The Language of Physics
Physics employs four denoting terms to cover what are usually referred to as forces: “gravity”, “electromagnetism”, “the weak force”, and “the strong force”. What is the semantics of these terms? Are they names, descriptions, or demonstratives? Do they function like standard natural kind terms?
It is sometimes said that Einstein’s theory of gravity in General Relativity shows that gravity is not really a force after all. The thought is that by replacing action at a distance with the geometry of space we have abandoned the idea that gravity is a force. There is nothing more to gravity than the structure of space. We used to think that gravity was a type of force acting across space, but it has turned out not to be. On this view, the term “gravity” does not analytically entail “force”. But there are other ways we could react to General Relativity: we could say that gravity has turned out not to exist, since the term does entail being a force and the theory shows that it is not a force; or we could hold that gravity does exist and has turned out to consist in the geometry of space, but it is still a force because force is reducible to geometry. So we have three possible positions: (a) gravity exists but has turned out not to be a force; (b) gravity exists and is a force because relativity theory tells us what this force is; (c) gravity does not exist because relativity shows it is not a force and “gravity” entails “force”. The first thing to say about these options is that it is completely unclear which view we should adopt; the meaning of the term “gravity” seems indeterminate between the three views. We could adopt any of them and not be accused of semantic (or scientific) impropriety.
Compare “magnetism”: what if we were to conclude that a charged particle distorts the space around it in a similar way to gravity? Should we then say that magnetism has turned out not to be a force? We could equally say that it is still a force but that the force reduces to geometry. Or we could simply announce that magnetism has turned out not to exist (like phlogiston). Again, the term “magnetism” seems too indeterminate to allow of a straightforward resolution; the decision about how to use the term in the light of the new theory seems purely pragmatic. It is easier practically to keep using the term and describing magnetism as a force, but this is hard to defend as the clear truth about the meaning of the term given the alternatives. We could equally declare that magnetism is not really a force after all, or that it doesn’t exist (we would then need a new term for what does exist in magnetic interactions).
Could there be fool’s gravity or fool’s magnetism? Could there be a Twin Earth that contains a force indistinguishable from gravity that isn’t gravity? That is hard to envisage: what could differentiate real gravity from its fake epistemic counterpart? The case is not like water and H2O, in which we have a clear idea of hidden structure in the shape of chemical composition, to be contrasted with superficial appearance. But if something acts like gravity or magnetism mustn’t it be gravity or magnetism? How could a force obey Newton’s inverse square law and not be gravity? So it is hard to see how the standard arguments apply to these terms of physics; they don’t fit the usual natural kind model. Or at least it is unclear that they do—we can describe hypothetical cases in different ways and not violate their meaning.
These physical terms are thus not readily classifiable as names or descriptions or demonstratives—that is, they don’t function in such a way as to fall clearly into any of those semantic categories. They are a bit like names, somewhat similar to descriptions, and akin to demonstratives—without being clearly any of the above. Thus we find semantic indeterminacy at the heart of the most exact of sciences. 
 Of course, we are only too familiar with epistemic indeterminacy at the quantum level, and possibly ontological indeterminacy. But it is another thing to find that “gravity” is neither a logically proper name of something, nor a definite description of something (“the force that acts at a distance”), nor a demonstrative term for something (“that physical phenomenon”). It seems to float between these alternatives. Thus we can’t say definitively whether Einstein abolished gravity in favor of spatial curvature or told us what gravity is (spatial curvature).