Two Additional Confirmational Strategies

A second strategy (S2) looks for support available to assumptions made by T from sources initially external to the theory T at hand, especially independently suc­cessful empirical theories.

Since elucidation springs from an independently supported theory T*, it raises the credibility of the assumptions and narratives it casts light on—hence its inter- estto realists. External support thus helps sharpen list L+. Strategy S2 has clear pluses. Since the ether posit never got any such support in the 19th century, ref­erences to the ether would have been filtered out (Cordero 2015). On the other hand, as a marker of probable truth, elucidation seems neither necessary nor suf­ficient for realism. Unsavory counterexamples give pause to granting any given theory-part high probability from elucidation alone. For example, when Kepler looked for broader theoretical support for his 2nd Law, he derived it from the Aristotelian laws of motion for sublunary bodies and some principles of optimal action. Kepler elucidated his law, but only by invoking as premises some of the most hopelessly wrong claims of Aristotelian physics. So, strategies S1 and S2 improve the realist project, although the job each can do seems less than ideal.

There is yet another policy relevant to the selectivist cause. A third evaluation strategy (S3) develops when a theory starts to wane, and continues for some time after it dies. It comprises efforts to explain why a discarded theory showed empirical success. The retrospective analyses involved are not automatically “self-serving.” Some S3 analyses unveil previously unappreciated causal or struc­tural justification for a theory's accomplishments. Recall, for instance, the expla­nation wave theorists gave for the success of corpuscularian optics regarding the phenomena of reflection, refraction, and polarization. Retrospective analyses fre­quently add precision to specifications of the parts a past theory gets right, as can be seen, for example, in recent attempts to show why erroneous fundamental posits in KirchhofPs theory of diffraction led to correct predictions (e.g. Brooker 2008). Realists can sharpen L+ by turning to the materials yielded by this form of ret­rospective elucidation, as indeed practicing scientists routinely do.

Unlike the versions of retrospective reading denounced by Stanford (2006), in S3 the elucidation of past truth is both informative and non-trivial. Past proposals may or may not show epistemic gains by S3 lights. Pre-modern theories fare poorly in this regard. The geocentric accounts of a planet's motion, for example, lacked cumulative content in terms of epicycles, eccentrics, and equants. In Ptolemaic models content growth was basically limited to (a) the most directly observable structures and (b) a set of claims held to be beyond doubt—notably the Principle of Uniform Circular Motion for heavenly bodies and the Aristotelian arguments for the absolute fixity and sphericity of the Earth. The main reason why the resulting orbits were denied realist interpretation is not because they failed the intelligibility requirement. If anything, at many levels, Ptolemaic constructions went out of their way to honor intelligibility—then guided by the Aristotelian theses just noted. Rather, the realist reason for rejection was that the epicycles, deferents, and equants invoked were multiply-realizable by the lights of existing knowledge (i.e. available data and commonly accepted principles). Similar comments apply to numerous other theories (think, for example, of the medieval approaches to physiology based on the doctrine of the humors). Here the point is that strategy S3 is not self-serving: Finding truthful content about unobservables in past theories can be a meaningful achievement. Moreover, finding such content is even central to getting realist inductions off the ground.

7