Is normative necessity distinct from metaphysical necessity?

In “Varieties of Necessity” (in Gendler and Hawthorne, Conceivability and Possibility, Oxford 2002), Kit Fine argued that we need to recognize that certain normative truths are in a sense necessary, and that the kind of necessity in question is sui generis, rather than being a special case of metaphysical necessity.

I shall not dispute Fine’s argument for the conclusion that there are normative necessities. However, I shall dispute his argument for the conclusion that these normative necessities are sui generis. On the contrary, as I shall argue, Fine does not give us a compelling reason to deny that normative necessity is a species of metaphysical necessity.

Fine’s central argument is this (§5, ¶2).

  1. Metaphysical necessity is that form of necessity that derives from the nature or essence of things.
  2. It is not part of the nature or essence of either naturalistic features or normative features that naturalistic features and normative features are connected in the way that they are.
  3. So, the connections that hold between naturalistic and normative features of things are not metaphysically necessary.

I shall not dispute the first premise (1). What seems more doubtful to me is the second premise (2).

Fine supports this premise with a specific example: “It is no part of what it is to be pain that it should be bad and no part of what it is to be bad that it should include pain.”

But is it really obvious that this premise is true? Why isn’t it part of the nature of badness that pain is bad? In general, why don’t the fundamental necessary general principles about what is good and what is bad partially define the very nature or essence of goodness and badness?

We can agree with Fine’s claim that it is not part of what it is to be a competent user of the concept ‘bad’ that one must either believe the general proposition that pain is bad, or be disposed to judge, of particular pains, that they are bad. This claim is, however, compatible with the further theses that the fundamental truth about the constitutive essence or nature of the property of badness itself is not itself a conceptual truth, and that this fundamental truth (together with the fundamental truth about the essence or nature of pain) entails that pain is bad.

In short, we can explain away Fine’s alleged intuition that “it is … no part of what it is to be bad that it should include pain”, by claiming that it does follow from the fundamental truths about the nature or essence of badness and of pain that pain is bad, but that these fundamental truths are “hidden”, in the sense that they are not themselves conceptual truths.

Since these truths about the essence of badness and of pain are “hidden” in this way, they are in a sense non-obvious. Since the constitutive essence of badness and of pain are non-obvious “hidden” truths, some people — like Fine — may misinterpret the fact that it is not obviously part of the essence of badness that pain is bad as showing that it is not part of the essence of badness that pain is bad.

Fine anticipates this response to his argument. But he assumes that the only way of developing this response is by arguing that the essence of badness is something that can only be discovered empirically. This is why he objects (§5, ¶7) that this response “is unable properly to respect the non-empirical character of ethical belief.” The rest of Fine’s discussion of normative necessity consists of a long and complicated argument for the conclusion that the fundamental normative necessities cannot be discoverable only empirically.

However, there is an obvious rejoinder to this objection. The proponents of this response to Fine’s argument do not have to claim that the essence or nature of badness can be discovered only empirically. They can claim that the essence or nature of badness can be discovered by synthetic a priori reasoning.

Because the kind of reasoning that could lead to an understanding of the essence of badness is synthetic, we can explain why it will involve more than just figuring out conceptual truths — and why the fundamental truths about the essence of badness are “hidden” and non-obvious. However, because it is synthetic a priori, we can agree with Fine’s argument that normative necessities can be discovered non-empirically.

Of course, the very idea of synthetic a priori truths is highly controversial. But so long as we hold that the fundamental truths about the nature or essence of normative features like goodness and badness can be discovered through synthetic a priori reasoning, we have an answer to Fine’s argument for distinguishing normative necessity from metaphysical necessity.

2 Replies to “Is normative necessity distinct from metaphysical necessity?

  1. Hi Ralph
    thanks for this – very interesting. Just a couple of quick comments:
    A) I wasn’t sure whether you wanted to draw from the conclusion that Fine’s argument to the conclusion that normative necessity is distinct from metaphysical necessity fails a further conclusion that normative necessity really is metaphysical necessity. It seems to me that, even if the failure of this argument means that as far as this argument is concerned normative necessity can be a form of metaphysical necessity, it would be still be nice to have positive reasons for thinking that it is. I wasn’t sure whether there were such positive reasons here apart from the rhetorical questions.
    2. I have a guess of what might be going on in Fine’s argument. Maybe he is thinking that normativity is always a relational property – that is, what is normative is something like a reason relation between a fact and an act (and properties like badness can be reduced to such relations). He might be then thinking that metaphysical necessities in 1 that are based on the essence of a thing must be monadic, non-relational intrinsic properties any duplicate would share no matter what relations it was in. This would be a reason to think that normative properties as relational properties are not a part of the metaphysical necessities. My inclination to resist this argument would be to argue that relational properties, especially dispositional ones, can be a part of a thing’s essence.

  2. You seem to accept Fine’s assumption,“It is no part of what it is to be pain that it should be bad”, otherwise there would be no problem to solve with the example. Could you say something about why we should accept this? Isn’t it plausible to say that pain is a thick concept denoting certain kinds of feeling that are, in at least one way, bad for the creature experiencing them?

Comments are closed.