The notion of truth is clearly central to our understanding of honesty. But we typically think of honesty and truth as merely indirectly related. We think of honesty as directly implying truthfulness but not truth itself.Footnote 1 And this is probably because truth is neither necessary nor sufficient for honesty. An assertion, for instance, can be honest without being true, and it can also be true without being honest. As Christian Miller (2020; 2021) expresses the conception of honesty that can seem to result, honesty concerns the facts merely as the agent sees them – being honest concerns “getting things right” merely subjectively, rather than objectively.

While such a subjectivist view of honesty is not unreasonable, I think we have good reason to find it ultimately inadequate. For though truth is neither necessary nor sufficient for honesty, to say that honesty implies truthfulness is to admit that honesty aims at the truth; and such an aim can have important consequences.

Consider the parallel case of belief. A claim’s being true is neither necessary nor sufficient for my believing it. Nonetheless, that belief constitutively aims at the truth has the implication that the best beliefs (what we might call “ideal beliefs”) are true – they do not just aim at the truth but, rather, achieve it. That belief aims at the truth, then, tells us something important about the nature of belief, as well as helping us to identify the best instances of it.

In a parallel way, the fact that honesty aims at the truth would seem to give us good reason to believe that ideal honesty achieves the truth as well. If it does so, then honesty does not concern the facts merely as the agent sees them. Though one could be honest to an extent while expressing what one merely believes to be true, one would fail to be ideally honest without expressing the truth itself. It would of course also follow, then, that a merely subjectivist account of honesty does not give us a full understanding of that notion, an understanding that fully illuminates the nature of honesty and the interest that it has for us.

In Section 1, I begin to argue for such a non-subjectivist view of honesty by explaining and defending the idea that honest acts “aim at the truth”. In Section 2, I discuss the implication that ideal honest acts “achieve” the truth, as well as how this sheds light on honest acts in general. In Section 3, I turn to honesty understood as a virtue, defending the claim that ideal honesty, as a virtue, reliably achieves the truth as well. And in Section 4, I discuss a similar view that has been discussed somewhat more frequently in the recent literature: the view that lying requires objective falsity. I conclude by noting some implications of the account of honesty that I offer, not only for honesty but for theories of virtue in general – and, specifically, for what we might call “internalist” theories of virtue.

Before beginning, I should note that throughout the paper my focus will be on “communicative honesty”. This is the kind of honesty involved in telling the truth and not misleading. Others have considered non-communicative honesty as well – involving acts such as giving others their due, properly complying with rules, keeping promises, and being respectful of property, as well the traits associated with these that might count as virtues (e.g., Baier (1990), Miller (2020; 2021), Roberts & West (2020)). Though I do not discuss non-communicative honesty here, I believe that the principal claim that I make about communicative honesty applies to these other cases as well.Footnote 2

1 Honest Acts Aim at the Truth

What does it mean to say that honest acts aim at the truth? That honest acts aim at the truth should be understood, first of all, as the claim that honest acts “constitutively” aim at the truth, which means that aiming at the truth is a necessary part of being an honest act. If an act does not aim at the truth, it is not honest. More contentfully, that honest acts aim at the truth means that a necessary part of being an honest act is that the performer of the act intends, in performing it, to be expressing the truth – thus, again, that if they do not intend to be doing so, their act is not honest.Footnote 3

That honest acts aim at the truth is apparent from paradigmatic cases of honesty and dishonesty. A paradigmatic case of an honest act would be saying, “Girish was playing trumpet at 10pm last night,” while believing that Girish was playing trumpet at 10pm and intending to express that he was. A paradigmatic case of a dishonest act, on the other hand, would be saying, “Girish was playing trumpet at 10pm last night,” while believing that to be false. In either case it might turn out that Girish was in fact playing trumpet at 10pm last night, but in the latter, the speaker’s lack of an intention to express the truth precludes their act from being honest.Footnote 4 In general, if someone “tells the truth”, they aim to express the truth; if someone lies, they aim to express a falsehood. And if a person does not intend to do so, their acts are not honest or dishonest, respectively.

Cases concerning being (and avoiding being) misleading – as opposed to direct lying and truth-telling – will be similar. The expression of truth involved in such cases will simply be implicit or indirect rather than explicit or direct. If one says to one’s partner, “I was at the ballet last night,” where that indirectly expresses that one was not anywhere else of interest to one’s partner, that will be non-misleading only if one believes that one was not anywhere else of interest to one’s partner, and intends for what one says to express that. If, on the other hand, one intends to be expressing such a thing while knowing it to be false, that is a dishonest act. As in cases of truth-telling, then, being non-misleading also requires intending to be expressing the truth.

What, more specifically, is involved in “intending” to be expressing the truth? In saying that the performer of an honest act intends, in performing it, to be expressing the truth, I mean not that they necessarily consciously think of their act as expressing some truth but, rather, that they necessarily intentionally perform an act expressive of a purported truth; that is, they necessarily intend to be doing something expressive of the truth. Think again of the claim that Girish was playing trumpet at 10pm. In order to be honest, the speaker of such a claim need not think, “In making the claim ‘Girish was playing trumpet at 10pm’, I am expressing the truth that Girish was playing trumpet at 10pm.” They simply need to intend to be saying what’s the case. That, we say, is the honest thing to do – so that is what one must be intending to do, in performing such an honest act.Footnote 5

To say that honest acts aim at the truth, then, is to say that a necessary part of being an honest act is that the agent performing it intends, in performing it, to be expressing the truth.Footnote 6 And we have seen that we have good reason to think that this claim is true. Why might one deny it? I’ll consider an objection on behalf of the subjectivist.

The subjectivist’s objection is that honest acts aim at expressing what their performer believes, not what is true. This objection is tempting at least in part because honest acts clearly are expressions of (at least) beliefs. It is thus initially plausible that in at least some cases, honest acts aim at expressing beliefs, rather than truths. But given that believing that p is believing that p is true, we should wonder what intending to express a belief but not a truth is supposed to look like. When might they come apart?

They most plausibly could be thought to come apart in cases of intending to express that one believes that p. I might tell you, for instance, that I believe that my friend Frida is innocent of some charge against her, and one might think that in such a case I express that I believe that p, rather than expressing that p – thus implying that it is possible also to intend to express that one believes that p while not intending to express that p. The problem with this proposal, however, is that even if “expressing that one believes that p” and “expressing that p” come apart in such cases, intending to express the former is still intending to express a purported truth: namely, that one believes that p, a purported truth about what one believes. Such an act, that is, though it does aim at expressing a belief, at the same time aims at expressing a truth. So this does not show that intending to express a belief and intending to express a truth can come apart. The subjectivist objection thus seems to fail.

2 Ideal Honest Acts Express the Truth

In the previous section, I argued that honest acts constitutively aim at the truth. If they do, we have the beginnings of an argument for the claim that ideal honest acts express the actual truth. For if honest acts constitutively aim at the truth, and if a thing’s constitutively aiming at the truth makes it the case that ideal instances of that kind of thing achieve the truth, then ideal honest acts achieve the truth. To evaluate this line of reasoning, we next need to evaluate the claim that a thing’s constitutively aiming at the truth makes it the case that ideal instances of that kind of thing achieve the truth. We will then see that ideal honest acts do achieve the truth.

The notion of “aiming” implies, first of all, that success and failure are both possible. If a thing aims at something and achieves that aim, it achieves success in that respect, while if a thing aims at something but does not achieve the aim, it fails in that respect. Moreover, if a thing achieves its aim, it is better in respect of that aim than a thing which has the aim but does not achieve it. What the idea of “constitutive aiming” adds to this is, again, that the aim is one which is a necessary part of being a given kind of thing – and, thus, that if a thing does not have that aim, it is not the kind of thing in question. Because “aiming” implies the possibility of being better or worse as concerns the aim, and because “constitutive aiming” determines a thing’s status as a given kind of thing, constitutive aiming also implies the possibility of being better or worse simply as an instance of the given kind of thing. It implies that achieving the constitutive aim makes instances of the kind better instances of the kind, whereas their failing to achieve it makes them worse instances of the kind. Where the constitutive aim is truth, then, better instances of the kind achieve the truth – meaning of course that ideal instances of the kind also achieve the truth.

Again, the paradigmatic use of this line of thinking concerns belief: because belief constitutively aims at the truth, ideal belief is belief that achieves the truth. This line of thinking tacitly relies on the claim that a thing’s constitutively aiming at the truth makes it the case that ideal instances of that kind achieve the truth. In the case of belief, the aim in question is to be true, so if a thing does not aim to be true, it cannot be a case of belief. Much like the similar aim of honest acts, this aim of belief can be cashed out in terms of an agent’s intentions: in believing that p, the believer intends to be believing the truth. And since doing so is constitutive of belief, if the believer is not believing the truth, their belief is not a case of ideal belief.

Given that honest acts constitutively aim at the truth, the claim that we have been considering thus warrants the conclusion that ideal honest acts achieve the truth as well. Whereas such success in the case of belief is the belief’s being true, such success in the case of honest acts is the honest act’s expressing the truth. So, whereas ideal beliefs are true, ideal honest acts express the truth.

What more could be said in support of this idea? I’ll consider a few points directly in its favour and then respond to a few worries about its counter-intuitiveness. First of all, the conclusion that ideal honest acts express the truth would seem to be positively supported by the basic interpersonal interest that we have in honesty: our basic interpersonal interest in the practice of truth-telling and not misleading is the interest of possessing actual truths. The interest we have in honesty (and plausibly a basic purpose of it) lies in our being able to rely on each other as concerns certain truths – either for coming to believe those truths or, more generally, for being able to secure their benefits. That is, our interest in honesty lies in being able to find out, through others, about the world – not merely in being able to find out about others’ minds, though that is also important. Honesty is important, that is, because accessing truth is important – because it is important that truths be known and, more generally, acknowledged.Footnote 7 For that reason, it makes sense that ideal honest acts enable our knowing or acknowledging the truth, rather than just what others believe to be true.

Next, despite being counterintuitive, the claim that ideal honest acts express the truth is in fact consistent with much of our pre-theoretical way of using the concept honesty. And the reason is that in pre-theoretical life, we typically have little need to think or talk about ideal honesty. Usually, it is enough if we can reliably differentiate acts that are honest from acts that are dishonest, sufficiently well that we can encourage the former and discourage and avoid the latter. In order to do so, we plausibly need a good grasp of the “threshold” for honesty, perhaps taking the form of necessary and sufficient conditions for acting honestly, or something more generic – but not the conditions for ideal honesty. The claim that ideal honest acts express the truth, then, is perfectly compatible with our usual ways of identifying honest and dishonest individuals.

Further, notice that another phenomenon that can seem at odds with this conception of honesty also is not: namely, the fact that when we say that we are being “completely” or “fully” honest, but what we thought was true then turns out to be false, we feel no need to retract our statement – we were being completely honest. Insofar as this is not a mistake, we should be able to explain how complete honesty and ideal honesty can come apart. I take it that complete honesty is relative to an individual’s current capacities, whereas ideal honesty is not. If someone has been completely honest without being ideally honest, they have been as honest as they could be, but not necessarily as honest as one can be. This distinction between completeness and perfection occurs with other types of virtue-related acts as well. One can be completely self-controlled, for instance, by remaining outside a shop in which one knows one would buy unhealthy food. But one is not ideally self-controlled unless one can go in and still resist the food. An analogous point is true of honesty.

It might also seem counterintuitive that the view being defended implies that if I gain a true belief, I become more capable of being honest. Notice, however, that this is not to say that true beliefs, all on their own, make one more honest. Expressing the truth is merely a necessary condition for ideal honesty, so one will need much more than true beliefs to be performing ideally honest acts. Conversely, of course, the view also does not imply that lacking true beliefs makes one dishonest. If one expresses a false belief, one thereby falls short of ideal honesty, but not necessarily of honesty simpliciter. So the person who lacks true beliefs will not therefore be doomed to dishonesty; they will simply be incapable of ideal honesty.

Notice, finally, that while my argument thus far can be seen as taking for granted that belief and truth are “norms of assertion” – that is, that the maker of an assertion should believe what they say (honesty being, at least minimally, the case in which the maker of an assertion does believe what they sayFootnote 8) and that what they say should be true (ideal honesty being a case in which what they say is true) – many hold more strongly that knowledge is a (or the) norm of assertion. According to such a view, one should assert that p only if one knows that p.Footnote 9 If that view is correct, then we also have a fairly direct route from the notion of honest acts to truth: if knowledge is a norm of assertion, then honest acts (as assertions) should express known facts and, hence, should express truths. And that, again, means that ideal honest acts do express truths.

3 Truth and the Virtue of Honesty

We have seen, then, that honest acts aim at the truth and that ideal honest acts achieve it. But honest acts are not the whole of honesty. There is also honesty “the virtue”, a character trait that contributes to the overall excellence of a human being and that manifests not only in the reliable performance of honest acts, but in being appropriately volitionally oriented toward such acts and otherwise appropriately emotionally attuned to their appropriate performance.Footnote 10So the honest person, amongst other things, reliably performs honest acts. My question here is whether the ideal honest person, similarly, reliably performs ideal honest acts – hence, whether they reliably express the truth, rather than what they merely believe to be true.

We can reach the conclusion that they do in much the same way that we reached an analogous conclusion in regard to honest acts: by recognizing that the virtue of honesty constitutively aims at the truth. We saw above that the claim that honest acts aim at the truth is the claim that the performer of an honest act intends, in performing it, to be expressing the truth. The analogous claim applicable to being an honest person is, in part, just a more general one: that the honest person generally intends to express the truth. They are generally truthful – they have, we might say, a standing intention to express the truth. And their having such an intention is in part constitutive of their being honest. As their virtue includes appropriate volitional and emotional orientation towards honest acts, however, this need not mean that the honest person has an intention to express the truth whatever the circumstances. Again, they intend to do so, as Aristotle says, only “at the right time, about the right things, towards the right people, for the right end, and in the right way” (Aristotle (2012) 1106b).Footnote 11

Given that achieving a constitutive aim implies being an ideal instance of the kind, we can thus conclude that the ideal honest person does not merely reliably express what they believe to be true; they reliably express the truth – “reliably”, that is, meaning not only regularly but also at the right time, about the right things, etc. And in that case, a straightforward subjectivism about the virtue of honesty must be incorrect as well. One can be honest to an extent while reliably expressing what one merely believes, but one cannot possess ideal honesty unless one reliably expresses the truth.

Thus, while some might find it tempting to think of honesty as a relatively easy virtue to acquire (in comparison to, say, justice or courage), it seems rather that, at least at its upper limit, it is actually quite difficult. Ideal honesty will plausibly require “seeking and honouring” the truth, as some have suggested of honesty generally (e.g., Williams (2002), Carr (2014), and Roberts and West (2020)), but it will require more. It will also require having found and reliably continuing to find the truth.Footnote 12

4 On Objectivist Accounts of Lying and Dishonesty

Though few philosophers defend the above sort of objectivist account of honesty, objectivist accounts of lying (and by extension, of the vice of dishonesty) are not quite as uncommon.Footnote 13 In the present section, I want to discuss how one might either assimilate or else distinguish extant objectivist accounts of lying and dishonesty from the account of honesty that I’ve defended.

Extant objectivist accounts of lying hold that lying requires objective falsity. That is, on such accounts, one is lying in asserting that p only if p is in fact false. This thesis has been defended by means of a broad range of purported evidence, from its necessity for understanding the liar paradox, to its necessity for understanding certain children’s stories, to its being the best explanation of ordinary people’s responses to questions about lie-involving scenarios.Footnote 14 Rather than attempting to evaluate all of this evidence, I want simply to offer two possible interpretations of the account itself, according to which one could either assimilate it to the account of honesty that I’ve defended or else distinguish it from that account.

To assimilate such a view of lying to the view of honesty that I’ve been defending, we must first suppose a particular relation between lying and dishonesty. Specifically, we must suppose that lying is not simply another word for being (directly communicativelyFootnote 15) dishonest but, rather, is a special way of being dishonest. In that case, the objectivist about lying could hold a view of dishonest acts that is structurally analogous to the view of honest acts which I’ve defended: a view on which, though not all dishonest acts express falsehoods, ideal dishonest acts do, and on which those ideal dishonest acts are all acts of lying. Since, on such a view, ‘lying’ is not simply another word for being dishonest, there will be acts of dishonesty that are not acts of lying – for example, dishonest acts that don’t express falsehoods but, rather, merely what is taken by the speaker to be a falsehood. Such acts will simply be called by a name other than ‘lies’ – perhaps ‘insincerities’. On such a view, one could be insincere and lie, but one could also simply be insincere, expressing what one believes to be false, even though what one believes to be false is in fact true. Lying and insincerity, on such a view, would thus be two ways of being dishonest, where these sometimes come together.

It is perhaps more common, however, to deny that lying is a special way of being (directly communicatively) dishonest and rather to hold that it is coextensive with such dishonesty – and to interpret extant objectivist views of lying as assuming this as well. In that case, we would rather have to distinguish extant objectivist views of lying from the view of honesty that I’ve defended. On such an interpretation, extant objectivist views of lying do not hold specifically that ideally dishonest acts express falsehoods (and, hence, are lies) but, rather, that dishonest acts as such express falsehoods. We might call such views “strongly objectivist” and spell out corresponding views of honest acts and of honesty and dishonesty as virtue and vice. Just as strongly objectivist accounts of lying and dishonest acts hold that such acts require objective falsity, strongly objectivist accounts of honest acts would hold that an act is only honest if it expresses the objective truth, and analogously for accounts of both honesty as a virtue and dishonesty as a vice.

Such accounts would clearly be distinct from the account of honesty which I’ve been defending, which we might rather call “weakly objectivist”. On such a view, only ideal honest acts must express the truth, which means that an act can meet the threshold for mere honesty while expressing an objective falsehood. Similarly, on the weakly objectivist account of honesty as a virtue which I’ve defended, the requirement of reliably expressing the truth applies only to ideal honest individuals. By comparison, on a weakly objectivist account of lying, the “best” acts of lying must express falsehoods (“best”, at least, relative to the aims and norms of lying), but an act could meet the threshold for being a mere lie while expressing a claim that happened to be true – and analogously for dishonesty as a weakly objective vice.

Now, for all I’ve said, it may be that a weakly objectivist account of honesty is true while a strongly objectivist account of dishonesty is. But it would be fairly surprising if that were so. Rather, it seems that if a weakly objectivist account of honesty is correct, a weakly objectivist account of dishonesty and lying should also be correct.

If a weakly objectivist account of lying is correct, the following point might explain why proponents of strongly objectivist accounts propose such a view: ‘lying’-expressions (expressions using ‘lie’ and its cognates) are ambiguous. The expression ‘That’s a lie’, for instance, can refer us either to the act of lying or to what is said in such an act. (An analogous ambiguity exists, for instance, in ‘assertion’-expressions, which can refer us either to the act of saying or to what is said in that saying.) In extant defences of objectivist accounts of lying, uses of ‘lying’-expressions sometimes alternate between these at potentially crucial points. Turri (2020), for instance, asks us to consider what can be inferred from the fact that someone believes a lie, and supposes that we can directly infer that the believer has a false belief. But the scenario he presents alternates between ‘lying’-expressions of the above two kinds:

Consider what can be inferred from the fact that someone believes a lie. A politician lies to his audience, “Iraq has weapons of mass destruction,” and they believe him. It seems to follow automatically that they have a false belief about Iraq. (ibid., p. 2103)

But plausibly, it follows automatically that the audience has a false belief about Iraq because ‘Iraq has weapons of mass destruction’ is (and is stated in the setup of the example to be) “a lie”. Plausibly, that false claim is both what is said to be a lie and what the audience believes when the politician lies to them. It’s of course correct that that claim’s being a lie is independent of whether the politician who expresses it believes it to be a lie. But that’s in principle compatible with holding that the politician only lies (i.e., only performs a dishonest act) if he believes that lie (i.e., that falsity) to be false.Footnote 16

So though in one of its uses, ‘lie’ can detach from any particular liar and become, at least roughly, synonymous with ‘falsity’Footnote 17, objective falsity may not be required for an act to be dishonest. An account of such acts, though, is what the objectivist account of lying appears to be. I don’t pretend to have shown that objectivist accounts of lying do falter on this point (not even that they should be construed as “strongly objectivist”), but if they falter, it’s plausible that they falter on the distinction between acts that are lies and the lies told in those acts.Footnote 18

5 Conclusion

I began this paper by reference to an account of honesty committed to subjectivism about honesty: Miller (2020; 2021) understands the virtue of honesty as, in part, “a character trait concerned with reliably not intentionally distorting the facts as the agent sees them”. What this means is that the honest person reliably does not intentionally distort the facts as they see them. It is the caveat ‘as the agent sees them’ that makes the account a subjectivist one. Understood as merely expressing a necessary condition or threshold for honesty, however, such a subjectivist claim seems unproblematic. Minimally, being honest does seem to require being reliable at expressing the facts as one sees them – my argument that ideal honesty achieves the truth has also relied on that claim. However, even if a subjectivist account is correct as concerns the conditions for minimal or average honesty, we have seen that ideal honesty requires more. Ideal honest acts express the truth, and agents with ideal honesty reliably do so as well. Thus, a complete conception of honesty – accounting for the whole range of its possibilities – outstrips its necessary and sufficient conditions.

What, more specifically, is the implication for subjectivist accounts of honesty? First of all, the implication is that they afford us only an incomplete understanding of the intended virtue. If otherwise correct, such accounts will ensure that we never misapply the concept honesty; but never misapplying a concept is insufficient for a full understanding of it. And if our aim in wanting a philosophical account of honesty is full understanding, rather than merely not making mistakes in applying the concept, possessing the necessary and sufficient conditions for honesty is only a step on the way to achieving our aim. (So, admittedly, is the treatment of honesty in this paper, but I suggest that it is a step further along the way.)

It is worth noting, finally, that the conclusion that ideal honesty achieves the truth presents problems not only for subjectivism about honesty but, predictably, for “internalist” conceptions of virtue more generally: if ideal honesty implies “external” success, then virtue in general cannot be understood as a matter merely of an agent’s internal state. Perhaps some virtue or other can be, but virtue as a whole, since such virtue will involve honesty, will involve external success. At least in the case of the ideally virtuous person, that is, aspects of their virtue will be world-involving. Interestingly, this conclusion is also at odds with certain conceptions of “character” in the virtue and psychological literature – namely, those which think of character traits as mere psychological mechanisms, metaphysically independent of the world otherwise.Footnote 19 On the contrary, the view of honesty that I have been defending implies that character, assuming that one’s virtue is a part of one’s character, cannot be fully understood in wholly internalist terms either. Thus, while my focus in this paper has been on a fairly narrow point concerning honesty and truth, the conclusions drawn have some important broader theoretical implications as well.Footnote 20