In lieu of an abstract, here is a brief excerpt of the content:

4 Archaeological Things Languages of Observation Robert C. Dunnell Introduction There is nothing more fundamental to science than the categories used to perceive the external world. They are so fundamental that most people are completely unaware of the influence such structures exert on understanding. In some respects archaeologists, continually faced with unfamiliar objects and object parts, have proved more perspicacious than many “hard scientists” in these matters. “Typology” and “classification” are recurrent, if somewhat sporadic , concerns. The Lower Mississippi Valley (sensu this volume, hereafter LMV) has figured heavily in those considerations and has served as the testing ground for many new approaches throughout the twentieth century. Even so, the lack of familiar models in common sense (where these processes are necessarily cryptic) and the failure to recognize the role of theory or to reach a consensus on a particular theory have left us chasing our tails at the expense of the ever-diminishing archaeological record. Here I explicate the role of classification in science generally and then in archaeology per se. A review of the history of “typological” constructs in the region precedes a critique of the dominant approaches today. I conclude by suggesting remedies. Background Language of Observation Contrary to its popular use as a synonym for hypothesis, here I use the word “theory” to mean the assumptive structure necessary to construct a system of explanation (Dunnell 2002 [1971]; Guralnik and Friend 1968:1511). Thus, there is no “theory” that people first entered the New World through the Bering Strait land bridge, for example. Such a statement is a hypothesis. The theory comprises the assumptions that allow one to write the hypothesis and integrate it with others into a body of knowledge. There is no point to arguing about the 46 / Robert C. Dunnell words—theory is this or is that. Like any other word, “theory” is what it is de- fined as, nothing more or less. The point is that assumptions are made in writing the Bering Strait scenario, and those assumptions are the kind of constructions that are the target of the word “theory” here. So conceived, theory has two components, a language of observation— what is often called classification—and a language of explanation. The former constructs the units of observation, things or kinds of things. The latter articulates observations into causal statements, explanations. All systems of explanation , sciences, religions, even cultures (common senses) can be shown to contain these two elements. And, despite what some writers contend (Binford 1977), they are not independent, but rather must be completely integrated, an iterative process well described in the sciences (e.g., Lewontin 1974a:8–12). Science, as a particular kind of explanatory system, imposes special conditions on theory construction. First, to be subject to challenge, revision, and replacement , it must be explicit rather than cryptic. Secondly, it is distinguished by its reliance on an empirical epistemological standard, i.e., for a hypothesis to be regarded as an explanation it must have been empirically tested and not falsified. Following from this, and of particular concern to the language of observation , units in a scientific theory must be empirically sufficient (Lewontin 1974a:9), i.e., measurable in the empirical world, to generate empirically testable results. These units also have to be explicable, i.e., they must be the terms in the processes used in explanations. Hence the iterative nature of theory construction : trying to define units that can be measured and explained simultaneously . The failure of processualism as science can be laid to the lack of appreciation of this fundamental process, among other things. Explanations were typically cast in transactional terms (e.g., behavior) even though the archaeological record is distinguished from other human records by its nontransactional character. Consequently, none of the lofty efforts of processualists were empirically sufficient. Most of what passed as theory was, in point of fact, circumlocution designed to camouflage this insoluble problem—hence the focus on methods for “interpreting” or “inferring meaning,” processes that converted archaeological units into those of common sense (our culture) and ethnology, either or both of which supplied the explanations. In this process, however, empirical sufficiency is lost because “inferences” and “interpretations ” can never be tested in a scientific sense. In the atheoretical condition of the discipline, archaeologists have frequently taken object-naming to be an act of classification. Thus chapters in substantive contributions labeled“classification”do not discuss the creation of the units used...

Share