The Normative Complexity of Virtues

Abstract

On what I will call the standard view, the distinction between the moral and the epistemic realms is both psychologically and conceptually prior to the distinction between any two given virtues. This widespread view supports the claim that there are moral and intellectual (or epistemic) virtues. Call this the fundamental distinction. In this paper, I raise some questions for both the standard view and the fundamental distinction, and I propose an alternative view on which virtues regain priority over the moral/epistemic divide. I suggest understanding them as normatively complex, distinctive sensitivities to both theoretical and practical reasons.

1. Introduction

Most claims about virtues are widely debated. Philosophers disagree on whether virtues are character traits [1,2,3,4,5,6], dispositions [7], faculties [8,9,10], or skills [11]. They discuss whether virtues’ normativity is basic or derivable from norms or consequences [12,13], and whether virtues must always come together [14,15,16] or can sometimes be possessed in isolation [17,18], or in clusters [19].

Compared to these largely disputed claims, there is one thesis that is strikingly accepted by most authors, namely that there are moral virtues and intellectual (or epistemic) ones. The former include benevolence, generosity, courage, kindness, honesty, trustworthiness, humility, and the latter open-mindedness, curiosity, attentiveness, intellectual humility, intellectual courage. Call this the fundamental distinction.

Although there are lively debates on exactly what determines a trait’s membership to one or the other group, e.g., their outcome, triggering stimulus, or function [20], and on the exact nature of specific traits [21,22,23,24], the fundamental distinction currently stands for the starting point of an increasing number of discussions, ranging across not only ethics and epistemology but also the history of philosophy, philosophy of education, philosophy of economics, cognitive sciences, and so forth [25,26,27,28,29,30,31]. Crucially, it is also the very pillar underlying the literature on virtue and vice epistemology: it is because “human beings have character traits, and [because] some of these traits are intellectual character traits” [32] p. 159 that a great number of questions emerge, which demand close philosophical scrutiny [33,34,35,36,37,38,39,40]¹.

By claiming that the fundamental distinction is largely assumed as a legitimate starting point, I do not mean to suggest that there are no notable exceptions². Despite being often invoked as an illustrious researcher into the nature and role of intellectual virtues, for instance, Zagzebski herself suggests that epistemic traits are a subset of the moral [6,41]. Similarly, Coady has argued that epistemology, properly understood, is a branch of ethics, and hence that intellectual virtues are also moral ones [42] pp. 23–24. Along similar lines, Hookway stresses the importance of conscientiousness in epistemology and the centrality of inquiry over justification [3], where inquiry cannot be assessed through a purely epistemological lens. Fricker talks of the virtue of “reflexive critical openness” to the word of others as a moral and intellectual trait [43], and Gardiner argues for the existence of attentional virtues and vices to then maintain that they have epistemic, moral, social, and political importance [44].

But accounts of this sort do not falsify my contention that the fundamental distinction is widely assumed across different debates, and that determining the exact relationships between moral and epistemic virtues and vices is currently considered less pressing an issue than settling specific questions within each debate. In this paper, I aim to resist this tendency by reflecting on whether, at a closer look, the fundamental distinction really is as solid as it is typically taken to be.

More precisely, I suggest that the fundamental distinction is part of a larger, and as widely held, account that I will call the standard view. On the standard view, the distinction between the moral and the epistemic realms is prior to the distinction between any two given virtues. That is: we start from an understanding of what it is to be a moral or an epistemic something, we compile a list of traits we recognise to be virtues, and we proceed to sort them into moral and epistemic categories. On a weaker version, the standard view describes our practices—how we learn and master virtue concepts, and how we employ them to assess ourselves and others. On this version of the view, our ordinary understanding of the distinction between normative realms is psychologically prior to our understanding of virtues. On a stronger version, the priority is conceptual and explanatory: a virtue is what it is in virtue of being such that it fits one or the other normative category. The distinction between the epistemic and the moral explains what a given virtue is.

In this paper, I argue that there are grounds to doubt the fundamental distinction and the standard view, in both its versions. Positively, I suggest that we operate a switch of direction: individual virtues are (both psychologically and conceptually) prior to “the moral” and “the epistemic” divide. Or, more robustly, each virtue is metaphysically distinctive and normatively complex, incorporating both what are traditionally understood as “epistemic” and “moral” elements in various—indeed, distinctive—ways. There are no “two kinds” of virtues and vices.

I begin by expanding on the standard view (§1) and making an initial case against its solidity. I then present my favoured alternative in more detail (§2) and provide three arguments for it. First, I resist the weaker version of the standard view by stressing that, if pressed, it is actually unable to account for our practices (§3). Then, I resist the stronger version with an argument to conceptual impossibility (§4). Lastly, I argue that there are delicate questions around commensurability that arise for the standard view but do not for my alternative proposal. Whilst much more work would be needed to prove that reversing the (psychological and explanatory) direction is the correct way ahead, I hope to convince the reader that it is at least a possibility worth investigating (§6). My goal is not to prove the standard view to be false but only to call into question the widespread consensus around it and to indicate a possible alternative.

2. The Standard View

On the standard view, the distinction between epistemic and moral normative domains is prior to the distinction between any two given virtues or vices. If this is true, the fundamental distinction seems to follow—virtues must come in two kinds. The locus classicus of the fundamental distinction from which several contemporary papers begin is this passage from Aristotle’s Nicomachean Ethics:

Virtue, as we have seen, consists of two kinds, intellectual virtue and moral virtue. Intellectual virtue or excellence owes its origin and development chiefly to teaching, and for that reason requires experience and time. Moral virtue, on the other hand, is formed by habit, ethos, and its name, ethike, is therefore derived, by a slight variation, from ethos³. (Aristotle 350AD, NE II 1103a 14–19) [45]

But whilst invoking it at the start of several discussions on the topic, the literature is quite selective in what it learns from this passage. On the one hand, virtually all retain Aristotle’s distinction between moral and intellectual virtues. On the other hand, however, they largely resist his behaviouralist criterion to draw it. Even more interestingly, they reject Aristotle’s deeper rationale for drawing it: his commitment to the partition of the soul. There must be different kinds of excellences since there are different parts to the human soul⁴. In fact, for Aristotle, our soul consists of an irrational and a rational part. The former is divided into a vegetative part and a part that somewhat shares in reason, where desires and emotions reside, whilst the second consists of theoretical and practical reasons. Intellectual and moral virtues have different behaviours and modes of acquisition insofar as they pertain to different parts of the soul. Their divergent characteristics are a direct implication of their belonging to theoretical reason and to the sensitive part of the soul, respectively⁵. Wary of the idiosyncrasies of this aspect of Aristotle’s account, contemporary philosophers dismiss it. It is then perhaps surprising that they maintain its resulting implication⁶.

That Aristotle’s larger picture is unpopular amongst contemporary philosophers is not surprising. For one thing, it is hard to understand the claim that our soul consists of different parts, and we are generally cautious to sever reason from feelings, intellect from emotions. It returns a picture of humanness that developments in psychology and medicine make us perceive as seriously outdated. More importantly, the functioning of each part of the soul seems to require the exercise of both kinds of virtues: there cannot be excellent rational deliberation without (reliable and accurate) feelings, desire, and perceptual understanding of a given situation. Conversely, there cannot be excellent choices without some ability to calculate one’s options, risks, and best courses of action [6,20,54,55].

But how are we to understand the fundamental distinction then? Although this is not the place to consider and assess the merits of all available proposals, a brief look at some of the most influential ones is necessary. First, because the solidity of the standard view essentially lies on the possibility of grasping the distinction between the moral and the epistemic, and thus between moral and epistemic virtues. Second, and relatedly, because highlighting that there is little to no agreement on how the task should be pursued may provide a first reason against the solidity of the standard view.

Here are some options:

a.

Modes of acquisition

Aristotle suggests a second discernibility criterion: moral virtues are formed by habit whilst intellectual ones by teaching. A variation on this view centres on voluntariness, whereby moral virtues are voluntary whilst intellectual virtues are natural. Yet, neither seems true: Were one to imitate kind behaviour for a lifetime but merely externally, without feeling the pull of reasons for kindness, we should not call them virtuous. And several intellectual virtues are properly mastered after years of habit and exercise, and perhaps less through teaching, e.g., open-mindedness [6]. At a closer look, both putative kinds of virtues seem to require natural inclination, teaching, understanding, and exercise.

b.

Realms of life

Some may insist that moral and intellectual virtues pertain to different spheres of life, namely actions and beliefs, respectively. It is quite easy to find counterexamples: One’s kindness, humility, or benevolence must be reflected on both one’s deeds and thoughts, and one’s intellectual courage or open-mindedness has to manifest in one’s actions as well. For instance, as Zagzebski highlights, compassion requires not just a compassionate motive but also reliable success in helping others [6] p. 149.

c.

Directionality

Some may think that moral virtues are “other-regarding” as opposed to “self-regarding”⁷. There may be more ways of understanding what this means. But if taken at face value, this proposal also lends itself to counterexamples. Whilst humility, respect, or generosity are quintessentially “other-regarding”, as they put others’ wellbeing in the centre, moral virtues like courage, prudence, temperance, self-respect, or patience regard the self. Conversely, intellectual virtues like truthfulness, open-mindedness, or intellectual humility cannot be understood by looking at the individual alone: they are motivated by, and manifest themselves necessarily in, a social environment, and they benefit others in a strong sense.

d.

Goals

A rather influential idea is that moral virtues aim to the wellbeing of others (and perhaps oneself), whilst intellectual virtues aim to truth. Julia Driver is a prominent advocate of this line⁸:

The intuition I would like to explore is that intellectual virtues have—As their source of primary value—Truth or, more weakly, justified belief for the person possessing the quality in question, and this is what ‘getting it right’ means for the intellectual virtues, whereas for the moral virtues the source of value is the benefit to others, the well-being of others, and for the moral virtues this is what ‘getting it right’ means. [26] p. 374

One problem with this line is that it is not clear how to suitably identify two clearly and fully distinguished goals: an epistemic and a moral one [57]. If we were to agree that the goal of moral virtues is happiness, eudaimonia, or human flourishing, it would remain unclear why goals like knowledge are not a constitutive part of it. If, as a reaction, we characterise the goal of moral virtues further so that it is more distinctly and unambiguously moral in nature, then it risks failing to fit all those traits that we would recognise as moral virtues: if we say that the goal of moral virtues is “doing the right thing”, traits like humility or wisdom are left out; if we say that it is “being wise”, then traits like courage or diligence would fail to count as moral virtues; if we say that it is “being moral”, we need a theory of “being moral” that is independent of a preliminary understanding of moral virtues, or else we cannot use reference to goals to qualify their nature. Finally, if we were to understand the claim that virtues have certain goals as saying that virtues aim at attaining certain goods, more would have to be said about the case of virtuous beliefs, as those cannot be explained by features of their motivational structure, given that they have no motivational structure at all. In other words, since we cannot believe at will [58,59,60,61], if being virtuous depended on hitting (or tending to) a certain goal, we could never determine whether a belief, rather than an action, is virtuous.

Of course, this brief excursus cannot aim to completeness. Its goal was only to put some flesh on the standard view and to bring to light the difficulties of providing a satisfactory criterion to ground the distinction between the moral and the epistemic.

3. An Alternative: The Primacy of Virtues

Let me outline an alternative to the standard view. Above, I spoke of “changing the direction”. This is what I mean: rather than conceiving of the distinction between the epistemic and the moral as (psychologically or explanatorily) prior, I urge we consider the distinctiveness of each virtue to be prior.

This calls for a certain understanding of virtues. Drawing on John McDowell’s idea that virtues are those “states of character whose possessor arrives at right answers to a certain range of questions about how to behave” [62] p. 331, I suggest we think of virtues as distinctive sensitivities to both theoretical and practical reasons. “Distinctive” in the sense that, in most situations, there will be one way in which the ideally (and exclusively) kind or honest person sees things. They will be attentive to, motivated by, and reacting to certain and not other aspects of a situation—aspects that are essentially different from the ones salient to the paradigmatic honest or generous person. On this view, virtues are specific lenses through which one sees the world and its dynamics. But as McDowell emphasises, virtues provide the right answers to a certain range of questions, which are, I want to say, those relevant to the specific way of seeing captured by any given virtue. Thus, generosity may remain silent in contexts that do not call for a generous act; not in all situations there is a fitting generous reaction. This is not a deficiency—the specificity of its salience area only further characterises the nature of the virtue.

Whilst virtues enable you to see the world in the right way and arrive at the “right answers”, vices lead you to the wrong answers to the questions of what to do and what to believe. This both determines the difference between virtues and vices and sheds some light on the source of virtues’ normativity. Someone is virtuous and thus commendable for getting things right, for seeing things the right way, for being a person who believes, acts, feels, fears well. Conversely, they will be criticisable for getting things wrong, for seeing things, believing and reacting cowardly, dishonestly, foolishly.

Of course, real, embodied people do not possess virtues of this sort, robust to the point that in each and every situation the person will react as the virtue would demand. And this is because each person possesses their personal cluster of virtues and vices, which influence each other. In fact, the exact ways in which different traits are integrated in someone explains exactly what facts will count for them as good reasons in any given situation. For instance, someone who is generous and kind but a bit obtuse would not react as someone who is generous, open-minded, but often unkind. Generosity will manifest differently in the two. Moreover, there is no univocal relationship between traits and responses⁹: the same response can manifest different traits in some contexts, without this threatening the distinctiveness of each.

Crucial for present purposes is this: From a developmental point of view, we are wired to grasp the distinctiveness of individual virtues from a young age [63,64,65]. With time, we recognise more and more acts as manifestations of kindness, even when they are tokens of types of actions we have never seen before and despite virtues’ lack of codifiability. We learn to become kind ourselves because we see the force of reasons of kindness. We train ourselves to finetune our sensitivity so that it recognises certain facts as good reasons, and we then find ourselves capable of acting kindly in new and varied situations. This should not be interpreted in an overly intellectualist manner. To “recognise certain facts as good reasons” is to see and respond to situations in sufficiently consistent ways, without that being necessarily accompanied by any explicit thought about reasons or about kindness [66].

What is crucial here is that, first, we typically understand (in a non-necessarily intellectualistic sense of “understanding”) how the sensitivity distinctive of kindness differs from that of generosity or honesty, and we do so (extraordinarily) reliably. Second, with mastering a virtue comes a sensitivity to both practical and theoretical reasons, a way of seeing the world that manifests itself in both actions and beliefs.

An account of virtues drawn along these lines will be one on which (1) they are sensitivities to both practical and theoretical reasons; (2) they are traits that constitute the essence of who we are, the evaluative orientation through which we see the world; (3) they can be manifested in actions, beliefs, desires, and inclinations; (4) they are not absurdly robust—one can count as possessing a trait even though they do not manifest it each and every time it would be fit; and (5) they do not need to come all together [67].

Although much more could be said, the key point is this: on this alternative approach, virtues are prior in the sense that they are the starting point. This is true first developmentally: we begin by understanding what virtues are, we get to master virtue concepts and only then (some people) get a grasp of what the ethical and epistemic realms are. More broadly, it is true psychologically, in the sense that the question of whether someone is generous or honest has a stronger pull within our practices than the question of whether they are doing well from an epistemic or from a moral point of view. Second, virtues are the starting point conceptually and explanatorily: it is the nature of virtues that can explain what it is to achieve the moral or epistemic good, or what it is to be morally or epistemically salient, rather than vice versa¹⁰. Put otherwise, that virtues are normatively prior means that it is only once we know them that we can ask ourselves what a morally good action or an epistemically good belief is. In a world emptied of virtues, those would be void questions. Relatedly, that virtues are normatively prior means that a belief or action is an apt response to a situation insofar as it is kind or honest to so believe or act¹¹, rather than because it is epistemically or morally good to so act.

4. Against the Weak Standard View: Practical Inaccuracy

In the rest of the paper, I will attempt to build a case against the standard view and show that my alternative does a better job at accounting for those delicate issues that are particularly challenging for the standard view. I begin by casting some doubt on the weaker version of the standard view in this section and discuss the stronger version in the following. Recall that on the weaker version the distinction between the moral and the epistemic is psychologically prior to the distinction between any two virtues.

To clarify, the point is not a simply empirical one or, as it were, accidentally true. It is not a matter of counting how many people happen to think one way or another or of finding a counterexample to the view I am trying to resist. Nor is it a merely developmental point, one that can be settled by looking at children’s learning patterns. Rather, it is a matter of looking for the best way of accommodating and accounting for our practices, understood as the naturally and culturally determined ways in which we understand and treat each other and the world. It is not an inquiry into the “simply empirical” or “accidental” because it regards what is essential to human beings as we happen to be, essential to how we conceive of the world and our place in it. By setting out to investigate what is “psychologically prior”, then, I mean to look at the role played by the moral/epistemic divide and the fundamental distinction in our practices.

If pressed, the standard view returns quite an odd picture of our practices. Suppose that contemporary philosophers are right in thinking that virtues (and vices) come in two kinds. How would this be reflected in our practices? First, presumably, there would be two sets of practices. Manifestations of moral or intellectual virtues would be met with different sets of positive reactions, whilst manifestations of moral or intellectual vices with different sets of negative reactions. In fact, there is a trend in the literature on responsibility for beliefs arguing precisely for this point [68,69,70,71]. The general idea is that there is a kind of blame (and perhaps praise) that is epistemic in nature, and it manifests in a reduction in epistemic trust, modifications of epistemic relationships, and the like. Second, if the fundamental distinction is true, then there would be no tension in manifesting intellectual honesty and moral dishonesty, for instance—unless the fundamental distinction automatically came with a set of interaction rules. Possessing a positive trait in a normative sphere and its opposite in the other one would elicit the corresponding normative kinds of praise and blame, and there would be no tension.

Assessing both claims is made harder by the lack of agreement on the nature of epistemic blame. For the sake of argument, I suggest we operate with a fairly thin account on which epistemic blame consists of a cognitive and/or emotional response to someone’s perceived failure to achieve an epistemic good (be it ignorance, misunderstanding, avoidable mistake, perhaps stupidity) that results in a lessened trust in that person’s capacity to do well epistemically. Given that no ill-will needs to be present, epistemic blame does not involve anything like resentment or indignation¹².

Now, if we look at the way we ordinarily react to manifestations of virtues and vices, both implications seem dubious. First, it is not clear to what extent the two sets of practices can remain distinct. Or better: it is not obvious how the distinction between our reactions to moral and epistemic virtues and vices is necessarily starker than the distinction between our reactions to two moral virtues or two epistemic virtues. Re virtues: take our reactions to manifestations of humility and intellectual humility on the one hand, and our reactions to manifestations of humility and courage on the other hand. Re vices: consider our reactions to manifestations of arrogance and intellectual arrogance on the one hand, and our reactions to manifestations of arrogance and cowardice on the other hand. In practice, the emotional and cognitive elements involved in assessing and interacting with someone manifesting moral and epistemic versions of the same virtue or vice will have much more in common than those involved in assessing someone manifesting two traits belonging, according to the fundamental distinction, to the same normative realm. That is, our reactions to manifestations of arrogance and intellectual arrogance will be much more cohesive a cluster than our reactions to manifestations of intellectual arrogance and intellectual cowardice.

Someone may insist that the key difference is that the epistemic kind of negative assessment does not involve resentment or indignation and rather consists of a modification in how we think of someone’s epistemic licence as a believer, so to speak. As it is often said, this is a modification in how much we can trust them. However, first not all negative reactions to a moral vice must involve resentment or indignation (e.g., cowardice), and thus resentment or indignation should not be taken as the marker of reactions pertaining to the moral realm. Second, it is not obvious that trust belongs to the epistemic rather than moral realm. To trust is to have confidence, to rely, to take on someone’s else point of view and accept it as a premise in our own practical syllogisms. It is not only a matter of calculating the probability of their opinion being a true belief in light of their past epistemic pedigree, which crucially is what trust would have to be if we had to insist that it is an eminently and exclusively epistemic reaction. The problem is that if we went down this path, we would risk picturing reactions to epistemic virtues and vices in a way that is dangerously close to what Strawson famously defined as issued from an “objective” rather than “participant” stance [72]. If trust really were an eminently epistemic reaction, we would be treating others like we treat a computer, merely calculating how reliable they, as sources of information, are likely to be.

To see the shortcomings of the second implication, just try to imagine someone who is intellectually honest and morally dishonest, or intellectually kind and morally unkind, or exclusively intellectually generous. If we met someone like this, it would strike us as odd—we would need an explanation for their inclinations and behaviours, e.g., cognitive anomalies, emotional traumas, difficult childhoods, borderline personalities. We would not limit ourselves to praising them epistemically whilst blaming them morally. Our reactions would be essentially characterised by a deep tension, which, I believe, speaks against thinking of virtues as coming in two possibly independent kinds. By contrast, a more integrated account accommodates the tension, since honesty is a sensitivity to both theoretical and practical reasons that manifests itself in degrees but does not abnormally develop exclusively within one normative realm¹³.

Finally, notice that in a large variety of ordinary cases we would not even be able to determine with certainty what virtue or vice a certain belief or action manifests and, consequently, whether our reactions count as instances of a morally or epistemically charged kind of blame. The advocate of the standard view would consider this to be a failing on our part and object that this only proves the messiness of our practices rather than a weakness in the fundamental distinction. Let me register a deep dissatisfaction with the nature of this objection. Our practices are undoubtedly complex. But exactly what good would the standard view be, as a view about an explanatory relationship between normative realms and virtues and, in turn, as a view about the very nature of virtues, if it is unsuitable to account for our ordinary practices? Again, whether philosophy should always aim to illuminate ordinary life is surely debatable. But I believe that in this case there is something about the nature of the subject matter that cannot prescind from it. We want to know more about the nature and functioning of virtues as we understand, use, and react to them. Or otherwise, if someone insisted that the fundamental distinction is true whilst not being able to capture our practices, I think they need to say more about what theoretical and practical role it is then meant to play within our framework.

5. Against the Strong Standard View: Conceptual Impossibility

In this section, I want to resist the stronger version of the standard view, on which the moral/epistemic divide is conceptually prior to the distinction between any two given virtues or vices. According to this view, the concepts “epistemic” and “moral” are necessary to the individual virtue concepts. Relatedly, the fundamental distinction states that there are virtues of two kinds according to whether they present characteristically epistemic or moral features, where these epistemic and moral features are necessary to determine the nature of each virtue. Necessary to determine what a given virtue is.

The point is that this, I think, gets things the wrong way around. A virtue concept is prior to the concepts of its epistemic and moral (alleged) kinds.

Let me elaborate. It is often remarked that there are intellectual and moral kinds of, for instance, honesty, curiosity, courage, or generosity. Conversely, many agree that there are intellectual and moral kinds of, for instance, dishonesty, carelessness, or inattentiveness¹⁴. Yet, crucially, identifying intellectual and moral kinds of honesty presupposes an understanding of what honesty tout court is. Being able to distinguish between different modes of something requires the capacity to track it amongst its different varieties. It is not that we learn how to master the concept “intellectual honesty” whilst engaging in conversations, how to master the concept “moral honesty” whilst acting in the world, and then realise that there is a remarkable similarity of intent between the two. Rather, we get to know honesty by learning to recognise honest persons—we see how they think, believe, desire, and behave; what drives them and how they “see things”. And this is because being honest is to take certain things to be true and certain actions to be worth pursuing.

The reason why I talk quite robustly of conceptual impossibility is that we couldn’t form the concept “intellectual honesty” if we looked only at someone’s beliefs, or at their sensitivity to theoretical reasons—where the “couldn’t” here does not refer to psychologically impossibility, though that is the case too. The point is that, even if we could psychologically, the resulting concept would not be our concept “intellectual honesty”: it would be closer to what we understand as sincerity, perhaps. The thickness and nuances of our concept “intellectual honesty” come precisely from it being a form of honesty, e.g., a virtue that in its essence comes with a regard for others’ right to be well treated, which includes, amongst other things, being told the truth. It is conceptually impossible to gain the concept of an intellectual or moral virtue as we understand it without mastering the concept of the virtue tout court.

To clarify, this is not meant to be a claim about concept acquisition. I am not suggesting that we begin by gaining a grasp of the general concept first thing, for people do generally begin with understanding particular cases and work up to general accounts. Rather, the claim is specifically about the acquisition of virtue concepts and the grasp of the fundamental distinction. My contention is that we learn, understand, and master virtue concepts tout court (by building our way up from particular cases) before we grasp the distinction between moral and intellectual virtues, for those particular cases involve specific actions, beliefs, emotions that are normatively complex, and incorporate elements from both what we call the epistemic and moral domains.

If anything, this raises once again the question of what prompts us to force the fracture between an epistemic and a moral dimension. Nothing in our lives and in our conceptual wiring seems to demand it. Or at least, if it may be needed to bring order and neatness to an admittedly very complex theoretical taxonomy, it is not obvious that it should play the prominent role that it currently does in the contemporary literature.

6. The Ravine Problem: Incommensurability and Interference

In this last part, I want to expand a little on a point I touched upon in Section 3 and that is receiving increasing attention in the debate on normativity, i.e., the problem of incommensurability. The idea discussed there is that, if we understand ethics and epistemology as being two normatively distinct realms governed by different sets of norms¹⁵, then we need to make sense of the cases in which moral and epistemic duties may conflict. Thinking about the fundamental distinction, the problem resurfaces in several guises: From a first-personal perspective, one may wonder what someone should do when considering whether to be morally virtuous or epistemically virtuous. From a third-personal perspective, we may wonder how we should evaluate overall someone who is morally virtuous or epistemically virtuous or vice versa. Although we may provide a neat response to at least this latter challenge by simply saying that we commend them morally whilst reprimanding them epistemically, in fact, it is not obvious how viable that is. Put simply, the fundamental distinction reinforces a ravine between two kinds of virtues (and vices) that does not exist in our ordinary lives and that the factual organicity of our mode of existence constantly demands us to bridge.

There are several reactions to the problem of incommensurability in the literature. Very briefly, some like Hall and Johnson [75], following Chisholm [76], hold that moral duties always trump epistemic duties. For instance, if you epistemically ought to do something, i.e., gather more evidence, and you morally ought to do something else, i.e., save a kid on your inconclusively supported belief, the moral ought always “wins” and you just plain ought to do that other (moral) thing. Similarly, Maguire and Woods “deny that really strong epistemic reasons will win the day over weak practical reasons” [77] p. 19 and Meylan claims that “[i]n tragic cases of doxastic divergence…, the dilemmas between the epistemic and the practical ought to believe are not sensitive to the reinforcement of the epistemic horn. No matter how many additional epistemic reasons the subject has not to hold practically right belief, there is no point at which the balances will tip on the side of the epistemic horn” [78]. Others argue for a lexical priority view, on which one kind of consideration typically prevails unless the other one is extraordinary; for instance, Reisner [79] p. 24 claims that “when pragmatic reasons for belief are strong enough, [epistemic] reasons for belief are silent, and…otherwise, pragmatic reasons for belief are silent in determining what one ought to believe, all-things-considered”. Yet others like Feldman [80] and Brown [81] voice some scepticism against the very idea that moral and epistemic oughts are commensurable¹⁶.

Once we accept the fundamental distinction, we are called to engage with these delicate matters. More precisely, the challenge is twofold: there is a question of commensurability and one of interference. The former concerns whether considerations pertaining to epistemic and moral virtues and vices belong to the same scale of measurement; the latter asks how they are to be weighed together. That is, suppose someone asks me how I feel about your action or belief. From within the standard view, I will have to first decide whether your action or belief belong to the moral or epistemic realm, and then proceed with the corresponding evaluation. In case your belief or action lent itself to both sorts of evaluations, I would have to first verify whether two normatively different kinds of evaluations can issue in an overall way of reacting and feeling towards you, and then to establish how exactly to do so¹⁷.

Things are smoother from a virtue-first approach: In answering for myself the very question of how I feel about your belief and action, I will weigh together considerations about your conduct that are not postulated as being of different sorts. Within the process of determining whether it is a virtuous or vicious action or belief and, most importantly, exactly what virtue or vice it manifests, I will have to take into account multiple factors regarding the values that your belief or action displays, the context you operate within, your usual inclinations, the effort you put in bettering yourself, and so forth. The power of a virtue-theoretic approach is that it does not need to find a generalisable way of bridging the ravine because it denies that there is a ravine to bridge to begin with.

My alternative view understands virtues as distinctive ways of incorporating certain, determinate ways of being sensitive to theoretical and practical reasons. When possessed, they make certain evidence salient, certain facts seem true, certain actions worth doing, certain outcomes desirable. So, in a way, virtues themselves are distinctive, complex, and refined “recipes” to weigh what counts as epistemic and moral considerations. Mastering virtue concepts enables us to evaluate complex circumstances: someone is not admirable morally and criticisable epistemically; they are humble. And someone else’s behaviour or belief is not admirable epistemically and questionable morally: they are curious. For a virtue-theoretic view, questions of commensurability and interference do not arise, as virtues themselves are sorts of rules of interferences, distinctive ways of weighing and assessing considerations.

That each virtue and vice stands for a specific way of weighing considerations is further indicated by the fact that each trait relates differently to all others: A moral virtue like generosity is less threatened by an epistemic vice like untruthfulness than honesty is; kindness is affected by gullibility to a lesser extent than courage is. Conversely, intellectual humility is tainted by cowardice less than open-mindedness is. This is true only because of the very nature of each virtue, and nothing else. The standard view has trouble accommodating this variety, since it aims to provide generalisable answers to the question of how to deal with epistemic and moral considerations.

To clarify, the “recipes” offered by virtue terms are uncodifiable and open-ended. Yet, this does not mean they are arbitrary. It simply indicates that they are not translatable into a complete list of prescriptions given the potential infinity of relevant situations. They are subject- and context-dependent, but there will be little doubt that a certain action or belief is a manifestation of generosity rather than honesty, or that a certain action or belief calls for pity, disgust, impatience; admiration, warmth, or esteem.

As a result, casting doubt on the standard view and fundamental distinction provides us with elements to build a unified theory of responsibility that does not split it into moral and epistemic responsibility, or into responsibility for actions and responsibility for beliefs, as the standard view invites us to do—somehow paradoxically, given their open rejection of Aristotle’s partition of the soul. Instead, we stress that it is persons that are responsible for their beliefs and actions. When we assess them, we assess them, and not only a part of them, or only from a certain, compartmentalised point of view.

7. Conclusions

In this paper, I have discussed a widespread way of thinking about the relationship between the epistemic and moral normative realms on the one hand and virtues and vices on the other. According to the standard view, the former distinction enjoys priority over the distinction between any two virtues. I proceeded to characterise this priority in a weaker and a stronger way and raised some questions about each. Further, I have highlighted that the standard view, and the fundamental distinction it generates, raises problems of normative and evaluative commensurability and interference between domains, which my alternative pre-empts. In fact, along the lines I have outlined, virtues offer normatively distinctive “recipes”, so to speak, to calibrate and weigh together different (morally and epistemically relevant, one may say) elements of any given situation.