bronze age

4.5.1 From raw material to reduction to treatment to craftsmanship

Bronze Age Metalwork in Scotland c. 2500-800 BC

This section reviews the current state of knowledge concerning metal production, circulation, use and deposition in Scotland during the Chalcolithic/Bronze Age (c. 2500-800BC). This encompasses copper, copper alloys such as arsenical copper and tin-bronze, gold, tin and lead. It does not seek to duplicate the research on the historical development of Bronze Age typo-chronological sequences (see section 1), propose typological connections beyond Scotland (see Gerloff 2007; Roberts et al. in press) or provide a historiography of metal and metallurgical research in Scotland. This section follows the evidence for the prospection of the potential ore and native metal sources through to the final deposition of the metal objects (see Ottaway and Roberts 2008). It builds on the recent review of research priorities for archaeometallurgy in Scotland which encompassed the Bronze Age (Hunter et al. 2006).

Prospection, Extraction and Ore processing

There is currently no conclusive evidence for metal prospection, extraction or ore processing dating to the Bronze Age in Scotland. Lead and copper ores are found throughout the country (Wilson and Flett 1921) and there are also widespread alluvial gold deposits.Most of these deposits occur as surface outcrops and could have been prospected for and worked in prehistory.Copper ore sources accessible to prehistoric prospectors and miners (Rice 2004) as are currently being surveyed at Tonderghie, southwest Scotland (Sheridan 2008, 69) have yet to be subjected to the intensity of fieldwork seen in Wales and Ireland (see Timberlake 2009). Museum collections and antiquarian accounts provide tantalising potential evidence for Bronze Age mining.There were also alluvial gold sources as at Leadhills and Wanlockhead, south Lanarkshire and Helmsdale, Sutherland exploited during the post-medieval period and later which have been extensively surveyed and accompanying micro-chemical characterisation led by Robert Chapman at Leeds University.

During the late eighteenth and early nineteenth centuries lead mining was a significant rural industry in parts of Scotland.  The principal lead mines were in Galloway and the Wanlockhead-Leadhills area of the Southern Uplands, in Islay and mid Argyll and at Tyndrum and Strontian in the West Highlands.  Copper ores are found in Galloway and the Southern Uplands, in the Ochills, in mid Argyll, around Loch Tay and on Shetland.  Most of these copper deposits were worked, albeit as trial prospects, during the eighteenth and nineteenth century but none were very productive.  The alluvial gold deposits of the Southern Uplands - principally in the Wanlockhead-Leadhills area and at Glengaber Burn close to St Mary's Loch - are some of the most extensive in Britain and were worked during the sixteenth and seventeenth centuries.  The small but locally rich alluvial deposits at Kildonan in Sutherland have also been worked and were the focus of a mini-gold rush in the late nineteenth century. See http://www.see.leeds.ac.uk/research/essi/people/chapman/index.htm

During the last twenty years there have been major advances in our understanding of Bronze Age metal mining in Britain and Ireland. Excavation at Ross Island, Co Kerry, has identified a major Early Bronze Age mine which is the probable source of A-Metal copper while fieldwork and excavation in mid and north Wales and the English Midlands have led to the discovery of eleven copper mines with radiocarbon dates in the Early Bonze Age (Timberlake 2003). All these sites have produced a distinctive range of stone tools used in rock breaking and ore crushing and which are often found in association with charcoal, burnt rock and other debris from firesetting.

A single nineteenth century account of stone and bronze tools from mines at Leadhills and a handful of coarse stone tools from Galloway have been proposed as putative mining implements. Additional indirect evidence for Bronze Age mining includes a now lost copper axe or ingot hoard found near a copper ore deposit at Tonderghie, Dumfries and Galloway and the Early Bronze Age lead beads from West Water Reservoir, Scottish Borders, which have a lead isotope signature suggestive of a Southern Uplands lead source (Hunter, Cowie and Heald 2006).There is no evidence for Bronze Age gold mining although the recent discovery of a looped spearhead within or close to the alluvial gold deposits at the Mennock Water, Wanlockhead is of potential interest and can be paralleled with the items of Bronze Age metalwork recovered from alluvial tin deposits in Devon and Cornwall.

Aerial image of the remains of Tonderghie copper mine, Dumfries and Galloway ©RCAHMS

The identification of ore sources exploited during the Bronze Age through the analysis of surviving objects has been the subject of fierce, and as yet unresolved, debate. The compositional analyses of Bronze Age copper and copper alloy objects found in Scotland is dominated by the Studien zu den Anfängen der Metallurgie, or SAM Project, (Junghans et al. 1960; 1968) which subsequently also included gold objects (Hartmann 1970, 1979, 1982). The original purpose of the SAM project was to trace objects back to ore sources and therefore reconstruct Chalcolithic and Bronze Age trade routes. The inability to address issues such as the complexities created by the recycling and mixing of metals or the heterogeneous nature of ore sources led to their findings being challenged. The introduction of more sensitive techniques also led to questions regarding the accuracy of the original measurements (see reviews in Northover and Rychner (1998) for copper alloy objects and Warner (2004) for gold objects). The recent excavation of early copper mines in Wales and Ireland has provided sufficient stimulation to re-analyse the older data in the light of new research. This has demonstrated the probable use of copper ore sources from Ireland for Chalcolithic-Early Bronze Age metalwork in Scotland (Peter Bray pers. comm.) implying that copper mines in Scotland would probably be later in date. The estimated quantities of copper ore being mined in Wales and Ireland during the Chalcolithic-Early Bronze Age exploitation would certainly have been large enough to produce the number of copper and copper alloy objects known from Scotland (Timberlake 2009).

Four striking gold discs were found with a cremation in the principal mound in a barrow cemetery in Orkney. Demonstrating beautiful craftsmanship, it has been suggested that the gold discs represent the sun. Image © National Museums Scotland.

Recent research and re-assessments of gold analyses in Ireland with reference to Scottish gold objects have demonstrated the potential differences that exist in the compositions of chronologically distinct object types. This has led to the suggestion that gold ribbon torcs (see Eogan 1983) may primarily date to the Iron Age rather than the Middle Bronze Age as previously supposed (Warner 2004). However, beyond the recent analysis of the Chalcolithic sun discs from the Knowes O' Trotty, Orkney indicating a probable Scottish provenance, there has been no comparable Bronze Age gold provenance project to that in Ireland (Warner et al. 2009). The use of lead isotope analysis to provenance copper and copper alloy objects has not been widely applied in Scotland in comparison to England and Wales (Rohl and Needham 1998). The isotopic analysis of a rare Early Bronze Age lead and cannel coal bead necklace found at West Water Reservoir, Peebleshire revealed a signature which is consistent with the southern upland sources (Hunter and Davies 1994, 2000; Hunter et al. 2006, 50). The highly probable movement of tin ore or tin metal from southwest England to Scotland, as evidenced not only by the tin-bronzes from the county but also the inlaid tin jet button from Rameldry, Fife (Baker et al. 2003) cannot be scientifically traced although a recent study demonstrates future potential (Haustein et al. 2010).

Geomorphology and geochemistry have great potential for indicating areas of early mining.Sediment analysis has already been used to identify medieval and later lead mining on Islay and at Leadhills and this technique could also be used to identify prehistoric extraction (Rowan et al.1995). The geochemical monitoring of heavy metal pollution has also been used successfully at some of the prehistoric Welsh mines (Mighall 2003) and is currently being tested at Leadhills.Provenancing studies are another source of information and the techniques used recently to demonstrate a correlation between placer gold deposits in the Mourne Mountains of Co Down and Irish Bronze Age gold artefacts could be applied to Scottish material.

Museum collections represent an underused source of information.The National Museum of Scotland and the Dumfries and Galloway Museums Service collections contain some possible stone mining tools and a review of other Scottish collections might be rewarding. This should be combined with a search of antiquarian and geological literature for accounts of the discovery of early mine workings and artefacts.

In the absence of coarse stone tools there can be a problem distinguishing between Bronze Age mines and other early or pre-industrial workings.The identification of Bronze Age mines is an important aim but should the question of early Scottish mining be included as part of a broader chronological priority. 

The broader research priorities for prehistoric and early mining in Britain have been outlined in a recent Historical Metallurgy Society research paper (Bayley, Crossley and Ponting 2008).The principal challenge facing the study of Bronze Age mining in Scotland is the recognition of the mines themselves.The apparent absence of coarse stone tools at Scottish mines suggests that other field techniques may be required to identify potential sites.The way forward would be a survey programme aimed at identifying those sites with potential evidence for pre-industrial mining (firesetting and firesetting debris, hand-picked workings, absence of gunpowder shot holes). This should be combined with detailed topographic survey to establish relationships between surface mining remains and dateable features in the immediate landscape, an approach currently being applied with some success to late prehistoric lead mining and settlement sites in the English North Pennines.The Leadhills area, for example, which combines multi period remains of lead and gold extraction with Bronze Age settlement, would be an ideal location for a trial survey examining the landscape setting of metal mining.

Ores to Metal: Smelting, Re-melting and Recycling


 Late Bronze Age hoard recovered in 1990 from Priestden Place, St Andrews, Fife.  This is the largest and most representative hoard of the Ewart Park assemblage from Scotland. In addition to bronze objects, amber beads and shale bracelets, organic remains of wood, textile, leather, fibre, boars’ tusks and horn or antler also survived.©NMS

There is currently no evidence for ore smelting dating to the Bronze Age in Scotland, let alone an actual workshop, reflecting a general paucity of evidence throughout northwest Europe (see Meurkens 2004). A copper droplet found at the Beaker settlement site of Northton, Isle of Harris together with the relatively extensive deposition of presumably used stone and clay moulds and mould fragments as in the landscape of the Migdale-Marnoch region and elsewhere during the Early Bronze Age (Cowie 1988; Needham 2004; Cowie and O’Connor 2009) represents the entire span of direct evidence for primary and secondary metal production. The location and organisation of these metallurgical activities in Scotland remains unknown as the findspots for moulds, as with metal objects, may well reflect intentional deposition rather than past production and consumption sites.

The absence of any ceramic vessel found in Scotland that is demonstrated through archaeometallurgical analysis to have been a crucible for smelting, re-melting or recycling means that any assessment has to rely upon compositional evidence from objects to explore recycling. This has been attempted for the Chalcolithic/Early Bronze Age in Britain and Ireland and has demonstrated that the first millennium (c. 2400/2500BC-1400/1500BC) of copper and copper alloy metal in Scotland was primarily re-melted from objects produced in Ireland(Peter Bray pers. comm). The Middle-Late Bronze Age compositional data has yet to be analysed. The contemporary evidence for high concentrations of moulds at settlement sites such as Cladh Hallan, South Uist and Jarlshof, Shetlandduring the Late Bronze Age (Hamilton 1956; Parker Pearson et al. 2004; Matthew Juddery pers. comm.) potentially implies that at least re-melting and recycling could have been carried out in the near vicinity. 

Metal to Object: Alloying, Casting and Working

The earliest metal objects found in Scotland are copper and gold from c. 2400/2500 BC (Coles 1969; O’Connor 2004). The evidence for arsenical copper, whether as a deliberate alloy or due to the exploitation of arsenical copper ores, had been identified in halberds by Northover (1989) and may simply be the result of smelting sulphidic ores and casting the halberds in closed moulds (Bray pers. comm.).Following the adoption of tin-bronze c. 2100/2200 BC, very few copper and arsenical copper objects are known. The early use of high tin-bronze to create a silvery coloured softer metal is clearly demonstrated in the Early Bronze Age Migdale-Marnoch regional metalworking tradition (Needham 2004). The early exploitation of lead is attested in the lead and cannel coal bead necklace found in an Early Bronze Age cist cemetery at West Water Reservoir, Peeblesshire (Hunter and Davies 1994, 2000). The alloying of copper, tin and lead to create a lead-bronze alloy which is widely evidenced in Wales and southern England during the Middle-Late Bronze Age (e.g. Northover 1980, 66-8) does not appear to have been as widely practised in Scotland although the St Andrew’s hoard, Fife revealed a greater complexity of alloying practices than had been anticipated (Cowie 1998).

The metallographic analysis of Bronze Age metal objects in Scotland is limited to a small number of objects and object types with the consequence that broader technological patterns have to be inferred. The casting, annealing and cold-working of copper and tin-bronze flat axes (Bray pers comm.), the closed mould casting and rivets in halberds and certain daggers (Gerloff 1975, Schuhmacher 2002) together with the delicate sheet-working of gold lunula and basket ornaments (Taylor 1980; O’Connor 2004) and evidence for embossed and incised decoration in gold (Needham 2000) and bronze (Coles 1969) indicates that the Chalcolithic/Early Bronze Age metal objects in Scotland were manufactured by craftspeople, potentially living in Scotland, using a relatively wide range of manufacturing techniques.

The increasing sophistication and widespread practicing of bronze casting and metalworking can be seen in the Middle Bronze Age with the expanded range and quantity of bronze objects (Coles 1964); the precision and length of the rapier blades (Burgess and Gerloff 1981); the adoption of socket technology in spears (Davis 2006) and axes (Schmidt and Burgess 1981); and finally in sheet metalworking of large objects with cast attachments such as shields, cauldrons and buckets (Gerloff 2010; Uckelmann in prep). The twisting and flange twisting of gold bars into torcs (Eogan 1967) as well as similar twisting of gold ribbons (Eogan 1983) represents a distinctly new technological approach to the metal (Roberts 2007). The Late Bronze Age represents a continuation and development of the earlier technologies but with a substantial expansion in the range of bronze and gold objects being produced and circulated (Coles 1060; Eogan 1994). Further developments in bronze-casting and working can be seen in the swords (Coloqhoun and Burgess 1988) and delicate gold wire-working in the striped penannular rings and lock-rings (Eogan 1994).

The absence of workshops means that the evidence beyond examining the objects derives from a small selection of bronze tools such as anvils (e.g. Ehrenberg 1980). The identification of individual metalworkers is restricted to drawing inferences from rare, distinct and technologically complex objects such as shields (Uckelmann in press; in prep and cauldrons (Gerloff 2010). This is exemplified by the discovery of two virtually identical shields - one from Coveney, Cambridgeshire and the other from Auchmaliddie, Aberdeenshire - which provide potential evidence for an individual craftsperson. The idea of an individual metal-smith has been proposed for Early Bronze Age gold in Wessex (e.g. Taylor 2005) and could also be relevant to objects from Scotland such as daggers.

Object to Artefact: Use, Circulation and Deposition

The use-wear on metal objects in Scotland during the Bronze Age has yet to be systematically analysed beyond specific object types and collections such as bronze flat axes (Moyler 1998), swords (Bridgford 2000; swords and spears Anderson 2011), shields (Uckelmann in press; in prep.) and socketed axes (Roberts and Ottaway 2004. Each case study has revealed far great complexity in the past uses of the object than had been anticipated. This is exemplified by the potential use of shields as defensive weapons rather than simply ceremonial ornaments (Uckelmann 2011; 2012) and the deliberate destruction of swords before deposition in rivers (Bridgford 2000). Use-wear has been used, often in conjunction with experimental replications, in the reconstruction of combat techniques during the Bronze Age (see papers in Uckelmann and Mödlinger 2011).

The circulation of objects, whether through an individual or community movement, exchange or recycling is difficult to reconstruct from the available evidence. Inferences tend to be based either on compositional or typological comparisons. Distribution maps reveal concentrations of a distinctive metallurgical trait or object type that, at least partially, reflects metal circulation (e.g. Coles 1960; 1964; 1969). For instance, the identification of regional metalworking traditions such as the Migdale Marnoch in northeast Scotland (Cowie 1988; Needham 2004) demonstrates a spatially and temporally restricted pattern of metalworking techniques and objects. In contrast, the St Andrew’s hoard, Fife revealed several objects which are not commonly found in Scotland such as a bugle shaped object, a ribbed-tanged knife and a peaked slide (Cowie 1991; Cowie 1998) potentially indicating the movement of metal objects from southeast England or northern France. However, as all metal objects could have been recycled as ingots, and the surviving objects were those deliberately placed and then subsequently recovered by chance several thousand years later, our perspective on metal circulation in the Bronze Age is seriously compromised.

The deposition of metal objects in land, riverine and maritime contexts has attracted considerable research whether at an individual site level as at Dail na Caraidh (Barrett and Gourlay 1999) and St Andrews (Cowie 1991; Cowie 1998); a regional perspective as at Migdale Marnoch (e.g. Needham 2004); or orientated towards specific object types such as swords and shields (e.g. Bridgford 2000; Uckelmann 2011; 2012). The geography of object deposition appears to reveal consistent patterns such as the placing of individual swords in rivers and shields in bogs (ibid) and the placing of Late Bronze Age hoards in association with higher ground as around the Eildon Hills (O’Connor and Cowie 1985). The potential for replicating the pioneering research by Yates and Bradley (2010a; 2010b) in southeast and eastern England in tracing the exact findspots of old and new hoards and then locating them in their landscape context is evident. The study of the placing of metal objects in graves as at Rameldry Farm, Fife (Baker et al. 2003) during the Chalcolithic/Early Bronze Age has benefitted substantially from the recent flurry of large projects such as the Beaker People Project, Beakers and Bodies and the National Museum of Scotland dating programme which are re-analysing and re-dating many funerary sites.

Visualisation of the copper working process and taphonomy (Ottaway 2001, Fig 1)

4.5.2 The Societal context of metallurgy

Writing Life Histories for Scottish Bronze Age copper-alloy Artefacts: Biography, Prosopography and weaving data

After a long theoretical gestation the concept of object biographies has become increasingly visible within the archaeology of Later Prehistory. Kopytoff's (1986) paper on the cultural biography of things from the edited volume The Social Life of Things can be seen as the beginning of the incorporation of the concept of artefact biography into archaeology. Over ten years later the call was taken up in other influential collections particularly Marshall and Gosden's (1999) edition of World Archaeology 'The Cultural Biography of Objects'. A recent paper by Joy (2009) explicitly tries to renew focus on the concept, though there are signs that he need not worry - biography has begun to enter the lexicon of archaeological material culture studies. However at its heart it is an idea that derives from anthropological fieldwork.

The copper-alloy objects of the Scottish Early Bronze Age (EBA) are an ideal case study for exploring the concept and assessing its usefulness for future researchers in this period of prehistory. Metalwork has always held an iconic, indeed eponymous status within Bronze Age studies, being central to understanding chronological and social change. The number of analytical, scientific and conceptual approaches that have been applied to bronze probably exceeds those for any other artefact category. This section will argue that this is crucial to the potential of biography and points towards its practical application using the concept of prosopography.

Biography as a framework is intuitively attractive for studying archaeological material culture. It focuses on a generational 'human' time scale for understanding how things and people interact. It unites the concepts of agency of things with archaeological ideas of geography, place and technology. Put briefly, biography encourages the study of fine-scale interactions of people and things. To identify how objects were being made and used; but also how people in the Bronze Age were being 'made' by the objects that they were making and using (Gosden 2005).

Being a weaver together of theory does not, however, make object biography straightforward to use in Bronze Age studies. The datasets available to archaeologists are incomplete, dispersed and imprecise when compared to the anthropological experience detailed in Kopytoff (1986) or Peers (1999). However, there is no need to be too pessimistic as many of the techniques available to the study of early metallurgy neatly mesh with object life events. Chemical composition and lead isotopes can be used to infer the origin of the ore used to produce the metal (Needham 2002; Rohl and Needham 1998), while recent work on chemical signatures of melting and casting, the distribution of moulds, and regional typologies can come together to pinpoint the birthplace of individual objects in their final form (Bray and Pollard in press). Metallography uses polished cross-sections of samples from objects under reflected light microscopy to investigate smithing processes as well as use-life (Scott 1991). Surface wear analysis can also show how copper-alloy objects were used in life and also, where applicable, how they were decorated (Wall 1987; Moyler 2008). Lastly the traditional archaeological strengths of excavation and burial archaeology can show how the object ended its social life and was removed from interaction. Many of the attributes of the lives of things can therefore be accessed using traditional and venerable archaeological techniques. Clearly, new theories can help breathe life into old datasets.

The practical limits of time, money, conservation concerns and quality of data constrain how widely object biographies can be written. Though elements of life history can be seen in our material science and archaeological datasets the chronological resolution within the Scottish Bronze Age is currently in terms of centuries (Needham et al. 1997; Needham 2004) . We can infer internal object sequences of action, but overall these sequences will float within broad periods of absolute time. Few objects have had multiple analytical techniques applied to them. Only the most important, rare or iconic can justify the expense and focus of broad based investigation. Therefore if a score is assigned to each Early Bronze Age copper alloy artefact from Scotland, (with a point gained for each technique applied to it, 1 for metallography, 1 for modern excavation, 1 for chemical composition and so on), only a few recently excavated daggers score highly (Bray 2009). Key examples would be the work on the Rameldry and Seafield West daggers (Baker et al. 2003; Cressey and Sheridan 2003). The majority of artefacts, even after over one hundred years of scientific attention, still score zero or one. Currently, no Scottish EBA metalwork has been sampled for lead isotope analysis, while the known chemical compositions were mostly derived from the Studies on the Beginnings of Metallurgy Project (SAM 1 and 2 project, Junghans et al. 1960, 1968) which were produced using the relatively imprecise, and now outdated analytical technique of Optical Emission Spectroscopy.

Overall, the vast majority of material classes have actually received less focused work than copper-alloy, which again argues against a widespread application of object biography. Where there are large available datasets, such as for metallurgy, representing a lot of known 'life-events', these data are widely spread across the assemblage. Rather than biography being a common, quotidian tool for material culture studies, it should be seen it as a feasible ideal that, in the near future, will only be applied to objects of national significance (for example see the study on the Danish Gundestrup Cauldron, a focal item in the collections of the National Museum of Denmark in Copenhagen Nielsen et al. 2005).

Data on artefact life events can be brought together under a slightly different framework that emphasises the analysis of fragmentary collections of data. Prosopography was developed in Ancient History to solve precisely the problem outlined above. Obviously people had life histories; however the written texts that survived are often fragmented and incomplete. Therefore prosopography collects all possible sources in parallel and draws biographical inferences on the group or assemblage level (Keats-Rohan 2007). When possible more detailed life histories are written, but overall all biographical datapoints or 'events' are used to their full potential to illuminate groups of people with a common origin, profession, age or fate. This is clearly the way archaeologists have been working for centuries, weaving together assemblages to infer relative chronologies and social systems. Building on the initial spark of object biography created by anthropology of current objects, prosopography may offer archaeologists a useful tool. The University of Oxford is exploring, through a project yet in its early stages,  ways that prosopography can be applied to archaeological datasets (Bray et al. forthcoming). Rather than individual stories about artefacts, this will be a model of data collection and interpretation to act as a bridge between archaeological information and material culture theory.


See also the ScARF Case Studies: Copper Mining at Tonderghie and Experimental Archaeology: Bronze Age Weaponry