2. Eudaimonist Virtue Ethics, Ancient and Modern

The locus classicus for both eudaimonism and virtue ethics is Aristotle’s Nicomachean Ethics (NE). Aristotle declares in the first book that flourishing (eudaimonia) is the end of actions (I.7, 1097b20–21), by which he means the best or highest good, that ‘end of things to be done [i.e., actions and crafts] which we will for itself, and for this [we will] the other things [to be done]’ (I.2, 1094a18–19).Footnote ⁴ Thus, Aristotle is a eudaimonist. After the famous Function Argument, he concludes that ‘the human good [i.e., flourishing] is activity of soul on the basis of virtue, and, if the virtues are many, on the basis of the best and most final’ (1098a16–18). So, Aristotle is also a virtue ethicist subscribing to a virtue conception of flourishing. And he goes on to tell us precisely what the relevant virtues are. At the end of the NE, Aristotle argues that the most flourishing human life is ‘the life based on the intellect’ (X.7, 1178a6–8), which is characterised by the exercise of theoretical wisdom in the activity of philosophical contemplation (theōria). ‘And secondly [most flourishing] is the life based on the other virtue’Footnote ⁵ (X.8, 1178a9), meaning practical character virtues like justice, courage and moderation, which are the subjects of the largest part of the NE. So already in Aristotle we see two distinct strains of virtue eudaimonism, one based upon the exercise of intellectual virtue, the other based upon the exercise of character virtues.

Many recent virtue ethicists follow Aristotle in being eudaimonists.Footnote ⁶ Sparking the contemporary revival of virtue ethics, Anscombe (Reference Anscombe1958) called for the abandonment of the modern ethics of obligation in favour of a classically inspired investigation into both virtue and human flourishing. Subsequent virtue ethicists have taken up Anscombe’s call. Hursthouse, for example, defines virtues as character traits that ‘a human being needs for eudaimonia, to flourish or live well’ (Reference Hursthouse1999, p. 167). Foot invokes ‘an understanding of the word “happiness” that is close to Aristotle’s eudaimonia in that operation in conformity with the virtues belongs to its meaning’ (Reference Foot2001, p. 97). Annas claims that ‘[h]appiness is (at least in part) constituted by virtuous activity’ (Reference Annas2011, p. 165). Each of these authors asserts that virtue is necessary for happiness in the sense of flourishing. And in each case the claim seems to be not simply that virtuous activity is instrumentally necessary for flourishing, but that it actually (partly) constitutes flourishing.

On the common Aristotelian view, a virtue of character is a robust psychological disposition to feel and act in correct ways and for correct reasons (NE, II.6, 1106b16–23; cp. Hursthouse, Reference Hursthouse1999, pp. 11–13). When these virtuous dispositions manifest to motivate corresponding feelings and actions, the virtue is exercised in character-virtuous activity. The ‘standard list’ of paradigmatic character virtues shared by Aristotle and most modern virtue ethicists includes justice, courage, moderation, honesty and generosity (Hursthouse, Reference Hursthouse1999, p. 28). Aristotle also recognises character virtues that contemporary virtue ethicists do not much discuss and which they may or may not accept as virtues, e.g., magnanimity, proper anger, proper shame, wittiness and friendliness (NE, II.7; cp. Eudemian Ethics, II.3, 1220b38–1221b3). And contemporary virtue ethicists often recognise virtues which Aristotle does not, such as hope, beneficence or charity, compassion or empathy, patience, humility and industriousness (Foot, Reference Foot1978; Adams, Reference Adams2006). Character eudaimonism takes flourishing to consist in the exercise of some number of such candidate virtues.

Interestingly, while Aristotle considers flourishing to be primarily the exercise of intellectual virtue, modern authors tend to ignore intellectual virtue eudaimonism and defend only character eudaimonism. Hursthouse (Reference Hursthouse1999) and Annas (Reference Annas2011), for example, only discuss character virtues and practical wisdom, never purely intellectual virtue or theoretical wisdom. To be sure, virtue epistemologists describe character-like intellectual virtues such as epistemic justice, intellectual courage and intellectual humility (e.g., Fricker, Reference Fricker2007; Roberts & Wood, Reference Roberts and Wood2007). But these are not the comprehensive theoretical wisdom which Aristotle prizes above all else. Modern Neo-Aristotelians, then, tend to spurn the central Aristotelian doctrine that flourishing is primarily theoretical contemplation. This could stem from an intuitive aversion to Aristotle’s apparent intellectual elitism. But this is speculation. All we find from modern virtue eudaimonists is silence on the topic. Since contemporary philosophers are primarily interested in character eudaimonism, that will be my focus here. I leave the prospects for intellectual virtue eudaimonism to other work.

3. Eudaimonism

Given the pedigree and prevalence of virtue-ethical forms of eudaimonism, a strong association between eudaimonism and virtue ethics is not surprising. Though he was not the first to investigate either eudaimonia or virtue, Aristotle’s admirably systematic treatment of each in the NE has made this the founding text for both eudaimonism and virtue ethics in the Western tradition. Modern virtue ethics is often formulated in terms of human flourishing. The result, in the words of Baril, is that ‘[e]udaimonia and virtue ethics are interwoven, many extant virtue-ethical theories being eudaimonist, and many (perhaps most) extant eudaimonist theories being virtue-ethical’ (Reference Baril and van Hooft2014, p. 17). Baril herself carefully distinguishes the two, but it would be easy to consider them equivalent, given the present state of the field.

Nevertheless, I urge to distinguish eudaimonism from its specifically virtue-ethical forms, first because there are other forms of eudaimonism which are not virtue-ethical. We can substantiate this claim before carefully defining eudaimonism, and this will help to motivate the definition to follow. At first glance, the feature of an ethical theory that most marks it as eudaimonist appears to be its organisation around the concept of eudaimonia. But virtue conceptions of eudaimonia, which identify flourishing with having or exercising virtue, have never been the only available substantive conceptions of the concept of eudaimonia.Footnote ⁷ If they were, ancient authors from Plato and Aristotle to the Stoics would not have felt so compelled to argue for their correctness. These authors considered lives of pleasure, honour, power or wealth to be serious enough contenders for the title of the best human life to warrant philosophical refutation.Footnote ⁸ Such refutation was indeed necessary, for philosophers like the Epicureans were hedonist eudaimonists who held that flourishing consists in pleasure and the absence of pain. To be sure, the Epicureans were also virtue ethicists, but they considered virtue to be instrumental to their hedonistic conception of flourishing, rather than identifying flourishing with virtue.Footnote ⁹

Some philosophers today still hold that flourishing consists in experiencing pleasure or satisfying desires or possessing objective goods like friendship, knowledge and achievement.Footnote ¹⁰ And nothing prevents an ethical theory from being organised around any such virtue-less conception of eudaimonia. Nor need such an ethical theory invoke virtues in the deontic, action-guiding part of the theory. That could focus instead on rules conducive to flourishing or on the consequences of actions for overall flourishing.

So then, if eudaimonism and virtue ethics are not equivalent, what is eudaimonism? To answer this question, we may look to Aristotle’s model. As we have seen, he establishes eudaimonia as the highest human good and the ultimate end or aim of human action (NE, I.7, 1097b20–21; I.13, 1102a14–15). The version of this claim that matters to ethical theory is not the psychological claim that everyone always desires flourishing (though Aristotle might have accepted this also). It is instead a normative claim about the ethical justification of action. Flourishing is the ultimate end of action in the sense that it is the final reason or purpose for which actions are ethically recommended, required or justified, when they are.

An ultimate end, in the normative sense, is the last good answer to a sequence of justificatory ‘why?’ questions. Why should someone work a tedious job? Perhaps, to earn money. Why should they earn money? Among other things, to purchase necessities. Why should they want necessities? To stay alive and fit. Fit to do what? We think such justifications must stop somewhere and that the justificatory force (the ‘value’Footnote ¹¹ or ‘goodness’) of any answer in the sequence depends upon the justificatory force of the answers which follow it, if any. That is, if earning money justifies working a tedious job, it is either because earning money is justified in turn by some other end or because earning money is an ultimate end (i.e., an end in itself).

It is a substantive assumption that sequences of justification do eventually terminate in an ultimate end, rather than continuing in an infinite regress (A for the sake of B for the sake of C…) or circle (A for the sake of B for the sake of A…). But it is certainly an attractive assumption. Aristotle notes that, if everything were justified by something else, then this endless series of justifications would be unsatisfying and meaningless, our desires ‘empty and vain’ (NE, I.2, 1094a18–22). Many of us expect a point to living and acting, some final goal at which to aim and which makes everything else worthwhile. Absent such justificatory bedrock, there seems no genuine justification to be had. As a theory of ethical justification, therefore, it is plausible for eudaimonism to presuppose that justification terminates in an ultimate end. Any non-terminating model of ethical justification would be at a serious theoretical disadvantage from the outset.Footnote ¹²

It is another substantive assumption that there is one single type of ultimate end rather than multiple (e.g., flourishing and pleasure and knowledge and friendship, etc.). The simplicity of positing only one ultimate end surely possesses theoretical and deliberative advantages, though there is no need at present to prove this assumption. Here, it is enough to establish this as a commitment of eudaimonism, which places human flourishing in the theoretical role of the sole ultimate end of human action in the justificatory sense. All ethically recommended actions are so ultimately because they contribute to flourishing, whether by constituting it, causing it or making it possible or probable. That is, an action’s contribution to flourishing is always the last good answer to justificatory ‘why?’ questions for the purposes of ethical recommendation. Equivalently, an ethical theory is eudaimonist if and only if, and because, it fundamentally recommends actions which contribute to flourishing. This recommendation is fundamental in the sense that all other, more concrete recommendations are derivable from, or conditional on, this one. The character eudaimonist recommends virtuous actions because these supposedly constitute flourishing; another kind of eudaimonist, with a different conception of flourishing, might make different lower-level recommendations altogether but offers the same kind of justification for them.

This is not the only available use of the term, ‘eudaimonism’. Some authors use it to refer to a specific, perfectionist theory of well-being as the fulfilment of one’s nature (Haybron, Reference Haybron2008, pp. 35–36) or perfection of one’s capacities (Besser-Jones, Reference Besser-Jones and Fletcher2016, pp. 187, 192). Other authors use ‘eudaimonism’ for claims about virtue. Crisp & Slote define ‘eudaimonism’ as the view ‘that genuine virtue must be in the interest of the virtuous individual’ (Reference Crisp, Slote, Crisp and Slote1997, p. 8). Parry & Thorsrud (Reference Parry, Thorsrud and Zalta2021) define it as a ‘position that links happiness and virtue’. In this spirit, one could employ ‘eudaimonism’ as the name for what I am calling ‘character virtue eudaimonism’ or else simply for the character virtue conception of flourishing. I will not claim that any of these are illegitimate uses of ‘eudaimonism’, but they are not how I employ the term.

Others do also employ ‘eudaimonism’ in my sense, however. Annas uses it to describe the ancient Greek philosophical ‘consensus that our final end is happiness (eudaimonia)’ (Reference Annas1993, p. 12). According to Russell,

LeBar is primarily interested in defending ‘Virtue Eudaimonism’, but he recognises the more general framework of ‘eudaimonism, the idea that all our reasons for action are for the sake of living well’ (Reference LeBar2013, p. 3). He explains Aristotle’s commitment to this framework much as I do:

Baril similarly explains that ‘according to eudaimonism[… r]ecommendations to φ are made on the condition that φ-ing is (alternatively) conducive to, or at least compatible with, the realization of eudaimonia’ (Reference Baril and van Hooft2014, p. 25, original emphasis). In this shared sense of the term, eudaimonism is structurally analogous to ethical hedonism, ‘the claim that all and only pleasure has positive importance’ (Moore, Reference Moore and Zalta2019), and to welfarism, ‘the view that morality is centrally concerned with the welfare or well-being of individuals’ (Keller, Reference Keller2009, p. 82). Such views agree that there is one fundamental ethical value or end and differ only in what they take that ultimate ethically justifying end to be.

This comparison is useful for distinguishing my sense of ‘eudaimonism’ from yet another, nearby sense: what Hirji calls a ‘a strong form of eudaimonism, the view that an agent ought to choose all actions for the sake of her own eudaimonia’ (Reference Hirji2019, p. 690, emphasis added). It is common enough for authors to think that a given agent’s ultimate justificatory end, according to eudaimonism, must be that agent’s own flourishing. This has formed the basis for a familiar charge of egoism on the part of eudaimonism, and specifically character eudaimonism: the eudaimonist agent is fundamentally concerned with their own flourishing and not with others’ good.Footnote ¹⁴ However, nothing about the justificatory structure of eudaimonism on its own restricts us to the view that only the agent’s own flourishing is their ultimate justificatory end. That end could instead be people’s flourishing in general. To saddle eudaimonism with egoism would be akin to asserting that ethical hedonism is essentially egoistic, when in fact universalistic hedonism is a live option, as Sidgwick (Reference Sidgwick1981) and other utilitarians attest. So, I will not treat egoism as essential to eudaimonism, and thus we can sidestep debates about whether such egoism is objectionable. This may strike some as a revisionary sense of ‘eudaimonism’, but I adopt it purely for dialectical convenience. My arguments below should apply to either an egoistic or universalistic form of character eudaimonism, so we may as well consider the logically minimal sense of ‘eudaimonism’ that is not committed one way or the other.

For present purposes, then, let us define ‘eudaimonism’ as the view that the ultimate justificatory end of action is flourishing (and not necessarily the agent’s own). This definition provides at least one criterion for evaluating eudaimonist ethical theories: whatever flourishing is taken to be must plausibly be an ultimate end, not requiring further justification.Footnote ¹⁵ This is not to say that whatever flourishing is cannot also be instrumental to other things – almost anything can sometimes be instrumental to something. But a candidate to constitute flourishing should not characteristically rely on its further effects for its justification.

What, then, makes something plausible as the ultimate end of action? As a heuristic, at least, an ultimate end should be something it makes sense to pursue for its own sake. Pain, for example, would be an implausible thing to seek for its own sake and money is only worth pursuing for what you can do with it, so neither is a good candidate to be an ultimate end. This conclusion, of course, rests on substantive normative judgements about the kinds of value which these things have or lack. But many of us are independently inclined to accept these judgements, and so there is no reason we cannot employ them in the course of evaluating candidate conceptions of flourishing.

Furthermore, for something to plausibly exhaust flourishing, it should make sense not just to realise that end when given an easy opportunity, but also to strive for it, to desire and seek and create and appreciate opportunities to realise it. That is, it should make sense to organise one’s life, including its more laborious activities like earning money and keeping healthy, around and towards that ultimate end. After all, an ethical theory is supposed to give us broad practical guidance about how to live and act, how to acquire and dispose of free time and resources, and all such guidance, by hypothesis, must bottom out in ultimate justificatory ends. To apply this heuristic, while it may be worthwhile, in some sense, to watch reality television even when it serves no further end, I take it that this is not a plausible candidate to exhaust flourishing, since (without saying more about the context or way in which a particular individual pursued this activity) it does not make sense to work, keep healthy, socialise, etc. ultimately for the sake of watching reality television. This, too, is a substantive normative judgement, but again one that many endorse.

One more example of how to apply this criterion of plausibility as an ultimate end will be relevant to what follows. What is a more plausible ultimate end, simply possessing virtue or exercising it, i.e., acting virtuously? Aristotle advocates the latter. Someone could possess virtue while asleep or otherwise incapacitated, in which case they seem not to benefit from it (NE, I.8, 1098b30–1099a7). The benefit of a dispositional state like virtue seems to come from its exercise.Footnote ¹⁶ Aristotle’s answer here strikes me as quite right. There is no point to developing a certain kind of character trait if one never has a chance to act on it. Courage would not benefit me, for instance, if I will never face risk or danger. If we should develop virtues for the sake of exercising them, then virtuous activity is more plausibly an ultimate end than simply possessing virtue, and therefore virtuous activity is a better candidate to be flourishing. I will assume this going forward. Thus, I will only object explicitly to the proposal that flourishing is character-virtuous activity and not merely being virtuous. I do not think a retreat from activity to a state of virtue will help the character eudaimonist overcome the objection to come.

4. The Non-Ultimacy Objection

The main problem for character eudaimonism is that the paradigmatic exercises of character virtues are not plausible as ultimate justificatory ends of action. On the contrary, character-virtuous activities are characteristically justified by the further ends they serve, the benefits they produce. In this section, I will attempt to establish this objection as simply and intuitively as possible. In doing so, I will elide some technical distinctions that might seem crucial to a defender of character eudaimonism. I will introduce these distinctions, anticipate corresponding objections and respond accordingly in the next section. But let us first put the basic argument on the table.

To begin, notice that activities which exercise character virtues fall along a spectrum between those benefitting oneself and those benefitting others – to employ some modern terminology, between ‘prudential’ and ‘moral’ activities, respectively. Paradigmatically prudential activities, which exercise virtues like moderation, include consuming adequate nutrition, physically exercising, saving money and undergoing therapy. Paradigmatically moral activities, which exercise virtues like justice and generosity, include feeding the hungry, donating to charity, keeping one’s promises and generally refraining from harming others for personal gain. Other exercises of virtues can fall somewhere between being fully self- and other-regarding. Exhibiting courage during a heated counter-protest, for example, might lead you to avoid excessive risk to yourself but also to intervene when appropriate to help others avoid harm.

Now, none of the paradigms of prudential, self-benefitting activities is worth doing solely for its own sake. We do not save money as an end in itself. We save money to have money, which in turn is only as good as how we spend it. Similarly, we consume and exercise moderately to maintain bodily health, and we utilise therapy to build or maintain mental health. No one with other options should organise their life around merely consuming, exercising or undergoing therapy.Footnote ¹⁷ And while health might in some sense be good for its own sake, it is, importantly, also good for the valuable things it enables one to do. We would not envy the life of someone who acted only to maintain health and did no more with their health than sit idle.

Neither are paradigmatically moral, other-benefitting activities ends in themselves. First consider negative moral actions (i.e., omissions) like refraining from theft or coercion. These omissions are ‘good’ in the weak sense that not refraining from harming would be positively bad. But they are not activities worth aiming at, which could justify other actions – they are not really activities at all. Even the idler morally refrains from harming others, but this does not turn their inactive life into a flourishing one.

Positive moral actions are not ends in themselves either. If they were, then (per the eudaimonist desideratum for a conception of flourishing described in §3) it would make sense to seek or create opportunities to engage in them: for example, to make promises simply in order to keep them, to hope people will become sick or endangered in order to assist them, or to create injustice for the purpose of justly rectifying it. But we do not and should not hope for or create such opportunities; we perform such moral actions for their effects. Making and keeping a promise is presumably only as valuable as is the promised act plus any additional benefit from building or maintaining mutual trust. The value in assisting someone comes from the benefit received by the assisted party. And distributing justly or generously derives its justification from the value of the relevant kind of distribution. Moreover, the goods typically distributed in just or generous transactions – money, dry goods, health, power and so on – are themselves instrumentally beneficial for what they enable people to do.

The main point here is that both prudential and moral actions are characteristically justified by the further activities which they help people to perform, whether the agents themselves or other people. And these further activities that beneficiaries are enabled to perform had better eventually include some ends in themselves and not just more prudential or moral actions, or else the justifications for these prudential and moral actions would never get off the ground. There would be a valueless regress of agents being prudent for the sake of being prudent or moral, and of agents being moral for the sake of others’ being prudent or moral. We need a more ultimate end, something besides yet more character-virtuous activity, to stop this justificatory regress.

Interestingly, some scholars of Aristotle have noted these points about paradigmatic character-virtuous activities. Hirji argues that ‘[v]irtuous actions are characterized in terms of the good ends or results they aim to achieve’ (Reference Hirji2019, p. 684). Courageous actions aim at safety and security, moderate actions health, generous actions benefits to others and just actions a fair distribution of benefits and harms (p. 679). Whiting similarly acknowledges that ‘virtuous actions typically aim at ends beyond themselves’ (Reference Whiting2002, p. 273). Moreover, Whiting notes that a ‘virtuous circle’, wherein virtuous activity was only valuable for making possible yet more character-virtuous activity, would be problematic. The virtuous circle’s ‘acceptability seems to depend on there being other eudaimonic values (besides virtue) that virtue promotes’ (p. 290). Whiting, then, has correctly identified the feature that poses a problem for character eudaimonism. But though these authors have correctly described Aristotle’s views on the matter, they have failed to appreciate how damning these facts are for character eudaimonism.

Now so far, we have only considered paradigmatic prudential and moral activities expressing the canonical character virtues of justice, moderation and generosity. But the same point about instrumental justification seems to apply to exercises of most other candidate character virtues. Exercising courage at a counter-protest or in a literal battle helps to advance one’s other ends: engaging in necessary risk helps to win the battle or to display solidarity and commitment to the cause. Courageous activity is justified by such ends and not worth doing solely for its own sake. That is, we should not seek battles, heated protests or other risky situations for the sake of exhibiting courage. Other Aristotelian virtues are similar. Exercises of proper anger like speaking out against a colleague’s offensive remarks are useful for changing people’s behaviour for the better or for defending the rights of injured parties. But better for everyone that the colleague, for example, had never engaged in the problematic behaviour in the first place.

Some exercises of Aristotelian character virtues which concern conversation and emotion, rather than action, may not characteristically be justified by the further ends they help to realise. For example, while being witty or friendly may help advance one’s ends in certain situations, it seems good to be witty or friendly even when nothing is to be gained by it. Similarly, it seems good to feel proper shame, not too much or too little, about the right sorts of things, even when nothing results from the feeling. However, while these exercises of character virtues are not obviously instrumentally justified, they still fail the heuristic test for being plausible ultimate ends. It would not make much sense to seek or create opportunities to feel shame to appropriate degrees in appropriate contexts. Neither would it make sense for most of us to organise our lives around exercising friendliness or wit.Footnote ¹⁸ We certainly do not think the point of prudent and moral actions is enabling ourselves and others to be funny or properly ashamed. Even if these are ends in themselves in one sense – the sense of being fitting responses – they are not instances of the ultimate end constituting flourishing and justifying all other actions.

Again, the same points obtain for modern candidate character virtues which Aristotle does not recognise. Exercises of beneficence, kindness and charity are characteristically other-benefitting, and therefore instrumentally justified, in just the same way as exercises of justice and generosity. Industriousness is only as good as the kinds of pursuits industriously engaged in, and many forms of ‘industry’ are only instrumentally justified, if at all. And again, character virtues concerning emotion rather than action – virtues such as hope, compassion, patience or humility – may not be characteristically instrumentally justified, but also do not justify most of the actions that make up a life. Even if such emotions are worth feeling (in the right situations) independently of their effects, it does not make sense to seek or create opportunities to feel hopeful, compassionate, patient or humble. It would be strange, for example, to seek achievements for the sake of being properly humble about them. Nor should we hope for people to experience pain or misfortune for the sake of feeling appropriately compassionate towards them.

If this is all correct and exercises of character virtues are characteristically justified by their effects – or if they are, at any rate, implausible candidates to be the ultimate end which serves to justify all other justified actions – then the exercise of character virtue is a poor candidate to be human flourishing, at least for the purposes of eudaimonism. One could insist, despite all the foregoing, that flourishing consists in character-virtuous activity, but this would divorce flourishing from its role in a eudaimonist ethical theory as the ultimate justificatory end of action. The present objection is a serious problem for character eudaimonism, if not for the character virtue conception of flourishing when separated from eudaimonism.

5. Responses Considered

Let me now clarify what this objection is not and does not depend on. For one, I do not make any claim that virtues themselves are best defined in terms of further ends or goods. I am not, for example, relying on a consequentialist account ‘which would define moral virtues as character traits that systematically produce more actual good than not’ (Driver, Reference Driver2001, p. 68), nor a ‘recursive’ account that ‘equates the virtues with appropriate attitudes to intrinsic goods and evils’ (Hurka, Reference Hurka2001, p. 20). Such views are advanced as competitors to the more Aristotelian accounts of virtues which character eudaimonists tend to accept. My claim is not that states of virtue necessarily relate to independent, ‘primary’ goods such as pleasure, knowledge or achievement. Rather, my claim is that the activities which exercise these virtuous states are justified by further ends. The version of character eudaimonism to which I am objecting holds that it is these virtue-exercising activities which constitute eudaimonia, not the mere possession of virtuous states. On this view, states of virtue are good for the sake of their exercise. I accept this justificatory claim, which is why I am focusing on versions of character eudaimonism with active, not static, virtue conceptions of flourishing, but I say nothing about defining virtues themselves in terms of ‘primary’ goods.

For another, I am not relying on collapsing the two distinctions Korsgaard (Reference Korsgaard1983) recognises: first, the distinction between ‘intrinsic’ and ‘extrinsic’ value deriving from something’s intrinsic and extrinsic properties respectively, and second, the distinction between ‘final’ and ‘instrumental’ value that belong to ends and means respectively. The non-ultimacy objection utilises only the latter distinction. My argument, in these terms, is that the value of flourishing is supposed to be final and not instrumental, but the value of character-virtuous activities is characteristically instrumental, and so flourishing cannot be character-virtuous activity.

This way of putting the argument, however, might mislead. Having final value in this sense is not the same thing as being the ultimate end posited by eudaimonism. The ultimate end – flourishing – is supposed to ultimately justify everything else. This means that it has final value, yes, but also that it lacks instrumental value (except perhaps contingently and accidentally, e.g., if someone offered you money to flourish). Moreover, eudaimonism supposes that everything else of value, even of final value, has instrumental value which eventually terminates in the final value of flourishing.

To avoid conflating ‘final value’ with being an ultimate end, I will henceforth instead refer to it as ‘intrinsic value’, though I mean this solely in contrast to ‘instrumental’ and not in contrast to ‘extrinsic’. In this, I follow Kagan, who defends Korsgaard’s distinctions but opts to ‘continue to use the term “intrinsic value” for value as an end’ on the grounds that alternative terms are ‘potentially misleading in their own ways’ (Reference Kagan1998, p. 290). To clarify, then, something is intrinsically valuable just in case it is an end in itself; something is instrumentally valuable just in case it is justified by further ends.

Bearing this in mind, I also note that I am not relying on an assumption that instrumental value precludes intrinsic value. Kagan (Reference Kagan1998), for one, argues that it is logically coherent for something to have both intrinsic and instrumental value, and indeed for intrinsic value to derive from instrumental value. My non-ultimacy objection is consistent with this possibility. I am not arguing that character-virtuous activity is instrumentally valuable and therefore cannot possibly have intrinsic value. I am instead arguing that character-virtuous activity is instrumentally valuable and therefore cannot plausibly be the ultimate end – for which intrinsic value is necessary but not sufficient.

This discussion nears ongoing debates in Aristotle scholarship, so, at this point, I wish to forestall the suggestion that this body of work has already identified and overcome the non-ultimacy objection against character eudaimonism. As noted above, scholars of Aristotle are aware that character-virtuous activity characteristically serves further ends. But Aristotle also suggests that the exercise of character virtue is choiceworthy or good for its own sake and that the virtuous agent is motivated by this intrinsic value. Some commentators find a tension or outright inconsistency here. How can virtuous actions be both ends in themselves and for the sake of further ends? And what does this mean for the virtuous agent’s motivation in acting? Scholars have attempted to address this interpretive puzzle (e.g., Ackrill, Reference Ackrill1978; Keyt, Reference Keyt, Anton and Preus1983; Whiting, Reference Whiting2002; Meyer, Reference Sauvé Meyer and Vasiliou2016; Hirji, Reference Hirji2018). Solutions tend to distinguish different aspects of character-virtuous activity that bear the different kinds of value. Hirji (Reference Hirji2018), for example, appeals to Aristotle’s distinction between the external act that the virtuous person would perform (described without reference to the agent’s actual motivation), which has instrumental value, and the virtuous (i.e., robustly well-motivated) performance of the act, which has intrinsic value as an exercise of virtue.

However, this interpretive problem is not the non-ultimacy objection that I am advancing. Nor are these scholars’ proposed solutions to the interpretive problem responses to the non-ultimacy objection. The interpreter is attempting to explain and render consistent Aristotle’s position and his various relevant comments. I am happy to grant that there is some such successful explanation. Any explanation proposed in the scholarly literature might do. For my part, I have no objection to a category of goods that are choiceworthy both for their own sakes and for their effects, a category which Plato and Aristotle both recognise (Republic, II, 357a–358a; NE, I.7, 1097a30–34). I am not charging the character eudaimonist with internal inconsistency for placing character-virtuous activity in this category.

Rather, the first charge of the non-ultimacy objection is that it is a substantive mistake to place character-virtuous activity in this category. And this is because, in light of our other, independent normative judgements and commitments, it is implausible that character-virtuous activity is an end in itself. (If the character eudaimonist is guilty of inconsistency, it is inconsistency with these other judgements.) Character-virtuous activity does not appear to possess intrinsic (i.e., final) value, as would be necessary to be a plausible ultimate end. If it did, then there would be at least some normative pressure to desire or seek or create or appreciate opportunities for people to exercise character virtue. But as I have said, this is implausible. For example, the chance to exercise courage is no good reason to court danger and the chance to exercise proper humility about an achievement is no good reason to seek to achieve it. To the contrary, there are good reasons to create conditions of justice, safety and plenty wherein most character-virtuous actions are no longer necessary – and intuitively, the virtuous person most of all will desire these more ideal circumstances and be pained by anything less. The real normative pressure, then, is to eliminate, as much as possible, opportunities to exercise character virtue. And even if our non-ideal, practical reality will always render character-virtuous activity necessary, that activity is merely making the best of a bad situation, not desirable for its own sake. In my opinion, the extant interpretive solutions in the scholarly literature on Aristotle simply do not address these substantive challenges.

Beyond scholars’ attempts to render Aristotle consistent, contemporary philosophers have devised several possible explanations for how character-virtuous activity could have some kind of intrinsic value in addition to its instrumental value and thus belong to the category of mixed goods. Actions can constitute, and not merely cause, valuable ‘results’, and actions themselves can bear value directly.Footnote ¹⁹ It might be intrinsically better to act from virtuous motivational dispositions than to act for other motives, holding fixed the external act, its consequences and, thus, its instrumental value. ‘Backwards-looking’ reasons for action (e.g., that I promised) might often motivate and partly justify some virtuous actions.Footnote ²⁰ As already mentioned, Kagan canvasses the logical possibility that some intrinsic value derives from instrumental value, even suggesting as an example that lives spent helping others could ‘possess intrinsic value by virtue of their instrumental value’ and could thus avoid an infinite justificatory regress (Reference Kagan1998, pp. 288–289).

Again, I have no objection to a category of goods with mixed value. I agree that there are plenty of conceptual tools to make sense of the logical possibility that character-virtuous activity could be both instrumentally and intrinsically valuable. But no version of this response will save character eudaimonism from the non-ultimacy objection, primarily because intrinsic value is merely a necessary condition and not a sufficient one for being a plausible ultimate end. The second and even more conclusive charge of the non-ultimacy objection is that, even supposing that character-virtuous activity has intrinsic value, it also undeniably has instrumental value.Footnote ²¹ And its instrumental value is not accidental to the activity, as it would be if, say, someone offered you money to exercise character virtue. Rather, instrumental value is essential to character-virtuous activity, in the sense of making the activity virtuous in the first place. I mean that part of what makes a given act, e.g., moderate or generous is its contribution to a further aim, e.g., health or benefits to others. (If, for example, a generous person learned that an intended beneficiary would not benefit because their gift would be destroyed en route, then sending the gift would not be generous but pointless, and the generous person would no longer desire to send it now.) This entails that character-virtuous activities are justified, at least in part, by their instrumental value towards other ends. Thus, even if an exercise of character virtue has some intrinsic value, this value provides at most a partial answer to a justificatory ‘why?’ question. A full justification of character-virtuous actions must also appeal to the other aims which those actions realise. The very nature of character-virtuous activity makes it cry out for this instrumental justification, in addition to any intrinsic justification it may have. This is the natural result of locating character-virtuous activity, as we inevitably should, in a category of goods choiceworthy, at least in part, for their effects.

Since character-virtuous activity is characteristically justified by further ends, regardless of whether it also has some intrinsic value, it cannot be the ultimate end of all other actions, that for the sake of which we do everything else that we do. An ultimate end permits no further justification; character-virtuous activity, since it has at least some instrumental value, requires such justification. Since eudaimonism assumes that flourishing is the ultimate end that requires no further justification and serves to justify all other justified actions, the character virtue conception makes for a poor conception of flourishing for the purposes of eudaimonism.

One final response on behalf of the character eudaimonist proposes that flourishing itself has only intrinsic value despite the fact that its sole constituent, character-virtuous activity, has instrumental value. It is logically possible for things to bear properties, including value, distinct from those of their constituents. For example, I possess the property of personhood and whatever value personhood confers, though the molecules that constitute me possess neither. One version of this response holds that flourishing consists not in particular instances of character-virtuous activity but, inspired by Aristotle (NE, I.7 1098a18–20), in the set of all character-virtuous activities performed across a complete life. On this proposal, instances of character-virtuous activities may possess instrumental value, partly for contributing to the valuable whole, but the set possesses only intrinsic and no instrumental value, as flourishing should.

I agree that these are logical possibilities, though I do not think they will save character eudaimonism. Essential to character eudaimonism as I have defined it is a character virtue conception of flourishing, which is the view that flourishing just is the exercise of character virtues (whether taken individually or collectively). I have sometimes employed language such as ‘consists in’ and ‘constitutes’ because these are often more idiomatic, but the view under consideration does not hold that flourishing and character-virtuous activity are separate entities, the one caused by or metaphysically grounded in the other. Rather, the character virtue conception specifies something in the world – character-virtuous activity – that fills the theoretical role of the concept, ‘flourishing’. This relation is more akin to the classical definition of ‘human’ as ‘rational animal’, with no ontological daylight between terms, than to the relation between humans and their constitutive molecules. It was for this reason that I have invoked the ‘finality’ criterion, that whatever flourishing is must plausibly be an ultimate end, not requiring further justification. As we will see in the next section, Aristotle himself originated this ‘finality’ criterion for conceptions of eudaimonia, so this is no straw characterisation of character eudaimonism.

Moreover, I do not think the character eudaimonist should treat flourishing and character-virtuous activity as distinct entities which could bear different kinds of value. For then flourishing would turn out to be something else besides character-virtuous activity. Perhaps character-virtuous activity is a necessary cause or essential ground of this further thing, flourishing, but flourishing itself must then either be analysable in some other terms (e.g., pleasure) or remain an unanalysable, ‘non-natural’ property. Either route departs from conceiving of flourishing as character-virtuous activity. These alternatives are coherent but outside the spirit of character eudaimonism. At any rate, I am satisfied if my argument against a largely unquestioned view forces its adherents to adopt a non-standard interpretation of that view.

As for the ‘complete life’ version of this response, I have not assumed that character virtue conceptions must identify flourishing with individual instances of character-virtuous activity nor with such activities taken collectively. (I have employed plural nouns, ‘exercises’ and ‘activities’, and mass nouns, ‘exercise’ and ‘activity’, that are ambiguous between individual and collective readings.) My goal is to object to character eudaimonism on either kind of character virtue conception. To be sure, I have emphasised the instrumental value of character-virtuous activity by considering particular examples of such activities. And again, I grant the logical possibility that a set of character-virtuous activities lacks instrumental value though its members possess it. But again, this view strikes me as substantively implausible. If individual character-virtuous activities are instrumentally justified by their effects, then these activities taken collectively seem also instrumentally justified by these same effects taken collectively. The collective effects provide reasons to engage in a complete life of virtue beyond the intrinsic value of such a life: if nothing else, the benefits which a virtuous person confers on others serve to instrumentally justify the virtuous person’s lifestyle.Footnote ²² In light of such appearances, denying instrumental value to a complete life of virtue seems unmotivated except to save character eudaimonism at all costs.

But whereas attempts to avoid the non-ultimacy objection appear ad hoc, there are further reasons to reject character eudaimonism. For if character-virtuous activity were the sole ultimate end, then every other intrinsic good would also need to be instrumental to character-virtuous activity. So, other ostensibly intrinsic goods such as loving relationships, cultural or religious observances, aesthetic creations and experiences, athletic achievements, play and intellectual contemplation, for example, would all need to contribute to character-virtuous activity and, thus, be partly instrumentally justified by these contributions. This, too, seems implausible. It is more plausible that we help people so that they have time and resources for such intrinsically rewarding pursuits, not that we engage in such pursuits so that we can better help people. Character eudaimonism gets the intuitive order of justification backwards. To my mind, a more plausible eudaimonist theory will explain how such candidate ultimate ends are themselves ways of flourishing. For reasons of space, however, I leave further discussion of this and other points to other work.