Fisher presents it as follows: "Imagine that we are trying to define 'neutron,' 'electron,' and 'proton' but in a way that does not rely on any theoretical physics. ... The basic Ramsey-Lewis method is to tell a story something like this":
There is one kind of thing, and another, and yet another; instances of the first of these orbit a clump of instances of the other two; instances of the first and the second are attracted to each other; instances of the first repel each other, as do instances of the second; instances of the third exhibit no attraction or repulsion to other instances of its kind; some strange force keeps the members of the second kind together in a clump despite their mutual repulsion; and so on.
So, we can apparently name the first thing a1, the second thing, a2, and so on. Having listed their properties and relations, we then rename a1 to electron, etc.
First of all, notice that we link names with things by ostension. This, we say pointing at an object, is going to be called a1. We are focusing on a real elementary particle. This seems less tractable for moral terms. Regarding "right," "wrong," "good," "evil," what object whose essence we want to grasp are we pointing at?
Second, we only know that a1 is in fact different from a2 once we have ascertained that they have different properties. We subject various particles we can't yet distinguish to many different empirical tests and see if they react to these tests differently. An assertion like "there is one kind of thing, and another, and yet another" is part of the conclusion of the Ramsey-Lewis method not its premises. For moral terms, we are supposed to distinguish between our m1 and m2 by conceptual analysis. But it seems to me that in the beginning our m1 is simply undefined, and the project cannot even get started. If, however, we assume that m1 is distinct from m2, as right is distinct from wrong, then one, this begs the question against both non-realism and non-naturalism; and two, why bother with the Ramsey-Lewis method rather than simply describe the nature of moral terms?