Moral Address, Moral Responsibility, and the Boundaries of the Moral Community

Ethics 118 (October 2007): 70–108� 2007 by The University of Chicago. All rights reserved. 0014-1704/2007/11801-0001$10.00

70

ARTICLES

Moral Address, Moral Responsibility, andthe Boundaries of the Moral Community*

David Shoemaker

When we hold someone morally responsible for her actions, we areacknowledging, reiterating, or perhaps even making it the case that,among other things, she’s a member of a particular sort of club, namely,the moral community. Members of the club are eligible for such as-sessments, and nonmembers are exempt. But what does it take to beor to become a member? Why are assessments of moral responsibilityintelligible and appropriate only for those on the inside? And what doesit take to lose membership or not to be a member in the first place? Itis my aim in this article to address these questions by advancing a theoryof the relation between moral responsibility, moral address, and moralcommunity.

I will simply take it for granted, following Peter Strawson, LawrenceStern, Gary Watson, Stephen Darwall, and many others, that there is animportant relation between these features.1 This is in part because ourpractices in voicing the praise and blame expressive of holding someonemorally responsible, in the paradigm case, consist of an interplay be-tween at least two agents, one who addresses a moral demand to the

* For extremely valuable criticism and advice on an earlier draft of this article, I amgrateful to Doug Portmore, Stephen Darwall, and Joshua Glasgow. I am also grateful toan anonymous referee and two anonymous associate editors at Ethics for their helpfulsuggestions. Finally, I would like to thank my stepdaughter Ashley Castillo for inspirationand insight.

1. See, e.g., Peter Strawson, “Freedom and Resentment,” in Free Will, 2nd ed., ed.Gary Watson (Oxford: Oxford University Press, 2003), 72–93; Lawrence Stern, “Freedom,Blame, and Moral Community,” Journal of Philosophy 71 (1974): 72–84; Gary Watson, “Re-sponsibility and the Limits of Evil: Variations on a Strawsonian Theme,” in Agency andAnswerability, ed. Gary Watson (Oxford: Oxford University Press, 2004), 221–59; StephenDarwall, The Second-Person Standpoint: Morality, Respect, and Accountability (Cambridge, MA:Harvard University Press, 2006).

Shoemaker Moral Address, Moral Responsibility 71

other via the praise or blame and the other who ostensibly hears, un-derstands, and either accepts or rejects the demand, and such an ex-change is possible only for those who have the capacity to enter into acertain kind of relationship with one another. Call those who share thiscapacity, then, moral agents, and call the collection of moral agents amoral community.

But what exactly does this capacity involve? To understand just whatmembership in (and with it the nature of) the moral community consistsof, the typical strategy is to explore the cases of those who are une-quivocally outside the community in an effort to find out what they lackthat we members have, that is, what prevents them from entering intothe requisite relationships with the rest of us. The consensus of mostcontemporary theorists is that what those outside the moral communitylack is the capacity to understand, apply, and/or respond to moral rea-sons. What provides intelligibility, a point, to our practice of praisingand blaming is that it involves genuine exchanges—moral conversations,if you will—and such exchanges, for these theorists, are exchanges ofmoral reasons. If one lacks the capacity for this sort of reason-basedaddress and exchange—a specific type of normative competence—thenone must be outside the boundaries of the moral community, beyondthe reach of a certain type of relationship.2 And while having this ca-pacity may not be sufficient for rendering actual praise and blame ap-propriate (there may be other excuses that cause us to suspend suchreactions in particular cases), it is nevertheless both necessary and suf-ficient for being a member of the moral community, rendering one atleast eligible for praise and blame.

My purpose in this article is to show that this theory, as it is currentlyespoused, is simply too crude and underdeveloped to deal adequatelywith several key marginal cases. And while it is true that we can learnmuch from consideration of those who are just outside the margins ofthe moral community, I hope to show that we can also learn much fromconsideration of those who are just inside the margins as well. I thusintend to build up a more fine-grained theory through considerationof four cases—two from outside the moral community and two frominside it—each of which will motivate the addition of one or moreimportant components to the theory. First, the case of psychopathy will

2. Beyond the theorists already mentioned, one can see explicit support for this stancein R. Jay Wallace, Responsibility and the Moral Sentiments (Cambridge: Cambridge UniversityPress, 1994); John Martin Fischer and Mark Ravizza, Responsibility and Control (Cambridge:Cambridge University Press, 1998), esp. 211–14 (the phrase “moral conversation” is theirs);Jeanette Kennett, “Autism, Empathy, and Moral Agency,” Philosophical Quarterly 52 (2002):340–57; Nomy Arpaly, Unprincipled Virtue (Oxford: Oxford University Press, 2003), esp.71–93; Susan Wolf, Freedom within Reason (Oxford: Oxford University Press, 1990); andGeorge Sher, In Praise of Blame (Oxford: Oxford University Press, 2006), e.g., 9.

72 Ethics October 2007

emphasize the underappreciated fact that a facility with recognizing,grasping, and exchanging moral reasons is not sufficient for moralagency. What is also needed is the motivation to comply with such rea-sons. Exploring the conditions necessary to such motivation will revealthe importance of certain emotional capacities. However, in consideringpossible exceptions to this criterion, found in various types of what Iwill call moral fetishists—moral fetishism is my second case—we will seethat these various types correspond to an important distinction betweenvarious types of moral reasons. As it turns out, it is only one type ofmoral reason that is relevant to moral responsibility, so the motivational(and emotional) restrictions on moral membership are more circum-scribed than typically thought.

By this point, susceptibility to emotional address will emerge as acrucial requirement for membership in the moral community, and thisrequirement would seem to demand that those susceptible in this waybe capable of empathy. A possible counterexample, though, and thethird case we will explore, is the high-functioning autistic adult, someoneseemingly capable of adhering to the right kind of moral reasons butwithout the capacity for empathy that seems necessary for having epi-stemic access to those reasons. To show how such autistics may be in-cluded in the moral community, therefore, I will develop an accountof the kind of empathy actually necessary for this component of moralagency—identifying empathy—and show how this is something of whichhigh-functioning autistic people are indeed capable.

Finally, we will look more closely at the relation between suscep-tibility to moral address and eligibility for moral responsibility by ex-amining the case of adults with mild mental retardation, a populationthat is simply not discussed much, if at all, in the philosophical literature.After taking some time to explain why their case is actually quite dif-ferent from that of children (contrary to what most theorists of re-sponsibility think), I will show that their susceptibility to emotional ad-dress from certain members of the moral community is enough toestablish their eligibility for (a limited range of cases of) moral respon-sibility. I will then demonstrate how this point generalizes to the rest ofus.

Before exploring these four cases, however, we need to understandboth the background view of moral responsibility and moral communityin play here, as well as the crude standard view of the conditions ofmoral agency I will be attempting to develop and refine herein. Let usbegin, then, with this expository material.

I. REACTIVE ATTITUDES AND MORAL REASONS

I will here be concerned with the notion of moral responsibility in thesense of “accountability,” which itself has a target, that is, one is ac-


countable to someone or other, namely, those eligible to hold one re-sponsible.3 I will also take as given the general approach to moral re-sponsibility set forth by Peter Strawson.4 Traditionally, the issue of moralresponsibility was raised in light of the threat of determinism: Is re-sponsibility possible if all events (including human actions) are causallydetermined by antecedent events, or is some form of indeterminismrequired instead? According to Strawson, though, as a matter of actualhuman practice, attributions of responsibility to one another are not—and could not be—undermined in any way by a belief in determinism.In other words, suspension of the “participant reactive attitudes” ex-pressed in holding someone responsible—attitudes such as resentmentand indignation, say—“is not the consequence of a theoretical convic-tion which might be expressed as ‘Determinism in this case,’ but is aconsequence of our abandoning, for different reasons in different cases,the ordinary inter-personal attitudes.”5 The practice of holding peopleresponsible is really an expression of moral attitudes, in other words,an expression of our concern that they live up to the basic moral demandfor goodwill in their actions (what I will refer to as the “basic demand”from here on out). When this expression is suspended, it is for one oftwo general reasons. On the one hand, the “offending” party may beexcused from responsibility in this particular case, either because shedidn’t really violate the basic demand (contrary to the initial interpre-tation of the action) or because she was justified in doing so (it was anemergency, say). Nevertheless, even if she is excused in this particularcase (where we suspend our reactive attitudes to her), she remains amember of the moral community, eligible for the reactive attitudes inother cases. On the other hand, the “offending” party may be exemptedfrom responsibility generally, given that she’s not an appropriate targetof the basic demand in the first place. This will be because the subjectis, according to Strawson, hypnotized, brainwashed, cognitively dam-aged, insane, a child, or otherwise temporarily or permanently inca-pacitated from being a functioning member of the moral community,“beyond the pale,” as it were, in terms of being able to engage in moraldialogue with the rest of us.6 And, on Strawson’s understanding of thematter, none of these excuses or exemptions is grounded on the truthor falsity of determinism.

3. I will, however, offer some reasons for why this is the right conception of moralresponsibility in Sec. III, titled “Moral Fetishism.”

4. Strawson’s is by no means an uncontroversial theory, of course. Nevertheless, it iseither accepted by, or at least not incompatible with, the views of the authors I will discusshere, so the fact that I am taking this particular theory for granted should cause nocontroversy with them.

5. Strawson, “Freedom and Resentment,” 82.6. Ibid., 77–80. See also Watson, “Responsibility and the Limits of Evil,” 223–25.


I want to set aside the determinism issue, however, to focus onStrawson’s remarks about the exemptions class, where we are left withsome important questions. First, Strawson draws the boundaries of themoral community to enclose those with whom we can intelligibly deploythe participant reactive attitudes, those generally defined as not exemptin the ways just articulated, those with whom we can engage in personalrelationships and successful moral address, and those whom we do nottreat “objectively.”7 What’s left open, though, are the precise positiveconditions rendering one in this nonexempt category. Second, howexactly does being nonexempt—whatever its conditions—render oneappropriately eligible for the reactive attitudes, for being morally re-sponsible at all? In other words, what is the precise nature of the re-lationship between membership in the moral community and moralresponsibility?

In addressing the first question, Strawson suggests that the natureof moral membership is complex but that there are nevertheless certainrecognizable features. First, it involves being able to enter into “inter-personal human relationships,” which include a wide range of feelingsand attitudes as constituent parts, among them “resentment, gratitude,forgiveness, anger, or the sort of love which two adults can sometimesbe said to feel reciprocally, for each other.”8 (But why are these sorts offeelings and attitudes a requisite key to moral membership? Strawsondoes not say; I will attempt to fill in this gap.) Second, one must beable to respond to certain sorts of reasoned appeals; someone incapableof this, someone to whom we hold an objective attitude, is someonesuch that “though you may fight him, you cannot quarrel with him, andthough you may talk to him, even negotiate with him, you cannot reasonwith him.”9

This way of putting it emphasizes the importance of reason-basedappeals, and it suggests an answer to the second question as well: if thepractice of holding responsible simply consists in the expression of thebasic demand via the reactive attitudes, then engaging in the practiceis intelligible only under the assumption that the target is capable ofunderstanding and responding to such demands. The expression ofresentment, after all, reflects “an expectation of, and demand for, themanifestation of a certain degree of goodwill or regard on the part ofother human beings towards ourselves.”10 Such expectation is simplyunintelligible, though, unless the entity one resents is capable of un-derstanding what goodwill is and requires and is also able to transform

7. See, e.g., Strawson, “Freedom and Resentment,” 79.8. Ibid.9. Ibid.10. Ibid., 84.


that understanding into an actual manifestation of it. The type of com-munication that is constitutive of the expressed reactive attitudes willbe pointless as a form of moral address when directed to those withoutthese epistemic and motivational capacities.11

But what precisely is involved in these capacities? In filling in theStrawsonian account, R. Jay Wallace offers the most detailed story, so Iwill take him as the paradigm case of the trend I will discuss in theremainder of this article.12 Wallace supposes first that, in expressing thereactive attitudes, we are demanding that the targeted agent complywith the moral obligations we accept, obligations supported by distinc-tively moral reasons. The question, then, is what conditions make it fairto demand that the targeted agent comply with these obligations? Wal-lace maintains that the theory offering the best explanation for ourvarious blaming practices is one presupposing the agent to have the“powers of reflective self-control,” consisting in both an epistemic com-ponent—“the power to grasp and apply moral reasons”—and a moti-vational component—“the power to control or regulate his behavior bythe light of such reasons.”13 Call this the “Moral Reasons–Based Theory”of moral agency, or MRBT.

The first element of MRBT, the ability to “grasp and apply moralreasons,” is, according to Wallace, rather complex. One must be ablenot only to understand the various concepts deployed in the moralprinciple in question but also to appreciate the justifications for theprinciple (or, if the principle is basic, one must be able to understandthat fact). The example Wallace considers is the principle of nonma-leficence. In order to grasp and apply the moral reasons expressed inthis principle, one must first understand the concept of “harm”—whatit means to harm someone generally and what kinds of specific treatmentwould count as harmful—and then one must appreciate the consider-ations bearing on what makes harming someone wrong. What this un-derstanding and appreciation provides is “the ability to bring the prin-ciple to bear in the full variety of situations to which it applies,anticipating the demands it makes of us in those situations, and knowingwhen its demands might require adjustment in light of the claims ofother moral principles.”14

Controlling one’s behavior in light of such reasons, however, is alsofairly complex, requiring three more particular capacities: (a) the ca-

11. See Watson, “Responsibility and the Limits of Evil,” 230–31.12. As already mentioned, others who agree with Wallace’s emphasis on the capacity

for grasping and applying certain sorts of reasons as the entry pass into the moral com-munity include Arpaly, Darwall, Fischer and Ravizza, Kennett, Sher, Watson, and Wolf.

13. Wallace, Responsibility and the Moral Sentiments, 157.14. Ibid.


pacity to critically reflect on one’s desires or impulses, assessing themin light of one’s moral reasons, (b) the capacity to make decisions thatdepend on one’s critical deliberations, and (c) the capacity to performactions that depend on one’s decisions.15

Reflective self-control, then, is quite sophisticated.16 But noticethat it does not require that one actually act on moral reasons to bemorally responsible; blaming anyone would be incoherent were thisthe case. Instead, it merely requires two capacities that may in factnever be exercised: a recognition-of-moral-reasons capacity and a mo-tivated-by-moral-reasons capacity. If we know an agent has both ca-pacities, therefore, MRBT implies that that is sufficient for us to viewsuch an agent as a full member of the moral community, eligible forboth interpersonal moral relationships and expression of the reactiveattitudes. Failing to have one or the other of these capacities, however,renders one ineligible on both counts. Putting this more formally,then, we have

MRBT Version 1:One is a member of the moral community, a moral agent

eligible for moral responsibility and interpersonal relationships,if and only if (a) one has the capacity to recognize and applymoral reasons and (b) one has the capacity to control one’s be-havior in light of such reasons.

Let us explore, then, how this theory fares with respect to the casesof psychopathy, moral fetishism, autism, and mild mental retardation.

15. Ibid., 158–59.16. I should point out that Wallace limits the Strawsonian connection between moral

responsibility and interpersonal relationships somewhat, citing Gary Watson’s discussionof the Robert Harris case (see Watson, “Responsibility and the Limits of Evil,” 235–48).Harris was a truly nasty man, someone who brutally murdered two teenagers and thencalmly ate the fast food they’d just purchased. If being a fair target of the reactive attitudeswere sufficient for rendering one eligible for interpersonal relationships, then Harris—whom we still blame (quite vociferously)—is someone eligible for interpersonal relation-ships, which seems quite unlikely: extreme evil typically disqualifies one from such rela-tionships. What Wallace maintains, therefore, is not that responsibility renders one fit forany old interpersonal relationship; instead, it renders one fit for moral relationships,defined by “successful exchange of moral criticism and justification” (Wallace, Responsibilityand the Moral Sentiments, 164). Of course, while one may not ever actually have to engagein this sort of exchange with the target of the reactive attitudes, it is enough that thetargets are at least candidates for this sort of exchange. Harris fits the bill here, even ifhe would not actually deign to reply to us: stony silence may still count as moral address,after all, just as a quiet, threatening form (see Michael S. McKenna, “The Limits of Eviland the Role of Moral Address: A Defense of Strawsonian Compatibilism,” Journal of Ethics2 [1998]: 123–42, esp. 132). But, in any event, for Wallace, there is a tight connectionbetween moral responsibility and moral relationships and both are dependent on theputative moral member’s having the capacity for reflective self-control.


II. PSYCHOPATHY

Most prominent defenders of MRBT point to paradigm cases of psycho-paths as agents who are outside the realm of the moral community, ex-empt from moral responsibility and to be treated “objectively.”17 Includedunder the same rubric, on Strawson’s analysis, would be young children,those with severe mental illnesses (e.g., schizophrenics), those with sys-tematically perverted minds, and compulsives (kleptomaniacs, perhaps,as well as those suffering from obsessive compulsive disorder).18 But whyis it that we suspend our reactive attitudes to these sorts of agents?

Before addressing that question, though, we need to get clear onjust what the nature of psychopathy is. It is now officially classified (along

17. Wallace does so explicitly on 177–78 of Responsibility and the Moral Sentiments.Strawson(“Freedom and Resentment”) includes as exempt those he variously calls “psychologicallyabnormal” (79), “warped or deranged” (79), and “moral idiots” (82), the last of which Watsonunderstands to be a reference to “being a sociopath” (Watson, “Responsibility and the Limitsof Evil,” 224). Jeannette Kennett, in Agency and Responsibility (Oxford: Oxford UniversityPress, 2001), allows that, while a true “sociopath” may be legally responsible, he is indeednot morally responsible (189, 209–14). Darwall (The Second-Person Standpoint) considers psy-chopaths to be exempt as well, although I will discuss his particular view in more detail inan upcoming footnote. Other defenders of MRBT are slightly more circumspect, but theytend to agree. Watson, while admitting that the “case of the sociopath is . . . complicated,”suggests that even so reactive attitudes are “pointless” with respect to them (“Responsibilityand the Limits of Evil,” 231 n. 12; on 239, he also clearly distinguishes the case of thepsychopath from the case of the murderer Robert Harris, whom he discusses in great detailas being a much more difficult, even paradoxical, case for the Strawsonian). Susan Wolfsuggests that, when it comes to eligibility for responsibility, sociopaths and normal humansare different (in Freedom within Reason, 152 n. 2), and she takes it as a basic pretheoreticintuition that her famous character Jo Jo (who has all the earmarks of a psychopath) is notresponsible (see Wolf’s “Sanity and the Metaphysics of Responsibility,” in Responsibility, Char-acter, and the Emotions, ed. Ferdinand Schoeman [Cambridge: Cambridge University Press,1987]). Fischer and Ravizza (Responsibility and Control, 78–80), along with Sher (In Praise ofBlame, 118 n. 3), agree that some psychopaths are exempt from moral responsibility, butthey also think that some are not. What’s important, though, is that they think that thepsychopath’s eligibility for responsibility corresponds directly to his ability to recognize andrespond to certain sorts of moral reasons, and that is all that is necessary for my point inthis section. Nevertheless, there are those who disagree with this general view, arguing insteadthat psychopaths are indeed morally responsible. For an early example, see Vinit Haksar,“The Responsibility of Psychopaths,” Philosophical Quarterly 15 (1965): 135–45 (although Hak-sar does admit that perhaps “when we know more about him, we may be able to show thathis responsibility was substantially impaired” [145]). Matt Talbert, in some unpublished work,also seems to think that psychopaths (the morally blind) are morally responsible. And whileT. M. Scanlon does not talk explicitly in terms of psychopathy, his stated view about warrantedmoral criticism suggests that he might agree (see What We Owe to Each Other [Cambridge,MA: Belknap Press of Harvard University Press, 1998], 287–90). Nevertheless, I will simplyadopt the view here that paradigm-case psychopaths are exempt, insofar as that is what theprominent defenders of MRBT do, in order to articulate what they take to be the preciseconditions of the exemption (and exemptions generally).

18. Strawson, “Freedom and Resentment,” 78.


with sociopathy) under the general term “anti-social personality disor-der,” which consists in “chronic antisocial behavior and violation of thelaw and the rights of others.”19 More specifically, the DSM-IV-TR statesthat the disorder is indicated when someone who is eighteen or oldermeets three of the following seven conditions (where these patterns ofdisregard have occurred since he or she was fifteen):

1. failure to conform to social norms with respect to lawful behaviorsas indicated by repeatedly performing acts that are grounds for arrest;

2. deceitfulness, as indicated by repeated lying, use of aliases, orconning others for personal profit or pleasure;

3. impulsivity or failure to plan ahead;4. irritability and aggressiveness, as indicated by repeated physical

fights or assaults;5. reckless disregard for safety of self or others;6. consistent irresponsibility, as indicated by repeated failure to

sustain steady work or honor financial obligations; and7. lack of remorse, as indicated by being indifferent to or rational-

izing having hurt, mistreated, or stolen from another.20

What, then, is the precise source of the psychopath’s exemptionfrom the moral community and moral responsibility? On MRBT Ver-sion 1, an entity is exempt insofar as it lacks one or both of the tworeasons-related capacities—either the recognition/application-of-moral-reasons capacity or the motivated-by-moral-reasons capacity—soobviously advocates of that view have to assert that the psychopath isimpaired in one of these ways. As it turns out, almost all MRBT theorists

19. Answers.com Web site, http://www.answers.com/topic/antisocial-personality-disorder.

20. American Psychiatric Association, Psychiatric News Web site, http://pn.psychiatryonline.org/cgi/content/full/39/1/25�a.While it may seem as if the bar forbeing diagnosed as a psychopath is set fairly low (only three of these conditions need bemet!), it is extremely important to the diagnosis that one exhibit repeated and chronicpatterns of these behaviors, which should be sufficient to distinguish genuine psychopathsfrom one’s deceitful and impulsive teenage neighbor. Nevertheless, psychopathy is esti-mated to be present in 4 percent of the U.S. population, so it is actually more prevalentthan anorexia (3.43 percent) and colon cancer (.04 percent); see Martha Stout, TheSociopath Next Door (New York: Broadway Books, 2005), 6–8. Of course, because the DSM-IV-TR conditions still seem as if they might let a number of garden-variety criminals throughthe net, there are those who disagree with the American Psychiatric Association’s incor-porating psychopathy under its more general “antisocial personality disorder” rubric andthink that, in order to make a distinction between ordinary criminals and psychopaths,we should preserve psychopathy as a separate disorder. See, e.g., the Wikipedia entry for“Psychopathy” (http://en.wikipedia.org/wiki/Psychopathy) for a discussion of this issue.I, however, am interested in what I have called the “paradigm case” psychopath, an in-dividual whom we can assume exhibits all, or nearly all, of the seven symptoms and sowill surely be distinct from the common criminal.


insist that what the psychopath lacks is some form of the first capacity.21

Thus Wallace, for instance, claims that the psychopath lacks the ap-plication aspect of that capacity: “It has been suggested that psycho-paths lack the qualities of imagination and practical understandingrequired to bring common moral principles to bear in new cases; forinstance, they often have great difficulty distinguishing between trivialand important moral concerns, and so lack the capacity to engage inintelligent critical reflection on moral issues.”22 Darwall, on the otherhand, seems to emphasize the other aspect of the first capacity, sug-gesting that the psychopath fails to recognize the moral reasons beingoffered to him in the first place.23 Fischer and Ravizza concur, insistingthat, insofar as the psychopath is not responsible, it is because he isnot receptive to moral reasons, failing to “recognize that other indi-viduals’ claims can generate sufficient reasons for action for him.”24

Finally, Watson agrees as well, directly implying that psychopaths ex-hibit a “lack of understanding” (presumably moral understanding).25

None of these theorists, then, specifies the source of the psycho-path’s exemption to be an incapacity for acting on, being motivated by,moral reasons.26 Instead, they all point exclusively to a purely cognitive,

21. This claim is contrary to Susan Dwyer’s assertion that “almost to a one, phi-losophers writing about psychopathy typically assume that . . . he lacks moral mo-tivation” (Susan Dwyer, “Evil and Moral Competence: What Psychopaths Can TeachUs,” Philosophers’ Magazine 9 [2000]: 32–33, 32, http://www.umbc.edu/philosophy/dwyer/papers/whatpschopathscanteachus.html). With respect to the philosophers Ihave cited to this point, though, Dwyer’s claim is just false. One possible exceptionare Fischer and Ravizza, but interpretation of their view is tricky, and so I’ll havemore to say about this possibility later on.

22. Wallace, Responsibility and the Moral Sentiments, 178. See also Jonathan Glover, Re-sponsibility (London: Routledge, 1970), 138, 177–78.

23. Darwall, The Second-Person Standpoint, 89. However, a word of caution is in orderhere. While Darwall explicitly takes himself to be discussing psychopaths and their “imper-viousness to moral demands” (89), matters are made more complex by the fact that he takesa paradigm case of psychopathy to be that of Robert Harris, the California murderer discussedat length by Watson. But it’s not at all clear that Harris was a psychopath (although Arpalylabels him one as well, claiming as support that Harris was “utterly unresponsive to moralreasons” [Arpaly, Unprincipled Virtue, 157, 170]). I actually think of Harris as a paradigm caseamoralist, someone who is entirely capable of acting on moral reasons but just doesn’t. Atany rate, not much hangs here on this point, but it is at least worth mentioning that Darwall’scategorization of Harris isn’t uncontroversial (e.g., Watson himself explicitly asserts thatHarrisis unlike psychopaths, noting that, while the latter lack moral understanding, Harris, bycontrast, merely “exhibits an inversion of moral concern” [Watson, “Responsibility and theLimits of Evil,” 239]).

24. Fischer and Ravizza, Responsibility and Control, 79.25. Watson, “Responsibility and the Limits of Evil,” 239.26. Indeed, Fischer and Ravizza even allow that, as long as the psychopath could rec-

ognize moral reasons and as long as he were weakly reactive to nonmoral reasons, then hewould actually be morally responsible, even if he weren’t weakly reactive to moral reasons.See Fischer and Ravizza, Responsibility and Control, 79–80.


epistemic deficiency. I disagree, and the basic argument I will advanceagainst this common view goes as follows. If we take the two conditionsof MRBT—the epistemic and motivational conditions—as exhaustive ofthe conditions of eligibility for moral responsibility and we begin withthe assumption that the psychopath is not eligible for moral responsi-bility, then the explanation for this assumption must be that he cannotsatisfy one or both of the conditions of MRBT. But the psychopath cansatisfy the epistemic condition, so he must not be able to satisfy themotivational condition. Further, insofar as the psychopath is psycholog-ically just like the rest of us except in virtue of his incapacity for complexemotions (e.g., guilt, shame, and remorse), this incapacity must be thesource of his motivational incapacity.

There are two main points of the argument in need of explanationand defense: (a) Why think the psychopath can satisfy the epistemiccondition? (b) Why think the psychopath lacks the capacity for complexemotions? Both are essentially empirical questions, but our discussion ofthese points will move us beyond purely empirical matters rather quickly.Consider the first point. Wallace explicitly claims that the psychopathlacks a purely cognitive sort of capacity, namely, the ability to figure outwhen moral principles apply in different cases. But, on most accounts ofpsychopathy, the psychopath is most assuredly able to do that much. NoteJeffrey Murphy’s description, for example: “Unlike the psychotic, the psy-chopath seems to suffer from no obvious cognitive or volitional impair-ments. He knows what he is doing (he has no delusions); and, since hetypically does just what he wants to do, it would be odd to call himcompulsive or to claim that he acts on irresistible impulses. . . . Psycho-paths know, in some sense, what it means to wrong people, to actimmorally.”27

Antony Duff also notes that psychopaths are “not intellectually in-competent,”28 and Herbert Fingarette agrees: “He knows the probableconsequences of his acts, social and antisocial, but does not care—veryoften does not care even enough to try to protect his liberty by con-cealing his doings; he has no pride to protect. . . . It would be mis-leading to call the psychopath morally incompetent or morally blind.He can, if he wishes, carry on a most intelligent, ‘insightful,’ and per-suasive moral discussion.”29

Finally, the psychopath’s ability to recognize and apply moral rea-sons should also be evident from his portrayal in various movies and

27. Jeffrey Murphy, “Moral Death: A Kantian Essay on Psychopathy,” Ethics 82 (1972):284–98, 285–86.

28. Antony Duff, “Psychopathy and Moral Understanding,” American Philosophical Quar-terly 14 (1977): 189–200, 192.

29. Herbert Fingarette, On Responsibility (New York: Basic, 1967), 26.


books on the subject. The Bruno Anthony character, in Hitchcock’sStrangers on a Train,30 is certainly an example of a psychopath capableof moral interlocution all day long, as are other fictional psychopathslike Hannibal Lecter, Alex DeLarge (from A Clockwork Orange), Tom“The Talented Mr.” Ripley, Patrick Bateman (from American Psycho), andeven Eric Cartman (from South Park).

Such psychopaths can, it seems, grasp and apply moral reasons andprinciples, so this cannot be the source of their exclusion from themoral community, as Wallace and others would have it. After all, theycan understand the consequences of their actions, engage in moralconversations, and grasp general moral principles. Now they may stilllack reflective self-control, of course, but not for the reasons Wallaceadvances. Instead, it must be, as Murphy emphasizes,31 the motivationalcapacity that’s missing, not any particular cognitive abilities: “Thoughpsychopaths know, in some sense, what it means to wrong people, toact immorally, this kind of judgment has for them no motivational com-ponent at all.”32

30. I owe the example to John Deigh, in “Empathy and Universalizability,” Ethics 105(1995): 743–63, esp. 743.

31. Murphy, “Moral Death,” 286. See also Shaun Nichols, Sentimental Rules (Oxford:Oxford University Press, 2004), 72: “A psychopath can be fully rational and judge that someaction is morally required without being motivated to do it.” Nichols goes on to cite a studyhe conducted that suggests that this may be a platitude of folk morality rather than the viewthat psychopaths make moral judgments only in the “inverted commas” sense, which is takento be the case by conceptual rationalists. Note also Stout’s remarks: “It is not that this group[psychopaths] fails to grasp the difference between good and bad; it is that the distinctionfails to limit their behavior” (The Sociopath Next Door, 9).

32. One might wonder how this analysis conforms to determinations of legal respon-sibility for psychopaths. Attempts to get psychopaths excused from punishment for theircrimes typically involve showing that they are insane, that psychopathy is a form of mentalillness rendering them unable either to understand the nature of their actions or to un-derstand the difference between right and wrong. The difficulty, of course, is in establishingeither disability, but in any event the thought is that psychopaths can’t be excused from legalresponsibility unless they have some sort of cognitive disorder. If their disorder were solelymotivational, after all, this wouldn’t serve to excuse them, for if they were at least able toconform their conduct to the law—regardless of their reasons for doing so—this would besufficient to establish their legal responsibility for failing to do so. The worry, then, mightbe that if the conditions for moral responsibility are essentially the same as those for legalresponsibility, I have, by focusing on the psychopath’s motivational deficiencies, given thewrong account of his being excused (especially since I have denied that he necessarily suffersfrom any cognitive deficiencies). Alternatively, if I am assuming that the conditions for moralresponsibility are not the same as those for legal responsibility, this might be viewed as asignificant departure from traditional understanding, in which case I owe the reader anargument for such a nonstandard view. (I am grateful to an anonymous associate editor forpressing this point.) I take the latter view here, but I don’t think it is much of a departurefrom the traditional view, nor do I think a great deal of argument is necessary to defend it.We hold one another morally responsible for a much larger set of things than we do legallyresponsible. For instance, beyond blaming, we praise people for the good things they do,


Turn, then, to the second issue: What is the source of this moti-vational incapacity? It is the psychopath’s particular sort of affectivedeficiency. This theme emerges throughout the psychological and phil-osophical literature as well. Start with Duff, who points out that psy-chopathy “involves an incapacity for such emotional and moral re-sponses as love, remorse and concern for others.”33 These are complex,mature emotions he is missing, as Fingarette also makes clear:

His feelings are various and can be strong, but are evanescent and“superficial.” Vexation, spite, quick, and labile flashes of quasi-af-fection, peevish resentment, shallow moods of self-pity, puerile at-titudes of vanity, absurd and showy poses of indignation are allwithin his emotional scale; however, mature wholehearted anger,true or consistent indignation, honest, solid grief, sustaining pride,deep joy, genuine despair are reactions not found. . . . Inexplicableminor cruelties, humiliations, inconveniences are dealt out freelyby him. . . . At times he very convincingly expresses “remorse” or“moral conversion”—though he may concurrently remain activelyengaged in his usual conduct. If this performance is exposed, heis quite unembarrassed, since for him there is nothing morally atstake. He does not care; hence no shame.34

I want to focus here on what I take to be the two crucial incapacitiesof the psychopath: (a) an inability to feel guilt, shame, or remorse, and(b) an utter lack of concern for the feelings of others.35 These are the

whereas criminal responsibility is simply about the assessment of legal blame and punishment.We also hold people morally responsible for their character, blaming or praising them fortheir vices or virtues, whereas there is no such analogous category in legal responsibility.Finally, and most important, even if we focus solely on blame for actions (so the case is mostclosely analogous to the criminal case), moral blame includes an assessment of whether ornot one did the right thing for the right reasons, whereas this last bit is (mostly) irrelevantfor criminal responsibility. In other words, in holding people accountable for their actions,we express the basic demand for goodwill, but that is clearly a demand for more than thatone simply conform one’s actions to the moral law; it is also a demand that one do so basedon the right sort of motive, a motive of goodwill (or at least not ill will). My point in thissection, then, is that if one is incapable of acting on such a motive, one is exempted frommoral responsibility, even if one is capable of conforming one’s conduct to the moral lawfor other reasons. In the next section, I will say more about the nature of the right reasons.

33. Duff, “Psychopathology and Moral Understanding,” 191.34. Fingarette, On Responsibility, 24–26. For an interesting discussion of the psychological

literature, which is in almost uniform agreement on these points, see Stout, The SociopathNext Door.

35. See also the Wikipedia entry (http://en.wikipedia.org/wiki/Psychopathy): “A psy-chopath is defined as having no concerns for the feelings of others and a complete disregardfor any sense of social obligation. They seem egocentric and lack insight and any sense ofresponsibility or consequence. Their emotions are thought to be superficial and shallow, ifthey exist at all. They are considered callous, manipulative, and incapable of forming lastingrelationships, let alone of any kind of love.” See also Stout: “Sociopathy [what we are calling“psychopathy”] is an aberration in the ability to have and to appreciate real (noncalculated)


aspects of the psychopath that distinguish him from us, so insofar as heis not eligible for moral responsibility, and insofar as it is the motivationalcomponent of MRBT Version 1 he is lacking (because he is capable ofthe epistemic component), the best explanation for this lack are theseincapacities. But why would being unable to feel guilt, or the failure tobe concerned for the feelings of others, lead to his being unable totranslate moral reasons into action?

I have elsewhere argued that caring for some X consists in thedisposition to experience mature, complex emotions corresponding tothe up-and-down fortunes of X.36 Roughly put, when I care about aperson, for example, I am disposed to respond with sadness, grief, ordespair when something bad happens to her and with elation, delight,or amusement when something good happens to her: that’s just whatcaring about her consists of. So if one is incapable of experiencing thesesorts of emotions regarding the fortunes of some object, this just meansthat one is incapable of caring about that object.

Furthermore, when it comes to (standard) nonwanton action—action preceded by some concern and critical reflection over the sourceof one’s will—motivation (both moral and nonmoral) typically dependson caring: my “winning” motive (a desire, say) to do A depends on mycaring about something that Aing will help preserve or promote (orthat failing to A will betray).37 If one lacks a care for X, on the other

emotional experience, and therefore to connect with other people within real (noncalcu-lated) relationships” (The Sociopath Next Door, 126).

36. See my “Caring, Identification, and Agency,” Ethics 114 (2003): 88–118. It might bethought that psychopaths—at least those homicidal ones we occasionally read about—canmeet this general definition insofar as they might experience delight in the suffering of theirvictims, so at least they would experience emotions in response to the fortunes or misfortunesof others, just the opposite of the emotions experienced by those who have benevolentconcern for others. (I am grateful to an anonymous referee for raising this point.) If it istrue delight, though, or some other complex emotion that this person is feeling at themisfortunes of his victims, he is not a psychopath in the common psychiatric sense underdiscussion, according to which psychopaths are simply defined in terms of their lack ofemotional capacity or at least their severe emotional shallowness. As Stout puts it, “The onlyemotions that sociopaths seem to feel genuinely are the so-called ‘primitive’ affectivereactionsthat result from immediate physical pain and pleasure, or from short-term frustrations andsuccesses” (The Sociopath Next Door, 127). (In addition to the above citations, see HerveyCleckley, The Mask of Sanity, 5th ed. [St. Louis, MO: Mosby, 1988], esp. 92, where Cleckleytalks about psychopaths’ “emotional emptiness.”) Indeed, on this understanding,psychopathsdo not even care about themselves: they fail to engage in prudential planning, they arepersistently indifferent to the prospects of imprisonment, they take and quit jobs on a whim,etc.; in general, they simply act on whatever egoistic impulse is strongest without having anyworries about any emotional ramifications that might stand in their way. See Stout, TheSociopath Next Door, 50.

37. I argue for this point in some detail in “Caring, Identification, and Agency,” esp.90–104. Exceptions include the actions of addictives, compulsives, and those with severecognitive impairments.


hand, one will not, other things equal, be motivated to perform any X-preserving or X-promoting actions; one will not be moved by X-reasons,as it were. And the converse holds true as well: if one fails to have anymotivation to act in accordance with X-reasons and X is something it ispossible to preserve or promote, then the explanation, others thingsbeing equal, is that one just does not care about X. Consequently, itmay seem as if the psychopath’s incapacity to be moved by moral reasonswould be a natural implication of his failing to care about morality. Nowthis explanation wouldn’t rule out his still being moved to act in a waythat looks moral, but his motivational ability in such cases will have itssource in his self-interested desires that acting like a moral person willhelp satisfy. So while he may act on reasons in conforming his actionsto the basic demand, he will not be acting on moral reasons.38

All of this might go to suggest, then, that we revise MRBT’s moti-vational condition to require that one be capable of caring about mo-rality and so be capable of being motivated to act on its moral reasonsas a result. Nevertheless, this revision would not yet capture what isdistinctive about the psychopath, for what he fails to care about isn’tnecessarily morality but other people.39 But if he fails to be motivatedto act on moral reasons, if care attaches to motivation as described above,and if the only relevant psychological distinction between him and therest of us is that he is incapable of caring about other people, thencaring about other people must be necessary for being motivated to acton moral reasons. In other words, other people must matter to you ifyou are to be able to act on the moral reasons attached to their ex-pression of the reactive attitudes. What looks to be the case, then, isthat moral reasons must get their normative motivational power fromtheir source in other people.

Think about it this way: there are surely other normative demandsto which the psychopath presumably can respond, so what is so specialabout the moral demand (and the moral reasons it provides) that ren-ders the psychopath incapable of responding to it? The best explana-

38. In an episode of the series South Park, the clearly psychopathic Eric Cartman triesdesperately to figure out how to make his “friend” Kyle like him so he’ll take him to anamazing Mexican restaurant he’s been dying to visit. He seeks advice from a disabled kidthat everyone likes and asks him, “How can I act nice so Kyle will like me?” The reply is thatthe best way to act nice is simply to really be nice. Cartman’s reply sums up the world of thepsychopath perfectly: “OK, so how can I act so he’ll think I really am nice?” For less “cartoony”support for this point, see the Wikepedia discussion of psychopathy at http://en.wikipedia.org/wiki/Psychopathy#_note�3: “Psychopaths understand that society expects them to be-have in a conscientious manner, and therefore they mimic this behaviour when it suits theirneeds.” Stout (The Sociopath Next Door) makes this point repeatedly as well.

39. As Stout puts it, “An obligation of any kind is something one feels toward beings,or toward a group of beings, who matter emotionally. And to a sociopath, we simply do notmatter” (The Sociopath Next Door, 126).


tion—and the explanation Wallace and the other MRBT advocates ig-nore—is the fact that the psychopath cannot care about the people hisactions affect in the way necessary to give rise to a care about their basicdemand and thus to the subsequent motivation to adhere to the moralreasons they provide. In other words, the psychopath simply cannot careabout the basic demanders, so their basic demand must strike him asboth puzzling and utterly irrelevant to the living of his life, and this iswhy expressions of reactive attitudes to him would thus be particularlypointless.

We are led, therefore, to our first significant revision of the con-ditions for moral membership:

MRBT Version 2: One is a member of the moral community, amoral agent eligible for moral responsibility and interpersonal re-lationships, if and only if (a) one has the capacity to recognize andapply moral reasons, and (b) one is capable of being motivated by thosemoral reasons because one is capable of caring about their source, namely,the moral agents affected by one’s actions.40

Nevertheless, couldn’t someone be motivated by moral reasons withoutcaring about any actual moral agents? Exploring this possibility will yieldanother important refinement to the theory.

III. MORAL FETISHISM

In this section I want to consider a couple of hypothetical exceptionsto MRBT Version 2. I suspect that there are no actual people like thoseI will describe, but they remain possible counterexamples to MRBT, andat any rate we can learn a great deal about the various types of moralreasons there are, as well as which ones do and do not matter for ourpurposes, by consideration of these cases. That is, by exploring hypo-thetical cases of those who are not capable of being moved by a certaintype of moral reasons, we will gain important insight into what therelevant type of moral reasons is, as well as what it takes to have accessto and be moved by it.41

There are a couple of different forms that this counterexample couldtake. On the one hand, there might those who care about morality, and

40. Duff seems to suggest a criterion roughly like this one, although he emphasizesonly the psychopath’s lack of concern for moral values. See his “Psychopathology and MoralPersuasion,” 198.

41. The similarity of my position in this section to the recent view about second-personalreasons espoused by Stephen Darwall will be obvious. In many ways, what I say here hasindeed been influenced by Darwall’s work, but I intend my remarks in this section to be,for the most part, freestanding and plausible in their own right via consideration of thevarious types of fetishists I introduce. In addition, I take a sharp detour from the work ofDarwall in the final part of this section, where I emphasize the crucial importance of emotion-based caring to the picture, something that Darwall explicitly resists.


so act on some recognizably moral reasons, but who nevertheless fail toact on the specific type of moral reasons relevent to moral responsibility.In other words, they would be people who fail to act on the moral demandin virtue of its being a demand, an authoritative claim on them by themembers of the moral community. Further, their failure to act on thisdemand would be explained by their failure to care about those membersof the moral community in the first place. On the other hand, theremight be people who care about morality and doing the right thing, butsolely in virtue of its being the right thing, so they act in accordance withthe basic demand, yet they do so without caring about the basic de-manders—the members of the moral community—at all.

Consider the first general type, what I will call the non–demand fetishist.There are two different kinds of this fetishist: the first I will call an agent-neutral fetishist, and the second I will call an agent-relative fetishist. The agent-neutral fetishist is someone who is motivated to adhere to the norms ofthe basic demand for goodwill but who does so for exclusively agent-neutral reasons. So the reason he refrains from stealing your money whenhe needs it and could get away with it, say, is “that it would underminea person’s interests” or “that it would fail to respect someone’s autonomy”or “that it would fail to maximize utility” or, most generally, “that it wouldfail to instantiate goodwill,” something it would be right for anyone inhis position to refrain from doing. Now one way in which he might cometo recognize such a reason would be via his sympathy for you, but this isnot the sort of fetishist I have in mind, for what should be obvious reasons.Instead, I want to focus on the possibility of someone who comes to acceptthis sort of reason without affect or care, perhaps simply because he takesit to be a demand of pure Reason or because it passes some sort ofuniversalizability test. The people he thus helps (or refrains fromharming) are not, for him, the source of the basic demand’s normativity;instead, their making of the demand or their position as demandersgenerally are irrelevant to his doing what he does.

The agent-relative fetishist, on the other hand, is also motivated toadhere to the norms of the basic demand, and while he does take thesenorms to be agent relative, he also does not take them to stem fromthe demands of moral agents. In other words, he accepts that he hasreason to refrain from stealing from you, and his reason for refrainingis “that I would be undermining your (or a person’s) interests and thuswould fail to instantiate goodwill,” but he denies that that reason hasits source in your authority as a member of the moral community todemand such goodwill from him.42 And again we shall focus on the

42 See Darwall, The Second-Person Standpoint, 6–8. See also his unpublished paper, “Au-thority and Second-Personal Reasons for Acting,” University of Michigan, Department ofPhilosophy, 2006.


version of such a character who comes by his agent-relative moral rea-sons without caring about any of the people involved.

We would, I suggest, think there is something wrong with both ofthese sorts of non–demand fetishists, but what exactly? It is that neithersort of person would take himself to be responsible to us. He would bemotivated to adhere to moral norms that are (at least coextensive with)the norms of the basic demand, but he would fail to recognize ourclaims on him to do so, or our rights against his refraining from doingso. Ultimately, then, he would fail to respect our authority as persons,as members of the moral community, to demand these things of him.43

We can see this rather clearly when we think about cases in which hefails to do what he takes himself to have moral reason to do. Suppose,to borrow an example of Darwall’s, that this person steps on my foot.Let us further stipulate that he is an agent-neutral fetishist and that themoral reason he has failed to abide by in not avoiding my foot is “thatit would cause someone pain.” Suppose, then, that I expressed resent-ment toward him, perhaps simply through an angry cry of pain. Forhim, this cry would serve only as a reminder, a signpost, that he hadfailed to adhere to his (agent-neutral) moral reasons. And if he felt thatmoral reasons should override in situations like this, then, while he mayexperience regret at having failed to do the right thing, he would notfeel remorse for his hurting me; while he may feel shame at havingfailed to live up to his moral ideals, he would not feel guilt for havingcaused me pain. In his eyes, of course, he would not have failed anyonebut himself; in our eyes, however, he would most certainly have failedme, his victim. There are default expectations we have of one another,and while these expectations may be coextensive with agent-neutralnorms, or even agent-relative norms, certain moral emotions (remorseand guilt, say) are coherent only if one takes those norms as havingtheir source in the expectations of actual persons, the fellow membersof one’s moral community. Someone without these emotions, however,a guiltless and remorseless person, would be just as foreign to us, justas excluded from our moral community and eligibility for interpersonalrelationships, as a psychopath (more on this comparison below).

The closest analogue for the rest of us is found by thinking aboutour relation to something like the environment. Suppose that I haveagent-neutral moral reasons to avoid spraying graffiti on a part of theGrand Canyon, say, but I do so anyway. Could I coherently feel remorseor guilt for doing so? Only with respect to how I have failed my fellowhuman beings; if there were no other moral agents, entities to whomI would be answerable for what I have done, these moral emotions would

43. Again, see Darwall’s The Second-Person Standpoint, e.g., 8, and his “Authority andSecond-Personal Reasons for Acting.”


simply make no sense. I cannot coherently feel remorse directed to thecanyon for what I have done to it; any remorse I do experience will bedirected to the other members of my moral community, insofar as Ihave let them down in some way. The agent-neutral and agent-relativenon–demand fetishists, then, when it comes to their dealings with us,are like the rest of us are when it comes to our dealings with the en-vironment, taking themselves not to be answerable to the affected partiesfor what they do and in so doing setting themselves outside the bound-aries of the moral community.

Turn now to consideration of the second general type of fetishist,what I will call the morality fetishist.44 This is a person who cares aboutmorality, who wants to do the right thing, but only in virtue of its beingthe right thing to do.45 This person thus does not respond to moralreasons when he acts. Moral reasons are facts capable of generating amoral ought, capable of determining (in the absence of countervailingconsiderations) what the right thing to do is. But the fact that A is theright thing to do is not itself a moral reason to do A. If so, it wouldhave to supply an additional moral reason to do A, once all the moreparticular moral reasons for A were in. But that would be quite im-plausible.46 The fact that A is the right thing to do simply consists ofthe facts (the moral reasons) x, y, and z that make A the right thing todo. To think that the fact that A is the right thing to do constitutes anadditional moral reason to A is to mistakenly think of this fact aboutmoral rightness as not wholly consisting in these more particular moralreasons, as being some sort of separately obtaining moral fact, whichborders on the incoherent.

Nevertheless, the morality fetishist does take himself to have reasonto A solely because it is the right thing to do. His reason for Aing, then,must be nonmoral (likely prudential, but perhaps aesthetic). In doingA, then, he is not responding to any moral reasons at all. But this iswhat makes his attitude toward morality a fetish, for he is devoted tohis object of worship for the wrong reasons.

Neither general sort of fetishist, then, responds to the right reasons.The non–demand fetishist (of both stripes) doesn’t respond to the rightmoral reasons, whereas the morality fetishist doesn’t respond to moral

44. This is the fetishist most closely akin to the one Michael Smith introduces in TheMoral Problem (Oxford: Blackwell, 1994), 73–76. There has been considerable reaction tothis case in the literature, but it is simply beyond the scope of this article for me to engagewith it here.

45. So, while there may be those who care both about morality as such and about themoral demanders and the particular moral reasons stemming from them, I am concernedhere only with the person who cares exclusively for morality as such, i.e., the fetishist.

46. I am grateful to Doug Portmore for emphasizing the need to conceive the moralityfetishist in this way.


reasons at all. What both are missing is a response to the particular sortof moral reasons derived from the basic demand conceived as a demand,namely, the valid second-personal reasons potentially addressable by themembers of the moral community.47 But what this also means is thatneither fetishist would take himself as being responsible to these basicdemanders; instead, insofar as he took himself to be responsible at all,it would be to an ideal, to the moral norms, or to Reason itself. Butthis simply isn’t the type of responsibility—accountability to—demandedvia the expression of the reactive attitudes. For the reactive attitudes toplay the role they do in moral responsibility, the reactor must have anessential authority in deploying them: he is expressing resentment toanother person, after all, and this is an expression nearly everyone takesseriously in and of itself. If, however, this authority were entirely deriv-ative, stemming from the “true” authority of rational consistency, say,or Morality generally, then in expressing my resentment to you I wouldmerely be a kind of tattletale, my expressions taken seriously only intheir role of drawing your attention to the fact that you have donesomething wrong, but I would remain merely a medium between youand Morality, and our exchange would thus fail to be an interpersonalexchange at all. Your proper initial response to wronging me would thusbe “Oh, thanks!” (for the reminder), not “Oh, sorry!” Indeed, if thiswere all my role consisted of, I could just as easily be replaced by adisinterested observer, or by your conscience, or by a theoretical con-clusion on your part, or by a computer-operated buzzer, or by a nagginglittle robot. Resentment consists in something more than this, however.It indicates not just that you have done something wrong but that youhave wronged me and that this bothers me and so ought to bother youas a direct result. My authority, then, stems from my being the wrongedparty,48 and so it couldn’t come from my being the mere site at whichwrongdoing occurred. I hold you responsible, via my expression of thereactive attitudes, and for you to take this as seriously as it is appropriatefor me to expect—for this form of address even to be intelligible atall—I have to matter to you. Our fetishists, though, insofar as the de-mand-expressers are unimportant to them, pursue their slavish devotionto morality absent the emotional commitment to the source of thespecific type of responsibility reason that renders their responsiveness

47. See Darwall, The Second-Person Standpoint.48. Matters are slightly more complicated in the third-person analogue case, where I

react with indignation over your treatment of someone else, but, as will become clear later,my authority in this case arises insofar as I empathize with my fellow harmed moral agent,so what you are doing to her is, in an important sense, something you’re doing to me aswell. Indeed, it actually seems to me that the phenomenology of resentment is qualitativelyidentical to that of indignation, suggesting that the difference is determined solely by whetheror not I was in fact the target of the original harm.


to that reason possible. It is thus in divorcing the demand from thedemander—failing to see that what makes moral demands important isjust the importance of the moral demanders—that our fetishists fail torespond like normal moral agents.

Are fetishists nevertheless morally responsible members of themoral community? This is, of course, an exotic, rather precious, case,but we may speculate a bit to make a more general point about actualcases. If these fetishists are truly incapable of acting on the specific sortof second-personal moral reasons at issue, it is hard to see how theywould be morally responsible. Instead, as mentioned earlier, they wouldbe just like psychopaths, stricken by a similar sort of tunnel vision, albeitwith respect to a different set of goals.49 The psychopath is focused onhis own immediate needs or desires, so the general reasons of moralitydo not matter to him and thus will always lose out when they conflictwith his needs or desires (as they often do). The fetishist, in contrast,does care about morality—he is obsessively devoted to it, after all—buthere it is the second-personal reasons of his fellows that do not matterto him, and thus it is those reasons that will always lose out when theyconflict with what he takes to be the demands of morality generally.Insofar as he would be a kind of bloodlessly calculating “morality ma-chine,” then, his incapacity for the kind of interpersonal relationshipsconstitutive of membership in the moral community would leave him,as with the psychopath, external to it.

How precisely is this case to be handled on MRBT though? It maybe implausible, as it was with the psychopath, to think that the fetishistwould fail to be able to recognize or grasp the relevant moral reasons.Nevertheless, it might be possible that, upon being presented with therelevant second-personal reasons, the fetishist is simply puzzled by them,in which case his failure would be epistemic. However, it may be difficultat first to see why the fetishist wouldn’t be able to respond to second-personal reasons if he could indeed recognize and grasp them. Afterall, being motivated by reasons whose source is solely the authority ofhis parent, boss, or leader of some other kind may be perfectly possiblefor him. So, if he were capable of responding to reasons generally, tosecond-personal reasons generally, and (in the case of the non-demandfetishist) some (non-second-personal) moral reasons specifically, whycouldn’t he also respond to the second-personal moral reasons specificto the basic demand and moral responsibility? The answer would besimple if he indeed suffered from a motivational incapacity: because hecannot care about the people whose authority is the source of thesesecond-personal reasons, he cannot care about the reasons they provide,and thus he cannot be motivated by consideration of such reasons (they

49. I am grateful to Doug Portmore for suggesting this way of putting it.


simply have no motivational grip on him). Thoughts of failing the de-manders produce no guilt in him, so such thoughts cannot be a mo-tivational source nor can his desire to avoid other kinds of emotionalpain associated with the failure or betrayal of someone he cares about.It is caring about the demanders, it seems, that is necessary for beingmotivated.

It might seem to be possible, however, for one to respond to thesecond-personal reasons relevant to moral responsibility solely in virtueof one’s respect for the authority of the demanders, where such respectis not an instance of the emotion-based caring I have discussed.50 Nev-ertheless, this possibility still fails to capture something essential to moralresponsibility and interpersonal relationships. Suppose that such an agentwere to respond to my expression of the basic demand solely on the basisof his respect for me (without having any emotional investment in meor the other members of the moral community). If he were to violate thebasic demand as expressed by me, he would presumably feel no guilt norwould he feel remorse were he to harm me in so doing. But that is stilla crucial failure all its own, and the reason is that, in addressing him viathe expression of reactive attitudes (either as the victim of his wrongdoingor simply as an indignant observer), I am not only presenting and drawinghis attention to the relevant second-personal reason for ceasing his currentwrongdoing. I am also inviting him to feel what he has done in wrongingme, inviting him to feel my pain as if it were his own, in a direct effortto motivate him to stop doing it (or to not do it in the future). In otherwords, the moral address that takes place via expression of the reactiveattitudes has two aspects, only the first of which has been noticed inthe literature: (a) it is an address of (second-personal) reasons, provid-ing a certain sort of claim against his wronging me, say; and, equallyimportantly, (b) it is an emotional address, urging the wrongdoer to feelwhat I feel as a result of his wrongdoing and then subsequently to feelthe guilt or remorse (at having caused that feeling) which I expect tomotivate him to cease his wrongdoing.51 Thus, while we could at leastcontrol the behavior of the respecter-sans-care by appeal to second-personal moral reasons, this would involve no more than having anobjective attitude toward him, and because he would be impervious tothe emotional aspect of our moral address, he would fail to be able to

50. I am grateful to Stephen Darwall for impressing the importance of this point onme. Note also his explicit distinction between respect and care in The Second-Person Standpoint,126–30.

51. The third-person analogues of the reactive attitudes work in the same way: as anobserver of his wronging someone else, say, I react with indignation, which serves to expressto him (a) a second-personal reason for ceasing the wrongdoing (or not doing it in thefuture) and (b) an invitation to feel what he’s done to the victim. I will have more to sayabout the relevance of this point in the final section.


engage with us in a fundamentally personal way. But if he lacks thecapacity for this sort of personal, emotional engagement, he thus lacksthe capacity for being responsible to us in the relevant way.

Of course, there could be some such fetishists and respecters-sans-care whom we would hold morally responsible, viewing them as full-fledged members of the moral community. If this were the case, though,it would have to be because we believe them to still be able to respondto both aspects of moral address and thus be capable of caring aboutthe addressers (or potential addressers) nonderivatively and capable ofbeing motivated by the right sort of moral reasons and emotional ap-peals if only they were to come to care about the source of those reasonsand appeals.

These remarks, then, suggest another important revision to ourgeneral criterion:

MRBT Version 3: One is a member of the moral community, amoral agent eligible for moral responsibility and interpersonal re-lationships, if and only if (a) one has the capacity to recognize andapply second-personal moral reasons and (b) one is capable of beingmotivated by those second-personal moral reasons because one iscapable of caring about their source (i.e., the affected party [orparties]), such that one is susceptible to the moral address expressible viathe reactive attitudes in both its reason-based and emotional aspects.

Nevertheless, there are two groups of humans who may seem to be in-capable of this sort of moral address but that we would nevertheless wantto retain within the moral community. It is to them that we now turn.

IV. AUTISM

What we have seen to this point is the necessity for revisions to MRBTinduced by consideration of a couple of outlier cases, the psychopathsand fetishists excluded from the moral community primarily for theirlack of the proper motivational component, their failure to be sus-ceptible to full-fledged moral address, and their inability to enter intogenuine interpersonal moral relationships. We turn now to discuss acouple of “inlier” cases, those included within the moral communitywho nevertheless seem to fail one or the other of MRBT’s conditions.I have just been arguing that the failure to be susceptible to moraladdress in both of its aspects—reason-based and emotional—consti-tutes a failure to be eligible for interpersonal moral relationships andthe kind of moral responsibility expressed by the reactive attitudes.Adults with high-functioning autism might, however, seem to be non-susceptible to the emotional aspect of moral address, whereas adultswith mild mental retardation might seem to be nonsusceptible to the


reason-based aspect of moral address. We will begin with the case ofautism.

One might put the worry as follows. It looks as if what is beingdemanded by the emotional aspect of moral address is empathy by thetarget of the reactive attitudes, imaginatively stepping into the shoes ofthe other person in order to feel what one has put him or her through.Indeed, many of those writing on the topic have (misleadingly, as wewill see) taken an inability for empathy to be the real failure of thepsychopath.52 But if the psychopath is (taken to be) exempt from mem-bership in the moral community by this inability, then so must be thosewith high-functioning autism (HFA), yet this would be the wrong result,for while those with HFA are indeed incapable of empathy, suffering aswell from a kind of emotional detachment, they are also clearly capableof recognizing and adhering to the basic demand, and thus they seemassuredly eligible for moral responsibility.

In addressing this worry, Jeanette Kennett starts by asserting that,while both psychopaths and those with HFA lack the capacity for em-pathy, those with HFA nevertheless “do seem capable of deep moralconcern.”53 This is because, she claims, those with HFA are capable,whereas psychopaths are not, of finding a way of seeing that the interestsof others give rise to moral reasons in the same way that their owninterests do. Their difficulties stem from the fact that they cannot runimaginative simulations of others’ mental states to figure out how theywould respond in various possible scenarios.54 But what they do instead,however, is work out moral reasons by reasoning, “as they would in other

52. See, e.g., Carl Elliott, “Diagnosing Blame: Responsibility and the Psychopath,” Journalof Medicine and Philosophy 17 (1992): 200–214; Gwen Adshead, “Commentary on ‘Psychopathy,Other-Regarding Moral Beliefs and Responsibility,’” Philosophy, Psychiatry and Psychology 3(1996): 279–81; Lloyd Fields, “Response to the Commentaries,” Philosophy, Psychiatry andPsychology 3 (1996): 291–92. See also Deigh, “Empathy and Universalizability,” for a discussionof the psychopath and “mature empathy.”

53. Kennett, “Autism, Empathy, and Moral Agency,” 354.54. See also Nichols, Sentimental Rules. Incidentally, I am granting Kennett this claim

for the sake of argument, but it turns out that matters aren’t nearly so tidy as this. Anotherperson with autism (who has actually been diagnosed as being low-functioning), Sue Rubin,self-reports as follows: “Another stereotype of people with autism is that we lack a theory ofmind, the ability to imagine how another person is thinking, what his perspective is, whathe is feeling—empathy. This is totally wrong. The ability to see something from anotherperspective is something I have been developing since I started typing [a form of facilitatedcommunication she uses]. . . . I actually enjoy looking at problems from various perspectives.As for understanding another persons [sic] emotions, I frequently overly react to the feelingsof others and often get sucked up in their emotions to such a degree that I can’t function”(Sue Rubin, “Castigating Assumptions about Mental Retardation and Low Functioning Au-tism,” http://soeweb.syr.edu/thefci/7�1rub.htm). Nevertheless, it will be fruitful to continueunder Kennett’s assumptions in order to explore a distinctly possible type of moral agent,one that brings out important points for our more general discussion.


matters, on the basis of patient explicit enquiry, reliance on testimonyand inference from past situations.”55 This thought leads Kennett tooffer a Kantian theory of moral agency: it is one’s disposition to dutythat makes one a member of the moral community, and one’s duty isdiscoverable solely by reasoning, and it is, further, capable of motivatingall on its own, independently of sympathy and empathy. Thus, “onlyindividuals who are capable of being moved directly by the thought thatsome consideration constitutes a reason for action can be conscientiousmoral agents.”56 Those with HFA are; psychopaths aren’t.

Both of her claims—one epistemic and one motivational—are false,however, or at least greatly exaggerated. Consider first the psychopath.He is, as already noted, capable of grasping and applying moral rea-sons—of seeing that the interests of others give rise to moral (or at leastto “moral”) reasons, and this capacity alone undermines Kennett’s at-tempted distinction. But if we grant her assumption that psychopathsare also incapable of empathy, how would these reasons be discovered?It would have to be indirectly, via facial and/or verbal cues, if anything,which would render the psychopath’s route to moral reasons nonstan-dard as well, albeit perhaps of a different nonstandard kind than thosewith HFA. But it also seems perfectly possible to have a psychopath whois autistic and who thus discovers his moral reasons by pure reasoning.At any rate, nonstandard route or not, having the ability to grasp andapply moral reasons alone tells us nothing yet about whether or notone is or ought to be exempt from moral agency.

Second, why think that the psychopath cannot be moved directlyby the thought that some consideration constitutes a reason for action?The psychopath does act for reasons, after all—even if they are onlyshort-sighted, instrumental reasons—and if thoughts about reasons canbe motivationally efficacious at all, then there seems nothing about thepsychopath’s behavior that should lead us to believe that he is exemptfrom that possibility. What is missing in him, of course, is the capacityto be motivated by thoughts about moral reasons, in particular, thoughtsabout second-personal moral reasons, but then it seems quite puzzlingto believe, as Kennett apparently must, that his incapacity for acting onthese moral reasons, if motivation is a matter of having certain thoughtsabout the reasons in question, is determined entirely and solely by thecontent of this sort of reason. In other words, the psychopath couldpresumably act on the following nonmoral reason: the fact that thestovetop is hot and will burn human flesh counts as a consideration infavor of not pressing one’s own hand down on the stovetop. However,the psychopath cannot, presumably, act on the following very general

55. Kennett, “Autism, Empathy, and Moral Agency,” 351.56. Ibid., 357.


moral reason: the fact that the stovetop is hot and will burn humanflesh counts as a consideration in favor of not pressing someone else’shand down on the stovetop.57 Now if the psychopath can presumablybe moved directly by the thought of the former reason, why can’t hebe moved directly by the thought of the latter reason? The only differ-ence is in the potential target of the principle, but why should a thoughtreflecting merely that difference, which translates into a mere differencein the propositional content of the reason, suddenly give rise to thecrucial boundary marking the limits of the psychopath’s motivationalcapacities?

What I have suggested, of course, is that the fact that moral reasonsare, or may be, expressed as moral demands—transforming them intosecond-personal moral reasons—gives rise to a distinction between therelevant type of moral and nonmoral reasons that, together with a cer-tain sort of lack of affect, explains the psychopath’s motivational inca-pacity with respect to the former but not the latter: insofar as the psy-chopath is incapable of caring about the source of such reasons, he isincapable of being motivated to act on the reasons in favor of protectingthat source. If one fails to recognize, however, that the relevant moraland nonmoral reasons are distinguished in virtue of the former’s rootsin the demands (or potential demands) of one’s fellows, one will simplybe unable to explain how the psychopath’s otherwise functional abilityto act on reasons ends at the borders of the (second-personally) moral.This is precisely why Kennett’s attempt to distinguish psychopaths fromthose with autism misses the mark.

But now consider the person with HFA, someone we take to bemorally responsible. If he is able to grasp, apply, and act on moralreasons and yet cannot empathize with his fellows, then it looks as ifeither (a) the emotional capacity I have emphasized is just not a nec-essary condition of being susceptible to moral address (on the assump-tion that susceptibility to moral address is sufficient for being eligiblefor moral responsibility and thus being a member of the moral com-munity) or (b) susceptibility to moral address is not sufficient for moralresponsibility (and, thereby, membership in the moral community). Ideny both purported worries.

First, it is crucial to point out that people with HFA typically careabout morality, about being morally good. As Kennett herself puts it,those with HFA can be “well-meaning, and wish like Brother Juniper

57. This is not yet, of course, the type of second-personal moral reason I have arguedis involved in moral address and responsibility, but (a) Kennett simply does not make anydistinction between the various types of moral reasons in the first place and (b) it is thoughtsof this more general type of moral reason anyway by which she thinks the psychopath isincapable of being moved.


to do the right thing.”58 But they also go beyond the moral fetishist byactually caring about the source of the basic demand, namely, othermembers of the moral community. Indeed, the high-functioning autistKennett repeatedly refers to, Temple Grandin, claims to experiencevery strong emotions, and these are affected in sync with the up-and-down fortunes of her cared-for objects, including other people.59 Whatis missing for those with HFA, remember, is just the ability to recognizemoral reasons via the standard route, by picking up on the emotionalcues of moral address directly and having an insider’s understandingof what others are feeling. Lack of empathy, then, does not imply lackof caring. As a result, their capacity to care about their fellow de-manders may undercut the motivation for Kennett’s move to a Kantiantheory of moral motivation as well, for it could turn out that themotivation of those with HFA to adhere to the moral demand has itssource in their affective capacities after all. It is perfectly possible, andthis is hinted at throughout the psychological literature, that thosewith HFA experience guilt when failing to live up to the basic demandand that they are thus motivated to discover what moral reasons thereare in order to avoid the guilt or remorse that comes their way in lightof such failures.60 We can thus easily distinguish them from psycho-paths in a way that preserves the central tenets of our latest revisionof MRBT. Those with HFA certainly have difficulty making the tran-sition from caring about other demanders to “hearing” their demands,but it seems that they can eventually figure out those demands via

58. Kennett, “Autism, Empathy, and Moral Agency,” 351; emphasis mine.59. See Temple Grandin, “Social Problems: Understanding Emotions and Developing

Talents,” http://www.autism.org/temple/social.html. It would be interesting to see whetheror not those with HFA like Grandin care more or less about people than about inanimateobjects. The strongest emotions reported by Grandin were those occurring after the Uni-versity of Colorado’s library flooded and nearly a million books “drowned”; Grandin re-sponded by crying and crying, grieving for the drowned books.

60. I say “hinted at” because, while there is clear evidence that those with HFA (andautism generally) experience emotions, both simple and complex, I have found no literatureexplicitly discussing autism and guilt or remorse. This may be because it’s simply taken forgranted that these emotions are indeed experienced by those with autism, or it may be thatno one has yet figured out a way to determine whether or not these are the precise emotionsthey experience. At any rate, I think I am quite safe in assuming that guilt and remorse areamong the many emotional experiences undergone by those with autism (especially thosewith HFA). For some examples of the “hinting” I have run across, see, e.g., Nurit Yirmiya,Marian Sigman, Connie Kasari, and Peter Mundy, “Empathy and Cognition in High-Func-tioning Children with Autism,” Child Development 63 (1992): 150–60; Lisa Capps, NuritYirmiya,and Marian Sigman, “Understanding of Simple and Complex Emotions in Non-retardedChildren with Autism,” Journal of Child Psychology and Psychiatry 33 (1992): 1169–82; andHidehiko Takahashi, Noriaki Yahata, Michihiko Koeda, Tetsuya Matsuda, Kunihiko Asai, andYoshiro Okubo, “Brain Activation Associated with Evaluative Processes of Guilt and Embar-rassment: An fMRI Study,” NeuroImage 23 (2004): 967–74.


alternate routes, and when they do, it looks like they can be just asresponsible to their fellows in the moral community as anyone else.What we actually want in the case of moral membership is that ourfellow agents be able to hear and understand our pleas in order toappreciate and respond to those pleas, and if how the agent gets thereinvolves a nonstandard route, it’s not at all clear why we should care.Thus, the person with HFA does not constitute a counterexampleeither to the claim that emotional capacities are necessary for suscep-tibility to moral address or to the claim that susceptibility to moraladdress is sufficient for being eligible for moral responsibility.

What I have said thus far is enough to defend MRBT Version 3from Kennett’s attack,61 but I actually want to defend a stronger claim,namely, that those with HFA are in fact capable of an aspect of empathythat psychopaths and other exempt agents are not, the aspect that turnsout to be essential to interpersonal engagement and membership in themoral community. Now Kennett discusses only one aspect of empathy,that of projective imagination. So what the person with HFA lacks, inbeing “incapable” of empathy, is the ability to understand what it’s likefor another person: she is unable to simulate others’ mental states, “totake on and understand the different perspectives and interests of oth-ers” in order to pick up on their emotional cues and thereby discoverthe appropriate moral reasons.62 But this is actually only one part ofwhat we think of as full empathy. After all, a good psychologist will beable to understand what it’s like for her patient—she will be able toimagine herself in his shoes—while nevertheless remaining, in an im-portant sense, detached from him. In addition, contrary to the as-sumption granted in the introduction to the issue earlier, psychopathsmay also be capable of this much; indeed, some murderous psychopathsmay get a certain kind of sick pleasure from imagining the terror theirvictims are going through.63 So this aspect of empathy is surely not whatwe think is sufficient for the kind of interpersonal engagement consti-tutive of intelligible moral address. After all, my demand that you stoptorturing me will simply be a joke—or perhaps a spark for even greateramusement—if you are getting pleasure in vividly representing to your-self what I’m going through.

In addition to a mere understanding of what the other personthinks or feels, then, the other aspect of full empathy, and the aspect

61. However, see a more nuanced articulation of Jeanette Kennett’s view in “Reasons,Reverence, and Value,” forthcoming in Moral Psychology, vol. 3, The Neuroscience of Morality:Emotion, Disease, and Development, ed. Walter Sinnott-Armstrong (Cambridge, MA: MIT Press,2008).

62. Kennett, “Autism, Empathy, and Moral Agency,” 346.63. See Deigh, “Empathy and Universalizability,” 761, for a somewhat similar discussion

of what he thinks of as the immature empathy of the sadist.


I think is actually sufficient for interpersonal engagement,64 is that oneshares the cares of the object of empathy (at least with respect to theevents giving rise to the empathy), that is, one is emotionally vulnerablewith respect to the fortunes of the items the person with whom oneempathizes cares about and vulnerable in a roughly similar way to theperson with whom one empathizes. In order to fully appreciate whatthings are like for you, I must not only be able to imagine what it isyou’re going through, but I must also be disposed to feel what it isyou’re going through in a similar fashion to the way you do. Considerthe typical case. In projectively imagining myself in your shoes, I seethe world through the filter of your beliefs, desires, and cares, and inso doing I am able to understand how you would be affected (how youwould be disposed to react emotionally) in light of certain events. Nowif I myself wouldn’t react in the way you do, I do not share your cares,and your emotional life will be, to me (at least with respect to thisparticular set of circumstances) like that of an alien: I might be able tosee how the world seems to you—and so I will, to that extent, have anintellectual understanding of the reasons you have and act on—but Iwill not be able to feel how the world feels to you, and so I will lack anemotional connection with your reasons and actions.

For example, suppose that you are a suicidal teenager. While I may,in putting myself in your shoes, understand your emotional reactions(and dependent desires), appreciating that they are intelligible (andperhaps to be expected) for a person like you given your set of caresin the situation in which you find yourself, I certainly won’t be able toexperience those emotions myself on your behalf, for they would notbe intelligible for me given my set of cares, even were I to find myselfin your situation.

I become engaged with you, however, when I can feel what youfeel, and this is rendered possible only by virtue of my caring aboutwhat you care about. When I understand what the world is like for you,the perspective I take on the world is filtered through your cares, so Iknow on an intellectual level how you would react emotionally to certainevents. I am capable of naming your emotional reactions, in other words.This much a psychopath might be capable of. When I also share yourcares, though, I am capable of feeling your emotional reactions myself:in the most vivid circumstances, it is as if what is happening to you ishappening to me (to some extent), such that we both react in emo-tionally similar ways to the fortunes of the objects we both care about,and so, in this respect at least, we are united. This is what is involved

64. It is sufficient, I should be clear, for interpersonal engagement, where this doesn’tnecessarily involve behavioral responses to that engagement. For that, one will need themotivational capacities discussed earlier.


when we speak of identifying with others: we not only can understandwhat the world is like for them, but we also can feel what they feel inthe way that they feel it.65 It is this sort of emotional engagement thatis at the heart of interpersonal relationships generally, and it is what Icall “identifying empathy.”

The psychopath is incapable of identifying empathy. Yes, he canimagine himself in the shoes of someone else, and perhaps he is ableto understand and label what that person is going through, but hecannot share the cares of anyone else, simply because he himself lackssuch cares altogether or at least he lacks the mature cares requisite ofparticipant demanders.66 Therefore, the psychopath is incapable of feel-ing what I feel in the way that I feel it when he hurts me or violates thebasic demand in other ways, so any resentment I express toward himwill strike him as unintelligible. The psychopath thus lacks the abilityto be party to the emotional address that is an essential component ofmoral address. Without this ability, one will be unable to identify withanyone else, but if one is completely unable even to appreciate (eventhough able to intellectually comprehend) what it’s like to be a moralagent, it is quite unclear how one could be such a moral agent.

Here is the picture thus far then. Given the nearly universal in-terest in decent treatment, all those capable of caring will typicallycare about not being wrongfully harmed, and so when such an eventoccurs, they will experience emotions produced by their perceptionof both the harm and the wrongfulness. The emotion linked to theharm will range from despair to disappointment to outrage. The emo-tion linked to the wrongfulness will most likely be resentment.67 When

65. This is not sympathy, however, which is a concern for the other person for her sake.See, e.g., Stephen Darwall, Welfare and Rational Care (Princeton, NJ: Princeton UniversityPress, 2002), esp. chap. 3, “Empathy, Sympathy, Care.” While I believe that sympathy typicallyfollows from full empathy, as described above, the two are certainly distinct: I may fullyappreciate what things are like for you and care about what you care about (to some extent)but still fail to care about you.

66. But isn’t it true that the psychopathic sadists and serial killers we read about oftenget a charge out of imagining the torture and suffering of their victims, suggesting that theiridentification with them is fairly robust? To the extent that they are true psychopaths, theysimply do not experience genuine emotions, so they are incapable of being emotionallyvulnerable—of sharing the cares of the victim—in the way that I have described it. (SeeStout, The Sociopath Next Door, 127.) Of course, they do feel something when imagining theplight of their victims, to be sure, but it is more akin to a kind of sexual pleasure than theemotional reactions that people who care undergo when the things they care about aredoing well or poorly. (I am grateful to an anonymous referee for raising this concern.)

67. I am focusing here on the second-personal reactive attitudes, but the analysis goesas well for the third-person and first-person analogues (typically expressed as indignationand guilt, respectively). In addition, I should make clear that I agree with Gary Watson thatthe emotional reactions expressing the basic demand are neither necessary to, nor consti-tutive of, the demanding itself. After all, there are certainly some extraordinary agents, like


this last emotion is expressed to the offender, it is an expression ofthe basic demand, and it is, at that point, a form of emotional address:“Can’t you see how you’ve made me feel?” This appeal is an invitationfor the offender to take a moment to appreciate what he has donefrom within the other’s perspective.68 If such demands matter to him,that is, if he cares about the demand in virtue of his caring about theemotional state of the demander, he will be moved to make the em-pathetic leap (so as to avoid betraying something or someone he caresabout). He will then picture what things are like for the other person,how the recent events feel to her through the filter of the relevantcares he takes her to have (and, given that we are talking about thebasic moral demand here, it should be no secret whatsoever what herrelevant cares are). If he also shares her relevant cares (and he should,as a fellow human with the same fundamental interest in decent treat-ment), he will fully be able to appreciate how she feels, and at thatpoint he will be in a position to respond to her original emotionalappeal in any number of ways: he may feel guilt or remorse and somay apologize or otherwise attempt to rectify what he has done. Al-ternatively, if he takes the other person’s resentment to be unwar-ranted, he may attempt to justify his actions, offer an excuse of somesort, plead an exemption, or simply scoff and walk on. But all of theseare subsequent to the establishment of an interpersonal relationship,an emotional communion between two moral agents. The identifyingempathetic response, spurred by caring about the demander, is theticket into the moral community.

The psychopath lacks the ability to care about the demander, sohe lacks the ability to respond in this way to the emotional address inwhich the basic demand consists. The person with HFA, on the otherhand, while incapable of projective imagination, is still capable of thekind of emotional exchange constitutive of moral agency; it’s just thatthe process of getting to the exchange is much more indirect. Theemotional appeal made by the victim via his emotional reaction to theautistic person’s violation of the basic demand, where the reaction in-dicates that something is wrong (perhaps realized through the autisticperson’s checking the facial reaction of the victim against similar pastreactions in her video database), may prompt the person with HFA to

Gandhi and King, who are able to hold their oppressors responsible without expressing theirdemand in so doing via negative reactive attitudes like resentment. (I remain agnostic,however, regarding whether or not such saintlike agents still do experience the negativeemotions but just do so without expressing them publicly.) See Watson, “Responsibility andthe Limits of Evil,” 255–58. For the first articulation of this point, see Stern, “Freedom, Blame,and Moral Community,” 78.

68. It is no coincidence that moral education starts with a question very much like this:“How would you feel if someone did that to you?”


engage in a circuitous reasoning route that will eventually get her tosee that she shares certain of the victim’s just-betrayed cares. Insofar asshe is capable of caring about both the victim and the victim’s demand,as well as sharing the cares of the victim (in a way sufficiently similarto the victim), she is capable of responding to the emotional appealsof the basic demand in a way that the psychopath cannot, so the personwith HFA gains entry into the moral community in a way the psychopathcannot. We may thus introduce one aspect of empathy into our latestrevision:

MRBT Version 4: One is a member of the moral community, amoral agent eligible for moral responsibility and interpersonal re-lationships, if and only if (a) one has the capacity to recognize andapply second-personal moral reasons one is capable of discovering viaidentifying empathy with the affected party (or parties) of one’s behavior,regardless of the method of identification and (b) one is capable of beingmotivated by those second-personal moral reasons because one iscapable of caring about their source (i.e., the affected party [orparties]), such that one is susceptible to being moved to identifyingempathy with that source by the moral address expressible via thereactive attitudes in both its reason-based and emotional aspects.

There is one last possible counterexample to contend with, however.

V. MILD MENTAL RETARDATION

Suppose one were susceptible to the emotional, but not necessarily thereason-based, component of moral address, and further suppose thatone were not quite capable of identifying empathy with the affectedparty of one’s behavior. According to MRBT Version 4, one would notbe an eligible candidate for membership in the moral community. Thiswould be an unfortunate result, however, for it would exclude adultswith mild mental retardation (MMR). I want briefly to explore this case,then, a case that is rarely discussed adequately or even at all in thephilosophical literature.

Most people (between 85 and 90 percent) who are mentally re-tarded are mildly so, diagnosed as meeting three general criteria: (a)IQ of between 50 and 70; (b) substandard adaptive functioning, that is,a lower level of independence and social responsibility than others ofhis or her age; and (c) onset of permanent condition by age 18.69 Thedistinction between various degrees of mental retardation (mild, mod-erate, severe, and profound) is entirely a matter of cognitive functioning,that is, IQ level. Viewed from the Piagetian framework, those with MMR

69. See, e.g., citation and discussion of the DSM-IV-TR definition in Mary Beirne-Smith,James R. Patton, and Shannon H. Kim, Mental Retardation, 7th ed. (Upper Saddle River, NJ:Pearson, 2006), 65–66.


go through the same stages of development as their unimpaired peers,but they simply do so more slowly, with most failing to develop to thehighest stage, the capacity to engage in formal or abstract operations(which unimpaired children typically reach by age 11).70 In what follows,I will focus on adults with MMR.

One might be tempted to treat such people, at least with respectto moral responsibility and moral community, as on a par with childrenages 7–11 (roughly the range of mental ages of those with MMR), andthis is what the theorists who actually address the topic do.71 Our treat-ment of children in this age range is anything but unambiguous, though,as Strawson himself notes in a revealing passage. As he puts it, parentsand other caretakers of children cannot have either a purely objectiveor a purely participant attitude toward those in their charge: “They aredealing with creatures who are potentially and increasingly capable bothof holding, and being objects of, the full range of human and moralattitudes, but are not yet truly capable of either. The treatment of suchcreatures must therefore represent a kind of compromise, constantlyshifting in one direction, between objectivity of attitude and developedhuman attitudes. Rehearsals insensibly modulate towards true perfor-mances. The punishment of a child is both like and unlike the punish-ment of an adult.”72

The motivation for this sort of “impure” back-and-forth treatmentis the “yet” in the first sentence, for it seems that, when it comes toexpression of the personal reactive attitudes toward unimpaired chil-dren of this age, they are expressed with one eye clearly fixed on thechildren’s future development. So while the children may not fully ap-preciate what is being demanded of them or the lessons their caregiversare attempting to teach them via expression of resentment now, thechildren will understand and be “grateful” one day, they are told, oncetheir full cognitive capacities have been actualized. In the meantime,they are treated as if they were responsible to prepare them for actuallybeing held responsible once they have grown into their more cognitivelydeveloped future selves. Indeed, on Wallace’s view, expression of thereactive attitudes may itself serve as a key developmental aid: childrenare treated as if they were responsible “partly . . . because [it] is believed

70. Ibid., 271–72. The other three stages of development, according to this model, arethe sensorimotor stage (birth to 2 years, standardly), the preoperational stage (2–7 years),and concrete operations (7–11 years).

71. Note, e.g., what Wallace suggests in Responsibility and the Moral Sentiments in 167 n.14: “Similar remarks [as those about children] apply to people who are mentally retarded,who may be seen as having perpetually undeveloped capacities for reflective self-control.”

72. Strawson, “Freedom and Resentment,” 88.


to be the most effective way to stimulate the development of their powersof reflective self-control.”73

But those with MMR will never cognitively develop beyond a certainpoint, so it seems that there just is no role that the reactive attitudes—in either their “pure” or “impure” form—could play for them. If it isthe presence of a certain level of cognitive capacity that warrants “pure”deployment of the personal reactive attitudes and full moral member-ship, then they, like children, should be exempt. If it is the expectation(or susceptibility to stimulation) of future cognitive development thatwarrants at least an “impure” deployment, involving treatment of themas if they were responsible moral agents, then they, unlike children,should still be exempt.

But, of course, as anyone who has been a caregiver to an adult withMMR knows full well, they are most definitely not exempt; rather, theyare often fully embraced as members of the moral community, as morallyresponsible agents—albeit with respect to a more limited range of ac-tions—with whom one is able to enter into interpersonal relationships.They are not treated merely as if they were responsible; instead, theyare responsible. What makes this possible, though? To answer this ques-tion, we need to say more to explain why adults with MMR are actuallydifferent in important ways from children and why these differenceslead to different treatment with respect to responsibility and moral mem-bership, contrary to what most philosophers have suggested.

First, many adults with MMR are capable of living on their own,whereas children (roughly ages 7–11) are not. This is in part due tothe fact that only the former are legally allowed to work, but I suspectthat this legal difference makes very little moral difference. After all,young children used to be allowed to work, yet they were still incapableof living on their own. This difference suggests that there’s a kind ofmaturity that we attribute to adults with MMR that is missing fromchildren. I believe that such maturity is a function of at least two deeperdifferences between children and adults with MMR. First, the latter arein general capable of learning in a way unavailable to normal children.In other words, because of the eventual cognitive development of nor-mal 7–11-year-old children (i.e., their soon being able to engage informal/abstract operations), their ability to learn with their pre-12-year-old cognitive tools is temporally limited. In other words, from the de-velopmental stage at which they can engage in concrete operations, theyhave only a short amount of time to learn about their world by meansof these cognitive tools before they develop the intellectual capacitiesthat enable them to learn about the world at both a different pace andin a different way from that point on. The mildly retarded are generally

73. Wallace, Responsibility and the Moral Sentiments, 167.


unable to “think about and deal with more than one or two aspects ofa complex situation,” nor are they able “to defer gratification so as tomake choices most likely to be of benefit over the longest run, and toforesee long-term consequences of present acts.”74 But then, given thesedifficulties, their ability to live on their own, to work, and to make theirway in the world generally indicates that they have learned a great dealabout how to adapt to a world that, in many ways, they do not fullyunderstand. This is a knowledge and general facility that, ironically, therest of us lack, given our normally paced cognitive development. Wehave not had to live (at least for very long) in a puzzling world that wecan only partially decode. As a result, the maturity at which adults withMMR arrive is hard won, and this marks a real difference between themand unimpaired children. In taking on adult responsibilities, then, theymay be able to take on responsibility as well.

Second, adults with MMR seem capable of a wider range of emo-tions than children. Of course, no one suggests that adults with MMRare capable of existential angst or ennui (are these actually emotions,though?), but they are certainly capable of fear, anger, sadness, and joy,as well as the more complex emotions of romantic love, despair, grief,guilt, and remorse, even if they may be unable to label the emotionsin question.75 While children are also capable of fear, anger, sadness,and joy, it is quite unclear, and actually seems rather unlikely, that theyare capable of the more complex group. This emotional range in adultswith MMR is another source, I suggest, of their specific kind of maturity,in this case an emotional maturity. But, as should be obvious by now,it is also the primary source of their membership in the moral com-munity. And insofar as adults with MMR are capable of the (nearly) fullrange of emotions, they are capable of caring about the (nearly) fullrange of things unimpaired adults can care about. They are surely able,then, to care about the agents who constitute the source of the basicmoral demand. What, then, is the problem?

I have been arguing that the moral demand does not wholly consistin reason-based address; in addition, and crucially, it is a form of emo-tional address, a plea for the target to step inside the offended party’sshoes to feel what his offense is like, an appeal for the offender toidentify with the offended, in order to fully appreciate just what he has

74. Daniel Wikler, “Paternalism and the Mildly Retarded,” Philosophy & Public Affairs 8(1979): 377–92, 381. Wikler is here citing Travis Thompson from an address made to theBehavior Control Group, Hastings Center, New York, 1977. See also Thompson’s “The Be-havioral Perspective,” Hastings Center Report 8 (1978): 29–32.

75. See, e.g, Mark J. Hauser and John J. Ratey, “The Patient with Mental Retardation,”http://www.psychiatry.com/mr/patientwmr.html. See also the fact sheet on mental retar-dation from the Texas Department of Aging and Disability Services, http://www.dads.state.tx.us/services/dads_help/mental/retardation/mr_fact_sheet.html.


put the offended through. Now, typically, one is led into identifyingempathy directly by picking up on the emotional cues of the otherperson. While adults with HFA cannot do this, they can still arrive atidentification via an alternate route and thus eventually be led to seeand appreciate the relevant second-personal reasons. But there mightbe thought to be two difficulties for adults with MMR here. First, whilethey can definitely pick up on the emotional cues of others (e.g., theycan tell directly when others are upset), they may not be able to makethe transition from there to identifying empathy. Second, even if theyare capable of empathy, they may not be capable of seeing and appre-ciating the relevant second-personal reasons thereby. And it certainlyseems as if they cannot engage in reason-based moral exchange a laWallace: reflection on, or application of, general moral principles orspecific moral demands seems beyond their ken. So the problem hereis, in a way, the opposite of what we encountered in the last section,for while the adult with HFA seemed to be unsusceptible to the emo-tional appeal aspect of moral address, the adult with MMR seems to beunsusceptible to the reason-based aspect of moral address.

What I want to suggest—and it is only a suggestion—is that, whilethose with HFA may take a nonstandard route into empathy, emotionalconnectedness, and eventually the relevant second-personal moral rea-sons, following a reason-based route, those with MMR also take a non-standard route into, eventually, the relevant second-personal moral rea-sons, following an emotion-based route. It starts with an external promptthat may eventually be internalized through repetition or rehearsal asthe person matures, a prompt that should seem familiar to all of us:“How would you feel if someone did that to you?” But the prompt herecomes from a loved one, a caregiver, someone the person with MMRcares about, and so the basic demand will be apprehended via this thirdparty in the following way. Suppose that the person with MMR has hurtsomeone else, a stranger. What caregivers will often do in such a caseis mimic, and even amplify, the reactive attitude expressed by thestranger, addressed firmly to the person with MMR. The caregiver mayalso express a reactive attitude of her own (along with the aboveprompt), perhaps disappointment, or even anger (if this sort of thinghas occurred repeatedly). These attitudes are expressions of the basicdemand on behalf of the stranger, and so while the person with MMRwill typically feel guilt or remorse as a result, it will be in response tothe negative reactions of his caregiver, not the stranger. It is the caregiverwhom he may feel that he has failed, in other words, so that, while hemay not in fact care about the stranger (or be able to take up hisperspective or be able to recognize or apprehend any second-personalreasons to which he is drawing attention), the person with MMR maystill be prompted to apologize and/or intend to avoid such hurts in the


future as a response to the emotional appeal addressed to him by thecaregiver, whereby he also thereby recognizes and apprehends the sec-ond-personal reasons of which she is the authoritative source (“that she,my caregiver, wants me not to hurt that man” is the reason on whichhe may act). Further, in responding to the caregiver’s emotional appeal,he does at least identify with her—he shares her cares—and insofar asshe identifies with the stranger, the person with MMR identifies withthe stranger by proxy.76

The hope of most caregivers, of course, is that eventually thosewhom they care for will internalize this process, so that when they areon their own in the world, a similar negative reactive attitude will promptthem to see the caregiver in the face of the stranger, such that thestranger’s expression of the basic demand will prompt the sort of re-sponse that they would have had to their caregiver. This is to allow thepossibility of extending the face of their familiar second-personal de-mander to an indefinitely wide range of people. But with successfulresponses of this sort, it becomes very difficult to distinguish betweenthis case and cases of ordinary moral responsibility and agency. Whileat first their responses to expressions of the reactive attitudes may bemuch more akin to those of children—responding for purely egoisticreasons to avoid the anger or disappointment of their caregiver—onceadults with MMR are able to respond to the emotional cues and addressof strangers (even if this response is mediated by their experiencing thedemand as if it were expressed by their caregiver), their reactions be-come distinct from those of children, and they might indeed be saidto be acting on the basic demand at that point, at least in a way that isjust as good as ordinary moral agency, and thus there seems no reasonto draw the boundaries of the moral community in a way that excludesthem.

What this example should drive home, though, is that eligibilityfor moral responsibility, in the sense of being accountable to, is notnecessarily a matter of being accountable to the moral patient, theaffected party of one’s actions. Instead, it is a matter of being account-able to a representative member of the moral community, someone withthe requisite authority qua fellow member to engage in moral addresswith the agent in question. So, while the adult with MMR may not besusceptible to the direct moral address of all of his fellows (includinghis moral patients), as long as he is susceptible to the direct (actual orimagined) moral address of someone (likely his caregiver), he is eligiblefor responsibility to us, the fellow members of the moral community ofwhich that someone is our representative. And here is where the im-

76. This may seem overly complicated, but this is because, I suggest, our reactionsto those with MMR are themselves quite complicated.


portance of the third-person analogues to the reactive attitudes becomesevident, for sometimes the wrongdoer’s moral patients have no voice,or don’t know how to use it, in which case it is up to the rest of us tohold that wrongdoer accountable by attempting to engage in moraladdress and exchange with him on behalf of his silent patient. And soit is here where we see how membership in the moral community istruly membership in a community—the “club” I referred to in the open-ing sentence of the article—a united collection of agents of equal au-thoritative voices, each of us equally capable of taking a stand on behalfof any our fellows, each of us accountable to, yet responsible for, oneanother.

We are now in a position to advance the final, complete formulationof moral agency:

MRBT Version 5: One is a member of the moral community, amoral agent eligible for moral responsibility and interpersonal re-lationships, if and only if (a) one has the capacity to recognize andapply second-personal moral reasons one is capable of discoveringvia identifying empathy with either the affected party (or parties)of one’s behavior or an appropriate representative, regardless of themethod of identification and (b) one is capable of being motivatedby those second-personal moral reasons because one is capable ofcaring about their source (viz., the affected party/parties or an ap-propriate representative), insofar as one is susceptible to being movedto identifying empathy with that source by the moral address ex-pressible via the reactive attitudes in both its reason-based and emo-tional aspects.

A final note. The four different agents I have discussed in this articleare merely representative “types,” wholly defined by their condition, all-or-nothing in their nonstandard ways. I did this in order to more easilyarticulate the conditions of moral agency. But, as should be obvious,any real-world tokens of these types are likely to be far more complex,exhibiting their “conditions” only in certain circumstances, to greaterand lesser extents. But this means that we should also take membershipin the moral community to be a more complex matter as well. So it willlikely be the case that those with HFA and MMR are perhaps full-fledgedmembers only within a limited domain of morally relevant circum-stances, namely, those in which they are indeed capable of meeting thereason-based and emotion-based conditions set forth. In other circum-stances, however, the objective attitude may simply be more appropriate.By contrast, some of those on the low end of the psychopath or fetishistscales may occasionally be able to recognize and respond to the relevantmoral reasons, and so they may be able to engage with us emotionally.In such circumstances they should be just as susceptible to the reactiveattitudes as the rest of us.


Of course, it would seem to be much more difficult to sustain aninterpersonal relationship with someone who drifts in and out of themoral community, but perhaps, as with other sorts of relationships thatwe enter into at the workplace and at the pool hall, say, we have roomand resources for relationships within contained pockets of our lives.So, in reality, matters will be much more complex than my final versionof MRBT may lead us to believe, but in a way this has been my overallpoint: the conditions for moral agency are far more nuanced and com-plex than they have typically been taken to be, and it actually short-changes the richness of our interpersonal relationships when we ignorethe variety of factors that make them what they are.

Moral Address, Moral Responsibility, and the Boundaries of the Moral Community

Documents