In 2021, Daniel Lapsley and Christian Miller proposed two versions of eliminativism regarding practical wisdom (phronesis). Despite being motivated by different considerations and employing different methodologies, these proposals converged on the claim that the concept of practical wisdom should be removed from our ethical framework as it is an untenable component of the otherwise credible Aristotelian view of virtuous character. This paper responds to these challenges by further advancing the “Aretai model”, an account of practical wisdom that we presented elsewhere (De Caro et al. 2021; Vaccarezza et al. 2023; De Caro et al. forthcoming). In Sect. 1, we illustrate Lapsley’s challenge, a form of soft eliminativism (i.e., reductionist in spirit), which we discuss and criticize in Sect. 2. In Sect. 3, we expose Miller’s hard eliminativist proposal (i.e., one that is non-reductionist in spirit) and his three concerns about practical wisdom as traditionally conceived by Aristotelians. In Sect. 4, we outline the Aretai Model with its three main tenets. Then, in Sect. 5 we respond to Miller’s hard eliminativism by appealing to the Aretai Model. Finally, we argue that if our arguments are correct, the Aretai Model can represent a significant paradigm shift in the way virtue ethics conceptualizes the relationship between practical wisdom and traditional ethical virtues.

1 Daniel Lapsley’s Soft Eliminativism

Daniel Lapsley’s article “The Developmental Science of Phronesis” (2021) makes some very reasonable claims about how philosophy and science should collaborate in the study of phronesis. The general spirit of such an enterprise, Lapsley writes, should be that envisioned by Piaget (1970), i.e., a dialogue “through mutual correction with the common goal of generating empirically responsible moral philosophy and philosophically responsible moral psychology” (140). The result of such a virtuous collaboration would be “theoretically robust, empirically well-attested, and philosophically tenable” (154).

Of course, philosophy and psychology have different roles to play in this joint enterprise. Philosophy has the task of exploring “fundamental human concerns about character, virtue, and how to live well the life that is good for one to live” (139). This inquiry, however, cannot ignore the empirical constraints that determine what human beings can and cannot do: “credible philosophical reflection must begin with the facts of human nature and accord with psychological realism” (140). Moreover, Lapsley notes, “there is an emotional resonance to phronetic moral perception” and “phronesis is importantly constituted by cognitive-affective mechanisms” (144) – and, of course, these issues are for the mind-brain sciences to illuminate.

To illustrate both the potential and the difficulties of the joint endeavor of philosophy and psychology in the study of phronesis, elegantly Lapsley draws on Robert Frost’s famous poem “Mending wall”.Footnote 1 In explaining this analogy, Lapsley notes that “repairing a common wall” is an opportunity “for constructive engagement to achieve a common purpose” and “is not unlike the give-and-take of interdisciplinary engagement across the philosophical and empirical boundary of moral psychology” (140). Here Lapsley clearly – and wisely, in our view – dissociates himself from two common complementary forms of intellectual chauvinism: that claiming (in the spirit of Heidegger 1951) that “science does not think”, and consequently cannot contribute in any way to philosophical inquiry, and that maintaining, conversely, that philosophy has become “historically insignificant” (Dyson 2012) and that, consequently, scientists should ignore whatever philosophers have to say, even regarding (seemingly) common matters. Open-mindedness is called for here: there are issues on which philosophy and science can and should collaborate – and phronesis, Lapsley claims, is one of them.

From this irenic perspective, Lapsley offers some compelling examples of how philosophy and psychology can engage in a “robust dialogue” in the study of phronesis. In this light, he argues that “social-cognitive theory provides constructs and mechanisms to account for perceptual sensitivity and discriminative facility credited to phronesis”; that “social-cognitive approaches to personality and… to moral self-identity and its development link the operations of phronesis to dispositional character”; and that “the development of metacognition and metalogical and meta-rational capacities provides a framework for understanding phronesis as a meta-virtue and as a developmental achievement” (Lapsley 2021: 155).

All these proposals are serious, and we not only agree with them all but have also previously defended some of them in our writings (also under Lapsley’s influence, we are happy to add). In particular, we agree that moral philosophy cannot be done without considering what kind of creatures human beings are (De Caro et al. 2007); what cognitive and emotional capacities they have (De Caro and Marraffa 2016); and what empirical constraints should be placed on philosophical views about ethics, and phronesis in particular, to make them plausible (De Caro, Marraffa, and Vaccarezza 2021) – all questions about which mind-brain science has much to say. In general, we believe that philosophy and psychology should cooperate in the study of phronesis, without either pretending to have priority over the other (De Caro and Vaccarezza 2021).

In some other places, however, Lapsley’s article seems to take a less conciliatory direction. On the one hand, he appears to endorse Owen Flanagan’s (1991) “principle of minimal psychological realism”, according to which possible moral theories must correspond to empirically credible moral psychologies. On the other hand, however, he seems much less enthusiastic about Jason Swartwood’s (2020) complementary principle – the “principle of minimal philosophical adequacy” – according to which psychological theories of phronesis should meet minimally adequate philosophical standards. According to Lapsley, such a principle might require psychologists to “stop doing what [they are] doing” since philosophers would dictate what to do instead – “a division of labor that certainly puts psychology in its place” (141).

In general, Lapsley perceives “palpable tension at the mending wall” and seems to place most of the responsibility for that on the shoulders of philosophers. Thus, he is unhappy both with Jason Swartwood’s (2020) claims that phronesis cannot be measured and that psychologists should consequently refrain from attempting such an unattainable feat, and with Kristján Kristjánsson’s contention that “psychology is rife with incorrect assumptions which must be corrected if it is to be a reliable partner on matters of virtue development” (Lapsley 2021: 141). More radically, at the end of his article, Lapsley appears to suggest that the Piagetian methodology of equal collaboration of philosophy and psychology regarding phronesis, mentioned at the beginning of the article, is only an interim step toward the elimination of phronesis. He writes: “My own approach… aims to understand phronesis in terms of well-attested psychological variable and processes, but where the ascription of phronesis is either optional or superfluous” (144). Thus, while we were originally told that the study of phronesis should involve philosophers and psychologists working together, the expected end result of this process is that phronesis will disappear (along with all the philosophical considerations about it), substituted by psychological constructs – so that, one might say, philosophy is finally put in its place (pace Piaget?).Footnote 2 Given how the article began, this is a surprising conclusion, but one that we suspect is shared by many scholars, both scientists and philosophers. Therefore, since Lapsley – a key figure in the contemporary debate on phronesis and an important source of inspiration for us – presents this claim with crystalline clarity, we will take some time to analyze it and explain, in a collaborative spirit, why we disagree with it.Footnote 3

2 Addressing Lapsley’s Soft-Eliminativist Challenge

Lapsley’s argument for his surprising conclusion has the structure of a dilemma in which both horns are very unpleasant for philosophy. The first horn is one in which the concept of phronesis is translated without residue into the current conceptual apparatus of psychology. In this way, Lapsley observes, the psychological counterparts of the philosophical concept take over and the latter becomes superfluous:

[The first possibility is that] phronesis is absorbed into psychological frameworks with no clear value-added explanatory role other than what is otherwise provided by psychological theory. In this case phronesis becomes something like the “luminiferous ether theory” once held critically necessary to explain the transmission of light until later Einstein’s special relativity found it superfluous (and the Michelson–Morley experiment disproved it). I want to suggest the possibility that the role of phronesis in Aristotelian virtue ethics is much like the role of ether in physics, once thought crucial but now expendable (154).

The second horn of the dilemma concerns the case in which phronesis cannot be translated into any standard psychological concept so that it holds out “as a psychological variable in its own”. However, writes Lapsley,

treating phronesis as a distinctive psychological variable will simply be unpersuasive when compared with extant theory and evidence. It would not be clear just what phronesis is supposed to do or what it would explain, in which case the use of the phronetic lexicon becomes optional, superfluous, or distracting, or a form of special pleading (154).

Let’s consider the two horns of the dilemma in turn, starting with the second. Of course, if phronesis is translated into a specific new psychological variable, this may be at odds with the current state of psychological research. But why should this necessarily be the case – assuming, of course, that the new psychological variable correctly accounts for the behaviors that would be considered phronetic at the moral level? When a scientist proposes the introduction of a new variable (or a new theory) in their field, their colleagues are often skeptical about it: Thomas Kuhn’s The Structure of the Scientific Revolutions (1970) has great pages about this. Over time, however, some of these new variables (or theories) prove their usefulness and are eventually accepted by the entire scientific community. Why should this be impossible in principle for the (hypothetical) phronesis psychological variable?

At any rate, the other horn of Lapsley’s dilemma is the crucial one. According to it, if phronesis can be completely translated into a psychological concept, it becomes useless, as it happened with the “luminiferous ether theory” – the alleged space-filling substance that for 150 years was believed to act as a medium for the transmission of electromagnetic waves – after Albert Michelson and Edward Morley conducted their seminal experiment in 1887. However, the fate of the luminiferous ether concept after Michelson-Morley’s experiment is not a fitting analogy for the purported translatability of the concept of phronesis into psychological language. In fact, that experiment did not translate the concept of the luminiferous ether into a concept acceptable to subsequent physical theory (to paraphrase Lapsley, the concept was not “absorbed into the [physical] framework”); rather, the experiment proved that the theoretical underpinnings of the concept of ether were entirely wrong, revealing that it had no correlate in the physical world and, consequently, should be abandoned altogether. In short, Michelson and Morley proved that the luminiferous ether had to be eliminated from our ontology, while in Lapsley’s scenario phronesis would be reduced to the psychological framework because psychology would offer a better account of its functions than philosophy and common sense, without eliminating them.Footnote 4

Of course, as it is sometimes pointed out, ontological reductionism can, in a sense, be considered a form of eliminativism, as the reduced notion is replaced by the notion into which it is reduced. However, this form of eliminativism is a soft one, as the replaced notion is substituted by an equivalent one so that, in some contexts, one is still entitled to use the original notion. Instead, in the case of hard eliminativism, such as that generated by the Michelson-Morley experiment, the eliminated notion is abandoned altogether, and there are no contexts in which one can continue to use it with any veridical pretense. For example, water is reducible to H20 and the latter could be taken to carry the entire ontological burden but water, as a concept, is not delegitimized; on the contrary, because of the Michelson-Morley experiment, ether was eliminated from our ontology altogether, since it was proved that nothing real corresponded to it. So Lapsley is a soft eliminativist insofar as he asserts the reducibility of phronesis to psychological constructs.

However, there is something more important about this horn of Lapsley’s dilemma. This is the idea that, if phronesis were shown to be completely translatable into psychological constructs, then it would become superfluous. We agree with this. But the crucial points here are why Lapsley finds this idea appealing and whether the idea of the full translatability of phronesis is plausible.

Regarding the why-question, Lapsley offers this justification for the translatability project:

Phronesis contains multitudes, and to wave it at the professions, at educators, at psychologists over the mending wall, without translation into well-attested theories and constructs, will bring pause to much needed interdisciplinary work (154).

If we understand this passage correctly, Lapsley is saying two things: (i) because of its heterogeneity, the concept of phronesis should first be translated into solid psychological constructs before being handed over to professionals and educators; (ii) if the lack of translatability of such a concept were insurmountable, this fact would jeopardize interdisciplinary work between philosophers and psychologists. Miller (2021) has generalized the criticism contained in point (i) by claiming that phronesis, because of its heterogeneity, is a concept that philosophers should abandon altogether (let alone hand it over to professionals and educators). We will address Miller’s criticism in the second part of this article; for now, let’s focus on Lapsley’s second point, which asserts that the translatability of phronesis into psychological constructs is indispensable for enabling interdisciplinary work between philosophers and psychologists.

First of all, upon examining the general history of interdisciplinary collaborations between philosophers and psychologists, it becomes apparent that much of the work, including some of the most significant contributions, has been accomplished without a precise translation (or any) of all philosophical notions into the language of psychology. Take issues such as consciousness, self-identity, free will, intentionality, and moral responsibility: there is certainly no satisfactory psychological definition of these notions (nor even a philosophical one, for that matter), but relevant interdisciplinary work has nonetheless been done anyway – and fruitfully communicated to professionals and educators as a blend of philosophical and psychological notions regarding, say, autonomy, personal identity, phenomenal experience, and leadership. So why not for phronesis?

Let’s then consider the most important question of Lapsley’s dilemma, the whether-question. Is it true that phronesis can (and even should) be translated without residue into psychological constructs? In this regard, we have strong doubts, which are supported by two main arguments: the first (the “Normativity Argument”) considers the status of normative items in general, while the second (the “Particularism Argument”) focuses on the particularistic character of virtue ethics. Let’s analyze these reasons in turn, starting with the Normativity Argument.

Here are three authoritative definitions of phronesis:

  1. 1.

    Aristotle (1999): “[Practical wisdom] is the capacity to deliberate well about what is good and expedient for [a man], not in some particular respect, e.g. about what sorts of thing conduce to health or to strength, but about what sorts of thing conduce to the good life in general” (NE VI, 5, 1140a 25-36).

  2. 2.

    Hursthouse and Pettigrove (2022): “Quite generally, given that good intentions are intentions to act well or ‘do the right thing’, we may say that practical wisdom is the knowledge or understanding that enables its possessor, unlike the nice adolescents, to do just that, in any given situation”.

  3. 3.

    Swartwood (2020: 72): “Practical wisdom is the understanding that enables a person to make reliably excellent decisions about how they ought to live”.

All these definitions of phronesis share a common element – they all incorporate normative moral terms, such as “the good life”, “doing the right thing”, and “how one ought to live” (some also incorporate epistemological normative terms, such as “understanding” and “knowledge”). However, there are good motives to think that a general problem arises with all attempts to reduce normative notions – i.e., those referring to what should be the case – to non-normative ones – i.e., those referring to what is the case. Consider, in this sense, the endless discussions about the alleged reducibility of semantic content, intentional properties, moral statements, and correct reasoning.

More popular, in some quarters, is the hard eliminativist approach to normative notions: this can be observed, for example, in the metaethical attempts by emotivists, sentimentalists, error theorists, and the like who seek to eliminate moral notions from our ontology. Later, in discussing Miller’s view, we will argue that hard eliminativism regarding practical wisdom fails. In any case, as we have seen, hard eliminativism is not the direction that Lapsley advocates, as he leans towards soft eliminativism, that is, reductionism (envisaging the possibility of translating phronesis into psychological constructs). Nevertheless, there are good reasons to think that soft eliminativism about normativity is not a satisfying route. It has been convincingly argued that normative concepts (and, if one is a realist, normative facts as well) have a character that non-normative concepts (and facts) simply do not have.Footnote 5 From this, it follows that all attempts to reduce normative concepts either explicitly or implicitly rely on other normative concepts or simply leave out what characterizes normativity, which, according to our preferred interpretation, is the offering of reasons for acting. (Dancy 2006). Derek Parfit offers an enlightening example in this regard:

Suppose that you are in the top storey of your hotel, and you are terrified of heights. You know that, unless you jump [into a canal], you will soon be overcome by smoke. You might then believe, and tell yourself, that you have decisive reasons to jump, that you should, ought to, and must jump, and that if you don’t jump you would be making a terrible mistake. If these normative beliefs were true, these truths could not possibly be same as, or consist in, some merely natural fact, such as causal and psychological facts (Parfit 2011, II: 327).

In this example, the fundamental gap between the normative reasons you have for jumping into the canal and all the causal and psychological facts concerning you is evident because – however good the reasons for jumping may be – you could always decide not to jump. And this strongly suggests that the normative level (i.e., what you should do) cannot be reduced to the factual level (i.e., what you in fact do).Footnote 6 In this light, the difficulties faced by the project of translating phronesis into psychological constructs should not be surprising, as this is just a specific instance of the extremely problematic project of reducing normative notions to non-normative ones.Footnote 7 Phronesis pertains to what a moral person ought to do, which not necessarily coincides with what one actually does or even to one’s tendencies or habits (habits are indicative of a phronetic character only as long as they contribute to the good life). In short, the Normativity argument indicates that phronesis – being a normative concept, not a descriptive one – cannot be reduced to non-normative notions such as those of cognitive psychology (at least if the latter are intended as intrinsically non-normative, as most psychologists tend to think).

In continuing our case against the reducibility of phronesis to psychological concepts, let’s now move to the Particularism Argument. According to a standard reading of Aristotle’s virtue ethics, which is accepted by most contemporary accounts (including ours), the specific excellence of virtue consists in finding the “right mean”, which is “relative to us” and defined “by reference to reason”, as Aristotle claims (NE II, 7, 1107a1). Reaching the mean in each specific situation is a challenging endeavor. It involves making a judgment to identify the action that represents excellence for a specific agent in a determinate situation. This evaluation depends highly on features that are unique and peculiar to that situation. Such things, as Aristotle has it, “are among particulars, and the judgment depends on perception” (NE II, 9, 1109b 22–23). Phronesis, therefore, cannot be reduced to a decision procedure that issues true judgment about action based on universal standards or criteria. Instead, it is a fine-grained “eye” that assesses the particular contingencies of a specific situation in relation to the equally contingent condition of the agent. What can count as a virtuous action for one agent could easily count as extremely deficient for another agent in similar conditions (as in cases like what would respectively count as a courageous action for an expert firefighter and as such for a non-expert passerby who is in front of the same fire).

All this amounts to the refusal of “ethical generalism” – i.e., it denies that to behave morally one just has to follow general principles (Sherman 1989; Broadie 1991; McDowell 1998).Footnote 8 On the contrary, in most of its expressions virtue ethics accepts “ethical particularism”, the view that assigns chronological and methodological priority to the perception of particulars and – differently from the deontological and the consequentialist approaches – makes ethics an uncodifiable fieldFootnote 9. This, as noted by many authors (Hursthouse 1991; Annas 2003; Hacker-Wright 2010: 220), arguably is the main difference between virtue ethics and its main rivals: being agent-centered rather than act-centered, virtue ethics does not aim to offer action-guidance conceived as a decision procedure. In this light, the uncodifiability of right action, which descends directly from the particularistic nature of phronesis, besides representing a peculiar way to offer a theory of action guidance, also has immediate implications on action and character assessment. What action an agent should perform is uncodifiable in principle, and the only criterion is that they should do “what a virtuous agent would, characteristically, do in the circumstances” (Hursthouse 1999: 79).Footnote 10

The particularistic approach, however, is at odds with a constitutive feature of cognitive psychology (to which Lapsley refers), and of empirical science in general – that of proceeding by the search of laws or, at least, generalizations of the phenomena under investigation. While this is possible for generalist ethical conceptions, it is not for the particularist ones, including virtue ethics. In the latter perspective, by definition, it is impossible to generalize about what agents should do in order to behave morally because, in each particular case, this will essentially depend on what the context and the agent are. Therefore, in a virtue-ethical framework, the project of reducing practical wisdom to a psychological construct appears to be unworkable.

To sum up, the Normativity and Particularism arguments strongly suggest that practical wisdom is not reducible to (to use Lapsley’s language, it is not translatable into) psychological concepts.Footnote 11 In our view, however, the irreducibility/untranslatability should not be interpreted as indicating that phronesis is a problematic concept or an impediment to the collaboration of philosophers and psychologists in its study. Nor should it be perceived as an attempt to put “psychology in its place”, at least no more than philosophy is put in its own place when science delineates the factual constraints that philosophers must respect to produce a conception of phronesis that is “empirically well attested”. As one can read in “Mending wall”, “Good fences make good neighbors”. Making the respective preconditions for cooperation between philosophers and psychologists explicit – sometimes “through mutual corrections”, as suggested by Lapsley – appears to us a necessary condition for its success.

One final issue needs to be addressed here. Differently from not a few contemporary virtue-theorists, we believe that one should refrain from stating that practical wisdom has components. This is because the term “components” suggests that there is a corresponding “compound”, much like, in chemistry, sodium chloride is a compound formed from the components of sodium and chlorine. Indeed, such terminology suggests that the components and the compound can be included in the same conceptual framework – a proposition that we reject in the case of phronesis.

In this regard, some caveats are in order. First, stating that practical wisdom does not have components does not imply that it does not have features, characteristics, or enabling conditions and that the latter cannot be translated into the language of psychology. Indeed, as we have argued elsewhere (De Caro et al. forthcoming) and we will explain in Sect. 4, this is precisely the case with four necessary conditions of practical wisdom: namely, moral perception, moral deliberation, emotion regulation, and moral motivation. However, while these features can be partly translated into psychological constructs, phronesis cannot.Footnote 12 This is because the above-mentioned features of phronesis have a hybrid nature. On the one hand, they belong to our “first nature”, which is our biological endowment and is subject to natural law; on the other hand, they have normative significance insofar as they imply perceptual, deliberative, emotional, and motivational skillfulness within the moral domain. To put it simply, the features of phronesis reflect how one is in fact disposed to do – cognitively and affectively – towards what one should do. This places them in-between the first-natural realm (the descriptive domain of one’s natural capacities and dispositions) and the realm of “second nature” (the normative, and consequently irreducible, “space of reasons”, which is acquired through moral education and concerns what should be done).Footnote 13 This hybrid nature makes the features of phronesis measurable as far as their descriptive dimension is concerned, so that it is possible to assess levels of moral perception, moral deliberation, emotion regulation, and moral motivation of an agent.

The second caveat is that the relationship that connects phronesis to its features is one of supervenience. That is, there cannot be differences in how phronetic two individuals are without there being differences at the level of the features, while there can be differences at the level of the features without any difference at the level of practical wisdom.Footnote 14

The third and final caveat concerns the question of measurability. In this regard, we want to stress two points. First, in our view, practical wisdom cannot be measured but only evaluated. That is, one cannot meaningfully say how much practical wisdom can be predicated of an agent (that is quantities are not at stake here), but one can say that an agent is wiser than another or that, all things considered, an agent is wise enough from a moral point of view. This means that qualitative evaluations are possible and indeed necessary for our moral lives: for example, one can say that Dr. Jekyll is wiser than Mr. Hyde or that helping someone in need is, generally speaking, more praiseworthy than ignoring them.

Second, if phronesis cannot be measured but only evaluated, its features – as said – to a certain extent can. In this regard, one objection may be raised: does not the measure of the features of practical wisdom amount to an indirect measure of practical wisdom itself? In our opinion, the answer to this question should be negative. One can score highest in all features of phronesis but still not act wisely due to akrasia. In other words, there may always be a gap between possessing the four features of phronesis, even at the highest level, and being phronimoi – that is, between having the necessary physiological and psychological conditions for doing the right thing and actually doing so.

3 Christian Miller’s Hard Eliminativism

Taking a different path from Lapsley, Miller (2021) puts forward a more radical eliminativist proposal. He identifies seven functions that Aristotelian ethicists have ascribed to practical wisdom over time: the End-Setting Function, the End-Specification Function, the Instrumental Function, the Justification Function, the Knowing Reasons Function, the Handling Conflicts Function, and the Perception Function. Then, in a subsequent work (Miller 2023), he introduces two additional functions: the Mean Function and the Emotion Regulation Function.

According to Miller (2021: 65), there is no justification for squeezing so many heterogeneous functions into a contrived single intellectual trait, such as practical wisdom. He articulates this view by raising three prima facie very reasonable concerns (2021: 58–59):

The Subsumption Concern What is left of a moral virtue once the various roles of practical wisdom are factored out? There does not seem to be anything else to having a moral virtue besides just having practical wisdom.

The Arbitrariness Concern What justification is there for ascribing certain functions to practical wisdom and not others when it comes to moral virtues, such that the list of functions does not end up being arbitrary or ad hoc?

The Unity Concern If there are multiple functions ascribed to practical wisdom, why [should one] think that they would all be carried out by a single character trait, given how diverse the functions tend to be?

From these premises, Miller draws a conclusion that, unlike Lapsley’s, lacks a reductionist character since ontologically it is a hard-eliminativist one. In fact, while Miller acknowledges that the reference to phronesis may hold practical value – as “a matter of convenient labeling” for the aforementioned functions –, he also posits that phronesis itself lacks any psychological correlate and, consequently, it simply does not belong, under any description, to the furniture of the world. In Miller’s proposal, the Michelson-Morley analogy is fitting: phronesis is akin to ether since there is no basis to posit “a distinct character trait of practical wisdom above and beyond the various capacities on the list. On metaphysical grounds, practical wisdom does not exist” (Miller 2021: 66). Therefore, Miller argues, virtue theory should shift its focus from the investigation of (non-existing) practical wisdom to identifying the distinct individual traits that can successfully explain how the functions traditionally attributed to phronesis are carried out.

When attempting to respond to Miller’s challenge, the first task is to clarify why practical wisdom goes beyond being merely a convenient label for a collection of independent functions. Additionally, one has to explain what these functions precisely are and understand their nature. An interesting proposal in this direction has been advanced by Darnell et al. (2019: 2). They have developed an account aimed at “appropriat[ing] Aristotle’s theory of practical wisdom for the contemporary world” by articulating “a four-function model that comprises a psychologically realistic, intellectual meta-virtue that guides excellent or virtuous action” (see also Darnell et al. 2022). This model may be used to challenge Miller’s eliminativism regarding phronesis by advocating for a perspective that is broadly Aristotelian yet informed and tested empirically. According to this account, phronesis can be understood via a four components model, which includes a constitutive function, i.e., the ability to perceive the salient ethical features of a given situation; an integrative function, which “involves integrating different components of a good life”; a blueprint function, i.e., a general conception of the good life, which includes at least a sketchy knowledge of “the place that different goods occupy in the larger context and how they interact with other goods”; and an emotional regulation function, which brings emotions in line with one’s moral judgment. Building on this model, Kristjánsson and Fowers (2022: 1) claim they can address both Miller’s and Lapsley’s eliminativist challenges defending phronesis as a “composite, yet integral capacity”. Their main line of defense relies on an analogy between phronesis and the decathlon, which, despite consisting of a sum of different athletic disciplines, is irreducible to them: not “everything that the decathlon theorists want to say about decathlon can be said in terms of the theories and constructs already available for the individual sports” (15). In Kristjánsson and Fowers’s opinion, this analogy,Footnote 15 combined with an appeal to ordinary language and with empirical evidence on the measurability of phronesis, should, at least, “shift the burden of proof from those who wish to defend and elaborate upon phronesis to those who want to challenge it” (Kristjánsson and Fowers 2022: 15).

While this is an interesting proposal, partly similar to ours, we do not fully endorse it for several reasons. First, it assumes the possibility of measuring not only the features of phronesis but phronesis itself and, as said, we harbor strong doubts about this claim.Footnote 16 Second, we partially disagree with Kristjánsson and Fowers’s four-components model regarding the assertion that phronesis is constituted of those alleged components.Footnote 17 Third, while we find the analogy with the decathlon brilliant, it presupposes that virtues are ontologically autonomous from phronesis; thus, it cannot be used to criticize views that refuse this claim, such as the Aretai model.Footnote 18 Finally, this analogy, while enlightening some features of phronesis, overlooks a very important one – its normative character. In our opinion, the normative character is an essential component of phronesis, and the meaning of this notion cannot be fully appreciated without considering it.

A more compelling comparison for phronesis, from this point of view, is with the work of critics, such as wine experts.Footnote 19 These individuals possess various abilities, including the capability to blindly identify the wines and discern their respective olfactory and gustatory properties. These abilities result from a combination of individual talent and proper training in wine tasting (Smith 2007). The exercise of these abilities is a prerequisite for the reliability of critics’ judgments regarding the quality of different wines. It is worth noting that, similarly to the situation with the decathlon analogy, the capacity for judgment in wine critics depends – more precisely, it supervenes – on the possession of some features (i.e., abilities), but it is not reducible to those capacities. However, unlike the decathlon, the judgmental capacity of wine experts belongs to a different logical space from that of the practical experts’ abilities – that is, the space of normativity.Footnote 20 It is also important to note that the enologists’ specific expertise can be applied in various ways: comparing different wines, suggesting how a particular vintage can be improved, detailing the olfactive and gustatory properties of, say, a Burgundy or a Barolo, and so forth. The capacity remains the same, but its manifestations can vary. And this, according to the view that we defend, is exactly what happens with respect to practical wisdom and the specific individual virtues that instantiate it.

In the next section, we will address Miller’s hard-eliminativist proposal by proposing a different model from Kristjánsson and Fowers’s (2022, 2024) that insists on the normative character of practical wisdom, conceives of it as virtuousness in general, and characterizes it as ethical expertise. We believe that this approach may help identify the functions of phronesis without losing sight of its normative role. Furthermore, through this reconceptualization, our aim is to demonstrate that not only can eliminativism be rebutted, but it can also be overturned. In other words, we propose to transcend the standard dualistic paradigm in virtue ethics by asserting that phronesis as ethical expertise constitutes the sole core virtue, with the more specific virtues being mere manifestations of it.

4 Redefining Phronesis: The Aretai Model

The Aretai model of phronesis, which we defend, provides in our view a robust response to Miller’s hard-eliminativist challenge.Footnote 21 Before considering how it can be applied to respond to the challenge, let’s briefly outline the key features of this model. In essence, the Aretai model pivots on three main tenets:

  1. (i)

    Virtue monism: ontologically, phronesis is the only moral virtue;Footnote 22

  2. (ii)

    Virtue molecularism: phronesis manifests itself through clusters of individual moral virtues;

  3. (iii)

    Phronesis is a particular kind of expertise (De Caro et al. forthcoming).

In this section we will briefly explore each of these ideas in turn; however, since we have argued extensively for the first two tenets of the Aretai model elsewhere, here we will focus more on the third.

According to the Aretai model, the possession of practical wisdom constitutes virtue in its entirety, transcending the diversity of situations and enabling an understanding of virtuous conduct across diverse practical scenarios (De Caro et al. 2021; De Caro et al. forthcoming). In this light, practical wisdom ensures both consistency and adaptability across different contexts, allowing for appropriate responses to a range of practical challenges (including those that are new or unfamiliar), all while fostering moral motivation. The Aretai model, therefore, encompasses virtue monism insofar as it contends that, ontologically, practical wisdom constitutes all the traditional virtues in their entirety – that is, it is their ratio essendi. From an epistemological point of view, however, practical wisdom can be identified only as long it manifests itself in the shape of the traditional ethical virtues. Hence, those virtues, serve as means to recognize practical wisdom, i.e., they are its ratio cognoscendi. Remember the analogy with the enologists: their expertise can be used in different contexts and recognized via different manifestations, but it remains one and the same.

Furthermore, De Caro and Vaccarezza (2020) have argued that acknowledging that an agent has performed a virtuous action requires, at least provisionally, ascribing them a significant amount of practical wisdom, according to the so-called “Principle of Phronetic Charity”. This assignment is then subject to confirmation (or disconfirmation) – which are qualitative evaluations, not quantitative measures – through the observation of that agent’s subsequent moral actions. Navarini et al. (2021) have termed this process “phronetic abduction”, describing it as an “Inference to the Best Moral Explanation” that occurs when encountering an instance of virtuous behavior in someone’s relational life. In Peircean terms, observing an agent’s virtuous action leads us to recognize that, since if someone is practically wise, they would behave in that manner, then that agent may be (provisionally) interpreted as wise. Attributing at least a minimal level of practical wisdom to our counterpart can thus be viewed as an abductive hypothesis.

The Aretai model refuses both “virtue atomism”, which posits that moral virtues operate as ontologically distinct global character traits, and “virtue holism”, which characterizes the “unity of virtue” thesis, according to which virtuousness must be conceived as the possession of all virtues simultaneously (De Caro et al. 2021). Instead, the Aretai model advocates for virtue molecularism. Contrary to atomism, virtue molecularism dismisses the notion that a person can be deemed wise by only possessing one moral virtue. For example, if an individual is courageous but lacks all other virtues, it would be legalistic to label that person as virtuous; rather, they should be considered immoral or at least amoral overall. Contrary to holism, virtue molecularism contends that one can still be virtuous even if they only possess some virtues, so acknowledging the obvious empirical fact that no individual is virtuous in all aspects. In this light, phronesis can be expressed to varying degrees across different domains – not solely in one, nor necessarily in all of them – where it manifests as a specific moral virtue. In essence, virtues are situated instances of an agent’s underlying practical wisdom and they come in clusters.

In a personal communication, Kristján Kristjánsson has raised an interesting objection against our virtue monism. In his view, from the ontogenetic point of view, individual virtues comes much earlier than phronesis:

Up to the age of approximately 8–10… the virtues develop fairly separately and conflicts between them are mostly overlooked or resolved unsystematically/haphazardly. Obviously, children need an intellectual virtue to know how to express an individual virtue (gratitude to their grandmother, say) but deinotes (calculation) can serve that function well.

In our view, however, what children really learn is how to be virtuous in general, that is, how to perceive, reason and be motivated morally, and how to regulate their emotions. In brief, children are phronetic agents in fieri – even if, just as we do with the adults, we tend to interpret their embryonic virtuous actions in terms of the individual virtues. What the apparent fragmentary nature of their virtues indicates, rather than an ontological independence of the virtues from one another, is the limited scope of their practical wisdom, that is, of their ethical expertise. In this regard, we should also remember that we all are always phronetic agents in fieri since none of us is really a phronimos: in fact, alas, even the best of us, sometimes act unvirtuously.

Finally, the Aretai model offers an expertise account of phronesis (and of all the traditional virtues through which it manifests itself). In Book 6 of the Nicomachean Ethics, Aristotle implicitly likens the possession of phronesis to “having an eye”, thereby equating it to a form of expertise: “The wise men, because they have an eye formed by experience, see correctly” (NE VI, 11, 1143b14). However, a question arises regarding how we can credibly account for this claim in contemporary terms. A response that has been gaining traction in recent years among philosophers and psychologists supports the identification of practical wisdom with ethical expertise.Footnote 23 One of the most interesting philosophical accounts of wisdom as expertise has been offered by Jason Swartwood and Valerie Tiberius (Swartwood 2013; Swartwood and Tiberius 2019). Their approach is based on the Recognition-Primed Decision (RPD) model, the foremost model of real-world expert decision-making skill, along with broader research on expertise. Their objective is to provide a conceptual framework for understanding practical wisdom as an expert decision-making skill in domains involving complex choices. More specifically, according to the RPD model, understanding how to conduct oneself in domain D is:

(a) an ability to identify (accurately, non-accidentally, and in a wide range of situations in D) what features in a situation require what response in order to achieve the goals of D, and, when there are internal obstacles to carrying out that response, (b) an ability to identify how to overcome those internal obstacles. Understanding how to conduct oneself can thus be described as an ability (Swartwood 2013: 515).

Swartwood’s strategy is, our view, very promising. It consists in arguing that wisdom can be an ability of such kind, and in particular a kind of ability in the domain of all-things-considered decisions, which, in turn, comprises a complex set of sub-abilities. More recently, Chen-Hung Tsai (2019, 2023) has challenged the argument of the asymmetry between wisdom and skill, which appeals to the idea that “a person with a particular practical skill cannot deliberate about the end that the skill is being used to pursue, and, even if she can, she is not required to do so” (2019: 3). Moving away from Stitcher Building on Swartwood’s account, Tsai maintains that the connection between phronesis and other skills is extensive. He contends that the fast, system-one decision-making and acting processes exhibited by the expert – any expert – are in fact always oriented towards the end and involve some level of deliberation.4 Indeed, in Tsai’s account this would explain why both the moral and any other expert are more likely to do the right thing when faced with unfamiliar situations. Rather than being identical or analogous to general skills, practical wisdom, in his perspective, is a species of practical skill. It shares with any skill, among other traits, the fact that “a further specification of the fixed but broadly specified end […]is relevant to the success of achieving the end” (2019: 10).

The Aretai model incorporates a view that aligns with the family of practical wisdom-as-expertise accounts. However, beyond conceptualizing practical wisdom as expertise in a distinctive way, this model takes a step forward in terms of the consequences of such reconceptualization for the traditional image of the virtuous agent and the relationship between practical wisdom and individual ethical virtues. In particular, it distinguishes itself from traditional perspectives by eschewing reliance on individual ethical virtues, conventionally seen as distinct character traits in terms of their ontology, psychology, and development. Instead, according to our view, achieving the kind of excellence associated with practical wisdom can be understood as a specific form of ethical expertise. This expertise involves the cultivation, application, and demonstration of four features (moral perception, moral deliberation, emotional regulation, and moral motivation), which are skills essential for navigating morally significant situations, and normally are ascribed to individual ethical virtues.Footnote 24 According to the Aretai model, these skills are not compartmentalized into different ethical domains or explained by the possession of context-specific ethical traits. Instead, in light of virtue monism, all these aspects depend on the possession of general ethical expertise, which is consistently exercised and applied throughout the moral domain.

As said, the Aretai model asserts that practical wisdom should be regarded as a form of ethical expertise that transcends specific situations. This entails the consistent demonstration of excellence in the ethical capacities mentioned above across various circumstances that demand comprehensive decisions about what one should do. We contend that an essential feature of practical wisdom lies in the capacities of sound judgment, thoughtful deliberation, and effective action in morally relevant situations. Importantly, this capacity extends beyond limited instances, requiring consistency and reliability across a broad set of situations and contexts, even those that are unfamiliar and intricate. In essence, ethical expertise should be transferable across different situations. The more extensive and diverse the range of situations an individual can effectively navigate, the more phronetic or ethically expert that individual is likely to be (De Caro et al. 2021). Hence, our definition of practical wisdom as ethical expertise is as follows:

An agent is an expert in the domain of morality with respect to a specific group if that agent possesses outstanding moral perception, moral deliberation, emotion regulation, and moral motivation in all-things-considered decisions across different situations compared to others in that group.

However, within the virtue-ethical field, there still is some reluctance to accept the interpretation of practical wisdom as expertise. Various concerns have been raised against the idea that practical wisdom – rather than individual ethical virtues – can be conceptualized as a skill or expertise. Three are the main reasons for skepticism. The first reason pertains to the domain. It is argued that, while skills are domain-relative, practical wisdom is general in scope; hence it cannot be categorized as any kind of skill or expertise (Hacker-Wright 2015). Secondly, it is suggested that, unlike skills, practical wisdom requires “a correct conception of worthwhile ends” (Hacker-Wright 2015: 986; see also Stichter 2016: 211). Finally, echoing Aristotle’s distinction between skill (techne) and practical wisdom, it is noted that someone can possess a skill without being motivated to act on it, a circumstance incompatible with wisdom, which inherently involves motivation at its core (Kekes 1995: 30; Zagzebski 1996; Stalnaker 2010: 408).

As for the first objection, our response hinges on treating the domain of all-things-considered decisions as a peculiar – no matter how broad – domain. While precisely delineating the boundaries of the moral realm remains a challenging task (Gert 2020; Machery and Stich 2022), certain characteristics can be identified that distinguish morally relevant situations from those that are not. It is widely accepted that morality entails a distinct and non-instrumental concern for the well-being or harm of those affected, whether actual or potential (Railton 2017), as well as for how benefits and burdens are distributed in social interactions (Brink 1989). Moreover, many moral questions and choices have straightforward answers that can be reasonably ascertained. However, in more contentious matters, principles, norms, and values clash, creating ambiguity regarding the optimal solution. In this regard, we claim that individuals with practical wisdom excel in making decisions of this nature. More specifically, we argue that those with practical wisdom outperform others in situations that demand comprehensive consideration of various factors and lack clear or absolute objectives – unlike endeavors like chess, where the sole objective is winning.Footnote 25 Critics of the practical wisdom as an ethical expertise model (and of ethical expertise in general) argue that the primary challenge lies in establishing precise normative criteria for all-encompassing decisions. Some even suggest that defining success conditions for such decisions is a futile endeavor (Hursthouse 1999; Swartwood 2020: 79–81, 90) since appropriate responses to specific situations, when considering all factors, are often highly context-dependent and genuinely contentious (McGrath 2008). While we acknowledge that there is no straightforward set of principles or rules that can easily dictate what should be done in every all-encompassing decision (Hursthouse 1999: 56), we maintain that elevated levels of development and proficiency in a few fundamental ethical skills can enable individuals to respond, reason, and act more effectively in morally relevant situations. This can be achieved by promoting less biased consideration of the interests at stake, enhancing deliberative abilities, clarifying objectives, and fostering better motives.

As for the second objection (that, unlike skills, phronesis requires well-defined ends), we assert, in line with research on non-moral expertise, that practical wisdom not only presupposes higher levels of moral perception and intuitiveness but also highly competent moral deliberation. Deliberation, in turn, implies the ability to represent different goals, outcomes, and courses of action, as well as the reasons supporting them, and to balance their relative worth. Moral deliberation stands out as qualitatively different from other types of reasoning, demanding skills not involved in other types of expertise (Hacker-Wright 2015; Stichter 2016). Above all, this kind of deliberation requires a critical evaluation and reflective comparison of ends – not just the means to achieve already established goals – in light of a broader conception of the good life.

Finally, there is the third objection according to which, while one can have a skill without being motivated to act on it, wisdom inherently motivates. We grant that one cannot possess practical wisdom without being motivated to act on it; however, we do not see this as an objection to interpreting practical wisdom as ethical expertise. Many influential models of expertise include motivational components (Dreyfus and Dreyfus 1991; Swartwood and Tiberius 2019). Accordingly, for the Aretai model ethical expertise does not merely demand dry reflection on what makes possible ends worth (or worthless) pursuing. Wise agents tend to act consistently with what they consider good or right, without encountering significant motivational obstacles preventing them from doing what they believe should be done.

5 Addressing Miller’s Hard Eliminativism

The Aretai model has the resources, we believe, to address Miller’s concerns about phronesis on which his eliminativist challenge is based. Let us start with his first concern:

The Arbitrariness Concern: What justification is there for ascribing certain functions to practical wisdom and not others, when it comes to the moral virtues, such that the list of functions does not end up being arbitrary or ad hoc? (Miller 2021: 58).

According to the Aretai model, it is possible – and indeed appropriate – to assign non-arbitrarily multiple functions to the single construct of ethical expertise. This does not mean that all the seven functions identified by Miller are necessarily correctFootnote 26: as said, our model identifies four fundamental capacities required for being an ethical expert. However, if this understanding holds true, the Arbitrariness Concern loses its ground.

The arbitrariness concern has been addressed at some length by other accounts of practical wisdom, such as the Standard model, and our response to this issue – despite diverging on the correct list of non-arbitrary features of practical wisdom – follows the same argumentative line. However, the Aretai model diverges from the more standard approaches when addressing Miller’s Subsumption Concern due to another of its key tenets – virtue monism. Here is again how Miller presents this concern:

The Subsumption Concern: What is left of a moral virtue once the various roles of practical wisdom are factored out? There does not seem to be anything else to having a moral virtue besides just having practical wisdom (Miller 2021: 58).

In this case, instead of attempting to rebut Miller’s concern, we are pleased to endorse it. Nothing is left besides practical wisdom, that is, no specific virtue exists independently. Therefore, the nature of the traditional virtues should be reconceived accordingly. More specifically, if the individual virtues serve as practical wisdom’s ratio cognoscendi, as we believe, one should consider them as highly useful labels that render practical wisdom visible to observers and can provide ex post practically valuable accounts of phronetic actions. In this regard, it is important to note that, even though from the ontological point of view we are advancing a monist account of virtuousness, we are not discarding the language of the virtues altogether. However, what is crucial to our account, as opposed to most rival interpretations of practical wisdom, is that the specific virtues enter the picture only in the post-factum phase of description/evaluation/justification of morally relevant actions. Be the virtuous action the product of intuitions or of reflexive, slower processes of deliberation, generally it is not the case that the practically wise operates a “choice” between competing virtues (“Should I be generous or fair in this context?”). Rather, the wise perceives or deliberates about what among the options at stake is the overall virtuous one. As Plato’s Protagoras has it, when we label someone as, say, courageous, what we really admire is one’s being wise (i.e., ethically expert) in the field of danger and fear, broadly conceived. Therefore, courage is the name of wisdom in that particular domain (356d–7b).

However, as mentioned, when we retrospectively seek to explain virtuous behavior – to others or to ourselves – virtues become relevant and useful. In fact, being “thick terms”,Footnote 27 they offer an efficient way to describe and account for the morality of actions, combining descriptive and evaluative elements. It is also important to emphasize that we do not view this process as a mere confabulatory rationalization; instead, it can provide accurate descriptions and valid justifications, making actions more intelligible to the observers (and potentially to the agents themselves). In summary, reasoning in reference to the specific virtues enables both the wise and their interlocutors to account for the morality of actions after they have been performed, rather than aiding in the decision-making process of how one should behave.

In a subsequent article (2023), Miller himself acknowledged that, unlike the Standard Model, the Aretai model effectively addresses his Subsumption and Arbitrariness concerns. However, he firmly contends that the Aretai Model lacks the resources to tackle his Unity concern.Footnote 28 Here it is again:

The Unity Concern: If there are multiple functions ascribed to practical wisdom, why [should one] think that they would all be carried out by a single character trait, given how diverse the functions tend to be? (Miller 2021: 59).

Furthermore, Miller (2023) vehemently claims that “what we get is a monstrously large and disjointed trait that is constituted by an incredibly diverse range of capacities. So here too we have a Unity Concern. Why think there is one trait constituted by all these capacities? And what unifies them all into this one trait?” (11). Summarizing, Miller argues that the Aretai Model is plagued by two Unity concerns: one regarding the multiple functions (or features) of practical wisdom, the other concerning the individual virtues. This is an astute objection, and we are pleased to address it since this allows us to clarify some fundamental issues regarding our model.

Let’s first discuss Miller’s Unity concern in regard to the constitutive features of practical wisdom. The first thing to emphasize in this respect is that in the same spirit in which the Unity Concern is raised against practical wisdom, it might be raised against other significant notions commonly employed to comprehend human behavior. Consider, for example, the notion of cognition. It covers a vast array of abilities, that include perception, attention, imagination, memory, judgment, reasoning, computing, problem-solving, decision-making, and understanding. Such a variety might invite to raise a “Unity concern” against the notion of cognition; but this concern would be unjustified. Situations exist in which it is advantageous and perfectly legitimate to talk of cognition in general, even though its different versions may appear significantly disparate. The same happens with practical wisdom.

Furthermore, as discussed earlier in relation to Lapsley’s view, the Aretai Model does not conceptualize practical wisdom as a trait directly investigable with quantitative measurements. Instead, it is seen as a construct supervening on the set of four features – moral perception, moral deliberation, emotion regulation, and moral motivation – that (i) characterize wise agents, (ii) have a psychological dimension, and (iii) (in principle) are empirically evaluable. As a result, it is not possible to pinpoint ethical expertise in itself within the agent’s psychological structure: in fact, this feature can only be identified at the level of second nature when attributing to others or to oneself the capacity to make complex moral decisions that require specific but diverse abilities. Also, the four features should not be seen as separate parts of practical wisdom nor as particular roles or functions played by it. Instead, they are constitutive conditions that must be met for an agent to be wise, that is, an ethical expert; and thus, Miller’s Unity concern regarding the features of practical wisdom is effectively addressed.

However, possibly anticipating this response, Miller (2023) raises another interesting objection. He refers to two possible interpretations of the relationship between practical wisdom and the sets of capacities that carry out the functions associated with it. One interpretation is reductive in character, viewing practical wisdom as identical to the capacities and, therefore, carrying no specific causal or explanatory weight. The other interpretation is non-reductive in character, as it does not conceive of practical wisdom as identical to the capacities but rather as constituted by them. In this reading, practical wisdom is expected “to do real causal and explanatory work above and beyond that done by its constitutive capacities”. We sympathize with the second interpretation; however, Miller (2023: 201) favors the first for two reasons: first, because it is “more parsimonious, avoiding having to posit a distinct metaphysically real trait”; second, because “it also avoids problems associated with causal overdetermination”.

The concern of parsimony is one with which we strongly disagree. At stake here is the autonomy of normative items in relation to non-normative ones. Nothing less than the legitimacy of the space of reasons, to which, in our view, certainly practical wisdom belongs, is at stake here, in contrast to the space accounted for by the nomological explanations of the natural sciences. Thus, we have no problem in granting that phronesis is ontologically autonomous from its constitutive features. However, the other concern raised by Miller – that regarding causation – is much more complicated as it delves into deep issues such as mental causation, the nature of causality, and even free will.

We intend to revisit this fundamental theme in future work; for now, then, we will only offer some basic ideas. First, it should be noted that all models of practical wisdom, not just ours, have deep problems with the issue of causation. The eliminativist model, in particular, deprives moral features of any causal (and even explanatory) relevance since, within that framework, all the causal work is done at the physical level: and this move is very contentious since, according to most conceptions, moral responsibility requires causation or, at least, is grounded in it (Sartorio 2007). Second, we can usefully return to the analogy between practical wisdom and wine expertise. At the level of description in which wine expertise is explanatorily relevant, there is no problem in saying that enologists’ general expertise is causally relevant in their professional actions. The problems only arise when one attempts to provide a general metaphysical account of the situation, seeking a unified account of the different levels of description, such as the mental to the physical. The same can be said of the issue of practical wisdom: when one confines oneself to the space of reasons – without entering into the metaphysical game of trying to offer a comprehensive synoptic view of the different levels of descriptions –, one can continue talking about the causal power of practical wisdom without particular problems. Finally, concerning the metaphysical problem itself (notoriously one of the harshest in philosophy), some views of causality, ranging from causal pluralism (Cole 2019; Martinez 2023) to interventionism (Woodward 2015), have been recently developed that seem to shed light on the causal role that mental items can play in our actions without encountering traditional problems such as causal overdetermination and the violation of the causal closure of the physical world.

Let’s now consider the concern regarding the unification of the individual moral virtues. As said, according to the Aretai model, individual virtues appear in moral behavior as manifestations of practical wisdom – which, for us, coincides with virtuousness. In this light, individual virtues exist at the same ontological level of practical wisdom since they represent instances of practical wisdom in specific domains and under specific circumstances; that is, they exist at the moral level, within the space of reasons, that is different from the physical and psychological levels. At the psychological level, instead, we find the four features that are necessary to phronesis (moral perception, deliberation, emotion regulation, and motivation); and, according to the Aretai model, practical reason supervenes on these four features. Thus, while the individual virtues simply are practical wisdom (since they are expressions of it), the four features are what make phronesis possible.

In this light, we can also address Miller’s concern about the unification of moral virtues under practical wisdom: there is, in fact, nothing to be unified because they are one and the same from the beginning – analogously to what happens with cognition. We believe that the Aretai model does not make the understanding of the relationship between practical wisdom and the individual virtues, or that among the virtues themselves, more difficult but easier. They all are one thing; thus, there is no need to look for a mysterious connection that links them all.

6 Conclusion

In our view, the Aretai model not only has the potential to refute soft and hard forms of eliminativism but, more radically, it stands to represent a paradigm shift in virtue ethics. Instead of adhering to a dualistic model that conceives of character as composed of two discrete categories of traits – namely, the traditional ethical virtues, on the one hand, and phronesis, on the other hand –, we advocate for an integrated view of the virtuous character. According to this outlook, the virtuous character is characterized by the possession of general ethical expertise (phronesis) as an all-encompassing and adaptable skill that consistently applies across diverse situations. In this light, the manifestation of virtuous behavior across different contexts can be ascribed to the existence of ethical expertise, encompassing a specialized set of skills (moral perception, deliberation, emotional regulation, and motivation), commonly observed in individuals with practical wisdom. We have concluded that character dualism should be replaced by an approach that acknowledges practical wisdom as the primary virtue governing ethical life. In this view, traditional ethical virtues serve as useful labels for describing how practical wisdom manifests in various situations and domains. Moreover, possessing practical wisdom is deemed not only necessary for virtuous behavior but also sufficient.

However, Aretai model does not dismiss the utility of discussing virtues; however, it strongly suggests that we should cease conceiving of them as ontologically, conceptually, and motivationally independent traits. Ethical virtue is one thing.