British Academy: The UK's National Academy for the Humanities and Social Sciences
Enquiry, Evidence and Facts: An Interdisciplinary Conference
Critical Distance: Stabilizing Evidential Claims in Archaeology
Professor Alison Wylie
Department of Philosophy, University of Washington
345 Savery Hall, Box 353350, Seattle WA 98195 U.S.
An abstract presented to the conference
‘Enquiry, Evidence and Facts: An Interdisciplinary conference’
at the British Academy, London, on 14 December 2007
Biography
Alison Wylie (Professor of Philosophy, University of Washington) is a philosopher of social science who works on epistemological questions raised by archaeological practice and by feminist research in the social sciences. In Thinking from Things: Essays in the Philosophy of Archaeology (2002) she develops an analysis of the strategies of triangulation by which archaeologists stabilize evidential claims. She is centrally interested here, and in analysis of feminist research practice, in reconceptualizing ideals of objectivity to take account of the ineliminable role played by contextual values in the research process. Her essays appear in the Sage Handbook of Feminist Research (2007), Episteme (2006), Embedding Ethics (2005), and Science and other Cultures (2003). She is editor of Epistemic Diversity and Dissent (Episteme 2006); co-editor of Value-Free Science? (2007), Doing Archaeology as a Feminist (JAMT 2007), and Feminist Science Studies (Hypatia, 2004). She is currently writing a book on feminist standpoint theory.
Abstract
The vagaries of evidential reasoning in archaeology are notorious: the material traces that comprise the archaeological record are fragmentary and profoundly enigmatic, and the inferential gap that archaeologists must cross to constitute them evidence of the cultural past is a perennial source of epistemic anxiety. And yet, archaeologists know a great deal about the cultural past. More to the point, they routinely find that, however radically a construct their evidence may be, it "resists appropriation"; it has a striking capacity to disrupt settled assumptions, redirecting inquiry and expanding interpretive horizons in directions archaeologists often had not anticipated. It is this capacity for constraining inference and interpretation that I am concerned to understand. In this paper I outline a model of evidential reasoning based on archaeological practice that integrates insights drawn from philosophical theories of confirmation and hypothesis testing that have much wider reach than this particular social/historical discipline.
First, I situate and further specify the problem about evidence that arises in archaeology, and that has precipitated crisis debates (mainly within North American anthropological archaeology) every two or three decades since the early 20th century (Wylie 2002). What drives these debates is, at bottom, a concern about the proxy status of archaeological evidence; there is no alternative but to depend on evidence that is, as one commentator puts it, “remote from and uncertainly coupled to the systems [archaeologists] seek to study” (Chippindale 2002: 606). Archaeological data radically underdetermine claims about their origins and cultural significance—they are, of necessity, so heavily theory-laden, and so interdependent—that vicious circularity seems inescapable. The worry is that archaeologists find all and only what their background assumptions predetermine; archaeological interpretation is reduced to the projection of contemporary preoccupations onto the past, vulnerable to the free play of contextual values. One reaction to the subjectivism and relativism that some feel is entailed by these concerns has been to declare archaeology a science, and limit the scope of inquiry to types of data and forms of inference that, it is hoped, will set archaeological claims on a firm foundation. But what counts as properly scientific is either so narrowly defined it drains the enterprise of anthropological interest, or admits forms of evidential reasoning that are vulnerable to just the kinds of insecurity they were intended to escape. The extreme positions articulated in the context of crisis debates are clearly untenable epistemically, but the concerns that give rise to them establish that there are no forms of archaeological data, or evidence based on these data, that can be treated as foundational in the epistemic sense that they require no further warrant; evidential claims, in archaeology as elsewhere, are products of inquiry rather than external impositions on it, and are therefore always defeasible.
On closer examination it is clear that the polarized positions articulated in the context of crisis debates do not exhaust the epistemic and methodological options available to archaeologists. Indeed, they systematically obscure a range of strategies for building and adjudicating evidential claims that archaeologists routinely deploy in practice, the merits of which they assess in more localized (problem and practice-specific) methodological debates that are strikingly free of the polemics generated by ideals of certainty and their relativist antithesis. These are strategies deliberately designed to elicit surprises and expose interpretive error in the context of non-experimental research, aligned with a disciplined responsiveness to the results of inquiry so designed; taken together, they make possible a nuanced, comparative appraisal of degrees of credibility and uncertainty in evidential claims.
To account for the resistant, corrective capacity of archaeological data-cum-evidence that sustains these mediating options, I propose a model of evidential reasoning that emphasizes the bootstrapping and triangulating strategies by which archaeologists play multiple lines of evidence off against one another. The point of departure is, to paraphrase Glymour (1980), the commonplace that evidential reasoning involves (at least) three components: empirical input in various forms; background knowledge (auxiliary hypotheses, in the broadest sense) that make it possible to describe, analyze, appropriate, represent empirical input as evidence; the claims (hypotheses, theories) on which this interpretively constituted input bears as evidence. Crucially, what counts as background knowledge or auxiliaries, an evidential claim, and a test hypothesis or theory based on (or evaluated in light of) this evidence is functionally defined and contingent on context of practice; it is not determined by intrinsic features of the content or epistemic status of the claims in question. I argue that it is various forms of critical distance between these constituents of evidential reasoning that archaeologists exploit in assessing the robustness of empirical claims and their evidential import. On this account, the resources for making these appraisals reside in the following features of evidential reasoning.
- Vertical independence. The theories that laden data, constituting them as evidence (background knowledge, auxiliaries), are not necessarily the same as (or components of) the theories on which they are brought to bear as evidence. The model of bootstrap confirmation articulated by Glymour (1980) can be extended beyond intra-theoretical “deductions from the phenomena”; I reframe it in terms of a principle of epistemic independence between theory and evidence that empirical inquiry of many forms (experimental and observational, quantitative and qualitative), is designed to exploit. Archaeologists make effective use of vertical independence to build and assess claims about the cultural past pitched at a number of different levels of specificity, ranging from reconstructive claims about particular past events and conditions, to expansive explanatory models of cultural process.
- Evidential robustness (assessed with respect to individual lines of evidence). Judgments of the reliability of, and weight to be accorded to, specific evidential claims depend on a jointly empirical/theoretical and pragmatic appraisal of the credibility of the sources on which archaeologists rely for the background knowledge (auxiliaries) they need to interpret archaeological data as evidence. This takes the form, in the first instance, of an appraisal of the credibility of these sources within their contexts of origin, and then of their reliability in extension beyond these contexts.
- Horizontal independence (robustness of a body of evidence). In addition, evidential reasoning typically depends on strategies of triangulation that exploit various forms and degrees of independence between lines of evidence: causal independence in the source and transmission of the signal (to use language from Hacking 1981, 1983; Kosso 1992), conceptual or methodological independence and, as a proxy for these others, disciplinary independence (Wylie 2002). This is one basis on which the robustness of evidential claims is assessed, where this turns on considerations of the variety of evidence (Wimsatt 1981); the principle is that cables are stronger than chains, given a capacity for mutual constraint between constituent strands of evidence.
Although evidential claims are purpose and problem-specific, and judgments about their robustness and relevance always open to reassessment, when well constructed (along these dimensions) they function as a contingent stopping point in warranting arguments; as such, they constitute an evolving foundation for adjudicating knowledge claims.
I expect that this model of evidential reasoning is generalizable; I make use of it (elsewhere) as the basis for a pragmatic reframing of ideals of objectivity in the context of debate about standpoint theory (Wylie 2003). For present purposes, however, my aim is to illustrate its chief components in archaeological terms. I focus, in particular, on the process by which radiocarbon dating has been continuously corrected and calibrated, its applications rendered both more precise and more narrowly circumscribed by bringing to bear multiple lines of evidence; ironically these often derive from precisely the dating techniques that radiocarbon dating was meant to displace. This is an example that brings into focus nuanced forms of evidential reasoning in archaeology that operate in the middle ground between (unattainable) foundationalist ideals of epistemic security, and their antithesis, the forms of wholesale methodological skepticism and relativism to which archaeologists are inclined when, in the context of crisis debates, instances of recognized error or insecurity are generalized to all forms of archaeologically grounded inference. My further objective is to draw out a meta-philosophical implication of this analysis: that in adjudicating questions the credibility of specific research tools, archaeologists rely appraisals of the disciplines from which they import background knowledge and evidence stabilizing technologies, and the social histories of trade in these resources between fields. This poses a challenge to standard philosophical analyses of evidential claims, and it suggests that the tacit knowledge that mediates evidential reasoning includes not only technical know-how but also a form of internal disciplinary sociology and history that is continuous with, and might be improved by the systematic integration of, philosophical and social-historical science studies.
References cited
Chippindale, C. (2002) Capta and Data. American Antiquity 65(4):605-612.
Glymour, C. (1980) Theory and Evidence. Princeton, NJ: Princeton University Press.
Hacking, I. (1981) “Do We See Through a Microscope?” Pacific Philosophical Quarterly 18:305-322.
Kosso, P. (1992) “Observation of the Past.” History and Theory 31(1):21-36.
Wimsatt, W. C. (1981) “Robustness, Reliability, and Overdetermination.” In Scientific Inquiry and the Social Sciences. M.B. Brewer and B.E. Collins, eds. Pp. 124-163. San Fransisco: Josey-Bass.
Wylie, A. (2000) “Rethinking Unity as a Working Hypothesis for Philosophy of Science: How Archaeologists Exploit the Disunity of Science.” Perspectives on Science 7(3):293-317.
---- (2002) Thinking from Things: Essays in the Philosophy of Archaeology. Berkeley, California: University of California Press.
---- (2003) “Why Standpoint Theory Matters: Feminist Standpoint Theory.” In Philosophical Explorations of Science, Technology, and Diversity. R. Figueroa and S. Harding, eds. Pp. 26-48. New York: Routledge.