The study of ancient Eurasian metallurgy has been suffering from (or preoccupied by) two conventional perspectives. One is that of the diffusion model emphasising the importance of the settled empires of ancient Egypt and Mesopotamia, of south-eastern Europe and of China (Shennan 1986, 1993; Kristiansen 1984). The supremacy of these 'cradles' of early civilisation is marked not only by social hierarchies, but also by technological inventions such as metal production. This view sees the mobile populations of the Eurasian steppe as occupying the "hinterland" of these early settled states in the south, believing that the emergence of metal technologies in the Steppe was the result of the expansions of "advanced" civilisations. The second perspective is rooted in the provenance study which traces metal objects back to their geological sources (Pernicka 2014). It assumes that chemical and isotopic composition of metal is static and only reflects a simple linear relationship between artefacts and specific ore deposits. Drawing from a legacy database of approximately 9,000 chemical analyses of copper-based artefacts, this thesis rejects the simplicity of both the diffusion and the provenance models. While admitting that the use of metal might have originated from western Asia, the development of metallurgy in the Eurasian steppe should be understood on its own terms. It is constantly re-shaped by vigorous circulation of metal artefacts across mobile communities on a regional or inter-regional scale. This observation is based on the application of a new innovative framework to interpret the patterns of compositional data (Bray et al. 2015). This novel method argues that metal can flow, quite literally, from one object to another as it is re-melted, re-mixed and re-cast in different shapes and colours, depending on different social contexts. Thermodynamic modelling and modern experiments have shown that during the copper melt, some volatile elements in copper alloys (e.g. arsenic, antimony, and zinc) are preferentially removed through oxidative loss. Instead, some elements, such as silver, nickel, and gold, tend to be preserved in metals. These predictable patterns of elemental losses provide valuable information to trace the directional flow of metal units between regions/cultures, if we combine chemical data of metal artefacts properly with archaeological context, landscape and chronology. By using this new methodology, several routes of copper supplies have been identified in the Steppe during different periods. They feature the exchange of metals within regional networks, fuelled by local copper sources. The Urals, central Kazakhstan, the Altai, and the Minusinsk-Tuva regions were the primary copper production centres that developed distinct trace-element chemistry and artefact typology. By contrast, alloying techniques employed by steppe peoples, generally demonstrate the long-distance connections based on two major metallurgical practices: arsenical copper in the western steppe and tin-bronze in the eastern steppe. Copper-arsenic production was concentrated in the Caucasus but the recycling of its arsenical copper became more apparent further away towards the Urals. On the other hand, the invention of tin-bronze metallurgy was triggered by the formation of the Seima-Turbino phenomenon (c. 2100- 1800/1700 BC) in the Altai, and this alloying tradition was amplified by the emergence of the Andronovo culture (c. 1700-1400 BC) in the Ural-Kazakh steppe. Tin-bronze ornaments, in particular, were exchanged between eastern and western mobile communities over a considerable distance, through the mechanism of pastoral seasonal movements. In conclusion, traditional views of diffusion and provenance theories cannot be uncritically applied to the inception of ancient metallurgy in the Eurasian steppe. Mobile pastoralists developed multi-regional production hubs based on the accessibility of ore resources and the variations in subsistence strategies. Although steppe metalwork revealed some technological borrowings from settled communities, steppe peoples had transformed them into locally adapted products that could fit into their socio-economic systems. That is, when dealing with the issues of Eurasian metallurgy, we should acknowledge the complexity of human engagement with metal and look into subtler differences in cultural context, landscape, and ideology.

This thesis considers the early copper and copper-alloy metallurgy of the entire Circum-Alpine region. It introduces a new approach to the interpretation of chemical composition data sets, which has been applied to a comprehensive regional database for the first time. An extensive use of GIS has been applied to investigate the role of topography in the distribution of metal and to undertake spatial and geostastical analysis that may highlight patterns of distribution of some specific key compositional element. The Circum-Alpine Chalcolithic and Early Bronze Age show some distinctively different patterns of metal use, which can be interpreted through changes in mining and social choices. But there are also some signs of continuity, in particular those which respect the use of major landscape features such as watersheds and river systems. Interestingly, the Alpine range does not act as a north-south barrier, as major differences in composition tend to appear on an east-west axis. Conversely, the river system seems to have a key role in the movement of metal. Geostastical analyses demonstrate the presence of a remelting process, applicable also in the case of ingots; evidence that opens new and interesting questions about the role of ingots and hoards in the distribution of metal at the beginning of the Metal Age. New tools and new analysis may also be useful to identify zones where there was a primary metal production and zones where metal was mostly received and heavily manipulated.

Doctorat en philosophie et lettres / info:eu-repo/semantics/nonPublished