Motivation: Essentially Motivation-Constituting Attitudes

Philosophical Review

Motivation: Essentially Motivation-Constituting AttitudesAuthor(s): Alfred R. MeleSource: The Philosophical Review, Vol. 104, No. 3 (Jul., 1995), pp. 387-423Published by: Duke University Press on behalf of Philosophical ReviewStable URL: http://www.jstor.org/stable/2185634 .

Accessed: 25/06/2014 10:30

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

Duke University Press and Philosophical Review are collaborating with JSTOR to digitize, preserve and extendaccess to The Philosophical Review.

http://www.jstor.org

This content downloaded from 185.44.78.156 on Wed, 25 Jun 2014 10:30:10 AMAll use subject to JSTOR Terms and Conditions

http://www.jstor.org/action/showPublisher?publisherCode=duke

http://www.jstor.org/action/showPublisher?publisherCode=philreview

http://www.jstor.org/stable/2185634?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp


The Philosophical Review, Vol. 104, No. 3 (July 1995)

Motivation: Essentially Motivation-Constituting Attitudes

Alfred R. Mele

The term 'motivation' has considerable currency both in moral philosophy and in the philosophy of mind. It appears in debates between internalists and externalists about moral judgments and moral reasons, in the related controversy over moral realism, and in explanatory schemes for purposive behavior offered in the philosophy of mind. But what is motivation? My aim in this paper is to elucidate a notion of motivation associated with a popular perspective on intentional conduct, a perspective that accords states of mind an important causal/explanatory bearing on such behavior. My strategy is to develop an account of a central species of motivation. Space constraints and my own limitations preclude my offering a complete account of motivation here; however, the account to be developed illuminates motivation as a whole.

1. What is the Question?

Compare the questions "What is motivation?" and "What is knowledge?" The latter admits of a reformulation-"What is it for S to know that p?"-that makes the structure of propositional knowledge explicit: such knowledge involves a subject, an object, and some relationship between them. Is there a similarly revealing reformulation of the question about motivation? Consider the question "What is it for S to be motivated to do x?" Here, we find structure, but we also find ambiguity. Being motivated to do x is sometimes understood as requiring a motivated doing of x: there is a success reading of 'motivated to do x'. Alternatively, when we say that someone is (or was) motivated to do x we sometimes mean only that she has (or had) some motivation to do x. One may also worry that this question, in mentioning doing, is too narrow as a reformulation of the original question. People desire such things as world peace and financial security; but even though such things are not doings, many such desires seem to be motivational states.

387



ALFRED R MELE

One may try to recast the original question as "What is a motivational state?" But there is ambiguity here, as well. Ann desires to acquire x and she believes that she will acquire it only if she does A. She does A, desiring to do A and being motivated to do precisely that. Is the belief itself a motivational state in virtue of the role it plays in Ann's motivated doing of A? That depends on how 'motivational state' is to be read. On a broad reading, any attitude that plays a major, typical role in producing a motivated action may count as a motivational state.' Interpreted more narrowly, motivational states are limited to states that themselves constitute motivation. Plainly, Ann's belief that she will acquire x only if she does A is not itself constitutive of motivation to A.

The question "What is a motivation-constituting state?"-or "What is it for a state to be motivation-constituting?"-avoids these ambiguities. The metaphysics of constitution need not be an issue here: for a readership lacking perfect linguistic pitch, the question might have been "What is it for something to be a motivation?" Assuming that all motivation-constituting items are representational states of mind-or "attitudes," for short-the question may be phrased in terms of attitudes. Later, I will raise and answer a more specific version of the question-about essentially motivation-constituting attitudes.

2. A Perspective on Action

Our commonsense notions of motivation and intentional action are intertwined. A thorough defense of a theory of motivation would require a rigorous defense of a theory of action. Given the magnitude of the latter project, here I simply adopt a popular perspective on intentional action that I have defended at length elsewhere (Mele 1992). At the heart of this perspective-"perspective P"-is a pair of theses: (1) all intentional actions are caused (but not necessarily deterministically so);2 (2) among the causes, in all cases, are mental items (events or states), including motivation-constituting attitudes.

Perspective P is an alternative to at least three others. In one (P1), intentional actions, or some of them (e.g., free actions), are

'For a broad reading of this kind, see Dancy 1993, 2. 2Probabilistic causation is an option.

388



MOTIVATION

not caused and are to be given noncausal explanations partly in terms of mental items (see, e.g., Melden 1961). In another (P2), all intentional actions may be caused, but explanations in terms of such things as beliefs and desires (normally) are noncausal explanations (see, e.g., Peters 1958 and Wilson 1989). In yet another (P3), intentional actions are caused, but mental states and events are involved neither in the causation nor in the explanation of intentional actions (see, e.g., Skinner 1953). Obviously, proponents of these perspectives are not egalitarian about them; lively, instructive debates have been staged over their relative merits. Having entered the fray elsewhere, I steer clear of it here. My aim is to illuminate a notion of motivation as it functions in perspective P. Incidentally, an approach to understanding behavior that features mental items in causal roles is by no means peculiar to philosophers. As psychologist Douglas Mook reports, the contemporary, "cognitive," approach to motivation in his field investigates "the- ories about internal mental operations as part of the causation of behavior" (1987, 53; my italics).

Each of the four perspectives identified faces familiar problems. Those who embrace perspective P usually do so as part of a naturalistic stand on agency. As they see it, mental items that play a causal/explanatory role in intentional conduct bear some important relation to physical states and events. Relations that have been explored include identity (both type-type and token-token) and various kinds of supervenience. Each kind of relation faces diffi- culties, and some theorists have argued that anything less than type-type identity (which encounters familiar problems) will fail to avert an unacceptable epiphenomenalism.3 There is a general problem about the causal relevance of the mental in a naturalistic framework. This is not to say, however, that the problem is insol- uble.4

Theorists attracted to P1 or P2 face the difficult task of devel- oping an adequate notion of noncausal belief/desire explanation of action. Three decades ago, Donald Davidson issued an important challenge to proponents of these perspectives. In a nutshell, it is this: Given that when we act intentionally we act for reasons,

3For a useful collection of essays on these themes, see Heil and Mele 1993.

4For my own attempt at a resolution, see Mele 1992, chap. 2.

389



ALFRED R. MELE

provide an account of the reasons for which we act that does not treat (our having) those reasons as figuring in the causation of the relevant behavior (Davidson 1963). Apply the challenge to a specific case. Ann has a pair of reasons for giving Bob a failing grade: first, his work merits that grade, as she knows, and she desires to give her students the grades they deserve; second, she knows both that her failing Bob would make him ineligible for a game against her son's team and that Bob's team would probably lose without Bob, and she would love her son's team to win. As it happens, Ann fails Bob for exactly one of these reasons, R. In virtue of what is it true that she failed him for R, and not for R*, if not that R (or her having R), and not R*, played a suitable causal role in her failing him?5 Although I know of no compelling argument that Davidson's challenge cannot be met, I do know of arguments- arguments I regard as compelling-that the most detailed attempts to meet it have failed in instructive ways (Mele 1992, chap. 13).

A major problem for P3, as part of a behaviorist perspective, is standard fare in textbooks on motivational psychology. An animal sometimes behaves in quite different ways under exactly similar external conditions. The link between stimulus and response in these cases is mediated by unobserved internal states (Mook 1987, 58-59). Considerable support may be found in most recent dis- cussions of animal learning (e.g., Toates 1986, chaps. 6 and 7) for the hypothesis that even in such animals as rats, relevant internal states include mental representations of goals and means. Of course, P3 is not solely a behaviorist perspective. Old-style epiphenomenalism (i.e., epiphenomenalism combined with substance du- alism) may endorse it as well; but a dangling mental life, one causally and explanatorily inert with respect to overt behavior, is not an attractive prospect.

I do not wish to suggest that a theorist must accept without qual- ification both of the central elements of perspective P to be attracted by the position on motivation that I will advance. However,

5One might urge that there is no fact of the matter about which reason the agent acts for in cases of this kind. But, surely, in at least some cases like Ann's, we do act for one of our reasons and not for the other. As it happens, Ann's moral character (like mine, I hope) is such that she would not fail a student for the sake of a victory by her son's team.

6A referee for this journal observed that although Ginet 1990 advances an alternative to perspective e, it is not clear that anything Ginet says there precludes his endorsing the view of motivation to be developed here. For

390



MOTIVATION

lacking the space here to defend the perspective I favor, and find- ing it impossible to tackle certain important questions about motivation from a neutral perspective, I deem it best to lay my cards on the table.

3. Motivation and Desire

Here are six popular theses in the philosophical and psychological literature on motivation.

1. Motivation is present in the animal kingdom, but does not extend throughout it.7

2. Motivated beings have a capacity to represent goals and means to goals.

3. A motivational attitude may have either a goal or a means to a goal as its object. (In common philosophical terms, there is both intrinsic and extrinsic motivation.)8

4. Motivation varies in strength. 5. The more strongly motivated agents are to do A, in compar-

ison to alternative courses of action, the more likely they are to do A, other things being equal.9

some relevant points about action-explanation on which Ginet and I dis- agree, see Mele 1992, 250-55.

7For an informative discussion of how the lines are to be drawn, see Toates 1986, chap. 2. For a statement of the prevailing view in motivational psychology, see Mook 1987, 104-5.

81n the philosophical literature, the distinction between intrinsic and extrinsic (or instrumental) motivation typically is framed in terms of desires or wants. An intrinsic desire is a desire of something for its own sake, or as an end. A wholly intrinsic desire is a desire of something as an end and not also as a means to, or as a constituent of, or as evidence of, something else. (S may desire her happiness wholly intrinsically while also desiring that she continue to be healthy, both as an end and as a constituent of, or means to, her happiness.) An extrinsic desire is a desire for something for its believed conduciveness to something else that one desires, where conduciveness is not limited to means/end relations and includes part/ whole relations. To the extent to which S desires living courageously as a constituent of living virtuously, the former desire is an extrinsic desire, in the specified sense. Of course, S might also desire living courageously for its own sake-that is, intrinsically. Not all Extrinsic desires are wholly extrinsic. For discussion of various notions of "intrinsic motivation" and "intrinsic behavior" in the psychological literature, see Heckhausen 1991, 403- 13. Also see Deci and Ryan 1985.

9PsychologistJohn Atkinson elegantly articulates an alleged connection between motivational strength and behavior: "The act which is performed

391



ALFRED R MELE

6. Whenever agents act intentionally, there is something they are motivated (in the "success" sense) to do.'0

Divergent conceptions of motivation may be at work in philosophical thought about motivational matters. My concern in this paper is a notion of motivation consistent with (among other things) these six claims. Readers who accept all six claims are likely to be more interested in the result than those who do not; but I conjecture that most readers do accept all of these claims, on some reading or other.

R. S. Peters concludes The Concept of Motivation by chiding motivational psychologists for their scientific aspirations and suggesting that "scrutiny of the conceptually illuminating start made by Aris- totle" would provide a more promising approach (1958, 157). One of the chief faults that Peters finds with motivational psychologists is their commitment to a causal approach to explaining behavior. I suspect that, when writing his conclusion, Peters forgot a striking claim in book 6 of Aristotle's Nicomachean Ethics "the origin of action-its efficient, not its final cause-is choice, and that of choice is desire and reasoning with a view to an end" (1139a31-32).

The term 'orexis', translated here as 'desire', also may be translated as 'motivation'. One reason that 'desire' is more convenient than 'motivation' as a standard translation for 'orexis' is that it al- lows more easily for the precise specification of the object of a motivation-constituting attitude. Another is that the noun 'orexis' has a cognate verb, 'orego', and both can be translated as 'desire', but

among a set of alternatives is the act for which the resultant motivation is most positive. The magnitude of response and the persistence of behavior are functions of the strength of motivation to perform the act relative to the strength of motivation to perform competing acts" (1957, 361). Com- pare this to a familiar thesis advanced by Donald Davidson: "If an agent wants to do x more than he wants to do y and he believes himself free to do either x or y, then he will intentionally do x if he does either x or y intentionally" (1980, 23). For criticism of this thesis and a defense of an alternative, see Mele 1992, chap. 3.

10Notice that claim 6 does not assert that every intentional action is a motivated action. It takes no stand, for instance, on whether certain un- wanted actional side-effects of intentional actions are themselves intentional actions. For example, when I play tennis in my tennis shoes, I knowingly shorten the "life" of my shoes, while having no desire to do that. Claim 6 is silent on the question whether I intentionally shorten the life of my shoes; but it insists that some relevant intentional action is motivated at this time: for example, my playing tennis, my playing tennis in tennis shoes.

392



MOTIVATION

not, of course, as 'motivation'. Setting classical Greek aside, related points may be made about philosophical uses of the terms 'desire', 'want', and 'motivation'. We can say that someone desires to swim today, or, more naturally, that she wants to swim today; and we can say that someone has a desire to swim today, or, less naturally, that she has a want to swim today. However, the expressions 'is motivated to swim today' and 'has motivation to swim today' are not only more awkward than their most elegant counterpart, they are also less precise. Motivation to swim today includes a desire to swim today, but it also is provided by other desires that support that desire. If Ann wants to swim today because she wants to exercise today (as she does every day) and believes that, given a tendon injury sustained in her daily running, swimming currently is the best form of exercise for her, her desire to exercise today provides her with motivation to swim today. Further, the claim that she is motivated to swim today lets in both her desire to swim today and her desire to exercise today as states motivating her to swim. Partly for this reason, some philosophers have adopted the convention of using the noun 'desire', or 'want', as a synonym for 'motivation'.

This practice occasionally troubles aficionados of ordinary usage. 'Desire' brings to mind affective tone or appetitive content, and 'has a want to A' is stilted. These worries can be quelled. As I noted elsewhere (Mele 1992, 47), distinctions blurred by this usage can be recaptured by differentiating among types of wants or desires (which terms may be used interchangeably as terms of art)-for example, egoistic versus altruistic desires, affective versus non-affective desires, appetitive versus non-appetitive desires (compare my desire to eat pizza and drink beer with your desire to continue reading this essay). However, there are other, more substantive worries, as I will explain later.

Not all desires for x constitute motivation for the agent to pursue x. Someone who desires world peace may be convinced that there is nothing she can do to promote it. Consequently, she may be in no way inclined to make efforts aimed at promoting it. Still, given that she desires world peace, news of progress on that front would be expected to bring her some pleasure; and partly because she desires world peace, she may be inclined to seek out such news, or at least to attend to it when she happens to encounter it in the media. But then there is something that the desire for world peace motivates her to do (at least in a sense of 'motivates' that does not

393



ALFRED R MELE

require success): namely, to attend to news of progress on that front. Not all of our desires are for actions of ours, and not every desire, D, constitutes or provides motivation for actions aimed at bringing about D's object. But desires may motivate actions that are not so aimed, as the preceding example illustrates.

The foregoing raises a problem for some attempts to characterize desire in terms of "direction of fit," an expression around which a cottage industry has grown." On one notion of satisfaction that ap- plies both to beliefs and to desires, a belief that p and a desire that p have the same satisfaction condition: its being the case that p.'2

But whereas many desires are functionally fit to contribute to their own satisfaction, in this sense, relatively few beliefs are. (Potentially self-fulfilling beliefs are the exception, not the norm.) The point is sometimes expressed in terms of "direction of fit": "belief aims at truth" (Williams 1973, 151), it aims to fit the world; desire aims at getting the world to fit it. In this connection, Michael Smith writes: "the difference between beliefs and desires in terms of direction of fit comes down to a counterfactual dependence of a belief and a desire that p, on a perception that not p: roughly, a belief that p is a state that tends to go out of existence in the presence of a perception that not p, whereas a desire that p is a state that tends to endure, disposing the subject in that state to bring it about that p" (1987, 54; cf. Pettit and Smith 1990, 574).

Smith's characterization of desire is unsuccessful. Standing at the airport, waiting for Angela's plane to arrive, it occurs to me that I should have called from home to see whether her plane departed at the scheduled time. I now desire that Angela's plane departed on time; but this desire in no way disposes me to bring it about that the plane left on time, since it does not dispose me to endeavor to change the past.'3 This failing of Smith's characterization

"There is a useful discussion of the literature in Humberstone 1992. '21t is controversial whether a belief and a desire can have the same

content. Alvin Goldman, for example, takes the contents of beliefs to be declarative propositions and the contents of desires to be optative propositions (1970, 101-2). For related views, see Castafieda 1975, chap. 10; Hare 1972, chap. 3; and Kenny 1989, 36-41. The point to be made can be re- formulated in a way consistent with these views.

13Smith, responding to the potential objection that "desire is not a suit- ably broad category of mental state to encompass all those states with the appropriate direction of fit," says that we may replace 'desire' with 'pro- attitude' (1987, 55). Notice that, standing at the airport, I had a pro-atti-

394



MOTIVATION

leaves it open, however, that all action-desires that p dispose their agents to endeavor to bring it about that p. My action-desires "that p9 are desires "that I A," where 'A' is an action variable. I return to this issue shortly.

Another problem for Smith's characterization of desire also is problematic for the idea that the noun 'desire' (or the noun 'want') is a synonym for 'motivation'. Although my desire that An- gela's plane left on time might motivate me to check a video mon- itor at the airport in order to reassure myself about this, desires are imaginable that do not motivate at all, in any legitimate sense. Connie is a fan of the New York Giants. The Giants are playing their traditional rivals next month and she hopes they will win. Hoping is a species of desiring;'4 but, as Connie knows, she has no chance of discovering how the game turns out: for she is about to begin piloting a spaceship to a remote galaxy and she knows that by game-time she will be permanently cut off from Earth's sporting news. Robert Audi has claimed that someone who wants p "tends to think (reflect, muse, or the like) or daydream about p at least occasionally, and especially in idle moments" (1973, 4). If this is true, we may plausibly suppose that this cognitive behavior is motivated by the want or desire that p. However, Connie is a special person; when she knows that she has no chance of learning how matters of relatively minor 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. Even so, if during her flight Connie were, per impossible, to learn that the Giants had won, she would be pleased by the news-an indication that it continues to be true of her that she desires a Giants victory. Appar- ently, Connie's desire for a Giants victory is motivationally inert (under her actual circumstances).'5

tude toward the plane's having departed on time; but that pro-attitude in no way disposed me to bring it about that the plane left on time.

14The American Heritage Dictionary reports, in the usage note on 'expect', that "to hope is to desire, usually with confidence in the likelihood of gain- ing what is desired." (Incidentally, I believe that the final clause is too strong.)

'51t may be observed that if, during the flight, Connie's co-pilot were to ask whether she wants the Giants to win the game, she would answer in the affirmative (other things being equal). However, her so answering is not motivated by her desire that the Giants win. When a friend asks me whether I want to complete this paper by the end of the century, and I truthfully answer yes, it is not my desire to finish the paper that motivates

395



ALFRED R MELE

Motivationally inert desires are exceptional. Most of our desires dispose us to do something or other. Many dispose us to try to bring about what we desire. Some dispose us instead to seek, or attend to, information about the desired items (e.g., world peace, when we deem ourselves incapable of promoting it), or to reflect or muse about the objects of our desires. Some desires dispose us to do all of this and more.

A distinction between "standing" and "occurrent" desires or wants merits attention (Alston 1967; Goldman 1970, 86-88). As the verb 'want' is standardly used, 'She wants her children to do well' may be true of a devoted mother even while she is wholly absorbed in writing a metaphysical treatise or dreamlessly sleeping. At such times, the mother is not thinking about her children; nor need some representation of her children's doing well be stored in a mental desire-bin. Rather, its being true of the woman, even at these times, that she wants her children to do well consists in her being disposed to have so-called "occurrent" desires for her children's welfare. Such a disposition is a standing want or desire.

Occurrent desires are not, as the name may suggest, occurrences or events: they are states. One alleged difference between occurrent and standing desires is that the former, and the former alone, are present to consciousness (Goldman 1970, 86). However, this idea is doubly mistaken. First, we are conscious of some of our standing desires. Ann, a coffee drinker, is disposed to acquire an occurrent desire to drink coffee when she smells coffee. While reading an article about the presence of this disposition in many coffee drinkers, it occurs to Ann that she has this disposition. She is aware or conscious of this disposition of hers-that is, of this standing desire. Second, part of what the distinction between standing and occurrent desires is supposed to mark is the difference between desires that influence behavior only if they become "activated" in some way and desires that are already active; but not all active desires are present to consciousness-at least if common sense and clinicians may be trusted.'6 Ann might have spoken as

my so answering, but a desire to speak truthfully (or something of the sort). One may suggest that desires that p, being intentional attitudes, are, in part, thoughts about p. However, this suggestion would not support the claim that Connie's desire for a Giants victory motivates her "to think . . . or daydream about" a Giants victory. Rather, the suggestion is simply that a desire that p is a certain kind of thought about p.

16For discussion of some relevant empirical work, see Vollmer 1993. If

396



MOTIVATION

she did to Bob because she wanted to hurt his feelings, even though she was not conscious at the time of her desire to do that.

Standing desires for x are dispositions to acquire occurrent desires for x. Occurrent desires are "active" desires-desires that do not need to be aroused in order to exert an influence on the agent's conduct. The dispositions that constitute standing desires have no explicit representational content. They are accorded a content-supplying description in virtue of the content of the occurrent desires in which their "activation" consists. Having no explicit representational content, standing desires are not explicit attitudes.i7 Rather, they are dispositions to have explicit attitudes of a certain kind. They are, we may say, "implicit" motivational attitudes-provided that explicit attitudes in which their activation consists are themselves motivational attitudes, and not motivationally inert desires. (Discussion of motivational attitudes in this paper should be presumed to be about "occurrent" attitudes unless explicitly noted otherwise.)'8

there are occurrent desires that are not present to consciousness, are there also occurrent beliefs that are not present to consciousness? I leave this issue open.

17Gilbert Harman writes, "one believes something explicitly if one's belief in that thing involves an explicit mental representation whose content is the content of that belief"; later, he speaks of something "explicitly represented in one's mind" as "written down in Mentalese as it were, without necessarily being available to consciousness" (1986, 13-14). This is the general picture I mean to evoke with the expression 'explicit attitudes'; but I venture no opinion about how the contents of attitudes are explicitly represented, nor do I claim that they are all represented in the same way (e.g., in a language of thought). Incidentally, my subsequent use of 'implicit' in 'implicit motivational attitudes' does not parallel Harman's use of the term in his expression 'implicit beliefs' (e.g., 13).

'8Suppose I compose a long shopping list, forget to bring the list to the market, and, owing to imperfect memory, fail to buy one of the items on my list-say, yogurt. When I stroll through the dairy section and pick up butter but not yogurt, I presumably have no occurrent desire to buy yogurt. It has been suggested to me (by one of the editors) that I do, however, have at the time a desire to buy yogurt-a standing one. I find this im- plausible. As I see it, just as my belief that yogurt is on the list is extin- guished, so is my desire to buy yogurt. To be sure, when I return home and compare my list with my groceries, I might say-truly-"Shucks, I wanted to buy yogurt"; but, of course, there was a time at which I had this desire-a time prior to my strolling through the dairy section. (A thorough defense of my way of distinguishing between occurrent and standing desires would require more space than the issue merits here. The main theses to be defended in this paper are consistent with more traditional ways of

397



ALFRED R. MELE

4. The Place of Desire in Essentially Motivation-Constituting Attitudes

Jonathan Dancy attributes to "Humeans" about motivation the thesis that desires "cannot exist without motivating"; "they are essentially or necessarily motivating states" (1993, 2).19 Although there is a significant element of truth in this attribution, it is a bit wide of the mark. Obviously, 'motivating' here is not used in its success sense; Dancy means 'motivation-constituting'. Even so, the thesis is false and should be eschewed by good "Humeans": some desires, as I have argued, are not motivation-constituting at all, much less essentially so. Here I turn to what may be termed, with Dancy, "essentially" motivation-constituting attitudes-or "EMC attitudes," for short.

Essentially and contingently motivation-constituting attitudes may be distinguished as follows. An attitude of an agent S is essentially motivation-constituting if and only if it is motivation-constituting not only in Ss actual situation but also in all possible scenarios in which S has it. An attitude of an agent S is contingently motivation- constituting if and only if although it is motivation-constituting in Ss actual situation, it is not so in some possible scenarios in which S has it. Thus, my current desire for a Giants victory over the Eagles next month is contingently motivation-constituting; for it is in fact motivation-constituting but it would not be so in a possible scenario in which I am like Connie in the pertinent psychological respect and am co-piloting her spaceship.

Detailed exploration of the nature and range of contingently motivation-constituting attitudes is a project for another paper. Some theorists have urged that some beliefs-construed as cognitivist or "truth-seeking" attitudes-are, or might be, motivation- constituting (Dancy 1993, McDowell 1978, McNaughton 1988, Na- gel 1970). The question whether there are essentially motivation- constituting beliefs is examined shortly. But it is worth noting that the existence of contingently motivation-constituting cognitivist beliefs is compatible with the "Humean" idea that essentially motivation-constituting attitudes are found only among desires and atti-

making the distinction, but some of the traditional details are objectiona- ble, for the reasons I offered.)

'9Cf. McNaughton 1988, 49, on "essentially motivational states."

398



MOTIVATION

tudes having desires as constituents. Although I comment further on this compatibility later, my primary concern is EMC attitudes.

In this section, I motivate a trio of theses about EMC attitudes.

Ti. Any desire having the agent's acting in some way or other as a represented object-any "action-desire"-is an EMC attitude.

T2. All EMC attitudes have the agent's acting in some way or other as a represented object.

T3. In human beings as they actually are, any EMC attitude is, or is partially constituted by, an action-desire. More specifically (letting 'A' be an action variable), in human beings as they actually are, any EMC attitude having the agent's A-ing as a represented object is, or encompasses, a desire to A.

I start with T1. An "action-desire," in the specified sense, is an agent's desire regarding how she herself will act; desires for others to act in certain ways are not "action-desires," in this sense. If desiring to A, where 'A' is an action variable, were not to preclude the presence of certain cognitive states, one might desire to A while having no motivation to A. For example, if an agent could desire to A while being convinced that she cannot A, she might not be the least bit inclined to try to A. However, desiring to A does preclude the presence of cognitive states of this kind. If I am convinced that I cannot travel faster than the speed of light, or 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. Achieving the represented objects of action-desires is doxastically open for the agent: if she does not explicitly believe that she can A, at least she is not convinced that she cannot A.20 Any desire to A (A being an action variable), by its very nature, inclines the agent, in some measure, to A intentionally, or to try to A, or to try to put herself in a position to A. This is part of what it is to be an action-desire.2' In so inclining an agent,

20For an argument that when desiring to A we do not always explicitly believe that we can A, see Mele 1992, 62.

211 assume a view of the attitudes according to which every belief that p is essentially a belief that p, every intention to A is essentially an intention to A, every action-desire to A is essentially an action-desire to A, and so on. Thus, for example, Ss belief that p in Ss actual situation is not something other than a belief that p in any other possible scenario, and Ss action- desire to A in his actual situation is not something other than an action- desire to A in any other possible scenario.

399



ALFRED R MELE

an action-desire constitutes motivation to A. Of course, as desires are commonly conceived, someone who desires to A may also desire to B and recognize that she cannot do both.22 But being inclined, in some measure, to A intentionally is compatible with also being inclined, in some measure, to B intentionally, recognizing that one cannot both A and B.23

Now, for T2. Take any desire whose represented object does not include the agent's acting: for instance, Al's desire, D, for the welfare of his children. If Al were to believe that he can contribute (in ways acceptable to him) to his children's welfare, then, given D's presence, he would have some motivation to try to contribute to it. But his having that motivation is not ensured by the presence of D itself. In Al, D might be accompanied by the conviction that there is nothing he can do to contribute to his children's welfare; for example, Al might be on Connie's spaceship, with no hope of influencing his children's life on Earth. In that case, Al would have no motivation to try to contribute to his children's welfare. Of course, even if Al is convinced that he can do nothing about their welfare, he may be motivated (i.e., have motivation) to think or wonder about how his children are doing; but the presence of D itself does not ensure the presence of such motivation, as the original spaceship example indicates. Al might be psychologically so constituted that he is not the least bit inclined to think about the object, x, of a desire of his unless he believes that he has some chance of contributing to x, or, at least, of acquiring information about x.

I have just argued that, among desires, only those having the agent's acting in some way as a represented object (i.e., only action- desires) are EMC attitudes. Should there be EMC attitudes other than desires, my argument would leave it open whether they-or

22Frank Jackson rejects this (1985). For a reply to Jackson, see Mele 1992, 48-50.

23Although desiring to A entails being inclined (in some measure) to A intentionally, or to try to A, etc., the converse is false. For example, my desire to run a mile this evening inclines me to run the fourth 220 yards intentionally, the third quarter-mile intentionally, etc. However, my being thus inclined to run, say, the fourth 220 yards does not require that I have a desire specifically to run that segment. My desire to run the mile is desire enough.

400



MOTIVATION

some of them-falsify T2. The question whether such attitudes exist is addressed shortly.

Given T1 and the existence of action-desires, at least some EMC attitudes are action-desires. That takes us part of the way to T3. Whether, in actual human beings, all EMC attitudes are, or encompass, action-desires remains to be seen. Do we have EMC attitudes toward A-ing that are not themselves desires to A and are not even partially composed of such desires? An argument for a negative answer may proceed by elimination.

Various "practical" attitudes having the agent's A-ing as a represented object may be considered. Hoping to A, as I have already indicated, is constituted in part by desiring to A. The same is true of wishing to A. How about intending to A?

Of course, only if intentions are motivation-constituting can they be a problem for T3; but intentions traditionally have been placed in the conative sphere. On a once-popular desire/belief reductionism about intentions, intentions are every bit as motivational as the desires that partially constitute them. However, such reductionism has encountered powerful objections in recent years, objections I take to be persuasive.24 Intentions, on my own view, are executive attitudes toward plans (Mele 1992, chaps. 8-11). Although the representational content of intentions is isomorphic with that of some desires, the intending and desiring attitudes differ. Intentions to A are states of being settled (not necessarily irrevocably) upon A-ing; but one can desire to A, and even desire to A more strongly than one desires not to A, without being settled upon A-ing. (Consider an agent who wants more strongly to respond in kind to a recent insult than to refrain from doing so, but who, owing to moral qualms, is deliberating about what to do and is, as yet, unsettled about whether to retaliate.) The settledness feature of intentions is an essential feature of the intending attitude, a feature not essential to the desiring attitude. But it does not follow from this difference that intention does not incorporate desire. Intending to A, as I view it, encompasses desiring to A (but without being reducible to a combination of desire and belief).25 Part of what it is to be

241n its simplest version, the view at issue reduces an intention to A to (i) a desire to A that is stronger than any competing desire and (ii) a belief that one will A. For criticism of this and various refined versions, see Brand 1984, Bratman 1987, and Mele 1992, esp. chaps. 8-10.

250n a standard view of desire, the psychological features of desires to A in virtue of which they contribute to intentional A-ings are their content

401



ALFRED 1 MELE

settled upon A-ing is to desire to A; lacking a desire to A, one lacks an element of a psychological commitment to A-ing that is intrinsic to being settled upon A-ing, and therefore to intending to A. To be sure, we can coherently say such things as "I intend to pay that unjust traffic fine, but I don't desire to pay it." But here we retreat to a narrow sense of 'desire' alluded to earlier. Even if I grudgingly pay the fine, I have an instrumental desire to pay it. If intentions to A encompass desires to A, intentions are no problem for the thesis (T3) that in real human beings, any EMC attitude having the agent's A-ing as a represented object is, or encompasses, a desire to A.

Do some beliefs undermine this thesis? A noncognitivist might urge that to form a belief that I ought to A is to issue a command to myself (the command to A), or, more prosaically, to form an intention to A.26 One might contend that in that case, what would satisfy the belief is what would satisfy the intention: namely, my A- ing. It is not my purpose here to close the door on this option. (I have criticized it elsewhere.)27 If intentions are EMC attitudes, a theorist who makes a case for the thesis that to form beliefs of certain kinds is to form intentions will have made a case for the thesis that some beliefs are EMC attitudes. Notice, however, that in constituting intentions, these alleged beliefs encompass desires, if I am right about intentions, and therefore raise no problem for T3. My present concern is "ordinary" beliefs-beliefs satisfied by their being true. That is, my concern is belief on a cognitivist construal.28 Noncognitivist construals of my belief that I ought to A as a desire to A, or as an attitude that encompasses an action-desire (e.g., an

and their strength. On my view of the contribution of intentions to A to intentional A-ings, the "settledness" feature of intentions is crucial and cannot be captured in terms of desire strength (and content), nor in terms of this plus belief (Mele 1992, 76-77 and chap. 9).

26For a putative connection between assenting to an ought-judgment and assenting to a self-command, see Hare 1963, 79. For the view that intention is a species ofjudgment, see Davidson 1980, 99-102; cf. Davidson 1985, 206. (I take Hare's references to self-commands to be metaphorical references to intention-formation.)

27Mele 1995, chap. 2. For criticism of the idea, in particular, that intention is a species of evaluative judgment or belief, see Bratman 1985 and Mele 1987, 43-44.

28"Cognitivism" about ought-beliefs places them among the truth-val- ued attitudes. Cognitivists need not hold that ought-beliefs are not motivation-constituting.

402



MOTIVATION

intention to A), obviously cannot falsify T3. And noncognitivists about such beliefs have no reason to hold that they constitute motivation to A without constituting or encompassing a desire to A.

So are any cognitivist beliefs EMC attitudes? The primary can- didates are cognitivist moral ought-beliefs-specifically, agents' (cognitivist) beliefs that they themselves ought, morally, to A. Some apparent problems for the thesis that such beliefs are (essentially) motivation-constituting have been explored in detail by opponents and proponents of internalist moral realism. The problems of amoralism and wickedness are cases in point.29 Neither side in the debate has been bowled over by the opposition's treatment of these deviations from moral health. However, another apparent difficulty has not received the attention it deserves. I call it "the problem of listlessness. "30

Consider an unfortunate woman, Eve, suffering from severe listlessness or clinical depression owing to the recent tragic deaths of her husband and children in a plane crash. It seems conceivable that she retains certain of her beliefs about what she morally ought to do while being devoid of motivation to act accordingly. Eve has aided her ailing uncle for years, believing herself to be morally required to do so. Her memory has not been damaged, and she may deem her circumstances not to have changed in such a way as to alter her moral obligations to her uncle. So perhaps she continues to believe that she ought, morally, to help him but now is utterly without motivation to do so, owing to her depression. Her depression, it seems, may consist partly in the utter absence of motivation to help her uncle, or it may deprive her of all motivation to help him. If this is possible, people may believe that they ought, morally, to A while having no motivation to A. Now, surely, no actual human being is essentially immune to clinical depression, even in its most severe forms. And if some possible (i.e., for us) instances of clinical depression or listlessness are such that people who believe that they ought, morally, to A devoid of motivation to A, then moral ought-beliefs are not essentially motivation-constituting for actual human beings. No matter what we believe we

29See McNaughton 1988, chap. 9; Milo 1984, chaps. 6 and 7. Cf. Brink 1989, 46-50, 59-60, 83-86; Dancy 1993, 3-6. Stocker 1979 provides a sem- inal critique of internalism.

30This problem leads Dancy to reject the thesis that moral ought-beliefs are essentially motivation-constituting (1993, 6).

403



ALFRED I?. MELE

morally ought to do, we would be devoid of motivation to do it were we to suffer from listlessness of the kind at issue while holding the belief. Perhaps, as Donald Davidson has claimed, there is a "tendency to say that if someone really (sincerely) believes he ought, then his belief must show itself in his behaviour (and hence, of course, in his inclination to act, or his desire)" (1980, 27). But if the tendency stems from myopic attention to garden-variety instances of believing that one ought, its existence cuts little philosophical ice. Perhaps people normally try to do what they believe they ought to do (and hence are inclined or motivated so to act). However, in normal cases, people do not suffer from clinical depression or listlessness.

Most readers, I conjecture, will find Eve's case conceptually coherent. Among readers who do, some may be willing to counte- nance contingently motivation-constituting moral ought-beliefs, and others may see this as mistaken.3' In any case, these readers understand (cognitivist) belief in such a way that whether an agent who believes that she ought, morally, to A has motivation to A is contingent upon features of her motivational condition (e.g., whether she suffers from listlessness of the kind at issue) that are conceptually independent of her so believing. There is nothing radical about this position, of course. It accords with a traditional, "Humean," division-of-labor view according to which motivation is constituted (even contingently) only by desires, and never by beliefs (see, e.g., Lewis 1988, Smith 1987). And the version of this position that leaves open the possibility of contingently motivation- constituting cognitivist beliefs makes a significant concession to anti-"Humeans," as I will explain shortly.

Are readers who deem Eve's case conceptually coherent begging the question against the thesis that there are EMC cognitivist moral ought-beliefs-thesis M? That is an odd question. They are merely reacting to the case in light of a familiar understanding-their own understanding-of (cognitivist) belief and motivation. Further, it is not as though pre-theoretical intuition favors M over, for example, the alternative thesis that our cogiiitivist beliefs that we ought, morally, to A tend, in conjunction with relevant desires (perhaps standing, generic desires to do what we morally ought), to generate EMC attitudes toward A-ing without the beliefs themselves consti-

31I take no stand here on the relative merits of these two reactions.

404



MOTIVATION

touting motivation. Indeed, to the extent that a "Humean" view of belief and desire-a view according to which (cognitivist) belief and desire inherently differ in their respective directions of fit-is reflective of pre-theoretical intuition, such intuition speaks against M. And if there are pre-theoretical intuitions about these matters, the popularity of the "Humean" view suggests that its way of carv- ing up the mind is more strongly supported by them than Ms way is. Proponents of anti-"Humean" theses like M are urging a reformative conception of cognitivist belief.

It can be said, in a rough and ready way, that any attitude that essentially constitutes a disposition of an agent to A intentionally, or to try to A, or to try to put herself in a position to A, essentially constitutes motivation to A. Assuming that Eve continues to believe that she ought, morally, to aid her uncle, we can ask whether that belief must constitute a disposition of this sort. Now, for the pur- poses of defending T3, there is no need to deny that some cognitivist moral ought-beliefs are themselves (nonessentially) dispositions to engage in appropriate intentional conduct and that they need no supplementation from distinct action-desires. Nor even need it be denied that all actual first-person cognitivist moral ought-beliefs are (nonessentially) dispositions of this kind.32 My claim about listlessness-cast in terms of dispositions-is that there is at least a hypothetical species of this malady to which human beings are not essentially immune, such that a human being suffering from it may believe that she ought, morally, to A without, in so believing, being at all disposed to engage in corresponding intentional action. Again, given a common understanding of cognitivist belief, this claim is utterly plausible. Its opponents need to explain why we should adopt a reformative conception of belief according to which some cognitivist beliefs are essentially motivation-constituting, or essentially constitute appropriate action-dispositions.

I know of two relevant species of argument-arguments from moral experience and arguments from the practical nature of mo-

321t is worth noting that some internalist moral realists view first-person moral ought-beliefs as being, at once, cognitivist beliefs that one ought, morally, to A and desires to A. (This, obviously, is consistent with T3.) See McNaughton 1988,106-10. Cf. McDowell 1978. McDowell attributes (1978, 15-17) a view of this kind to Nagel (in Nagel 1970). For an alternative reading of Nagel, see Dancy 1993, 8-9.

405



ALFRED RI? MELE

rality. David McNaughton has advanced a representative argument of the first sort (1988, 47-50; cf. 1988, 23 and McDowell 1978). He contends (1) that part of what it is to have the (veridical) experience of being morally required to A is to apprehend a reason to A. On the assumption (2) that all reasons for action are motivation-constituting, he concludes (3) that having motivation to A is internal to the experience of being morally required to A.

Assume, for the sake of argument, that (1) is true. Then it may be supposed that in having the experience of being morally required to assist her uncle, Eve apprehends a reason to do so. The supposed reason may simply be that she is morally required to do this; or it may be constituted by considerations on the basis of which she apprehends this moral requirement. But part of what is at issue, on the assumption that (1) is true, is precisely whether Eve's apprehending the pertinent reason suffices for her having motivation to continue aiding her uncle. I have argued, in effect, that at least from the perspective of a traditional view of the attitudes, Eve's depression may be such that although she apprehends a moral reason (in some sense of 'reason') to continue helping him, she is utterly unmotivated by that reason (and her apprehen- sion of it): more precisely, she has no motivation to help him. What McNaughton needs in this connection is an argument that apprehending a reason for A-ing-in precisely the sense of 'apprehending a reason' in which having an experience of being morally required to A is supposed to include apprehending a reason for A-ing- suffices for having motivation to A. He produces no such argument. Even if some analyses of reasons for action treat motivation- constituting items as essential constituents of such reasons, this does not provide grounds for confidence that readings of 'reason' on which (1) is true are readings on which (2) is true as well. Eve believes on some grounds or other that she ought to help her uncle; simply calling those grounds "reasons" to help him cannot render them motivation-constituting.

Now, beliefs that one ought, morally, to A are practical beliefs- beliefs about what to do. As R. M. Hare has put it, "moral judgments, in their central use, have it as their function to guide conduct" (1963, 70). It may be alleged that to reject the idea that there are EMC cognitivist moral ought-beliefs is to drive a wedge between moral judgments and intentional conduct, thus taking the

406



MOTIVATION

practical bite out of morality. I call this "the argument from toothlessness."33

The chief problem with this argument is that the practical bite of a species of attitude may be underpinned in ways that do not require attitudes of that type to be essentially motivation-constituting. Recall that T3 is consistent with the anti-"Humean" thesis that first-person cognitivist moral ought-beliefs normally are motivation- constituting and even with the thesis that all actual beliefs of this kind are, in fact, motivation-constituting. (Our not being essentially immune to listlessness of the sort featured in Eve's case is compatible with our never actually suffering from it, of course.) Thus, T3 is consistent with a broad range of ways of underpinning the practical bite of moral ought-beliefs. Some ways may involve the anti-"Humean" idea that some cognitivist beliefs are (nonessentially) motivation-constituting. But the underpinning may be done even from a "Humean" perspective. It may be argued, for example, that moral agents typically are possessed of a standing generic desire to do whatever they morally ought. This desire, so as not to put too fine a point on it, is a generic motivational concern to act in accordance with one's judgments about what one morally ought to do. Even when the concern is not an occurrent one, it may be activated by a judgment that one ought, morally, to A, thus giving rise to occurrent motivation to A. In agents with such a desire, judgments that they themselves ought, morally, to A would naturally issue in occurrent desires to A, which desires may play a familiar role in motivating conduct.

On a traditional view of the attitudes, Eve's story is utterly coherent. This is not to say that a theorist cannot propose a reformative conception of the attitudes according to which Eve's possessing a cognitivist belief that she ought, morally, to aid her uncle is incompatible with her suffering from listlessness of the kind de- scribed. In attempting to understand and explain intentional behavior, we standardly advert to a host of attitudes, including desires and cognitivist beliefs. Traditional conceptions of this battery of attitudes undoubtedly fall short of perfection; but if there is no defect in these conceptions that would be remedied by altering the battery to include the pertinent reformative notion of cognitivist

330f course, Hare himself is a noncognitivist; but the argument from toothlessness is available to noncognitivists and cognitivists alike.

407



ALFRED R MELE

belief, then there is little to recommend the revision. I see no theoretical need, in moral philosophy or the philosophy of mind, for a conception of cognitivist moral ought-belief that would render Eve's story incoherent. In the following section, in accordance with T3 (the thesis that, in us, any EMC attitude is, or encompasses, an action-desire), I defend a precise account of EMC attitudes.

First, an objection requires attention.34 Someone might contend that even if cognitivist moral ought-beliefs are not EMC attitudes, action-desires are on no firmer ground in this respect. It may be claimed, for example, that a jogger might desire to run another lap but be too tired even to try, so that this desire is not motivation- constituting. Again, an agent who takes his A-ing to be physically impossible might hope or wish that he could A, but he does not desire to A. However, someone might believe that he can A and desire to A although, in fact, even an attempt to A is beyond his power. We can desire to do what, unbeknown to us, we are unable to do, and unable even to try to do. (To take an extreme case, although S desires to A, an irresistible demon will not permit some prerequisite of Ss trying to A: for example, Ss acquiring some suitable intention.) But this possibility, rather than entailing the possibility of action-desires that are not motivation-constituting, illustrates the possibility of causally impotent motivation-constituting attitudes-or, more precisely, of motivation-constituting attitudes that cannot motivate, in the success sense, even a corresponding attempt (in the circumstances). This last possibility is markedly different from the one highlighted in my discussion of listlessness. If my little story about Eve is coherent, it is not as though she is possessed of a motivation-constituting belief on which, owing to her listlessness, she is unable to act. Rather, her listlessness consists partly in the absence of motivation to help her uncle (or, alternatively, deprives her of all such motivation), even though she believes that she ought, morally, to help him.

I should emphasize that I am not making the empirical claim that in actual human beings, depression sometimes "works" in the way imagined in Eve's case. Perhaps all actual depressed human beings who believe that they ought, morally, to A do have some motivation to A. Perhaps, when they do not even try to A, that motivation is overridden by stronger motivation that is explained

340ne of the editors articulated essentially this objection.

408



MOTIVATION

by-or partly constitutes-their depression (e.g., a desire not to exert oneself); or their depression might block the formation or acquisition of intentions that accord with this moral motivation. But even if depression does in fact work in some such way in actual human beings, that plainly is compatible with my claim that actual human beings are not essentially immune to depression or listlessness of the kind imagined in Eve's case-hence, with my claim that moral ought-beliefs are not essentially motivation-constituting in actual human beings. If the actual mechanics of depression in actual human beings leaves no room for cases like Eve's, part of what I have argued, in effect, is that this is a contingent fact-more specifically, this supposed truth about the mechanics of depression is not grounded in any essential motivational properties of cognitivist moral ought-beliefs, nor in any essential psychological features of human beings.35 Should depression work in an acquaintance of mine in the way I have imagined it works in Eve's case, that is compatible with my acquaintance's being a human being; and, I have urged, it is compatible with anything that we have good reason to believe about moral ought-beliefs.

5. Essentially Motivation-Constituting Attitudes Functionally Characterized

Intentional attitudes may be decomposed into a content (e.g., "The Giants will win today") and a mode (e.g., belief, hope, fear) (Searle 1983, chap. 1). At a familiar level of description, action- desires have three internal resources for contributing to behavior:

35I do not claim to have produced a knock-down argument for this. I have not demonstrated that no new argument can establish, for example, the thesis (MA) that the nature of moral agency itself conceptually requires that moral ought-beliefs be essentially motivation-constituting cognitivist beliefs and, therefore, that cases like Eve's are conceptually impossible. I believe that MA, in placing the psychological demands it does on moral agents, would place our moral agency in serious jeopardy. For all we now know, we are in fact psychologically constituted like Eve and thus are ca- pable of possessing cognitivist moral ought-beliefs that are not motivation- constituting at all, much less essentially so. For all we now know about human psychology, our actual psychological constitution is such that, were MA true, we would be devoid of moral ought-beliefs and would not be moral agents. The question whether MA and its ilk license skepticism about our moral agency requires a paper of its own; I cannot do it justice here.

409



ALFRED I?. MELE

their mode, their content, and their strength. The mode of an action- desire (i.e., desire), together with its content, places it in the motivational arena.36 The content provides the agent with a target or goal in that arena. An action-desire's strength is, very roughly, the force it exerts upon the agent to try to achieve the goal. (I have examined this last resource at length elsewhere, arguing, among other things, that agents often are not helpless victims of motivational forces-that the strengths of many desires are subject to the agent's control. Unless a desire is irresistibly strong for an agent, it is in some sense up to her whether she acts on it.37 Space does not permit an examination of motivational strength here.)38

Earlier, I noted that "S desires to A" is more precise than "S has motivation to A." Can similar precision be secured in terms of "motivation-constituting attitudes"? Let us say that an "A-repre- senting" motivation-constituting attitude is any motivation-constituting attitude having the agent's A-ing as a represented object, where 'A' (again) is an action variable. Represented objects include both "positive" actions and refrainings. The represented object of Al's desire to attend today's Giants game is his attending that game; the represented object of Ann's desire to refrain from attending the game is her so refraining. "Negative" motivation-constituting states-for instance, Beth's fear of flying-may be construed as being at least partly composed of motivation-constituting attitudes having refrainings as represented objects. Beth's fear is constituted, in part, by a motivation-constituting attitude whose represented object is her refraining from flying.

36The relevance of content is illustrated by the observation that although some desires-that is, some attitudes having the mode desire-are motivationally inert (as in the spaceship example), all action-desires are motivation-constituting.

37On motivational control, see Mele 1987, chaps. 2-6; Mele 1992, chap. 4; and Mele 1995, chap. 3. For an analysis of irresistible desire, see Mele 1992, chap. 5; a principle linking motivational strength to intentional action is defended in Mele 1992, chap. 3.

38Another issue that I must set aside is the nature of "motivational" attitudes that are contingently motivation'-constituting. These include all desires that are neither motivationally inert nor action-desires. A plausible hypothesis is that a central function of motivational attitudes of this sort is to give rise to appropriate EMC attitudes; but a proper exploration of the hypothesis would require extensive discussion. Doing justice to the question whether there are, or can be, contingently motivation-constituting cognitivist beliefs would require a paper devoted to that topic.

410



MOTIVATION

All EMC attitudes, as I have argued, have the agent's acting in some way as a represented object (T2). Thus, we can say that every EMC attitude is an EMCA attitude, where 'EMC' designates the species of attitude and 'A' is a place-holder for the pertinent represented object. Ti, T3, and the observation that any attitude encompassing a desire to A is itself an EMCA attitude (in virtue of encompassing an EMCA attitude) jointly imply the following:

T4. In human beings as they actually are, an attitude x is an EMC attitude if and only if x is, or encompasses, an action- desire, and x is an EMCA attitude if and only if x is, or encompasses, a desire to A ('A' being an action variable).

Although the word 'desire' is of no special interest to me in this connection, a certain functional role associated with action-desires is a pressing concern. I want now to suggest the following as an approximation of a thesis to be endorsed here:

T5. In human beings as they actually are, an attitude x is an EMCA attitude if and only if (1) the agent's A-ing is a represented object of x ('A' being an action variable) and (2) an essential function of x is to contribute to its own satisfaction by inducing the agent to try to bring it about that she A-s.

This thesis accords a certain "direction of fit" to EMCA attitudes (in actual human agents). They have the function of getting things to fit them (by inducing the agent to make a suitable attempt).

Four notions at work in T5 require brief comment. First, I understand contributing, in this context, as a causal notion; I have not abandoned perspective P Second, satisfied attitudes should not be confused with satisfied agents. An EMCA attitude is satisfied by the agent's A-ing; but the agent herself might find her A-ing quite un- satisfying. Third, an attitude x essentially has a function Fin an agent S only if x has that function not only in Ss actual situation but also in all possible scenarios in which S has x. An attitude x nonessentially has a function F in S if, for example, x's having that function in S is contingent upon Ss having an attitude of type y and Ss having an attitude of the latter sort is neither entailed nor presupposed by Ss having x. Fourth, trying, as I understand it, requires no special effort. When I intentionally type the word 'type', I am

411



ALFRED I MELE

trying to do that even if I encounter no special resistance.39 Fur- ther, trying to A is one way of trying to bring it about that one A- s: in trying to A one is trying to bring it about that one A-s. But there are other ways. Someone who wants to exercise daily but fears that he will succumb to laziness may try to bring it about that he exercises daily by telling his friends and loved ones that he has resolved to do so. This person tries to bring it about that he exercises daily by increasing his motivation to do so. As he sees it, once he makes the announcement, failing to exercise daily would carry an extra cost-his friends and loved ones would think less of him for backing out. In making the announcement, he is not trying to exercise daily; but in making the announcement, he is trying to bring it about that he exercises daily. The notion of trying to bring it about that one A-s at work in T5 includes, but is not limited to, trying to A.40

Is T5 true? This question amounts, of course, to the question whether the following two claims are true:

T5a. In human beings as they actually are, if an attitude x is an EMCA attitude, then x has the properties identified in clauses (1) and (2) of T5.

T5b. In human beings as they actually are, if an attitude x has the properties identified in clauses (1) and (2) of T5, then x is an EMCA attitude.

Start with T5a. T4 implies that in real human beings, an attitude is an EMCA attitude only if it is, or encompasses, a desire to A. I take it to be a conceptual truth that any action-desire has the function of contributing to its own satisfaction. If this is right, all action- desires essentially have this function. Further, if all action-desires have this function essentially, and T4 is true, then in real human beings, all EMCA attitudes essentially have the function of contributing to their own satisfaction. Now, our ordinary concept of desire accords action-desires a more specific function than this. A professor's desire to gain the attention of her fidgeting students may so distract her that she unintentionally places a piece of chalk between her lips while scrawling on the blackboard with her cigarette. Even though these antics gain the full attention of her students,

39This conception of trying is defended in Adams and Mele 1992. 40The discussion a bit later of Ann and Bob also illustrates this point.

412



MOTIVATION

the desire has functioned defiantly in the production of its satisfaction. Our seeing this as an instance of deviant causation indicates that we accord action-desires a more precise function than merely contributing to their own satisfaction. We view desires to A as having, more fully, an essential function of contributing to their own satisfaction by way of their contribution to intentional conduct of the agent that is directed toward A-ing. Trying is essentially intentional; and any instance of trying to bring it about that one A-s is directed toward A-ing. Hence, a plausible way to capture this fuller function is in terms of trying, as in T5.41

Cases in which an agent wants to A unintentionally may seem to undermine T5a. Ann is promised $10,000 for offending Bob unintentionally; and she knows that there is no reward for intentionally offending him. Wanting the money, Ann desires to offend Bob unintentionally. A straightforward, successful attempt to offend him will not gain her the money, as she realizes; for such an attempt would amount to intentionally offending Bob. Does this imply that Ann has a desire to A-an action-desire-that lacks the function of contributing to its own satisfaction by inducing her to try to bring it about that she A-s?

Two possibilities need attention. First, Ann's desire to offend Bob unintentionally may not be an action-desire at all (i.e., as I put it earlier, a "desire having one's acting in some way ... as a represented object"). Ann realizes that people sometimes take offense at things she says or does even when she does not at all desire to offend them. (Sometimes she learns only months later that they have been offended by her words or deeds.) Ann might hope- and, again, hope encompasses desire-merely that this sort of thing will happen in Bob's case; the represented object of her desire may be a mere happening rather than a prospective action. Ann's attitude toward her unintentionally offending Bob may be on all fours with certain of her non-action desires: for example, her desire to live to be a hundred years old and her desire to sleep peace- fully tonight. T5a is a thesis about essentially motivation-constituting

41A theorist might try the alternative tack of maintaining that all action- desires-all desires to A-are desires to A intentionally. If this were true, the professor's desire to gain her students' attention would not have been satisfied (since she did not gain it intentionally). However, since some agents possessed of action-desires lack the conceptual sophistication required for desiring to A intentionally, this tack fails.

413



ALFRED RI? MELE

attitudes having (prospective) actions of the agent as represented objects. If Ann's desire to offend Bob unintentionally is not an action-desire, it fails as a counterexample to T5a on two counts. The desire lacks an object of the right sort; and given T4, it is not essentially motivation-constituting.

Second, Ann's desire to offend Bob unintentionally may be a genuine action-desire. But then, in virtue of possessing this desire, Ann will be inclined, in some measure, to bring it about that she offends Bob unintentionally. In one relevant scenario, she knows that she tends to offend Bob unintentionally when she is extremely busy: when she is preoccupied with her work, for example, she tends, without then realizing it, to speak more tersely than she ordinarily does to people who phone her at the office; and, when Bob calls, her terse speech tends to offend him. Knowing this, Ann may undertake an engrossing project-say, writing a paper on motivation-with the hope that her involvement in it will render her telephone conversation at the office sufficiently terse that, should Bob call (as he frequently does), she will unintentionally offend him. In pursuing this strategy, Ann is trying to bring it about that she offends Bob unintentionally. This is a coherent attempt; in particular, it is not intrinsically self-defeating. An action-desire of Ann's to offend Bob unintentionally may have the function of contributing to its own satisfaction by inducing her to try to bring it about that she offends him unintentionally. Further, if a desire of hers to offend Bob unintentionally were not to have this function, it would plausibly be regarded as a non-action desire, a desire whose represented object is something's merely happening rather than her performing an action of a particular type.

I turn to T5b. The thesis has intuitive grip. An attitude having the agent's A-ing as a represented object and having the function of contributing to its own satisfaction by inducing the agent to try to bring it about that she A-s certainly seems to be a prime can- didate for a motivation-constituting attitude. All tryings are motivated; and, at least at first glance, attitudes having this functional relationship to trying apparently constitute motivation to try. Fur- ther, an attitude having the specified function essentially seems well suited to be essentially motivation-constituting. However, appear- ances sometimes mislead; potential objections require attention. Are there attitudes that have the essential function of contributing to their own satisfaction-and, more specifically, of contributing to

414



MOTIVATION

this by inducing the agent to try to bring it about that she A-s- but, nevertheless, are not essentially motivation-constituting?

Assuming that such attitudes as hoping to A and wishing to A encompass desires to A, I have already argued, in effect, that they are essentially motivation-constituting. If counterexamples are to be found to T5b, a natural place to look is the "cognitive" sphere, the sphere of beliefs. Some action-regarding cognitivist beliefs may help bring about their own truth-hence, their own satisfaction. They merit consideration.

Bob is a place-kicker. His kicking coach has informed him that, regarding any field goal he attempts, his believing that he will make the field goal will increase his chances of making it. Bob, who has heard of the power of positive thinking, decides to augment his preparation for each kick with an attempt to get himself to believe that the kick will succeed. Eventually, he regularly gets himself to believe this, and his confidence in the success of his kicks signifi- cantly boosts his success rate. Bob's beliefs that he will make particular field goals are satisfied by his making them; and his believing that he will make them causally contributes to his making them. Does Bob's belief that he will make his next field goal have the function specified in T5b? If so, T5b is false; for this belief is not motivation-constituting.

Plainly, the pertinent belief does not have the function of contributing to its own satisfaction by inducing Bob to try to bring it about that he makes the field goal. Bob is already settled on trying to make the field goal prior to acquiring the belief that he will make it. His having the belief increases his chances of making a successful kick, given that he is settled upon trying to do so. But the belief is functionally irrelevant to whether Bob tries to make the kick.

Consider a case designed to avoid this problem. An exceptionally risk-averse person, Rex, is particularly averse to making unsuccessful attempts. He will not attempt anything unless he believes that, should he make an attempt, the attempt would succeed. Rex is confident about his chances of A-ing, something that he would like very much to do. He believes that he will A if he tries to A. Be- lieving this, he believes, as well, that he will A; for possessing the conditional belief, and realizing that he would like very much to A, Rex believes that he will try to A. If Rex were to lack the belief that he will A, his lacking it would be explained by his lacking the

415



ALFRED R MELE

conditional belief-in which case, he would not even attempt to A and would not A. Rex's making an attempt thus is contingent upon his believing that he will A: were he to lack the belief, he would not try to A. Does this belief have the function of contributing to its own satisfaction by inducing Rex to try to bring it about that he A-s?

Given Rex's peculiar form of risk aversiveness, his trying to A is contingent upon his believing that he will A. But Rex's belief that he will A does not induce him to try to A. It is true that if he were to lack this belief, he would not try to A and would not A. But although the attitudes that contribute causally to his trying to A include attitudes that partly constitute the basis for this belief, they do not include this belief itself. Rex's belief that he will A is epi- phenomenal with respect to his trying to A.42

Consider a third case. Sally, a Stoic, is disposed to want to perform any action that she believes she will perform. She believes that happiness is found (in part) in discovering what fate has in store for one and in desiring to act as one believes oneself to be fated to act; and she desires to be happy. Sally now believes that

42Rex's cousin, Dex, is psychologically so constituted that he is not the least bit inclined to try to A unless he is convinced that he would A should he try. Might an agent like Dex undermine my earlier claim that any desire to A ('A' being an action variable) inclines the agent, in some measure, to A intentionally, or to try to A, or to try to put himself in a position to A? Suppose that Dex would like to A but that he does not believe that, should he try to A, the attempt would succeed. Then, given the kind of person Dex is, he is not the least bit inclined to try to A (nor, we may suppose, is he at all inclined to try to put himself in a position to A). If, nevertheless, he desires to A, my claim is false. Notice, however, that from the fact that someone would like to A it does not follow that he desires to A. I would like to sink the winning basket in an NBA championship game and to hit a home run in a World Series game: doing these things would be exciting, and I sometimes daydream about doing them. But I fully recognize that at my age and with my athletic talents, these heroic deeds will never be open to me. It would be an abuse of language to say that I desire or want to do them. Similarly, whereas Dex would desire to A if he believed that he would A should he try to A, in the case as it stands he lacks the desire to A although he would like to A. Dex and I differ in that I occasionally am inclined to try to do things I believe I will not succeed in doing (e.g., sink a basket from mid-court or throw a dart into the bull's-eye), and he never is. When he lacks the belief that he would A should he try to, his attitude toward his A-ing is comparable to my attitude toward my hitting a home run in a World Series game; he no more desires to do the former than I desire to do the latter, although each of us would like to do these things.

416



MOTIVATION

she will A. As a consequence, she desires to A, which desire issues in a successful attempt to A. Does her belief that she will A have the function identified in T5b?

Even if this belief has, in some sense, the function of contributing to its own satisfaction by inducing Sally to try to A, it does not essentially have this function. The belief has this function (if it does) because Sally is disposed to want to perform actions that she believes she will perform. In the absence of that disposition, the belief would lack the relevant function. However, the presence of the disposition in Sally is neither entailed nor presupposed by her believing that she will A. Hence, that belief does not essentially have the function at issue.

Are there nondoxastic attitudes that have the essential function identified in T5b without being essentially motivation-constituting? One might suppose that such attitudes can be found somewhere in the animal world, if not in us. Consider the following:

The wasp Sphex builds a burrow ... and seeks out a cricket which she stings in such a way as to paralyze but not kill it. She drags the cricket into the burrow, lays her eggs alongside, closes the burrow, then flies away.... In due course, the eggs hatch and the wasp grubs feed off the paralyzed cricket.... To the human mind, such an elaborately organized and seemingly purposeful routine conveys a convincing fla- vor of logic and thoughtfulness-until more details are examined. For example, the Wasp's routine is to bring the paralyzed cricket to the burrow, leave it on the threshold, go inside to see that all is well, emerge, and then drag the cricket in. If the cricket is moved a few inches away while the wasp is inside making her preliminary inspec- tion, the wasp, on emerging from the burrow, will bring the cricket back to the threshold, but not inside, and will then repeat the pre- paratory procedure of entering the burrow to see that everything is all right. If again the cricket is removed a few inches while the wasp is inside, once again she will move the cricket up to the threshold and re-enter the burrow for a final check. The wasp never thinks of pulling the cricket straight in. On one occasion, this procedure was repeated forty times, always with the same result. (Wooldridge 1963, 82-83)

It might be suggested that the wasp possesses an attitude that has the insect's "checking the burrow" as a represented object and has the essential function of contributing to its own satisfaction by inducing the wasp to try to check the burrow. It might be claimed, further, that this attitude is not motivation-constituting: the wasp's

417



ALFRED I?. MELE

"checking" behavior has all the marks of an unmotivated execu- tion of a fixed subroutine.

Grant, for the sake of argument (and as is plausible), that the wasp has no motivation-constituting attitude with the relevant object. We have a counterexample to a more inclusive version of T5b only if the wasp possesses an attitude having the function of contributing to its own satisfaction by inducing the wasp to try to check its burrow. But precisely those considerations that lead us to believe that the wasp has no motivation-constituting attitude toward checking its burrow should lead us to believe that the wasp is not trying to check its burrow. Its behavior does not include burrow checking at all. Rather, what might look initially like behavior aimed at checking to "see that all is well" is simply the mindless unfolding of a fixed routine. The wasp, in entering its burrow, is no more trying to check it for safety, or for its (continued) suitability to egg laying, than a flower, in turning toward the light, is trying to gain energy.

There is a general moral here. Trying to A is, essentially, intentional behavior; and, it is generally agreed, intentional behavior is found only among motivated beings. Beings lacking motivation will not try to do anything, and beings lacking motivation to A (and motivation to try to A) will not try to A.43

Although T5 survives the objections I have considered, a modi- fication is required-fortunately, one that leaves T5's spirit intact. Sometimes agents desire to try to do something; they are possessed of an attitude having the mode desire and content of the form "I try to A." Ann believes that it will be difficult to defeat her tennis opponent, and she desires to try hard to win the match. She realizes that sometimes, although she wants to win a tennis match, she does not exert herself fully to win. This time, she wants to make an unrelenting effort to achieve that goal. T4, T5, and the assumption that Ann's desire to try to win the match is an action- desire jointly imply that Ann possesses an attitude having the function of contributing to its own satisfaction by inducing Ann to try to bring it about that she tries to win the match. A desire to try to win the match might induce Ann to do this. Wanting to try to win the match, she might, for example, try to talk herself into trying to

430n the possibility of scenarios in which an agent who has no motivation to A nevertheless has motivation to try to A, see Mele 1990.

418



MOTIVATION

win it. However, Ann's desire to try to win the match might instead have a more direct effect on her: wanting to try to win, she might try to win, without trying to bring it about that she tries to win.

In some cases, we try to A without making any attempt to get ourselves to try to A. This is as it must be. If, concerning everything that we try to do, we first had to try to bring it about that we try to do it, there would be an impossible regress of tryings. There are, as I will put it, some "basic tryings" to A-tryings to A that do not proceed from tryings to bring it about that we try to A. Now, if every desire of ours to try to do something were to have the function of contributing to its own satisfaction by inducing us to try to bring it about that we try to do that thing, then no desire to try to do something would be functionally geared to the production of a basic trying. That would be an unfortunate result. But given T4, and assuming that any desire to try to do something is an action-desire, T5 yields that result.

T5 was framed partly in terms of trying in order to capture a precise function of EMC attitudes. As I put it, we see desires to A as having a more precise function than merely contributing to their own satisfaction; we view them as having the function of doing this by way of their contribution to intentional conduct directed toward A- ing. And, as I noted, trying to A and trying to bring it about that one A-s are precisely such conduct-intentional conduct directed toward A-ing. Any desire to try to A (or, more broadly, to try to bring it about that one A-s), in having, essentially, the function (FE) of contributing to its own satisfaction, has, essentially, the function (F2) of inducing the agent to perform an intentional action directed toward A-ing, where A-ing is such that its occurrence is not ensured by the desire's satisfaction; for trying to A is both essentially intentional and essentially directed toward A-ing (as is trying to bring it about that one A-s). This is an important difference between action-desires that are, and action-desires that are not, desires to try to do something. In the former case, essentially having F1 suffices for essentially having F2; in the latter case it does not. (In the latter case, a desire to A has the function of inducing the agent to perform an intentional action directed toward A-ing, but the desire's satisfaction obviously ensures the occurrence of an A-ing.)

These points having been made, T5 may be modified as follows:

419



ALFRED R MELE

T5*. In human beings as they actually are, an attitude x is an essentially motivation-constituting attitude if and only if (1), there is some action-type A such that the agent's A-ing is a represented object of x and an essential function of x is to contribute to its own satisfaction by inducing the agent to try to bring it about that she A-s, or (2) there is some action-type A such that the agent's trying to A is a represented object of x and an essential function of x is to contribute to its own satisfaction by inducing the agent to try to A.

As I observed earlier, trying to A is a species of trying to bring it about that one A-s. In a sense, attitudes characterized, respectively, by conditions 1 and 2 in T5* have the same essential function- the function of contributing to their own satisfaction by inducing the agent to try to bring it about that she A-s. A salient difference is that an EMC attitude whose represented object is the agent's trying to A-but not an EMC attitude having the agent's A-ing as its represented object-is satisfied by such a trying, whether or not the trying issues in an A-ing. The chief point to be made is that T5* has T5's virtues while avoiding its insensitivity to basic trying. T5* does not preclude there being EMC attitudes geared to the production of basic tryings.

Operating within a perspective on intentional action that accords mental states an important causal/explanatory bearing on intentional behavior and that views all motivation-constituting items as representational states of mind, I have motivated T5* and under- mined what appear to be the weightiest lines of objection to it. If T5* is true, we now have an account of an important species of motivation in real human beings: essentially motivation-constituting attitudes. Such attitudes have, essentially, a functional connection to intentional conduct, a connection of an eminently motivational kind; and they differ functionally from "truth-seeking" attitudes. Essentially motivation-constituting attitudes are paradigmatic motivational attitudes. On the assumption that attitudes are the motivational units, essentially motivation-constituting attitudes are paradigms of motivation. Although a proper understanding of this species of human motivation does not yield a complete theory of motivation, a successful account of paradigmatic motivational atti-

420



MOTIVATION

tudes is a promising point of departure for a fuller account of motivation and its place in the lives of intelligent agents.44

Davidson College

References

Adams, Frederick, and Alfred Mele. 1992. "The Intention/Volition De- bate." Canadian Journal of Philosophy 22:323-38.

Alston, William. 1967. "Motives and Motivation." In The Encyclopedia of Philosophy, ed. Paul Edwards. New York: Macmillan.

Aristotle. Nicomachean Ethics. In Works of Aristotle, vol. 9, ed. W. D. Ross. London: Oxford University Press, 1915.

Atkinson, John. 1957. "Motivational Determinants of Risk-Taking Behav- ior." Psychological Review 64:359-72.

Audi, Robert. 1973. "The Concept of Wanting." Philosophical Studies 24:1- 21.

Brand, Myles. 1984. Intending and Acting. Cambridge: MIT Press. Bratman, Michael. 1985. "Davidson's Theory of Intention." In Actions and

Events, ed. Ernest LePore and Brian McLaughlin. Oxford: Basil Black- well.

.1987. Intention, Plans, and Practical Reason. Cambridge: Harvard University Press.

Brink, David. 1989. Moral Realism and the Foundations of Ethics. Cambridge: Cambridge University Press.

Castafieda, Hector. 1975. Thinking and Doing. Dordrecht: Reidel. Dancy, Jonathan. 1993. Moral Reasons. Oxford: Blackwell. Davidson, Donald. 1963. "Actions, Reasons, and Causes." Journal of Phi-

losophy 60:685-700. 1980. Essays on Actions and Events. Oxford: Oxford University Press,

Clarendon Press. . 1985. "Replies to Essays I-IX." In Essays on Davidson, ed. Bruce

Vermazen and Merrill Hintikka. Oxford: Oxford University Press, Clar- endon Press.

Deci, Edward, and Richard Ryan. 1985. Intrinsic Motivation and Self-Deter- mination in Human Behavior. New York: Plenum Press.

Ginet, Carl. 1990. On Action. Cambridge: Cambridge University Press.

44 am grateful to Paisley Livingston and Piers Rawling for motivating me to write this paper and to George Ainslie, David McNaughton, Paul Moser, Natika Newton, Steve Sverdlik, Bob Ware, and the editors of the Philosophical Review for comments on earlier versions.

421



ALFRED g MELE

Goldman, Alvin. 1970. A Theory of Human Action. Englewood Cliffs, N.J.: Prentice-Hall.

Hare, R. M. 1963. Freedom and Reason. Oxford: Oxford University Press. . 1972. Practical Inferences. Berkeley: University of California Press.

Harman, Gilbert. 1986. Change in View. Cambridge: MIT Press. Heckhausen, Heinz. 1991. Motivation and Action. Berlin: Springer-Verlag. Heil, John, and Alfred Mele, eds. 1993. Mental Causation. Oxford: Oxford

University Press, Clarendon Press. Humberstone, I. L. 1992. "Direction of Fit." Mind 101:59-83. Jackson, Frank. 1985. "Internal Conflicts in Desires and Morals." American

Philosophical Quarterly 22:105-14. Kenny, Anthony. 1989. The Metaphysics of Mind. Oxford: Oxford University

Press. Lewis, David. 1988. "Desire as Belief." Mind 97:323-32. McDowell, John. 1978. "Are Moral Requirements Hypothetical Impera-

tives?" The Aristotelian Society, supp. vol. 52:13-29. McNaughton, David. 1988. Moral Vision. Oxford: Basil Blackwell. Melden, A. I. 1961. Free Action. London: Routledge and Kegan Paul. Mele, Alfred. 1987. Irrationality. New York: Oxford University Press.

1990. "He Wants to Try." Analysis 50:251-53. 1992. Springs of Action. New York: Oxford University Press. 1995. Autonomous Agents. New York: Oxford University Press.

Milo, Ronald. 1984. Immorality. Princeton: Princeton University Press. Mook, Douglas. 1987. Motivation: The Organization of Action. New York: Nor-

ton. Nagel, Thomas. 1970. The Possibility of Altruism. Oxford: Oxford University

Press, Clarendon Press. Peters, R. S. 1958. The Concept of Motivation. London: Routledge and Kegan

Paul. Pettit, Philip, and Michael Smith. 1990. "Backgrounding Desire." Philo-

sophical Review 99:565-92. Searle, John. 1983. Intentionality. Cambridge: Cambridge University Press. Skinner, B. F. 1953. Science and Human Behavior New York: MacMillan. Smith, Michael. 1987. "The Humean Theory of Motivation." Mind 96:

565-92. Stocker, Michael. 1979. "Desiring the Bad." Journal of Philosophy 76:738-53. Toates, Frederick. 1986. Motivational Systems. Cambridge: Cambridge Uni-

versity Press. Vollmer, Fred. 1993. "Intentional Action and Unconscious Reasons." Jour-

nal for the Theory of Social Behavior 23:315-26. Williams, Bernard. 1973. Problems of the Self Cambridge: Cambridge Uni-

versity Press.

422



MOTIVATION

Wilson, George. 1989. The Intentionality of Human Action. Stanford: Stan- ford University Press.

Wooldridge, Dean. 1963. The Machinery of the Brain. New York: McGraw Hill.

423



Documents

Motivation: Essentially Motivation-Constituting Attitudes