Fig 1.
Increase in metal consumption c. 2400–1500 BC.
The histogram is based on standard values of metal weight calculated for each main artefact type (i.e. flanged axe, chisel, shafthole axe, and so on [15, 19, 26–44] per 100 year hence compensating for the differing length of the periods. LN II (2000–1700 BC) emerges as the first period of growth. NBA IA (1700–1600 BC) has slightly more metal in circulation. NBA IB (1600–1500 BC) is the breakthrough period, with plentiful metal. The two interpolation maps illustrate the development from LN II (B) to NBA IA (C) on a geographical scale, with orange areas denoting the highest densities. The total number of objects, on which Fig 1 is based, is 1879 from Denmark, Sweden and Norway. Reprinted from [19] based on map images provide by Natural Earth (public domain) under a CC BY 4.0 license, with permission from ArkIT and Helle Vandkilde, original copyright [2017].
Fig 2.
Model of multi-scalar connectivity in northern Europe around 2000 BC from local to super-regional levels and consisting of four overlapping spheres of interaction.
The Baltic coast of Mecklenburg-Vorpommern was the pick-up zone for metals to Scandinavia, but the three nodal hub regions–southern Britain with Wessex, eastern Denmark/Scania and the Circum-Harz region with Halle–were directly linked. Remarkably, the main resources on which these hubs relied–namely copper, tin and amber–were located outside their geographical precinct. Map images are provided by Natural Earth (public domain) under a CC BY 4.0 license. The map, first published by H. V. [19], is with permission reworked by L. H. and H. V. using the software Adobe Illustrator. Contains data from [15, 49–59].
Fig 3.
Metalwork in LN II (Fig 3A) and NBA IA (Fig 3B) quantitatively arranged according to the main types, local production or foreign import.
Although local production is by far the most common, it includes a category of hybrids. This indicates that imports were routinely remelted and some were recast in an ‘in-between’ style. This hints at copper mixing. Regional inventory studies were used to calculate the total number of objects in circulation [15, 26–41, 64].
Table 1.
Average trace element composition in mass percent of the major clusters defined via average-link cluster analysis based on logarithmic concentrations of arsenic (As), antimony (Sb), silver (Ag), nickel (Ni) and bismuth (Bi), executed on an extended dataset of 450 objects dating from LNII to NBA IB.
Fig 4.
Bivariate analysis of Ni/Sb and Bi/Sb values. The three fahlore types of clusters 10–11 stand distinct.
The Ni-free copper (cluster 12) is a separate type, while the medium-Ni (cluster 10) and high-Ni (cluster 11) copper show overlap, indicating a common ore source or possible mixing.
Fig 5.
Copper types and their association with artifact styles.
(A) The primary copper groups in Scandinavian LN II and their association with imported, local-style and hybrid artifacts. Únĕtician and British inputs differ markedly from one another, as is evident in both artifact style and copper type. (B) Continuity and change in copper type 2000–1600 BC in Scandinavia. Ni-fahlores and low-impurity copper occur throughout the period. Ni-free fahlore is a LN II phenomenon, while the minor-impurity copper of cluster 4 dominates in NBA IA.
Fig 6.
Lead isotope ratios of the artifacts from LN II compared with possible ore sources.
A) compares the most likely copper sources with the complete dataset for LN II while B) displays the different clusters during LN II. The ore data are from: Mitterberg ore district [48]; Hron valley, Slovakian Ore Mountains [59, 80]; Inn Valley, Alpine region [9]; Buchberg, Inn Valley, Alpine region [81]; Great Orme mining region, Wales [82–86]; Alderley Edge mining region [85, 86]; Valais valley, Switzerland [87]. The analytical uncertainties are comparable with the size of the symbols.
Fig 7.
Lead isotope ratios of the artifacts from NBA IA compared with possible ore sources.
A) compares the ore data with the complete dataset from NBA IA while B) displays the different copper groups during NBA IA. The ore data are from: Mitterberg ore district [48]; Hron valley, Slovakian Ore Mountains [59, 80]; Inn Valley, Alpine region [9]; Buchberg, Inn Valley, Alpine region [81]; Great Orme mining region, Wales [82–86]; Alderley Edge mining region [85, 86]; Valais valley, Switzerland [87]. The analytical uncertainties are comparable with the size of the symbols.
Fig 8.
Lead isotope ratios of artifacts of secured foreign origin during LN II.
The typologically British and Únĕtician artifacts (A) show a wider distribution than the typologically local-crafted artifacts (B). The comparison with ore bodies having pure copper highlights the comparability of British ores with British artifacts, more precisely with Great Orme metal. The ore-related data are from: Ross Island [85, 86, 88]; Great Orme mining region, Wales [82–86]; north and central Wales mining regions [86, 89]; Alderley Edge mining region [85, 86]; eastern Alpine ores [9, 81].
Fig 9.
Lead isotope ratios of artifacts of secured foreign origin during NBA IA.
The few typologically foreign artifacts (A) can be still correlated with cluster 8 metal, though different in type from the British ores. The alignment of new artifact types including spearheads supports the assumption of mixing. The typologically local-crafted artifacts (B) are consistent with cluster 4 metal and in this case can be correlated with British ores. The ore-related data are from: Ross Island [85, 86, 88]; Great Orme mining region, Wales [82–86]; north and central Wales mining regions [86, 89]; Alderley Edge mining region [85, 86]; eastern Alpine ores [9, 81].
Fig 10.
Principal Component Analysis of the main trace elements (Ni–Sb–Ag–As) of the LN II (dots) and NBA IA (squares) artifacts from Denmark, including the LN II Pile hoard in Scania.
Generally, colors indicate the artifact category, the data set of Pile [19] is highlighted with open symbols. The extended data set is compared with the median values for copper types defined by Rassmann [91, 93], where Nieder–Neuendorf copper is Ni-free, copper types Bresinchen and Bennewitz have medium Ni values, and copper type Dederstedt has higher Ni values. Reinkupfer (pure copper) is defined, on the basis of several finds including the Helmsdorf burial, as very low-impurity copper [91].
Fig 11.
Trace element ratios of As–Ni, As–Sb and Ag–Sb of the fahlore ore copper groups compared to the relevant ore bodies (normalized values) in the Slovakian Ore Mountains and the eastern Alps.
Ore values are normalized to copper and based on regional, interdisciplinary investigations of the specific mining regions [9, 59, 80, 81].
Fig 12.
Lead isotope ratios of the different fahlore groups from LN II defined in this study compared with relevant ore regions.
The plots clearly visualize the relation of (A-B) the Slovakian Ore Mountains with Ni-fahlore copper (the data represents the Hron valley, Slovakia [59, 80]), and of (C) the Inn Valley, especially the Buchberg, with Ni-free fahlore [9, 81].
Fig 13.
Logarithmic scatter plot of the trace elemental concentrations As–Ni, As–Sb and Ag–Sb of cluster 4 compared with the Mitterberg ore district and the Hron valley in the Slovakian Ore Mountains.
Ore values are normalized to copper and based on regional interdisciplinary investigations of the specific mining regions [48, 59, 80].
Fig 14.
Histogram of the tin percentage within (A) LNII and (B) NBA IA artifacts found in Denmark.
Fig 15.
(A) British-developed bronze flat-axe from Selchausdal, northwest Zealand (NM B5310, photo: Nørgaard). The 20-cm-long axe has a geometric decoration covering the surface. Low-impurity copper is alloyed with 10% Sn. Scandinavia holds the largest proportion of British type axes outside the British Isles 2000–1700 BC. (B) British axe interpreted as hack-metal found at Limfjord, Aalborg. Reprinted from [106] under a CC BY license, with permission from the Trustees of the British Museum.
Table 2.
Lead isotope ratios of artifacts of interest concerning the discussion of typological hybrids and foreign imports.
The artifacts presented here are plotted in Fig 16.
Table 3.
Trace element values of the artifacts of interest concerning the discussion of typological hybrids and foreign imports.
The artifacts presented here are related to Fig 16.
Fig 16.
Lead isotope ratios of artifacts that can typologically be described as hybrids and locally crafted artifacts made of low-impurity copper (clusters 1, 8 and 9).
(A) LN II artifacts with British lead isotope signatures (MLXX a, MLXXc, CM142, B294, B9819, B10789, VMÅ139 and VHM 5384) and local style artifacts (B1335, SØM 3197, NM3887) considered to be hybrids–(compare with Fig 8). (B) NBA IA artifacts that due to their isotopic signature are to be regarded as made (partly of) British high-tin metal (NM B644, B5557, B4077, NM26073). Conversely, the spread of cluster 8 artifacts suggests that the low-impurity copper used in the earliest Scandinavian Bronze Age derives from several different locations. The ore-related data are from: Ross Island [85, 86, 88]; Great Orme mining region, Wales [82–86]; north and central Wales mining regions [86, 89]; Alderley Edge mining region [85, 86]; eastern Alpine ores [9, 81]; Hron valley, Slovakian Ore Mountains [59, 80];.
Fig 17.
Trace element ratios (Ni/As, Sb/As) of the mining sites at Ross Island, Ireland, and Great Orme, Wales, compared with British and pseudo-British axes in addition to other artifacts in clusters 1–5, 8 and 9.
Ore values are normalized to copper and based on regional interdisciplinary investigations of specific mining regions [82–86, 89].