Wolfgang Schwarz

Causation and the Ramsey-Carnap-Lewis account of theoretical terms

Peter Menzies and Huw Price, in their forthcoming "Is Semantics in the Plan?" have spotted a mistake in Lewis's "Psychophysical and theoretical identifications". But they don't spot that it's a mistake, and rather think it shows that the Ramsey-Carnap-Lewis-account of theoretical terms is severly limited.

The mistake is that Lewis identifies "theoretical role" with "causal role":

... a general hypothesis about the meanings of theoretical terms: that they are definable functionally, by reference to causal roles. (p.249)

This is what I have called a functional definition. The T-terms have been defined as the occupants of the causal roles specified by the theory T (p.255).

Menzies and Price take this for granted and conclude that the RCL-account is inapplicable to any term that can't be analysed by causal roles, and also to causation itself because here a causal analysis will be uninformative.

But I think it is very clear from what Lewis writes elsewhere that theoretical roles need not be causal roles. This restriction doesn't occur in the more rigorous "How to define theoretical terms", nor in later presentations of the RCL-account in "Putnam's Paradox", pp.58-60, "Naming the Colours", pp.346f. and "Ramseyan Humility", §3. Indeed, the restriction doesn't even fit Lewis's own example in "Psychophysical", where a detective proposes the following theory:

X, Y and Z conspired to murder Mr. Body. Seventeen years ago, in the gold mines of Uganda, X was Body's partner ... Last week, Y and Z conferred in a bar in Reading ... [...] (p.250)

Lewis continues to explain that X, Y and Z are the theoretical terms, or T-terms implicitly defined by the theory, whereas all the other terms are O-terms, which, Lewis explains, can be "any old terms", including mathematical terms, causal terms or whatever. And it would be odd to claim that the detective's story defines X, Y and Z by causal roles: being a partner of Mr. Body in the gold mines of Uganda seventeen years ago isn't a causal property in the straightforward sense in which being typically caused by tissue damage is.

I think Lewis made the mistake because "Psychophysical" is mainly about mental states which are, according to Lewis, defined by their causal roles.

Menzies and Price wonder what other concepts could replace causation as general O-terms for the definition of any T-term. The only candidate they can think of are semantic terms: any T whatsoever is definable as "the referent of T". But this, as they note, isn't very helpful if, say, we want to use the RCL-account to solve Frank Jackson's physicalist location problem.

The obvious solution, I think, is to follow Lewis and simply refrain from specifying, once and for all, some limited vocabulary for O-terms. Semantic terms might be definable by mental (and other) terms, mental terms by causal (and other) terms, causal terms by counterfactual terms, and so on. For any T-term there have to be some O-terms that define it; but this does not mean that there have to be some O-terms that define any T-term. .