Mill-Ramsey-Lewis

ABSTRACT: Functionalists about truth employ Ramsification to produce an implicit definition of the theoretical term true, but doing so requires determining that the theory introducing that term is itself true. A variety of putative dissolutions to this problem of epistemic circularity are shown to be unsatisfactory. One solution is offered on functionalists’ behalf, though it has the upshot that they must tread on their anti-pluralist commitments.

Much recent philosophical attention has been devoted to variants on the Best System Analysis of laws and chance. In particular, philosophers have been interested in the prospects of such Best System Analyses (BSAs) for yielding *high-level* laws and chances. Nevertheless, a foundational worry about BSAs lurks: there do not appear to be uniquely appropriate measures of the degree to which a system exhibits theoretical virtues, such as simplicity and strength. Nor does there appear to be a uniquely correct exchange rate at which the theoretical virtues of simplicity, strength, and likelihood (or *fit*) trade off against one another in the determination of a best system. Moreover, it may be that there is no *robustly* best system: no system that comes out best under *any* reasonable measures of the theoretical virtues and exchange rate between them. This worry has been noted by several philosophers, with some arguing that there is indeed plausibly a set of tied-for-best systems for our world (specifically, a set of very good systems, but no robustly *best* system). Some have even argued that this entails that there are no Best System laws or chances in our world. I argue that, while it *is* plausible that there is a set of tied-for-best systems for our world, it doesn't follow from this that there are no Best System chances. (As I will argue, the situation with regard to laws is more complex.) Rather, it follows that (some of) the Best System chances for our world are *unsharp*.

Standard objections to the notion of a hedged, or ceteris paribus, law of nature usually boil down to the claim that such laws would be either 1) irredeemably vague, 2) untestable, 3) vacuous, 4) false, or a combination thereof. Using epidemiological studies in nutrition science as an example, I show that this is not true of the hedged law-like generalizations derived from data models used to interpret large and varied sets of empirical observations. Although it may be ‘in principle impossible’ to construct models that explicitly identify all potential causal interferers with the relevant generalization, the view that our failure to do so is fatal to the very notion of a cp-law is plausible only if one illicitly infers metaphysical impossibility from epistemic impossibility. I close with the suggestion that a model-theoretic approach to cp-laws poses a problem for recent attempts to formulate a Mill-Ramsey-Lewis theory of cp-laws.