R. Jay Wallace is a specialist in moral philosophy and practical reason. He wrote the SEP entry on practical reason.

Practical reason involves personal reflection about moral principles, behavioral norms, and one's values. One considers a set of alternative possibilities for action, none of which has yet been performed, and decides what one ought to do, or what it would be best to do.

In arguments reminiscent of Kant, Wallace argues that practical reason must be separated from theoretical reason, which is about explanation and prediction.

Having the capacity to reflect and decide has implications for the structure of the will.

Wallace concludes that a "volitionalist" position is required, as opposed to the conventional "naturalist" view. Naturalism, he claims, does not have the capacity - the reflective self-control - required for moral agency, which requires the ability to choose independent of one's conscious desires, inclinations, yearnings, and various longer-term dispositions.

This power of choice, he says, is primarily a first-personal, deliberative phenomenon.

Wallace's power of reflective self-control seems to be "reasons-responsive" in John Fischer's and Susan Wolf's sense. This power makes agents morally responsible.

The powers of reflective self-control are forms of general competence or capacity. They involve the capacity to grasp and apply the reasons expressed in moral principles, and to control what one does by the light of one's moral understanding. My first question is the following: what does this capacity for control imply about the structure of the will?

The answer, I would suggest, is that the capacity for control is best understood in what might be called volitionalist terms. According to the volitionalist position I have in mind, the will is not merely a susceptibility to motivating states of desire, but a capacity for active self-determination. Underlying this interpretation is the assumption that our motivations divide fundamentally into states of two different kinds. There are, first, motivations with respect to which we are basically passive, such as conscious desires, inclinations, yearnings, and various longer-term dispositions. I shall call such motivations given desires. Second, there are motivations that are not merely given, but that directly express our activity as agents, such as choices, decisions, and intentions to act. The power of self-determination relevant to moral responsibility, I maintain, is a capacity for motivations of this second, volitionalist kind.¹¹

11. See my Responsibility and the Moral Sentiments, sect. 5.2.

Let us take a closer look at this contrast between different kinds of motivating state, starting with given desires. These motivations are rather like sensations, in that they are presented to us in experience rather than being things that we ourselves do. Of course, desires and sensations differ in a number of respects. Most importantly, desires are conceptually structured in a way sensations are not; they have propositional objects, for instance, and are typically accompanied by evaluative thoughts. The conceptual structure of given desires makes it possible for them to respond to our deliberated judgments about what we have reason to do, much as Aristotle thought that the appetitive and other desires of the virtuous would obediently fall in line with their rational verdicts about the good.¹² But as Aristotle would be the first to admit, given desires are not necessarily sensitive to judgments about the good in this way.¹³ Furthermore, even when our given desires respond to our reasoned verdicts about action, their doing so is not something that is directly under our control. In this respect, given desires are not to be classified as voluntary phenomena.

12. This aspect of desires is emphasized by Angela M. Smith, in 'Identification and Responsibility', in Ton van den Beld, (ed.), Moral Responsibility and Ontology (Dordrecht: Kluwer Academic Publishers, 2000), 233 46.
13 See Watson, 'Free Agency', as reprinted in Watson (ed.), Free Will, 96-110.

Generalizing, we may say that the class of given desires encompasses all those motivating states that are potential objects of self-awareness in the processes of reflection leading up to a determination to act. The 'given-ness' or 'passivity of such desires can thus be traced to the structure of reflective consciousness. ¹⁴ Given desires are states we take account of in deliberating about what we have reason to do (such as being attracted to the prospect of going to a party tonight), and they are also states we find ourselves in after we have reached a settled verdict through such reflection (such as remaining attracted to the party option even after one has concluded that it would be better to put in an appearance at the philosophy convention). But they are not the only kinds of motivation of which we are capable. There are, in addition to the sorts of desires that we can reflect on in deliberation, the distinctively active states of intention, choice, or decision, which characteristically bring deliberation to a final conclusion.

14 On this point, compare Korsgaard, The Sources p. 92 lecture 3, esp.92-4.

The volitionalist approach that I favor thus turns on the distinction between these active motivations and the given desires to which we are subject. In saying this, I do not mean to imply that all of the motivations conventionally referred to as desires are — like given desires — states with respect to which we are basically passive, objects that are presented to reflective consciousness. In philosophical practice it is customary to operate with an extended concept of desire as pro-attitude, encompassing not only the phenomenologically significant states of attraction I have called given desires, but also such volitional states as intentions and choices, as well as evaluative and normative judgments.¹⁵ I have no quarrel with this extended usage, as long as it is kept in mind that the various motivating states collected under the term desire have different roles to play in the processes of deliberative agency.

15 Compare G. F. Schueler, Desire (Cambridgc, Mass.: MIT Press, 1995), ch. 1.

In particular, it is important to be clear that volitional motivations are independent from our given desires, in the following sense. What an agent chooses or intends to do is not a function of the given desires to which the agent is subject at the time. Indeed, because our given desires as a class are not necessarily responsive to deliberative reflection, we need a capacity for motivation that is independent of those desires if we are to possess the powers of reflective self-control necessary to moral responsibility. The potential unresponsiveness of given desires shows itself in the fact that they are the kinds of psychological states that present us with temptations to violate the requirements of reason. We will be in a position to act rationally in the face of temptation, then, only if we are equipped with a basic capacity for self-determination in the face of competing desire, as it were. A general capacity for choice of this kind, for instance, is what we implicitly impute to blameworthy agents when we assume that they have the power to resist the immoral desires to which in fact they give in. That persons possess this form of competence is thus a necessary condition for their being directly subject to moral requirements that they flout. ¹⁶

16 It is not alone sufficient, of course: in addition, the person must have the capacities for moral understanding and reasoning that are also included under the powers of reflective self-control.

This same assumption is reflected in our deliberative practice as agents. In thinking about what to do, we take it for granted that we have a capacity for self-determination that outstrips the desires that are objects of deliberative reflection. The discovery that one is strongly attracted to the option of staying in bed, for instance, does not yet answer the practical question of whether that is what one is going to do. Indeed, it seems a discovery of this kind cannot answer the practical question if our activity as practical reasoners is to be fully intelligible. In deliberation, we try to determine what to do by reflecting on the question of what there is reason for us to do. This whole process makes sense, however, only on the assumption that we are capable of determining ourselves to act in ways that align with our deliberated verdicts about what we ought to do. But as we have seen, it is in the nature of our given desires that they are not automatically in alignment with our verdicts about our reasons. It follows that our competence to comply with the conclusions of moral deliberation involves the power to choose what we shall do in ways not laid down by the desires to which we are passively subject. ¹⁷

17 For persons whose given desires were always automatically in alignment with their deliberated verdicts, the volitionalist power of choice might not be a condition of moral agency and accountability in this way. But we are not such persons.

My next question is this: given that moral agency requires the capacity for volitionalist motivation, how should we regard the various approaches to the ontological issues that I sketched at the start of this contribution? Let us take the naturalist approaches first. Clear examples of this kind of approach are theories that accept what I earlier referred to as psychological determinism, the thesis that our actions are causally determined by our beliefs together with our given desires. ¹⁸ If this thesis is accepted, it seems to me, then we shall have to conclude that there are no moral agents in the sense I have been talking about. Such agents are distinguished by, among other things, the capacity for volitionalist motivation, and this cannot be reconciled with the thesis of psychological determinism. Accepting this thesis, naturalists would seem to have two options: either they may deny altogether the distinction between given desires and volitions that I have maintained is central to our understanding of ourselves and others as accountable agents. Or, they may admit some surface distinction between these two classes of motivation, but then offer a reductionistic account of volitionalist motivations, analysing these as combinations of beliefs and given desires, or insisting that they are themselves determined causally by such beliefs and desires. ¹⁹

18 Strictly speaking, psychological determinism would require that there be exceptionless laws linking these psychological states with the intentional actions and bodily movements they cause. For further discussion of psychological determinism, see Christopher Mark Grau, 'Moral Responsibility and Wolf's Ability', in van den Bell (ed.), Moral Responsibility and Ontology, 129-42
19 Even those who reject psychological determinism can maintain, of course, that such volitional states as intention and choice in some sense cause the actions we perform (see the following section for further discussion of this point). What they deny is that these volitional states can be either reduced to, or causally explained by, combinations of beliefs and given desires.

These characteristically empiricist strategies seem to leave no room for the genuine capacity for self-determining agency. Understood in volitionalist terms, this capacity involves the power to choose what one does independently of the given desires to which one is subject; but psychological determinism in the forms just mentioned amounts to the denial that persons are equipped with such a power. Defenders of the naturalist approach traditionally insist that persons could have done otherwise, even in a world in which their actions are determined by their given desires, insofar as they would have acted otherwise, had they been subject to different configurations of desire. ²⁰ But conditionalist analyses of this variety do not capture the distinctive kind of capacity that seems to be required for moral agency, which is a capacity for choosing in ways that are independent from the given desires to which one is actually subject.

20 See e.g. Donald Davidson, 'Freedom to Act', as reprinted in his Essays on Actions and Events (Oxford: Clarendon Press, 1980), 62-81.

Considerations of this kind have led some philosophers to reject naturalist accounts in favor of the third kind of approach I sketched at the start of this paper, the theory of agent-causation. This theory begins from the thought, which I have endorsed, that the capacity for choice is an active power of persons to determine what they shall do independently of the desires to which they are subject. In exercising this capacity, it seems, we ourselves fix an answer to the question of what we are going to do, in a way that cannot be traced back to the operation of psychological states and forces in us. Proponents of agent-causation conclude that persons are equipped with a special causal power different in kind from the ordinary form of causality linking events in the natural world, a power that is effective precisely at those points at which such ordinary causal relations break down. The hallmark of this approach is thus the contention that our ontology can accommodate moral agents only if and to the extent that we need to postulate such agents in order to explain events that cannot be accounted for in ordinary scientific terms.

Now I share the sense of many critics that this way of thinking about moral agency is incredible. The theory of agent-causation is problematic, because it interprets the capacity for choice as a causal power of persons whose rationale lies in its direct contribution to explaining what goes on in the natural world. The theory holds that agency is possible only on the condition that there are natural events (such as firings of neurons in a person's brain ²¹) that cannot be explained causally in terms of other events, but that instead must be accounted for by appeal to persons. This is to understand the human capacity for choice as part of an explanatory theory of the world, something we are led to postulate by our interest in causal explanation and prediction—an interest that is definitive of what might be called the theoretical point of view. But this is not the only way to see matters. On the practical reason approach sketched at the start of this paper, the power of choice has its natural home in the context of the practical point of view, a distinctively first-personal perspective defined by our interest in the deliberative question of what we ought to do.

21 Compare Chisholm, 'Human Freedom and the Self', 29.

Understood in this way, the practical point of view is the standpoint of practical reason, a standpoint we adopt when we reflect about what we have reason to do. To say that the power of choice has its natural home in the context of this perspective is to say that it is our preoccupation with deliberative questions of practical reason, and not our interest in explanation and prediction, that leads us to postulate that agents are equipped with this power.²² We inevitably assume that we ourselves have this capacity when we undertake to deliberate about what we should do. As I said above, the idea that we have it in our power to determine what we shall do in ways independent of our given desires provides the natural context for our own deliberative activity. In this sense, the power of choice is primarily a first-personal, deliberative phenomenon. ²³

22 For a different but complementary justification of the capacity for decision by reference to its role in the context of practical deliberation, see Michael E. Bratman, Intention, Plans and Practical Reason (Cambridge, Mass.: Harvard University Press, 1987).
23 This is not to say that it is an essentially private activity — there is, for one thing, the utterly familiar activity of collective deliberation, conducted from the perspective not of I' but of 'we'. For another, I assume that our reasons must be capable of being made intelligible to other persons whose deliberative problems are very different from our own. This is the role of advice, which involves a kind of hypothetical deliberation, detached from the corresponding capacity for self-determination that accompanies practical reflection in the first person. One asks what a given person in a given situation would have reason to do, where the person in question is not oneself, and the situation described is not one's own.

The proponents of agent-causation correctly grasp that moral agency cannot be rendered fully intelligible unless we abandon the thesis of psychological determinism. What they fail to see is that the phenomenon of moral agency also cannot be rendered fully intelligible without articulating an alternative to the theoretical standpoint of explanation and prediction. Theorists of agent-causation interpret the insight that persons have the capacity for active choice as a thesis about their causal powers, where causal powers are in turn understood as items whose significance is primarily explanatory. In effect, it is assumed that agency will be possible only if it contributes positively to the explanation and prediction of events that fall within the purview of scientific theorizing, but are inexplicable in terms of such theorizing. Agent-causation theorists thus maintain that the activity of persons presupposes that the chains of causation linking ordinary events are inherently gappy, the gaps marking points at which explanation requires an irreducible reference to an agent, construed as a kind of unmoved mover. ²⁴ The resulting picture extracts from the familiar phenomenon of self-determination an ontological commitment that is hard to take seriously, suggesting that our status as agents confers on us a divine prerogative of intervention in the causal order of nature from a position mysteriously outside it.

24 See Chisholm, 'Human Freedom and the Self', 32.

Difficulties of this sort seem likely to persist so long as we continue to interpret agency as an essentially explanatory phenomenon. To avoid them, we could follow the naturalist route sketched above, which would mean abandoning the volitionalist conception of motivation. Alternatively, we could attempt to rescue the volitionalist conception by removing it from the context of theoretical reason and placing it within the normative framework of practical reason. Viewed in this new light, the power of choice is not understood as a causal force in competition with other forms of causality, something for which a space would need to be cleared in the order of nature. Rather it speaks to the normative interests regulative of the practical point of view, a standpoint into which we project ourselves when we deliberate about what we ought to do. Deliberation requires us to assume that the question as to what we are going to do is not settled by our given desires, but that we have it in our power to determine the answer to that question for ourselves. And the fact that the capacity for self-determination is presupposed by practical reason in this way is the point and rationale of the volitionalist conception.