Moral Development and Character Formation
Larry Nucci
University of Illinois at Chicago
Copyright © 1997, Larry Nucci.
In Walberg, H. J. & Haertel, G. D. (1997). Psychology and educational practice. Berkeley: MacCarchan. p. 127-157.
As in many areas of educational research, the field of moral education is rife with controversy. These disputes are not limited to psychological accounts of the nature of moral development or character formation, but extend to the very definition of educational aims in this area. Arguments surrounding the aims of values education capture the essential quandary for any pluralist democracy attempting to construct a shared civil society without privileging the particular values of any one group. At the heart of the matter is whether we can point to a set of moral values that would form the basis of an "overlapping consensus" that would permit approaches to moral education that appeal to more than local or particularistic values. Without such consensus the incommensurable qualities of local values would render shared notions of a moral community impossible. A related issue is whether there are features of individual psychology which can be appealed to in fostering the development of children who would act in accordance with such common or transcendent moral values. The contribution which educational psychology can make with regard to these issues is to clarify how moral and social values are formed, and to address the social and psychological factors which contribute to the tendency of individuals to act in ways that are concordant with their own well-being and the welfare of others. Controversies notwithstanding, the past several decades have witnessed a great deal of progress with regard to our understanding of these issues. This chapter will address those aspects of what has been learned in the areas of developmental and educational psychology that can help educators engage in meaningful moral and character education.
Historically, these issues have been approached from two perspectives with divergent, though overlapping, interests and differing sets of assumptions about the nature of social development and socialization. On the one hand have been traditional character educators (Ryan & McLean, 1987) whose emphasis has been on processes of internalization and self control that would ostensibly result in virtuous conduct. On the other hand have been cognitive developmentalists (Power, Higgins & Kohlberg, 1989) whose emphasis has been on the development of structures of moral reasoning which ostensibly underlie action choices. These two points of view, one (character education) emphasizing non-rational mechanisms of self-control and behavioral follow-through, and the other (cognitive developmental) emphasizing rationality in the form of moral decision-making, are irreconcilable in terms of underlying philosophy. Moreover, at the level of educational policy, these two perspectives have tended to take turns over the past thirty years in dominating the attention of educators interested in fostering children's social as well as academic growth. Yet, neither point of view, in their traditional form provides a sufficient basis for guiding educational policy. There can be no meaningful moral action in the absence of moral judgment since morality by definition requires choice and intent. Thus, proponents of educational policies that ignore the development of moral decision making have generally absented themselves from offering suggestions in the area of moral education. Conversely, the development of moral judgment, though necessary, is not a sufficient aim of moral education (Power et. al., 1989). For moral judgment, in and of itself, does not lead to a particular course of action. It would appear then, that the divergent aims of these two points of view for the development of moral reasoning, and the development of characterological propensities for moral follow through are both desirable and necessary components of any educational contribution to children's moral growth. This chapter, therefore, will pull information from diverse areas of research to construct a coherent picture integrating both sets of concerns. In order to reconcile these points of view, I will need to appeal to contemporary cognitive theories of personality formation (Sarbin, 1986), and recent work on the development of the moral "self" (Noam & Wren, 1993), rather than restrict discussion to the behavioral and social learning perspectives that have been traditionally used in support of character education. We begin this discussion by looking at the issue of moral judgment.
What Do We Mean by Morality? A large part of the controversy surrounding moral or character education has to do with how morality is to be defined. In everyday discourse morality refers simply to the norms of right and wrong conduct. At issue, however, is what is meant by moral right and wrong, and whose criteria shall be used to judge the wrongness of actions. As it turns out, this diversity at the level of public opinion, has a corollary in the underlying heterogeneity of the structures of the individual's social concepts. Within the individual, concepts of social right and wrong are not all of one type, but are organized within distinct conceptual and developmental frameworks. In research conducted over the past twenty years, it has been discovered that individuals treat some forms of social behavior as moral universals, other forms of social conduct as subject to determination by local cultural or social norms, and still others as matters of personal choice (Turiel, 1983). More specifically, these conceptual differences emerge when formal criteria for morality are employed which define morality as those interpersonal behaviors that are held to be right or wrong independent of governing social rules, and maintained as universally binding (Turiel, 1983). Prescriptions which meet these criteria are those which refer to actions, such as hitting and hurting, stealing, slander, which have an impact on the welfare of others. Accordingly, concepts of morality have been found to be structured by underlying conceptions of justice and welfare (Turiel, 1983). Morality, then, may be defined as one's concepts, reasoning, and actions which pertain to the welfare, rights and fair treatment of persons.
Morality (defined in terms of justice, welfare, rights) can be distinguished from concepts of social conventions, which are the consensually determined standards of conduct particular to a given social group. Conventions established by social systems such as norms or standards of dress, how people should address one another, table manners and so forth derive their status as correct or incorrect forms of conduct from their embeddedness within a particular shared system of meaning and social interaction. The particular acts in and of themselves have no prescriptive force in that different or even opposite norms (e.g., dresses for men, pants for women) could be established to achieve the same symbolic or regulatory function (e.g., distinguishing men from women). The importance of conventions lies in the function they serve to coordinate social interaction and discourse within social systems. In keeping with this definition, concepts of social convention have been found to be structured by underlying conceptions of social organization (Turiel, 1983).
The distinctions which have been drawn between morality and social convention have been sustained by findings from over 50 studies conducted since 1975. This research has indicated that children, adolescents, and adults treat violations of morality, such as harming another, as wrong whether or not there is a governing rule in effect, and generalize these judgments of wrongness to members of other cultures or groups which may not have norms regarding such actions. Conventions, on the other hand, are viewed as binding only within the context of an existing social norm, and only for participating members within a given social group. While there is some controversy over whether the distinction between morality and convention is made by members of all cultural groups, a number of studies have demonstrated that subjects from a wide variety of the world's cultures do differentiate between matters of morality and convention. Evidence in support of the morality/convention distinction has been obtained from subjects in Brazil, India, Israel (Arab and Israeli subjects), Korea, Nigeria, Virgin Islands, and Zambia. Moreover, recent research has demonstrated that something parallel to the distinction between morality and social convention operates within the moral and normative conceptions of religious children and adolescents with respect to their conceptions of religious rules. It has been found (Nuuci, 1989) that children and adolescents from observant religious groups (Amish-Mennonite and Orthodox Jews) judged certain religious norms (e.g., day of worship, work on the Sabbath, baptism, circumcision, wearing of head coverings, women leading worship services, premarital sex between consenting adults, keeping Kosher) in conventional terms in that they regarded these as contingent on religious authority or the word of God, and as particular to their religion. In contrast, moral issues (e.g., stealing, hitting, slander) were regarded as prescriptive (wrong to do) independent of the existence of a rule established by religious authority or by God's word, and as obligatory for members of all other religious groups.
The discovery of these psychological distinctions between moral and conventional forms of social right and wrong provides an empirical basis for beginning to address some of the definitional issues vexing moral education (Nucci, 1989). In differentiating what is moral from what is socially "proper", these findings can allow educators to focus the discussion of moral education on questions of how best to develop children's moral understandings (i.e., concepts of welfare and fairness), and their tendencies to act in accord with such moral principles, rather than being captured by heated arguments over which set of local conventions or religious norms ought to be included within the collection of values to be addressed by the curriculum. In keeping with the broad cross cultural generalizability of this research, the identification of morality as centered around issues of justice and human welfare is consistent with common sense construals of the basic task of values education as fostering the development of people who don't lie, cheat, steal, or hurt others. At the same time, the grounding of these definitions of the content of morality in basic developmental research, avoids falling into the trap of what Lawrence Kohlberg so aptly and perjoratively labeled the "bag of virtues" approach to establishing the aims of moral education. These core moral concerns for fairness and welfare are not virtues in the usual sense, but constitute the central issues for moral judgments and consequent actions.
The distinction between morality and convention also allows the educator to give convention its due. Earlier analyses of children's moral development, such as Kohlberg's stage theory (see Power et. al., 1989), interpreted attention to convention as characteristic of the reasoning of persons at lower stages of moral development, and therefore as something to be overcome through moral education. As stated above, concepts of conventions are now understood as distinct from moral understandings, and structured by children's and adolescents' emerging conceptualizations of social systems and social organization. Conventions are constituent elements of social systems. Just as morality is fundamental to interpersonal interaction, conventions are essential to the operation of society. The development of children's and adolescents' understandings of the functions and purposes of social convention, therefore, have educational worth in their own right (Nucci, 1989). In sum, current research on the structure of children's social concepts, provides an empirical basis for differentially addressing development within each of these conceptual systems rather than reducing either morality or conventional norms to a single framework.
Context, and the Inevitability of Controversy. While the discovery of distinct domains of social knowledge can help to focus the aims of moral education by identifying the core content of morality (Nucci, 1989), the heterogeneity in people's social understandings, and the contextual overlap of moral and non-moral normative components in everyday life means that an honest approach to moral education will always need to contend with contradiction and controversy. Such overlap is inevitable given that all social interactions take place within societal systems framed by conventions. Thus, although many everyday issues are straightforward instances of either morality or convention, many others contain aspects from more than one domain. In such cases, people may differ from one another in terms of the information they may bring to a situation, or the weight they may give to one or another feature of a given issue. Two basic forms of overlap occur between morality and convention. In one form, called domain mixture, conventional norms sustaining a particular organizational structure are in harmony or conflict with what would objectively be seen as concerns for fairness or rights. Examples of such overlap would be conventions for lining up to purchase tickets, or gender role conventions which proscribe areas in which men or women may participate. In the former case, the convention (lining up), while a morally neutral and arbitrary way to arrange people, could be used to serve a distributive justice function (turn taking), and cutting in line would, therefore, become unfair. In the latter case, the convention (gender role) may be in conflict with fairness if the convention prevents members of one gender from obtaining opportunities afforded the other. The second type of moral/convention overlap, labeled second order moral events, occurs when the violation of a strongly held convention is seen as causing psychological harm (insult, distress) to persons maintaining the convention. In our culture, for example, attending a funeral in a bathing suit would generally be seen as insensitive toward the deceased and the grieving family, and not merely an instance of unconventional conduct.
In responding to issues that involve elements from more than one domain, individuals may either subordinate the issue to a single dimension and reduce an issue of overlap to one that is primarily either moral or conventional, or engage in an effort to coordinate the multifaceted nature of the issue taking the moral as well non-moral aspects of a given situation or event into account. These responses to overlap at an individual level help to account for the inconsistencies we observe within people as they respond to events in different contexts (a subject I will take up later). They also help to explain how cultural groups, or subgroups arrive at different readings of social issues they consider to be morally neutral or charged with moral meaning. Within western society, those instances of overlap in which convention and morality are in harmony account for the moral component of what is generally viewed as mannerly and respectful conduct, and for the moral aspect of norms and procedures that sustain participatory forms of government. For the most part, within democratic societies, these areas of overlap are non-controversial inasmuch as they represent values concordant with morality and the conventional status quo. Consequently, values education programs purporting to foster such conventional values enjoy wide public support. Controversies are likely to emerge, however, whenever the relations between moral and non-moral components of issues are not in accord, and, therefore likely to be viewed differently by the affected parties. In the case of second-order issues such as disputes over what constitutes modesty in forms of dress (e.g., women's skirt length), the essentially conventional nature of such issues generally allows for local consensus to settle the matter. While some civil libertarians might protest any constraint on student choice with respect to personal conduct, and some conservative religious people might protest as immoral, any alteration in the norms of public conduct, such second-order issues are generally resolved by elected school boards, or school policy. More problematic for curriculum designers and educational policy makers are potential conflicts between morality and convention embedded within the norms that sustain existing social order, and by implication benefit members of the privileged social classes.
An illustrative example of this type of issue is nicely captured in the following incident described by Maya Agelou (1971, pg. 39) in her novel, "I Know Why the Caged Bird Sings." The passage recalls an incident in which a local judge mistakenly refers to Maya Angelou's grandmother by the title, Mrs. The use of the title was a mistake, because the conventions of the depression era south decreed that whites, but not blacks be referred to by titles. Through the discriminatory use of titles, whites symbolically maintained their social supremacy over blacks. In the situation described by Maya Angelou, her grandmother was subpoenaed to give testimony before the judge. She writes:
The judge asked that Mrs. Henderson be subpoenaed, and when Momma arrived and said she was Mrs. Henderson, the judge, the bailiff and other whites in the audience laughed. The judge had really made a gaffe calling a Negro woman Mrs., but then he was from Pine Bluff and couldn't have been expected to know that a woman who owned a store in that village would also turn out to be colored. The whites tickled their funny bones with the incident for a long time, and the Negroes thought it proved the worth and majesty of my grandmother.
From the vantage point of our current understanding of racial prejudice and segregation within American society, the treatment accorded to Maya Angelou's mother was clearly unjust and immoral. That is, despite the arbitrary and conventional nature of titles, we now generally acknowledge that their discriminatory use as depicted in this particular context served the immoral purpose of symbolically subjugating and consequently humiliating African-Americans. What is interesting in this example is how the parties at that time viewed the employment of the titles Mr. and Mrs. In general situations these titles serve to convey hierarchical relations between adults and children, and in formal situations establish respect between adults of equal status. In the situation described by Angelou, they were used to establish the socially inferior position of African-Americans (Negroes) in relation to whites. From the white position of power, the judge's "gaffe" was a source of humor, because the judge could not have intended to elevate a Negro woman to the same status as a white. For those whites who viewed the use of first names, when addressing adult African- Americans, as sustaining the social system, and for those African- Americans who might have accepted the status quo, the issue was one of conventionality, and the judge's gaffe, simply a humorous error. From that vantage point, the judge as a white man, would have been right, and had the right to refer to Maya Angelou's grandmother by her first name. On the other hand, for those African Americans in Maya Angelou's community, who viewed it as unjust to employ titles to symbolically maintain them in an inferior social position, the issue was one of morality, an understanding that could only be arrived at by coordinating concerns for fairness with concepts of conventions as constitutive elements of the social order. From that vantage point, the judge's implicit acknowledgment of her grandmother's social accomplishments (being a store owner) put her on an even footing with whites, and served as a source of pride, and confirmation that the discriminatory social practices they endured were artificial and unsustainable in the light of an objective view of the situation. From the point of view of the African-American community, the judge's "gaffe" inadvertently correctly captured the right of Maya Angelou's grandmother to be called Mrs.
While this incident from America's past is easy to look at in the cool light of history, such issues are not easily dealt with when they concern contemporary practices. Two of the elements that lend to the difficulty in dealing with these issues are captured in the above example. First, the conventionalized practices (e.g., forms of address, modes of dress) in and of themselves are morally neutral and play themselves out in the course of everyday life. Thus, people steeped in a particular way of life may not be cognizant of the moral implications of their particular social system. The problem here, of course, is that inequitable systems may simply perpetuate themselves, and educational curricula based on transmitting the values of the community may become the hand-maidens of immorality. Second, because such issues are multifaceted, they lend themselves to more than one interpretation, raising the specter of controversy for the educational system. This is particularly problematic for public schools since such controversies tend to have political ramifications. Generally, people in positions of relative power and privilege are more likely to view such issues in conventional terms and favor maintenance of the status quo, since the conventional system serves their personal interests. People on the receiving end of such conventionalized inequities, on the other hand, are less likely to subordinate such overlapping issues to convention, and more likely to be cognizant of their moral status. The moral dilemma these overlapping issues pose for educators is how to allow students to address the moral contradictions posed by some of society's conventions in areas such as gender and race relations without themselves becoming subject to the positions held by political groups that inevitably align themselves with one or another side of such issues.
Moral Diversity and Informational Assumptions Variations in the moral meanings people attribute to particular actions stem not only from the areas of overlap between morality and the conventions of social systems, but also arise as a result of the differences in factual assumptions people make about given acts. Within our own culture, for example, people hold different views about whether it is morally wrong or all right to engage in the physical punishment of children. In her research on this issue, Wainryb (1991) found that pro corporal punishment parents held the view that this behavior was all right because it was a highly effective, educative act rather than one of unprovoked harm or abuse of the child. When such parents were presented with information that spanking is no more effective than other methods of disciplining children, significant numbers of parents shifted in their view of corporal punishment and maintained that it was not all right for parents to engage in the behavior. Conversely, when parents who maintained that it was wrong to engage in corporal punishment were presented with information that experts had found spanking to be the most efficient method to teach young children, there was a tendency for such parents to shift toward a view that corporal punishment would be all right.
In the above example, the morality of an action shifted as a function of the informational assumptions people had regarding the effect of the act. In other cases, informational assumptions can alter people's views of the moral culpability of the actor. Many people in our culture, for example, have been found to hold to the view that homosexuality is an immoral lifestyle choice (Turiel, Hildebrandt, & Wainryb, 1991). From that perspective, being a homosexual entails a conscious decision to engage in behavior which they consider to be offensive and indecent. Leaving aside such questions as to whether homosexuality should be viewed in such normative terms or as a matter of private, personal conduct, the issue of choice is central to whether the individual may be held accountable for his or her sexual orientation. Information that would bear on that issue (e.g., findings of a substantial genetic component in determining sexual orientation) would undoubtedly impact the moral evaluation many people would make of homosexuals, even if it had no impact on their view of homosexual acts.
In sum, the moral worlds within which people act out their lives are affected by informational as well as contextual variables which enter into the evaluations people generate about particular courses of action. As with issues of domain overlap, the impact of new information regarding the causes or effects of social behaviors both complicates and enriches the role of education in preparing students to deal with social and moral issues. From a policy standpoint, we are once again confronted with the need to recognize that values education within a pluralistic, information rich, democratic society means preparing students to coordinate fundamental moral understandings of fairness and human welfare with potentially changing conventions and informational assumptions.
Understanding Inconsistencies in Individual Conduct The existence of domain overlap and differences in informational assumptions helps to account not only for disagreements between people about the moral meaning of social issues, but also helps to explain some of the inconsistencies we observe within individuals. Just as different groups of people may disagree over the moral meaning of contextualized social issues, individuals may differ in their attribution of moral meaning of actions within different contexts. In relatively unambiguous cases, deciding upon the moral or conventionally correct course of action is fairly straightforward. However, in cases where moral and conventional expectations are in conflict, where the information regarding the meaning of the action is ambiguous, or when moral concerns run counter to highly salient pragmatic or personal desires of the actor, individuals display inconsistency in their social judgments and subsequent actions. From the point of view of domain theory, such inconsistencies are the inevitable consequence of applications of a multifaceted conceptual framework in the context of varying heterogeneous social contexts. This is not a case of situational ethics. People do not make up their morality on the spot. The place of morality within a given context, however, will vary as a function of the person's application of the totality of their social understandings and concerns to a given situation.
Reconceptualizing the Developmental Aims of Moral Education Given what we currently know about moral cognition, it is sensible to propose that the core focus of moral education be upon students conceptions of fairness, human welfare and rights, and the application of those moral understandings to issues of everyday life (Nucci, 1989). Research on the development of children's moral understandings has shown that morality begins in early childhood with a focus upon issues of harm to the self and others. Preschool aged children are very concerned with their own safety, and understand that it is objectively wrong to hurt others. Even three - year - olds, for example, understand that it is wrong to hit and hurt someone even in the absence of a rule against hitting because, "When you get hit, it hurts, and you start to cry." Young children's morality, however, is not yet structured by understandings of fairness as reciprocity. Fairness for the young child is often expressed in terms of personal needs and the sense that one isn't getting one's just desserts. "It's not fair.", often means, "I didn't get what I want.", or that someone's actions caused the child to experience harm. By age 10 nearly all children have constructed an understanding of fairness as reciprocity (treating others as one would wish to be treated), but have difficulty in coordinating their sense of fairness as equality with notions of equity. Expanding the sense of fairness to include compassion, and not raw justice, and to tie that sense of compassionate justice to a conceptually compelling (logically necessary) obligation to all people and not just the members of one's community is the developmental task of adolescence and adulthood.
Similar research on the development of children's understandings of social convention (Turiel, 1983) indicate that constructing an understanding of why conventions matter is a long process. Unlike morality, there is nothing intuitively obvious about the functions of convention. Even though most children have learned the content of their society's conventions by early elementary school, the purpose of such rules is not easily understood. In fact it is not until middle to late adolescence that children develop a coordinated understanding of conventions as constituent elements of social systems. It is little wonder then that children so often seem disconnected from society's rules even when their normative content (e.g., "Don't talk with your mouth full.") has been repeatedly presented to them.
This developmental research can be of enormous value to educators interested in developing "good" children. It provides curriculum designers and classroom teachers a framework from which to direct educational efforts at moral education which are appropriate for students at different points in development, and provides a basis from which to differentially address both the moral and conventional dimensions of social values. In doing so, educators will contribute to the development of a fair and compassionate moral citizenry that also understands and respects the need for convention. As we have just seen, however, contextualized moral judgments may often call upon the person's ability to weigh or coordinate moral and non-moral considerations. Defining the aims of moral education in such circumstances becomes more complex. Put simply, the moral educator is not simply interested in developing the students moral and conventional understandings in such contexts, but is also interested in whether or not the student will be aware of, and prioritize the moral elements of such issues when deciding upon a course of action.
In the past this issue was dealt with rather neatly by Kohlberg's six stage sequence of moral development (see Power et al., 1989). According to Kohlberg's standard account, moral development moves from early stages in which moral understandings of fairness are intertwined with prudential self interest and concrete concerns for social authority, to conventional moral understandings in which morality (fairness) is intertwined with concerns for maintaining social organization defined by normative regulation. Finally, at the highest, principled stages of morality attained by a minority of the general population, morality as fairness is fully differentiated from non-moral prudential or conventional considerations, and morality serves as the basis from which the individual not only guides personal actions, but is able to evaluate the morality of the conventional normative system of society. This progression has been appealing to moral educators for several reasons. First, because the sequence was empirically based and purportedly described a universal developmental progression, this description of development offered educators an "objective" non-political basis from which to engage in moral education. Second, because the stages were presumably "content free" in that they do not pertain to particular issues, but instead refer to structures of reasoning, educators did not need to be concerned about the specific positions students take with respect to given issues. Finally, the sequence moved ultimately to a principled moral resolution of the kinds of complex issues of overlap discussed above. In other words, from the teacher's point of view, philosophical and political conundrums were resolved by the natural logic of the developmental process.
What we have now come to understand is that the progression identified through Kohlberg's paradigmatic research program does not adequately capture the ways in which people make socio-moral judgments, and cannot therefore, serve as the sole guide to moral education. Kohlberg described the sequence of age-related changes in the ways in which moral and non-moral (especially conventional) concerns are typically integrated in overlapping contexts. For example, Stage 4 (conventional) moral reasoning as described in the Kohlberg system reflects the emergence in middle to late adolescence of understandings in the conventional domain that social norms are constituent elements of social systems (Turiel, 1983). Although these age-typical integrations are captured by Kohlberg's stage descriptions, they do not represent the full range of socio-moral decision-making patterns that individuals present. For example, in the process of conducting their careful and extensive research aimed at standardizing moral stage scoring, the Kohlberg group discovered that individuals at all points in development may respond to Kohlberg's moral dilemma's by reasoning from a perspective of either rules and authority or justice and human welfare. From the vantage point of our current understanding of the domain related heterogeneity in people's social cognition, such within-stage variation can be accounted for by recognizing that the Kohlberg tasks generate reasoning employing knowledge from more than one conceptual system.
In moving beyond Kohlberg's landmark research on children's moral development, educators are both liberated in terms of how they might conceptualize the opportunities they have to engage students in moral reflection and behavior, and are also vested with greater responsibility for stimulating students to think and act in such moral terms. If in fact, children at all points in development are capable of considering moral issues from a moral perspective of justice and welfare, then it becomes important to increase the likelihood that children will "read" and prioritize the moral component of contextualized social issues, rather than simply attempting to move students toward a distal "principled" stage of moral judgment where such moral prioritization becomes a matter of course. This is not say that development doesn't matter. Achieving principled moral understandings in the full sense of Kohlberg's theory, presupposes a fully developed understanding of societies as social systems, an understanding that is only arrived at in middle to late adolescence (Turiel, 1983). In addition, the ability of children to "see" the morality of certain actions requires a similar, late adolescent, level of sophistication in the area of convention. For example, when we presented the issue of a person wearing a bathing suit to a funeral, many young adolescents (12 to 14 years of age) failed to see any problem. In their minds, since conventions (such as those regarding dress) were nothing other than the arbitrary dictates of authority, and since the important issue in this case was attendance at the funeral, there was nothing wrong in going to a funeral in a bathing suit. Older adolescents, however, who had constructed an understanding of conventions as constituent elements of social systems, were able to see the second-order moral implications that might arise from violating this social convention, and tended to view it as wrong to wear a bathing suit to a funeral, since within the societal framework of the funeral party, dress conveyed a sense of respect for the deceased, and sensitivity to the feelings of the grieving family (Nucci & Weber, 1991).
The point being made here is that attention to develop within moral education needs to be accompanied by attention to students' reading of overlapping social issues. In focusing upon development within each of these conceptual (moral and conventional) frameworks, educators contribute to students' capacity to understand and function within their social and moral worlds. However, because these systems interact in context, strictly developmental aims as set forth in the traditional Kohlbergian position need to be reconsidered. Since it has been found that students' reasoning about such complex issues is not the result of reasoning structures within a single system, the weight that students give to moral and non-moral considerations, and not just their reasoning within the moral domain becomes of interest. For example, whether students viewed the Maya Angelou story described above in moral or conventional terms might well be as significant as whether the students were at early or advanced points in their social development. As was illustrated in the Maya Angelou excerpt, there is no guarantee that individual development alone will lead to such a reading of overlapping social issues. Unfortunately, even those individuals judged within the Kohlberg framework to be at post-conventional stages of reasoning have been found to be subject to social pressures and situational cues in their reading of the moral meaning of actions (see the discussion of the Milgram study in Turiel, 1983).
There is research which demonstrates that teachers can impact the ways in which students read social issues, and the tendencies of students to attempt to address both the moral and conventional aspects of complex social issues. In their study, Nucci & Weber (1991) divided students into three discussion groups which met for 4 weeks. During these groups, students discussed issues that were primarily moral, conventional, or overlapping both morality and convention. Throughout the course of these weekly discussions, one group was directed to treat all issues in terms of moral concerns for fairness and human welfare; a second group was directed to treat all issues as matters of social convention and social order, and the third group was directed to treat moral issues from a moral perspective, conventional issues from a conventional perspective, and to coordinate moral and conventional components of multifaceted issues. Following this intervention, students levels of moral and conventional reasoning were assessed through interviews. Students in all three groups were able to clearly respond to unambiguous moral or conventional issues. However, when asked to write their views about the values contained in an incident which had both moral and conventional features, subjects in the moral only group subordinated complex issues to moral concerns, and subjects in the convention only group subordinated complex issues to matters of norms and social organization. Only the third group spontaneously looked at both features of issues and attempted to coordinate them. As this relatively benign and short-term treatment illustrates, education can be influential in framing the meaning individuals will give to complex social situations.
Given that many of the moral issues of everyday life are enmeshed within conventionalized norms and contexts, it seems imperative that children be given the intellectual and attitudinal tools necessary to deal with these realities. What this means in practice is that students not only be given opportunities to develop their understandings and ways of reasoning about morality and convention, but that they be engaged in the more complex task of evaluating and coordinating the moral and social organizational elements of multifaceted social issues. These processes will necessarily be different at differing points in development. Adolescents with complex understandings of societies as systems will address such issues in ways that are more integrative and complex than will children. Nonetheless, meaningful discourse about the moral component of multifaceted social standards, can be addressed across a broad age range. Furthermore, the contexts in which morality and convention may overlap are not confined to the distal world of adult society, but also arise in contexts structured by the norms of children and adolescents. For example, engaging children in "seeing" the moral implications of group norms of exclusion which might arise within cliques which don't want "geeks" as members, or in helping them to deal with masculine norms of toughness in the context of playground disputes can help them to formulate ways of constructing "societies" that are non-discriminatory, just, and safe. Thus, helping students come to terms with the difficult task of integrating what is moral with the need for social order and organization need not be seen as far afield from the straight forward business of raising "good" children who don't lie, cheat, steal, or hurt others.
If we have learned anything over the past 30 years, it is that moral education cannot be isolated to one part of the school day, or to one context, but must be integrated within the total school experience. Bringing the approach described here to the curriculum as a whole, however, may prove unsettling for some. Engaging children in critical moral reflection about issues raised in literature, or about existing or historical social standards contained within social studies or history texts may seem threatening to those who maintain the recapitulation view of the mission of public schools to foster citizenship. From that point of view (see Ryan in Nucci, 1989), values education serves the purpose of bringing the young into the existing social order so that society may be preserved and perpetuated. Unfortunately for that perspective, pluralist democracies are dynamic, and efforts to stifle critique run counter to the very nature of the democratic society such people hope to preserve. Whether we wish to engage the resources of our schools to develop the ability of our citizenry to engage in thoughtful moral critique of our culture is a political and moral decision. As I have laid out, the knowledge base from which to construct such an approach to moral and social education is available.
The preceding discussion has highlighted the basic reasons why moral education must attend to issues of social cognition and moral reasoning. Knowing right from wrong is more than a simple process of being aware of specific social rules, and doing the right thing is not a simple matter of putting those rules into practice. Social contexts are not fixed and, therefore, do not always lend themselves to habitual or formulaic ways of responding. Moreover, extant social rules may themselves require changes to bring them in line with morality. Reading and evaluating what is morally right, therefore, entails judgment. Being a good person, however, is more than a matter of understanding what is morally right. In philosophy a distinction is made between deontic judgments of what is morally right and aretaic judgments of responsibility which involve a commitment to act on one's deontic judgment. In everyday language we use the term "character" to refer to the tendency to act in ways that are consistent with what one understands to be morally right. A person of good character is someone who attends to the moral implications of actions and acts in accordance with what is moral in all but the most extreme of circumstances. This everyday usage of the term character captures an important feature of what is ordinarily meant by a good person. The question for us as educators becomes one of understanding how these common sense notions of character map onto actual human psychology, and what aspects of the educative process can contribute to character formation. Unfortunately, most of the current rhetoric about character education has little to do with what people are actually like, and more to do with a political agenda that would return us to mistaken practices of the past. It is important to remember as we move forward in our efforts to engage schools in meaningful moral and character education just why character education fell out of favor in the first place.
Limitations of traditional forms of character education. Traditional character education, which had its heyday in the early part of this century, had as its central aim fostering formation of elements of the individual's personality and value structure which would constitute socially desirable qualities or virtues. In the late 1920s a major research effort was undertaken by Hugh Hartshorne and Mark May to identify the factors that contributed to the formation of character. The design of their research was based on the reasonable premise that the first step should be to identify those individuals who possessed moral virtues. What they had expected to find was that the population of 8000 students they studied would divide up into those who displayed virtuous conduct nearly all of the time, and those who would not. To the surprise and disappointment of the researchers they discovered that few students were virtuous, and that instead, most children cheated, behaved selfishly, and lacked "self control" a large amount of the time. Virtue, according to their data, seemed to be context dependent as students cheated, or lied et cetera in some situations and not in others. As Clark Power (1989, p. 127) noted: Hartshorne and May concluded that there were no character traits per se but "specific habits learned in relationship to specific situations which have made one or another response successful."
The reference to habit by Hartshorne and May is concordant with traditional views of character formation. Since Aristotle, the development of virtue has been thought to emerge out of the progressive building up of habits. Contemporary character educators (Ryan & McLean, 1987; Wynne in Nucci, 1989) likewise rely heavily on psychological theories that emphasize punishment and reward systems to reinforce desired behavior, and systems of inculcation which are presumed to instill values and virtues in the young. It is worth remembering that in response to their findings, Hartshorne and May concluded that such traditional approaches to character education through the use of didactic teaching, exhortation, and example probably do more harm than good since such practices do not take into account the practical demands of social contexts. In other words, such rigid instruction runs counter to the evaluative and contextualized nature of moral life. By focusing solely on efforts to instill proper values and habits, such approaches fail to develop students' capacities to make the social and moral judgments that contextualized actions require. Moreover, these rigid approaches run counter to the multifaceted and complex nature of human personality. Research on personality conducted over the past 30 years (Sarbin, 1986) has served to confirm the view of character offered by Hartshorne and May by demonstrating that people cannot be accurately described in terms of stable and general personality traits since people tend to exhibit different and seemingly contradictory aspects of themselves in different contexts.
The moral self. Findings that individual personality and character are multifaceted, complex and responsive to contextual cues, seems to comport with such common experiences as knowing people who are shy in some contexts and gregarious in others, and fits our general common sense understanding that people are not always consistent in their moral positions or actions. On the other hand, our awareness of such inconsistencies also runs counter to shared experiences that people are more or less shy than others, kinder and more trustworthy than others, and so forth. In other words, there seems to be a sense in which human personality or character is consistent Resolving this apparent contradiction in the nature of persons has been the task of contemporary personality and social psychology. Resolution with respect to issues of morality and character seems to rest on a recognition that judgments and not just habits are operating when people respond to social contexts. In this light, observed consistencies within individuals across contexts may be accounted for with reference to the ways in which individuals address moral consistencies or inconsistencies within themselves. In other words, if individual moral actions are guided by choices and not simply the result of unreflective habit, then the issue for character education rests not with inculcation and habit formation, but in understanding how it is that people judge the worth of their own actions in relation to their world view and sense of themselves as moral beings. We need to move away from the notion of character as a set of externally provided traits and habits to a view of the moral self as constructed rather than absorbed and as being updated and reconstructed continuously (Sarbin, 1986).
Self in this view is not so much an entity as it is a story or a narrative we tell ourselves in which we are the featured character. Who we are emerges as we engage the social world and attempt to provide ourselves an account of how we initiate actions (a sense of agency), and of who that agent is (a sense of identity), and who we wish that agent to be ( a combination of agency and identity). What we call the self, is a psychological construction which we form in social contexts. Before we are born, aspects of the content of who we will become are already laid out for us. Each of us lives in a particular time period, cultural and historical context, and family situation. We are given a name, assigned a gender, and live in a society in which race matters or doesn't matter. All of which comes without our asking, and none of which comes with prepackaged understandings. Personal development, then is in part a function of how one interprets the hand one is dealt at birth, and the meanings and ways in which one enacts the different roles (e.g., boy, girl, athlete, scholar, gang member, professor, someone named Larry or Maria) which we assume in context. In a sense such social roles imply scripts, and some social learning theorists have mistakenly reduced social conduct to knowledge of social scripts. Social life, however, is not rigidly scripted, and to the extent that one can use this metaphor it would be more in the sense of a broad outline in which persons present an interpretation of a given role (e.g., mother) which they enact and modify in social context. In addition, social roles are not simply accepted by individuals, but are evaluated and modified to comport with individuals' constructions of what a social role should be as it relates to themselves. Finally, personhood and a sense of agency requires personal choice, and individuals engage in choices which would establish their uniqueness (see Nucci & Lee in Noam & Wren, 1993).
The connection between this drammaturgical or narrative view of self, and the present discussion of character has to do with how individuals construct a view of themselves as moral beings, or what some have called the moral self (Noam & Wren, 1993), and the relation between this moral self and the more general narrative we construct which constitutes our personal identity. Self as singular, that is who we refer to when we speak about ourselves, look at our own baby pictures, and experience the sense of agency when we engage in actions, is in fact multifaceted. Those facets of who we are vary in terms of their salience or importance both as a function of the general narrative we have constructed about ourselves, and the particular situation we find ourselves in. Our "moral selves", what some have been calling character, is only a part of and functions in relation to the totality of who we are (see Blasi in Noam & Wren, 1993). When we act in context, our reading of the social situation may or may not engage our moral understandings. And when our moral understandings are involved, it may or may not be the case that the moral part of who we are is the most salient. Richard Nixon, for example, argued that being President was different from being an ordinary citizen in that the Presidency required one to act in extra- legal and amoral ways when the pragmatic self interests of the United States were at stake. In essence, Nixon's understanding of the role of President meant that morality was secondary to political pragmatics. Implicit in his argument was the notion that Nixon's moral self remained intact, but on hold while he was acting from pragmatics. We see similar forms of argument in the self reports of adolescents who engage in aggressive acts to steal from others (Guerra, Nucci, & Huesmann, 1994). Often these adolescents explain their actions (e.g., hitting a woman to steal her purse) in means - end terms in which the moral consequences of their actions (hurting another person) are placed well below the pragmatic goal of obtaining goods. While these adolescents will describe themselves in moral terms (e.g., fair, respectful) in relation to general dealings with people , especially family members, they define their actions in very "business" like terms (e.g., taking care of business) when describing their actions on the street.
Character and the moral self. In Blasi's (Noam, 1993) work on the moral self, he makes the point that morality may or may not be a central element of the general narrative we construct about who we are. In other words, morality may or may not be a salient issue in constructing one's personal identity . The fact that virtually all children construct basic moral understandings about fairness and human welfare does not mean that being a person who acts on that knowledge in relation to others is necessarily an important part of how one self defines. For the adolescents described above, or for some businessmen for that matter, being moral may not be as integral to their self definition as are other facets of their personal identities (e.g., gang member, successful businessman). According to Blasi, the experience of "guilt" or moral responsibility emerges in those situations in which one acts counter to what one knows to be morally right only for those for whom morality is an integral part of personal identity. In other words, from Blasi's work, we can infer that a central feature of what we mean by moral character is the degree to which being a moral person attains salience as a part of one's self definition. Acting in consonance with one's deontic moral judgments is for someone of "good" character important for that person's sense of intrapersonal coherence in the vast majority of contexts.
From an educational standpoint this means that character formation is not a curricular issue in the usual sense of a course or program designed to teach a particular content. Character emerges from the more general individual environment interactions from which students construct their sense of themselves. There is no simplistic model or formula for "building" character. And, as much as those of us who each year brave Chicago's character building Winters would like to believe, no specific set of experiences that lead to good or strong character. Schools contribute to character to the degree to which they constitute environments conducive to more general social and emotional development, and more specifically moral environments in which students are treated fairly and with respect, and which convey and enact through teacher behavior and school policy a general climate in which morality (as opposed to arbitrary adult authority) is valued. Having said that, there are some policies and practices which schools can engage in that raise the likelihood that schooling will contribute to students moral development and character. I will end this chapter with a brief summary of some of those policies and practices. Before doing so, however, I think it is important to recognize the limitations for public policy of any attempt to address larger social issues of crime or violence solely through educational efforts designed to alter the morality and character of individuals.
The limitations of reliance on individual responsibility. Much of what we see in present day society by way of criminal activity, and juvenile crime in particular needs to be understood as a rational response to objective social conditions rather than simply a lack of morality or character of individuals. A study which we (Sapiro & Nucci, 1991) conducted in Brazil of adolescents' and young adults' conceptions of everyday forms of corruption is highly instructive. Nearly all of our young subjects across social classes and economic levels engaged in what they considered to be corrupt social practices (e.g., paying a police officer to avoid a ticket, paying for physician services without receipt to enable the physician to avoid taxes and charge a lower fee) at least some of the time. When asked to evaluate these practices nearly all of our subjects argued that they were wrong. However, lower class subjects irrespective of educational level were five times as likely as upper middle class young people to state that engaging in such practices was justified in the face of an overwhelmingly corrupt social system. In contrast upper-middle class university students were more than twice as likely as lower class subjects irrespective of educational level to argue that it was important to not engage in such practices in order to offer individual resistance to the corrupt social system and thereby change it. What is instructive for us at the policy level is to recognize that these observed class and educational differences in orientation to the immorality of corrupt public behavior did not reflect a difference in the morality of individuals (nearly all subjects saw the acts as objectively wrong), but rather social class differences in the sense of political and social empowerment to effect change in the objective social situation, and the belief on the part of the poor and uneducated that such actions constituted a rational form of self protection from victimization by the general system. While the US. is not Brazil, the lesson to be drawn is that we should not expect school approaches to moral and character education aimed at individual responsibility to completely compensate for broader changes that need to take place with respect to social policies which impact America's poor and disenfranchised.
In conclusion, let me summarize some of the main points of this chapter and indicate some of the implications of recent research for educational practices and policies with respect to moral development and character formation. These practices divide more or less into those which concern academic or intellectual content and reflection, and school policies or practices which affect general school climate or student activities. I begin first with academic practices.
Issues of School Climate and Student Activity
Angelou, M. (1971). I know why the caged bird sings. New York: Bantam Books.
Guerra, N., Nucci, L., & Huesmann, L. R. (1994). Moral cognition and childhood aggression. In L. Rowell Huesmann (Ed.), Aggressive behavior: Current perspectives.(pp. 13-33). New York: Plenum.
Noam, G., & Wren, T. E. (Eds.). (193). The moral self. Cambridge, Mass: MIT Press.
Nucci, L. (Ed.). (1989). Moral development and character education: A dialogue. Berkeley: McCutchan.
Nucci, L., & Weber, E. (1991). The domain approach to values education: From theory to practice. In W. Kurtines & J. L. Gewirtz (Eds.), Handbook of moral behavior and development (Vol. 3: Applications) (pp. 251-266). Hillsdale, NJ: Erlbaum.
Power, C., Higgins, A., & Kohlberg, L. (1989). Lawrence Kohlberg's approach to moral education. New York: Columbia University Press.
Ryan, K., & McLean, G. F. (Eds.). (1987). Character development in schools and beyond. New York: Praeger.
Sapiro, C., & Nucci, L. (April, 1991). Brazilian adolescent and young adults' conceptions of corrupt social practices. Paper presented at the biennial meetings of the Society for Research in Child Development, Seattle, WA.
Sarbin, T. (1986). Narrative psychology: The storied nature of human conduct. New York: Praeger.
Turiel, E. (1983). The development of social knowledge: Morality and convention. Cambridge, Mass: Cambridge University Press.
Turiel, E., Hildebrandt, C., & Wainryb, C. (1991). Judging social issues. Monographs for the Society for Research in Child Development (Vol. 56, pp. 1-103).
Wainryb, C. (1991). Understanding differences in moral judgments: The role of informational assumptions. Child Development, 62, 840-851.