Part I: The Problem
What does it mean to do the right thing? Each day we face hundreds of decisions with moral implications. In deciding how to act on these matters, how do we determine what the moral option is? What metric should we use to measure the respective morality of each prospective action?
Not everybody agrees on the answer to that question. Today, I’d like to propose two competing moral frameworks, and then compare the benefits and problems with each. Essentially, the basic question at hand is this: are actions themselves categorically moral or immoral, or does the morality of an action depend on the context and the intended results of the actor? Should we establish a fixed code that says which actions are moral or immoral first, and then apply that code to each situation as it arises? Or should we take each unique situation as it comes and act so as to attain whatever outcome is most moral?
Those are kind of abstract ideas, so let’s apply them to scenario. Let’s say there is an innocent child in front of you, who has done nothing wrong. And let’s say there are three other innocent children whom somebody else is holding at gunpoint. You don’t know any of these four children. The gunman offers you a choice: if you shoot this innocent child in the head in the next two minutes, I will let these three children go. If you don’t, I will kill these three children and let the other one go. Whoever this twisted bastard is, it is impossible for anybody to stop him (yourself included), and you are 100% certain that everything he says is true. So…do you shoot the child?
The first moral framework I discussed would say no. If morality is determined by actions, then the action of killing an innocent child is clearly immoral. The second moral framework I discussed would say yes. If morality is determined by results, the result of one dead child is clearly morally superior to the result of three dead children. Which is correct?
Part II: The “outcomes” perspective
There is no “right” answer to that question; either decision would be justifiable. On the one hand, when it’s presented in this abstract manner, I think most people’s first inclination is to look at it practically: save three kids by killing one. Logically reading this blog, without actually holding the gun or seeing the face of the child in front of you, it seems like simple arithmetic. Ultimately, whichever children are killed don’t much care who is pulling the trigger. If you allow those three children to die, how could you live with yourself knowing that you had the opportunity to save two lives, but didn’t? And what if it weren’t three children that stood to die by your inaction, but five? Or ten? Or fifty, or a hundred, or a hundred thousand? No matter how high the number becomes, the first moral framework would say you can’t kill the child in front of you, because killing is wrong. After a certain number, that viewpoint becomes downright unacceptable. And if we separate ourselves from the agent even further by imagining that somebody else was presented with this decision, our preference becomes even clearer. Which news story would you rather read: that a hundred children were killed today, or just one? It seems like a no brainer: the just outcome is the minimization of death, so you should take whichever action is necessary, however horrible that action may be, to bring about that outcome.
In philosophy, this school of thought is called “consequentialism”. Consequentialism argues that the consequences (or at least intended, foreseeable consequences) of one’s actions are the only metric by which the morality of those actions should be judged. This is the “results” side of the argument. Consequentialists feel there are no moral or immoral actions without viewing those actions in the context of their results. There may be moral or immoral intentions, but those intentions must be defined by the outcome they intend to bring about. One variation of this is “utilitarianism”, which essentially advocates whatever decision is likely to bring the greatest amount of happiness and pleasure to the largest amount of people or living beings. These theories were historically advanced by John Stuart Mill and, more recently, by Peter Singer.
Another thought group supporting the “results” argument is the political realism movement. As a whole, realism contends that truth exists independently of human perception, and thus independently of purely human constructs like ideals or morals. Like consequentialism, it is very results oriented. Unlike consequentialism, realism rejects human motives, morals, or intentions as irrelevant to these results, whereas consequentialism views them as the metric by which to evaluate those results. Political realists claim that morality is so subjective, and that human motivations for any action are oftentimes so bad and selfish anyway, that morality should essentially be ignored. Practical goals should be prioritized over subjective principles of right and wrong, and agents should simply pursue their own objectives without regard for ideology. Historic advocates of political realism include Morgenthau and Sun Tzu, but Niccolo Machiavelli is often cited as the original political realist. In his most famous treatise, The Prince, he applied this perspective to the life of a ruler. Machiavelli argued that virtually any means to keep oneself in power were justifiable, so long as they served the purpose of protecting the ruler’s ability to effectively rule. In the section of The Prince titled Of Maintaining a Princedom, he summarized this outlook:
“Anyone who would act up to a perfect standard of goodness in everything must be ruined among so many who are not good. It is essential therefore for a prince to have learnt how to be other than good and to use, or not to use, his goodness as necessity requires…the prince, in short, ought not to quit good courses if he can help it, but should know how to follow evil courses if he must.”
Machiavelli summarized this viewpoint by coining one of the most famous, hotly debated philosophic phrases of all time: the ends justify the means. To assert that the ends justify the means is to assert that an otherwise immoral action is justified if it is undertaken to attain a moral result. That statement bears enormous implications on how we ought to live. Is it right?
I think it’s clear that at least sometimes, it is. It can’t be the case that the ends never justify the means, because we need some mechanism to minimize the damage caused by those who don’t follow the rules. Not everyone agrees with me; there are some who say that the ends never justify the means at all. This is perhaps the origin of modern pacifism and anarchism, the notion that using force on others is the ultimate wrong and can never be justified. People of this line of thought cite men like Jesus or Gandhi as examples of how to fight injustice without resorting to immoral actions. Practically, however, such alternatives oftentimes do not exist without incurring preventable negative consequences that most people are unwilling to accept. For instance, in the example I gave above about the gunman, most people agree it would be a mistake to not kill one child to save 1,000. Absent context, it is immoral to kill; but if a gunman enters a school and opens fire in the cafeteria, and you have the power to stop the massacre by killing him, you should do it. Absent context, it is immoral to wield force on others; but if a policeman finds a rapist or pedophile or burglar and has the power to arrest him and prevent him from harming others, he should do it. And absent context, it is immoral to lie, but sometimes white lies in response to questions like “does this dress make me look fat?” or “is Santa Claus real?” or “where do babies come from?” can do more good than harm. Humans are flawed. They sin, and make mistakes, and screw things up. If it is possible to correct those mistakes, sometimes that must be done, even if doing so requires straying from otherwise ideal behavior. In these situations, the ends justify the means.
Part III: The “actions” perspective
But there are also situations in which they do not. Just as there were unacceptable extremes of the “actions” framework, there are unacceptable extremes of the “results” framework too. For instance, take my earlier example with the gunman. What if you had to brutally behead 100 children with a kitchen knife in order to save 101? Would you yourself physically commit such a vicious genocide to save just one life on sum? How could you live with yourself for brutally murdering so many innocent, scared, crying, pleading human children? How can anyone with a conscience actively, willingly participate in something so horrible? To many people, there is something about certain actions that, regardless of context, is just unilaterally unacceptable. There is a certain line between action and inaction, and in a choice between the lesser of two evils, people are understandably reluctant to cross that line. If you do nothing, then at least you can be certain that whatever happens will not be your doing. If the maniac decides to kill those children, it will be horrible, but it will be entirely his fault. You’ve washed your hands of it. If we live our lives by a strict moral code and only do things which we know for certain to be moral and right and virtuous and just, then the onus of responsibility for the world’s ills is stripped from us. That’s a comforting thought. Terrible things may happen due to those who do not abide by our code, but at least we ourselves will be free of blame for the outcome. And if everybody followed the rules and shunned immoral actions, the outcomes would eventually take care of themselves.
In philosophy, this school of thought is called “deontology", "moral absolutism", sometimes "idealism." This is the “actions” side of our debate. Idealists clash with realists because they contend that perception defines belief, and that without something being perceived it does not exist. For example, if a tree falls in the forest without ever being heard or seen, they’d argue that it didn’t really fall at all, since no living being perceived it’s falling. Since morality and principle are human constructs, to an idealist they exist just as vividly as do the tangible world our minds perceive, and thus cannot be dismissed for practical convenience. As it relates to our debate, deontologists would not support the rejection of ideals in order to attain a desired end state. Certain components of deontological ethics have been advanced by Immanuel Kant, George Berkeley, Johan Gottlieb Fichte, George Wilhelm Friedrich Hegel, Francis Herbert Bradley, and Josiah Royce.
Religion is sympathetic to this viewpoint. For example, the Bible does not say “Thou shalt not kill unless thou think the killing is justified.” It says “Thou shalt not kill”, period. That is for good reason, because in real life, the decisions we’re faced with are almost never as simple as the situation provided above. In that example, I stipulated that you are 100% certain of what will happen if you don’t kill the child. In real life, very rarely is the result that clear. When the foreseeable consequences of each option become murky, consequentialist judgments become confusing. And the more the outcomes of each option become obfuscated, the more the intuition and impartiality of the decision maker become suspect. It’s easy for brilliant philosophers to sit back in a comfy chair and muse on hypotheticals, but it’s much more difficult for everyday people to justly make high-stakes decisions in the heat of the moment. For this reason, universal adherence to a set of simple, succinct, memorable, guidelines can often be much more effective at producing just outcomes than leaving decisions to the open interpretations of billions of individuals.
There is a certain humility in raising eternal ideals above flimsy human judgment and selfish human motives. Important decisions often must be made by people who have a huge personal stake in the outcome, and the outcome that favors them may be different from the outcome which is morally superior. When asked to make complex case by case evaluations under tremendous pressure, stress, and temptation, are human beings really noble enough to prioritize the greatest good for the greatest number over their personal desires? And if they are, are they really knowledgeable, clearheaded, insightful and wise enough to accurately foretell the long term consequences of each option, and then justly evaluate the morality of each result? History is rife with examples to the contrary. To assert that we can weigh all these factors objectively and fairly, that we possess the mental capacity and selflessness to accurately diagnose when it is okay to do terrible things (like murder), is not only dangerous and unrealistic, but also a bit arrogant.
For this reason, if one is to do something with the justification that the ends justify the means, he has to be damn sure he’s right. The bar has to be set pretty high on his degree of certainty. To take an immoral action on anything less than concrete assurance that it’s justified is to take a mighty big risk with other people’s rights. As a people, we should receive these claims with suspicion and a critical eye. We must make sure they are the exception rather than the rule.
So…what qualifies as an exception?
In what circumstances is it okay to take an immoral action in pursuit of a morally superior outcome?
When do the ends justify the means?
Part IV: When do the ends justify the means?
We need a mechanism to draw the line separating the times they do from the times they don’t. Naturally different people will draw that line in different places. But in most circumstances, I believe we can use objective, logical observations to help us draw it.
After much thought, I have identified four conditions which must be met if the ends are to justify the means. Each condition deals with a variable that affects the morality of each respective option. In order for each condition to be met, the variable it deals with must be at a certain level. To establish a basic, memorable framework to help the decision maker in the heat of the moment, I have associated each variable with a question. By answering these four questions, the decision maker should be able to choose more wisely.
The first question is, “How moral are the ends desired compared to the morality of the means employed?” Clearly, this can be subjective. The underlying variable here is the relative goodness/badness that each option stands to bring about. If the ends are to justify the means, they must be of such dire importance that not achieving them would be an even greater moral offence than the means would be. For example, while the ends may justify the means of you’re trying to save ten lives by killing one person, the ends do not justify the means if you’re trying to win $10 by killing one person. The immorality of the means is the same in both instances, but the ends only justify them in the first case. This is the most obvious and frequently cited justification for employing the maxim.
The second question is, “How direct is the link between the means and the ends?” The underlying variable here is the agent’s certainty that taking said action will bring about the desired end state. For the ends to justify the means, that certainty must be very high; there must be little doubt that the means will directly and infallibly lead to the ends. Otherwise, the actor risks merely adding additional immorality, without ameliorating the immorality he seeks to avoid. For example, if one is faced with a button that will kill one person, and in exchange ten other lives will definitely be saved, it makes sense to push the button. But if one is faced with a button that will kill one person, and in exchange for pushing it those ten other people’s chance of surviving goes from 0% to 5%, pushing it may not make sense, because there’s a 95% chance that doing so will result in eleven deaths instead of ten. It is not certain, in fact not even likely, that taking the immoral means of killing one will actually bring about the desired ends of saving the ten.
The third question is similar to the second, but not identical: “How direct is the link between not using said means and not attaining the ends?” The underling variable here is the agent’s certainty that the even-more-immoral result will definitely come about if he takes no action at all. Returning to our example above, this is the opposite percentage: if you do not push the button, what is the chance that those ten people will die, irrespective of how that chance will be affected by pushing the button? In order for the ends to justify the means, we must be able to affirm with a high degree of certainty that if this immoral action is not taken, an even more immoral event will take place.
The last question is “Are there any alternate means of lesser immorality which might obtain the same ends?” Up until this point, we have assumed that there are only two options: doing the immoral means, or not doing them. Real life is oftentimes more complicated, and multiple courses of action must be considered. For example, killing one life to save ten may be justified under the circumstances above, but not if the ten will also be saved by pressing a second button that doesn’t kill anyone. The distinction between this question and question three is that question three deals with inaction, while question four deals with a previously unconsidered action, a sort of “Plan C”. If there exist any alternate means of lesser immorality which will attain the same ends, the morally inferior means are not justified.
That’s a mouthful, but it gets even more complicated because these conditions and variables are not wholly independent of one another. Oftentimes, answering one question requires knowing the answer to another. For example, condition three is sometimes dependent on question one. If the potential moral consequences of inaction are severe enough, the standard of certainty in question three lowers, because we simply cannot risk those consequences regardless of how small the percent chance is that they occur. For instance, if the chance that 10 people will die is 20%, it may not make sense to kill 5 innocent people in order to send that chance back to 0%. But if the chance of a nuclear holocaust is 20%, killing five people in exchange for returning that chance to 0% is probably a safe way to hedge ones bets. Similarly, condition two and condition four are related. There are oftentimes many different means which might bring about the desired effect, but the relative probability of success is different for some than for others. So in our example on condition four, if button one kills one person and has a 100% chance of saving the other 10, and button two kills nobody but has just a 75% chance of saving the ten, those odds need to be weighed.
The last three conditions are mostly objective; in theory, we could assign a % to the possibility of various different outcomes. On occasion, the first condition is also objective; if all the outcomes deal with one and only one fixed numerical unit (money lost, lives saved, people displaced, etc.), and the agent doing all the action is the same, we can make objective determinations of which outcome is preferable using the percentages from the final three conditions (in philosophy, this is called “decision theory”.)
But oftentimes, the first condition is subjective. Much like the opening scenario with the child and the gunman, deciding which actions are acceptable in pursuit of outcomes can be a deeply personal decision. Analyzing the likelihood of various outcomes resulting from various actions can help, but ultimately the decision may need to be predicated on moral beliefs that differ from person to person. When agent making the decision is an individual, that person has to do some soul searching and act accordingly.
But what if the agent itself is not an individual person?
What if the agent has no moral preferences?
What if the agent is an organization that claims to represent millions of people all at once, and claims to act on their behalf, but each of those people have their own unique moral beliefs?
In other words, what if the agent is the government?