As the 2016 US presidential election vividly illustrated, political campaigns are frequently rife with questions of truthfulness and lying. This fact is hardly surprising: it’s perfectly reasonable to want one’s political representatives to be trustworthy, and a liar seems by definition to be untrustworthy. Whether or not it’s true that liars are necessarily untrustworthy, it is certainly true that lying carries a negative connotation. Like “coward”, “liar” is a term of condemnation: neither term can be used in a purely descriptive way; each conveys an element of criticism. This is a fact that Donald Trump used to great effect in the 2016 presidential race. One of his most repeated epithets during the Republican primaries, for example, was used to disparage “Lyin’ Ted Cruz.” In the general election, the phrase “Crooked Hillary” condensed more than three decades’ worth of attacks against Hillary Clinton into a sound bite. An effective one, as it turns out.

The irony of Trump’s victory is that he won the election by calling into question his opponents’ honesty while making far more claims that were either false or unsubstantiated than any of those opponents. And this despite the fact that the media regularly called attention to Trump’s mendacity. For example, just 30% of Trump’s statements that Politifact fact-checked were determined to be at least half true, compared to 74% of Clinton’s. Somehow, though, the electorate’s perception does not appear to have tracked the reality of circumstances.

We think philosophy has a role to play in identifying and correcting the disconnect between perception and reality with regard to politicians’ trustworthiness. By providing a theory of lying and truthfulness that is sensitive to lived experience, philosophers can help people to avoid talking past one another – as the media and wide swaths of the electorate seem to have done during the last election – when discussing such important issues. Armed with a satisfactory theory of lying, people can avoid simply relying on gut intuitions about what to believe and whom to trust. This can help people to avoid falling for it when others try to play on their intuitions, as political campaigns so frequently do.

The locus classicus for much of the philosophy of lying is Augustine (354-430) who was one of the first to define a lie explicitly as consisting in the intent to deceive – or as he poetically phrases it, “having one thought in the heart and another on the lips”. Even if many philosophers today do not share Augustine’s other commitments – such as his teleological view that lying is wrong because it violates the purpose for which God granted us language – the basic conception of a lie as involving a deceitful intent continues to shape (indeed, to constrain) contemporary thinking on the subject.

Many of the rich thought experiments Augustine uses to motivate his view have become archetypal in the literature. In one scenario, he describes a magistrate outside your door, inquiring whether you are harbouring a wanted murderer, which ex hypothesi you are. Despite acknowledging what he sees as a Christian duty to shelter anyone in need and to help our fellow sinners, Augustine insists that a false statement about the murderer’s whereabouts is a lie and a sin. Indeed, even if you know that the accused murderer is innocent or the magistrate corrupt, intentionally deceiving the latter is still a lie.

One thousand two hundred years later, Kant further develops the Augustinian line that lying is always impermissible. He repeats Augustine’s argument that it is contrary to the nature and purpose of language to use words to communicate thoughts that are contrary to what one actually believes. But unlike Augustine, Kant attempts to establish this position without relying on scripture. Kant argues that the moral law – including the absolute prohibition on lying – is derivable from reason alone, and not merely from divine revelation.

In the Groundwork for the Metaphysics of Morals, Kant uses the example of an insincere promise to illustrate what he takes to be wrong with lying more generally. The success of the institution of promising is predicated on the fact that people generally keep their word. A moral law that permitted people to make insincere promises in order to get out of inconvenient circumstances would thus undermine itself, as the universal practice of such a policy would lead to promises never being believed. The fact that a policy of insincerely promising cannot be universalised demonstrates that doing so necessarily violates the moral law.

For Kant, the problem with insincere promises is not unique to promising: any lie is going to be similarly self-undermining. In the 1799 essay “On a Supposed Right to Lie from Philanthropic Concerns”, Kant defines a lie as an “intentionally untruthful declaration”. A declaration, as Kant uses the term, is an utterance that purports to reflect one’s thoughts, and in so doing invites someone to take said utterance as truthful. Essentially, a declaration is an utterance that avows truthfulness. In The Metaphysics of Morals, Kant identifies truthfulness with honesty and sincerity: an utterance is truthful if it accurately represents one’s thoughts (including one’s beliefs), regardless of whether those thoughts are themselves accurate. An utterance is untruthful if it misrepresents one’s thoughts. To make an untruthful declaration – that is, to lie – is thus self-contradictory because it involves avowing that one’s utterance contains precisely that which it does not contain: it amounts to saying, “I believe what I do not believe”, or “I intend what I do not intend”.

In “On a Supposed Right to Lie”, Kant echoes Augustine in arguing that it is impermissible to lie even to a person who wants to kill your friend who is hiding in your home. While Kant is roundly criticised for this view, such criticism often misses its mark by overstating the reach of Kant’s prohibition. His view may be rendered less implausible when the narrow definition of lying that Kant has provided is taken into account. He does not, for instance, insist that you must tell the would-be murderer where your friend is; only that you mustn’t lie. So perhaps you should simply refuse to answer; or call the police; or try to talk to the person who wants to harm your friend and see if you can change their mind; or, if you think you can, wrestle the weapon away from them. Another option that Kant allows for is engaging in deception that falls short of lying – through careful word choice or evasion, for instance.

With this added clarification, Kant’s (and Augustine’s) prohibition on lying does seem to align with many people’s intuitions. Faced with a situation in which truthfulness would cause harm or be otherwise objectionable, it is not uncommon for people to respond by telling a half-truth or making use of Gricean conversational implicature in order to mislead without telling an outright lie. Similarly, it is uncommon to find people who are willing to admit openly to a lie. One is reminded here of Ryan Lochte’s admission, after the incident at a Rio de Janeiro gas station during the 2016 Olympic Games, that he “over-exaggerated” about what happened. For many of us, at least, there is a psychological barrier to admitting that one has lied. This psychological barrier is consistent with a belief that lying is not just presumptively wrong, but morally unjustifiable.

But not all philosophers hold such a pessimistic view of lying. Or, rather, it may be better to say that not all philosophers hold such an optimistic view with regard to truthfulness. Two scenes from the Mahābhārata, the great Indian epic, provide powerful illustrations of what can happen when the insistence on truth-telling as a virtue is taken to its logical extreme. In so doing, they provide a powerful argument against the Augustinian-Kantian view.

In the first scene we will consider, the hero Yudhiṣṭhira is fighting a war to defend his status as rightful heir to the throne. One necessary condition for winning the war involves convincing the commander of the opposing army, Droṇa, that Droṇa’s son Aśvatthama has been killed. Unable to kill Droṇa’s son, Yudhiṣṭhira’s brothers kill an elephant of the same name and then yell for Droṇa to hear, “Aśvatthama’s been killed!”. Because he suspects a ruse, Droṇa demands to hear the truth from Yudhiṣṭhira’s mouth: only from Yudhiṣṭhira will he believe it, because Yudhiṣṭhira is widely known for being honest: indeed, his chariot has always floated a few inches above the ground because of his unfailing honesty. Thus faced with a dilemma – tell the truth and lose the war or lie and keep alive the hopes of victory – Yudhiṣṭhira tries to extricate himself by choosing his words carefully, so as to deceive without lying: he says, “Lord, he is killed,” subsequently uttering under his breath the words, “the elephant.” Having uttered the literal truth but in a deceptive manner, Yudhiṣṭhira succeeds at deceiving Droṇa – but at the expense of losing his status as someone who is unfailingly honest. As he answers Droṇa’s question, Yudhiṣṭhira’s chariot falls to the ground.

The second Mahābhārata scene that is relevant here argues directly against Kant’s analysis of the would-be murderer case. The scene tells the story of Kauśika, an ascetic who has gained fame for upholding a vow to always tell the truth. One day some bandits come along, chasing travellers through the area of the forest where Kauśika lives. When the bandits appeal to Kauśika’s famed truthfulness and ask him where the travellers went, he tells them. According to the Mahābhārata, this act of truth-telling leads not only to the travellers’ deaths, but also to Kauśika’s punishment in the afterlife. His sin, the argument goes, lies in valuing his own reputation for truthfulness above even the lives of innocents. In telling the truth, Kauśika violated the more important role he had as a human being who has the opportunity to protect others. On this analysis, an unwillingness to lie in order to protect others – whether innocent travellers in the forest or your friend hiding from a murderer – reveals a sort of perversity, a fetishisation of the truth.

Kauśika and Yudhiṣṭhira have both mis-valued truth. They’ve failed to recognise that the value of truth lies not in uttering true words, but rather in truth’s conduciveness to trustworthiness. Each has been untrustworthy: Yudhiṣṭhira despite his truth-telling, and Kauśika in virtue of it.

One need not look far in the recent election to find people who have made the same mis-valuations of the truth as Yudhiṣṭhira and Kauśika. Consider, for example, FBI Director James Comey. On October 28, 2016, just eleven days before the presidential election, Comey wrote a letter to the members of Congress, explaining that the FBI had reopened the investigation into Hillary Clinton’s email server that in July he’d said was completed. In doing so he went against two longstanding FBI policies: not to comment on ongoing criminal investigations, and not to make any announcements within 60 days of an election if doing so could reasonably be interpreted as interfering with said election. The widespread criticism he received for deciding to make the announcement notwithstanding, doing so was the only way for him to avoid violating the pledge he’d made in July to keep Congress updated of any changes in the status of the investigation into Clinton’s email server.

Once the potentially relevant emails were found on Anthony Weiner’s computer, Comey found himself in a no-win situation. If he revealed the information, he stood to be criticised for attempting to influence an election. If he withheld the information, he stood to be criticised for not fulfilling his pledge to keep everyone informed of new developments. This is Kauśika’s dilemma: there is no way for him to be both true to his word and true to his role. Ultimately, Kauśika’s mistake lay not merely in speaking the truth to the bandits, but in recklessly making the vow to begin with: it’s his vow that presents him with the dilemma. So, too, for Comey. His problem was not his response to the emails, it was the fact that he made a commitment that could so easily come to be in conflict with other commitments that he already held. So while it’s clear that Comey told the truth, there’s a good case to be made that in so doing he failed to remain true to his role as FBI Director.

If Comey is a real-life Kauśika, Donald Trump can be understood as analogous to Yudhiṣṭhira: saying what he needs to in order to accomplish his goals, but carefully selecting his words in an attempt to maintain deniability when called to account. One of the most notorious examples of this occurred in Trump’s criticism of Megyn Kelly after the August 5, 2015 Republican presidential debate. Kelly had questioned Trump at length about his positions on women’s issues and comments he’d made that many take to be sexist. Referring to this line of questioning, Trump later said on CNN that Kelly had “blood coming out of her eyes, blood coming out of her wherever.” When people criticised Trump for making such a statement, he denied that he was making a veiled reference to Kelly’s sexual anatomy, claiming in a tweet that he “obviously” was referring to her nose. The denial notwithstanding, Trump’s meaning is clear. But it must have also seemed clear to Droṇa whom Yudhiṣṭhira was referring to by saying, “he is killed.” Both speakers, in choosing their words so carefully, have attempted to preserve plausible deniability.

As long as we stick to the Augustinian-Kantian approach, wherein whether one lies depends wholly on the properties of the utterance in question, we have to admit that Trump didn’t lie. The words he uttered never declared that Kelly was on her menstrual period. And if we understood his utterance to be conveying such a message, the responsibility for doing so is ours for being careless enough to draw the inference. But if such an analysis would be unsatisfactory with regard to Yudhiṣṭhira, as the Mahābhārata indicates it is, so too is it unsatisfactory with regard to Trump. If there is a moral presumption against lying, then you cannot be exempted from criticism for an obviously deceptive utterance merely because there is a non-standard interpretation of that utterance according to which it is true. If you could, then no utterance could ever be soundly criticised as a lie, for there is always an alternative interpretive scheme that could enable the utterance to come out as true. Neither Kant nor Augustine appears able to avoid this untoward consequence of what we might call the linguistic approach to lying.

While Augustine and Kant may be correct in the abstract about the significance and ethical impermissibility of the intent to deceive, the Kauśika and Yudhiṣṭhira examples (as well as the analogous anecdotes from the recent election cycle) reveal the absurdity of this position when taken to its logical extremes and applied to particular cases. What we need, therefore, is not further refinement concerning what technically counts as a “declaration”, or other such semantic nitpicking, but a way of resituating the concept of lying within the more realistic roles that real people actually occupy. A possible framework for this may be found in Confucianism, where roles and relationships are taken to define a person, her duties, and the context and meaning of her actions and utterances.

In one passage from the Confucian Analects, for instance, Confucius is asked at different times and by two different disciples what the proper course is regarding fulfilling one’s duties. He gives quite different advice to each disciple, instructing one to always act promptly and the other to move slowly and only after other filial obligations are fully discharged. At this point, we can easily imagine an Augustinian or Kantian either accusing Confucius of intentionally deceiving, and thereby lying to, at least one of the disciples (even if we forgive this deception by appeal to pedagogical effectiveness), or else finding some way of interpreting his assertions as technically not lies in the first place. In contrast to these options, Confucius goes on to explain his discrepancy by saying that he tailored his recommendations to what he knew of the different personalities of the two disciples. Rather than lying and truthfulness hinging on the impersonal logic of assertions and declarations, Confucius’s words reflect an appreciation of the different particularities of each disciple and a respect for the intimate and unique teacher-student relationship he inhabits with each individual.

Similarly, in another passage from the Analects, when Confucius is confronted by a certain court official and asked point-blank whether the Duke of his home province properly understands ritual, Confucius replies with a curt affirmative and then promptly retires. As the rest of the anecdote makes clear, the Duke seems not to have properly understood ritual, and Confucius knew that he didn’t. Yet Confucius was also under an obligation to formally respect his liege lord. Again, the Augustinian-Kantian impulse may be to cast Confucius’s affirmative as a lie, or to find a logic-chopping way of getting it off on a semantic technicality as not a “declaration” per se. Attending to the importance of roles, however, allows us to see that it was the initial court official who erred, by putting Confucius on the spot with his tactless and public question. Confucius responded with ritual appropriateness to a question that was itself ritually inappropriate, and in a way that was true to his role as subject to the Duke. Yet the terseness and subsequent retirement also unambiguously communicate his real negative assessment of the Duke. As we say, such actions “speak louder than words,” and in a way that only those attuned to the particular relational dynamics of the situation – the conflicting duties, the different audiences and witnesses, etc. – can truly appreciate.

In a more recent context, consider the way that Hillary Clinton’s hacked speeches were weaponised against her during the 2016 election as evidence of her being “crooked”. In one famous leaked passage (from a speech to the National Multi-Housing Council in April 2013), Clinton said that “you need both a public and private position”. This quickly became a rallying cry for her opponents, who construed it as evidence of a penchant for deceptiveness. Saying one thing publically while another thing privately does, after all, bear striking similarity to “having one thing on the lips and another in the heart”. In this way, both her opponents and even her own defenders (who quickly tried to divert attention from the content of the quotation to the unlawfulness of the hack) shared in the intuitions bequeathed by the Augustinian-Kantian tradition. A Confucian alternative would instead focus on Clinton’s role during those speeches, and the difference between speaking as a private citizen (she had left the State Department two months prior) and speaking as a representative of the government. Too often when we complain about being “taken out of context”, the reference is solely to other parts of the text in question. But another essential dimension of “context” is the particular roles we and our audiences occupy.

The flip side of this can be seen in the comments given by Corey Lewandowski, then campaign manager for Donald Trump, in defence of accusations that his candidate had lied. Lewandowski (at a panel at Harvard’s Kennedy School of Government on 1 December 2016) complained, “You guys [the press] took everything Donald Trump said so literally… The American people didn’t. They understood it. They understood that sometimes – when you have a conversation with people, whether it’s around the dinner table or at a bar – you’re going to say things, and sometimes you don’t have all the facts to back it up.” While Augustinians and Kantians would target the formal inconsistency of differing assertions made to the press and those made at a bar, Confucians would instead emphasise the two very different roles operative in these contexts. A presidential candidate, in any of their utterances, is ipso facto speaking in a particular role, defined by multiple audiences, one of which is the press. No matter how we might wish our candidates to feign populist accessibility by visiting the bars regular citizens frequent, this is performance: they are not – there and then, at least – private citizens, but are speaking as political representatives, as Confucius was qua Duke’s subject. Within such a role, the American people whom Lewandowski says “understood” Trump’s utterances only did so by misunderstanding the role of their candidate.

Appreciating the way in which roles contextualise the ethics of speaking will presumably not remove all partisan misinterpretation, and it certainly won’t make candidates suddenly stop deceiving. The hope, though, is that by broadening our vocabulary regarding lying to include more than merely the intent to deceive, and by attending to the broader context of the utterance at least as much as to the formal properties of the utterance itself, our political discourse will at least be able to move past the otherwise too-easy condemnatory rhetoric of simplistic sound bites and tweets.

Kevin DeLapp is the Fleming Professor of Philosophy at Converse College (South Carolina). He is the author of Moral Realism and, with Jeremy Henkel, Lying and Truthfulness: A Reader . His main interests are in metaethics and Chinese philosophy.

Jeremy Henkel is an assistant professor of Philosophy at Wofford College (South Carolina).He is co-editor, with Kevin DeLapp, of Lying and Truthfulness: A Reader. His main areas of interest include Buddhist philosophy and the intersection between philosophy of language and ethics.