¶ 1 Kant’s paradigm of duty
Truthfulness is, so to speak, Kant’s go-to duty.Footnote 1 He invokes it in a wide range of philosophical settings, such as his discussion of free will in the Critique of Pure Reason, in his argument for a pure moral theory in the Groundwork, in the detailed moral philosophy of the Metaphysics of Morals and in his late lectures on education. Even though its scope and its theoretical foundation vary, the duty not to lie remains Kant’s prime example of a strict and unequivocal obligation through the decades. The following are some examples.
In the antinomies chapter of the first Critique, Kant needs to cite an action that is freely chosen and at the same time causally determined in the empirical world. His choice falls on “a malicious lie, through which someone has brought about a certain confusion in society” (C1 A 554/B 582).Footnote 2 We can investigate this kind of behaviour psychologically, referencing the agent’s natural tendencies, upbringing, environment and the circumstances that prompted it. Yet we view the malicious lie as a culpable wrong; and that is possible, Kant maintains, only if we see it as proceeding from the agent’s underlying non-empirical character. This new perspective is open to us only if we embrace the transcendental idealism set out earlier in the first Critique. Kant’s novel metaphysics allows us to treat a lie as a free act for which the agent is responsible.
Now, one of the factors that, as Kant points out, help us account for lying as a psychological phenomenon is “the malignancy [Bösartigkeit] of a natural temper insensitive to shame” (C1 A 554/B 582). This is important. He emphasises the shamefulness of lying in his lectures on pedagogy, edited by Friedrich Theodor Rink and published by Nicolovius in 1803. Educators should nurture the sense of shame that children naturally feel when they see others lie or when they lie themselves:Footnote 3
[O]ne must never shout “Shame on you!” to them, except in case they are lying. Nature has given the sense of shame to the human being so that he betrays himself as soon as he lies. Hence if parents never talk to their children of shame except when they lie, the children will then keep this blush of shame with respect to lying for their entire lifetime.
Lying is a way of depriving a human being “of the respect and credibility for himself that everyone should have” (IX 489.14–15). Time and again, Kant asks his students and readers to impress upon their children the need to be honest with themselves and with others. In these lectures, he also rejects the permissibility of so-called necessary lies, that is, lies that in emergencies appear to offer the only way out (cf. LP IX 490.15–18).Footnote 4
The duty to be truthful takes centre stage in the Doctrine of Virtue, the part of the Metaphysics of Moral concerned with ethics as opposed to the philosophy of law. Kant draws on that duty in his argument against multiple proofs for what is – on the face of it – one and the same duty:
If, e.g., someone wants to devise a proof for the duty of truthfulness first from the harm a lie causes other human beings but then also from the worthlessness of a liar and his violation of respect for himself, what has been proved in the first case is a duty of benevolence, not one of truthfulness, and hence not this duty, for which proof was required, but another one.
It is in this book that his unrelenting stance towards lying is most explicit. Any untruthful declaration whatsoever is seen as an act of self-degradation. Consequently, veracity is classified as a duty not to others but to one’s own self. Interpreting the Book of Genesis, Kant goes so far as to indicate that he takes lying to be worse than murder: “It is worth noting that the Bible dates the first crime, through which evil entered the world, not from (Cain’s) fratricide but from the first lie (since Nature, after all, rises up against the former), and names as the author of all evil the liar from the beginning and the father of lies” (MM VI 431.5–9).Footnote 5
Evil entered the scene with the Fall, not with Cain’s murder of his brother Abel (Genesis 4:8). The Fall was provoked by a lie. According to Christian doctrine, Satan took on the guise of a snake to tell Adam and Eve that, if they ate fruit from the forbidden tree at the centre of the garden, they would not die but rather be like gods, knowing good and evil (Genesis 3:5). Kant echoes the biblical account in his 1796 essay on perpetual peace in philosophy, calling lying “the actual rotten spot [der eigentliche faule Fleck] in human nature” (CT VIII 422.4).Footnote 6 Everything would be possible if only we could wipe it away. He even suggests that the commandment not to lie, “if adopted with ardour into philosophy … would not only bring about in it perpetual peace but also secure it for all time to come” (CT VIII 422.10–12). In Kant’s eyes, there is evidently something deeply corruptive about untruthfulness.
Finally, the Groundwork. Kant makes his case for a pure foundational ethics rest on the universality and necessity of moral norms.Footnote 7 Everyone must concede, he tells us, that the command ‘thou shalt not lie’ is not restricted to human beings “as if other rational beings did not have to heed it” (GM IV 389.14–15). He infers from this that a moral system cannot be built upon anthropology, that is, an empirical account of human nature, but rather must be based on purely rational concepts. What we need is a metaphysics of morals; and the Groundwork is meant to lay the foundations of such a novel philosophical discipline. This intuition – that morality is essentially universal and necessary – underpins all of Kant’s mature moral philosophy.Footnote 8 It is thus significant that Kant should choose the duty not to lie to motivate this project.Footnote 9 Kant also uses a specific version of the duty of veracity – that of the lying promise, not of lying as such – to illustrate his newly formulated moral principle in Section I (GM IV 402.16–403.17). It resurfaces as one of the four examples employed to demonstrate the division of duties according to the categorical imperative in Section II (GM IV 422.15–36; see ¶ 8).Footnote 10
In addition, the duty not to lie serves to illustrate the all-important distinction between autonomy and heteronomy at the end of Section II. Heteronomy, Kant argues, lies at the foundation of hypothetical imperatives, of an ought that derives its rationale from ‘something else’, an object of desire, something I want, in short: matter. His example is: “I ought not to lie if I want to maintain my honourable reputation” (GM IV 441.13–14). By contrast, the categorical imperative tells me to do something regardless of what I want: “I ought not to lie even if it did not bring on me the least disgrace” (GM IV 441.14–15; cf. IV 419.21–25). The duty not to lie is purely formal. As such, it must be self-imposed if it is to have any moral authority whatsoever.
However, the far and away most prominent reference to truth and truthfulness occurs in Kant’s 1797 essay, in which he addresses the question of whether it is legitimate to lie about the whereabouts of someone pursued by a would-be murderer. Let us have a closer look at the example and its history.
¶ 2 Constant’s example: a brief history
In 1797, Benjamin Constant and Carl Friedrich Cramer conspired to elicit an unprecedented confession. Responding to Cramer’s translation of one of Constant’s essays, Kant concedes that untruthfulness is criminally wrong even if it is the only way to save a friend from being murdered. Being truthful in all declarations, he says in the essay on the ‘supposed right’, is a “sacred, unconditionally commanding command of reason” (VIII 427.24–25).Footnote 11 In Constant’s problem case, a murderous individual asks a householder about the hiding place of his intended victim, who happens to be the householder’s friend. The exchange is generally assumed to take place on the doorstep of the house to which the friend has fled. That is why, particularly in English-language scholarship, the case has come to be known as that of the ‘murderer at the door’.
Note, however, that we do not know the location of the exchange between the two men. Neither Constant nor Kant mentions a door. In principle, the murderer could run into the householder anywhere, for example, in the street, which is where by tradition the conversation takes place. It is true that the two parties need to be near the friend’s hiding place for the murderer to pose an immediate threat. But that need not be the householder’s doorstep. The expression “at home” (zu Hause, SR VIII 427.8, 11) suggests that the murderer wants to know whether the pursued party is ‘in’, which is compatible with the encounter taking place near the house, by the front door or indeed in the house. The third scenario is at least as plausible as the second.Footnote 12 For that reason, the example should not be called that of the ‘murderer at the door’.Footnote 13
Now, Constant did not, as is sometimes assumed, invent the example of the murderer, the victim and the person in the know.Footnote 14 Variations of the case – which has its roots in the Bible and in Graeco-Roman antiquity – had been dividing scholarly opinion for centuries.Footnote 15 The Old Testament contains the story of Rahab, an innkeeper or prostitute, who assisted the Israelites in capturing the city of Jericho by giving refuge to two spies whose task it was to gather intelligence about the military resources of the city. When the soldiers sent out by the king to capture them reached her house, Rahab lied. She told them the spies had already left:
And the woman took the two men, and hid them, and said thus, There came men unto me, but I wist not whence they were: And it came to pass about the time of shutting of the gate, when it was dark, that the men went out: whither the men went I wot not: pursue after them quickly; for ye shall overtake them.
Rahab’s stratagem was successful. The spies survived and the Israelites took the city. As the story raises the prospect of justifying lying on scriptural grounds, it remained popular with theologians well into the eighteenth and nineteenth centuries.
What is more, ancient philosophy is replete with similar casuistical questions. In the Republic, Plato broaches the question of whether promises are to be kept under all circumstances. Should we return weapons to a friend who has gone insane since he entrusted us with them for safekeeping? This is not a question about lying, since Plato assumes the promise was made in good faith. But the structure of the underlying difficulty is the same, which is why the two cases are often discussed together: Do we need to adhere to an otherwise stringent obligation when it is likely to cause great harm? In fact, Plato mentions the question of truthfulness in the same breath: the interlocutors agree that “we ought not to return them in that case and that someone who did would not be acting justly; nor would he who chose to speak nothing but the truth to someone who was in that state” (Republic, 331c).
More often than not, in Plato’s time as in ours, those who pose a threat are not friends gone mad but our avowed enemies. Julius Ebbinghaus reminds us that Plato’s contemporary Xenophon insists that a general who deceives an enemy is just (Ebbinghaus, Reference Ebbinghaus1954: 79; cf. Memorabilia IV.II.15 and, for a historical precedent, Cyropaedia 1.6.31). The question of whether it is permissible to deceive the enemy in times of war took centre stage in the early modern debate (see ¶ 4).Footnote 16
Constant’s example of the murderer, his innocent victim and the person who knows the victim’s hiding place first occurs, with minor variations, in Augustine’s De mendacio, the earlier of his two short treatises on lying (see Annen, Reference Annen1997: 9 fn. 17). Augustine there recounts a tale about a certain Firmus, who held the see of Thagaste, Augustine’s birthplace in Northern Africa:
When, by an order of the emperor delivered to him by messengers sent for that purpose, a man was sought who had fled to him and whom, to the best of his ability, he was hiding, he answered the messengers that he could neither lie nor betray the man.
True to his name, the bishop stood firm even under torture, as Augustine tells us he must.Footnote 17 Augustine proceeds to modify the case as follows. We may, he says, be asked not where the intended victim is – which allows us to refuse to comply – but rather whether the victim is to be found in a certain place. If so, we cannot refuse to answer because any response other than a credible lie is likely to put the victim’s life at risk. Augustine’s proposed solution is ingenious:
If, then, you do not know where the man is, there is no need of concealing the truth, but you must acknowledge that you do not know. If, however, you do know where he is …, when you are asked whether he is there or not, you must not say: ‘I will not answer your question’; you must rather say: ‘I know where he is but I will never point it out.’ [Scio ubi sit, sed numquam monstrabo.] For if you do not answer in regard to one place and say that you will not betray him, it is as though you were indicating that place with your finger, since a definite suspicion is aroused.
The modified example is much like the scenario later discussed by Wolff, Michaelis, Fichte, Constant and Kant. Note, however, that Kant compounds the problem by stipulating that the householder cannot evade saying either yes or no. That is why he cannot avail himself of any of the evasive manoeuvres devised by Augustine, Michaelis and other champions of truthfulness (see ¶ 20).
In the Enlightenment debate about (allegedly) ‘necessary lies’, philosophers by and large frowned upon Augustinian absolutism. Honesty and sincerity are important. But when it comes to emergency situations, most thinkers of the period favoured a certain pragmatism.Footnote 18 Samuel Johnson’s verdict is representative of the general mood of the time. According to Boswell, he answered the casuistical question of whether it is ever admissible to depart from the truth as follows:
The general rule is, that Truth should never be violated, because it is of the utmost importance to the comfort of life, that we should have a full security by mutual faith; and occasional inconveniences should be willingly suffered that we may preserve it. There must, however, be some exceptions. If, for instance, a murderer should ask you which way a man is gone, you may tell him what is not true, because you are under a previous obligation not to betray a man to a murderer.
Note that Johnson justifies this departure from the general rule by positing a prior obligation to the victim, even if he does not tell us what exactly he takes that obligation to be.
It would, however, be a mistake to conclude that either philosophers or the eighteenth-century public – even on the British Isles – support utilitarian solutions to quandaries of this kind. Adam Smith’s analysis of another ‘trite example’ (as he calls it) can serve as a warning. The case has its origin in Cicero’s De officiis.Footnote 19 This is how Smith introduces it in his Theory of Moral Sentiments:Footnote 20
[A] highwayman, by the fear of death, obliges a traveller to promise him a certain sum of money. Whether such a promise, extorted in this manner by unjust force, ought to be regarded as obligatory, is a question that has been very much debated.
Smith emphasises that it would be “the highest injustice” to force the traveller to pay up by law. In fact, holding the traveller to his promise would come close to saying that the highwayman had a right to extract it in the first place:
To suppose that a judge ought to enforce the obligation of such promises, or that the magistrate ought to allow them to sustain action at law, would be the most ridiculous of all absurdities. If we consider this question, therefore, as a question of jurisprudence, we can be at no loss about the decision.
Things are different, however, when we look at the case from an ethical point of view or, as Smith puts it, “as a question of casuistry”. A gentleman’s dignity and honour will fuel the desire to keep his promise:
If we consider the matter according to the common sentiments of mankind, we shall find that some regard would be thought due even to a promise of this kind; but that it is impossible to determine how much, by any general rule, that will apply to all cases without exception.
Smith adds the caveat that a gentleman would be justified in breaking his promise if he had pledged a very significant sum to the highwayman or if paying up “would entirely ruin the family of the promiser” (Smith, Reference Smith1767: 420). But we would not, Smith claims, want to be friends with someone who takes such a promise lightly. If the traveller can honour the promise without making too great a sacrifice, he should. Needless to say, the traveller should not make a lying promise, that is, a promise he does not even intend to keep. While Smith steers clear of Kant’s rigorism, there is still a surprisingly strong virtue ethical presumption in favour of honesty, a presumption that may or may not, as Bernard Williams suspects, be “a hangover from an age before the modern world [Smith] did so much to inaugurate” (Williams, Reference Williams2002: 116).
Our survey falls short of a comprehensive history of the example of the would-be murderer, its many permutations and related moral predicaments. (That would be an interesting project in its own right.) It is, however, sufficient to show that it was a standard question discussed by generations of philosophers and theologians before Constant and Cramer hurled it at Kant in 1797. We shall return to how Kant’s contemporaries tackled the problem throughout the book.
¶ 3 Kant’s response: reactions
In the 1797 essay, Kant owns up to a controversial position he has, he tells us, already defended in print, whilst that is evidently not the case (see ¶ 13); the position in question – that there is an unconditional duty of right to be truthful – is at odds with the line he takes in his 1770s ethics lectures (see ¶ 6) and in the Doctrine of Right, a book published in the same year as the reply to Constant (see ¶ 9); and the obscurity of the argument reaches almost Hegelian proportions (see ¶¶ 34– 39). It is not easy to take the piece seriously.
Unsympathetic readers are inclined to heap ridicule upon Kant. Thomas De Quincey is one of them. In his On Murder Considered as One of the Fine Arts, he uses Kant’s essay on lying as a foil to make his own rather peculiar undertaking look less objectionable. He would, De Quincey says,
so far from aiding and abetting him by pointing out his victim’s hiding-place, as a great moralist* of Germany declared it to be every good man’s duty to do, … subscribe one shilling and sixpence to have him apprehended, which is more by eighteen-pence than the most eminent moralists have subscribed for that purpose.
De Quincey does not wish to discuss murder from a moral perspective, which he declares to be his “weak side”. Rather, it is to “be treated aesthetically, as the Germans call it, that is, in relation to good taste”, which lies at the heart of his undertaking (De Quincey, Reference De Quincey1851: 21). It is hardly an accident that the identity of the ‘German moralist’ is once again revealed in a footnote:Footnote 21
Kant – who carried his demands of unconditional veracity to so extravagant a length as to affirm, that, if a man were to see an innocent person escape from a murderer, it would be his duty, on being questioned by the murderer, to tell the truth, and to point out the retreat of the innocent person, under any certainty of causing murder.
De Quincey comes close to suggesting that Kant saddles the householder with a positive obligation to volunteer the victim’s location, which is evidently not the case. Kant’s view is that the householder has no choice in the matter. If he must speak – as Kant himself stipulates, see ¶ 20 – he must tell the truth.
More sympathetic readers are naturally reluctant to attribute to Kant a view they consider indefensible. In recent decades, some eminent Kantian ethicists and Kant scholars have thus been trying to make the essay, in Robert Benton’s words, “more acceptable” (Benton, Reference Benton1982: 136).Footnote 22 Otfried Höffe, for instance, denies outright that the piece contains a “dubious rigorism”.Footnote 23 He says, “As the title of the piece shows, it treats a problem of rights, while the moral problem (the duty to be honest) is set aside” (Höffe, Reference Höffe and Farrier1994: 155).Footnote 24 Now, it is true that Kant brackets lying as an ethical problem in a footnote (SR VIII 426 fn.), thus restricting the scope of the essay to the sphere of law or ‘right’ (see ¶ 15). But that does not help us purge the essay of its apparent absolutism. First, the juridical impermissibility of lying renders it – morally – impermissible overall. Second, as the same footnote demonstrates, Kant is an equally uncompromising opponent of lying in the realm of ethics. His explicit restriction of the 1797 essay to the sphere of right does not exonerate him.Footnote 25
Philosophers who do not deny Kant’s manifest moral rigorism often seek to discount it. Allen Wood, for instance, claims that we should not pay too much attention to his hard-line prohibition of untruthful declarations. Kant’s repeated statements, he tells us, are not to be “taken at face value as expressions of his considered views”; they are “rhetorical exaggerations on the part of a moralist who is not only motivated by the philosophical aim of systematizing moral rules for theoretical purposes but also – or even instead – concerned to have what he sees as the proper effect on his audience” (Wood, Reference Wood2008: 252). Wood’s analysis is attractive, but it exacts a high price: it saddles Kant with the paradox of lying about lying to prevent further lies.Footnote 26 For asserting something without qualification what one does not fully believe to be true is also, in Kant’s books, a lie.Footnote 27
The most popular escape route, however, is to dismiss Kant’s rigorism as superficial.Footnote 28 If the reply to Constant could be detached from the core tenets of the underlying moral theory, there would be no need to accept the conclusions of the essay. Kant would simply be guilty of an error of judgement. This kind of revisionism can claim Kantian credentials, since he himself famously applies it to Plato’s theory of forms in the first Critique:
I note only that when we compare the thoughts that an author expresses about his subject, in ordinary conversation as in his writings, it is not at all unusual to find that we understand him even better than he understood himself, as he did not determine his concept sufficiently and consequently sometimes spoke or thought contrary to his own intention.
Yet those who decide to put Kant’s absolutism to one side are still left with the task of accounting for it in historical or biographical terms. The most outspoken representative of this tradition, H. J. Paton, attributes Kant’s uncompromising defence of veracity to the “circumstances and motives of the time” (Paton, Reference Paton1954: 202). For Paton it is a momentary aberration. The essay was written, he argues, “in a bad temper in his old age” (Paton, Reference Paton1954: 201), which prompted Kant to abandon the somewhat less rigid view that surfaces in the earlier ethics lectures:
[I]t is a melancholy fact that at the age of 73 a man may become more forgetful, and perhaps more petulant, than he was in his prime. Kant’s forgetfulness led him to accept as his own a doctrine he had never taught. Is it going too far to suggest that his petulance may have led him to defend it?… It looks as if Kant may have taken Constant’s criticism as a reflection both upon himself and upon his country, and we need not blame him if he was a trifle irritated.
Paton also calls attention to Kant’s upbringing to account for the alleged lapse. In old age, a philosopher may “tend to fall back on cruder views than he would have supported in his heyday”, particularly on “the less sophisticated beliefs in which he was brought up”. In a “moment of pardonable irritation”, an old man “can push his central conviction to unjustified extremes under the influence of his early training” (Paton, Reference Paton1954: 202).Footnote 29 In the same spirit, Wallace Matson wryly remarks that the 1797 essay forces us to place Kant, along with Plato, in the category of “philosophers … who lived too long” (Matson, Reference Matson1954: 859); and Mario Brandhorst concludes that Kant’s absolutism is “in more than one respect the expression of a moral pathology” (Brandhorst, Reference Brandhorst, Bacin, Ferrarin, La Rocca and Ruffing2013: 81).
But Paton’s reference to the 1770s lectures is more than problematic. Alasdair MacIntyre is right that “the Vorlesung should not be used as reliable evidence for Kant’s developed views” (MacIntyre, Reference MacIntyre and Peterson1995: 345).Footnote 30 What is more, it would be unwise to cast aside the essay too readily on account of its unwelcome absolutism. There is, as Manfred Kuehn notes, every reason to believe that Kant’s rigorism about truthfulness “represents his considered view on the subject” (Kuehn, Reference Kuehn2001: 403). If so, we had better take it seriously. No matter how alienating the 1797 piece may seem, we will never fathom it if we dismiss Kant’s reply to Constant too quickly. This is not to say that we should go out of our way to defend it, a task that has rarely been tackled, and with not much success.Footnote 31 We need to examine Kant’s argument patiently within its historical context. If it turns out to be unsuccessful, we need to pinpoint the reasons why it fails.Footnote 32 We must approach the essay with an open mind.
