Matthew Bedke

Having no recourse to ways of knowing about the natural world, ethical non-naturalists are in need of an epistemology that might apply to a normative breed of facts or properties, and intuitionism seems well suited to fill that bill. Here I argue that the metaphysical inspiration for ethical intuitionism undermines that very epistemology, for this pair of views generates what I call the defeater from cosmic coincidence. Unfortunately, we face not a happy union, but a difficult choice: either ethical intuitionism or ethical non-naturalism, but not both.

This paper considers normative naturalism, understood as the view that (i) normative sentences are descriptive of the way things are, and (ii) their truth/falsity does not require ontology beyond the ontology of the natural world. Assuming (i) for the sake of argument, I here show that (ii) is false not only as applied to ethics, but more generally as applied to practical and epistemic normativity across the board. The argument is a descendant of Moore's Open Question Argument and Hume's Is-Ought Gap. It goes roughly as follows: to ensure that natural ontology suffices for normative truth, there must be semantically grounded entailments from the natural truths to the normative truths. There are none. So natural ontology does not suffice for normative truth.

Historically, the most persuasive argument against external reasons proceeds through a rationalist restriction: For all agents A, and all actions Φ, there is a reason for A to Φ only if Φing is rationally accessible from A's actual motivational states. Here I distinguish conceptions of rationality, show which one the internalist must rely on to argue against external reasons, and argue that a rationalist restriction that features that conception of rationality is extremely implausible. Other conceptions of rationality can render the restriction true, but then the restriction simply fails to rule out external reasons.

You know the story. You have a few intuitions. You propose a few theories that fit them. It’s a living. Of course, things are more complicated than this. We are sensitive to counterexamples raised by others and wish to accommodate or explain away an ever-wider base of intuitive starting points. And a great deal of the action occurs in rational reflection that can alter what is intuitive, and in theorizing that overturns formerly justified beliefs and moves us to new justified beliefs. Details aside, this method in ethics and elsewhere—of first relying on intuitions to form justified beliefs, and subsequently using best-fit (or reflective equilibrium) theorizing on all justified beliefs to move to other justified beliefs—has received a lot of critical attention lately. But it is not a bad method. It is a good method caught in a bad relationship. For its presumptive metaethical companion, realism, would have us believe that intuitions support beliefs about real, stance-independent facts of the matter. That strikes many as dubious. After sorting through some relevant concerns in this vicinity, I argue that the solution is not to reject intuitional methods but to embrace quasi-realism.

Here I discuss the conceptual structure and core semantic commitments of reason-involving thought and discourse needed to underwrite the claim that ethical normativity is not uniquely queer. This deflates a primary source of ethical scepticism and it vindicates so-called partner in crime arguments. When it comes to queerness objections, all reason-implicating normative claims—including those concerning Humean reasons to pursue one's ends, and epistemic reasons to form true beliefs—stand or fall together.

It often seems that what one ought to do depends on what contingent ends one has adopted and the means to pursuing them. Imagine, for example, that you are applying for jobs, and a particularly attractive one comes your way. It offers excellent colleagues in a desirable location, the pay is good, and acquiring a job like this is one of your ends. If practicing your job talk is a means to getting the job, the following seems true: (1) If you want1 to get the job, then you ought to practice your job talk. Let us call conditional ought sentences that purport to express an end in the antecedent and a means to the end in the consequent end-given oughts. Some end-given oughts run into the problem of detachment; i.e., some end-given oughts seem true, and yet we do not think the consequent by itself, detached from the conditional, is true even if the antecedent is true. Consider a case where you want revenge on Bill for some slight offense. You happen to have the opportunity to poison Bill’s drink while he is away, which is the only thing that would lead to his demise. What of: (2) If you want to kill Bill, then you ought to poison his drink. (2) runs up against the problem of detachment because of the following modus ponens argument: • If you want to kill Bill, then you ought to poison his drink. • You want to kill Bill. • Therefore, you ought to poison his drink. In fact, you ought not poison Bill’s drink. You ought to avoid him and seek counseling.