2. Pragmatic paradoxes and self-attribution

I follow Sainsbury (Reference Sainsbury2009: 1) in understanding a paradox to be “an apparently unacceptable conclusion derived from apparently acceptable reasoning from apparently acceptable premises.” Citing Sainsbury, Huemer (Reference Huemer2018) requires that the reasoning for the unacceptable conclusion have widespread and robust appeal that leads to persistent disagreement among experts. To borrow Huemer's example, a statement such as “I am nothing” fails to qualify as a paradox because it is either false, or has an alternate meaning such as “I am unimportant.” The air of contradiction in the statement is better explained by a more widely appealing intuition (i.e., that “I” is in fact something), or a lack of robustness (i.e., that “nothing” is ambiguous). Sometimes, too, we use “paradox” in a loose sense to simply describe an odd phenomenon when there is nevertheless agreement among experts. For example, the so-called “paradox of happiness,” in which we decrease our odds of achieving happiness by directly pursuing it, does not present us with any unacceptable conclusion; we would not need to reevaluate what happiness means when considering this fact, because happiness is not typically defined by whether we directly aim toward it or not.Footnote ⁵ To be a genuine paradox, then, a puzzle must (1) have an apparently unacceptable conclusion, (2) apparently acceptable reasoning from apparently acceptable premises, and (3) one can reasonably expect (1) and (2) to yield persistent disagreement among experts; the more a puzzle satisfies these conditions, the more intractable and paradoxical it is. These conditions should be enough to guide us through our discussion.

With these preliminaries in place, do philosophers think that humility is genuinely paradoxical? It seems they do; persistent disagreement among philosophers about how to explain humility's puzzling features is one indication that it generates a problem that is rather intractable. Brennan (Reference Brennan2007: 126), for instance, takes himself to be offering a “competing account of why saying ‘I am modest’ or thinking it is paradoxical,” in response to Driver's (Reference Driver1989) view. While Brennan argues that one can have knowledge of one's worth by comparing oneself to the ideally virtuous person, Driver (Reference Driver1989: 376) contends that the utterance “I am modest” is self-contradictory because it implies that one is knowledgeable about one's self-worth.Footnote ⁶ Statman (Reference Statman1992) for modesty, and Hare (Reference Hare1996) for humility, follow this line as well by arguing that an accurate assessment of one's worth generates paradoxes for those who are also egalitarians about human worth (e.g., Ben-Ze'ev Reference Ben-Ze'ev1993). On their views, it is seemingly impossible for someone to be humble by accurately judging that they are morally superior to most, because this belief directly undermines one's commitment to egalitarianism.Footnote ⁷ I will address this specific worry in the following section. For now, I only wish to highlight how philosophers find assertions of modesty and humility to be self-contradictory or in conflict with other widely held commitments, indicating that authors do find something seriously paradoxical about these traits. The precise way that purported solutions attempt to account for these paradoxical features will become evident as we progress. Let us begin then with the self-attribution puzzle, which will lay the foundation for considering our other puzzles.

SAP presents us with a pragmatic contradiction as opposed to a logical contradiction. Paradoxes of self-reference, such as the liar paradox (“This sentence is false”) are logically self-contradictory, appearing false if true, and true if false. By contrast, pragmatic paradoxes are “falsified by their own utterance,” meaning that they contradict a fact that the speaker implies by making the utterance, often the propositional attitude that the speaker expresses (Cohen Reference Cohen1950: 85). Ebersole (Reference Ebersole1953) notes examples such as “there are no sentences” and “I believe nothing” as instances of pragmatic paradoxes; the speaker who utters “there are no sentences” has contradicted a fact she implies by making the utterance, which is itself a sentence; similarly, the speaker who asserts “I believe nothing” expresses a belief that is inconsistent with her utterance. Most authors consider Moore's paradox, often expressed as “It is raining, but I do not believe it is,” or, “It is raining, but I believe it is not raining,” to be another example of a pragmatic paradox.Footnote ⁸ Moore (Reference Moore and Schilpp1944) explained the absurdity of the speaker's utterance by noting that when a speaker asserts that p, they imply that they believe that p. The intractability of Moore's paradox is affirmed in part by the disagreement among philosophers about various solutions, which have additional consequences for understanding norms of assertion, belief, and rationality.Footnote ⁹

Does self-attributing humility generate a pragmatic paradox? I do not think so. While philosophers converge upon the belief that, in normal cases, the humble person would not utter SAP, a speaker who utters SAP doesn't imply anything about their attitudes or the truth of their utterance. Let me illustrate by first attempting to articulate the reasoning that is motivating the puzzle. I suspect it goes something like this:

1. (1a) If a speaker S is humble, then he will not assert SAP.

2. (2a) S asserts SAP.

3. (3a) Therefore, S is not humble.

Robinson (Reference Robinson, Alfano, Lynch and Tanesini2020: 27) describes this reasoning in a manner that is telling of its lack of robustness, writing, “if I truly possess that characteristic, I (typically) will not have made the utterance in question.” Note the parenthetical “typically” here, which is playing a more substantial role than is often admitted. Robinson's own solution to the puzzle requires that the humble person rarely face circumstances where they need to self-attribute humility, because most of the time, what it means to be humble is to be inattentive to one's own merit.Footnote ¹⁰ This strikes me as a reasonable explanation, but I worry that it does not resolve any genuine paradox given its attention to exceptional circumstances. For instance, Robinson would permit a humble person to utter “I am humble” to appease an annoyingly insistent complimenter. Yet, such an exception means that the utterance in question is not a pragmatic paradox, because the contextual factor of an annoying complimenter is unrelated in principle to a humble speaker's being inattentive to their own self-worth; notice that such contextual factors could not similarly absolve “there are no sentences” from being paradoxical. I reckon that more emphasis ought to be placed on taking the quality of being inattentive to be part of humility's definition, as opposed to emphasizing the exceptional circumstances that might sometimes lead a humble person to attend to their own merit.

If hedging on typical cases is only possible when something unrelated to the utterance stands to make it false, a better case is one that considers whether self-attribution undermines humility's constitutive features. Here is one problem, again invoking Robinson's proposed solution: if being inattentive to one's worth is part of humility's definition, one may worry that a self-attribution always implies attentiveness. But I do not think that this is right. For example, suppose that a PhD student, Laura, is searching for a humble dissertation advisor. Laura has some evidence that Professor Smith is humble, having seen him engage in characteristically humble behaviors such as downplaying his achievements, attempting to change the topic of conversation away from himself, and so on. She also has some testimonial evidence from Smith's colleagues and students who find Smith to be humble. But imagine that Laura wants to hear Smith's assessment of himself, so she directly asks her prospective advisor, “Are you humble?” – to which Smith replies, “Yes, I am humble.”

How should this self-attribution affect Laura's assessment of Smith? Laura should, I think, bear in mind that trait attributions suggest many facts about an individual's psychological states and behavior. These qualities are much murkier than the facts a speaker implies by uttering other paradoxical statements (e.g., “there are no sentences” is obviously a sentence). It might be true that the humble person typically does not self-attribute, but we mean this as something that follows from humility's definition as one fact among an uncountable number of interrelated facts that Smith implies about his character. Among these, importantly, are the characteristically humble behaviors that Laura has already observed and gathered from Smith's colleagues’ testimony. More, however, Smith implies that he has characteristically humble psychological attitudes – inattentiveness to his worth, desires to consider others’ worth more than his own, a lower estimation of his worth, or other attitudes along these lines. Without simply begging the question, one could not say that X is a quality necessary for being humble, which a speaker implies by their self-attribution, and which is also defeated by that self-attribution. Here are three reasons why. First, note that it would be odd for Laura to take Smith's self-attribution as evidence that Smith is not humble, and much less reasonable for her to think that it counts against her previous observations and testimonial evidence of Smith's characteristically humble behavior. This is further motivated, second, by the fact that the evidence that Laura possesses of Smith's humility could in principle be obtained by Smith, such that his self-attribution amounts to an ordinary instance of self-reflection. Just as the generous person may reflect on their charitable contributions, a humble person may reflect on their own willingness to shift attention away from themselves. Third, and finally, the facts that most conclusively falsify Smith's utterance are unrelated character flaws. Self-attribution is unjustified if, for example, Smith really views himself as great, admirable, and unbeatable as a dissertation advisor. But in that case, Smith has said something false only because he views himself as great, admirable, and unbeatable. There is nothing paradoxical about self-attribution if typically, one refrains from uttering SAP because of considerations unrelated to the utterance in question. Insofar as exceptional cases exist in virtue of Smith's other characteristically humble states and behaviors, self-attribution is not contradicted by whatever Smith implies by his utterance.

What's needed is a claim to the effect that asserting SAP constitutes a kind of obvious immodesty – that is, that self-attribution strangeness is in fact part of humility's meaning. Once more, however, I am unsure how one could hold this view without begging the question in favor of strangeness. Presumably, virtue theorists should want to explain away strangeness at least to the extent that it no longer generates wider problems for one's virtue theory. Consider one such wider problem: should one's virtue theory favor a doxastic explanation (i.e., in terms of our having or lacking certain beliefs) or non-doxastic explanations (e.g., in terms of emotions, desires, attention, etc.) of strangeness? As I see it, the answer is “neither,” since both types of account offer comparatively reasonable explanations of strangeness – something one would not expect if there was live disagreement about a genuine paradox. For instance, consider Robinson's non-doxastic explanation which locates the problem of SAP in a speaker's cognitive attention.Footnote ¹¹ Aside from exceptional cases of being prompted by a third-party in which S might reluctantly admit that he is humble, in premise (2a), S implies that he is attending to his humility too much to count as genuinely humble. But as we said, the issue with this explanation is that exceptional cases show that we cannot draw any conclusions about S's attitudes from his assertion, because in some contexts, S might say “I am humble” to appease an annoying complimenter who will not go away until S finally admits that he is humble. This explanation is not a unique solution to the problem; a doxastic account of humility could alternatively hold that S would not self-attribute only if he has the right sorts of beliefs about his worth and the worth of others. Hence, it may be reasonable for S to assert that he is humble if it is in fact true, and he is in a context with an interlocutor who would recognize the truth of the assertion as well. Emphasis on context leads us back to the kind of situation between Laura and Smith, where we admit that humility's definition can turn on a theoretical commitment to doxastic or non-doxastic attitudes – but the alleged strangeness of the utterance itself is not relevant to that definition.

Since both doxastic and non-doxastic explanations agree that S is not humble in ordinary cases when he asserts SAP, we must admit that definitions of humility do not turn on apparent strangeness. And as we would expect with a non-paradox, we find more consensus among experts about the reasoning and conclusion than originally thought. We can, of course, make the above reasoning seem more contradictory by inserting the premise that S is humble. But this will only help to further solidify the consensus about the puzzle.

1. (1b) S is humble.

2. (2b) If S is humble, then he will not assert SAP.

3. (3b) S asserts SAP.

4. (4b) Therefore, S is not humble.

We can now see the contradiction here, and why both doxastic and non-doxastic explanations would be correct to reject (1b) here. The problem with (1b), in short, is that it neglects widely shared norms about virtue confirmation and disconfirmation. For example, we could initially judge that Smith is compassionate but one day find him torturing puppies. In such a case, either Smith must never have been compassionate, or he suddenly became cruel (since the compassionate person would not torture puppies). Therefore, it is either false that Smith is compassionate, or true that he is compassionate in one sense (perhaps prior to torturing the puppies) and not another (upon torturing the puppies). Similarly, we can reject the claim that S was humble in the first place given (1b).

Perhaps the problem lies with humility's conflict with a norm of assertion. So, maybe we have a problem that looks like this:

1. (1c) If a speaker S possesses an attribute, then S can truthfully assert they have that attribute.

2. (2c) If S possesses the attribute of humility, then S cannot truthfully assert SAP.

3. (3c) Therefore, it is not true that if S possesses an attribute, then S can truthfully assert they have the attribute.

We should recall that, as Robinson pointed out, (2c) is false. It is only typically the case that the humble person will not self-attribute. Once we admit that an individual can sometimes be in a context that permits a truthful assertion that they are humble, we give up on the prospect of locating a genuine paradox.Footnote ¹² Pragmatic paradoxes involve contradictions, not exceptions. Frequency of context aside though, a more important concern is that (2c) misses the distinction between humility and the kind of speech act that could in principle cancel itself. That is, it is implausible that the utterance “I am humble” is strong enough to show, all by itself, that the person uttering it is not humble, or worse, that they are a braggart. In their account, Alfano and Robinson (Reference Alfano and Robinson2014) define bragging as asserting something about oneself that aims to impress one's addressee. Since bragging depends on a speaker's communicative intentions, SAP cannot be paradoxical simply because humility is an excellent or impressive quality. It can, however, be implicated in what Alfano and Robinson (Reference Alfano and Robinson2014: 268) would call “the braggart's version of Moore's paradox” if a speaker pairs their brag with a disclaimer, leading them to humblebrag. A proposition of this sort is considered below.

There is something strange about making the disclaimer of not attempting to impress while simultaneously attributing an impressive quality to oneself. What makes SAP-D more robust than SAP is this seemingly inconsistent demand imposed upon the addressee. Self-attributing humility demands an addressee both accept that the speaker has an impressive quality and that they not be impressed by it. Thus, humility may be the sort of impressive quality that cannot easily be cancelled by a disclaimer when self-attributed.Footnote ¹³

There are a few things to say about this. First, note the apparent contradiction here is not in humility, but rather, in the meaning that the disclaimer is supposed to communicate given the speaker's intentions. Alfano and Robinson take bragging to involve a speaker aiming to produce the emotion of being impressed in the hearer. Therefore, we could replace humility with any other admirable quality and obtain a similarly contradictory result, as Alfano and Robinson illustrate with the statement “I'm not trying to impress you by saying this, but I am a genius.” In this case, we see that the contradiction arises in large part from the impressiveness of being a genius. Yet, if something like “I am a concert pianist” is not bragging, and we assume this to be an impressive quality, it is not clear why “I am humble” would be bragging either.Footnote ¹⁴ One might still worry that humility, unlike intelligence or being a concert pianist, is somehow disconfirmed by the utterance; one can be a genius and a braggart, but we do not think one can be humble and a braggart. This is true. However, this is owed to the fact that the speaker has bragged by having the intention to impress, which she implies by her disclaimer “I'm not trying to impress you.” So, it is the ‘D’ part of SAP-D that generates a paradox. As Alfano and Robinson note, an agent cannot intend to assert what she takes to be an impossibility. Hence, a speaker who asserts SAP-D implicitly rejects the apparent strangeness of her assertion.

Second, notice that if bragging depends on an agent's intentions, there are cases which may look like bragging but fail to qualify as such because the speaker intends something else. This means that even apparent brags are insufficient for undermining one's humility unless the speaker's intentions are definitively anti-humble. Consider the statement, “I think I'm much more humble than you would understand.”Footnote ¹⁵ In typical cases, of course, we should think that this utterance reflects negatively on a speaker's humility, since the speaker goes beyond self-attributing humility and openly disparages the addressee's intelligence. Such a statement might negatively reflect on a person's humility, but they also may intend to convey (rather rudely) that others are failing to appreciate the positive features of their character. Moreover, such a statement seems an unlikely candidate for a humblebrag as well, since the speaker is not merely presuming unintelligence or insensitivity in the addressee, but rather states it outright.Footnote ¹⁶ Thus, the utterance by itself is insufficient to undermine one's humility independently of the speaker's intentions. So long as self-attribution permits these exceptions, there is nowhere for a genuine paradox to take hold.

One might wonder why we have not tried to appeal more to the propositional attitudes of the speaker to explain what has gone wrong with their utterance. The reason is that SAP, by itself, does not purport to be saying much about the speaker's attitudes – at least not in the way that we would expect with pragmatic paradoxes such as “I believe nothing” or “It is raining, but I do not believe that it is.” Because virtues presuppose that the agent have certain beliefs, desires, and intentions, we are unlikely to find a contradiction that will be illuminating for our understanding of humility as a virtue unless we invoke these propositional attitudes explicitly. While the self-attribution case does not pose a genuine paradox, it still shows that humility might be in tension with other requirements for being virtuous, such as accurate or true beliefs about oneself, an ability to cultivate virtue, and an intention to perform virtuous actions. Of these, I think belief poses the most formidable challenge. So it is to the epistemic problem that I now turn.

3. The epistemic puzzle and its errors

Recall from our discussion of pragmatic paradoxes that a speaker's utterance contradicts a fact implied by the utterance, often the propositional attitude of the speaker. With Moore's paradox, for instance, we find the speaker to be asserting ~p, even though we think that the speaker should be rationally required to believe that p. In that case, we locate a problem with the attitudes of a speaker who ought to be guided by a rational norm to believe that p. So, the best way to make SAP more paradoxical is to explicitly state that the person who utters it believes it, in a manner akin to Moore's paradox. Driver (Reference Driver1989: 377) draws a direct comparison between SAP and Moore's paradox, writing:

So, we can simply add the conjunct “I believe that I am humble” to form an apparent contradiction:

Driver's case for the paradoxicality of EP rests on a single intuition: if one knows or believes one is modest, then acts of modesty are insincere or false modesty – i.e., not modesty at all. Essentially, Driver's view generates the following dilemma: either modesty is not a virtue, or modesty requires ignorance of one's self-worth. According to Driver, one could be modest, but it would nevertheless be paradoxical to believe or know that one is modest.Footnote ¹⁷

While we are dealing with some competing intuitions, I find that none of them can successfully generate a paradox. These intuitions are as follows:

1. (A) Ignorance of one's worth is necessary for modesty or humility.

2. (B) Ignorance of one's worth is compatible with modesty or humility.

3. (C) Ignorance of one's worth is incompatible with modesty or humility.

To my knowledge, Driver is the only philosopher to fully endorse (A), and I believe that only (A) can underwrite EP. Virtue theorists usually reject (A) because virtues are excellent qualities that seem to require that an individual have the skillful know-how virtue called practical wisdom. So, if EP depends on something like (A) to be paradoxical, then it will not succeed since (A) lacks widespread appeal. Among non-philosophers, Driver's intuition fares better. However, non-philosophers might endorse an externalist view about the value and attribution of virtue in general.Footnote ¹⁸ That is, for non-philosophers, behavior is often what matters for attributions of virtue, not the internal states of the agent. This would not show that ignorance is necessary for modesty as (A) states, but rather, that humility is an ordinary instance of the intuition that virtues can permit ignorance, as (B) states. Therefore, Driver's intuition that the person who knowingly understates their worth is falsely modest is weak compared against broader intuitions we have about how virtues are attributed. So, this leaves us with either (B) or (C) to get a paradox.

The problem is that intuitions (B) and (C) force us to confront virtue theoretical commitments about propositional knowledge while not generating any paradoxical effects on their own; namely, we consider whether propositional ignorance makes one more or less virtuous. Assuming the standard view that virtues are excellent traits of character, virtue theorists generally agree that virtuous dispositions require knowledge-how, not knowledge-that, a view widely endorsed regardless of whether one favors intuition (B) or (C). Accordingly, (B) and (C) are not doxastic requirements about knowing or believing something about one's modesty or humility. They arise instead from broader considerations about whether propositional knowledge or ignorance makes one's virtue more or less valuable. Consider (B): just as we are not morally or intellectually better or worse off for not knowing exactly how many stars are in the sky, or the recipe for a delicious soufflé, we don't seem better or worse off simply not knowing some good quality we possess. This is why Driver (Reference Driver2001) adopts the view that propositional ignorance is neither intrinsically good nor bad in response to the worry that modesty as ignorance would apparently violate the general constraint that virtues be voluntarily acquired (since we assume that one cannot be voluntarily ignorant of one's self-worth).Footnote ¹⁹ Those who accept such a general constraint on the voluntary acquisition of virtue would favor (C); however, proponents of (C) could adopt the view that modesty or humility requires non-overestimation (e.g., Richards 1988 (for humility); Flanagan Reference Flanagan1990 (for modesty)). Importantly, one's account of modesty or humility need not turn on intuitions (B) or (C) because neither entail that believing or knowing that one is humble means that one is not humble. One could simply hold, as Flanagan (Reference Flanagan1990: 427) does, that “[modesty] need not involve, ideally it should not involve, the relatively rare defect of underestimation [emphasis mine].” In short, proponents of (B) and (C) agree that propositional knowledge circumstantially matters for humility, independently of one's first-order judgements about one's own humility. However, nothing about (B) or (C) suggests that beliefs about our own humility generate any paradoxical effects.

I take it that paradoxical effects can only arise if propositional knowledge or ignorance specifically includes attitudes toward one's own modesty or humility, as opposed to one's other merits and achievements. This “strong-ignorance claim” as Flanagan (Reference Flanagan1990: 423) calls it, “requires ignorance at the second level,” about one's own virtue. Since becoming virtuous usually involves being sensitive to evidence about how we ought to cultivate virtue, one worry is that our higher-order attitudes about our own virtue will be in tension with our attitudes in expressing the virtue itself. Consider this problem: can one believe that one is humble and believe that this belief is reasonable? The straightforward answer is yes – indeed, some philosophers might argue that such coherence is rationally required (e.g., Huemer Reference Huemer2011).Footnote ²⁰ It may be incoherent to assert something like “I am humble, but humble people do not believe they are humble,” but it is an open question whether EP is incoherent in this way; the second conjunct may simply be false. This open question does not arise when we are confronted with other pragmatically paradoxical propositions, which cannot be truly believed. One cannot appropriately assert “there are no sentences” precisely because it is not an open question whether “there are no sentences” is a sentence.

What about humility's conflict with virtues that have clearer descriptions? Admittedly, practical wisdom still requires that a person have some propositional knowledge in order to act virtuously. For instance, an honest person cannot be honest without uttering propositions that they believe to be true; similarly, a generous person may need to know that their generous act is going to produce good states of affairs. Perhaps then, the humble person must have certain attitudes about their worth. And, maybe these attitudes conflict with some other important normative commitment that the virtuous person ought to have. Statman (Reference Statman1992: 426) develops this worry into a paradox:

1. (1d) It is possible that there exists a person P who is genuinely admirable and worthy, as well as modest.

2. (2d) Necessarily, either P knows he is genuinely worthy, or he does not.

3. (3d) If P knows he is worthy, he cannot be modest, since his worth is exceptionally high.

4. (4d) If P is ignorant of his value, his alleged modesty cannot be considered a virtue.Footnote ²¹

A similar, but more precise formulation of the problem is described by Hare (Reference Hare1996: 236):

According to Statman and Hare, however, the epistemic contradiction is now embodied by the following:

Flanagan (Reference Flanagan1990: 423) says that it is “extremely odd, counterintuitive, and moralistic if it disqualified a person from being modest if she knew she was.” I am inclined to agree. However, if knowing that one is humble conflicts with egalitarian beliefs, then Flannagan is mistaken, because one is apparently led to believe that they are better than others in order to be humble. In EEP, the first conjunct implies that the second conjunct is false, because the speaker really is morally superior to others and believes it. So, the epistemic paradox seems to create a normative contradiction that undermines one's commitment to egalitarianism.

This could work in generating a pragmatic paradox, but only if we thought that an individual's propositional knowledge concerning others’ merits matters for humility. But I don't think it matters. For example, consider an Olympic swimmer who has just won gold and believes the proposition “I outperformed my competitors.” Does this belief reflect negatively on the swimmer's humility? I don't see how. It seems that one's humility is not defeated by this knowledge any more than it is constituted by a lack of such knowledge. Here are two reasons to support this claim. First, beliefs about performing better in comparison to other members of the moral community do not suggest nasty beliefs; from “I outperformed my competitors,” it does not follow that the agent believes “I am better than them,” and it is a proposition of this latter sort that would undermine one's humility or commitment to egalitarianism. Indeed, there are many ways to explain why one's worth as a swimmer is not determined by the performance of other swimmers. A humble person could compare herself to an ideal swimmer (e.g., Brennan Reference Brennan2007), or think of herself with respect to humanity as a whole (e.g., Tanesini Reference Tanesini, Alfano, Lynch and Tanesini2020). However one wants to put it, the important, but often overlooked, insight is that humility is essentially noncomparative – like other virtues, its value is not owed to how we compare ourselves to other moral agents, but rather, how we grow in accordance with the ideal agent to become a better version of ourselves.Footnote ²³ This view is intuitively more plausible than any view based on the alleged tension generated by EEP.

Second, we often look to general constraints on virtue to avoid character fragmentation and acquire a coherent value system. We would think it strange if someone were to frequently make racist remarks and say they are committed to egalitarianism, not just because egalitarianism precludes racism, but because this fact can, in principle, be grasped by the racist.Footnote ²⁴ If we judge that a racist is unable to grasp this fact, we would likely judge them to have diminished responsibility, which in turn, changes whether they possess a given trait at all. Likewise, if humility is a moral virtue, and virtues are traits, it is fair to think that there must be a normative reason to be humble that one could grasp in principle, even if one does not currently recognize such a reason. On the assumption that humility is a virtue, we would expect properly humble beliefs to cohere with egalitarian beliefs (or other beliefs that are involved in one's other virtuous character traits), unless given a reason to suppose otherwise.

To summarize, I identified two errors by considering EP and EEP. The first error involves confusing knowledge-how with knowledge-that. Virtues are not explained by an agent's propositional knowledge or lack thereof, and thus, there are no pragmatic contradictions between an agent's first-order attitudes, or coherence between these and her higher-order attitudes. The second error involves mistakenly supposing some propositional attitude to have normative significance for humility when it does not. Beliefs or knowledge about one's worth do not entail that an agent has anti-humble attitudes, as we considered among the potential conflicts with the attitudes implicated in one's other traits or broader normative commitments.

4. The axiological puzzle

So far, we have assumed that humility is a virtue. We make this assumption because SAP and EP try to force us into a dilemma: to account for humility's apparent strangeness, we must either reject humility as a virtue, or accept some feature that is incompatible with traditional virtue theory. Among several features potentially at odds with traditional virtue theory, we considered whether all virtues can be self-attributed, whether humility's main feature is non-doxastic, whether virtue requires ignorance or propositional knowledge, and whether humility is at odds with egalitarianism. To resolve SAP and EP, authors generally proceed by arguing that humility is a virtue and one of these features can be accepted without damaging consequences to one's virtue theory.

The axiological problem, however, challenges the assumption that humility is a virtue. Sidgwick puts this problem well when he writes:

Here is one way that this view could be formulated into a paradoxical dilemma: either humility is not a virtue, or a virtue can direct us to have a low opinion of ourselves even if our merits are high. Sidgwick thinks that humility's strangeness privileges the first horn of the dilemma since humility is in tension with the virtue of self-respect, given one's “comparatively high” merits. Indeed, apart from those working in the Judeo-Christian tradition, philosophers have historically rejected modesty and humility as moral virtues given their apparent tension with another virtue.Footnote ²⁵ Aristotle (Reference Ross and Brown2009: 1125a33) understood humility to be “both commoner and worse” than vanity, believing instead that the ideally virtuous person, who possesses pride, accepts the goods and praises they deserve. Hume (Reference Hume and Schneewind1751 [1983]: 73–4) includes humility among the monkish virtues which “serve to no manner of purpose; neither advance a man's fortune in the world, nor render him a more valuable member of society.” Nietzsche (Reference Nietzsche, Clark and Swensen1887 [2006]: 26–7) roughly thought that the weak turn “timid baseness” into humility, where humility is a type of false self-righteousness in which one chooses mediocre inoffensiveness over curiosity.Footnote ²⁶ Although Kant (Reference Kant, Gregor, Guyer and Wood1797 [1996]: 6: 434) believed humility to be a virtue, he worried that interpersonal humility risked being “false” or “lying humility” in which one tries to “equal or surpass others … believing that in this way one will get an even greater inner worth.” Drawing from Kant, Dillon (Reference Dillon, Timmons and Johnson2015: 45) contends that humility must be constrained by self-respect, and she views humility's status as a great moral virtue on its own puzzling given the comparison between humility's low estimation of oneself against the fortitude of traditional virtues such as wisdom and courage.

For the sake of argument, I will put the axiological problem into a single utterance by placing it in tension with another virtue, using pride for illustration:

To think that AXP presents a genuine paradox, I believe, is to rely more upon one's virtue theoretical commitments as opposed to widely shared intuitions. If humility is not an excellence, one likely believes, like Dillon, that humility is on par with humiliation or self-deprecation, or that it yields few or no benefits for humanity on its own. Dillon takes such a view to be the more intuitive one – yet, in order for AXP to be paradoxical, one must show that humility and another value are mutually exclusive, be it pride, self-respect, or something else, requiring us to make additional judgements about these values. It is a fair generalization, I think, to say that philosophers have historically worked from their broader normative commitments to make sense of ambiguous traits such as modesty and humility, but to the detriment of identifying these traits’ virtuous qualities. For instance, consider the ancients’ unity of the virtues thesis: the view that you can have one virtue if and only if you have all virtues. If one accepts this thesis and the view of pride as Aristotle has characterized it, then one is essentially forced to exclude humility from the virtues. We can expect similar issues to arise when considering just how complex and nuanced the above philosophers’ moral psychological projects are – complexities that cannot be thoroughly explored in this paper. However, not only are some of these background commitments implausible (such as a strong unity of the virtues thesis), but I see no reason why they should bear upon our intuitive assessments of humility's value.Footnote ²⁸

A similar critique could be made against the Judeo-Christian tradition for treating humility as a virtue because of our special relationship to God. However, it seems to me that when humility takes a positive role in a coherent value system, its value does not strictly depend upon the system itself; the qualities that stand to make humility a virtue for Christians can be valued regardless of whether one also adopts the Judeo-Christian doctrine. I don't think that this can be said when humility is framed negatively, because we must accept theoretical commitments such as the unity of virtues thesis, Hume's category of monkish virtues, Kant's conception of self-respect, and so forth. I do not deny that the Judeo-Christian tradition similarly assigns humility positive value for its role in Judeo-Christian doctrine. However, we could imagine, for instance, a once devout Christian who inculcates humility for religious reasons but who continues to find humility morally valuable after becoming an atheist; it would not be abnormal for an atheist to view Jesus's humility as morally valuable. For worse, Jesus is rarely considered in the secular philosophical literature on humility, even though he would be a paradigm example of virtuous humility and the perfectly virtuous person.Footnote ²⁹ Interestingly too, Jesus is not shy about attributing humility to himself. He says, for example, “Take my yoke upon you, and learn of me; for I am meek and lowly in heart: and ye shall find rest unto your souls.”Footnote ³⁰ What is stranger is that Jesus's claim that he is the son of God is apparently far less humble than the phrase “I am humble” because the former implies that he is perfect and possesses all the virtues. Yet, such a claim to virtue does not seem to pose any paradox either.

What explains this? I reckon we find utterances such as “I am God's humble servant” reference broader commitments to religious doctrine but need not reference any character-level merit of the speaker. Hence, having a complete value system – one that can be represented by some ideally virtuous person – does not entail that a speaker's self-attribution must coincide with a character-level evaluative judgement. Instead, the speaker might just be describing themselves, as one could when saying “I am open-minded.” Or, it might be a higher-order value judgement, for instance, “if we want to reason well about p, we should defer to expert authority.” Neither of these propositions suggest that an agent has any evaluative attitudes toward their own merits. Returning to the apparent tension in AXP, nothing would suggest that the relevant sort of pride is an evaluative judgement as opposed to a descriptive one, or that it is an evaluative judgement of the sort at the same level of one's merits as opposed to being a higher-order judgement of oneself. As far as I can tell, these are complex theoretical commitments that are to be sorted out after engaging with intuitions about humility's value, which are by far positive, even if apparently strange.

If reasoning from the ideally virtuous person is not a convincing illustration of humility's place in a complete value system, we can also consider real world concerns about description and measurement. On the generally accepted view that modesty and humility are positive personality traits (e.g., Peterson and Seligman Reference Peterson and Seligman2004: ch. 20; Gregg et al. Reference Gregg, Hart, Sedikides and Kumashiro2008), the empirical literature would also direct us to be more cautious about treating assertions such as AXP as inherently contradictory. Consider the so-called modesty effect hypothesis, which suggests that the humbler individuals are, the more modest they are rating their own humility on self-reports. Davis et al. (Reference Davis, Placeres, Choe, DeBlaere, Zeyala, Hook, Worthington, Davis and Hook2017) note two competing approaches to this hypothesis: guilty until proven innocent, and innocent until proven guilty. Since people are likely to self-enhance on evaluative traits, some researchers opt for the highly cautious guilty until proven innocent approach, which mistrusts self-reports because they appear inherently paradoxical; as SAP and EP aimed to show, self-reports of humility amount to bragging that undermines one's humility. This conservative approach has fallen by the wayside, however, because it prevents researchers from relying on those reports to describe and assess levels of humility. Researchers who adopt the more liberal innocent until proven guilty approach are still able to locate instances in which self-reports are correlated with exaggerating one's own intelligence (Hilbig et al. Reference Hilbig, Heydasch and Zettler2014; Kajonius Reference Kajonius2014), but not because the self-reports themselves are discredited. The liberal approach allows researchers to discover the various contextual factors that can influence self-reports of humility to explain why some people might be more prone to the modesty effect than others. These valuable insights are not reached from any a priori notion that humility is paradoxical, or a theoretical framework that places humility in tension with another positive trait of character.

AXP rests on a kind of skepticism about the value of humility that, I think, shows that neither SAP nor EP are genuinely paradoxical. To accept AXP, one must believe that SAP and EP wrongly assume that humility is something excellent, because the only way to find a plausible contradiction involved considering how virtuous humility conflicts with some other thing we care about: e.g., avoiding bragging, commitment to egalitarianism, avoiding deception and vicious desires, and so on. But we generally agree that humility is excellent, or at least, that the other puzzles have merited discussion on that assumption. Once more, the relevant kind of disagreement in a pragmatic paradox is found in a fact that the speaker implies by their utterance; for good reason, we do not disagree about whether a speaker who utters “I believe nothing” expresses a belief. By contrast, when we reject AXP as genuinely paradoxical, we admit that disagreements about the nature of virtuous humility are disagreements worth having. Intuitions about humility's value ought to come first, and revisions to one's virtue theory, if any, ought to come second.