by Chris Fowler
FIG. 37.2. Copper objects from the mid-fourth millennium BC hoard found at Bygholm, Skanderborg, Denmark. These metal artefacts were found with a fragment of funnel beaker pottery.
(Photograph by Lennart Larsen, the National Museum of Denmark).
In northern Europe and the Alpine area, a substantial break in the supply of copper objects around 3200–3000 BC apparently severely reduced the quantity of copper in circulation. This hiatus may reflect a disruption in access to a primary ore source (Ottaway 1989). The reduced quantity of archaeologically visible metal objects may have been partly due to changing depositional practice in light of the increasing rarity of imported metal (Klassen 2000, 273). This hiatus in use—or deposition—of metal objects is noticeable throughout north-west Europe, where metal, or at least its archaeologically visible deposition, goes entirely out of fashion during the first half of the third millennium BC.
The situation in the British Isles and Ireland was quite different, as metal was not circulated in any consequential—or archaeologically visible—amounts until the mid- third millennium BC (Needham 1996; Fitzpatrick 2011). Copper and gold objects were placed in funerary contexts and hoards whilst the Corlea 6 wooden trackway in the Irish Midlands, dendrochronologically dated to 2259±9 BC, sports marks produced by a metal axe (O’Sullivan 1996). This circulation of gold alongside copper challenges the broader observation that the earliest gold objects beyond the Mediterranean date only to the late third millennium BC (Hartmann 1970, 1979, 1982; Elèure 1982; Primas 1995) (Fig. 37.3).
FIG. 37.3. Irish style hammered sheet gold lunula (found at Gwithian, Cornwall, Britain).
(Photograph by the British Museum).
Metal production technology
From the mid-third millennium, metal production becomes archaeologically visible throughout western Europe, as copper, copper-alloy, and gold objects were produced and deposited in much greater numbers. However, early copper smelting only left very ephemeral traces and we may well be unable to identify the earliest generations of smelting or experimentation archaeologically (Timberlake 2005). The evidence for metal production is restricted to copper mining, ore refining, and smelting, and is only found at a handful of geographically distant sites. The earliest and by far the richest copper production traces have been excavated in Languedoc, southern France: the Vallarade copper mine dates from c. 3100 BC and is virtually contemporary with the earliest copper smelting at nearby La Capitelle du Broum (Mille and Carozza 2009) and ore processing at Roque Fenestre, Pioch Farrus 448 (Espérou et al. 1994). Production evidence elsewhere in France is limited to two copper droplets from a hearth at Val-de-Reuil, Seine Valley, dating to the late third millennium BC (Billard et al. 1991; see Meurkens 2004). Elsewhere, much of the best production evidence comes from Ireland and Wales, but consists primarily of copper ore mining activity, with several sites in use during the mid- and late third millennium BC. Of particular interest is the copper sulphide ore extraction and possibly smelting in south-west Ireland at Ross Island c. 2400 BC onwards (O’Brien 2004) and mining traces at Copa Hill, Cwmystwyth, Wales c. 21/2000 BC and potentially slightly earlier nearby at Erglodd (Timberlake 2009). The only other broadly contemporary evidence for metal production is a splash of arsenical copper in a midden at Northton (Isle of Harris) in Scotland, dated to the late third millennium BC (Simpson et al. 2006).
The limited nature of the fourth and third-millennia metal corpus must be taken into account when discussing the spread, type, and variety of early metal production. For instance, there are c. 150 fourth-millennium BC copper objects from southern Scandinavia (Klassen 2000, Katalog); c. 700 mid to late third millennium BC copper and bronze objects in Ireland (O’Brien 2004), and c. 1,400 late fourth to mid-third millennium BC copper objects in eastern Languedoc, south-east France (Gutherz and Jallot 2005). Metal may well have been extensively recycled, though this is difficult to detect (Bray 2012; Bray and Pollard 2012). The metal objects we study represent only those items which were carefully deposited and neither decayed nor were looted or recycled over the past 4,000–6,000 years (cf. Taylor 1999). Thus, all conclusions drawn from this fragmented and non-representative corpus are biased by the choices made by people in prehistory (whether to recycle/curate metal objects and where/how to deposit them), as well as the decisions of farmers, antiquarians, and archaeologists.
The chemical and isotopic compositions of these earliest metal objects often indicate patterns of circulation and points of origin for the objects and raw materials in circulation. In south-east France the evidence indicates the exploitation of a single ore source over several centuries, since high silver and antimony compositions match lead isotope data and point to the ores at Cabrières (Ambert 1999; Prange and Ambert 2005). The copper objects of the North European Plain demonstrate a signature of high arsenic and low silver (Ottaway 1973, 1989). Scandinavian copper is compositionally comparable to the rich Alpine ores (Klassen 2000, 2004); and lead isotope analysis links one specific metal type, Riesebuch copper, to Alpine sources (Klassen and Stürup 2001). Finally, the vast majority of copper objects in Ireland and western Britain are high in arsenic and antimony, which together with the objects’ lead isotope signature suggest raw materials from the Ross Island ore body (Coghlan and Case 1957; Case 1966; Rohl and Needham 1998; O’Brien 2004). In contrast, the recent re-analyses of compositional data in eastern Britain and north-west Europe suggests that the main copper in circulation, originally termed ‘Bell Beaker metal’, was formed by mixing metals from two obviously distant geological sources, one perhaps as far as northern Spain (Needham 2002). Hence, it is important to note that whilst metal types can be tentatively identified, the chemical composition of the finished object gives no information as to whether the raw material or the finished product was being circulated.
THE SPREAD OF NEOLITHIC METAL USE IN NORTH-WEST EUROPE
Metal travelled extremely long distances from where it was mined and the development of local secondary production clearly shows that at least some metallurgical knowledge went with it (Roberts 2008b). Although distribution maps reflect only those objects which, through taphonomy and happenstance, could be recovered in archaeologically meaningful ways, recent continental syntheses have pointed towards the importance of river systems in facilitating long-distance movement (e.g. Klassen 2004, fig. 145; Sherratt 1996). Moreover, wheeled transport and animal traction probably developed throughout much of north-west Europe in the late fourth and early third millennia BC. Wooden carts—presumably ox-drawn—were in use from the early third millennium, as a solid wooden wheel found in Kideris Mose, Denmark and dated 2880–2700 BC implies (Nielsen 1993). Similar wooden wheels, cart fragments, and iconographic representations of animal traction in Italy and central Europe slightly pre-date this example, placing the advent of animal traction squarely in the late fourth millennium BC (Sherratt 2006). It is more difficult to assess how the technology itself travelled. A process of learning and teaching at an exploitable ore source must have been necessary to communicate the various stages of metal production. An individual or a group able to initiate metal production would then move to another ore source. There are comparable metallurgical practices at earlier dates to the east, not only negating any concept of independent invention in western Europe but also implying that this was the continuation of a punctuated, pan-European transmission process (see Roberts 2008b; Ottaway and Roberts 2008; Roberts et al. 2009). The techniques of extracting, preparing, and smelting ores as well as casting, working, and shaping objects were similar throughout north-west Europe. However, one must doubt the uniformity of this training across western Europe. Presumably learning metallurgy was structured in ways similar to learning other technical skills (e.g. Brodie 1997; Högberg 2008), and it likely varied in significance from one region to the next.
The initial influx of metal into north-west Europe was not immediately followed by a growing metal industry, a marked increase in metal production and deposition, or even a developing appreciation for
metal as a material with unique and novel properties. Rather, after several centuries of extremely sporadic deposition (and probably varying local secondary production), metal fell out of fashion and presumably circulation in northern Europe from 3200–3000 BC and north-west continental Europe from 3000–2500 BC (cf. Krause 2002, 34). Similar patterns exist in other materials; for example, in the British Isles during the fourth millennium BC, jet and amber saw an initial flush of popularity, then fell out of use for several centuries before regaining their significance in the mid-third millennium BC.
South of the Loire river, the appearance of early metal objects—such as at Roquemengarde—was swiftly succeeded by growing centres of primary and secondary metal production elsewhere which remained in use through the early third millennium BC, perhaps due to connections with the larger Mediterranean communication and exchange network. In fact, the presence of large quantities of copper beads, awls, pendulums, and flat axes in the late fourth to mid-third millennia BC contrasts with far fewer copper objects produced in the region during the second half of the third millennium BC, when the earliest metal objects and metal production become archaeologically visible in Ireland, Britain, and the near continent. These trends cannot solely be explained by the development of new metallurgical techniques, but also involved the transmission and adoption of practices, objects, and presumably ideologies related to the so-called ‘Beaker package’ (see Vander Linden, this volume). Thin-walled Beaker pottery with an S-shaped profile and banded decoration was frequently accompanied in graves by polished stone bracers, finely made flint arrowheads, v-perforated buttons in diverse materials (possibly part of special adornment for the dead), and/or daggers in copper, flint, and occasionally other materials (Fig. 37.4). The ability to recognize and acquire these raw materials, the craft required to form them and to suitably prepare or ornament their surfaces, would have been fundamental for replicating and recreating this new set of practices and beliefs. During the late third millennium BC, the exploitation of gold, copper, and, later, bronze was inextricably tied to this social focus on the manipulation of diverse, exotic, and often eye-catching raw materials to make a socially circumscribed set of objects presumably used in many ways but found primarily in funerary contexts.
FIG. 37.4. Bell Beaker burial assemblage with copper tanged dagger, bone/ivory toggle and polished stone wristguard with sheet gold caps (found at Barnack, Cambridgeshire, Britain).
(Photograph by the British Museum).
THE ROLE OF METAL AND METAL TECHNOLOGY IN NEOLITHIC SOCIETY
The number of copper objects produced in north-west Europe over the centuries prior to 2500 BC was not large, and more indicative of an occasional rather than continuous production process. Metal was used differently at different times for conveying certain messages; and its frequent association with exotic, eye-catching, and presumably highly valued materials in socially significant contexts (i.e. hoards and burials) marks it out as part of a special and distinct class of material culture. Only small numbers of metal objects circulated at any one time. Yet, copper and gold objects continue to be accorded an overly special significance in many publications and are assumed to have held inherent, frequently unspecified value for prehistoric communities. This is implied in familiar interpretations such as its use for prestige and control by elites (see Bartleheim 2007 for review and discussion). The tendency has been to look for detectable changes in Neolithic societies due to the influence of metal production and consumption. However, metal probably did not carry quite the importance or prestige in prehistoric communities that we imagine. The lack of evidence for metal production at least partly reflects its sporadic nature. Existing evidence suggests metal working was engaged in by part-time practitioners and not dedicated ‘professional’ smiths until the mid-second millennium BC. This non-specialist production underlines the status of metalwork as a normal, not exceptional, part of the larger material assemblage (cf. Bartelheim 2007). Moreover, early metal tools did not provide an advantage over existing materials in performing everyday tasks. It should not be assumed that the flat axes were simply regarded as tools, rather than valued for their appearance as metal trinkets were. Not only were they less effective than their polished stone and flint counterparts (Mathieu and Mayer 1997), but many of the earliest were not even fully work-hardened and demonstrate few signs of intensive use.
Metalworking techniques such as casting, cold working, and annealing, or the repetition of metal forms, demonstrate a marked technological and aesthetic conservatism. Metalworking in the fourth and third millennia BC was not a dynamic or innovative technology, but practised sporadically and at small scale to traditional specifications and in order to meet relatively simple requirements (i.e. for the production of copper flat axes or trinkets). There is no evidence of a linear appreciation in the amount of metal from the fourth millennium BC: rather, the hiatus circa 3000–2500 BC must be accounted for. Choices about the forms and uses of metal objects reflect existing social desires (Roberts 2008a). For example, Irish copper flat axes mimic in form and depositional context the polished stone axes in circulation in preceding centuries (e.g. Cooney and Mandal 1998). In a nearly identical process, late third millennium BC copper axes in southern Scandinavia took on many of the social functions of thin-butted flint axes (Vandkilde 1996). In this region, copper and later bronze were almost entirely restricted to one form (the flanged axe) and depositional context (the hoard) for at least the first 200–400 years of their (reasonably) widespread use, in contradistinction to flint (particularly daggers) which continued to be deposited in settlements, burials, and hoards well into the second millennium BC (Frieman 2012a, 2012b; Vandkilde 1996). Copper, gold, and lead were first used in southern France at a time when beads of various materials, including horn, bone, variscite, and shell, were a significant part of the dominant funerary rite (Barge 1982). Furthermore, the mid-third millennium BC rationalization and standardization of the metal corpus can be tied to the social needs expressed in the widespread Bell Beaker funerary rite (Ambert 2001; Vander Linden, this volume).
It is commonly believed that metalworking was complex and required the development of complex social structures—such as the emergence of status distinctions or formal apprenticeships—not necessary for other prehistoric technologies (Strahm 1994, 2005; Krause 2003). However, many of the other materials used during this period, and used far more frequently than metal, required multi-stage production processes of comparable complexity. Knapping flint requires significant experience of the material’s unique physical properties; and producing the complex and regular tools characteristic of the fourth and third millennia BC necessitates a period of learning, probably facilitated by an older or more capable flint knapper (see De Grooth, this volume). From the fourth and into the third millennia BC, flint was mined in deep shafts across north-west Denmark and beyond (see Capote and Díaz-del-Rio, this volume). Mining and prospecting for special rock sources was not limited to flint, as axes of jadetite quarried from isolated outcrops in the Italian Alps circulated throughout western Europe from the sixth millennium BC (Pétrequin et al. 2008; Klassen 2004; Sheridan 2007), variscite was mined for use as ornaments at sites like Can Tintorer, Spain (Blasco et al. 1998), and haematite sources across the Black Forest region of Germany were exploited from the fifth millennium BC (Goldenberg et al. 2004). None of the mid- and late third millennium BC copper mines known in western Europe reflect the levels of technical and logistical complexity or scale of many of these earlier mines.