PREVIOUS   NEXT   CONTENTS   HOME 

4.3 Social distribution

One of the biggest problems with metal-detected finds is that they usually come without context details (i.e. while many come from the ploughsoil, others are removed from archaeological contexts), although sometimes they are found in areas where there are known sites. A total of 579 figurines out of the 1000 under consideration come from known sites, and the site types were divided into three broad categories: urban (289 figurines), rural (155 figurines) and military (162 figurines). The urban category includes London, the coloniae, civitas capitals and small towns, while rural includes small settlements, villas, religious sites and hoards. As Eckardt (2002, 29-30) noted, not only are there problems defining what constitutes sites such as small towns or rural settlements, but also determining how to quantify sites with both military and civilian occupation. On such sites, finds from contexts that are specifically military or civilian were counted within their category, while the value of finds with no contextual information was split between the appropriate military and urban categories. In order to determine what constituted the various site types, such as the civitas capitals, small towns and forts, a number of sources were used including Gurney (1995a), Mattingly (2006), Wacher (1995) and the Ordnance Survey Map of Roman Britain (1956, 3rd edition).

Figure 10 demonstrates that, although the scatter of (primarily metal detector) single finds between the Humber and the south coast is greatly thinned out, figurines were being used on all three site types. As expected the greatest numbers come from the major urban centres in the Romanised south – London, Colchester and St Albans – with slightly smaller groups from centres such as Cirencester, Silchester, Wroxeter and York. Moderate assemblages have been found on the major military sites, such as the forts along Hadrian's Wall, Chester, Caerleon and Richborough.

 

Figure 10: Distribution of the different site types
Figure 11: Proportion of figurines from urban sites (total = 289 figurines)

Figure 11 shows that, among the urban sites, London forms by far the largest collection. What is perhaps more surprising is the particularly small number of figurines from coloniae other than Colchester. In fact, no figurines have been found at Gloucester at all, and only one at Lincoln. There are 14 figurines from York, but almost all of them are without contextual information; thus it is not possible to say whether they are from contexts in the civil settlement or military contexts associated with the legionary fortress.

The figurines from civitas capitals comprise almost a quarter of the total from urban sites, but there is great variation in the numbers from the different towns (Figure 12). The figurines from Silchester, Cirencester, Leicester, Exeter and Wroxeter comprise over half of the assemblage from the civitas capitals, while only small groups were found from the remaining sites. However, six of the figurines from Exeter were found in a group together in the High Street (586, 587, 588, 589, 590 and 591; Milles 1782). No further details are available, so it is not known whether they were a hoard or in any way associated with a temple or shrine. The variation in the numbers of figurines from these sites may in part be due to the level of excavation they have been subjected to, although it should be pointed out that many of the figurines are antiquarian finds that were uncovered by chance, and therefore must in part reflect the differences in the use of figurines at these sites. It is also clear from Figure 12 that while a military presence can add significantly to the number of figurines from a site, as at Wroxeter, sites without a strong military presence, such as Silchester, may have large assemblages.

 

Figure 12: Figurines from civitas capitals (total = 82 figurines)
Figure 13: Figurines from small towns (total = 80 figurines)

The small towns are represented by figurines from 27 sites and comprise a group of similar size to the civitas capitals. Single figurines have been recovered from 12 of the small town sites, and details are given in Figure 13. While the levels of excavation at these sites is variable, and in fact many have seen very little investigation at all, sites such as Bath and Alchester have been subject to considerable excavation yet still have produced only small numbers of figurines. In fact the numbers are similar to those from the rural settlement and villa sites at which single figurines are also more common. Eight (nine if Bitterne is included) of the sixteen sites with more than one figurine have military associations (the three figurines from Castell Collen were found in the vicus associated with the fort), and most of the sites with large numbers of figurines such as Caerleon, Corbridge and Kirkby Thore had a significant military presence. The large assemblages from Great Walsingham and Springhead are both directly associated with temples or areas of ritual deposition within the small towns. The assemblage from Caerleon in particular stands out. Many of the figurines were recovered during excavations, and six are known to have come from the vicus and five from the fortress itself. Thus figurines were in use by both the military and civilian populations, although one must remember that these groups were intimately linked.

Boar 742 from Camerton is thought to have been found in association with a large quantity of military equipment and ironwork, possibly representing a hoard, which was recovered by metal detector (Jackson 1990). The four figurines from Kenchester unfortunately have no associated context details. It is not even known whether any were found together, and the status of Kenchester regarding the military is also unsure, although Mattingly (2006, 329) suggests the small town there might be a vicus associated with a military installation. The sites of Godmanchester, Sea Mills and Water Newton have each produced only one figurine, but are also the site of military installations and it cannot be established whether these finds are from civilian or military contexts.

London (the provincial capital), Colchester (a colonia), and St Albans (a possible municipium) are all large and cosmopolitan sites at which one would expect to find prestige goods associated with 'Roman' culture, and thus the large numbers of finds from these sites is not unexpected. What is perhaps more surprising is the lack of figurines at some of the other urban settlements. Certainly almost all of those sites with larger groups of figurines also have military origins. Silchester and Cirencester stand out as urban sites with large figurine collections and little in the way of military connections. Although there was a fort at Cirencester, it was only occupied for a short time in the 1st century AD (Wacher and McWhirr 1982, 22), and none of the figurines for which we have location details come directly from the area of the fort. However, like Colchester and St Albans, both Silchester and Cirencester were associated with Iron Age oppida and were important sites at the heart of Roman occupation in Britain.

Unfortunately, very few of the finds from any of these urban sites come from known contexts, which makes it difficult to assess how they were being used. The context details that we do have are listed in Table 7. From this it can be seen that figurines have been found in both domestic and public locations. Many of the figures found in association with houses come from floors or layers in rooms, and do not appear to have any ritual function in their deposition.

Others were found in public spaces including amphitheatres at Cirencester and London, the forum and basilica at Silchester and the theatre at St Albans. Two came from possible mansio sites in Chelmsford and London, and one from a bath house in St Albans.

The remainder are from sites of religious or ritual activity such as temples at St Albans, Caerwent, Colchester, Wycomb, and Great Walsingham and cemeteries at Poundbury, St Albans, York, Colchester and London. However, it should be pointed out that only the figures from Poundbury and York actually came from within graves. The remaining figurines from cemetery sites were found in other features such as ditches; thus metal figurines do not appear to be a major part of the burial rite at these sites. Only one figurine, Lar 104 from Silchester, is actually thought to have come from a lararium, although Venus 132 from St Albans was found placed in a wood-lined pit under the cellar stairs (Frere 1972, 144). Although Frere (1972, 105) suggested it may have been discarded as scrap metal after the owner converted to Christianity, others believe the reasons for its presence in the box may be more prosaic, such as loss of interest in the piece (Henig 1995a, 170). Eagle 428, Arm 457 and Leg 458 were also found in a cellar at St Albans, and Perring (1989, 284) suggests that niches in cellars may have been used as lararia. Colchester in particular has a large number of figurines and other votive objects that are directly related to temples found at the Balkerne Gate (Crummy 2006) and one figurine from St Albans was found at the South Gate. Lead figure 1007 from the vicus at Castlefield, Manchester may represent one of the few figurines that appear to have been deliberately deposited as it was found inserted vertically into a hearth in Enclosure G (Cool 2007b, 134).

Figurines from military sites (162 in total) have been separated into four categories: legionary fortresses (41%), vexillation fortresses (3%), forts (46%) and Saxon shore forts (10%). While the numbers of figurines from the legionary fortresses at Caerleon (5), Chester (7), Colchester (22), Wroxeter (11) and York (14) do seem quite high, the majority of the finds from Wroxeter and York are without context and so the numbers largely result from adding the figures of unknown context to those with a known military context. It should also be noted that the military occupation at Colchester, Alchester, Wroxeter and Exeter quickly gave way to long-lived civil settlements (Crummy 1984, 3-9; Sauer 2000; Webster 2002a, 17-37; Bidwell 1980, 16), and that while a large number of figurines have been found at Colchester, it is unlikely that many are from military contexts. Thus the real distribution of the figurines from these sites cannot be known. Only Caerleon and Chester have a high proportion of figurines from known military contexts, and it seems likely that the three unprovenanced figures from Chester are also from military contexts.

A small group of five figurines has been found at the vexillation fortresses of Chichester (581, 765), Kingsholm (254, 510) and Lake Farm (561), but unfortunately there is no contextual information regarding the discovery of any of these pieces. However, it is perhaps likely that the two figurines from Chichester are from civil contexts since military occupation there was again short-lived and the town was predominantly a civil settlement (Mattingly 2006, 137).

As might be expected, the figurines from forts form the most numerous group from military contexts, in part due to the larger number of sites involved. In all some 75 figures were recovered from 36 sites. Single figures were recovered from 23 of these sites (Ambleside, Bewcastle, Birdoswald, Birmingham, Brough under Stainmore, Burgh by Sands, Caernarvon, Catterick, Chesters, Great Chesterford, Godmanchester, Holme St Cuthbert, Loughor, Manchester, Moresby, Neath, Ribchester, Sea Mills, Stanwix, Stoke Abbott, Swanton Morley, Wall and Water Newton), and two figures from a further four sites (Binchester, Carrawburgh, Great Chesters and Lanchester). As usual, these low numbers may largely be the result of lack of excavation at these sites, although some, such as Birdoswald and Stanwix, have been investigated (e.g. Wilmott 1997, and various reports in Bidwell 1999). Sites from which three or more figurines have been recovered are: Benwell (4), Carlisle (3), Carzield (3), Corbridge (7), Kirkby Thore (5), Papcastle (4), Piercebridge (6), South Shields (7) and Vindolanda (5). Thus some forts contain numbers of figurines that rival those of the civitas capitals, and it is unfortunate that the groups from South Shields and Piercebridge do not have accompanying contextual information.

The Saxon shore forts form another small group, and figurines have been recovered from Burgh Castle, Portchester Castle, Reculver and Richborough. Richborough was first used as a military centre in the 1st century, but was then refortified in the late 3rd century (Cunliffe 1968, 3-6). While single figurines have been found at Burgh Castle and Portchester, and there are two from Reculver, a large group of 12 figurines has been recovered from Richborough, but unfortunately little contextual information is available for most of them.

The context details that we have for figurines from military sites show that they have been found in a variety of contexts within the fortresses or forts (Table 8). A number have been found within barrack buildings, often in contexts associated with the abandonment of the site, and others in various features such as pits, ditches and gullies. The only figurines from a temple or shrine associated with a fort are from Carrawburgh: Genius 197 from the Shrine of the Nymphs and Genius and Dog 235 from Coventina's Well. Three figurines were found in a furnace pit within a barrack block at Carzield (Birley and Richmond 1942, 158), and while they were perhaps originally due to be melted down for reuse, one cannot discount that they were deliberately left together in the pit.

The final group comprises finds from rural sites, which are concentrated south of the Wash. They were further split into four groups – rural, villa, hoard and religious – details of which are given in Table 9. The rural group comprises small settlements, except for the pottery workshop at Stibbington, Cambridgeshire, from which Jupiter 1165 was recovered. As can be seen in Table 9, the majority of the settlement and villa sites produced only one figurine, although two or three figurines were occasionally found. Rural sites with more than one figurine are Woodcock Hall, Norfolk; Stonea, Cambridgeshire; Dragonby, Lincolnshire and Aust, S. Gloucestershire; while villas are Boxmoor and Gorhambury, Hertfordshire; Hartley, Kent; Brail Wood, Wiltshire, and Frocester Court, Gloucestershire.

The rural settlement group is rather diverse, and little is known about many of the sites. Eleven figurines were recovered during excavations, but only seven have any context information, and these are listed in Table 10. The small number of finds with context information means little can be said about their deposition. However, it should be noted that two, Jupiter 1165 and Lar 1186, were found in contexts dating to the abandonment of the settlement, and that Lar 1186 does seem to have been specifically placed within the limekiln at the end of its use.

A moderately sized group of 39 figurines have been found on 34 villa sites, and while many may have come from within the villa building itself, almost none have known contexts. Eagle wing 1010 was found in a cellar at Gorhambury, Hertfordshire, and unknown male Deity 801 in a floor at Thurnham, Kent. Finds from other locations at villa sites include Bacchus 370 from a pit at Frindsbury, Kent, and Bound captive 529 from a ditch at Frocester Court. None of the finds are from contexts with overtly ritual associations. However, it is of interest that the only direct evidence for the production of figurines in Britain comes from a building associated with the villa at Gestingthorpe. This appears to have been a workshop used for producing a variety of bronze objects (Draper 1985).

Figure 14: Distribution of figurines at religious sites and in hoards: 1 Southbroom, 2 Ashwell, 3 Barkway, 4 Willingham Fen, 5 Felmingham Hall, 6 Lydney, 7 Uley, 8 Henley Wood, 9 Pagan's Hill, 10 Lamyatt Beacon, 11 Maiden Castle, 12 Farley Heath, 13 Woodeaton, 14 Thornborough, 15 Brigstock, 16 Thistleton, 17 Harlow and 18 Gosbecks Farm

While the majority of the rural settlement sites and villas are located in southern Britain, the hoards and religious sites are concentrated in a narrow band across the south (Figure 14). Although there are figurines associated with urban and military sites in other parts of the country, there are none from sites with an exclusively religious function outside this area. In addition, unlike the settlement sites, many of the hoards and religious sites produced groups of figurines. Figurines have been found in five hoards, the largest of which was from Southbroom near Devizes, Wiltshire. This hoard is exceptional not only because of its size, comprising 17 figurines, but also the mixture of highly classical and provincial, probably locally produced, pieces. Unfortunately the circumstances of this discovery are not known, nor is there any indication of a settlement or temple site in the vicinity.

The smaller hoards from Willingham Fen, Cambridgeshire; Felmingham Hall, Norfolk and Barkway, Hertfordshire, similarly have unknown associations. However, unlike the Southbroom hoard, the figurines form only a small part of the hoards from these sites. The group from Willingham Fen, which was found in a box, contained Mars 28, Horse and riders 161 and 162, Raven 431 as well as three heads (447, 448 and 449), and bull and owl attachments (357, 430). Evans (1984) has suggested a link between this hoard and a shrine located some four kilometres away on the basis of unusual ribbed baton terminals found in the hoard and at the shrine, as well as the combination of horse and rider figurines with sheep/goat burials, also seen at the temple at Brigstock. The hoard from Felmingham Hall was similarly found within a container, this time a pottery vessel, and included Lar 111, Heads 343 and 345, mount 421 and Raven attachments 610 and 611. Meanwhile the Mars 27 from Barkway was found with a group of silver plaques, four of which bear depictions of Mars and two of Vulcan, while a final plaque is inscribed with a dedication to Mars Toutatis (Toynbee 1964, 329). This is also thought to be a temple treasure, but again no site has been found in the vicinity of the hoard. Certainly the nature of Romano-British hoards has been discussed in recent years (e.g. Guest 2005; Hobbs 2006; Johns 1996; 2010; Millett 1994) with a variety of reasons put forward for their possible deposition. Johns (1996) in particular points out that material in hoards may not have been deposited ritually, but represented personal or temple wealth that was deposited for safekeeping.

A final small hoard from Ashwell, Hertfordshire comprised numerous plaques in silver and gold like those from Barkway, as well as Figurine 753, Arm fragments 807 and 808 and gold jewellery (Jackson and Burleigh 2007). This important group was dedicated to the goddess Senuna but, unlike the previous hoards, was found in association with a Romano-British settlement that also has evidence for substantial ritual activity (Jackson and Burleigh 2007, 49-52). Thus while the other hoards may well have been associated with a temple or other ritual site, the Ashwell group is the only one for which this is attested. It should also be noted that a few figurines from other rural settlement sites have been found in association with temples such as the Helmet crest 997 and Horse 998 from Stonea, Cambridgeshire.

Temple sites themselves have produced a number of large groups, many of which are located towards the south-west of Britain including the temples at Lydney (8 figurines), Uley (16 figurines) and Lamyatt Beacon (9 figurines). Slightly further east and north are the temples at Woodeaton (14 figurines) and Brigstock (8 figurines). Two figurines are also thought to be associated with the Romano-Celtic temple at Maiden Castle (Wheeler 1943, 75). One is unidentified fragment 593 found by Cunnington (Wheeler 1943, 75) and the second the highly unusual Three-horned bull 167 which is mounted by a goddess on its back, a human-headed bird on its tail and a now headless bust behind the horns (Toynbee 1964, 103).

Unfortunately many of the figurines from these sites do not come from excavated contexts (Table 11), and only the site of Uley has contextual details for the majority of the finds. The information that we have shows that while a number of figurines do come from within the temples themselves (Structure II at Uley), others are found in the general area of the temple complex and are not necessarily in structured deposits but are often associated with the late phases, and perhaps abandonment or destruction of the site. Although three of the Uley pieces were recovered from pit F19 within the temple (703, 705 and 711), a feature which may have belonged to the early phases of the building, it was highly disturbed and contained no primary deposits (Woodward and Leach 1993, 34). The figurines from Uley were also fairly widely scattered throughout the northern and western areas of the site, a pattern matched by many of the other votive objects such as the altars, statuary, and votive leaves (Woodward and Leach 1993, fig. 225).

There are also important differences between the finds from these religious sites, and it should be noted that some important religious sites which have been subjected to intense excavation, such as Bath, Somerset; Lowbury Hill and Marcham, Oxfordshire and Nettleton, Wiltshire, have not produced any metal figurines at all. Some of the temple sites contained very homogeneous figurine groups, as indicated by the predominance of dogs at Lydney, Mercury at Uley and horse and riders at Brigstock, while others such as Woodeaton and Lamyatt Beacon contain a far more varied collection. Overall it appears that on rural settlement and villa sites, figurines were regularly being used in small numbers, presumably for personal worship, and that at some religious sites they formed an important part of the ritual activity being undertaken.

The lack of more specific information on the contexts from which figurines have come, even when excavated, is frustrating as it does not allow a closer analysis of the way in which they may have been used. However, from the little information that we do have, we can see that figurines were being used on both civil and military sites in a variety of ways. The fact that they have been found within barracks, married quarters, houses and villas shows that they were being used in domestic contexts. They are often found in demolition layers, and their use in abandonment rituals is hinted at in their deposition in the limekiln at Castle Cary, the workshop at Stibbington and perhaps the furnace pit at Carzield. Meanwhile a stamped lead figure, possibly representing Juno, buried in the arena floor at the eastern entrance to the London amphitheatre (Wardle 2008, 194 and 199) represents a deposit made at the construction of the building.

The figurines from the Castle Cary limekiln and Carzield furnace pit highlight the fact that a number of figurines have been deposited in burnt contexts. Lead deity 1007 was found in a hearth within the Castlefield vicus in Manchester. It was placed vertically and Cool (2007b, 134) suggests that the better modelling of the upper body may have reflected the fact that only this part of the figure would have been visible once it was inserted in the ground. Scholar 212 was also recovered from a hearth in married quarters in the fort at Vindolanda (Birley 1973, 122). A further two figurines were recovered from just outside the south gate of the fortress at Exeter. Victory 771 was recovered from one of three cremations at the site, two of which contained a variety of ceramic vessels including locally produced 'Fortress wares', which indicate a military date (Salvatore 2001). Panther 772 was also found in the vicinity and may have come from a cremation. The contexts of both finds are dated to the third quarter of the 1st century AD (Allason-Jones 1991, 257; Ling 1991, 257). The Castle Cary Lar is from a rural site dated to the 2nd and possibly early 3rd century, but the other figurines from burnt contexts all have a military connection and are dated to the 1st or 2nd centuries.

The most obvious function of figurines as religious items is further indicated by their presence at temples. However, this is more common on urban and rural religious sites, as only two figurines have been found in association with military shrines. Of course military personnel could have frequented shrines or temples on civil settlements attached to military sites, but perhaps their worship often took a different form, as is in part attested by the more frequent appearance of stone dedications and altars at these sites (Jones and Mattingly 2002, maps 8:1, 2, 4 and 8). Since figurines are not regularly found at burial sites, they do not seem to form a regular part of the burial rite.

 PREVIOUS   NEXT   CONTENTS   HOME 

© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions | File last updated: Tue Mar 20 2012