Introduction: The Debate - Oxford Scholarship Jump to ContentJump to Main Navigation
Reasons from WithinDesires and Values$

Alan H. Goldman

Print publication date: 2009

Print ISBN-13: 9780199576906

Published to Oxford Scholarship Online: February 2010

DOI: 10.1093/acprof:oso/9780199576906.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE ( (c) Copyright Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see Subscriber: null; date: 25 September 2016

Introduction: The Debate

Introduction: The Debate

(p.1) 1 Introduction: The Debate
Reasons from Within

Alan H. Goldman (Contributor Webpage)

Oxford University Press

Abstract and Keywords

This chapter describes the nature of deliberation, and defines the debate between internalists and externalists about reasons. Internalists hold that all reasons derive from existing motivations. Externalists believe that objective values provide reasons for acting. Internalism will be defended on the grounds that it is simpler, more intelligible, and more unified in its accounts of both theoretical and practical reasons. Topics to be addressed in later chapters are outlined here.

Keywords:   nature of deliberation, internalists, externalists, existing motivations, objective values

I. Deliberation

“It was a once‐in‐a‐lifetime opportunity, and I've had a couple of those.”

Days, weeks, months go by in which I engage in no real deliberation about what to do. I do not think I am unusual in this regard. On an average day in which I am not teaching, I read or write for an hour or two, drive to my office, check correspondence, go to lunch with colleagues, work for another hour or two, then return home for some exercise, and so on. I might briefly ask myself what to work on next, whether to play tennis or bike, whether to eat in or out, but noting the alternatives, I simply opt for one or another (with permission from my wife). It is similar if I more rarely pause to ask myself whether I should be doing what I initially planned, whether, for example, the laundry would be done on time for me to keep my lunch appointment. Here I simply make sure that my action does not prevent the satisfaction of some concern more important to me.

In none of these ordinary situations is there deliberation in anything like the way philosophers typically describe: listing and weighing of reasons on each side of a contemplated action, assigning rough numerical values to reflect the weights, summing and reflectively forming the intention to perform the action with the greatest weight of reasons behind it. In my case something remotely like this happens maybe a few times a year, if that much. In all their writing about practical reasoning and weighing reasons in a detached and reflective frame of mind, philosophers have ignored the fact that the process they attempt to describe almost never occurs. But this fact is significant. I will attempt a more accurate description of the process (p.2) below in preparation for a proper interpretation of it in later chapters. But before considering the process as it actually occurs, it is worthwhile to reflect on why it is so rare. Are we irrational in so rarely considering the reasons for our actions? I assume not.

Practical deliberation is the exception and not the rule, but that does not mean that agents ordinarily act without reasons for what they do. In ordinary contexts of action we are rational in doing what we are immediately motivated to do. Simply doing what we feel like doing ordinarily reflects our coherent motivational states at the time. Our inclinations line up with our implicit judgments about what it would be good to do, and action automatically follows.1 Because unreflective action is typically rational, we normally do not need to consciously consider our reasons, stand back from our desires, and deliberately choose to do what is most rational, what we have most reason to do. We need not do this because acting without deliberation nearly always reflects the reasons we have or the values of the actions and their likely outcomes at the time for us. And this is already significant for the main argument of this book.

Either we are somehow (automatically, magically, evolutionarily?) in tune with the objective values our actions can embody or produce and motivated to maximize such value in the world, or our reasons reflect the motivations we have, what we are concerned to do at various times, what we care about, what has value for us at any given time simply because we value doing that at that time. The second explanation sounds less mysterious and initially more plausible. We can trust our actions in general to be rational neither because of some magical connection to objective values that generate our reasons, nor because we must view agents as rational in order to interpret or explain their actions (many people act irrationally at least some of the time, and we have no problem understanding and explaining their actions—we see, for example, that they were overwhelmed by emotion or temptation). We can trust our unreflective decisions because they reflect our motivations and hence the reasons that derive from our motivations. We need not deliberate because our actions automatically flow from our desires or concerns, and (p.3) our reasons too derive from these motivations, reflecting shifting priorities among immediate desires, which in turn derive from and specify our longer standing, deeper, and more general concerns.

States of affairs that present themselves as opportunities or threats, as reasons for or against certain actions, themselves already reflect our motivations, what we care about. If we care about enough, have a variety of major concerns, we will have a couple of once‐in‐a‐lifetime opportunities. But even on those occasions the same situation may present itself as both opportunity and threat, an opportunity to fulfil certain important desires and a threat to the fulfilment of others. This is when deliberation typically occurs. Serious deliberation occurs when one is pulled in opposite directions by conflicting reasons affecting important concerns and relating to the same prospect.

Consider an example of such deliberation in which something important hangs on a decision and conflicting considerations present themselves. Not too long ago I had to decide whether to accept a position at another university or remain where I was and had been for many years. Then I certainly did have to consider carefully the pros and cons, the reasons for and against the move. On the pro side, it seemed that I would be joining a lively and productive department and leaving a meddlesome administrator (little did I anticipate a change for the worse in administration at the new university), would be gaining on average brighter and more serious undergraduate students, would be living in a more pleasant environment with more trees, changing climates, and fewer rude people and traffic jams, would be closer to my sons, and would be near a city with cultural attractions second only to New York. On the con side, I would be leaving close friends and colleagues, would be giving up graduate students, stone crabs, and late night restaurants, would have to find new tennis opponents on my level, would have to sell and buy a house and arrange a move. To these I had to add my wife's concerns (of course counted above mine if she is reading this).

Did I assign rough numerical weights to these reasons and attempt to sum so as to maximize net benefit? No. What I did was try to make sure my list of important considerations was complete (did I need to find out if the new location has good Thai food or level surfaces for rollerblading?) and to picture as vividly as I could what each outcome would be like. I also considered whether I might integrate other actions (p.4) that might compensate for the negatives or add to the positives. Might I move to a nicer neighborhood in my current city instead of moving to a new town? Might I move to a golf as well as tennis community in the new location? (For long‐range planning, integration is crucial—making plans that will allow for the satisfaction of my sports, aesthetic, family, and gustatory interests, as well as my academic interests, whatever the secret priorities among these concerns might be.) Finally, I attempted to gather some further information. How much would the move cost? What was the housing market like in both locations? One point to emphasize about this process is that at no point did I call any of my desires or concerns into question: I took for granted that the pros were pros and cons were cons. I was not standing back and reflecting on my desires (as some philosophers describe deliberation). I did not focus on my desires at all, but on these possible future states of affairs that immediately struck me as positive or negative, reflecting my motivations without foregrounding them. And then I simply made a decision.

Before considering yet a third sort of case, we can note the similarities and differences in the two contexts briefly described. In the first and most common, we simply act without deliberating. Here action, like belief, is more or less automatic. Here we can assume that there are no major conflicts in motivations, that inclination lines up with implicit evaluative judgment about what it would be good to do, leading to action without conscious forming of intention or choice. In the second less common but not altogether rare context there are conflicting reasons that must be brought vividly to mind, informed, prioritized, and integrated in deliberation. The prioritizing is again more or less automatic, however, once the other steps are complete. When the priorities are clearly felt, decision again follows without meditative choice, much as belief follows from full consideration of the evidence for and against a proposition when there is evidence on both sides. And just as evidence for belief is typically clearly for or against, indicative of truth or falsity, so reasons for action are rarely ambiguous, but instead clearly in line with one's concerns or opposed to them.

The only ambiguity in the list of reasons for and against the move was the status of changing college administrators: was it a reason for the move, as it appeared to be at the time, or a reason against, as it turned out? The question arises not because of any lack of clarity in my preferences or in the relations of the possible states of affairs to them, but simply because of a (p.5) lack of relevant information—I had no clue about the impending change in administrators where I might go. It seems that I was subjectively rational to count the new administration as a reason in favor of the move, given the information available to me at the time, but in the next chapter I will argue that in such a case I only appeared to have a reason in favor of the move; in fact, I had a reason against it of which I was not aware.

I have been emphasizing the automaticity of action in the absence of deliberation and of decision after deliberation once the relevant factors are brought to mind, and I have suggested a comparison of the context of action to that of belief. Belief probably would not be as reliable as it is, would not be true as often, if we had to choose what to believe after considering evidence. We are better off for being programmed to automatically believe what strikes us as true. Similarly, action would probably not be as successful, would not as often satisfy our desires and concerns, if we had to constantly stand back and assess our motivations and reasons and choose what to do. It may be more common in the case of action for neither of the opposing decisions to be clearly indicated by the opposing reasons, and then decision itself will determine the priorities for the future. But in both cases it is normally clear what counts for and against belief or action, clear whether reasons once brought to mind are pro or con, and equally clear what to believe or how to act once the evidence or reasons are in.

In the case of action as opposed to belief, some of these pros and cons seem clearly relative to personal preferences. I prefer teaching more motivated and brighter students; better teachers than I might welcome the challenge of teaching less independently motivated students. For them the prospect of “better” students would be a reason against a move; for me it was a reason in favor of one. Good tennis facilities are important to me; they count for nothing for others. Thus many of the considerations or possible states of affairs that count as reasons for me seem to do so only by reflecting my personal concerns.

By contrast, there may seem to be other reasons that are not so personally relative. It might be thought that moral reasons should not have been absent from my list of reasons for and against the move. Should not I have considered the hardship caused by abandoning graduate students whose half finished dissertations I was advising, or the demand for loyalty to an institution that had treated me well for much of my long tenure there? In point of fact, I did arrange not to abandon the dissertations, but loyalty to institution (p.6) counted for nothing for me. Being unconcerned about it, I did not consider it a reason against the move. But did that lack of concern imply that I had no such reason? Many would argue that there is no such implication, that the moral reasons I have are independent of whether I recognize or care about them. I cannot escape my moral obligations by not caring about them, and therefore I have these reasons whether I recognize them or not.

The same point of view might be supported by a third type of case, much rarer still, in which one does question one's desires or concerns, in which they are brought to the fore instead of being automatically reflected in what strikes one as pro and con reasons. Somehow I seemed to have missed my midlife crisis (was it hiding behind the move?), but those who do not miss that landmark ask themselves such questions as, “Do I really want to continue in this career (or this marriage)?” “Is what I spend my life pursuing really worthwhile; is there real value in what I am doing, or am I wasting my life chasing insignificant goals?” Such questions are raised in periods of crisis (or after a particularly bad day in the classroom). When we are reasonably content, our deliberations leave structuring deep concerns in the background, or our actions flow smoothly from more superficial desires. But these periods of discontent do occur, and they indicate that deep concerns and projects too can be called into question. The worries expressed seem to call into question whether our subjective values, even those most fundamental to the course of our lives, measure up to what is truly or objectively valuable, what we ought to be pursuing as opposed to what we are pursuing. As in our attitude toward moral demands, we seem to be presupposing here objective values or reasons independent of our current motivations, reasons that we ought to care about and take into account whether we are doing so or not. We seem to be holding up our subjective values to such objective standards.

But again such questions can be given a different interpretation. Even when I call the ongoing course of my life into question, even when I ask whether what I am doing is really worth my concern, I may not be asking for a comparison between my values as so far expressed in my actions and objective values as what I or anyone else ought to be pursuing, but instead asking whether what I am doing, what I continue to do perhaps out of force of habit, inertia, or lack of energy, is what I really care about most deeply, what will continue to satisfy me in the long run by answering to these genuine concerns. When I ask whether continuing my current life (p.7) plan will really satisfy me, I am not asking whether it would satisfy anyone independently of their dispositions and concerns. I am not interested in whether it would satisfy an investment banker or whether it is valuable from the point of view of the universe as a whole, but whether it is really what I want out of my life. Do I really deeply care about what I am doing any more, or are there other things that I care about more, or things that I would be disposed to care about if I took the time to think about them?

Those who favor the former interpretation would insist that our actions need and can be given a deeper justification. Whatever our deepest concerns, they argue, we can always ask whether satisfying them is objectively worthwhile, whether objective values and reasons ultimately justify our concerns and pursuits. For reasons must not only explain our actions, and must not only justify them in our own eyes, but must justify them tout court. Genuine reasons are normative: they determine what we ought to do if we are rational. What we really ought to do cannot simply be a function of what we want to do. Typically the point of asking what we ought to do is to call into question what we want or feel like doing. I feel like going back to sleep in the morning when the alarm rings, but I know that I ought not to. I have many reasons not to go back to sleep and no reason to do so. Hence, according to these philosophers, if reasons are to be genuinely normative, they must be independent of our motivations and reflect objective values.

Reasons are what we deliberate about on those relatively rare occasions in which we do deliberate. They are what motivate us if we are rational. They indicate how our actions can satisfy our concerns. If my moving to a new job would satisfy my desire for more income, the higher salary is a reason for me to apply. We do not deliberate about our desires, but about such things as living in a better climate versus losing friends and colleagues. Desires are not what motivate us: they are states of being motivated.2 And particular desires in themselves do not seem to create reasons. As suggested in the alarm clock example, just feeling like doing something is not in itself a reason for doing it. If you wake up on a morning when you have a busy schedule before you, and you feel like going back to sleep, that does not give you a reason to do so. The fact that you have appointments is a reason (p.8) for you not to go back to sleep, and it would be a reason for anyone in similar circumstances, a seemingly objective reason independent of what you feel like doing.

But again on the other side, how would the claim that there is a reason to do something motivate me to do it if it did not connect to any of my concerns? If someone reminds me that there is a religious reason not to work on a Saturday or Sunday, that would not strike me as a reason for me, precisely because I do not care about the demands of religions. Reasons must be capable of motivating us. They explain the actions of rational agents as well as justifying them. We are rational agents most of the time, so we must be motivated by reasons. But how could they motivate us if they do not reflect or connect with what we care about?3

From the second person point of view too, when we try to convince another person that she ought not to act in the way she seems to intend, we try to appeal to what we know she cares about. If she is about to act irrationally by taking some unnecessary or pointless risk, we do not argue that aversion to this risk is included in some list of objective rational requirements, but instead point out that the likely outcome will threaten the continued satisfaction of her other concerns, that it will endanger completion of cherished projects, or will likely harm her or those she cares about in some way she finds unacceptable. An agent will be convinced of irrationality only when incoherence with her concerns or deepest motivations is demonstrated to her, not when she is presented with some purported objective requirement that might not appear to her to be a requirement for her.

First person deliberation, as described above, does not typically directly address desires of the agent, but focuses instead on states of affairs that present themselves as good or bad, as opportunities or threats. But their so presenting themselves may be relative to an agent's concerns, including personal and idiosyncratic concerns. When these concerns themselves are called into question, they seem to be held up to a more objective standard. Is what we are doing really worthwhile? But this question too may be addressed to the relation between day‐to‐day routine or superficial desires and deeper concerns. If deepest concerns—for oneself or one's family, or (p.9) for those projects that have structured and given meaning to so many of one's other pursuits—themselves come into question, this again seems to call for an objective validation in terms of reasons or values independent of these concerns. But it may reflect instead a despair of there being any such ultimate reasons, or of there being any point to doing anything at all. Fortunately, few of us ever do call these fundamental concerns into serious question. We do not seem to need a reason to care for our children or ourselves.

Negotiations with others, as opposed to first person deliberation, generally also take the concerns of others as fixed, once known. And we appeal to those concerns or to states of affairs that we know to relate to them when trying to convince others of alternative courses of action. We aim to find shared values and base our negotiations on them. Does such sharing point to a deeper objective basis, or does it end there? Is appeal to what we value the ultimate appeal, or is there an independent truth about what we ought to value?

II. Internalism versus Externalism

“Nobody goes there any more because it's too crowded.”

The debate to which the last question points is that between internalists and externalists regarding reasons, or between subjectivists and objectivists regarding value. Internalists hold that the reasons we have for acting are limited by our pre‐existing motivational states, by our desires and concerns. A reason is not a reason intrinsically: in itself it cannot demand on pain of irrationality that agents be motivated by it. It is because agents have certain concerns or desires that they have reasons; their motivational states limit what reasons they can have. Externalists deny such limitation, holding that reasons determine what we ought to do whether we care to do so, indeed whether we have any cares that require us to care to do so, or not. They hold that we are rationally motivated because we recognize the independent force of reasons, while internalists believe that reasons are such because they motivate us or relate to previously existing motivations. The question, in short, is whether, ultimately, motivations generate reasons or reasons generate motivations. Do I have a reason to exercise because I care (p.10) about my health, or simply because it is good for me, independently of my concerns? Are reasons internal or external?

Internalists do not claim that reasons always result in actions in rational agents. They need not be overriding reasons. Motivation can consist in a disposition that is not actualized most of the time. I am concerned about music and therefore motivated to go to concerts, but I do not go constantly even when opportunities arise, because other motivations are often stronger. Indeed, internalists can allow that agents can have reasons even though they are not motivated by them at all at particular times. They may be in irrational frames of mind at those times, too depressed, for example, to care about any of their usual concerns. Or, again irrationally, they may not realize that they need to specify their deeper and broader concerns into specific desires and intentions if they are not to go unsatisfied.

If I am concerned about my health, I need to be motivated to do some kind of exercise and to translate that motivation into an intention to jog or play tennis or . . . Many people are concerned about their health but irrationally never become motivated to exercise. For the internalist there must be such concerns in relation to which reasons are reasons, before a lack of motivation can count as irrational. But agents may not be motivated by the reasons they have. Again, perhaps faultlessly this time, they may not be motivated because they are not aware of certain reasons or opportunities for satisfying their desires, unaware of and so once more unmotivated by reasons they have. Rational agents are motivated by the reasons they have of which they are aware, but agents can fail to be motivated by reasons they have because of ignorance or irrationality.

On the other side, externalists can posit desires whenever we are motivated to act on reasons, but they will see many of these desires as themselves motivated by the reasons, as are the actions that result. If I exercise voluntarily, then in some sense I must have desired to do so, but that desire can spring from the recognition that exercising is good for me. On this account, we come to desire things because we recognize the independent values in them, and those independent values provide the reasons for the desires. Desires themselves can be justified or not, and their justification derives from those reasons that give rise to the desires. Desires that lack reasons should not be acted on and provide no reasons themselves. Again, desiring to go back to sleep does not in itself give me a reason to do so.

(p.11) The question regarding external reasons is equivalent to the question whether there are objective values. Since reasons are considerations that count for or against actions, they are indicative or constitutive of positive or negative values to be pursued or avoided. Reasons exist independently of persons' motivations or concerns if and only if values or normative facts exist independently of persons' evaluations or valuings, as what they ought to value whether or not they do so. Reasons imply values. But values also imply reasons. If values are to be relevant to our lives, they must have practical import. They must be normatively significant: they must give us reasons to act. They must be what we ought to pursue and protect. This is so whether or not they are independent of our current motivations.

The subjectivist, of course, has an obvious explanation for why recognition of value, expressed by certain evaluative judgments about what it would be good to do, should be motivating. Such judgments on that account already express our motivations. The values we recognize according to this view reflect what we care about: there is no question why we should be motivated to achieve or protect what we care about. Caring is just being motivated. Conversely, what seems valuable to us when we have coherent motivations and are relevantly informed is valuable for us. For the subjectivist, values exist only for and because of valuing beings, beings motivated to seek or avoid certain states of affairs. Value subjectivists will be reason internalists. Because values are relations between objects or states of affairs and subjects who value them, and because reasons reflect values, reasons are also relative to subjective values or concerns. The demand to value various objects or objectives follows only from the demand for coherence with deeper existing concerns and motivations. If I ought to exercise, it is because I am concerned about my health, and coherence demands sometimes acting in ways that satisfy that concern.

Likewise, for the current objectivist, objective values reduce to external reasons, reasons to desire certain objects or outcomes.4 This reduction depends on positing an impersonal perspective from which such reasons are grasped. From this perspective, which corrects for our partial valuings, value requires or merits motivation. Rational agents viewing the states of affairs in question properly, will recognize and respect these values. The impersonal perspective must still be one that we can adopt, from which (p.12) we can still recognize reasons for us. But once we recognize objective value, we recognize that it provides reasons for our desires and actions. This claim may not be quite as transparent as the counterpart claim of the subjectivist. If values exist independently of our motivations, they might connect to none of our concerns. But if they connect to none of our concerns, possibly demanding instead that our concerns be other than they are, how and why should these purported values motivate us? The question admits of no obvious answer.

Internalism these days, however, is much like Yogi Berra's restaurant that no one goes to any more because it is too crowded. Philosophers call it the dominant position, but then they all attack it. Among the most prominent philosophers, only Bernard Williams has argued for the position in one or two articles, and Harry Frankfurt has eloquently defined and endorsed it, but without addressing many of the arguments for and against.5 On the (p.13) other side are a slew of the most famous names in philosophical circles, all arguing for objective values and external reasons.6

Bernard Williams opened the current debate for the internalists by arguing that there is no rational route to an agent's future motivation except through reasoning from her present motivational set. Reasoning can motivate us only by proceeding from prior motivation, and reasons must be capable of motivating us. If we do not begin from something we care about, we will not end by caring about something. If I do not care about playing golf, for example, then the only way I can be argued into being motivated to play is by appeal to something I do care about—my health, my desire for relaxation, exercise, competition, or business deals, my aesthetic appreciation of the outdoors. No objective features of the game will motivate me unless connected to such concerns or dispositions to enjoyments.

Externalists have replied to Williams that an agent can come to be motivated by coming to see the true values and disvalues in things,7 much as we can come to have new beliefs and even new concepts by perceiving things we had not seen before. In dramatic cases a person can undergo a character conversion, acquiring entirely new sets of motivations and losing old ones, by vividly experiencing others' needs to which she was formerly insensitive, or by coming to see that what she was formerly pursuing, for example wealth or fame, was not really worth the effort. Once again for externalists the sources of reasons that will motivate rational agents are such facts about what is or is not worth doing or promoting. My reason to play golf is not that I desire to or have other desires, but simply that I would enjoy it, and enjoyment is objectively good.8 On the other side, suffering ought to be relieved, and a person unmotivated to do so can come to be so motivated by coming into close contact with another's suffering. We need no prior motivations in order to be able to discover true value or disvalue.

(p.14) Both sides to this debate can agree that we can care about what is not truly valuable for us, that which we have no reason to care about. Both can agree also that we can fail to care about that which is good for us, what we have reason to pursue. Finally both can agree that reasons are not necessarily what motivate us, but what ought to motivate, what motivates if considered correctly or rationally. Here the externalist's interpretation of these claims may strike us as simpler and more straightforward, since he sees value as independent of all our actual motivations. If value exists independently of people's concerns, it is clear why it can come apart from those concerns.

The internalist, by contrast, views irrational desire in terms of incoherence within motivational sets. We may be motivated to pursue what is bad for us when an urge to do so opposes a more stable evaluative judgment reflecting a deeper concern. I may desire my next cigarette, but this urge opposes my deeper concern for avoiding terminal illness. Conversely, we may be unmotivated to pursue what is good for us because we lack information or awareness or because we suffer from depression or other irrational frames of mind. I may fail to work on an article or book despite caring deeply about a successful career because I am unaware that I could succeed in publishing it or because I am too lazy or depressed to make the effort. Correct consideration that reveals the true value for a person of various objects or objectives is that which derives from coherent and informed sets of motivational states. Ultimately, values reflect our deepest concerns.

The crucial point regarding the argument of Williams and the externalist's reply to it is that the internalist can always argue that whatever motivations we come to have through encounters with new situations, we must have been disposed to become so motivated. Whenever externalists employ the perceptual model, holding that we can come to see normative facts as they are, come to see the true values in objects and thereby become motivated to pursue them, the internalist can reply that we could not have become motivated had we not been so disposed, had we not already had related concerns or perhaps hidden character traits. A rational action may seem to be out of character to an observer who does not know the details of an agent's dispositions or motivational sets, but it remains true that we do only what we are disposed to do. Trivially, when we act for reasons, we (p.15) must be the sorts of persons who are motivated to act for those kinds of reasons.

If I see people suffering and become motivated to help them when I took no interest in them before, the internalist will believe that I was nevertheless of such a character as to be sensitive to that kind of suffering—after all, some others would remain insensitive in the presence of such pain. The externalist will believe that I acquired a new disposition from the novel experience—after all, people undergo character conversions prompted by new life‐changing experiences. And even if we respond only in ways in which we were disposed to respond, this does not refute externalism about reasons, since we could be disposed to respond to objective reasons or values.

But if the internalist can claim that we are always disposed to do what we do and to value what we come to value, and if these dispositions count as part of the motivations to which reasons are relative, does this not trivialize his central thesis that reasons depend on prior motivations? Does it not trivialize the entire debate, since there will no longer be any difference between his position and the externalist's, for whom it can also be said that we desire to do what we voluntarily do (at least desire the likely outcome more than the available alternatives)? No. While it is true that in some sense we must have been disposed to react positively to anything to which we do react positively, this does not trivialize the internalist's thesis or her debate with the externalist. It implies only that we cannot have an empirical test to differentiate these theories, either from the agent's or the observer's point of view. (An agent may learn of her own character only by noting how she reacts in various circumstances.) But the question remains whether rationality can demand of us that we be motivated by what does not in fact motivate us or connect to our current concerns. This again is equivalent to the question whether value is objective, whether it ought to motivate us however we happen to be disposed. Can objective states of affairs in themselves require motivation from a rational agent?

The answer may be not only of philosophical interest and importance, but may have practical import as well, not only in terms of our broad attitudes toward ourselves and the world, but from the point of view of a moral educator or of someone trying to convince another to change his (p.16) behavior. Can such advice succeed by getting the other person to see the normative facts or rational requirements as they are? Or must one aim at a gradual change of attitudes and feelings through appeal to present concerns for self and others? Ethics teachers who hope to affect their students' behavior and not just their beliefs may well favor the second approach.

If there is no empirical evidence clearly supporting one side or the other to this debate, then it must be adjudicated in terms of broader theoretical considerations. Some might want to include pragmatic or practical effects of belief among the criteria for choice as well, such as the possible positive effect of belief in internalism noted in the previous paragraph. On the other side, and perhaps more significantly, objective reasons and values play the role of God for contemporary philosophers. Their pursuit gives our lives meaning and purpose without our having to create them. Their presence validates the ways we value things. They provide a standard for our subjective values and pursuits to meet and provide reasons for our having the motivations we have. Such considerations, however, aside from providing a contentious explanation for the inclination to belief in objective values in terms of a quasi‐religious yearning, are epistemically irrelevant. They provide no evidence for the existence of objective values, no reasons to believe in them. Aside from removing one bad argument in favor of externalism (namely, why should so many philosophers believe in objective values if they do not exist?), these considerations do not provide evidence against belief in objective values either (Just ’cause you're paranoid doesn't mean they're not out to get you; just ’cause you believe for bad reasons doesn't mean your beliefs aren't true.)

But there is a related argument that is epistemically relevant, if not sound. A philosopher might compare belief in objective values to belief in real physical objects. Realism about physical objects provides a deeper explanation of our perceptual experiences than does phenomenalism, the theory that countenances only perceptual experiences and views objects as regular patterns or collections of experiences. Similarly, it can be claimed, realism about values, appeal to values independent of our desires, provides a deeper justification for our desires and subjective values. Appeal to physical objects explains not only when, but why our perceptual experiences occur. Phenomenalists cannot provide that explanation. Likewise, appeal to (p.17) objective values justifies even our deepest concerns. Subjectivists cannot provide that justification. Motivations, the source of values for the subjectivist, can themselves be supported by reasons, and, lacking such support, can be irrational. Objective values provide the reasons for those deep concerns that otherwise would lack them.

There is, however, an important disanalogy between these arguments and a closer analogy from the epistemology of belief to the debate over values that supports internalism. The disanalogy is the difference between appeal to explanation in the one argument and the appeal to justification in the other. The analogy relates to justification itself in the two domains. If we stick to explanation in both domains, the analogy disappears. While appeal to physical objects explains our perceptual experiences (my desk causes me to have visual experiences by reflecting light, and so on), and natural selection perhaps explains how we have all come to perceive physical objects as we do, the explanation of motivations in terms of objective values remains mysterious. Natural selection might explain the common instinctual desires with which we begin life, but what we come to value differs among individuals and often has little to do with survival or reproduction. And while the physical causal chains involved in perceiving objects can be specified in detail, the causal chains presumed to connect objective values to desires remain unspecified.

An objectivist can reply that, since objective values are instantiated or realized in ordinary physical objects or states of affairs, the causal chains linking such values to our motivations and actions are the usual physical causal chains. If it is then asked how the values themselves enter into the causal explanations, how they, and not simply the physical states of affairs in which they are realized, are causally efficacious, the objectivist replies that the explanation in terms of value is better because we react in the same ways to the same values however they are realized physically. The same values will have multiple physical realizations. Broader explanations of our motivations in terms of the values themselves instead of their particular physical instantiations are preferable if we would react in the same way to any instantiations of the same values. If a baseball batter cannot hit any curve ball, then his striking out is normally best causally explained by the pitch having been a curve ball, and not by the particular physical path it took. Broader explanations are often better than narrower when the (p.18) narrower or more specific explanations are part of a common class of causes any of which would have worked in the circumstances.9

In this way the objectivist can claim that values enter best causal explanations for motivations and actions, even though the particular causes of these desires and actions are the physical causes we all believe in. But the answer is not entirely satisfactory. The question remains why our motivations track these particular sets of physical states of affairs that are said to realize the same values. It still seems that we must be magically tuned in to these values if we do not reverse the explanation and say that we see them as values because we care about them or are motivated by them. And the latter explanation also explains why people disagree over what is objectively valuable. If I admit that I cannot hit a curve ball, we understand why I cannot hit all pitches of that type (perhaps I am always swinging over them). But if I say that I like baseball and other competitive sports because they are good, it remains mysterious why goodness attaches to all of them. And when I specify what for me are their good‐making properties, the question remains why all these properties are good, and why other people do not react in the same positive way to them.

Turning to justification in both domains, a different analogy appears but ultimately points in different directions in the two domains. The analogy is best brought out by extending a point made by Harry Frankfurt. Bernard Williams, we noted, had argued that we cannot reason to new motivation except from other current motivations. Frankfurt argues in similar fashion that we cannot assess claims of value except in terms of what we find valuable, in terms of what we value or care about.10 Just as we cannot look beyond our set of beliefs to compare them to the world as it is in itself, so we cannot stand back from all our concerns at once in order to compare them to our impersonal grasp of objective values.

Traditional foundationalists in epistemology deny the former claim. They hold that we can check our perceptual beliefs against “the given,” unconceptualized raw perceptions. Such non‐conceptual content might exist as a source of empirical concepts. One cannot acquire the full concept “red” except by first perceiving red things. But once conceptualized, once (p.19) perceived objects appear to be some way or to have certain properties, perceptions immediately give rise to beliefs. It is then doubtful that we can bracket these beliefs, strip away the concepts, and return to the raw experience to verify the truth of the beliefs. Once I have the concept “red,” in certain conditions I cannot help believing that I am seeing something red. Justification here, as in the case of values, must be a matter of coherence within the belief system.

But if depth of explanation adds to the coherence of a set of empirical beliefs, if beliefs about the physical world are more coherent if they receive an explanation by appeal to physical objects as the causes of our perceptual beliefs, then the criterion of coherence in itself justifies a realist stance toward the physical world.11 If physical objects and their properties are to explain the ways we perceive, then these objects must really exist. In the case of values, by contrast, we add only mystery and not deeper understanding by appealing to objective values as the cause of our subjective values. As noted, the full account of the causal connections remains unspecified, as opposed to the detailed and complete causal explanations we have in the case of perceiving physical objects.

Realism in both domains posits a set of objects or objective states of affairs existing independently of our subjective states. It implies that those subjective states that represent those objective states of affairs as being some way or having certain properties could misrepresent them. If objects and values are independent of our beliefs and concerns, then our beliefs about them can be mistaken and our values misguided. Realism implies a wholesale possibility of error. But while the logical possibility of wholesale error in our beliefs about the physical world is intelligible (although probably we would not have survived such universal misrepresentation), there seems to be no sense to the claim that we all ought to care about things completely unrelated to any of our present concerns and ought to give up all of those concerns. Could the true nature of the world and of the values objectively in it demand that we not care for ourselves or our children? Would you respond to or care about such a demand? If not, what practical relevance could objective values have?

(p.20) III. Advantages of Internalism: A Modified Humeanism

“If you can't imitate him, don't copy him.”

This book, you may have by now surmised, will defend reason internalism, the view that all practical reasons derive ultimately from our concerns. There are no objective values or external reasons. This position is favored by its transparent intelligibility as well as its metaphysical and epistemological economy. While your judgment of its intelligibility will have to await its full presentation in subsequent chapters, it certainly enjoys an initial advantage in this regard. Reasons on this view are states of affairs that motivate rational agents, and values are states of affairs that are rationally evaluated positively. Rationality consists in being relevantly informed and coherent, in which case one acts on the reasons one has. We have at least an initial grasp of what it is to desire, to evaluate, and to be informed and coherent, although all these concepts will be analyzed fully in what follows. We come to see certain facts as reasons when they indicate how to satisfy our desires. And indeed they normally are then reasons for us. Here it is not mysterious how reasons motivate us and how we know they are reasons.

The full analysis of desire to be presented in a later chapter will be controversial—somewhat more so than a parallel analysis of emotion—but both analyses are rooted in our intuitive grasp of what it is to have a desire or to be in the grip of an emotion. And this intuitive grasp is in turn grounded in our common experiences of desiring things, having emotions, and ascribing these states to others. We understand what subjects' desires and evaluations are and how they perceive reasons to act in terms of them. It is not mysterious that if I desire to play tennis and think that playing is good for me, and if this desire coheres with my other concerns for health, competition, and so on, then I see that I have a reason to play. We understand the claim that reasons are states of affairs that motivate rational agents, and we see why they should motivate if they reflect already existing motivations or concerns.

By contrast, I have pointed out that appeal to objective values as the causes of our desires remains mysterious. Objective values or brute, irreducible normative facts involve non‐natural properties, properties not (p.21) countenanced by any scientific theories. Furthermore, these properties demand motivation from us, demand to be pursued and protected. The mysterious nature (or “queerness,” as J. L. Mackie put it)12 of such properties is perhaps reduced somewhat by holding first that they need not necessarily or intrinsically motivate us, but only determine what we ought to value or desire, only provide external reasons. Second, it is widely claimed by believers in them that objective values supervene on natural properties. This means that there can be no differences in value without some differences in ordinary natural properties. If two objects share all their other properties, they must have the same value. The ordinary natural properties provide our reasons, not the additional fact that they are valuable. If golf provides exercise, and exercise is objectively good, then you have a reason (perhaps easily defeasible) to play golf. The fact that it provides exercise and promotes health, a natural fact, is your reason to play, not the additional non‐natural fact that health and exercise are good. Their being good just is their providing such (external) reasons; it is not an additional source of the reasons.

But, as in the case of causal explanations discussed earlier, while this “buck‐passing” account of objective value13 perhaps reduces the mystery somewhat, it does not eliminate it. The question remains why these properties provide value or constitute reasons, and the natural answer that they do so by connecting to our concerns or motivations is unavailable to the objectivist. For her it remains a brute inexplicable fact that values as non‐natural properties attach to certain physical objects and states of affairs. By contrast, the subjectivist or internalist need not ask how motivations create values, if to desire an object is in part just to judge or see it as valuable for oneself (part of the analysis of desire to be defended below). What we desire and judge to be valuable is so if we are coherent and informed of the nature of the object. When fulfilment of desire fails to create value for a subject, it is because the subject was not informed of what its satisfaction would be like, or because the satisfaction blocks that of more central concerns; that is, is inconsistent with them.

Furthermore, as others have pointed out, the concept of supervenience itself is mysterious if values do not reduce to natural properties. Why (p.22) should there be no difference in values or reasons without a difference in natural states of affairs if the former do not somehow consist in the latter? The answer might be offered that values supervene on ordinary natural properties because they are always realized or instantiated in those other properties. Value is realized in the form of an artwork, in the courage of an heroic act, in the pleasure of a relaxing day at the beach. But it remains mysterious why objective values attach to just these properties from which they are distinct and why they should demand motivation from us.

Thus the intelligibility of the subjectivist account is tied to its metaphysical economy, since the excess metaphysical baggage of the appeal to objective values and brute, non‐natural normative facts as causally efficacious renders the objectivist account mysterious. Epistemological clarity and economy are also connected. If there are non‐natural normative facts, and if we are to know of them, then our ordinary perceptual, affective, and cognitive abilities would seemingly need to be augmented by a special faculty of value intuition. Once more, even if values reduce to other natural properties in their instantiations, we would have to intuit that these properties are valuable. It can be claimed that intuition is not a special faculty, but simply a matter of certain propositions that are not inferred from others seeming true to us when we think about them.14 Certain value judgments, such as “pain is bad,” seem true to us in that way. But the subjectivist can explain why this proposition seems true, namely, because those who judge it to be true are motivated to avoid pain: they desire its absence. The objectivist or intuitionist cannot provide any such explanation, and so her intuition remains mysterious.

Objectivists can reply that we need a faculty of intuition anyway to apprehend logical or mathematical truths. We do not literally perceive that two plus two is four, but must intuit the truth of this proposition. And we do not know how we do that; certainly it has nothing to do with desires. But even if that is correct for mathematics, mathematical truths seem quite different from values. It cannot be the same faculty that apprehends both, although it often goes by the same name. It seems instead that intuition is just a name for some way of knowing things that we cannot understand or locate. There must be mathematical truths that we know whether or not we can understand how we know them, but the same is not true of objective (p.23) values. Being more controversial, it is a strike against belief in them that we cannot understand how we could know or be motivated by them.

By contrast, we generally know our own desires and concerns. For the subjectivist, to know whether something is valuable or good for me, I only need reflect on what I find valuable or am motivated to pursue when relevantly informed and coherent. If reasons are relative to desires, and if desiring something is in part to be disposed to bring it about (again part of my analysis), then we understand both why states of affairs that reflect our desires are perceived as reasons and why apprehension of reasons results in action. Once again reasons simply indicate ways to satisfy our rational desires or concerns.

In addition to being more intelligible and economical, the internalist account to be defended presents a unified view of the entire domain of reason. As indicated, both theoretical and practical rationality, rationality in belief as well as desire and action, consist mainly in coherence. Coherence will be analyzed as the avoidance of self‐defeat. Believing what is unlikely to be true based on other beliefs and evidence is self‐defeating, since the aim of belief is truth. Similarly, failing to act on the strongest reasons one has, those which reflect one's deepest concerns and evaluative judgments at the time, is also self‐defeating, since the aim of action is to satisfy those desires or concerns. Thus rationality in belief and rationality in desire and action will be given similar analyses.

While there are both reasons for beliefs and reasons for action, there can be one account of what reasons are. They are states of affairs that motivate rational subjects, either to believe or to act. If we are rational, we are all motivated to believe what strikes us as true, and so we will base our beliefs on evidence or indications of truth. Epistemic reasons are the same for all since we all aim at truth, and truth is not relative to our beliefs or desires. And if we are rational, we will form intentions to act on what will satisfy our coherent and informed concerns and desires: we will act on the strongest reasons we have at the time of which we are aware. In fact, as pointed out at the beginning, since most of us are rational most of the time and tend to act on our desires, we will tend to act on reasons automatically. But since we do not all have the same desires or concerns, practical reasons will not be the same for all.

Nevertheless, although practical reasons are not all shared, the nature of reasons, both practical and epistemic, admits of a single analysis. Reasons (p.24) motivate rational agents, and agents are rational if coherent and relevantly informed. While my account will assimilate theoretical to practical reason with the aim of avoiding self‐defeat as central in both domains, the objectivist will tend to go in the other direction, assimilating practical to theoretical reason. Practical rationality for her will consist first in forming true beliefs about what is valuable. But there are two problems here. First, she will have to explain how we come to have mostly true beliefs in this domain: how we come to track independent values through their disparate instantiations. Then there will be the additional and different step of explaining why we should be motivated to act so as to achieve what is objectively valuable, if it connects to none of our current concerns. At best the theory will be less simple and unified.

Given all these theoretical advantages of the internalist account of reasons and subjectivist account of values, why do so many philosophers these days, not all of whom are not as smart as I am, argue for the opposing position?15 First, despite its simplicity, unity, and intelligibility—all criteria for theory choice—they believe that it fails to satisfy the one prior criterion: they believe that it cannot account for all the data. According to them, it cannot account for reasons that seem to exist independently of our motivations. These include reasons for desires or basic concerns themselves, reasons that seem to give rise to new motivations instead of deriving from existing desires, moral reasons that we seem to have whether we care to have them or not, and certain seemingly rationally required prudential reasons, concern for our own welfare. In regard to the latter classes of reasons, a person who is banging his head against a wall or ignoring his own child's acute suffering appears irrational no matter how much he desires to do the one or is unconcerned about the other.

Given the theoretical advantages of internalism, the burden of proof is on the externalist to show that there are reasons for which the internalist cannot account. But once the externalist produces these purported examples, that burden shifts to the internalist to show that he can indeed explain or reasonably dismiss these cases. That is a burden I cheerfully accept for the fourth chapter. I will argue that there is no universal rational requirement to accept moral or prudential reasons as one's own, and that the internalist can explain both reasons for desires that have reasons and the acquisition of (p.25) new motivations in the course of experience. Basic concerns, by contrast, need no reasons to explain or justify them. I need no reason to care for myself or my loved ones.

A second reason for favoring objectivism and externalism is the phenomenology of seeming to encounter values or reasons in the world and the experience of deliberating. As noted earlier, we do not normally focus on our desires in deciding what to do, but instead focus on states of affairs or possible outcomes that immediately strike us as good or bad. We evaluate objects and not our own motivations. And these objects seem in themselves to be of positive or negative value. Whether a book should be recommended for a prize is determined by the value of the book, not by how many people desire to read it. Reasons, as what we deliberate about, are themselves good and bad states of affairs—that the book contains new ideas or is written in a dull style. Often we discover our own motivations and the priorities among them only by seeing what strikes us as more important or valuable in the world, by how we react to various objects or opportunities.

We know, however, from science and philosophical reflection in other domains that phenomenology can be misleading as a guide to metaphysics. Colors seem in experience to be not only real, but objective—out there on the surfaces of objects independent of our visual systems. But we know that colors are relational properties, analyzed only by appeal to the ways they appear to us, to subjects with our visual systems. Similarly, reasons are states of affairs out in the world, but their status as reasons may still be relative to subjects' motivational sets. That a tennis racquet is on sale is a reason for me to go to the sporting goods store, but only because of my interest in playing tennis. Although such concerns remain in the background as various states of affairs present themselves as opportunities or threats, it still seems that every evaluative judgment is made from the perspective of a person with a set of concerns more or less widely shared. We need not think about our desires or their connection to reasons in order correctly to see certain facts as reasons, or in order for the connection to determine their status as reasons.

The most philosophically astute reader might already question my claim of transparent intelligibility for the position I defend, in that she might already have noted some apparent contradiction or at least tension in my earlier claims. For one, it was claimed that “just feeling like it,” a single (p.26) desire in itself, provides no reason to do anything, although it is desires or concerns that give rise to reasons. My claim will be that only some desires give rise to reasons, and the third chapter will clarify which ones do so. I will claim that only coherent and informed sets of desires anchored by deeper concerns give rise to reasons. Furthermore, the plausibility of the claim that any desires give rise to reasons depends on a proper analysis of desire. If desires were simply felt urges or purely affective states, they would not generate reasons. But desires include implicit evaluative judgments as components that reflect deeper concerns with which they cohere. Such desires do create reasons that indicate how to satisfy them.

The student desires to party, and seemingly does not desire to study, but she has every reason to study and no reason to party. This is because she is concerned not to fail and has other concerns that cannot be satisfied if she does fail, although none of these concerns may be felt as urges; none may involve any yearning sensations or purely affective states. These concerns may be reflected only in her sober judgment that she had better not fail. But such evaluative judgments can be themselves parts of desires or motivational states. Usually they reflect deeper concerns and more stable or standing desires that do not always generate the sensations that attach to urges. These cognitive components of desires are normally better indications of reasons than are the purely affective components. But again, only some evaluative judgments count as motivational, and the third chapter will once more clarify which ones do.

Thus desires, like emotions, will be seen to be complex states involving different affective, dispositional, cognitive, and functional components none of which is necessary or sufficient in itself for ascribing the states. The concepts of desires, emotions, and attitudes are all what philosophers call cluster concepts, different clusters of components in each case, some of which may be absent in particular instantiations. My desire for a piece of cheesecake involves a yearning sensation, but not an evaluative judgment that eating one would be good, while my desire to finish this chapter involves the judgment that finishing it would be good, but not a felt urge. Motivational states that generate reasons include not only such occurrent or superficial desires, but deeper concerns that organize and explain the more specific and superficial desires. These concerns almost always include evaluative judgmental components. My concern for my own welfare includes the implicit judgment that it would be good for me to promote it, (p.27) and this concern generates more specific desires related to my health and comfort.

The position I will defend, the thesis that all value and practical reasons derive from desires or “passions,” derives from Hume. But I advocate a very much modified Humeanism. Transforming the quotation at the start of this section, we should imitate Hume's position but not copy him. Hume himself believed that practical rationality reduces to rationality in belief and instrumental rationality. It demands only that we act on true beliefs and adopt efficient means to our ends. Ends themselves are beyond criticism. I will argue that this requirement to adopt effective means to our ends is itself part of the demand for coherence as the avoidance of self‐defeat. This demand generates broader concepts of practical rationality and irrationality than Hume recognized.

Most contemporary Humeans view reasons as complexes of desires and beliefs. According to this view, my reason to go to the sporting goods store is my desire for a new racquet and my belief that I can buy one there. I reject that analysis. Desires are not reasons even when combined with beliefs, and single desires in themselves do not provide reasons. Reasons are what we deliberate about, and they justify and explain our actions. While desires and beliefs are parts of the explanatory causes of our actions, they are not typically objects of deliberation and do not singly in themselves justify our actions. If I want to find out whether I have a reason to go to the sporting goods store, I inquire whether there is a racquet on sale there. That there is one on sale is my reason, and if there were not, I would have no reason to go there, whatever my desires and beliefs about racquets may be. But that there is one on sale there is a reason for me because it indicates how to satisfy my desire for a racquet, which connects to my concern to play better tennis, to exercise and compete successfully, and so on.

The claim that motivational states include cognitive components in the form of implicit evaluative judgments, and that these are generally more reflective of our reasons than are the purely affective components of desires, may seem to be anti‐Humean. The complex analysis of desire so as to include judgment is indeed opposed to much of what Hume says about passions. In regard to judgment itself, however, he held only that beliefs about or judgments of objective facts lack motivational force unless connected to desires. For him, evaluative judgments, specifically moral judgments in themselves, are motivating or motivational (components of (p.28) or reflective of motivations). These are not judgments of objective facts, which are motivationally inert. I will disagree with the claim that moral judgments always express motivations, but I will argue that certain other evaluative judgments do. And of course I agree with the crucial thesis that objective facts in themselves provide no reasons and rationally require no motivations.

I share Hume's skepticism about objective values, and therefore I am skeptical of external reasons as well. Reasons are states of affairs and not desires, but they depend on subjects' motivations, including their deepest and broadest concerns. This account fits well with a plausible analysis of desires themselves. Suitably amended and expanded, as it will be in the chapters to follow, Humean internalism provides our most economical and unified theory of practical reason. This theory informs us that, although reasons for action are states of affairs in the world, what makes them reasons lies within valuing subjects.


(1) If we typically act for reasons without deliberation, when, if ever, do we act for no reason or without reason? We do so when we act on a pure whim, just because we feel like it, when this feeling is unconnected to any more stable concerns, or when we falsely believe we have a reason. Such action is irrational, as opposed to arational, only if it blocks the satisfaction of those other deeper concerns.

(2) As pointed out by Jonathan Dancy, Practical Reality (Oxford: Oxford University Press, 2000), pp. 85–94.

(3) This argument resembles one given by Bernard Williams, “Internal and External Reasons,” in E. Millgram (ed.), Varieties of Practical Reasoning (Cambridge, MA: MIT Press, 2001).

(4) See, for example, Thomas Nagel, The View from Nowhere (Oxford: Oxford University Press, 1986), p. 139.

(5) Bernard Williams, “Internal and External Reasons”; Harry Frankfurt, The Reasons of Love (Princeton: Princeton University Press, 2004). After publishing the articles on which this book is based and writing most of its first draft, I discovered a just published book by Mark Schroeder, Slaves of the Passions (Oxford: Oxford University Press, 2007) that also defends the internalist position. We share the basic thesis that reasons derive from desires, but his arguments for the position, his analyses of desires and reasons, and his corollary theses are very different from mine. In his initial argument for the position, for example, he assumes that all philosophers agree that some reasons derive from desires and that we should therefore seek a unified account according to which all reasons share this source. But the premise is false: most objectivists will claim that when one person has a reason to go to a party, desiring to dance there, and another person who does not like to dance does not have that reason (Schroeder's example), this difference in reasons derives not simply from a difference in desires, but from the fact that the first person will derive pleasure from the party and the second will not, pleasure being an objective value that we all have reason to pursue. Objectivists hold that we need reasons to satisfy our desires, and these reasons derive from objective values. And the inference that if some reasons depend on desires, we should seek an account in which all do simply seems to assume that there is one source of all reasons. I too will argue for a unified account, but the unity in my theory derives from an analysis of rationality that encompasses both epistemic and practical rationality. Our different analyses of reasons leads me to explain away what Schroeder accepts and to accept what he attempts to explain away. Specifically, he accepts that every desire in itself can generate reasons that explain why certain actions tend to satisfy the desire. By contrast, I agree with objectivists that desires to go back to sleep or to count blades of grass in themselves generate no reasons, but I provide an internalist explanation for this. Schroeder attempts to provide an internalist explanation for the central claim of objectivists that we all have moral reasons of equal weight. But his explanation fails, as I will argue later, and I reject the claim. Finally, Schroeder attempts to reduce the normative concept of reasons directly by an analysis is non‐normative terms. Once more his analysis implies that every desire can generate reasons, which I reject. I share the goal of naturalizing the normative, but proceed to do so indirectly by analyzing reasons in terms of rationality and then providing a naturalistic account of rationality. I provide further argument in the text for why this order of analysis is preferable. All these major differences in our approaches will become clearer in my subsequent chapters. I recommend the reader to compare our defenses of the central internalist thesis that reasons derive from desires, and especially recommend reading his excellent discussion of the nature of metaphysical reduction, a meta‐theoretical discussion of the sort that I do not attempt to provide here.

(6) e.g. Thomas Scanlon, What We Owe to Each Other (Cambridge, MA: Harvard University Press, 1998), pt. I; John Searle, Rationality in Action (Cambridge, MA: MIT Press, 2001); Joseph Raz, Engaging Reason (Oxford: Oxford University Press, 1999); Derek Parfit, Climbing the Mountain, unpublished but widely circulated manuscript; Warren Quinn, Morality and Action (Cambridge: Cambridge University Press, 1993); Nagel, The View from Nowhere; Dancy, Practical Reality.

(7) See e.g. Brad Hooker, “Williams' Argument against External Reasons,” in Millgram (ed.), Varieties of Practical Reasoning.

(8) The internalist replies that I have reason to pursue only those enjoyments I care about, if I do not care about pursuing every possible enjoyment.

(9) For expansion on this point, see Alan Goldman, Empirical Knowledge (Berkeley: University of California Press, 1988), p. 34; for its application to values, see Ralph Wedgwood, The Nature of Normativity (Oxford: Oxford University Press, 2007), p. 195.

(10) Compare Frankfurt, The Reasons of Love, p. 26.

(11) This is not an epistemology book, and so not the place to defend these claims made by way of comparison. But for a defense and an alternative kind of foundationalism, see Alan Goldman, Empirical Knowledge (Berkeley: University of California Press, 1988), chs. 4, 7, 9.

(12) J. L. Mackie, Inventing Right and Wrong (New York: Penguin, 1977), pp. 38–42. I use the term “mysterious” because I take it to be no longer politically correct to call properties queer.

(13) The account is found in Scanlon, What We Owe to Each Other, pp. 95–100.

(14) Michael Huemer, Ethical Intuitionism (New York: Palgrave, 2005), p. 101.

(15) I am discounting Yogi Berra's explanation, that they simply made too many wrong mistakes.