Explanation

The nature of scientific explanation is probably the philosophical topic that has provoked the greatest number of interchanges between philosophers of science and archaeologists. The new archaeologists believed that archaeologists could make their explanations scientific by formulating them according to the pattern of Hempel’s deductive-nomological (D-N) model, which says that for an event to be explained scientifically, its description must be shown to be the conclusion of a deductive argument that has as its premises statements of initial conditions and at least one universal law. To take a rather old-fashioned but simple example of D-N explanation, one explains why inserting a penny in place of a blown fuse restores electrical current by citing the initial condition that the penny is made of copper, along with the universal law that all copper conducts electricity. Critics were quick to point out that Hempel never claimed that all scientific explanations fit this model; it is one model among many.

Many archaeologists, concerned about the difficulty of finding universal laws required for D-N explanations, looked for less stringent explanatory models. Philosophers of science, already heavily engaged in trying to understand the nature of scientific explanation, were drawn into the debates about the nature and possibility of scientific explanation in archaeology, and found that archaeology provided a good testing ground for some of their theories. Instead of merely claiming that the new archaeologists had failed to understand Hempel, some philosophers tried to make a more positive contribution by considering philosophically acceptable alternative models of explanation. They also distinguished between formal models that focused on the structure of explanation and its ontological and epistemological basis, and substantive models of explanation in a discipline, which look at appropriate generalizations and facts that can be mustered to explain archaeological phenomena. Hempel himself had already proposed an alternative inductive-statistical model that used statistical laws and allowed for probabilistic rather than deductive support of the conclusion/event to be explained, and had argued that this model was better suited to history and the social sciences.

Systems explanations of the archaeological record are popular with many archaeologists. The term ‘systems’ can refer to both formal and substantive features of explanations, and the failure to recognize this distinction caused some confusion in early discussions of the topics. Archaeologists proposed substantive systems explanations as alternatives to diffusionist explanations of cultural phenomena such as the spread of agriculture. Diffusionist models account for cultural change and innovation by looking for an external source of an idea or trait and then tracing its spread (diffusion) to other cultures. Systemic explanations focus on the following internal features: (1) multiple interacting causes that can reinforce or counterbalance one another, (2) negative feedback mechanisms that promote stability in the face of pressures for change, and (3) positive feedback mechanisms that amplify some factors, allowing for change within the system.

Systems explanations are sometimes called ‘neofunctionalist’. Whereas contemporary systems explanations bear some resemblance to Radcliffe-Brown’s structural-functional explanations, they differ in several important ways. (1) They treat the archaeological record on a par with historical documents as a source of evidence. (2) They provide a mechanism - positive feedback - to explain change. Although structural-functional explanations recognized negative feedback mechanisms, such as joking relationships that maintained social stability through release of tensions, they could not account for beneficial changes in societies and tended to regard any change as harmful. (3) Systems explanations emphasize the significance of humans’ physical environment as well as their social environment. A central tenet of the new archaeology is that societies cannot be considered in isolation from the physical surroundings that both shape and are shaped by them. By supplementing negative feedback mechanisms with positive feedback, by broadening the notion of a system to include the physical environment of a society, and by analyzing the archaeological record for appropriate evidence, systems explanations offer a theoretical framework for organizing and extending archaeological knowledge.

Some archaeologists argued that ‘the systems model’ of explanation was a superior alternative to the D-N model because it did not require laws. However, although the systems explanations that archaeologists use do not address the problem of laws in an explicit way, their reliance on laws is implicit in their requirement for causal mechanisms. The ‘systems model’ uses different kinds of substantive principles from those adopted by diffusionist models, but it simply is not a model of explanation in the same formal sense that the deductive-nomological model is - indeed, some systems explanations can be construed as examples of the D-N model or of the inductive-statistical model.

The point to be made about philosophical models of explanation is similar to that about confirmation. The problem of finding a set of general criteria that characterize genuine scientific explanation is a serious concern for philosophers, who have argued for alternatives that focus on various aspects of successful scientific examples of explanation. It is difficult to describe briefly the differences among the philosophical models, but a few distinguishing features can be mentioned. Some models, like Hempel’s, emphasize the logical connection between explanatory factors and that event which is to be explained. Other models emphasize the causal relationships that are important to sound explanations, and still others emphasize the ways in which good explanation serve to unify scientific knowledge. A common feature of these models is that all of them invoke laws. As with philosophical studies of confirmation, we can note that while philosophical analysis of scientific explanation can deepen our understanding of science and our appreciation of it, doing good scientific work does not depend on having the ‘right’ model.

A diverse group of critics of the ‘new archaeology’ characterize themselves as ‘postprocessual archaeologists’. Many of them are unhappy with the new archaeologists’ concern with laws and scientific explanation, as well as their preoccupation with environmental and ecological matters. Postprocessualists emphasize instead the symbolic function of prehistoric human behavior, and decry its neglect by new archaeologists. Moreover, they believe that it is possible to recapture this symbolic behavior from archaeological remains. Their work involves distinguishing function from style in cultural remains and studying carefully the stylistic features. They analyze and explain their results using ‘general principles’ that govern certain kinds of symbolic behavior, such as the pattern of distinguishing the ‘raw’ (natural) items from the ‘cooked’ (culturally transformed). By invoking general principles, however, their own interpretations also appeal implicitly to laws, even if these are employed in the service of so-called interpretation rather than explanation. From the viewpoint of the new archaeologists, the postprocessualists use principles that depend on questionable anthropological theorizing and lack empirical support. It is fair to say that the theories of both new archaeologists and post-processualists involve lawful explanation or interpretation, although the aims of the two groups and the content of the laws they invoke are very different (see Postprocessual Archaeology; Processual Archaeology).