SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. 1. Doxastic Freedom
  4. 2. Freedom of Action, Voluntary Control and Doxastic Freedom
  5. 3. Freedom of Intention and Doxastic Freedom
  6. 4. Conclusion
  7. References

This paper defends the possibility of doxastic freedom, arguing that doxastic freedom should be modelled not on freedom of action but on freedom of intention. Freedom of action is exercised by agents like us, I argue, through voluntary control. This involves two conditions, intentions-reactivity and reasons-reactivity, that are not met in the case of doxastic states. Freedom of intention is central to our agency and to our moral responsibility, but is not exercised through voluntary control. I develop and defend an account of freedom of intention, arguing that constitutive features of intention ensure that freedom of intention cannot require voluntary control. Then I show that an analogous argument can be applied to doxastic states. I argue that if we had voluntary control of intentions or of doxastic states, this would actually undermine our freedom.

Are our doxastic states free? This question appears to matter for the possibility of epistemic responsibility. But it is not clear what the criteria for doxastic freedom should be. One approach is to try to model doxastic freedom on freedom of action. I will argue that the non-voluntary character of our doxastic states is a major stumbling block for such an approach. I will develop and defend an alternative approach, modelling doxastic freedom on freedom of intention. This is a form of freedom that, for principled reasons, is exercised other than through voluntary control. I will argue that if we had voluntary control of intentions or of doxastic states, this would actually undermine our freedom.

1. Doxastic Freedom

  1. Top of page
  2. Abstract
  3. 1. Doxastic Freedom
  4. 2. Freedom of Action, Voluntary Control and Doxastic Freedom
  5. 3. Freedom of Intention and Doxastic Freedom
  6. 4. Conclusion
  7. References

1.1. Doxastic Freedom and Epistemic Responsibility

When we act in the world with our bodies, we often exercise a kind of freedom—or so we suppose. Freedom of action1 appears to be a ground of the possibility of moral responsibility for our actions. To say that we are morally responsible for our actions is to say not only that our actions are evaluable as morally good or bad, but that we are subject to moral demands or obligations when we act, that it is up to us to meet those demands or obligations, and, perhaps, that we are apt to be praiseworthy or blameworthy according as we meet or fail to meet them in our actions.

Many philosophers hold that we are epistemically responsible for our doxastic states—for our beliefs and withholdings of belief.2 On this view, we are subject to epistemic demands or obligations with respect to our doxastic states—obligations not to believe propositions for which we lack adequate evidence, for example. It is up to us to meet those demands or obligations, and we are, perhaps, praiseworthy or blameworthy according as we meet or fail to meet them in our believings and withholdings.3

If freedom of action is a ground of the possibility of moral responsibility, then, it might be thought, freedom of doxastic states is a ground of the possibility of epistemic responsibility. It is commonly accepted that we enjoy freedom of action and are morally responsible for our actions (even if the nature of freedom of action and moral responsibility are matters of dispute). But it is far from obvious that we enjoy freedom of doxastic states—or ‘doxastic freedom’, as I will call it. Thus, it is not clear that we can be epistemically responsible for our doxastic states.

In this paper I will argue that we have to recognise two different forms of freedom—or two different ways in which freedom can be exercised. If we try to model doxastic freedom on freedom of action, then we will find that we lack doxastic freedom. But it would be a mistake to try to model doxastic freedom on freedom of action. I will offer an alternative account, on which doxastic freedom is analogous to freedom of intention. Freedom of action is exercised through voluntary control; doxastic freedom and freedom of intention are not.

For convenience, I will take for granted that we do indeed enjoy freedom of action. Perhaps this commonly held view is false; perhaps we lack freedom of any kind. My aim is not to establish with certainty that we have doxastic freedom. My aim is to provide a characterisation, perhaps partial, of what doxastic freedom would be, and to argue that there is no special reason for denying the existence of doxastic freedom, as distinct from any other form of freedom.

1.2. Measuring Doxastic Freedom

Our doxastic states are free, I will take it, if we exercise freedom in the normal process of forming, retaining, revising and extinguishing our doxastic states. I will call this the process of regulating our doxastic states. Regulation can take place by means of conscious mental events such as judgments, as when you form the belief that p by judging p. It can, but need not, involve reflective reasoning about doxastic states and their contents. It can also occur without the occurrence of any conscious mental events, such as when you automatically believe what you see, without the occurrence of any event of judgment. Even in these automatic cases, the process of regulation engages with the level of conscious judgment in a certain way. The beliefs formed are typically ones whose contents would be endorsed in judgment by the subject if she were to consider them, and they are disposed to be regulated by events at the level of conscious judgment. You see what looks like a dog, and automatically form the belief that there is a dog in front of you; then you remember that your neighbours own a large cat that can look like a dog, and you judge that what you see may not be a dog, thus extinguishing the belief you had formed automatically.

What would it be for us to exercise freedom in regulating our doxastic states? It is not obvious.4 One way to approach this issue is to consider how we exercise some form of freedom that is more familiar as such. The most familiar, and philosophically discussed, form of freedom, is that which we exercise in our free actions. A natural starting-point, then, is to examine how we exercise freedom in our actions, and see if the features that constitute the exercise of freedom in the case of free action are also present when we regulate our doxastic states.

It might be objected that the criteria for doxastic freedom may differ from the criteria for freedom of action. Matthias Steup presses just this objection:

“Assessing the freedom of actions calls for one yardstick, assessing the freedom of doxastic attitudes for another. It’s a mistake to think that the freedom of actions and the freedom of doxastic attitudes can be gauged using one single yardstick.” (Steup 2008, 389–90.)

Ultimately, I will argue that Steup’s point here is correct, though not for the reasons he offers. But I think that it is nevertheless helpful to begin with freedom of action. That’s because we do not have much of an independent grip on doxastic freedom; we are not going to be able to find criteria for doxastic freedom without starting from something more familiar. By considering freedom of action, we can come to see what one model of doxastic freedom would look like, and to get a sense of what alternative models there might be.

I will not here be doing anything so ambitious as pursuing a general account of freedom, of freedom of action, or of doxastic freedom. Rather, I will try to identify some necessary conditions on how agents like us exercise freedom in certain domains. When I talk of ‘criteria’ for freedom, it should be understood in this sense. I want to show that there are no special problems for seeing ourselves as free in the doxastic domain.

2. Freedom of Action, Voluntary Control and Doxastic Freedom

  1. Top of page
  2. Abstract
  3. 1. Doxastic Freedom
  4. 2. Freedom of Action, Voluntary Control and Doxastic Freedom
  5. 3. Freedom of Intention and Doxastic Freedom
  6. 4. Conclusion
  7. References

We exercise freedom of action only when we control what we do.5 We control what we do by means of the will. It is only by acting in accordance with, or guided by, our wills, that we act freely. So much is truistic, I take it, given a suitably liberal understanding of ‘the will’. But what is involved in acting in accordance with, or guided by, your will?

There is a tradition in philosophy of understanding the will as constituted by certain pro-attitudes. According to this tradition, freedom of action crucially involves acting in accordance with, or guided in the right way by, those pro-attitudes. Historically, the pro-attitude of desire took centre stage in this tradition; more recently, philosophers have tended to focus on the pro-attitude of intention.6 It is on intentions that I will concentrate.

On this view, then, the paradigmatic way to act according to your will is to act as you intend to act, because you so intend. Here we have a partial conception of the exercise of freedom of action.

This conception might be regarded as too demanding. Arguably, we often act freely, and indeed intentionally, without forming any intention to act in the particular way we do. Think of habitual actions that are performed thoughtlessly, and automatic actions that are performed during some skilled activity, such as engaging the clutch while driving.

Even in these habitual and automatic cases, however, it seems that there is a dispositional involvement of intentions. That is, were you to form some relevant intention—to perform the action, or not to perform it, say—then that intention would be efficacious in bringing you to act accordingly, barring the intervention of some abnormal factor such as external manipulation or paralysis. (This latter qualification is intended to deal with Frankfurtian cases in which you act freely even though you could not have acted otherwise.7 For brevity I will sometimes omit it in what follows.) When you engage the clutch, you may not have formed any intention to do so. But it is part of the action’s being under your control that, had you intended to keep the clutch disengaged, you would have done that instead. Likewise, if you had formed an intention to engage the clutch more or less quickly, or a second earlier or later, you would have acted accordingly. If these counterfactuals do not hold, then your action is unreactive to your intentions. In that case, you could not be said to be in control of what you are doing, or to be acting freely.

It seems, then, that the exercise of freedom of action by agents like us is subject to a requirement of intentions-reactivity. This is a requirement on an act of Φing, such that, across a range of counterfactual scenarios in which you have some suitably related intention to perform a particular action, such as an intention not to Φ, you act as specified by that intention. In sum, even if free action is not always intended action, agents like us act freely only when exercising a capacity to do what we intend to do.

Doxastic states are not intentions-reactive. Intentions to be in particular doxastic states are not normally involved, even dispositionally, in the process of regulating our doxastic states. Such intentions have no significant role in our doxastic lives. If we were to form such intentions, they would normally be ineffective. Suppose you intend to believe that it is raining. If the evidence plainly fails to indicate that it is raining, you will not be able to form the belief that it is raining. If the evidence does indicate that it is raining, then you may form the belief that it is raining, but you will do so because of the evidence, not because of your intention.

Doxastic states fail even a very weak version of the intentions-reactivity requirement, according to which an act or state of Φing counts as free only if acts or states of that kind are normally intentions-reactive.

The point is not that intentions have no effect on our doxastic states. The point is that intentions do not govern our doxastic states as they govern our actions. You may have an intention to form a belief about the weather, and you may, as a result of executing that intention, come to believe that it is raining. But the formation and retention of that particular doxastic state, the belief that it is raining, will not be intentions-reactive. In the case of actions, by contrast, intentions can control which particular actions we perform, and not just which more general types of action we perform.

Perhaps we can sometimes bring ourselves to form particular beliefs as an indirect result of intending to form those particular beliefs. The intention, for example, can cause us to attend selectively to the evidence, and engage in various other forms of self-manipulation. This is not what we normally do, and it is not something we can succeed in doing with any reliability. Such cases will not form the paradigm for a kind of freedom that we normally exercise in regulating our doxastic states.

Intentions-reactivity is a condition on the exercise of freedom of action by agents like us. Doxastic states are not intentions-reactive. Should we conclude that we have no doxastic freedom? That depends on whether intentions-reactivity is an appropriate ‘yardstick’ by which to measure doxastic freedom.8 In order to begin pursuing the question whether it is an appropriate yardstick, we can consider intentions-reactivity in the light of another putative condition on our exercise of freedom of action.

It is helpful here to turn to another kind of approach to freedom: one that emphasises the connection between the will and reason. In this tradition, the emphasis is not on desires or intentions; rather, the will is seen as a capacity to act as you see most reason to act, or as you judge there to be most reason to act.9 Reasons themselves are construed not as attitudes of the subject but as considerations that favour acting in certain ways. (Of course, a consideration can be seen as, or judged to be, a reason for doing something, when it is in fact not so. It is often supposed, however, that freedom requires the subject to be receptive, to some extent, to the reasons she in fact has; this will be important later on.)

Here, then, we have a second partial conception of the exercise of freedom of action. According to this conception, the paradigmatic way to act according to your will is to act as you see, or judge there to be, most reason to act, because you see or judge there to be most reason to act in that way.

Again, this conception may be regarded as too demanding. Arguably, some habitual, thoughtless or automatic actions are not performed for any reason. But even in these cases, there will be a dispositional involvement of reasons. Suppose you break a glass and (freely) swear in frustration. Perhaps you do not thereby act for a reason. But it is part of that action’s being under your control that, had you seen or judged there to be reason not to swear, or reason to swear differently, you could have reacted to that reason by acting differently.10 If this condition does not hold, then your action is unreactive to your reasons as you see them. In that case, you could not be said to be in control of what you are doing, or to be acting freely.

There are various kinds of reason that we can recognise or perceive as favouring acting in particular ways: moral reasons, prudential reasons, aesthetic reasons, etc. Our control of our actions is such that we are capable of reacting to any kind of reason that we recognise as favouring acting in some way, by acting accordingly. There is no principled restriction, among the reasons we can recognise, on the kinds of reasons we can react to in action.

It seems, then, that the exercise of freedom of action by agents like us is subject to a requirement of reasons-reactivity.11 This is a requirement on an act of Φing, such that, across a range of counterfactual scenarios in which you see most reason to Φ or not Φ, or to ΦF-ly, you act accordingly, no matter what kind of reason it is that you see.12 Even if free action is not always action for reasons, agents like us act freely only when exercising a capacity to act for (perceived) reasons in this unrestricted way.

I have outlined two apparent conditions on the exercise, by us, of freedom of action: intentions-reactivity and reasons-reactivity. Are these two conditions connected?13 It would seem so. Consider what intentions do. You form an intention to Φ at t in order that, when t comes, you will Φ.14 In the case where your intention is consciously formed, you come to form the intention by considering what to do. And you typically do this by considering your reasons for doing one thing or another—your reasons to Φ at t versus your reasons to Ψ at t, say. (Agents need not deploy the concept of a reason to do this. They may simply ask themselves what to do, and rely on considerations that are in fact reasons in order to decide.) Thus, intentions are paradigmatically formed as a result of recognition of reasons for action, and cause reaction to those reasons in action. Intentions translate recognition of reasons for action into the actions putatively recommended by those reasons. What’s more, any kind of reason—moral, prudential, aesthetic, etc.—for Φing can play a role in your coming to intend to Φ. Just as you can Φ for any kind of reason that you recognise as favouring Φing, so also you can intend to Φ for any kind of reason that you recognise as favouring Φing.

Failure to do what you intend will be failure to do what you see or judge there to be most reason to do, in so far as you intend to do what you see or judge there to be most reason to do. Intentions-reactivity of actions is a necessary condition of reasons-reactivity of actions, for planning creatures like us whose agency involves intentions.

I will say that an action is under voluntary control when it meets the conditions of reasons-reactivity and of intentions reactivity.15 Voluntariness is a matter of exercising a capacity to select and execute what you do on the basis of what you see or judge yourself to have most reason to do. For us, this capacity is often exercised by forming and executing intentions. This seems to me to capture what we ordinarily mean by ‘voluntary’.16

I claim, then, that we exercise freedom of action through voluntary control. Only by acting voluntarily do we act freely.

I have already noted that doxastic states are not intentions-reactive. We can also see that doxastic states are not reasons-reactive in the way that actions are. Consider:

  •  Someone offers you €500,000,000 to believe that the U.S. is still a colony of Great Britain.
  •  Someone offers you €500,000,000 to believe that it rained on Aristotle’s 30th birthday.17

In neither case can you win the reward (note that you are being asked to form an outright belief). In these cases, you may take yourself to have an excellent reason to have the target belief, but you will be unable to react to that reason by forming the belief.18 The reason you (take yourself to) have is a practical rather than epistemic reason. As it happens, it is a financial reason, but that is not crucial: you would also be unable to form the target belief because it would make you happy, would save the world, or whatever. What these and similar cases illustrate is that we are, in our doxastic lives, systematically unreactive, or only very restrictedly reactive, to practical reasons. In this respect, doxastic states stand in contrast with actions, which are reactive to any kind of reason you can recognise.

The point is not that our doxastic states cannot be influenced by practical considerations. Of course they can be so influenced. For example, our beliefs about ourselves tend to be biased in ways that are self-serving. The point is that such considerations typically cannot be reasons for which we hold beliefs—considerations whose probative force we can acknowledge in deliberation about what to believe, and form beliefs in reaction to.19

The failure of doxastic states to be intentions-reactive is closely linked to their failure to be reasons-reactive. Intentions typically ensure that we do what we have determined ourselves to have most overall reason to do. But we don’t regulate our doxastic states by considering what particular doxastic states we have most overall reason to be in. We regulate them primarily by considering evidence for their contents. We don’t have to ask ourselves the further question whether to believe what the evidence recommends, rather than, say, what will make us happy. There is no role for intentions to be in particular doxastic states, in the process of regulating our doxastic states.

In sum, we do not have voluntary control over our doxastic states. Our doxastic states are not under the command of the will in the way that our actions are. We do not exercise doxastic freedom in the way that we exercise freedom of action.

I have shown that we do not control our doxastic states as we control our actions. This salient difference between control of doxastic states and control of actions cannot be ignored by any adequate defence of doxastic freedom. In particular, I think it is a major stumbling block for any attempt to model doxastic freedom on freedom of action.

One might try to employ some conception of freedom of action that makes no mention of intentions-reactivity or reasons-reactivity. But, as we have seen, it is very plausible that intentions-reactivity and reasons-reactivity are deeply involved in our exercise of freedom of action. The strong presumption must be that the failure of doxastic states to meet those conditions counts against their being free in the same way that actions are.

What’s more, the difference is likely to show up, in other forms, within at least some other conceptions of freedom of action. For example, suppose we adopt a conception of freedom that makes crucial appeal to the notion of doing what you desire to do, perhaps including harmony with higher-order desires (Frankfurt 1971). Well, in the cases described above, in which you are offered a financial reward to adopt some belief, you will probably desire very strongly to adopt the belief, and there is no reason to suppose that that desire won’t be in harmony with your higher-order desires. Still, your doxastic states will fail to respond.

We could, instead of modelling doxastic freedom on freedom of action, state criteria for doxastic freedom that do not include voluntary control, and that are different to the criteria for freedom of action. This approach faces the challenge of offering some non-question-begging justification for treating doxastic freedom as something whose criteria differ from those for freedom of action. Without such a justification, this approach will be open to the accusation that it is simply equivocating on the term ‘freedom’; or at least that doxastic freedom is so radically different from freedom of action that it is not fit to ground the possibility of epistemic responsibility, as freedom of action is supposed to ground the possibility of moral responsibility.20

I think that existing defences of doxastic freedom fail. Either they try to claim that doxastic states are free in just the same way that actions are free, and thus fail to do justice to the fact that we don’t control our doxastic states as we do our actions. Or, they propose criteria for doxastic freedom quite different to the criteria for freedom of action, without showing that what they are giving an account of is really a kind of freedom.21

In the next section, I will propose and defend the view that doxastic freedom should be modelled on freedom of intention. I will argue that this is genuinely a model of freedom, and that doxastic freedom so construed can plausibly ground epistemic responsibility.

3. Freedom of Intention and Doxastic Freedom

  1. Top of page
  2. Abstract
  3. 1. Doxastic Freedom
  4. 2. Freedom of Action, Voluntary Control and Doxastic Freedom
  5. 3. Freedom of Intention and Doxastic Freedom
  6. 4. Conclusion
  7. References

3.1. The Need for Freedom of Intention

I have been emphasising the role that intentions have in the exercise of freedom of action by agents like us, and thus in our moral responsibility. But it is a familiar thought that your actions can be controlled by your intentions, and yet you fall short of full moral responsibility for your actions. Consider, for example, an agent who compulsively forms intentions to pick up red objects, intentions which he then executes normally. Or consider an agent whose intentions are, unknown to him, controlled by some external agent by means of a neural implant. Consider, finally, an agent who forms intentions on the basis of his own consideration of reasons, but whose grasp of his reasons is so utterly unsystematic and incomprehensible that we would consider him insane. For example, Fischer and Ravizza (1998, 65) describe a case in which an agent slays his fellow passengers on a boat with a sabre, and is so disposed that the only consideration he would recognise as a reason not to slay his fellow passengers is that there is a passenger in the lower cabin smoking a Gambier pipe. In such a case, unless further investigation could yield some way of understanding this agent’s take on his reasons (perhaps in terms of some idiosyncratic set of values or background beliefs), the agent is not fully (if at all) morally responsible, but rather insane.

In some of the kinds of cases just described, the agent will count as acting voluntarily. He exercises a capacity to select and execute actions on the basis of what he sees or judges himself to have most reason to do. But in all of the cases described, the agent lacks something in the way of freedom. The intention on the basis of which the agent acts is not free. That helps explain why the agent is not fully morally responsible, even if he enjoys some weaker form of responsibility.

Now, we might say that, if you act on the basis of an unfree intention, then your action is itself unfree, even if it may be voluntary. Or we might say that, in such cases, you may act voluntarily and freely but you lack a free will.22 I don’t mind which of these things we say. For my purposes, the important point is that freedom of intention is something that goes beyond voluntary control of action, and is a form of freedom that is important in our lives, because necessary for full moral responsibility.

I have taken for granted that we have freedom of action. It might seem more controversial to suppose that we have freedom of intention. But I don’t think it should be so controversial. In defence of this I offer two considerations. First, as I have just argued, freedom of intention is a necessary condition on full moral responsibility for actions, and I take for granted that we can be fully morally responsible for our actions (the alternative is bleak, to say the least). Second, we hold people responsible directly for their intentions, and not only for the actions that flow from them.23 For example, suppose your comrade in war forms the intention to betray your position, believing that the enemy will offer him certain rewards to do so. Your comrade may never receive the expected contact from the enemy, and thus may never execute his intention. Nevertheless, you will blame and resent your comrade for his treacherous intention. It seems very plausible that part of what grounds your comrade’s responsibility for his intention is that he holds it freely. If his intention were compulsive, or externally imposed, or the result of insanity, he would not be responsible for it and you would not blame and resent him. Freedom of intention seems to ground responsibility for intentions, just as freedom of action is a ground of responsibility for action.24

I think that these connections between freedom of intention on the one hand, and freedom of action and moral responsibility on the other, also make clear that freedom of intention does indeed merit the name of freedom. Freedom of action and freedom of intention are species of the same genus. If freedom of intention were not a species of freedom, then it is quite unclear how it could play the role it does in grounding full moral responsibility, or why we should think that there is an absence of freedom in the cases I described at the start of this subsection.

3.2. Freedom of Intention and Voluntary Control

How is freedom of intention exercised? Unlike freedom of action, freedom of intention is not exercised through voluntary control.

This is demonstrated by Kavka’s toxin puzzle (Kavka 1983). You are offered a large reward to intend today to drink a mild toxin tomorrow. The toxin’s effects are very unpleasant but short-lived and harmless. You are offered nothing to actually drink the toxin. Whether you win the reward or not depends entirely on whether you have the requisite intention today. Can you react to the prospect of the reward by intending to drink the toxin? It seems not. You cannot intend to drink it, knowing that when the time comes you will have have no reason to drink it and strong reasons against doing so. You would need some reason for actually drinking it, in order to be motivated to intend to drink it.25

This and similar cases show that we are not, in our intendings, reactive to any kind of reason that we can recognise as favouring having certain intentions. In particular, we are very restricted in our reactivity to reasons that do not favour acting in one way or another, but merely favour intending to act in one way or another. The prospect of a reward for intending to drink the toxin is an example of such a reason. Intentions are not reasons-reactive in the way that actions are.26

Correspondingly, intentions are not themselves intentions-reactive. Intentions to have particular intentions are not normally involved, even dispositionally, in the process of regulating our intentions. If we were to form such intentions, they would normally be ineffective. You cannot, for example, form at t1 an intention to form at t2 an intention to Φ, and then at t2 execute the former intention by forming the latter one.27 This failure of intentions-reactivity is not surprising. Intentions-reactivity has a role when you are determining whether to do something on the basis of whether you have most overall reason to do it. We don’t regulate our intentions by considering what particular intentions we have most overall reason to have. We regulate them primarily by considering the prospective objects of our intentions—by considering what to do, in view of our reasons for doing one thing or another. We don’t have to ask ourselves the further question whether to intend to do what we have most overall reason to do.28

Thus, freedom of intention is a form of freedom that we exercise otherwise than through voluntary control. In what follows I will offer more of a positive characterisation of freedom of intention, and argue that an analogous account can be given of doxastic freedom.

3.3. Characterising Freedom of Intention

To work towards a positive characterisation of freedom of intention, we can consider what is lacking in the case of an agent who acts voluntarily but is not fully morally responsible. Take the crazed killer, who does what he intends, and does what he sees reason to do, but whose take on his reasons is so incomprehensible that his moral responsibility is diminished, at least. This agent, we can suppose, is reactive to the reasons he recognises. But he is profoundly disordered in his receptivity to reasons—that is, in his capacity to recognise the reasons that he in fact has. According to Fischer and Ravizza, it is this failure of reasons-receptivity that explains why the agent is not morally responsible. Fischer and Ravizza defend the view that moral responsibility requires regular reasons-receptivity, where this involves the agent’s displaying “an understandable pattern” (Fischer and Ravizza (1998), 71; emphasis in original) of recognition of reasons across actual and counterfactual scenarios. They further characterise this as follows:

“An agent’s reasons-receptivity must exhibit a suitable correspondence to the objective (relative to a given set of preferences, values, and beliefs) grading of the strength of reasons, in order for the pattern in the agent’s reasons-receptivity to be understandable. [...]

“When “reasons” do not connect and relate to each other in appropriate ways, they do not generate a minimally comprehensible pattern.” (Ibid., 72.)

I assume that Fischer and Ravizza are right that regular reasons-receptivity is necessary for full moral responsibility. What I want to claim is that freedom of intention does not require reactivity to reasons merely for intending, but does require regular receptivity to reasons for action, as well as reactivity in your intentions to recognised reasons for action. That is, it requires that, when you engage in regulating your intentions, you exhibit a regular pattern of recognition of reasons for action, as well as the capacity to implement that recognition in your intentions, by forming and retaining intentions to act as those reasons recommend, and by not forming or retaining intentions to act contrary to what those reasons recommend.29

I want to claim, further, that doxastic freedom is analogous to freedom of intention, in a way that I will explain.

I will spend the rest of this subsection arguing for the first of these claims, and the next subsection arguing for the second.

There is a distinction, we have seen, between reasons for intending to Φ that are such because they are reasons for Φing (e.g., a reward for drinking the toxin), and reasons merely for intending to Φ (e.g., a reward for intending to drink the toxin). This is an instance of a more general distinction between two kinds of reasons for being in propositional attitude states. Another instance is the distinction between epistemic and practical reasons for doxastic states—that is, the distinction between reasons for being in some doxastic state that are such because they pertain to whether its content is true or false (e.g., evidence for p), and reasons merely for being in some doxastic state, independently of whether its content is true or false (e.g., a reward for believing p).30 Another instance of the distinction would be: considerations that make X desirable, versus considerations that merely count in favour of desiring X.31

The general distinction can be made as follows. On the one hand, there are considerations that pertain to the world’s being as it is represented in the state’s content. These can be considerations that make the world’s being that way something to be enacted, considerations that make it likely that the world is that way, or considerations that make the world’s being that way desirable. On the other hand, there are considerations that merely pertain to the subject’s being in the state—considerations in the light of which the subject’s being in the state would be a good or bad thing.

I will call the first kind of reasons object-directed reasons. I will call the second kind state-directed reasons. What we have seen, with regard to both doxastic states and intentions, is that they are severely restricted in their reactivity to state-directed reasons.

My first claim, then, is that freedom of intention requires regular receptivity and reactivity to object-directed reasons for intentions, but does not require reactivity to state-directed reasons for intentions. Freedom of intention thus stands in contrast with freedom of action, which involves regular receptivity to reasons for action, and reactivity to any kind of reason for action to which the subject is receptive.

It might be said that this way of viewing the matter undermines my claim that there is an important difference between freedom of intention and freedom of action. Actions, after all, are not reactive to state-directed reasons, for the very good reason that there are no state-directed reasons for action. Freedom of action thus cannot require reactivity to state-directed reasons, any more than (according to me) freedom of intention does. Wherein, then, lies the difference?32

Primarily, the difference is that, in the case of intentions, there is a kind of (putative) reason to which we are (or can be) receptive, but to which we are not reactive, whereas, in the case of actions, freedom is exercised through reactivity to any kind of reason to which we are receptive. It would be a mistake to think that this is because freedom in general requires reactivity to object-directed reasons and not to state-directed reasons, and there are no state-directed reasons for action. Rather, the distinction between object-directed and state-directed reasons does not apply to actions at all. Reasons for actions are neither object-directed nor state-directed reasons for those actions. (They are, however, object-directed reasons for the corresponding intentions.) This fact does not undermine or trivialise the difference between freedom of intention and freedom of action. It is a further asymmetry on which that difference partly rests.

In fact, the relation of a reason for action to the action it recommends is, in an important respect, more like the relation of a state-directed reason for intention to the intention it recommends, than like that of an object-directed reason for intention to the intention it recommends. A reason for action is a reason that favours performing a certain action by indicating the desirability of performing it. This does not depend on the reason’s bearing on the object of the action (if any) or on some content represented by the action. This point is connected to another important aspect of the difference between freedom of intention and freedom of action—namely, the role of intentions themselves in exercising that freedom. We exercise freedom of action by determining what actions it would be desirable to perform, and by then ensuring through the mechanisms of decision and intention that we perform those actions. We do not control our intentions by determining what intentions it would be desirable to have (as indicated by state-directed reasons), and ensuring that we form and maintain those intentions. Because of this, there is no role, in the regulation of our intentions, for intentions to have particular intentions.

Thus, it seems to me that, by drawing the contrast between state-directed and object-directed reasons, we do not obliterate or trivialise the difference between freedom of intention and freedom of action. Rather we gain a fuller understanding of it. This difference is sufficient to mark two different species of freedom—one exercised through voluntary control, and the other not—but it is not so great that we cannot recognise them as members of the same genus, namely freedom. So I will argue.

On the view I am proposing, there are at least two types of situation in which an intention to Φ will not count as free. The first is a situation in which the agent is unreceptive to reasons for or against Φing. For example, like the crazed killer, the agent would recognise only some very specific and arbitrary consideration as a reason not to Φ. The second is a situation in which the agent lacks the capacity to react to the reasons she recognises as favouring or counting against Φing, by, accordingly, intending to Φ, not intending to Φ, or intending not to Φ. For example, the agent who compulsively intends to Φ lacks the capacity to react, in her intentions, to the reasons she recognises for or against Φing.

What this view denies is that an agent who is unreactive to the reasons she recognises as merely favouring intending to Φ will thereby count as unfree in her intentions with respect to Φing.

That is a statement of the view. Now I will state my argument for it, before going through the argument in detail:

  • 1
     Freedom of intention is freedom exercised in the process of regulating your intentions.
  • 2
     The process of regulating your intentions is, constitutively, directed towards a certain aim, namely the aim of acting correctly.
  • 3
     Freedom exercised in a process that is constitutively directed towards a certain aim is freedom in pursuing that aim when engaged in the process, rather than freedom with regard to whether to pursue that aim when engaged in the process.
  • 4
     Freedom of intention is freedom exercised in pursuing the aim of acting correctly, rather than freedom with regard to whether to pursue that aim when regulating your intentions. (From 1, 2, 3)
  • 5
     Freedom in pursuing an aim requires regular receptivity and reactivity to considerations that recommend conduct conducive to attaining that aim, but does not require reactivity, in pursuing the aim, to considerations that recommend conduct irrelevant to or counter to attaining the aim.
  • 6
     Object-directed reasons for intentions recommend conduct conducive to the aim of acting correctly; state-directed reasons often recommend conduct irrelevant to or counter to that aim.
  • 7
     Freedom of intention requires receptivity and reactivity to object-directed reasons, but does not require reactivity to state-directed reasons. (From 4, 5, 6)

Now I will go through the steps of the argument.

I have been assuming throughout that (1) is true. It is hard to see what else freedom of intention would be, if it were not freedom that is exercised in the process of regulating your intentions.

(2) may seem more controversial. But I think the difficulty is mainly to do with how we should spell out the notions of aim-directedness and of correct action. Fortunately, I can largely sidestep these thorny questions, and rely on what I take to be the very plausible core point.

To bring out this plausibility, let me first note that the process of regulating your intentions is a process of selecting, coordinating and controlling how you act. If you are not engaged in selecting, coordinating and controlling how you act, then you could not be said to be regulating your intentions, since intentions are states that represent your selection of how to act, and that you are disposed to execute.

This process is certainly aim-directed. It aims, first, to determine how you act. If you are not aiming to determine how you act, then you are simply not engaged in that process. But if you are aiming to determine how you act, then you are also aiming to act in the way that you determine; otherwise, you wouldn’t be aiming to determine it. Thus, the process of regulating your intentions has as an aim that you do whatever it is that you ultimately form the intention to do.

But it would be wrong to say that, when we engage in this process, we aim merely to do whatever we happen to form an intention to do. If that were the aim, then we could satisfy it just by forming some arbitrary intention, which we go on to execute. We don’t form intentions arbitrarily. We care about or are sensitive to what we ought to do, or what would be correct to do. We aim not merely at acting as intended, but at acting correctly.33

By ‘correctness’ here I mean to express a notion that is familiar (albeit the terminology for expressing it varies) and (I think) essential, but notoriously difficult to explicate in an illuminating way. It is the notion of what is objectively a right or good thing for you to do, or what you are objectively permitted to do, in a given situation. It is the property that we are looking for in an action, when we ask ourselves whether to perform that action. This is not specifically moral correctness, but all-things-considered correctness. Moral considerations are one kind of consideration that contribute to determining or indicating what is correct to do; there are also, for example, prudential considerations and aesthetic considerations. In asking yourself what to do, you might satisfy yourself that, morally, the thing to do is to Φ, while, prudentially, the thing to do is to Ψ. But you will still face a further question, namely what is the thing to do. It is in answering this question that you are concerned with all-things-considered correctness.34 If there were no such all-things-considered standard, then you could not even begin weighing, say, moral considerations and prudential considerations.

On the other hand, there being such a standard does not entail that there is always one unique correct action available to an agent. It may sometimes or often be the case that there are several correct actions available. Correctness is more like permissibility than like obligatoriness. (It follows that intention-regulation will sometimes have to be influenced by more than just what is correct, in order to fix a determinate intention. This is compatible with my argument: it is one thing to aim at correctness, it is another to be influenced by nothing but what is correct.)

It would be nice to try to say what this property of all-things-considered correctness (goodness, rightness, to-be-doneness) amounts to and why we care about it, but any such attempt would be ill-advised given the scope of this paper.35 Fortunately, it is possible to pick out the property (as I hope I have just done) without giving an account of it.36 In particular, it can be picked out as a property that has a special connection to reasons for action: it is the property that all genuine reasons for action are conducive to—the property the having of which is determined or indicated by the totality of an agent’s reasons for action in a given situation. It is very plausible that we aim at this property in our intention-regulation.37 That you so aim would explain why you are moved, when you regulate your intentions, by (what you take to be) reasons for action.

I claim that this aim of acting correctly is not merely associated with the process of regulating your intentions, but is constitutive of it. You could not be engaged in this process, if you were not aiming to intend and perform correct actions.

One reason to think that this is so is, precisely, that it does seem impossible to engage in the process and be indifferent to considerations that bear on how to act. If it were possible to engage in the process and be indifferent to such considerations, then we should be able to form the intention to drink the toxin. The impossibility of being indifferent to considerations that bear on how to act is easily explained if the aim of acting correctly is constitutive of the process.

A second consideration in favour of the constitutive claim is that it seems to capture something important about the notions of action and agency. Consider a creature who engaged in selection, control and coordination of its behaviour, but who did so in a way utterly indifferent to what would be correct for it to do. The resulting behaviour, and the intention-like states that generated it, would be arbitrary and incomprehensible. The creature would not, for example, intend and enact means to its ends. It would not try to bring about what it took to be good or valuable, or to prevent what it took to be bad or disvaluable. I think that we would not consider the arbitrary behaviour of this creature to be action, and we would not consider the creature an agent. Thus, it seems that selection, coordination and control of how you act constitutively involves aiming to act correctly.

Before moving on, let me make a brief qualification. Shah (2008, 12) has argued that intentions have an ‘aim’, but resists understanding the notion of an aim literally, as a purpose towards which all intention-formation is directed. Rather, he proposes to understand it as a standard of correctness for intentions, contained in the very concept of intention. On Shah’s view, practical deliberators who deploy the concept of an intention are thereby committed to applying to any intention they may form the norm of being correct iff it is not the case that the intended action is one that the agent ought not perform. Shah claims that this proposal best explains the privileged role of what I have called ‘object-directed reasons’ in our intention-formation, without falsely implying that intentions cannot be influenced, unconsciously for example, by other considerations. If one is moved by Shah’s arguments, one can understand everything that I say below about aims in his terms. That is, understand an aim as a norm that you are committed to being motivated by. The point is that, in engaging in the process of intention-regulation, we are, as a constitutive matter, guaranteed to tend to be moved by considerations (that strike us as) conducive to correct action, and to be unmoved by considerations that run counter to correct action.38

Premise (3) distinguishes between the exercise of freedom in your pursuit of an aim and the exercise of freedom with respect to whether to pursue an aim while engaged in a process. The claim is that, if an aim is constitutive of a process, then only the former, and not the latter, can be required for the exercise of freedom while engaged in that process. If it is constitutive of a process’s counting as A-ing that it be directed toward aim T, then it is no restriction on freedom of A-ing that it is impossible to A and yet not pursue T. It is just a fact about what constitutes A-ing. It may still be perfectly possible to A freely, so long as you are free in your pursuit of T. This will involve, for example being reactive to reasons that you see as recommending conduct conducive to T.

It is important here to distinguish freedom of A-ing from freedom with respect to whether to be an A-er at all. It may be that the process of regulating our intentions is something that we cannot but sometimes engage in. If so, then we are not free with respect to whether to be intenders (although we may be free with respect to whether to engage in it at a particular time, and with respect to what prospective actions). But we may nevertheless exercise freedom in the process of regulating our intentions, and thus be free intenders.

(4) follows straightforwardly from (1)-(3).

I move on to (5), the premise that freedom in pursuing an aim requires regular receptivity and reactivity to considerations that recommend conduct conducive to attaining that aim, but does not require reactivity, in pursuing the aim, to considerations that recommend conduct irrelevant to or counter to attaining the aim.

The first conjunct of this premise is in line with what was said about freedom in sec. 2 and at the start of this subsection.39 The second conjunct is very plausible. Freedom in pursuing T surely does not require reactivity, when pursuing it, to reasons whose recommendations are irrelevant or counter to T. To react to such reasons would not be to pursue T; it is thus hard to see how freedom in pursuing T, as opposed to freedom in doing something else, could have anything to do with reactivity to such reasons.40

According to (6), object-directed reasons for intentions recommend conduct conducive to the aim of acting correctly, but state-directed reasons often do not. The first part of this follows from how I characterised object-directed reasons and the aim in question. By contrast, state-directed reasons for intentions, by definition, have nothing to do with whether, in acting as prospectively intended, you would be acting correctly. Often, state-directed reasons will recommend intentions to act in ways that are pointless or incorrect. For example, to drink the toxin would be incorrect. To form the intention to drink the toxin would be to give up on the aim of acting correctly.41 It would be to intend to do what you are well aware you ought not do.42

The conclusion, (7), is that freedom of intention requires receptivity and reactivity to object-directed reasons, but does not require reactivity to state-directed reasons. This follows from (4), (5) and (6). Note that reactivity to state-directed reasons is reactivity in general to such reasons; thus, to secure the conclusion it is enough that state-directed reasons often recommend conduct irrelevant to or counter to the aim of acting correctly.

I have argued that freedom of intention does not require reactivity to state-directed reasons. This argument is also supposed to be an explanation of why freedom of intention should not require reactivity to state-directed reasons. In sec. 2 and 3.2 above I argued that, if doxastic states and intentions are not reactive to state-directed reasons, then they should not be reactive to intentions either. Thus, we have here an argument that freedom of intention requires neither reasons-reactivity nor intentions-reactivity, and an explanation of why that should be so. In short, we have an argument that freedom of intention does not require voluntary control, and an explanation of why that should be so. All of this is compatible with the point that freedom of intention is a genuine species of freedom, apt to ground responsibility.

I have claimed that both freedom of action and freedom of intention are elements in the overall freedom that we, as agents, enjoy—the freedom that grounds our responsibility for our actions. I have claimed, however, that freedom of intention is not exercised as freedom of action is; we do not control our intentions as we control our actions. This runs counter to a common assumption: that if freedom of some will-based progenitors of our actions, such as intentions, has a crucial role in our overall freedom, then we must control those progenitors in just the same way that we control our actions.43 It seems to me not only that this common assumption is false, but that something entirely opposite is true: if we controlled our intentions in the same way that we control our actions—i.e., if we had voluntary control over our intentions—that would actually compromise our freedom of action. That is because we would thereby be liable to act in ways that, by our own lights, were not recommended by our reasons. The reasons-reactivity of our actions would be diminished. Consider the toxin puzzle. If you had voluntary control over your intentions, you would form the intention to drink the toxin. You would not plan to give up the intention, or even to reconsider it, before the time came to drink the toxin: if you plan to give up or reconsider an intention, you are not resolved on executing it, so you have already given it up. So, having formed the intention, you must be resolved on drinking the toxin and disposed to do so. Now you go ahead and execute your intention by drinking the toxin. But, even by your own lights, you have little or no reason to drink the toxin and strong reasons against doing so.44 Acting in a way that you take yourself to have strong overall reason not to act is hardly an exercise of freedom.

The role of intentions in controlling action is such that our lack of voluntary control over intentions enriches our freedom of action, and hence our overall freedom, rather than taking away from it.

3.4. Characterising Doxastic Freedom

I now want to show how the account I just gave of freedom of intention, and the argument for it, can be carried through for doxastic freedom.

The claim is that doxastic freedom requires regular receptivity to object-directed reasons for doxastic states, and reactivity to (recognised or perceived) object-directed reasons, but does not require reactivity to state-directed reasons for doxastic states. Object-directed reasons for doxastic states, recall, are considerations having to do with the truth or falsity of the contents of those states. I will assume, for brevity, that such considerations are always evidence. State-directed reasons for doxastic states are considerations that count in favour of being in certain doxastic states, independently of the truth or falsity of their contents—rewards for believing certain propositions, for example.

On this view, there are at least two types of situation in which a doxastic state with the content p will not count as free. The first is a situation in which the agent is unreceptive to evidence for or against p. For example, an individual with a paranoid delusion that p might be such that she would recognise only some very specific and arbitrary consideration as evidence against p. The second is a situation in which the agent lacks the capacity to react to the (putative) evidence she recognises for or against p, by, accordingly, believing, disbelieving or withholding on p. Freudian cases of repressed beliefs may provide an example of this. Beliefs of this sort are thought typically to persist in the face of the subject’s own take on the evidence for their contents.

What this view denies is that an agent who is unreactive to the reasons she recognises as merely favouring believing, disbelieving, or withholding on p, will thereby count as unfree in the regulation of her doxastic states with respect to p.

Here is an argument for the view, parallel to the argument in the previous subsection:

  • 8
     Doxastic freedom is freedom exercised in the process of regulating your doxastic states.
  • 9
     The process of regulating your doxastic states is, constitutively, directed towards a certain aim, namely the aim of truth.
  • 10
     Freedom exercised in a process that is constitutively directed towards a certain aim is freedom in pursuing that aim when engaged in the process, rather than freedom with regard to whether to pursue that aim when engaged in the process.
  • 11
     Doxastic freedom is freedom exercised in pursuing the aim of truth, rather than freedom with regard to whether to pursue that aim when regulating your doxastic states. (From 8, 9, 10)
  • 12
     Freedom in pursuing an aim requires regular receptivity and reactivity to considerations that recommend conduct conducive to attaining that aim, but does not require reactivity, in pursuing the aim, to considerations that recommend conduct irrelevant to or counter to attaining the aim.
  • 13
     Object-directed reasons for doxastic states recommend conduct conducive to the aim of truth; state-directed reasons often recommend conduct irrelevant to or counter to that aim.
  • 14
     Doxastic freedom requires receptivity and reactivity to object-directed reasons, but does not require reactivity to state-directed reasons. (From 11, 12, 13)

(8) seems true for reasons exactly analogous to those that support (1).

(9) is the analogue of (2). Once again, I think that the core point is very plausible, even if it is controversial how to spell out the notion of aim-directedness and how to characterise the relevant aim.

First, note that the process of regulating your doxastic states is aim-directed. It aims, first, to fix your doxastic states. But it would be wrong to say that, when we engage in this process, we aim merely to fix our doxastic states some way or another. If that were the aim, then we could satisfy it just by forming some arbitrary doxastic states. We don’t form doxastic states arbitrarily. We aim not merely at forming doxastic states, but at getting at the truth.45

What exactly is it that we aim at when we aim at truth? I am going to assume that, in forming a doxastic state with respect to p, you aim that the following state of affairs obtain: you believe p iff p. Again, I think it is not crucial that this be exactly correct; it is plausible that something very like this is what we aim at if we aim at anything. This would explain why, in regulating your doxastic states, you are moved by (what you take to be) evidence.46

I claim that this aim of truth is not merely associated with the process of regulating your doxastic states, but is constitutive of it. You could not be engaged in this process, if you were not aiming at truth.

This claim is supported by similar considerations to those that support its analogue for the case of intention. First, it does seem impossible to engage in the process of regulating your doxastic states, and be indifferent to evidence. If it were possible to engage in the process and be indifferent to evidence, then we should be able to form evidentially unsupported beliefs for financial rewards or for other practical reasons. The impossibility of being indifferent to evidence is easily explained if the aim of truth is constitutive of the process.47

As I noted, there are beliefs that are unresponsive to the subject’s own take on her evidence, namely Freudian repressed beliefs. Do these constitute a counterexample to my claim here? No, because these beliefs are surely not regulated by the normal process that regulates doxastic states. They are regulated by different psychological processes. For example, they fail to connect normally with the subject’s weighing of her evidence, and with her explicit judgments about what is the case; they are instead held in place by, for example, certain unconscious fears and desires. Such cases do not count against the claim that the process of regulation of doxastic states, as I characterised it, is one in which you cannot be indifferent to evidence.

The second point in favour of the constitutive claim is that it seems to capture something important about the notion of belief and of doxastic states. Consider a creature who formed stable representations of the world, and who was disposed to act on the contents of those representations, and so on—but who formed those representations in a way utterly indifferent to evidence, even by its own lights. Imagine, for example, that it formed these representations by ‘accepting’ propositions generated by a random-proposition-generator, and never ‘accepted’, ‘rejected’ or ‘withheld on’ a proposition in any other way. The creature would not be disposed to revise its representations when their contents were mutually contradictory—which they may often be, since formed with no regard for truth. Nor would it be disposed to form new representations with contents entailed by the contents of its existing representations. I think that we would not consider these representations to be beliefs, or the creature to be a believer. Thus, it seems that the process of regulating your doxastic states constitutively involves aiming at truth.

Let me note that the qualification I made in the previous subsection, concerning whether we should regard the aim of intention as a literal purpose or rather as a norm that we are committed to, applies mutatis mutandis in this case.48

(10) is identical to (4). Once again, we should distinguish between being free believers and being free with respect to whether to be believers.

(11) follows from (8)-(10), just as (4) follows from (1)-(3). And (12) is identical to (5).

(13) is the analogue of (7) and is true for similar reasons. It is straightforward that object-directed reasons for doxastic states recommend conduct conducive to the aim of truth. By contrast, state-directed reasons for a doxastic state, by definition, have nothing to do with whether, in forming that doxastic state, you would achieve the aim of truth. Often, state-directed reasons will recommend belief in falsehoods, or withholding on truths.

(14) follows from (11)-(13), analogously to before.

As before, we have here an argument that, and an explanation why, doxastic freedom does not require reactivity to state-directed reasons. As before, this also amounts to an argument that, and an explanation why, doxastic freedom does not require intentions-reactivity. In short, we have an argument that doxastic freedom does not require voluntary control, and an explanation of why that should be so.

I have offered an account on which doxastic freedom is exercised rather differently to freedom of action. Earlier, I worried that any such account would face the challenge of showing that the ‘freedom’ it posits is of the same genus as freedom of action, and is fit to play a role in grounding epistemic responsibility, as freedom of action is fit to play a role in grounding moral responsibility. The account that I have offered meets this challenge. Doxastic freedom, I have argued, is analogous to freedom of intention. The latter is certainly a species of freedom, and is fit to play a role in grounding moral responsibility. By analogy, doxastic freedom as I have characterised it is a species of freedom, and fit to play a role in grounding epistemic responsibility.

I have focused only on what I take to be a central necessary condition on doxastic freedom. I have not tried to offer sufficient conditions for doxastic freedom. It could still be argued that there is some further necessary condition that our doxastic states fail to meet, or that there is some other relevant respect in which our control of our doxastic states differs from our control of our intentions. It seems to me, however, that until such an argument is provided, there is no special obstacle to seeing our doxastic states as free.

The more general condition on freedom, that covers the various species of freedom I have discussed, is a condition of reasons-responsiveness. (Here, ‘responsiveness’ incorporates both receptivity and reactivity, as per Fischer and Ravizza’s (1998) terminology.) It is in the nature of intentions and of doxastic states that a certain kind of reason—the object-directed kind—has a canonical or privileged status. It is what is sometimes called ‘the right kind of reason’. One could say, then, that freedom always requires responsiveness to the right kind of reason. In the case of actions, almost any reason can be the right kind of reason.

On the account I have given, the constraints on the reason-responsiveness of belief are a function of the constitutive aim of belief (and likewise for intention). One might wonder if this account does not suggest that freedom of belief is, after all, exercised in the same way as freedom of certain kinds of action—namely, instrumental actions. Suppose you adopt the goal of making your friend happy, and then you deliberate about what to do in order to bring about that result. Being free in how you act in order to make your friend happy surely does not require that, in the course of this deliberation, you be responsive to considerations that you see as irrelevant or counter to the goal of making your friend happy.49

But this does not show that freedom of belief is after all just the same as freedom of (instrumental) action. It’s true that, in deliberating about what to do in order to make your friend happy, you cannot be responsive to certain considerations (e.g. that making her sad would distract her and help you to beat her at scrabble). But, when you deliberate about what to do in order to make your friend happy, you deliberate about whether to perform some or other more determinate action that is not itself intrinsically or constitutively aimed at that goal. For example, you might deliberate about whether to bring your friend flowers. If so, your deliberation about whether to bring your friend flowers will be framed by the goal of making her happy, and so you will be interested in considerations such as how much your friend likes flowers, whether there isn’t some more efficient way of making her happy, and so on. But bringing your friend flowers is something that you could do, or not, for any kind of reason that you recognised as favouring or disfavouring doing so, including reasons that have nothing to do with making your friend happy. You could do it with any sort of end in view, or none at all. That is why, if you bring her flowers, this action, though instrumental, is nevertheless voluntary.

Of course, bringing your friend flowers in order to make her happy is not something that you could do for the reason that doing so would make her sad. But this does not mean that the action you perform is not one that could have been performed for any kind of reason you recognised. When we pick out your action as bringing your friend flowers in order to make her happy we do not thereby specify an action to which the goal of making your friend happy is intrinsic. Rather, we pick out an action of the type, bringing your friend flowers, and we make a further claim about the (extrinsic) goal in the pursuit of which that action was performed. The relevant individuation of the action you performed, as an action of bringing your friend flowers (in a certain time and place, etc.), is such that you could have performed or refrained from performing that action for any kind of reason you recognised. Accordingly, it is also an action that is reactive to your intentions.50

The situation with belief is very different. The aim of belief is constitutive of belief-regulation as such. An episode doesn’t count as a judging that p, for example, if it is not directed at this aim.51 That’s why, when you judge that p, it’s not true to say that you could have performed that action for any kind of reason you recognised. And that’s why freedom in the regulation of belief does not require reactivity to truth-irrelevant reasons. It’s also why intentions are not involved in the regulation of belief. In this respect, judging that p is unlike giving your friend flowers. It’s also unlike giving your friend flowers in order to make her happy, since the latter is merely an action of giving your friend flowers, whose (extrinsic) goal we specify in picking it out in this way. And, unlike the action of doing something in order to make your friend happy, you do not judge that p by performing some more determinate, voluntary action.

You do form a belief about whether p by doing something more determinate—for example, by forming the specific belief that p. But this more determinate thing that you do is not under voluntary control. Again, there is an important disanalogy with instrumental action here.

Thus, the freedom that we exercise in belief-regulation is not, after all, analogous to the freedom that we exercise in performing instrumental actions. Parallel points apply with respect to freedom of intention.

It might be said that action, just like belief, has a constitutive goal, namely the good. Indeed, this would be congenial to what I have said in this paper about intention. But to say that action aims at the good is not to make a distinction between reasons for action that are conducive to that aim, and reasons that are not. It is to characterise all reasons for action. If action aims at the good, then any consideration that you can recognise or perceive as a reason for action, is a consideration that you see as conducive to the good. So, attributing a constitutive aim to action will not drive a wedge between reasons for action that you can react to, and those that you cannot react to. It will not establish a symmetry between regulation of action and regulation of belief. (On the other hand, attributing a constitutive aim to intention does establish a symmetry with belief, in so far as the aim of intention pertains to a property of the object of intention, and not of the state of intention itself.)

I argued in the previous subsection that voluntary control of intentions would actually compromise freedom of action. I think that voluntary control of doxastic states would compromise freedom of intention. Accordingly, it would undermine responsibility for action, as well as for intention. Consider: one role that doxastic states play is to provide us with access to prospective reasons for action. When you believe that some consideration obtains, you can take that consideration as a reason for acting some way. In so far as doxastic states attain the aim of truth, they tend to provide us with access to the reasons that we actually have. But if you had voluntary control over your doxastic states, you would sometimes form and retain beliefs in ways indifferent to truth and evidence. Such beliefs, like any others, would tend to guide your intentions and actions. You would therefore be liable to take, as reasons for action, considerations that you in fact had no evidence to think obtained, and perhaps strong evidence against—even by your own lights. The regular reasons-receptivity of your intentions would thereby be compromised.

For example, if you had voluntary control over your beliefs, you would form the belief that the US is still a colony of Britain, if offered an immense reward to believe it. Suppose you were then planning a trip to the US from Britain, and had to decide whether to take your passport. If the US is a colony of Britain, you don’t need your passport. And it’s a pain to bring your passport if you don’t have to. So, you would form the intention not to bring your passport, and, other things equal, you would not bring it. All this despite the fact that you have excellent reasons to bring your passport. This would not be cause for congratulation.

The role of doxastic states in the regulation of intention is such that our lack of voluntary control over our doxastic states enriches our freedom, rather than taking away from it.

My account links epistemic responsibility to responsiveness to evidence, via a notion of doxastic freedom. The idea of grounding epistemic responsibility in our dealings with evidence is not new. Richard Moran (2001) has argued that epistemic responsibility is something that emerges in our evidentially-informed deliberation about what is the case. Pamela Hieronymi (2008) has defended a similar view, on which we are responsible for attitudes that result from or constitute our settling a question—a question of the form whether p in the case of beliefs, and a question of the form whether to Φ in the case of intentions. One problem with such views, it seems to me, is that it is not entirely clear why, in general, deliberation or question-settling should give rise to responsibility.52 A second problem is that many of our doxastic states are not the outcome of deliberation, nor are they constituted by or the result of the settling of a question (unless the notion of question-settling is understood so thinly that it seems to lose any explanatory force). I think that the account I have offered here captures what is right about these views, while avoiding these two problems. Deliberation and question-settling (normally) give rise to responsibility when they are reasons-responsive in the specific way that I have set out, because such reasons-responsiveness is a crucial condition on freedom, and there is a general connection between freedom and responsibility. However, neither deliberation nor question-settling (in any strong sense) need have a causal or constitutive role in bringing about or sustaining a doxastic state, in order for that state to meet the conditions for doxastic freedom.

4. Conclusion

  1. Top of page
  2. Abstract
  3. 1. Doxastic Freedom
  4. 2. Freedom of Action, Voluntary Control and Doxastic Freedom
  5. 3. Freedom of Intention and Doxastic Freedom
  6. 4. Conclusion
  7. References

If epistemic responsibility requires doxastic freedom, this can make us doubt the possibility of epistemic responsibility. I have argued that such doubt is misplaced. We do not have voluntary control over our doxastic states, but voluntary control is not required for doxastic freedom. Doxastic freedom is analogous to freedom of intention, not to freedom of action.

On the account I have given, certain kinds of doxastic states, such as paranoid delusions and Freudian repressed beliefs, are unfree. If freedom is required for responsibility, then we are not epistemically responsible for such beliefs. That seems to me a plausible consequence of the account.53

Footnotes
  • 1

     For the purposes of this paper, ‘freedom of action’ refers to the kind of freedom we enjoy in our bodily actions. We may also enjoy this kind of freedom with respect to some mental actions.

  • 2

     More precisely, a doxastic state is a mental state with a propositional content and involving a doxastic attitude of belief, withholding or disbelief. I will focus only on outright attitudes, not on graded belief. When I talk of the claim that you are epistemically responsible for your doxastic states, I mean not that you are responsible for whether you have a doxastic attitude towards a proposition, but that you are responsible for which attitude you have towards a given proposition, given that you have any attitude towards it.

  • 3

     Philosophers who write about epistemic responsibility differ over its relation to praiseworthiness and blameworthiness. Some authors (e.g., Hieronymi (2008)) treat full responsibility as bringing with it both praise- and blameworthiness. Other authors treat one or the other as primary. For example, Weatherson (2008) argues that epistemic justification is tied in the first place to praiseworthiness. In his influential treatment, Alston (1988) argues that if justification is to be understood in terms of obligations (a view he ultimately rejects), then it would have to be a kind of blamelessness. Yet other authors don’t mention praise or blame at all. In any case, what is crucial is that epistemic responsibility is a stronger notion than mere evaluability: it is not merely that our doxastic states are evaluable against certain standards, but that we are subject to certain obligations or demands that we can be held to.

  • 4

     Different conceptions of doxastic freedom are offered by Pettit and Smith (1996), McDowell (1998) and Steup (2008). If my arguments in this paper are correct, then all of these conceptions are importantly lacking.

  • 5

     Under an appropriate description. There are botched actions, and attempts to do things that have limited chances of success, that will not count as controlled under certain descriptions (e.g. ‘potting the black ball’). These, if free, will nevertheless count as controlled under other, often more basic, descriptions (e.g. ‘striking the cue ball with the cue’).

  • 6

     This tradition is associated with the classical compatibilist views of Hume (1978) and Hobbes (1997). Among more recent work, Frankfurt (1971) focuses on desires, including higher-order desires. Bratman (1987) seminally emphasises the importance of intentions in understanding human agency. Although this sort of view is associated with compatibilism, an incompatibilist could just as well hold that an action is free only when it is in accordance with the agent’s desires or intentions.

  • 7

     Such cases appeared in Frankfurt (1969). One way of developing the qualification is as follows. In the actual scenario, your action is causally produced by a certain psychological mechanism. Call it the ‘actual mechanism’. The claim, then, is roughly this: in those counterfactual scenarios in which the actual mechanism is operative, you act as you intend. In Frankfurtian cases, the problematic counterfactual scenarios are ones in which the actual mechanism is not operative. For a development of an actual mechanism view, though not one that focuses on intentions, see Fischer and Ravizza (1998).

  • 8

    Steup (2008, 387) denies that it is. See Booth (2009) for criticisms of Steup’s argument.

  • 9

     Influential views of this sort, which is Kantian in spirit, have recently been developed by Wolf (1990) and by Fischer and Ravizza (1998). See also Owens (2009).

  • 10

     How exactly to spell out this counterfactual is a slightly tricky issue—see n. 12 below. The swearing case is discussed by Weatherson (2008); see McHugh (forthcoming c) for some criticism of Weatherson’s treatment. For an interesting discussion of actions not performed for reasons see Hursthouse (1991).

  • 11

     It is important to be clear that reasons-reactivity is a matter of having the capacity to act on the reasons you see yourself as having. This is distinct from reasons-receptivity: the capacity to recognise the reasons you in fact have. Together, these two conditions constitute reasons-responsiveness. The terms are Fischer and Ravizza’s (1998). In what follows I will sometimes talk of ‘reacting to the reasons you recognise’. This should be understood as shorthand for ‘reacting to the reasons you recognise or perceive’. That is, reasons-reactivity is reactivity both to reasons that you correctly take yourself to have and to reasons that you incorrectly take yourself to have or see yourself as having.

  • 12

     Again, this requirement may have to be restricted to those scenarios in which the actual mechanism is operative. See n. 7 above. A further complication has to do with akratic action. By ‘akrasia’ here I mean action that is contrary to your impression or judgment of what there is most reason to do. You can Φ freely but akratically. You can also Φ freely even though, had there been reason not to Φ, you would have akratically Φed anyway. In view of this, Fischer and Ravizza (1998) claim that reasons-reactivity requires only that there be at least one possible scenario in which the actual mechanism operates, you recognise sufficient reason to do otherwise, and you react accordingly. They claim that this condition ensures that the actual mechanism has the ‘executive power’ to react to any kind of reason to do otherwise. I argue against this in my (McHugh forthcoming c, n. 28). In any case, the crucial point for me is that there be a capacity to react to any kind of reason you recognise—however this point about capacities is spelled out in counterfactual terms. In the akratic case you fail to react to the reason you recognise yourself as having, but you are not thereby unreactive in the relevant sense: you have the capacity to react to the reason you fail to react to. Indeed, it is in part because you have this capacity that the case counts as one of akrasia.

  • 13

     One might ask if the conditions are so much as distinct. Isn’t an intention to Φ itself a reason to Φ? No: a reason to Φ, as I am using that term, is a consideration that favours Φing, not an attitude of the agent. We will see later that intentions-reactivity and a certain sort of reasons-reactivity are doubly dissociable.

  • 14

     This is discussed at length by Bratman (1987). See also Pink (1991).

  • 15

     This use of the term ‘voluntary’ is broadly in accordance with McHugh (2011, forthcoming c) and Hieronymi (2009).

  • 16

     See McHugh (forthcoming c) for more defence of this.

  • 17

     The first example comes from Alston (1988). The second comes from McHugh (forthcoming c), in which there is extensive further defence of this argument against doxastic voluntarism.

  • 18

     It might be claimed that you cannot take practical reasons to be reasons for forming beliefs. But this seems very implausible. Even if practical reasons cannot be reasons for forming beliefs, it would not follow that we can’t take them to be so. We are in general capable of being wrong about what is a reason for what. Furthermore, that practical reasons cannot be reasons for forming beliefs is highly controversial. Those philosophers who deny it would appear to be counterexamples to the claim that we can’t take practical reasons to be reasons for forming beliefs.

  • 19
  • 20

     The sort of dilemma I have just outlined, for any defence of doxastic freedom, is pressed by Booth (2009). See also Nottelmann’s (2006) criticisms of what he calls the ‘analogy argument’ for doxastic voluntarism.

  • 21

    Steup (2008) sometimes exemplifies the first strategy, although at one point seems to endorse the second. McDowell (1998) seems to exemplify the second. I am not sure which way to read Pettit and Smith (1996).

  • 22

     For example, Frankfurt (1971) distinguishes between free action and freedom of the will.

  • 23

     As pointed out by Pink (2009, 111).

  • 24

     Perhaps it is worse and more blameworthy to act on a wicked intention than merely to hold such an intention. My point is only that you can be morally responsible for intentions themselves, independently of acting on them.

  • 25

     This verdict is shared by Kavka (1983), Shah (2008), Hieronymi (2009), Pink (2009), and many others. Mele (1992) has some quibbles but his position is still congenial to the present point. Note that you do not get the reward if you take some measure to give yourself an independent reason to drink the toxin, such as making a bet on whether you will drink it.

  • 26

     It is sometimes suggested that the fundamental explanation of why you cannot intend to drink the toxin is that you know you will not do so, and you cannot intend to do what you know you will not do. But note that if you know you will not drink the toxin, that is because your reasons count against drinking it. In any case, the present point is that there are certain restrictions on the reasons-reactivity of intentions. What’s more, certain other phenomena, such as the fact that intentions are not themselves intentions-reactive (see below), are closely connected to this point, and do not seem to be explained by the point that you cannot intend to do what you know you will not do.

  • 27
  • 28

     As in the case of doxastic states, we can have and execute intentions to form intentions of certain types—e.g. intentions regarding whether to Φ. As Anders Nes pointed out to me, such intentions to form intentions play an important role in planning. What there is no role for, however, is intentions to form particular intentions, e.g. the intention to Φ.

  • 29

     Again, the reactivity component is reactivity to recognised or perceived reasons (see n. 11 above). There is a question how the requirement of reactivity to reasons for action should be spelled out, given the possibility of akrasia. The correct formulation will have to capture the idea that the agent (or her mechanism) has the capacity, to react to any kind of reason for action that she recognises, by intending to act accordingly, even if she is akratic, or would be akratic in various counterfactual scenarios. See n. 12 above.

  • 30

     This and the following would have to be refined slightly, in order to deal with the following kind of case: someone offers you a reward for believing propositions that are false. Here, evidence against p will constitute a practical reason to believe p. As Reisner (2009) points out, whether something counts as a practical or an epistemic reason is not an intrinsic feature of the fact or consideration that constitutes the reason. I leave this complication aside.

  • 31

     For discussion of this distinction as it pertains to doxastic states, see Olson (2004), Rabinowitz and Rønnow-Rasmussen (2004) and Reisner (2009). For discussion of the instance pertaining to intentions, see Shah (2008). Hieronymi (2005) discusses the distinction in generality.

  • 32

     Thanks to an anonymous referee for pressing this question.

  • 33

     The agent need not consciously represent this aim, or even be capable of representing it (conceptually, at least). Perhaps young children and animals have this aim; in that case, it would be an aim in the sense of being a function of the processes by means of which they regulate their intentions. See Velleman (2000), McHugh (forthcoming a, forthcoming b), on this notion of an aim.

  • 34

     On the need for a notion of what is the thing to do all-things-considered, tout court, sans phrase, all-out, or without qualification, see for example: Davidson (2001), Wedgwood (2007, 24–25), Alvarez (2010, 15–16). Coming to the conclusion that Φing is the thing to do does not always lead to the intention to Φ, as akrasia shows. On my account, akratic intentions are a sort of failure or malfunction of intention-regulation. As Shah (2008, 11) points out, the possibility of akrasia does not show that, in normal cases, some further piece of reasoning, or some further psychological factor, is required to effect the transition from concluding that Φing is the thing to do to intending to Φ. Rather, in cases of akrasia there is some interfering factor that obstructs this transition.

  • 35

     For recent discussions see Wedgwood (2003), Alvarez (2010, 13ff).

  • 36

     See Wedgwood’s discussion of the ‘formal recognitional view’ of correct choice (Wedgwood 2003).

  • 37

     Although I will not argue it here, I think something like the following is plausible: an intention is itself correct iff it is an intention to perform an available correct action. If this is right, then we can say that the regulation of intention aims at correct (and executed) intending. This would make the analogy with doxastic states, to be developed below, even closer, if the regulation of doxastic states aims at true belief, and if true belief is correct belief.

  • 38

     Shah’s account applies only to intention-regulation that is carried out through deliberation framed (perhaps implicitly) by the question what to intend, since it is only in such deliberation that the concept of intention, and its constitutive norm of correctness, is ‘activated’. This is not a problem, since it is precisely within deliberation about what to intend that a state-directed reason might strike you as favouring intending to Φ, and that you will nevertheless be unable to react by forming the intention to Φ for that reason. The same point applies, mutatis mutandis, to the corresponding account of the ‘aim’ of belief, which I will mention below.

  • 39

     In sec. 3.1 I described a case of reasons-reactivity without regular reasons-receptivity (the sabre killer). I noted that one could treat such cases as involving a kind of freedom of action, without the richer kind of freedom, freedom of the will, that is required for full moral responsibility. On this treatment, there are two kinds of freedom, one of which does not require regular reasons-receptivity. If one wishes to treat the cases in this way, then one might wonder why freedom in pursuing T should require regular receptivity to considerations relevant to attaining T, as I claim here. If there are these two kinds of freedom, then they apply equally to aim-directed activities. My claim here can be understood thus: in pursuing T one enjoys the second, richer kind of freedom only if one is regularly receptive to considerations conducive to T. Accordingly, what I am offering an account of is freedom of intention in this rich sense. There may be a weaker kind of freedom of intention that requires only reactivity to reasons for action, but not regular reasons-receptivity. Likewise for doxastic freedom. Note that the claim I am primarily concerned to defend is that freedom in pursuing T does not require reactivity to considerations counter or irrelevant to attaining T. This claim is independent of the idea that such freedom requires regular receptivity to considerations relevant to attaining T; I take the latter idea simply to be part of a full understanding of freedom in the rich sense.

  • 40

     Here, I am arguing that freedom in pursuing T does not require reactivity to considerations not conducive to T. This should be distinguished from a stronger claim: that pursuing T requires you to be reactive exclusively to considerations (that you take to be) conducive to T. This latter claim appears to be false. However, it has been claimed that, in deliberating about whether to believe p, you are reactive exclusively to considerations (that you take to be) conducive to truth. Thus, the fact, if it is a fact, that you aim at truth in regulating your doxastic states, will not explain why you are exclusively reactive to (what you take to be) truth-conducive considerations in deliberation about what to believe (if, indeed, you are exclusively so reactive). For discussion see Owens (2003), Shah and Velleman (2005), Steglich-Petersen (2009) and McHugh (2011, forthcoming a). My present purpose is not to explain all the facts about how doxastic states can be motivated, but to argue that freedom of intention and doxastic freedom do not require reactivity to state-directed reasons.

  • 41

     The relation between state-directed reasons and acting correctly can be more complicated. When considering what to do, you may have a reason to simply form some intention one way or another, in order to facilitate planning. This would appear to be a state-directed reason that does not bear directly on which of the courses of action under consideration would be correct, but does bear on the best way for you to pursue the aim of acting correctly. It is an interesting question to what extent we are reactive to such reasons. It seems clear that we can, because of pressure on cognitive resources, say, go ahead and form the intention to do whatever at that time seems best to do, even when we know that we might make a better decision if we took more time to reflect. But can we opt for a particular intention because of state-directed reasons to have that particular intention (as opposed to reasons to form some intention)? Pink (1991) argues that we can. In any case, my point stands as long as state-directed reasons are not in general conducive to correct action.

  • 42

     We must bear in mind here that drinking the toxin is not instrumental to having the intention to drink it, nor to winning the reward.

  • 43

     For example, see Owens (2009, 121). Compare Pink (2009).

  • 44

     Perhaps the mere fact of having an intention gives you some reason to execute it. Resolution has its merits. But only a crazed fanatic of resolution could think that such a reason outweighed the strong reasons not to drink the toxin.

  • 45

     Or perhaps, as I argue in (McHugh forthcoming b), what we aim at is not merely truth, but knowledge. For the purposes of the present paper I will stick with the less controversial idea that we aim at truth. My arguments would have to be reformulated somewhat in order to accommodate the idea that doxastic regulation aims at knowledge, but they would go through all the same. The reason for this is that, from your point of view, considerations relevant to whether you know p will largely overlap with considerations relevant to whether p (see McHugh 2010 for more on this). In particular, reactivity to what I am calling ‘state-directed reasons’ will not be conducive to knowledge, any more than to true belief. So, aiming at knowledge won’t require reactivity to state-directed reasons. For more on what is involved in doxastic regulation having such an aim, be it truth or knowledge, see Velleman (2000), McHugh (forthcoming a, forthcoming b). See n. 33 above.

  • 46

     In (McHugh 2011) I defend the claim that when you judge p you do so with the aim of judging p iff p, and use this claim to explain the pattern of motivation characteristic of judgment. Given the relation between judgment and belief, and the role played by judgment in what I have called the process of regulating your doxastic states, the arguments of that paper would, I think, carry over to the present case.

  • 47

     Here I am using the impossibility of being indifferent to evidence as the premise of an inference to the best explanation (IBE), whose conclusion is that the regulation of doxastic states is constitutively aimed at truth. This conclusion I will in turn use to defend doxastic freedom. There might be suspicions of question-begging about the IBE. After all, the opponent of doxastic freedom is likely to think that the impossibility of being indifferent to evidence simply shows that we do not exercise freedom in the regulation of our doxastic states, rather than that there is some aim we are pursuing in the regulation of our doxastic states. But I don’t think there is any question-begging here. Even the denier of doxastic freedom will accept that the impossibility of being indifferent to evidence requires an explanation. The claim that we lack doxastic freedom is not itself an explanation of this phenomenon. The IBE thus cannot be disqualified on dialectical grounds. It is up to the denier of doxastic freedom to offer some alternative account of why it is impossible to be indifferent to evidence. Furthermore, the claim, that the regulation of doxastic states is constitutively aimed at truth, does not by any means presuppose doxastic freedom; there is plenty of work to do to get from this claim to the conclusion of my argument.

  • 48

     That it should be regarded as a norm is argued for by Shah (2003), and Shah and Velleman (2005). This ‘normativist’ view, like the corresponding view about intention, would have implications only for doxastic regulation that is carried out through deliberation—see n. 38 above. The claim that belief literally has an aim is rejected by Owens (2003). For responses to Owens see Steglich-Petersen (2009) and McHugh (2011, forthcoming a).

  • 49

     This is pointed out by Steglich-Petersen (2006), who argues that deliberation about what to believe is indeed analogous to deliberation about what to do in order to achieve some aim you have adopted.

  • 50

     There are other ways of picking out actions that also have implications for the sorts of reason that motivate them. For example, if we pick out an action as an act of generosity, it arguably follows that the action was not performed for reasons conducive to some malevolent end. But to perform an act of generosity you must perform some more determinate action, that could itself have been motivated by any sort of reason you recognised. This issue was raised for me by Joshue Orozco.

  • 51
  • 52

     Of course Moran and Hieronymi have plenty to say on this issue. It seems to me that neither of them establishes a case for responsibility in a sense that goes beyond mere evaluability (see n. 3 above). Unfortunately I do not have space to discuss this. The main point of the present paragraph is to distinguish my account from theirs.

  • 53

     Thanks to Matthew Chrisman, Tony Booth, Daniel Whiting, Anne Meylan, Kirk Michaelian, Joëlle Proust, Claudine Tiercelin, and several anonymous referees, as well as to audiences at the Institut Jean Nicod’s Knowjust seminar, the Epistemology and Philosophy of Mind conference at KU Leuven in January 2010, and the Justification Revisited conference at the University of Geneva in March 2010. The preparation of this article was supported by the Agence Nationale de la Recherche, under the contract ANR-08.BLAN-0205-01.

References

  1. Top of page
  2. Abstract
  3. 1. Doxastic Freedom
  4. 2. Freedom of Action, Voluntary Control and Doxastic Freedom
  5. 3. Freedom of Intention and Doxastic Freedom
  6. 4. Conclusion
  7. References
  • Alston, W.P. (1988). The deontological conception of epistemic justification. Philosophical Perspectives, 2, 257299
  • Alvarez, M. (2010). Kinds of reasons: an essay in the philosophy of action. (Oxford: OUP)
  • Booth, A.R. (2009). Compatibilism and free belief. Philosophical Papers38, 112
  • Bratman, M.E. (1987). Intentions, plans and practical reasons. (Cambridge, Mass.: Harvard UP)
  • Davidson, D. (2001). How is weakness of will possible? (In his Essays on actions and events. Oxford: OUP)
  • Fischer, J.M. and Ravizza, M. (1998). Responsibility and control: a theory of moral responsibility. (Cambridge: CUP)
  • Frankfurt, H.G. (1969). Alternate possibilities and moral responsibility. Journal of Philosophy, 66, 82939
  • Frankfurt, H.G. (1971). Freedom of the will and the concept of a person. Journal of Philosophy, 68, 520
  • Hieronymi, P. (2005). The wrong kind of reason. The Journal of Philosophy, 102, 437457
  • Hieronymi, P. (2008). Responsibility for believing. Synthese, 151, 35373
  • Hieronymi, P. (2009). Two kinds of agency. (In L. O’Brien and M. Soteriou (Eds.), Mental actions. Oxford: OUP).
  • Hobbes, T. (1997). Leviathan (ed. R.E. Flatman and D. Johnston). (New York: Norton)
  • Hume, D. (1978). An enquiry concerning human understanding (ed. P.H. Nidditch). (Oxford: Clarendon)
  • Hursthouse, R. (1991). Arational actions. The Journal of Philosophy, 88, 5768
  • Kavka, G.S. (1983). The toxin puzzle. Analysis, 43, 3336
  • McDowell, J. (1998). Having the world in view: Sellars, Kant and intentionality. Lecture 1: Sellars on perceptual experience. The Journal of Philosophy, 95, 431450
  • McHugh, C. (2010). Self-Knowledge and the KK principle. Synthese, 173, 231257
  • McHugh, C. (2011). Judging as a non-voluntary action. Philosophical Studies, 152, 245269
  • McHugh, C. (forthcoming a). Belief and aims. Philosophical Studies
  • McHugh, C. (forthcoming b). What do we aim at when we believe?Dialectica
  • McHugh, C. (forthcoming c). Epistemic deontology and voluntariness. Erkenntnis
  • Mele, A. (1992). Intentions, reasons and beliefs: morals of the toxin puzzle. Philosophical Studies, 68, 171194
  • Moran, R. (2001). Authority and estrangement: an essay on self-knowledge. (Princeton: Princeton UP)
  • Nottelmann, N. (2006). The analogy argument for doxastic voluntarism. Philosophical Studies, 131, 559582
  • Olson, J. (2004). Buck-passing and the wrong kind of reasons. Philosophical Quarterly, 54, 295300
  • Owens, D.J. (2003). Does belief have an aim?Philosophical Studies, 115, 283305
  • Owens, D. (2009). Freedom and practical judgement. (In L. O’Brien and M. Soteriou (Eds.), Mental actions. Oxford: OUP).
  • Pettit, P. and Smith, M. (1996). Freedom in belief and desire. The Journal of Philosophy, 93, 429449
  • Pink, T. (1991). Purposive intending. Mind, 100, 343359
  • Pink, T. (2009). Reason, voluntariness and moral responsibility. (In L. O’Brien and M. Soteriou (Eds.), Mental actions. Oxford: OUP.)
  • Rabinowitz, W. and Rønnow-Rasmussen, T. (2004). The strike of the demon: On fitting pro-attitudes and value. Ethics114, 391423.
  • Reisner, A. (2009). The possibility of pragmatic reasons for belief and the wrong kind of reasons problem. Philosophical Studies, 145, 257272
  • Shah, N. (2003). How truth governs belief. The Philosophical Review, 112, 447482
  • Shah, N. (2008). How action governs intention. Philosophers’ Imprint8.
  • Shah, N. and Velleman, D. (2005). Doxastic deliberation. The Philosophical Review, 114, 497534
  • Steglich-Petersen, A. (2006). Voluntarism and transparent deliberation. South African Journal of Philosophy, 25, 173178.
  • Steglich-Petersen, A. (2009). Weighing the aim of belief. Philosophical Studies, 145, 395405.
  • Steup, M. (2008). Doxastic freedom. Synthese, 161, 375392.
  • Velleman, J.D. (2000). On the aim of belief. (In his The possibility of practical reason. Oxford: OUP.)
  • Weatherson, B. (2008). Deontology and Descartes’s demon. The Journal of Philosophy, 150, 540569.
  • Wedgwood, R. (2003). Choosing rationally and choosing correctly. (In S. Stroud and C. Tappolet (Eds.), Weakness of will and practical irrationality. Oxford: OUP.)
  • Wedgwood, R. (2007). The nature of normativity. (Oxford: OUP.)
  • Wolf, S. (1990). Freedom within reason. (Oxford: OUP)