Realism About What?
What historical generalization (if any) can we reasonably induce concerning posits licensed by the joint application of confirmational strategies S1 and S2? There seems to be a historically supported expectation of projectability of retention for theory-parts that make to L+ by passing both the S1 and S2 filters.
Admittedly, compared to fuller counterparts in their mother theories, the parts that make into L+ are comparatively abstract (less precise), restricted and coarse-grained, as are the descriptive theoretical schemes built on them. On the cheerful side, the postulated entities, processesand accounts presently sanctioned by S1 and S2, far from making a meagre picture of reality, provide a remarkably thick and highly textured array of claims about the world beyond the reach if unaided observation. Examples range from detailed cosmological histories from 13,000 million years back to the present, to descriptions at various levels of generality of the composition, structure, and interactions of matter, to organic life and its diversity, to histories of the rise of humans and human nature, and more. The resulting theoretical corpus is not a haphazard quilt of dubious significance but a body of abstract and perspectival, finite-range, coarse-grained assertions that, nevertheless, display astonishing (and growing) levels of integration into a detailed and textured picture of the world.If the suggestions made in this second part are on the right track, then, contrary to many thinkers (realists and antirealists), more than thirty years after Laudan (1981), it seems reasonable to claim that the divide et impera strategy of selective realism can succeed—or so I have tried to argue.
Taken separately, neither S1 nor S2 can accomplish the required task. Strategy S1 does a specially good job of filling list L— with theory-parts that face compelling specific doubts against them; also, S1 eliminates candidates marred by underdetermination or conceptual conundrums.
In turn, S2 brings both external support and novel intelligibility to conjectures introduced by a theory. Strategy S3 sharpens the assessment of the epistemic success achieved by past accounts. Selectivism, therefore, should require prospective theory-parts to have passing grades from both S1 and S2. The ensuing naturalist realist proposal comprises two complementary criteria:Refutational Criterion (R-): A theory-part will reveal itself as “doubtful” if multiple pieces of recalcitrant data converge inferentially in that particular part, and saving the part in question is consistently accompanied by degeneration of the whole system, as measured by current scientific criteria.
Positive Criterion (R+): A theory-part will reveal itself as appropriate for realist commitment (i.e. for being deemed “very probably approximately true”) if it passes muster by both S1 and S2.
Note that these criteria can be applied while a theory is in full flight. The historical record seems to support both criteria well.
8
More on the topic Realism About What?:
- Realism About What?
- Verificationist Anti-realism
- What is Scientific Realism?
- Do Electrons Exist? that is not the Question
- Introduction
- Robustness and Hacking’s Scientific Realism
- Realism of Properties, Realism of Entities and Their Role in Microphysics
- Laudan’s Historical Critique
- Arguments for Realism
- Facets of Realism