1
Motivation-Encompassing Attitudes
Part of Alfred Mele’s defense of the basic Humean claim that desire is a necessary condition of action is the claim that intentions and action-desires – desires that take the agent’s possible actions as their object – essentially encompass motivation to act. In contrast, state-desires – desires other than action-desires – can be contingently, but not essentially, motivation encompassing. Beliefs, in further contrast, cannot motivate at all; they are neither essentially nor contingently motivation-encompassing attitudes.[1]
In this paper I argue that Mele’s argument for the claim that state-desires are only contingently motivation-encompassing attitudes also supports the claim that action-desires are only contingently motivation-encompassing attitudes. If Mele’s argument is cogent, then there are no essentially motivation-encompassing desires.
I also argue that Mele’s argument against the claim that beliefs are motivation-encompassing – again, if cogent – disproves the claim that desires are motivation-encompassing. Consequently Mele fails to draw a distinction between desires and beliefs that might be used to support the view that desire is a necessary condition of action.
A. State-Desires and Action-Desires
Before I get to Mele’s arguments, some terminological clarifications are needed. Mele defines ‘motivation-encompassing attitudes’ as attitudes that “are, constitute, or include motivation” (14).[2] For an attitude to encompass motivation is for the attitude to dispose an agent to act under certain circumstances.[3] A desire for a coffee, for example, disposes me to stop in to the café if I pass one, all other things being equal.
A motivation-encompassing attitude can encompass motivation either essentially or contingently. If the attitude encompasses motivation essentially, then it encompasses motivation “not only in [the subject’s] actual situation but also in all possible scenarios in which [the subject] has that attitude” (15). If the attitude encompasses motivation contingently, then it encompasses motivation in the actual situation, but not in all possible scenarios.[4]
Mele refutes the view that all desires are essentially motivation-encompassing. He gives the example of Ann,whose desire that the Pistons win does not encompass motivation to do anything at all.[5] Because she knows that she cannot do anything to bring about a Pistons’ victory, she is not disposed to make it the case that the Pistons win. She is not disposed to check on the score during the game, because she believes that she will be unable to check on the score. She is not even disposed to think or daydream about the game, because Ann “is a special person”:
when she knows that she has no chance of learning how matters of minor to moderate importance to her have turned out, she has no tendency at all to think or daydream about those matters, even though she desires a particular outcome (26).
In this case her principle of not spending time thinking about something she wants to happen when she will never know whether or not it occurs prevents her from being disposed in the ordinary way. Put more generally, beliefs that certain actions are impossible, on the one hand, and an agent’s principles, on the other, are capable of eliminating or preventing any disposition that a desire for some state of affairs might produce, at least in the case of state-desires.[6]
Mele considers an objection to his view. One might argue that the desire for a Pistons’ victory does dispose Ann to act, but her beliefs and principles obstruct those motivations from manifesting in behavior (26-7). This sort of response, in Mele’s words, “has the ring of a desperate move to save a theory” (27). Any attempt to defend this account will be based on the assumption that desires necessarily motivate in the first place. That is, it will be circular.[7]
Mele then argues that another type of desire – action-desires – do essentially encompass motivation to act. If the object of my desire is a possible action of mine, then I am essentially motivated to do it. “Any agent’s desire to A, given the very nature of action-desire, is an inclination of the agent – sometimes only a very weak inclination – to A. This is part of what it is to be an action-desire” (136-7). Hence if an attitude does not incline the agent to act it cannot be an action-desire, by definition.
One problem with this argument is that his appeal to “the folk concept of action-desire” (137) is a move that seems equally open to the proponent of the view that all desires are essentially motivation-encompassing – a view that Mele rejects. A proponent of that view might claim that it is part of the folk concept of all desires that they are essentially motivation-encompassing. She might then use this claim as a premise in the argument that Mele dismisses as circular – the one that “has the ring of a desperate move to save a theory.” On this account Ann’s belief that she cannot affect the game does not eliminate her disposition to affect the game, it merely prevents the disposition from manifesting in behavior. The desire essentially disposes; it is merely blocked from manifesting in action by the belief.
Mele avoids this result by arguing that an agent cannot desire to do something and believe that it is impossible. When I desire to A and then find out that I cannot A, I cease desiring to A. “If I am convinced that I cannot travel faster than the speed of light, change the past, or defeat the current heavyweight champion of the world in a fair fight, then although I might wish that I could do these things, I do not desire to do them” (135). It might be the case that I wish that I could A or that I would desire to A, if it were possible, but neither of these states are or entail a desire. Indeed, a rational person will cease to desire to do that which she knows she cannot do.
In contrast, a state-desire – like the desire that the Pistons win – can both cease to dispose the agent to act in light of beliefs that she cannot bring about the desired state of affairs and persist as a desire (26). An action-desire cannot both cease to dispose the agent to act in light of beliefs that she cannot do what she desires to do and persist as a desire. If she is rational, her belief will eliminate her desire and – perhaps – replace it with a wish. If she is irrational, then her belief and action-desire can coexist, and her action desire continues to dispose her to act anyway.
If this argument is cogent, then Mele proves that one explanation for how a state-desire might not encompass motivation cannot apply to action-desires. There is a reason, however, to think that the argument is not cogent. At the very least, it is not consistent with some of Mele’s other claims. He gives another example of a case in which an agent believes some action is impossible:
It may be claimed, for example, that a jogger may desire to run another lap but be too tired even to try, so that his desire does not constitute motivation to run. Again, an agent who takes his A-ing to be physically impossible may hope or wish that he could A, but he does not desire to A (136, first emphasis added).
This suggests that hoping, like wishing, does not entail desiring. Yet in his consideration of Ann’s state-desire that the Pistons win, he claims that “hoping that p entails that desiring that p” (26).
Aside from this pretty straightforward evidence of a double-standard with regard to state- and action-desires, Mele also fails to explain why it is not the case that action-desires, like state-desires, might not encompass motivation because of principles of the agent. Remember that Mele explains that Ann might both desire that the Pistons win and yet not be disposed to think or daydream about whether they have won because of her principle of not bothering to think about such things if she can never know the facts. Surely other properly formulated principles can eliminate the motivation that an action-desire might encompass as well. Suppose, for example, that Ann desires to use cocaine, but has the principle of not acting on desires to do what is illegal. Under these circumstances, she is not disposed to use cocaine any more than she is disposed to daydream about the game under the first set of circumstances. Her principles explain the lack of motivation in the action-desire just as it does in the state-desire.
The only obvious way of avoiding the conclusion that action-desires too are only contingently motivation-encompassing, is to insist that the action-desire continues to encompass motivation, but that the principle prevents the motivation from manifesting in behavior. If this is an acceptable reply in this case, however, more needs to be said about why the same strategy rings of a desperate move to save a theory when used to defend the claim that all desires are essentially motivation-encompassing attitudes.
All of this proves that Mele does not offer a cogent argument for his claim that action-desires are essentially motivation-encompassing but that state-desires are not. In the case of state-desires, Mele argues that since we can conceive of a state-desire that does not encompass motivation even contingently, it follows that no state-desire encompasses motivation essentially, because factors like those that make Ann’s desire that the Pistons win motivationally inert can equally affect any state-desire. If this is right – and I think it is – then the conceivability of an action-desire that does not encompass motivation even contingently entails that no action-desire encompasses motivation essentially because the factors that make Ann’s desire to use cocaine motivationally inert can equally affect any action-desire.
B. Beliefs and Desires
In a similar way, Mele’s argument for the claim that beliefs do not encompass motivation – if cogent – can be used to argue that desires do not encompass motivation either. Mele offers the following scenario. Imagine that Eve “has aided her ailing uncle for years, believing herself to be morally required to do so.” When her immediate family is killed in a tragic accident, she loses her motivation to help her uncle even though she still believes that she is morally required to do so. Mele concludes that “[i]f we can imagine this, [then] agents may believe that they are morally required to A and yet have no motivation to A” (111).
This sounds like an argument meant to parallel the argument that state-desires are not essentially motivation-encompassing. This time Mele gives an example of a normative belief that does not encompass motivation. From this he concludes that no normative belief is essentially motivation-encompassing. After all, the factors that make Eve’s belief that she should help her uncle motivationally inert can equally affect any normative belief. Hence no normative belief is essentially motivating.
At this point, then, Mele has said nothing that successfully supports a distinction between beliefs and desires. Nor has he said anything that supports a Humean account of motivation. As Mele admits when considering state-desires, a contingently motivation-encompassing attitude can still motivate. If Ann does not have her ‘no daydreaming about unknown possibilities’ principle, then her desire that the Pistons win can motivate the daydreaming, even if it does not essentially encompass motivation.[8] Likewise, the conclusion that normative beliefs do not essentially encompass motivation is consistent with the claim that they can motivate.
Instead of offering an additional argument, however, he simply assumes that the Eve scenario proves that normative beliefs are not contingently motivation-encompassing either. A contingently motivation-encompassing attitude “is such that, in forming or acquiring it, one acquires motivation” (127).[9] Yet
in at least some… instances of listlessness, it is not the case that an agent’s moral judgment that he ought to A initially has motivational force directed at A-ing and then loses it; rather, at the time at which the judgment is made… the agent is utterly devoid of motivation to A (127).
Again, this argument seems to parallel that advanced for the claim that state-desires do not essentially motivate. There are cases in which an agent has a normative belief but no corresponding motivation. Since the factors that explain the lack of motivation might apply to any normative belief, no normative belief is essentially motivation-encompassing.
The same goes for Ann’s desire to use cocaine. As long as her principle of not using cocaine is firmly in place, her desire to use cocaine will not dispose her to do so even at the moment it arises. Other principles and beliefs can preclude her disposition to do any other thing as well. Hence no desire is essentially motivation-encompassing.
Mele, however, takes this as evidence that beliefs can never be contingently motivation-encompassing. Rather than thinking “that in those cases in which an agent believes he ought, morally, to A and does have motivation to A, the motivation is built into the belief,” an “alternative hypothesis” (namely the one that Mele endorses) “is that, in these cases… something is missing such that, were the ought-belief to occur in its presence, motivation to A would result” (127-8). Mele adds that a desire is the most obvious candidate for filling this motivational gap, and that the desire to do what one should is particularly well-suited to the task.
The problem with all of this is that the argument he offers for the claim that beliefs never by themselves provide motivation can be used just as well against the claim that desires provide motivation, and if this is right, the claim that a desire is the “obvious candidate” for filling in the motivational gap in the case of normative beliefs is untenable. Let make this as clear as possible.
Mele argues that the possibility of normative beliefs that do not encompass motivation supports the conclusion that none do. If this is right, then the possibility of desires that do not encompass motivation supports the conclusion that none do. Mele then argues that since no beliefs are motivation encompassing, we should assume that in cases in which an agent believes she should A and is motivated to A, the desire to A, or the desire to do what one should, provides the motivation, and that the belief does not. The parallel of this is that since no desires are motivation encompassing, we should assume that in cases in which an agent desires to A and is motivated to A, some other desire – perhaps the desire to do what one desires – provides the motivation, and that the original desire does not.
This is problematic for two reasons. First, the same problem arises for the second-order desire. It is still conceivable that an agent desires to do what she desires and desires to A but is not motivated to A. In this case, another desire is needed, and so on ad infinitum. Second, even if there were reason to think the regress ends at the second-order desire, the account would not be consistent with the claim that in some cases an agent’s first-order desire can motivate action. It would always be this second-order desire to do what one desires that does the motivating, and this seems patently false.
Conclusion
Mele argues that the conceivability of a normative belief that does not encompass motivation to act supports the conclusion that no normative beliefs encompass motivation. He also gives an example of a state-desire that does not encompass motivation to act. By his own reasoning, we can conclude that state-desires do not encompass motivation either.
Mele also argues that action-desires are importantly different from both state-desires and beliefs because they essentially encompass motivation to act. Yet his argument for the claim that state-desires do not essentially encompass motivation to act also supports the conclusion that no action-desire essentially encompasses motivation to act.
It is conceivable that an action-desire does not encompass motivation to act. Consequently – again by Mele’s own reasoning – action-desires do not encompass motivation. Since action-desires and state-desires together exhaust the category of desires, it follows from Mele’s arguments that no desires encompass motivation. This consequence is unacceptable. Hence Mele’s arguments fail to prove a distinction between desires and beliefs that might be of use in defending a Humean account of motivation.
Works Cited
Dancy, Jonathan. 1993. Moral Reasons. Oxford: Blackwell.
Davidson, Donald. 1963. “Actions, Reasons, and Causes.” Journal of Philosophy 60:
685-700.
Mele, Alfred R. 2003. Motivation and Agency. Oxford: Oxford University Press.
[1] One might object that this is too strong a statement of Mele’s view. He says that one of his central theses is that “first-person cognitivist moral ought-beliefs do not essentially encompass motivation to act accordingly,” and even with regard to this, he only claims that “the grounds for holding that we have such beliefs are much weaker than the grounds for denying that we do” (107). This suggests that Mele’s thesis is quite different – and much less controversial – from the one that I ascribe to him. Below I argue that Mele argues for the claim that beliefs are not contingently motivation-encompassing either, and by means of the same argument that he uses for the first claim. Mele’s view is that the “grounds for holding that we have” even contingently motivation-encompassing beliefs “are much weaker than the grounds for denying that we do,” and this claim is central to his (and any) defense of a Humean account of motivation.