Findings on care development by the Ethic of Care

A series of studies have investigated care development using the Ethic of Care Interview, constructed by Eva Skoe (Skoe & Marcia, 1991; Skoe, 1993). In addition to a real-life conflict, the measure consists of three care-focused hypothetical conflicts surrounding (1) unplanned pregnancy (cf. abortion study), (2) marital fidelity and (3) care for a parent (see Appendix A). These dilemmas were initially selected because they were expected to represent

“frequently occurring, real-life situations of interpersonal concerns where helping others could happen at the price of hurting oneself”

(Skoe & Marcia, 1991, p. 292). Open-ended interviews, cautiously

22Besides, Flanagan and Jackson (1987) regarded Gilligan’s gestalt-shift metaphor as logically misleading, because it confuses visual perception with moral construal. While it is impossible to see the duck and the rabbit at the same time in the gestalt-shift, it is not impossible to see both the justice and care saliencies in a moral problem.

Consequently, there is no logical reason why both care and justice considerations cannot be used in the same reasoning episode. Noticing saliencies of both, or using both orientations, however, do not necessarily make a moral problem clearer, and easier to solve, but might point out some hidden relevant aspects. Gilligan’s more recent research program seems to share this view; care and justice narratives are seen as intertwined with each other (Brown, Debold, Tappan & Gilligan, 1991).

probed by interviewer, are scored according to five levels, derived from In a Different Voice.

The validation study by Skoe and Marcia (1991) comprised 86 female university students, aged 17-26 years, focusing on the question whether women’s care and identity development are closely related. They found that the sequence of care levels paralleled the sequence of identity development, measured by Marcia’s Identity Status Interview that is based on the expansion of Erikson’s theory of identity development. More specifically, students with diffused identity, uncommitted to any definite directions in their lives, scored the lowest at the self-oriented level, along with students with foreclosed identity. It is worth noting that those foreclosed students indicated unquestioned commitment to childhood-based norms and values. By contrast, students with moratorium status, undergoing exploration of alternatives in their lives tended to score at Level 2.5 that correspondingly indicates questioning of the conventional care ethic. Finally, women with achieved identity status, having made both exploration and well-defined commitments, tended to evidence the highest level, balancing care for self and others.

Hence, women’s care and identity issues were found to be closely related, as can be predicted from Gilligan’s theory.

This research pattern was replicated in a sample of 76 female and 58 male high school and university students, aged 16-30 years (Skoe & Diessner, 1994). Findings about women were replicated, and men’s identity was also found to be related to care development, but to a lesser extent. Interestingly, care development was more related to identity development than justice development for both genders, but only for women was this difference significant. Recently, Skoe and Lippe (2002) have explored how care development is related to another ego development measure, Loevinger’s Sentence Completion Test, with a Norwegian sample of 144 participants, aged 15-48 years.

Again, care development was more strongly related to ego development (r = .58) than was justice development (r = .20), measured by Rest’s Defining Issues Test (DIT). Thus, the study gives further support to the assumption that care reasoning is of more relevance to personality or identity development than justice reasoning. Furthermore, androgynous gender role orientation (having both traditional feminine and masculine characteristics in

self-description) has been found to be related to higher levels of care reasoning, but only for women (Nicholls-Goudsmid, 1998;

Söchting, Skoe & Marcia., 1994; Skoe, 1995). It seems that in order to release themselves from the ethic of self-sacrifice, women have to adopt more assertive characteristics typical for traditional masculinity (Skoe, 1995). To summarize so far, these results indicate that women’s care development towards highest levels requires giving up the traditional feminine gender role and questioning tradition-based values, in line with Gilligan’s (1982) contention.

With regard to Gilligan’s claim of female-specific care reasoning, none of the aforementioned studies found proposed gender differences in care reasoning. Interestingly, however, they emerged in two Canadian mature adulthood samples, aged 40-84 years, women scoring higher than men (Skoe, Pratt, Matthews &

Curror, 1996). Studies among early adolescents point out the similar gender difference in Canadian (Skoe & Gooden, 1993) and U.S. samples (Meyers, 2001), whereas it was not found in a Norwegian sample (Skoe & al., 1999). These results suggest that gender role stereotyping is stronger in North-America than in Northern Europe, especially among older age cohorts, and furthermore, they imply that Gilligan’s observation of female-related care reasoning had some empirical basis in her own culture sphere. These results also underscore the relevance of cultural factors to care development, as well as point out that North American findings cannot be generalized to represent all Western cultures (Skoe, 1998).

Studies have found positive correlations for care and justice reasoning, ranging from .21 to .63, (Skoe & Marcia, 1991; Skoe &

Diessner, 1994; Skoe & al., 1996, Skoe & Lippe, 2002). There is some support as well to the assumption that care and justice reasoning are not related to each other. Skoe and Lippe (2002) found that the relationship between the ECI and the DIT was not significant any more after controlling for verbal intelligence, and Skoe et al. (1996) found that justice and care reasoning were not significantly related on real-life dilemmas (n = 27, was however small). These correlations nevertheless indicate that care and justice reasoning partly share an underlying developmental path, rather than being separate alternative developments. Both modes of moral development describe progress from the initial

self-oriented concern towards other-concern, and thus may reflect certain elements of ego development, such as cognitive style, impulse control and character development (Skoe & Lippe, 2002;

Snarey, 1998). In line with this assumption, fearful attachment style, based on self-reporting, has been found to be associated with the lowest levels of care, and secure attachment style with the other levels, indicating that a sufficient inner security is a prerequisite for moral progression beyond self-oriented level (Söchting, 1996).

Furthermore, both modes of moral reasoning have been found to be positively related to such cognitive-developmental indexes as role-taking (Selman, 1980; Skoe & al., 1996) and integrative complexity of reasoning (Skoe & al., 1996). Söchting (1996) found that sophisticated cognitive appreciation of others’

subjective states was related to higher care levels, whereas affective-based appreciation was more related to Level 2. These findings indicate that care development really concerns cognitive processes of reasoning. With regard to empathy-related capacities, care reasoning has been found to be positively related to self-reported perspective-taking, and negatively related to personal distress, whereas empathic concern was related to Level 2, and only for women (Skoe, Hansen & Nickerson, 2001). To date, cognitive-based role-taking seems to be important in care-reasoning, whereas emotional capacities might have a curvilinear relationship, somewhat incongruent with the previous theory-building (Gilligan, 1982; Gilligan & Wiggins, 1988; Noddings, 1984).

Skoe (1998) argues that development in moral reasoning is based on questioning the previous positions and the formulation of a new, more inclusive position. These processes are similar to the Piagetian cognitive processes of disequilibration and accommodation. Still, Skoe and Lippe (2002) do not regard cognitive processes as central to care development, but refer to Loevinger’s concepts of internalization of interpersonal relationships, and motivation for mastery, derived initially from psychoanalytical theory. Care development appears to progress with crisis experiences, including interpersonal conflicts and losses (Gilligan, 1982; Skoe & Marcia, 1991) that also transform the experience of self in the interpersonal context. The systematic investigation of determinants of development is lacking, however,

with the exception of Pratt, Arnolds and Hilbers’s (1998) study.

They examined the relationship between parenting style and adolescents’ care reasoning, and found that the mothers’ emphasis on care in socialization narratives was positively related to adolescents’ care reasoning. In addition, the mothers’ emphasis on care was positively related to the girls’ personal adjustment in terms of lesser loneliness and higher self-esteem, whereas it was negatively related to boys’ adjustment. This implicates that even if there are no gender differences in terms of developmental levels, care reasoning may operate differently in men and women, as argued by Skoe and Diessner (1994).

To conclude, several findings confirm that the ECI is a valid measure of care development. The developmental trend is parallel to age. The youngest participants in studies have been 10 years old, demonstrating a self-oriented level (Meyers, 2001) and no adolescents have been found to score at the postconventional levels of 2.5 and 3 (Skoe & al., 1999).²³ The distribution of late adolescents and adults across developmental levels has been found to be roughly similar in Canada and Norway; about 15% reaching Level 3, 35% scoring at Level 2.5, 20-25% scoring at Level 2, and 20-30% scoring at Levels 1 and 1.5 (Skoe, 1998). Moreover, the ECI has been found to be positively related to volunteer helping work (Skoe, Pedersen & Hansen, 1997), consultation with others, sense of availability of cognitive support, self-reported health and positivity about one’s aging, whereas it is negatively related to authoritarism (Skoe & al., 1996). The ECI score has recently been found to be distinct from verbal intelligence, which is relevant for its validity as an open-ended interview measure (Skoe & Lippe, 2002).

Further research is needed to explore specific determinants, as well as general characteristics of care development. The only existing longitudinal study has been conducted in a small Canadian sample, showing that care reasoning was relatively stable in middle and mature adulthood over a 4-year period (Skoe, Pratt, Matthews & Curror, 1996). That study was methodologically restricted, however, because the standardized measure was not yet available and only real-life dilemmas were

23For adolescents, revised dilemmas, surrounding family and friends have been used (Skoe, 1998).

used as a measure. The goal of this study is contribute to these issues by exploring care development in the context of justice development. Consequently, it addresses the question whether care reasoning progresses in invariant and irreversible sequence, forming qualitative different wholes and being hierarchically integrated, i.e. satisfies the criteria for development put forward in Kohlberg’s theory (Colby & al., 1987).