Which is better conceivable or teachable

Georg Lind. Is morality teachable? Results of modern research on moral psychology

Transcript

1 Georg Lind Can morals be taught? Results of modern moral psychological research

2 Can morals be taught? Results of modern moral psychological research Georg Lind Second, supplemented edition 2002, Berlin, Logos-Verlag

3 The only real force against the Auschwitz principle would be autonomy, if I may use the Kantian expression, the force for reflection, for self-determination, for not participating. Theodor W. Adorno

4 For Gisela, Gregor, Antonio and Glenda

5 Contents Foreword to the Second Edition On an Educational Theory of Moral Development The Two-Aspect Model of Moral Judgment Behavior and Moral Development Kohlberg's Moral Judgment Interview A Critique The Moral Judgment Test An Experimental Questionnaire Moral Education versus Moral Maturation: The Problem of Regression Education versus Socialization I: Can moral judgment be simulated like a political attitude? Education versus Socialization II: Cultural Influences on Moral Judgment Promotion of Moral Judgment through Educational Processes: An Intervention Experiment Conditions for the Effect of Interventions to Promote Moral Judgment: A Meta-Analysis Moral Stereotypes and Pluralistic Ignorance: The Perception of Moral Needs of Students by Teachers Educational experiences as a condition for moral development Conclusions Bibliography and sources Appendix: Moral atmosphere questionnaire (MAF)

6 Acknowledgments Many people were involved in the preparation of this book. I would like to thank Tino Bargel, Peter Dobbelstein-Osthoff, Dr. Monika Kuhn, Ottfried Lind, Leonore Link, Manfred Scholz, Christa Kolbert-Ramm, Johann-Ulrich Sandberger, Dr. Heinz Schirp, Karin Tritt. Patricia Knoop was responsible for editing and proofreading. The studies described here were funded by the German Research Foundation (Bonn), the Research Group Society and Region e.v. (Konstanz), the Ministry of Culture of North Rhine-Westphalia, EMNID (Bielefeld) and the University of Konstanz.

7 Foreword to the second edition 1 This is an attempt to answer the millennia-old and now more relevant than ever question: Can morals be taught? Plato reports that Socrates already raised this question. He couldn't find an unequivocal answer, but realized that it was related to another equally important question: Is morality a skill (or just an attitude and value)? Morality is teachable, I understand his dialogue with Menon, if it were a virtue or, as we would say today, a moral ability. The reverse is also true: if it is a skill, then it can also be taught. Many authors deny both, giving different reasons for this. Some say morality cannot be taught because it is a question of innate personality traits (so-called traits). One can only try to separate the morally good people from the morally less good people by selection, in order to ensure that morality comes into play or at least does not go under entirely. Morality cannot be taught, say the others, because it is not a question of ability, but of value, attitude, motivation or the right level of awareness. But such things cannot be taught or taught in the traditional sense. They can only be changed by making people understandable or, if that doesn't help, by social coercion. These ideas about morality result in clear recommendations for upbringing and education. If morality is an innate quality of human beings, then moral education is pointless. Society can limit itself to classifying people into morally good, less good and bad and treating them according to this classification. From the assumption that morality is just a set of values ​​or an attitude or the like, it can be deduced that moral education and upbringing are limited to this. 1 Compared to the first edition of this book under the title Moral and Education (published by Asanger-Verlag, Heidelberg) numerous changes added and many graphics revised. The second edition of this new edition was mainly used to correct spelling and to bring this foreword and the list of literature up to date. The page numbering has changed slightly as a result. For technical reasons, the old spelling was retained. 7th

It is possible to make the adolescent understand the given moral rules of society or to get him to accept them (at least verbally) in some other way. Such ideas are hardly compatible with the conviction that man can be formed, and with the principles of democracy, which renounce social coercion and indoctrination. So far, both ideas have hardly been supported by empirical research. Personality traits exist more in the eyes of the beholder than in people's actual behavior (see e.g. Cohen, 1969). Moral attitudes and values ​​hardly show any systematic reference to human behavior (Hartshorne and May,) and can, if necessary, be simulated in almost any way (Emler et al., 1983; see also Chapter 6). These findings do not mean that a person's genetically determined makeup and their values ​​and attitudes have no significance for their behavior. However, they point out that these are not consistent and unchangeable properties and that the results of personality scales and attitude tests are of little help for the question of the teachability of morality. Since the beginning of the 19th century, studies in moral psychology have produced numerous indications that morality is not just a question of correct attitudes or values, but rather a question of ability. The Berlin forensic psychiatrist Levy-Suhl (1912) found that delinquent juveniles have the same moral values ​​as non-delinquent juveniles, and that there must therefore be other reasons for immoral behavior. This early finding was subsequently confirmed many times. Hartshorne and May (1929) showed in a large-scale experimental study that fraudulent behavior is hardly tied to certain moral attitudes and is also in no measurable connection with moral instruction in religious education. Piaget (1932 / 1973a) made the question of the relationship between moral judgment and behavior the subject of revealing studies of child behavior in playgroups. Kohlberg's (1958) studies on moral development in young people led him to the concept of moral judgment, which, as I will show in this book, bridges the gap between individual moral values ​​on the one hand and behavior on the other (Chapter 2). For a long time, however, these tips received little attention and their importance is still often underestimated today 8

9 or misunderstood. The so-called judgment-action problem is still being tried to solve by excluding the concept of judgment and the research carried out on it, which, as expected, cannot succeed (Garz et al., 1999). Moral abilities have been an explicit subject of moral psychological research since the early 1960s. This is mainly thanks to the American psychologist and educator Lawrence Kohlberg. Kohlberg was one of the first to try to define, measure and promote moral abilities. This groundbreaking research achievement has received a lot of attention from both academia and the general public. But it will m.e. Still too little appreciated, even if, as my own research shows, some of Kohlberg's basic assumptions, especially the assumption of an invariant rising moral development, can no longer be maintained today. A loss of moral judgment is possible and can be observed in many people. It seems that such regressions are mainly due to inadequate learning opportunities (see above all Chapter 5). I first became aware of Kohlberg's work in the early 1970s, at the end of my studies. They immediately fascinated me. I was impressed by the fact that Kohlberg scientifically dealt with a topic that, according to the prevailing doctrine, hardly seemed scientifically suitable. I was particularly impressed by his assumption that morality is, to a high degree, a skill that can be promoted by creating favorable learning conditions. He therefore called his theory cognitive-development-oriented. I sensed at the time that this assumption, if it were to stand up to scrutiny by scientific research, would have great significance for the question of moral education in the broadest sense. Until then, many of us could only imagine morality as a question of attitude and values. If Kohlberg were right, and morally good behavior was really more a question of ability and training and less of the right attitude or genetic disposition, we would have to refrain from some traditional methods of moral education that aim at changing attitudes and values ​​or selection. There is also no longer any reason to evade the task of moral education with reference to the democratic educational mandate of our schools. Gerhard Portele (1978) described the problem of (traditional) moral education in democracy as that of you-should-want-9

10 len paradox: social coercion and democratic maturity are mutually exclusive, and thus democracy and (moral) education are also mutually exclusive. A cognitive-development-oriented moral education could resolve this paradox: The goal of promoting the ability to think and act according to universal moral principles accepted by the individual is not only compatible with the principles of a democratic society, but a central task of democratic upbringing and education . Kohlberg's theory is rightly criticized for the fact that, despite his impressive, 20-year longitudinal study, it can only refer to a relatively narrow database, not least due to the fact that he only had a very complex interview process at his disposal to obtain data. Implementation and evaluation of this procedure are time-consuming and costly. In addition, despite the strong standardization through a comprehensive evaluation manual, the evaluation is based heavily on subjective assessments and on specifications that could raise the suspicion that the data obtained with it are distorted in favor of the theory (see Chapter 3). Thanks to my many years of collaboration in the international longitudinal study of the Konstanz Collaborative Research Center 23 Educational Research from 1973 to 1986, I had the opportunity to put the Socratic hypothesis and Kohlberg's research information to the test with the help of a large, objectively experimentally obtained data set. This research was supplemented by numerous smaller studies and experiments that I was able to carry out together with students or colleagues. In the last ten years, more studies from other cultures and educational intervention studies have been added. Initially, most of the studies revolved around measuring morale. I have given top priority to this question of the adequacy and validity of the measurement. What use are good theories and practices to us if the data basis is not sufficiently valid and it is left to the individual to believe in their correctness or not. I therefore spent several years checking the validity of the method we had chosen to measure moral judgment (Lind, 1978; 1984; 1985a; 1995, 2002; Lind & Wakenhut, 1985), and I have also worked on translations in other languages ​​have consistently urged to devote more attention and energy to this question than is normally the case in the social sciences. Translated versions of the MUT can only be 10

11 are considered valid and equivalent if they are experimentally validated on the basis of theoretically well-founded criteria (Lind, 2002). After Kohlberg and his group lost sight of the concept of moral judgment more and more and their interest concentrated on the apprehension of moral consciousness, it seems that MUT is the only method that can apprehend moral judgment. Other methods primarily document moral attitudes, such as the defining issues test (DIT), the P-value of which expresses how strongly people prefer post-conventional moral principles over other forms of moral argumentation. After its author, Jim Rest, for a while claimed that the DIT could also be used to measure moral judgment (Rest, 1979; 1986b), he indirectly withdrew this claim. In several experiments it was shown that the measured value could easily be cheated upwards (which should not be possible with indicators for a skill): We [...] have eliminated the faking study from our set of the validity criteria (p. 115 ). The MUT's ability indicator, the C value, has shown itself to be dizzying in such experiments (see Chapter 6; Lind, 2002 / in press). This research situation demands, as I try to show in more detail elsewhere (Lind, 1985d; 1989a; 1989b; 1993; 1996a; 1996b; 2000a; 2000b; 2000c; 2002 / in press) a new understanding of morality and democracy. Piaget and Kohlberg already anticipated this new understanding in their early writings, even if they could not prove it as clearly as we can today with the help of MUT. First, solving moral problems actually depends on acquired skills (and not just on the right attitude or motivation) in a similar way to solving other types of action problems. The previous (and unfortunately still widespread) idea that intentions and wills were sufficient to act morally (or democratically) can no longer be maintained. Second, moral abilities, like moral attitudes, must be understood as a property or aspect of moral (judgment) behavior and not as separate components or behavioral domains. Just as the color, shape and elasticity of a ball can be clearly differentiated, but not separated and examined separately from one another, so can the cognitive and affective aspects of moral behavior

12 clearly distinguish (and measure logically independently of one another), but do not separate them into different behavioral classes. Because this insight still receives too little attention, I have devoted a separate chapter to it. Third: This new understanding has consequences for the practice of education and also for the design of our educational system. Above all, it opens up new dimensions in education for us. As teachers and parents, we are no longer faced with the choice between total abstinence from moral questions on the one hand and indoctrination of values ​​on the other, as was the case for a long time, but we see ourselves faced with the also democratically legitimate moral task, ability to encourage adolescents to apply their own moral principles consistently and in a differentiated manner. I was able to experience the educational energies this insight can unleash in the eighties when, with the help of Lawrence Kohlberg, Ann Higgins and Fritz Oser in North Rhine-Westphalia, I initiated the practical school test project Democracy and Education in Schools (DES) and together with Fritz Oser, Wolfgang Althof and Sibylle Reinhardt scientifically accompanied (Lind & Raschert, 1987; Lind, 1990; Lind & Althof, 1992, as well as here Chapter 8). In the meantime I have had various opportunities to study the practical school fertility of the ability theory of morality. The teaching method of the moral dilemma discussion proposed by Blatt and Kohlberg plays a key role here, which I developed further in cooperation with schools and students (Lind, 1999c). The results of this work are currently being used to design new training opportunities for teachers in the fields of ethics, education for democracy 2 and mentoring 3. In the first edition from 1993, under the title Moral und Bildung, published by Asanger-Verlag, this book has a found quite positive feedback. It was soon sold out. The moral psychologist and pedagogue Günter Schreiner from Göttingen described the research results presented in the book as a milestone in the moral pedagogical discourse. For Roland Wakenhut, social and organizational psychologist at the Catholic University of Eichstätt, the work summarized here provides important 2 See 3 See 12

13 impulses for moral psychological research which, after the death of Lawrence Kohlberg and the dissolution of the Harvard Center for Moral Education, shifted its focus to Europe. Peter H. Rossi, Professor of Sociology at the University of Massachusetts, Amherst, writes about the methodology proposed here for measuring cognitive structures in the area of ​​moral thought and action, the Experimental Questionnaire (EQ): I am pleased to know your EQ method to have learned. Your method is much more theory-based than our factorial survey idea, and with your consent, I'll borrow some of your ideas. Dr. John M.Bynner (Professor at Open University, Milton Keynes, UK) comments: I think your arguments about attitude measurement [described in more detail in Lind, 1985c; G.L.] are extremely convincing. They are in accord with the general concern Lamiell makes for ideothetic measurement. Since the first edition, further important investigations have been carried out which are based on the research reported here and which deal intensively with the educational theory of moral development presented in this book. I can only name a few here: Studies to check the validity of the MUT (Heidbrink, 1995; 1996; Lind, 1996c; 2002; Rest et al., 1997; Biggs et al., 1999), to simulate the ability to make moral judgments (Wasel, 1994; see here Chapter 6), on the connection between morality and political activism in an international perspective (Gross, 1997), on the moral judgment of medical staff (DuBois, 1997; Lind, 2000a), on the dependence of moral judgment on educational opportunities in vocational school students ( Beck, 1995; see here Chapters 5 and 11), on the connection between moral judgment and opportunities to take on responsibility (Lind, 1996a; Comunian & Gielen, 1999), on the effect of the dilemma discussion on juvenile prisoners (Lind, 1994) and on the pluralistic Teachers 'ignorance of students' moral values ​​(Lind, 1995). In addition, the knowledge gained from this was used for the further development of pedagogical methods to promote moral and democratic skills (Lind, 1999c). The Moral Judgment Test (MUT) 4 developed for our research is now available in several languages. Most of the translations have since become strict. 4 The English name is Moral Judgment Test (MJT). 13th

14 gen validation tests (English, French, Hebrew, Dutch, Italian, Portuguese, Spanish, Basque, Turkish, Macedonian, Czech, Finnish). Further versions are in the works (Japanese, Chinese, Korean). Despite the extensive data situation (there are currently more than individual surveys with the MUT) and an intensive and in some cases very critical examination of this method, there has so far been no finding that fundamentally supports the two-aspect theory and thus the existence of an aspect of the ability of morality questions. On the contrary, this theory and the moral judgment test developed on this basis are confirmed in new ways (Biggs et al. 1999). I hope that the educational theory of morality presented in this book, which is largely based on studies with the MUT, will help us to recognize the importance of the ability aspect of morality for everyday educational life and education policy and to reflect on the necessary consequences. Georg Lind Konstanz, January Address: 14

15 1 Towards an educational theory of moral development Is the development of morality inherent in our genes or is it forced by society through pressure and favor? Does education in the family and in schools therefore have to focus on waiting or, if one follows the second perspective, on applying pressure and sanctioning? The central finding of the studies presented here is that neither of the two explanations applies and therefore both recommendations for pedagogical practice are unfounded. Rather, all studies suggest that moral behavior and development are the result of (or lack of) educational processes. Closely related to the empirically suggested educational theory of moral development is the two-aspect theory of moral behavior. According to this, it is not enough to see moral behavior solely from the point of view of moral affects (attitudes, feelings, values, motives, etc.). In addition, there must be the study of cognitive aspects or skills (Lind, 1985a; 1985b; 1986a; 2002/1995). If it were just what we want and feel, almost all of us would be morally perfect. Nobody (or at least hardly anyone) wants to be immoral or act unfairly. Most, if not all, people are sensitive to injustice and immorality. But from moral and volition there is often a further path filled with many difficulties to moral action (see Levy-Suhl, 1912; Piaget, 1973a). In order to bring our moral ideals into harmony with behavior, we therefore need certain skills, especially moral judgment (Kohlberg, 1964). I consider this to be the most important discovery of the 20th century, especially as it has to prevail and assert itself against powerful opposition. These have their origins in the theological postulate that good will alone makes one happy, and in the psychological separation of affect and cognition into two different behavioral domains (Bloom et al., 1956). The most important result of our investigation is that there are two equally important, inseparable aspects to be distinguished in moral behavior, the affective and the cognitive aspect. The affective aspect (our moral ideals, attitudes and motives) is just as necessary and indispensable a component of moral behavior as the cognitive one, but unlike the latter, people differ only little in this, so that it is us 15

16 also says little about why people behave so differently in morally relevant situations. The cognitive aspect (above all the skills that are necessary for resolving moral conflicts) seems to be the one that is often very different in different people and leads to very different behavior, and is of great importance for the educational process. These theses represent the core of my educational theory of moral development. They are not new. They go back to the ancient Greek philosophers. Socrates has already thought about it in his famous dialogue with Menon on the question of whether virtue can be taught. If virtue is a skill then it is also teachable, argues Socrates, and if it is a skill then it is also teachable. At the beginning of this century, the Berlin forensic psychiatrist Levy-Suhl (1912) made the astonishing finding that juvenile offenders have the same moral values ​​as non-offenders. But the cognitive educational theory of morality has so far hardly been empirically or experimentally proven in such a way that it can stand up to competing and far more popular theories of moral development, for example alongside the socialization theory, according to which the development of morality is the result of social pressure to adapt, or of maturation - or emergence theory, according to which moral judgment develops with age, practically without external influences, almost by itself, as a result of a genetically controlled maturation process. According to socialization theory and social learning theory, moral development occurs through learning of the norms and values ​​of the group to which one belongs. One of the fathers of socialization theory was the French sociologist and educational theorist Emile Durkheim. For him, humans naturally live in an unstable state, beset by their instincts, which need to be limited and redirected in the interests of society. According to Durkheim (1984/1902), this requires making social norms and values ​​understandable through coercion and authority. According to this theory, the goal of moral development is the acquisition of moral attitudes, in which, in more recent theoretical drafts, cognitive processes can be involved. The cognitive abilities that are necessary for moral 16

To translate values ​​or principles into behavior appropriate to the situation and to resolve the moral conflicts that may arise. According to the theory of maturation, on the other hand, a person is good from birth or has the capacity to be good; Society, on the other hand, is the root of evil in this system of thought and, with its norms and compulsions, represents a hindrance to moral development. The spectrum of the theorists of maturation of morality is diverse, but there are some common convictions. All is well as it comes from the hands of the Creator; everything degenerates in the hands of man, leads one of the great exponents of this school of thought, J.-J. Rousseau (1971/1762, p. 9), his epochal work Emile or On Education. He insists that man is naturally good ... [but] society corrupts and makes man unnatural (p. 241). Sigmund Freud and his psychoanalytic school repeat Rousseau's thesis in a modified form: the prevailing social norms led to the suppression of human instincts; this drive suppression is jointly responsible for immoral behavior. Psychoanalysis therefore relies on the fact that, as Arno Plack (1969) sums up, human nature, if left alone and not bent from an early age, develops a behavior that best suits the individual and society ( P. 345). As a result, even institutionalized education cannot have any positive effects on individual moral development and pedagogy, as von Braunmühl (1975) put it in a nutshell, can only be concerned with its own abolition. The cognitive development theory of moral development by J. Piaget and L. Kohlberg occupies a special position among the previous approaches. It has elements of both theories. According to Kohlberg (1984), the development of cognitive structures is more the result of processes of interaction between the structure of the organism and the structure of the environment than the direct result of maturation (p. 8). But the direction and sequence of development are, according to this theory, (genetically?) Determined. With reference to Piaget, Kohlberg assumes that the stages of moral development form an invariant sequence, order or succession in individual development (p. 14). While cultural factors can speed up, slow down, or stop this development, they cannot change its sequence. The prognosis of the cognitive development theory is also optimistic. Instead of 'man is good', this theory assumes, 'man becomes good in the course of 17

18 of his cognitive-moral development '. At least he's not getting any worse. Due to the inner logic of moral development, social influences can only speed up or slow down moral development, but they cannot reverse it. This assumption of the invariant sequence or non-regression forms a central pillar of the theory and, in contrast to some of their statements about the role of the social environment, dominates the research of Kohlberg and his colleagues. So your theory is essentially a maturation theory. Numerous studies in recent years have underpinned Kohlberg's (1969) sharp criticism of socialization-theoretic approaches as well as naive-maturation-theoretic approaches and apparently also collected clues for a confirmation of the cognitive development theory. On the other hand, they have also found indications that some of these confirmations are based on inadequate and sometimes biased research methods or test arrangements (Chapter 3). The main task of the study presented here is therefore to remedy these deficiencies on the basis of cognitive development theory. The result shows that Kohlberg's theory does not adequately describe moral development either. Rather, such a theory must first be developed, which we have referred to above as the educational theory of moral development. Developing this theoretical foundation was the main aim of my investigation. 1.1 Practical significance The present study is primarily oriented towards the fundamentals; it wants to bring about a decision between three alternative explanations. This does not rule out that these explanatory approaches or theses have practical significance, especially for pedagogical, educational and social therapeutic measures. On the contrary, the question of the role that educational processes have for moral development seems to be of the highest relevance for the goals and methods of any moral or social education, indeed for education in general. A confirmation of the emergence thesis would mean that children become morally more mature beings if they are left to their own devices and if they are educated as little as possible. Indeed, many reform educators and anti-educators such as Alexander S. Neill (1969) in his famous 18th

19 Summerhill School justified its methods by pointing to the natural development of moral abilities. If this view could be empirically proven, educational policy could be limited to the provision of free spaces in the concrete as well as in the figurative sense of the word. If Kohlberg's cognitive development theory is correct, the worst that can happen in the absence of support would be a slowdown or stagnation in moral development. But what if this theory does not apply, if the moral development is just reversed in the absence of support? We will show that, contrary to the Kohlberg theory, this is actually the case. If the educational stimulation by general educational institutions (usually schools) ceases, people's moral judgment decreases significantly (Chapter 5). Confirmation of the socialization theory would mean the exact opposite. It would mean that the educational authorities of the family and school would have to pay special attention to the teaching of social norms and the control of their observance in order to turn the initially amoral child into a morally competent member of society. The adolescent only needed to understand the moral norms to the extent that this made it easier for him to adhere to them (Durkheim 1976/1924; Brezinka 1988). In contrast, the educational theory of moral development predicts that more is needed than understanding social norms and adapting to them, that, as Piaget and Kohlberg assumed, developed cognitive skills are essential for moral development. Furthermore, this theory predicts that the demands that can be derived from the socialization theory are largely in vain, since almost all children, even at a very young age, have a high sensitivity for fundamental moral principles and these are not first imparted to them have to. 5 What children, as well as many young people and adults, lack is the ability to apply moral principles in concrete situations and to resolve moral conflicts between conflicting principles. Educational theory therefore implies a third way of upbringing. It says that morality is a task field of a similar structure and with similar requirements as any other task field we deal with in life. 5 Recent studies on moral development in childhood are referred to in Chapter 2; see also Lind 1989a. 19th

20, and that therefore very similar didactic efforts are necessary in order to optimally promote moral judgment. Of course, moral problems cannot be solved with the same methods and skills as mathematical or physical problems. But here, too, the following applies: a) knowledge of practical life contexts (moral knowledge) must be imparted and b) opportunities for testing this knowledge on practical tasks must be created. Effective moral education requires special didactics, well-trained educators and appropriate curricula. The first approaches for this are available. 6 Of course, Kohlberg's warning about the psychological fallacy must also be remembered here. No matter how true, empirically tested and perfect a psychological theory may be, it cannot offer a comprehensive plan of action for all actors in a field of application. Acting and making decisions in relation to the world of life is more than the application of a theory, it is the coordination of a variety of forms of knowledge and theories of different origins at a specific moment in a specific situation by a specific person. It would be wrong to want to replace the comprehensive competence that people, even children, bring to such decisions with a single, monolithic theory about a relatively small aspect of the world in which they live. The moral psychologist is always dependent on cooperation with those affected in the field, with teachers, parents and children. According to the educational theory of moral development, democratic-discursive processes are not only an additional condition for the development of moral judgment, but their indispensable prerequisite (see Chapters 8 and 9). The results presented here underline the central importance of moral judgment in the individual development process. The fact that only cognitive interventions, which deal with thinking and discussions about moral conflicts, have a significant and lasting effect of considerable magnitude, speaks more for the validity of cognitive development theory than for maturation and socialization theories to explain the changes in the moral behavior of young people and young adults. The finding that educational institutions, but not the professional environment, have a strong overall 6 Power et al. 1989; Lind & Raschert 1987; Oser & Althof

21 have a positive influence on moral development, speaks for the importance of ability-promoting measures for moral development, and against maturation or adaptation theses and their theoretical and practical implications.7 Moral discourses that promote development can in principle take place anywhere, including in the family, among friends and at work. However, they obviously find the most favorable conditions in general schools. If moral development is not sufficiently stimulated by educational processes, the development of moral judgment stagnates or even regresses. Young people who leave the school system at an early age and enter the professional world, where there is less room for dialogical processes, show a clear decrease in moral judgment (Chapter 5). Kohlberg and many of the educators inspired by him have shown that schools could do their job of promoting moral and democratic skills better than before if they made targeted and methodical efforts to do so (Chapters 8 and 9). Philosophers such as Jonas (1979), Böhler (1992) and Mittelstraß (1992) substantiate with convincing arguments that the promotion of such competencies is urgently required today, in view of a complex environment with its diverse problems. Why, then, has the importance of educational experience in moral psychology been almost entirely overlooked? In addition to the fact that there is no room for the variable education in either socialization or the theory of maturation, the reason seems to me to be that educational experiences have become so general today and so taken for granted by many that we are hardly aware of their effects. In many societies there is compulsory schooling for many years, so that the effects of educational experience can only be distinguished with difficulty from purely age-related (maturation) processes. What is really educational development is often attributed to age-related maturation. 7 This thesis does not contradict the studies by Lempert (1988) and his colleagues, who found progress in the moral development of the industrial workers they interviewed. There may also be educational opportunities in the workplace. 21

22 1.2 Definition and operationalization After it has become clear that a clarification of the empirical validity of the three theories of moral development dealt with here is not only of academic interest, but also of highly practical interest, the question arises as to how an empirical or experimental evaluation can be brought about leaves. A prerequisite for this is obviously that these theories concern an actual fact that can be observed by everyone, so that a consensus on correctness or falsity can be reached on the basis of intersubjective examination. The demand for intersubjective verifiability sounds simpler than it is. James B. Watson (1970/1924), a founder of modern psychology, saw it fulfilled by the fact that every theory is limited to those things that can be observed and that laws are formulated only with regard to these things. According to Watson, what the organism does or says can be observed (p. 6). Only if theories adhere to this seemingly trivial principle can they be understood intersubjectively and assessed on the basis of empirical data. Unfortunately, this methodological principle was taken ad absurdum by its early proponents, the behaviorists, because they went far beyond the goal and stipulated that all those psychological phenomena that are not observable (and therefore do not exist) are not observable (and therefore do not exist) that do not correspond to those known up to now Above all, physical measuring methods were measured. For many psychologists this meant blocking any research in central areas of psychology. Those who did not want to go quite as far as Watson say that psychological dispositions cannot be directly observed, but can be inferred on the basis of indicators and certain additional assumptions. One speaks here of hypothetical or latent constructs. But because of the psychological premises on which it is based, this position is on a highly uncertain, scientific basis. 8 The methodological position of our investigation is that psychological dispositions such as moral judgment and moral attitudes can be understood and measured as manifest properties of behavior (see Chapters 2 and 4). According to this position, known as methodological behaviorism, it must by no means be assumed that moral 8 For a more detailed presentation of my criticism see Lind, 1985a, and chap

23 principles are merely latent constructs that can only be assumed (hypothetical constructs) or grasped intuitively, but cannot be proven empirically or experimentally. Nor is it necessary to take it merely as an arbitrary description of certain acts of behavior and therefore as largely self-referential (in the sense of the statement: Person A eats, so I call him hungry). The focus of the present study is the concept of moral judgment and its operationalization as the structure of judgment behavior in morally relevant situations. According to Kohlberg (1964, p. 425), moral judgment is the ability to make moral decisions and judgments that are moral (i.e. based on moral principles) and to act in accordance with these judgments. This definition is in contrast to two traditional definitions, which it must be shown that they are less adequate. One defines moral behavior as an act that outwardly conforms to social conventions and norms. Accordingly, he would be called moral who carries out an act defined as moral or who neglects an immoral one. This definition of morality is inadequate for two main reasons: a) An act viewed in isolation almost never allows a clear conclusion to be drawn about the morality of the agent. An act can be viewed as moral by others without the person wanting to be moral, or it can serve the sole purpose of being viewed as moral by others in order to serve an underlying possibly immoral purpose. For example, someone gives away his coat to a beggar because he no longer likes it (= non-moral motive) and ensures that the press reports on the gift (= overriding purpose). b) The definition of moral as the conformity of an action with social norms has as many meanings as there are socio-cultural norm systems. One and the same action would have to be judged as moral at one point and as immoral at another time, depending on which social norm is being used as the criterion. For example, the following behavior: One student lets another during a class 23

24 write off sen work. From the teacher's point of view this is a violation of the school's rules, but from the student's point of view it is a fulfillment of the commandment of help among friends. The definition of morality as simple norm conformity is missing two important determinations, namely that the act is based on moral principles of the individual and that these principles can be universalized. The second, linguistic definition of morality as the justification of behavior with moral terms (= moral statements) assumes a reference to action, but cannot substantiate it. In her case, for example, it cannot be decided whether statements with moral content have a determining function for a person's behavior, or whether they are only an expression of an intellectual exercise or a subsequent justification (rationalization) of opinions and decisions. 9 Moral statements offer access to a person's moral dispositions, but viewed in isolation they do not allow a valid measurement of moral judgment. Only the psycho-social context in which moral arguments are used determines their psychological reality. An actor, for example, who recites moral sentences, will not be called judgmental for this reason alone. But if a person is convinced of an important matter by a moral argument, even though it is contrary to his own opinion, we are ready to assume that he has moral motives and that he is able to control these motives on his own To apply doing. The problems of both approaches are avoided in Kohlberg's definition of moral judgment. On the one hand, action is only called moral insofar as it can be proven to be determined by moral principles. On the other hand, moral principles (or their verbal representations) are attempted to establish an empirical connection between moral statements and action with the help of the construct of the judgment of responsibility. Thomas Wren (1990) has this in an excellent study of metaethical positions in psychology Attempt discarded because it implies an infinite regress of judgments (how can the bridge be built between responsibility judgments and action?). Wren, on the other hand, postulates an intrinsic connection between any seriously expressed moral argument and a person's actions. But this hypothesis of seriousness can also not checked in individual cases and thus leaves the problem unsolved

25 presentation in the form of arguments) are only regarded as an expression of this ability to the extent that they prove to be effective in behavior. Moral judgment can neither be defined by behavior alone nor by the underlying thinking, but only by both. According to the Kohlberg definition, on which we, by and large, also rely, the question of how judgment and action are empirically connected makes no sense, strictly speaking. They cannot be defined and measured independently of one another. 10 This shows how closely content-related and methodological issues are linked. Without the possibility of measuring the construct of moral judgment as it is theoretically conceived, the question of which of the theses on moral development is applicable cannot be decided empirically. For a long time, in accordance with the socialization thesis, morality was only measured as adherence to or deviation from social norms (someone who knows and adheres to the norms is morally competent). 11 However, Piaget and Kohlberg were the first to recognize that the core of morality is a cognitive ability that goes beyond the learning of socially recognized norms and tries to translate this knowledge into suitable measuring instruments. The greatest attention was paid to Kohlberg's Moral Judgment Interview (MJI), which is described in more detail in Chapter 3. As shown there, the MJI in its latest revision fully meets the criteria of the so-called classical test theory, which was an important goal of this revision. But to distinguish from this is the question that is of interest in everyday life, whether knowledge of a certain person's disposition to behave (e.g. Mr X is not a cheater) can be used to predict their decision in a certain moral conflict situation (Mr X does not take any money out of the cash register when he is alone in the store). However, this is not a judgment-action problem, but an action-action problem, which in principle can be empirically investigated. In practice, however, not all activities are accessible to research for ethical and other reasons. Only as a literary figure could Dostoyevsky Raskolnikov kill a person in order to study the effect of this act on his moral feelings. The discussion of the Milgram experiment (Milgram 1974; for an overview see Meeus & Raajmakers (1989) .11 See the thorough criticism of Pittel & Mendelsohn (1966) shows how narrow the limits of experimentation in the field of morality can be this kind of operational definition and measurement of morality

This success has, it should be shown, had a negative effect on the theoretical validity of this measuring instrument. In addition, certain instructions in the manual (Colby, Kohlberg et al. 1987) lead to preference being given to the invariance theory, which makes a fair comparison of theories difficult. This bias in favor of the theory of maturation is further exacerbated by the widespread tendency in moral judgment research to overlook the level of education as a condition factor and to examine almost only young people who are in education, i.e. where age-related maturation and educational experience coincide. To overcome these problems, a new measurement approach is used in this study, the moral judgment test (MUT), which allows an undistorted measurement of regressions (Chapter 4). MUT is closely based on Piaget's and Kohlberg's theory of morality, in particular on the structural properties of moral judgment behavior, but not on their methods. The operationalization of the structure is based primarily on the methods of psychological attitude research (Likert scales), experimental psychology (multifactorial designs), judgment scaling and the forgotten diacritical method for checking disposition hypotheses by Egon Brunswik. On the other hand, in addition to young people who remained in training for a long time, those who finished their training early were also examined, so that the influence of age and level of education on moral development can be studied at least partially independently of one another. 1.3 Criticism of the Kohlberg approach The Kohlberg approach made two important discoveries possible: First, almost all people, regardless of gender, class or culture, recognize moral principles such as justice and helpfulness. Almost everyone prefers higher forms of moral argumentation over lower ones in the sense of the Kohlberg levels (Rest 1973; Lind 1985a). Second, the ability to understand moral concepts in one's own behavior. on the diacritical method, see Brunswik 1955; also Chapter 4. 26

To apply the tenets consistently and in a differentiated manner, i.e. what we call moral judgment, is still poorly trained at a young age and needs to be developed. Children often apply moral principles in an inconsistent and undifferentiated way. When your mind is preoccupied with pursuing one goal, other goals have little chance. At the lower levels of individual moral development, people are seldom influenced by moral arguments, especially when they are trying to get their opinion through. Only in the course of development do they learn to combine commitment to a cause with an integrated and differentiated moral judgment. Only then can you consider several goals, moral and non-moral, at the same time when making a decision and integrate them into your self. The present study shows that the Kohlberg approach is in great need of revision as regards the relationship between affect and cognition in moral behavior, the definition of structure and, as has already been pointed out, the assumption of the invariant sequence, according to which: a) Moral affects and cognitions must not be understood and measured as separate categories of behavior, but rather as aspects of one and the same behavior. In his first work on this, Kohlberg (1958, p. 89) defined this relationship in very different ways: (A) as an inseparable unit, (B) as distinguishable but parallel aspects of moral behavior, (C) as the stage and level of moral development and (D) as moral and logical judgment. Kohlberg later rightly dropped the equation of levels with affects (attitudes) and levels with cognitions. Equating the cognitive aspect of morality with amoral, logical thinking skills was arguably an embarrassing solution, since he had no methods at his disposal to conceptualize and measure cognitive aspects of morality independently of the affective ones. Definitions (A) and (B) remain. In a somewhat modified version, they form the basis of our two-aspect model of moral-cognitive development and our measurement approach (see Chapters 2 and 4). With the two-aspect model, central contradictions in the concept of level and structure can be resolved in Kohlberg. For a long time Kohlberg saw an extension of the Piaget phases in his stages. The equation of levels 1 and 2 with Piaget's heteronomy and autonomy was, however, untenable, as was the equation of moral autonomy with postconventionality. 27

28 With the identification of sub-levels (A and B) Kohlberg later tried to define the Piaget phases independently of its levels. This attempt, however, only made the problems in the operationalization of structure and structural wholeness come to the fore even more clearly. How can sub-levels be imagined when levels form a structural whole? b) The structure of moral behavior must not be understood as a specific behavioral content (to which the structure definition in the evaluation manual fits), but must be defined and measured relationally. Structure usually means the set of relations that exist between individual elements. For example, the relational information on how consistently a person uses a moral argument is important for the attribution of moral motives. Consistency cannot be measured by looking at individual acts of a person in isolation, as classical psychometrics does, but only by analyzing the relational characteristics of response patterns. In Kohlberg's interview method, structure is not defined relationally, but formally.Although Kohlberg and his colleagues have constantly tried to clarify this definition of structure, it is still controversial today (Eckensberger & Reinshagen 1980; Lind 1985b; Keller 1990). Remarkably, Kohlberg defines structure relationally in his theoretical work, namely as the manifest pattern of individual judgment behavior. A relational definition of structure is also suggested by the postulate of structural wholeness, which Kohlberg sees confirmed by the high consistency of his respondents' answers. But Kohlberg and his colleagues have opted for an a-structural methodology (the so-called classical test theory) as the basis of their research, which excludes the collection of relational information, such as consistency (Chapter 3). c) Moral development is characterized by a relatively slow build-up of skills (instead of a rapid change in attitudes) as a result of educational processes and an at least just as slow decline in these skills after the elimination of educational and exercise opportunities. Moral development is therefore not, as Kohlberg assumes, characterized by an invariant sequence of stages; Moral judgment declines if it is not stabilized and promoted through discursive processes and educational interventions (see Chapters 5 and 11). 28

29 1.4 The Social Psychological Perspective of Moral Judgment Research Social psychology, Kumpf notes, has become noticeably more a-social in the USA and also in Germany, in the sense that it has turned completely away from the study of individuals and social phenomena (Kumpf 1990, p. 123). This reproach is also made of the cognitive development theory (see Schreiner 1992), but this is only partially justified. Kohlberg (1969) tried early on to theoretically relate individual characteristics on the one hand and characteristics of social institutions on the other, as well as their encounter and interaction. However, little of this has so far been translated into empirical research. The present study therefore takes up the socio-psychological dimension by including communication processes, phenomena of social perception, democratic interactions and processes of social perception stereotyping and pluralistic ignorance in the analysis (Chapters 8 and 10). In addition, as already mentioned, it ties in with the classic attitude concept of social psychology or expands it structurally. As Scott (1968, p. 208) rightly pointed out, the concept of hiring has the invaluable advantage that its precise definition and empirical orientation reveals its inappropriateness. For a long time before Kohlberg, this concept was used almost exclusively to analyze people's morality. 13 setting scales provide immediate verbal behavioral material. They are therefore not affected by the problems (halo effect, anchor and reference stimulus problems) of self-descriptions or self-diagnoses in personality tests. 14 The shortcoming of the classic attitude concept lies in the fact that it does not offer the possibility of ascertaining structural differences, so that it is not possible, for example, to distinguish undifferentiated attitudes from neutral attitudes (Scott 1968; Converse 1970). The structural expansion of the setting 13 May & Hartshorne 1926; Hartshorne & May 1928; Krathwohl et al. 1964; Shaw & Wright 1967; in summary Pittel & Mendelsohn It seems reasonable to assume that people have systematic tendencies when judging others (R. Cohen 1969), even when it comes to judging their own personality. 29

This problem has been largely solved within the framework of our two-aspect model (Chapters 2 and 4; Lind 1982). The attitude concept, which has been expanded to include the structural feature, is apparently the only way to adequately examine theories that are important to us, such as regression, culture-specificity or affect-cognitive parallelism. In addition, it fulfills Ockham-Russel’s demand for intersubjectivity and for theoretical transparency and thrift 15, especially in comparison to so-called qualitative measurement approaches, which are often based on ad hoc, difficult-to-understand assumptions The work presented here is based on a general process model of changing moral judgment. This process model roughly reflects the basic assumptions that we want to put to the test here, and at the same time forms a summary of our investigation (Fig. 1.1). All of our analyzes are based on openly observable human behavior and not on principally unverifiable (hypothetical) constructs or unprovable assumptions about human motives. Where we use mental terms when describing the moral 15 Wilhelm von Ockham's demand entia non sunt multiplicanda, sine necessitate is also passed down as entia non sunt multiplicanda praeter necessitatem. Loosely translated: One should not increase the number of constructs without necessity. Because of the cutting criticism with which Ockham () used this sentence against the scholastic naughtiness of countering every difficulty that arises in a theory by inventing new constructs, it is also known in philosophy as Ockham's razor. The modern version of this requirement comes from B. Russel (1972): If everything in a science can be interpreted without accepting this or that hypothetical thing, then there is no reason to accept it (p. 472; unless otherwise noted, all foreign languages ​​are Quotes in this work translated by the author). 16 Kohlberg has only recently revised his skepticism about the attitude concept and has recognized that structural attitude tests such as the moral judgment test (MUT) can achieve significant improvements: With the MUT, according to Kohlberg (1985), an evaluation algorithm is possible , with which a pure structural value can be assigned to each individual (p. xvii). 30th

Using the behavioral process model, we try to operationally define them as precisely as possible as properties of the behavior of an individual, instead of relying solely on the uncertain basis of our intuition. But at no point do we associate our investigations with the claim to predict a very specific behavioral act of a person, as so many have tried in vain. This renunciation results from an ethical self-restraint as well as from the insight into the fundamental impossibility of such an undertaking. We can only assume that a person's moral judgment, as well as moral attitudes, have an important influence on their personal decision. But we must also assume that this decision, if it is to be appropriate and just, depends just as much on unforeseeable situational circumstances and sensitivities. The present study is therefore limited to the moral judgment and the conditions of its development. This does not rule out that moral judgment is of great importance for the quality of individual decisions. There is even a lot to support this assumption. But it should be left to the reader whether he or she considers the available evidence to be sufficient for this. Our model therefore shows the personal decision of the individual as a blank. 31

32 The chapters of this thesis can be assigned to the components of our process model. The following chapters begin with the central aspect of moral judgment (Fig. 1.1, Box 3). Chapter 5 examines the maturation thesis on the basis of the regression phenomenon. Chapters 6 and 7 focus on the hypotheses that follow from the socialization thesis. The following chapters contain partial investigations on various aspects and mechanisms of changing moral judgment in social interactions in the context of educational processes, specifically on the possibility of effectively changing the moral climate of an institutional context (Chapter 8), on the effectiveness of social interventions that are specifically aimed at direct a stimulation of the cognitive structure of moral judgment behavior (chapter 9) and to the accuracy of the perception of moral facts in the institutional context school (chapter 10). Finally, Chapter 11 examines the question of what influence educational experiences have on moral-cognitive development compared to other factors such as socialization, age and selection. 1.6 Technical information The database The hypotheses of this study are tested on various types of empirical and experimental data. The majority of the data reported here comes from studies in the context of larger research projects or our own projects. The hypotheses were also checked on the results of other studies, for which they were recalculated in many cases or made comparable by means of suitable conversions. The original reports or sometimes the raw data were used for these calculations. In some cases, the results have only been translated into a graph in order to better assess which thesis they support. Further details and the sources are referred to at the relevant point. A substantial part of the database of this work comes from the international comparative 32

33 Longitudinal study by the University Socialization Research Group in Konstanz. 17 From this, various subgroups were analyzed, depending on the question (Chapters 5, 7 and 11). The definition of these subgroups is given in the corresponding chapters. A laboratory experiment to simulate moral judgment (Chapter 6) and a field experiment to change the moral atmosphere in school (Chapter 8) serve as additional data basis for this investigation. The latter was planned and carried out in cooperation with Fritz Oser (Friborg) and the State Institute for Schools and Further Education of the State of North Rhine-Westphalia. The data of the representative study on the moral judgment of adolescents and young adults (Chapter 5) come from an unpublished vocational training study carried out in 1990 by EMNID, Bielefeld. On the statistical evaluation of correlations The present study is relatively strongly influenced by methodological and research methodological considerations. This turned out to be necessary, since our entire prior understanding with which we approach the subject of our investigation has melted into these methods. They should therefore be explained in more detail here (see Lind 1985a). If we base our research on inadequate methods, i.e. on false prior understanding, it cannot achieve much. Unfortunately, many social scientists believe that certain historically evolved methods are absolutely valid, and thus the guarantee for scientificity and the basis of professional affiliation with which they bring all creative research, which is always based on the development of new methods, to a standstill. The real 17 members of the research group were next to the author Tino Bargel, Barbara Dippelhofer-Stiem, Gerhild Framhein, Hansgert Peisert (project leader), Johann-Ulrich Sandberger and Hans Gerhard Walter. At the international level, the study was part of the international comparative longitudinal study Diplomés de l'université: leur Formation et leur Conception de la Vie (for short: FORM) coordinated by the European Coordination Center for Social Science Research in Vienna, under the direction of Hansgert Peisert and W adis aw Markiewicz. See Sandberger et al. 1982, Framhein & Langer 1984, Dippelhofer-Stiem & Lind

34 Collegial cooperation is based on an interest in a common topic. The right attitude to scientific research manifests itself in the ideal of setting up as informative as possible or, as Popper (1968) says, risky hypotheses, and subjecting them to a fair and intersubjectively comprehensible test through observation. The methodological standpoint of anything goes is not represented here. On the contrary, it is required that the methods used not only guarantee a substantive and objective examination of theories, but also that they are a valid operationalization of the question. In the context of our investigation, which focuses on the examination of three theories about the nature and the conditions of moral development, the main question is whether and to what extent these theories and the hypotheses that can be derived from them are in accordance with our observations. This question is very different from studies that are carried out with the aim of obtaining data on a small group that can be generalized to larger groups using inferential statistics. Therefore, as the statistician Kruskal (1958, p. 815) suggests, study groups are treated here not as samples, but as populations. Measures of statistical significance such as t, F or 2 do not generally, as their name suggests, secure the significance or generalizability of a finding. 18 In addition, the level of statistical significance does not provide any information about the quality of a theory or a prediction, since it is primarily dependent on the size of the sample, i.e. a freely selectable variable. Instead of compensating for this disadvantage with complicated measures to ensure the test strength, as suggested in some places, it seems more obvious to directly calculate measures for the effect size (Cohen 1988). From a mathematical point of view, these measures represent nothing more than a standardization of the significance measures for the sample size. These are essentially indicators such as the correlation coefficient r, proportions of variance or d-values ​​(Glass et al. 1978; Cohen 1988). In this work we will mainly focus on r as measure 18 These measures only indicate to what extent a finding is representative of the population of cases from which a sample was selected. The sample results are only representative if the totality is precisely defined and the selection is strictly random (Hays 1963). 34

35 for the strength of an effect or the size of a relationship or a difference (see Chapter 9). But effect size measures also have disadvantages. Contrary to the often used designation as a measure of practical relevance, they are not always suitable for formulating practically relevant statements about the size of an effect. They are not independent of random characteristics of the respective study group that are irrelevant for the statement. For example, if the variance of a variable whose effect is to be investigated happens to be very low or zero in a sample, then no effect can show up, even if it were present. An obvious way of solving this problem is to take the absolute differences in measured values ​​registered in a study directly as a measure of the effect size. If two measured values, e.g. 4.3 and 5.1, the effect is the difference, i.e. 0.8. Here one will rightly object, however, that the number 0.8 can only be meaningfully interpreted if one e.g. knows how large the range of values ​​the scale allows. On a scale of 4 to 6, a difference of 0.8 is a relatively large value; but a very small value on a scale from 100 to +100. The meaning of absolute differences therefore depends crucially on the values ​​that can theoretically be achieved with a certain scale due to its construction. Some authors such as Fend and his colleagues (Fend et al. 1976) have therefore converted all the scales they used in a study to the same length (0 to 20) and thus made comparisons. Taking up this idea in a slightly modified form, I would like to suggest two steps to achieve a standardized effect size (independent of irrelevant conditions): First, the calculation of a relative difference (the absolute difference divided by the absolute scale width) and, second, an agreement on how big a must be relative difference in order to be labeled significant or very significant. As far as I know, no such agreement has yet been explicitly made by anyone, but it is in fact present in empirical research. A review of very different social science studies shows that the range of difference values, which are described as very significant or highly significant in the literature, is large; we found the designation highly significant for relative differences of 1.5 to 15 percentage points. But around three quarters of all difference values ​​that were assessed in this way were between 8 and 10 percentage points. That means, as a rule, difference values ​​that have a 35

36 size of approx. 8 to 10% of the theoretically possible scale width, designated as highly significant or very significant. Hence, in this paper I will use the term very meaningfully for effects that are equal to or greater than 10% of the scale on which they were measured. A relative difference of 5% should only be described as significant. With these two benchmarks we have two practicable criteria for the interpretation of difference values, where the isolated evaluation of effects is called for. In cases where a direct comparison of two or more predictions with the observations is possible, such interpretation aids are unnecessary anyway. 36

37 2 The Two-Aspect Model of Moral Judgment Behavior Our two-aspect model of moral judgment behavior and moral development is based on Piaget's and Kohlberg's cognitive development theory, but modifies it in essential parts. This model reveals which ideas our observation and measurement methods are based on, by means of which we want to test the correctness or falsity of the three theories of moral development in question. It is dealt with in detail here because the question of how the object of investigation is defined, ie what is meant by it in detail, is of crucial importance for the construction and evaluation of the observation or measuring instruments, with the help of which suitable data for the assessment of theories to be found. The two-aspect model was developed primarily to solve some fundamental problems of the Kohlberg approach. As will be shown in the next chapter, there is a contradiction between the theoretical model and Kohlberg's measurement method (the Moral Judgment Interview), since central predictions of his theory cannot, or not adequately, be checked with his observational data. This includes the prediction of an invariant developmental sequence, which will be discussed in more detail in Chapter 5, and the prediction of an affective-cognitive parallelism, which will be examined in detail here. What is meant by the affective-cognitive parallelism? Just a definition or also an empirical fact? Although the assumption of parallelism appears in many works by Kohlberg, but also by Piaget, both have always only expressed themselves very vaguely. Hardly anyone, as William Damon (1983) criticizes, has proposed a satisfactory concept for the relationship between affect and cognition. Gertrud Nunner-Winkler even comes to the conclusion that there is no parallel development of both aspects. The findings, writes Nunner-Winkler (1989a), can only be interpreted consistently if one abandons Piaget-Kohlberg's thesis of cognitive-affective parallelism and reads the different results in connection with the respective survey methods used (p. 593; also Nunner - Winkler 1990a, p. 4). 37

38 This chapter aims to show that a) the concept of cognitive-affective parallelism provides the basis for an adequate definition of moral behavior (and thus also moral development), b) that it supports the hypothesis of a close correlative relationship between affective and contains cognitive aspects of moral judgment behavior, which can be confirmed empirically, and c) that it also implies a hypothesis about the developmental interaction of both aspects, which is consistent with other research findings. 2.1 The problem of recording behavioral aspects The disagreement between the two standpoints is based on the question of how one should imagine the relationship between affect and cognition and how one can adequately methodically record both in research. The hardest part of this problem is probably to understand affect and cognition as aspects of the same behavior and to operationalize both aspects in such a way that they do not suddenly become names for two completely different components. The difference between aspects and components is often overlooked or viewed as unimportant in psychology, although it is common and important in everyday life. The aspects of an object, for example the elasticity or roundness of a ball, cannot be viewed as separate components and cannot be removed from the ball; Components, such as the outer jacket and the bladder in soccer, can easily be broken down into parts and examined separately. Similarly, there are various properties that are characteristic of a person's behavior, but which cannot be materially separated from him. For example, the assumption that a person is helpful but unreliable can be tested against the same part of their behavior at the same time. It is not necessary or even permissible to capture both behavioral characteristics in completely different behavioral segments in different contexts. Because the hypothesis also implies that this person is generally unreliable, especially in help situations: they quickly agree to help, but later notices that they even 38

39