The Promise and Perils of Prototyping: Can We (and Should We) Define Character Education? Free

The general idea of McGrath’s proposal sounds useful to me, but I am a virtue theorist and researcher rather than a character educator or program evaluator. So, the first question that I have regards how widely the concern about domain specification is shared by the people involved in CEPs, including character educators, effectiveness researchers, educational administrators, and policy makers, among others. If domain specification is not a widely held worry, then we can stop and thank McGrath for a well-articulated, but unnecessary proposal. My guess is that many people will see this as an important issue, but this ought to be demonstrated through a survey or a vote or some means of establishing a widespread perception of the need for greater definition of the field. Otherwise this proposal becomes another academic exercise that does not go beyond one person’s assessment, even if that evaluation is entirely correct.

The second form of participation would be for the character education community to endorse a proposal for addressing domain boundary specification. McGrath has given us a good rationale for approaching it through prototyping. Doing so effectively will require buy-in from the relevant stakeholders. Field definition is only worthwhile to the extent that it is shared by those inhabiting the field. Short of that, statements about what does or does not belong in a domain are only pronouncements that will, at best, cultivate camps within a

discipline. In addition, prototyping requires more than general agreement about its potential usefulness. It should also include developing a consensus about its contents.

I used the gerund form in the section title because consensus is more a process than a result, and McGrath did recognize that “the development of a technical prototype is an evolving process that is open to debate” (p. 33). Yet he did not describe how that process could unfold. We can, of course, reach consensus on a question at a given point in time, but consensus at any given point in time needs to be nurtured, revisited, revised, reconsidered, and re-established because domains of effort change, people come and go, and needs evolve. I suggest that, in working toward consensus, it is important to start at the beginning. In this case, that means gathering input on the crucial features of CEPs. McGrath’s proposal seems like a reasonable starting point, but it should be vetted by an expert panel. To what degree does a representative sample of the best minds in the character education community agree with his formulation? How might they want to amend it? How many features should this prototyping process have? What should those features be? There are many questions to be answered.

I recognize that this is no simple task, and there is no guarantee that anything like consensus can be reached. Moreover, even with a basic consensus, differences of opinion and priority will inevitably continue. But without a reasonably consensual approach to prototyping, McGrath’s proposal cannot rise above being an idiosyncratic suggestion. If, however, a respected group of leaders can agree, in the main, about what defines CEPs as CEPs, then there is hope that a prototyping process may accomplish the aims for which it was designed.

It seems to me that consensus building would take an iterative form, with an initial proposal, such as the one McGrath provides, followed by rounds of feedback and revision until the expert panel can endorse the proposal. A valid prototyping process would continue to be reviewed and revised throughout its useful life to remain relevant and informative as conditions change. It is also important, however, for this review and revision process to be guided by more than opinion and practical concerns. In my view, it is also vital that it be guided by good virtue theory and research (Cokelet & Fowers,2018; Fowers, Carroll, Leonhardt, & Cokelet, 2018).

McGrath refers to virtue theory twice in his article, including a differentiation of character and skills development, and a comment about “essential elements of character” (p. 15). He states that theory can enhance CEPs, and, I assume he believes that theory can enhance the development of a prototype model as well. I certainly believe that theory is important for both and that there is far more theoretical work to be done in both programing and prototyping endeavors. It seems to me that prototyping is one of the most important forms of theorizing because the aim of prototyping is to set a conceptual definition on an entire field. If that is true, then the theory guiding the prototype model must be made explicit and it must be vetted by experts.

Virtue theory is a very large topic, with many books and articles devoted to it, so I can only make a few broad recommendations about some salient points here. It seems to me that there is a set of essential concepts that should inform prototyping. McGrath does refer to my top priority of situating discussions of virtue within discussions of human flourishing, as all of the ancient Greek philosophers— the most prominent sources of virtue theory— did. But he mentions flourishing very briefly, and he intimates tension between individual and communal flourishing. The relations between the individual and communal good is a vital theoretical question that requires more attention. There are some useful resources that provide helpful ways to conceive the relationship between individual and communal goods (e.g., Fowers, 2005, 2012; Marujo & Neto, 2017), so we need not start from scratch on this issue.

McGrath also incorporates my second theoretical priority by including training in practical wisdom as one of the elements of his prototype, although he was more tentative than I would be on this point. The vital role of practical wisdom is often overlooked by scholars and educators interested in character, but it is widely understood as central among virtue theorists (e.g., Fowers, 2003; Kristjânsson, 2015; Schwartz & Sharpe, 2006). McGrath also recognizes the important point that practical wisdom is not simply a virtue like any other, avoiding the mistake that Peterson and Selig-man (2004) made in seeing wisdom as no more central to character than courage or temperance. Rather, practical wisdom is the capacity to decide which virtues are relevant to a given situation, what actions would constitute those virtues in the specific circumstances, and in prioritizing virtues in one’s actions.

Third, character development is seen by most virtue theorists as a process of habitua-tion, to which McGrath refers as the development of habits, and elaborates with the prototypical feature of identity. Although his definition of character development in terms of developing the whole person in a narrative identity is useful, narrative identity is a formal, and therefore, contentless definition. After all, there are many kinds of narrative identity. Contrary to McGrath’s belief that narrative identity differentiates character habituation and skills development, there is no reason that a narrative identity could not be formed around an individual’s development as a highly skilled person who has progressed from being a novice to an expert through a series of experiences and training and now identifies as a master of a domain (any domain at all). An even more telling example of the lack of content specification of the concept of narrative identities is that they can be formed for the development of vicious persons as well. Indeed, one of the most interesting questions about vicious character is how it develops in a life. McGrath may have been better served by including something more specific, such as the concept of a moral identity, which is defined by a commitment to acting morally (e.g., Blasi, 1994). The important point is that the identity ideal for character development that he or any other character education specialist must be spelled out. To his credit, McGrath indicates that this identity will include “positive psychological attributes” and “moral growth” that will contribute to “social good” or “personal good” without harming others. The trouble is that the meaning of these terms are not obvious and are in need of definition and elaboration. There are many ways to conceptualize each of these terms, and some versions will be easier to endorse and more useful than others. There are and should be many voices in these core definitional questions, which means that there will be variations in identity ideals across CEPs. From a virtue theory perspective, an identity ideal should be recognizable as a vision of a distinctly good human being, with the meaning of goodness spelled out in reasonable detail. There are different ways that this can be done, but we would expect a clear family resemblance among these character ideals.

The concept of habituation raises a fourth theoretical issue on which many versions of character theory and education falter. There is a great deal of interest in the concept of self-control among many scholars and educators (e.g., Baumeister & Exline, 2000; Peterson & Seligman, 2004). This is understandable because self-control is often crucial to acting properly for imperfect creatures. Yet self-control is neither a virtue nor a goal in most versions of virtue theory (Fowers, 2008; Woodruff, 2001). The reason for this is that the goal of character habituation is to become the kind of person who wants to act in the best way. Having a virtue means that an individual is characteristically motivated to act according to that virtue when it is appropriate to do so. Self-control is generally unnecessary to the virtuous because engaging self-control means that one has fallen short of the proper motivation attendant to virtue. Self-control can be seen as an essential element along the way to the development of a virtuous character because acting as if one has a virtue is the way to become a person with that virtue. Self-control is a core feature of continence, a character state wherein one acts well despite contrary desires. Continent action is, of course, more desirable than acting poorly, and it can be seen as a step toward developing a virtuous character, but self-control is not an endpoint. McGrath cites his own work and that of others endorsing self-control as a virtue (McGrath, 2015; Park, Tsukayama, Goodwin, Patrick, & Duckworth, 2016), but this falls short of the ideal of an individual who identifies with virtue sufficiently to be characteristically and spontaneously motivated to act virtuously. It is important to recognize that the inclusion of self-control as a factor in McGrath’s factor analytic results is an artifact of the inclusion of self-control in the Values in Action Inventory of Strengths (VIA-IS) rather than a demonstration of its importance for character.

McGrath advocates a whole-person approach to character education, which is consistent with most virtue theorists’ views. Yet this consensus is not very deep because some virtue theorists hold a strong unity of the virtues position (e.g., Socrates, Aristotle, Wolf, 2007), which suggests that having one virtue means having all the virtues. Others hold a weak unity of the virtues view (e.g., Adams, 2006; Sherman, 1989), which means, roughly, that having one virtue makes having others more likely. Others hold that there is no unity of the virtues because a person can hold one or a few virtues independently (e.g., Peterson & Seligman, 2004) or even that virtues are highly localized to specific kinds of situations (Miller, 2013). CEPs and a prototype for them merit more in-depth discussions of the unity of virtue question, because it is central to any concept of character.

McGrath notes that “efforts at an integrated character education program can be enhanced by theory about the essential elements of character” (p. 15). He then cites his factor analytic results (McGrath, 2015) as a theoretical resource indicating a possible three-virtue structure for CEPs. Clearly, factor analysis can be one source of data to inform theory, but it is not theory. In addition, the three-factor result he cites emerged from self-report responses to the VIA-IS and is therefore only applicable to the extent that one conceptualizes character based on that measure. Few virtue theorists would accept the VIA-IS as a paradigmatic model of virtue, and most virtue lists differ significantly from both the original VIA-IS model and from McGrath’s results. Of course, there is no consensus list of virtues (and there never has been a single list). There is not even a consensus view on how such a list is best formulated or on what sort of research could establish agreement about a particular list. This means that those who design, implement, and study character education need to formulate the set of character strengths they foster in a careful and informed manner, but without the assumption that they are working from a canonical list. Rather than debating the contents of a virtue list, most virtue theorists focus more on discussing how and in what contexts a character strength contributes to a good life.

This short and far from exhaustive list of theoretical issues indicates that there are vital theoretical questions attendant to creating a prototype for CEPs. Other topics, such as the skill/character strength distinction McGrath discusses, the place of vice in virtue theory, whether it is possible to “overuse or underuse” virtues, and components of virtues (knowledge, behavior, motivation, and disposition) also merit theoretical exploration on the way to a useful prototype. This discussion indicates that the role of theory in McGrath’s prototype proposal has been underplayed. A proposed prototype meant to guide the evaluation of many programs requires a stronger conceptual basis to justify the judgments that are made within it. A prototype that is used to evaluate programs should also have empirical justification.

In addition to participation, consensus building, and theory incorporation, the potential importance of a prototype model necessitates a thorough empirical assessment of its validity. The participation of stakeholders and the iterative development of a prototype model would form the initial stages of this validity assessment by assessing its content validity. Once the prototype model is formulated, it should be further evaluated before it is used for any consequential appraisals of CEPs.

The first question is whether individuals using the prototype model can conduct the ratings in a reliable way. What kind of training or documentation is necessary to facilitate reliable ratings? Once there is evidence for the reliability of the ratings, it is necessary to investigate whether the prototype ratings function in the intended manner. At a basic level, can the ratings be used to discriminate between bona fide CEPs and other educational programs designed to increase prosociality? Does the prototype model perform this function better than a categorization scheme based on a reasonable set of essential features? This question would be particularly important in borderline cases, wherein the prototype model should be superior to one based on essential features.

Another way to look at prototype validity is to ask whether the prototyping process is beneficial to people designing, implementing, revising, and studying CEPs. Will a single prototype model serve all these needs or do these users require different versions of it? Can people designing or revising a program use the prototype model as a guideline? Does the prototype model discriminate helpfully as an inclusion criterion for studies of CEPs? What is an appropriate demarcation for discriminating CEPs from other programs that promote prosociality? Although such discriminations would not be absolute because the underlying understanding of prototyping is that there will be resemblances among the full spectrum of prosociality programs, a stated purpose of the prototype model was to be able to make defensible discriminations among programs. McGrath exemplifies this with two hypothetical programs that both obtain a rating of 4/7, but that is just a hypothetical. What is needed is a good rationale and supportive evidence that the number that is used for demarcation has the meaning that is attributed to it. Although interpretation of ratings at the extremes will be simple, users will need to know how to interpret the results of prototype scores in the middle of the range. Clarity on interpreting middling ratings is vital because that is the real point of a prototype model. Discriminating fully prototypical CEPs from programs that do not have anything to do with character education does not require a prototype model.

My intention in these comments has been to outline the steps that I see as necessary for the successful development of a prototype. In my view, McGrath makes a good case for its potential value and he makes an intriguing beginning with his proposal. Yet I believe that a good prototyping model will require considerably more effort to include character education stakeholders and to develop a consensus about a prototype model that is theoretically and empirically defensible. I hope my comments are useful in the development of this potentially useful tool.