhaps lararium-like in form. If indeed an altar of raised alcove form, the width of the columns, combined with its location in the corner of the building, thus limiting the extent of the altar, suggests it may have had a shelf of up to c. 1.8 m wide placed on the columns for the holding of prominent features for the related deity/deities of the shrine. It is speculated that there could have been multiple shelves or an alcove within the columns which was utilised for the placement of further effigies or associated candles and burners. If the altar took the form of a household lararium then part of the worship may have involved the offering of plants and animals.Footnote 53 The limited space, while insufficient to house any concerted accumulation of offerings, would have been sufficient for the burning of offerings such as cereal grains, beans, grapes and cakes of cereal,Footnote 54 with herbs and incense also used as offerings. If this was the common form of offering, with perhaps further distribution of the burnt material including onto the barrow and its perceived mound, then it would explain the relatively small quantity of items indicative of ritualised deposition, with these deposits not surviving in the archaeological record. This also correlates with the domestic pottery recovered across the site and is suggestive of a rural shrine.
No clear doorway is evident in the plan of the building, though given the possible presence of the altar in its south-east corner, this might reasonably be speculated to have been located either in the west or north wall. The smaller size of the central post-hole along the north wall may be an indication of this, though equally a doorway on the west facing onto the barrow mound would be feasible. Underlying the shrine and extending beyond it by 2.0 m or more, chalk foundation deposit G56 may have constituted an external walkway around the foot of the building, or else acted as a visual ‘exclusion zone’ around it, its whiteness emphasising its sanctity or purity. Given the cella-like form of the shrine itself, such a path might mirror an ambulatory, a common feature of Romano-Celtic temples. The 3 m square post-built shrine structure identified at Biddenham Loop, Bedfordshire,Footnote 55 was enclosed by a small ditch, perhaps acting like the chalk walkway at Red Lodge in demarcating the sacred focus and separating it from the rest of the site. Similarly, the temple structure excavated at Scole, Suffolk, was a c. 3 m square four-post structure with a concentric ‘sill footing’ on at least three sides interpreted to define an ambulatory,Footnote 56 but perhaps was instead simply a minor boundary around a shrine. The circular shrine at Eriswell, Suffolk, spanning the first to mid-third/fourth century is noted to have contained a possible crushed chalk floor surfaceFootnote 57 , and a small rectangular roadside shrine at Springhead measuring c. 3 m by 4 m had a chalk surround interpreted as wall footings.Footnote 58 While the practicalities of the availability of construction materials cannot be discounted, the use of chalk in a number of these structures may have been a deliberate choice driven by religious custom or prescription.
Overall, this structure appears typical of the small buildings identified elsewhere as shrines, whether occurring in rural isolation, in roadside settlement locations or as ancillary buildings within the bigger temple/sanctuary complexes (Fig. 11). Larger temple complexes often contained multiple ancillary buildings, with smaller buildings within the complex at Uley thought to represent shops and hostelries for visitors.Footnote 59 The sunken room and smaller shrine associated with the temple at Lamyatt Beacon are of a similar size to the Red Lodge shrine, and whilst containing no deliberate deposits are of comparable size.Footnote 60 Additional structures from Springhead suggest a ‘portico’ structure, perhaps used as a separate shrine, representing a further deity.Footnote 61 Temple 2 at ChanctonburyFootnote 62 was associated almost exclusively with deposits of pig teeth and skull bones, suggesting that the temple use differed from that of the primary temple structure within the temenos. The size of the Red Lodge shrine means that it cannot be compared to buildings within larger temple complexes, which would have contained multiple buildings to address the needs of visiting worshippers. The fact that small shrines are mostly identified solely by special deposits or location means that there may be further structural comparisons which have not been identified as shrines, with those at Monkton (Kent), South Perrott (Dorset), Great Dunmow (Essex) and Ortons Pasture (Rocester, Staffordshire) identified by deposits and demonstrating the varying forms potential shrines may take.Footnote 63
Places mentioned in discussion (east of England only).
Fig. 11 Long description
Six archaeological site plans arranged in a 3 by 2 grid compare shrine layouts defined by post holes. Each plan uses filled shapes for features such as deposits, ditches, or other mapped areas and a thin outlined square with corner points to mark the post-hole structure. One plan includes a north arrow at the top right of the overall figure. A scale bar labeled 0 and 5m appears in the bottom-right plan. The plans are labeled with site names only; no wider geographic map or coordinates are shown. The image A showing Red Lodge. Labels include “Votive deposits,” “Chalk path,” and “Structure - Post holes.” A cluster of small filled shapes is labeled “Votive deposits” to the left of the structure. A path labeled “Chalk path” is shown adjacent to the structure. The image B showing Scole. Labels include “Flint path” and “Structure - Post holes.” A path labeled “Flint path” is shown near the structure. A dashed line segment appears near the lower left of the plan. The image C showing Biddenham Loop. Labels include “Surrounding ditch” and “Structure - Post holes.” A continuous band labeled “Surrounding ditch” encircles the structure. The image D showing Great Dunmow. Labels include “Votive deposit” and “Edge of hollow delimiting shrine.” Multiple filled areas are labeled “Votive deposit.” A curved boundary line is labeled “Edge of hollow delimiting shrine.” The image E showing Springhead. A single irregular filled feature is shown without additional internal labels. The image F showing Orton’s pasture. A rectangular enclosure-like outline is shown with openings along the outline. A scale bar beneath the plan is labeled 0 and 5m.
Other structures
In addition to the barrow mound and the shrine building, identified remains of structures occupying the religious precinct are sparse. The small pond-like feature located to their south may perhaps be viewed as an active structure performing a functional role. This could indicate that the religious complex had multiple areas for worship and/or deposition, potentially with several deities being revered and in different ways. The potential of a water source in the form of a pond is significant as it may indicate the creation of a water source within the temenos. The association of shrines with springs, for example, can be seen at Buxton, Bath, Springhead and Ratham Mill with specialised religious sites sited adjacent to shrines.Footnote 64 The site at Springhead shows further association with water, with finds recovered from the riverbed, and posited terraces though to have acted as a viewing platform. This use of waterways, bogs, rivers, etc. as sacred spaces for the deposition of objects is a well-established phenomenon from the Iron Age into the Roman period,Footnote 65 though it is noted that there are fewer examples from the third century onwards,Footnote 66 perhaps representing a cultural shift away from natural sources as sanctified sites. While a small concentration of coins was recovered from the Red Lodge pond, they were generally from the top of its infilling and so are likely not representative of regular deposition within the pond during its lifetime. This implies it was not specifically used for the deposition of objects, though degradable foodstuffs may have been cast into the water.
The oven/flue to the north of the barrow was a later addition to the precinct, probably constructed at the same time as the shrine. While its remains offer little insight into its potential use, its above-ground structure having been entirely lost, it is surmised that these denote the likely presence of an oven or hearth. Bakeries and ovens are not uncommon on Roman religious sites, with bakeries in the form of clay-floored circular structures with hearths at Springhead,Footnote 67 and a flint-lined oven at Chanctonbury Ring.Footnote 68 The vicinity of these bakeries to the shrines suggests they played a crucial role within the worship occurring at the shrine; it is posited that this oven was perhaps used either in the production of ritual foodstuffs either consumed or left as offerings, or else to burn offerings, possible deliberately shaped before being ritually used.Footnote 69
Structured and votive deposition, and other ritual practices
As previously noted, it is likely that the multiple and varied instances of structured deposition identified within the investigated portion of the religious precinct constitute a general continuum of devotional activity occurring throughout its lifespan. Although their chronology is not entirely clear, it appears that in its earlier Roman manifestation this primarily takes the form of the burial of a variety of single animals, often at potentially key locations within the sacred precinct (Fig. 12).
Distribution of animal burials, coins and other votive objects.
Fig. 12 Long description
The map illustrates the distribution of various objects within a religious precinct. It includes personal objects marked in purple, repurposed iron heads in orange, votive objects in red and animal husbandry objects in green. Key locations are marked with animal burials, such as pigs, cows, dogs and cockerels, each labeled with the number of instances. The map also highlights specific areas with coins and votive objects, including 27 coins near the shrine and 11 coins in a demolition spread. The spatial arrangement shows clusters of objects around the central circular feature, with directional markers indicating north. A scale bar is present, showing distances up to 20 meters.
While the placed head-and-hoof deposits represent a large portion of the structured deposits identified at the Red Lodge site, this is not a significant quantity compared to some larger temple sites elsewhere and, due to the paucity of remains and associated deposits, are difficult to affirm confidently to a specific cult or deity. The large quantity of pig deposits at Chanctonbury Ring, Sussex, have been speculatively associated with either a hunting cult or a cult of the boar,Footnote 70 while others at Hayling Island, Wanborough and Walbrook also contain significant deposits of pig bones, with the Walbrook temple being recognised as representative of a Mithraic temple.Footnote 71 It is possible that the Red Lodge shrine was used as a religious centre for either of the above deities, though the choice of coins placed with the deposits may suggest other associations; a sestertius with the Imperial Virtue of Annona, the personification of the Roman grain supply, and an as with Pietas, representing the Virtue of Dutifulness, both deliberately placed. These are not associated with a specific deity, and the available evidence does not allow a clear focus of worship to be established, as different objects and structured depositions from across the site have multiple known associations. While there is a specific focus of activity around the shrine, the site as a whole contained a relatively small number of structured deposits which could be used to suggest a specific deity. It is perhaps likely that multiple deities were worshipped at many Roman shrines,Footnote 72 with the rural location of the Red Lodge shrine maybe indicating a more fluid nature of worship, including multiple deities or that of specific seasonal events.
While limited in number, the structured cow, dog and cockerel deposits within the precinct were probably deliberately placed to imbue the area with significance, though the meaning behind them can only be speculated. The general lack of other features within the area, restricted to the well, suggests that these structured animal deposits had a role in the use of the worship at the shrine, with the well perhaps indicating an area of cleansing on entry to the temenos centre.
The single cow burial, found on the opposite side of the shrine, represents a likely isolated deposition. The lack of any butchery or consumption evidence from the bone indicates that it did not have a role in feasting. While limited, cow bone remains from other sites, such as Heybridge, suggest that cows were used for consumption, with product utilisation resulting in few articulated bones recovered.Footnote 73 Larger sites, such as the Roman temple at Chanctonbury, while containing multiple domestic ox remains, show clear use of locational deposition of animal parts, indicating different use for consumption and deposition, with a large number of cranial elements recovered but little other bone.Footnote 74 Chanctonbury also shows clear zonation of depositions of species, with sheep and ox bones primarily in the temenos ditch, with pig remains concentrated at the temple, and oyster shells in a midden elsewhere within the temenos.Footnote 75 The single example of a cow burial at the Red Lodge site means that any spatial comparison of deposits will be accentuated, though the location, spatially separated from the shrine by the barrow mound, is likely deliberate and significant. While there was a deliberately placed cow scapula recovered from within a pit alongside cockerel remains, with the scapula showing evidence of butchery, again, the limited sample size means both may be representative of one-off depositions.
The dog burial from Red Lodge, while limited to a single example, is significant in its location, at the perceived entrance to the temenos, and with its association with horse remains. Dog burials have associations with guarding, with early comparisons with Cerberus, the ancient Greek guardian of the underworld, with fertility, and even healing and agriculture.Footnote 76 Dog and stag imagery is associated with Silvanus, a god of wild places, while the cult of Nodens at Lydney included dog votivesFootnote 77. The close association of dog and horse remains, as seen at Red Lodge, has also been suggested to be of importance, the fact that both animals are neither wild nor domesticated meaning that their sacrifice and deposition is imbued with greater significance.Footnote 78 The deposit is very similar to the cockerel burial, which is associated with a cattle scapula, showing hook-hanging damage, and a shed deer antler. The association of the cow and horse scapulae with articulated remains, of cockerel and dog respectively, is of note as they denote joints of meat which would otherwise be used for consumption. The deer antler is also of note due to its associations with Silvanus, the god of wild places. While small, fragmented pieces of antler were also recovered from the recut of the ring-ditch G12, potentially implying further deposits, the lack of confirmed complete structured deposits of antler is suggestive of an isolated deposit.
While the presence of only a single example of placed cockerel remains means ascertaining any associated deity or meaning is difficult, singular deposits of domestic fowl have been recovered from the Mithraeum temples of Walbrook and Carrawburgh,Footnote 79 with domestic fowl burials also known to be linked both with Mithras and Mercury.Footnote 80 Associations with Mercury have been previously identified at Uley, where domestic fowl remains were found alongside altar fragments depicting the deity.Footnote 81 Chicken/cockerel remains, and a figurine deposited outside the temple at Heybridge, were potentially also linked with Mercury.Footnote 82
The nature of depositional activity seems to change in the later Roman precinct, after or alongside the construction of the shrine, perhaps marking an evolution of devotional practice or even a change of deity worshipped. The most conspicuous deposits from the site are those of six pig head-and-hoof burials located in close association with the shrine. Where found, the articulation apparent in head-and-hoof deposits of various animal species, and of various dates and contexts, is often interpreted to indicate that the hide was attached at the time of deposition,Footnote 83 which may have been significant to the role of the remains in the ritual activity occurring, though the lack of use-wear on the Red Lodge remains may suggest that any use was of a limited nature prior to their burial. It is noteworthy that the shrine site at nearby Lackford included pig head-and-hoof deposits,Footnote 84 as did those at Eriswell,Footnote 85 the former also producing numerous votive plaques. Similarly, the shrine at Snow’s Hill, Haddenham, featured head-and-hoof deposits with coins placed in the jaws, though in this instance they were of sheep,Footnote 86 demonstrating that while the sacrificed animals may have varied between religious sites, their votive treatment could be remarkably similar.
The additional presence of the pond and oven might be inferred to indicate the worship of multiple or non-specific deities at this shrine, with its rural location and historic connections with the landscape maybe allowing more general worship for the local populace. A comparison to this is the larger temple sanctuary sites with their smaller shrines on the periphery of the main temple, allowing multiple areas of worship within a single precinct, with the multitude of temples from Springhead allowing multiple areas of worship, with the variety of votive deposits suggestive of different deities.Footnote 87
Post-Roman
Although no osteological evidence for hanging trauma was discerned on the medieval inhumation inserted into the former temenos ditch, and no remains of a gallows structure identified on or near the barrow site, it is not unreasonable to posit that this was the location of the execution place alluded to in contemporary documentary sources. While the burial seems unrelated to, or unaware of, the previous sacred use of this place, it at least serves to emphasise the enduring significance of barrows in the post-Roman landscape, often as prominent markers on estate and parish boundaries, their liminal locations often being utilised either as meeting places or for dispensing penal justice.Footnote 88
Conclusions
Comparison with other structures interpreted as shrines, as well as the reuse of prehistoric monuments, has demonstrated the variety of locations and forms that they can take, with the Red Lodge example having similarities with different examples, thus exhibiting the diversity in this type of religious setting during the Roman period. The Red Lodge shrine would have been a significant feature in the local landscape, focused on the existing Bronze Age barrow. The lack of Roman features from surrounding archaeological work implies it lay in a clearing, within a widely wooded landscape. The various depositions hint that the barrow mound and adjacent shrine were the central focus of the religious area, with controlled access through from the temenos leading to the centre of religious focus. The perceived route, past the well and oven, implies activities related to cleaning and ritual burning or the acquiring of foodstuffs were pertinent to religious acts on the site. The number of special deposits from the site is generally small and it is posited that these, particularly the head-and-hoof deposits, may only have been deposited during specific religious ceremonies, or by nominated religious figures/individuals within the community. The range of objects which have been placed hints that it may have been the act of deposition, as opposed to the object itself, which was more pertinent at this site. It is assumed that the focus of the religious activity took place on the barrow mound, as a prominent feature in the temenos, with its elevated position in relation to the surrounding landscape emphasising its role within the inner sanctum of the enclosure. The isolated nature of the shrine suggests it was dedicated to either a local deity, or, more likely, had multiple aspects of dedication.
Acknowledgements
The project was commissioned and funded by RPS Consulting Ltd on behalf of Crest Nicholson (Eastern). Rachael Abraham of the Suffolk County Council Archaeological Service is thanked for her guidance and monitoring of this project. The excavation was directed by the author, with assistance from ASE archaeologists. The contributions of the following specialists to the analysis and interpretation of this site are duly acknowledged: Stacey Adams (plant macrofossils), Isa Benedetti-Whitton (ceramic building material), Agnese Benzonelli (painted wall plaster), Trista Clifford (metalwork), Anna Doherty (prehistoric pottery), David Dunkin (shell), Hayley Forsyth-Magee (animal bone), Kayt Hawkins (Roman pottery), Karine Le Hégarat (worked flint) and Paola Ponce (human remains). Jude Plouviez is thanked for kindly providing information on the Lackford shrine site ahead of its publication. Andrew Lewsey produced the plan figures and Fiona Griffin the artefact figures. Mark Atkinson edited this report and provided advice and additional insight.
Open access
