Early Bronze Age metallurgy: a newly discovered copper manufactory in southern Jordan.

Levy, Thomas E. ; Adams, Russell B. ; Hauptmann, Andreas 等

The Early Bronze Age (EBA, c. 3600-2000 BC) represents a time of fundamental social change in the southern Levant when the first fortified towns and urban centres evolved. Since the 1930s, scholars have linked advances in metallurgy with the emergence of urbanism and the rise of some of the earliest civilizations (Childe 1930). Recent excavations at the site Khirbat Hamra Ifdan (KHI) in the copper-ore-rich Faynan district of southern Jordan have revealed the largest Early Bronze Age metal workshop in the Middle East, and have yielded thousands of finds related to ancient copper processing. This unique assemblage of archaeometallurgical remains includes crucible fragments, prills and lumps of copper, slags, ores, copper tools (e.g. axes, chisels, pins), copper ingots, a few furnace remains and an extensive collection of ceramic casting moulds for ingots and tools. The archaeometallurgical data from this site provides vital information for accurately reconstructing EBA metal processing as well as some of the dimensions of trade relations that were linked to significant changes in social evolution in that period. The `manufactory' (Costin 1991) complex at KHI survived in a remarkable state of preservation due to the structures being sealed by wall collapse as a result of earthquake activity at the end of the Early Bronze III (c. 2700-2200 BC). Thanks in part to this `Pompeii effect', KHI represents the first near-complete EBA metal workshop in the ancient Near East. Use of on-site Geographic Information Systems (GIS) data recording has greatly streamlined and facilitated the recognition of activity areas linked to ancient metal production at the site (Levy et al. 2001). The analysis of the metallurgical data from KHI will enhance our understanding of the copper production processes and provides an important analytical lens for monitoring the oscillations in social change in a region traditionally viewed as a periphery to the ancient EBA centres of civilization--namely Egypt and Mesopotamia.

The Faynan district, ~50 km southeast of the Dead Sea, was the most important resource area for copper in the southern Levant (FIGURE 1). The Jabal Hamrat Fidan (JHF) represents the western `Gateway' to Faynan, a region which contains evidence for numerous single and multiple occupation sites dating to the entire span of the EBA and was the largest source of copper ore during this period in the southern Levant. Sites such as KHI provide ideal `open-air laboratories' for investigating the social dimensions of these ancient production centres.

[FIGURE 1 OMITTED]

The Faynan district is best known from biblical texts as the location of Punon, one of the 40 stations mentioned in the book of Numbers (Numbers 33: 42-43; generally identified with the Greek Phaino, the present Khirbet Faynan, Jordan) and in the archaeological literature for extensive evidence concerning ancient mining and smelting sites spanning the Chalcolithic (c. 4500-3600 BC) through the Islamic periods (Hauptmann 1989; 2000; Hauptmann et al. 1992).

KHI is situated on a naturally defended mesatop plateau or `inselberg' aproximately 25 m above the wadi channel, and located in the middle of a seasonal drainage known as the Wadi Fidan. The area was first visited by Frank (1934), in part by Glueck (Glueck 1935; Adams 1992), and reported on by several other survey projects (Knauf & Lenzen 1987; MacDonald 1992). The site was first probed by Adams (1999; 2000), whose early excavations demonstrated the rich potential of the site for examining the technological and scalar changes in copper production during the EBA. In 1999-2000, the first large-scale excavation work was carried out at the site under the direction of T.E. Levy and R.B. Adams aimed at providing a foundation for examining the social context of EBA metallurgy (Levy et al. 1999).

EBA metallurgy in context and at KHI

The earliest metal production sites in the southern Levant date to the Chalcolithic period (c. 4500-3600 BC), when production was carried out near large corporate buildings within village settlements found primarily in Israel's northern Negev desert far away from the source areas of copper in the Arabah valley and Sinai Peninsula (Levy & Shalev 1989). Chalcolithic metallurgy was linked to the emergence of the first chiefdom organizations in this part of the Near East and seems to have been controlled by social elites (Levy 1998). By the beginning of the EBA I (c. 3600-3300 BC), with the collapse of Chalcolithic settlement over much of Palestine, for the first time, metal production took place near the ore sources of Faynan in Jordan. With the rise of fortified towns in the southern Levant during the EB II-III (c. 2900-2200 BC), and within the contexts of developing complexity in the Levant and surrounding cultures (i.e. Old Kingdom Egypt, see below), metal production increased dramatically in scale. In the copper district of Faynan, for the first time large-scale EB II-III mining activities were carried out in galleries and shafts. To smelt the ore, 13 contemporary smelting sites were discovered where some 5000 tons of slag reveal a metal production of perhaps several hundred tons of copper during the EBA (Hauptmann 2000). This is by far the largest evidence of copper production during this period in the ancient Near East. The recent excavations at KHI provide a powerful lens for examining the more general processes of social evolution seen through the development of craft specialization in ancient metallurgy. Scalar differences in metal production between the Chalcolithic period and the EB III are marked and point to significant differences between metal production for chiefdom elites in the former period, and more `commodity driven' production in the latter (Adams 1999).

Three main occupation phases have been identified at KHI. Stratum I represents later occupations from the Islamic, Byzantine and Iron Age periods. Stratum II dates to the EB IV occupation and Stratum III to the EB III period when the site was occupied most extensively. There are indications of an EB II occupation in Stratum IV; however, the excavation sample size is too small to make meaningful observations. Radiocarbon dates from the region, coupled with ceramic and stratigraphic data provide a solid chronological framework for evaluating changes in metal production during the last phases of the EBA at Faynan (TABLE 1).

The main occupation phases at KHI are Stratum III (EBA III, c. 2700-2200 BC) that coincides with the consolidation of the earliest fortified sites throughout the Levant and Stratum II (EB IV, c. 2200-2000 BC) that is often linked to a period of social collapse. The reasons posited for the collapse of EBA urbanism are varied and include: general climatic deterioration in the 3rd millennium Near East (Weiss et al. 1993), Egyptian conquests during the Old Kingdom Fifth Dynasty (Redford 1992), invasions of West Semitic `Amorite' tribes from Syria, internal social upheaval and other factors. It is against these formative processes of social change that EBA metallurgy evolved.

Early Bronze Age metal workshops have been found across the Near East in Oman at Maysar 1 (Weisgerber 1981); in Iran at Shahr-i Sokhta (Heskel 1982), Tepe Hissar (Pigott et al. 1982), and Shahdad (Pigott 1989; 1999); in Anatolia at Arslantepe (Palmieri et al. 1993), Kestel (and Goltepe) (Yener 2000), Norsuntepe (level VIII --Muller-Karpe 1994), Tepecik, Cudeyde (Goldman 1956), Gozlu Kule/Tarsus (Goldman 1956), Degirmentepe (Esin 1988), Hisarlik (Level II, Troy--Schliemann 1881). Interestingly, the nearby (c. 100 km south) copper mines at Timna in southern Israel have produced little evidence of EBA metal processing and the island of Cyprus, the most important source of eastern Mediterranean metal during the Middle and Late Bronze Age (c. 2000-1300 BC), has virtually no evidence of contemporary EBA metal production or processing. More significant is the relatively small scale of production and treatment of raw metal reflected at these other Near Eastern EBA metal production sites noted here. Until the recent excavations at KHI, the largest assemblage of metallurgical remains of that period came from Schliemann's excavations at Troy (Hisarlik) where 70 crucible fragments, 70 casting moulds and some other metal production remains were found. In addition to the small sample sizes from these sites, little data has been quantified, making comparative studies difficult.

The recent excavations at KHI (FIGURE 2) shed new light on the scale and organization of Near Eastern EBA metal processing due to the unusually large sample of artefacts related to production activities. The excellent preservation of these materials at the site and the relatively undisturbed contextual data of these finds point to `factory-like' production technologies. A total of 3782 archaeometallurgical finds were recovered from the EBA strata at the site in an excavation exposure of c. 977 sq. m. The total volume of deposits excavated for the two main EBA strata are Stratum II--EB IV = (160.58 cu. m) and Stratum III--EB III (405.91 cu. m). The finds can be divided (TABLE 2, FIGURE 3) into:

a copper metal objects (axes, chisels, pins, ingots)

b copper ores (most identified with specific source areas in Faynan)

c copper prills (droplets of metal obtained through smelting) and copper lumps (produced by re-melting smaller batches of raw copper)

d slag (extremely metal-rich slags transported as an intermediate product from smelting sites in the Faynan area to KHI for further metal extraction, and slag from remelting in crucibles)

e casting moulds (for axes, chisels, pins, blades, and ingots)

f remains of pyrotechnical facilities (small crucibles for melting copper-rich slag, batches of copper, partly for smelting ore; some clay rods or `Lady Fingers' that functioned in smelting furnaces) and

g spherical and elliptical grooved stone hammers used for further treatment of metal objects or breaking up casting moulds.

[FIGURES 2-3 OMITTED]

The organization and scale of metal production

Identification of the distribution patterns of the thousands of metallurgical remains discovered at KHI was facilitated through GIS and the ArcView 3.2 program. As highlighted in the flow chart (FIGURE 4) and the graphic presentation of data through GIS maps (FIGUREs 5A, 5B), by identifying clusters of archaeometallurgical data across the site in courtyards, rooms and other architectural features, it is possible to reconstruct all of the stages in metal processing at this site. Accordingly, ore extracted through mining activities was mostly smelted near source; however, the distribution of small quantities of ores and slag at KHI points to some on-site smelting of ore in crucibles. The bulk of pre-cast metal arrived in the form of prills (droplets of smelted metal), small metal batches and copper-rich slag from the specialized smelting sites (Hauptmann 2000) (FIGURE 1). In terms of pyro-techniques, the concentrations of metal prills and metal lumps across the site, and evidence of re-cycling based on clusters of re-melted copper prills, ingots, and tools `frozen' in the negative impression of a crucible base (referred to as the KHI `cupcake'; FIGURE 3:2), point to the centrality of melting and re-melting activities for casting at KHI.

[FIGURES 4-5 OMITTED]

It is possible to identify an EBA `factory-line'-like operation for the production of copper tools and ingots at the KHI settlement. As seen in the GIS map for the EB III period (FIGURE 5A), specialized metal-processing activities were concentrated in a variety of the more than 80 rooms, courtyards and other spaces excavated at the site. The primary melting and casting of metal took place in the largest (southern) courtyard (No. 54; FIGURE 2) exposed at the site. Here hundreds of moulds were used to cast copper ingots and final products such as metal axes, chisels, pins and blades. The moulds appear to have been broken (probably with hammerstones) to retrieve the metal objects and then discarded in refuse areas surrounding the courtyard. Metal products were then moved north through the site to other rooms (e.g. in the northern portion of the excavation area and Rooms 25, 43 and other areas; FIGURE 5A) where hammering on anvils, grinding, polishing and other final production activities took place. Using GIS and Nearest Neighbour analyses ArcView plug-ins such as `Calculate Density' and `Kernel', a wide range of other artefacts (stone hammers, grooved hammerstones, ceramics, etc.) linked to metal production have also been analysed to clarify the `Metallurgical Chain' at KHI. Although lack of space precludes including all these studies here, the full range of metal production activities is highlighted in the flowchart (FIGURE 4).

The magnitude of the KHI metallurgical assemblage for the EB III (TABLE 2), its comparison with other contemporary Near Eastern production sites, and the highly specialized nature of KHI during this period point to a `quantum leap' in the scale of early metal production at a time when the earliest Levantine urban centres were growing and solidifying their power. GIS plots comparing the EB III (Stratum III; FIGURE 5A) distribution of metallurgical remains and architecture with those of the EB IV (Stratum II; FIGURE 5B) highlight the nested and structured organization as well as scale of metal production in the EB III urban period. In the following (EB IV) period, a time of general social dissolution in the southern Levant, there is a distinct absence of architectural features and infrastructure associated with metal production. The quantity of metallurgical remains drops markedly from c. 3356 in the EB III to only c. 404 in the EB IV. Regional surveys in the research area show (with the exception of KHI) a paucity of EB IV sites that would have provided a local subsistence base for the inhabitants of the site. However, in the EB III there was a network of small hamlets, agricultural installations, cemeteries and other sites linked to KHI. Thus, the JHF settlement pattern data, coupled with on-site metallurgical production data at KHI indicate a much less intensive, periodic and unsystematic scale of production in the EB IV compared with the EB III. These results confirm the generally accepted view of a social `collapse' in the EB IV; however, by focusing on craft production as an analytical lens for examining change, the foundations of a more quantitative method for measuring change has been achieved.

Processing and trade of copper at KHI

With the discovery and analyses of the metal workshop at KHI, for the first time in Old World archaeometallurgy research there is now strong evidence that provides better insight into the craftwork of copper processing during the 3rd millennium BC. Until now, reconstruction of these processes was based on simple speculative chemo-physical models of modern metal technology. Such models indicate a refining of raw copper, i.e. a deliberate purification from impurities such as lead, arsenic, iron or nickel by blowing air into the liquid metal bath. Accordingly, impurities get oxidized until the liquid copper itself is transformed to copper-oxide. Hence, a typical feature of refining is a surplus of cuprite ([Cu.sub.2]O) in the crucible. The new discovery reported here changes this view by providing fresh empirical data from the Early Bronze Age. The finds from KHI show that a variety of metallurgical materials were delivered from outside to this EB settlement/manufactory. Among these were slags with copper contents up to 10% weight or more (`furnace conglomerate') and thousands of prills and small batches of copper with adhering slag. They are proof of the partially limited success of smelting processes carried out at the smelting sites in the Faynan area (FIGURE 1). In addition, high-grade pieces of malachite, chrysocolla and copper-chlorides were transported from the nearby Faynan mines to KHI. At the KHI processing centre, they were all repeatedly smelted and melted in crucibles in order to remove slag, and to gain larger lumps of metal of sufficient quantity to cast into an ingot or tool (FIGURE 3). The smelting of ores perhaps followed a traditional workshop recipe that goes back to the Chalcolithic period, when smelting activities were carried out inside the settlement. The hypothesis that ancient metal workshops had their own `recipes'--which might have been characteristic for certain regions--has been suggested earlier (Junghans et al. 1960).

The multiple re-melting and re-cycling of copper lumps and prills into larger units are probably identical with the production of washed or purified copper as described in 3rd-millennium BC cuneiform tablets from Ebla in Syria (Reiter 1997). This was called `urudu-luhha', in contrast to urudu, which means un-purified copper. Before a more in-depth knowledge on EBA copper processing was available, it was suggested that such purification would mean a deliberate refining of metal in the modern metallurgical sense (Waetzold & Bachmann 1984). As shown by the phase contents of slags and copper, this was not the case at KHI, and there is no evidence as yet for refining at any other of the ancient Near Eastern EBA workshops. The skill and expertise of the KHI metallurgists was remarkable as they produced high-quality copper perfectly suitable for a flawless casting and ready for export in the form of both ingots and finished tools (FIGURE 3). Alloying was not performed in the KHI manufactory. This is an important observation, as tin bronzes are extremely rare in the EBA II/III southern Levant. It seems that this metallurgical process was carried out at urban centres in the north and the northwest (Hauptmann 2000).

In addition to technological questions, archaeometallurgical studies concerning provenance provide insights on where the copper from Faynan was traded. Sourcing metal is based upon a chemical `fingerprinting' of impurities in ore (the source) and a comparison of finished objects with this ore. Methods to determine these impurities comprise both chemical analyses of trace elements, and the analyses of lead isotope abundance ratios (LIA).

For Faynan, reference compositions of ores and (EB) raw copper were previously determined (Hauptmann et al. 1992). Similarities between the two were found (FIGURE 6), but a large range of trace element variation seems to be characteristic. The chemical composition of crescent-shaped bar ingots from KHI indicate a considerable bunching of trace elements that results from the metal processing as described above: i.e. re-melting and re-cycling led to a homogenization of varying compositions of copper that now may be seen as a representative average composition.

[FIGURE 6 OMITTED]

This chemical composition is also reflected in a number of late EBA crescent shaped bar ingots that were found at several localities in the southern Levant (FIGURE 7). These data point to the origin of the ingots in the area of Faynan, an observation that is also supported by the stylistic attributes of the ingots (Adams 1999; 2000). Unmistakable, however, are lead isotope abundance ratios, which do not change between ore and metal, no matter how many (s)melting processes the ore/metal was exposed to. These observations are based on the analyses of some 70 south Levantine crescent-shaped bar ingots at the Max-Planck-Institut fur Chemie at Mainz, and at the Geochronological Laboratory, University of Munster. All compositions match exactly the range of ores and raw copper from Faynan. Admittedly, we cannot exclude that bar ingots derived from Timna copper, but such a suggestion lacks any archaeological evidence. The bar ingot trade from Faynan supports previous observations of a pronounced `metallurgical drift' of Faynan copper to the north, northwest (Hauptmann 2000) and southwest since the Chalcolithic period. Finally, the new data from KHI in Jordan support hypotheses that relate the extensive `EB IV' settlements in the Negev desert of Israel with an ancient copper trade network that ultimately reached Old Kingdom and Intermediate Period Egypt (Haiman 1996; Adams 1999). Recent petrographic analysis of ceramics from these EB IV sites in the Negev suggest the possibility that in fact some of these sites have a longer period of occupation and probably originated earlier during the EB III. This is confirmed by both typological and petrographic analysis of the KHI ceramics and comparison with the evidence from the Negev (Goren 1996; Adams 1999). The combination of the ceramic and metallurgical similarities make a compelling case for the re-dating of the some of these Negev sites.

[FIGURE 7 OMITTED]

Conclusion

The evidence for large-scale and multi-phased manufacture in copper production at KHI during the later phase of the EBA presents a unique opportunity to examine the evolution of a highly specialized craft at a time when the Levant `periphery' and neighbouring `core' civilizations such as Egypt were undergoing major sociopolitical changes. In particular the rise of fortified Early Bronze Age towns in the Levant suggests developing complexity in social organization, which parallels the rise of Pharaonic Egypt during the Old Kingdom, albeit on a smaller scale. In the Levant, a part of these social changes were the development of elites and corporate authorities in the new urban centres that controlled specialized craft and subsistence production (Costin 1991). The recent excavations reported here are different from the normative EBA fortified town pattern in that KHI is

a a naturally fortified, non-urban site that lacks evidence of palatial and large scale architecture (FIGURE 5A)

b located in the desert periphery rather than the Mediterranean `centre' of EBA urban settlement, and

c contains evidence for highly organized craft production from an as yet to be determined organizing body.

Current models for understanding the role of pre-industrial elites suggest they manipulated production and distribution of staples (e.g. cereals and other commodities) and/or wealth objects (craft products) as a mechanism to enhance and consolidate their social status (Johnson & Earle 1987). It is against this background that our assessment of the implications of the large-scale production of copper at KHI can be made. Three primary factors are key to understanding the development of the metal trade in the region during the EBA: production, distribution (trade) and consumption (Costin 1991). Since our site at KHI is the key EBA site in the largest Levantine copper resource zone, we will have, perhaps for the first time, a fuller picture of the socio-economic processes, which until now have been dominated by assessments based upon the archaeological view as seen from evidence for end-of-the-line trade and consumption. Forthcoming in-depth research on the EBA metallurgical remains from KHI will provide the first comprehensive assessment of the role of metal production on the emergence, maintenance and collapse of the earliest urban phenomenon in the Levant from one of the region's major ore resource zones in this part of the ancient Near East.

TABLE 1. Radiocarbon determinations for JHF research area, during
the Chalcolithic-Early Bronze Age.

lab no. period site [sup.14]C date
 BP

HD-12337 Chalcolithic Tell Wadi Faynan 5740 [+ or -] 35
HD-12336 Chalcolithic Tell Wadi Faynan 5375 [+ or -] 30
HD-16380 EBA I Wadi Fidan 4 4702 [+ or -] 37
HD-16327 EBA I Wadi Fidan 4 4718 [+ or -] 25
HD-13776 EBA I Wadi Fidan 4 4684 [+ or -] 50
HD-16379 EBA I Wadi Fidan 4 4576 [+ or -] 45
HD-16378 EBA I Wadi Fidan 4 4422 [+ or -] 51
HD-13975 EBA II Barqa el-Hetiye 4376 [+ or -] 57
HD-10577 EBA II-III Faynan 9 4140 [+ or -] 109
HD-10573 EBA II-III Wadi Gwair 4 4059 [+ or -] 55
Beta-143811 EBA III JHF 120 (KHI) 4020 [+ or -] 70
HD-16533 EBA III JHF 120 (KHI) 4044 [+ or -] 40
HD-10994 EBA III Faynan 9 3973 [+ or -] 85
HD-10574 EBA III Ras en-Naqb 1 3971 [+ or -] 67
HD-10993 EBA III Faynan 9 3981 [+ or -] 50
Beta-143813 EBA III JHF 120 (KHI) 3960 [+ or -] 50
HD-10579 EBA III Faynan 16 3923 [+ or -] 61
Beta-143810 EBA III JHF 120 (KHI) 3870 [+ or -] 40
HD-16529 EBA III Wadi Gwair 3 3919 [+ or -] 26
HD-16534 EBA III JHF 120 (KHI) 3914 [+ or -] 45
HD-10584 EBA III-IV Faynan 9 3812 [+ or -] 77
Beta-143812 EBA IV JHF 120 (KHI) 3650 [+ or -] 60

lab no. [sup.14]C dates provenance publication
 cal BC (1[sigma])

HD-12337 4675-4575 Square A, Najjar et al. 1990
 L 23, -1.4 m
HD-12336 4330-4165 Square A, Najjar et al. 1990
 L 8, -1.05 m
HD-16380 3610-3375 Area D, L 4-14 Adams & Genz 1995
HD-16327 3615-3380 Area D, L 4-9 Adams & Genz 1995
HD-13776 3610-3365 Area A, L 50 Adams & Genz 1995
HD-16379 3360-3165 Area A, L 5 Adams & Genz 1995
HD-16378 3255-2920 Area A, L 22 Adams & Genz 1995
HD-13975 3080-2910 House 1, Fritz 1994
 27/91 L 13
HD-10577 2880-2500 smelting site Hauptmann 1989
HD-10573 2840-2490 smelting site Hauptmann 2000
Beta-143811 2605-2465 L 1236 this article
HD-16533 2585-2490 Trench 1, Adams 1999
 L 114
HD-10994 2575-2345 smelting site Hauptmann 1989
HD-10574 2565-2400 smelting site Hauptmann 2000
HD-10993 2560-2455 smelting site Hauptmann 1989
Beta-143813 2555-2535 L 1602 this article
 2490-2445
HD-10579 2465-2315 smelting site Hauptmann 2000
Beta-143810 2465-2205 L 1010-14758 this article
HD-16529 2460-2345 smelting site Hauptmann 2000
HD-16534 2460-2320 Trench 2, Adams 1999
 L 209
HD-10584 2395-2135 smelting site Hauptmann 1989
Beta-143812 2125-2075 L 1010-14208 this article
 2055-1935
TABLE 2. Summary of Early Bronze Age metallurgical finds from
Khirbet Hamra Ifdan, Jordan.

artefact category Stratum Stratum Stratum
 II III IV

metal objects
axes 0 3 0
chisels 0 0 0
pins 12 47 0
ingots 2 58 0
lumps 15 169 3

copper ores
OE (ore) 62 230 4
OE-DLS 10 35 0
 (dolomite-limestone-shale)
OE-PR (ore and prill) 0 3 2
OE-PSO 52 216 3
 (ore mixed with prill and slag)
OE-PSO/DLS 24 80 1
 (combination of both)
OE-SL (ore and slag) 7 37 0

prills
PR (prill) 21 280 1
PR-PSO 1 29 0
 (prill, mixed with slag and ore)

slag
SL (slag) 8 32 0
SL-SMS 19 82 1
 (shiny, glass-like black slag)

casting moulds
axes 1 25 0
chisels 0 58 0
chisel w/butt 0 29 0
pins 1 25 0
small ingots 0 3 0
ingots 16 345 0
indeterminate 51 366 1
blade 0 15 0

smelting facility remains
crucible fragments 40 625 6
Lady Fingers 2 117 0
furnace fragments 8 44 0

stone tools
hammerstones 45 342 0
grooved `mining or 7 61 0
 metalwork' hammer

total 404 3356 22

Acknowledgements. The authors are grateful to Dr Fawwaz Al-Khraysheh, Director General, Department of Antiquities of Jordan and Dr Ghazi Bisheh, the former Director General, for their support. Thanks also to Dr Pierre Bikai, Director, American Center for Oriental Research (ACOR), Amman, for his help. Special thanks to UCSD students Adolfo Muniz, Neil Smith, Yoav Arbel, Michael Homan, Sarah Melina and Celia de Jong for their help with data processing in the field and laboratory, James D. Anderson for surveying expertise, and to Mr Steger, Deutsches Bergbau-Museum, for performing chemical analyses. We are grateful to the C. Paul Johnson Family Charitable Foundation, the National Geographic Society, and the Deutsche Forschungsgemeinschaft for generous financial support and to our UCSD Summer Session Archaeological Field School students for their hard work. Finally, we thank the Bedouin tribes of the Qurayquira village (Jordan) for their friendship, hard work and hospitality.

References

ADAMS, R.B. 1992. Romancing the stones: new light on Glueck's 1934 survey of Eastern Palestine as a result of recent work by the Wadi Fidan Project, in P. Bienkowski (ed.), Early Edom and Moab: The beginning of the Iron Age in Southern Jordan: 177-86. Sheffield: J.R. Collis Publications. Sheffield Archaeological Monographs 7.

1999. The development of copper metallurgy during the Early Bronze Age of the Southern Levant: evidence from the Faynan Region, Southern Jordan. Unpublished Ph.D dissertation, University of Sheffield.

2000. The EBA III-IV Transition in southern Jordan: evidence from Khirbet Hamra Ifdan, in G. Philip & D. Baird (ed.), Breaking with the past: ceramics and change in the Early Bronze Age of the Southern Levant: 379-401. Sheffield: Sheffield Academic Press. Levantine Archaeology 2.

ADAMS, R.B. & H. GENZ. 1995. Excavations at Wadi Fidan 4: A Chalcolithic village complex in the copper ore district of Feinan, Southern Jordan, Palestine Exploration Quarterly 127: 8-20.

COSTIN, C.L. 1991. Craft specialization: issues in defining, documenting, and explaining the organization of production, in M.B. Schiffer (ed.), Archaeological Method and Theory 3: 1-56. Tucson (AZ): University of Arizona Press.

CHILDE, V.G. 1930. The Bronze Age. Cambridge: Cambridge University Press.

ESIN, U.H.S. 1988. Degirmentepe (Malatya) Kutarma Kazisi, IX Kazi Sonuclari Toplantisi I: 79-125.

FRANK, F. 1934. Aus der Araba I: Reiseberichte, Zeitschrift des deutschen Palastina-Vereins 57: 191-280.

FRITZ, V. 1994. Vorbericht uber die Grabungen in [Barqa.sup.-] el-[Heti.sup.-]ye im Gebiet von [Fe.sup.-][na.sup.-]n [Wa.sup.-][di.sup.-] el-`Arabah (Jordanien) 1990, Zeitschrift des Deutschen Palastina-Vereins 110: 125-50.

GLUECK, N. 1935. Explorations in Eastern Jordan II, Annual of the American Schools of Oriental Research 15: 1-288. New Haven (CT): American Schools of Oriental Research.

GOLDMAN, H. 1956. Excavations at Gozlu Kule, Tharsus II. From the Neolithic to the Bronze Age. Princeton (NJ): Princeton University Press.

GOREN, Y. 1996. The Southern Levant in the Early Bronze Age IV: The Petrographic Perspective, Bulletin of the American Schools of Oriental Research 303: 33-72.

HAIMAN, M. 1996. Early Bronze Age IV settlement pattern of the Negev and Sinai Deserts: view from small marginal temporary sites, Bulletin of the American Schools of Oriental Research 303: 1-32.

HAUPTMANN, A. 1989. The earliest periods of copper metallurgy in Feinan, Jordan, in A. Hauptmann, E. Pernicka & G.A. Wager (ed.), Old World Archaeometallurgy: 119-35. Bochum: Deutsche Bergbau-Museum.

2000. Zur fruhen Metallurgie des Kupfers in Fenan, Der Anschnitt 11.

HAUPTMANN, A., F. BEGEMANN, E. HEITKEMPER, E. PERNICKA & S. SCHMITT-STRECKER. 1992. Early copper produced at Feinan, Wadi Araba, Jordan: the composition of ores and copper, Archeomaterials 6: 1-33.

HESKEL, D. 1982. The development of pyrotechnology in Iran during the fourth and third millennia BC. Unpublished Ph.D thesis, Harvard University.

JOHNSON, A.W. & T. EARLE. 1987. The evolution of human societies--from foraging group to agrarian state. Stanford (CA): Stanford University Press.

JUNGHANS, S., E. SANGMEISTER & M. SCHRODER. 1960. Metall-analysen kupferzeitlicher und fruhbronzezeitlicher Bodenfunde aus Europa. Berlin: Mann-Verlag. Studien zu den Anfangen der Metallurgie Band I.

KNAUF, E.A. & C. LENZEN. 1987. Edomite copper industry, in A. Haddidi (ed.), Studies in the history and archaeology of Jordan 3: 83-8. Amman: Department of Antiquities of Jordan.

LEVY, T.E. 1998. Cult, metallurgy and rank societies--Chalcolithic period (c. 4500-3500 BC), in T.E. Levy (ed.), The archaeology of society in the Holy Land: 226-44. Leicester: Leicester University Press.

LEVY, T.E., R.B. ADAMS & M. NAJJAR. 1999. Early metallurgy and social evolution: Jabal Hamrat Fidan, ACOR Newsletter 11: 1-3.

LEVY, T.E., J.D. ANDERSON, M. WAGGONER, N. SMITH, A. MUNIZ & R. ADAMS. Digital archaeology 2001: GIS-based excavation recording in Jordan, The SAA Archaeological Record 1: 23-9.

LEVY, T.E. & S. SHALEV. 1989. Prehistoric metalworking in the southern Levant: archaemetallurgical and social perspectives, World Archaeology 20: 353-72.

MACDONALD, B. 1992. The southern Ghors and Northeast `Arabah archaeological survey. Sheffield: J.R. Collis Publications. Sheffield Archaeological Monographs 5.

MULLER-KARPE, A. 1994. Altanatolisches Metallhandwerk. Neumunster: Wachholtz. Offa-Bucher 75.

NAJJAR, M., A. E.-S. ABU DAYYA, E. SULEIMA, G. WEISGERBER & A. HAUPTMANN. 1990. Tell Wadi Feinan: the first pottery Neolithic Tell in southern Jordan, Annual of the Department of Antiquities of Jordan 34: 27-56.

PALMIERI, A., K. SERTOK & E. CHERNYKH. 1993. From Arslantepe metalwork to arsenical copper technology in eastern Anatolia, in M. Frangipane, H. Hauptmann, M. Liverani, P. Matthiae & M. Mellink (ed.), Between the rivers and over the mountains: 573-99. Rome: University La Sapienza.

PIGOTT, V. 1999. The development of metal production on the Iranian plateau, in V.C. Pigott (ed.), The archaeometallurgy of the Asian Old World: 73-106. Philadelphia (PA): University of Pennsylvania. University Museum Monograph Volume 89.

PIGOTT, V., S. HOWARD & S.M. EPSTEIN. 1982. Pyrotechnology and cultural change at Bronze Age Tepe Hissar, in T.A.S.F. Wertime (ed.), Early pyrotechnology. The evolution of the first fire-using industries: 215-36. Washington (DC): Smithsonian Institution Press.

REDFORD, D.B. 1992. Egypt, Canaan, and Israel in ancient times. Princeton (NJ): Princeton University Press.

REITER, K. 1997. Die Metalle im Alten Orient unter besonderer Berucksichtigung altbabylonischer Quellen. Munster: Ugarit-Verlag.

SCHLIEMANN, H. 1881. Ilios. Stadt und Land der Trojaner. Forschungen und Entdeckungen in der Troas und besonders auf der Baustelle von Troja. Leipzig: F.A. Brockhaus.

SEGAL, J., L. HALICZ & R. COHEN. 1999. A study of ingots and metallurgical remains from `Ein Ziq and Beer Resisim, Central Negev, Israel, in S.M.M. Young, A.M. Pollard, P. Budd & R.A. Ixer (ed.), Metals in antiquity: 179-86. Oxford: Archaeopress. BAR International series S792.

WAETZOLD, H. & H.G. BACHMANN. 1984. Zinn-und Arsenbronzen in den Texten aus Ebla und aus dem Mesopotamien des 3. Jahrtausends, Oriens Antiquus 23: 1-18.

WEISGERBER, G. 1981. Mehr als Kupfer in Oman, Der Anschnitt 5-6: 174-263.

WEISS, H., M.-A. COURTY, W. WETTERSTROM, F. GUICHARD, L. SENIOR, R. MEADOW & A. CURNOW. 1993. The genesis and collapse of third millennium north Mesopotamian civilization, Science 261: 995-1004.

YENER, K.A. 2000. The domestication of metals--the rise of complex metal industries in Anatolia. Leiden: Brill.

THOMAS E. LEVY, RUSSELL B. ADAMS, ANDREAS HAUPTMANN, MICHAEL PRANGE, SIGRID SCHMITT-STRECKER & MOHAMMAD NAJJAR *

* Levy & Adams, Department of Anthropology, University of California, San Diego, CA 92093-0532, USA. Hauptmann & Prange, Deutsches Bergbau-Museum, D-44791 Bochum, Germany. Schmitt-Strecker, Max-Planck-Institut fur Chemie (Otto-Hahn-Institut), D-55020 Mainz, Germany. Najjar, Department of Antiquities, Amman, Hashemite Kingdom of Jordan.

Received 19 September 2001, accepted 19 December 2001, revised 30 January 2002