In the field of psychology, the construct of moral identity emerged as one response to the perceived gaps between moral judgment and action [Blasi, 1983, 1984]. As such, moral identity was conceptualized as the extent to which being a moral person is important to people's sense of who they are and who they want to be, and posited to act as an intervening psychological structure between moral thinking and conduct. In other words, it was thought that, for people for whom morality is central to and integrated within their identity, the desire to live in a manner consistent with their sense of self would serve as a key moral motivation [Aquino & Reed, 2002; Colby & Damon, 1992; Hardy & Carlo, 2005; Matsuba & Walker, 2005]. In the target article, Krettenauer and Hertz [this issue] place the construct of moral identity in its right intellectual context and briefly summarize the main findings of three decades of research: Measures of moral identity, framed largely in terms of the self-reported centrality a person places on possessing and enacting moral values, have repeatedly yielded findings of individual differences that, while theoretically consistent (inasmuch as moral identity centrality has been shown to be associated with readiness to engage in prosocial action and to abstain from antisocial behavior), have also been relatively small and not unique.
Krettenauer and Hertz, however, are not much concerned in this article with the staying power of the moral identity construct as a means for predicting individual differences in moral behavior. Their main concern, one that has also been articulated by other researchers in the field [e.g., Hardy & Carlo, 2011; Lapsley & Stey, 2014], relates to the paucity of findings regarding age differences in moral identity. Indeed, Krettenauer and Hertz are faced with a quandary. The most broadly used measures of moral identity have consistently and reliably yielded findings of no age differences. Yet, the authors assert, and we also agree, that moral identity must develop in some way. In an effort to address this quandary and advance the field, Krettenauer and Hertz draw on McAdams' three-level framework [McAdams, 1996] to identify three component processes of moral identity where developmental change is most likely to occur. Research into these component processes, they reason, might be more successful at revealing age differences, thereby illuminating the developmental trajectory of moral identity.
A focus on component processes takes the current “centrality” approach to conceptualizing and measuring moral identity as fundamentally sound. However, a lack of developmental change in moral identity conceptualized in terms of its relative centrality could also arise because that model has serious limitations that get in the way of understanding the development of moral identity (and also, to some extent, the nature and development of individual differences in moral identity). A number of significant critiques of the centrality approach to moral identity [e.g., Nucci, 2004; Pasupathi & Wainryb, 2010a; Proulx & Chandler, 2009; Turiel & Smetana, 1984] have been proposed, on various grounds. Furthermore, the field of identity research [e.g., Erikson, 1968; Marcia, Waterman, Matteson, Archer, & Orlofsky, 1993; McLean, Syed, Yoder, & Greenhoot, 2014] suggests important considerations in studying identity of any kind - considerations that are overlooked in the centrality approach. In the remainder of our commentary, we consider how one might build from some of the existing critiques and findings an alternative way of thinking about moral identity - one that allows for examining normative developmental changes as well as a broader range of individual differences and, ultimately, also the development of individual differences and specific forms of moral identity. The latter, though not the primary focus of this commentary, is nonetheless significant because we believe that the authors' interest can be understood as a specific question about the development of a particular form of moral identity.
What Is the Endpoint of Moral Identity Development?
To look at the normative development of anything, it is essential to understand what the outcomes of such development are for healthy, typical adults. As stated at the outset, the centrality approach to moral identity that Krettenauer and Hertz are building from conceptualizes moral identity in terms of the varying degrees to which people experience unity between their moral values and their personal goals. In fact, much of the existing literature on moral personality and moral motivation places this subsuming of self to morality as the ideal outcome of moral identity development [e.g., Colby & Damon, 1992; Frimer, Walker, Dunlop, Lee, & Riches, 2011; Walker & Frimer, 2007]. If a person feels that moral values such as being honest or generous are central for defining her identity, she is said to possess a strong moral identity; other outcomes are viewed as entailing a weak or diminished moral identity and, therefore, not the desired normative endpoint. And yet, this ideal according to which people subsume or sacrifice their personal goals to moral values, while being something that some individuals may aspire to, is not a feasible or likely endpoint for normative development, for several reasons.
Current understandings of the processes of identity development provide a first challenge to this proposition. As originally postulated by Erikson [1968], identity development entails processes of exploration and commitment that unfold in multiple domains and have multiple potential outcomes. Thus, as adolescents or adults engage in the process of exploring, for example, potential occupational identities, the outcome of this process is not more or less occupational identity, or even a stronger or weaker occupational identity, but rather different types of occupational identities. Research has shown that forms of identities, occupational and otherwise, differ significantly in terms of structural features and contents [e.g., Josselson, 2003; McLean et al., 2014]. Applied to the development of moral identity, this suggests that people will ultimately develop distinct types or forms of moral identity rather than merely more or less of it. While Eriksonian perspectives do not specify which forms those varied moral identities might take, moral philosophers as well as data from psychological research on morality offer guidance in this regard and, in so doing, further challenge the expectation of moral identity conceived only as the subsuming of personal goals to moral values.
In pondering right and wrong, moral philosophers working from diverse perspectives have underscored the pervasive and inevitable complexity of moral life [e.g., Appiah, 2006; Dworkin, 1986; Hampshire, 1983; Nussbaum, 2001; Scheffler, 1992; Wolf, 1982]. This complexity involves several layers, including the fact that, in ordinary moral experience, competing moral claims or principles sometimes collide, making it unclear not only what the right decision is but even that one right answer - an answer without moral costs - exists. Also, alongside their moral commitments, people also hold genuine and legitimate nonmoral concerns, concerns that involve personal goals as well as special responsibilities such as those created by love and friendship. Such concerns cannot be wholly neglected while still claiming to live a moral life, and they also cannot (and should not) be merely subsumed to moral concerns in every situation. Such irreducible complexity, then, makes the simplistic monism of the centrality approach to moral identity problematic in at least two ways. First, there are many occasions in which even “saints,” and certainly the vast majority of healthy adults, will have to do wrong because many moral decisions and actions will have real and inevitable moral costs. Second, an identity configured solely around moral values, with no room for the personal and the particular, has the potential to be of real harm and is fundamentally inconsistent with healthy development.
Moral philosophers' considerations of the complexity of moral life - and the complexity of what it means to be a good, moral, person - are borne out in the considerable psychological data available on morality. Social domain theorists have established that most people are concerned with justice and care for others, value those principles highly, and believe them to be prescriptive and not contingent on common practice or culture [for a review, see Smetana, Jambon, & Ball, 2014]. But research has also shown that when moral values come into conflict with one another, people negotiate between competing moral claims, inevitably compromising some of them [e.g., Helwig, Ruck, & Peterson-Badali, 2014]. Furthermore, research has shown that most people have many nonmoral considerations they hold dear, including pursuing deeply held personal goals [Nucci, 1996; see also McGregor & Little, 1998; Schindler, Staudinger, & Nesselroade, 2006]. Thus, normal adults are frequently engaged in crafting compromises between moral values and other considerations. Finally, research has shown that normal adults sometimes behave in ways that violate their moral principles, and oftentimes adults acknowledge this; in such occasions, people struggle to reconcile the contradictions between their actions and principles [Pasupathi & Wainryb, 2010a; Wainryb, 2011; Wainryb & Pasupathi, 2010]. Given these documented realities about the moral lives of normal, functional, healthy adults, the notion of “sainthood” or exemplar status as an endpoint - as the single, desirable endpoint - of moral development is not viable. Rather, most of us - the vast majority of us - live and function (and develop) in a moral muddle. It may thus be more productive to try to characterize that moral muddle within which typical adults construct and manage various shapes of moral identity in order to have clarity about what it is children and adolescents are developing towards.
How to Capture the Various Forms That Moral Identity Might Take in Adulthood?
What is known about the processes of identity development in general, along with the compelling philosophical and empirical arguments concerning the complexity of moral life, underscore the profound limitations of framing questions about individual differences in moral identity in terms of the self-reported centrality of moral values. As an alternative, we offer two nonmutually exclusive dimensions, firmly rooted in conceptual and empirical work, along which questions about individual differences in moral identity may be more productively framed.
One dimension refers not to the self-reported centrality with which people rate moral values but to the extent to which moral values are prioritized in the context of multiple competing moral and nonmoral concerns and the extent to which moral values inform individuals' responses to (real or hypothetical) events. This way of asking about variation in forms of moral identity is actually also likely to capture variation in self-reported centrality, consistent with findings that moral exemplars' narratives prioritize moral themes [e.g., Colby & Damon, 1992; Matsuba & Walker, 2005; Walker & Frimer, 2007], and may in fact allow for more variability in this regard than do self-report scales, in part because it comes closer to tapping into the real-world tradeoffs that people make between varying priorities and considerations. However, such an approach also allows for variability in moral identities that goes beyond centrality.
For example, this approach may also unveil individual differences in the nature of the particular events that catalyze a strong investment in moral concerns [Matsuba, Pratt, Norris, Mohle, Alisat, & McAdams, 2012]. It may also allow us to observe individual differences in the breadth or narrowness of what people view as relevant targets for moral concerns [e.g., Kahn, Friedman, Perez-Granados, & Freier, 2006], as well as in the ways they manage tradeoffs between moral and nonmoral concerns [e.g., Smetana et al., 2014].
We wish to underscore, however, that in looking at individual differences in moral identity in this way, three issues must be kept in mind. First, having moral concerns dominate one's life story constitutes one kind of moral identity, but people who do not saturate their life stories with exclusively moral concerns also have a moral identity. Second, having moral concerns dominate one's life story is not the only developmental endpoint - it is merely one way in which some individuals come to tell their life story. Third, it is problematic, in our view, to value a life story saturated with moral concerns over other types of life stories. Such strivings for martyrdom may not improve the world in measureable ways [e.g., Appiah, 2006; Wolf, 1982], can compromise our capacity to meet special obligations to those closest to us [e.g., Pasupathi, 2014], and, ultimately, can undermine healthy developmental needs [e.g., Nucci, 1996; Smetana, 2011].
The other dimension along which questions about individual differences in moral identity may be fruitfully posed relates to the various ways in which people manage the common experience of doing the morally wrong thing even though they know it was wrong. Although individual differences are likely in how people grapple with the contradictions created by their own moral wrongdoing, to date there is little empirical evidence about the forms such grappling might take. We have begun to examine possible continuities in the way people construct their own agency around experiences of having harmed another person, asking whether some people are consistently more likely to, for example, view themselves as having acted badly due to forces beyond their control whereas others are consistently more likely to hold themselves responsible. Our data also suggest that some individuals tell elaborated, coherent, and psychologically framed stories about their own wrongdoing while others tell more fragmented and sparse accounts that leave out the explanations and meanings that make their own wrongdoing make sense [Pasupathi & Wainryb, 2010a; Wainryb, 2011; Wainryb, Komolova, & Florsheim, 2010; Wainryb & Pasupathi, 2010]. We have speculated about the implications of these differences for both moral agency [Pasupathi & Wainryb, 2010a; Recchia & Wainryb, 2014; Wainryb, 2011; Wainryb & Pasupathi, 2010] and moral identity [Pasupathi & Wainryb, 2010a; Wainryb & Recchia, 2015], but testing those speculations is an ongoing project.
How Does This Help Us to Understand the Development of Moral Identity?
Our broader approach to conceptualizing moral identity allows more variability in adult forms of moral identity and is firmly connected to empirical findings. Importantly, it also does not exclude the majority of healthy adults from having a moral identity. More relevant to the concerns of Krettenauer and Hertz, it conceives of moral identity in a way that is amenable to study in its younger and developmentally earlier forms.
Competing Values and Moral Identity Development
The approach focused on self-reported centrality of moral values has left young children entirely outside the realm of experimental examination and has also failed at uncovering reliable age differences in cross-sectional and longitudinal samples of adolescents and adults. We suggest that an approach focused on exploring how people negotiate and make sense of their own multifaceted experiences - those that encompass competing moral goals or competing moral and nonmoral goals - is bound to uncover many significant age differences, thereby illuminating some aspects of the developmental trajectory of moral identity. Although the life story in its full form is a developmental achievement of late adolescence and early adulthood [Habermas & Bluck, 2000], research on children's autobiographical narration has shown that children as young as 5 can construct meaningful accounts of their experiences [Fivush & Nelson, 2004; Pasupathi & Wainryb, 2010b; Reese, Yan, Jack, & Hayne, 2010]. Moreover, work in our laboratory has shown more specifically that, starting at least at the age of 5, children can construct fairly coherent narrative accounts of their own morally-laden experiences [Recchia, Wainryb, Bourne, & Pasupathi, 2015; Recchia, Wainryb, & Pasupathi, 2013; Wainryb, Brehl, & Matwin, 2005; Wainryb, Komolova, & Brehl, 2014].
Examining the extent to which life stories and “smaller stories” are imbued with moral themes at different ages represents one avenue for studying developmental changes in moral identity, as does exploring children's versus adolescents' relative privileging of morally relevant events and actions over other kinds of experiences. Another promising avenue entails exploring the ways children, versus adolescents, link morally relevant experiences with their sense of self. Krettenauer and Hertz made particular note of self-event connections, which become increasingly prevalent in narratives across adolescence and early adulthood. Importantly, self-event connections, regardless of their outcome (i.e., whether they construct a sense of growth from an event or dismiss the event as irrelevant to the self), entail a process of grappling with the identity implications of experiences [Pasupathi, Mansour, & Brubaker, 2007]. Further, such connections also rest on earlier developmental precursors, including the elaboration and exploration of psychological features of experiences [Pasupathi & Wainryb, 2010b]. Ongoing work in our laboratory has begun examining these issues using several hundreds of narrative accounts that children between the ages of 5 and 18 made about their own morally-laden experiences.
Finally, and extending our earlier suggestions, narrative approaches also permit looking at how specific patterns of moral investment in some domains rather than others, or how more and less expansive targets of moral regard, might develop from childhood through adulthood. Although research suggests that young people often imbue conversations and disagreements that bear on nonmoral topics (e.g., science, chores) with moral concerns [e.g., Callanan, Valle, Luce, & Rigney, 2014; Smetana, 2011], a careful developmental exploration of these issues with an eye towards understanding changes in moral identity has not yet been done.
Managing Moral Wrongdoing and Moral Identity Development
The second dimension along which we suggested conceptualizing moral identity involved examining the various ways in which people manage the contradictions between their moral values and their occasionally hurtful or unfair behavior. This approach is also amenable to developmental exploration.
Our own work reveals that, across childhood and adolescence, children begin to think about their own wrongdoing with increasing sophistication and complexity [Recchia et al., 2013, 2015; Wainryb et al., 2005; Wainryb & Recchia, 2014]. That complexity includes an increasing emphasis on the psychological aspects of experiences [Pasupathi & Wainryb, 2010b] such as the goals, beliefs, and emotions that undergird their actions and help them make sense of them. It also includes an increasing capacity to link experiences to broader conclusions about self and identity [Pasupathi & Wainryb, 2010a; Recchia & Wainryb, 2014; Recchia, Wainryb, Bourne, & Pasupathi, 2014; Wainryb & Recchia, 2015]. These developmental shifts permit adolescents to hold an increasingly more complex and multifaceted moral identity, one in which they remain good, though imperfect, people. But our research also shows that even young children are grappling with these issues (e.g., with the ways in which their harm of others was their fault or not their fault), and are challenged to build accounts for their wrongdoing [Bourne, Wainryb, & Pasupathi, 2015; Recchia & Wainryb, 2014; Wainryb et al., 2005, 2014]. Those accounts are the foundation for increasingly mature forms of moral identity.
The Development of Individual Differences in Forms of Moral Identity
Questions concerning the emergence and development of individual differences are distinct from questions about normative developmental shifts. However, normative development can inform the way individual differences are assessed and interpreted. Indeed, individual differences can be most meaningfully understood in the context of normative developmental trajectories. While this topic is beyond our present scope, we suggest that conceiving of moral identity in terms of how people manage both the tradeoffs between competing concerns and their own moral wrongdoing also allows for examining the emergence and development of individual differences along these dimensions. Of course, one could attribute these individual differences in moral identity to inborn traits or temperament and treat them as not developmental. However, it is likely that many of the differences we discussed above emerge and develop over time and in the course of childhood and adolescent experiences and are constrained by the co-occurring developmental changes. To make sense of the development of such individual differences, one may draw on broad conceptual models that ground the emergence of individual differences on processes such as canalization [Waddington, 1971], developmental pathways [Bowlby, 1980], equifinality and multifinality [Cicchetti & Rogosch, 1996], and differential susceptibility [Belsky, 2005]. In addition, work more closely related to the moral domain can also inform these processes. For example, there are emerging data from narrative and conversational research in our and others' labs that suggest maternal socialization processes as one route to individual differences in the ways children manage their own moral wrongdoing [Laible & Thompson, 2002; Recchia & Wainryb, 2014; Reese, Bird, & Tripp, 2007; Wainryb & Recchia, 2014; Wang, Leichtman, & Davies, 2000]. In other work, researchers have begun asking about the kinds of experiences that might be explicitly connected to attunement to the moral concerns embedded in particular domains, such as environmentalism [Matsuba et al., 2012]. Furthermore, one could also ask about the kinds of experiences, reflections, maternal and other socialization processes, and broader sociocultural factors that lead some individuals to view moral concerns as so identity-central that conflicts between the personal and the moral are attenuated [e.g., Colby & Damon, 1992; Frimer et al., 2011].
Concluding Remarks on Saints and the Rest of Us
Taking the above arguments into account, let us return to the target article. Krettenauer and Hertz, rightly in our view, suggest that to see moral identity development unfold will require a broader lens. We take this further and argue that a broader definition of moral identity is what is needed. We outlined some ways in which moral identity might be more broadly conceptualized but want to note that our approach incorporates the concerns of those researchers interested in moral identity defined in terms of value centrality. One way to think about the target article is this: Krettenauer and Hertz want to understand the development of a particular form of moral identity - one in which personal concerns and moral values are so intertwined that doing the right thing simultaneously addresses personal goals. We argue for the importance of also understanding the development of varied forms of moral identity, forms that are more commensurate with the experience of the majority, rather than limited to that of moral exemplars.