In his well-known piece ‘Autonomy and Benevolent Lies’ (in Thomas Hill, Autonomy and Self-Respect [New York: Cambridge University Press, 1991]) Thomas Hill argues that out of respect for people's autonomy, we ought not to tell benevolent lies. He argues that we are obligated to tell the truth, especially when asked directly for it, even if we know it will cause a person more pain. This is because truth-telling is tied to respecting autonomy, which involves giving people a realistic picture of their situation, however rosy or bleak, and letting them decide what to do with the information given. Telling people benevolent lies, because one thinks they will be happier believing something else, is an instance of paternalism, which Hill wants to resist. In this paper I will raise a case that seems at first glance to be an objection to Hill's theory i.e. a case in which the right information to give someone is not what he has requested. In offering an explanation for how Hill can respond, I will suggest that a revision to his view would allow him to justify certain benevolent lies, while still avoiding his worry that benevolent lies are necessarily paternalistic.
A benevolent lie is a concealment of truth in order to make a person experience less pain. When a person tells a benevolent lie, she is assuming that the one lied to would be happier, or somehow better off, not knowing the truth. The underlying paternalism in such behavior is what bothers Hill, so he argues that it is never right to tell a benevolent lie to a normal, rational adult, even if the truth will cause the person more pain, whether temporarily or in the long run. Because Hill thinks telling a benevolent lie is being paternalistic, it constitutes a failure to respect a person's autonomy.
Hill discusses several ways that people define the term ‘autonomy’, one of which is autonomy as an ideal for rational life. He argues that ‘opportunities for rational, self-controlled living are restricted when one does not know the realities of one's choice situations’. Hill brings up the example of a woman named Mary who asks her roommate whether the roommate thinks that Mary's ex-boyfriend would be willing to strike up a relationship again. The roommate, knowing that the ex is willing, but also believing that the relationship would bring more harm than good to Mary, chooses to lie. Mary might be a rational, psychologically competent person, and in this sense she is autonomous, but when certain highly relevant facts are hidden from her, this prevents Mary from having a realistic picture of her life, a picture that Hill thinks is necessary ‘for genuine rational control over one's life’. In so far as knowledge helps paint this picture, knowledge is a key factor in securing or boosting autonomy. This explains why failing to provide someone with the knowledge she asks for, when one could do so, is a failure to respect autonomy.
But what if it seems that providing knowledge will have fatal consequences, whereas telling a lie will increase chances of survival? Consider the following case: a man injured from a gunshot wound is rushed to the hospital by emergency medical technicians (EMTs). The wounded man asks the EMT what his chances of survival are. The EMT knows that in dire medical situations, optimism plays a big role in increasing chances of survival, so she chooses to say, ‘You'll be fine’, rather than a more truthful answer such as, ‘I can't know for certain’ or ‘Your chances are slim’, which may increase the likelihood of the patient's dying. This EMT case seems to be a slam-dunk instance in which every reasonable person agrees that the right thing to do is to tell a benevolent lie (though perhaps revealing the truth if and when the patient recovers would also be called for). Thus we have a challenge for Hill: is even this EMT case one where telling a benevolent lie is impermissible? If not (that is, if it is permissible to tell a benevolent lie in this case), then does that undermine his theory?
Hill can admit that the EMT ought to lie, while still upholding his theory, if he can show that the EMT case is importantly different from the types of benevolent lie cases he considers. Upon closer examination of Hill's examples (the aforementioned Mary-roommate case is one of several), it is evident that what Hill is worried about is the paternalistic attitude of those who are tempted to tell a benevolent lie. It is questionable, however, whether the EMT's attitude towards the patient is paternalistic in the same way that, say, Mary's roommate is being paternalistic. One line of explanatory defense Hill could use to tackle the EMT challenge is to ask what the imagined reaction of the deceived would be upon finding out the truth. If it seems that the deceived would be upset, then that is evidence to believe that one is being paternalistic and that the withholding of truth is a failure to respect the person's autonomy. If, however, the deceived is grateful for the harm he has been spared from being told the truth, that is reason to think that his deceiver has not been acting paternalistically, because the deceiver behaved in a way that the deceived would have approved of and later endorses. Therefore, benevolent lies may sometimes be warranted; it's only when the imagined reaction of the deceived is one of anger or disapproval that we have reason to think that a person who tells a benevolent lie is guilty of being paternalistic.
In Mary's case, we could imagine that she would very reasonably be angry upon finding out that her roommate had lied about the ex-boyfriend's willingness to start over, even if the ex was a no-good-man and the relationship was doomed to fail. What Mary could be angry about is the ‘I-know-what-you-want-better-than-you-do’ attitude of her roommate, which causes the roommate to conceal information. In the EMT case, however, I can't see the formerly injured adult, now alive and well, reasonably becoming indignant with the EMT for withholding the truth about the gravity of the situation. Instead, the patient is likely to say something such as, ‘thank you for giving me a better chance to live, even if it meant lying to me’. The expressed gratitude of the patient would be an indication that he didn't feel patronized and that the EMT had acted on his behalf in a way he (wounded man) would have endorsed himself, if he could have. Of course, the patient could not in fact have authorized the lie at the time, because then he would no longer be genuinely deceived, and the lie would lose all of its medically optimizing effects.
The EMT does precisely what the injured man later approves of. The case thus raises the question of whether we should privilege a predicted future state of mind over a present one. Hill doesn't consider this, because in his examples he wants to stipulate that the deceived would never find out about the lie. But if Hill would agree that a benevolent lie is justified in this EMT case, which I think he ought to, then that implies favoring an imagined future state of mind over a present state.
Mary's roommate might imagine that a future-state Mary would not endorse the past benevolent lie. Even if Mary later agrees with her roommate that the ex-boyfriend would have been no good, we can nonetheless imagine Mary reasonably saying to her roommate, ‘but you still shouldn't have lied to me. You should have let me decide for myself whether to pursue the relationship.’ By focusing on the imagined reaction of the deceived, Hill can distinguish his Mary case from the EMT case. He can acknowledge that a benevolent lie is warranted from the EMT, and he can make that concession without undermining his entire theory about lies and autonomy, because the lie in the EMT case is one that the patient who was lied to later approves of.
I am sympathetic to Hill's idea that in the name of respecting autonomy, the default position should be truth-telling, regardless of the pain it might bring. Sometimes, however, we might be justified in telling a benevolent lie when truth was asked for, without suffering the charge of being paternalistic. I have suggested that Hill could build this into his theory if he considered the idea that would-be deceivers ought to imagine the future reaction of those they plan to deceive. In particular, if the imagined reaction of the deceived (and perhaps the later actual reaction) is one of gratitude for and approval of the past benevolent lie, then that is reason to think that although one has lied, one is not guilty of failing to respect autonomy.
