Loevinger’s Abductive Method and Her Debate with Kohlberg

Loevinger’s Abductive Method

Many psychologists (Habermas, 1979; Broughton & Zahaykevich, 1977; Snarey et al., 1983) have correctly pointed out that, unlike many other post-Piagetian stage theorists, I have not resorted to ratiocination or “philosophy” as a way of fleshing out the structure of each stage of ego development.  As Snarey et al. put it, the several stages of my conception are “ideal types” a la Max Weber.  This label is evidently considered a criticism.  Let us set aside the question of whether empirical stages are any more aptly characterized as “ideal types” than logical stages such as Kohlberg’s.

If, instead of classing stages under a conception proposed by Weber, one classes them under a more modern title, the sting of the criticism is withdrawn.  The most a propos descriptions are prototypes (Evans, 1967), family resemblances, or “fuzzy sets” (Wickelgren, 1981).  Indeed, the idea of categories defined by a strict logic of necessary or invariant characteristics is now widely recognized, even by biologists, as often so uncharacteristic of the real world as to sound slightly dated.

The identifying feature of prototypes, for example, is precisely that they do not have any invariant feature.  Rather, they comprise a loose assembly of features, any sufficient number of which identifies a member of the class.  “Frequently, only a tiny subset of all the characteristic attributes of a concept will be sufficient to cause us to think of that concept.  There is nothing common to all of the sufficient cue sets for a given concept…, but there are many attributes that appear frequently in these cue sets.  Following Wittgenstein (1953), we say that the cue sets for a given concept have a “family resemblance to each other” (Wickelgren, 1981, p. 33).  This logic corresponds exactly to the Brunswikian logic on which my method is based, namely, that for any “deep structure” of personality, there are many alternative, mutually substitutable manifestations (Loevinger, 1983).

Loevinger, Jane:  On Ego Development and the Structure of Personality, in Developmental Review 3, 339-50 (1983), page 344.

This article was in response to the article by John Snarey, Lawrence Kohlberg, and Gil Noam:  Ego Development in Perspective:  Structural Stage, Functional Phase, and Cultural Age-Period Models, in Developmental Review 3, 303-338 (1983), where they write:

[Loevinger’s] approach has been a trade-off between significant empirical gains and obvious philosophical shortcomings.  (page 321)

I think Loevinger has a fitting last word on the question of philosophy when she concludes:

Humans are marvelously various in their ways.  I am more comfortable than Kohlberg and his colleagues with the idea that the path of the pilgrim is at least partly ineffable.  (page 348)

 

Chronology of the spat between Kohlberg and Loevinger

After publishing his opus maximus in 1981, The Philosophy of Moral Development, Kohlberg moved to consolidate his “invisible university” of developmental research by assimilating various research areas into his method without much accommodation to other approaches.  He had already bypassed Vygotsky’s research with weak arguments (1968); now he was ready to sideline Loevinger’s research.

This attempt was made in a 35 page article in Developmental Review (3, 1983, pp. 303-338), Ego Development in Perspective:  Structural Stage, Functional Phase, and Cultural Age-Period Models, by John Snarey, Lawrence Kohlberg, and Gil Noam.

There are many problems with the content and process of this article.

Process:  it was common practice in the field of developmental psychology at the time to get feedback from an author about the characterization of their work while the article was being developed.  As it happened, Loevinger was left to respond to the mis-characterizations after the fact.  (Jane Loevinger On Ego Development and the Structure of Personality, Developmental Review, 3, 1983, pp. 339-350).  So much for Loevinger’s “membership” in Kohlberg’s “invisible university.”

Content:  There are many assumptions in the Kohlberg et al. article that Loevinger addresses.  I will mention three here, concentrating on the third.  First, they assume Loevinger is writing in the psychoanalytic tradition, a tradition she had clearly diverged from.  Second, they carelessly line up her stages with Kohlberg’s and others in ways that are not clearly justified.  Third, they condemn Loevinger as a philosophical amateur with conclusions such as the following:  “[Loevinger’s] approach has been a trade-off between significant empirical gains and obvious philosophical shortcomings.”  (pg. 321)

Loevinger’s response to this critique:

Many psychologists (Habermas, 1979; Broughton & Zahaykevich, 1977; Snarey et al., 1983) have correctly pointed out that, unlike many other post-Piagetian stage theorists, I have not resorted to ratiocination or “philosophy” as a way of fleshing out the structure of each stage of ego development.  As Snarey et al. put it, the several stages of my conception are “ideal types” a la Max Weber.  This label is evidently considered a criticism.  Let us set aside the question of whether empirical stages are any more aptly characterized as “ideal types” than logical stages such as Kohlberg’s.

If, instead of classing stages under a conception proposed by Weber, one classes them under a more modern title, the sting of the criticism is withdrawn.  The most a propos descriptions are prototypes (Evans, 1967), family resemblances, or “fuzzy sets” (Wickelgren, 1981).  Indeed, the idea of categories defined by a strict logic of necessary or invariant characteristics is now widely recognized, even by biologists, as often so uncharacteristic of the real world as to sound slightly dated.

The identifying feature of prototypes, for example, is precisely that they do not have any invariant feature.  Rather, they comprise a loose assembly of features, any sufficient number of which identifies a member of the class.  “Frequently, only a tiny subset of all the characteristic attributes of a concept will be sufficient to cause us to think of that concept.  There is nothing common to all of the sufficient cue sets for a given concept…, but there are many attributes that appear frequently in these cue sets.  Following Wittgenstein (1953), we say that the cue sets for a given concept have a family resemblance to each other” (Wickelgren, 1981, p. 33).  This logic corresponds exactly to the Brunswikian logic on which my method is based, namely, that for any “deep structure” of personality, there are many alternative, mutually substitutable manifestations (Loevinger, 1983, pg. 344).

The problem is not Loevinger’s understanding or appreciation of philosophy (see her book, Paradigms of Personality, Freeman, 1987); the problem is Kohlberg’s unswerving loyalty to Kantian philosophy that excludes considerations of caring and consequences.  Loevinger concludes her critique of Kohlberg’s critique with:  “Humans are marvelously various in their ways.  I am more comfortable than Kohlberg and his colleagues with the idea that the path of the pilgrim is at least partly ineffable.”  (1983, pg. 348)

In my mind, Robert Kegan, part of Kohlberg’s “inner circle” of his “invisible university”, salvaged Kohlberg’s reputation in regards to Loevinger with this elegantly written remembrance of Loevinger:

In the 1970’s when I was a doctoral student at Harvard, in Lawrence Kohlberg’s shop, Jane Loevinger’s visits were anticipated with something like the eagerness, curiosity, and trepidation a family might have awaiting the arrival of an outspoken, stern but loving aunt whose tough-minded integrity concealed a sympathetic heart.  Ordinary colleagues say highly critical things about your work behind your back and respectful things to your face.  Loevinger was just the opposite.  She would leave a trail of overturned vanity in her wake, and then months later you would hear from a colleague how highly she spoke of what you were up to.

I remember one visit when, after beating us all up for several days with her unerring eye for weak links in someone’s thinking or method, she joined Carol Gilligan and me as we accompanied Kohlberg to a talk he was going to give to a local psychoanalytic institute .  While Kohlberg spoke, amid the dark Victorian woods and high ceilings of the institute, the analysts grew more and more restless (and no doubt “interpretive”) with what they took to be his preoccupation with the superego.  Loevinger, Gilligan and I watched as they proceeded to actively misunderstand and demean his life project in the kindest terms.  We departed with a crestfallen Kohlberg (but not before one analyst had sniffingly asked Jane if she was “the Loevinger of the scales” as if there were something slightly fishy about her).  In the ride home, though, it was Loevinger, much more than his closer colleagues, who shrewdly and sensitively found ways to restore Kohlberg’s diminished spirits.  (“From Taxonomy to Ontogeny:  Thoughts on Loevinger’s Theory in Relation to Subject-Object Psychology”, in Personality Development:  Theoretical, Empirical, and Clinical Investigations of Loevinger’s Conception of Ego Development, Westenberg et al. eds., Psychology Press, 2009, pg. 39)

This is the description of an up-front, confident and caring character, not a muddle-headed empiricist encased in her lab coat with the tunnel-vision of a positivist, or a leering psychoanalyst making subjective interpretations.  [In a personal communication (9/25/16), surviving family members of Jane Loevinger have confirmed Kegan’s description of Loevinger’s character, and thanked me for bringing it to the fore on this website.]

In the end, Kohlberg died at age 59, and is best known for his moral dilemmas, particularly the Heinz dilemma.  [For a pithy synopsis of moral development stages in relation to a moral dilemma, click here for a Calvin & Hobbes version.]

Loevinger died at age 89, and is best known for her research that concluded (in the wake of WWII and Hitler) that the “authoritarian personality” was associated with an intolerance of ambiguity rather than some specific psychosexual aberration.  For a tribute by one of her former students, Dr. Susanne Cook-Greuter, see the following link.

As for Loevinger’s developmental research:  her studies were done with adults and extrapolated back developmentally.  She left it to her students, particularly Michiel Westenberg, to devise sentence completion methods for younger subjects.  (see Westenberg, Jonckheer, Treffers, Drewes, Ego Development in Children and Adolescents:  Another Side of the Impulsive, Self-Protective, and Conformist Ego Levels, in Personality Development:  Theoretical, Empirical, and Clinical Investigations of Loevinger’s Conception of Ego Development, Westenberg et al. eds., Psychology Press, 2009, pp 89-112)  Westenberg’s findings tend toward age ranges more aligned with my conclusions than Kohlberg’s did in his 1983 article.