Korsgaard's basic idea is that so long as the murderer doesn't realize that you know him to be a murderer, then lying to (people you believe to be) inquiring murderers will still be efficacious, even if universally practiced. You can publicly abide by the maxim of lying to (believed) murderers, and this murderer will still believe you -- since he assumes that you take him to be an ordinary person, to whom you wouldn't lie.
So that helps avoid absurdity in one kind of case, at least. What about transparently inquiring murderers (like the inquiring Nazi)? Korsgaard writes (p.330, fn.4):
On the other hand, suppose that the murderer does, contrary to my supposition, announce his real intentions. Then the arguments that I have given do not apply. In this case, I believe your only recourse is refusal to answer (whether or not the victim is in your house, or you know his whereabouts).
But this is just obviously inadequate in the Inquiring Nazi case. If you refuse to answer his inquiry, he will (we may suppose) storm inside and check things out for himself. You're effectively signing the hidden Jew's death warrant. Surely this is no less disqualifying than Kant's original lunacy.
Reference:
Korsgaard, C. (1986) 'The Right to Lie: Kant on Dealing with Evil' Philosophy and Public Affairs 15 (4):325-349.
What if we included in the maxim something about the justness of the murderer's cause? So, perhaps it is okay to lie to murderers if their intent to kill is taken by you (the liar) to be unjust. Even if the murderer is transparent about his intentions, presumably he can't be sure of your moral beliefs. So a maxim to lie in such cases could still be efficacious.
ReplyDeleteI hadn't realized that Korsgaard had published about this (I'd encountered her approach to opaque murderers in conversation after a colloquium). I wonder if a Kantian can make use of Kant's theory of retribution here. Of course, the big problem is that Kant thinks you can't lie because of the harm you think someone might do with the information; you can't pre-emptively treat people badly in response to bad things they may choose to do in the future, no matter how confident you are about what their future choices will be. But perhaps the person announcing an evil intention is sufficient to establish that they have it, and that you're not just making unacceptable assumptions about their choices, and so perhaps it might justify punishing them for harboring this sinister intention partly by misinforming them.
ReplyDeleteAaron - that might defang objections to lying based on the idea that it is 'unfair punishment' or some such. But what about the objection that it violates the Formula of Universal Law? To get sane results, it seems that one must reject the Kantian categorical imperative.
ReplyDeleteJoshua - this is just more pushing about the lump under the rug. The problem will re-emerge in cases where the murderer knows that you would disapprove of his murderous aims. I don't see any barrier to constructing such cases.
It seems that any Kantian response to the visiting murderer is going to fail the third step. If it becomes a universal law that visiting murderers are thwarted, why would a murderer ever come visiting?
ReplyDeleteAt least as I understand Kant (and I realize no two people understand him the same way), the way punishment in general works under the universal law formula is that you are respecting the wrongdoer as a moral agent by applying something like his own (defective) law to him, so you're still using the universal law idea in applying punishments. Hence executing murderers, etc. is supposed to all fit. In many cases, there's no exact equivalent, so Kant argues for proportional responses. I'm suggesting that while it's extremely murky what a proportional response is (one of the reasons I'm not a Kantian, another being that I don't believe in retribution anyway), punishment can certainly include things that would never be acceptable under any other circumstances; I am not sure why that couldn't include lying (it helps with fittingness if the wrongdoer's wrong action could be seen as including some form of deception, but that can probably be done with a little imagination).
ReplyDeleteThe hypothetical itself is dubious: real nazis smash in the door and might ask someone a question like that, but they hold a luger to your neck the entire time. With a luger at your head, you would probably tell the truth (or many would) even if it meant someone else would take the fall--but the real point of Kant's point on lying was that, not knowing the eventual outcome (you might lie, and they find the jew anyway) one should intend to do good (really the cat. imp. itself--) and thus tell the truth, regardless of consequences (which are often unknown, even in a luger in the head situation).
ReplyDeleteWhy can't we just change our maxim to something like: I will lie to the nazi at the doorstep SO AS NOT TO GIVE any information about the Jew's whereabouts. Is this maxim effective when universalized? Looks like it. All people lie to nazis about Jew's whereabouts so as not to give them any information. The nazi now knows this, so when he asks where a Jew is in the universalized world, whatever I say will be meaningless (I think this is fair, because the very nature of a lie is that I get to change around the meaning of my words--and it doesn't just have to be a false statement).
ReplyDeleteNow, a problem might be that I don't get to DECEIVE the nazi. What we really think is imperative is not just that I get to lie to the nazi but that I get to lie to him so that the chances of him finding the Jew go down, and not just stay the same. For example, I could tell him the Jew went to the park, when in fact he is upstairs. The nazi believes me, and so I improve the chances that the Jew gets away because I lead him down the wrong trail.
Right, we can suppose that the Nazi's default action is to search upstairs. So merely giving no information is insufficient. You need to positively deceive him, which isn't compatible with publicizing/universalizing the maxim of your action.
ReplyDeleteIt appears that the only solution to the inquiring murder is Anscombe's conception of Kant's objectivism, namely, that lying to certain people at certain times could be willed as a universal law. For instance, if the nazi situation described occurs, we could all will it to be the case.
ReplyDeleteKant makes no prohibitions against this kind of categorical, although in a response to a critic he said that lying to the inquiring murder while the victim (jew in this case) is escaping could result in the nazi bumping into the jew. But this is an implausible scenario of probability, and probability is insufficient for categoricals due to the inherent instability and hypothetical nature of them. Especially in a scene of probability as improbable as this scene in which the jew would be avoiding the nazi at all costs while the nazi is unaware that the jew is nearby.
I wrote a post about this a while ago. My basic conclusion was that you should try to minimize the murderer's chances of finding his victim by being as misleading as possible. If you universalize that maxim, you'll find that (depending on some facts about the world and some details about how you define universalization) you'll either provide no information (and have no impact on the murderer's chances) or you'll be actively misleading (and reduce the murderer's chances). That maxim passes muster with the categorical imperative, and the specific details of how you carry it out are just a matter of strategy. You have no duty to speak the truth, just a duty to be as misleading as possible, using any means at your disposal (including lies, silence, etc.).
ReplyDeleteThere are some ways to construct the scenario which make your situation hopeless - the murderer wants to look upstairs and nothing you can say can dissuade him. In those cases, it's not even clear why he bothered to ask you the question. If the categorical imperative is construed as making your decision procedure transparent, and the murderer's default action is to look upstairs, then it looks pretty hopeless.
Just to clarify, I take Korsgaard's interpretation of the categorical imperative to be that you ought to act only in ways that wouldn't be undermined by your decision procedure being universal / public knowledge. This precludes deception (since you would no longer succeed in positively deceiving the murderer if he knew that this was what you were aiming at -- he'd just ignore your utterances as so much empty noise, and go look upstairs). That is: Korsgaard's interpretation of the CI precludes deception (in the cases I've described).
ReplyDeleteBut (we may suppose) the agent in this situation could successfully deceive the murderer (since their decision procedure isn't actually transparent) -- in which case this is precisely what they ought to do. That is, the agent (obviously) ought to lie in this situation even though this act couldn't be universalized/publicized (in Korsgaard's sense). From this we can conclude that Korsgaard's moral theory is just plain wrong.
(But maybe there's some alternative interpretation of Kantian universalizability that would do better...)
Suppose the policy is to deceive those who are intending to choose wrongly? That could be interpreted in such a way that the murderer couldn't know it to apply to himself, because if he did, he'd choose otherwise.
ReplyDeleteThat's a fun suggestion. Two worries:
ReplyDelete(i) It's possible to intentionally do something one knows to be wrong. The murderer might just not care (enough). Even motivational internalists think akrasia is possible, after all.
(ii) More to the point, the agent's application of a maxim is always filtered through their beliefs, so what really matters here is whether (the murderer knows that) the agent believes the murderer has bad intentions. We can build into the case that the murderer does know this about the agent. (But for some reason his default expectation is that the agent won't lie to him -- e.g. maybe he thinks the agent is a Kantian.) In this case, again, the agent should lie even though this wouldn't succeed if he first publicized his maxim of 'deceiving wannabe wrong-doers'.
Why does the murderer get to search upstairs if you refuse to answer?
ReplyDeleteWhy not? There's a possible scenario where this is his default course of action (i.e. what he will do unless positively convinced that his victim is elsewhere), and a moral theory is meant to apply to all possible scenarios.
ReplyDeleteSo, the murderer has to coerce you then, doesn't he? You haven't given permission for him to enter the house, right?
ReplyDeleteIt seems to me that sort of coercion may very well forfeit the same kind of immunity that his deception does...
After all, Kant did say that a hindrance to a hindrance to freedom is universally acceptable...
Right, that would be a sensible view, but it doesn't seem to be Korsgaard's (if I'm reading her correctly). She seems to take it as an absolute constraint that you may only act in ways that wouldn't be undermined by your general decision procedure being universal / public knowledge -- there's no exception to the publicity constraint for dealing with bad guys; no 'forfeit'. She's quite explicit that you can only lie to the murderer in the case where he doesn't realize that you know he's a murderer -- i.e. the case in which your general policy of lying to murderers could be public without it undermining the efficacy of this lie -- otherwise, she thinks (as quoted) that "your only recourse is refusal to answer".
ReplyDeleteis kant right then to distinguish between the wrong of lying and the wrong of deception?
ReplyDelete