Costs and benefits in technological decision making under variable conditions: examples from the late Pleistocene in southern Africa

Citation: Alex Mackay, Ben Marwick Costs and benefits in technological decision making under variable conditions: examples from the late Pleistocene in southern Africa.
Internet Archive Scholar (search for fulltext): Costs and benefits in technological decision making under variable conditions: examples from the late Pleistocene in southern Africa
Tagged:

Summary

Context

Mackay and Marwick explore the changes in technological strategies as a response to risk. Using a model that examines the relationship between resource returns as a function of time invested in technology. More specifically they examine changes in material of flaked artifacts, complexity of flaked artifacts, and rates of discard in three rockshelter sites in South Africa. Using global climate temperatures as a proxy for environmental resources, colder temperatures being associated with poorer environments and fewer resources while warmer environments are associated with more resources. The three sites examined through their model of technological time-costs: Diepkloof Rock Shelter, Klein Kliphuis Rock Shelter, and Elands Bay Cave, all lie within the current southern African Winter Rainfall Zone and were occupied during the Late Pleistocene. The Western Cape of Africa where these sites lie went through two cold glacial periods (MIS 2 and 4) and two warmer periods (MIS 3 and 5).

Methods and Materials

Mackay and Marwick use a model of technological time-costs. Technological time-cost refers to the amount of time required to procure, produce, and maintain stone artifact technologies. The model examines the relationship between total time outlay and the resources gathered within a patch. Incorporated in the model is the idea of multiple encounters in a patch and the capture rate, together these are called the success rate. This model is a curve that follows a von Bertalanffy growth function. In an archaeological context of hunter-gatherers, this function represents how return rate from a patch will decrease as the patch is exploited since depletion of resources continues to occur over time.

Results

This model is applied to the assemblages at the three rockshelters mentioned above. The expectation is that during colder periods hunter-gatherers would be in higher risk situations and would therefore be associated with more costly technologies seen through the manufacture of complex implements and use of high quality material. Oppositely, during warmer periods, assemblages would be characterized by lower quality material and less complex implements. At Diepkloof Rock Shelter (DRS), from 70-74 ka bifaces and bifacial points are present while less complex forms such as backed artifacts occur from 70 ka through 60 ka and are subsequently replaced with unifacial points. This pattern is also observed at Klein Kliphuis Rock Shelter (KKH). At DRS, quartz and quartzite predominate the earliest unit with few occurrences of silcrete. During 70-74 ka unit, silcrete has replaced quartz while quartzite is still fairly dominant. From 65-70 ka quartz becomes dominant again with quartzite also being present. Expectations of high cost technologies are only met once during the late MIS 4. During MIS 5, which is seen as the period with the lowest overall risk, is represented by low cost technologies. These observed periods are the only ones that conformed to the predictions of the model.

Theoretical and Practical Relevance

Connection

Torrence (1989) states that technology is a solution, in the case of these three rockshelter sites, the material type and technological complexity are the variables that are solutions to the problem of environmental change. However, the observed cases do not follow the expected costs. Torrence’s idea that optimal technology may not always be chosen may be a factor in why these cases did not better follow expected costs. Mackay and Marwick’s model assumes that the hunter-gatherers will choose the optimal technology in a given environment, however, as Torrence suggests this may not be the case.

Judgment

Although the model did not perfectly predict the technological changes at these sites, the authors examined the shortcomings of the model and ways that it should be further developed and refined. This positive outlook on the models limited ability for prediction is inspiring to others particularly when faced with a similar situation to this.