"In the mid-1950s a profoundly pessimistic discussion of "The Limitation of Inference in Archaeology" appeared in the (British) Archaeologocal Newsletter. The author, field archaeologist M. A. Smith, concluded that it is "a hopeless task" to attempt to attempt to move from one to the other "by argument"; between "the human activities we should like to know about" and the "visible results which survive from them" there is "logically no necessary link" (Smith 1955: 4). The standard of epistemic credibility to which Smith appeals is deductive certainly, and the focus throughout is on the vagaries of isolated inferences from fragmentary material "finds". Just a few years later Stephen Toulmin published Uses of Argument (1958), a philosophical critique of the preoccupation, among logicians, with just the kind of idealised (deductive) argument Smith had invoked. His central objection: the formalism that had bewitched logicians led them to systematically read out of account a wide range of warranting conditions that are, in fact, crucial to justificatory argument. Although Smith's conclusions were extreme, she made explicit premises that continue to structure debate about the credibility of evidential reasoning in archaeology. In this paper I explore the road not taken, outlining a model of evidential reasoning in archaeology that focuses on the kinds of scaffolding, and strategies for exploiting epistemic independence between multiple lines of evidence that, in practice, put archaeologists in a much stronger epistemic position than Smith, and many since, have been prepared to recognise".

For further information contact: Duncan.Wright@anu.edu.au or Jack.Fenner@anu.edu.au