Obviously the ends justifies the means

last edited 2024-11-05

There are two versions of the idea that "the ends can't justify the means" that I want to refute. One is basically an extreme version of political libertarianism, and the other is popular with conservatives and especially Catholics, as it's based on part of Thomas Aquinas's philosophy.

The libertarian version

So, libertarianism is based on freedom of association. There's a fundamental right to be left alone, and all other "rights" are only valid insofar as they're corollaries of this. But really extreme libertarians say that nothing can ever justify violating it. For example, you couldn't steal $5 from a millionaire to save a life, because that would violate the millionaire's right to not associate with you.

Besides being absurd, this belief has a coherency problem, which is that it doesn't account for the concept of probability. What about an act with some probability of resulting in a rights violation? For example, taking $5 that might or might not belong to someone else to save a life.

There are three possible answers such an extreme libertarian could give to this, all of which are bad:

An act is absolutely immoral if it has *any* probability of being a rights violation. This is untenable because it would prohibit basically everything. For example, if the millionaire gives you permission to take his money to save a starving person, you still can't do it, because there's *some* possibility that you misheard or misunderstood the millionaire and they didn't actually give you consent.

An act is only absolutely immoral if it has *100%* probability of being a rights violation. This would defeat the point of believing this ethical theory in the first place, because real world probabilities are *never* 100%. For example, if the millionaire says they don't consent to you using their money to save a starving person, taking their money anyway doesn't have a 100% chance of being theft, because there's *some* possibility that you misheard or misunderstood, or that the money actually belongs to the starving person and the millionaire stole it.

There's some threshold of probability between 0 and 100% that makes an act absolutely immoral. This one is subtly incoherent because whether an act crosses that threshold depends on *how you count actions*. For example, let's say the threshold is 50%, and you're a judge and there's a 10% chance that the defendant is innocent. Then, you can sentence them. But! If you sentence 10 people that each have a 10% chance of being innocent, the chance of sentencing at least one innocent is 65%. Which according to the rule is absolutely immoral, regardless of the consequences. So this view would mean that in each of the 10 cases it's okay to sentence the person, but when looking at the 10 cases as a whole, it's not okay to sentence them... this is incoherent.

The Catholic version

This version's based on Thomas Aquinas's principle of double effect, which says that good outcomes *can* outweigh evil outcomes (even rights violations), justifying the act, *if* the good outcomes aren't *because of* the evil ones. But besides being counter-intuitive (why should it matter which outcome causes the other?), this position suffers from its own logical problem: you can't always objectively define whether an act counts as having an evil *means* or an evil side effect.

Let's look at two scenarios often used as examples, the organ transplant and trolley problem. The organ transplant is:

You are a doctor with 5 patients in need of various organ transplants. Without the organs, all 5 patients will die today. Time is almost up when an innocent bystander arrives at the hospital who has the required organs to save all 5 patients. You can either let this bystander walk away while your 5 patients die, or murder the bystander and use their organs to save all 5 patients. Assume the bystander doesn't consent to being sacrificed.

And the trolley problem:

A train is about to hit 5 people are tied to the track. You can either do nothing and let 5 people die, or you can switch the track, sending the train on a different path where only 1 person is in the way. Again, assume this person doesn't consent to being sacrificed.

Proponents of this ethical theory would say you can't murder the bystander to do the organ transplant because that would be an evil means, which can't be justified by the good outcome of saving 5 people, even though 5 is more than 1. But many of them would say you *can* switch the track in the trolley problem because that kills 1 person *as a side effect*, not as a *means* to saving the 5 people.

Now, let me offer an alternate analysis with the same ethical theory: killing the hospital bystander is okay because you're not actually killing them *as a means* to getting the organs. The bystander dying is just a side effect of cutting them open, which allows you to get the organs to save the 5 patients.

(Incase you're thinking that cutting the bystander open is itself an evil means because it causes them pain: no, the pain you inflict on the bystander is a *side effect* of cutting them open, not a means to cutting them open.)

Why is this analysis wrong? How can you determine that killing the bystander counts as an indisivible action which is the means, rather than cutting them open being the means and everything else, including the death and pain you cause, being side effects and thus eligible to be outweighed by saving 5 people?

I call this the "arbitrary line" argument, because it shows that there's no objective way to draw the line between means and ends, so we can't use that as part of an ethical framework.

By the way, a consequentialist doesn't have to think that murdering the bystander is acceptable. It's possible that the ratio needed to justify a consent violation is greater than 5:1 (maybe it would be acceptable if it saved *100* patients), or it's possible that the act is wrong because it doesn't actually have good consequences in the long term (once this incident becomes public knowledge, people might lose trust in hospitals and refuse to go to them for fear of being sacrificed, and maybe this would lead to many more than 5 deaths).

Justifying versus purifying

Finally there's a misunderstanding about consequentialism that's important to dispel. Good ends can *justify* evil means, but can't *purify* them. For example in the organ transplant scenario, even if it's justified to murder the 1 person, it would *still* be murder, and if you were to meet your victim in another world, you'd owe them your life as restitution. Even though it was justified for you to murder them. If this seems bizarre, consider that if it had been possible to save the 5 people by sacrificing your own life instead of someone else's, you wouldn't have been justified in sacrificing someone else. The ends can *justify* the murder, but they can't make it not murder.

With that distinction illuminated, there's a fascinating observation I want to share about the effects of different ethical frameworks on the mind: *People who claim to think consequences can never justify evil means are more likely to make appeals to consequences than consequentialists.*

Examples: every conservative argument against anarchism, many conservative or libertarian arguments against communism (anything about incentives or historical results), and many right-wing "libertarian" arguments for national borders. I can't think of a single time I've heard an avowed consequentialist commit this fallacy.

Why you should be an Anarchist

Communism works

Borders are not defensible

There's a profound but straightforward psychological explanation for this. Their position doesn't work; its implications are too far from aligned with their intuition. They intuitively know that in some situations consequences can outweigh justice, but can't admit it, so the only path open to them is to say that injustice was actually justice because of the consequences. They're unwilling to separate justice from consequences, so they pollute them with each other.