Chris Reus-Smit has done an exemplary job in presenting my argument despite its complexity. Having shown the situatedness of my explorations, I obviously cannot present a “grand” theory without contradicting my argument concerning the inappropriateness of subjecting praxis to “theoretical” standards. But Chris’s “further questions” show their bite. Pointing to the diversity of meanings that make “semantic fields” much less orderly than assumed, he questions my preference for semantics over social epistemes, notions of causality, and so on. His second query concerns habits and nonlinguistic ways of acting. Finally, there is the issue of whether my conventionalist account, mediating between the “is” and the “ought,” allows for conceptualizations of what is “good and right,” exemplified by the human rights discourse.
As to the semantics: the primacy of language does not involve the primacy of any specific semantics in which historical experiences, folklore, and so on, are sedimented. I agree that semantic fields are messy, because most of our concepts develop through analogies and metaphorical extensions. Consequently, our concepts are not “covering,” but are more like clusters with overlaps, as suggested by Wittgenstein. Conceptual analysis is therefore an attempt at ordering— given specific questions and interests— that cannot refer to an “external” preexisting world. Even an assumed “perfect order” will not do, because principles and values cannot be ordered once and for all. “Order” must therefore be understood as “ordering,” not as a realization of an “ideal order.”
As to the nonlinguistic ways of “doing,” habits (Michael Oakeshott) and “drills” (Wittgenstein) enter, as have recent attempts at “incorporating” emotions into action accounts. Reus-Smit surmises that my preference for semantics might indicate a claim to the primacy of communicative practice. I think he is correct, but something more needs to be said. I am neither a Habermasian (in his earlier emphasis on transcendental pragmatics) nor do I necessarily espouse the “logic of the better argument.” My point is that we must understand (collective) action differently: as performances rather than as the execution of preexisting plans.
As to Wittgenstein and his emphasis on “drills” in learning how to “go on,” already Hume stressed that interacting in the social world is “learned” not by speculation and demonstration but through “learning by doing.” This sounds like a behaviorist program of conditioning, mistaking all actions as “re-actions” performed by a Pavlovian dog. But that cannot be correct. Following rules—for example, learning from a master painter—is always accompanied by verbal glosses and thus quite different from a conditioning program. Similarly the “dos” and “don’ts” in moral training are also transmitted not only by imperatives but also by saying how a “nice child” would act and by teaching appropriate emotional responses: not to get mad, not to sulk, but to face up to new challenges despite past failures. Thus, not only is “thinking” an eminently social affair but also our emotions are not purely idiosyncratic responses, which a “theory” of action could ignore.
As to Reus-Smit’s suggestion that the human rights discourse provides a more fruitful approach to mediate between the “is” and the “ought,” here he and I probably differ. I voiced my doubts in my previous book. Viewing the entire problematique of choice as realizing “subjective rights— making even constitutional arrangements a subjective “right to democracy” —remains for me problematic.
Finally, concerning “grand theory,” no theory of action can tell you what to do, as Wittgenstein’s example of the signpost shows. The signpost points in a direction, but should you therefore go that way? This is an issue of embracing a choice by making a commitment that cannot be left to an “objective” observer, because our life and our freedom are at stake. Second, the main issue in this exchange is whether our understanding of practical choices will be better served by “theory,” even if it is no longer limited by positivism and has become attentive to “types,” “cases,” and culture. Or is a different mode of thinking required to meet not only the challenges to theorizing that have been around since Aristotle, Hume, and Weber but also those raised by pragmatism and language philosophy (Wittgenstein, Searle)? It seems that Reus-Smit tends toward the former, whereas I espouse the latter position.
