Summary of Kohlberg (1971) From Is to Ought: How to commit the naturalistic fallacy and get away with it in the study of moral development

Summary of Kohlberg 1971: From is to ought.Summary by Topher Hunt, October-December 2010.

Kohlberg, L. (1971.) From is to ought: How to commit the naturalistic fallacy and get away with it in the study of moral development. In Mischel, T., ed. Cognitive Development and Psychology. New York: Academic Press. 151-235.

MAIN POINTS

GENETIC EPISTEMOLOGY AND MORAL PSYCHOLOGY

The articles in this book [Cognitive Development and Epistemology, ed. T. Mischel, 1971] explore questions of how psychological research can inform epistemology and philosophy, and vice versa. It's pretty clear why psychology needs philosophy, illustrated by "two generations" of "inadequate epistemology" dominated by behaviorist stimulus-response theory, where "learning" did not involve "knowing" and only learning mattered, and where the process of learning in epistemology was the same as the process of learning a dance or learning psychotic behavior patterns. Knowledge is an important concept, and Piaget's "fundamental contribution" was to "observe children's development in terms of the categories (space, time, causality, etc.) which philosophers have deemed central to knowing". (152)

Despite the erroneous emphasis of most psychologists in this field, morality is primarily a philosophical, rather than behavioral, concept. Kohlberg has let philosophy concepts inform his research, which is why he has been able to unprecedented research on morality. Philosophies differ and it's not clear which are more adequate than others; but one can come from different philosophies and achieve the same research results. Any classical philosopher ("Kant, Mill, Hare, Ross, or Rawls") is [implied] a better starting point than behaviorist beliefs such as "conscience is a condition avoidance reaction to certain classes of acts or situations". (152)

However, inversely, empirical research on morality can also help to solve some sticky philosophy problems; the "implications I now believe my genetic studies have for philosophic ethics" are the main topic of this article. (Hence - "is" to "ought".) This is "much more controversial ground"; Alston [same volume] is quoted arguing that Kohlberg's moral reasoning development research can't ever escape his "arbitrarily selected criterion" of moral good, and thus can never make normative "pronouncements as to how people ought to reason". Despite such critiques, developmental psychologists must make claims about "ought" ethics based on "is" research, for 2 reasons, namely:

1) any "ethically justifiable educational or other practical application of his research findings" must be informed by ethical philosophy, to steer the research away from producing ethically unacceptable consequences. This is a major concern of Kohlberg's, and his position is that "the stimulation of development is the only ethically acceptable form of moral education". (153)

2) as argued for Kohlberg in a critique by Peters [same volume], "a psychological theory of ethics (or of cognition) is incomplete, even as a psychological theory, if its philosophic implications are not

spelled out. I claim, persuaded by some of my philosopher friends, that an ultimately adequate psychological theory as to why a child does move from stage to stage, and an ultimately adequate philosophical explanation as to why a higher stage is more adequate than a lower stage are one and the same theory extended in different directions." (154)

Piaget "takes his theory of cognitive stages to be a theory of genetic epistemology, rather than to be a purely psychological theory." An adequate psychological explanation must also explain / address the universal human aspects of these concepts, which transcends pure psychology to make some philosophic claims; an "is" is incomplete without some "ought" to it. To attempt to divorce epistemology from psychology research, or to try to conduct psychology research which is epistemologically "neutral", is to commit to an implicit epistemological position which is probably erroneous and thus shouldn't be kept undeclared. Moreover, when a researcher's epistemology is explicit, research findings can "lead to both partial validation and partial correction" of its position. Dewey, Mead, Baldwin, and the other early philosopher-psychologists saw philosophy and psychology as naturally intertwined, but since "behaviorists, logical positivists, and analytic philosophers" these have become estranged. The task of closing the gap between these two disciplines is huge, and involves discriminating between and addressing several forms of "naturalistic fallacy" ("the fallacy of deriving ought statements from is statements). The concept that positive and normative statements should "be based on mutual awareness", is argued to not be a fallacy, and to be essential to re-integrating psychology and philosophy.

UNIVERSALS AND RELATIVITY IN MORAL DEVELOPMENT

- Most social scientists subscribe to a "cultural relativity of ethics", where (socio)moral development is seen primarily as "the direct internalization of external norms of a given culture". This (in his eyes) misleading belief has led to an inaccurate picture of morality and moral learning as "fundamentally emotional and irrational processes based on mechanisms of habit, reward and punishment, identification, and defense." Kohlberg will argue that this ethically relativist position makes statements which are empirically verifiable and, as such, may be falsified by observation.

A. RELATIVISM, TOLERANCE, AND SCIENTIFIC NEUTRALITY: SOME CONFUSIONS OF SOCIAL SCIENCE

As Brandt (1961) points out, ethical relativism typically involves 3 beliefs:

a) "that moral principles are culturally variable in a fundamental way"

b) "that such divergence is logically unavoidable, that is that there are no rational principles and methods which could reconcile observed divergencies of moral beliefs"

c) "that people ought to live according to the moral principles they themselves hold". (156)

a) and b) are distinct propositions; the first is an "is" statement, "everyone has their own values". The second is an "ought" statement, "everyone ought to have their own values". Believing solely a) makes you a cultural relativist; believing a) and b) makes you an ethical relativist. Kohlberg

sees frequent confusion and inconsistent application of these propositions a) and b) at the heart of one form of the naturalistic fallacy - for example, a woman observes that everyone has their own values, then cites some universal value in making a judgment on the situation; yet then denies her own right to make any judgment which goes beyond herself, since everyone ought to have their own values. [The "is" observation becomes reappropriated into an "ought" statement, the implications of this shift apparently being invisible to the person making it.]

Social scientists tend to "think relativism is required by attitudes of (a) questioning the arbitrary or conventioral nature of the morality of their own culture, (b) fairness to other cultures and to minority groups, and (c) scientific value neutrality or objectivity in studying values", and as noted above, frequently confuse cultural relativism (empirical) with ethical relativism (normative). Kohlberg will argue here that "cultural relativism neither gives nor receives logical support from" ethical relativism, the latter exemplified in the three attitudes / values just mentioned. (157)

Fallacy 1 - Quotes Feuer arguing that ethical judgments about values "can be directly derived from establishing empirical truths about their origins". This is a form of the naturalistic fallacy where empirical facts are inappropriately extended as ethical judgments; we may empirically observe some trait about a phenomenon (ie "expressing sexuality diminishes frustration and repression") yet may have no grounds to make value judgments about the phenomenon (such as "sexual expression is right and good"). [See note1.] Another exemplary instance of this form of naturalistic fallacy is worth quoting: "respect for differences between cultures is validated by the scientific fact that no technique of qualitatively evaluating cultures has been discovered." (159) The fallacy: an observation of difference in moral perspectives (ie between cultures) does not amount to normative testimony that no ethical adjudication can occur.

Fallacy 2 - The is --> ought confusion can also work in the other direction: an ethical-relativism norm of tolerance can "lead to confusions about the facts" of cultural relativism. Empirically ungrounded assertions of cultural relativism (a) are based on a confusion between ethical relativity (b) or the relativistic position that no morals are absolutely valid, and ethical tolerance (c) or the non-relativistic position that all human beings should be accorded "liberty and respect" regardless of their beliefs (this being held to be an absolutely valid principle). Failing to take your moral beliefs of tolerance and relativism as an object of analysis and duly determine whether they are held (by you) as universal, can lead you to misrepresent the facts of moral diversity by asserting cultural relativism as a fact, when it is in fact your moral position. When the logic is carefully studied, ethical relativism (b) and the non-relativistic position of tolerance (c) are mutually exclusive, as the anti-universalism of (b) undermines or is undermined by the assertion of tolerance as a universal principle. [This is what Wilber and others refer to as the "performative contradiction" of much postmodern ideology.] Another way to distinguish ethical relativism (b) from ethical tolerance (c) is that ethical relativism asserts that all morals are relative and non-universal, whereas ethical tolerance only asserts that the appropriate response to violations of the (posited) universal morals, are relative and non-universal (ie whether to blame, punish, etc). (160)

1 # p 158 - This form of naturalistic fallacy can also be seen as idealizing certain empirical traits or conditions, here "diminishing repression", as good and right in and of themselves. The fallacy here is that Feuer does not own up to having universalized these values a priori; (at least in Kohlberg's illustrative conversation with Feuer) he claims to be only drawing on empirical science, yet is embedding his comments in a universally applied ethical judgment. He's making an implicit (universal) ethical judgment and calling it a scientific fact; this is logically dishonest.

Norms of ethical relativism, masquerading as ethical tolerance, have lead many social scientists to deny the observed empirical facts of moral development research (eg. that middle-class students attain higher moral levels, more quickly, than lower-class students of same age) - norms steamrolling facts. The appropriate response to these facts is not to call them middle-class-biased (which they aren't), nor to attempt to relativise moral constructs just because these show a cultural skew; but to consider contextually appropriate ways to respond to (eg.) the higher crime rates in poor, lower-class areas. (161)

Between the two types of naturalistic fallacy observed above (facts provoking ungrounded judgments, and judgments provoking ungrounded facts), we see how social scientists can "reject both fact and the natural intuition that there is something universally 'rational', 'ethical', or 'mature' in principles of justice which prohibit stealing".

Fallacy 3 - A third form of naturalistic fallacy is "the confusion of ethical relativism with 'value neutrality' or 'scientific impartiality'". Using the (important) ideal of scientific impartiality as justification for the stance of ethical relativism, is in fact precisely NOT impartial or value-neutral because it "prejudges the facts" in rejecting a priori any "culturally universal criteria which might aid in defining the field of the moral". Just as it would be inappropriate to define "scientific beliefs" as those beliefs of the majority of members of a society, it is equally misleading to define moral beliefs in terms of those held by the majority of a collective (as Berkowitz, 1964 has done). Since "most people or cultures do not agree that morality is defined by the values of the majority", Berkowitz's claim to the contrary is arbitrary and, indeed, hypocritical. [See my reactions to this scientific beliefs argument2.]

Fallacy 4 - Confusing scientific neutrality with ethical relativism can be traced back to Weber's distinction of the rational sphere and the sphere of values, "toward which a rational man or a scientist must take a stance of value neutrality". This has produced a fourth fallacy: "the confusion between the rational as the scientific or factual, and the rational as the value neutral". Value neutrality unacceptably "assumes ethical relativity rather than justifies it. To assume ethical relativity is to rule out the possibility of rational methods of coming to ethical agreement without considering the validity of such

2 # p 161 - I find myself balking a bit at Kohlberg's argument by analogy here. As I understand it, he's saying that just as "scientific" beliefs can't acceptably be defined in terms of the belief of the collective, neither can beliefs in the domain of morals. While I agree with the heart of this point, his way of making it suggests that the domain of moral inquiry and the domain of scientific inquiry (which he has so carefully distinguished up to this point) have the same relationship to collective beliefs. Science, at least, is defined in part in contrast to, and in contrast to a reliance on, the backdrop of conventional beliefs.

However, this character of "scientific beliefs" seems to me to be grounded in peculiarities of the domain of scientific inquiry, rather than grounded in traits shared by all domains of knowledge, or in the nature of the barriers between these domains. Kohlberg doesn't justify why he thinks the moral domain must operate in the same way as science does, and as such, this analogy seems to be a red herring to some extent. The more important question in my mind would be: Just as something inherent to science makes it unacceptable to define it in terms of collective belief, what character inherent to the domain of moral judgment makes it unacceptable to define morality in terms of collective belief?

Kohlberg's response to this (p 162) is to bootstrap a universal definition of morality off of research subjects' own definitions of morality and use this to turn Berkowitz's (1964) own, culturally relative definition of morality against him. Since "most people or cultures do not agree that morality is defined by the values of the majority", Berkowitz's claim to the contrary is arbitrary and, indeed, hypocritical. While this response does undermine the position (attributed to Berkowitz) that cultural relativism is grounded in value neutrality, it does not seem to justify the equation of the structure of scientific beliefs and moral beliefs which was strongly suggested on the previous page.

methods in actual detail". Weber himself claimed that values in the moral sphere cannot be ranked, and by doing so, he confined the idea of "progress" to the technical or rational sphere - but did so in the context of designing legislation, and didn't acknowledge Brandt's (1959) insight that "moral philosophers can define methodological criteria of moral judgment and argument with about as much agreement and clarity as philosophers of science can define methodological criteria of scientific judgment and argument". Weber never distinguished valid and consentual methodological criteria for adequacy of moral inquiry, from arbitrary and value-laden bias; this is a distinction we must be careful to make. In conclusion: "It is illogical to claim that something is impossible in advance of inquiry." (163)

B. EMPIRICAL STUDIES OF MORAL DEVELOPMENT AND THEIR IMPLICATIONS

Despite anthropologist friends' predictions that Kohlberg would have to abandon his moral sequence when he explored moral development in other cultures, Kohlberg found that the form and structure of moral reasoning (ie, the characteristics of instrumental and exchange logic) were present in Taiwanese and Malaysian aboriginal children just as they are in American children, despite different content of justifications. This consistency in form, and the ubiquitous presence of the same elements or dimensions to moral reasoning (listed in table 2, below) justify calling moral development "universal".

- TABLE 1 (p 164) - presents and explains Kohlberg's six moral stages.

- TABLE 2 (p 166) - lists the aspects of, or considerations in, moral judgment, which Kohlberg found pretty universally considered (regardless of what a person thought of that element or how they accounted for that dimension of morality, these elements tended to show up consistently in moral reasoning) - 3 categories are given: modes or "functioning kinds" of judgment (ie obligation, right, standard), elements or maxims of obligation and value (ie social welfare, respect, justice), and issues or relevant social institutions (ie property, norms, truth).

The idea of "moral categories" as delineated in table 2 comes to us through Piaget from Kant, but is better articulated by Dewey. Dewey is quoted at length explaining that moral categories can be extracted from the functional terms which appear in moral discourse, just as the categories of physical science can be extracted from looking at terms used in physics discourse allow us to infer where epistemological divisions exist between functional concepts like space, time, mass, energy, etc.

- TABLE 3 (p 168) - describes how each moral stage might view the moral issue of "Value of human life", with exemplars for each. To demonstrate stage sequentiality, these exemplars were taken from the same individuals, and the name and age of each exemplar are mentioned.

Defining "stages": more than just "age trends". Have an invariant sequence, through which development may proceed at varying speeds. Stages are "structured wholes" meaning they represent "total ways of thinking" about morality rather than representing specific attitudes or beliefs. Insofar as there are various aspects in moral development (ie sub-lines such as cognitive and motivational elements of morality), these aspects seem to develop in relative cohesion. Individuals are typically not 100% encased in a single stage; a common distribution pattern is for the dominant stage to produce 50% of a person's judgments, while the next higher and next lower stage appear each 25% of the time. A stage concept "implies universality of sequence under varying cultural conditions". (167-171)

- TABLE 4 (p 170) lists arguments both for and against taking moral action in the Heinz dilemma, for each of Kohlberg's 6 moral stages, in order to demonstrate that each stage represents ways of thinking and reasoning rather than confining people to specific values or attitudes.

- FIGURE 1 (p 172) and FIGURE 2 (p 173) present line graphs with the frequency of appearance of each moral stage plotted vertically for each of 3 age groups studied (10, 13, and 16 years of age), one graph for each of 5 communities in different countries varying from middle-class urban school children in US, Mexico, and Taiwan, to "isolated villages" in Turkey and Yucatan. In all 5 graphs, from the age-10 to the age-16 groups, frequency of stage-1 and stage-2 responses falls and frequency of stage-4 through stage-6 responses rises. However only in the American children do stages 5 and 6 become most prominent by age 16, and in the two "isolated villages" stage 1 stays most prominent throughout the ages studied.

Kohlberg notes that "we have found no important differences in development of moral thinking between Catholics, Protestants, Jews, Buddhists, Moslems, and atheists", and for religious groups, the same stages of moral reasoning seem to surface, embedded in religious themes and content.

"Basic hierarchies of moral values are primarily reflections of developmental stages in moral thought" - the ordering or hierarchy of values, as well as the basic moral values themselves, is not culturally relativizable, but is found to be the same ordering regardless of culture studied. [See my thoughts on this3.]

Extreme relativism denies any universally meaningful morality, and also asserts: a) that there's no way to legitimately evaluate moral adequacy of different value standards, and b) that "value differences cannot be explained by a theory of morality, but must be explained by a theory of

3 # p 174 - Having established that basic moral values are universal, he then addresses the relativist critique of "the ordering or hierarchy of these values is idiosyncratic and relative". His response to this is that, Look, we've studied moral reasoning and development in all sorts of cultures and found the exact same invariant sequence everywhere.

I want to be clear in my mind that this response hasn't yet reached beyond the empirical data to justify normative assertions about what is mature morality. It's relying on an unstated argument of "Well, everyone we've talked with seems to develop through the same sequence" and in it, the argument of "Everyone, given enough time, points towards these and these traits as indicating mature morality", which is really a meta-form of "Mature morality is defined by the beliefs of the collective". He hasn't yet tied these findings back into his discussion of how to (and how not to) link empirics and norms, and thus his conclusion of "Therefore we can empirically find moral universals" is left somewhat ambiguous in meaning.

Isn't this reminiscent of Berkowitz' argument, which Kohlberg just tore apart (p 161)? No, because Kohlberg's critique was pointing to Berkowitz' hypocrisy in pre-rejecting any universal criteria for defining morals beyond the scope of the local culture and community - a rejection which leads to logical inconsistencies. In contrast, Kohlberg is starting out with the opposite suggestion, namely that there are universals in moral content and development, and indeed this is a position consistent with the beliefs / experience of the majority - including, as he wryly points out, most cultural and ethical relativists.

So in this sense Kohlberg is defining morality by drawing on the eventual beliefs of the global collective, rather than the present beliefs of the local collective (in isolation from a larger context). But I'll acknowledge that he is still committing a naturalistic fallacy here, and has yet to justify it.

I'm curious to see what happens to this argument if Kohlberg finds one tiny exception to part of it. If he someday studies a community or context where, say, the "desired" and the "desirable" aren't distinguished, does this undermine the normative conclusion that a mature morality distinguishes those ideas? In other words, what conclusions can be made from observing universals isn't nearly so interesting to me as what conclusions may be made from observing almost-universals.

psychological need" or of culture. External moral judgments thus are arbitrary and have no meaningful value. [Unsure how to interpret this...4]

"A more popular view" is a "moderate sociological relativism" where a universal form and function of morality are recognized, but morality is seen as relative in content and specifics. This moderate relativism is "sociological" in that it frames morality as "originating at the social-system level, not the individual level". Form of moral reasoning is universal, but the rules and values (both those we find universally, and those we consider more arbitrary) are shaped by sociocultural particularities, and are then internalized to the individual. Examples of this view of morality include Sumner (1906) and Durkheim (1925). Sociological relativism treats evaluations within a culture differently from evaluations across cultures: The former is possible, based on the extent to which an individual's attitude aligns with and respects that of the culture - thus children may be morally judged. However, as (eg. for Durkheim) the content and values of morality do not have meaning beyond the norms of the group, behavior of people in other cultures may not be evaluated by my culture's morals. (176)

Challenging both extreme relativism and sociological relativism, Kohlberg's research finds much less difference in the "concepts, values, or principles" of morality than often portrayed, both in the presence of the 30ish moral categories previously discussed, and in the sequence of stages of reasoning. Moreover, many of the "marked differences between individuals and cultures" which relativists cite as evidence for moral plurality, are in fact better interpretable as the appearance of different developmental stages - these differences, being developmental, aren't arbitrary. Moral differences, both between individuals in a group, between a child and an adult, and between members of different groups / cultures, can be evaluated "as being more or less adequate morally" - thus we reject both ethical relativism, and the more moderate "sociological relativism", as morality has not only universal form but also universal content.

This does not mean, as Asch (1952) would suggest, that moral principles are held the same by all people or by all adults. Differences in morality are not merely differences in perception or comprehension of a situation, leading the same moral principles to be applied differently; there are also "fundamental differences in principles or modes of moral evaluation" on top of the mere differences in interpretation of the world. Moral reasoning itself differs, in addition and somewhat independently of other types of reasoning, but not arbitrarily so much as developmentally.

Notes that "sociological relativism" (which accepts universal moral categories and concepts, but denies universality of moral content) is inconsistent with the empirical finding of a single stage factor which influences moral thinking across culture and produces a universally "invariant order of development". The related Durkheimian assertion that "the basic meaning of 'X is wrong' is 'X violates

4 # p 175 – Kohlberg notes an extreme relativist position which argues that moral judgments imposed from externally don't account for the specific psychology and culture of a circumstance and thus are useless. He comments that this view suggests that one's action is always the 'best' action, from the perspective of the actor. But it's possible to hold this “best action at the time” view, taking a psychological empathic perspective, and still level moral judgments as these are distinct perspectives of judgment. (I'm assuming that Kohlberg wouldn't similarly throw out the considerations of psychology and culture in evaluating an action.)

Put from the perspective of the object of study, then, under this extension of Kohlberg's logic it is possible to be doing the best you can, be acting as conscientiously and morally as you know how, and still to be acting immorally - not only in the eyes of a 2nd perspective but in a more objective, universal sense. That troubles me.

the rules of my group' " is only valid at the conventional level, and applies neither at preconventional nor postconventional levels. The reasoning pattern of conventional morality often allows it to "stretch" to interpret behavior in new ways depending on social consensus and pressures, but underlying this seemingly arbitrary extension of values is a dedication to the norms of the group. (177-178)

Kohlberg proposes this summary view of moral universality: Moral stage 6 is a universally obtainable, though empirically rare, stage which imparts a fixed set of universal abstract moral principles. At lower stages, morality is not fully principled; thus it is more subject to redefinition by specific context and by one's social frame of reference. At lower stages, for example conventional, the universal moral principles (not apprehended fully until stage 6) tend to 'pull' at values such that "the more generalized and consistently held content" approximate some of these and are 'relatively' universal.

"A culture cannot be located at a single stage, and the individual's moral stage cannot be derived directly from his culture's stage." [Brownian drift?] Stage-achievement of individuals takes the form of a frequency distribution / probability distribution: it's easier, and more probable, to develop to stage 6 morality in modern America than 5th-century Athens. Moreover [and underlying cultural moral development], the moral pioneers in any era create a "historical horizontal decalage" effect where it gradually becomes easier to adopt higher moral stages. (178)

Hobhouse's (1906) early research on moral development - interesting that "cultural relativists have never attempted to refute the basic facts" of Hobhouse's argument for social evolutionism. Rejections of Hobhouse's evolutionism, such as the response of Westermarck (1960), have generally taken the form of 'Social moral evolutionism leads to an unacceptable telos which must somehow be arbitrary to the speaker's own perspective', rather than an actual logical or scientific response to the argument. But "developed consciousness" can be defined in non-arbitrary terms using "objective measures of ontogenetic or historical sequence", as both Kohlberg and Hobhouse have done. (179)

The confusion of ethical relativity (morals are non-universal in form and content) and cultural relativity (morals appear differently in different cultures, possibly supported by a universal structure and content of morality) - this confusion itself is "characteristic of a transitional phase in the movement from conventional to principled morality". And longitudinally, all such extreme relativists eventually move on to full moral principlation. "This suggests that relativism, like much philosophy, is the disease of which it is the cure; the very questioning of the arbitrariness of conventional morality presupposes a dim intuition of non arbitrary moral principles". The insight of relativity itself, presupposes an underlying non-relativity. (180)

THE COGNITIVE-DEVELOPMENTAL THEORY OF MORALIZATION

Having already presented the empirical evidence of universal morality, next will present a psychological theory explaining why the universal content of each stage is universal, and why the developmental sequence of morality is also universal. The psychological explanation for upward sequentiality in moral development is "broadly the same as our philosophical justification for claiming that a higher stage is more adequate or more moral than a lower stage". Kohlberg stresses he is not simply asserting that later moral stages must be better than earlier ones; he is pointing out that if there

is a better, it must come later: relative adequacy can explain ontogenesis (time order of moral stages), rather than inferring adequacy from ontogenetic evidence. (181)

A. EVIDENCE FOR AN ORDER OF PSYCHOLOGICAL ADEQUACY IN OUR STAGES

We can bootstrap off of people's a) interpretations and b) evaluations of arguments at multiple stages, in order to support our assumptions about the relative philosophical adequacy of moral stages. Research by Rest (1971) demonstrates that, when presented with arguments from each moral stage, people at a given moral stage will comprehend stages up to and including their dominant stage, but will fail to comprehend (distort, downward assimilate) stages beyond their dominant stage. (Where they do comprehend higher stages, it's usually the next higher stage, and their score usually shows them in transition to that stage.) Thus moral stages have a "hierarchy of cognitive difficulty" to them, and people seem to only be able to correctly comprehend moral stages below theirs. (181)

Increasing complexity does not prove increasing adequacy. In addition, Rest's subjects also ranked stages by moral adequacy: they had high accuracy (ie consistency with Kohlberg's stage order) in ranking the moral stages which they already comprehended, and less accuracy in ranking the moral stages beyond their dominant stage. Subjects are most likely to assimilate arguments one stage above or below their dominant stage. [Interesting example here.5]

B. STAGE ORDER AND STAGE MOVEMENT IN A COGNITIVE-DEVELOPMENTAL THEORY

In what way can this analysis be a psychological theory? In the cognitive-developmental framework, moral stages have a sequence to them for the same reason that any cognitive stages have a sequence to them: "because cognitive-structural reorganizations toward the more equilibrated occur in the course of interaction between the organism and the environment".

The assumptions of our psychological theory: Piaget, Dewey, Tufts, Mead, Baldwin, and Kohlberg have all indicated a view of cognitive development involving stages "representing cognitive-structural transformation in conception of self and society", each stage implicating successive modes of role-taking, and thus opportunities for role-taking (aka perspective-taking) are what drive stage development. All these theorists also assumed "an active child who structures his perceived environment", and thus moral stage development represents the "the interaction of the child's structuring tendencies and the structural features of the environment" - implying "successive forms of equilibrium in interaction" - and implying that an optimal combination of equilibrium and disequilibrium is required for development.

Notes some assertions we can make about moral theory based on our psychological theory. If it can be proven that "forms of moral judgment clearly reflect forms of cognitive-logical capacity, the emotivist notion that moral judgments are the expressions of sentiments is wrong." In addition, our

5 # p 182-183 - a fantastic example of the obsession with "moving to the top" that Zak always talks about: "I am not sure whether you will score my responses as stage 5 or stage 6. They may sound stage 5 in content, but I believe they have an underlying stage 6 rationale. However, if I am mistaken, will you please follow the Rest and Turiel procedures and send me some stage 6 responses so that I may move one stage up and be saved." Being at moral stage 6 is being saved...?

evidence for qualitatively different reasoning at different moral stages, and the invariant sequence of their adoption, allows us to refute a) the Butlerian / Kantian idea that moral principles are inborn intuitions, and b) the view of morals as direct apprehension of facts. Moral judgments are active constructions, not "passive reflections of external facts". [p 184 - I wonder if this could be extended?6]

We can also say, based both in psychology and philosophy, that "the more mature stage of moral thought is the more structurally adequate" - structural adequacy being an evaluation of form which does not evaluate "truth value or efficiency". In specific, the empirical formal criteria for a mature moral stance, namely the extent of differentiation and integration (aka consistency), are analogous the normative formal criteria for "adequate moral judgments", namely the extent of prescriptivity and universality. These "map into" each other respectively: increasing prescriptivity of morality is reflected in increasing differentiation of development, and increasing universality (aka consistency) of morality is reflected in increasing integration of development. Increasing differentiation / prescriptivity, one facet of moral development / adequacy, also represents a gradual differentiation of morality from other judgment spheres (ie the "ought" differentiating from the "is"). Table 3 (p 168) mentions these formal moral criteria and shows their gradual adoption reflected in changing moral values.

Developmental theory considers a more differentiated and more integrated structure to constitute "a better equilibrium" in that it can handle more challenges / tasks, more self-consistently. We see decreasing contradiction and inconsistency (in moral definitions etc.) as we move towards the higher moral stages. (185)

C. THE COGNITIVE COMPONENTS AND ANTECEDENTS OF MORAL DEVELOPMENT

The "psychological assumption" that moral judgment development involves cognitive development, does not mean that moral dev is about increased knowledge; first-grade children, in general, already know the basic moral conventions of their society. Rather, as Alston (1968) has argued, the cognitive component means that morality is a form of "active judgment" rather than a product of "emotional or volitional states". Thus morality, being a form of judgment, has a "definite cognitive structure" which influences "the child's use and interpretation of rules in conflict situations, and his reasons for moral action" rather than simply his awareness of social conventions. (185-186)

By using the term cognitive, Kohlberg asserts that the "characteristic form" of moral judgment at any given stage, is "parallel to the form of intellectual judgment at a corresponding stage". This "isomorphism between the development of the forms of logical and ethical judgment" means that each new moral stage will imply, and make use of, new logics never used before (logics described by Piaget

6 # p 184 - Kohlberg argues that the qualitatively variant reasoning at each moral stage, and the invariant sequence of adoption of these, are grounds for rejecting the views of morals as either a) inborn intuitions or b) (imperfect) direct apprehension of facts. Essentially he's using empirical observations of moral ontogeny and development, in order to justify eliminating candidate views on the form of morality. I find myself wondering, could the same thing be done in the domain of reflective judgment? What would this entail? Identifying ways that developmental (fundamentally longitudinal) data can allow us to eliminate candidate views of the form of epistemology - a topic I have so often seen discussed. Is this even a reasonable comparison?

in his analogous logico-mathematical stages). [p 186...7]

As stated earlier, "true stages" of development are defined as having: a) invariant order across different contexts; b) a "structured whole" or deep structure underlying superficial differences, stages being separated by qualitative rather than merely quantitative differences; c) hierarchical integration, meaning that lower stages are somehow included and reorganized in the higher stages, and also meaning that anyone with access to higher stages can also access / comprehend lower stages, though higher stages are hierarchically preferred.

Because the empirical data shows moral stages to fit with these traits, and because Piaget's cognitive stages also meet these criteria, "it is logically necessary that the two sets of stages be isomorphic". Thus moral stages need logical underpinnings - and "each aspect of the stage must logically imply each other aspect". ["aspects", I think, refer to different elements of moral reasoning, which are linked by logical connectors which increase in complexity over development.]

This isomorphism of moral with cognitive development, and our explanation of moral sequentiality through logical necessity, can explain why moral development always goes in one direction, without requiring an explanation of it through neural hardwiring, external natural forces, or invariant elements of cultural influence or educational systems.

However this does not mean that "moral judgment is simply the application of a level of intelligence to moral problems"; Kohlberg asserts that moral development is its own process, its own line, for which equivalent cognitive-logical development is a necessary but not sufficient precursor. (187)

Based on this necessary-but-not-sufficient relationship, we would expect that "all children at a given moral stage will pass the equivalent-stage cognitive task" but not vice versa. Cites research from Kuhn, Kohlberg, Selman, and others which supports this prediction. (187) In addition, notes that IQ tests ("mental age" tests) correlate with moral maturity, though not as well as Piagetian-stage-based tests.

D. THE AFFECTIVE-VOLITIONAL COMPONENTS AND ANTECEDENTS OF MORAL JUDGMENT DEVELOPMENT

Distinguishes between the state of having a concept available, and actively using it. In many cases it may seem not worth the effort for individuals to strive for as optimal of moral reasoning as their cognitive potential would allow -- after all, many highly moral figures became martyrs. Thus part of the developmental gap between cognitive development (which sets the potential for the moral) and actual moral stage of reasoning, may be explained by volitional and affective factors.

The cognitive-developmental view sees cognition and affect as two aspects or perspectives present in all mental events. The varying intensity of the affective element does not reduce or undermine the centrality and continued relevance of the cognitive element of reasoning. Moral thinking

7 # p 186 - Kohlberg's treatment of cognitive and moral stages, specifying the relationship of these as the cognitive being necessary-but-not-sufficient for the moral, and various other developmental "lines" which influence moral development in varying ways - somewhat uncritical approach? not as philosophically grounded as his other ideas? what would I offer in its place?

is grounded in reference to "sentiments", but access to and differentiation of these sentiments is a cognitive-developmental process which is a sub-part of moral development. Thus "the quality of affects involved in moral judgment is determined by its cognitive-structural development, and is part and parcel of the general development of the child's conceptions of a moral order". People at different moral stages may experience the same affect, but interpret it / respond to it differently (ie, anxiety --> "being chicken" vs. anxiety --> "warning of my conscience") (p 189-190)

Notes that insofar as there's "slip" between stage in intellectual and moral development, this area of cognition (which Kohlberg calls "concepts of moral sentiments") is very closely tied to moral development more broadly defined; it too tends to lag from "pure" cognitive development. (189)

E. SOCIAL ROLE-TAKING COMPONENT AND ANTECEDENT OF MORAL JUDGMENT DEVELOPMENT

The act of "role taking" (aka perspective taking) makes "social cognition" a distinct area of cognition, separate from cognition about the physical world. Moral judgment relies on this ability to take roles. Thus, part of the gap between pure cognitive development and moral development can be attributed to the "amount of opportunities for role taking" available to the person. Moral role-taking can only be practiced in the context of participation in a community context.

Different amounts of opportunities for role taking may also help to explain the different rates of moral development in lower-class vs. middle-class children. Peer-group popularity vs. unpopularity status of American school children "partly arise from, and partly add on to, prior differences in opportunities for role taking in the child's family". (190) Kohlberg notes that providing role-taking opportunities (in order to optimize moral development) does not mean providing the most affectively warm and positive environment possible. While a degree of emotional warmth and acceptance is required for a child's comfortable participation in social discussion, an environment of optimal social role-taking can be relatively warm / affectionate or relatively impersonal.

F. JUSTICE COMPONENTS AND ANTECEDENTS OF MORAL JUDGMENT DEVELOPMENT

Stage of moral reasoning isn't a statement about how much of the world you are encompassing in your moral values; it's about the structure of your reasoning and judgments given your values / beliefs. Example: a stage-1 vegetarian says it's OK to kill and eat Eskimos because they kill seals. There are conceptions of empathy and justice throughout moral development, though these are applied differently at every stage, and these conceptions slowly become more "differentiated, integrated, and universalized" throughout development. (192)

A "conception" of justice doesn't become a "principle" of justice until it moves from being about role taking, to being about the resolution of conflicting roles. True principled morality involves recognizing multiple, valid and conflicting moral claims. (Where this conflict is absent, you don't have a truly moral question.)

There are multiple types of justice principle, including equality and reciprocity (ie commutative,

contractual exchange - punishment & reward). Arguments about what is just involve either a conflict between equality (egal) and reciprocity (merit), or between equal opportunity and equal benefit. (193)

[ *** FAVORITE CONCEPT *** ]

A "justice structure" is a tacit principled moral consensus which underlies shared expectations and values in a community and institutionally coordinates the social needs of reciprocity and equality and allocate the "benefits and burdens" within a community. Social environments with better-organized justice structures facilitate role-taking opportunities more than living with a poorly-organized justice structure will do. Thus, participating in an effective justice structure promotes healthy moral development.

G. COGNITIVE CONFLICT, EQUILIBRIUM, AND MATCH IN THE DEVELOPMENT OF MORAL JUDGMENT

The idea that "moral principles are cognitive structural forms of role taking, centrally organized around justice as equality and reciprocity", supports the assertion / assumption that moral principles aren't external rules nor inherent to our biology, "but rather the interactional emergents of social interaction". Just as logic is an "ideal equilibrium of thought operations", justice represents an equilibrium of social interaction. Both logic and justice rely on the "core conditions" of reciprocity and reversibility. And both are interactional: the justice concept is not merely an internalization of social relations, any more than logic is an internalization of language. (193)

Saying that logical & moral stages (and their conceptual proxies, logic and justice) are forms of equilibrium, leads to the conclusion that they are also interactional. Opportunities to role-take are opportunities to experience and resolve the conflicts (interactions) between one's own and another's actions / perspectives, and find the optimal equilibrium among these. Social environments stimulate development by providing experiences of sociomoral conflict which push one to find a resolution at or above one's present developmental stage.

Exposure to higher stages of thinking as modelled by others "is probably neither a necessary nor sufficient condition for upward movement". "Passive exposure" studies show minimal change in stage. People at a given stage don't tend to comprehend the messages from higher stages, presumably because each stage transition involves a major "cognitive reorganization" rather than just an addition of new factors. Turiel (1969) sees cognitive conflict as a "motor" for such developmental reorganization. He hypothesizes passive exposure works by making the grower aware of better solutions than the ones currently used, forcing them to rethink their current organization - meaning that exposure to higher arguments is only one of several factors in moral development, alongside exposure to conflicts resolvable only at higher levels, and debate with different-minded peers. (194)

MORAL STAGES AS A HIERARCHY OF FORMS OF MORAL INTEGRATION

Summary of our cognitive-developmental theory: "moral judgment is a role-taking process",

with "a new logical structure at each stage" parallel to those of Piaget; the structure is a justice structure which "is progressively more comprehensive, differentiated, and equilibrated than the prior structure". This section will summarize structures at each stage, highlighting these changes. Observing increased equilibration at each stage cannot itself demonstrate increased "ethical adequacy", but can help us identify criteria which can demonstrate this (to come after this section). (196)

STAGE 1

defines sociomoral order as "differentials of power status and possessions", lacking a grasp of logical (and moral) reciprocity. Sociomoral order involves obeying those stronger than you, or being punished by them if you don't.

STAGE 2

has a sense of "fairness as quantitative equality in exchange and distribution". Reciprocity is conceived as tit for tat (whether positive or negative) and conceives of non-interference, "leave me alone and I'll leave you alone".

STAGE 3

can role-take in Selman's (1971) theory-of-mind logical perspectival tasks, which require a recognition that an "other child would anticipate that the finder would try to anticipate him" - stage 2 is incapable of this. Adapting this task into the moral domain, asking children what the Golden Rule tells you to do if someone hits you in the street, will differentiate stage-2 responses (literal reciprocity: hit them back) from stage-3 responses (ideal reciprocity - involves taking both self's and other's roles simultaneously). Stage 3 can see the problematic vicious circle of the "strict equal exchange" of stage 2 -- this is replaced by an ideal, eventual reciprocity which might require unilaterality in some situations. However this ideality is restricted to taking those roles one more directly identifies with; simplistic role stereotypes limit role-taking. The Golden Rule doesn't require the object of role-taking to also role-take, meaning that in developmentally unequal situations it can result in an anemic "turn the other cheek" ethic.

STAGE 4

extends beyond merely orienting around "concrete dyadic interpersonal relationships", to orienting around a "social order of roles and rules" which constitute a larger community. Justice isn't about reciprocity between pairs, so much as a relationship of individual to the social order. Social inequality is seen as an acceptable reflection of effort and talent. Justice is easily conflated with "maintaining social order", and as such, stage-4 morality isn't good at imagining / creating new social orders; better at maintaining the existing system.

However, there's an important distinction between stage-4 ability to maintain order, and the stage-1 ability to merely obey the existing order. Stage 4 is Durkheimian in flavor, although as discussed above, it can only be attained through cognitive and moral development, not through a simple socialization process or internalization / learning of rules (as Durkheim, Freud, and Piaget would say). It's more "rational" in nature than these other views would suggest. (199)

STAGE 5

Many moral philosophers and social scientists argue that it's impossible to develop a meaningful morality grounded in anything beyond the relative context of one's specific community beliefs. The demonstration that a "substantial body (25%)" of society has developed a morality which builds off of stage-4 conventional morality and which "more adequately handles moral problems than does stage 4", undermines this claim. (200)

Limits to stage 4: it gives no guidelines for moral interaction with anyone outside the community context; and it only maintains, cannot easily author or adjust, the social order. This causes friction in times of social change where the creation of new rules is called for.

Stage 5, in contrast, is "rule utilitarian" in that it can step back from the current order enough to evaluate what rules would best serve society's fundamental goals. Stage 5 differentiates the means of social order (laws, private property, etc.) from the ends of social welfare (health, safety, liberty, etc.), and as such, can see rules of social order as mechanisms to an end. For example, a stage-5 "rational member" might put faith in the abstract arrangements created by "Constitutional Democracy" not because of the concrete rules which this generates, but because the procedures of generating those rules take certain social welfare maxims into account (representation, liberty, etc.) (202)

Stage 5 lawmaking procedures always seem to revolve around the idea of social contract (conceived as a procedure for generating rules, rather than as the act of accepting rules already fixed). Because contracts rely on voluntary consent, personal liberty tends to be the most fundamental principle of stage-5 morality. Lawmaking is discussed foremost here because ontogenetically, the lawmaking perspective must be established before moral rules can be made. But in social history, "critical moral philosophy and constitutional democracy" have arisen simultaneously.

Stage 5 morality sees laws as "exemplifications of universal logical possibilities" and as such, becomes aware of the arbitrariness of one's own rule system. This is the basis for the abilities of critical thinking and reflection (which involves thinking about one's own thought, norms, etc.) This reflection raises the question of "Why should I be moral?": the relativism resulting from this question can push people around the 4 --> 5 transition to a pseudoregressive instrumental-egoistic view of morality, reduces stage 5 morality in terms of conformity with the motive of maintaining personal comfort.

The "metaethical questioning" and reflection characteristic of the transition out of stage 4, doesn't always lead directly to stage 5. Sometimes it results in ideologies exalting the self or a given group as the grounding for all morality. "Such college student ideologies" are generally short-lived, but sometimes solidify into long-term orientations which at best resemble spin-offs of stage-3 or stage-4 systems, at their worst produce such "amorality" structures as that of Stalin or Hitler. These exceptions aren't considered stages because they aren't universally present, and mostly remix elements from earlier

stages with no clear higher-order organization present. (203)

Kohlberg claims that stages 5 and 6 are the only two higher modes of moral judgment which may succeed stage 4 - allowing, of course, that the particular content of morality derived from these moral stages may vary greatly. Stage 5 resolves the problem of the arbitrary-ness of laws by grounding these in procedures of lawmaking which draw on non-arbitrary criteria of social welfare. However, stage 5 morality will break down when the contract lawmaking procedures underlying it aren't morally adequate; it has no universal moral grounding beyond its legitimate procedures of developing morality. Stage 6 improves on this.

STAGE 6

In some situations, the constitutional contract procedures sacrosanct to stage 5 don't yield clear morals, or provide laws which are clearly immoral. Example: "civil disobedience of constitutionally legitimate laws" is treated relativistically, or up to personal preference. These situations need a grounding for a universal morality outside of the sphere of law. That's because stage-5 procedural principles of deriving morality (such as the rule-utilitarianism, above) are methodological principles rather than true moral principles. Kohlberg quotes the mutually contradictory arguments of 2 stage-5 philosophers who were asked the Heinz dilemma, to demonstrate that the structure of stage 5 morality can lead to equally valid but contradictory judgments, and thus stage 5 produces non-universal morality in contexts where a legitimately established law can be questioned (the Heinz dilemma being one such context). (206)

Stage 6 is a moral orientation which achieves more universality than stage 5 by grounding morality in "what may alternately be conceived of as (a) the principle of justice, (b) the principle of role taking, or (c) the principle of respect for personality." Kohlberg quotes (p 208) a stage-6 philosopher who grounds his answer to the Heinz dilemma in 2 universal moral principles: the value of human life, and justice. This response acknowledges rule-utilitarianism and the value of social contract, but sets these in their place as institutionalized means to these universal principles which cannot account for all conceivable situations. This stage-6 response is the only response (contrasted with the 2 stage-5 responses) which defines moral obligation as independent from one's social role (and context).

The stage-6 philosopher applies the concept of universality as a principle of role-taking: any person in the same situation should apply the same values and considerations to their moral decision. Notes that this stage-6 form of universalization is distinct from Kant's categorical imperative (ie, only act on maxims which you would have everyone else act on). The categorical imperative can lead to conclusions like "it is wrong to lie to save a life because to universalize lying for good causes is to negate the meaning of truth telling" - because maxims refer to concrete rules rather than abstract principles, in the context of stage-6 universal principles it makes no sense to universalize maxims; and indeed, doing so can produce conservativism which frustrates a post-conventional morality. Stage-6 "primary principles", on the other hand, "explicitly refer to humanity in the person of yourself and every other" and is more universalizable and reversible than are the conclusions generated by a literal adherence to the categorical imperative. (211)

Rather than just promoting formalistic procedures to developing morality, stage 6 seeks "substantive moral principles" themselves: fully universalizable, reversible groundings which state

what's fundamentally morally worthy. This contrasts with Kantian justice which prescribes the formal character of justice, but lacks substantive principles. Indeed, the substantive principle of justice is required to make Kant's " 'ends in themselves' formulation workable" as it lets us specify that the treatment of humans as ends should be considered a right or claim of all people, which correlates duties with rights and makes duties specific to a person, distinguishing Kant's morality of ends from a utilitarianism which creates the "heap of unconnected obligations" critiqued by Ross (1930). (212)

Justice (reciprocity and equality) was a fundamental obligatory theme at every stage just described. Stage 5 defines justice in terms of civil rights, equality of opportunity, and contract. Stage 6 defines justice in terms of human rights based in the value of human beings as ends in themselves - and redefines those stage-5 concepts in terms of this fundamental human worth.

Notes that the stage-6 philosopher equated universalizability with reversibility. Reversibility implies that duties are correlative to rights, and also implies that having a right requires the holder to recognize that right in others. Rawls (1971) offers a thought experiment to concretize these conclusions: A just solution (ie to the Heinz dilemma) is one that all parties would accept if described to them beforehand, assuming they could be placed in any role and didn't know which role they would take. Similarly, the valid claims of each role are those acknowledged by each party in this experiment, and those valid claims define valid duties. "In the sense just outlined, a universalizable decision is a decision acceptable to any man involved in the situation who must play one of the roles affected by the decision, but does not know which role he will play." His [or hers] is not a greatest-good perspective nor an "ideal spectator" perspective (which is impartial and uninvolved); it's a shared perspective, product of the overlay of all relevant role-takings. (213)

OUR STAGES FORM AN ORDER OF MORAL ADEQUACY: THE FORMALIST CLAIM

Thus far this article has shown evidence for a "culturally universal, invariant moral sequence" and evidence that "this sequence represents a cumulative hierarchy of cognitive complexity perceived as successively more adequate by nonphilosopher subjects." Then it followed the logical structure of each stage, showing how each stage adds new logical "features", builds on the old logics, and handles moral problems more adequately than previous stages. It argued that a "justice structure" grounds morality throughout development, though interpreted differently at each stage. These conclusions make up "a psychological theory of moral judgment, a theory which assumes certain philosophical postulates for the sake of psychological explanation." (214)

As Alston clarifies (quoted, p 273 this volume), Kohlberg has claimed that higher moral stages are philosophically "better" in some ways, and must clarify how this is meant. Additionally, Kohlberg's claim here is a form of converting an "is" into an "ought" and thus must be justified, since some forms of this conversion constitute a "naturalistic fallacy".

Some basic clarifications:

First, Kohlberg's claim does not grade the "moral worth of individual persons"; it merely ranks

adequacy of forms of moral thinking. Kohlberg sees "morals" as primarily a kind of judgment, not a kind of behavior or institution.

Second, Kohlberg's stage-6 morality is a deontological theory of morality. This means it deals in duties, rights, and justice, rather than in ultimate aims or ends (teleological) or in virtue (approbation). As such, the claim of moral adequacy Kohlberg makes is that stage 6 moral reasoning produces superior judgments about duties and rights, than other modes of moral reasoning do. No claims are made about the adequacy of this mode of reasoning over teleological or approbative theories. For that matter, in Kohlberg's view, teleological theories don't really fit into the domain of morality. (214)

Third, the claims of moral superiority are just that: claims of MORAL superiority - not claims of greater cognitive complexity, efficiency, parsimony, scientific truth, philosophic adequacy, or epistemological sophistication (although some of these may be tightly linked to moral development). For Kohlberg's claims of moral adequacy to be intelligible, morality must be viewed as an "autonomous domain, with its own criteria of adequacy or rationality" - and not be judged primarily by other criteria of adequacy. For clarity, the metaethical assumption is taken that such moral judgments can't be measured directly and "are not true or false in the cognitive-descriptivist sense" (ie are not empirically verifiable scientific truths); morality is its own domain with its own evaluative criteria distinct from those of science and philosophy.

Specifically, those criteria describe the form of moral judgment (the judgment's ideality, universalizability, impersonality, etc.) rather than its content (ie the conclusion, maxim, etc). But "the formal definition of morality only works when we recognize that there are developmental levels of moral judgment which increasingly approximates the philosopher's moral form." The formal criteria which determine how moral a judgment is, are "only fully met by the most mature stage of moral judgment" -- thus, when evaluated on formalist criteria, more mature stages of moral judgment are "more moral" than less mature stages. (215)

"More moral" is meant both in the evaluative sense, and in the sense that the judgment fits more precisely into the moral domain, such that immature moralities can be described as having many of the qualities shared by judgments in other, morally neutral domains (eg, behave more like aesthetic judgments than like mature moral judgments). In this sense, moral development is seen teleologically (in Kohlberg's treatment of it, not in the content of the theory), as increased access to an ideal morality: moral development consists of approximating a fixed endgoal. This fits in with Kohlberg's earlier comment that moral development involves an increasing differentiation of the moral sphere from other judgment spheres. (215-216)

"Developmental theory assumes formalistic criteria of adequacy", namely differentiation and integration. These criteria are parallel to the (also formalistic) criteria in the moral domain, of prescriptivity and universality. Each moral stage "represents a successive differentiation of the moral from the nonmoral and a more full realization of the moral form". Kohlberg's developmental definition of morality, and associated judgments, are not meant to judge moral worth; they merely observe "a progressive developmental clarification of the function" we call morality, in the same way that Piaget's cognitive stages observed differentiation of cognitive functions rather than rating the intelligence of individuals' acts. Similarly to how Piaget would define "intelligence", "moral judgment" is a function,

present throughout life (at least, starting around age 4); and yet is an endgoal, which remains incompleted throughout most of development - it is "only fully defined by its final or principled stages". (216)

Asserting that more mature stages of moral judgment are "more moral" is a developmental metaethical statement. This statement is not a normative ethical principle, and thus, it's not something we can use to generate moral judgments themselves. Some philosophers (eg Kant) have attempted to convert deontological theories into "rules for praise, blame, and punishment", methods of determining virtue, etc. Kohlberg doesn't think this can be done with his moral theory: there's nothing about stage-6 principles which mandates blame or punishment, nor inherently evaluates virtue. Similarly, saying that stage-6 morality is the most "moral" does not entitle us to make claims about "Why be moral?" or "What good is justice?". Kohlberg's formalist claim is merely that stage 6 represents a more moral (more morally clear / focused) mode of judgment than earlier stages do; and that to the extent that it matters to "play the moral game", stage 6 is the way to do so most correctly. [See my confused thoughts.8]

Quotes Alston re: the ubiquitous disagreement between moral philosophers. In fact moral philosophers who accept formalist definitions of morality largely agree about what the formal properties of morality are - although they may vary greatly in their particular moralities. Moreover, alternatives to the formal view are either arbitrary or relative; no "positive alternative" to formalist morality has been suggested. Thus, Kohlberg's conception of morality (defined formally, as it is) is strongly supported in philosophy. (218)

THE CLAIM FOR PRINCIPLES OF JUSTICE

In addition to the formalist definition, Kohlberg has offered a substantive definition of stage 6 morality "in terms of principles of justice". Principles of morality have a different, more un-usurpable quality than principles in aesthetics and other domains: a good moral judgment is defined by its adherence to principles more than in other domains - that is, according to a universal method of making judgments, which we expect everyone to follow. [p 218 - see my doubts about this definition.9]

8 # p 217 - I'd like to hear Kohlberg's thoughts on the relationship between "moral as opposed to amoral" (or his term "morally neutral") and "moral as opposed to immoral". He never uses negating terms like these, and I really wish he had for clarity. What it sounds like he's saying on p 217 ("our developmental metaethical conception of the higher or later as the more moral is not a normative ethical principle generating moral judgments") is that his "is to ought" claim is merely that higher stages are more "moral as opposed to amoral" - ie, more differentiated from other domains of judgment - but is saying that our understanding of higher and lower stages canNOT inform us as to distinguishing "the moral from the immoral" - nor justify moral action, etc.

Or, perhaps more precisely, Kohlberg is commenting that the FORM of morality (ie, his formalist claim that higher stages are formally more adequate and thus "more moral") can only justify certain claims. That would make sense.

9 # p 218 - I wonder where Kohlberg is grounding his assertion that "the whole notion that there is a distinctively moral form of judgment demands that moral judgment be principled". He is saying that in our very conception and definition of a "moral" domain, given the unique function we allocate to that domain, we've prescribed an absolute value to the principles contained in it. For him, to hold to "principles" of morality means to use a universal method of judgment and choice and to expect all others to adopt given a similar situation.

Does the acceptance that principles are paramount in morality, necessarily follow from all reasonable

Kohlberg is defining principles as "a general guide to choice" rather than a concrete immutable rule - considerations in moral choice.

Categories of principles found in Kohlberg's data include (corresponding to Sidgwick's categories): prudence / self-realization, benevolence / welfare of others, respect for authority / society / others, and justice. These principles appear at all stages (though prudence and respect for authority disappear from consideration by stage 6). These principles are all good in themselves prima facie (not means to a higher end), and they refer to "states of affairs which are involved in all moral situations and are potentially relevant to all people".

Principles are neither means nor ends, neither rules nor values; they are guides to thorough moral consideration. "They reduce all moral obligation to the interests and claims of concrete individuals in concrete situations" and help to resolve moral conflicts / competing claims. When moral principles are thusly reduced to guides for considering claims, we can then say that justice is the most fundamental principle, underlying the others. (219-220)

Benevolence may also be considered relatively universal, as "the content of moral concerns and claims is always welfare"; however, it can never adjudicate between welfare claims. Welfare concern is impetus, "precondition for experiencing a moral conflict" rather than a "mechanism for its resolution". The most obvious method of applying benevolence is through "quantitative maximization" - Kohlberg discussed this earlier, and now dismisses it as "no true moral principle". [see rule utilitarianism, discussion of stages 5 & 6, above]

To review - Kohlberg's argument for justice as the fundamental moral principle: [Some big ungrounded claims here, that I'm left to trust Kohlberg on. If any of these are meant to follow logically from earlier arguments or be redundant to earlier conclusions, they don't / aren't.] (220)

1 - conceptions of justice and empathy are present throughout development, in successively more differentiated & integrated forms

2 - only justice, not empathy, is a principle at the highest stage of development - in that it trumps "all other considerations, including welfare"

3 - other principles, when made fundamental, either don't resolve moral conflicts effectively, or result in wrong-seeming and unsatisfying resolutions.

4 - many philosophers have argued similarly that justice is most fundamental.

5 - in situations of moral conflict with legitimacy such as civil disobedience, only justice can resolve

definitions of a domain of morality? I question that. Intuitively I feel that this jump lacks a consideration of development, particularly development in perception, which would seem to demand of us a more sophisticated handling of the concept of "principles". I personally make different moral judgments of others based on my sense of their development (what they are able to perceive and understand) - which also requires me to differentiate an "ideal" moral judgment from a "person-applied" moral judgment.

Maybe a highly developmentalized approach to morality such as mine is one of the unsustainably unformalistic moralities which Kohlberg cursorily dismisses as marginal to honest moral inquiry (p 218). I can't evaluate that right now. In any case, that's an example of a potential morality which doesn't apply its moral "principles" evenly to everyone.

It's worth noting on this note, Kohlberg's list of principles (p 219 - both empirically derived and philosophically confirmed) omits reference to growth. This doesn't bode well for Wilber's "Basic Moral Intuition", the main inspiration for my moral views.

the conflict adequately.

6 - philosophers have only questioned justice as the fundamental moral principle because of a "refusal to accept a formal deontological concept of morality". (But again, no strong alternative positive definition of morality has been offered in place of this.)

"Formalists who disagree with the primacy of justice usually do so because they wish to keep morality completely content-free." Kohlberg asserts that "the moral form of universality, tied to the notion that obligations are to persons, logically implies the principle of justice" but he can't fully justify that claim in this article; he simply "points to the fact that no principle other than justice has been shown to meet the formal conception of a universal prescriptive principle". Quotes Alston's critique (this volume) that universal presciptivity can be a feature of any moral principle and thus there's nothing special about justice. Kohlberg acknowledges that he has not proven justice to be the only possible universal moral principle, but responds that no one has suggested anything better: "no philosopher has ever seriously attempted to demonstrate that an alternative substantive principle to justice could function in a universal prescriptive fashion in a satisfactory way".

In summary: although a (1) formalistic definition of morality, (2) based most fundamentally in the principle of justice, are not proven as the only possible basis for morality, no one has proposed serious alternatives. Thus "my definition of stage 6 as the way people ought to reason" is not arbitrary nor based on Kohlberg's own preference, as Alston would argue.

FROM "IS" TO "OUGHT"

2 forms of "naturalistic fallacy" Kohlberg is NOT committing: 1) "deriving moral judgments from psychological, cognitive-predictive judgments or pleasure-pain statements", and 2) assuming that moral maturity is part and parcel of biological growth. What Kohlberg IS claiming is that "any conception of what moral judgment ought to be must rest on an adequate conception of what it is". In order for a moral theory to be philosophically adequate, it must "work" empirically - it must be reconfirmed in the moral judgments of the public.

Kohlberg's claims so far about "is" and "ought": (223)

1 - It is scientific fact that "there is a universal moral form successively emerging in development and centering on principles of justice".

2 - This universal moral form sees its judgments as based on "conformity to an ideal norm" rather than to fact.

3 - Science can say whether a philosopher's view of morality "phenomenologically fits the psychological facts". Science can NOT say what morality ought to be (as Durkheim tried to do) because science and morality use different discourses.

4 - "Logic or normative ethical analysis" can however make observations about the relative adequacy of different forms of moral judgment at resolving their own self-declared goals. In that science can contribute insights about the goals and performance of different moral stages, it can aid this analysis.

5 - (the exciting and contested claim): A scientific (psychological) explanation for why people prefer a

higher stage to a lower and thus develop upwards, "is broadly the same as" a moral theory for why they should do so. This linking of is and ought is what allows us to make (meta)ethical claims from empirical observation. Adds that this linkage is only allowable for interactional developmental sequences and not for maturational sequences (such as Freud's psychosexual stages) [I don't understand this distinction, which Kohlberg claims to explain in depth elsewhere, Kohlberg 1969].

References Scheffler (1953) for refuting the position that facts cannot be used in developing principles or "oughts". Claiming that "normative theories need to be grounded on a firm view of the facts" means that "philosophical analysists are justified in asserting universal features" (given Kohlberg's evidence) and highlights that one cannot merely establish "moral common sense", since each moral stage has different common sense [and since presumably the most adequate is demographically a slight minority]. This claim also implies that the ideal moral theory must include the positive features of each moral stage, since the higher (more adequate) stages retains these features. This claim alone "rules out most ethical theories". (224)

However Kohlberg also claims that psychological theory and normative ethical theory are isomophic or parallel - that is, an adequate psychological and adequate normative analysis of a moral judgment "will be made in similar terms" because the psychological description "corresponds to the 'deep structure' of systems of normative ethics" along which moral stages develop. This reiterates Kohlberg's earlier comment that the "formal psychological developmental criteria of differentiation and integration, of structural equilibrium, map onto the formal moral criteria of prescriptiveness and universality". Only "fruitfulness of its results" can justify this "parallelist" claim; Kohlberg encourages philosophers to adopt elements of moral psychological theory, such as equilibration, into their arguments in order to gain new insights and demonstrate this fruitfulness.

Examples of how a stage psychology theory can inform moral philosophy: this can clarify philosophy's role as constructing, rather than merely clarifying or analyzing, morality -- since the research doesn't support the views of morality as naturalistic/Bethamite or intuitionist/Kantian in form. This insight shows that clarifying and codifying the most common or conventional moral form (ala Hare) "misses the true task" of normative analysis. Stage psychology also implies that "arguments for a normative ethic must be stepwise" - ie that arguments for a mode of morality should be addressed to the logics and typical values of the next lower level. [p 226 - I happen to agree, but the argument here is unconvincing. See my critique.10]

A developmental "deep structure" of morality, straddling psychology and philosophy, provides "a new definition of the moral philosopher's task, a definition more exciting than that implied by much

10 # p 226 - Kohlberg comments that stage psychology implies "that arguments for a normative ethic must be stepwise" and gives enough example of this to give me a pretty good sense of what he means by that, but fails to really justify or argue this point. For example, why specifically should an argument to stage 5 be addressed to stage 4 reasoning rather than stage 2? In what specific ways do arguments directed towards the "instrumental egoistic man"-directed arguments for social contract institutions "fail to be fully convincing"?

Kohlberg's relying on the reader either to fully accept his described Piagetian mechanics of his stage theory as given, or he's counting on the reader to know more than I do about the pitfalls of reducing political theory to instrumental stage-2 concerns. Either way, by doing so, he fails to highlight the actual benefits of stage theory for normative philosophy. The benefits of a theory only become clear to me when it leads to conclusions which you can verify as valuable independently of a reliance on the theory, and which you could not have easily reproduced without the theory.

recent philosophic work".

FROM THOUGHT TO ACTION

People are not either honest or dishonest; everyone cheats some of the time, probably distributed on a bell curve. Moreover there's no reason to assume that some people will cheat reliably, since situational cheating tests yield very low correlations. Moreover moreover, everyone has different "bags" of values, and taking actions in conflict with your literal values does not necessarily imply inconsistency, just a contextuality. The way people attribute praise and blame to others doesn't reflect the way they make their own moral decisions. Many times the same behaviors may be taken for highly moral or immoral reasons. Behaviors are not directly morally evaluable, only through one's moral judgment and principles. (227)

Drawing on the earlier (empirical) observation that "the major general individual and group differences in moral judgment are developmental differences", Kohlberg suggests that "maturity of moral thought should predict to maturity of moral action". Mature thought is prerequisite to mature action. Thus "the judgment-action relationship is best thought of as the correspondence between the general maturity" of their thought and that of their action. Thinking of the relationship between thought and behavior in terms of maturity, provides "a broader developmental notion or moral action" than the "bag of virtues" approach. [Really?11]

Mentions studies where principled undergraduate students were much less likely to cheat [on a trivial topic, see text for details] than conventional students, whereas the student's attitude towards cheating (whether it's OK) did not predict whether they chaught. The results imply that "moral judgment determines action by way of concrete definitions of rights and duties in a situation" - if a person's primary defining logics for rights and duties don't find anything in a situational context to anchor onto, their incentive to "act morally" evaporates. To stage 1, "Cheating is always wrong" means "you always get caught"; to stage 3, "it's good to be honest because nice people are honest", and these 2 interpretations will be triggered in different situations. (229)

Moral stages define structure of values, not content of values. The end choice (ie content, action) can vary. Structure varies with each stage: logics of each stage "bring sensitivity to new aspects of the moral situation" to the foreground. However one's morality is a "stable cognitive disposition", despite and throughout affective changes; "the moral force in personality is cognitive". Affect itself is morally neutral, though affect may produce the energy to pursue a chosen moral conclusion. (230)

The image of morality as a "conflict between the flesh and the spirit" (ie id and superego) is misleading. The real moral crises happen when the "superego" internalized moral expectations break down - example: Lord of the Flies. Studies by Kohlberg and related have shown that conventional-stage values were "less independent of what other people think" and thus more uprootable than principled-stage values. If (as Kohlberg's research suggests) a belief's influence on action is dependent

11 # p 228 - Does any serious moral philosopher use a "bag of virtues" approach to describe how moral thought and action are connected? This discussion of "There are different ways to interpret immoral actions" and "Everyone cheats to some extent" seems very simplistic, almost a different writing style, compared to the previous sections. Maybe this is the article's age (40 years old) showing through.

on "the cognitive adequacy of the belief", we must "start theorizing about thought and action in a new way." Inquiry into this "new way" cannot necessarily rely on the analogies of physical or social thought-action, because the realm of morality involves a concept of sacrifice which "has no parallel in the realm of scientific and logical thought". (230-231)

Ends with a quote by Socrates stating 1) that virtue has fundamentally one universal form, 2) this form is justice, 3) virtue is "knowledge of the good. He who knows the good chooses the good.", and 4) knowledge of the good is obtained through philosophy or intuition, not "correct opinion or acceptance of conventional beliefs".

SELECT REFERENCES

Alston, W. P. (1968.) Moral attitudes and moral judgments. Nous, 1968. 2(1): 1-23.

Asch, S. E. (1952.) Social psychology. Englewood Cliffs NJ: Prentice Hall.

Berkowitz, L. (1964.) Development of motives and values in a child. New York: Basic Books.

Brandt, R. B. (1959.) Ethical theory. Englewood Cliffs NJ: Prentice-Hall.

Brandt, R. B. (1961.) Toward a credible form of utilitarianism. In M. D. Bayles, ed. Contemporary utilitarianism. Garden City NJ: Anchor Books.

Durkheim, E. (1925.) Moral education: A study in the theory and application in the sociology of education. (MMM ed.) New York: Free Press, 1961. (1st ed., 1925.)

Hobhouse, L. T. (1906.) Morals in evolution. London: Chapman & Hall.

Kohlberg, L. (1969.) Stage and sequence: The cognitive-developmental approach to socialization. In D. Goslin, ed. Handbook of socialization theory and research. New York: Rand McNally.

Rawls, J. (1971, in press.) Justice as fairness. Cambridge MA: Harvard Univ. Press.

Ross, W. D. (1930.) The right and the good. New York: Oxford Univ. Press (Clarendon).

Scheffler, I. (1953.) Anti-naturalistic restrictions in ethics. Journal of Philosophy 1953 1(15).

Sumner, W. G. (1906.) Folkways. Boston MA: Ginn.

Turiel, E. (1969.) Developmental processes in the child's moral thinking. In P. Mussen, J. Langer and M. Covington, eds. New directions in developmental psychology. New York: Holt.

Westermarck, E. (1960.) Ethical relativity. Paterson NJ: Littlefield, Adams, 1960. (1st ed. 1932.)