Moral Identity and Developmental Theory

The notion that self-identity and morality are deeply implicated has long-standing roots in both ethical theory and psychology. In ethical theory it is evident in Harry Frankfurt's [1971] account of what it means to be a person: A person (as opposed to a wanton) is someone who cares about morality. A person cares about the desirability of one's desires (second-order desires) and then wishes to will them all the way to action (second-order volitions). Similarly, Charles Taylor [1989] argued that identity is defined by reference to things that have significance for us. It is the result of strong evaluation about what is worthy or unworthy, and these discriminations are made against a horizon of significance that frames and constitutes who we are as persons. He writes, “My identity is defined by the commitments and identifications which provide the frame or horizon within which I can try to determine from case to case what is good or valuable, or what ought to be done or what I endorse or oppose” [Taylor, 1989, p. 27].The affinity of selfhood and morality is a theme in several psychological traditions as well. Erikson [1968, p. 39] argued, for example, that an ethical capacity is the “true criterion of identity,” but he also noted that “identity and fidelity are necessary for ethical strength” [Erikson, 1964, p. 126]. This suggests that moral identity is the clear goal of both moral and identity development and that in the moral person the two developmental tracks are ideally conjoined. Similarly, Damon and Hart [1982] showed that, within each domain of the “Me Self” (physical, active, social, psychological), the highest level of self-understanding implicates a moral point of view. This suggests that the moral self is the clear outcome of self-development [Lapsley, 2005]. Indeed, recent research has shown that morality is considered indispensable to selfhood; it is the moral self that is essential to our identity, more than personality traits, memory, or desires [Strohminger and Nichols, 2014]. It may well be that, for all the contingent facts about ourselves, it is our moral integrity that is the necessary fact of the “real me” [Carr, 2001].Of course few have done more than Blasi [1984, 1985] to elevate the importance of moral identity in post-Kohlberg moral development research. For Blasi, the moral person is someone whose very selfhood is constructed on moral grounds; it is someone whose desires reflect a wholehearted commitment to morality. Morality is essential, important, and central to self-understanding; and, to the extent that not everyone prioritizes morality in this way, it is also a dimension of individual differences.Blasi's account of moral self-identity struck a chord. It encouraged reflection on the link between personal agency and the construction of moral ideals. It raised questions about how best to understand moral character. It opened up possibilities for engaging other psychological literatures, particularly those regarding personality and cognition, with the goal of deriving robust integrative models of moral functioning [Lapsley & Stey, 2014]. It is implicated in research on the moral ideal self [Hardy, Walker, Olsen, Woodbury, & Hickman, 2014], moral exemplars [e.g., Colby & Damon, 1992; Walker & Frimer, 2007], and social-cognitive accounts of moral personality [Aquino & Reed, 2002; Lapsley & Narvaez, 2004].From these perspectives, Krettenauer and Hertz [this issue] assembled what they call the “standard model” of moral identity: Moral identity is the degree to which being a moral person is important to an individual's identity [Hardy & Carlo, 2011]. The goal of moral identity development, on the standard account, is the integration of self and morality; it is the integration of values with motivational and emotional systems. The authors add an important codicil to the standard account: The integration of self and morality is not available to children. Childhood, on this account, is “void of identity.”But the authors think that Blasian moral identity has been given enough time to show its mettle and has not delivered all that it promised. It promised to offer a resolution of the judgment-action gap, but there are doubts about whether moral identity is a better predictor of behavior than moral affect. It implied a developmental model that has not been adequately tested, let alone vindicated. For example, indices of moral identity are uncorrelated with age over the course of adolescence and emerging adulthood. Krettenauer and Hertz are concerned to understand why the developmental claims of the standard model have proved unavailing and to offer a number of recommendations by way of remedy.In their view one reason for the paucity of developmental research is the “top-down logic” that has characterized research on moral identity. This means that extant research focuses on the measurement of moral identity in adolescents and young adults and has neglected charting the developmental trajectories that get us to adult manifestations. A second reason is that current (adult-centric) measurement strategies are insensitive to the developmental features of moral identity even if developmental research were undertaken. So, on their view, we need both a new conceptual understanding of moral identity and new ways to measure it. Certainly, both are welcome.Krettenauer and Hertz propose a way of rescuing the developmental claims of moral identity by linking it to three domains that show more developmental promise. One might find evidence of development in literatures that track the increasing differentiation and integration of self-development. Harter [2012] showed, for example, that self-conceptions are both increasingly differentiated as one moves from early to late adolescence but also more deeply integrated. Hence the rhythm of development bids the adolescent to differentiate real versus ideal selves, or display multiple selves across different contexts, or create self-evaluations that differ across domains, but, at the same time, integrate disparate selves into a higher-order generalization, create a global sense of self-worth, or coordinate disparate and contradictory aspects of the self into a coherent self-system. Perhaps something like this can be deployed to understand the rhythm of moral identity development.Self-determination theory is also held out as a promising source of developmental claims. The continuum of self-determination moves from several forms of extrinsic regulation (external, introjected, identified, integrated) to authentic internal self-regulation where the self is experienced as the locus of causal agency [Ryan & Deci, 2000]. Perhaps this continuum is a model of how moral identity development might proceed. Finally, one might look for developmental themes in life story narratives [e.g., McAdams, 2009; McAdams & Pals, 2006]. For example, moral identity development could be a matter of creating a life story but one that follows either an essentialist (where the focus is on self-traits) or narrative line. In the narrative approach, a life story is constructed in a way that brings coherence to past and future conceptions of self-identity.This is perhaps enough to reprise the major themes of Krettenauer and Hertz's thesis. What is attractive about the proposal is just how strongly integrative it is. It folds the “self-importance” aspect of moral identity within the context of what Harter [2012] has to say about differentiation and integration in the development of the self. It locates moral identity as a moment in the growth of internal motivation as outlined in self-determination theory. It holds out a place for moral self-identity in the way we make sense of our lives through personal narrative.Hence, self-important moral traits will show differentiation and integration; behavior will be self-regulated increasingly by internal sources of motivation and be connected to larger narratives of the responsible self-as-agent. Indeed, the narrative understanding of the responsible self anticipates connections to important new constructs [e.g., moral agency; see Pasupathi & Wainryb, 2010]. Moreover, as the authors point out, their scheme aligns with three layers of personality as conceptualized by McAdams [McAdams, 2009; McAdams & Olson, 2010; McAdams & Pals, 2006], a conceptualization that has strong heuristic value.These are entirely plausible and welcome suggestions. Indeed, Krettenauer and Hertz propose a highly interesting and possibly field-expanding contribution to moral identity theory to the extent that it burnishes the developmental claims of the theory. Yet there are grounds for caution. My concern is that the proposal overstates its developmental promise and gives slight notice to the real strengths of extant theoretical approaches which seem to offer a much more promising basis for building the developmental case for moral identity.One problem is that, of the three alternatives, only the differentiation-integration angle is explicitly developmental. The authors themselves seem deeply ambivalent about the developmental bona fides of the narrative level of abstraction; and the external, introjected, identified, and integrated modes of external regulation in the continuum of self-determination are characteristic adaptations to social contexts and are not stages of development, as the authors note. Moreover, the fact that all four modes can be motivating to the self-same individual seems like a problem for a developmental theory (without further exposition) and not a mere caveat (as the authors would have it). And in self-determination theory, the importance of autonomy-supporting contexts cannot be overstated, so a moral identity theory premised on self-determination would have to include a contextual or situational element that is thinly described in the present account. A social-cognitive approach does much better on this score, as I will argue below.The authors blame the top-down logic of current approaches to moral identity for the current lack of developmental specification. But this cannot be correct. The so-called top-down logic is, in fact, required of developmental accounts that accord with the orthogenetic principle [Werner, 1957]. The authors implicitly commit to an orthogenetic account of development in their embrace of the differentiation-integration of self-development. Indeed, Harter [2012] appeals to differentiation-integration just to underscore the cognitive developmental aspects of self-construction.But, if any developmental process accords with the orthogenetic principle, it is cognitive development. Here a conception of the end-point is crucial for making sense of the developmental processes. The top-down logic makes developmental explanation possible. The end-point makes reference to a standard that allows one to distinguish progressive development from mere change, and the standard is instantiated in the conceptualization of the end-point. Put differently, developmental change, if it is to count as an instance of development, is evaluated in terms of how closely it approximates the final stage of the developmental process [Kitchener, 1983]. Indeed, “the developmental end-state is a normative standard of reference by means of which we can evaluate the direction of development and its degree of progress towards this goal” [Kitchener, 1986, p. 29]. Far from an impediment to developmental analysis, a conceptualization of the end-state is utterly required. Hence, the authors cannot simultaneously reject (what they call) top-down logic while also touting an approach to self-development (e.g., in terms of increasing differentiation-integration) that requires it.There is also a sense in which the authors appear to have given up on moral identity development, or else have overstated the developmental promise of the approach they favor. More than once they remind readers that there is no moral identity in childhood and that all the action is in adolescence and young adulthood. Apparently there is not much action for all that given that moral identity does not appear to be correlated with age over this period. But surely correlations with age over a restricted range have little probative value and are an ambiguous indicator of development in any event.Yet even if we accept the authors' conclusion that the lack of correlations with age during adolescence counts against the standard model, it would seem odd that the remedy is to completely abandon childhood in the search for developmental precursors. To do so leaves us in a far worse place than the standard model that does not give up on early developmental precursors of moral identity. The authors do acknowledge some evidence of moral-self integration in childhood but do not follow the implications of this for establishing a developmental line for moral self-identity in young adulthood. Instead, we are referred to the personological approach favored by McAdams [McAdams, 2009; McAdams & Pals, 2006] for our developmental intuitions.The difficulty is that the “new Big Five” is not a theory of development but rather a taxonomy of perspectives that frame a unified science of personality. It is more metatheory than theory. It tells us where to look for moral personality (e.g., in dispositional traits, characteristic adaptations, self-defining life narratives) but not with any specificity for the purposes of theory-building. I agree with the authors that personal life story narratives are a hugely important component of any theory of moral identity, but we get no closer to a developmental story by waving our hands at “characteristic adaptations” or “dispositional traits.” Here we are simply naming things that must have preceded construction of life story narratives. What we have not done is the hard work of articulating a theory of development that yields moral identity (even as life story narratives) as an outcome. With the approach favored by the authors, we have substituted a taxonomy of categories relevant for personality for developmental theory and hence are no closer to a developmental analysis of moral identity.But we are closer to a developmental analysis of moral identity than the authors acknowledge. Although Krettenauer and Hertz mention Blasi's sadly neglected account of identity modes, they pass over his own seven-stage developmental account of moral character that ranges from early life to adulthood [Blasi, 2005]. At the highest stage, there comes into clear focus the notion of “wholeheartedness” [a notion derived from Frankfurt, 1988] that refers to a commitment to structure the will around moral desires. Wholehearted commitment to moral desires, to the moral good, becomes an aspect of identity to the extent that not to act in accordance with the moral will is unthinkable. It is hard to say whether this sequence has empirical promise, but we should also not conclude too hastily that childhood is a theoretical void with nothing of interest to contribute to a developmental story for moral identity.In my view, the social-cognitive account of moral self-identity has more resources for achieving what Krettenauer and Hertz say they want, which is a theory that accounts for behavior and has a developmental story to tell (and one that reaches all the way down into early childhood). Aquino and colleagues, for example, have developed an impressive research program that documents the many powerful ways that moral-identity centrality moderates behavior of all kinds and in many settings [e.g., Aquino, Freeman, Reed, & Lim, 2009]. This research program has shown that our self-concept is a mélange of different identities of which morality is but one and that the key variables of interest are not so much differentiation-integration but rather the availability and accessibility of moral identity as a social-cognitive construct. The chronicity of social-cognitive-moral schemes is the defining feature of moral personality [Lapsley & Narvaez, 2004]. Moreover, social-cognitive theories of moral personality readily account for the dynamic interaction between persons and contexts, for “situationism” in the display of moral behavior, and for individual differences. And there is a plausible developmental story, too.A plausible developmental account of the moral personality would not begin in adolescence, as Krettenauer and Hertz might prefer it, but in early childhood [e.g., Lapsley & Hill, 2009; Lapsley & Narvaez, 2004]. This account must have several features. It must track the sort of developmental change that yields adult moral identity as an outcome. It must pick out notions of wholeheartedness but also key social-cognitive variables such as availability, accessibility, and chronicity. It should also anticipate novel facts, some of which are corroborated [following Lakatos, 1978].Where should we find the developmental source of wholeheartedness that characterizes the standard model of adult moral identity? Perhaps it is found in the emergence of conscience and the committed compliance of young children to sensitive and responsive adult caregivers [Kochanska, 2002a]. Children with a strong history of committed compliance to parents are likely to see themselves as embracing parents' values and, on this basis, construct a moral self that regulates future behavior [Kochanska, 2002b]. Indeed, in one study Kochanska, Koenig, Barry, Kim, and Yoon [2010] showed that toddlers with a strong history of internalized “out-of-sight” compliance (at 25-52 months) were competent, engaged, and prosocial at 80 months of age. But this effect was mediated at 67 months by the moral self. Kochanska et al. surmised that the self-regulation of the moral self might be due to the high (chronic) accessibility of moral schemas, which is precisely the mechanism anticipated by social-cognitive accounts of moral self-identity [Lapsley & Narvaez, 2004; Narvaez & Lapsley, 2009].We have also attempted to articulate a developmental account of the moral self from a social-cognitive perspective [Lapsley & Hill, 2009; Lapsley & Narvaez, 2004]. On this account, moral chronicity is built on the foundation of generalized event representations that characterize early sociopersonality development [Thompson, 1998]. These prototypic knowledge structures are progressively elaborated in the early dialogues with caregivers who help children review, structure, and consolidate memories in script-like fashion [Fivush, Kuebli, & Clubb, 1992]. Autobiographical memory is also constructed within this “web of interlocution.” Parents help children organize events into personally relevant autobiographical memories which provide action-guiding scripts that become frequently practiced, habitual, and automatic. A sense of the morality can become a part of the child's autobiographical narrative to the extent that parents reference norms, standards, and values in their dialogic interactions. In this way, parents help children identify morally relevant features of their experience and encourage the formation of social-cognitive schemas that are chronically accessible. One attraction of this model is that it provides direct developmental lineage between the autobiographical narratives of concern to adult moral identity with the emergence of autobiographical self-narratives in early childhood.Krettenauer and Hertz make a real contribution to the ongoing articulation of moral identity theory. Every theory can use a little help, and their appeal to self-determination theory, to the orthogenesis of self-concept development, and to life-course personal narratives are useful emendations. I hope it leads to productive new lines of research. But the at the of their that it is all by an for a developmental theory, is The personological approach that is requires developmental specification. It to show how developmental in childhood the of moral identity in adolescence and and it seems to do I think the social-cognitive approach that I is an It is deeply integrative with personality theory, it is a robust predictor of behavior, it accounts for contextual and and there is a plausible developmental that us to adult forms of moral self-identity development. Although there are for moral identity theory to the three lines of research by the authors, with the of developmental research is not one of