September 27, 2008

Exceeding Our Grasp: Science, History, and the Problem of Unconceived Alternatives

By Love, Alan C

STANFORD, P. KyIe. Exceeding Our Grasp: Science, History, and the Problem of Unconceived Alternatives. New York: Oxford University Press, 2006. xiv + 234 pp. Cloth, $55.00-"Should we really believe that our best scientific theories simply tell us how things stand in the various inaccessible domains of nature they purport to describe" (p. 5)? According to explanationist defenses of scientific realism (namely, nothing other than realism makes sense of the profound success of science), yes. According to KyIe Stanford's novel argument against a realist construal of science, no. Stanford's tactic is to fuse two traditional objections to scientific realism: the pessimistic induction over the history of science and the underdetermination thesis. The former argues against realism using the preponderance of failed theories that were understood as successful at their time. The latter seizes on the possibility of empirically equivalent alternatives that also meet the strictures of an explanationist defense. Through an examination of the purported difficulties facing these objections (for example, isolating genuine rather than contrived empirically equivalent alternatives), Stanford weaves together a new induction over the history of science that operates out of the situation of "recurrent, transient underdetermination" within scientific communities; that is, not being able to conceive of viable alternatives. Over time scientific theories are replaced by (very different) alternatives that are at least as well confirmed by the evidence available to support earlier theories. The earlier theory was favored because no alternative had been conceived. Once the alternative is articulated and further evidence is procured, it becomes better confirmed than the earlier theory, thereby effecting the replacement-but then is immediately subject to the same situation. The history of science provides many reasons to believe that present scientists have not ruled out all relevant possibilities (unlike many cases of abductive reasoning in everyday life), and therefore the eliminative inference that licenses talking of current theories as approximately true should be resisted. Differences between past and present theories are irrelevant to the problem of unconceived alternatives, which concerns theorists not theories. Why think current individuals or communities are better able to explore the space of possible alternatives than those of the past? The historical muscle behind this argument is Stanford's appraisal of the explanatory reasoning of Darwin, Francis Galton, and August Weismann in the later nineteenth century regarding the nature of heredity and development. For example, Darwin simply did not conceive of common cause explanations for inherited characteristics. Instead of parental characteristics making a direct material or causal contribution (as in Darwin's theory of pangenesis), the appearance of a characteristic in both parent and offspring could be the result of a shared, underlying (transmitted) factor. This alternative was a serious possibility given the evidence available at the time and was advocated by Darwin's cousin and regular correspondent, Francis Galton. Thus, a realist interpretation of pangenesis as approximately true was unwarranted. Similar situations are demonstrated for Gallon's stirp theory and Weismann's germ plasm theory-whole classes of alternatives remained unconceived. Realist responses to this historical record and its implications for current theories, whether focused on the (lack of) maturity of past sciences, causal (or causal-descriptive hybrid) theories of reference, approximate truth, or selective confirmation of some components of past theories, all fail to blunt the difficulty in Stanford's estimation.

Realists will object that if beliefs cannot be sorted into those confirmed by predictive and explanatory success in connection with specific evidence and those of a more mundane sort, then Stanford's position reduces to a form of general skepticism (the problem of unconceived alternatives in everyday eliminative inference). Stanford appeals to the need to recover the heterogeneity of scientific practice and recognizes that limning these differences is a formidable challenge. But these differences also portend the selective confirmation of theory components (retrospectively and prospectively), thereby rescuing realism from the dilemma of unconceived alternatives. Another difficulty pertains to how historical changes in scientific community structure transform the problem of unconceived alternatives. This is germane because (following Stanford) it concentrates on theorists rather than theories; the dynamics of individual and social cognition under varying conditions must be considered.

Does antirealism leave us without a workable understanding of science, especially the life sciences, which have an intuitive realist appeal? Most discussions of realism have concentrated on physics, where the mathematical representation of theoretical content facilitates the consideration of diverse empirically equivalent alternatives. In biology, a form of realism about theoretical content seemingly undergirds the scientific discourse. Recognizing that the reinterpretation or elimination of theoretical language with realist overtones is not a plausible option, Stanford begins advancing a new perspective on instrumentalism in the final chapter. Although these positive components require much more elaboration, they routinely provoke further investigation. Are instrumentalists more inclined to explore potentially fecund alternative conceptualizations rather than simply confirm and extend extant theories?

Stanford has genuinely advanced the philosophical discussion about scientific realism with his careful articulation of the problem of unconceived alternatives. While his instrumentalism may be resistible, there is no question that proponents of realism will have to regroup to successfully meet the challenge.-Alan C. Love, University of Minnesota.

-Alan C. Love, University of Minnesota.

Copyright Review of Metaphysics Sep 2008

(c) 2008 Review of Metaphysics, The. Provided by ProQuest LLC. All rights Reserved.