Jump to ContentJump to Main Navigation
Kantian Ethics and EconomicsAutonomy, Dignity, and Character$

Mark White

Print publication date: 2011

Print ISBN-13: 9780804768948

Published to Stanford Scholarship Online: June 2013

DOI: 10.11126/stanford/9780804768948.001.0001

Show Summary Details
Page of

PRINTED FROM STANFORD SCHOLARSHIP ONLINE (www.stanford.universitypressscholarship.com). (c) Copyright Stanford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in SSO for personal use. Subscriber: null; date: 23 September 2021

Individual in Essence, Social in Orientation

Individual in Essence, Social in Orientation

(p.86) Chapter 3 Individual in Essence, Social in Orientation
Kantian Ethics and Economics
Stanford University Press

Abstract and Keywords

This chapter examines character, which includes both judgment and the will, in relation to a general Kantian view of the economic individual. It explores how this individual differs from the homo economicus of neoclassical economics and other ideas introduced by sociologists, philosophers, and heterodox economists (particularly social economists). It argues that the economic agent is individual in essence and social in orientation. The chapter first looks at individualism and sociality before comparing the Kantian-economic model to atomism and its opposite, social embeddedness. It defends a limited sense of atomism as an implication of autonomy that itself implies sociality rather than an asocial orientation. It then considers several concepts of identity as they apply to the Kantian individual and argues that an individual is defined, individuated, and identified over time by his/her unique character. It also analyzes the recent work of Christine Korsgaard on practical identity and self-constitution to explain how character is self-created and maintained by the individual's choices and actions.

Keywords:   character, judgment, will, economic individual, individualism, atomism, sociality, social embeddedness, Christine Korsgaard, Kantian-economic model

The first two chapters of this book focused on the more technical details of Kant's moral philosophy, such as the categorical imperative, the nature of duties, and the operation and strength of the will. In this chapter, I want to pull back a bit from the trees and bring the forest into view, eschewing the technical details and focusing instead on character, which I regard as including both judgment (from Chapter 1) and the will (from Chapter 2). I do this in order to steer the discussion toward a general Kantian view of the economic individual, and explain how this individual differs from both the homo economicus of neoclassical economics and other conceptions brought forward by philosophers, sociologists, and heterodox economists (especially social economists). The thesis that I defend in this chapter is that the economic agent is individual in essence and social in orientation. A person's autonomy implies that she is an individual who makes choices of her own judgment and will, and, recognizing the equal dignity of other persons, she is led to take those other persons into account in her autonomous decision-making. To that end, the first part of the chapter will focus on individualism, and the second will focus on sociality, both implied by the same reading of Kantian dignity and autonomy that I emphasized in the first two chapters (and which continues throughout the book).

(p.87) Individual in Essence

I have three goals for this part of the chapter. First, I will defend the way in which the Kantian-economic model represents the individual and compare it to atomism, an aspect of the standard economic individual which is widely derided by heterodox economists, and its opposite, social embeddedness. I will defend a limited sense of atomism as an implication of autonomy that itself implies sociality rather than an asocial orientation in which the agent has no regard for other people. True to autonomy, an agent's decision is her own, independent from any inside or outside controlling influence, but she may—and often should—allow influences from many sources, especially a concern for her fellow human beings. In other words, atomism describes how you make choices, and sociality comes from what choices you make. Properly understood, atomism describes the process of choice rather than its content; the Kantian-economic agent autonomously and atomistically makes socially oriented decisions (consistent with a specific understanding of social embeddedness).

Second, I will discuss several concepts of identity as they apply to the Kantian individual. In 2003, John B. Davis published The Theory of the Individual in Economics, a groundbreaking work in which he analyzes and rejects the neoclassical conception of the individual. Using concepts of identity theory from philosophy, Davis argues that this neoclassical conception, which identifies the individual with his preferences, can neither individuate agents at any given time (because two persons may share the same preferences) nor identify an agent throughout time (because, contrary to standard heuristic assumptions, preferences can and do change over time). More constructively, he argues for an essentially socialized conception of the individual and argues that, ironically, social embeddedness provides a better basis for identifying the economic individual than the atomistic conception of neoclassical economics.

I agree with Davis that an agent's social bonds and roles are very important, particularly to her sociality (to be discussed in the second part of this chapter), but they fall short of defining her identity. Rather, building on the discussion in Chapter 2, I argue that an individual is defined, individuated, and identified over time by her unique character, made up of her judgment and will; after all, what can be more intrinsic to a person than her faculties of deliberation and choice? In the final section of this part of (p.88) the chapter, I will borrow from the recent work of philosopher Christine Korsgaard on practical identity and self-constitution to explain how an individual's character is self-created and maintained by her choices and actions, and how this interacts with her preferences as well as the social aspects of her identities in a reflexive process, which leads into the discussion of sociality in the second half of the chapter.

Individualism, atomism, and social embeddedness

As we know, Kantian dignity endows persons with an immeasurable, incomparable value or worth, which demands respect from others as well as from themselves. But dignity is not primary; it derives from autonomy, the capacity for self-governance that separates humans from beasts. An autonomous person is a self-aware agent: she acts in the world, she knows she acts in the world, and she chooses how she acts in the world. More precisely, she determines for herself the laws that guide her decisions and her actions, without undue or unreflective influence from either external or internal factors, and then chooses to act according to those laws. To the extent that others deceive or coerce her, they are failing to respect her dignity; to the extent that she allows “alien” influences to corrupt her judgment or her will, she fails to respect her own dignity (and therefore herself).

Based on this sense of autonomy, I maintain that the Kantian economic agent is strictly individual in terms of her capacity for free choice, which in turn must conform to the laws of her own cognition and conation, and through which all other influences must be filtered. These influences, the myriad factors that may affect her choice, may be deeply social in nature. She may consider the feelings and wellbeing of her family, her friends, her co-workers, the members of her community—and indeed, in the appropriate context, she should (as we shall see in the second part of this chapter). She may consider how she herself fits into the web of sociality which she has woven around her; she may value her family ties more than her work-related ties, or seek to strengthen her community ties at the risk of weakening some ties with friends from “back home.” And she may be urged by various persons in her circle to do certain things and refuse to do others. But regardless of the salience of these aspects of choice, the (p.89) moral law must be supreme, and she must make a choice which is consistent with her judgment, whether it is in conformity with these social factors or in opposition to them. Particularly in the case of perfect duty, she cannot allow any other considerations to affect her resolve to perform that duty (except when it indicates a conflicting obligation). Only in cases of imperfect duties (such as beneficence), which do not demand a specific act but rather a general but sincere attitude, or an apparent conflict of duties, can these obligations be fulfilled with a mind to empirical social factors.

Autonomy demands that we reflect on our preferences and incentives, subjecting all the various aspects of them to our judgment (based on the moral law) when appropriate, and endorse them only when they are not in opposition to our moral character.1 When a person makes a choice, she must be sure that the choice is truly hers, and that she has made it in the best way she can, consistent with the spirit of the moral law. If someone asks her to do something, she cannot automatically do it like an animal following a command or a computer executing a program. If the request is innocuous—“please pass me the sports section”—then her judgment is not invoked; but if the request is morally questionable—“if my wife asks, don't tell her you saw me here”—she must reflect on whether the requested action is something she could do in good conscience. If she blindly “obeys,” she is allowing herself to be used as a tool, a mere means to another's end. If, after reflection, she chooses to go along with the planned ruse, then she has made her own choice, voluntarily inserting herself into the causal chain of events, and she is therefore responsible for the consequences. Autonomy implies responsibility, not just to make the right choices, but for the consequences of the bad ones.

The Kantian-economic person's choice is hers and hers alone, and in this sense she is an individual agent. How does this conception compare with atomism, which John Davis describes as “the idea that individuals are fully autonomous beings in the sense of possessing independent choice sets”?2 If atomism is taken to deny any other-regarding preferences or inclinations, where external factors are limited to impersonal constraints (prices, opportunities, and the like), then this does not apply to the Kantian-economic individual at all. She is certainly not blind to others, their pleasures and pains, their needs and desires—nor is she blind to her own desires and needs concerning others. But she does put these needs and (p.90) desires, others' as well as her own, in the proper context, and does not let them unduly influence her choices when they conflict with the moral law. Part of this is a matter of judgment—knowing when such factors should count and when they should not—as well as her will—being able to resist such factors when they threaten the execution of her best judgment. But the Kantian-economic agent is not atomistic in the sense of making choices in isolation from the concerns of (or regarding) other people.

Nonetheless, there is an essential atomistic element to the Kantian individual implied by autonomy: the process of choice is atomistic in that, at bottom, the individual makes her own choices, making use of outside factors only when she judges it appropriate. Furthermore, the second part of this chapter emphasizes that Kantian duty requires that external influences and concerns play an important role much of the time, orienting the essential individualistic agent in a very social way. Remember that even the universalization procedure based on the Formula of Autonomy depends on social knowledge; for example, lying is shown to be self-defeating because of the effect of universalized lying on trust, an inherently social concept. Social knowledge is also essential for the proper operation of judgment in choosing applicable duties (or resolving conflicts between obligations) and then tailoring the chosen action to the particular (social) situation. Finally, as Andrews Reath writes, autonomous action in the world “presupposes a background of rules and social practices, or better, a system of reasoners able to exercise the same capacities, and limited only by the principle of using their reason in ways that other agents can accept while at the same time continuing to view themselves as autonomous.”3 But regardless of the social context of any specific decision situation, and how much the agent may incorporate social information about it into her deliberation, at the end, her choice is hers and her alone—by virtue of her autonomy, it is essentially and necessarily atomistic.4

Atomism is often contrasted with social embeddedness; these are the two opposing views of the individual analyzed by Davis in The Theory of the Individual in Economics and other work, representing the views of mainstream and heterodox economics (especially social economics). In Davis's words, “the difference between these two conceptions rests on whether individuals and their behavior are explained ‘externally’ in terms of their social relationships or ‘internally’ in terms of their private tastes (p.91) and preferences.”5 But I regard this as a false dichotomy, or at least a non-exhaustive pair. Choices, behavior, and actions can certainly be explained, at least in part, by social factors such as relationships, roles, and norms, as well as by the agent's own preferences, beliefs, and values, which themselves may be influenced by social factors.6 But the process by which all of these enter deliberation, as implied by Kantian autonomy, is strictly internal. For instance, Mark Lutz, another prominent social economist, writes that “persons as social individuals are embedded in a constitutive web of social relations: they value persons and evaluate institutions as to their responsiveness to people.”7 Once again, this is no problem for the Kantian individual, since it does not address the process of decision-making, but only its domain, which can (and should) include social factors, subject to endorsement by judgment.

There are obvious dangers in failing to exercise judgment in the face of social factors. Economic sociologist Mark Granovetter, in a seminal paper on social embeddedness, cautions that this concept can be taken too far, and can result in just as much atomistic choice as under-socialized individualism can:

A fruitful analysis of human action requires us to avoid the atomization implicit in the theoretical extremes of under- and oversocialized conceptions. Actors do not behave or decide as atoms outside a social context, nor do they adhere slavishly to a script written for them by the particular intersection of social categories that they happen to occupy.8

Granovetter's point is very Kantian in spirit: persons can certainly take their social roles and relationships into account when making choices, but to be autonomous they must make these choice themselves, after adequate reflection, rather than blindly following what is expected of them by others. Davis agrees, writing that “the idea of pure embedding is an unsustainable conception,” as does Lutz, who writes that social economics properly “includes decision makers who function neither as mechanical atoms nor as subordinated cells nourished and controlled by social processes.”9

But perhaps the process of judgment is not so atomistic after all; we have seen that social factors inform judgment, but they may go even deeper than that. As Hodgson writes,

Individual choice requires a conceptual framework to make sense of the world. The reception of information by an individual requires a paradigm or cognitive (p.92) frame to process and make sense of that information. The acquisition of this cognitive apparatus involves processes of socialization and education, involving extensive interaction with others. The means of our understanding of the world are necessarily acquired through social relationships and interactions. Cognition is a social as well as an individual process. Individual choice is impossible without these institutions and interactions.10

The ideal would be for the autonomous person to reflect on these social factors as they are incorporated into her cognitive processes, but doing so is clearly unrealistic: many if not most of this activity occurs at a subconscious level, and long before we reach any level of mental maturity. Nonetheless, once maturity is reached, the autonomous agent will reflect on her ways of thinking, and in the process she may discover that some categories she uses (perhaps ones based on racial stereotypes, for example) are improper or immoral, and then try to revise her conceptualization of these issues in respect of the equal dignity of all. (We will discuss reflection and endorsement in the context of the self-constitution of character further below.) And even if social factors run deep into the process of cognition, as long as choice is not determined by any antecedent psychological causes (including those operating on cognition itself, itself a psychological process), there must always be an element of choice that is completely free, and that part is autonomous and atomistic in nature.

In conclusion, the Kantian individual can be very socially embedded in terms of her preferences, influences on her judgment, and her cognitive processes themselves, but at bottom her choices are still her own, which, as I will argue next, defines the nature of the individual that distinguishes her and persists over time. And since this individualism is based on autonomy and dignity, it demands respect and also implies normative or political individualism, as detailed in the next two chapters.

Identity and character

The preceding discussion of atomism and social embeddedness primarily dealt with the proper inclusion of external (social) factors in one's moral deliberation. But, as we know, autonomy also requires an agent to allow her own preferences and inclinations to influence her decision-making only after being endorsed by her judgment (in concert with the moral (p.93) law). But mainstream economists typically identify the individual with her preferences, so denying her preferences when they conflict with duty would mean denying that which makes her who she is.

This conception of the individual defined by preferences, however, has been thoroughly criticized. In chapter 3 of The Theory of the Individual in Economics, Davis argues convincingly (and conclusively) that an individual cannot be identified by her preferences, for reasons drawn from two philosophical identity conditions, identification over time and individuation of persons at any one time. At the risk of oversimplifying Davis's detailed account, simply put, preferences change (contrary to economists' standard heuristic assumption), rendering identification over time using preferences impossible; and preferences are not necessarily unique to each person, making it impossible to use preferences to distinguish between individuals. So preferences are not stable enough to represent who we are over time, nor do they pick out individuals as unique. It follows that if and when a person makes choices without considering her preferences, or even against them, she should not feel she is denying part of herself. In fact, I would argue (as I do in the next section) that she is expressing her true self, the self that lies beneath her mere desires and preferences—in other words, her character (made up of her judgment and will).

But maybe we are setting up an economic straw man with all this talk of preferences; despite economists' methodological strategies, surely no one thinks that a person's favorite flavor of ice cream defines who she is.11 Many people would be more likely to identify themselves with their positions on weighty social issues: abortion, torture, affirmative action, universal health care, the role of the state, and the like. Certainly opinions on such issues, which are more likely to manifest themselves in principles than in “enlightened” preferences, say much more about who a person is than ordinary commodity preferences do. But these opinions, if they are well-informed and reflective, are based on values that derive from one's judgment, and the determination to act on those values—or even simply to express them in front of an unsympathetic audience—reflect one's strength of will. For both reasons, opinions do show more of who a person is, but only because they are so closely related to her judgment and will, which is to say, her character.

But others reject such an internalist view of the individual altogether, (p.94) and look outward, rather than inward, to identify the individual. They may identify her by her social roles and ties: she is a daughter, sister, wife, partner, mother, aunt, grandmother; she is an employer, employee, partner (in a professional sense), associate, intern; she is a friend, confidante, acquaintance, rival, or enemy. In other words, it is claimed, individuals are not identified by who they are inside, but by where they fit in society's web: identity through social embeddedness.12 After all, when we meet a new person, typically the first things we ask about her are what she does for a living, if she is married, and if she has children. Suppose we find out that our new acquaintance is a divorced lawyer with a son; certainly there is likely more than one divorced lawyer with a son in our community, perhaps even in our social circle. But if we determine her “neighborhood” in the social web in enough detail, it is possible we could identify her uniquely in terms of her social roles and relationships.

This idea certainly has its appeal, but also its problems, one common and one unique. First, one's placement in the social structure can change over time, often abruptly, such as when changing jobs, getting married or divorced, falling into a new set of friends, and so on. Of course, the importance of this shortcoming may be lessened if we consider that our various social links rarely change at once, making possible a “family resemblance” of our social states over time (the same could be said about preferences themselves). But in the extreme, if these factors did change simultaneously—upon entering the witness relocation program, for instance, an extreme case that completely refines our social status—we would not think of ourselves, deep down, as different people. So social links seem to have the same instability problems as preferences, but nonetheless even as our positions in society change, we remain who we are.

Second, social links are not primary; they are based both on circumstances that are out of our control and choices that are in our control. Those choices in turn will be based on a combination of preferences and values, which the autonomous agent will first filter through her judgment, and will therefore be an expression of her character. A married doctor with two kids who belongs to a local bird-watching club arrived at that social status through a combination of luck, desire, and effort. If we want to identify this person, it is more accurate to use what is primary—her judgment and character underlying the effort that brought her desire to (p.95) fruition given her luck—than to use the consequences thereof. We can certainly use her social achievements (and failures) as proxies for her more essential attributes, but then we have to parse out the role played by good luck and circumstances (although perseverance in the face of bad luck or negative circumstances certainly demonstrates strength of character).

Please keep in mind that even if we end up rejecting the idea that our social relationships define us, we do not have to deny their influence on our decision-making, choices, and actions. As we saw above, Kantian autonomy does not imply this, any more than it forbids taking preferences into account—when appropriate. Just as a person should not lie to gain material advantage, she should not lie simply to please a friend. External and internal influences do and can influence our choices, but only to the extent that no duty is being violated by doing so. Furthermore, they must enter into our decision-making when doing so serves to fulfill a duty: if you mean to help a friend, you must know what sort of help she needs, and if the needs of two friends conflict, you are free to judge which you needs help more, or which friend is closer to you. As we saw in chapter 1, this represents the essential role that real-world, human information and context play in moral action, and speaks against one of the perpetual misunderstandings of Kant's ethics.

So, rather than identifying individuals with their preferences or their social roles and ties, I maintain that we should identify a person with her character. The understanding of character developed here can solve the problems of temporal continuity and individuation in a more constructive and intuitive way than preferences or social roles can. While one's memories, feelings, or preferences may be transitory, or may be shared among several persons, each agent's character—her judgment and will as reflected in her agency—is necessarily her own (as discussed above). Each agent has her own sense of the moral law as derived from her own reason and based on her own experiences, and as a result she will resolve moral dilemmas and conflicts of obligations differently from anyone else. Furthermore, her judgment may change with experience (as described in Chapter 1), and her will may strengthen and weaken over time (as described in Chapter 2), but these changes are usually neither sudden nor drastic, and they maintain enough consistency in character to identify her over time. And if a person's character were to change abruptly—for example, following severe (p.96) physical or psychological trauma—it is both common and appropriate to say the person is no longer the person who was before.

But if every person, individually and independently, realizes the moral law within her—the categorical imperative, as Kant described—in her judgment, would that not make all persons identical rather than individuated and distinct? Harry Frankfurt forcefully makes this objection, which is worth quoting at length:

The autonomous will can only be one that incorporates what Kant calls a “pure” will. It must conform, in other words, to the requirements of a will that is indifferent to all personal interests—that is entirely devoid of all empirical motives, preferences, and desires. Now this pure will is a very peculiar and unlikely place in which to locate an indispensable condition of individual autonomy. After all, its purity consists precisely in the fact that it is wholly untouched by any of the contingent personal features that make people distinctive and that characterize their specific identities…. The pure will has no individuality whatsoever. It is identical in everyone, and its volitions are everywhere exactly the same. In other words, the pure will is thoroughly impersonal.13

True, the basic contours of the moral law will be the same—the respect demanded for humanity, the general duties outlined by the categorical imperative, and so on—but the nuances and subtleties will be very different. While the same formal structure of the categorical imperative would (ideally) be realized by every rational person, each person will nonetheless implement it according to her own judgment based on her own experiences and perspective, and will execute those decisions to a degree based on her own will, combining to form her own character, which makes her unique.

We can draw an analogy to judicial decision-making, borrowing (as we did in Chapter 1) from the jurisprudence of Ronald Dworkin. Ideally, every judge in the United States follows the same Constitution, the same statutory law, and the same precedents, but nonetheless each judge may come to a different conclusion as to the “right answer” in any given case due to her own unique interpretation of the legal and political system.14 Over time, her decisions build up a judicial record, and observant legal scholars can certainly differentiate between one judge's jurisprudence and that of another. And while any one judge's jurisprudential approach may shift or evolve over time, such shifts are rarely abrupt enough to lead one to say “is this really the same judge?” And if it were, that question would certainly be justified; we would say (p.97) she is acting out of character, though the judge herself may disagree, based on our differing conceptions of her judicial character. Furthermore, her justification of her decision may offer new insight into her jurisprudence—or even her moral character, if her answer is regarded as insincere.

Now think of a person you know well—well enough for you to claim to know her character, how she thinks, and the decisions she is likely to make (and carry out). When such a person surprises you with a decision or action you did not expect, you may say “that's not like her” or “that's not the Maria we know.”15 Or, when confronting her, you may say “you've changed” or “it's like I don't know you at all,” or, most pointedly, “it's like you're a different person.” And this is not because her preferences, memories, or feelings have changed; all of that would be understandable, or, at the most, temporarily puzzling (“I didn't think you liked jazz”). Rather, it is because the essential elements of her character, her judgment and will—her very self—seem different.16 For these reasons I argue that it is a person's character, her judgment and her will, that identifies her, distinguishing her from other persons as well as consistently picking her out over time (regardless of gradual change).

On the surface, this idea is similar, though not directly based on, Kant's own conception of disposition or character, which he termed Gesinnung.17 As Henry Allison writes,

Kant makes it clear that he recognizes that the choices of rational agents … must be conceived in relation to an underlying set of intentions, beliefs, interests, and so on, which collectively constitute that agent's disposition or character. Otherwise these choices and maxims could be neither imputed nor explained; they would have to be regarded as completely arbitrary expressions of a “liberty of indifference,” without any “sufficient reason.” … In addition, by enabling us to regard a person's specific acts and decisions as expressions of an underlying set of intentions or pattern of willing, which can itself be the object of a moral evaluation, it provides a means for thinking about the moral life of a person as a whole.18

Though my conception of character is not specifically drawn from Gesinnung, a much more subtle and nuanced concept with intricate links to related ideas in Kant's moral psychology, they are certainly of the same spirit, both representing the myriad of background influences and experiences that inform one's choice and contributing to a process of self-constitution, a concept to which we now turn.

(p.98) Practical identity and self-constitution

When discussing identity, philosophers often distinguish between two senses of the concept: personal identity and practical identity. Personal identity, discussed above, is a matter of metaphysics: how is a person to be individuated from other persons, as well as identified over time as the same person? Practical identity, on the other hand, is a matter of action, morality, and responsibility, and arises from the common-sense realization that, regardless of metaphysical debates over the reality or illusion of the self, when a person acts she feels that it is she who is acting. Practical identity represents the standpoint, the idea of her self, from which she acts. As Christine Korsgaard, one of the foremost writers on personal identity, writes, “from the practical point of view our relationship to our actions and choices is essentially authorial: from it, we view them as our own…. We think of living our lives, and even of having our experiences, as something that we do.”19 This perspective demands that we see ourselves as actors in the world, inserting ourselves in the causal chain of events or initiating new ones—again, regardless of metaphysical debates over the nature of the self. Furthermore, we have to treat persons as actors in the world if we are going to ascribe moral status to their actions, as well as responsibility for them.

Korsgaard considers practical identity to be a different approach to discussing personal identity, and one that can answer both the individuation and continuity issues.20 For instance, debates regarding personal identity usually focus on mental states of the person, such as memories in the version common attributed to John Locke. But this framework regards the person as a passive receptacle of these feelings or experiences, rather than as an active agent in the world. For instance, in response to Derek Parfit's contention that personal identity and self-perceptions are “nothing more than connections and continuity between events in the life of a person,” philosopher Stephen Darwall argues that

persons have the capacity themselves to affect just what those continuities and connections are…. The capacity to choose our ends, and rationally to criticize and assess even many of our desires, means that our future intentions and desires do not simply befall us; rather, they are to some degree in our own hands. If this is true, there is a sense in which we cannot simply consist in connections and continuities, because we are ourselves capable of affecting these.21

(p.99) In response to Parfit's normative claim that the illusion of selfhood renders unimportant the distinction of persons, Darwall argues (consistent with the discussion above), “Though there is a sense in which a person may have changed so much that we may wonder whether, for certain purposes, we should regard him as the same person, this is not the sense which underlies the moral distinction of persons. What underlies that idea is that we are distinct choosers, distinct centers of rational decision making.”22 And as he writes elsewhere, “the rational person is not constituted by whatever ends or preferences he happens to have at any given moment. Rationality consists, at least partly, in our capacity to make our ends and preferences the object of our rational consideration and to revise them in accordance with reasons we find compelling”—reasons which, in the case of the Kantian-economic model, are embodied in the moral law.23

In Self-Constitution, Korsgaard builds on this conception of practical identity, arguing not only that agency informs the general concept of practical identity, but also, more specifically, that persons are self-constituted through their actions, which both reflect, and contribute to, who they are:

The task of self-constitution involves finding some roles and fulfilling them with integrity and dedication. It also involves integrating those roles into a single identity, into a coherent life. People are more or less successful at constituting their identities as unified agents, and a good action is one that does this well. It is one that both achieves and springs from the integrity of the person who performs it.24

In the Kantian-economic model, character works in much the same way, as the agent's judgment and will are both expressed and shaped by the choices she makes. Each decision she makes is a product of her judgment and will: the consequences of that decision (as well as her reflections upon it) inform her future judgment, and her strength or weakness of will in executing that decision affects her resolve in future decisions. In this way, the agent's character continually evolves as she moves through—and acts in—the world.25

Korsgaard nicely contrasts this self-constituting agent with static conceptions, which take the agent's essential properties to be fixed and immutable, and also posit a one-way, determinate relationship between her character and her actions.26 But she argues that the idea “to be a person is to be constantly engaged in making yourself into that person,” and (p.100) therefore your character, your identity, is never settled; it is constantly in a state of change, which is usually smooth and gradual, but is occasionally abrupt in response to life-changing events.27 As Joel Feinberg writes, “self-creation in the authentic person must be self-re-creation, rationally accommodating new experiences and old policies to make greater coherence and flexibility.”28 Importantly, Feinberg emphasizes the gradual, never-finished nature of self-determination, starting from the rudimentary character of the child, who cannot begin forming her character from scratch, but must instead be “implanted” with some principles from the start. Gerald Dworkin raises the same point regarding early agency: “to the extent that the self uses canons of reason, principles of induction, judgments of probability, etc., these also have either been acquired from others or … are innate. We can no more choose ab initio than we can jump out of our skins.”29 But as we approach maturity, we can and should reflect on those early influences, as much as possible, and ensure that we agree with them in light of our increasingly developed autonomous character (as described above in response to the Hodgson quote).

It is in this way that an agent is self-constituting: every person is a product of not only her experiences but also her choices, and each new choice either confirms who we are or changes it. Furthermore, the very process of deliberation, of considering all of the agent's preferences, social roles, responsibilities and obligations, and choosing an action, unifies her identity.30 Korsgaard contrasts what she calls the Combat Model of the soul, in which reason and passion are (often) contrasting forces battling over control of the agent's choices and actions, and the Constitution Model, in which the agent herself stands above—but not separate from—her reason, passions, and all of the other incentives that may influence her choices, and she decides which force will be her will.31 Drawing from Plato's description of the just state in the Republic, Korsgaard argues that an agent constitutes her self from the various aspects of her identity, and in the end “she identifies with her constitution,” which if drawn up well will unify her as an agent who can act efficaciously and autonomously in the world.32 As she notes, we commonly use language such as “pull yourself together” and “make up your mind” to describe conflicted choice situations, and in this case such idioms are rather accurate.33 In the final paragraph of Self-Constitution, she expresses this (p.101) idea nicely: “in the course of this process, of falling apart and pulling yourself back together, you create something new, you constitute something new: yourself.”34

But at the same time, this language should not be taken too literally. Consider the arguments that the self is not unified, but that instead persons are composed of multiple selves, either at one time or over time, which battle or bargain with each other (or with the central, active part of the person) to control choice and action.35 Phenomenologically speaking, though we may speak of various selves within our minds and fighting each other, each of us cannot help but feel that she is deliberating, choosing, and acting as one person. We are certainly not shunting off those tasks on someone else; if we are, then unity of agency is hardly our most serious problem. Some may claim that even though choice is eventually issued by one overarching self, it is the result of the struggle between various lesser selves. But while we may struggle with competing urges, drives, preferences, ends, goals, dreams, duties, and obligations, there seems no reason to name these various things “selves,” for to do so is to grant those selves some of the status we ascribe to persons, such as dignity, autonomy, and agency, and therefore normative status that would demand respect. But is it realistic (if even useful) to suppose that these various selves have agency? I think Jon Elster sums it up best when he writes that

barring pathological cases … we ought not to take the notion of “several selves” very literally. In general, we are dealing with exactly one person—neither more nor less. That person may have some cognitive coordination problems, and some motivational conflicts, but it is his job to sort them out. They do not sort themselves out in an inner arena where several homunculi struggle to get the upper hand.36


To be sure, social roles, links, and responsibilities also enter into this deliberative self-constituting process, and as with other experiences and choices, the agent is not a passive subject of her social identities.37 As Korsgaard writes,

you are a human being, a woman or a man, an adherent of a certain religion, a member of an ethnic group, a member of a certain profession, someone's lover or friend, and so on. And all of these identities give rise to reasons and obligations.

(p.102) Your reasons express your identity, your nature; your obligations spring from what that identity forbids.38

But before these identities can become a part of an agent's practical identity, her sense of self (or character) from which she acts, she must take an active role in endorsing these roles by choosing what groups to join, what people to associate with, and what social responsibilities to assume. Even the aspects of your social identity you are born into—being a child of your parents, a member of your community, a citizen of your nation—must be endorsed by you before they become part of you and reasons on which you can act autonomously. However the social identities come about, they “remain contingent in this sense: whether you treat them as a source of reasons and obligations is up to you. If you continue to endorse the reasons the identity presents to you, and observe the obligations it imposes on you, then it's you.”39 So like preferences, social identities, along with their constituent roles and responsibilities, are subject to the endorsement of an agent's judgment based on the moral law; as important as those features are to the agent's life, they are nonetheless secondary to her character.

The necessity of endorsement implies that the agent is reflective, in particular regarding her incentives for action; as Korsgaard writes, self-consciousness “transforms incentives into what Kant calls inclinations,” which can be motivating in a way that mere incentives cannot.40 The agent uses her faculty of judgment to assess her desires and then transform desiring into having a desire, thereby taking possession of it and deciding whether to indulge it; or as Korsgaard paraphrases Plato, “having an appetite for something and giving that appetite the nod are not the same thing.”41 Amartya Sen has also discussed the reflective nature of the person, which he calls self-scrutiny: “A person is not only an entity that can enjoy one's own consumption, experience and appreciate one's welfare, and have one's goals, but also an entity that can examine one's values and objectives and choose in the light of those values and objectives.”42 For the perfectly autonomous agent, this process of reflection will be decisive; her judgment will be sound and her will unwavering. The imperfect agent, however, will fail occasionally in this reflective process, either judging incentives or preferences incorrectly (relative to the moral law), or judging well but lacking the willpower to follow through on its recommendations.43

This also relates to John Davis's argument that individuals may be (p.103) socially embedded, or defined by their place in the social web, but still be individual by virtue of being self-reflective: “by describing how individuals actively form self-conceptions—precisely because of social influences operating upon them—it succeeds in introducing agency into the conception of the individual as socially embedded in a nonarbitrary manner.”44 Indeed, in later work Davis has referenced Korsgaard's concept of the “reflective structure of consciousness” (though disavowing the explicitly Kantian aspects of her analysis), linking it to earlier discussions of metapreferences.45 He acknowledges that social factors are open to endorsement, because while “social factors influence how individuals form self-concepts,” reflexivity “implies that individuals can detach themselves in some degree from the determining effects of social factors influencing them,”46 “rather than simply serving as passive repositories of those influences.”47

Perhaps the only point of disagreement here regards to what degree they can do this, because the ability for “complete” endorsement is implied by autonomy, reaffirming the atomistic choice faculty of the individual. Indeed, Davis writes that “individuals may also fail actively to form self-conceptions (or have their self-conceptions determined for them by society.”48 Philosopher Marina Oshana, too, is doubtful:

Who persons are, how they define themselves, and the content of their motivations, values, and commitments are essentially fashioned by connections to other people, to cultural norms, rituals, tradition, and enterprises. We cannot refigure these phenomena at will. Indeed, given their enormous centrality to our lives, they are phenomena that might even elude our scrutiny, our attempts to direct a critical lens upon them and render them self-made.49

But if the Kantian agent failed in either of these ways, it would be her choice to do so, and in this way it still affirms the primacy of the individual's character. As Herman explains, a person forms herself to some extent by choosing how—and for whose sake—to fulfill positive duties such as beneficence: “part of what I do in satisfying imperfect duties is shape the relationships that make claims on me, and in so doing, shape myself.”50 True, the Kantian standard sets a high bar, but if autonomy (in the form of authenticity and self-realization) is to be understood as a normative goal (as suggested in Chapter 1), then realistic skepticism may be warranted, but defeatism is not.

This criticism also extends to Davis's endorsement of collective (p.104) agency as a possible conception of social embeddedness.51 Collective or plural agency is a philosophical framework for understanding the coordinated behavior of more than one agent. Philosopher Margaret Gilbert gives a simple example of two persons taking a walk: each person is walking, of course, but more importantly, they are walking together, as opposed to two strangers who just happen to be walking side-by-side.52 Views on plural agency fall into three camps: the most extreme (and least held) maintains the existence of a “group mind” that issues intentions that are by their nature collective.53 Others, in the “holistic” camp, deny the existence of a group mind, but nonetheless maintain that collective intentions are not reducible to the individual intentions of the participants in the shared action.54 Yet others, sometimes referred to as “individualistic or atomistic,” maintain that all intentions are individual, even if they concern collective action and are shared among others in the group; Raimo Tuomela, from whose account on plural agency John Davis draws, is among this group.55 While affirming that “only individuals form intentions,” Davis argues that “alongside those intentions expressed from a first-person singular point of view, individuals also express shared or collective intentions from a first-person plural point of view”; in other words, “I-intentions” co-exist with “we-intentions.”56 But Davis affirms that “a we-intention is an individual attribution of an intention to the members of a group to which the individual belongs, based on that individual both having that we-intention and also believing that it is held by other individuals in the same group.”57 Besides Gilbert's stroll, other common examples that illustrate this point are the players on the sport team working together towards victory, or the members of an orchestra playing a symphony, neither of which make sense if each individual in the group does not believe that the other members share the same intention.58

If we accept Kantian autonomy and the atomistic process of choice it implies, it is clear we cannot accept any sort of “group mind” hypothesis where choice is made at a supraindividual or collective level. As we have argued, autonomy implies that one's choices are her own, her unique contribution to the causal chain of events in the world; even if she lets another person unduly influence her actions, she alone chooses to let him do this. But as long as choice remains hers, there is no reason to deny that, in agreement with Tuomela and Davis, agents can self-consciously act in (p.105) concert with other agents in groups or collectives. In Korsgaard's terms, the agent would have to endorse the shared intention and thereby make it a reason or an obligation for her. That collective intentions create obligations for the constituent individuals is a key theme for Tuomela, and another point of agreement with the conception I present herein. I can take a walk together with a friend, but despite the contention of the holists, I am choosing to take a walk with her, with the understanding that she has chosen to take a walk with me, and that both she and I understand each other's intentions (similar to the assumption of common knowledge in game theory). In that sense and that sense alone we together are taking a walk, but a Kantian view denies that there is a distinct and separate “we” making this choice and acting according to it. Tuomela prefers to refer to this coordination of intentions as reciprocal rather than shared, which supports my argument (reinforced below) that even individuals with atomistic choice processes can be socially oriented in various ways, such as according to reciprocity (which can be considered, after all, the normative force behind the universalization formula of the categorical imperative).

To summarize, the Kantian-economic approach maintains that agents are essentially individual, but at the same time they can be—and, ethically speaking, must be—social in orientation. I hope to have reassured those with concerns about sociality or community that individualism need not be threatening. In fact, if we treat persons as individuals imbued with autonomy and dignity, social harmony takes on much more meaning because it will be the result of free, individual choices, rather than coercively enforced order.

Social in Orientation

Having argued that the Kantian agent is essentially individualistic by virtue of her autonomy, we now turn to her social orientation. As I said in the first part of this chapter, her choices are hers and hers alone, and (ideally) are made in accordance with the moral law to the exclusion of undue internal and external influences. But despite this metaphysical and practical agency-based isolation, the moral law itself demands that she take social factors into account—especially the needs and desires of others. (p.106) Autonomy is not simply the inner freedom that comes with the capacity for free choice; it is just as importantly a responsibility towards others, and it is this aspect of autonomy that lends Kantian ethics, and the Kantian individual, a fundamental social nature.

To be sure, Kant is not the first name people associate with sociality; as we have seen, his ethics are commonly regarded as cold, unfeeling, and better suited for transactions between strangers than friends or family, so any resulting sociality would likely be regarded by most as very “arms-length.” As Louden writes (and then refutes),

Kant is often portrayed as an extreme moral individualist, one who holds that each moral agent is an end in itself, a discrete individual owed respect for its autonomy, an autonomy that is safeguarded by inviolable rights. Such an individualism, it is alleged, views person as atomistic, and cannot readily accommodate larger social units such as the family which transcend mere atomism.59

As with many misconceptions, this one nonetheless contains a kernel of truth. In fact, Kant spoke at length about friendship, family, and love; if he did not emphasize such close relationships when discussing moral duty, it was because, as with the pursuit of self-interest, we do not normally regard imperatives as necessary to generate kindness and affection among friends and family.

But the impression that Kant's is an ethics for strangers, while meant as a derisive indictment, is instead an insightful complement, for it implies that, to a significant degree, we owe the same kind of consideration, respect, and “love” to strangers as we naturally do towards our family and friends. As Sullivan writes:

Earlier Western philosophers had thought of morality as originating within the personal and private relationships of the family and extending outward from there to the public order. Kant, by contrast, situates morality primarily within human public life, which he defines in a formal and impersonal way.60

But nonetheless, we are perfectly free to pay more attention (and devote more resources) to those closer to us (in denial of yet another common misconception concerning Kant; more on this below).

In this part of this chapter, I argue that Kantian ethics, especially as embodied in the Kantian conception of the economic individual, is not only consistent with sociality of the highest degree, but it actively supports (p.107) this sociality, and that it does so through the essential concepts of dignity and respect. And what's more, I think it does this in the context of the minimal government of Kant's political philosophy, so that the individual can flourish and prosper with minimal interference (and support) from the state, due to a mutual respect and support from her fellow persons. (This will lead us into the policy discussion of the last two chapters of the book.)

Perfect and imperfect duties again

Let us start from what we know (from Chapter 1): perfect and imperfect duties. Recall that perfect duties are usually negative and strict, allowing no latitude in execution (unless the duty is overruled completely by another), while imperfect duties are usually positive and wide, demanding no particular action but rather the sincere adoption of ends (and action in accordance with them when feasible). For the purposes of this chapter, I think we can safely overgeneralize and say that perfect duties are duties of noninterference (do not steal, do not injure), while imperfect duties are duties of beneficence (do help others).

Put this way, there is obvious similarity with Isaiah Berlin's distinction between negative liberty, which involves rights to noninterference, and positive liberty, which involves rights to assistance.61 While perfect duties certainly imply rights to noninterference, imperfect duties do not imply any correlative rights, because they do not demand any particular action that could support such rights, and therefore do not generate a right to assistance in the sense of an enforceable claim. As Kant wrote, “no one is wronged if duties of love are neglected; but a failure in the duty of respect infringes upon one's lawful claim.”62 However, in a society of Kantian individuals, we will see that a person who is in need may expect assistance, if we understand expectation in the positive sense of prediction, rather than the normative sense of a demand or claim. But she can demand noninterference, especially if the perfect right that guarantees it is also a juridical right (one enforceable by the state).

Of course, this is a familiar idea from classical liberalism, and certainly noninterference is a minimal conception of sociality. After all, a hermit practices noninterference, but would hardly be considered a paragon (p.108) of sociality; he does no harm, which is good, but neither does he help anyone, which is not. Nonetheless, we should not overlook the importance of noninterference; without a strong sense of, and respect for, personal boundaries, persons will be too concerned with protecting themselves, their loved ones, and their property to have much time or energy to devote to what we consider the more social virtues. (This is the same sense in which a wealthy country can better afford to invest in environmental safeguards; if a nation's people are starving and disease-ridden, they will not have much will to evaluate and adjust their environmental impact.) Only when our personal boundaries are secure against hostile transgressors do we feel comfortable enough to relax them, especially to strangers. So we can say that noninterference—or perfect duties—are a necessary though clearly insufficient condition for a flourishing society.

Recall how perfect duties are usually derived: either from the consistency-in-conception test generated by the Formula of Autonomy or from the negative part of the Formula of Respect. That perfect duties demand respect is obvious; the prohibitions on murder, theft, assault, and lying all result from not treating humanity merely as means while not at the same time as an end—in other words, respect for the autonomy and dignity of fellow persons. But it does not demand anything more than respect; the hermit, after all, is nothing if not respectful. That perfect duties result from the consistency-in-conception test—an eminently logical and cold moral standard—is also very telling, for a person following only perfect duty is satisfying the minimal social requirements. A society built on that foundation may persist, but we would not expect it to prosper and flourish; more on this point below when we discuss the Kingdom of Ends (and a philosopher named Adam Smith).63

As with perfect duty, there are two standard sources of imperfect duties, both of which take a more humanistic point of view. The more direct derivation is from the positive half of the Formula of Respect—treat people always as ends—which impels us to take the well-being and happiness of our fellow persons into account as we travel through life. But the conception-in-the-will test also generates imperfect duties, by supplementing the logical test of internal consistency with the doctrine of ends-in-themselves—specifically, other persons (and the agent herself). Remember that Kant himself regarded the true domain of ethics to be (p.109) imperfect duty; any minimally decent person knows not to steal or assault others, but he may need help remembering to keep others' well-being in mind.64 Both of these formulae emphasize that acknowledging the dignity of persons requires not just that they be respected in a negative fashion, but that they be considered ends-in-themselves and treated in a positive fashion as well, included as an end among other ends in personal decision-making (as in the model developed in Chapter 1). Kant explicitly linked beneficence with a broader sense of sociality: “the maxim of common interest, of beneficence toward those in need, is a universal duty of human beings, just because they are to be considered fellowmen, that is, rational beings with needs, united by nature in one dwelling place so that they can help one another.”65

Kant was also rather poetic on the necessity of both of these attitudes, respect and love: “by analogy with the physical world, attraction and repulsion bind together rational beings (on earth). The principle of mutual love admonishes them constantly to come closer to one another; that of the respect they owe one another, to keep themselves at a distance from one another.”66 Later in the same passage, Kant emphasizes the opposite nature of the two duties: “a duty of free respect toward others is, strictly speaking, only a negative one (of not exalting oneself above others) and is thus analogous to the duty of right not to encroach upon what belongs to anyone…. The duty of love for one's neighbor can, accordingly, also be expressed as the duty to make others' ends my own (provided only that these are not immoral).”67 As we will see in the next section, Kant also held both attitudes to be essential to achieving the kingdom of ends.

When discussing Kantian beneficence, there is a danger of letting utilitarianism in through the back door. After all, we are saying that each agent has a moral responsibility to take into account the well-being of other persons based on equality of dignity, and that is precisely how some would justify utilitarianism. But there are several crucial differences, the most important being that in the Kantian conception, no particular beneficent or helpful act is required—and certainly not the utility-maximizing act.68 As we saw in the discussion of the prisoners' dilemma in Chapter 1, Kantian ethics is not perfectionist, being much more specific about what not to do than what to do. Accordingly, taking into account the well-being of others does not imply extreme self-sacrifice (as exemplified by Peter Singer), nor (p.110) does it require calculative wizardry (as mocked by Thorstein Veblen, as we saw in the last chapter).69 This aspect of Kantianism is much more in the spirit of virtue ethics, which (at least in its Aristotelian guise) would counsel practicing kindness, benevolence, and charity in moderation, not through extreme self-sacrifice (nor through any particular act).

In addition, the Kantian idea of beneficence is much more nuanced than the standard utilitarian version, if only because respect for dignity remains paramount. Kant was very emphatic about the spirit in which help was to be given:

We shall acknowledge that we are under obligation to help someone poor; but since the favor we do implies that his well-being depends on our generosity, and this humbles him, it is our duty to behave as if our help is either merely what is due him or but a slight service of love, and to spare him humiliation and maintain his respect for himself.70

This is a natural impulse; it is common, in response to expressions of gratitude, to say “no problem,” “it was nothing,” or “anyone would have done the same.” These express not only modesty on the benefactor's part, but are gestures that help preserve the self-respect of the recipient. What's more, Kant is clear that when one practices beneficence, it must always be geared towards the other person's own view of his well-being, not the benefactor's external judgment of it: “I cannot do good to anyone in accordance with my concepts of happiness (except to young children and the insane), thinking to benefit him by forcing a gift upon him; rather, I can benefit him only in accordance with his concepts of happiness.”71

So Kantian beneficence is less specifically demanding, and more subtle regarding how it is practiced, than utilitarianism, but does it go too far in the other direction? Is this understanding of beneficence too permissive? Only if you need a moral philosophy to tell you exactly what to do, when to do it, how to do it, and to what extent to do it.72 Setting aside the wonderful irony of contemplating the possibility that Kantian ethics may be too permissive, this perception reflects the often neglected but very important role that judgment plays in actual real-world decision-making. Recall that the categorical imperative is purely formal, helping us derive general guidelines for moral action, as well as reminding us why they are important (the equal dignity of all persons). Imperfect duties, if they are to be followed wholeheartedly—as they must—cannot be merely (p.111) paid lip service, for the attitudes they prescribe must be sincere. The duty to keep others' well-being in mind is not something to be taken lightly; even if one cannot provide aid or assistance to others at any particular time, one must always be mindful of the circumstances her fellow human beings are in.73 That, I think, is more than enough moral burden for most, who may prefer to write a check and feel comfortable putting the plight of others out of their minds for a while. Understood this way, the duty of beneficence, and the attention and care it demands we pay to each other, ties all persons—family, friends, neighbors, and strangers alike—together as one people, inextricably linked by the moral law in what Kant termed the kingdom of ends.

The kingdom of ends (and a man named Smith)

In Chapter 1, we mentioned the broad, teleological nature of Kant's moral theory as reflected in the third formula of the categorical imperative, the Formula of Legislation for a Moral Community (also known as the Formula of the Kingdom of Ends): “every rational being must so act as if he were through his maxim always a legislating member in the universal kingdom of ends.”74 While this formula is not meant to guide moral decision-making or the formation of maxims as such, the kingdom of ends does represent the final goal of moral endeavor, a utopian state of the world in which all persons can pursue their own ends in cooperation with each other:

For all rational beings stand under the law that each of them should treat himself and all others never merely as means but always at the same time as an end in himself. Hereby arises a systematic union of rational beings through common objective laws, i.e., a kingdom that may be called a kingdom of ends (certainly only an ideal), inasmuch as these laws have in view the very relation of such beings to one another as ends and means.75

It is worth emphasizing that Kant regarded the kingdom of ends as “certainly only an ideal,” an analogy, of course, to the Kingdom of God envisioned by Christians. But as with all ideals, it points us in the direction of progress, improving society through mutual respect and beneficence.76

In detailing the kingdom of ends, he also shows how the basic concept of duty is based on relationships between persons of equal dignity (as (p.112) well as how the final formula of the categorical imperative incorporates the first two):

The practical necessity of acting according to this principle, i.e., duty, does not rest at all on feelings, impulses, and inclinations, but only on the relation of rational beings to one another, a relation in which the will of a rational being must always be regarded at the same time as legislative, because otherwise he could not be thought of as an end in himself. Reason, therefore, relates every maxim of the will as legislating universal laws to every other will and also to every action toward oneself; it does so not on account of any other practical motive or future advantage but rather from the idea of the dignity of a rational being who obeys no law except what he at the same time enacts himself.77

This relates back to the sociality inherent in the Formula of Autonomy; the reason we must test our maxims for universalizability is based on equal respect for, and reciprocity toward, all rational (human) beings, which also corresponds to the Formula of Respect in both its negative and positive aspects.78

Just as he was realistic about human weakness and frailties (Chapter 2), Kant was just as realistic about persons' “unsocial sociability, i.e., their tendency to enter into society, combined, however, with a thoroughgoing resistance that constantly threatens to sunder this society.”79 While we have a natural propensity to live in society with others, he argues, we also have a desire to have things our own way, which we recognize others may resist (as we resist the same in others). But this self-centered aspect, while “unworthy of being loved,” also fuels a competitive, ambitious drive in us, without which “all of humanity's excellent natural capacities would have lain eternally dormant.”80 Despite this practical advantage, our unstable coexistence is also threatened by the “ethical state of nature … in which the good principle, which resides in each human being, is incessantly attacked by the evil which is found in him and in every other as well.”81 In a Hobbesian spirit, humankind's “unsocial sociability” necessitates both a public law, arising from the state, and a moral law, arising from the wills of individual rational agents, both of which lead to the ideal of the kingdom of ends.

Corresponding to these two needs, achieving the kingdom of ends involves two stages, a moral civil community and an ethical community.82 In Kant's terms, an ideal civil community is achieved when people follow (p.113) their perfect duties, chiefly the juridical duties enforceable by the state, such as the duties (codified in laws) prohibiting murder, assault, and theft. At this stage, persons do not follow these duties for the sake of duty, as is the Kantian ideal, but rather follow them merely out of self-interest, that is, to avoid punishment.83 However, an ethical society is characterized by citizens following both their perfect and imperfect duties, as well as doing so for the sake of duty, for imperfect duties cannot be enforced at all, and many perfect duties are impractical for the state to enforce (such as the duty not to lie, especially in noncommercial contexts).

While the moral civil community is a crucial step towards the kingdom of ends, it is merely an intermediary step; as we mentioned above, a world in which persons only fulfill their perfect duties towards each other may operate on some minimal level, but it would hardly be a world in which persons flourished and maximally furthered their ends. Recall that, according to Kant's consistency-in-conception test, a world in which no persons observed their imperfect duties (such as beneficence) could exist without internal contradiction, but it could not be willed rationally by any person because it would not be consistent with the recognition of persons as ends-in-themselves. Indeed, Kant referred to persons inhabiting such a society, however orderly it appears to be, as existing in an “ethical state of nature,” in which they obey laws out of fear of punishment, not out of respect for the law (or the duties underlying it).84 Only a truly ethical community—a world in which people observed both perfect and imperfect duties, treating each other (positively) as ends to be furthered and not simply avoiding (negatively) using others as means, both out of respect for the moral law—will allow all persons to pursue their ends in cooperation with each other.85

Perhaps a parallel to the thought of Adam Smith would give some perspective on the kingdom of ends and help show its applicability to sociality. A world in which agents follow only Kant's perfect (or negative) duties, such as duties not to harm others, would be much like the impersonal marketplace described in Smith's Wealth of Nations. This minimally ethical environment would certainly serve its purpose in facilitating trade, but certainly not as a model for a complete society in which people can not only survive but also prosper and flourish. Such a world needs more, which in Kantian terms would be a world in which agents also followed (p.114) imperfect duties, his kingdom of ends; it would also resemble the world of Smith's Theory of Moral Sentiments, one in which persons exercise their capacity for sympathy, or fellow-feeling, resulting in benevolent sentiments.

In the twenty-first century, of course, Smith is mainly regarded as the father of modern economics (even if most modern economists treat him like the crazy uncle who lives in the attic), while Kant had very little of significance to say regarding markets or commerce. But it should not be a surprise that, as eighteenth-century moral philosophers, Adam Smith and Immanuel Kant were strongly linked, intellectually and historically.86 Kant was exposed to both Wealth of Nations and Theory of Moral Sentiments, and evidence of their influence (especially the latter) can be seen throughout his work. Both scholars strongly emphasize impartiality as a core element of their moral systems, they were both strongly influenced by Stoic thought, and they shared a concern for human dignity and freedom from tyranny. However, while their substantive ethical thought was very similar, they differed in their positions on the basis of morality; Kant disagreed with Smith's (and Hume's) sentimentalism, preferring to ground his moral system in the respect for dignity and autonomy that issues from reason alone, regardless of feeling or inclination (see Chapter 1). More specifically, Kant held that beneficence should come from respect for the moral law rather than natural inclination, while Smith wrote that it arises out of sympathy, the capacity to imagine oneself in the circumstances of another. To Smith, it is this sympathy of persons towards each other that generates sentiments of benevolence: “it is, that to feel much for others and little for ourselves, that to restrain our selfish, and to indulge our benevolent affections, constitutes the perfection of human nature; and can alone produce among mankind that harmony of sentiments and passions in which consists their whole grace and propriety.”87 Setting aside details of moral psychology, however, both Smith and Kant recognized the need for other-regarding motivations and actions to create a complete social order, representing the best we can do, as opposed to the least we can live with.

With regard to Smith, there has been an enormous amount of work on the (apparent) contrast between his emphasis on benevolence (arising from the capacity for sympathy) in Theory of Moral Sentiments and the emphasis on self-interest or self-love in Wealth of Nations. It seems that the issue of consistency, which comprised the original “Adam Smith problem,” (p.115) has been resolved in the positive, and what remains is the “‘new’ Adam Smith problem concerning the precise nature of their relationship.”88 As I read him, in focusing on self-interest in Wealth of Nations, such as in the famous butcher-brewer-baker passage, Smith was outlining the minimal requirements for the operation of markets only—in particular, impersonal market exchanges amongst strangers or mere acquaintances, rather than close friends or family relations. Rather than recommending that market participants be motivated solely by self-interest, or endorsing such attitudes, Smith was arguing that even if they are so motivated, markets can operate smoothly; Samuel Fleischacker writes

that human beings can pursue even their individual interests together, that even society without benevolence need not be a hostile society, that economic exchange, even among entirely self-interested people, is not a zero-sum game. The emphasis is on the “even” in each case.89

Smith was not making a moral or prudential argument for self-love or egoism, nor was he arguing that self-interest was sufficient for a flourishing society outside of the market realm. He was merely making a case that a market can operate based on the participants' pursuit of their own self-interest, not that it should operate on such a basis.90

But at the same time, he did recognize that the majority of economic transactions in a developed commercial society would be between persons with little personal connection, for whom each need have no special concern.91 Furthermore, he did say that often, benevolent actions will interfere with the proper operation of markets, but this is one extreme, the opposite of pure self-interest, and does not argue against motivations marginally deviating from self-interest. He was also making what we think of now as a Hayekian efficiency-of-information argument, that each person knows his or her own interest better than anyone else does, and is therefore better placed to pursue that interest, as opposed to the more distant—and therefore less informed—actions of policymakers.92

In Theory of Moral Sentiments, on the other hand, Smith was describing (and prescribing) appropriate conduct in a broader context, social interaction in general, outside the narrow confines of anonymous market exchange. The capacity for sympathy or fellow-feeling becomes essential to generate benevolence (and beneficence) towards others, which Smith deems necessary for a truly flourishing society. This is not to deny (p.116) the obvious importance of some degree of self-love, of course, but rather tempers it with the sentiments arising from persons' sympathy for others' circumstances, especially those close to us, to whom general benevolence becomes more specific.93 In market circumstances, sympathy does not play as significant a role, not only because of the relative lack of personal connection between buyer and seller, but also because participants can rest assured that all involved are tending to their own affairs to the best of their abilities, and do not need assistance or aid unless they ask for it (recalling the passage quoted from Kant above regarding respectful beneficence). But in more general social contexts, seeing someone in need naturally generates, through our capacity for sympathy, sentiments of benevolence, which manifest themselves in good deeds.

Smith's description of self-interest as the minimal precondition for the operation of the market resembles Kant's limited endorsement of a civil society based on state enforcement of perfect duty alone. In such a world, persons would respect each other merely in a negative fashion: they would not cheat, would not steal, would not harm each other, and so forth. This behavior, of course, is what we expect—at the minimum—from market participants, as they pursue their own self-interest within the broad constraints of justice: as Smith writes, “every man, as long as he does not violate the laws of justice, is left perfectly free to pursue his own interest his own way.”94 In the terms of the Formula of Respect, in such situations, persons do use each other as means to their own ends, but not merely so, for they also treat others as ends-in-themselves by following basic rules or duties of respect. I do use the baker to get my bread and feed my family, but I do not steal the bread from him, nor do I cheat him out of it. Voluntary transactions, absent coercion or deceit, satisfy the negative standard of treating people as means while also as ends (as will be emphasized in Chapter 5).

But this neglects the positive aspect of the Formula of Respect, which generates imperfect duties through the requirement to treat others as ends (not just abstain from using them merely as means). When considering a world of only negative duties, Kant was clear (and somewhat moving):

Would it not be better for the well-being of the world generally if human morality were limited to duties of right, fulfilled with the utmost conscientiousness, and benevolence were considered morally indifferent? It is not so easy to see what effect (p.117) this would have on human happiness. But at least a great moral adornment, benevolence would then be missing from the world. This is, accordingly, required by itself, in order to present the world as a beautiful moral whole in its full perfection, even if no account is taken of advantages (of happiness).95

Likewise, as Jonathan Wight notes, “Smith would be appalled by a world that holds wealth above human connections, a world of markets unsupported by a social undergarment of moral fabric.”96 Both Smith and Kant envisioned a society where love for oneself and for one's neighbors coexist, supported to some extent by positive law, but primarily by morality (whatever its ultimate nature).

Generally, Smith and Kant would agree that the impersonal arena of the market is not an appropriate framework or mechanism for ensuring a flourishing society as a whole. Nonetheless, it is appropriate for organizing economic activity, especially transactions between anonymous strangers. As we saw above, the market does not represent the ideal kingdom of ends, which needs agents to observe both perfect and imperfect duties and take each others' ends as their own. But we are not required to take literally everybody's ends into account (as in utilitarianism); this requirement is flexible in terms of whom to favor or not, and how much. Even though I may not care at all for the well-being of my local baker, even if I think him a scoundrel and a cad, I must still treat him with the respect he is due as an autonomous person, and this is the essence of morality within the market: mutual respect if not mutual beneficence. Indeed, beneficence between strangers may even be inappropriate, given each person's lack of knowledge about the other, but what strangers can (and should) give each other is respect. So the market represents a kingdom of ends which is limited but nonetheless complete within its scope—ensuring the maximal freedom from interference consistent with the same freedom for all. It is the institutionalization of choice within the confines of justice, which is necessary in the context of interaction between strangers, where it may not be between friends.97 (We will have more to say about the market and choice in the next two chapters.)

Relations and reciprocity

As we just discussed, the market represents a minimal level of ethical behavior, in which perfect duties are followed but imperfect duties little (p.118) apply. This framework may be acceptable among strangers, but is definitely not appropriate when dealing with family or friends, or even neighbors or co-workers. Beneficence for strangers is simply a more moderate version of beneficence (or care) for friends and family. Since beneficence should always be respectful of the other person's dignity and autonomy, and we know much less about strangers and casual acquaintances than we do about close friends and family, the degree to which we can and should be beneficent, as well as the kind of kindness we can practice, is naturally more limited in such cases. In addition, an excessive degree or type of beneficence between strangers would generate a proportionate amount of gratitude in response, and Kant felt that excessive gratitude was demeaning (as we described above).

As we discussed, Kant's ethics are often criticized as being an “ethics between strangers,” which would seem to imply that he dismissed friendship, which is patently untrue; indeed, Allen Wood argues that “one would have to go back to Aristotle to find a major philosopher for whom friendship is as important to ethics as it is for Kant.”98 However, Kant certainly endorsed the biblical injunction to “love thy neighbor,” although he understood this “love” as a general benevolence, a well-wishing that could be extended to everyone: “when I say that I take an interest in this human being's well-being only out of my love for all human beings, the interest I take is as slight as an interest can be. I am only not indifferent with regard to him.”99 But then he considers that “one human being is closer to me than another, and in benevolence I am closest to myself. How does this fit in with the precept ‘love your neighbor (your fellowman) as yourself’?”100 The answer lies in the difference between mere benevolence, or well-wishing, and beneficence, or such feelings put into practice or action, “for in wishing I can be equally benevolent to everyone, whereas in acting I can, without violating the universality of my maxim, vary the degree greatly in accordance with the different objects of my love (one of whom concerns me more closely than another).”101 Recall that the duty of beneficence is literally the duty not to be indifferent to others, which leaves a great deal of latitude to give more consideration and aid to those closest to you (including yourself), as long as you keep distant strangers in mind as well.

(p.119) Korsgaard casts this topic in terms of reciprocity and responsibility, stressing that

to hold someone responsible is to regard her as a person—that is to say, as a free and equal person, capable of acting both rationally and morally. It is therefore to regard her as someone with whom you can enter the kind of relation that is possible only among free and rational people: a relation of reciprocity.102

In this understanding, reciprocity, and the responsibility implied thereby, is an expression of each person's recognition of equal dignity in others. Because of this dignity, we respect other persons, both in terms of living up to our commitments to them, as well as expecting them to live up to their commitments to us.103 To do any less would be to treat them as less than rational, or as children. But reciprocity of respect in a negative sense is not enough (outside the market arena), for respecting the dignity of other persons also implies treating them as ends-in-themselves, and taking their ends as our own (to some extent). As Korsgaard continues, “to join with others as citizens in the Kingdom of Ends is to extend to our inner attitudes and personal choices the kind of reciprocity that characterizes our outer actions in the political state.”104 And finally, relations between strangers and between friends must both embrace reciprocity, but to a different degree; again, our duties of beneficence need not be exercised to the same extent with strangers as with family and friends, though all must embody a certain level of mutual respect due to each person's dignity. But, of course, reciprocity between friends goes much deeper, involving the sharing of intimate thoughts, feelings, and concern that is not necessary, not even appropriate, between strangers.105

The reciprocity that figures so strongly in Kantian ethics is a conscious attitude that each person rationally and morally takes toward others. But in recent years, many social scientists have written more naturalistically on the origins of reciprocal behavior, as well as its importance in maintaining social order. The simplest versions of this go back to Hume, who based reciprocal action on long-term self-interest:

I learn to do a service to another, without bearing him any real kindness; because I foresee, that he will return my service, in expectation of another of the same kind, and in order to maintain the same correspondence of good offices with me or with others. And accordingly, after I have serv'd him, and he is in possession of the advantage (p.120) arising from my action, he is induc'd to perform his part, as foreseeing the consequences of his refusal.106

But modern evolutionary thinkers, including Herbert Gintis, write of strong reciprocity, in which a person is “predisposed to cooperate with others and punish non-cooperators, even when this behavior cannot be justified in terms of self-interest, extended kinship, or reciprocal altruism.”107 Gintis argues convincingly that this disposition evolved to promote group survival, and provides ample experimental and anecdotal evidence for its continued existence and influence over decision-making.

There is no reason to suppose, however, that this evolved reciprocity is incompatible with the Kantian version (recognizing that they are not identical in origin or implication). As I indicated above, Kantian reciprocity is a conscious attitude, while Gintis's version is an unconscious disposition. Kant values such inborn drives or feelings insofar as they help the agent adhere to dutiful action in the face of weakness, but would not regard action performed on their basis to be truly moral without a conscious realization of the rationale for doing so. This is not Gintis's concern, though; his goal is to explain the “high level of sociality [among human groups] despite a low level of relatedness among group members.”108 Nonetheless, agents who are naturally inclined to behave reciprocally in Gintis's sense may (or, for the sake of autonomy, should) eventually reflect upon this disposition and justify it normatively to themselves. (Or perhaps some of them will dismiss it as foolish sentimentalism, in which case they will long battle their basic drives, the difficulty of which those who have tried to lose weight or stop procrastinating know all too well.) Amartya Sen reconciles the two sources of behavior well, after lauding the advances made in evolutionary explanations of norms and values:

But, once evolutionary survival is taken into account, must the burden of selection fall entirely on that process (with conscious selection reduced to simple endorsement of natural selection)? Why can't the two means of selection be both actively at work? Since human beings are reflective creatures who take their values and critical powers seriously, the role of conscious and scrutinized selection will not be obliterated merely because evolutionary selection is also going on. Critical reflection does not give immunity from evolutionary selection, but nor does evolutionary selection convert reflective beings into thoughtless automatons.109

So while evolved reciprocity can explain some degree, perhaps a large degree, (p.121) of observed sociality, it stops short of recognizing the reflective capacities (also the result of evolution) that were emphasized in the first half of this chapter. If agents consciously endorse their instinctive dispositions toward reciprocity, it may lose whatever conditional nature it has, and become a more solid foundation for a prosperous, flourishing society.

We started this chapter by consolidating judgment and will from the first two chapters into a conception of character, which identifies the Kantian-economic individual. But while she is essentially individual by virtue of her autonomy, she is socially oriented by virtue of the moral law which she sets to herself through exercising her autonomy. The chapter ended with a discussion of the same Kantian-economic agent's sociality, how autonomous agents interact in an atmosphere of mutual respect and concern, including the limited but essential role played by the market in preserving autonomy and dignity (within the bounds of justice). In the final two chapters of this book, we turn to matters of policy, or how the actions of the state are delimited by the respect owed persons because of their dignity. In Kantian ethics, the state has the same moral responsibilities that individuals do, because a person must be treated with respect by anyone. We will see that this realization has serious implications for the theory and practice of consequentialist welfare economics.


(1.) Gary Watson puts this in terms of acting on the basis on one's valuations (arising from judgment) rather than from one's motivations (arising from desire); see his “Free Agency,” especially 215–6.

(2.) Davis, “Atomism, Identity Criteria, and Impossibility Logic,” 83.

(3.) Reath, “Legislating for a Realm of Ends,” 192.

(4.) I do not have much to say regarding methodological individualism, the modeling strategy that attempts to explain all social phenomena by reference to individuals, since my immediate concerns are not methodological, and so much in this debate hinges on the understanding of the individual, especially whether that understanding includes relationships with other persons. Generally, I am sympathetic to Geoffrey Hodgson's view that “all satisfactory and successful explanations of social phenomena (including in economics) involve interactive relations between individuals,” but that methodological individualism hardly seems like an appropriate name for such theories (“Meanings of Methodological Individualism,” 217). Nonetheless, according to Hodgson, Friedrich von Hayek held such a view under that very name, maintaining that “society consists not merely of individuals, but also of interactions between individuals, plus interactions between individuals and other aspects of their environment including, presumably, both the natural world and other socio-economic systems” (“Meanings,” 215, referring to Hayek, Studies in Philosophy, Politics and Economics, 70–1). See also Zwirn, “Methodological Individualism or Methodological Atomism,” who makes similar claims regarding Hayek, and argues that the term methodological atomism should be used for the mainstream view, which omits social factors, while methodological individualism should be reserved for accounts like that of Hayek.

(5.) Davis, “Conception of the Socially Embedded Individual,” 92. See also Davis, Theory of the Individual in Economics, chap. 6.

(6.) See Hargreaves Heap, “Individual Preferences and Decision-Making,” for a brief survey of social preferences. For more on the compatibility of autonomy with concern, care, and community, see Hill, “Importance of Autonomy,” and Oshana, “Autonomy Bogeyman.”

(7.) Lutz, Economics for the Common Good, 6.

(8.) Granovetter, “Economic Action and Social Structure,” 58.

(9.) Davis, Theory of the Individual, 108; Lutz, Economics, 6.

(10.) Hodgson, “Meanings,” 218 (emphasis mine).

(11.) To be sure, on methodological grounds, Vilfredo Pareto did say that “the individual can disappear, provided he leaves us [a] photograph of his tastes” (Manual of Political Economy, 120), quoted in Rizvi, “Adam Smith's Sympathy,” 249.

(12.) For a survey, see Davis, Theory of the Individual, 117–27.

(13.) Frankfurt, Necessity, Volition, and Love, 132.

(14.) Dworkin, “Hard Cases.”

(15.) What can I say—once you're a Jet …

(16.) In no way am I implying that these changes are necessarily bad; I imagine every new parent must have said, at some point, “I was never the same after my child was born.” (And they do not mean just extra pounds or less hair!)

(17.) On Gesinnung, see Allison, Kant's Theory of Freedom, 136–45, and Munzel, Kant's Conception of Moral Character. Munzel chooses the translate the term—“one of the most difficult words to render in English”—as “comportment of mind,” which “is intended to convey the specific sense thereof as one's mental bearing, informed by principles one consistently adopts in setting and pursuing one's purposes and in guiding one's choice making” (xvi–xvii).

(18.) Allison, Kant's Theory of Freedom, 136–7. Note the relevance of the second sentence to my brief comments regarding situationism at the end of the last chapter.

(19.) Korsgaard, “Personal Identity and Unity of Agency,” 378.

(20.) For more discussion on the relationship between the two topics, see the essays in Mackensie and Atkins, Practical Identity and Narrative Agency. In this discussion I do not rely on Kant's own discussion of personal identity, which is largely metaphysical and not relevant to my treatment of character; on these neglected issues, see Ameriks, Kant's Theory of Mind, chap. 4; and Kitcher, “Kant on Self-Identity” and “Kant's Philosophy of the Cognitive Mind.”

(21.) Darwall, “Scheffler on Morality,” 254, responding to Parfit, “Later Selves and Moral Principles.” (Lest there be any confusion, the bulk of Darwall's article deals with Scheffler's arguments in “Ethics, Personal Identity, and Ideals of the Person” regarding Parfit; only the last paragraph, quoted above, deals with Parfit directly.) See also Shoemaker, “Utilitarianism and Personal Identity.”

(22.) Darwall, “Scheffler on Morality,” 254–5.

(23.) Darwall, Impartial Reason, 101–2.

(24.) Korsgaard, Self-Constitution, 25 (emphasis mine). Korsgaard is not the only philosopher to have argued for self-constitution, though her conception of it is unique; for a list of other adherents, see Velleman, “Self as Narrator,” 203n1, and for a critical survey, see Berofsky, “Identification, the Self, and Autonomy.”

(25.) Munzel uses similar language in describing “the role of maxims as in fact constituting principles formative of character, as being (literally speaking) ‘character-building devices’” (Kant's Conception of Moral Character, 68).

(26.) Korsgaard, Self-Constitution, 42–4.

(27.) Ibid., 43.

(28.) Feinberg, Harm to Self, 35; see more generally 33–5 on autonomy as self-determination.

(29.) Gerald Dworkin, “Autonomy and Behavioral Control,” 24. (See also Oshana, “Autonomy and Free Agency,” 196–8, and Herman, “Responsibility and Moral Competence,” 91–7.) Note also the similarity with economic conceptions of business competition (to be discussed more in Chapter 4): mainstream economists treat competition as a fixed equilibrium determined by certain initial conditions, whereas Austrian economists see it as a continual process of discovery, innovation, and equilibration that never actually ends in an equilibrium but is always heading toward one until the next change in market conditions alters where the hypothetical equilibrium is. (For instance, see Ikeda, “Market Process.”)

(30.) In this sense, she achieves integrity in the formal sense of coherence and continuity, as opposed to more substantive understandings (implying specific moral values); see McFall, “Integrity,” on the various meanings of the term, and Minkler, Integrity and Agreement, for a comprehensive discussion of substantive integrity (specifically, honesty) within economics. Korsgaard can be read as arguing that formal coherence as an agent implies substantive morality according to Kantian principles.

(31.) Korsgaard, Self-Constitution, chap. 7.

(32.) Ibid., 135, drawing on Plato's Republic, book 1.

(33.) Korsgaard, Self-Constitution, 126.

(34.) Ibid., 214.

(35.) The most well-known conception of this, perhaps, is George Ainslie's picoeconomics (see Picoeconomics and Breakdown of the Will). See also the papers in Elster, Multiple Self (including an early presentation of picoeconomics by Ainslie); Lynne, “Divided Self Models”; Davis, Theory of the Individual, chap. 4; and Bazin and Ballet, “Basic Model of Multiple Self.” Etzioni argues that Kantian decision-making itself is best represented by multiple selves: “by Kantians, the self is divided, one part standing over the other, judging it and deciding whether or not to yield to any particular desire” (“Toward a Kantian Socio-Economics,” 140). It is unclear how literally he meant this (a problem common to much of the multiple selves literature), but certainly (p.216) Kant did not posit separate selves, but a single will which strikes the appropriate balance between duty and inclination (see note 40 in Chapter 2).

(36.) Elster, Multiple Self, 30–1 (emphasis in original).

(37.) This can be taken in the context of the debates between communitarians and John Rawls over the extent that society shapes the individual; see Bell, “Communitarianism,” section 2.

(38.) Korsgaard, Sources of Normativity, 101.

(39.) Korsgaard, Self-Constitution, 23.

(40.) Ibid., 120. As Kant writes in the Religion, “freedom of the power of choice has the characteristic, entirely peculiar to it, that it cannot be determined to action through any incentive except so far as the human being has incorporated it into his maxim” (23–4).

(41.) Korsgaard, Self-Constitution, 139, based on Plato, Republic, 437c; this also relates to the discussion of volitionism and “gaps” in Chapter 2.

(42.) Sen, Rationality and Freedom, 36; on this, see also Charles Taylor's conception of “strong evaluation” in Sources of the Self, and Davis, “Identity and Commitment.”

(43.) On the latter point, see David George's frank description of his acknowledged love for junk food in Preference Pollution, 8–11.

(44.) Davis, Theory of the Individual, 114.

(45.) Davis, “Identity and Individual Economic Agents.”

(46.) Davis, “Conception of the Socially Embedded Individual,” 94.

(47.) Davis, Theory of the Individual, 114.

(49.) Oshana, “Autonomy and Free Agency,” 196.

(50.) Herman, “Moral Improvisation,” 298.

(51.) Davis, “Collective Intentionality and Individual Behavior,” and Theory of the Individual, chap. 7.

(52.) Gilbert, Walking Together.

(53.) See, for instance, Schmid, “Rationalizing Coordination.”

(54.) See, for instance, Gilbert, Walking Together, chap. 6: “When a goal has a plural subject [as opposed to the shared personal goal of the participants], each of a number of persons … has, in effect, offered his will to be part of a pool of wills that is dedicated, as one, to that goal” (185). But she explicitly disavows any notion of a group mind: “human beings create joint commitments together and thereby constitute plural subjects” (Sociality and Responsibility, 3).

(55.) See Tuomela, Importance of Us. The characterizations of the last two viewpoints comes from Schulte-Ostermann, “Agent Causation and Collective Agency,” 192.

(56.) Davis, Theory of the Individual, 130.

(57.) Ibid., 134.

(58.) I would be remiss if I neglected to mention team reasoning, a conception of collective action that emphasizes processes of rational decision-making over the nature of agency; see Gold and Sugden, “Theories of Team Agency,” for a comprehensive summary of the various approaches to team agency, as well as their relationships to the conceptions of plural agency described above.

(59.) Louden, Kant's Impure Ethics, 172. See also Herman: “if autonomy is a source of dignity, it seems equally to be the source of a kind of autarchic individualism, supporting a conception of persons as radically separate from one another” (“Cosmopolitan Kingdom of Ends,” 52).

(60.) Sullivan, Immanuel Kant's Moral Theory, 199.

(61.) Berlin, “Two Concepts of Liberty.” (The role of negative liberty in Kant's political philosophy will be addressed in the next chapter.)

(62.) Kant, Metaphysics of Morals, 464.

(63.) In other words, the Vulcan salutation “Live long and prosper” from Star Trek was particularly ironic; a race ruled by logic may live long, but true prosperity would likely elude it.

(64.) Gregor too considers that “Kant's ethics is primarily a study of [imperfect] duties,” and blames the characterization of his ethics as “legalistic” for overlooking this (Laws of Freedom, 95).

(65.) Kant, Metaphysics of Morals, 453.

(66.) Ibid., 449 (emphasis removed).

(67.) Ibid., 449–50. It is important to note also that the duty of beneficence does not extend to concern for another person's moral character; we certainly have a duty not to corrupt others, but no positive duty to look out for others' virtue. (See ibid., 386, 393–4, and Sullivan, Immanuel Kant's Moral Theory, 205.)

(68.) At times, however, Kant seemed to lean toward a maximizing conception; see Baron, Kantian Ethics Almost without Apology, 95–8, where she rejects this as an inaccurate and inconsistent picture of Kant's views.

(69.) Singer, “Famine, Affluence, and Morality”; Veblen, “Why Economics Is Not an Evolutionary Science.” Kant does, however, say that the rich have a stronger obligation to help others, though their beneficence is less meritorious than a similar act on the part of a less wealthy person (Metaphysics of Morals, 453). Nonetheless, beneficence is an imperfect duty for all, rich and poor.

(70.) Kant, Metaphysics of Morals, 448–9.

(71.) Ibid., 454; this is also relevant to the critique of “libertarian paternalism” given in Chapter 5 of this book.

(72.) On the degree and kind of latitude allowed with respect in imperfect duty, see Baron, Kantian Ethics Almost Without Apology, chap. 3; Herman, “Scope of Moral Requirement”; Sherman, Making a Necessity of Virtue, chap. 8; and Baron and Fahmy, “Beneficence and Other Duties of Love” (as well as the references (p.218) given in note 55 in Chapter 1). Related to this is the issue of supererogation (acting above and beyond the demands of morality) and whether Kant can accommodate the concept; this is a primary theme of Baron's Kantian Ethics (and references therein).

(73.) This includes being purposefully attentive to the sufferings of others: “It is therefore a duty not to avoid the places where the poor who lack the most basic necessities are to be found but rather to seek them out” (Kant, Metaphysics of Morals, 457).

(74.) Groundwork, 438.

(75.) Ibid., 433.

(76.) See Paton, Categorical Imperative, 194–5, on moral progress and the kingdom of ends; in particular, “one great merit of Kant's system is that it puts into a true perspective the spirit, as opposed to the letter, of the moral law” (194).

(77.) Groundwork, 434. The reader may note the similarity between the Kantian concept of universal legislation based on impersonal concerns and John Rawls's concept of hypothetical decision-making behind the “veil of ignorance” (Theory of Justice).

(78.) Indeed, Herman suggests that “if one thinks of Kantian morality as inherently social it is hard to see what the idea of a kingdom of ends adds” (“Cosmopolitan Kingdom of Ends,” 66).

(79.) Kant, “Idea for a Universal History,” 20; also, “the human being is a being meant for society (though he is also an unsociable one)” (Metaphysics of Morals, 471). (For more on this, see Wood, Kant's Ethical Thought, 213–5, and Kneller, “Introducing Kantian Social Theory.”)

(80.) Kant, “Idea for a Universal History,” 21.

(81.) Kant, Religion, 97.

(82.) Sullivan, Immanuel Kant's Moral Theory, 214–6.

(83.) Of course, as we shall see in Chapter 4, this is the standard assumption made in neoclassical law and economics, corresponding to Oliver Wendell Holmes's “bad man” (from “The Path of the Law”); see Cooter, “Models of Morality in Law and Economics,” and Gordon, “The Path of the Lawyer” (including the wonderful term Homo law-and-economicus).

(84.) Kant, Religion, 95.

(85.) For an application of this idea to the improvement of public debate (about which Kant was passionate), see Rossi, “Public Argument and Social Responsibility.”

(86.) The similarities between Smith and Kant have been explored most intensively and thoroughly by Samuel Fleischacker; see his “Philosophy in Moral Practice,” “Values Behind the Market,” and Third Concept of Liberty.

(87.) Theory of Moral Sentiments, I.i.5.5. See Evensky, Adam Smith's Moral Philosophy, 12–6, for more on Smith's vision of the ideal progress of humankind, (p.219) corresponding (politically) to “the liberal plan of equality, liberty, and justice” (Wealth of Nations, IV.9.3).

(88.) Young, Economics as a Moral Science, 25.

(89.) Fleischacker, On Adam Smith's Wealth of Nations, 91 (emphasis in original).

(90.) See ibid., 55–7, on Smith's moral/political assessment of capitalism. Also, Smith's conception of self-interest (or self-love) is very different from that of the modern economist; on this see Wight, “Adam Smith and Greed,” and McCloskey, “Adam Smith.”

(91.) Furthermore, as Deirdre McCloskey explains in Bourgeois Virtues, actual markets also support virtue, not just in the minimal aspect of respect described in the sketch of markets above, but in a strongly positive sense, incorporating not just prudence, justice, temperance, but also faith, hope, and love.

(92.) See Fleischacker, Third Concept of Liberty, 137–8, and references to WN therein.

(93.) See Theory of Moral Sentments, VI.ii.1, for Smith's description of the diminution of sympathy and benevolence as social distance increases, what Young calls “a kind of inverse square law” (Economics as a Moral Science, 71), which is also very similar to Kant's views on social distance and beneficence, discussed below (see also Herman, “Scope of Moral Requirement,” 205–8).

(94.) Wealth of Nations, IV.9.51. This is also what David Gauthier means in Morals by Agreement when he refers to the market as a “morally free zone”: “in understanding the perfect market as a morally free zone we shall be led back to its underlying, antecedent morality” (84–5)—that is, mutually agreed-upon constraints on behavior corresponding to what Smith called the “laws of justice” that thereby define the boundaries of the market. How Gauthier breaks from both is by claiming that the market is the ideal model for an ethical society, and morality is necessary only where markets are not possible (see Morals by Agreement, chap. VIII).

(95.) Kant, Metaphysics of Morals, 458.

(96.) Wight, “Adam Smith's Ethics,” 156.

(97.) As Aristotle wrote, “if people are friends, they have no need of justice” (Nicomachean Ethics, 1155a); however, Sherman claims that Kant would have said the opposite, based on our “unsocial sociability” and the possibility of betrayal of respect between friends (“Virtues of Common Pursuit,” 20). See also Korsgaard, “Creating the Kingdom of Ends,” 189–97 on the close relationship between Aristotle's and Kant's conceptions of friendship, as well as the references in the next note.

(98.) Wood, Kant's Ethical Thought, 275 (more generally, 275–82); see also Kant, Metaphysics of Morals, 469–73, and Paton, “Kant on Friendship.” This criticism of Kant regarding friendship often comes from virtue ethicists, who have written much more about friendship than Kant has; Aristotle's Nicomachean Ethics (p.220) (books 8 and 9) is the starting point (and, frequently, the stopping point as well) of any philosophical discussion of friendship. (See also Sherman, Fabric of Character, chap. 4; Badhwar, Friendship [especially her introduction]; and McCloskey, Bourgeois Virtues, chap. 8. For a popular account informed by philosophy as well as other fields, I highly recommend Joseph Epstein's Friendship: An Exposé.) Consequentialists, naturally, have the hardest time explaining friendship; see, for instance, the critical work of Kapur, “Why It Is Wrong to Always Be Guided by the Best,” and Stocker, “Schizophrenia of Modern Ethical Theories.”

(99.) Kant, Metaphysics of Morals, 451.

(101.) Ibid., 452. For a more thorough treatment of impartiality and love in a Kantian context, see Velleman, “Love as a Moral Emotion.”

(102.) Korsgaard, “Creating the Kingdom of Ends,” 189.

(103.) This applies to commitments we make to ourselves also; see the discussion of duties of self-respect and weakness of will in Chapter 2.

(104.) Korsgaard, “Creating the Kingdom of Ends,” 192.

(105.) For a detailed look at how autonomy plays into family relations and friendships, see Kupfer, Autonomy and Social Interaction, chap. 4.

(106.) Hume, Treatise of Human Nature, 521.

(107.) Gintis, “Strong Reciprocity and Human Sociality,” 169; see also various chapters in Gintis et al., Moral Sentiments and Material Interests, in particular the first chapter (by the editors).

(108.) Gintis, “Strong Reciprocity and Human Sociality,” 169.

(109.) Sen, “Foreword,” x.