C. Why should we be skeptical about naturalistic theories? Read this paragraph by David Hume until makes sense to you:
In every system of morality, which I have hitherto met with, I have always remark'd, that the author proceeds for some time in the ordinary ways of reasoning, and establishes the being of a God, or makes observations concerning human affairs; when all of a sudden I am surpriz'd to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is however, of the last consequence. For as this ought, or ought not, expresses some new relation or affirmation, 'tis necessary that it shou'd be observ'd and explain'd; and at the same time that a reason should be given; for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it. (Treatise of Human Nature, book 3, part 1, §1)
Hume is here making a pretty good case for why it is impossible to deduce values from only facts: It seems that whoever tries this is always forced to change the subject from is-es to oughts. A way of restating this conclusion is to say that is-es never entail oughts.
In the 20th century, G.E. Moore expanded on this pattern of reasoning with what he called the Open Question Argument. I will begin with a presentation of the argument that follows closely Moore’s original formulation.
Suppose you provide a naturalistic analysis of a term like “good” or “right action” – as I did in B above. Suppose, for example, that you say that a right action is whatever the Pope says to do. It seems that without any internal contradiction, I can ask the following question: “I know the Pope said to do it, but is it the right thing to do?” In other words, I can agree with the natural facts in the analysis (that the Pope said to do it), but even when I do, it remains for me an open question whether that action is right. This is meant as a general point from which we should conclude that any naturalistic analysis of “good” or “right action” cannot be correct, because even if the factual matters in the analysis are in place, it’s still an open question about whether or not there’s also goodness, or rightness there.
Moore’s own example invokes the following analysis: What’s good is what is pleasurable. Since what is pleasurable is a matter of natural facts, this theory is naturalistic. He then proposes the following question: “Sure it’s pleasurable, but is it good?” and observes that the mere fact that something is pleasurable [or satisfies any other factual criterion] is never enough to establish that it is good. Since a thing’s being pleasurable [or its having any other natural-fact property] leaves open the question of whether it is good, Moore took this to mean that goodness impossible to reduce to mere natural facts.
A less confusing way to state this argument would be the following: Describe an action using only facts (the “is-es” of the left column). Try to choose your facts in such a way as to best make the case that the action is the right thing to do. Now imagine another person who agrees with the truth of all the facts you listed, but wonders whether the action really is the right thing to do. Is she making a logical mistake? Do the facts you listed on their own settle whether it’s the right thing to do? Suppose your partner in the debate asks you: “What’s the relevance of the facts you listed to the ethical question at hand?” You would be tempted to answer this with a value statement, but that would only show that you were implicitly making normative assumptions all along, and your normative conclusion didn’t come from pure facts after all. However, if you resist this trap and answer her question with a factual statement, your interlocutor could ask you for the relevance of that new fact to the ethical question. “Fine,” she would say, “add that fact to the list. But you’re no closer to explaining the relevance of this longer list of facts to the normative question.” No matter what facts you list about some action, it still seems like it will be an open question about whether the action is right.
The Open Question Argument is a (highly disputed2) reason for thinking that you can’t get evaluations from facts alone. As Wittgenstein would say, facts entail only facts, not ethics. There are two separate controversies here. One is whether facts perhaps can entail evaluations after all. Rachels, in his “unfair test” example, suggests that he thinks they can (but I think he makes a mistake; see above). More interestingly, perhaps some subtle naturalism could be right even if facts never entail values. As you know, water = H2O. However, the fact that some stuff is H2O does not entail that it is water. Entailment is a logical relationship, and the identity of H2O and water was a scientific (and not logical) discovery. The point is, some analysis like “water=H2O” could be right even if there is no entailment, and that leaves the door open for some values-to-facts analyses being potentially right (though still undiscovered).
A further controversy is whether there even are normative facts. Most contemporary ethicists are Cognitivists, which is to say, they believe that normative (“right-column”) statements do have truth values: All those statements are either true or false (though it’s not always easy to tell which). Naturalism would be a plausible account of how they get truth values: Normative claims are true if the facts to which they reduce are the case. There is one further intuition worth taking seriously: We think that if two objects, or two situations or two people are absolutely indistinguishable in all natural facts, then they also match in their moral qualities. Moral facts (if there are any), may be difficult to reduce to natural facts, but they also don’t seem to float free of natural facts. Now imagine two entire universes that are identical in all natural facts – that is, every thought and every electron in one universe has an exact duplicate in the other. Both would be perfectly described by the same Wittgenstinian world-book. Would it make sense for these two universes to differ with respect to any normative facts? Can you imagine such a thing? If you can’t (and I can’t), that strongly bolsters the claim that normative truths do somehow depend on natural facts. If they do then some sort of naturalism is true, though it’s very difficult to say more about it than this. Whenever we propose a specific naturalistic theory, like “right action = whatever action maximizes the pleasure in the world,” we can always apply the Open Question Argument and realize that there is conceptual room for an action which does indeed maximize global pleasure, but is nonetheless not right. It seems that this conceptual gap will always exist.