The Boiling Baby Redux
Jeremy GayedIn my last essay, I promised to work through the “Boiling Baby” dilemma to reach a morally justifiable course of action for the actor in the hypothetical. It’s been awhile since the original post, so I’ll start by restating the dilemma:
Imagine that you are a federal law enforcement agent in New York. In the course of your duties, you capture an Islamic terrorist just before he can set off a nuclear explosive in the heart of the city. The terrorist tells you that another bomb has been planted in Chicago, where your family happens to live, that he knows where it is, and, unless it’s stopped, it will be detonated within the hour. The catch is that he won’t tell you the specific location. Other agents attempt to torture the information out of the terrorist, but he knows that he only has to hold out for an hour to prevail. Torture is therefore ineffective.
Another agent locates the terrorist’s newborn son. The baby is brought to the interrogation room and handed to you. One of the other agents prepares a large pot of boiling water.
You are certain of four things: (1) the terrorist is telling the truth about the existence of the second bomb and his knowledge of its location; (2) there is no way to discern the bomb’s locations before it detonates without being told; (3) if the bomb is not located and defused, it will detonate and kill hundreds of thousands of people; and (4) if you begin to boil the baby, the terrorist will reveal the location of the second bomb.
The first step in the methodolgy we’ve been using is to determine against what metestick to measure the morality of our actions. Although our options may seem limitless, there are really only three general approaches: a nihlistic ethic, in which the will of the actor is the only moral imperative; a consequentalist ethic, in which the morality of an action is judged by its outcomes; or a deontological ethic, in which the morality of an action is judged regardless of its outcome. I concluded previously in “Jesus, Jeremy Bentham, and Ann Frank” that a Christian is compelled by scripture to follow a deontological ethic. I will proceed from that conclusion.
Deontologically, our decision whether to boil the baby is judged independently of its consequences. That is, the fact that deciding not to boil the baby may contribute to hundreds of thousands of death is simply irrelevant to the morality of the decision.
Boiling the baby is self-evidently wrong, taking the action in a vacuum. We cannot, then, boil the baby, regardless of the consequences of failing to do so. (Boiling the terrorist himself is a closer question–reasonable people can disagree over what morally may be done to an aggressor to stop his aggression). We are prohibited from boiling the baby, and, by the terms of the hypothetical, can only watch helplessly as New York counts down to destruction.
Some of you will, undoubtedly, find the conclusion hard to swallow. How can the rights, or even the life, of one baby–and a terrorist’s baby, at that–outweigh the lives of hundreds of thousands of other innocent people? The answer lies in the fundamental difference between consequentialist and deontological thought.
Consequentialism, by its own terms, holds actors responsible for the results of their actions, and asks only that those results cause more good than harm. Consequentialist virtue requires accurate prediction of the results over time of any action–a Godlike knowledge that lies beyond human ability. Although strains of consequentialism recognize and seek to limit this “omniscience problem” (such as by holding people accountable only for what they actually foresaw, or what a reasonable person would foresee), these limits are unprincipled, and inherently contradict the central premise of consequentialism.
In the Boiling Baby dilemma, the consequentialist must seriously consider boiling the baby, because the harm of doing so would be outweighed by the good of saving all those lives. The chain of causality of the decision does not, however, stop with the disarming or explosion of the bomb. If the morality of the decision is a function of its consequences, moral responsibility for the boiling baby decision must continue beyond the immediate consequence of detotonation or disarmament. It is possible–arguably likely–that boiling the baby would outrage the Islamic world and inspire further attacks on the United States, resulting in greater loss of life than if the New York bomb had been allowed to detonate. Boiling the baby might thus result in greater death than refusing to do so. Alternatively, the bomb could be defective. Boiling the baby would then be reprensible, because it caused significant harm with no countervailing benefit. On the other hand, it is possible–arguably likely–that a successful nuclear attack on New York could result in nuclear retaliation by the United States, magnifying the consequences of refusing to boil the baby.
There is no way to predict with certainty the primary effects of the decision, let alone the secondary and further removed consequences, or to divine the final balance of harms and goods that will result. Consequentialism thus makes intentional virtue-and even intentional vice–impossible. We cannot know whether we have acted for good or evil until all the consequences of our actions have played out, and the benefits and harms can be weighed.
Deontological reasoning, by contrast, assigns moral responsibility only to intentional actions. We are not morally responsible for the terrorist’s decision to place a bomb in New York. We should try to stop him, certainly, but if we fail, moral culpability for the deaths lies only with he who intended them to occur.
While a deontological actor knows no better than a consequentialist the results over time of boiling–or refusing to boil–the baby, the deontological actor knows that boiling the baby is wrong. He knows that if he chooses to boil the baby, he is morally responsible for that wrong. He also knows that while he should use his best efforts to stop the bomb from exploding, he is not morally responsible if it does.
So, how can one baby outweigh the lives of hundreds of thousands of other innocent people? Deontological ethics tells us that we lack the ability to weigh those lives accurately against each other in the context of a moral decision affecting them. More importantly, we have no moral responsibility or right to weigh lives in that fashion. Our responsibility is something more attainable for mere mortals–to judge the intrisnic good or evil of our own intentions and acts, to complete those which are good, and to avoid those which are evil.

February 6th, 2008 at 8:34 am
And then, once you’ve done your best to locate the bomb in New York, you can begin to try to locate the bomb in Chicago, where your family is.
Your point about the unprincipled nature of consequentialist philosophies that emphasize the bound nature of human reason is well taken. It makes little sense to advocate an ethical theory that explicitly demands that we take ALL the consequences of our actions into account, and then declare that we really don’t have to take into account ALL the consequences of our actions. Yes, it’s impossible for anyone to know the chaos-like ripple effects that his actions will cause, but consequentialism, to be consistent and legitimate, must require that he do just that. If we find that task literally impossible, I think that says more about consequentialist philosophy than it does about us.
February 9th, 2008 at 12:07 pm
Your well-made point about deontological ethics, is that the interior act of the will, as regards the means of an action or the external part of it, are entirely unaffected by downstream considerations of consequence, which are in the reason. I.e. circumstances and calculations, cannot make an evil will good. This is very true. But the converse idea, that circumstances and calculations cannot make a good will evil, is not true. They can, and this is where deontological ethics sort of depart from classical Aristotelian ethics, and medieval Christian ethics. For instance, I think Kant (the extreme deontologist) is going insane when he argues that lying is always wrong, even when a murderer asks you for information! So we need a system to balance both deontology and externalist-consequentialism, and their respective errors. I happen to think Aristotle and Aquinas are the cure, primarily because they intellectualy protect the non-dualistic nature of goodness and evil.
They do this by maintaining that the circumstance or external object of an act can corrupt an action with good intent, because for an action to be good, it must be good by all four classical ways, by material, by object, by circumstance, and by intent. Conversely, for an action to be evil, it has only to fail by one of the four. This sort of thinking helps preserve a unitive theory for all human actions. That our reason, our will, our bodies are accountable as one unitive whole. And if you think about it, for any ethical or rational action to be truly “vountary” and include the body, the intellect, and the will, those parts must be connected in a unitive whole.
“On the contrary, Augustine says (Contra Mendac. vii), that ‘there are some actions which neither a good end nor a good will can make good.’ … I answer that, As stated above we may consider a twofold goodness or malice in the external action: one in respect of due matter and circumstances; the other in respect of the order to the end… Now it must be observed, as was noted above (19, 6, ad 1), that for a thing to be evil, one single defect suffices, whereas, for it to be good simply, it is not enough for it to be good in one point only, it must be good in every respect. If therefore the will be good, both from its proper object and from its end, if follows that the external action is good. But if the will be good from its intention of the end, this is not enough to make the external action good: and if the will be evil either by reason of its intention of the end, or by reason of the act willed, it follows that the external action is evil.” St. Thomas, Summa Theologica, I-II Q20 A2
February 9th, 2008 at 12:21 pm
It seems a fallacy to say something is self-evidently wrong, and is therefore wrong within a deontological framework. Isn’t deontological morality assigned externally? If it is not, then one agent might say, “I can’t boil this baby, it is self-evendently wrong.” And another says, “It seems like a neutral action to me…” Could you offer an argument that we have a duty that prevents the boiling of the baby?
February 9th, 2008 at 2:20 pm
Thomas Aquinas’s view of deontological right and wrong is compelling, but is not the only principled way to apply the ethic. Kant’s problem is that he pays too little attention to the meaning of “right” and “wrong” before applying his categorical imperatives. Aquinas’s solution, as you noted, is to set up a useful, if complex and formalistic, methodology for assessing human conduct.
An alternate way to address the problem is simply to pay more attention to where we are getting our notion of right and wrong. The Christian–at least the Protestant–has access both to the Bible and to the means and discretion to interpret it. Doing so closely and carefully will positively resolve most instances of Kantian absurdity.
For example, Kan’t categorical imperative against “lying” prohibits making a false statement even to prevent a murder. As noted in “Jesus, Jeremy Bentham and Ann Frank,” a close examination of Scripture shows that this definition of “lying” is overly broad, and overly simplistic. The conduct prohibited by God in Scripture is much narrower, and afford believers significantly greater latitude of action.
February 9th, 2008 at 2:41 pm
Ben, I noted that boiling the baby was self-evident only to save time. “Self-evidence” is not a valid ground for concluding that something is right or wrong deontologically; I merely took the liberty of presuming that my present audience would, applying any principled method of moral determination, easily conclude that boiling the baby was wrong.
I’m happy to use boiling the baby to demonstrate the application of the principles of the essay, however. The actor–we’ll call him “you”–is faced with the choice of boiling or not boiling the baby. You rightly reject a consequentialist approach and resolve to use a deontological ethic. You are a Christian. Your deontological approach tells you that you must do what is right and avoid what is wrong; your faith tells you what is in fact right and wrong.
You determine right and wrong by consulting the hierarchy of Christian moral authorities, starting first with Scripture. Throughout Scripture, you would find unambiguous prohibitions against harming the innocence, such as Deuteronomy 19:10 (ordering Israelites to set aside sanctuary cities so that “innocent blood will not be shed in your land”); 2 Samuel 4:9-12 (David executes two men who, seeking to gain his favor, bring him the head of one of his mortal enemies, saying, “When wicked men have killed an innocent man in his own house and on his own bed–should I not know demand his blood from your hand and rid the earth of you?”); 2 Kings 24:1-5 (listing the sins of Manasseh, which include “the shedding of innocent blood”); James 5:1-8 (condemning the rich for a multiplcity of wrongs, including murdering “innocent men, who were not opposing you”)
Based on the clear and frequent Sciptural prohibitions against harming the innocent (of which the above is only a sampling), you would conclude that you were prohibited from harming this particular innocent. Of particular interest to you would be the passage in 2 Samuel. In that passage, a man who had tried to kill David, and likely might do so again in the future, is murdered. David condemns the murderers for killing an innocent man–even though their murder likely removed someone wou harbored intentions to kill him and his children.
February 9th, 2008 at 8:57 pm
A fine response for the Christian. I am more interested in your analysis of the question independent of Christianity. From the basis of rationality alone, can you make such a claim? A related question would be whether there is any such thing as a self-evident truth without some outside source like Christianity to give you a starting point for rationality to work from. (The next question this leads to is whether any true moral framework may exist within a materialist philosophy).
February 13th, 2008 at 8:07 am
I’m assuming, Benji, that you’re referring to moral truth when you ask whether there exists any self-evident truth. Otherwise I could just say something like “You can have one of something,” the most basic assumption of mathematics. By “something like Christianity,” I take you to mean some sort of meta-framework: some system based on a source of meaning outside of the physical world. My initial response would be, “No, there is no self-evident moral truth which finds its basis entirely in the physical universe.”
The essence of positivism/materialism is that only that which we are able to physically observe exists. If we can’t observe something with our physical senses — not if we don’t observe it at the moment, but if we in principle can’t observe it — then the consistent materialist or positivist must say that this thing doesn’t exist. But given the difficulties with attempting to construct any moral system solely on the basis of materialist principles (which I discuss in my Gibberish article), it seems that a materialist self-evident moral truth would be a contradiction in terms.
I’m curious, Jeremy, in your implication that boiling the terrorist might be justified using deontological ethics. It would seem, at least at first, that such actions could only be justified by employing at least a partially consequentialist ethic. But you seem to deny that such an ethic is morally justified. Am I correct in reading this implication into the end of your argument? If so, how would you deontologically justify boiling the terrorist but not his child?
February 14th, 2008 at 4:33 pm
This reminds me of playing High School football. Our defensive coach taught that the initial defender on the ball should go for the tackle, while subsequent defenders should go for the ball while assisting on the tackle. The reason being that going for a fumble while tackling reduces the likelihood of a good tackle, and a successful fumble recovery is rare. “Of course,” said coach, “if you go for the fumble and it works, you have my blessing.” Translation: we teach acts deontologically, we judge acts consequentially. Ignorant at best, hypocritical at worst.
I agree with Paul that the tone of Jeremy’s piece is to suggest consequentialism falls short. I read the same thing.
Jeremy: doesn’t a deontological standard mean we can’t lie to the SS who asks about Jews in the attic?
February 16th, 2008 at 6:41 pm
Benji, the question concerning moral reality and moral realism is, I think, so skewed by the society and culture that we have all grown in and drawn deductions from. The reality of a true moral framework comes from social perception. As a child one of the first phrases uttered “mine” shows the ultimate human, moral framework. We are predisposed to desire that which will advance ourselves, an innate “Social Darwinism”, which drives the human feelings of greed, anger, guilt, shame, love, and desire. The change arrives when we implement an outside philosophy provided by our religion, government, culture, or family. The lens in which we look at the aforementioned human emotions is provided not by ourselves, but by the world around us. For me my family, the church, and my friends; for the terrorist his family, his Islamic faith, and his friends; for Hitler his family and his pride. The question mark appears when you ask, “What allowed Hitler to see genocide as okay?”
This one man had a skewed perspective on what was right and wrong and formed an entire culture’s social morality around his faulted perspective. The point in which the true, human, moral framework appears, occurs when our culture, religion, and government spew an ideology that not only agrees with our innate moral duty, but promotes it. The holocaust would have gone nowhere had the people of Germany, the government of Germany, and even the church of Germany not supported Hitler. Hitler saw a world filled with a perfect race as right, and he was willing to kill of innocent Jews to achieve that “rightness”. Does the boiling of a baby, in order to save a city and our family, satiate our need for “rightness” provided by our innate, human, moral framework? Does Hitler’s innate desire for “rightness” become satiated by the extermination of Jews, Gypsies, Catholics, and homosexuals? I am not sure that we can really answer this question other than by saying what Jeremy initially said, “Our responsibility is something more attainable for mere mortals–to judge the intrinsic good or evil of our own intentions and acts, to complete those which are good, and to avoid those which are evil.” If we follow our own desires, provided us by human nature, where would we be? How many would we have hurt to advance ourselves?
February 17th, 2008 at 5:38 am
Trying to find a one-size-fits-all moral framework for every situation isn’t using the full scope of human intelligence and our ability to work on a situation-by-situation basis. A moral framework is like a fossilized template - might work in some situations. It’s a comforting thing to do because it seems to diminish responsibility (I was acting according to my beliefs).
The only dilemma I have with the anesthetized baby situation is keeping the incident away from the public.
Give the baby an anesthetic and make the terrorist fully aware that he’s responsible for the boiling water.
You said: “resulting in greater loss of life than if the New York bomb had been allowed to detonate. ”
I doubt it. There are over 8 million people in New York city itself. If anything was likely to trigger a war then a nuclear explosion in NYC would probably be top of the list.
Boil an anesthetized baby or burn over a million children?
Boil the baby. Am I the only one to have made a decision?
February 17th, 2008 at 7:46 pm
If your decision to boil the baby is based on a purely rational calculus, by considering the sheer number of people saved, that still doesn’t explain why you decided to boil the baby. Nor does it begin to explain why boiling the baby was the right thing to do.
In order to “have made a decision”, and in order to explain those things, you have to presume two things: 1) that your action is rationally motivated, not just an emotional instinct (i would hope this to be the case, or i fear for your soul in quite a new way, as yet undiscussed) 2) that preserving the human race, or saving 8 million lives, or whatever, is something that ought to be done.
We can calculate all you want about the number of lives saved, but if you say that humanity ‘ought’ to be preserved in this case, you are introducing an idea which is not the result of calculation or considerations of consequence. You have gotten this idea (the ‘ought to’ part) from your listener’s conscience, and from your own conscience. Thus, your usage of this ‘felt’ idea is without rational justification.
The issue here is one of proportion: our conscience is an irrational organ, a little disorganized cluster of good instincts. The ‘practical reason’ decides which one most applies to individual cases. By boiling the baby, you have taken something in the conscience which is minor, and amplified it out of proportion to everything else. To refrain from killing innocent babies, is simply a stronger and more applicable part of the conscience, than is the duty to preserve the human race.
“If we did not bring to the examination of our instincts a knowledge of their comparative dignity we could never learn it from them.” -C. S. Lewis
February 17th, 2008 at 9:32 pm
Benji,
I had a long, long reply drafted to your question, but most of the essential points have been at least implied by the intervening comments. I will add only this: reason is only a method. Granted, it’s the method we think is the most likely to allow us to discern falsity, but reason cannot itself provide the subjects that we seek to classify as true or false, right or wrong. The subject matter of morality has to come from somewhere else. Reason tells us that it can’t come from just anywhere; it has to come from a source that’s not demonstrably false or error-prone.
February 17th, 2008 at 9:38 pm
Paul,
You asked how I would deontologically justify boiling the terrorist, but not the baby. Just to clarify, I didn’t say that boiling the terrorist was justified; I said it was a closer question. So all I can tell you at the moment are my thoughts on why boiling the terrorist might be justified.
Deontological ethics requires that I first ascertain the appropriate moral standard, then apply it. I use the Bible as the primary source for moral imperatives. As I noted in the essay, the Bible prohibits harming innocents. It does not, however, prohibit all harmful acts.
To justify builing the baby, I would have to, from scripture, conclude that it was morally permissible to harm an aggressor, or a party in league with an aggressor with unclean hands himself, in order to stop the aggression. If this was the appropriate moral standard, and there weren’t any complicating circumstances, boiling the terrorist to locate the bomb could be morally justified.
February 17th, 2008 at 9:42 pm
Tom,
We certainly can lie to the SS officer. The method I advocate asks that we first ascertain the appropriate moral imperative, then apply it (essentially with blinders on). In the Ann Frank essay, using the Bible as the primary source of moral truth, I conclude that not all deliberate falsehoods are morally prohibited; only those delivered by one who has a duty to speak the truth, to one who has a right to hear the truth spoken. The modern tendency to equate “lying” with sin automatically is the result of sloppy language, and, I fear, sloppy thinking, on the part of generations of Christians.
February 18th, 2008 at 10:13 am
I’m calling bull on you here, Jeremy. Your last sentence exhibits some pretty stunning historical revisionism.
Namely, the “modern tendency to equate ‘lying’ with sin automatically” is not a tendency (in the commonly understood meaning of that phrase) nor is it modern. Christians have understood lying as such to be sin since at least the Didache (latest commonly agreed upon date of writing: 120 AD): “My child, be not a liar … the way of death is … bearing false witness … hating the truth, loving a lie”. One could just as easily speak of Christians’ “modern tendency to equate murder with sin.”
It is, of course, entirely possible that you’re right, and that equating lying with sin is sloppy thinking. If so, though, such thinking has been a hallmark of virtually all Christians since at least the early 2nd century.
Don’t be coy about it, Jeremy: when you make your argument about lying, you’re explicitly contradicting at least 1,900 years of Christian moral teaching on the subject. It’s not something to be done on the sly.