eli5 What is Kant’s categorical imparative?


I have to understand it for a debate in class about the death penalty, but I can’t wrap my head around it just right

In: 6

For Kant, the perpetrator suffers roughly the same amount as their victim(s). So if their crime instigated death, then death as a punishment would be morally correct (through this school of thought). Eye for an eye, basically. Murderers should be subject to the death penalty per his line of morals.

Deals with the category of the action rather than the outcome.

The majority of us can agree that lying and stealing are bad, on their own.

Some people will say that lying and stealing become necessary when one is left with no other option (e.g. to feed their family when they have no income).

Some people will say that it doesn’t matter what conditions resulted in the need to lie and steal, the fact that a person lied and stole on its own is morally wrong. Therefore, the whole category, i.e. lying and/or stealing is morally wrong no matter what reasons are used to justify it.

Under Kant’s categorical imperative, the individual conducts themselves by a set of rules that they wish others to hold as evidently true.

For example: I won’t lie and steal, and if I don’t lie and steal, I expect others will do the same.


They believe having consistency and predictability of justice will allow individuals to make better choices, knowing full well the outcome of actions x+y=z every time.

It’s an almost impossible standard to hold, since humans are very hypocritical and tend to dismiss perspectives other than their own or ones they can relate to, but that doesn’t mean we should throw it out altogether.


Ofc, this is as basic as I can manage to explain it. I am very sure I had to omit or sacrifice a lot of nuance and detail, but for what I expect out of an ELI5, this should suffice.

The easiest way to understand Kant’s categorical imperative is that any rule one person follows should be able to be applied to everyone else under any circumstance.

If a rule can’t be applied to every person under every circumstance, then it’s not a good rule.

So, for instance, Kant argues that a rule which fits these requirements is “you should not lie”. Because it applies to any person and any situation (according to Kant at least), it’s valid as a moral law. If one could imagine a situation where lying were good (Kant can’t), then it wouldn’t be valid.

This is an oversimplification of course, but that’s the jist of it.

(This is as far as I remember and understand it. I could very well misremember or have misunderstood some things.)

There are two different possible kinds of moral imperative:
1. Hypothetical imperative. Imperatives of this sort will depend upon the outcome that is being sought. So for example, if the goal is to help your friend in his marriage, there would be certain actions that would lead to that goal. And so those actions would be imperative to get to that goal. The actions you should take are dependent upon the goal that you want to achieve. Now at this point, the question naturally is “what is the goal that we want to achieve?” The most common answer (but not the only one) that you’ll get is the hedonistic one, that we should maximise pleasure and minimise pain.

2. Categorical imperative. Imperatives of this sort depend purely on the action itself. So for example, the action of going through a red light. That action itself is imperative for drivers to not take. It doesn’t really matter what the goal is. What matters is the category that the action itself falls into. The question then is, how do we decide what actions are good (and thus should be done) and which actions are bad (and thus should be avoided). And for this, Kant turns to his universalising principle, that we judge actions based on imagining what would happen if everyone acted in that way. If everyone was to go through red lights, that would obviously be problematic. And so, everyone should stop at them. Even if your individual circumstances might make it seem like you should break the rule, what is morally correct is to always follow the rule and act consistently.

Please do correct or clarify anything I am mistaken on or poorly worded!

Basically the golden rule, a fancy wording of “do unto others as you would have them do unto you.”

I think it goes like “act only on that maxim which you could make a universal law.”

In other words, say you cheat on a test. By doing that, you are saying by your actions that it is permissible for everyone to cheat on a test. But what would happen if that was a “universal law,” like everyone should cheat on tests if they feel like it? That would be bad. Do you want to see a doctor who cheated on medical school tests? Do you want a teacher who doesn’t know what they’re talking about because they cheated on college tests? Do you want an accountant who cheated on their math test? No, so you couldn’t make “cheating is okay” a universal law. Therefore you shouldn’t cheat.

But let’s look a good rule. Suppose you refuse to rape people. What would happen if everybody followed your example by not raping? There would be a lot less trauma, and probably nobody would rape you. So that would be a good universal law, so it is morally acceptable to refuse to rape people.

Suppose you were to act on the maxim that “you should kill people who have killed people.” What would the consequences of that be? Could that be a universal law? By one argument, no. If you kill a murderer, then somebody has to kill you. Then somebody has to kill them, and so on and so forth until everybody is dead.

The counter argument would be to add more specific conditions: “you should kill people who have killed people who haven’t killed people.” But then, of course, revenge murderers would be free from the death penalty. So you could be like “you should kill people who have killed people but only if you have permission from a jury.” But if those jurors grant permission, aren’t they also acting on the maxim of “kill people who have killed people, as long as there are 13 between jurors and executioners?” And doesn’t that kind of condone gang violence as long as there are 2 gangs have at least 13 members each?

So where you stand on how the categorical imperative impacts the morality of the death penalty depends on how many “except for” clauses you’re willing to add.