Descartes famously argued that we can technically be wrong about anything. We can even doubt that the world isn't actually real in some sense and is just a simulation. But why does doubt imply the possibility that you could be wrong? Isn't doubt just a psychological state?

What exactly makes it irrational to think you can't be wrong about a particular thing? For example, is it irrational to be certain of the fact that there is no God? Or to be certain that the sun isn't going to turn into a unicorn tomorrow? Is it irrational to not just assign something to be of a non zero, but a literally zero probability?

In Bayesianism, this is often defended against, by the notion that if you assign a zero probability to something, it means that no amount of evidence can change your belief. But shouldn't a non zero probability, even if infinitesimal, be justified as well? If I am assigning a non zero probability to something like the sun turning into a unicorn tomorrow, aren't I basically implying that there is a non zero chance that could happen? But what evidence do I have that it could? If I have none, how could I assign a non zero probability to a statement if I don't have any indication to suggest that it is possible in the first place?

Suppose I assert certainty in X. One can argue that my certainty is not rational simply because I can't prove that for sure since I may be wrong. But the very possibility of me being wrong is an assumption. What if I'm simply not wrong? Until and unless I've been shown that I'm wrong, there is no way to actually falsify that certainty. But if there is no way to falsify it, how can it be considered to be irrational?

Note that my question isn't about whether we should think we can't be wrong about everything, but whether it is rational to think that we can't be wrong with respect to certain things. Historical examples of thinking we can't be wrong with respect to X and having it turn out that we were wrong about it does not imply that we can be wrong about Y.

  • A few of your questions seem to share a quality of being answerable by something like, "By examining evidence". "Can it be rational to think you can't be wrong about something?". Sure. If you share an opinion that has been demonstrably proven - via examination of evidence - to be correct. Mar 11, 2023 at 10:02
  • 1
    It depends on how you answer the question "Is anything ever certain?". If your answer is "yes", you will no doubt also believe that there is sometimes enough evidence about some things to prove them certainly true. When you have that much evidence for something, it would be rational to believe that you cannot be wrong about it. But, I would say, unwise. To do that is a bit like marking your own homework. A wiser course would be to check your belief with other people. If you get confirmation from people you trust, you can be confident that you are rational to believe it.
    – Ludwig V
    Mar 11, 2023 at 10:15
  • how do you think out of rational, are you able to think irrational? what this mean rational? logic based or what? Mar 11, 2023 at 11:24
  • 1
    Sometimes it is argued that there are incorrigible beliefs, ones that our minds simply cannot change, however hard we try to will them to. If we have an incorrigible belief about the nature of beliefs, this can be taken for evidence about that nature, but beyond such a domain, it is easy to see that an incorrigible belief can be detached from the facts. Still, this doesn't mean we're at fault/in the intellectual wrong, for holding the belief. Maybe we could externally change it by inflicting brain damage on ourselves, but that wouldn't seem justified, depending on the belief being changed... Mar 11, 2023 at 19:08
  • 1
    @thinkingman You don't need to be so hard on yourself and force yourself to choose between those 2 options only. You can also believe whatever you want, rational or not, and be happy with that.
    – Frank
    Mar 12, 2023 at 0:13

2 Answers 2


If you don't have an external validation and think you cannot be wrong on some of your claims just because you know, or are convinced that you cannot be wrong, or don't see a reason to question that you could be wrong, then this is the door open to making any claim and not convincing anybody. You just have an inner conviction that others may not share and you have no reason to expect others to go along with your claim. Someone may have the opposite inner conviction to yours with equal legitimacy. And you run the risk of being very wrong. What if you have a strong conviction that you can fly because you are on some drug and put that to the test?

  • "Until you spread your wings, you'll have no idea how far you can walk."
    – Scott Rowe
    Mar 12, 2023 at 3:56

Descartes was perfectly sure for the cogito (cogito, sum) and consequently for any other impression as clear and distinct as the cogito (Meditations on First Philosophy).

If an impression as clear and distinct as the cogito is false, then cogito can be false as well and this cannot be. So impressions of the same quality have to be true as well and one can be sure of that.

In these statements of Descartes, there is nothing fundamentally irrational.

PS: It is, in general, impossible to avoid assigning zero probability, at least in some cases. For example, always avoiding assigning zero probability to events, one is, in fact, assigning zero probability to the event "assign zero probability to other events".

You must log in to answer this question.