Weigh our specific evidence to get closer to the truth
Kratochwil 2008 (Friedrich Kratochwil, professor of international relations at European University Institute, 2008(Friedrich, “The Puzzles of Politics,” pg. 200-213)
In what follows, I claim that the shift in focus from “demonstration” to science as practice provides strong prima facie reasons to choose pragmatic rather than traditional epistemological criteria in social analysis.
Irrespective of its various forms, the epistemological project includes an argument that all warranted knowledge has to satisfy certain field- independent criteria that are specified by philosophy (a “theory of know- ledge”). The real issue of how our concepts and the world relate to each other, and on which non-idiosyncratic grounds we are justified to hold on to our beliefs about the world, is “answered” by two metaphors. The first is that of an inconvertible ground, be it the nature of things, certain intuitions (Des- cartes’ “clear and distinct ideas”) or methods and inferences; the second is that of a “mirror” that shows what is the case.¶ There is no need to rehearse the arguments demonstrating that these under- lying beliefs and metaphors could not sustain the weight placed upon them. A “method” à la Descartes could not make good on its claims, as it depended ultimately on the guarantee of God that concepts and things in the outer world match. On the other hand, the empiricist belief in direct observation forgot that “facts” which become “data” are – as the term suggests – “made”. They are based on the judgements of the observer using cultural criteria, even if they appear to be based on direct perception, as is the case with colours.4¶ Besides, there had always been a sneaking suspicion that the epistemological ideal of certainty and rigour did not quite fit the social world, an objection voiced first by humanists such as Vico, and later rehearsed in the continuing controversies about erklären and verstehen (Weber 1991; for a more recent treatment see Hollis 1994). In short, both the constitutive nature of our concepts, and the value interest in which they are embedded, raise peculiar issues of meaning and contestation that are quite different from those of description. As Vico (1947) suggested, we “understand” the social world because we have “made it”, a point raised again by Searle concerning both the crucial role played by ascriptions of meaning (x counts for y) in the social world and the distinction between institutional “facts” from “brute” or natural facts (Searle 1995). Similarly, since values are constitutive for our “interests”, the concepts we use always portray an action from a certain point of view; this involves appraisals and prevents us from accepting allegedly “neutral” descriptions that would be meaningless. Thus, when we say that someone “abandoned” another person and hence communicate a (contestable) appraisal, we want to call attention to certain important moral implica- tions of an act. Attempting to eliminate the value-tinge in the description and insisting that everything has to be cast in neutral, “objective”, observational language – such as “he opened the door and went through it” – would indeed make the statement “pointless”, even if it is (trivially) “true” (for a powerful statement of this point, see Connolly 1983).¶ The most devastating attack on the epistemological project, however, came from the history of science itself. It not only corrected the naive view of knowledge generation as mere accumulation of data, but it also cast increasing doubt on the viability of various field-independent “demarcation criteria”. This was, for the most part, derived from the old Humean argument that only sentences with empirical content were “meaningful”, while value statements had to be taken either as statements about individual preferences or as meaningless, since de gustibus non est disputandum. As the later dis- cussion in the Vienna circle showed, this distinction was utterly unhelpful (Popper 1965: ch. 2). It did not solve the problem of induction, and failed to acknowledge that not all meaningful theoretical sentences must correspond with natural facts.¶ Karl Popper’s ingenious solution of making “refutability” the logical cri- terion and interpreting empirical “tests” as a special mode of deduction (rather than as a way of increasing supporting evidence) seemed to respond to this epistemological quandary for a while. An “historical reconstruction” of science as a progressive development thus seemed possible, as did the specification of a pragmatic criterion for conducting research.¶ Yet again, studies in the history of science undermined both hopes. The different stages in Popper’s own intellectual development are, in fact, rather telling. He started out with a version of conjectures and refutations that was based on the notion of a more or less self-correcting demonstration. Con- fronted with the findings that scientists did not use the refutation criterion in their research, he emphasised then the role of the scientific community on which the task of “refutation” devolved. Since the individual scientist might not be ready to bite the bullet and admit that she or he might have been wrong, colleagues had to keep him or her honest. Finally, towards the end of his life, Popper began to rely less and less on the stock of knowledge or on the scientists’ shared theoretical understandings – simply devalued as the “myth of the framework” – and emphasised instead the processes of communica- tion and of “translation” among different schools of thought within a scien- tific community (Popper 1994). He still argued that these processes follow the pattern of “conjecture and refutation”, but the model was clearly no longer that of logic or of scientific demonstration, but one that he derived from his social theory – from his advocacy of an “open society” (Popper 1966). Thus a near total reversal of the ideal of knowledge had occurred. While formerly everything was measured in terms of the epistemological ideal derived from logic and physics, “knowledge” was now the result of deliberation and of certain procedural notions for assessing competing knowledge claims. Politics and law, rather than physics, now provided the template.¶ Thus the history of science has gradually moved away from the epistemo- logical ideal to focus increasingly on the actual practices of various scientific communities engaged in knowledge production, particularly on how they handle problems of scientific disagreement.5 This reorientation implied a move away from field-independent criteria and from the demonstrative ideal to one in which “arguments” and the “weight” of evidence had to be appraised. This, in turn, not only generated a bourgeoning field of “science studies” and their “social” epistemologies (see Fuller 1991), but also suggested more generally that the traditional understandings of knowledge production based on the model of “theory” were in need of revision.¶ If the history of science therefore provides strong reasons for a pragmatic turn, as the discussion above illustrates, what remains to be shown is how this turn relates to the historical, linguistic and constructivist turns that preceded it. To start with, from the above it should be clear that, in the social world, we are not dealing with natural kinds that exist and are awaiting, so to speak, prepackaged, their placement in the appropriate box. The objects we investi- gate are rather conceptual creations and they are intrinsically linked to the language through which the social world is constituted. Here “constructivists”, particularly those influenced by Wittgenstein and language philosophy, easily link up with “pragmatists” such as Rorty, who emphasises the product- ive and pragmatic role of “vocabularies” rather than conceiving of language as a “mirror of nature” (Rorty 1979).¶ Furthermore, precisely because social facts are not natural, but have to be reproduced through the actions of agents, any attempt to treat them like “brute” facts becomes doubly problematic. For one, even “natural” facts are not simply “there”; they are interpretations based on our theories. Secondly, different from the observation of natural facts, in which perceptions address a “thing” through a conceptually mediated form, social reality is entirely “arti- ficial” in the sense that it is dependent on the beliefs and practices of the actors themselves. This reproductive process, directed by norms, always engenders change either interstitially, when change is small-scale or adaptive – or more dramatically, when it becomes “transformative” – for instance when it produces a new system configuration, as after the advent of national- ism (Lapid and Kratochwil 1995) or after the demise of the Soviet Union (Koslowski and Kratochwil 1994). Consequently, any examination of the social world has to become in a way “historical” even if some “structuralist” theories attempt to minimise this dimension. [. . .]¶ Therefore a pragmatic approach to social science and IR seems both necessary and promising. ¶ On the one hand, it is substantiated by the failure of the epistemological project that has long dominated the field. On the other, it offers a different positive heuristics that challenges IR’s traditional disciplin- ary boundaries and methodological assumptions. Interest in pragmatism therefore does not seem to be just a passing fad – even if such an interpre- tation cannot entirely be discounted, given the incentives of academia to find, just like advertising agencies, “new and improved” versions of familiar products.