I think in some ways it’s an artificial problem - like a ‘thought experiment’, the idea of which is to bring out the nature of the dilemma inherent in Kant’s injunction. It’s somewhat like the ‘trolley problem’ which likewise poses an apparent dilemma in order to make you think through the issues involved. Pragmatically, there are, as you say, many courses of action, including not saying anything at all. But I think the idea’s value inheres in the way it makes you think through the implications of Kant’s dictum, rather than its being a practical instruction or course of action.

(Actually a couple of artworks come to mind. One is a novel that was enormously popular in the 1960’s - The Magus, by John Fowles. Another is Sophie’s Choice, which I think won an Oscar in the 1980’s. Both stories revolve around similar kinds of forced choices, and what the characters do in response to being forced to make an appalling choice.)

Kant makes almost exactly this argument, but goes even further. (You can tell them whatever you want, outright lie, since they are only interested in hearing what they want).

Alas, I cannot remember where it is, but I recall him suggesting in situations of such coercion, the person demanding the answer is only concerned with what they want to hear, rather than gathering truth. Thus, it's not really a lying to the coercier because they were never asking for turth in the first place.

you could ask the murderer for a cup of tea and a chat about how he sees his future panning out and what he is doing with his life.

I'm not sure that lying to the murderer would be using him in any way really.

Who cares if you have to lie to someone threatening your kids? Is this something you're going to feel guilty about? No. Is it something society will judge you for? No. Will there be any legal ramifications. No. Will God deduct brownie points for getting into heaven because you lied under duress?

What good does it serve to always follow Kant's maxim anyway? You can say under ordinary circumstances it's best not to treat people as a means to an end. That's a nice ideal. But it's just that, an ideal that someone came up with.

Is the point that Kantian ethics are impossibly ideal? Probably. So is the golden rule and the ten commandments.

I see it as lying/dishonesty if one isn't forthright about what one has in mind, and one instead diverts, manipulates, etc. But, I don't see lying as a categorically bad thing. In fact, I think that lying is sometimes a good thing.

You can lie when you do not do what you say, because you do not know. The murderer can say he is not a murderer if he does not know he is one (is it lying ?). Then you can learn to lie. If you know the murderer will come, you can learn what you can answer: "that your children are not there". You can also learn not to answer questions... So since we are all liars (we have learnt), I do not think Kant is right. But if you look back in history, it was a problem for the people living in south america when the spanish men came: they were not liars. They learned quickly.

We are not a remote controlled robot, we are an autonomous robot. We can do what we have learned to do (to pray god, to vote, to figth the climate...), and when we do not know, we can learn. If you have not learnt to lie, Kant is right.

Lying is only bad when it causes harm.

Lying usually causes harm but it still causes nowhere near as much harm as murder.

Anyway, to deal with the OP's attempt at a rebuttal of the common critique.

It's true that it's a false dichotomy to say that you either have to lie to the murderer or let him murder someone. But the thought experiment specifically deals with a situation where the only way to stop the person being murdered is to lie to the murderer. In such a case it is certainly preferable to lie because lying causes a whole lot less harm than murdering does.

Consequentialism is ultimately the only metaethical position that makes sense because if following certain rules or having certain virtues led to more harm than good then we wouldn't think such rules or virtues were ethical at all.

It's best to read Kant. Here, for starters:

http://bgillette.com/wp-content/uploads/2011/08/KANTsupposedRightToLie.pdf

Generally speaking, when someone resists a thought experiment by trying to argue that actually you have further options, it's possible to modify the thought experiment so as to remove those options.

For example, in this case, let's say that instead of a murderer at the door you're chatting to a random person. Unless, during that conversation, you tell that person at least a small lie, a mad genie who is watching the whole thing via a crystal ball will kill your family. The genie is unstoppable and will do this thing unless you tell a lie. The lie must be a lie in every sense.

You can't 'get out' of thought experiments by pleading other options. So long as there is at least one possible situation where your two options are lying or your family being murdered, than the thought experiment stands.

The implicit assumption in these types of thought experiments is that what we would choose to do is always ethically pure.

But surely there are times when we would choose to do an ethically wrong thing in order to save someone. Just because we can imagine a scenario where we might want to murder an innocent person (say, a chance to shoot Hitler's mother before he's born) doesn't mean we can't assert that the murder of an innocent person is still morally wrong. Likewise, we can say lying is morally wrong but still choose to lie to prevent a tragedy. It's a sort of self-sacrifice in that you accept moral guilt and shame (a small amount for a lie, a large amount for a murder) in order to save others. If we tried to redefine it to always be blameless then it would no longer be a sacrifice.

Ethics is an element of deciding what to do, but it's not the complete determinant. There are other values. They often contradict. A human being has to decide which is most important to them.

No guys. If you're talking about Kant, then you've got to read the thing. It makes a difference! If you're talking about anything else, then you're not talking about the Kant, especially about his "Murderer at the Door." And if you mean to be talking about Kant without having read, then you're both making a fool of yourself and possibly being annoying, and at the same time probably promulgating wrong information, or at the least fostering wrong impressions. Example:

But the thought experiment specifically deals with a situation where the only way to stop the person being murdered is to lie to the murderer. — luckswallowsall

And this is exactly wrong.

And also, part of the idea is that it isn't anyone's children. Another part is that you don't know where your friend is. And as to categorical imperatives (CI), they're an art not a science. If you find you're subject to more than one, you make your best choice and the other falls away. So writes Kant in his Metaphysics of Morals.

Reading between the lines, it seems to me that Kant notes that, altogether apart from the harm done by lying-in-itself, in situational lying you're taking possession of a situation, in a sense, and likely you have neither requisite knowledge, power, nor authority. As to good intentions, those pave the road to hell. Therefore, not good intentions, but right intentions, a whole other animal, and nowhere near as easy.

As to good intentions, those pave the road to hell. Therefore, not good intentions, but right intentions, a whole other animal, and nowhere near as easy. — tim wood

I have to interject here and point out that right intentions are by definition good intentions. When we say "good intentions" are no excuse in a non-technical-philosophic context (which Kantianism is not), what is actually being meant is that "plausible deniability well-wishing" is no excuse.

For instance, if someone really did not intend to cause the death of another individual, even if they physically did, it's ruled an accident; the intentions where good and so there is no liability. If someone caused the death due to creating circumstances that they could have easily avoided, even though colloquially we understand the meaning of "I didn't mean that to happen", the liability comes precisely because their irresponsible actions leading up to the accident betray an "insufficient intention to not kill someone by avoidable accident".

Though I agree otherwise with your rebuttal, there are some points you maybe missing that can add further weight to your position (which I would tend to agree is sufficient to show Kantianism isn't fatally menaced by this line of criticism; as even if it was, then that more important maxim by which the maxim not to lie is constrained would just be the more important maxim; Kant has hard time of conceiving of a more important maxim than the truth, but for good reason).

The specific murderer at the door scenario Kant addresses is first and foremost in terms of legal doctrine against a criticism that this sort of truth telling would collapse society.

Today, the best defense isn't Kant's original defense, but to point out Kant's legal doctrine is extremely popular, most people and governments accept it; so, clearly society can function with such a doctrine.

If a parent, with a gun pointed in their face, told a true answer to a criminal about the whereabouts of their children, we do not hold the parents liable. Likewise, if someone tells the truth in court under oath and this leads to harms, even if those harms can be argued to be greater than whatever the court accomplished with the truth, the truth teller is not only never liable for damages due to the truth telling, but always liable for telling a lie in court regardless of how good the excuse is. These legal doctrines are direct near mirror reflections of Kant's position.

The last case, of lying to a criminal creates a liability, is also true in legal doctrines of most countries. If one makes a lie to guide events to a better destination, one must get it right! If a lie is irresponsible and actually causes far worse outcomes, that creates a liability. Though in criminal activity under duress things get complicated and extreme and most lying would be excusable on the same grounds that telling the truth is excusable under duress. However, an adjacent example elucidates the basic principle, telling someone a lie that then leads to harm from that person acting on the assumption of that lie creates a liability, but telling the truth almost never creates a liability, if the person causes harm based on that true information the responsibility lies with them.

Also of note, Kant only says, at least as that essay goes, that the lie creates a liability, that does not by extension mean it is immoral, just one is now responsible (minimizing responsibility is not a Kantian principle).

For instance, one really can have a good enough excuse to lie under oath (in ones moral system, that may or may not be Kantian), and accept that the courts can never accept that as "good enough to excuse the crime of perjury" and must punish all cases of perjury, including one's morally justified perjury, not merely for the practical aim of maintaining credibility but the much more important reason of principle that if "perjury can be ok, then all cases of perjury can be adjudicated to determine if it's ok or not-ok instance of perjury, in further court proceeding in which perjury may also be legal and need to be adjudicated, and so on and so forth". In other-words, one can commit perjury on moral grounds while simultaneously support society's prosecution of the perjury if one is found out; this is not a contradictory position and highlights how and why morality does not equate to legality.

Lastly, an important point on this issue is that society really can function with only truth telling. Even if everyone tells the murderer the truth, that does not stop people tracking the murderer down, apprehending the murderer and removing the threat from society and deciding what to do with the murderer in a truth-telling based system of discourse. Kant was responding to a criticism that society would completely fall apart, not a criticism that in some edge cases it will seem to us justifiable to lie; since the criticism is about society functioning at all, Kant responds from a legal point of view. Would it happen sometimes that the murderer does more harm from people telling them the truth on their murder spree? probably. Likewise, does a society dedicated to the truth that doesn't tolerate lies breed less murderous rage? probably too. And furthermore, is a society so dedicated to the truth less likely to collectively murder far more people in organized and unjust wars? arguably, yes. So, like most (possibly all) points consequentialists make, it does not really resolve anything, and it's easy to defend Kant even in radical truth telling (it does not lead to absurd results, it's just uncomfortable to a society comfortable with lying in marketing, to get ahead, in politics; and we are seeing today that that comfort doesn't stay comfortable for long etc.).

I think the myth that there is a dichotomy stems from there being too many defenses of Kants view to choose from.

I have to interject here and point out that right intentions are by definition good intentions. — boethius

Hmm. I'll leave it at that.

For instance, if someone really did not intend to cause the death of another individual, even if they physically did, it's ruled an accident; the intentions where good and so there is no liability. — boethius

That's not necessarily the case. The could be liable due to negligence. It depends on the situation. Basically, you're not off the hook no matter what just because you didn't intend to kill (or maim or whatever) anyone.

That's not necessarily the case. The could be liable due to negligence. It depends on the situation. Basically, you're not off the hook no matter what just because you didn't intend to kill (or maim or whatever) anyone. — Terrapin Station

Did you read my next sentence?

Yes, but it was written in such a convoluted way that I didn't catch that you were acknowledging negligence.

It's not convoluted, I am trying to highlight the difference between the colloquial "I didn't intend that to happen" and the legal technical requirement to find fault in intention to determine criminal liability. If one can really show one's intentions where completely responsible and what seems like criminal liability is due to incompetence that oneself didn't have the competence to realize, it's possible to shift the liability up the chain to whoever hired you.

If "I didn't intend it" really is true, and really has facts to back that statement up with all the responsible steps taken that align with that intention, most legal systems don't find criminal liability; one really did not wrong in our legal doctrines regardless of the events that one "physically caused" (in at least a proximate sense).

Making this technical distinction with "I didn't intend to" in the common sense of just "well I didn't think that would happen, even if, yes, I realize it was a big risk to drink and drive and I could have taken steps to avoid that", is not a straightforward clarification, since the same words are used.

It's not convoluted, I am trying to highlight the difference between the colloquial "I didn't intend that to happen" and the legal technical requirement to find fault in intention to determine criminal liability. If one can really show one's intentions where completely responsible and what seems like criminal liability is due to incompetence that oneself didn't have the competence to realize, it's possible to shift the liability up the chain to whoever hired you. — boethius

That's convoluted, too.

The basic idea is that if you didn't take normal, "reasonable" precautions, you're going to have some degree of liability due to negligence.

The basic idea is that if you didn't take normal, "reasonable" precautions, you're going to have some degree of liability due to negligence. — Terrapin Station

Yes, that's the basic idea of negligence, but not the point I'm trying to make. The point I'm making is that the negligence is in intention despite not "wishing for bad things to happen" in our legal doctrines. Someone unfamiliar with our these legal frameworks would not immediately realize this, because society doesn't talk this way. We usually say "sure, you didn't intend for the crane to fall, but you were irresponsible in managing the crane, so much so it's criminal"; what's left out is that the determining of "irresponsible" requires intentional faults (cutting corners to save money, drinking on the job, or just laziness, the intention to provide minimal effort, at a level incompatible with what the task demands etc.).

To clarify to you and other readers why I'm teasing out that point, it's to show how much our system is Kantian.

Criminal liability is never on the person who really does have good intentions, if that can be demonstrated. Likewise, truth telling in court never creates a criminal liability. These liabilities can almost never be created due to these things regardless of consequences that then ensue. These are very Kantian principles and society functions with them.

Truth telling under duress (for ordinary citizens) almost never creates a criminal liability (nearly all cases we can imagine where it's debatable are going to be agents of the state, not the "murderer at the door" scenario Kant addresses; maybe we can put our minds together to create a situation for an ordinary citizen that has similar problems, but it's not easy to do and it will be far removed from an ordinary crime of "gun in face followed by truth divulged"). And I know of almost no argument that not only is one liable to not-tell-the-truth to the murderer, but one is liable to make up a crafty lie. So even people who defend the liar on moral grounds, I have never seen an argument that there is also legal liability to come up with a good lie in such situations (and this is the point Kant's making in the murderer-at-the-door essay).

Kant's defense (in terms of criminal liability) of the truth teller to the murderer is pre-ambled with "a yes or no answer cannot be avoided" and so Kant's playing the "what if game"; all situations I can think of where criminal liability for truth telling emerge, it is next to the option of telling the truth that one may not answer that question due to various pre-existing duties (and usually, but not always, duress can excuse those duties; i.e. if you phone up a banker and he just passes out the true details needed to access client accounts this is certainly a crime of the banker and likely civil liability of the bank, but if you kidnap the bankers family and coerce the banker to giving these true details, no criminal liability is created with the truth telling; for such kinds of situations to create criminal liability on the truth-teller, are going to require agents of the state protecting nuclear weapons, or something similarly super important, to, maybe, make those agents criminally liable for not resisting coercion, at some extremely high threshold, that they've likely previously agreed to; and again they would usually not be liable for failing to make up a good enough lie, even more extreme situations are needed for that to be in up for debate).

Our legal system is certainly Kantian, or rather Kantian ethics is modeled after the legal system. The laws of a country are intended to be so constructed that society in general should work as well as possible. In individual cases, however, the law may fail to give a person what he really deserves. It couldn’t be otherwise since a law that tried to cover all individual contingencies would be no law at all, but it also means that the law is not really moral. It says that like cases should be treated alike, but no two cases are really alike. Kantian ethics ignores this and permits individual horrible things to happen (like causing a murder by telling the truth) as long as society in general is better off if everyone follows the rules. It allows us to cause injury with open eyes and it’s therefore an immoral system.

Kantian ethics ignores this and permits individual horrible things to happen (like causing a murder by telling the truth) as long as society in general is better off if everyone follows the rules. — Congau

Kantian ethics is not really concerned with making society "better off" in some material sense. Rather, it's about being free in a positive sense, while being a member in a society. It is actually a personal system.

It allows us to cause injury with open eyes and it’s therefore an immoral system. — Congau

It allows us to allow others to cause injury. Because it takes the freedom of others seriously.

It allows us to allow others to cause injury. Because it takes the freedom of others seriously. — Echarmion

Do we really want others to be free to cause injury? Is that sort of freedom moral?

Do we really want others to be free to cause injury? Is that sort of freedom moral? — Marchesk

They're not free to do so under Kantian morals. But we are not responsible for making them into moral beings.

Yes, that's the basic idea of negligence, but not the point I'm trying to make. The point I'm making is that the negligence is in intention despite not "wishing for bad things to happen" in our legal doctrines. Someone unfamiliar with our these legal frameworks would not immediately realize this, because society doesn't talk this way. We usually say "sure, you didn't intend for the crane to fall, but you were irresponsible in managing the crane, so much so it's criminal"; what's left out is that the determining of "irresponsible" requires intentional faults (cutting corners to save money, drinking on the job, or just laziness, the intention to provide minimal effort, at a level incompatible with what the task demands etc.). — boethius

I don't think that many people really intend to be lazy, for example.

My issue lies within the dichotomy: you either have to lie, or you have to tell the murderer where your children are. — Clint Ryan

No matter how much I like Kant's "Critique of Pure Reason" for its ability to detect interesting epistemic patterns in the world of knowledge, I find his publications on ethics, such as "Critique of Practical Reason" to be much, much weaker. It doesn't seem to correspond particularly much with how morality is done in practice.

Furthermore, I personally relegate ethics and morality to the axiomatic domain of religious law, because it has a much more elaborate infrastructure for determining jurisprudential questions, i.e. the "What is right and what is wrong?" type of questions on human behaviour. In that sense, I find Kant's work on ethics also a bit irrelevant to me.

In this particular case, Kant simply fails to take the doctrine of the lesser evil into consideration.

For example, most activity in medicine revolves around administering poisons and liberally cutting into people's bodies. Therefore, absolute rules such as "You shall not administer poisons" or "You shall not chop off other people's limbs" are nonsensical. In such case, you could as well close all hospitals, because that is pretty much all they do.

When faced with a dichotomy between two evils, you simply choose the lesser one.

So, yes, there is an interdict on bearing false witness, but it will trivially take a back seat on avoiding to do something worse than that. If Kant's views on ethics cannot handle that in all obviousness, then they are simply unworkable.

For example, most activity in medicine revolves around administering poisons and liberally cutting into people's bodies. Therefore, absolute rules such as "You shall not administer poisons" or "You shall not chop off other people's limbs" are nonsensical. In such case, you could as well close all hospitals, because that is pretty much all they do. — alcontali

Where do you get the idea that only absolute, abstract maxims pass the categorical imperative?

you could ask the murderer for a cup of tea and a chat about how he sees his future panning out and what he is doing with his life. — wax

:lol: I like tea :zip:



Firstly a thought experiment specifically denies the wriggle room to the worm to highlight a real problem with a concept. So I don't think we have any other choice than the ones provided. In the murderer at the door scenario if Kant is right then we can't lie at all and must give away the location of the victims as his principle prescribes. People's feelings/intuitions about this is that to tell the truth to the murderer is downright wrong. It's this clash between Kant and moral intuition that the thought experiment attempts to expose. So, though you can say "There is a million other things you could do beside those two things." you'd be missing the point of the thought experiment.

That out of the way I'd like to give you my two cents...


If you take a fish out of the water, it's going to die.

Everyone knows that. Kantian morals, if you think it's wrong and the murderer at the door looks like it wants to prove precisely that, you have to realize that it, like the fish on land, is not in its proper environment.

What do I mean?

Take two aspects of Kantian morals:

1. People are ends in themselves. They have worth which is not depended on what can be achieved if they are used only as means

2. Categorical imperative : don't do anything that will be pointless if everyone does it

Now, imagine a society where these maxims are in full effect i.e. EVERYONE is following these basic Kantian rules.

I'm excluding the insane (psychopaths/sociopaths/mentally retarded) from this discussion. Do you think there'll be a murderer at the door looking for a victim? No! There will be no murderers, no thieves, etc. In fact there will be no immorality in such a society.

So, while the murderer at the door thought experiment does expose an issue about Kantian morals it's relevance is limited to the present moral state of society as we know it. Had it been that Kant's morals were implemented in full in ALL societies there would be no such murderer at the door situations at all.

Do you agree that Kant's morals in the world, at present, is like a fish out of water, destined to die?

We have to put Kantian morals in its proper environment - one in which all Kantian principles are in effect.

Consequentialism, despite its problems, is more suited to the present world's moral standards. It factors in, quite oddly, the absence of Kantian morals.

I see it as lying/dishonesty if one isn't forthright about what one has in mind, and one instead diverts, manipulates, etc. But, I don't see lying as a categorically bad thing. In fact, I think that lying is sometimes a good thing. — Terrapin Station

I would concur.

To further support Kant's notion of Categorical Imperative, I'd say that it's one of the best rules of thumb that I've been fortunate enough to come across. It's a way to 'measure' and/or otherwise determine whether or not some thought, belief, and/or behaviour is moral.

If everyone did 'X' what would the world look like? What would the result be? Would the world be a much better place?

The Golden Rule, which I see talked about more than I care to say, is also a good general rule of thumb. However, it suffers the fatal flaw of mistakenly presupposing that everyone likes being treated the same way...