Giorgio Buccellati, Critique of Archaeological Reason
Excerpts and Synopses

Salmon Merrilee


Excerpts from 1982 "Philosophy and Archaeology"


Chapter: 2: Laws in Archaeology page: 8-30 Scientific laws when the connections are persistent and significant they or the statements that describe them are called scientific laws.
Logical generalizations Divided in: universal generalizations - i.e. all kinship relations that are important in a society are expressible in the language of that society. The numerical proportion is either 0% or 100%; and statistical generalizations - sometimes called probabilistic generalizations that state a certain proportion of the members of a class included or excluded from another class.
Statement A statement maybe general in the sense that is not restricted to particular times, places or individuals and still fail to be either a statistical or a universal generalization. The law statement must have: logical generality, generality with respect to content, and must be true.
Laws in Archaeology the laws that seem to be most properly characterized as laws of archaeology are the regularities or empirical generalizations that relate various items of material culture with pattern of human behavior.

Chapter: 3: Confirmation in Archaeology, page: 31-56 Confirmation is a relation between the hypothesis statement and the statement that expresses the observations or results of experiments.
Deductive reasoning is important in archaeology, as it must be in any science that employs statistics. Nevertheless inspite of the support given by the mathematics and deductive reasoning the problem of confirming statements about the past on the basis of contemporary archaeological evidence is basically a problem of inductive knowledge.
Inductive logic is concerned with the support of some statements can provide for others when the statement neither deductively entail nor contradict one another. Inductive logic tries to analyze and appraise arguments in which the premises lend some support to the conclusions. Good inductive argument can be distinguished from their deductive counterparts by three factors: a) valitidy of an deductive argument; b) Ampliative inductive arguments and c) deductive arguments known as "truth per serving".
The Hypothetico-Deductive Method according to the standard accounts it follows three phases: a) formulate the hypothesis (H); b) deduce the some prediction (P), which is amenable to observation and c) perform the observation to see whether the stated prediction is true or false (p.34).
Significance
is the matter of substance not form and knowledge of the subject matter is required to separate the significant from the insignificant.
Prior probability
of a hypothesis is simply the degree of plausibility that it has prior to the testing situation being considered. Probably the most pragmatic consideration for an archaeologist who is trying to assign a prior probability to a hypothesis is the status of available alternatives.
The Bayesian model
is the determining an unknown probability when certain other probabilities are known.

Chapter: 4: page 57-83 Analogy and Functional Ascription An interesting chapter inferring a topic not taken as seriously into account: the archaeological context and to what extent we ascribe function to various types of evidence at their immediate placement. Salmon maintains a plausible balance between the use of analogy as an indispensable source that highlights potentially the function and the crucial importance of the complexities of the archaeological context which in itself serves as a unique resonator of the function. Highly critical on the sporadic or evasive parallels driven analogical between the ethnographic and archaeological data. Particular priority is given to: the context association, systematic process of data collection, multidimensional analysis, careful comparison with analogical cases (synchronic and ethnographic), designation of prior probabilities, quantification; all these parameters serve as the conceptual spine of the chapter which can assist in the testing and confirming of an hypothesis and reach to solid conclusions, if not deductive generalizations.
Chapter: 5: page 84-112 Functional Explanation Archaeology is about why! The chapter infers the explanation in archaeology employing a wide range of philosophical and theoretical concepts. Answering the questions why and how is crucial and providing explanations remains a pivotal matter while exploring a phenomenon. Salmon borrows from philosophy two models: 1) Deductive-Nomological (D-N) defined as: the relationship between explanatory premises and event to be explained as one of deductive entailment. The model is developed in three conceptual stages: law, initial condition and phenomenon to be explained. 2) Inductive-Statistical (I-S) - defined as a model which allows explanations whose laws are statistical generalizations instead of the universal laws required by the D-N model. In the I-S the explanatory premises are supposed to render the explanadum phenomenon highly probable (p. 98-99). Salmon treats systematically the benefits and limitations of both models neglecting however their cohesive benefits on the explanation of function. On the consideration of such limitations alternatively she proposes as more effective the Statistical-Relevance model in its crudest form says that an event is explained when all the factors that are statistically relevant to its occurrence or nonoccurrence have been assembled, and the appropriate probability value has been determined for its occurrence in light of those factors, (p.109).
Chapter: 6: page 113-138 Structure of Archaeological Explanation The chapter extends on the structure of explanation focusing on particular archaeological case studies. Salmon addresses two tools of explanation respectively the probabilistic causes leaded by statistical operations that measure the degree of significance yielded from a dataset and the casual causes where the statistical regularities are indirect through a common cause (p. 136). Again the Statistical-Relevance model is used as a comprehensive tool to capture the full set of data (casual relations). Repeatedly Salmon emphasizes the D-N and I-S as models heavily depended upon logical relations rather than causal relations to provide relevant connections between explanans and explananda.
Chapter: 7: page 140-176 Theory Building in Archaeology The chapter mostly offers conclusive remarks. Certainly situates the archaeological theory under the umbrella of the social science defining theory as a set of interrelated, rather high - level principles or law that can provide an explanatory framework to accommodate a broad range of phenomena . Salmon synthesizes the theoretical building in archaeology within a hierarchical framework that includes the observational patterns, explanation and the theoretical concepts. Part of the picture also is the high-level theory, which considers large amount of archaeological data subjecting the materials to detailed study and thoughtful consideration developing a set of empirical generalizations and then trying to construct laws and higher principles. Salmon argues for a the theoretical scheme developed within a hierarchical framework that includes the observational patterns, explanation and the theoretical concepts. Part of the picture also is the high-level theory defined as large amount of archaeological data subjecting the materials to detailed study and thoughtful consideration developing a set of empirical generalizations and then trying to construct laws and higher principles. Salmon recognizes the contribution of Dunnell on the foundation of a pure and hermeneutic archaeological theory Dunnell 1971. However his approach on the systematics is considered as highly vague, lacking clarity on terminology. The conceptual framework of the classification scheme is heavily criticized. Salmon sees the connection between the artifacts, classification and phenomena developed through an incomplete framework which creates rather a parody of theory that relies on a strategy of guessing imposing to the data an intuitive behavior. The employment of the mathematical models is particularly treated. The description and explanation of the equations however remains rather challenging. Salmon goes in favor to the processual theoretical model pioneered by Binford who develops an integral cohesive and functional approach based on the general assumptions, common-sense, hypotheses, induction and theories. Even at this case however Salmon claims to emphasize a careful execution of this scheme in the case of archaeological record.