Download file in PDF format: TRAC 2004: Metalworking and Late Roman Power: A Study of Towns in Late Roman Britain (pp. 27–38)
Metalworking and Late Roman power: a study of towns in later Roman Britain
Introduction There is much evidence for activity within towns in the later Roman period in Britain but this has not received sufficient attention from a theoretical perspective. This is problematic since, as Esmonde Cleary points out, the ‘Golden Age’ and the fourth century cover similar lengths of time (1989a: 235). It is also symptomatic of the social constructs within which archaeologists operate. The concept of decline and fall was initially advanced by Edward Gibbon’s The Decline and Fall of the Roman Empire (published between 1776 and 1788). Here it will be argued that, rather than representing past reality, this concept is in fact a result of both modern social constructs, derived from eighteenth and nineteenth century ideas, and classical notions that a ‘golden age’ was replaced by ‘iron and rust’ (a Roman elite concept). In this context, this paper examines the trend of metalworking activities within the disused or demolished public buildings of towns in the later Roman period of Britain (late-third and fourth centuries). This obvious focus within these central locations must have had a greater importance than that suggested by functional arguments and ones of decline. This evidence is often neglected because it has been seen as an inferior activity occurring in the context of declining grandeur. It will be suggested here that the metalworking may not only have had practical uses but also had a far more important ritual motivation, associated with the regeneration and rebirth of the significance of central locations as represented by the public buildings.
Interpretations past and present Decline has been the central theme for most interpretations of towns in the later Roman period in Britain. Wheeler described later Roman Verulamium as a bombarded town and a nucleated slum (Wheeler and Wheeler 1936: 30) whilst Frere, also in Verulamium, envisaged some Roman-style buildings of a late date but these were contemporary with timber buildings of lesser and uncivilised occupation (Frere 1966: 87, 97). Similar biases towards views of civilised living without a consideration of a different, yet equally important, town life in the later Roman period were expressed by Wacher (1975: 411-422) and Reece (1980: 77-85). For Esmonde Cleary, too, there was a “decline in standards” and then an end that was “nasty, brutish and short” (1989b: 132, 161). The concept of decline is also central to Faulkner’s work (1994; 1996; 2000) in which he envisaged “police towns of an age of blood of iron” (2000: 124).
These attitudes have influenced the way in which metalworking within the public buildings in the later Roman period has been considered. This activity is often merely interpreted as evidence for the decline of the public buildings and town life in general. Amongst more considered approaches Faulkner sees towns replacing forts and vici as military workshops with the metalworking under the control of the government (Faulkner 2000: 128). Perring, referring to London, suggested that the city was attempting to find revenue for its public sites through renting out the land to industry (Perring 1991: 113). He also suggested that the state might have intervened to ensure the maintenance of essential services (1991: 103). Likewise, Fulford and Timby, on Silchester, suggested that control is implied through the use of the forum-basilica for metalworking (Fulford and Timby 2000: 579). Inherent within these interpretations, however, is the concept of decline from the preceding golden age but as an alternative we can interpret this stage in the development of towns in a similar light to interpretations of earlier periods.
The origins of the construct of decline and fall Archaeologists have inherited reconstructions of the past that are firmly rooted within the culture and period that produced them (Terrenato 2001: 72). The concept of decline and fall used in Romano-British studies is one such example and, in order to aid understanding of the archaeology, we need to reconsider its relevance to the period in question. The search for the origins of these negative views of late Roman Britain, decline and fall, must include a brief examination of the life and work of Edward Gibbon. The Decline and Fall of the Roman Empire (1776-1788 ) has decline – leading to fall – as its dominant theme and this work has had a great impact on the way in which academics, politicians and others have thought about and conceptualised change. As a result of his public school and Oxford education and the Grand Tour (Burrow 1985: 5, 8, 13), Gibbon developed a great admiration for civilisation. He also relied heavily on, and was influenced by, classical texts (Matthews 1997: 30) in which the concept of a ‘golden age’ decaying to ‘iron and rust’ has its origins (Dio Epit. 72.36.4; Hesiod: 13). Traditionally the later Roman period has been judged by the standards of the earlier period without consideration of any positive attributes (McKitterik and Quinault 1997: 5; Porter 1988: 139); these attitudes and methods have influenced the way in which the metalworking within the public buildings has been studied. The idea of the golden age, then, was simply a social construct, a product of the elite, and thus it becomes apparent that decline and fall was also an elite conception with no necessary relevance to the masses; classical authors recognized the decline of their section of society (Pocock 2003: 18).
Gibbon’s work, however, was also a product of eighteenth century ways of thinking (McKitterick and Quinault 1997: 2) which, stemming directly from the Renaissance, were firmly rooted on concepts of Roman civilisation and the classical past. The Renaissance period saw a new focus of interest in Roman history and culture and drew upon classical ideals (Barkan 1999). By the seventeenth and eighteenth centuries, when Gibbon was writing, the demand for classical antiquities, the concept of civilised living and the admiration of all things Roman had reached a new height (Moatti 1989: 21, 59). It was the beginning of a particular association between classical civilisation and the modern elites which was to remain into the twentieth century and impact upon archaeological thought (Ayres 1997: 165; Hingley 2001: 149). The theory of romanisation in the early-twentieth century, for example, developed from the concept that the Roman ways of life were more civilised and brought benefits to the ‘barbarian natives’. This also meant that at the end of ‘civilisation’ and the ‘golden age’ there was simply ‘decline’ to ‘iron and rust’ and this whole framework of thought has had a detrimental affect on the way in which towns in the later Roman period have been studied. Throughout the twentieth century Romano-British archaeology has envisaged romanisation as progress (Hingley 2000: 140–142). Only recently has a post-colonial approach begun to address the issues of elite and non-elite and civilised and uncivilised (e.g. Webster and Cooper 1996). There is still a long way to go, however, since ideas, myths and social attitudes within society are difficult to break.
The industrial activity within public buildings Evidence for metalworking within towns in the later Roman period has been systematically collected by the author (Fig. 1). This shows a clear trend towards the location of such activities within public buildings, the central sites of towns such as fora-basilicae and the public baths (Fig. 2). Data from excavations was gathered for the twenty-one ‘large’ towns which feature in Wacher’s The Towns of Roman Britain (1995, second edition). The number of excavations that have taken place in each town vary greatly and it should be acknowledged that more evidence may remain undiscovered. It will be argued here that as well as having a practical side, to supply metal equipment to the townspeople at a time when supplies of newly-produced metal was low, the metalworking also had a far greater significance in ritual terms and to the lifecycle of the public buildings and the power of central sites. This symbolic aspect was perhaps derived from the growing importance of metalworking within central locations demonstrated in the late pre-Roman Iron Age oppida (e.g. Blockley 1995: 51; Fulford and Timby 2000: 12; Niblett 2001: 43).
Town No. of public buildings with metalworking evidence Aldborough 0 Brough-on-Humber 0 Caerwent 1 Caistor-by-Norwich 0 Canterbury 3 Carmarthen 0 Chelmsford 0 Chichester 0 Cirencester 2 Colchester 0 Dorchester 1 Exeter 1 Gloucester 1 Leicester 1 Lincoln 0 London 2 Silchester 1 Verulamium 1 Winchester 1 Wroxeter 2 York 0 Figure 1: Table showing the number of public buildings with metalworking within each town. The evidence corresponds well with the level of archaeological investigations in each town.
0 1 2 3 4 5 6 7 8 Forum-basilica Public baths Temple Theatre Amphitheatre Other Public building Frequency Figure 2: Graph showing the frequency of public buildings with identified metalworking from the twentyone towns considered in the table of Figure 1.
Metal production in Roman Britain It is necessary to examine the history of the extraction industries and the evidence for the availability of metals for recycling. Firstly, the iron industry will be examined since there is evidence for iron smithing in a significant number of the public buildings in the later Roman period including the Silchester basilica (Fulford and Timby 2000: 576), Caerwent basilica (Brewer 1990: 86) and within the laconicum of the public baths at Canterbury (Blockley 1995: 190). The Weald was one of the main iron-producing centres in Britain during the Roman period. Production probably started immediately after the invasion but most sites had reduced or ceased production by the late-third century (Cleere and Crossley 1995: 39). Likewise it seems that in the iron producing area of East Yorkshire few sites were still in operation by the late-third century (Halkon and Millett 1999: 48). Iron production in the Forest of Dean area continued into the fourth century but, although demand for iron would not have decreased, sites produced far lower quantities than those of the Weald (Fulford and Allen 1992). Lead-working has been identified at a number of sites including within the Macellum at Wroxeter (Macalister 2000) and the fora-basilicae at Cirencester (Wacher 1962: 7), Leicester (Hebditch and Mellor 1973: 40) and Dorchester (Holbrook 1998: 121) in the later Roman periods. Few leadproducing sites have been excavated but those at the small town of Charterhouse in the Mendips (Todd 1993; 1994) show that lead production had begun before the Roman conquest but by the fourth century production appears to have ceased (Burnham and Wacher 1990: 211). Lead pigs with stamps cease after the mid-second century but this may simply suggest that industries were now in private hands (Tylecote 1962).
From around A.D. 250 onwards, however, lead was also needed for the manufacture of pewter (Beagrie 1989: 175; Brown 1976: 25–26), the moulds for which have been found in the forum-basilica at Silchester (Blagg and Read 1977; Fulford and Timby 2000: 576) and within
the bath building at Gloucester (Heighway and Garrod 1980: 78). Bronze-working, which would have needed lead and tin, for casting, has also been identified on a number of public building sites including the demolished Exeter forum-basilica (Bidwell 1979: 110–111), the Ridingate in Canterbury (Blockley 1986) and the forum-basilica in Wroxeter (Atkinson 1942; White and Barker 1998: 117). It is highly probable that the metal for these activities was obtained from recycling. Scrap metal, then, would have been an important resource for metalworkers in late Roman Britain (Manning 1976: 143) and it is certain that the metal elements of public buildings would have been subject to salvage. Iron clamps were used with stonework and their systematic removal is evident from the rubble of the Temple of Sulis Minerva at Bath (Cunliffe 1984: 212) and from the excavations at the Ridingate at Canterbury (Blockley 1986). Bath buildings used iron clamps to attach box-flue tiles to the wall and these, as well as lead pipes, were robbed from the St. George’s Street bathhouse, Canterbury (Frere and Stow 1983: 38) and the Huggin Hill Baths in London (Marsden 1976: 20–21). Figure 3 shows that five bath buildings contained evidence for metalworking in the later Roman period; few bath buildings have been investigated to an extent that enables this evidence to be identified. Fora-basilicae could also have contained cast bronze statuary as well as furniture fittings and structural decoration.
Category No. of bath houses With metalworking 5 Without metalworking 1 Insufficient evidence 16 Figure 3: Table showing the frequency of bath buildings with metalworking in the later Roman period.
This activity is comparable with the re-use of stonework which is suggested by the deliberate demolition of the fora-basilicae at Verulamium and Exeter. Such an undertaking is likely to have demanded considerable organisation and manpower (Bidwell 1980: 86; Niblett 2001: 143). Rather than representing decline, this activity was opportune, systematic and dynamic and thus a form of town life. The material may have been used to invoke the power of the past and relate it to the present time (cf. Eaton 2000). It is evident that it was not simply salvaged scrap from buildings that was recycled but also objects such as chain fragments, pins, toilet instruments, studs and hinges as found associated with the metalworking in the Wroxeter Macellum (Ellis 2000: 56). In the Gloucester bathhouse, too, there were hinge-fittings, clenchbolts and a lock clasp associated with the later-fourth century metalworking (Heighway and Garrod 1980: 78). All this evidence is connected with people living their lives in positive ways. Much of the metalworking occurred on demolished sites of the public buildings (e.g. Exeter forum-basilica and the London Huggin Hill baths) or with evidence for ritual deposition and feasting connected with the metalworking inside the shell of the public building (e.g. the forabasilicae of Silchester, Cirencester and Wroxeter and the Wroxeter baths). This suggests, possibly, the coming together of people for the special event; analogies with African societies show that people gather together for metal production and working because they are important activities (e.g. the Ekonda of Zaïre, Herbert 1993: 63–65).
Memories of the importance of public buildings continued to have significance through their role as locations for metalworking and their demolition and re-use in the Late Roman period. The sites would have “tugged memories which gave access to the past which in turn could have given potency to the present” (Hope 2000: 79). There was a late Roman component to the buildings’ lives and to the towns in general. To interpret this as decline and lesser occupation is judgemental since people lived their lives in the context of this period of history just as fully as in the ‘golden years’ of the empire. Metalworking within these public buildings would have been an important activity and cannot be ignored; the ritual significance of the metalworking intensifies this.
The symbolic meaning of metalworking There is a growing realisation that pre-modern industrial activities, especially metalworking, cannot be understood through thinking in functional terms alone (e.g. Aldhouse-Green 2002; Budd and Taylor 1995; Hingley 1997). Most of the work, however, has focused upon prehistory rather than the Roman period, but since it is likely that much of the pre-Roman way of life continued through the Roman period it can be suggested that metalworking retained a symbolic and ritual significance in late Roman Britain. For the Iron Age, Hingley (1997: 9–10) argues that there was an association between iron production and generation/regeneration and that the life-cycle of objects was viewed in similar ways to the life-cycle of people, crops and animals. Metalworking was thus associated with regeneration, re-birth and productivity. The analogies with African iron-production and working used by Hingley (1997) are useful here. Many smelting rituals in Africa are often connected with the view that such activities are similar to human regeneration through sexual reproduction (Herbert 1993: 234; Reid and MacLean 1995). Smelters of the Ader Hausa of Niger wash and decorate their furnaces, treating them as if they were virgin brides. The ‘bride’ is impregnated by the fire during the smelt and then ‘gives birth’ to a ‘child’ in the form of new iron. The slag is treated like the placenta and buried (Herbert 1993: 58). Although iron production will have had different associations in different periods it is possible that the act of recycling metals in the public buildings in the later Roman period was seen as an act of regeneration and rebirth. The context in which it occurred may have formed part of an attempt to regenerate the power of the public building.
There is certainly much evidence for ritual activity connected with the metalworking in the public buildings of the later Roman period. This includes both ritual sacrifice and feasting whilst the pewter vessels produced were perhaps associated with ceremony and power. Within the forum-basilica at Silchester, Fulford and Timby (2000: 578) have suggested that an altar, indicated by a tiled area, may have been part of a shrine associated with metalworking while the high number of fowl bones found in the vicinity could represent sacrifices. The forum basilica at Cirencester produced evidence of oyster shells associated with the metalworking (Holbrook 1998: 109). At Wroxeter, the disarticulated remains of a foetus or newborn child were found within a casting-pit connected with the metalworking within the Annexe of the baths-basilica (Barker 1997: 82). Baby burials may represent the ritual practice of revitalisation (Scott 1991: 119) and if so this would complement the metalworking activity. Renovation, and thus regeneration, of the forum-basilica of Cirencester involved lead- and bronze-working. Room One, which was blocked off from other rooms, had a deposit of oyster shells, a skeleton of a dog and some pottery. These deposits all contribute to the argument that metalworking was a significant activity and this is increased by the fact that it was taking place within central locations: the public buildings.
As highlighted earlier, pewter moulds have been found on the forum-basilica site at Silchester (Blagg and Read 1977; Fulford and Timby 2000: 576) and at the baths site in Gloucester (Heighway and Garrod 1980: 78). Poulton and Scott (1993) have discussed the ritual significance of pewter artefacts and their possible use as status items as represented in their hoarding; this is part of a growing literature concerning the deposition of metalwork and other non-functional uses of metal items in the Roman period (e.g. Clarke 1997; Dungworth 1998; Manning 1972; van Driel-Murray 1999). It is clear, then, that the metalworking within the public buildings would have been an important activity and should not be neglected or interpreted as representing the decline of the towns. Religious symbolism of renewal, regeneration and rebirth dominated activities taking place on the sites of the public buildings.
Central sites, regeneration and the metalworker It is clear that there was an obvious association between central places and metalworking in the urban centres of the later Roman period. Metalworking does occur in other places such as townhouses, Winchester and Cirencester being notable examples (Holbrook 1998; Zant 1993) along with the recently excavated Building 8 in Insula IX at Silchester dating to the earlier Roman period (Clarke and Fulford 2002: 152), but this practice, from the existing evidence, seems not to have been very common. During the earlier Roman period in Britain, when metal extraction and production was at its height, there was more evidence for metalworking within towns but, as Millett suggests (1990: 127), this begins to fall off at the beginning of the later Roman period due to decentralisation of the settlement hierarchy. It can be suggested that this metalworking within public buildings may have been a symbolic renewal of industrial activity within these central places, the centres of the civitates, and emphasised by its occurrence in the centres of the towns.
This can be supported further with evidence from the late pre-Roman Iron Age which saw the development of central places: the so-called oppida sites. It is being increasingly suggested that the LPRIA marked great social changes with the rise of central sites, political centres of a territory, which did not exist before this time (Creighton 2000; Hill 1995; contra Cunliffe 1991). Significantly here, Creighton (2000: 10) suggests it is gold and gold-working, with the introduction of coinage, that led to these changes and the rise of hierarchies. The adoption of coinage and minting must represent significant social changes (Haselgrove 1996: 67). It is possible that the role of these central sites in gold-working may have enhanced the position of other metalworking and metalworkers. Gold was related to authority but iron, with its military and agricultural significance (Hingley 1997: 13–15), must also have been connected with this. Iron-working has been evidenced at such sites as Verulamium, Camulodunum and Silchester. Discussions of the status of the metalworker have to take into account their religious significance (Fitzpatrick 1984: 184); metalworkers, especially it seems iron-workers, may have possessed sacred skills linked to secret processes (Henderson 1992: 119). Creighton suggests that gold- and bronze-workers would also have had similar powers (Creighton 2000: 37). Studies of early Irish and Scottish literature suggest that the smith was regarded as possessing supernatural powers with the ability to protect, heal and harm which would have set them apart within society (Aldhouse-Green 2002: 16; Gillies 1981: 73; B. Scott 1986: 154–155). The ritual deposition of metalworkers’ tools may support these arguments (Aldhouse-Greene 2002: 16; Manning 1972: 246).
The Roman invaders may have recognised the significance of these central sites, and the special status given to metalworkers within Iron Age society, and attempted to suppress and dominate it. It is likely that the banning of weaponry and war-related artefacts, after the conquest, disrupted traditional modes of display and the role of native metalworkers (Potter and Johns 1992: 85). The special status and mythical powers of the metalworkers, however, may have continued amongst the native peoples of Roman Britain; the processes involved were perhaps remembered through rituals and ‘spells’ (Budd and Taylor 1995: 139). It is possible, then, that the metalworking within late Roman public buildings represented a revival not only of the buildings themselves, in acts of regeneration, but also of pre-Roman practices which were symbolically placed within the centre of the towns. Collingwood, in the 1930s, was the first to suggest a Celtic revival highlighted through metalwork and pottery designs (Collingwood and Myres 1936: 256–258); his explanation for its survival, however, was rather weak. The importance of history and myth within societies may have been more significant (Gosden and Lock 1998: 11). The survival of the power of the central sites and the role of the metalworker through rituals, myths and stories would enable the revival, although more work is required to document how this potential survival may have come about.
Conclusion More stimulating approaches to the later Roman period in Britain and the later Roman period in general, are needed in order to revitalise our understanding of the archaeology. Studies of towns in the later Roman period in Britain have been preoccupied with socially constructed assumptions of decline and fall in which Gibbon’s The Decline and Fall of the Roman Empire has played a major part. Through examining the evidence for metalworking within late Roman public buildings it is clear that there is rich evidence for activity in towns during the late Roman period and the concept of decline should not be used to evaluate it. Metalworking, and metal recycling within the public buildings may be seen as an act of regeneration and revitalisation of central sites that had been so powerful in the late pre-Roman Iron Age and earlier Roman period. It may also represent a revival of the important religious and ritual role of the metalworker.
Department of Archaeology, University of Durham
Acknowledgements This study represents research for my undergraduate dissertation at the University of Durham. I would like to thank the following people for their much appreciated advice, comments and assistance: Dr. R. Hingley, Dr. G. Milne, Dr. J. Shepherd, Mrs. P. Gamm, my mother and father, Catrine and my brothers Graham, Matthew and Jonathan.