A Cognitive-Developmental Approach to Moral Attachment

A cognitive-developmental approach to moral attachment is presented, subsuming processes commonly termed attachment and identification, and drawing on both James Mark Baldwin’s theory of the imitative “ideal self” and Ausubel’s theory of “satellization.” Five components of attachment and five components of identification are described to form the moral self. Moral identification arises from (1) natural tendencies to imitate the parent or other model, (2) a desire to conform to the parent’s normative expectations, (3) a perception of similarity to the parent (intensified by imitation), (4) a perception of the greater competence or higher status of the parent, and (5) an idealization of the parents’ competence or virtue. Moral attachment is comprised of (1) an emotional dependency on parents and empathy with them, (2) vicarious self esteem derived from the parents’ competence or status, (3) the ability to derive self esteem from the parent’s approval and affection so as to forego other sources of success or competence, with associated security or self esteem, in the absence of direct signs of success, (4) reciprocity and complementarity in this relationship, and (5) a feeling of obligation to persons and relationships characterized by attachment processes. The relationships that characterize the attachment processes are not necessarily limited to the biological parent, but may be to any significant other. The moral attachment process begins in early childhood, usually with a parent as the object, develops with experience, advances in cognition in the 2- to 8-year age span, and later is found aimed toward admired others (e.g., peers, teachers, coaches). This concept of moral attachment attempts to supplement Kohlberg’s structural-developmental stage theory of moral-judgment development with a theory of the moral self that Kohlberg deliberately excluded from his earlier work.

For the last 25 years, there have been three prevalent approaches to moral development. The most recent approach is that of situational social learning 230theory initiated in the mid-1950s by Jack Gewirtz (1956, 1961, 1969, 1976) and others in relation to social dependency and attachment but extended to moral behavior by Burton (1976), Casey and Burton (1986), Gewirtz and Peláez-Nogueras (chapter 7, this volume), Mischel (Mischel & Mischel, 1976), and Bandura (1969, Bandura & MacDonald, 1963). An earlier approach is the structural-developmental approach initiated by Piaget (1932, 1965) and elaborated by Kohlberg (1958, 1984). The third and oldest tradition is the “socialization of character” approach usually based on neopsychoanalytic theory of attachment as integrated with a functional anthropological theory (Levine, 1982). There have been several important studies based on the neopsychoanalytic developmental character-type theories of Horney (Peck & Havighurst, 1960) and Fromm (1947, 1958), as well as the biographical studies by Erikson (1962, 1969) based on his neopsychoanalytic “functional” stage theory (Kohlberg, 1987).

All the above-mentioned approaches took as their point of departure Hartshorne’s and May’s (1928; Hartshorne, May, & Mailer, 1929; Hartshorne, May & Shuttleworth, 1930) classic studies of the organization or consistency of moral behavior with regard to the “virtues” of honesty, service (prosocial behavior), and self-control, and the reactions to them. The results of Hartshorne’s and May’s monumental study disproved common sense assumptions about character and its basis, as first formulated by Aristotle in his Ethics (1962). Aristotle viewed character as a set of cognitively steered habits formed by guided opportunities to practice such virtuous conduct as honesty. Hartshorne and May found no consistent habits or behavioral virtues, and found that behaviors denoting honesty were not influenced by traditional “character education” (i.e., the preaching of virtue along with the opportunity to practice these virtues in public schools, churches, or Boy Scout organizations). They concluded that moral behavior represented situational conformity to classroom or group norms, and to the contingencies of social reinforcement like those elaborated by B. F. Skinner and J. L. Gewirtz. A review of Hartshorne’s and May’s initial assumptions, findings, and conclusions is presented in Kohlberg (1984, chapter 7; 1987 chapter 7).

As his point of departure, Kohlberg (1958) took several considerations into account. These included Hartshorne’s and May’s failure to find character and its determinants as initially assumed by them; American concerns in the tradition stemming from Aristotle; the replication of neopsychoanalytic studies done in the 1950s of Hartshorne’s and May’s findings of the situational specificity of “resistance to temptation” or “prosocial behavior”; and their failure to find consistent correlates between child-rearing practices, protective measures of “superego strength,” and moral behavior (Kohlberg, 1963, 1964).

Accordingly in 1958, Kohlberg decided to focus on the development of moral judgment in the stage-structural tradition and to relate it to conduct, first through teacher behavior ratings and comparisons of delinquent with nondelinquent adolescents (Kohlberg, 1958), and later through a variety of other studies (reviewed by Blasi, 1980; Kohlberg, 1984, chapter 7). Following Piaget (1932/1965), this 231structural-developmental approach to moral judgment has systematically ignored the “self” or “ego” in favor of a rational-epistemic or moral subject. It is from the prescriptive judgments to hypothetical dilemmas of that rational-epistemic subject that a stage progression can be reconstructed rationally (Kohlberg, 1984, chapter 3). Such a hypothetical epistemic subject is the author of the moral judgments offered in response to hypothetical dilemmas (Colby & Kohlberg, 1987), responses philosophers term “deontic” judgments (i.e., of lightness, duty, and justice). In order to link stages of deontic judgment to observations of moral conduct, we have found it necessary to postulate judgments of responsibility in addition to deontic judgments (Kohlberg, 1984, chapter 7). Having such a judgment of lightness or justice, the moral actor must make a further judgment of the self’s responsibility to act in accordance with this deontic judgment of justice. Presumably a judgment of responsibility to act morally also implies a judgment of guilt (or shame) at failure to carry out the act for which the actor is responsible.

The idea of a judgment of responsibility points in two directions that go beyond structural deontic judgment. First, responsibility is defined by membership in a social community that has shared or collective norms that the actor is expected to uphold and for which he is the subject of group sanction for failure and of group solidarity for success. This direction, that of neo-Durkheimian anthropological theory, has been applied to our research in school communities and the moral atmosphere and collective norms of the school or community (Higgins, Power, & Kohlberg, 1984; Kohlberg, 1987; Power, Higgins, & Kohlberg, 1989).

The second direction in which our study of moral responsibility and moral action carries us is to the resurrection of the notion of moral character, conceived as the development of the ego or the “moral self.” The moral self is the subjective side of the organization or unity of moral behavior postulated by the notion of character. It is the sense of the self’s “integrity” or “identity” that becomes at stake in moral action. This is the core of Erikson’s (1962) idea of Martin Luther’s identity in his statement, “Here I stand, I can do no other” (Blasi, 1984, 1985). In Erikson’s terms, prior to moral identity is moral identification, which Erikson interprets in classical Freudian terms of the resolution of the Oedipal conflict through superego function.

A review of a vast body of research on parent–child relations, conscious and unconscious guilt, and moral behavior has failed to confirm the usefulness of classical Freudian superego theory as a general account of moral character (Kohlberg, 1963, 1964). In this paper, we elaborate an alternative theory of the formation of the moral self and the birth of the sense of responsibility based on three sources. The first source is the theories of the moral self of the American symbolic interactionists: J. M. Baldwin (1906), John Dewey (1939), and G. H. Mead (1934). The second source is the neo-psychoanalytic character-typological theories or schools of ego or self-development of Ausubel (Ausubel, Sullivan, & Ives, 1980), Loevinger (1976), and Kohut (1977). The third source is Piaget’s 232(1932/1965) theory of moral development before he developed a “pure structural approach.”

My colleagues and I have written a general introduction to the social-self theories of Baldwin and Mead (Kohlberg, Hart, & Wertsch, 1987). In the present paper, we focus on the child’s relations to adults in the years two to eight, the period of the formation of a moral self and a sense of moral responsibility in the theories of Baldwin, Mead, Ausubel, Kohut, and Piaget, as well as the period of superego formation in Freudian theory. Our theory shares with Freudian theory a concern with identification in the formation of the moral self or ego, but construes identification in a very different way than does a Freudian theory of unconscious drives and defenses. It also shares with Freudian theory a concern with love or attachment to parents as related to the development of the moral self, but again in a very different form than the Freudian theory of superego formation.