Author:Fox, Philip

THE PHILOSOPHICAL LITERATURE witnesses a by now fairly large debate about the degree to which normative notions like "ought" or "reason" are perspective dependent. Roughly speaking, objectivists believe that what one ought to do depends on all the facts, whereas perspectivists believe that what one ought to do depends instead on one's epistemic perspective on the facts (and so only on facts that are epistemically available in some yet-to-be-specified sense). (1) Their dispute traditionally revolves around examples like the following:

Doctor: Jill, a doctor, must decide how to treat her patient. She only has two options: prescribe pill A or pill B. After careful research, her evidence decisively suggests that A is the cure and B is lethal. In fact, however, the reverse is true: A is lethal and B is the cure. Jill knows that the patient will die soon if she does nothing and that she has no time for further research. What ought Jill to do? (2) The answer is not obvious. On the one hand, one could side with the objectivist and say that Jill ought to prescribe B. After all, this is the only way of saving the patient's life. On the other hand, one could side with the perspectivist and say that Jill ought to prescribe A. After all, choosing any other option seems completely irresponsible.

What makes cases like Doctor interesting is that both positions seem to have significant intuitive appeal. Even if one is initially more attracted to one of them, it is easy to imagine how someone might have ended up on the other side. Partly for that reason, there is little ground for thinking that intuition alone will help advance the debate between objectivists and perspectivists much further. At best, appealing only to our intuitions looks like a recipe for a stalemate.

Fortunately, a less intuition-based, more theory-driven approach is available. In order to pursue such an approach, this paper draws together two separate themes from the recent literature: first, that the normative is in some sense action guiding, and second, that a discussion of perspective dependence in the practical domain benefits from a discussion of similar issues in the epistemic domain. (3) The main thesis is that objectivism is incompatible with a very plausible assumption about the possibility of being correctly guided by a normative theory.

Put succinctly, my argument against objectivism is as follows:

The Argument from Action Guidance

  1. It is sometimes possible for a normative theory to correctly guide action.

  2. If objectivism is true, this is never possible.

  3. Therefore, objectivism is false.

    The central advantage of this argument is that it only makes an extremely weak assumption about the action-guiding role of normative theories. In particular, it does not assume that normative theories are always capable of guiding action, and--as I will explain in more detail below--it thereby avoids recent criticisms of guidance-based arguments against objectivism.

    With the main argument on the scene, let us pause briefly for a general observation about arguments of this kind. I take it that all guidance-based arguments face the same choice. First, they could employ a fairly demanding notion of action guidance, in which case it will be relatively straightforward to explain why objectivism is inconsistent with it, but also correspondingly more difficult to show that this notion does not beg the question against objectivists right away; this approach is taken by Errol Lord, and it has been criticized accordingly by Clayton Littlejohn and Jonathan Way and Daniel Whiting for its strong assumptions about the possibility of action guidance. (4) Or second, guidance arguments could employ a fairly modest notion of action guidance, in which case it will be relatively straightforward to show why it does not beg the question against objectivists, but correspondingly more difficult to explain why objectivism is inconsistent with it. (5) This paper takes the second approach, and so for the most part I will be concerned with defending premise 2 above.

    I proceed as follows. Section 1 introduces objectivism and perspectivism as two substantive views about the perspective dependence of "ought." In section 2, after setting out two central assumptions of this paper, I develop and defend an account of what it is for a normative theory to correctly guide action. Section 3 anticipates a number of potential objections to this account. In section 4, I use this account to explain why objectivism rules out that normative theories are ever correct guides to action. Section 5 concludes. Throughout, I aim to advance the debate over the perspective dependence of "ought" beyond merely intuitive considerations and to shed light on a notion of action guidance that should be of independent interest to normative theory.


    I have already mentioned Doctor, a representative example that illustrates the disagreement between objectivists and perspectivists. To repeat, Jill has two pills available to treat her patient: pill A, which would be lethal, and pill B, which would cure the patient. She also has decisive but misleading evidence about which is which. What ought Jill to do?

    Objectivists will say that Jill ought to prescribe B, because, on their view, what one ought to do depends on all the facts. This includes the fact that pre-scribing B will actually cure the patient, which--if there is no epistemic filter that reason-giving facts must pass--we can plausibly assume to be a decisive reason to prescribe B. Perspectivists, however, will say that Jill ought to prescribe A, because, on their view, what one ought to do depends on one's epistemic perspective on the facts. Since Jill's available evidence decisively suggests that only A will cure the patient, it is this pill that she ought to prescribe according to perspectivism.

    As I have indicated, there is a question about what exactly it means for a normative notion like "ought" to depend on one's epistemic perspective. (6) Since my argument against objectivism does not depend on this question, I will not say much about it here. To simplify the presentation and without further argument, I will adopt Benjamin Kiesewetter's elaborate version, which has been particularly influential in the recent literature and whose central thesis is the following: (7)

    Evidence-Relative Perspectivism: S ought to [phi] if and only if S's available reasons decisively favor [phi]-ing. For our purposes, it will suffice to work with an intuitive notion of "available reasons." (8) In Doctor, Jill has a body of evidence available to her--such as a review of recent articles from medical journals and the opinion of her well-respected colleagues--that gives her strong reasons in favor of prescribing A. Furthermore, all facts that might count against prescribing A, including the fact that it will kill the patient, are epistemically unavailable to her in any intuitive sense: by assumption, she does not know that these facts obtain, nor is she in a position to know or even justifiably believe them. On a perspectivist reading, any plausible normative theory will thus entail that Jill's reasons to prescribe A are overall decisive and so she ought to prescribe A.

    We have seen that objectivism and perspectivism entail conflicting verdicts about what Jill ought to do. Before I discuss the relative plausibility of these views, however, let me try to preempt (though perhaps not decisively) a suspicion that might have occurred to the reader by now: that the dispute between objectivists and perspectivists is merely verbal.

    In particular, we could follow the lead of what we can call disambiguationist views, which recommend distinguishing between two different notions of "ought." On such a view, Jill ought objectively to prescribe B, but ought subjectively to prescribe A. With such a distinction at hand, we might then say that objectivists and perspectivists do not have a substantive disagreement. They would both be right about Jill in different senses.

    Like most other parties to the debate, however, I do not think that the problem can be solved so easily (9) The reason for this lies in the nature of the deliberative "ought," which is the kind of "ought" that we are interested in here. A judgment of the form "I ought to [phi]" is supposed to answer the deliberative question, "What ought I to do?" in a way that concludes my deliberation and allows me to make a rational decision on the basis of my deliberation. Importantly, an "ought" judgment can do this only if it employs a univocal notion of "ought." For I might know in a given situation that I [ought.sub.subjectively] to prescribe A and that I [ought.sub.subjectively] not to prescribe A. (10) But in order to conclude my deliberation and make a decision on this basis, I still need to know what I ought to do, all things considered. There is no point in disambiguating, again, between a sense in which I ought all things [considered.sub.subjectively] to prescribe A and a sense in which I ought all things [considered.sub.objectively] not to prescribe A, for the same question will arise all over again. Unless there is, at some point, a univocal answer to the question of what I ought to do period, I cannot conclude my deliberation in the way that is necessary for me to make a rational decision on the basis of such deliberation. What objectivists and perspectivists have a substantive disagreement about is whether what one ought to do in the univocal deliberative sense is perspective dependent or not. (11)


    This section has two aims. First, I set out two central assumptions of my argument, one about the possibility of action guidance and the other about the relationship between objectivism in the practical and the epistemic domains. Second, I offer a substantive but still sufficiently generic account of what it is for a normative theory to be correctly action guiding.


To continue reading