From Metallurgy to Bronze Age Civilizations: The Synthetic Theory NISSIM AMZALLAG Abstract and the emergence of elites who controlled copper During the past few decades, evidence for the ancient production and trade. At the same time, the increas- smelting of copper has been discovered in areas isolated ing demand for metallurgical artifacts was regarded from one another. In most of them, the beginning of as a driving force that encouraged inventiveness and metallurgy had no substantial social and cultural conse- technical progress among the smiths. This scenario quences. Accordingly, the diffusionist theory (assuming the existence of a single homeland for metallurgy and its was mainly based on archaeological investigations in central importance in cultural development) has been re- Mesopotamia and Egypt, where copper artifacts were placed by a localizationist theory, in which the emergence found from the earliest stages of the development of of metallurgy is simply a continuation of the working of complex society.1 Consequently, the Near East came native copper. But neither of these theories is able either to be regarded as the most ancient homeland of met- to correlate similarities observed among disparate Bronze Age civilizations or to explain the status of the smelter allurgy, whence radiated both metallurgy and a civi- 2 as civilizing hero in ancient mythologies. The problem, lizing impetus. I argue, arises because previous scholars did not distin- During the second half of the 20th century, how- guish properly between two modes of copper production: ever, scholars identified many other homelands of crucible metallurgy and furnace smelting. According to metallurgy (the Balkans, the Iranian plateau, Spain, the localizationist theory, crucible metallurgy appears as a 3 spontaneous extension of the melting of native copper but South America, Thailand). Moreover, the autono- does not result in any substantial cultural change, whereas mous development of copper metallurgy discovered the general principles of a diffusionist theory would regard in the Aegean refuted the idea of local diffusion from the emergence of furnace metallurgy as a unique event neighboring areas (the Balkans or the Near East).4 For that spread rapidly and spurred on vast cultural changes these reasons, the diffusionist theory was replaced by (if diffusionists had ever actually understood the differ- ence between the two production methods). I propose a localizationist theory postulating a polyphyletic ori- instead a synthetic theory in which the spread of furnace gin of metallurgy. metallurgy—which was fundamentally different from The diffusionist theory associated metallurgy with crucible metallurgy and depended on complex technical the emergence of social complexity, which is consid- knowledge—from the southern Levant generated a wide ered the earliest stage in the development of civiliza- network linking Bronze Age societies. This has important implications for our understanding of the international tion. But this linkage is somewhat spurious: on the network of exchanges in technology, artifacts, and ideas Iberian peninsula, for example, metallurgy remained— during the Bronze Age.* for at least a millennium—a secondary activity (mainly producing ornaments) without substantial cultural and social influence.5 Likewise in Thailand, where introduction metallurgy focused on the production of utilitarian Scholarship in the first half of the 20th century held artifacts from the earliest stages of its development, that the emergence of copper metallurgy was a central the prehistoric society did not evolve toward a central- factor in the development of the earliest civilizations. ization/concentration of power.6 It was assumed that metal tools improved agriculture, Nor is the mastery of metallurgy directly related to which, in turn, led to a rapid population increase any civilizing advancement, as assumed by the diffu- * I want to thank Beno Rothenberg and Yehudit Gavish for manuscript. their help and for their fruitful discussions. The two anony- 1 Wenke 1991; Rothman 2004. mous reviewers and the editors are also acknowledged here 2 Wailes 1996, 5. for their pertinent observations, their objections, and their 3 Solheim 1968; Patterson 1971; Renfrew 1973; Bayard encouragement. They helped me to clarify many points that 1981; Jovanovic 1985; Glumac and Todd 1991. remained obscure in the submitted version. Thanks to Pat- 4 Renfrew 1972. rick Jean-Baptiste for helping to conceptualize and draw 5 Rovira 2002. the maps and to Sara Fine-Melzer for English editing of the 6 Bayard 1980; White and Pigott 1996. 497 American Journal of Archaeology 113 (2009) 497–519 498 NISSIM AMZALLAG [AJA 113 sionist theory. The Nahal Mishmar hoard, a unique crucible smelting collection of southern Levantine prestige artifacts Melting (liquefaction by heating) of native copper from the early fourth millennium B.C.E., for example, in crucibles for its casting was known before heat con- is characterized by a complex method of manufacture version of ore in the metal (smelting). But in the earli- (the lost-wax casting technique) and a surprising knowl- est stages, smelting was frequently done in crucibles. 7 edge and practical application of alloying processes. This encourages us to investigate the link between Nonetheless, this outstanding technical advancement melting and smelting. is unrelated to any social hierarchy and concentration of power in the southern Levant.8 Moreover, during From Melting to Smelting the third millennium B.C.E., no significant differences The smelting of copper in a crucible has been noted have been noted between techniques used by peoples in many areas of Asia, Europe, and South and Central from Central Asia (e.g., the Andronovo culture, with America between the fifth and the second millennia a low level of social organization) and by their neigh- B.C.E. (table 1). Such a distribution suggests that this bors living in the city-states on the Iranian plateau.9 mode of copper smelting appeared independently in In the localizationist scheme, metallurgy becomes at least seven areas: the Iranian plateau, the northern only one of a series of factors (e.g., human migration, Euphrates, the Balkans, Central Europe, the Iberian ecological change, introduction of new species, new peninsula, Thailand, and South America. Accordingly, discoveries) contributing to the emergence of highly the emergence of crucible smelting of copper should organized societies.10 Theoretically, this process is ex- not be considered an exceptional event. pected to generate a great diversity among the Bronze An analysis of information from the Anarak region Age societies, but their comparison points on unsus- (in the Iranian plateau) suggests a sequence for the pected similarities remain unexplained in a localiza- development of crucible smelting. In this area, mala- tionist context. chite was mined from the ninth millennium B.C.E. as a The diffusionist and localizationist theories integrate semiprecious stone and/or pigment, and native copper separate parts of the real situation. From an epistemo- was worked from the seventh millennium B.C.E., first logical view, this cognitive dissonance suggests that by cold hammering, a technique later replaced by heat these theories suffer a common flaw in their basic hammering and annealing. From the early fifth millen- statements. At the earliest stages of metallurgy, cru- nium B.C.E., copper (probably of native origin) was cibles and furnaces are used to produce copper, but melted and then cast in open/bivalve molds.11 At Tepe these processes are not delineated in the diffusionist Ghabristan, in addition to casting molds, crushed mala- or localizationist theories because scholars frequently chite has been found near heavily slagged crucibles, do not devote enough attention to the way the copper confirming that copper ore was indeed smelted.12 The is produced. Taking note of the differences between same temporal sequence of events (cold and heat ham- crucible and furnace processes of copper smelting, mering, annealing, and casting of native copper) pre- however, enables us to integrate these contradicting ceded the emergence of crucible smelting of copper views into a common framework, a synthetic theory that in the northern Euphrates area.13 It seems, therefore, has broad implications for understanding technology, that crucible smelting was discovered in the context of trade, and culture in the Bronze Age. extraction of native copper from its mineral gangue 7 Bar-Adon 1980. 10 Ruiz (1993, 56) even concluded that “explanations that 8 Gilead 1994; Fletcher 2008. This fi nding was considered give great causal weight to metallurgy in the dynamics of so abnormal within the classical perspective of development copper-bronze age change in Iberia need to be replaced by of civilizations that the artifacts were for a long time consid- alternatives that give metal its appropriate value. It is diffi cult ered to originate from Mesopotamia. The Nahal Mishmar to suppose that the formation and consolidation of an elite cave was therefore interpreted as a repository or hiding place could be based on a small-scale industry that was not essential for the trade between Mesopotamia and Egypt (Gates 1992). for subsistence and that depended on raw materials whose But recent analyses have attested the local origin of these ob- availability could not be easily restricted. The causes of social jects (Goren 2008). differentiation must be sought in the intensifi cation of other