Neolithic beginnings of metallurgy in the central Mediterranean region?

In this post we present a summary of information on the apparent beginnings of metallurgy in the central Mediterranean region.

Summary Copper and silver are the earliest worked metals in Sardinia: there is evidence of their use from the first half of the 4th millennium BC, in the sphere of the Ozieri facies of the Final Neolithic. The use of gold is represented by a solitary artefact belonging to the Late Eneolithic (Beaker) period. Lead, however, appears roughly around the mid 3rd millennium BC.

Silver represents approximately 22 % of artefacts recovered in Sardinia from the Neolithic and Eneolithic periods; of these 4% are attributable to the first half of the 4th millennium BC and 8 % to the second half of the 4th millennium BC; the rest belong to the 3rd millennium BC. It is interesting to note that, as for copper, in the early phases of metallurgy (4th millennium BC) a consistent quantity of silver finds come from settlements, whereas from the middle phases of the Eneolithic, silver and copper artefacts are almost exclusively found in funerary contexts. Further proof for consideration is provided by the presence of a solitary silver artefact among the metal finds from the multi-layered shrine site at Monte d’Accoddi: a small disc recovered amongst finds from the Middle Eneolithic.

The archaeological data The earliest Sardinian silver artefact was found in tomb V of Pranu Mutteddu and dates to the first half of the 4th millennium BC (facies Ozieri I). This small necropolis is of particular importance because it is evidence of individual burials in monumental tombs during a cultural facies of predominantly collective burials in rock-cut tombs. The presence of a small necropolis of rock-cut tombs close to the individual burials reflects social inequality represented through the different treatment reserved for eminent personages. This is confirmed by the wealth of grave goods of tomb V: a ceramic dish, two silver disc rings, a flint dagger and blade, and obsidian arrowheads (Atzeni/ Cocco 1989). Stone continues to hold an important role which it will maintain in some geographical areas even into the middle Eneolithic during the 3rd millenium BC. In the second half of the 4th millenium BC (Ozieri II or Sub-Ozieri) the number of sites and silver artefacts doubles. An »allée couverte« placed opposite the rock-cut tombs of the necropolis of Montessu contained the remains of a burial with grave goods including pottery, a round threaded silver ring and a copper disc ring (Atzeni 1987). Two more rings and a silver lamina were discovered in the Su Coddu/Canelles settlement (Melis 2005). One was a disc ring similar to that of Pranu Mutteddu. Given its smaller dimensions, it was probably used as part of a necklace. The second ring was an incomplete circle. They were found in the Canelles l. Badas area (Stratigraphic Unit 1058), in a cylindrical storage pit (structure no. 47), reused as a rubbish dump. The lamina was found in an area to the north (Su Coddu) in an irregular-shaped structure (no. 51) that was only partially investigated (Ugas et al. 1985). Slag deposits were discovered in the same area as the lamina, and were interpreted on site as waste from smelting silver. Remains of marine molluscs, pottery, and stone tools were recovered from the same layer. In the tomb 12 of the Cannas di Sotto necropolis some silver rings with materials of the Final Neolithic and the Early Eneolithic were recently discovered (Salis 2013).

During the 3rd millennium BC the presence of silver artefacts grew notably, alongside a more widespread use of copper.

The archaeometric data The Greeks called Sardinia »The island of silver veins« (scholiast of Plato’s Timaeus) because of the abundance of this metal. Such richness must have been noticed by prehistoric Sardinian human groups, and probably, by those who came into contact with them. The Sardinian deposits were subdivided into three families according to lead isotopic ratios: Cambrian deposits, Hercinian vein-type and polymetallic mixed sulphides (Valera et al. 2003). Here, the archaeometric analyses of Nuragic artifact suggest a preference for Cambrian silver deposits, located in Sulcis-Iglesiente (south-west Sardinia).

The beginning of metalworking in Sardinia is linked to the presence of copper and silver artefacts in the context of the first half of the 4th millennium BC. However, the first direct evidence of metalworking belongs to the second half of the 4th millennium BC, with the find of a crucible at Su Coddu/Canelles (Manunza 2005). The slag discovered in the structures of the first half of the 4th millennium BC could be attributed to the reduction of copper and silver, according to Ugas et al. (1985), but this slag has never been scientifically analysed. The results of the analysis of other slag, later discovered in the Canelles l. Badas sector, rule out the possibility of them being related to metalworking (Melis 2005; Melis et al. 2012). More than 50 silver artefacts from the Neolithic and Eneolithic have been discovered in Sardinia. Despite this, archaeometric research has been applied only to a limited number of objects. The authorisation of destructive analysis has only very rarely been permitted; the necessity to preserve very ancient and precious objects prevails. Such analyses are nearly always of very small, and in general, much oxidised objects. Despite the absence of data it is reasonable to presume that the silver, considering the availability of raw material from which the artefacts discovered in Sardinia were manufactured, originated on the island (Lo Schiavo 1996): in fact, although we cannot exclude that the oldest artefacts could have been introduced from outside, the presence of argentiferous ore deposits, of native silver, and of a high number of silver artefacts, permits us to formulate this hypothesis. It will need to be confirmed by archaeometric analyses in the future. The rarity of native silver (Atzeni 2005) might suggest an early exploitation of the numerous deposits of silver-rich galena. The absence of an evidence for smelting argentiferous ores or of cupellation may be justified by a lack of discoveries.

The picture is still too incomplete to recreate the »châines opératoires« of the earliest Sardinian metallurgy, as there is no data on extraction nor on the instruments used for mining, while there are also no specific studies of crucibles. The oldest Sardinian artefact made of gold and silver dates back to the Late Copper Age; a torque, discovered in a Bell Beaker grave at Bingia ‘e Monti. An analysis by using a scanning electron microscope with energy dispersive X-ray spectroscopy (SEM-EDX) revealed a higher percentage of sil- ver in the outer layer of the necklace than underneath; the surface contained 65% gold, 32% silver and 2% copper, whereas the metal beneath the surface layer contained 51% gold, 45% silver and 3% copper (Atzeni et al. 2003; Lo Schiavo et al. 2005).

Discussion The silver artefacts discovered in Sardinia dating to the 4th and 3rd millennium BC had generally uniform morphological characteristics, in particular simple band or spiral rings that were also extremely widespread in more recent periods (Bartelheim et al. 2012). The most common categories were rings (44%), followed by necklace pieces (40%) and bracelets (8%). These were all small objects, with the exception of the disc from Padru Jossu and the electrum torques of Bingia ‘e Monti, from the Final Eneolithic.

Chronology and cultural contacts The archaeological data suggests the precocious appearance of silver and copper in Sardinia in comparison with the rest of the Mediterranean. On the origins of metallurgy in the central-western basin of the Mediterranean there exists a wealth of literature, recently reprised in Dolfini 2013 and 2014.

A silver artefact was discovered in the Alepotrypa cave in southern Greece and is dated to the mid 5th millennium BC to early 4th millennium BC (Muhly 2002; Roberts 2009); however, in Cyprus the metallurgy of copper, which would be of fundamental importance during the following millennia, developed in an embryonic form in the second half of the 4th millennium BC (Guilaine 2011). Looking towards the central-western Mediterranean, in which Sardinia was part of the obsidian network since the Neolithic, we see that in the south of France silver was almost completely absent in the Eneolithic pre-Beaker and Beaker period. Metallurgy began in France during the second half of the 4th millennium BC and involved copper, lead, and gold. The number of metal objects found in Sardinia and Italy in the last quarter century has changed the picture considerably when compared with the data examined by Primas (1995): for the 4th and 3rd millennia BC the author had indicated only four finds in silver for Sardinia and two for northern Italy. Recently Bergonzi (2012) has published a paper with eleven such sites for the Italian peninsula and ten for Sardinia (Bergonzi 2012). The number of sites for Sardinia has now reached 16, including Bingia ‘e Monti where a necklace in electrum has been found.

In the absence of direct radiocarbon dating for Tomb V at Pranu Mutteddu, in which the most ancient silver artefacts were discovered, its date is assumed to fall within the first half of the 4th millennium BC. The entire 4th millennium BC represents an embryonic phase of metallurgy, characterised by a gradual growth in production that was not exclusively reserved for funerary rites. The early 3rd millennium BC signalled the beginning of a new phase, distinguished by important changes. The presence of numerous silver artefacts amongst grave goods emphasises this development.

On the Italian peninsula the metallurgy of silver developed predominantly in the Rinaldone facies, with particular concentration in northern Latium and in the area around Rome (Giardino 2012).

The introduction of metallurgy should be searched for in the preceding Neolithic phase of central-western Italy (Liguria and Tuscany), and in particular in Chasséen-type contexts. Considering the intense activity of exchange between Sardinia and Tuscany during the Neolithic within the obsidian network, a comparison between the two areas is necessary.  During the final Eneolithic Bell Beaker phase the relationship between Sardinia and the mid-Tyrrhenian Italian coast is confirmed through various discoveries.

Conclusions The number of silver finds underlines its growing impact in Sardinian prehistory. The use of gold, present in the electrum necklace of Bingia ‘e Monti, remained only marginal. This situation extended to the entire Italian peninsula, where before the Middle Bronze Age gold remained a great rarity (Bergonzi 2009).

The coeval appearance of silver and its importance in Sardinia and in the Rinaldone area, together with the possible Sardinian origin of silver in some peninsular objects, suggests an exchange of knowledge between the two groups. Considering the most striking Eneolithic characteristics of Rinaldone (e. g. most advanced metallurgy, growth of territorial competitiveness, rich grave goods with many arms) compared to the Ozieri we can even suppose that contact between the two groups sparked the »Eneolithisation« of Sardinia. This hypothesis does not conflict with the model proposed by Dolfini (2013) on the origins of metallurgy: according to this model knowledge of metallurgy would have spread from the Balkans to northern and central Italy and from there to Corsica and Sardinia. While there are presently insufficient elements to confirm or reject the hypothesis of an independent origin of metallurgy in Sardinia, it is, however, plausible to imagine that there would be an exchange of knowledge via the various contacts with central Italy.

The second half of the 4th millennium BC sees the adoption of a more opportunistic approach to the manufacture of stone tools, pottery, and the use of hard animal materials (e. g. bone, horn, shell, ivory, teeth), paired with a reduction in production times, which may be associated to the development of metallurgy, as well as agriculture (Melis et al. 2012).

During the 3rd millennium BC the use of silver increases, and it seems to become an identifying symbol of Neolithic groups, in contrast to the Monte Claro society, that used it only very rarely.

(Source: “Silver in Neolithic and Eneolithic Sardinia”, by Maria Grazia Melis)

Introduction Interest in the beginnings of metallurgy in Italy and surrounding areas is, one might say, as old as archaeology. However, only recently has the antiquity of the local metallurgical practices and products been fully recognised (but see Barfield 1966 and Renfrew & Whitehouse 1974 for notable exceptions). Skeates (1993), in particular, was the first to draw attention to the growing amount of data concerning Neolithic metalwork from this region. Grounding his study in a detailed review of the evidence and chronology then available, he claimed that the first metal-using horizon in the central Mediterranean was to be ascribed to the last centuries of the Neolithic, in the late 5th and early 4th millennia cal BC. His view was further confirmed by Pearce (1993; 2000; 2007) and Barfield (1996), who mainly researched the northern Italian evidence.

Curiously, most of the Italian scholarship took for a long time a very different stance. With few exceptions (e.g. Cazzella 1994; De Marinis 1997), Italian prehistorians argued that the first metallurgical ‘cultures’ found in the Italian peninsula (namely Remedello in the north, Rinaldone in the centre, Gaudo in the south-west and Laterza in the south-east) had not actually flourished until the late 4th or early 3rd millennia cal BC, in the developed Copper Age (Bianco Peroni 1994; Carancini 1993; 1999; 2001; 2006; Peroni 1971; 1989; 1996). They also controversially dated a large section of copper and arsenical-copper metalwork, which was previously thought to be Copper Age, to the first phase of the Early Bronze Age.

Radiocarbon has now disproved both elements of this hypothesis, thus reinstating early Italian metallurgy (and especially the key Rinaldone evidence from ore-rich west-central Italy) within its Late Neolithic and Early Copper Age context (Dolfini 2010; 2013; 2014; Giardino 2009–2012; Passariello et al. 2010). However, one fundamental question still remains unanswered: when did metallurgy begin in the central Mediterranean? In the Middle Neolithic, as Barfield (1966) once controversially proposed, or in the Late Neolithic as most scholars seem now keen to believe? And if the latter proves true, when exactly can we date its inception considering that the Italian Late Neolithic lasted for about 700 years?

Ore deposits and prehistoric mining The central Mediterranean region features three major ore-mineral districts in the Alps, west-central Italy, and Sardinia. These are complemented by a plethora of relatively minor deposits and outcrops scattered across most of the region with the notable exception of the circum-Adriatic area, the Maltese archipelago, and most of Sicily. The nature of the ore deposits is extremely varied, comprising as it does primary and enriched copper and iron-copper sulphides (e.g. chalcocite, chalcopyrite and bornite), fahlores (i.e. polymetallic copper sulphides such as tetrahedrite and tennantite, which may be naturally rich in arsenic and antimony), and galena (lead sulphide), the latter occasionally argentiferous. Secondary (or supergene) deposits resulting from the weathering of primary ores are also found abundantly; these may include copper oxides and carbonates such as malachite, azurite and cuprite (Baumgarten et al. 1998; Carobbi & Rodolico 1976; Cavinato 1964; Pearce 2007; Preuschen 1973; Valera & Valera 2005; Valera et al. 2005). Stibnite (antimony sulphide) and cassiterite (tin oxide) are also found in Tuscany and Sardinia. Whereas the former was almost certainly smelted in the Copper Age to make antimony necklace beads, prehistoric exploitation of the latter is still debated (Giardino et al. 2011; Penhallurick 1986: 79–83; Tanelli 1983; 1989). Significant iron, zinc and mercury deposits are also present in the region, but their exploitation began at a later stage.

The oldest evidence for metallurgical copper mining in this region (and, for that matter, in the whole of western Europe), is documented at Libiola and Monte Loreto in Liguria, northwest Italy. Here, extensive workings of the local chalcopyrite have been brought to light by an Anglo-Italian team, whose findings included: artificial galleries backfilled in prehistory once the cupriferous veins had been depleted; evidence of fire-setting to loosen the rock; and extensive dumps of graded mining debris, which are informative of in situ beneficiation (i.e. rock grinding and sieving aimed at separating the high-grade ore from the gangue). Radiocarbon dating indicates that these mines were worked from the Early Copper Age onwards (Maggi & Pearce 2005; Pearce 2007). Interestingly, an isolated date obtained from viburnum charcoal (a short-lived Mediterranean shrub) found in the dumping grounds at Monte Loreto calibrates in the Late/Final Neolithic (Beta-203528; 5010±40; 3944–3704 cal BC: Campana et al. 2006). However, it is unclear whether this should be considered proof of the Neolithic sourcing of chalcopyrite, an isolated experiment in the procurement of native copper (or supergene ores), or simply an unreliable outlying date. Thus, in spite of their richness, diversity and overall accessibility, the central Mediterranean ore deposits still defy any attempt to date their exploitation (at least for the purpose of copper smelting) to periods earlier than the Copper Age.

The state of the evidence indicates that claims for Middle Neolithic metalwork are at best unproven, and most probably incorrect. Significantly, this was recognised by the later Barfield (1996: 67), who tempered his earlier statements by noting that archaic copper axes are normally found in later contexts in the circum-Alpine and circum-Adriatic areas. It is therefore to later phases that we must turn our attention if we are to understand the emergence of metallurgy south of the Alps.

The first metal artefacts from the central Mediterranean In contrast with the controversial evidence for Middle Neolithic axes, some twenty copper and silver objects from Italy and Sardinia can be securely assigned to the Late or Final Neolithic; these are all dated by radiocarbon or associated pottery. A good deal of early metalwork is also found along the eastern Adriatic coast, but this is not discussed in this paper as these objects display unmistakable eastern European traits, thus suggesting origins within the Balkan metallurgical sphere (Žeravica 1993).

Assessing local production Where were these objects made? In the metal-rich districts of the central Mediterranean, either by working nuggets of native copper or by smelting the local ore? Or were they all imported from neighbouring areas where metal technology had already set foot?

Such was the state of the evidence with regard to Neolithic metal production until the recent, comprehensive publication of excavations carried out at Botteghino, a Neolithic open settlement near Parma (Mazzieri & Dal Santo 2007). Two copper awls were found at this site along with ‘slags’ (not yet analysed) and a fragmentary crucible with copper lumps still adhering to it. A radiocarbon date taken from the same context as the crucible indicates that metal was worked here in the third quarter of the 5th millennium cal BC (Hd-25298; 5619±25 BP; 4500–4370 cal BC), while another date from a different context calibrates slightly later (Hd-25297; 5456±25 BP; 4350–4260 cal BC). This chronology seems further confirmed by the abundant pottery in the Chassey style, which is widespread in northwest Italy in the mid/late 5th millennium cal BC. If future analysis confirms that these ‘slags’ are metallurgical in nature, this would testify to south Alpine copper smelting in the third quarter of the 5th millennium cal BC. Interestingly, this is the time period when metal production is first documented north of the Alps at Brixlegg in the Inn Valley (Höppner et al. 2005; contra Gleirscher 2007).

Copper axes of the Late and Final Neolithic Further evidence suggests that copper axes were utilised, and almost certainly made, in the Italian peninsula during the Late and Final Neolithic. This consists of three elongated tools with stout body, quasi-parallel margins and splayed cutting edge that have been found, in two separate instances, in the Late Neolithic deposits at the Bocca Lorenza cave near Vicenza (Caddeo & Giacobbi 1961; Pellegrini 1910). Bagolini (1984), Bianchin Citton (1988) and De Marinis (1992) argued that these objects related to the Copper Age burials also found at this cave, for these cut down deeply into the underlying Neolithic stratification.

However, Skeates (1993), Barfield (1971; 1996) and Pearce (2007: 42–4) highlighted that attentive examinations of the excavation diaries and publications do not confirm this reading. On the contrary, in both instances the excavators pointed out that the axes were found in the undisturbed Neolithic deposit, and that Neolithic artefacts including VBQ III (incised and impressed) potsherds were closely associated with them. Significantly, this interpretation was later accepted by De Marinis (1997: 36).

The Copper Age developments of Neolithic metallurgy have been the subject of recent works of mine (Dolfini 2010; 2011; 2013; 2014; Dolfini et al. 2011), and I shall summarise the principal results here. First, long-standing claims for the gradual development of metalworking and metal-using in the course of the Copper Age are to be rejected. Indeed, quite the opposite holds true: all stages of the metallurgical chaîne opératoire are now unequivocally documented at the very beginning of this period (c.3600–3300 cal BC), and were probably carried out with little variation until the introduction of tin-bronze technology. Copper Age ore mining has been brought to light at several sites including Monte Loreto in Liguria and Saint Véran in the French Alps (Bourgarit et al. 2008; Maggi & Pearce 2005); copper smelting is attested at open settlements such as Terrina IV in Corsica, San Carlo in the Tuscan Colline Metallifere and Neto-Via Verga north of Florence, but also at numerous rock shelters in the eastern Alps (Camps 1988; Fedeli 1995; Pearce 2007; Pedrotti 2001; Perini 2001; Sarti & Volante 2002); casting and smithing were probably carried out in most of the central Mediterranean region as the widespread crucibles, tuyères, and casting residues would suggest (Dolfini 2014); and metal objects were frequently deposited in burials following utilisation in a wide range of tasks including tree-felling, wood-working and possibly combat (Dolfini 2011).

Furthermore, mastery of polymetallic technology was perfected during this time period. Sulphidic ores were routinely smelted including iron-copper compounds and fahlores, from which arsenical and arsenical/antimonal alloys were naturally obtained. Metals other than copper were also exploited based on locally available sources. These encompassed silver, either from galena or argentiferous fahlores, lead (in Sardinia), and the rare antimony (in Tuscany). Gold was also used in Sardinia and possibly Malta during the 3rd millennium cal BC, but it is doubtful whether it was sourced locally (Atzeni et al. 2005; Giardino 2000; Trump 2004: 233; Valera et al. 2005). Importantly, claims for the gradual transition from an allegedly simpler arsenical-copper technology to allegedly more advanced ternary (copperarsenic-antimony) alloys in central Italy have been disproved (De Marinis 2006). On the contrary, objects with both alloys (as well as pure copper) were cast from the beginning of the Copper Age, thus further demonstrating the rapid and precocious developments of metal technology in this area (Dolfini 2010; Dolfini et al. 2011).

It is my contention that metal-working and metal-using intensified during the Final Neolithic c.3800–3600 cal BC. This is a short-lived and as yet poorly understood transitional phase in which the Late Neolithic world and some of its most distinctive practices (e.g. the use of portable figurines and the exchange of obsidian) dwindled and finally disappeared, or at any rate lost their prior centrality. New practices and spheres of action emerged in their place, in which the central Mediterranean communities rooted a novel sense of belonging and being in the world. Most prominently, these encompassed formalised burial practices and ancestor rituals in which new concepts of gender and personhood were brought about, and new artefacts were employed in their materialisation (Barfield 1998; Robb 2009; Whitehouse 1992a; 1992b). Interestingly, radiocarbon dating now indicates that Copper Age funerary customs were first elaborated in the centuries 3800–3600 cal BC (Cazzella & Silvestrini 2005). New ways of making and decorating ceramics also emerged in this time period from the pulverisation of the wide-ranging pottery styles of the Late Neolithic. In all likelihood, metalworking partook in the same, fast-changing technological and cultural milieu.

Conclusions For a long time, early central Mediterranean metallurgy stood in splendid if rather uncomfortable isolation. This was due to two contrasting forces that pulled the subject, and especially its chronology, in opposite directions. On the one hand, it was argued that the first metal objects would have made their appearance in the early to mid 5th millennium cal BC – a time when most European societies had not yet made their first encounters with metalwork. On the other hand, it was claimed that metallurgical practices and products did not significantly spread through the Italian peninsula until the developed Copper Age – much later than in neighbouring areas north and east of the Alps. The evidence discussed in this paper has disproved both claims. As for now, not a single metal artefact can be securely assigned to as early a time as the Middle Neolithic (c.5000–4500 cal BC). At the same time, it is now evident that not only did the first metal objects appear south of the Alps in the ensuing Late Neolithic (c.4500–3800 cal BC), but also that the technology to make them spread in the same time-span. Crucially, this enabled the earliest smiths to locally cast large implements including axes, and to extend the then-developing extractive technology of copper to new ores and metals including lead, silver and antimony. Such an upsurge in metal-working and metal-using, which probably took place in northern Italy, west-central Italy and Sardinia in the Final Neolithic (c.3800–3600 cal BC), created the technological and social condition for metal to be fully incorporated into central Mediterranean society – with the notable exception of southernmost Italy, Sicily and the Maltese archipelago – from the Early Copper Age (c.3600–3300 cal BC).

(Source: “The Neolithic beginnings of metallurgy in the central Mediterranean region”, by Andrea Dolfini)

Abstract This paper presents a review of our knowledge of the earliest phases of metal use and working in Sardinia, based where possible on radiocarbon chronology. It covers the cultural periods from the late Neolithic Ozieri to the Copper Age Monte Claro phase. I suggest that in contrast to continental Italy, the earliest metalwork in Sardinia seems to be used for display. It is striking that metal use and working seems to emerge in Sardinia later than in mainland Italy, despite the island’s rich metal resources.

Introduction although the island is rich in ores, evidence for early metallurgy and metalwork is relatively rare in the archaeological record, indeed they do not become widespread until the Bronze Age. Much of our archaeological evidence for the later Neolithic and Copper Age comes from multiple burials in rock-cut tombs, which are locally known as domus de janas, and in megalithic tombs and cairns, so that it has been difficult for archaeologists to work out the material culture sequence. Tombs with multiple burials were regularly used into the Bronze Age, with previous interments being cleared out or moved to make room for new burials and their grave goods, so that it is often difficult to be sure about the chronological relationships among the material found in these tombs, and indeed their contents may only date to the final period of their use (Tanda, 2009, pp.67-68).

Until very recently a secure radiocarbon chronology did not exist for Sardinia (and indeed radiometric dates for material found in multiple burials are sometimes of doubtful worth, since they likely date nothing other than the object that was determined), but an increase in the number of available dates means that we can now begin to understand the relationships between the various cultures of the late Neolithic and Copper Age and their chronology.

Discussion Perhaps the most striking aspect of the evidence for the earliest metallurgy in Sardinia is its paucity, at least until the Monte Claro phase, especially given the presence of abundant metal resources on the island. However, one of the key affordances of metal as a material is the ease with which it can be recycled, so the archaeological adage that absence of evidence is not evidence of absence is best kept in mind when assessing the incidence of metal finds, especially as it is precisely in those phases when metal was most rare that it is most likely to have been recycled.

Sardinian metallurgy seems to begin in the Ozieri phase, which we have seen starts in the last quarter of the fifth millennium, with the succeeding sub-Ozieri phase beginning around 3600 cal BC. Until dated contexts for the earliest metalwork are available, the parsimonious hypothesis is that the scarce evidence for Ozieri metalwork is likely to date to later in the timespan covered by the phase. On present evidence, and pace Melis (2014, p.488), this suggests that metallurgy in Sardinia appears somewhat later than in continental Italy, where metalworking first appears around 4300 cal BC (Pearce, 2015, pp.48-51). Since in the second half of the fifth to the later fourth millennium cal BC Sardinia was at the centre of a flourishing central Mediterranean obsidian trade, with its long distance networks (Tanda, 2009, p.64), it seems odd that metalwork arrives late and develops slowly in Sardinia, and we must ask why this might have been so.

Early metalwork in continental Italy is characterised by copper awls, with copper axes then appearing in the north-east around 4000-3800 cal BC, on the fringes of a network covering the eastern Alps, the middle Danube and the Carpathian basin; it is also possible that copper axes appear in northern and west central Italy around the mid fifth millennium, though this latter hypothesis is controversial (Pearce, 2015, p. 51). The Ozieri metalwork of Sardinia, on the other hand, is characterised by ornaments (beads and rings). Awls (a class of object that is variously described in the literature as points, pins, awls or punches) seem to only appear in sub-Ozieri times, when they become the most common artefact class. The first dagger is attested in sub-Ozieri times, while the report of an axe at Su Coddu remains to be confirmed (and dated). Daggers and axes are certainly documented later, in the third millennium BC Filigosa phase. The different pattern of incidence of artefact classes, with ornaments having an important early role on the island, may indicate that in Sardinia display was an important factor in the adoption of metalwork, unlike continental Italy (Pearce, 2007, pp.37-51). Indeed, it is interesting to note that after the sub-Ozieri awls become less common relative to other artefact classes, only increasing again in the Monte Claro phase; Monte Claro metalwork is dominated by daggers and awls, with few ornaments, suggesting a new social role for metalwork.

Melis (2009, p.517) suggests that awls only become common again in the Beaker period, but this is clearly wrong as the high relative number of Monte Claro awls shows. The production of long blades by pressure flaking with copper points, documented in layers dated to the end of the fourth / early third millennium BC at Contraguda (Perfugas SS), provides useful proxy evidence for the specialist use of copper points (Costa and Pelegrin, 2004; cf. Pearce, 2000; 2007, pp.48-51). Finally, it must also be noted that since there is cultural continuity from Ozieri through to Abealzu, with Monte Claro appearing perhaps to be intrusive, the introduction of metal-use and working does not seem to be associated with any major cultural change.

Both copper and silver artefacts and working occur from the Ozieri phase, silver initially in the south of the island, and the early attestation of silver may be explained by Sardinia’s rich silver resources, comprising outcrops of acanthite, cerargyrite and argentiferous cerussite and galena ore, as well as the likely presence of native silver. It is noteworthy that silver artefacts are also concentrated in the Copper Age Rinaldone culture of west-central Italy, particularly in northern Latium and in the Rome area (Bergonzi, 2012). For example, a silver bead was found in the multiple burial rock-cut tomb 23 at Selvicciola (Ischia di Castro VT), which has radiocarbon dates in the mid fourth millennium cal BC, though its final phase of use remains undated, so in the absence of a secure association we must remain sceptical about its dating (Petitti, et al., 2002; Petitti, Persiani and Conti 2016: 182-184). Bergonzi (2012, pp.573-574) notes that silver artefacts in Sardinia and the Rinaldone culture do not share strong formal similarities implying that while contacts between the two zones are likely, they may not have been strong. Based on lead isotope analysis, Carboni, et al., (2016) suggest that a fragment of silver sheet found in tomb 12 at Osteria del Curato – via Cinquefrondi (Rome) was made of metal from the Iglesiente in Sardinia It is striking that metallurgy in Sardinia seems much earlier than that of Corsica, where the pit at Terrina IV (Aléria, Haute-Corse), with the earliest radiocarbon-dated evidence for metalworking on that island, was filled around 3250-2400 cal BC (Pearce, 2013; cf. Camps, 1988, pp.82-85, 1992). Moreover, silver metallurgy does not appear to be attested in Corsica, despite the fact that the island had functioned in the Neolithic as a stepping-stone between Sardinia and the Italian mainland (Melis, 2014, p.490).

A difficult question to resolve regards how metal use and working originated in Sardinia. It seems unlikely to have developed autonomously, but given the extensive trade in Sardinian obsidian in the second half of the fifth to the later fourth millennium cal BC (Tanda 2009, p.64), it is more likely the result of transmission of know-how, perhaps even by metal prospectors, but the archaeological record provides no real evidence as to where these may have come from. Traffic in Sardinian obsidian was directed towards central and northern Italy and southern France (Tykot 2004, Lugliè 2012), which may provide an indication of the areas from which knowledge of metallurgy is most likely to have originated. Southern France is unlikely to be the source of such know-how, as here too metallurgy developed quite late (in the late fourth/early third millennium BC in the Languedoc, later in Provence: Bartelheim and Pearce 2015, p. 703), so northern or north-central Italy (Pearce 2015) seem the most likely candidates, though the lack of attestations of early metalworking in Corsica may make the central Italian hypothesis less likely, as the island traditionally acted as a ‘stepping stone’ between Sardinia and peninsular Italy.

Conclusions An anonymous scholiast to Plato’s Timaeus (25, B; Greene 1938: 287) tells us that Sardinia was once called ‘the island of silver veins’ (ἡ ἀργυρόφλεψ νῆσος) and silver is certainly an important feature of the early metallurgy of the island. In this paper I have provided an overview of our knowledge of metal use and working from the Ozieri to Monte Claro phases, attempting to establish some order in the data, notwithstanding the many difficulties caused by the lack of a secure radiocarbon chronology for early metallurgy, and the fact that most early metal artefacts were found in multi-period, multiple burial tombs. What emerges is that metallurgy seems to appear rather later than in continental Italy, and has a distinct character, being used for display. What is now needed is a programme of analysis, and in particular lead isotope analysis, so that we can learn more about the character of the early metallurgy and the likely ore sources.

(Source: “The ‘island of silver veins’: an overview of the earliest metal and metalworking in Sardinia”, by Mark Pearce)

Research-Selection for NovoScriptorium: Philaretus Homerides