Empathy and Moral Development

Empathy and Moral Development IMPLICATIONS FOR CARING AND JUSTICE Martin L. Hoffman New York University PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge, United Kingdom CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge CB2 2RU, UK http://www.cup.cam.ac.uk 40 West 20th Street, New York, NY 10011–4211, USA http://www.cup.org 10 Stamford Road, Oakleigh, Melbourne 3166, Australia Ruiz de Alarcón 13, 28014 Madrid, Spain q Cambridge University Press 2000 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2000 Printed in the United States of America Typeface Palatino 10.25/14 pt. System DeskTopPro/UX [BV] A catalog record for this book is available from the British Library. Library of Congress Cataloging in Publication Data Hoffman, Martin L. Empathy and moral development : implications for caring and justice / Martin L. Hoffman. p. cm. Includes bibliographical references and index. ISBN 0–521–58034–X 1. Moral development. BF723.M54H64 2000 155.2'5 – dc21 2. Empathy. ISBN 0 521 58034 X hardback I. Title. 99–29669 CIP Contents Acknowledgments 1 page ix Introduction and Overview Part I 1 Innocent Bystander 2 Empathy, Its Arousal, and Prosocial Functioning 29 3 Development of Empathic Distress 63 4 Empathic Anger, Sympathy, Guilt, Feeling of Injustice 93 Part II Transgression 5 Guilt and Moral Internalization 113 6 From Discipline to Internalization 140 Part III 7 Relationship Guilt and Other Virtual Guilts Part IV 8 10 175 Is Empathy Enough? Empathy’s Limitations: Over-Arousal and Bias Part V 9 Virtual Transgression 197 Empathy and Moral Principles Interaction and Bonding of Empathy and Moral Principles 221 Development of Empathy-Based Justice Principles 250 vii Contents 11 Multiple-Claimant and Caring-Versus-Justice Dilemmas Part VI 12 13 Culture The Universality and Culture Issue Part VII 263 273 Intervention Implications for Socialization and Moral Education 287 References 299 Author Index 319 Subject Index 325 viii CHAPTER 1 Introduction and Overview When I tell people my field is moral development, the first response is usually silence, sometimes ‘‘Oh!’’ They think I mean religion, telling the truth, the decline of traditional family life, and prohibitions against drugs, alcohol, and teenage pregnancy. When I tell them my interest is people’s consideration for others, they perk up at first but then say something like it must be frustrating to study that because everybody is interested in themselves; who cares about anyone else, except maybe their family? But when I say humans could not have survived as a species if everyone cared only about himself, they pause, think about it, and then say something like ‘‘You might be right.’’ The evolution argument carries weight, as though it were self-evident that hunters and gatherers had to help each other to survive, so humans must have helping genes. In any case, it is in this end-of-millennium, first-world context of competitive individualism and little caring for others that some of us study prosocial moral behavior – knowing full well that however much a person cares about others, when the chips are down, the individual thinks of himself first: He or she is not the other. People do make sacrifices for others, however, sometimes big sacrifices, and they help others in small ways all the time. This adds to the quality of life and makes social existence possible. So there is something to study. Indeed, the topic has preoccupied philosophers at least since Aristotle and has been a topic of research interest in psychology for almost a century. The topic’s staying power, I think, lies in its selfevident importance for social organization and the fact that it epitomizes the existential human dilemma of how people come to grips 1 Empathy and Moral Development with the inevitable conflicts between their egoistic needs and their social obligations. Philosophy and religion have various answers to this dilemma, and their answers have parallels in contemporary psychological theory. One answer, the ‘‘doctrine of original sin,’’ which assumes people are born egoistic and acquire a moral sense through socialization that controls egoism, is paralleled in early Freudian and sociallearning theories that stressed the importance for moral development of reward and punishment by parents, especially giving and withholding affection. The diametrically opposed and more interesting ‘‘doctrine of innate purity,’’ associated with Rousseau who viewed children as innately good (sensitive to others) but vulnerable to corruption by society, has a rough parallel in Piaget’s theory, not that children are innately pure but that their relation to adults produces a heteronomous respect for rules and authority which interferes with moral development. This corruption by adults can only be overcome by the give-and-take of free, unsupervised interaction with peers, which, together with children’s naturally evolving cognitive capability, enables them to take others’ perspectives and develop autonomous morality. The resemblance to ‘‘innate purity’’ is that the free and natural interaction of premoral children produces moral development, whereas interaction with (socialized) adults prevents it. Philosophers like Immanuel Kant and his followers, who attempt to derive universal, impartially applied principles of justice, helped inspire Kohlberg (and to a lesser extent Piaget) to construct an invariant sequence of universal moral stages. And the British version of utilitarianism represented by David Hume, Adam Smith, and others for whom empathy was a necessary social bond, finds expression in current research on empathy, compassion, and the morality of caring. Contemporary theories of prosocial moral development tend to focus on one dimension, each with its own explanatory processes. Social-learning theories deal with helping behavior and specialize in the processes involved in reward, punishment, and imitation. Cognitive-developmental theorists deal with moral reasoning and employ concepts like perspective-taking, reciprocity, cognitive dis2 Introduction and Overview equilibrium, progressive construction, and co-construction. Theories of emotional and motivational development employ concepts like parent identification, anxiety over loss of love, empathy, sympathy, guilt, and moral internalization. I have long written on the emotional/motivational dimension, especially empathy development, guilt, and moral internalization. To me, empathy is the spark of human concern for others, the glue that makes social life possible. It may be fragile but it has, arguably, endured throughout evolutionary times and may continue as long as humans exist. In this book, I update my previous work and frame it in a comprehensive theory of prosocial moral behavior and development that highlights empathy’s contribution to moral emotion, motivation, and behavior but also assigns special importance to cognition. The aim is to elucidate the processes underlying empathy’s arousal and its contribution to prosocial action; to throw light on the way empathy develops, from preverbal forms that may have existed in early humans and still do in primates, to sophisticated expressions of concern for subtle and complex human emotions. My aim is also to examine empathy’s contribution to the principles of caring and justice, to resolving caring–justice conflicts, and to moral judgment. I have been working on the theory for three decades. It includes elements of the philosophical and psychological approaches mentioned earlier but also makes use of contemporary cognitive psychology – memory, information processing, causal attribution, and especially the synthesis of affect and cognition. Its primary focus is consideration for others, often called ‘‘caring’’ morality, but also includes ‘‘justice’’ and the mutually supportive though sometimes contradictory relation between caring and justice. The theory attempts to account for human action in five types of moral encounters or dilemmas, which I believe encompass most of the prosocial moral domain. In the first, the simplest type, one is an innocent bystander who witnesses someone in pain or distress (physical, emotional, financial). The moral issue is: Does one help and how does one feel if one does not help? In the second type, one is a transgressor, who harms or is about to harm someone (accidentally, in fights, arguments). The moral issue is: Does one refrain from 3 Empathy and Moral Development harming the other or at least feel guilty afterward? In the third type of moral encounter, which combines elements of the first two, one is a virtual transgressor, who, though innocent, believes he or she has harmed someone. The fourth type is more complex: It involves multiple moral claimants among whom one is compelled to make a choice. The moral issue is: Whom does one help and does one feel guilty over neglecting the others? The fifth type, caring versus justice, involves multiple moral claimants but also a clash between considering others and more abstract issues such as rights, duty, reciprocity. The moral issue here is: Which principle prevails, caring or justice, and does one feel guilty for violating the other? Multiple claimant and caring–justice dilemmas are especially important in societies like ours that are becoming increasingly diverse culturally. All five types share an empathic motive base – empathy defined as an affective response more appropriate to another’s situation than one’s own. Each type features empathic distress – one feels distressed on observing someone in actual distress – and one or more motives derived from empathic distress: sympathetic distress, empathic anger, empathic feeling of injustice, guilt. The book begins with an analysis of innocent bystanders. The bystander model attempts to answer these questions: What are the motives that predispose innocent bystanders to help victims? What are the psychological mechanisms that underlie the arousal or activation of these motives? What is the developmental course of the motives? It takes the first three chapters (part I) to answer these questions. I begin in chapter 2 by defining empathy as an affective response that is more appropriate for another’s situation than one’s own. Empathic distress is the focus, as bystanders are typically in a position to respond to someone in distress. I review the evidence from a variety of sources that empathic distress functions as a prosocial moral motive, but most of the chapter is taken up with various modes of empathic arousal. If empathy is the product of natural selection, as I have argued elsewhere (Hoffman, 1981), it must be a multidetermined response that can be aroused by cues of distress coming from the victim or the victim’s situation. Empathy is indeed multidetermined, and I discuss 4 Introduction and Overview five distinctly different modes of empathic arousal. These include three that are preverbal, automatic, and essentially involuntary: motor mimicry and afferent feedback; classical conditioning; direct association of cues from the victim or his situation with one’s own painful past experience. The empathy aroused by these three modes is a passive, involuntary affective response, based on the pull of surface cues, and requires the shallowest level of cognitive processing. This simple form of empathic distress is important, however, precisely because it shows that humans are built in such a way that they can involuntarily and forcefully experience another’s emotion – that their distress is often contingent not on their own but someone else’s painful experience. The three preverbal modes are crucial for arousing empathy in childhood especially in face-to-face situations, but they continue to operate and provide empathy with an important involuntary dimension throughout life. They not only enable a person to respond to whatever cues are available, but they also compel him to do it – instantly, automatically, and without requiring conscious awareness. There are two higher-order cognitive modes: mediated association, that is, association of expressive cues from the victim or cues from the victim’s situation with one’s own painful past experience, where the association is mediated by semantic processing of information from or about the victim; and role-or perspective-taking, in which one imagines how the victim feels or how one would feel in the victim’s situation. These modes may be drawn out over time and they may be subject to voluntary control, but if one is paying attention to the victim they can be involuntary and triggered immediately on witnessing the victim’s distress. What they contribute to a person’s empathic capability is scope; they also enable a person to empathize with others who are not present. The existence of multiple arousal modes bears on my definition of empathy as not requiring, though often including, a close match between observer’s and victim’s affect. The many modes of empathic arousal assure a certain degree of match, even across cultures (as will be discussed), for two reasons: mimicry, which may be automatic and neurally based, assures a match when observer and victim 5 Empathy and Moral Development are in face-to-face contact; conditioning and association assure a match because all humans are structurally similar and process information similarly and are therefore likely to respond to similar events with similar feelings. But there are times when empathy does not require a match and, indeed, may require a certain mismatch, as when a victim’s life condition belies his feelings in the immediate situation. These are the times when verbal mediation and role-taking may take center stage. My theoretical framework for the development of empathic distress is presented in chapter 3, which is a key chapter in the book. In it I argue for a developmental synthesis of children’s empathic affect and their development of a cognitive sense of others as distinct from themselves. The synthesis results in five ‘‘stages’’ in the development of empathic distress: (a) reactive newborn cry; (b) egocentric empathic distress, in which children respond to another’s distress as though they themselves were in distress; this happens during the developmental interval in which they can feel empathic distress (from early preverbal arousal modes) but still lack a clear distinction between self and other; (c) quasi-egocentric empathic distress, in which children realize the distress is the other’s, not their own, but confuse the other’s inner states with their own and try to help by doing for the other what would comfort themselves; (d) veridical empathic distress, in which children come closer to feeling what the other is actually feeling because they now realize that the other has inner states independent of their own; (e) empathy for another’s experience beyond the immediate situation (e.g., chronic illness, economic hardship, deprivation), when children realize that others have lives that may be generally sad or happy; and a subcategory, when children can empathize with an entire group (homeless; Oklahoma City bombing victims). I also present evidence for my hypothesis that beginning with stage (c), children’s empathic distress is transformed in part into a feeling of sympathetic distress or compassion for the victim, and from that time on when children observe someone in distress they feel both empathic and sympathetic distress. My use of the term empathic distress throughout the rest of the book refers to this empathic/sympathetic distress combination. 6 Introduction and Overview In this developmental scheme, each stage combines the gains of the previous stages. At the most advanced stage, one is exposed to a network of information about the victim’s condition, which may include verbal and nonverbal expressive cues from the victim, situational cues, and knowledge of the victim’s life condition. These sources of information are processed differently: empathy aroused by nonverbal cues is mediated by the largely involuntary, cognitively shallow processing modes (mimicry, conditioning, association). Empathy aroused by verbal messages from the victim, a third party’s description of the victim’s state or condition, or one’s personal knowledge about the victim requires more complex processing (mediated association, role-taking). At the most advanced stage, observers may act out in their minds the emotions and experiences suggested by the above information and introspect on all of it. In this way they gain understanding and respond affectively to the circumstances, feelings, and wishes of the other, while maintaining the sense that this person is separate from themselves. When their information about the other’s life condition contradicts the other’s behavior in the immediate situation, their empathy can be as influenced, possibly more influenced, by the other’s life condition than by his or her immediate behavior. It should be clear by now that cognition plays an important role in development of empathic distress. Cognition is highlighted even more in chapter 4, where I note the human tendency to explain events causally and show how attributions about the cause of another’s distress can shape empathic distress into four empathy-based moral affects. When the cause is beyond the victim’s control (illness, accident, loss), observers’ empathic distress is transformed at least partly into sympathetic distress which is like the developmental transformation of empathic into sympathetic distress discussed in chapter 3. If someone else is the cause, one’s empathic distress is transformed into empathic anger, which consists of either empathy with the victim’s anger or a dual feeling of empathic sadness or disappointment (if that, rather than anger, is how the victim feels) and anger at the culprit. The latter, dual type of empathic anger may be prevalent in societies like ours in which, owing to socialization, 7 Empathy and Moral Development direct anger is not easily felt. It is also another case in which empathy involves a mismatch between observer’s and victim’s feelings. When a discrepancy exists between the victim’s character and the victim’s fate (a good person fares badly), the apparent violation of reciprocity or justice may transform an observer’s empathic distress into an empathic feeling of injustice. And, finally, when observers do not help, or their efforts to help fail, even for legitimate reasons, their view of themselves as causing the victim’s continuing distress may transform their empathic distress into guilt over inaction. It goes without saying that people’s empathic distress can be reduced by blaming the victim for his or her own distress. An important point about the bystander model is that to respond with empathic distress and the various empathy-based affects does not require the victim to be physically present. Because of the human capacity to represent events and imagine oneself in another’s place, and because of the power of represented events to evoke affect, to feel empathic distress one need only imagine victims, as when reading about someone’s misfortune, arguing about economic or political issues that involve victims or potential victims, or even making Kohlberg-style judgments about hypothetical moral dilemmas. One can also turn an abstract moral question into an empathy-relevant one by imagining a victim, say, of corporate downsizing, and how he feels. The ability to represent thus expands the importance of empathic morality beyond the face-to-face encounters of children and members of primary groups, which has been the focus of most of the research. It expands the bystander model to encompass a variety of situations limited not by the victim’s presence but by the observers’ imagination. As the bystander model is the prototypic moral encounter for empathy, especially empathic distress, the transgression model is the prototypic moral encounter for empathy-based transgression guilt (in contrast to bystander guilt over inaction). The transgression model also highlights children’s early socialization at home and is the prototypic encounter for moral internalization. The moral issues are these: What motivates a person to avoid harming others and to consider their needs, even when their needs conflict with his or her 8 Introduction and Overview own? When one does harm another, does one feel guilty afterward? When one contemplates acting in an instrumental, self-serving way that one realizes may end up harming someone (though that was not one’s intention), does one anticipate feeling empathic distress and guilt? Exactly what is meant by moral internalization? These issues are dealt with in part II of the book, which includes chapters 5 and 6. Guilt and moral internalization are chapter 5 topics. There I describe empathy-based transgression guilt, adduce evidence that there is such a thing and that it functions as a prosocial moral motive, and speculate about the developmental processes in its formation. I also point up the importance of moral internalization, which I define simply as follows: A person’s prosocial moral structure is internalized when he or she accepts and feels obligated to abide by it without regard to external sanctions. That is, the rewards and punishments that may have previously motivated one to consider others have lost most of their force and one now experiences the motive to consider others as deriving autonomously from within oneself. The various conceptions of moral internalization – Freudian, social-learning, cognitive-developmental, attributional, information-processing – are reviewed in this chapter. Those that I found most useful are integrated into the theory of guilt development and moral internalization presented in chapter 6. My definition of an internal moral motive is that it: (a) has a compelling, obligatory quality, (b) is experienced as deriving from within oneself, (c) makes one feel guilty when one acts or considers acting in ways that may harm others, (d) disposes one to consider another’s needs even when they conflict with one’s own. When such a conflict exists, the empathy arousing processes that work in bystander situations may not be powerful enough to motivate one to act prosocially. To create prosocial motives that are powerful enough to operate in conflict situations requires parents to actively socialize the child to consider others. Parents interact with children in many ways but only in discipline encounters do they make the connections necessary for guilt and moral internalization: that is, connections between the child’s egoistic 9 Empathy and Moral Development motives, the child’s behavior, and the harmful consequences of the child’s actions for others. And only in discipline encounters do parents put pressure on children to control their behavior and consider the needs and claims of others. If parents do this right, they can give children the experience of controlling their behavior through their own active processing of information about the consequences of their actions for others, which contributes to their developing an empathybased internal motive to consider others. Doing it right means using inductions when the child harms or is about to harm another. Induction highlights both the victim’s distress and the child’s action that caused it and has been found to contribute to the development of guilt and moral internalization in children. My explanation is this: Most parental discipline has powerassertive and love-withdrawing components that put pressure on the child to attend to the parent: too little pressure and the child may ignore the parent; too much, and the emotions aroused (hostility, fear) may prevent children’s effective processing of inductive information and direct their attention to the consequences of their action for themselves. A salient induction that fits the child’s cognitive level and puts just enough pressure on the child to process the induction’s information and attend to the consequences of the child’s action for the victim may arouse empathic distress and guilt (through the arousal mechanisms described earlier). In this way parents can exploit and build upon an ally that exists within the child – his or her empathic proclivity – and create a moral motive that may compete with the child’s egoistic motives. When the child experiences, repeatedly, the sequence of transgression followed by parent’s induction followed by child’s empathic distress and guilt feeling, the child forms Transgression → Induction → Guilt scripts, which have motive properties due to their empathic distress and guilt components. When a script is activated for the first time in an actual situation involving conflict with others, its motive component may not be strong enough to overcome the prospect of egoistic gain. But it may become strong enough with repetition, and when combined with cognitive development and peer pressure it may be effective. That is, peer pressure compels children to realize 10 Introduction and Overview that others have claims; cognition enables them to understand others’ perspectives; empathic distress and guilt motivate them to take others’ claims and perspectives into account. These prosocial moral scripts are not passively acquired but actively formed by children in a continuing process of constructing, synthesizing, and semantically organizing inductive information and relating it to their own actions and the victim’s condition. This active mental processing makes the child’s internal cognitive and affective processes salient to the child, and the child experiences the scripts and their implicit norm of considering others as the child’s own construction and part of his or her internal motive system. Parental intervention is no longer necessary and the scripts, now Transgression → Guilt scripts, can be activated by the child’s own awareness of harming someone. When activated, a script’s associated guilt and motivation to make amends is felt by the child as coming from within him-or herself. The script can be activated in advance by the child’s thoughts and images about the harmful effects of his or her acts. The resulting anticipatory guilt is a motive against committing the act, and if the child does commit the act, he or she will feel guilty. In short, what chapter 6 suggests are the antecedent factors that can lead to the development of an early moral motive to consider others even when one’s needs conflict with theirs. Later experiences of various kinds expand this motive to areas of life not dealt with at home. These experiences also provide skills and competencies that serve the motive and help make connections between it and relatively abstract moral principles like caring and justice. The chapter also summarizes empirical evidence for the theory and takes up the issue of direction of effects. Once a child acquires Transgression → Guilt scripts, it should perhaps not be surprising that the scripts can be activated and trigger guilt feelings in the child whenever the child thinks he transgressed, even when he or she did not. I call this virtual guilt, and the presumed harmful acts, virtual transgressions. Virtual guilt is not a new concept: A definition of guilt in Webster’s Ninth New Collegiate Dictionary is ‘‘feelings of culpability especially for imagined offenses.’’ In chapter 7 I describe and try to explain several variations 11 Empathy and Moral Development of virtual guilt. One, ‘‘relationship’’ guilt, may be endemic to close relationships because they provide endless opportunities not only for hurting one’s partner but also for thinking one has. That is, relationship partners become so dependent on each other that their feelings and moods depend heavily on the feelings, moods, and actions of the other. More importantly, each partner knows the other is similarly dependent on him or her, and each partner may as a result develop a keen sensitivity to the potential impact of his or her words and deeds on the other. It may therefore seem reasonable when one’s partner is sad or unhappy and the cause is unclear not only to feel empathic distress but also to blame oneself for the partner’s state. One might not feel guilty if certain of one’s innocence, but that requires keeping accurate mental records of previous interactions with one’s partner, a type of emotional bookkeeping rare in close relationships. A related type of virtual guilt – ‘‘responsibility guilt’’ – stems from having responsibility for someone who is harmed, even when the facts clearly indicate that one was not at fault. What seems to happen is that one empathizes with the victim’s pain, reviews the situation in one’s mind, realizes that one could have acted differently and prevented the accident, shifts from I could have to I should have, blames oneself, and feels guilty. Whereas close relationships and positions of responsibility are the context for relationship and responsibility guilt, pursuing one’s normal developmental goals and interests can provide the context for virtual transgressions involving ‘‘developmental guilt.’’ A person may feel that by leaving home for college he will damage his parents – separation guilt; and by achieving more than his peers he contributes to making them feel inadequate – achievement guilt. A person may also feel guilty over his relative affluence, that is, over benefiting from privileges that others lack – guilt over affluence. Though adults may feel guilt over affluence, I class it with developmental guilt because it seems more prevalent in adolescents (at least it was in the 1960s) and may be a significant part of the prosocial moral development of those who experience it. 12 Introduction and Overview It is known that people who experience the traumatic death, injury, or other misfortune of someone else (in war, acts of terrorism, natural disaster, corporate downsizing), while they remain unharmed, often feel guilt over surviving. The guilt is compounded by conflicting emotions of joy at surviving and empathic sorrow for the victims. Add to this the hidden relief that the worst happened to someone else, and one can have a painful case of guilt – survival guilt. Guilt may be the survivor’s answer to the question, ‘‘Why me – why was I saved and not somebody else?’’ What this person is saying, and what survival guilt may have in common with guilt over affluence is that one cannot justify the advantage one has over the victim. One’s advantage therefore violates the principle of fairness or reciprocity; and the awareness of being advantaged may transform empathic distress for the victim into an empathic feeling of injustice and a feeling of guilt. The prevalence of guilt over one’s relative advantage, survival, and the other types of virtual guilt confirms my belief that humans, at least in our society, are ‘‘guilt machines.’’ In chapter 8, I shift from empathic motivation’s contributions to prosocial moral action, to its limitations, which result from empathy’s dependence on the intensity and salience of distress cues and the relationship between observer and victim. One limitation is that although we expect more intense empathic arousal with more salient distress cues, extremely salient distress cues can be so aversive that an observer’s empathic distress is transformed into an intense personal feeling of distress. This empathic over-arousal can move observers out of the empathic mode, cause them to be preoccupied with their own personal distress, and turn their attention away from the victim. An exception is that for people who are committed to a helping relationship (therapist–patient; parent–child) empathic overarousal may intensify empathic distress and motivation to help the victim. The second limitation is empathy’s vulnerability to two types of bias: familiarity bias and here-and-now bias. Although people tend to respond empathically to almost anyone in distress, they are vulnerable to bias in favor of victims who are family members, members 13 Empathy and Moral Development of their primary group, close friends, and people who are similar to themselves; and to bias in favor of victims who are present in the immediate situation. Empathy’s vulnerability to over-arousal and the two types of bias may not be a significant problem in small homogeneous ‘‘primary group’’ societies, or in bystander, transgressor, and virtual transgressor moral encounters involving one victim. Indeed, these limitations may have a hidden virtue: If people empathized with everyone in distress and tried to help them all equally, society might quickly come to a halt. Seen in this light, empathic bias and over-arousal may be empathy’s ultimate self-regulating, self-preserving mechanisms, which fits with the increasing evidence that the ability to regulate one’s emotions correlates positively with empathy and helping behavior. Still, empathic over-arousal and especially empathic bias may pose problems in encounters involving multiple claimants and encounters in which caring for another conflicts with the demands of justice. These problems can be reduced, I hypothesize, when empathy is ‘‘embedded’’ in a moral principle with which it is congruent, because this allows empathy to gain structure and stability from the principle’s cognitive dimension. Chapter 9 relates empathic affect to Western society’s prevailing moral principles: caring and justice. Empathy’s congruence with caring is obvious. It is also congruent with aspects of criminal justice, which involves victims, and this is briefly discussed. Most of the chapter deals with distributive justice, which pertains to how society’s resources should be allocated – ‘‘equally,’’ or according to one’s ‘‘need,’’ ‘‘effort,’’ or ‘‘merit’’ (competence, productivity). Empathy is congruent with all of these justice principles, but less so with competence and productivity. My argument that empathic arousal may alter one’s views of distributive justice can be summarized thus: If a person thinks about how society’s resources should be distributed, a self-serving perspective will make him prefer principles that coincide with his own condition: high producers will choose merit and low producers will choose need or equality. If empathy is aroused, the welfare of others will be considered and even high producers may 14 Introduction and Overview choose need or equality – or, more likely, merit regulated to prevent extreme poverty (need) and vast discrepancies in wealth (equality). Regulated merit is at the heart of the philosopher John Rawls’s theory of justice, notably the ‘‘difference principle,’’ which assigns great weight to how society’s ‘‘least advantaged’’ are treated. Rawls uses a ‘‘veil of ignorance,’’ which compels people whom he imagines are constructing a society from a rational, totally self-serving perspective – but without knowing their place in that society – to ensure that the least advantaged’s needs will be taken care of. Rawls’s other purpose in using the veil of ignorance is to rule out empathy, so that the difference principle would be derived on purely rational, selfserving grounds. I applaud Rawls’s approach but spend a lot time in this chapter arguing that empathy and the veil of ignorance are actually functionally equivalent, though operative in different contexts. Owing to empathy’s congruence with justice, people will empathize with victims of justice violations (someone cheated out of his earnings or whose rights are violated). When they do this they may be aware of both their empathic feeling for the victim (empathic distress, guilt, empathic anger, empathic feelings of injustice), and the activated justice principle. The resulting concurrence of empathic affect and a moral principle creates a bond between them, a bond that is strengthened by subsequent concurrences. In this way, moral principles, even when originally heard about in ‘‘cool’’ didactic contexts, may acquire empathy’s affective and motive properties and become emotionally charged representations or prosocial ‘‘hot cognitions.’’ The implications of this hot-cognition concept are twofold. First, when a moral principle is subsequently activated in a moral encounter or even in didactic or research contexts, empathic affect is aroused. This empathic affect will have two components: a stimulusdriven component (victim’s distress) and a principle-driven component. The principle-driven component will have a ‘‘heightening’’ effect or a ‘‘lowering’’ effect on the intensity of the stimulus-driven component. This should reduce the likelihood of empathic overarousal (and under-arousal) and thereby help stabilize the individ15 Empathy and Moral Development ual’s empathic affect across situations. The second implication is that the bystander and transgression models must be enlarged to include not only empathic affect aroused by a victim’s distress, but moral principles that may also be activated by the victim’s distress and that may help stabilize the bystander’s or transgressor’s empathic affect. Reciprocity underlies most justice principles: Good deeds should be rewarded, bad acts punished; punishments should fit crimes. I suggest reciprocity is not inherently prosocial, as it encompasses ‘‘eye-for-an-eye’’ as well as ‘‘hard-work-should-be-rewarded’’ thinking. But it can become prosocial when it is associated with empathy, as when reciprocity is violated by someone’s being treated unfairly. When that happens, reciprocity can intensify the observer’s empathic distress and transform it into an empathic feeling of injustice. Finally, empathy, alone or embedded in a moral principle, can play an important role in moral judgment. The basic argument for this was made over two centuries ago by David Hume: We obviously applaud acts that further our own well-being and condemn acts that may harm us; if we empathize with others we should therefore applaud or condemn acts that help or harm others; and, unless we are abnormally callous, we will feel indignant (empathic anger) when someone willfully inflicts suffering on others. I would add that most moral dilemmas in life may arouse empathy because they involve victims – seen or unseen – of one’s own actions or actions by someone else whom one is judging. Empathy can influence one’s moral judgment of oneself or of the other directly, or indirectly through the moral principles it activates. The developmental research on distributive justice, in which children are asked to allocate rewards to recipients who differ in productivity and other respects, is clear on what children of different ages view as fair. The research, discussed in chapter 10, shows a developmental trend from allocating rewards based on self-interest in preschoolers, to a strong preference for equal division of rewards at 4 or 5 years; to an increasing emphasis on reward in proportion to productive output or output integrated with need (poverty) among children 8 or 9 years and older. Older children also apply different 16 Introduction and Overview justice principles according to the context: They favor a productivity rule in reward-for-work situations, equity in voting situations, and equality combined with need in charity situations. By 11 or 12 years, they favor ‘‘productivity’’ over ‘‘need’’ for strangers but equate the two principles for friends, and they allocate as much to a needy friend as to a productive stranger – not unlike adults. There is little developmental research on empathy’s contribution to justice. I suggest that parental inductions around sharing and turn-taking start the socialization for ‘‘equality.’’ ‘‘Equality’’ is also fostered by preschool and kindergarten teachers and by direct pressure from peers who want to have their share. I suggest socialization for ‘‘effort’’ also begins at home but is not systematic until early elementary school when children’s academic performance is assessed and rewards given for self-improvement, which, above all, requires effort. Socialization for ‘‘productivity’’ and ‘‘competence’’ begins in earnest when academic performance is based on comparison with one’s classmates, in grade four or five, and continues through the rest of one’s education and on into the world of work. These socialization experiences are integrated with children’s direct justice-relevant experiences such as feeling distressed when treated unfairly (not rewarded for hard work), their observations that others feel distressed when similarly treated, and their empathic responses to that distress. These direct justice-relevant experiences build on the children’s empathically charged transgression-guilt scripts about sharing acquired at home. The result is a network of integrated experiences that provide raw material from which children can construct an increasingly complex empathy-based sense of fairness and concern for others. With language, they can classify certain acts as morally wrong, unfair, and (eventually) form them into more general, abstract but still empathically charged principles of justice. Language also enables children to begin, on their own and in conversations with others, to make their own moral inferences in light of the interpretations, explanations, and emotional reactions of adults and their own cognitive and emotional reactions as bystanders and victims. Each child does not construct a moral code anew, as 17 Empathy and Moral Development some cognitive developmentalists claim, but is active nonetheless in reconstructing and understanding moral rules, using information communicated by adults and his own experience. The above can be put in terms of a division of labor between parental inductions that communicate rules of fairness and carry the force of authority, children’s ability to decenter and preference for reciprocity, and peer interactions which highlight equality: Peers advancing their own claims compel one to realize that one’s desire is not the only thing that must be considered; decentering and reciprocity enable one to understand the basis of another’s claims; inductions, acting on one’s natural empathic proclivity, make one receptive to those claims. The resulting empathy-based fairness concepts are shaped further by the values communicated by parents, peers, teachers, religion, media. Children with these experiences are well versed in rudimentary forms of our society’s caring and justice principles. These processes are haphazard until adolescence, when children are more ‘‘formally’’ introduced to moral principles that are supposed to guide behavior. It is then, if ever, that the individual’s active role in constructing a moral code, evident throughout childhood, takes center stage. The raw materials continue to be the products of socialization, as discussed. These include empathy-charged justice/fairness scripts generated in discipline encounters by inductions bearing on sharing and effort, which are enhanced by emotionally salient personal experience as bystander and victim, and by exposure to the media. One thinks and reasons about these, and in debates, especially with peers, one may analyze, interpret, compare and contrast, and accept or reject them and thus construct one’s own set of general, to some extent abstract though emotionally charged moral principles. When one has internalized and committed himself to caring or justice principles, realizes one has choice and control, and takes responsibility for one’s actions, one has reached a new level. One may now consider and act fairly toward others, not only because of empathy but also as an expression of one’s internalized principles, an affirmation of one’s self. One feels it is one’s duty or responsibility to 18 Introduction and Overview consider and be fair to others. This connection between self, principle, and duty may in some cases result from an emotionally powerful ‘‘triggering event’’ (extreme injustice) that causes one to reexamine one’s life choices and leads to a new moral perspective and sense of social responsibility. Behaving in accordance with a moral principle is not always a simple matter of lining one’s actions up with the principle. Moral encounters often involve multiple claimants, situations in which bystanders must choose which victims to help, and some encounters involve conflict between caring and justice. Both types of moral encounters are discussed in chapter 11. Multiple claimant dilemmas in the caring domain that come to mind are people drowning or caught in a burning building, when one must choose whom to help; a doctor deciding whether to perform an abortion, when the claimants are the fetus, the pregnant teenager, and the teenager’s parents; a lawyer deciding whether to defend someone he believes is guilty of murder, when the claimants are the defendant who has a right to a trial, his future victims if he goes free, and the victim’s family who want him punished; Kohlberg’s hypothetical World War II air-raid warden who had the choice of remaining at his post or leaving to help his family whose part of town had just been bombed; the similar but real dilemma of a nurse who was helping an Oklahoma City bombing victim when she heard the second bomb blast. The moral issue for ethics in these dilemmas is which claimant should one help. The issue for science is who will one help. Evolutionary biology’s answer is simple: One helps those with whom one shares the most genes. Psychology’s answer is that when there is one claimant, bystanders empathize with virtually anyone in distress (chapter 2). When there are multiple claimants, one will probably empathize with family members and others who fit empathy’s familiarity and here-and-now biases (chapter 8), although one may feel guilty over those one does not help. In other words, evolutionary psychology says we choose to help those who share our genes; psychology says we choose to help those in our primary group. But we share more genes with those in our primary group, which raises certain questions. Is psychology’s answer fundamentally the same as 19 Empathy and Moral Development evolutionary biology’s? Is empathic bias the functional equivalent of sharing another’s genes? The answer to both questions may be yes, given the argument that empathy derived from natural selection pressures in human evolution (Hoffman, 1981). In any case, in multiple claimant situations empathy may not be enough. Kant and his followers, including Rawls and Kohlberg, claim that caring is subordinate to justice because caring is usually personal and particularistic, involves decisions that are affectively rather than rationally based, and lacks the formal properties of justice. I prefer to view caring and the different types of justice as ‘‘ideal types’’ that may occur in varying degrees in all situations. When caring and justice co-occur they may be congruent. They may also conflict, as when a professor is convinced by a student’s plea that his ‘‘life will be wrecked’’ if he does not get a higher grade; as when a student is asked by a friend for the questions on an exam he just took; as in Kohlberg’s famous dilemma based on Les Miserables, in which a man steals a drug to save his wife’s life. The last two examples of caring dilemmas given earlier (Kohlberg’s air-raid warden and the Oklahoma City nurse) can be considered caring-versus-justice dilemmas if we classify violations of a person’s duty or responsibility as criminal acts or as instances of nonreciprocity between role demands and behavior. To illustrate multiple claimant and caring–justice encounters in depth, I use the dilemma of a professor who is asked to write a letter of recommendation for one of his students who is applying for an important job. The student is good but not outstanding. If the professor has some friendship with the student and knows other things about him (such as that there is a sick child in his family), he might write a strong letter of support. But things get complicated if the professor also empathizes with the colleague who needs an especially outstanding applicant or with the other unknown candidates who also need the job. The dilemma so far is confined to the caring domain, but justice issues are also relevant: The academic system places high value on merit (scholarly output, competence) and the integrity of the system rests on recommenders’ candid assessments of job applicants, which the professor’s colleague expects from him. 20