Inference and evidence in archaeology: a discussion of the conceptual problems

Citation: Sullivan, A. (1978) Inference and evidence in archaeology: a discussion of the conceptual problems. In Advances in Archaeological Method and Theory (Vol. 1), edited by M.B. Schiffer. New York. (RSS)
Internet Archive Scholar (search for fulltext): Inference and evidence in archaeology: a discussion of the conceptual problems
Tagged:

Summary

Sullivan pointed out the most common problem for making inference in archaeology is the uncertain linkage between the inference and evidence. The first part of the article is to review the scientific inference in archaeology, and then address some problems and seek solution. Sullivan discusses many archeologists concerned the procedural matters of archaeological inference, for example, ethnographic materials is one of the procedural matters; however he thinks archaeologists should focuses more on how to justify their conclusion.

The limitations for inference in archaeology results from the nature of archaeological data, inadequate connection between data and argument, and incoherent concept from material remains to data. For archaeological evidence, there are two types of concept. One is phenomenological concept, which viewed archaeological evidence as data, and tends to make single interpretation based on data. Another is problem-oriented approach, which also has some problems; for instance, if the required remains for generating particular data do not exist, then it is hard to answer the research question.

Sullivan thinks the links between data and past behavior depends on archaeological theories. Based on the theory, a model is needed to specify how information about the past transmitted to the present by material remains. The criteria for adequate model includes: 1. Specify what and how information from the past circumstances surrounding is likely to be represented. 2. Specify the conditions which affect the likelihood of archaeological remains. 3. Provide the method to distinguish the traces of production and trances of use of material remains.

Sullivan then proposed three different theories which relate to above criteria respectively. They are formation theory, recovery theory, and analytic theory. The formation theory focuses on the traces which portrays information of past. He thinks that trace production is context-dependent, which provides us a basis to examine the past information. The recovery theory focuses on fieldwork practice, including sampling, research design, and site-building process, which guild the direction of evidence. The analytic theory is to provide appropriate methods to separate archaeological materials in terms of their distinct contexts.