Randolph Clarke

Let F be a fact in virtue of which an agent, S, is blameworthy for performing an act of A-ing. We advance a slightly qualified version of the following thesis: (Reason) F is (at some time) a reason for S to feel guilty (to some extent) for A-ing. Leaving implicit the qualification concerning extent, we claim as well: (Desert) S's having this reason suffices for S’s deserving to feel guilty for A-ing. We also advance a third thesis connecting desert of feeling guilty with the fittingness of this response. In light of our three theses, we address several claims that have been made regarding responsibility and desert. We take issue with the divorce of desert from responsibility. We find acceptable a claim regarding blameworthiness and reason to induce guilt, and we defend the idea that it is noninstrumentally good that one who is blameworthy be subject to a fitting feeling of guilt. We argue against a view on which desert of blame has a teleological dimension.

Part of our ordinary conception of our freedom is the idea that commonly when we act—and often even when we don’t act—it is up to us whether we do this or that. This paper examines efforts to spell out what must be the case for this idea to be correct. Several claims regarding the basic metaphysics of agential powers are considered; they are found not to shed light on the issue. Thinking about agents’ psychological capacities provides some illumination, though the idea of freedom remains puzzling.

Several philosophers claim that the phenomenology of one’s own agency conflicts with standard causal theories of action, couched in terms of causation by mental events or states. Others say that the phenomenology is prima facie incompatible with such a theory, even if in the end a reconciliation can be worked out. Here it is argued that the type of action theory in question is consistent with what can plausibly be said to be presented to us in our experience of our agency. Several routes to a claim that there is nevertheless a prima facie incompatibility are examined, and all are found wanting. The phenomenology of agency, it is argued, is no threat to a standard causal theory of action.

Agent-causal accounts of free will face two problems. First, such a view needs an account of rational free action, that is, of acting for reasons when one acts freely. And second, an intelligible explication of causation by an agent is required. This paper addresses both of these problems. Free actions are seen as caused both by prior events and by agents. Reasons (or their mental representations) can then be seen as figuring causally when one freely acts for reasons. It is suggested that agent causation can be given a non-reductive theoretical definition, as on some views event causation is.

Freedom of the will is intrinsically valuable in deliberation and action, and it is a necessary condition for moral responsibility. Free will is not a sufficient condition for the latter; other abilities, and the absence of certain responsibility-undermining conditions, are also necessary. The free will requisite for moral responsibility is a self-determination in coming to have a particular intention in action. It does not consist even partly in an ability to do otherwise. ;Deliberation requires believing that one is free to choose and to act as one chooses. But deliberation does not require believing that one can perform each of the actions one is considering. Nevertheless, it is desirable in deliberation and action to determine, oneself, what ends one will pursue, and this self-determination appears to be incompatible with determinism. ;A leading compatibilist view identifies free will with a capacity to direct one's behavior by reflective practical reasoning. The freedom that agents have in akratic behavior and in some kinds of wrongdoing are especially difficult to account for on this view. A general consideration against the adequacy of this view is that an agent in a deterministic world, even if she acted with the capacity in question, would be in a thorough sense a product of the distant past. Though it might be appropriate to make moral judgements of such an agent, she would not deserve moral praise or blame. ;Incompatibilist views that simply add an element of chance to human action are also unsatisfactory. What must be secured is that generally when an agent acts, there are several different courses of action that are genuinely open to her, and she makes it happen that she performs one of these actions in particular. The most promising account of this kind of self-determination is an agent-causal view on which agents cause the orderings of reasons for which they act. It can be allowed that free actions are probabilistically caused by prior events. However, the agent causation that is required for free will is not reducible to event causation.

This paper examines the view of abilities to act advanced by Kadri Vihvelin in Causes, Laws, and Free Will. Vihvelin argues that (i) abilities of an important kind are “structurally” like dispositions such as fragility; (ii) ascriptions of dispositions can be analyzed in terms of counterfactual conditionals; (iii) ascriptions of abilities of the kind in question can be analyzed similarly; and (iv) we have the free will we think we have by having abilities of this kind and being in circumstances that are propitious for their exercise. I raise doubt about each of these claims. Further, I argue that even if abilities of the kind in question are dispositions, and even if determinism is compatible with our commonly having unexercised abilities of this kind, the compatibility of determinism and free will remains in question.

The chief difficulty for agent-causal accounts lies in defending the notion of agent causation. Either of two types of realist account of causation can be drawn on to explicate the claim that enduring substances are among the causes of things. But there remains the objection that, although this claim is intelligible, it is necessarily false. Several objections to the possibility of substance causation are considered, and it is concluded that there are, on balance, good reasons to reject this possibility.

Agent-causal accounts aim to secure greater control than can be secured by any event-causal libertarian account. Assuming that all it requires is possible, an integrated agent-causal view succeeds at this goal and adequately characterizes free will. Such a view captures well the common idea that free agents are originators of their free actions. Responses are offered to Peter van Inwagen’s challenge to agent-causal views and to Galen Strawson’s argument that free will is impossible. A claim that free will requires downward causation—of a sort involving the supersession of microlevel laws—is rejected.

This chapter begins an assessment of agent-causal libertarian accounts, which require causation by agents, where this is construed as causation by enduring substances and not reducible to event causation. Timothy O’Connor’s agent-causal view is examined. Like most such accounts, it does not require, in any case of acting freely, that events such as the agent’s having certain reasons cause the event that the agent is required to directly cause; the view consequently fails to provide for the reason-explanation of free actions. An alternative agent-causal account is advanced, one that integrates agent causation with an event-causal theory of action.

Here I examine the charge that the indeterminism required by event-causal accounts is at best superfluous; if free will is incompatible with determinism, then, it is said, no event-causal libertarian account adequately characterizes free will. The distinction between broad incompatibilism and merely narrow incompatibilism is brought to bear. If the latter thesis is correct, then an event-causal account can secure all that is needed for free will. However, if broad incompatibilism is correct, then no event-causal account is adequate, though such views can still secure some things of value that cannot exist given determinism, such as distinctive types of difference-making and attributability, and the truth of the presumption of open alternatives that we commonly make while deliberating.

This chapter examines the charge that the indeterminism required by standard event-causal libertarian accounts would diminish the control that is exercised in acting. The objection has been advanced with an ensurance argument and an argument from luck. Both arguments are rejected; nondeterministic causation of an action by its immediate causal antecedents need not diminish at all the type of control relevant to free action. This chapter further assesses the account of free will advanced by Robert Kane, which imposes certain special requirements meant to address the problem of diminished control. These special requirements provide no help; but, in any case, none is needed.

Deliberative libertarian accounts allow that basic free actions may be causally determined by their immediate causal antecedents; indeterminism is required only at earlier points in the processes leading to free actions. Accounts of this type proposed by Daniel Dennett, Laura Ekstrom, and Alfred Mele are examined here. Given the assumption of incompatibilism, deliberative accounts fail to provide for the sort of difference-making that is distinctive of free action. Further, they fail to evade the problem of diminished control that they are meant to evade.

A growing number of philosophers now hold that agent causation is required for agency, or free will, or moral responsibility. To clarify what is at issue, this paper begins with a distinction between agent causation that is ontologically fundamental and agent causation that is reducible to or realized in causation by events or states. It is widely accepted that agency presents us with the latter; the view in question claims a need for the former. The paper then examines a “disappearing agent” argument from Derk Pereboom that is aimed at showing that free will requires agent causation that is ontologically fundamental. It is argued that the argument fails. Further, it is argued that, contrary to Pereboom’s claim, the issue raised by his disappearing agent argument is distinct from the problem of present luck that libertarian theories of free will face. The paper concludes with an assessment of the prospects for success of a disappearing agent argument showing that agent causation that is ontologically fundamental is required for agency tout court.

This paper defends a minimal desert thesis, according to which someone who is blameworthy for something deserves to feel guilty, to the right extent, at the right time, because of her culpability. The sentiment or emotion of guilt includes a thought that one is blameworthy for something as well as an unpleasant affect. Feeling guilty is not a matter of inflicting suffering on oneself, and it need not involve any thought that one deserves to suffer. The desert of a feeling of guilt is a kind of moral propriety of that response, and it is a matter of justice. If the minimal desert thesis is correct, then it is in some respect good that one who is blameworthy feel guilty—there is some justice in that state of affairs. But if retributivism concerns the justification of punishment, the minimal desert thesis is not retributivist. Its plausibility nevertheless raises doubt about whether, as some have argued, there are senses of moral responsibility that are not desert-entailing.

An autonomous reason for intending to A would be a reason for so intending that is not, and will not be, a reason for A-ing. Some puzzle cases, such as the one that figures in the toxin puzzle, suggest that there can be such reasons for intending, but these cases have special features that cloud the issue. This paper describes cases that more clearly favour the view that we can have practical reasons of this sort. Several objections to this view are considered and rejected. Finally, it is considered whether the existence of such reasons would conflict with an attractive coherence principle linking the rationality of intending with that of acting as intended. The paper concludes with a qualified affirmation of autonomous reasons for intending.

A disposition mask is something that prevents a disposition from manifesting despite the occurrence of that disposition’s characteristic stimulus, and without eliminating that disposition. Several authors have maintained that masks must be things extrinsic to the objects that have the masked dispositions. Here it is argued that this is not so; masks can be intrinsic to the objects whose dispositions they mask. If that is correct, then a recent attempt to distinguish dispositional properties from so-called categorical properties fails.