Metalworking and other industrial activities within public buildings in Britain are usually assumed to form part of the general decline of the towns and civilisation, with the buildings being taken over by degenerate activity. Rarely has it been given detailed attention and analysis that could lead to alternative possibilities.
7.3.1 Production
Very little is known about the actual items produced by the metalworking activities. At Silchester, a large number of nails were associated with the iron-working evidence, although it is uncertain whether these were a product of the activity or if they were intended for recycling (Fulford and Timby 2000: 74). A large quantity of nails was also
FiGURE 7.8. The excavated areas of the Winchester Palace complex in London, Southwark, with traces of possible fourth-century activity within the building (drawn by A. C. Rogers; adapted from Yule 2005,
Found contemporary with the metalworking debris at Lincoln (Steane 2006), although again their relationship to the process is uncertain. It has been suggested that iron nails were amongst the products of the iron-working in the basilica at Caerwent, because there were nails in the debris here (P. Guest personal communication). The production of nails would clearly have been of considerable importance in the maintenance and construction of town buildings including new timber structures. The ritual connotations of nails (see Section 7.3.3) may also suggest that the character of production had symbolic elements.
In a few rare instances mould fragments have been found. Brown (1976) has highlighted the lack of surviving mould debris from known bronze-casting workshops of the Roman period. The small number of finds need not reflect the level of activity. Within the basilica at Silchester, though unstratified, half a mould for casting copies of the coins of
Tetricus II (a. d. 273-4) was recovered (Fulford and Timby 2000:72).82 The oolitic limestone moulds from the temple building at Gloucester were for casting pewter vessels (Heighway and Garrod 1980). Limestone moulds also came from the site of the Huggin Hill baths (Museum of London Archive and Research Centre record: DMT88) and from within the basilica at Silchester (Fulford and Timby 2000: 73), which in both cases seem to have been for the production of bronze or pewter vessels. A fragment of a pewter vessel mould was found amongst the metalworking waste in the principia at York (Phillips and Heywood 1995: 66-7). It was not possible to determine the type of object made within the casting pits of the baths-basilica annexe at Wroxeter (Barker et al. 1997: 81-6). The limited data available do not provide much evidence for the objects that were manufactured. However, it seems that the metalworking involved the production of items for use - iron tools, bronze objects, pewter vessels, and even coins - and structural parts for repairing or constructing buildings.
Most of the evidence for iron-working within public buildings is for smithing rather than smelting, which can be identified through the waste material (Richards 2000: 421). This suggests that iron objects and structural parts were being recycled. The collection of scrap metal has implications for the role of towns, and recycling iron requires resmithing at very high temperatures, which demands significant organisation and skill, whereas bronze can be remelted and reshaped at lower temperatures, making this technologically easier (Tylecote 1976: 22, 162-3). Iron smelting has been identified in some late contexts such as at Silchester in insula IX (Fulford et al. 2006), although not on the site of the basilica. Only one piece of bloomery slag was found on the basilica site, indicating that smelting is unlikely to have taken place within the building (Richards 2000: 421); this suggests that old iron objects were resmelted or new material was brought in and used.
The recycling of metal is also indicated by collections of scrap metal found in towns. The objects from Room 1 of the west range of the basilica at Wroxeter most probably represent a collection of scrap metal for recycling. The same is also likely for the large collection of metal objects found in the temple near the forum-basilica at Wroxeter in the early twentieth century (Bushe-Fox 1914: 9). Metalwork was also ritually charged (see Section 7.3.3), perhaps especially if it was intended for reworking, and may well have been treated with caution. Hoards of metal items and currency bars that could be reworked were ritually deposited in prehistory, the transformative processes of the metalworking adding to the symbolism attached to the objects (Hingley 2006b). Some of the bronze fragments from the collection seem to have originated from statuary, because there were patterns representing hair and drapery (ibid.: 2-9). Certain public buildings may have been chosen as safe depositories for scrap metal and the material may then have been taken to the central forum-basilica for reworking. Other bronze statue fragments associated with metalworking evidence have been found in the basilica buildings at Cirencester (Holbrook 1998:108-9) and Silchester (Fulford and Timby 2000: 72). It seems most likely that towns were able to make use of a range of sources of metal and were able to adapt when new circumstances arose, taking advantage of material when it became available.
7.3.2 Commerce, fabricae, and the Church
One interpretation of the industrial evidence within the public buildings is that of commercial activity. Perring (1991a: 103) argues that the sites of public buildings may have remained in the ownership of the town after their disuse and that they were rented out to craft-workers to generate revenue. Without documentary evidence, however, it is not possible to know for certain whether this occurred. It is equally possible that the public buildings remained in use for public activities whilst some areas were used for metalworking. A number of cases where commercial activities might be suggested include the activity within the portico and laconicum of the St. Margaret’s Street baths at Canterbury and the porticoes of the baths at Wroxeter. The hearths and metalworking identified within the portico of the forum at London also faced onto the street. The metalworking within rooms in the east range of the forum at Lincoln could have exploited its close proximity to the street, as might the metalworking in the Vine Street building in Leicester.83
Some writers have argued that government control may have been an important factor in industrial activities. Mattingly (2006a: 336-7), for example, emphasises that the metalworking evidence within the basilicae at Silchester, Caerwent, and Exeter was relatively well ordered and indicates a regulated reuse of public space, whereas Fulford and Timby (2000: 579), discussing the Silchester basilica, also suggest that centralised control was important; they point to the possibility that the basilica was an imperial fabrica producing weaponry and recycling metal. This is the argument adopted by Faulkner (2000a: 128) in his discussion of urbanism in the later Roman period, which sees towns becoming centres of the state’s 'total war mentality’, with the establishment of military workshops making them 'gloomy police towns of an age of blood and iron’ (ibid.: 130). Much of our understanding of state fabricae comes from late documentary sources, especially the Notitia Dignitatum (Seeck 1876). The Notitia indicates that the fabricae were highly organised establishments, many being devoted to one specific area of production, but there is nothing from any of the buildings in Roman Britain to indicate that production was geared towards any specific item of military equipment. Arcuaria, for example, referred to workshops making bows whereas ballistaria made artillery, hastaria made spears, and sagittaria produced arrows (S. James 1988). The Notitia does record some more general workshops, fabricae armorum, which produced a mixture of weapons and armour (ibid.), but the available evidence does not prove the existence of these in Britain.
The state-controlled fabricae are also a popular explanation for the industrial activity identified within public buildings in other parts of the Empire. At Amiens (Gallia Belgica), for example, part of the forum was levelled, probably in the second half of the fourth century, and then used for metalworking (Bayard and Massy 1983: 239, 252; Bayard and Piton 1979:162). Parts of the building lying outside the excavation could also have been in use at this time, although further excavation would be required to provide this information. This evidence has been interpreted as a possible fabricae (Bayard and Massy 1983: 239, 252), but despite the Notitia Dignitatum mentioning the presence of a shield and sword factory in Amiens there is no positive evidence in the form of inscriptions, documents, or products from the site that can support this. The apparent continued use of the forum also suggests that the metalworking, as at Silchester, was only one part of the function of the complex at this time.
The 'small town’ of Argentomagus (Gallia Aquitania) has evidence of late metalworking in the theatre, consisting of hearths and debris, with most of the evidence concentrated within two corridors (Dumasy 20oo: 218-23). It has also been interpreted as representing the presence of a military workshop for the production of armour and weaponry (ibid.) but, again, there is no positive evidence for this. It is possible that the hearths were placed within the corridors rather than the arena because they produced a lot of smoke (ibid.). This would suggest that the arena was still being used for other purposes at this time and that there was an intensification of use of the building rather than decline. Some kind of officially run factory is often favoured, because this fits in with the Empire’s greater concern for defence in the later Roman period, especially in the centre of the Empire such as Italy (Christie 2006: 216, 312-13). Though fabricae undoubtedly existed, there is a danger in assuming the British evidence can be interpreted in this way.
The Church is another organisation that could well have taken over buildings to carry out industrial activities. Often based on the close proximity of a church, evidence suggests that they took advantage of the spaces to make goods for their own use and as an income (cf. Christie 2006: 214-17). Leone (2003) has demonstrated that, in North Africa, areas of production including olive presses and pottery kilns, often in public buildings and other public spaces, were located near churches. This suggests that the activities were regulated and organised. Moreover, these structures need not have been removed from public use entirely and may indeed have encouraged people to use these spaces more. Excavations within the theatre at Italica in Spain indicated partial robbing of the structure in the fourth century, but at the same time it continued to be used for many activities. Evidence included an oven, a series of hearths, an olive mill, and a bone workshop (Rodriguez Gutierrez 2004). The arena was also used as a cemetery, which indicates a religious focus here and possibly a church.
It is more difficult to consider Church involvement in Roman Britain because of the problematic identification of churches, as we have seen. A lead sealing from one of the rooms in the west range of the Silchester basilica had a chi-rho symbol and the letters ‘PMC’ (for Provincia Maxima Caesariensis) inscribed on it. This might suggest a Christian presence within the building at the same time as the metalworking, although the object may not necessarily have been connected with the Church (Fulford and Timby 2000: 580). The frigidarium at Wroxeter, which leads directly into the baths-basilica, may have been converted into a church (Ellis 2000: 55; White and Barker 1998: 125), but this is uncertain and there are also the problematic timber buildings within the Lincoln forum where there was metalworking in the east range (Steane 2006: 54-7).
7.3.3 Metalworking, space, and ritual
Stirling (2001: 69) has argued that public buildings provided convenient locations for metalworking because oftheir stone walls, high roofs, and large spaces, and their proximity to roads and water. These are important considerations, but in Britain it appears that many instances of the metalworking took place within rooms or areas on the edges of the buildings or concentrated at the ends of main rooms within them, such as in the basilicae at Caerwent, Cirencester, London, and Silchester. This implies that central parts of the buildings were still being kept for other purposes. The laconicum of the St. Margaret’s
Street baths in Canterbury lay in the furthest southeast corner of the baths and the piscina lay on the western side, so other parts of the baths may well have remained in use or have been put to different functions. People will have continued to congregate and interact in these buildings at the same time as the metalworking was being undertaken. They were not simply derelict spaces for reuse. Local governments may have continued to use the buildings and perhaps the metalworking was an aspect of their continued power. Its concentration within the public buildings, especially the forum-basilica, may also indicate a social role for the metalworking, perhaps drawing on pre-Roman practices, where people gathered, interacted, carried out ceremonies, and feasted, with the metalworking forming a focus to the activities.
Linked with this is the potential religious significance of the metalworking, which could have had an impact on the way in which the activities and their spatial context were perceived. Studies of Iron Age metalworking have emphasised its religious symbolism, moving beyond purely technological considerations (e. g., Aldhouse-Green 2002; Giles 2007; Hingley 1997b). This aspect has not received so much attention in Roman studies, especially of the late Roman period. Recent work on iron deposition and the significance of iron in the Roman and late pre-Roman periods is an important exception (Haselgrove and Hingley 2006; Hingley 2006b). The evidence from Roman public buildings can be analysed in the light of reappraisals of industrial processes in the past as meaningful beyond the basic parameters of economic production.
Ethnographic studies, especially on iron production in Africa, demonstrate there are alternative ways of thinking about metalworking that contrasts with Western technological and economic understanding. Though, of course, not directly applicable to the Roman West, studies have shown that iron production often involved complex rituals, with smiths sometimes holding the status of religious leaders. Haaland’s (1985) work on iron production in the Sudan and Reid and MacLean’s (1995) in East Africa demonstrated links between iron production and human fertility and procreation, with smelting considered a procreative act in which furnace and bellows took the roles of the sexual partners (ibid.: 149). The cognitive link between iron production, transformation, and human procreation was also the major theme of Herbert’s work Iron, Gender and Power (1993) on the smelting rituals of a number of groups in Africa.
Other ethnographic studies connected with metalworking include Barndon (2004) and Gansum (2004), but there are far fewer studies of these aspects of technology within archaeology. Though it is not wise to use ethnography to provide 'answers’ about the past, it can give examples of how actions and processes can be conceptualised in different ways, which is vital for archaeology. Moving beyond modern Western assumptions can help to demonstrate that people in the past will have experienced their worlds differently. Hingley’s (1997b) study of British Iron Age iron production drew on Herbert’s (1993) work to examine possible ritual activity associated with smelting and smithing sites. He examined especially the possible symbolism of passage and rebirth in the location of the metalworking on sites. Aldhouse-Green (2002) has considered links between collecting iron ore, producing iron and agricultural activity, cooking, and nourishment;4 whilst also working on the Iron Age, Giles (2007) has examined the symbolic links between metalworking, agricultural production, and procreation. The significance of iron in pre-Roman Western Europe is
4 There are also hoards of deposited metalwork, including iron sickles and cauldrons, which might indicate this
Link to agricultural activity and cooking, such as that from the Llyn Fawr lake in South Wales (Aldhouse-Green
2002: 13).
Also implied in classical texts. Tacitus records in the Germania (XL.2-5), for example, that a number of the tribes worship Nerthus, a goddess of agriculture, and they could only carry out their annual festival once all iron objects had been hidden away (Aldhouse-Green 2002: 9). Blacksmiths may have had special status in pre-industrial periods, as suggested in texts from early historic Ireland.84
The inclusion of iron-smiths’ tools within ritual hoards indicates the special significance of iron-working in the Iron Age and the continuation of these attitudes and beliefs in the Roman period. Relevant hoards include those at Llyn Cerrig Bach on Anglesey (C. Fox 1946), Waltham Abbey in Essex (Manning 1972:231), and Fiskerton in Lincolnshire (Field and Parker Pearson 2003). The tools are usually in good condition, suggesting that they were not thrown away because they were broken; nor was the act likely to relate to changes in technology. The finds do not consist of complete sets of smith’s tools, but they are often accompanied by other types of objects, which argues against their being deposited for safekeeping by craftsmen (Manning 1972: 238-9). The significance of iron-working and its products is also inferred by the large number of Iron Age and Roman date iron-work hoards catalogued and studied by Manning (1972) and more recently by Hingley (2006b). Given the symbolism attached to iron objects through their manufacture, the number of accidental losses would have been small; even iron nails had religious meaning and they were used in religious activities such as defixio (Dungworth 1998:153).85
Iron artefacts are not the only metal items known in hoards: Poulton and Scott (1993) have documented a large number of late Roman pewter hoards across Britain often found in wells, rivers, and pits, suggesting a conscious association of religiously imbued metalwork with watery places. These hoards are often regarded simply as dinner services deposited for safekeeping (e. g., Brown 1973: 201-4; Lethbridge and O’Reilly 1933: 166),86 but the collections rarely make sense as dinner services and they often appear unused even though this material can be easily marked or damaged. The pieces may even have been deliberately produced for deposition, which has implications for the significance of metalworking within public buildings. Likewise, Budd and Taylor’s (1995) examination of bronze-working has emphasised the need to put the 'magic’ back into studies of ancient metalworking and attempt to conceptualise the activities as ritual and symbolic rather than economic. The procedures of bronze-working would have been passed on through rituals and spells, and metalworking would have had a special place within prehistoric society. Like other ritual activity and ideas about space and place, this undoubtedly continued in the Roman period (cf. Merrifield 1987). Metalworking activities were bound up with ritual and belief, their presence within public buildings bringing notions of regeneration and renewal and forming a focus for associated activities.
Within the annexe of the baths-basilica at Wroxeter, contemporary with or directly succeeding the bronze-working, the bodies of four foetuses or newly born babies were deposited, one of which was directly within a casting pit and another next to a post-hole of a timber structure that was associated with the metalworking (see Figure 7.7; Barker et al. 1997: 8i-6). Metalworking continued here after the baby burials, with new hearths constructed that were associated with metalworking debris (ibid.: 83-4). A further burial was in the floor near the exit from the northern half of the annexe to the southern half, whilst the fourth was that found by Kenyon in her 1930s excavations in the northwest corner of the north room (Kenyon 1938: 188). There are some classical sources (e. g., Plin. HN VII.15; Plut. Mor. 612a) referring to infant burials being placed within settlements rather than in cemeteries with graves and burial rites, because it was believed that as children the infants did not possess souls and they were not yet part of the earthly community; nothing survived them after death.
In Britain, however, there were few infant burials known in either settlement or external contexts (Philpott 1991: 97), and it is likely that local practices also continued and infant bodies were treated in a variety of ways.87 At Silchester, for example, the bones from a minimum of four infants were discovered, apparently having been deliberately placed within pits not as complete skeletons and combined with animal bones (Snelling 2006: 200-5). The clear concentration of burials within the baths-basilica at Wroxeter associated with the metalworking does, then, appear to be significant, demanding explanation, perhaps relating to ideas of regeneration attached to the metalworking. At Leicester two baby burials were found in the same location as metalworking within the Vine Street building (T. Higgins personal communication), which may also support the connection between ritual and metalworking. This association may have had symbolic and ritual meanings linked with regeneration on a wider scale, the practice drawing on pre-Roman beliefs, which Scott (1991: 119-20) saw as integral to the infant burials that she studied on Roman rural sites.
In the basilica at Silchester (see Figure 7.5) there is also evidence that indicates religious activity associated with the metalworking. Contemporary with the phase of iron-working that concentrated at both ends of the hall, was a tiled area, laid down within the hall, now represented only by a few remaining tiles (Fulford and Timby 2000: 74-5). There were also traces of slots indicating a timber structure that may have enclosed it (ibid.). There was markedly less iron slag and other iron-working debris in this vicinity, which would suggest that it was deliberately kept clear. There was also a large number of late-third - and early-fourth-century coins in this area and two oak leaves cut from sheet lead (although one was unstratified; ibid.: 72-5, 578). These finds indicate that a domestic interpretation for the structure is unlikely and that the evidence may represent a shrine, in use at the same time as (and perhaps associated with) the industrial activity.
Well F127 and adjacent pit F107 in the nearby north range of the basilica contained a large number of bird and fish bones as well as sheep bones, the remains of two neonatal pigs, and iron slag (ibid.: 69-71). Pit Fi8 within the northern area of the metalworking, near the tiled area, and the general occupation layer within the building (Phase 7.13) also produced a high incidence of bird bones.9 Fulford and Timby (2000: 577-8) argue that this may relate to feasting and sacrifices taking place within the basilica and is unlikely to represent solely domestic occupation. Its association with the metalworking evidence points to these religious and ritualised activities being intertwined with the industrial processes. Such an interpretation may also be supported by the evidence from the rear range of the principia at York, where there was a late-fourth - or fifth-century horizon of young pig bones contemporary with the metalworking debris (Phillips and Heywood 1995: 64). The large assemblage of young pigs may suggest elite activities, perhaps feasting, rather than simply butchery on the site (ibid.). The deposition of waste material into pits at the end of the metalworking process, as found in the Exeter basilica (Bidwell 1979: 110-11), may also relate to the religious significance of the material (Giles 2007; Hingley 1997b) and tie into the broader practice of ritual pit deposits (e. g., Fulford 2001; J. D. Hill 1995a). Within the basilica at Cirencester, Room 1 was blocked off around the same period as the metalworking activity within an adjacent area of the building. Nineteenth-century excavations within the room by Wilfred Cripps uncovered a large number of oyster shells together with a dog skeleton (Holbrook 1998: 108). Dog burials have been associated with ritual deposits both in prehistory and the Roman period (e. g., Smith 2006), and it is possible that there was some kind of religious activity taking place that was connected with the life of the building and involving the metalworking.
The industrial activity indicates that the public buildings in the heart of the towns were real and symbolic centres of people, production, and regeneration, and that metalworking, and the rituals associated with it, represented renewal and continuity. It is even possible that the industrial activity, and its symbolic significance, was a response to perceived changes to the towns. It would also have altered movement and enriched experiences within the buildings, contrasting with earlier organisations of space, but other parts of the buildings may have been used for their original functions, indicating the intensification of use of the structures, including meetings and the curia. A controlling authority is likely, suggesting continued leadership within the towns as a whole. Metalworking within public buildings in the later Roman period could be a symbolic representation of the vitality of towns at this time. In the late pre-Roman Iron Age, metalworking and other activities were integrated with the elite presence and religious nature of sites (cf. Haselgrove and Millett 1997). Combining practical, political, and religious considerations, the forum-basilica and other Roman public buildings could have now been adapted to fit local traditions where the forge or hearth was central to meetings, networking, social activity, power, and ritual. These were places of events and actions and not the decaying vestiges of one-time greatness.
9 Serjeantson's (2000: 500) study of the bird bones from Period 7 highlights the high proportion of these bones in
The assemblages compared with other animals and suggests this is unlikely to reflect overall food consumption
Patterns across the town, thus indicating special activity.
World History
