NEW CHALLENGES TO CHARACTER EDUCATION

The recent growth of character education worldwide has met with a number of criticisms. This article applies a humanities methodology to investigate three of these criticisms in the belief that a growing movement can only benefit from attention to its critics. The first criticism is that character education depends on flawed or unreliable social science methods. In response, the article recommends more focused attention on the philosophical foundations of empirical research on character. The second criticism is that character education leaves unjust systems unchallenged. In response, the article recommends increased academic dialogue between character education discourse and social justice discourse. The third criticism is that character education violates the purpose of a university. In response, the article denies the objection, appealing to historical and philosophical sources to argue that character has been and should be at the heart of all Higher Education.


Introduction
Research and teaching on character and virtue has been growing over the last two decades. Many new institutes, courses, and research centres have sprung up around the world. 1 Every growing and successful movement attracts criticism, and every healthy and respectable movement engages that criticism and seeks to learn from it. This has already been done once for character education. Thirteen years ago Debating Moral Education was published, an edited volume containing both critiques and defences of character education . But since then, many more critical voices have been raised, some of which directly respond to arguments made in the aforementioned book. A survey of these criticisms shows three to be the most frequent and prominent. This article examines each of these criticisms in programmatic fashion. It does not pretend to have a comprehensive or satisfying response to any of them. Its purpose is to map out what kind of research would be needed to provide an adequate response. I will propose directions for future study in character education by attention to the points of greatest vulnerability. This will enable character education to adapt and grow into new forms that overcome its weaknesses.

First Criticism: Research on Character Depends on Faulty Empirical Methods
In order to prove its effectiveness, character education is frequently observed and measured by social scientific studies, which evaluate whether they are having an impact on the character growth of students. Two attacks have been levelled against the social scientific research on character, from opposite angles. Some criticise the methodologies of the research as unreliable; others take the exact opposite route, and use empirical research to refute the effectiveness of character education. By juxtaposing these two critiques I wish to show the importance of assessing the broader role of empirical studies in character theory. Let us begin with the latter critique.
As is well known, in the medieval period the Aristotelian ethical framework, modified by the Christian tradition, was pretty much the only framework available. The idea that the goal of life is the pursuit of virtue, and that virtue is best pursued by practice and cultivation of habits, was commonplace. But when and why did that change?
It changed with Martin Luther, one of the most significant political, theological and philosophical figures of the modern era, whose ground-breaking ideas gave birth to the Protestant Reformation. Luther rejected Aristotelian virtue ethics wholesale, along with the its later Catholic developments, seeing the whole idea of virtue as a misguided attempt to achieve righteousness by one's own efforts that can only be achieved by faith in Christ. Luther wrote, for example, that "the righteousness of God is not acquired by means of acts frequently repeated, as Aristotle taught" (Cited in Zahl, 2019, p. 201).
Luther can be considered the greatest and most influential opponent of virtue and character in the Western world. Luther's arguments are of course not new, but they are being used in a new way to attack the resurgence of virtue ethics in the modern university. According to Simeon Zahl, Professor at the University of Cambridge, "one of the strongest arguments behind the early protestant protest against virtue-based paradigms … has not yet been refuted, and the recent theological revival of virtue ethics is on less solid ground than it appears" (Zahl, 2019, p. 222).
Zahl's basic argument is that the Lutheran critique of virtue ethics is not a philosophical critique but an empirical one, and that this makes it far more powerful than is commonly recognised (Zahl, 2019, p. 205). Luther did not reason his way into rejecting virtue ethics as an inadequate paradigm, he experienced his way into it. Zahl writes that "Luther supports his views about the bondage of the will and the inability of human beings to fulfil the Law through arguments explicitly derived from his personal experience of the insuperability of sinful affections and desires: Years of ascetic practice, community life, and prayer, under ideal early modern conditions for the production of virtue -a scrupulous Augustinian cloister -simply had no substantial effect in diminishing sinful desire that he could discern" (Zahl, 2019, p. 212). Luther believed he was not virtuous, in spite of years of trying. Because virtue ethics did not work for him, he abandoned it.
Zahl's argument is solely rooted in Luther's own self-assessment. Zahl does not engage any contemporary empirical studies of virtue and the efficacy of character education to see whether their results com-

FEATURE ARTICLE
pare with Luther's. Zahl's method raises a number of weighty questions. Should other opinions be founded on Luther's experience? What makes Luther's experience more definitive than the experience of any human being alive today? What happens when two experiences suggest contradictory theses? None of these questions can be addressed here. The question that matters for the purposes of character education is the following: how much weight should we place on empirical studies to demonstrate the fruitfulness of character education?
If we think that Luther's self-assessment was unreliable, then we have to be consistent and ask about the reliability of contemporary self-assessments as well. The problem is that most empirical studies of character are based on the exact same measure that Luther used: self-assessment of the subjects of study. As Kristján Kristjánsson notes: The great majority of existing instruments to measure character [are] simple self-report questionnaires.
[There are] possible response biases in such measures caused by self-deceptions and self-fabulations. Even if I consistently think I am a duck, this does not make me a duck. (Kristjánsson, 2013, p. 283) Similarly, Joshua Hordern, Professor at Oxford University, worries that a self-assessment of character is a fallible foundation on which to make moral reasoning depend. He writes that the "endemic human tendency to overestimate, underestimate, or misunderstand entirely one's own moral state' means that 'one's own character is normally little known to oneself " (Hordern, 2012, p. 102). Our ability to self-assess accurately is itself dependent on certain virtues, like honesty, humility, self-awareness, and someone who lacks these may go dramatically wrong in their self-assessment, yet how do we determine whether they lack these virtues or not if self-assessment is what we're relying on?
Hordern also worries that the use of virtue in moral decision-making can insulate an individual or a community from critique, since it obscures the rationale for a course of action, making it instead a function of (purportedly) virtuous person's intuitive sense about what is right. He writes, "making one's moral understanding dependent on one's virtue of character habituated by one's society and a highly specified account of eudaimonia may foreclose the possibility of experiencing moral correction" (Hordern, 2012, p. 102). In other words, an entire community could become a self-congratulatory echo chamber, where everyone affirms everyone else's virtuousness and nobody realises that some key virtues are being entirely missed. To be sure, using 360 degrees of assessment will help significantly, but it doesn't solve the underlying problem, which is the absence of an objective benchmark for character that the social sciences can base their studies on. As Kristjánsson points out, we "may suggest triangulation via reports of peers and significant others (teachers, parents, siblings, etc.), but the snag is now that even if not only I think I am a duck but other people too, this still does not make me a duck" (Kristjánsson, 2013, p. 283).
This dependency on subjective measures, which are affected by the character of the subjects in an epistemological circle, makes character education vulnerable to the changing tides of social scientific studies. To change the metaphor, it makes character education into a house built on sand. There are as many empirical studies that deny the effectiveness of character education as ones that affirm it. The situationist critique of character also uses empirical studies (see, e.g., Doris, 2002;Harman, 1999).
What needs to be explored is a robust account of the proper role of empirical study in the broader efforts to cultivate virtue and character. We need empirical data, but these empirical data are always interpreted by a philosophical framework that evaluates its significance. Empiricism is itself a philosophical position about how we get knowledge, a position that reaches its full flourishing in the natural sciences. As Daniel Little points out, "if social scientists are captivated by the scientific prestige of positivism and quantitative social science, they will be led to social science research that looks quite different from what would result from a view that emphasizes contingency and causal mechanisms" (Little, 2009, p. 175).
It is a mistake to think of science as a body of knowledge. Science is actually a mode of enquiry, and all knowledge it claims to have is always provisional, open to being overturned by the next laboratory experiment. In order to respond properly to Zahl, Hordern, and Kristjánsson, we need not only more methodologically robust empirical studies, we need a stronger account of the philosophical position that is affected by, but not wholly dependent on, whatever social scientific research currently exists.
The questions that need answering in order to provide a robust response to critics of character education theory are as follows. Firstly, what are the philosophical presuppositions behind empirical research on character? Secondly, what is the rightful role of empirical data in measuring character development? Thirdly, how should empirical data be interpreted by its broader theoretical framework? To answer these questions requires engaging with the literature on the underlying philosophical presuppositions behind social science research. 2 Attention to the philosophical theory behind both virtue ethics and social science can help show how they relate and what role social science can play in forming a strong foundation for character education.

Second Criticism: Character Education Leaves Oppressive and Unjust Systems Unchallenged
In the aforementioned 2010 volume, Debating Moral Education, Romand Coles writes the following strident criticism of character education. In his view, its proponents: tend to focus on renderings of honesty, courage, character, respect, fairness, generosity, and so forth that are framed as if they could be achieved without doing much at all to question and change the basic parameters of our political economic relationships, practices, and the associated theodicy of history that has bound our ethical-political imaginations. It is as though, if we just walk our paths with moral rectitude and perhaps a little bit of tinkering at the edges of things-that might be enough. But what if many of these sanctioned paths are directly corrupt for the evil they do, or indirectly corrupt for the responsibilities they deny? (Coles, 2010, p. 228) Coles is worried that character education can silently legitimate structural injustices by drawing attention away from them, implicitly suggesting that the structures are not really the problem. Coles' objection received no response from other contributors to the volume and remains unanswered in today's literature on character education.
In 2018, a special issue of Sociological Research Online appeared that was devoted to attacks on UK character education initiatives. Spearheaded by Bull and Allen, it offered a series of scathing critiques about the vulnerability of character education to manipulation by a rightwing political agenda. They expressed concern that a focus on character places an undue burden on oppressed individuals to change their character instead of changing the system that oppresses them: By occluding the social context, individualised character 'traits' become located as the primary cause of social mobility or 'success' in life, when in fact, they instead provide a means for the FEATURE ARTICLE rationalisation or justification of unequal outcomes. Such a focus on individualised attributes is particularly insidious because it outlaws political anger at structural inequities and injustices, focusing it inwards instead. (Bull & Allen, 2018, p. 396) They see this as evidence of an "implicit theory of capitalism visible in character and resilience discourses, which requires individuals to uphold morality because the economic system cannot" (Bull & Allen, 2018, p. 396). Similarly, another contributor, Nick Taylor, writes that "emphasis on individual virtues, psychological traits, or skills as markers of success … risks perpetuating existing discourses that individualise responsibility for a highly unequal society and economy" (Taylor, 2018, p. 403).
There is much in these criticisms that misses the mark. For example, character theory is not individualistic, but gives a social dimension to the very essence of virtue. Cameron, Bright, and Caza, aware that virtue theory is frequently associated with political conservatism, draw on ancient discussions in order to affirm that virtue is "synonymous with the internalization of moral rules that produce social harmony" (Cameron et al., 2004, p. 767).
Character theory is also not unaware of the impact of social systems on the development of virtue. Research at the Oxford Character Project has produced seven strategies of character development, one of which includes awareness of the ways in which systems and structures impact virtue capabilities (Lamb, Brant, et al., 2022).
Nonetheless, the way society and systems are referenced in these approaches is insufficient to answer this objection, because the focus is still on the ways in which character is shaped, or character growth inhibited, by such systems, not on the social inequalities embedded into the system. Therefore, I suggest that proponents of character education take seriously this critique as some-thing that does not undermine the value or importance of training in virtue, but that might point out ways to strengthen and improve it.
It is a remarkable fact that academic debates on character and virtue have very little overlap with social justice discourse, even though both inhabit the broader field of ethics. This absence of dialogue between the two may be partly explained by different focal points: one orients its discussions around the transformation of personal qualities, the other around the transformation of political structures and social contexts. If the broader goal of both is human flourishing and well-being, then we can see how both start from opposite ends of the spectrum in pursuing that goal. But the two focal points are inextricably intertwined: the character of individuals both shapes and is shaped by the systems they inhabit in circular fashion. This means that each of the two ethical foci is only impoverished by a lack of attention to the other. What is needed is a fuller picture of the conditions for human flourishing that takes into account both aspects of what is needed to engender change. This can take place at both the research and the pedagogical levels.
Firstly, research. Let us consider two examples of engagement between virtue ethics and social justice. James Hankins' Virtue Politics: Soulcraft and Statecraft in Renaissance Italy argues for a foundational connection between virtue and society in Western political thought (Hankins, 2019). The seminal political theorists of Renaissance Italy saw the goal of politics as the formation of virtuous citizens, and conversely considered virtuous citizens to be an essential factor in the maintaining of a just and peaceful society. This book could be seen as approaching virtue from the side of politics and society. Conversely, Lisa Tessman's Burdened Virtues: Virtue Ethics for Liberatory Struggles starts from the opposite end (Tessman, 2005). Tessman explores what Aristotelian virtue ethics can offer to the struggles for justice and equality in today's society, with a focus on gender equality. These contributions are necessarily limited by the scope of a single academic monograph.
The particularity of their foci shows how much more needs to be done to bring these two discourses together fruitfully.
Secondly, pedagogy. Those who engage in character education should make systems change an explicit focus, highlighting it as a desired consequence of the character formation of those in positions of power. They should also raise awareness of the privileged status that belongs by necessity to anyone with the leisure to receive character education. Programmes like the Oxford Character Project are offering character education for the privileged elite, those with the training, capability, and financial support to study at graduate level at Oxford. This should have an impact on the kind of virtues on which such a programme chooses to focus and on how those virtues are manifested in leadership. Educators of character need to be explicit about the ways in which character qualities can help transform unjust systems and create fairer and equitable structures for the benefit of all.
Character education discourse needs to show more awareness of the structural injustices and inequalities, which pervade our world -not just the way those systems inhibit character growth, but the way they inhibit all kinds of human flourishing including opportunity and provision of basic material needs. The problems we see in the world can be addressed from the top down and from the bottom up simultaneously, and we should not prioritise one approach over the other. The most powerful and effective social action will be the kind that harnesses the insights of both virtue ethics and social justice to address the problem at both the micro-individual level and the macro-structural level.

Third Criticism: Character Formation Violates the Purpose of the University
We now come to an objection that considers character education inappropriate for the university, however appropriate it might be for primary and secondary education contexts. David Carr of Birmingham University has three concerns about the principles behind character education. Firstly, he worries that teaching character to over-18-year-olds could be seen as an unwarranted restriction of their moral freedom to choose for themselves what is right and wrong. He writes: Up until a certain point, we have to make decisions for young people, but after that they may be leftare, indeed, entitled -to decide for themselves. But, by much the same token, what scope or justification might remain for formal moral education beyond generally accepted years of discretion? (Carr, 2017, p. 114) He concludes that "the key question about teaching moral virtues is not that of whether we are cultivating morally questionable qualities, but of how we might teach them is a way that does not undermine or inhibit the freedom of choice that is a sine qua non of virtuous action or conduct" (Carr, 2017, p. 112).
A few qualifications to make sure this objection is understood. Carr is not arguing against the legitimacy of ethics as a topic of higher education. He recognises that an ethicist may teach ethics to students. But to avoid indoctrination, the ethicist is expected to teach ethics in an open-ended manner, giving students tools for enquiry rather than imposing the teacher's own answers on students as if they were the right answers. This means, necessarily, that the ethical debate is framed and assessed as a purely intellectual enquiry without reference to the actual decisions made by the students outside the classroom. While virtue and character could be and sometimes are taught in this way, proponents of character education typically argue that it ought not to be taught in this way. They frequently quote Aristotle's dictum that "we are inquiring not in order to know what virtue is, but in order to become good, since otherwise our inquiry would have been of FEATURE ARTICLE no use." 3 Character education is not about intellectual debates on the relative merits of virtue, or rather, it is only about them as a means to the end of becoming virtuous. Proponents of character education would consider a course on virtue ethics a failure if all it did was increase students' intellectual knowledge of virtue theory without transforming their character. But to aim at transforming their character in a particular way implies that the teacher is pushing a particular ideological stance on the students.
Carr has identified an unavoidable tension between freedom of opinion and learning objectives. The teacher of character formation cannot avoid having a set of character qualities whose development in the students is one of the goals of the course. If it were open to debate whether or not these qualities were desirable, the teacher could not make them a goal of the course, since the students might legitimately decide that such qualities were not desirable and thus choose not to develop them. The ethical value of a particular virtue cannot at once be open to disagreement and a development objective.
Carr's second objection to making character a learning objective is that it breaks down the separation between professional and private, since the teacher is aiming to change the way the student behaves everywhere, not just in class. It is one of the core definitions of a virtue that it is only a virtue if it is consistent across different spheres of life. This means that to cultivate virtue successfully in a student, they must become courageous, honest, empathetic, humble, etc. not only in the profession they are training for but also in their private lives as well. Yet while such interventions may be appropriate for children, nobody has the right to interfere with the private life of an adult. "Trainee doctors, lawyers or business executives, " Carr writes, "may reasonably protest that so long as they are conducting themselves justly or virtuously in professional role, it is no-one's business but theirs whether they conduct themselves virtuously in their private lives" (Carr, 2017, p. 117). A teacher may give students a failing grade in virtue ethics if they cheated on the exam. But he or she may not give them a failing grade because he or she saw them having a drunken brawl on the street a week before the exam. In short, if Higher Education is about professional training, then according to Carr, character education is not appropriate for it.
Thirdly, Carr objects that character education violates an older, more traditional goal of the university -the disinterested pursuit of knowledge. He contends that the goal of virtuous formation of university students undermines the university's "traditional commitment to the pursuit of knowledge for its own sake" (Carr, 2017). This means that "while it may be that [some] studies involve significant character-forming discipline, the adoption by teachers of any more explicit character-forming agenda might well be regarded as so much personally intrusive imposition" (Carr, 2017, p. 120). In short, whether the university is seen as a professional training school in the newer model, or as a sanctuary for the disinterested pursuit of knowledge, in both cases the goals and methods of character education are inappropriate and should not be used.
What follows will begin as a simple refutation of Carr's critique, before exploring what can be learned from it for the benefit of character education. We start with his third objection, that character formation violates the pursuit of knowledge for its own sake.
There are two difficulties with Carr's position: one historical and the other philosophical. From a historical perspective, the notion of the University as committed to disinterested knowledge is by no means 'traditional' . It arose as part of the Enlightenment and found its full flourishing in 19th-century Germany, with the ideal of Wissenschaft, or rigorous critical enquiry. This notion replaced the older and more traditional ideal of Bildung, or intellectual and moral formation. Several scholars have pointed out that the university was in fact in the business of character formation right from its origins, even if such formation was focused more on intellectual than on moral virtues. As Rowan Williams writes: "a closer look at the origins of the university might give us pause before we simply oppose modern pragmatism to ancient contemplation." For the premodern mindset, he tells us, the purpose of a university was "to create "public people" -people who, whatever their specialism, are committed not only to reasoned argument … but to a responsibility to the ideal of rational governance and rational public discourse." The university's role in society was to "nourish … honest and hopeful speech, for the sake of a properly reasonable culture and politics" (Williams, 2004) -in other words, the formation of intellectual virtue. Similarly, Nigel Biggar concludes a brief historical survey with these words: "Universities were never simply the child of an ivory-tower love of knowledge for knowledge's sake. They were always partly fuelled by practical concerns, whether the concerns of private individuals or of those with public responsibility" (Biggar, 2022, p. 99).
But there is also a philosophical difficulty with the idea that the university should pursue disinterested knowledge, which is that such a pursuit is impossible for human beings as we are.
Anyone who spends time reading the greatest philosophers of the twentieth century -people such as Merleau-Ponty, Heidegger, Gadamer, Wittgenstein, Ricoeur, Polanyi, among others -is met with the very strong unifying theme that there is no such thing as disinterested knowledge, no such thing as the pursuit of knowledge for its own sake, no such thing as neutral enquiry, no such thing as an objective point of view. For such thinkers, these treasured notions of the 19th century are a myth and must be exposed as such. All knowledge is interested, in the sense that it is formed in response to implicit questions that arise from a particu-lar cultural and historical situation. All knowledge is situated in the mind of a knowing subject, which 'knows' in a unique way due to his or her upbringing, history, and concerns. All intellectual pursuit is driven by the needs and interests of a particular cultural milieu, and has certain goals in mind. This is very easy to see if we read the supposedly neutral scholarship of an era other than our own -the particular interests, presuppositions, and perspective stands out to us as if it was written in blood (Gadamer, 1977). It is only the scholarship of one's own time and place that can ever give the illusion of seeking knowledge for its own sake.
Moreover, Carr would be unlikely to deny that even the "pursuit of knowledge for its own sake" (if such a thing exists) requires formation in order to be done properly. A scholar doing rigorous scholarly research needs to be possessed of the intellectual virtues -honesty, thoroughness, fairness to all sides of a debate, not presenting the work of others as if it was one's own, not twisting the evidence to fit prior opinions, etc. The same tension Carr identified between learning objectives and freedom of opinion applies to all scholarly training. One cannot penalise a student for plagiarism at the same time as keeping the question of whether or not plagiarism is wrong a matter of debate in the classroom. Teaching cannot but assume the truth of the standards by which students are measured, and this means that all teaching has at its core some principles that are not open for debate. Why, then, should these be limited to the intellectual virtues?
What the defenders of disinterested knowledge were trying to get at is better expressed through the notion of freedom of enquiry, the liberty to pursue any line of questioning without having to justify its immediate relevance, without dogmatic restrictions on what one is allowed to question or propose. This notion does rightly belong in character education because it is one of the intellectual virtues necessary for honest rational debate. Carr is right that students' freedom of choice in this area should be

FEATURE ARTICLE
protected. If a student of character education wants to ask whether honesty or humility or courage really are desirable virtues and not vices, or whether virtue can be acquired, or whether there is any point in pursuing virtue, the teacher should not shut down these questions as illegitimate. But even while the debate is going on, the student needs to be made aware that the course assumes that virtue is worth pursuing and that to question that is to question the very purpose of the course. This principle applies to all other kinds of teaching as well. A student may be free to argue that dishonesty, plagiarism, and sloppiness are appropriate forms of scholarship, but if they try to do scholarship in that way, they will fail the course. If Carr wants to call that educational compulsion, we must only reply that such compulsion is a condition for the possibility of any university education of any kind.
This last point also refutes Carr's first objection, that character formation should not be imposed on adults. All teaching imposes values, whether implicitly or explicitly. Because character education does so explicitly, it draws attention to this fact, but this does not make it unique or distinctive in any way. Carr's objection to character education in Universities falls apart, then, both because character formation is a more traditional goal than the pursuit of disinterested knowledge, and because insofar as the latter has any meaning, it includes character formation at least at the intellectual level.
Let us turn, finally, to Carr's second objection: that character formation violates the separation between professional and private. Insofar as he means that that teachers can't follow students around 24/7 and grade them on the virtuousness of their lifestyle, no proponent of character education is suggesting such a method. A student will always be able to do well in a character formation assessment while remaining unvirtuous. This unavoidable disjunct between the classroom and the rest of life only reveals the limits of a teacher's ability to formally measure and assess student learning. Such limits are already known and compensated for in the ordinary course of employment. Employers know that a student with top grades in medicine or law may for some other reason be unsuitable for a medical or legal job. That is why businesses hire people not only on the basis of their resumes or curriculum vitae (CV), but conduct interviews and ask for character references.
What, then, can we learn from Carr for the future of character education in Universities? Two points are worthy of note. First, we need to remember the limits of our ability to measure students' character. This point is reinforced by our first criticism about the limits of empirical study on character more broadly. Without any objective guarantee that character formation is having its desired effect, teachers are compelled to cultivate the virtue of hope in themselves. Hope is unnecessary where there is certainty. But without certainty, the need for hope becomes manifest. Teachers may reasonably hope that their teaching is having some effect outside the classroom. Secondly, we need to encourage critical debate about virtue itself in the classroom, instead of presenting a pre-decided framework as dogmatic and unquestionable. Students must always feel free to challenge the premises of the education they are receiving, and the teacher must always be open to changing his or her mind on the basis of something the student has said.

Conclusion
Those invested in character education can only benefit from paying attention to its critics. If in the worst case they are correct, and if this exposes a fatal flaw in character education, it is better to be honest and admit this sooner than later so our energies can be directed elsewhere. But if, as is more likely, they do not undermine the entire project, they can still serve to highlight flaws and point to areas of improvement in how we understand, teach, and communicate about character theory. If character theory is to grow and develop, it can only do this by learning what its weak points are. Addressing these weak points will enable it to adapt and change until the criticism no longer applies. And even if the criticism misses the mark completely and is already fully addressed by existing accounts of character theory, still, the criticism may draw attention to common misconceptions about virtue, or to a prevalent position concerning it, and thus help its proponents to direct their efforts towards countering that misconception. If a critique represents the common opinion of millions of people, then it is a critique worth knowing about and developing resources for a public response.