In his landmark paper, ‘The Schizophrenia of Modern Ethical Theories’, Michael Stocker introduces an affliction that is, according to his diagnosis, endemic to all modern ethical theories.Footnote ¹ This is the famous moral schizophrenia that is created when ethical theories require agents to adopt a divided moral psychology in order to achieve special goods like friendship. Stocker claims that this condition is profoundly unhealthy: he describes it as a ‘malady of the spirit’ and considers it to be a reductio of any ethical theory that makes it necessary for its representative agents to integrate two distinct types of thinking within an aggregate moral psychology.

Stocker's paper is well known and often cited, yet moral schizophrenia remains an obscure diagnosis. This is partly due to the fact that ‘Schizophrenia’ led to a stalemate between defenders of modern ethical theories and their critics. On one hand, defenders of modern ethical theories tend to respond to Stocker's challenge by emphasizing the merits of acting on the prescriptions of duty and, in some cases, by noting that acting in ways prescribed by duty need not preclude the achievement of goods like friendship.Footnote ² In doing so, however, most sidestep the psychological challenge of moral schizophrenia by either ignoring the fact that acting for the sake of duty requires a divided psychology or by denying that a divided psychology is burdensome.Footnote ³ The result is an unwarranted level of complacency with respect to the psychological mechanisms required to implement ethical theories in a realistic human context. On the other hand, supporters of Stocker tend to rely on the moral schizophrenia diagnosis as if it provides a decisive blow to the credibility of any theory that requires a divided moral psychology. This is a conclusion in need of further argument, and critics of modern ethical theory display their own type of complacency by assuming that moral schizophrenia is intolerably burdensome compared to the alternative moral psychologies available to human agents.

My aim in this article is to argue that moral schizophrenia, properly understood, is not as disruptive as Stocker and his supporters suggest. Thus, my aim is to argue that the type of psychology associated with moral schizophrenia is not exceptionally pathological when compared to a non-schizophrenic alternative.Footnote ⁴ However, I also intend to argue that moral schizophrenia represents a non-trivial practical challenge for most ethical theories in current circumstances. My aim, then, is to break the stalemate above and suggest that the psychological burdens of moral schizophrenia are both acceptable and deserving of serious consideration.

This conclusion, if correct, will be significant for a broad range of ethical theories despite the fact that moral schizophrenia is disproportionately presented as an objection to consequentialism. The fact that the most influential forms of consequentialism involve indirect measures forces its advocates to be especially forthright about a division existing between the theory's prescriptive justifications and the motives it recommends for moral agents to cultivate.Footnote ⁵ As a result, consequentialism is sometimes assumed to be specially vulnerable to moral schizophrenia despite Stocker's indictment of modern ethical theory as a whole. Consequentialists therefore have an added incentive to consider the analysis of moral schizophrenia provided here, but advocates of other ethical theories ought to be similarly interested in the substance of Stocker's challenge.

The value of my analysis is that it suggests a different way of responding to the moral schizophrenia challenge than is normally pursued. Instead of accepting that moral schizophrenia is pathological and then attempting to demonstrate that it does not apply to one's chosen ethical theory, my analysis allows one to accept the possibility that moral schizophrenia is a necessary feature of implementing an ethical theory without having to accept the conclusion that it is intolerably disruptive. The difference between these two responses to the moral schizophrenia challenge is important, because the latter invites further investigation into the psychological burdens that are associated with living out one's chosen ethical theory. It acknowledges that there may be tolerable but non-trivial psychological burdens for human agents seeking to live according to the demands of an ethical theory in a realistic social context.

The article is structured as follows. In the first three sections I rule out a number of incorrect interpretations of Stocker's position and then offer what I take to be the most accurate description of moral schizophrenia and the reasons why Stocker believes it is disruptive. This is important, because the argument in ‘Schizophrenia’ is not nearly as clear as it is provocative, and the ambiguity in this widely known presentation of moral schizophrenia invites misinterpretations of what turns out to be a subtle diagnosis when less commonly read sources are taken into account. In the fourth section of the article I argue that Stocker does not offer sufficient evidence to conclude that moral schizophrenia is pathological compared to his preferred moral psychology. Finally, in the fifth section of the article, I refer to the paradox of friendship to argue that moral schizophrenia is a noteworthy challenge even if it is not harmful enough to constitute the moral equivalent of a serious psychiatric disorder.

I

Stocker claims that moral schizophrenia is a condition that occurs when agents are forced to accept a lack of harmony between their reasons and their motives.Footnote ⁶ Reasons here refer to the values or justifications that determine what agents ought to do according to their chosen ethical theory, i.e. considerations of duty, obligation and rightness. Motives, on the other hand, refer to the desires that lead agents to pursue things that they consider to be good. This distinction is surely a rough account of the psychological mechanisms responsible for human moral agency, but the claim that there is something harmful about a lack of harmony between an agent's reasons and motives is based on the plausible intuition that it is unhealthy to be motivated to seek what one does not value or to value what one is not motivated to seek. Stocker acknowledges that some forms of this disharmony, like weakness of the will, may be inevitable, but he maintains that an agent ought to at least be moved by her major ethical values and that she ought at least to value the things that her major motives identify as being good. Otherwise, she will exhibit a divided moral psychology – one that is, according to Stocker, sufficiently disruptive to prevent her from leading a fulfilling moral life.

Given this abstract description of moral schizophrenia, it is no wonder the term is commonly cited in conjunction with a memorable example that Stocker provides to show that modern ethical theories precipitate the lack of harmony that he has in mind. In the example, Stocker asks you to imagine that you are in the hospital recovering from a long illness when a friend, named Smith, drops by to boost your spirits.Footnote ⁷ You are initially very pleased by Smith's visit and further convinced that he is a true friend. However, as you talk to Smith about why he has come, you slowly discover that he has not been moved to visit you because he cares specifically about you as a friend. Instead, his decision to visit you was motivated by duty. Stocker suggests that Smith may have felt a duty to visit you as a fellow Christian or communist, but the example is set up to be easily adapted to any modern ethical theory. For example, Smith may have determined that visiting you in the hospital was his best available opportunity to maximize the good, or he may have chosen this opportunity to fulfill an imperfect duty to support the autonomy of others. The point is that if Smith's duty is so generated, then the nature of his decision to visit you seems to undermine the authenticity of the act understood as an act of friendship. A true friend would presumably visit you in the hospital because he cared about you and not merely because duty obligated him to do so. Smith's visit seems to be missing the intimacy that makes friendship and other types of personal relationships valuable.

The Smith example is persuasive enough to have become a well-known thought-experiment in contemporary ethical theory, yet it is not immediately obvious how the example is meant to prove Stocker's point. This much is clear: Stocker provides Smith as an example of a harmonious moral agent – that is, one whose reasons coincide with his motives – who is subsequently cut off from a significant source of moral value because he is not able to act from motives that are consistent with friendship. This is an important point to remember, because it is easy to conflate the two separate obstacles that Stocker presents for modern ethical theories: (1) the inability to accommodate sources of moral value like friendship, and (2) the inability to provide agents with a source of moral justification that is consistent with a healthy psychology. The challenge associated with moral schizophrenia is an exclusive disjunction: modern ethical theories either require harmonious agents to forsake moral goods like friendship or they require that agents lack an important kind of psychological harmony. The problem with Smith is that he exhibits a ‘healthy’ moral psychology that, when combined with a modern ethical theory, ends up preventing him from participating in authentic personal relationships. This makes Smith an unusual representative of moral schizophrenia, since a lack of psychological harmony arrives on the scene only if we imagine a new version of Smith who is motivated to act for the sake of his friend without necessarily having his motive for the visit coincide with his reasons for acting in a morally justified manner. The disparity between this new agent's motive for his visit and the reasons that justify his visit according to his ethical theory is the root of moral schizophrenia. Stocker's claim is that modern ethical theories create moral schizophrenia because the only way they can capture vital sources of value, like friendship, is by resorting to a divided moral psychology of this kind.

The disjunction in Stocker's challenge leads to two different questions that one can ask about the origins of moral schizophrenia. First, what is it about Smith's visit that undermines his friendship with you? Second, what kind of disharmony is created in the psychology of agents capable of succeeding where Smith fails because their motives for visiting you are not unified with reasons provided by modern ethical theories? I discuss each of these two questions in turn.

II

There are several tempting but incorrect ways of interpreting what is disturbing about the Smith thought-experiment. One is to interpret the source of alienation at stake as a lack of emotional warmth exhibited by Smith. It may seem as if Smith's confession that he is visiting you out of a sense of duty reveals a cold psychological motive rather than a more sensitive motive that we normally associate with personal relationships. However, a lack of warmth is not the heart of the problem with which Stocker is concerned. A lack of emotional warmth might make matters worse, but it is not responsible for the kind of alienation with which Stocker is concerned. To see this, one need only note that agents can act for the sake of duty in passionate ways. Moreover, agents can act for the sake of friends in ways that are calm, deliberate and devoid of the warm emotional characteristics that accentuate our relationships but are not necessary for their integrity. For example, if Smith visits me in the hospital because he is overcome by a passionate devotion to communism and I happen to be a fellow comrade, then his motive is steeped in emotion but nevertheless fails as an act of friendship. Alternatively, if he visits me because of a direct concern for my well-being but he happens to be a subdued, emotionally frigid individual, then his action is still one of a friend. Most of us prefer that our friends not be this way, but a lack of emotional warmth is not the source of the problem at hand.Footnote ⁸

Another tempting but incorrect way of interpreting the Smith example is to view the source of alienation as a lack of spontaneity on Smith's part. In other words, the fact that Smith planned out his reason for visiting you in the hospital in advance may be seen as alienating because he seems to have reflected, deliberated and then ultimately chosen visiting you as his best available option, as opposed to simply acting on an immediate motive to cheer you up. The concern here is reminiscent of Bernard Williams’ claim that ethical theories, like Kantianism, force agents to perform ‘one thought too many’ in situations where they ought instead to act from immediate motives.Footnote ⁹ A true friend, one might think, would not need to engage in critical reflection before deciding to visit you in the hospital; he would simply realize that this is an appropriate action for him to perform for the sake of a friend. However, that this interpretation of what is principally wrong in the Smith example is mistaken is revealed in at least two ways. First, if a lack of direct, spontaneous action were the root problem in the Smith example, he could step back and deliberate about what any ethical (or scientific, mystical, etc.) theory required of him and the good of friendship would be equally precluded. Stocker's target is presumably more restricted than this result would imply. Second, there are cases in which actions resulting from reflective deliberation still qualify as authentic acts of friendship. For example, if Smith reflected for hours (using elaborate formulas and reference information about your preferences) on the best way to cheer you up and then finally settled on a visit (over, say, sending flowers), his act would surely be one of a friend.Footnote ¹⁰ He would be a particularly neurotic friend, but his direct concern for you would still manifest itself in an authentic act of friendship.Footnote ¹¹

Finally, it is tempting to assume that impartiality plays some critical role in what makes the Smith example problematic. Most modern ethical theories provide agents with impartial reasons for fulfilling their duties, and one might think that this impartial origin for the agents’ duties is what creates an obstacle for authentic relationships. Friendship and other personal relationships are prime examples of goods that require partiality to be authentic, so acting on impartial reasons would seem a sure way to undermine the value of these relationships. And so it is; however, it is not the defining feature of what makes the Smith example problematic. Performing an act of friendship for the sake of impartial reasons is sufficient but not necessary to undermine the value of the act. To see this, we need only imagine a case where Smith visits you in the hospital out of a sense of duty that is generated from partial considerations. For example, imagine that after Smith reveals that a sense of duty led him to visit you in the hospital, you ask what kind of duty it was, exactly, that motivated his visit, and he responds: ‘It is my duty as a friend to visit you. I take my personal relationships seriously, and so I work hard to be a good husband, father, brother or friend to those I owe special consideration.’ To twist the knife, imagine that he also confesses he would really rather be golfing if it were not for his obligation to fulfill his special duty to you, as a friend. We now have a case where impartiality is not a factor in Smith's reasons for visiting you in the hospital, yet the objectionable aspect of his visit remains. Thus, impartiality is not the essential component of what Stocker identifies as a barrier to acting in a way that is consistent with friendship.Footnote ¹²

This brings us to the heart of the problem in the Smith example. The problem is that Smith cites an impersonal reason for visiting you in the hospital – a reason that does not indicate that he values you for your own sake.Footnote ¹³ This is the core element of Smith's motivation that is toxic to the authenticity of friendship: his reason for visiting you treats you as an instrumental good. If Smith were a true friend, he would be moved to visit you because he cares about you for your own sake and not because he views visiting you as a means of achieving some further objective. As soon as Smith discloses that he has come because of a desire to discharge some underlying obligation, the ordinary assumption that he is there because of a direct concern for your welfare seems to evaporate. If Smith is motivated by impartial considerations, then his failure to act for your sake is that much more noticeable, but any underlying obligation is enough to undermine the visit as an act of friendship if the obligation serves as a motive that treats you as an instrumental means of discharging his duty.

The fact that acts of friendship can be undermined in this way reveals something interesting about the social relations that Stocker identifies as special goods that modern ethical theories have a problem capturing, e.g. love, friendship, affection, fellow feeling, and a sense of community.Footnote ¹⁴ These goods are unique because they must be achieved by acting for the sake of the relevant persons involved to be valuable. It is not possible, for example, to achieve goods like love or friendship by seeking out such relationships with the aim of realizing them as moral goods. To do so is self-defeating, because pursuing a personal relationship as an instrumental means of promoting the good will succeed only in tainting the authenticity of the relationship and thus undermining its goodness.

Personal relationships are curious moral goods in this respect, but they are not the only goods that, to retain their moral value, must be promoted via motives that are not directly influenced by instrumental considerations. Virtues like sympathy, benevolence, courage and modesty are also considered to be character traits that lose their moral value if they are performed from a sense of duty.Footnote ¹⁵ If we praise someone for her benevolence and she responds by saying that she does her best to live up to this description because she feels that she is morally obligated to do so, then we may continue to view her actions as laudable but we will no longer see this person as genuinely benevolent in the sense that the virtue requires. She will be acting in a way that is consistent with benevolence, but it will no longer be obvious that she possesses the virtue if her actions are guided by ethical obligations rather than an immediate concern for others. Similarly, an agent who self-consciously strives to become modest because she believes she is morally obligated to do so may fail to capture the virtue of modesty. If she knowingly alters the way she presents herself to others exclusively because she thinks that doing so is morally required, she will fail to exhibit the character trait we associate with the virtue of modesty. Hence, personal relationships are not the only goods undermined by directly seeking them for the sake of their instrumental contribution to one's general moral aims. Stocker is therefore correct to think there is a broad and important class of goods that present a challenge to modern ethical theories because these goods must be captured indirectly with respect to the way they are commonly justified.

III

Having established what is responsible for the alienation present in the example of Smith, we are now in a position to examine the second half of the disjunction associated with moral schizophrenia: the alleged pathology for the divided moral psychology required for modern ethical theories to successfully capture ethical goods like friendship. Imagine a different friend, Smithers, who visits you in the hospital because he cares about you for your own sake. What kind of disharmony exists within Smithers’ moral psychology if his motive for the visit is not unified with the reasons provided by his ethical theory? Smithers, we will assume, is able to avoid making his motives the transparent instruments of his reasons; thus, he visits you out of a direct concern for you as a friend and not exclusively because his ethical theory tells him that he has an obligation to do so. His motives are, in this respect, somewhat independent of his reasons for acting in a fully justified manner.

The first thing to note about Smithers is that he is someone any modern ethical theory will recommend as a representative moral agent. This ought to be obvious, but one sometimes encounters caricatures of modern ethical theories in which they are limited to endorsing agents like Smith, i.e. agents whose motives are merely transparent vehicles of their moral obligations. For example, utilitarianism is sometimes depicted as if it recommends that agents self-consciously seek to maximize happiness at every available opportunity, and Kantianism is sometimes depicted as if it recommends that (human) agents always attempt to act for the explicit and exclusive sake of duty. Both caricatures should by now be obsolete. These theories, and others, are fully capable of endorsing agents with a separation between the motives associated with their decision-making procedures and the reasons that reflect the standards of rightness they accept as indicative of what is morally justified.Footnote ¹⁶

Hence, the pressing question for modern ethical theories is not whether they can endorse agents like Smithers and thereby avoid the trap that precludes Smith from securing goods like friendship; the question is how the interface between the motives and reasons of these agents ought to be structured and whether it creates an unacceptable level of disharmony within the psychology of agents who strive to live up to their chosen theory's demands. Stocker initially seems to offer a straightforward answer: any split between an agent's motives and her reasons constitutes a lack of harmony worthy of the name moral schizophrenia. This answer is suggested in a revealing passage where Stocker considers a sophisticated version of egoism that has agents pursue their personal relationships from appropriately intimate motives while periodically reflecting on how well their life is going in terms of fulfilling their egoistic goals. Stocker claims that even if such a theory is psychologically possible and avoids the trap of alienation in the agents’ relationships, it will not touch his criticism that these agents ‘will not be able to embody their reason in their motives; that they will have to lead a bifurcated, schizophrenic life to achieve what is good’.Footnote ¹⁷ This passage is important, for it shows that Stocker objects to reasons not being unified with motives at a fundamental level, so that even agents who successfully avoid alienation in their relationships are morally schizophrenic if their ethical theory provides an independent justification for these relationships.

Why does Stocker make this strict claim about the division between reasons and motives, and what makes this division psychologically taxing? Again, there are tempting but incorrect ways to answer this question. First, one might think the problem is created by the mere fact that an agent with a division between her motives and reasons will have two different sources of input affecting her decision-making process. The idea here is that it is not possible for a healthy psychology to be driven by two distinct sources of input – even in cases where the two sources recommend a single course of action. For example, the egoist who acts for the sake of her friends while her acts are simultaneously endorsed by her ethical theory (since authentic friendships will most effectively promote her egoistic goals) will have two distinct sources of input to guide her decisions.Footnote ¹⁸

This interpretation of moral schizophrenia seems to follow from the simplicity of Stocker's diagnosis in ‘Schizophrenia’, but it is ruled out by other sources in which he claims that acts of friendship can be overdetermined by multiple sources of normative control working in concert within the agent to endorse a single action.Footnote ¹⁹ Specifically, Stocker discusses the possibility of one source of normative control taking the form of a sine qua non condition on an agent's actions, i.e. a necessary condition that serves as a governing principle but not as a direct motive to perform a particular action. An example Stocker uses to describe this type of condition is sailing. If he sails to amuse himself, but he would not sail if he thought that doing so would kill him, then he does not sail in order to amuse himself and remain alive.Footnote ²⁰ Similarly, if he would not borrow a boat unless he had been given prior permission to use it, this does not imply that while using the boat he is acting with a conscious intention to avoid trespassing.Footnote ²¹ The underlying justifications for one's actions can therefore function as limiting conditions without necessarily acting as direct motives to perform particular actions. Hence, Stocker's moral schizophrenia diagnosis is not based on the mere psychological presence of an agent's impersonal moral reasons, for these reasons can act as governing principles that justify the agent's actions in a way that is safely independent of her immediate motives.

A second misinterpretation of Stocker's position is that certain motives (e.g. to act for the sake of a friend, as opposed to going sailing) should not be limited by reasons that determine how we ought to act. In other words, one might think Stocker is claiming that any variance between the motives necessary for authentic friendships and the governing conditions imposed by considerations of duty will create an intolerable lack of harmony in the psychology of a moral agent. This reading of Stocker accepts that having two sets of synchronous inputs is acceptable in some contexts, but it maintains that giving reasons priority over one's motives will improperly hold the motives of friendship hostage to the limits determined by considerations of duty. If you are Smithers, on this reading, it will be difficult to have authentic motives to act for the sake of a friend if you know that these motives can at any moment be overridden by your ethical obligations. The worry is that this relationship between reasons and motives creates psychological burdens in the form of a tendency for agents to think of their commitments to their relationships in impersonal terms – a consequence of being continuously aware of the limiting conditions responsible for determining when such relationships ought to be overridden by broader concerns.

This interpretation of Stocker's argument is tempting, but it incorrectly attributes to Stocker the extreme view that one's motives to act for the sake of one's friends should never, in a healthy moral psychology, be limited by one's impersonal reasons to act in an ethically justified manner. Stocker clearly distances himself from this extreme view and acknowledges that endorsing friendship no matter what the moral cost is a reductio that no one ought to accept.Footnote ²² He is wise to do so: if you find yourself committed to the claim that you ought to lie to protect a friend from being justly convicted of murder, then you know something is wrong with your analysis of friendship and obligation.Footnote ²³ Moreover, Stocker explicitly recognizes that moral considerations can act as limiting conditions without undermining one's authentic motives, and he specifically applies this observation to the case of Smith:

Stocker also distances himself from the view that agents should not consciously reflect on their moral duties.Footnote ²⁵ Thus, he is not proposing that moral schizophrenia occurs as soon as an agent accepts the possibility that she will sometimes have to reflect on the legitimacy of the motives associated with her friendships and potentially allow these motives to be overridden by her moral obligations. Stocker claims the goodness of friendship is not contingent on its being sanctioned by an agent's moral duty, but he does not endorse the romantic fiction that a healthy moral psychology requires friendship to be impervious to critical reflection and one's broader ethical obligations.Footnote ²⁶

What is left, then, to separate a modern ethical theory from Stocker's position? Both accept that considerations of duty can act as governing conditions that justify goods like friendship without acting as direct motives, and both accept that considerations of duty should sometimes take priority over an agent's motives to act for the sake of her friends. The remaining difference, I believe, is that modern ethical theories propose determinable methods for deciding when goods like friendship ought to be outweighed by considerations of duty. This is the key component in modern ethical theories that Stocker finds objectionable. Modern theories are not committed to the view that friendship must be motivated by an underlying sense of duty, so they can avoid endorsing agents like Smith, but modern theories are committed to the view that there is some determinable point where moral obligations ought to take priority over the motives that lead agents to act for the sake of their friends. Specifying when this point will occur is a difficult, often empirical, question to which modern theories will not have simple answers, but for each theory there is in principle an answer. This is not the case for the sine qua non conditions proposed by Stocker.

Stocker's view is that duty and friendship are each components of a realm of moral values that is not harmonious – a realm where one component may take precedence over another in certain cases without such cases implying a general priority relationship between the two components. Thus, Stocker claims that duty ought to limit friendship in some cases but that friendship ought to limit duty in others.Footnote ²⁷ According to Stocker, we can expect these two components of the more general domain of ethics to conflict without the need for a determinable method that decides in advance how the relevant sine qua non conditions adjudicate such conflicts. Rather than viewing ethics as synonymous with duty and then deciding whether there are cases where friendship qualifies as a legitimate exception to one's duty, Stocker's view is that friendship reveals how ethics extends beyond the scope of duty and is composed of a plurality of conflicting values.Footnote ²⁸ He thus rejects a hierarchical conception of the relation between the reasons that agents have for acting in accordance with duty and the motives they possess in order to sustain authentic personal relationships. The ambitions of modern ethical theory are misguided, according to Stocker, for no methodology can be established in advance to balance the significance of obligations generated by friendship compared to those generated by duty.Footnote ²⁹

Thus, Stocker believes that modern ethical theories lead to an unhealthy moral psychology because they specify our global ethical obligations in terms of a standard of right action that subsequently imposes a hierarchical relationship between one's duty and one's commitment to goods like friendship, i.e. goods that one cannot pursue for the sake of duty without self-defeating results. This hierarchical relationship between the agent's reasons for acting rightly and her motives for pursuing the good creates a psychological problem because the agent is presented with a single, overriding criterion of right action that she cannot straightforwardly reconcile with her motives when she inhabits a world of plural and conflicting ethical values. In effect, Stocker's objection is that modern ethical theories seek to implement a harmonious conception of rightness within an unharmonious realm of ethical goods. This leaves agents in the awkward position of being committed both to an overriding duty to act according to a unified conception of rightness and to their own independent motives to promote goods like friendship that can only be indirectly incorporated with their reasons to act for the sake of what is right.

The chain of events leading up to moral schizophrenia is therefore as follows: the presumption that ethics is synonymous with duty creates the hierarchical relation between reasons and motives, the need to reconcile the predetermined method that specifies this hierarchical relation with an agent's personal motives makes it necessary for indirect measures to be incorporated within what is otherwise an overriding obligation to act in an ethically justified manner (i.e. according to duty), and the indirect measures that prevent self-defeating results create a lack of harmony in the agent's psychology because they force the agent to balance two incompatible modes of thinking, i.e. thinking in terms of one's own personal motives versus thinking in terms of an overriding commitment to reasons that perceives the objects of one's motives (e.g. one's friends) as impersonal components of a broader ethical agenda. The key component in this chain of events is the determinable algorithm that specifies the relation between the limits of duty and the agent's motives. Being aware of this algorithm either forces the agent to see her personal relationships as contingent parts of a more general commitment to duty, or it forces her to suppress the commitment to duty in her moral psychology because, according to Stocker, ‘it is unclear both how and even whether, for example, the trust needed for friendship is compatible with an overriding, or even a very strong, concern for duty’.Footnote ³⁰ The trust in question is compatible with recognizing that duty must sometimes override friendship, but this trust is allegedly incompatible with governing conditions that require agents to approach the conflicts in the realm of moral values with a psychology containing sub-levels that treat friends as impersonal moral goods.

Thus, a healthy moral psychology, according to Stocker, is one that approaches the unharmonious subject matter of ethics with a unified, and so appropriately indefinite, point of view. Moral schizophrenia, by contrast, is the condition an agent suffers from if she attempts to approach the unharmonious subject matter of ethics with a determinate theory of the reasons she has to act in a justified manner, because this theory will pass on the conflicting elements of ethics to the agent's psychology and force her to partition her authentic motives from the caustic effects of her justificatory reasons.

IV

If this is an accurate description of moral schizophrenia, then it is clearly a more complex ‘malady of the spirit’ than one might otherwise assume after reading Stocker's original ‘Schizophrenia’ paper. It is no wonder, then, that proponents of modern ethical theories tend to ignore the psychological challenge presented by moral schizophrenia considering the subtle distinction that separates it from Stocker's favored moral psychology and the ways in which it is vulnerable to misinterpretation. The misinterpretations of Stocker's view (e.g. the claim that duty should never outweigh friendship) are easy to dismiss, and I expect that those working with an accurate interpretation of his view often find it so difficult to see why he thinks that theory-driven governing conditions create psychological problems that they simply lose interest.Footnote ³¹

More importantly, if one successfully identifies the subtle distinction between the governing conditions required by modern ethical theories compared to those endorsed by Stocker, it is not clear that he is correct to think that modern ethical theories are specially vulnerable to psychological difficulties that arise because of the conflicting obligations associated with one's commitments to goods like friendship and one's broader ethical duties. Instead, it is reasonable to demand further evidence to support his claim that a divided moral psychology is specially problematic compared to the undivided psychology he tacitly assumes to be a healthy default option. No one will claim that divisions within one's moral psychology are inherently desirable, and those who defend modern ethical theories will admit that challenges exist for agents seeking to reconcile their impersonal duties with their friendships. But it is not obvious that the challenges associated with a divided moral psychology are both (a) challenges that do not arise in some comparable form within a unified moral psychology, and (b) challenges that are so overwhelming that they preclude agents from experiencing a healthy mental life. There is, by Stocker's own admission, an inevitable tension between goods like friendship and our broader ethical obligations. Given this fact, it is not self-evident that the unified moral psychology Stocker recommends will be decidedly less taxing than a divided psychology when it comes to balancing our conflicting values.

Stocker's support for his conclusion rests on the claim that the need to separate a type of moral thinking from the other parts of one's decision-making ability is specially problematic because it requires that one periodically consult an otherwise suppressed set of reasons for action while normally keeping this set inactive when acting for the sake of one's friends. In Stocker's words, ‘It is bad enough to have a private personality, which you must hide from others; but imagine having a personality that you must hide from (the other parts of) yourself.’Footnote ³² This support for his indictment of a divided moral psychology seems compelling, but it is not decisive when Stocker fails to put his own alternative forward for the purpose of comparison. How much less disruptive is it for an agent to navigate a world of plural and conflicting ethical values if she proceeds without a fixed method for balancing her competing ethical commitments? She will not need to hide aspects of her personality from the other parts of herself, but she will feel a substantial amount of anxiety regarding the prospect of facing difficult ethical decisions without any predetermined theory to provide principles or guidelines for her to follow. There are, for Stocker, important sine qua non conditions that specify situations where considerations of duty ought to outweigh one's commitments to goods like friendship. But these sine qua non conditions cannot be determined in advance or else Stocker ends up facing the same structural problem he diagnoses for modern ethical theories, so agents are left without this source of guidance. That is, they are left without this source of guidance unless they possess a sufficiently well developed sense of practical wisdom that allows them to act on the substance of Stocker's sine qua non conditions without the benefit of predetermined, theory-driven principles. For those of us without this highly developed sense of practical wisdom, we will experience a lack of psychological harmony in the form of anxiety, or at least uncertainty, when faced with the prospect of dealing with conflicting commitments in the realm of ethical values without being able to consult some determinable source of guidance about how to adjudicate these conflicting commitments.

In effect, Stocker's method of dealing with the inevitable tension between ethical goods like friendship and our broader ethical obligations is to bury the tension underneath the surface of a unified psychology. This method avoids the complications that arise for modern ethical theories that employ a psychological division of labor to moderate the conflicts between our reasons and our motives, but the method provides no promise that these conflicts will disappear. It then becomes an empirical question how much is gained by dealing with the problem in this psychologically ‘harmonious’ way.Footnote ³³

To make the difference between these two methods more concrete, let us consider yet another person who comes to visit you in the hospital to boost your spirits. (It is, after all, supposed to be a long illness.) Call this person Smythe. Smythe has a unified moral psychology, but, unlike Smith, he is not motivated to visit you because he is acting on an impersonal duty. Smythe visits you because he cares about you for your own sake. Yet Smythe is aware that he must respect the appropriate sine qua non conditions that specify his competing moral obligations. These conditions need not be as demanding as those that are associated with maximizing theories like (most forms of) consequentialism, but they must be sufficiently stringent for Smythe's actions to be plausibly justifiable. Thus, Smythe cannot run over elderly pedestrians in a rush to see you before visiting hours are over, nor can he steal taxi fares from unsuspecting tourists to finance his transportation to the hospital. In less straightforward cases, Smythe will face difficult decisions about the relative priority of his commitments to his friends compared to his commitments to other competing ethical values. Will the relevant sine qua non conditions prevent him from briefly slipping away from work or taking time otherwise devoted to his children in order to visit you? The difficulty for an agent operating with the unified moral psychology Stocker proposes is knowing how to settle hard cases of this kind without any principled form of ethical guidance. Unless Smythe acquires an impressive capacity for practical wisdom that allows him to internalize the appropriate sine qua non conditions, he faces an arduous task recognizing the appropriate limits of his friendships.

The kind of moral agent Stocker proposes is therefore vulnerable to a potentially serious type of disharmony in his moral psychology. This disharmony will be created by psychological tensions that are different from those associated with moral schizophrenia, but the former could be just as disruptive, i.e. the psychological problems facing Smythe could be just as psychologically demanding as those facing Smithers. At the very least, there is no clear evidence suggesting that the problems facing agents exhibiting moral schizophrenia are so much worse, compared to the problems facing agents endorsed by Stocker, that only the former are prevented from leading a fulfilling moral life.

V

Thus, ethical theories are not necessarily destined to produce the equivalent of a mental disorder if they require agents to adopt a psychology that is divided into independent parts. Nevertheless, the fact that Stocker has not met the burden of proof necessary to establish that modern ethical theories are pathological should not lead proponents of such theories to ignore the possibility that he has identified an interesting psychological phenomenon. Balancing our broad ethical obligations with the authenticity of goods like personal relationships is a considerable challenge, and knowing when to adopt different perspectives within a divided moral psychology might well be a very strange and onerous task. We must therefore avoid dismissing Stocker's challenge too quickly. Instead, we ought to confront the details of the moral schizophrenia challenge in order to assess the genuine psychological burdens that it implies. This will be a constructive step toward breaking the stalemate between supporters and detractors of the original ‘Schizophrenia’ paper. My recommendation is thus that we dispense with the debate about whether moral schizophrenia is a reductio of modern ethical theories and instead begin to investigate the messy implications of living with a divided moral psychology.

By way of conclusion, then, I will draw attention to one important way in which living with a divided moral psychology presents a challenge for agents who subscribe to any moderately demanding ethical theory. The challenge arises in the context of what I will call the paradox of friendship. It is a challenge that has been of special interest to consequentialists, but it is by no means confined to this theory.

Recall that Smithers is our example of an agent who has reasons for acting in a justified manner and authentic motives to act for the sake of his friends. His motives are independent of his reasons in the sense that his motives are not transparent instruments of his reasons, but his motives are nevertheless constrained by the limiting conditions that are specified by his ethical theory's account of what is justifiable. Thus, his motives to act for the sake of his friends can coincide with his ethical theory's justification for the general value of friendship, but there will be at least some cases where the bonds of friendship ought to give way to his broader ethical obligations. To think otherwise is to give in to the romantic fiction that friendship is immune to one's ethical obligations, and as we have seen this is a conclusion no one ought to endorse. But when, exactly, should Smithers scrutinize his motives to ensure that the actions they recommend fall within the limits of what his ethical theory sanctions? The challenge for Smithers is being able to recognize when this ought to occur, i.e. when to set aside his commitments to his friends without undermining the authenticity of his relationships by scrutinizing his motives so often that they are chronically superseded by his impersonal reasons for acting in a justified manner.

What makes the challenge psychologically taxing is that at any given moment when Smithers has an opportunity to act in a way that is justified by his chosen ethical theory, it is unlikely that this theory will sanction acting for the sake of a friend rather than acting to benefit others in much more dire need of aid. For example, few ethical theories will proclaim that Smithers visiting you in the hospital is preferable to him volunteering at the burns unit of an understaffed children's hospital if these alternatives are part of a single, isolated decision. If an ethical theory did reach this conclusion, it would give us reason to doubt the theory.Footnote ³⁴ But, of course, the logic of this decision can be expanded to any instance where Smithers faces a similar set of opportunities, and if Smithers repeatedly decides against acting for the sake of his friends he will eventually undermine his relationships with them. Whether it is visiting a friend in the hospital, spending a quiet evening with his wife or buying gifts for his children, it is likely that at every given moment there will be an overriding demand on Smithers’ time and resources if he chooses to live up to his ethical obligations. Unless his friends and family are among those most deserving of aid,Footnote ³⁵ Smithers will be constantly tempted to abandon his commitments to them despite the fact that his ethical theory will endorse his relationships if they are globally evaluated, because they are necessary for him to develop into, and continue to be, the kind of properly socialized agent who is capable of responding to moral obligations at all.Footnote ³⁶

This is the problem that Peter Railton has memorably illustrated in his defense of consequentialism against the threat of alienation from one's personal relationships.Footnote ³⁷ Call it the paradox of friendship to accentuate the parallel with the paradox of hedonism – the latter being the fact that attempting to produce the most pleasure at every opportunity will result in self-defeating outcomes if this strategy is employed for any significant length of time.Footnote ³⁸ The paradox of friendship is a curious phenomenon. The capacity to engage in authentic relationships is a vital part of having a healthy human identity, and any ethical theory that fails to respect this capacity is ethically and psychologically indefensible. Yet at each opportunity for decision-making, the limits set by a moderately demanding ethical theory will prevent agents from honoring the level of commitment required for their personal relationships, because agents will be instructed to choose against the comparably trivial acts that, when taken as a whole, are necessary to sustain these relationships.Footnote ³⁹

For example, if Smithers is a Kantian agent he will find himself in situations like the one Railton imagines for his model agent, Juan, i.e. a choice between using resources to cheer up a despondent spouse or using these resources to help others in desperate need of aid.Footnote ⁴⁰ Smithers ought sometimes to privilege his friends and loved ones, but to do so he must make decisions that, when taken in isolation, are unjustifiable.Footnote ⁴¹ At least this is so if Kantians want to avoid the accusation that their ethical theory is implausibly lax because it can be interpreted so as to allow agents to discharge their imperfect duties by helping friends (or strangers) in trivial ways rather than helping others in more serious need of aid.Footnote ⁴² Similarly, in an isolated choice between trivial benefits for one's friends and vital benefits for strangers, the virtue of beneficence ought to outweigh the virtues of loyalty and trust associated with friendship, yet most forms of virtue ethics will want to claim that the virtues associated with friendship ought to be preserved for a virtuous life on the whole. Even if significant ‘agent-centered prerogatives’ are built into one's ethical theory,Footnote ⁴³ the paradox of friendship will arise as long as (a) circumstances are sufficiently ‘non-ideal’ and (b) the ethical theory is sufficiently demanding to place a higher priority on the desperate needs of strangers than the value of the comparatively trivial individual acts that collectively form a basis for agents’ special relationships. Neither one of these two conditions is unlikely.

The paradox of friendship is therefore a practical dilemma with nearly the same basic, ubiquitous character as the paradox of hedonism.Footnote ⁴⁴ It is a predictable result when certain widely shared ethical assumptions are implemented in non-ideal circumstances by creatures with our cognitive and emotional capacities. It can be partly alleviated by our assumptions about the priorities required for personal relationships changing so that those of us living in affluence need not lavish our friends and loved ones with resources to the extent we currently do for our relationships to be considered authentic. Nevertheless, the facts of human socialization are, I am assuming, such that a significant priority on friends and loved ones is essential for us to thrive and lead ethically justified lives; hence, the paradox of friendship will not be eliminated by the (appropriate) move to relationships with less drastic discount rates placed on the needs of strangers.

Inevitable though it may be, if the paradox of friendship is understood as an in-principle objection to ethical theory, then the indirect measures that have been proposed by consequentialists like Railton are, I think, highly convincing.Footnote ⁴⁵ However, successfully demonstrating that ethical theory can, in principle, avoid the paradox of friendship does not entail that the psychological implications of doing so are trivial. To prevent oneself from acting on the moment-to-moment recommendations of one's ethical theory in order to achieve the total good of friendship, one must temporarily suppress the part of one's moral psychology that provides moment-to-moment reasons for acting in an ethically justified manner. This could be a wholly or only partly subconscious process, but if the normative authority of one's ethical justifications must be suspended then this will be a unique psychological challenge. A careful balance is required for an agent like Smithers to cultivate a capacity for authentic relationships and an ability to know when, and for how long, to suppress his critical ethical thinking about the limits of friendship and its overall (impersonal) justification. Ideally, we want Smithers to receive both the moral education necessary to participate in authentic relationships and a finely tuned sense of when circumstances are urgent enough for him to stop insulating his personal motives from his justificatory reasons and potentially act against the bonds of these relationships. This is a difficult problem because Smithers will presumably need to consult his ethical theory to determine whether he is in circumstances urgent enough to stop suppressing his moment-to-moment ability to consult his ethical theory. But, of course, if this part of his moral psychology is being suppressed, then the question of whether his circumstances are urgent enough to reactivate the reasons supplied by his ethical theory is exactly the kind of question Smithers will be unable to ask.

It is hard to deny that overcoming this dilemma will be no ordinary psychological challenge. The challenge will not be exactly the same for every ethical theory: the level of demandingness, among other factors, will make the details of each theory's burdens unique. Nevertheless, each theory will face a shared need to temporarily suppress agents’ moment-to-moment access to justificatory reasons without cutting agents off from these reasons completely.

It is in the context of this issue that Stocker sardonically asks whether agents will have to be equipped with ‘mental alarm clocks’ to periodically wake them from intuitive moral thinking so they can step back and evaluate their actions from a reflective point of view.Footnote ⁴⁶ This may not be far from the truth. Using (at least) two levels of moral thinking apparently requires special mechanisms that will allow agents to periodically take up a reflective standpoint. For example, Philip Pettit suggests that consequentialists rely on what he calls moral ‘triggers’ designed to snap agents out of their ordinary moral thinking in exceptional circumstances.Footnote ⁴⁷ Pettit's suggestion essentially strips away the sarcasm from Stocker's mental alarm clocks and presents them as a viable option, though it does so without ignoring the fact that implementing a divided moral psychology using mechanisms of this kind is a delicate business. This is an intriguing proposal – one that is underappreciated within the consequentialist literature on the paradox of friendship, let alone that of other ethical theories that provide agents with justificatory reasons that are potentially caustic to the authenticity of agents’ personal motives.

The question of whether moral triggers are psychologically feasible will have to wait for another occasion, since my aim here is not to present a detailed treatment of the paradox of friendship or the psychological mechanisms necessary to avoid it. My point in closing with this paradox is to suggest that Pettit's strategy is representative of exactly the right way for modern ethical theorists to respond to moral schizophrenia, i.e. by stripping away the rhetorical insinuation that a divided moral psychology is pathological without ignoring the fact that it constitutes a legitimate challenge. Moral schizophrenia is a noteworthy psychological problem for all moderately demanding ethical theories even if falls short of constituting a decisive, in-principle objection. Thus, instead of assuming that moral schizophrenia is disruptive enough to be pathological and then focusing on whether it can be avoided, proponents of modern ethical theory ought to investigate the psychological obstacles associated with this condition and the extent of the burdens that it poses for human agents. In doing so, I suspect that we will learn important lessons about what is necessary for creatures like us to successfully balance our personal motives with our broader ethical obligations. This, in turn, will be a useful step towards achieving the kind of practical wisdom required for human agents to lead a life that is both harmonious and ethically justifiable.Footnote ⁴⁸