Love, knowledge (wisdom) and justice: Moral education beyond the cultivation of Aristotelian virtuous character

ABSTRACT

There could hardly have been a more influential twentieth-century philosophical essay than Elizabeth Anscombe’s ‘Modern moral philosophy’, in which she condemned the prevailing and competing ethics of duty and utility of her day and urged moral philosophers to abandon the search for any general conception of ‘morality’ in favour of return to Aristotle’s more particular focus on virtue and virtues as powers or qualities of good human character. While moral philosophers have not been slow to rally to this banner—with timely and useful attention to a wide range of character virtues—it was inevitable that moral educationalists would also soon turn towards understanding virtuous character to be a prime if not main concern of moral education. However, without denying that recent moral philosophical attention to virtue and virtues has been worthwhile, it is crucial to appreciate that in and of themselves such qualities mostly fall short of moral status and value and that some independent criterion of the moral cannot, as Anscombe recommended, be renounced. In this spirit, the present paper also attempts to comprehend such moral and educational significance by reference to interlinked concepts of love, knowledge and justice.

KEYWORDS:

• Virtue
• character
• morality
• love
• knowledge
• justice
• education

‘People divide off vice and virtue as though they were two things, neither of which had with it anything of the other. This is not so. There is no useful virtue that has not some alloy of vice, and hardly any vice, if any, that carries not with it a little dash of virtue; virtue and vice are like life and death, or mind and matter—things which can’t exist without being qualified by their opposite … That vice pays homage to virtue is notorious; we call this hypocrisy; there should be a word found for the homage which virtue not infrequently pays, or at any rate would be wise in paying, to vice.’ (Butler, 1966, p. 113)

Anscombe’s ethical revolution

Elizabeth Anscombe’s paper ‘Modern moral philosophy’ (1958) virtually changed the landscape of modern analytical ethics. In the space of a modest journal article, Anscombe argued that neither of the prevailing (and rival) ethical theories of mid-twentieth-century analytical moral theory—namely, the ethics of duty (ultimately, Kantian) deontology and utility (utilitarianism or consequentialism)—could provide a satisfactory account of the rational grounds of moral agency. More specifically, she observed that neither of these general ethical theories was well equipped for due appreciation of the affective dimensions of human character and motive as significant constituents or determinants of action in the real-world circumstances of moral conduct. In this light, Anscombe made her influential game-changing proposal that contemporary moral philosophers might do well to redirect their attention from the unpromising search for some general principled basis for moral life in favour of close (psychological) attention to particular notions of virtue and virtuous character of precisely the sort defended by Aristotle in (primarily) his Nicomachean Ethics (Aristotle, 1941)—with, perhaps, special attention to the practical reason or wisdom of his concept of phronesis. Many of Anscombe’s philosophical contemporaries and many more of later generation were not slow to rally to this banner and so modern neo-Aristotelian virtue ethics was born. Moreover, while this moral perspective had its initial impact on mainstream moral theory, the turn to Aristotle, virtue and moral character has inevitably found its way to various fields of applied ethics, not least that of educational theory (see, for example, Carr, 2012; Carr & Steutel, 1999; Harrison & Walker, 2018; Kristjansson, 2007, 2015).

Virtue ethics departs crucially from such rival ethical theories as emotivism (also known as non-cognitivism or sentimentalism), utilitarianism (or consequentialism) and deontology in several key respects. Like utilitarianism, and unlike deontology and forms of emotivism, virtue ethics is a form of ethical naturalism that seeks empirical grounding for moral agency; but unlike consequentialism it regards pleasure and/or happiness to be a far from adequate or clearly determinate basis for this. Like emotivism, and unlike utilitarianism and deontology, virtue ethics takes sentiment and emotion to be significant for moral life: but unlike emotivism, it strongly affirms the place of reason in moral agency. Lastly, like deontology, but unlike utilitarianism, virtue ethics locates moral reason or deliberation in a particular conception of practical reason; but it nevertheless denies the independence of such reason from the (biological and material) facts of natural human life. In short, the virtues of virtue ethics are states of character, ultimately constructed upon a natural basis of human affect or disposition, by the formative operations of a distinctive mode of practical reason or wisdom (phronesis) in the light of a naturalistic conception of human flourishing (eudaimonia). On this view, it may appear that virtue ethics gets the very best of all ethical worlds by providing a compelling account of the origins of human moral life and association in empirically discernible, if educationally cultivated, states of human character that are also allegedly conducive to collaboration and rapprochement with modern psychological and other social scientific research and therefore apt for empirical research and enquiry (see Kristjánsson, 2013). In this light, much contemporary work on professional and vocational ethics as well as on basic moral education has also turned in a virtue ethical direction (see, for example; Carr, 2018; Walker, 2009).

However, the present paper aims to question whether the virtuous character of virtue ethics can do the work that Anscombe seems to have thought with particular view to the educational implications of this: to ask, in short, whether appeal to virtue can, as she supposed, effectively displace reference to morality in moral life and education. To this end, whereas the next section will consider the issue of whether the character virtues of Aristotelian ethics are sufficient in and of themselves to ensure moral conduct without further appeal to some independent moral principle, following sections will examine the (perhaps more interesting) question of whether any purported character virtue may be deemed even necessary for the achievement of (what would be commonly accepted as) moral ends. While the present enquiry will deliver a negative verdict on the issue of the moral sufficiency of character virtues, it will nevertheless argue that while some of the capacities that virtue ethicists regard as virtues—namely, wisdom, justice and charity—may be considered key dimensions of morality, they are so only insofar as they are much more than qualities of human character.

The moral status of neo-Aristotelian character virtues

According to the Aristotelian virtue-ethical mainstream, reflection upon that range of character traits commonly called virtues offers the best route to understanding moral life and agency—and, in that case, moral education may be largely considered a matter of character formation under the guidance and direction of the Aristotelian practical reason or wisdom of phronesis. However, understanding moral life and agency as phronesis-shaped character is far from unproblematic, especially for any modern moral educational purpose. First, despite widespread modern association of the term ‘virtue’ with morality, it seems generally doubtful that the character virtues of Aristotelian conception have any necessary moral import. The basic difficulty here is that the Greek arete which has come (via Latin virtus) to mean ‘virtue’—signifies no more than ‘excellence’. As such, it was applied to a wide range of (non-human no less than human) qualities, as well as to human capacities or abilities that have little or nothing to do with morally good character or conduct. Even where the term is applied to humanly admirable or beneficial character, as in Aristotle’s Nicomachean Ethics, it is far from clear that such application always refers to anything that we might confidently call moral character, or that the conduct consequent upon such character is devoted to what we should consider to be moral ends.

Of course, we need not doubt that ancient Greek philosophers regarded the virtues of wisdom, temperance, courage, justice and so forth, as good or desirable human character traits. Plato’s Socrates—arguably the founding father of western philosophy—argued that such so-called ‘cardinal virtues’ were ultimately reducible to wisdom as knowledge of the truth of things—of ourselves, our relations with others as well as of the world in general—and while the more post-Socratic Plato seems to have regarded justice as the master virtue, it also seems (from the cave allegory and his divided line account of knowledge) that any accurate appreciation of this would still need the cultivation of appropriate knowledge. That said, Aristotle’s understanding of such virtues departs significantly from that of his philosophical forebears in effectively denying much if any epistemic role to such traits—largely via his pioneering distinction between various forms of theoretical and practical reason (Carr, 2023a). Here, while he does not entirely deny the epistemic significance of all forms of practical reason—for example, the exercise of techne may indeed give rise to forms of practical knowledge or expertise—Aristotle nevertheless insists that the virtuous deliberations and judgements of his phronesis or practical wisdom are not primarily concerned with the discovery of knowledge. Thus, while he clearly considers phronesis to be an intellectual virtue—as Socrates and Plato also clearly regarded wisdom—he is no less clear that it is not, in the manner of episteme, a virtue concerned with knowledge production. On the contrary, phronesis requires deliberation and judgment precisely because it cannot aspire to the rational certainty that we might seek through episteme (Aristotle, 1941, book 3, section 3).

But if courage, temperance, justice and so on are human excellences—but not of knowledge production or skill acquisition—what is the distinctive realm of expertise in which their excellence may shine forth? Evidently, it is the sphere of personal and interpersonal conduct and association of consequence to human harm and benefit that we are also commonly given to evaluate in moral terms: hence the familiar latter-day endowment of Aristotelian or other so-called qualities of virtuous character with moral significance. But this identification is clearly too quick and far from unproblematic. First, just as epistemic judgements are justifiable on the basis of truth to the facts, and the expertise of techne must submit to tests of practical efficiency or effectiveness, there must be some principled yardstick whereby the deliverances of practical wisdom may count as reasonable or appropriate if they are to be of genuine moral benefit. But while Aristotle does propose a rational steer for the deliberations and judgements of phronesis—namely the principle of the mean in the service of human flourishing (eudaimonia)—this is not always or inevitably to the end of what would usually be considered (in either past or present times) moral ends. Moreover, apart from other levels of difficulty here, there is a general problem of apparent circularity. For while there is evidently more to Aristotle’s notion of human flourishing than virtuous conduct—such as the contingent blessings of human fortune—it seems that Aristotle is at one with his immediate philosophical forbears in taking a flourishing human life to be a life guided by the conduct of such virtues as courage, temperance, honesty and justice. But since there is a quite significant sense in which a flourishing life is a life of virtue, and a virtuous life is a flourishing one, there is also small room here for any real moral evaluation of either virtue or flourishing that does not fatally beg the key ethical question.

To spell out the difficulties more precisely, we should again recall that the practical deliberations and judgements guiding virtuous conduct fall well short of epistemic certainty. Human affairs are invariably prey to moral dilemmas with regard to which the epistemically uncertain deliberations and judgements of phronesis provide no unequivocal guide to right conduct. For example, take the case of a generally honest and compassionate husband and father who knows that a missing son is dead but cannot decide whether to reveal this truth to a wife and mother in the grip of suicidal depression. On the one hand, if the husband honestly admits what he knows, he may risk his wife’s further distress or suicide; on the other hand, if he conceals the truth out of compassion, he may yet prolong her anxiety and suffering—in addition to also being dishonest (and thereby disloyal) to his spouse. Thus, in the case of this or other conflicts in which even virtuous agents cannot be simultaneously honest and compassionate, one agent may choose (with some moral loss) to act honestly, and the other (with equal loss) to act out of compassion. But here, apart from the fact that the conduct of both agents is morally compromised, insofar as they choose differently in much the same circumstances, it would appear that neither of them—even if they hold both honesty and compassion in significant regard—can share quite the same visions of moral benefit or flourishing.

In that case, however, it would not seem that there is any useful general or common conception of either virtue or flourishing. In this light, Aristotle’s proposal to provide a clear-cut all-purpose objective touchstone of what is virtuously (or morally) right or good in terms of what conduces to his naturalistic concept of human flourishing (eudaimonia) is questionable. First, in formal terms, it has been noted that Aristotle’s apparent inference at the start of his Nicomachean Ethics from the commonplace that all human activity has some end to the conclusion that there is some (common) end at which all human activity aims is logically suspect (Geach 1972; see also Carr, 2021), for it is quite clear that there are many and diverse ends (not all of which need be moral) at which human activity aims. But secondly, from a more cultural or sociological perspective, the prospect of identifying any commonly agreed conception of virtue to which human agents do or have (historically) subscribed has been seriously challenged by (for one) the modern virtue-ethicist Alasdair MacIntyre (MacIntyre, 1981, 1987)—a major influence on contemporary educational and professional ethics—who has compellingly argued (despite some momentary ‘naturalist’ deviation from this more general constructivist position in MacIntyre, 2009) that notions of virtue and flourishing are essentially variable and contested normative and cultural constructs. In clearly direct criticism of Aristotle, MacIntyre speaks tellingly of ‘rival traditions’ of virtue which cannot be supported by any naive pre-cultural ‘metaphysical biology’ (MacIntyre, 1981, pp. 58, 163, 196) of natural human function of the kind upon which Aristotle evidently seeks to base them.

But even if we allow, with MacIntyre (and other latter-day neo-idealists), that there is much social or cultural diversity to conceptions of human good, benefit or flourishing might we not still insist that there are certain empirically grounded commonalities of human nature that must impose significant constraints upon what we can fairly count as good or beneficial human conduct (for much this point, see M. Nussbaum, 1993)? This might seem especially so with regard to those ‘executive’ virtues concerned with the control of appetites, or the stiffening of agent resolve: in this regard, since human beings have to eat to live—though not too much as well as not too little—and also stand up to threats and dangers, they will evidently require temperance and courage in any society or culture. Thus, as Geach asserts—in explicitly ‘naturalistic’ terms: ‘Men need the virtues as bees need stings’ (Geach, 1977, p. 17). Still, this point evidently falls well short of any conception of virtue as inevitably directed to moral ends. Indeed, even if we are now quite clear about what courage is or what form it should invariably take (if, that is, discretion is sometimes the better part of valour), it seems that agents may be quite courageous or temperate in the performance of bad or wicked no less than morally good or humanly worthwhile acts. Indeed, the effective prosecution of what is morally untoward often requires such basic ‘good’ character.

To be sure, philosophers have sought to deny this: thus, Geach insists that seemingly courageous but wicked conduct—such as that of (say) a terrorist bomber who sacrifices his own life to destroy others—cannot count as virtuous in the absence of some broader (moral) deliberation that would preclude any such action (Geach, 1977). But aside from the fact that this objection is hardly compelling, if not question-begging, it also imposes an implausibly high bar on much common attribution of courage. Thus, wicked terrorists aside, there is little reason to deny that insofar as many less than virtuous agents may be considered highly courageous in pursuit of quite morally indifferent personal ends—such as rock-climbing or sky diving—any specific moral purposes hardly seem necessary for much common ascription of courage. This also seems true of temperance, since it is evident that agents may exercise much self-control in pursuit of thoroughly wicked or morally indifferent ends. But much the same applies to such more social or ‘other-regarding’ virtues as generosity or gratitude: one may be quite generous to those one knows to be morally derelict or sincerely grateful to people who have done bad or wicked things on one’s behalf. Thus, in the kind of terms that most folk would nowadays consider to be of genuine ‘moral’ import, it seems that few if any Aristotelian character virtues could or should be considered more than accidently or incidentally moral. Such fully phronesis-governed virtues as courage, temperance, perseverance, generosity, gratitude, cooperativeness—even honesty—might all be exercised in ethically dubious contexts or to morally suspect ends.

On the other hand, with regard to the so-called cardinal virtues, there might seem to be exceptions to this rule in the cases of justice and wisdom—since such qualities would appear inherently moral. While fuller exploration of these qualities is postponed until the next section, some brief observations are presently in order. First, with regard to justice—which Aristotle evidently regarded as both a character virtue and as a general moral or social principle—it is clear from contemporary moral theory that its status as a general moral principle has been both defended (by, for example, Rawls, 1971) and denied. On the previously noticed view of MacIntyre—notably in his influential 1987 work Whose Justice, Which Rationality—justice is both a contested and contestable notion. Indeed, Aristotle’s own conception of justice would hardly compel much civilised assent nowadays: not only did it endorse slavery but would also have condoned treating slaves quite badly. And, of course, this cannot but have a serious knock-on effect for any conception of justice or fairness as an associated character virtue. As in the case of other ‘virtuous’ dispositions as generosity and gratitude, fairness might well be exercised to dubious moral ends: thus, a slave owner might seek to treat his slaves fairly (by, say, punishing them equally) or a mafia boss might be scrupulously just to his gang members in equally dividing spoils. Moreover, to object here again that such conduct could not count as fairness or justice unless it served moral ends can only be question-begging in the absence of some independent criterion of morality—which is what is precisely here in question.

What, however, of that virtue of Aristotelian wisdom that appears to shape the proper conduct of such other virtues as courage, temperance and justice or fairness? The difficulty now, however, is that if one gives Aristotelian definition to such wisdom, it is not a moral—or even a practical—capacity, but rather an intellectual virtue. Just as the intellectual virtue of episteme governs reliable knowledge acquisition and techne guides effective cultivation of skills, the practical wisdom of phronesis guides conduct towards the personal and/or interpersonal benefit of agents in light of the naturalistic Aristotelian ideal of flourishing. But now, the status of phronesis as an intellectual power governing correct or reliable practical deliberation still falls short of any distinctively moral end or purpose. Indeed, if we again appreciate that virtues (arete) are but ‘excellences’ variably conducive to flourishing—which means generally succeeding or faring well in human affairs—it would again seem that few if any character virtues need have any moral significance. On this view, virtues would simply be those qualities of character that assist agents to succeed at this or that project according to personal and perhaps inescapably local standards of satisfaction, achievement or prosperity. Indeed, this conception of virtue seems to have been largely that of the ‘sophists’ of such Platonic dialogues as Gorgias and Republic, who specifically held that the manipulative ‘art’ or skill of rhetoric—which Socrates sought to expose as quite bogus or fraudulent—conduced to the flourishing (or advantage) of its practitioners.

To be sure, while it might now be objected that sophists could only be demonstrably mistaken on this score—insofar as any deliberate rhetorical deception or exploitation of others would inevitably fall short of any and all virtue—it is unlikely that such complaint would have cut much ice with sophists. On the contrary that rhetoric might require some dishonesty or lying as a means to assisting an agent’s manipulation or exploitation of others, would likely have counted in its favour for sophists as effectively serving the interests of its practitioners. Of course, insofar as the basically epistemic Socratic or Platonic objection to rhetoric was that ignorance, falsehood and deception must inevitably damage the health of the soul (see Plato, 1961a, p. 462), there might seem to be more of a moral case against rhetoric and its practitioners. But this strategy seems less available on any Aristotelian conception of practical wisdom as significantly divorced from more theoretical or epistemic concerns and primarily devoted to the practical cultivation of states of character ordered to the end of personal or local success or prosperity. In this regard, it is notable that some influential contemporary virtue ethics of basic Aristotelian tendency (though with also some utilitarian drift: Driver, 1989, 2001) seriously claims that there are genuine ‘virtues of ignorance’ (such as modesty) in which disregard for truth or facts of the matter is an inherent feature of their virtuous status.

At all events, insofar as we can clearly distinguish courage and temperance in the service of moral ends from their no less genuine occurrence in pursuit of morally suspect goals, morality cannot, by that very token, be an inherent feature of virtues that they wear, as it was, on their sleeves. In this light, while Aristotle rightly distinguishes the practical habits of good character from the skills of craft and technique, it seems that these are nevertheless not dissimilar insofar as either may be directed towards immoral as well as moral ends. In the last resort, all that substantially distinguishes the skilled from the virtuous is that whereas techne requires discipline of feeling, reason and action for the efficient performance of arts or crafts, phronesis requires it for the promotion of successful (or advantageous) personal and/or inter-personal conduct. For the most part, however, temperance, courage, compassion, generosity, gratitude and the rest are little more than positive personal or social skills of incidental or more or less moral significance. Thus, while we should not in the least deny (in company with such major modern moral theorists as Kant, 1967, p. 305; see also Baxley, 2010) that Aristotelian or other character virtues may provide significant practical or psychological support for the promotion of moral goals, it seems that the defining concepts and principles of moral life are not themselves inherent features of even commendable human character and conduct. Still, even accepting that such character virtues as wisdom, fairness and so on cannot be of themselves of moral worth, it might yet be queried whether some of the principles, powers or qualities with which some of these capacities are associated are nonetheless integral to moral life or association.

The moral and educational significance of wisdom and justice

The later virtue ethics of Christian (most notably of medieval and Thomist) tradition identifies seven key virtues. Thus, in addition to the four so-called ‘cardinal virtues’ of wisdom, temperance, courage and justice of Greek philosophical antiquity (called ‘cardinal’ because, while such virtues may not be defined in terms of other virtues, other virtues are often definable in terms of these: see Carr, 1988), subsequent Christian thought supplemented these with the three ‘theological virtues’ of faith, hope and love taken directly from Saint Paul’s First Letter to the Corinthians (Corinthians 1: 13). Of the cardinal virtues—construed as qualities of good character—we have noticed that temperance, courage and fairness are insufficient in themselves to guarantee moral conduct and that wisdom is not strictly (certainly on Aristotelian construal) a moral virtue as such. Nonetheless, justice—on some understanding of this notion—might reasonably be supposed a significant dimension of any conception of moral life, and that there could be no genuine moral agency without some form of wisdom. Again, with regard to the theological virtues, while faith would seem too dependent on local religious tradition for wider moral espousal, and hope is a more context-relative attitude than moral character trait (whereas one might always reasonably aspire to courage there may be circumstances in which it is hardly reasonable to hope), there might seem more scope for love in the wider economy of human moral life and association. Moreover, this has been seriously entertained by modern moral philosophers of high renown (for example Murdoch, 1970, 1993; M. C. Nussbaum, 1992). Hence, this section will attend more closely to the moral and educational significance of wisdom, justice and love on less particular virtue ethical interpretations.

We may start with some re-consideration of the notion of wisdom. First, it is hard to avoid rational, if not epistemic, construal of wisdom or to deny that it must—on some reading of this rather loose term—play a large role in any serious understanding of moral conduct. For if moral conduct is a genuine form of agency, it is difficult to deny—in the manner of moral non-cognitivists or sentimentalists—that human moral actions are guided by reasons for which agents can be held responsible. It therefore seems implausible to define wisdom in other than cognitive or rational terms or apart from reason, knowledge and understanding—and, crucially, there could hardly be moral education without it. Indeed, Aristotle’s philosophical forbears, Socrates and Plato, were evidently inclined to strong epistemic interpretation of moral agency, effectively identifying moral wisdom with knowledge of the truth of things. To be sure, this strong epistemic reading of moral wisdom may well receive its most striking expression in Plato’s cave allegory, according to which ultimate grasp of (moral) good is attainable only via emancipation from ignorant entrapment in the realm of sensible experience (Plato, 1961c, pp. 514a-520a). This said, it is also this strong epistemic interpretation of moral wisdom that appears to be opposed by Aristotle’s strict formal distinction of the practical intellectual virtue of phronesis from the theoretical virtue of episteme and in his insistence that the practical deliberations and judgements of his practical wisdom have little if anything to do with the discernment of knowledge.

Still, while his pioneering distinctions between different species of rationality are not entirely without philosophical merit, the best case against any over-sharp separation of practical wisdom from theoretical knowledge may be that Aristotle does not consistently observe it himself. Indeed, in his own celebrated treatment of the problem of akrasia or weakness of will (Aristotle, 1941, book 7)—the puzzle of how an agent can know what is morally right in some circumstance but fail to act in accordance with such knowledge—Aristotle hardly strays far from Socrates, effectively claiming that such agents are like drunken men who fail to know or understand what they should do in their inebriated state. (It is also worth noting here that the account of moral virtue of the major contemporary neo-Aristotelian virtue ethicist, John McDowell (1997) is influenced more by this epistemic construal of moral error.) Hence, insofar as agents require significant knowledge of the world, of themselves and of their relations with others for sound practical deliberation and judgment, such reasoning cannot but have wider epistemic implications. (Moreover, agents who deliberate unwisely may yet often learn much of moral value from their mistakes). That said, while the accounts of wisdom of Socrates and Plato are more epistemically directed than that of Aristotle, it may appear that these also fall somewhat short of identifying what is distinctively moral about such knowledge. For example, while Plato’s Republic correctly appreciates that the idea of justice—of giving to others what is rightly due to them—is central to any concept of the (moral) good, his familiar analogy between the proper state of soul and that of optimal political order has proven far from ethically helpful. (Other difficulties aside, this has clearly been the source of enormous later philosophical confusion, insofar as his Republic has been widely misinterpreted as a political tract, rather than—as Plato explicitly intends—an account of good order in the soul: this point is well made in Jonas, 2016b)

All the same, it is arguable that the cave allegory and the epistemology of the divided line of Plato’s Republic point in the right normative direction. Both of these ideas aim essentially towards the same point: that the sort of normative reasoning required for any accurate grasp or vision of good or just human life cannot be reduced to or derived from the generally accepted ‘facts’ or data of empirically conditioned human sense experience. That said, both these notions have been much misunderstood. First, the main point of the cave allegory is evidently that any clear perception of what is (morally) right and good requires significant rational liberation—in this world—from the subjective egotistical attachments (well identified in his Laws: Plato, 1961d, p. 1318) and from the wider social and cultural conditioning that distort fair and accurate perception of self and others. The point is therefore not—as often held—that any vision of the good requires relocation to some other-worldly intellectual realm of purely abstract ideas (see Murdoch, 1993, pp. 298–399; Jonas, 2016b). Likewise, the epistemic taxonomy of Plato’s divided line—the placing of normative reasoning about goodness, rights and justice to a ‘higher’ position than opinion, sensible knowledge or (even) the a priori reasoning of mathematics—is not a matter of its elevation to some transcendental realm of impractical contemplation: on the contrary, it is to make the perfectly proper point that normative reasoning—while essentially practical—is of a significantly different order from these other (empirical and logical) forms of reflection and enquiry. Crucially, for present concerns, it affirms that the ethics of Plato, unlike that of Aristotle, is not any form of naturalism: that fundamental normative concepts such as good and justice, cannot be directly derived from empirical reflection.

Surprising as it may seem, the ethical direction in which these Socratic-Platonic insights would seem to point most clearly is towards the eighteenth-century deontology of Rousseau and Kant. In fact, Rousseau explicitly acknowledged Plato as a key influence and his observation in The Social Contract that ‘those who regard themselves as the masters of others are even greater slaves than they’ (Rousseau, 1973, p. 181) is very much in the spirit of Socrates. The essence of Rousseau’s thought is that in the course of economic transition from a more primitive (and perhaps egalitarian) ‘state of nature’ to a more complex yet socially segregated civil order, human agents come to define themselves in terms of superior and inferior status, falling prey to precisely the same delusion, vanity and hubris (to which he gives the label ‘amor propre’ and to which modern existentialists refer as ‘mauvais fois’ or bad faith) of the prisoners of Plato’s cave. By the same token, Rousseau questions the rational coherence and moral significance of the accidents of good fortune by which the powerful and privileged deny basic human rights, liberties and benefits to less well-favoured others—but which they nevertheless all too readily claim for themselves. For while many have held that some humans should be considered ‘slaves by nature’ (including, here notably, Aristotle) on what natural evidence could any such judgment be made? Rousseau’s answer is that there can be no such evidence insofar as questions of human freedom and rights are not for empirical but normative determination. Thus, it is not by virtue of skin colour, gender or even IQ that agents are owed freedom or rights, but only insofar as such liberties are bestowed and acknowledged by others. In that case, however, what is sauce for the fortunate goose must also be so for the unfortunate gander: if rights and liberties are not matters of accident of birth, but of moral recognition, the slave owner is no better placed for claiming such consideration than the slave—and to suppose otherwise is precisely to be a spiritual captive of Plato’s delusional cave. Once we subscribe to the normative discourse of human freedom and rights, we cannot coherently deny to others those basic human liberties that we claim for ourselves. In more deontological terms, our own claims to freedom and rights entail obligations or duties to respect such claims on the part of others.

Rousseau’s effective socio-political reworking of the Socratic or Platonic idea that something like moral justice requires eschewal of self-focused and self-interested vanities and prejudices to the end of more common and general interest received (with due acknowledgement of his influence) more formal and technical attention in the work of his younger contemporary Immanuel Kant. In his great philosophical critiques, Kant argues: first, that moral reason or deliberation should be distinguished from the knowledge of empirical enquiry (so that the ‘ought’ or obligation of moral deliberation cannot be derived from the ‘is’ of empirical fact): second, that reasoning to moral justice must therefore require impartial rational detachment from personal desires, interests and inclinations to the end of more general or universal principles of consideration for the interests of all human (rational) agents, regardless of status differences (see, especially, Kant, 1967). Here, Kant resembles Aristotle in construing reasoning about the justice of human relations as essentially practical—hence, separate and distinct from the concerns of epistemic enquiry; but he also differs from him with regard to the logical relationship of such reflection to the natural circumstances of human life. But there is again much potential scope for confusion on this matter. Indeed, virtue ethicists usually consider Aristotle to have the edge over Kant by reason of his ultimate grounding of moral or virtuous conduct in a naturalistic conception of human flourishing that is guided by a context-sensitive practical wisdom in a way that Kant’s practical reason is not. However, insofar as Aristotelian practical wisdom is precisely shaped by regard for what may appear to be given human circumstances, such circumstances cannot but condition our view of how we should treat others. In that case, we might not accord Aristotle’s ‘natural slave’ the same rights as free citizens, insofar as his ‘natural’ worth is perceived to be less. On the other hand, Kant’s moral reason dictates how all rational agents should be treated irrespective of—indeed, frequently at variance with—their given natural or social status. But, of course, this hardly implies that Kantian moral deliberation can or should be indifferent to individual differences. On the contrary, the fact that this person is denied the basic human rights of that one precisely gives us a moral reason for opposing such inequity. It is in precisely this sense that Kant seems to provide something far closer than Aristotle’s naturalistic practical wisdom to a distinctive moral (rather than prudential) conception of practical reason for human association.

The moral and educational significance of love

That said, insofar as Kant’s categorical imperative seems to require that moral actions need to be performed entirely for their own sake, or without regard for any ulterior or prudential purpose, this may seem to introduce a further puzzle about moral motivation. Here, by the way, insofar as some connection is commonly made between the moral law of deontology and the so-called ‘golden rule’ of Christian and other religious ethics, the notion that one should ‘always treat others as you should like them to treat you’ (Matthew 7: 12) actually would provide agents with a self-interested motive for so doing—and the same also seems to be true of the twentieth-century contractarian ethics that has often claimed to follow in the wake of Rousseau and Kant (see, notably, Rawls, 1971). It should be clear, however, that we cannot equate such reasoning with Kant’s categorical imperative or moral law which appears to require moral agency to be more disinterested, impartial or impersonal. In this regard, the Aristotelian focus on virtue and virtuous character might appear closer to real life than Kantian or other ethics, precisely by virtue of providing a more credible naturalistic or down-to-earth account of familiar motives for good conduct. On this view, agents are self-controlled because this is the most effective or disciplined way to achieve their practical goals: they are fair and cooperative because this is by and large the way to win the cooperation of others; and so forth. But if, as often, moral action seems to run counter to ordinary human interests or advantage, the deontological account of moral obligation may seem to raise the perennial ethical question: ‘Why should I be moral?’ All the same, the idea of moral action for its own sake is not so easily dismissed and far from obviously at odds with ordinary human motivation.

To begin with, people obviously do act or engage in activities for their own sake—from doing crossword puzzles, writing philosophical essays or poetry to climbing Everest. Here, moreover, while it might be objected that agents do such things as a means to personal satisfaction, this really adds little or nothing to the plain truth that such actions are done for their own sake. Relatedly, indeed, it should not be supposed that motives are always desires or drives—such as hunger, vanity or status-seeking—that are, as it were, external or additional to the inherent point of such actions. Thus, that a given action—one that promotes universal human benefit and justice—seems fine and noble in its own right, may be quite enough for a right-minded agent to perform it. But further, while actions in accordance with Kant’s moral law or the categorical imperative have sometimes been characterized as disinterested, impartial and impersonal, it should be noticed that ‘disinterested’ does not mean the same as ‘uninterested’ and impartial is not at all the same as impersonal. On the contrary, strictly impartial actions—such as the scrupulously equitable treatment by parents of their equally loved children—can be highly personal and vice versa. Indeed, this observation could hardly be more to the present point insofar as it naturally invites some consideration of love—the ultimate theological virtue of neo-Aristotelian scholasticism—as a dimension of any complete account of human moral life and education. For the theological virtue of love—or in the somewhat milder term of some New Testament translations (such as the King James Bible) ‘charity’—is precisely addressed to the sort of general positive attachment to other human beings that is most notably expressed (for example, in the Matthew Gospel account of Jesus’ Sermon on the Mount) as ‘loving one’s neighbour as oneself’.

Still, this idea is also not without problems. In the first place, the substantial semantic ground covered by the simple English term ‘love’ is prone to much more precise discrimination in the Greek of the New Testament and other ancient literature. As notably explored in a classic modern treatment of the topic (Lewis, 1960)—the Greeks distinguished between four sorts of love under the labels eros (sexual desire), storge (family affection), philia (friendship) and agape (broader human brotherhood). In this light, the universal Christian love of neighbour (and also, notably, enemies) is usually denoted by the term agape (or later Latin caritas). However, a trouble now is that it is not easy to square such universal love with the more personal human affection and attachment characteristic of relationships with partners, family or friends (see Carr, 2022). Indeed, while it might be clear how to entertain some such ‘higher’ love of agape as a general (disinterested) moral principle, it seems less evident how any agent could hope to feel such universal attachment to every other human person. It may also be a further psychological problem that since some people do not appear to love themselves, any injunction to treat others as one would treat oneself could hardly be generally sound advice. In short, there would seem to be considerable psychological shortfall between the diverse affective states that we commonly associate with (at least) the Anglophone term ‘love’ and the more disinterestedly principled respect for agents as ends-in-themselves associated with deontological moral law, the Kantian categorical imperative or more universalistic interpretations of arete.

Once again, however, it may be that a philosophically compelling connection between more basic or ‘lower’ erotic and familial human loves and something closer to the higher love of agape is available courtesy of Plato. For, in Plato’s much-loved dialogue The Symposium, we find Socrates debating the human significance of the ‘lower’ sexual attraction of eros with a range of carousing companions (Plato, 1961b, p. 1318). Allegedly reporting the views of the prophetess Diotima, Plato’s Socrates seeks to show how even the superficial, wanton and (at least from any modern moral perspective) sexually exploitative relationships of ancient Greek times between older and younger males might nevertheless provide opportunity—under the enlightened mentorship of an older and wiser partner—for growth of respect for those qualities of virtuous character that underpin deeper friendship (philia) and, perhaps beyond this, to appreciation of other-regarding human love of closer approximation to agape. Thus, in terms of what has been called ‘the ladder of love’ (see, for example, Blondell, 2007), even erotic attraction may operate as a vehicle of ascent from natural physical desire or lust to more morally or spiritually elevated appreciation of and/or regard for others. Indeed, it seems that Socrates here seeks to encourage eventual replacement of the dubious homoerotic sexual liaisons of his day with something more like the chaste moral and spiritual associations that he himself appears to have successfully cultivated with his own youthful disciples. At all events, on this view, agents may be assisted to move beyond such more carnal attachments, either by renouncing them entirely or adopting a more detached or balanced perspective on them. But the ladder of love of Plato’s Symposium certainly points to significantly higher development of self-focused amorous sentiment in the light of deeper knowledge of others and/or of one’s personal relationships with them.

So considered, even the more basic loves of sexual desire (eros) and natural familial attachment (storge) may provide springboards for development of more morally principled or altruistic regard for other persons that is closer to friendship (philia) and love of neighbour (arete). In this light, however, such further personal attachments cannot be matters of mere blind affect and must involve significant epistemic engagement. Thus, again, despite common talk of the ‘blindness’ of human (notably erotic) love, and any and all claims that non-human creatures may also be capable of love (or, indeed, of knowledge), it is clear that the affectionate attachments of dogs, cats and rabbits (and it is significant that the list does not usually extend much beyond this to tortoises, goldfish and stick-insects) must fall well short of the cognitive or rational grasp or appreciation of what is loved by human agents. While Fido may doggedly follow me around or lick my face, he cannot appreciate or admire me for my skill at chess, sense of humour or taste in visual art. This is also why animal or pet attachments cannot grow, develop or deepen (or perhaps even wane or fade) in the way that human love both can and does: the more one gets to know a topic of interest (such as Baroque music), an activity (such as Alpine mountaineering) or a person (such as a friend or spouse) the more one may come to love this ‘object’—or, perhaps, by the same token, become bored or sated with it. In much this spirit, the distinguished twentieth-century philosopher and novelist Iris Murdoch precisely claims—speaking specifically here of the love of one human agent for another—that ‘love is knowledge of the individual’ (Murdoch, 1970, p. 28). To be sure, while this cannot be quite correct as stated—since some more basic human loves may indeed be ‘blind’ to the truth about their objects and there may also be deep knowledge of others that is far from loving—Murdoch may seem generally on track in taking other-regarding knowledge to be significantly implicated in or required for love of much human worth.

Thus, on the story so far, the key dimensions of human moral life and agency may be understood in terms of an eternal triangle of love, knowledge and justice. Of these, justice—understood as a general concern for the wellbeing of others alongside or no less than one’s own is informally expressed in the religious injunction to love one’s neighbour as oneself and more formally in the categorical imperative of Kantian or other deontology. For reasons already aired, however, neither of these formulations quite hits the mark. For present purposes, the main shortcoming of the generally promising deontological concept of moral conduct as impartial respect for other rational agents as ends in themselves is that it lacks grounding in any very credible human motivation. Indeed, neither the highly formal norms and imperatives of deontology, nor the purposes of consequentialism—to which the ends more than the motives of moral conduct are held to be of prime concern—leave much room for the personal engagement that we commonly associate with genuine moral agency. However, we have seen how moral theorists from Socrates and Plato to Murdoch and Nussbaum have sought to find a significant role for love in the general economy of human moral life. Indeed, Jorge Garcia has more recently made a yet bolder attempt to place love firmly at the centre of contemporary ethical theorising (Garcia, 2019).

Still, in light of recent arguments, while Garcia’s case for the central moral significance of love is not without merit, his explicit construal of love as a character virtue and more general theoretical inclination to virtue ethics is less compelling and seems a stretch too far. To begin with, love is a far too diverse and protean human quality to admit of any unitary or unambiguous interpretation and many sentiments, attitudes and dispositions commonly regarded as love have little or no ethical concern whatever (Carr, 2022). But secondly, as so far argued, even where love may be fairly considered a quality of virtuous character (as in cases of philia and arete), it may fall well short of either moral purposes or consequences. With regard to purposes, a particular, even self-effacing, attachment to friend or partner may be quite personally commendable without much wider moral regard. With regard to consequences, even more inclusive loves may simply miss the moral mark. Even if, as Garcia maintains, (other-regarding) love is necessary for moral agency, it may not be sufficient and the best loving intentions may—through ignorance, carelessness or whatever—have negative moral consequences. For even if consequentialism falls ethically short via neglect of moral motives in favour of outcomes, this hardly means that only motives matter and that actual positive or beneficial outcomes are of no moral significance whatsoever for successful moral agency. In that case, the question of what is or is not morally justifiable cannot rest with what we may regard as virtuous—or fit or excellent for this or that human purpose. For since it remains an open question whether so-called virtuous conduct is in fact morally justifiable in this or that circumstance, some independent measure of moral worth is evidently needed to decide the issue. With particular regard to love, then, if not any but only the right sort of love is crucial for moral life or conduct, how may we distinguish such true love from the (many) sorts of (even virtuous) love that have no such evident concern or outcomes?

As also seen, however, while there are clearly many different kinds and varieties of so-called human love—such as eros, storge, philia, agape—the upshot of the Symposium ‘ladder of love’ is that they need not be conceived as entirely disconnected or discontinuous: that, indeed, ‘lower’ forms of amorous attachment may be preconditions of or give rise to rather ‘higher’ forms. With the right guidance, what starts off as primitive infantile need or dependence upon parental care can soon develop, with proper nurture, into relatively selfless childish desire to return such care and concern to loving parents; what begins as purely erotic adolescent attachment to the physical charms of another person may often develop (as indicated by Socrates) into deeper respect for the personality and character of that individual and to shared interests with them; what begins as a love of others of one’s local community or tribe can be extended to other human agents more widely in the name of love of one’s neighbour (or even enemy) as an end in his or her own right. In this light, the ladder of love may reach to the very skies and the smallest acorns of human attachment may eventually sprout into the largest and most widespread oaks of altruistic care and concern for their fellow creatures. To this end, however, the main engine of such development is no less clearly the third key dimension or component of moral agency and association—the distinctive human capacity for wisdom or knowledge—under the guidance of human reason (if this is also held to include the rational cultivation of sentiment and emotion). In short, the route to the highest human love of something like agape lies through appropriate (moral) education or tuition.

Conclusion: the twists and turns of modern moral education

Modern philosophical and psychological reflection upon moral education, at least over the past century or so, seems to have progressed through a succession of fairly distinct phases. First, the experimental psychology of learning theorists of earlier years of the twentieth century—which had seriously neglected the all-important role of reason and rational choice in moral life and agency—gave way in due course to the cognitive developmental accounts of Piaget, Kohlberg and others which came to dominance after the second world war. Still, the closing years of that century saw some reaction to (particularly Kohlbergian) cognitive developmentalism on the grounds that this, in its turn, seemed to ignore or neglect the practical dimensions of moral development of character and actual conduct. While the first moves in this direction seem to have been made by American educationalists located more in traditions of modern experimental psychology, these were soon followed by modern virtue ethicists who have urged return to Aristotle’s conception of good character for significant reconciliation of empirical character and practical conduct with (practical) moral reason. More recent days, however, have seen some attempt to revive the work of Plato as an alternative source of potent moral educational insights by the educational philosopher, Mark Jonas. In this regard, a major complaint of Jonas (2016a) is that modern or contemporary moral educational neglect of Plato may derive largely from the acknowledged influence on and interpretation of this philosopher’s work by the influential cognitive developmentalist Kohlberg (1984). Here, Jonas takes issue: first, with what he regards as Kohlberg’s false ‘intellectualist’ reading of Plato’s account of moral reason which neglects its affective and motivational dimensions; second, with his failure (as suggested by his dismissive ‘bag of virtues’ comments) to appreciate the significance of habituation and role-modelling in psychologically and practically sound moral development. Indeed, Jonas argues that Plato’s own close attention to these matters at numerous places in his work clearly indicates that his thought was very much closer to Aristotle’s virtue ethics than sometimes nowadays assumed.

All the same, while Jonas’s moral educational return to Plato is timely and commendable, it is not in the present view quite on target and his particular attempt to disassociate Plato from Kohlberg (or vice versa) is in some danger of throwing out the Platonic baby with the Kohlbergian bathwater. For despite the overall shortcomings of Kohlberg’s account of moral education (as explored in an extensive modern critical literature) his very real debt to Plato—albeit via the deontology of Kant and Piaget—cannot and should not be ignored or denied. Thus, Kohlberg explicitly and correctly discerns (again via Kant and Piaget) Plato’s strong leaning—clearly evident in his metaphors of the cave and the divided line—towards an empirically transcendent or deontic form of moral reason and reflection that is considerably less apparent (if not, to be sure, actually opposed or repudiated) in the work of his pupil Aristotle. Indeed, as we have seen, the pivotal mode of reasoning of Aristotle’s ethics is fundamentally a form of hypothetical practical reason (phronesis) centred upon the promotion of certain naturally (or empirically) determinable human benefits of personal or social flourishing. This is what crucially classifies Aristotle’s account as a species of ethical naturalism (like utilitarianism) and what significantly distinguishes it from the reason and reflection of Socrates and Plato on the form of moral good which evidently aims to shape or transform our empirical experience and character more than act in its service. Thus, while we can agree with Jonas that the ultimate aim of Socratic or Platonic normative reflection is to foster human wisdom in this (rather than some other metaphysical) world, what the metaphors of the cave and divided line nevertheless show is that this cannot come to pass without some disinterested normative detachment or distancing from the world—from the vain and delusive snares of ego (Plato, 1961d, p. 1318)—in the name of some more abstract principle of reciprocal moral regard between rational human agents. Indeed, this could hardly be clearer from Jonas’s own robust defence of Socrates’ ‘first city’ in the Republic (Jonas et al., 2012) wherein truly virtuous citizens exhibit an ego-free commitment to the common good which is quite purged of the worldly ambitions whereby the agents of Aristotelian flourishing might yet be driven.

What, then, of Jonas’s argument that Plato’s position cleaves nearer to Aristotle’s by virtue of like insistence on the developmental processes of moral habituation and exemplification? Again, while the moral significance of such concerns need not be gainsaid, they should not also be overstated. To be sure, as microcosms of wider society—and usually as the initial larger social contexts beyond the family circle to be encountered by children—schools have ever played an important role in the basic discipline of young people via initiation into the rules of personal and interpersonal conduct and association. Thus, most if not all school discipline, in both earlier and later stages of pupil development, focuses upon ensuring that pupils do not speak out of turn, that they put their hands up to ask questions, are respectful of common or other personal property, are reasonably courteous towards or tolerant of others, do not violently interact in classroom or playground and so forth. Moreover, while the purpose of such school or wider social discipline is not educational as such—since it matters less whether the young comprehend its ends than that they abide by its dictates—it is nevertheless usual for such constraints to be attended by some promotion of rational understanding of its personal and social purposes. So considered, such basic discipline is the basis of much good character formation and of moral instruction more generally. In this light, it is also only proper that social and professional agencies should try to ensure, as far as possible, that those (such as parents and teachers) charged with custody and care of children and young people are themselves persons of good or exemplary character. That said, it seems likely that this or other basic character formation is neither necessary nor sufficient for ultimate educated moral growth: while it is not uncommon for those of exemplary moral nurture to go to the bad, others have survived appalling upbringing to become agents of exemplary moral character. As for role-modelling, one has only to glance at the state of (western and other) contemporary populist politics to appreciate that the most admired and effective (political and other) role models can often be those of the most deplorable moral character (Carr, 2023b).

Thus, while it is much to the good that parents and teachers do all in their power to ensure that children and young people are disciplined in accordance with some defensible conception or ideal of personal and social good, such (perhaps Aristotelian) character training—even under the guidance of the practical deliberation of phronesis—evidently falls well short of the morally transformative wisdom of virtue envisaged by Plato in the cave allegory and elsewhere in Socratic and other dialogues. In this light, the pressing moral educational question would seem to be that of how any such more ego-transcendent vision of moral virtue might come to follow or emerge from the more egotistic interests and attachments of natural human development. How can captives of the cave or of the ‘fat relentless ego’ (Murdoch, 1970: Plato, 1961c, p. 1318) move towards the relative moral detachment and selflessness of Socratic virtue? This is where the ladder of love of Plato’s Symposium clearly comes into its moral educational own. And here, though there is likely to be much devil in the detail, the general moral drift is not too hard to grasp. Education towards such higher attachment and commitment may be expected to follow by exposure to influences and mentors (such as parents in the home or teachers in schools) to that wider knowledge and understanding of the world, themselves and their relations with others that enables young people to move beyond self-focused infantile needs and dependencies to engagement with more mature or adult interests and activities. Moreover, while examples of the kinds of interests and activities that may exercise ultimate moral influence of this kind are clearly limitless, one common practice of particular moral educational significance is that of the reading to children from earliest years by parents and teachers of various kinds of (factual and fictional) literary narratives. Thus, while the teaching of tennis, advanced mathematics or nuclear physics may not be a personally fulfilling interest or enthusiasm for all and sundry, it is more generally (and wisely) held that there is much of general personal and moral worth to the teaching of history and imaginative literature. Thus, Jane Austen’s Emma seems a near perfect example (though such examples, from Greek antiquity to the present day, may be multiplied ad infinitum) of Platonic metanoia whereby the heroine of Austen’s title achieves significant emancipation from the deluded and hubristic cavern of the fat relentless ego to a more morally objective and enlightened understanding of herself and her relations with others. Moreover, while the moral insight into herself that Emma gains in the course of Austen’s novel—and by which her readers may also stand to benefit—is quite Platonic (and/or Murdochian), it is clearly also not of the other-worldly or out-of-the body kind for which Plato has often been criticized, but the quite familiar and everyday knowledge of fair and decent human association.

Still, as this paper has sought to show, we should beware of supposing that such moral progress boils down—in the manner of Aristotelian neo-naturalist virtue ethics—to allegedly natural but morally ambivalent qualities of human character and it is important to avoid the dubious elevation of such so-called ‘virtues’ to primary moral status as currently canvassed by many contemporary character educationalists. Here, to be clear, the problem is not that we cannot reasonably speak of some qualities of human character as virtues or identify differences of moral significance between good and bad character (though some serious shortfall between the ancient and modern senses of this term should be noted and the somewhat archaic or superannuated character of this term has also gone surprisingly unremarked). We need not also doubt, as Kant (1967; see also Baxley 2010) and other past moral theorists have held, that human qualities of so-called good character may have an important supportive, though secondary and auxiliary, role in the conduct of moral life and agency. All the same, in the spirit of the quote from Butler at the start of this paper, we should understand that such qualities are all too often morally protean or ambivalent and that in default of some independent and principled normative benchmark of moral significance they cannot provide any very reliable guide to that life. In this regard, we must conclude by re-asserting—against Anscombe and recent neo-Aristotelian virtue ethics—the primacy of the moral over the strictly subordinate notion of virtuous character.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

David Carr

David Carr is Emeritus Professor of the University of Edinburgh and was until recently Professor of Ethics and Education at the University of Birmingham (UK) Jubilee Centre for the Study of Character and Virtues. He is author of four books, editor or co-editor of several major collections of essays on philosophy and/or education and his papers have appeared in such journals as Mind, Philosophy, Philosophical Quarterly, Proceedings of the Aristotelian Society, Philosophical Studies, Journal of Value Inquiry, British Journal of Aesthetics, Educational Theory and Oxford Educational Review. Much of his work has explored aspects of virtue ethics and, more recently, the impact of literature and various other arts on moral character.