Metal sources and Trade networks in Bronze Age Europe

The Bronze Age was the first long period in human history when widespread trade networks connected Europe and the wider Eurasian continent, defining a pre-modern era of globalization, or, ‘bronzization’. One of the driving forces behind the inter-regional trade was the constant need for metals, and other raw materials and goods available only in certain parts of Europe. An important factor in the development of the widespread trade networks were the geographical distances between metal producers and consumers. In the Nordic realm, a new maritime economy supported the development towards long-distance trade networks. Some scholars argue that Bronze Age travellers, i.e. warriors and traders played a crucial role in this new economy, as competition must have arisen between different regions in Europe, or even between different local communities, to control the flow of goods and people.

In the context of theories about long-distance trade of metals, swords were considered particularly suited for analysis. First, the distribution of swords is often held to reflect trade routes. Second, swords are useful as indicators of group identities and craftsmanship as they were prestige objects of high symbolic and cultural value, and with high demands on quality and functionality. The Hajdúsámson-Apa swords dated to c. 1600–1500 BCE are a case in point; being high-end weapons that were exchanged across cultural borders. The flange-hilted swords of Naue type were, on the other hand, functional warrior swords widely distributed across the Continent c. 1300–1100 BCE. While cross-regional stylistic repertories influenced sword typologies, a high degree of local production may also be noted.

Recycling is also an important factor to consider. For the period in question 1600–1100 BCE, there are two competing opinions on recycling. One perspective goes in favour of seeing recycling as a standard phenomenon. Another perspective considers recycling to be of less importance. In this context, swords are generally held as objects subject to minimal recycling, because of the need for technological control, and control of the alloy, and perhaps also because they were personal objects with histories of use, with related associations and memories which may have prevented recycling. The exceptional character of swords means that their production or recycling cannot be viewed in terms of routine practice. That said, swords were clearly also subject to regular damage, and recycling should therefore not be completely excluded. The absence of swords from Bronze Age graves in many parts of Europe emphasizes the extent to which these objects remained in circulation over generations. Another frequently used method for withdrawing swords from circulation were hoarding practices, and a preference for wetland depositions after c. 1500 BCE shows that swords often ended their use-lives by being handed over to water-bodies and not by being recycled.

Analytical data were obtained from 145 samples taken from 118 swords and daggers covering the period 1600–1100 BCE. It should be stressed that most swords have already been analysed and published. However, 36 of the total number of 118 have recently been sampled and analysed within this project or are previously unpublished.

There are 101 chemical and lead isotope data from 83 swords/daggers from Scandinavia, 27 from 20 swords from Northern Germany seven from six swords from Southern Germany and 10 from 9 swords from Northern Italy.

The following main sword types are comprised by this study: Full-hilted swords dated to c. 1600–1100 BCE, hilt-plated swords (1600–1100 BCE), flange-hilted swords (1500–1100 BCE) and octagonal-hilted swords from 1500 to 1300 BCE. All these occur in various sub-types. Nordic full-hilted swords are generally distributed only in the Nordic zone, while the other sword types have a more widespread geographical distribution.

Scandinavia

Out of the total of 83 swords from Scandinavia there are 33 full-hilted swords and daggers. Eleven of the earliest type of full-hilted swords and daggers from Scandinavia date to around 1600–1500 BCE (IB) and are considered to be associated with the Hajdúsámson-Apa type found in Hungary. While some swords from Denmark, e.g. Stensgård are potential import pieces from Hungary; other swords found in Denmark and Norway should rather be seen as derivates of the Hungarian sword types. There are, however, also swords and daggers from the indigenous Valsømagle horizon. Additionally, eight full-hilted swords of typical Nordic type dated to 1500–1300 BCE, and nine dated to 1300–1100 BCE have been analysed.

Early forms of short swords and daggers in Scandinavia (and Central Europe) tend to be dominated by the hilt-plated type with an organic hilt. Imports from Central Europe as well as local replicas date to about 1700–1600 BCE.

Northern Germany

For comparison lead isotope and chemical data for 20 swords is included. This material includes five hilt-plated swords (one of Sögel/Wohlde type, one of Wohlde type) dated to 1600–1500 BCE. However, a majority of the swords are of the full-hilted type and can be dated to 1500–1300 BCE. Five of the full-hilted swords belong to the subtype octagonal-hilted sword; the remaining three to the typical Nordic type. There are four flange-hilted swords dated to 1500–1300 BCE. From the succeeding phase, 1300–1100 BCE, there are three full-hilted swords of the Nordic type.

Southern Germany

The sampled swords from southern Bavaria are assigned to the types Kirchbichl and Hausmoning with circle eye decorations on the octagonal handle and the type Leonberg with paragraph rows on the handle. Two of the metal-hilted swords cannot be associated with one type. The six southern Bavarian full-hilted swords with decorated octagonal grip cover geographically the core zone of the north-alpine distribution of these swords, as well as the characteristic types. Chronologically, they are dated to the later 15th and the 14th century BCE.

Northern Italy

The samples (10) from Italy are from 9 swords and 1 rivet from one of them. All the objects come from Northern Italy, in particular from the Veneto and Friuli-Venezia Giulia regions. The age of the swords varies from the Middle Bronze Age 1 (MBA1) to the Recent Bronze Age (RBA) that in absolute date corresponds to 1650–1150 BCE.

Ιnterpretations of lead isotope and trace elemental data are based on comprehensive databases (i.e. the Alpine ArchaeoCopper Database for the Alpine area (Nimis et al., 2012; Artioli et al., 2016a), which has been merged into the extensive databases based on OXALID, n.d. and expanded as described in Ling et al., 2014, which are constantly updated with new evidence, most recently from the Italian Alps, Spain (Montero-Ruiz, 2017) and Mitterberg (Pernicka et al., 2016a). All ore data used for comparisons with the lead isotope ratios of the swords were either measured by TIMS or MC-ICP-MS, both with the same overall analytical error of 0.1%, even if the within run errors of the latter are mostly much smaller than in the former.

The analytical results show that the small number of the earliest swords, dated to 1600–1500 BCE, have large spread of lead isotope compositions indicative of several sources of copper. Swords dated to this period generally have low concentrations of nickel (Ni) and/or arsenic (As), but exceptionally up to about 1% and 0.8%, respectively. Antimony (Sb) and silver (Ag) are present in low concentrations in a few swords, but characteristic fahlore signatures, with significant As, Sb and Ag, are not observed. Seven different deposits can be identified as possible sources of copper for the swords dated to 1600–1500 BCE.

The swords dated to after 1500 BCE show more homogeneous compositions, both in terms of lead isotope and trace elemental data. The main impurities, approximately up to 1% in the majority of the swords dated to 1500–1100 BCE are Ni and As, but some swords are nearly free of impurities. In general, for the bulk of the samples, the concentrations are slightly shifted towards lower concentrations of these elements for the later swords, dated to 1300–1100 BCE. The same feature can be noted for Sb, but its concentrations are lower by an order of magnitude. The concentrations of Ag are generally much lower than of the other three elements (<0.1%), and there seem to be a shift towards somewhat higher levels for the later swords.

The swords dated to 1500–1300 BCE show consistency with five ore deposits but only three of them are predominant, however, during the latest phase, 1300–1100 BCE, most swords seem to be made of copper from one dominant region. It was observed that all analysed swords can be divided into several groups with very similar chemical and lead isotope compositions, therefore single samples, or swords, will not be discussed separately. However, some of them have very different isotopic and/or chemical compositions and they merit individual assessment.

The chemical and lead isotope compositions of the sixteen (18 samples) Scandinavian and German swords dated to 1600–1500 BCE are consistent with the geochemical characteristics of several mining regions known to be exploited at that time in Europe, including the Great Orme in Wales, Italian Eastern Alps, Mitterberg in Austrian Alps, Slovakian Ore Mountains, South Spain and Cyprus. Additionally, one sword from Italy dated to this period, has lead isotope and chemical compositions consistent with the ores from Tuscany, Campiglia Marittima, Temperino mine.

Copper ores from the British Isles are potential sources for five swords dating to 1600–1500 BCE. All of them are full-hilted swords from Denmark. Among these are three swords of the Nordic Hajdúsámson-Apa type from Dystrupgård and the Hajdúsámson-Apa sword MA-135003. These, and also the Valsømagle type sword MA-135001, have lead isotope ratios consistent with the copper ores from the Great Orme mine in Wales, which was exploited throughout the Bronze Age, however with a peak in production 1600–1400 BCE. The As concentration in these swords are somewhat low in relation to Ni compared to typical signatures for ores from Great Orme but within the Great Orme range. Also, the low Sb impurity levels (maximum in these swords is 0.04%) are coherent with copper from Great Orme. There is, furthermore, a clear distinction between the two Scandinavian countries, Denmark and Sweden, since the ores from the Great Orme in Wales are not represented at all among the swords from Sweden.

One sword, a hilt-plated dagger, both from Sweden, and a hilt-plated sword from Norway (B5469a) are isotopically consistent with the ores from the Italian Eastern Alps (Southalpine AATV).

The full-hilted Danish sword, MA-135005, low in impurities, is consistent with the ores from Mitterberg, in the Austrian Alps. Two hilt-plated swords from Northern Germany, with mutually similar contents of Ni (ca 0.3%) are different in terms of lead isotopes and content of As and Ag. One (MA-071243), has lead isotope ratios characteristic of the radiogenic lead, consistent with the ores from Mitterberg. The other, with slightly higher Ag (MA-071222), has the LI ratios consistent with the ores from Slovakia. Also, the lead isotope compositions of two samples from the full-hilted sword from Norway (ALM 25 and 26; blade and hilt) are consistent with the Slovakian ores from Spania Dolina, although not identical. Both alloys have low concentrations of Ni and As but are different in terms of Sb. Two other full-hilted swords are consistent isotopically with the ores from Slovakia (Banska Štiavnica): from Denmark (MA-134999; Hajdúsámson-Apa), and from Sweden (UM29218_458A).

The Hajdúsámson-Apa swords MA-135005 and MA-134999 from Denmark, analysed by Bunnefeld (2016a), and an associated sword from Norway (ALM 25 and 26) fit well with lead isotope and chemical data of objects, including swords, from the Hajdúsámson and Apa hoards, probably suggesting that the copper derived from ores in the Mitterberg area or the Slovakian Ore Mountains.

Some swords have lead isotope and elemental compositions that are also consistent with other ore sources than these major sources listed above. One of these is a sword from Vreta in Sweden (FG 050575) with a very low concentrations of impurities. In terms of manufacture, the sword from Vreta is quite extraordinary and shows strong typological links with the swords from the Nebra hoard as well as the ones from the Hajdúsámson-Apa hoards, which were identified as produced from the copper from Mitterberg. However, its lead isotope ratios seem consistent with the copper ores from an ancient copper mine of Aznalcollar near Seville, in the copper rich region Huelva in southern Spain where copper has been exploited in the Bronze Age. Another sword that is made of copper geochemically different from the majority of the analysed swords, is a full-hilted sword of the Valsømagle type from Sweden (SHM 617282) which has lead isotope ratios that are consistent with the Cypriot copper ores and slags as well as with the copper of oxhide ingots from Uluburun. The unusually low concentration of trace elements in this sword further supports such an interpretation.

Overall, the swords dated to 1600–1500 BCE show a notable variation in ore sources. They indicate a very dynamic diffusion of metal across Europe. Furthermore, metalworking styles and metal in most cases seem to have different origins. The period 1600–1500 BCE is traditionally thought to represent the ‘breakthrough’ of the Nordic Bronze Age, with the development of Nordic metalworking styles, emerging from established connections with Bronze Age cultures in the Carpathian Basin and the Mediterranean. In contrast with this, the sword data indicate that metal was brought to Scandinavia from several different sources, and not mainly from the Slovakian Ore Mountains. This may be a reflection of traditions going back to the preceding period, when Scandinavia was supplied with copper from both Central Europe and the British Isles.

The large group of 58 swords dated to 1500–1300 BCE shows consistency with five ore deposits, and of these three are predominant: Italian Eastern Alps, Austrian Mitterberg and Slovakian Ore Mountains.

The trace elemental compositions for the majority of the 43 analysed swords (48 samples) dated to 1300–1100 BCE are characterised mainly by Ni and As (ca. 0.2–1% Ni and 0.1–0.6% As). Generally, Ag is below or much below 0.10%; the same is generally noted for Sb. However, some of the swords have very low impurity levels, distinguishing them from the majority. The lead isotope ratios show that the great majority has lead isotope ratios consistent with the ores and slags dated to 1300–1100 BCE from the Eastern Alpine deposits of Trentino–Alto Adige and are also consistently well aligned with the isochron of the south-eastern Alpine deposits.

However, a few of the swords found in Denmark have lead isotope ratios suggesting other copper ores; e.g. the sample from the blade of a full-hilted sword of Nordic type (5: Ke 5023 A) is consistent with the ores from the mine of Calabona in northern Sardinia, but also very close to some ores in the Slovakian Ore Mountains. The hilt of the same sword (6: Ke 5023 B) seems more consistent with the ores from the Italian Alps but also with the mines of Linares in southern Spain. Also, two other flange-hilted swords (10: Ke 5077 and NMK142: B2295) and a dagger (NMK 659: B3810) have LI ratios within the overlapping isotopic fields of the Italian Eastern Alpine deposits and ores in south Spain, but also with some ores of the French Massif Central.

To conclude, the chronology and geography of the deposited swords indicate that the different regions relied on different metal supplies, that changed during the course of the Bronze Age. Furthermore, this changed from a disparate pattern in the earlier phases, to a more homogenous pattern in the later phase, with fewer suppliers covering larger areas.

(Source: “Moving metals IV: Swords, metal sources and trade networks in Bronze Age Europe”, by Johan Ling et al.)

Research-Selection for NovoScriptorium: Philaretus Homerides