I consider cases where you increase the risk that, e.g., someone will die, without increasing the risk that you will kill them: in particular, cases in which that increasing of risk is accompanied by a decreasing of risk of the same degree such that the risk imposition has been offset. I defend the moral legitimacy of such offsetting, including carbon-offsetting. A companion piece to [3]. [Journal]

Building on ideas introduced in [2] and [3], I propose a new formalist account of legal (/proximate) causation—one that holds legal causation to be a matter of amoral, descriptive fact. The account starts with a metaphysical relation, akin to but distinct from common-sense causation, and it argues that legal causation aligns exactly with the relation; it is unified and principled. [Journal]

We can cause windows to break and we can break windows; we can cause villages to flood and we can flood villages; and we can cause chocolate to melt and we can melt chocolate. Each time these can come apart: if, for example, A merely instructs B to break the window, then A causes the window to break without breaking it herself. Each instance of A breaking, melting or burning, etc. something, is an instance of what I call making. I argue that making is an independent, theoretically important notion—akin but irreducible to causing—and metaphysicians should pay attention to it. [Journal]

The three central chapters have been superseded by the published [3], [4] and [5]. The remaining chapter, "The Bystander doesn't Kill," expands on an argument given in [3] for the conclusion that the bystander in Thomson's famous case doesn't kill the lone worker onto whom she diverts the runaway trolley.

Geach and Thomson argued that while many things might be good for or as a so-and-so—a good toaster, for instance—there is no sense to the idea of anything being plain good. And if there is no sense to the idea of anything being plain good, there can't be any sense to the idea of maximising (plain) goodness and so consequentialism is a nonsense. Almotahari and Hosein recently sought to plug a hole in Geach's and Thomson's argument. I show that their plug doesn't hold and that Geach's and Thomson's argument is invalid. [Journal]

W. D. Ross said that A objectively ought to ɸ just if she ought to ɸ were she omniscient. Ross's idea remains, in outline, the received understanding of moral objectivism. Yet it returns false directives. For example, there is no doubt some solution to the global energy crisis. No one now knows what it is, but if you were omniscient, you would. By Ross's lights, then, you ought—right here! right now!—to solve the global energy crisis. We take it as a datum that this is not so, and in turn we reject Ross's idea. In its place, we offer a new account of objectivism, according to which A objectively ought to ɸ only if: if A were to try her best to ɸ, she would. It's not the case that you ought to solve the global energy crisis because if you tried your best to, you'd fail.

There is a difference between person A killing person B and A merely letting B die. In that same vein, there is also a difference between person A saving person B and A merely letting B live. If I resuscitate you, then I save your life; if I move my car out of the way, enabling the fire engine to reach your house in time, it might be that you would have died but for my moving my car, but I don't save you, the fireman carrying you through the flames does—I merely let you live. This paper argues that that metaphysical difference between saving and letting live gives rise to a moral difference. It then puts that moral difference to work. E.g. it accounts for the long-felt moral difference between failing to rescue a drowning child and failing to donate $4000 to Oxfam (sufficient for them, in the aggregate, to prevent a child's death). A companion piece to [3] and [5].

In [4], I proposed a new account of legal causation, one that had nothing to do with common-sense causation whatsoever. Here I extend that account beyond so-called principal liability to so-called accessory liability (though my account will ultimately disavow that breakdown). The account proposed here aims to capture the settled, central tenets of criminal liability for result crimes with two, simple metaphysical concepts: whether, e.g., defendant killed victim and/or whether she increased the risk that victim would die.