Until recently, the different stages of producing, using and distributing stone tools were studied in relative isolation, rather than as an integrated system. In the early 1970s, with the advent of sourcing technology, particular attention was paid to the study of distribution networks. One of the earliest and most influential of these studies was Renfrew's analysis of obsidian distribution in the Aegean, using trace elements to correlate obsidian sources with their areas of distribution (Renfrew et al. 1966; Renfrew 1969; 1973; 1977a; Dixon et al. 1971). By comparing obsidian consumption rates with distance from the relevant obsidian sources, Renfrew identified 'supply' and 'contact' zones, areas that were considered to have differential access to the obsidian sources (Renfrew 1969; Dixon et al. 1971, 43).
Based on these studies, Renfrew proposed his Law of Monotonic Decrement, which determined that sites would show a linear fall-off pattern in consumption rates where people had practised simple, direct-access procurement strategies. Deviations from this fall-off pattern were considered to be the result of different exchange systems operating (Renfrew 1969, 157; 1977b). Exchange systems with distinctive fall-off patterns included down-the-line exchange, reciprocity, central place redistribution, market exchange and middleman or freelance traders (Renfrew 1975, 42-3).
Since Renfrew's initial models were proposed, further work on distribution networks using petrographic or chemical characterisation of raw materials has been completed globally (Bettinger 1982; Binns and McBryde 1972; Bradley and Edmonds 1993; Cobean et al. 1968; Findlow 1982; Harbottle 1982; McBryde 1978; 1984; McBryde and Watchman 1976; Nocete et al. 2005; Rath and Torrence 2003; Spence 1981; Spence et al. 1984; Swete Kelly 2001; Tykot 2003; Weigand et al. 1977). However, many of these more recent studies have identified problems with the results gained from applying a simple univariate comparison of consumption against distance (Earle 1982, 7; Ericson 1977, 110), recognising that other factors may have influenced distribution. These include archaeological factors such as site formation processes, the contemporary nature of sites and different collection strategies. For example, during the collection of stone artefacts, sampling strategies may have been applied, for instance focusing upon particular raw materials or finished tools. These sampling strategies would affect the overall composition of the stone tool assemblage. Subsequent analysis of consumption rates of particular raw materials or tool types within the assemblage would therefore be unavoidably biased.
Apart from archaeological considerations, other factors may have influenced distribution. These include geographical factors such as difficulty of terrain (Earle and Ericson 1977, 6; Findlow 1982, 71), transport methods (Ericson 1982, 131), territorial boundaries (Bettinger 1982; McBryde 1978) and local availability of other raw materials (Ericson 1977, 118; Wiant and Hassen 1985, 105; Blanton 1985, 116). Differences in site hierarchy, population and function will also affect consumption rates at sites, thus influencing the appearance of how past distribution networks may have operated (Bradley and Edmonds 1993, 7; Sidrys 1977, 96; Torrence 1986, 16-17). Context of the distribution – that is, whether the goods are considered to have elite or ritual significance or are purely utilitarian – may also affect how goods are distributed (Bradley and Edmonds 1993; Hodder 1982).
Social factors have also been identified as an important influence on how distribution networks are structured, with some researchers moving away from examining distribution purely as an economic process (Earle 1982, 2). For example, McBryde's (McBryde and Harrison 1981; McBryde 1978; 1984; McBryde and Watchman 1976) study of the distribution of greenstone axes from quarry sites in Victoria, Australia, used ethnographic, archaeological and petrographic information to discover that axes were widely distributed among groups that shared a common religious and language system, despite the local availability of raw material of equal quality. From this distribution, McBryde inferred that the reaffirming of social ties through the exchange of axes was considered more important than the economic transaction of exchange. The social importance of distribution was also identified in the exchange of axes in Neolithic Britain (Bradley and Edmonds 1993) and Papua New Guinea (Burton 1989; Chappell 1966); raw material exchange in Australia (Gould 1978) and in the distribution of obsidian in the Pacific (Rath and Torrence 2003; Pavlides 1988; Swete Kelly 2001). In most of these examples, raw material and stone tools were distributed into areas where high-quality material was already available, indicating that the distribution was operating for reasons other than economics.
Ideally, any approach used to understand distribution would investigate both economical and social/political/ideological factors. However, as a starting point, I would argue that a model focusing upon economic factors should be used. By understanding the optimal procurement strategies for stone tool users, an economic baseline is created. Variations from this baseline may indicate whether other factors, including social or political reasons such as territorial boundaries, gift exchange, or perceptions of quality or value, were contributing to the distribution patterns. However, without the initial understanding of the optimal procurement strategies for stone tools based on economics, the social, political or archaeological factors behind the visible patterning may not be visible to the observer.
Based on this reasoning, an approach based on economic factors was selected to identify the different choices made in the production and distribution of stone tools at Huizui. There were several reasons why this approach was decided upon. Firstly, there has been little research completed on how production-distribution systems of utilitarian goods operated within the regional Erlitou economy. Without this basic understanding of how the different utilitarian goods production-distribution systems operated, it is difficult to identify if, and to what extent, political and social factors influenced their structure. The problem of identifying political and social influences is compounded by the issue that these types of influences are more likely to be unique cultural factors. Economic factors, however, tend to be more universal among human societies and therefore can be more easily distinguished through the use of models that examine best practice or economic ideals.
Economic factors are also more likely to be visible within the archaeological record. While the unique nature of all human societies presents myriad possibilities of how production-distribution systems may have operated, economic responses are limited and may be more accurately deduced. Previous stone tool studies using an economic approach include Torrence (1986), Findlow and Bolognese (1984) and Gould (1978). These studies examined what was the most economically ideal situation with regard to production-distribution systems and then compared this ideal to archaeological assemblages. Variations from this ideal could then be interpreted by considering other factors such as social, ideological or political. Therefore, while the initial approach towards analysing the stone tool production-distribution systems is economic, the interpretation of the results allows for an integration of all potential contributing factors.
One of the most significant contributions to our understanding of the relationship between production sites and distribution networks was Torrence's (1984) study of the Melos obsidian sources in the Aegean. Torrence's research was one of the first studies to provide a framework that made it possible to integrate the analysis of procurement and production sites with studies of distribution systems. Torrence proposed that all stages of a production-distribution system were interrelated; by studying one part, invaluable information could be gained as to how other parts of the system may have operated. The crucial parameter used to link the different parts of production-distribution systems was identified as the overall level of efficiency of the system. As the distribution network became more complex, Torrence proposed that methods of production and raw material procurement had to become more efficient to make these networks economically viable, a situation which has since been recognised by other researchers, especially in relation to the production of utilitarian tools (Costin 1991, 11). Measures of efficiency were identified through evidence of attempts to reduce costs, such as the use of standardised or specialised technology and method. A significant result of Torrence's model was the ability to predict or test possible distribution mechanisms through the analysis of quarry or production sites. This middle range theory therefore has the possibility of providing a link between the behaviours seen at quarry and production sites with different types of distribution networks, and vice versa.
While Torrence's model has since been criticised for applying contemporary capitalist concepts of efficiency and cost-benefit ratios to prehistoric production-distribution systems (Bradley and Edmonds 1993, 9), it is still an important rationalisation of the relationship between production and distribution. Through the identification of optimal production-distribution systems by analysing cost-reducing strategies, Torrence has provided an economic baseline, removed from culturally specific issues, against which archaeological production-distribution systems can be compared. Variations between archaeological systems and the economically optimal systems can then be used to help understand what other influences may have contributed to how the archaeological production-distribution systems operated, including social and political factors.
Efficiency is also a flexible measure that can be used to understand all aspects of the systems, including raw material procurement, production, consumption and distribution. Finally, efficiency in some parts of the system may also allow for predictive modelling of how other parts of the production-distribution system operated.
© Internet Archaeology/Author(s) URL: http://intarch.ac.uk/journal/issue26/15/3.html
Last updated: Wed Jul 1 2009