Identifying Mixtures

Identification of Mixtures

It is an unanswered question to which extent recycling or mixing of bronze or its components (copper, tin) was a common practice in the Early Bronze Age and late stages. However, this is an important question since it affects the ability to investigate the sources of the bronze owing to the changes of the chemical and/or isotopic composition. Recent publications using the holistic approach described above yielded indications that mixing of metals played a role from the Early Bronze Age onwards and showed that the issue can be investigated at best using a combination of isotope systems and chemical information. Building on these results, the Sögel Project, starts from the main hypothesis of intentional mixing of raw materials or batches of bronze in the Early Bronze Age. In particular, for objects with typologically close relationships like Sögel-Wohlde blades, such practice can be assumed. The Sögel project aims to reveal the mixing of materials in the Early Bronze Age to answer the question of whether the different blade types and single objects among each other are related. In addition, the relationships between blades in different European region pursued to answer the unsolved question of how the metallurgical tradition of making these blades developed in Northern Europe.

In principle, multiple components might have been mixed when bronze objects were produced in antuquity, but the higher the number of mixed components (tin, copper, bronze, lead) the lower the chance to recognise mixing and to identify the source of the single components. In the simplest way, two components were mixed, and their chemical and isotopic values behave strictly mathematical, where it is possible to reconstruct the source values of the single components by a mass balance calculation. In a bivariate diagram of two parameters, these mixtures plot on a straight mixing line between the source materials, the so-called endmembers, and are described by the relative composition of the two components. The resulting correlation can be distinguished in positive and negative. In the first, both parameter change relative to each other, while in the negative correlation, one parameter increases as the other decreases. Possible parameters are elemental concentrations and isotope values. To identify potential correlations in the data of the Sögel-Wohlde blades is the challenge of the project.

Theoretical mixing relationships between source and mixtures. Brief explanation: (a) represents a two-component system with components A and B. The blue solid line represents a simple mixing line with elemental concentrations, while the dashed (hyperbolic) lines may be produced when isotope values are used. (b) and (c) shows three- and four-component systems, which create mixing triangles or quadrangles. (d) shows the case of sequential mixing of (b), where the mixing of component C happened in a step after a mixing of A and B. From Berger et al. (2022).

Mixing of components can show up in different ways on a multi-parameter plot. For example, mixtures of two components produce straight lines in a ratio vs. ratio plot when the measure for the curve curvature is r = 1 (i.e. when the ratios of the concentrations of the two denominators are equal for all data points; otherwise it will be a hyperbolic plot), on element concentration vs. element concentration plots, when the denominator is the same for both ratios (e.g. ²⁰⁴Pb for ²⁰⁸Pb/²⁰⁴Pb vs ²⁰⁶Pb/²⁰⁴Pb) and when plotting a ratio against the reciprocal of a concentration (e.g. ²⁰⁶Pb/²⁰⁴Pb vs. 1/Pb). A hyperbolic curve is shown on ratio vs. ratio plots if (r ≠ 1, which determines the extend of the curvature depending on the relative concentration of the denominator in the endmembers) and on ratio vs. element plots. A ratio can be either element/element or isotope/isotope.

Hypothetical behaviour of isotopic data in case of the mixtures of two materials. (a) hyperbolic relationships of mixtures when r ≠ 1 in a ratio vs. ratio plot. The curvature of the hyperbole depends on the ratio of the concentrations of tin and copper in the bronzes responsible for e.g. ⁶ẟ⁶⁵Cu and ẟSn in Bz1 Bz2. If a bronze Bz2 is mixed to a bronze Bz1 the isotopic signature of Cu and/or Sn may change. If either  ẟ⁶⁵Cu or ẟSn of the added component is identical to the starting material, the respectively parameter does not undergo any changes, thus plotting parallel to the x- or y-axis, respectively. The same would happen, if pure copper or tin is added to Bz1. (b) shows several hypothetical mixtures: Bronze Bz1 had been mixed with copper Cu1 with an identical Cu but different Pb isotope composition. Another bronze (Bz2) had been mixed with different copper (Cu2) with identical Pb but different Cu isotopic composition, which might happen when copper of the same deposit but different ore minerals (e.g. chalcopyrite and malachite) have been mixed. Adding a bronze (Bz3) with a different lead and copper isotope composition to another bronze (Bz4) this mixture would result in an oblique or even curved mixing line. From Berger et al. (2022).

Determining a mixture, simple mixing calculations can be performed. However, for a reconstruction of the sources of a mixture, higher computational models like the Bayesian mixing calculation are used. Nevertheless, a clear reconstruction of the sources is not possible when more than two components were mixed, only possibilities and for the most likely sources and their portion to the mixture can be calculated. With such mixing models, based on correlations between several artefacts it might be possible to determine multiple mixing steps, what can be used to relate single objects to each other and to trace their manufacturing. In the archaeometallurgy, multi parameter approaches as used in this project are still rare, although recently several promising approaches have been conducted. For example, indications of mixing bronzes have been identified on artefacts from Apa-Hajdúsámson type swords and the findings from Salcombe. In these studies, it turned out essential to make a reasonable preselection of artefacts to be analysed and assessed for potential mixing practices. Analysing an arbitrary selection of objects from different regions, periods or artefact types will most likely produce datasets that are very difficult to interpret or have no relations at all. That is why the Sögel project concentrates on the different blade types of the Söhlde-Wohlde district district and potential precursors in other regions or artefact assemblages from graves and hoards.