Wolfgang Schwarz

Local Descriptivism

To be an electron is to roughly satisfy our electron theory; to be a banana is to roughly satisfy our banana theory. To say that electrons or bananas are such-and-such is equivalent to saying that things (roughly) satisfying a certain theoretical role are such-and-such.

Thus our Total Theory of the world is arguably a priori equivalent to its "electron" ramsification or its "banana" ramsification, in which all occurrences of "electron" and "banana", respectively, have been replaced by existentially bound variables. What Total Theory adds to those Ramsey sentences is only the Carnap sentence for "electron" and "banana": the material conditional with the Ramsey sentence in the antecedent and Total Theory in the consequent. And this conditional is arguably analytic.

On the other hand, Total Theory is not equivalent to its total ramsification, in which all non-logical terms have been replaced by existentially bound variables: any Ramsey sentence like that is logically equivalent to a sentence stating only how many things there are; but Total Theory goes far beyond that.

Does it follow that while "electron" and "banana" are definable by their theoretical roles, some terms -- "consciousness", "causation", "left-hand side", "natural property"? -- are not and therefore can't be ramsified away?

Or should we conclude that even the meanings of "electron" and "banana" are not quite captured by any theoretical roles because external constraints affect their interpretation, so that even if we believe that "banana" applies to things iff they are Phi, it might happen that in reality it applies to things iff they are Psi, because Psi is the causal origin of our "banana" use, or because Psi is more natural than Phi.

(Another possibility why "banana" etc. might be exempt from ramsification is that English isn't rich enough to capture the relevant property. Imagine a fragment of English with "banana", "apple" and "nice" as the only predicates: certainly our Total Theory restricted to this language is not equivalent to its "banana" ramsification. But that can't be the entire story. Even in a language in which every property is expressible (Lagadonian comes to mind), Total Theory isn't equivalent to its Ramsey sentence. So something else must go wrong in the process of global ramsification.)

I think we needn't conclude anything like this. From the assumption that for all terms t, Total Theory is equivalent to its t-ramsification, it simply doesn't follow that Total Theory is equivalent to its total ramsification. Example: since mereological terms are interdefinable, you can ramsify out any one of them without loss of information. But if you ramsify them all at once, you're left with only the formal structure of mereology that is satisfied by any boolean algebra without zero, no matter if it has anything to do with parthood.

The same might be true on a larger scale: maybe ramsifying out a single term never makes any difference, but once you start replacing several terms simultaneously, you run the risk of losing information.

This possibility might not be very interesting if one holds that terms like "electron" literally get their meaning from their place in Total Theory. Then the other terms in that theory can hardly get their meaning via some theoretical role that contains "electron"; otherwise the process is unfounded and nothing could get any meaning to begin with.

But if one prefers a different kind of meta-semantics on which meanings are constituted by, say, causes, dispositions, conventions or intentions, then I think this possibility is not unattractive. It allows one to maintain that absolutely all terms are exhausted by their theoretical roles without succumbing to Global Descriptivism.