Who is committing a fallacy in this discussion, and what kind of fallacy is it?

A: There's nothing wrong with killing people at random

B: If everybody did it, you would too most likely be a random victim.

A: Well, if everybody was killing random people, I would surely be an idiot not to do the same (in order to defend myself).

I feel like they are both wrong for the same reason: the world in which "everybody does it" is not a realistic world, it's imaginary.

And both discussants use this imaginary world to defend their stance: B uses the imaginary world to point out the potential self-harm in A's stance, while A uses the imaginary world to potray his stance as necessary.

  • 5
    Doesn't A's second response completely misunderstand the notion of 'random'? Killing people in self-defense is not random. So he is at that point in a different argument altogether. – jobermark Aug 24 '17 at 15:05
  • 8
    But no, 'the world in which "everybody else does it"' is Kant's best moral world, and it is a quite reasonable state to consider, or Kant himself would not be nearly so famous... You have to consider it as an ideal, not an actual state, of course, but Kant's reason to proscribe killing is basically B's argument. If there are rules, and you need to be special, you are doing something wrong. – jobermark Aug 24 '17 at 15:05
  • 6
    There's nothing wrong with attempting to buy the third can of Coke from the left on the second shelf in the convenience store at Punkydoodle Corners at 12:14 PM on September the 3rd, 2018, right? But what if everyone did that? There would be riots and deaths as 7 billion people all descend upon the store at once. Therefore plainly it is immoral to make the attempt. Is that argument fallacious for the same reason your given argument is fallacious, or is the fallacy different in mine? – Eric Lippert Aug 25 '17 at 6:15
  • What about "A: There's nothing wrong with killing people at random as long as not everybody is doing it."? – Trilarion Aug 25 '17 at 7:21
  • @EricLippert That is indeed exactly the problem, your argument only seems trite because, of course, not everyone is going to want to buy that can. But then the issue becomes one of speculation on the consequences. If I do X is it reasonably likely that everyone else will want to do so too, and would the consequences of that be desirable? This is one of the difficulties with deontological ethics, almost any maxim can be made universalisable by adding detail (or conversely prohibited by removing detail). – Isaacson Aug 25 '17 at 8:44

Logicians distinguish two kinds of bad arguments, unsound and invalid, see IEP's Validity and Soundness. A fallacy is a flaw in logic that makes an argument invalid. Unsound arguments, on the other hand, may well be logically valid, but argue from false premises.

The problem pointed out in the OP, that the world referred to by the arguers is unrealistic, means that the premise is false in our world. Given that, however, B's point is well-taken, i.e. plausibly valid, it is similar to the logic of "in the land of the blind the one-eyed man is king". A, on the other hand, has a problem with his logic on top of the unsoundness. Defending yourself does not amount to killing people at random even if others are killing people at random. It would at most amount to killing people who are a threat, and even that is not given, since self-defense does not necessarily require the killing.

There is also another problem with the A's reasoning. "There is nothing wrong with killing people at random" is a value judgement, not a description of fact. Even in the world where "everybody" is killing people at random that alone does not make it right, does not mean that they ought to be doing it. After Hume many philosophers consider inferences from is to ought statements to be generally invalid, it is known as the is-ought guillotine or the naturalistic fallacy.

This last point touches on a connection between logic and ethics. "If everybody did it" hypotheticals are important in what is called meta-ethics, the study of ethical systems. They express what is called universalizability of ethical norms, and many ethicists consider it an important requirement that a reasonable system of norms must meet. Perhaps, its most famous expression is Kant's categorical imperative:

"Act only according to that maxim whereby you can at the same time will that it should become a universal law."

In other words, any norm of ethical behavior must be such that it would be rationally acceptable if everybody acted according to it. A's "norm" of killing people at random apparently fails to meet this requirement.

  • So, according to Kant, is killing at random immoral due to his Categorical Imperative or due to his "never use rational beings as means to an end, only as ends themselves" philosophies? I have never quite figured out which one it should fall under. – Daishozen Aug 25 '17 at 19:43
  • @Daishozen the formula of humanity (never treat people as just means) is intended to be equivalent to the universal law formation. They're both the categorical imperative, just looked at in different ways. – Canyon Aug 31 '17 at 19:51
  • 1
    @Daishozen Kant's position is indeed that the categorical imperative and the formula of humanity are equivalent, but his argument to this effect (assuming free will as the source of rational action) is not very convincing. On modern view the two are independent. One can imagine a society of individuals pursuing material self-interest (survival, wealth, etc.) and using others as means to their ends, while expecting to be treated in kind. Such conduct is universalizable and may even support some forms of cooperation. It will not support killing at random since that threatens one's own survival. – Conifold Sep 7 '17 at 3:31

I find that your question is very ambiguous, as well as the arguments.
There is nothing wrong with "if everybody did it" argument. What makes it wrong, if anything, is what it refers to. If instead of "killing people at random," we use "making random acts of kindness," then:
A would say, "There's nothing wrong with making random acts of kindness."
B would say "if everybody did it, most likely I would be a recipient of kindness."
There is nothing "wrong" and there are no fallacies.


Because of their variety of structure and application, fallacies are challenging to classify so as to satisfy all practitioners.

-- Wikipedia - List of fallacies

Deontological ethics concerns making a morality judgement of an action based on rules. The word "Wrong" in premise A signals to me a matter of ethics. Per Wikipedia:

Ethics or moral philosophy is a branch of philosophy that involves systematizing, defending, and recommending concepts of right and wrong conduct.

I would encourage you to apply a Practical Syllogism and see where that takes you.

A practical syllogism is an instance of practical reasoning which takes the form of a syllogism, where the conclusion of the syllogism is an action.


First of all, you have to specify the situation to the question to this answer.

Secondly, there is no possible way in the universe that one individual can answer for everyone. Therefore, the question alone is moot.

In order to calculate an accurate response, one would have to ask a specificifaclly articulated situation to each individual in the universe. An individual's reply would largely depend on his/her experience. And his/her response would then be further predicated upon the experience of those whom were present in their formative years. So, there can never be a true collective answer to any situation posed by this question because every individual's experience with any given situation is always predicated upon another's experience to a given situation and that original individual's exposure to another's response.

Therefore, although many might agree on a general answer,the collective response would still be a fallacy because it is not collectively definitive.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.