I read a lot about people saying that it is impossible to be certain about anything with 100% certainty, but that means that this rule in and of itself must be true 100%.

I mean the rule that "we cannot be certain about anything" must itself be certain, which breaks the rule. So in order for this rule to be applied it needs to be broken at least once which is for itself.

This circular reasoning makes it hard for me to grasp.

Also some people divided the things that we can know for certain and the things we cannot (e.g functional certainty: I am certain I am writing this, not sure how, but certain it's happening), but even that has been doubted and counter arguments where proposed to refute it. So how exactly do we interpret this? Does it mean that this rule simply isn't all around true and that we can know somethings with absolute certainty?


100% doesn't mean a numerical representation that is mathematically practical, it just means "absolute"

  • 1
    Statements like "we can not be 100% certain" have no numerical "value" of certainty attached to them, nor does it make sense, there is no meaningful sample space. The "100%" in the sentence does not mean anything more than a superlative for practical confidence. Since the "rule" is not meant in an abstract generalistic sense it can not be applied to itself, and even if it is, it means nothing more than "I am pretty sure". Sometimes it is even jokingly stated as:"There is only one thing that is 100% certain, that nothing else is 100% certain".
    – Conifold
    Mar 7, 2019 at 23:47
  • Humility, and pragmatc skepticism are cornerstones of real learning. If you believe anything with abdolute certainty then how can you change your mind? And if you can't change your mind, how can you learn? It's just easier to keep an open mind. I'm 'pretty sure' that the sun will rise tomorrow though.
    – Richard
    Mar 7, 2019 at 23:52
  • There are knowns and there are unknowns, but then there are known unknowns and unknown unknowns. When you are certain of something, are you automatically certain you are certain of it? If I am uncertain that I am certain, is that certainty? If not, I can be certain of something, but uncertain that I am certain. If you are at least uncertain about the certainty of everything you know, then you have not created a contradiction by being certain that you are not truly certain of anything.
    – user9166
    Mar 9, 2019 at 1:30
  • @Conifold you can replace that 100% with absolute, I am not trying to quantify certainty, I am trying to assert the statement of "we can never be certain about anything"
    – engma
    Mar 9, 2019 at 3:12
  • @Richard Being humble and pragmatic is good in the scientific research realm, where we look at the "how" or the "actuality" or "essence" of the object under test, but I am talking about being certain that this thing "exists", not discussing it's "form" but the fact that it is "truthful" or "real" like me and you, not discussing our "form" of existence, just that it is, leaving an open mind for the actuality of it
    – engma
    Mar 9, 2019 at 3:14

7 Answers 7



'Nothing is certain' is a proposition. If it is true, then no proposition is immune from error. But that's problematic since 'Nothing is certain' is a proposition which refutingly applies to itself.

No paradox

Our being certain of anything is a propositional attitude, not itself a proposition. It is the attitude towards a proposition of believing it to be immune from error. In the present case it is the propositional attitude of believing that no proposition is immune from error. I don't think that refers back to the attitude itself, since (as said) an attitude is not a proposition. Believing can't be immune from or vulnerable to error since it is a merely a psychological phenomenon which obtains or doesn't. Of course, what we believe can be immune from or vulnerable to error. The former in principle, the latter all too common.


The point isn't (to respond to another answer) that we can't be 100% certain - absolutely certain, though the '100%' logically adds nothing to 'certain' any more than my 'absolutely' does. Rather, ask yourself what sense does it make to say that we are 99.37% certain or 55% certain? Certainty is a condition we're either in or out of - we can't be in it to a numerically determinate degree (<100%).

  • So from that we deduce that certainty is subjective and not objective?
    – engma
    Mar 10, 2019 at 21:43
  • I am also not sure I understand what you mean by "propositional attitude", just in terms of usability? if it's an "attitude" then it has limited usability and will be mostly in terms of being an incentive to continue researching in science, but for example not very useful in terms of criminal indictment, otherwise every criminal would go free, right?
    – engma
    Mar 10, 2019 at 22:44
  • Geoffrey, if I may ask: why does anyone insist that the proposition is a paradox rather than a hasty expression?
    – user38026
    Sep 9, 2019 at 12:46

It is useful to distinguish truth from certainty: specifically, a statement being true and a person being certain about the statement. For instance, it might be true that there is an odd number of fish in the ocean right now but perhaps I cannot be certain of this. So, some things may be true even though we cannot be certain that they are true.

The same goes for the 'rule' you mention. It might be true even though we cannot be certain of it. The rule also applies to itself: it says that we cannot be certain about it. That is perfectly consistent, as, again, it is possible for a statement to be true and for us to not be certain of it.

What may seem nonetheless problematic about this rule is that it might be interpreted to say that we shouldn't believe it. But that depends on what exactly it means by 'certainty'. If it's certainty as in a subjective probability of 1, it might be true but uninteresting: we'll do fine with being almost-certain. And we could be almost-certain about the rule itself, as it doesn't preclude that. But if by 'certainty' it simply means high probability (not just 1), in the sense that we cannot be sure to any degree of anything, then it does undermine itself: if it's true we shouldn't believe it.


We most often use the term truth for indicating facts and often these facts seem to exist for a long period. But these also disappear after a certain period. We always experience something immutable even though there happens growth and development to our body. The only thing that can be certain is the true nature of oneself and it is the Ultimate Truth. One becomes completely satisfied with that only thing. Other things cannot satisfy him always. And the satisfaction is only because it is immutable and he is 100% certain about it. Actually I should use more than 100%. Since there is no usage like 1000%, 10000% (with certainty), I don't wish to use it.

The term 'Ananda' certainly conveys the certainty of something. https://en.wikipedia.org/wiki/%C4%80nanda_(Hindu_philosophy)

So I don't agree with this statement--"we cannot be certain about anything". IMHO, this is a wrong notion.

Also some people divided the things that we can know and the things we cannot, but even that have been doubted, so how exactly do we interpret this?

Read the description of the verse 7.7 of the Bhagavad Gita: https://www.bhagavad-gita.us/bhagavad-gita-7-7/

If this is true, our individual perception need not necessarily be 100% certain. There is a great possibility for misapprehension. http://www.mahavidya.ca/2015/06/25/maya-the-concept-of-illusion/

So your doubt is reasonable, but except in the case of one thing and it is the thing that I have mentioned already. All other things come under the second category.

Does it mean that this rule simply isn't all around true and that we can know somethings with absolute certainty?

If you agree with the idea mentioned in the first para, the answer to this question must be "Yes". But unfortunately those who realized this truth are rare.

  • +1 For a non-Western perspective.
    – J D
    Mar 15, 2023 at 16:53

It's absurd to say we cannot be 100% certain.

In fact we are 100% certain, existentially certain, of everything.

A simple example: You see someone across the street. Is that Joe? You're not sure.

So are you uncertain? Yep. 100%. Are you certain you're uncertain? Yep. 100%.

Conventionally you're not certain and there's no way to prove anything is 100% certain because how would that be determined? How can something be 'more than certain' to measure something is certain?

Existentially you're always 100% certain (that you are uncertain).

"It might rain tomorrow."

Are you sure? Absolutely.

Well that's 100% as "might" accounts for all possibilities.

  • 1
    +1 For a certain response about certain responses. Very meta.
    – J D
    Mar 15, 2023 at 16:54

I mean the rule that "we cannot be certain about anything" must itself be certain, which breaks the rule. So in order for this rule to be applied it needs to be broken at least once which is for itself.

Certainty is immeasurable. A person is certain about something which he considers indisputable, true. Certainty is not a characteristic of a thing, but rather a mindset of someone who perceives that thing.

I wouldn't consider that "truism" a rule, nor a maxim, nor an axiom. Where is its proof? It is a meaningless good-sounding phrase.

  • Certainty is measurable. Chance of rain is 30%. As for certainty as a mindset, please consider 'existential' instead of mindset - a specific vantage. Please see my answer for examples. Mar 10, 2019 at 21:53
  • @Randy Zeitman, you equate chance with certainty. People can have different certainties about the "30%" chance of rain.
    – user96931
    Mar 14, 2019 at 16:21

Randy Zeitman in his answer says:

It's completely absurd to say we cannot be 100% certain.

He and others here have danced around a fact that needs to be made explicit. Certainty is a psychological state, and truth is generally assumed to be a logical state. Thus, we can be certain even when we are wrong, and uncertain even when we are right in matters of truth.

A simple example of that is a crackpot who pushes pseudoscience seemingly without regard to moderation or reason. Some have proposed tounge-in-cheek that there are essential characteristics or properties of such arguments and people that can be measured. Another example of misplaced certainty are the historical examples of entire groups of well-educated people who turn out to have been wrong in retrospect. Miasma theory was prominent when Ignaz Semmelweis built up a scientific argument that hand-washing before delivering babies saved lives. The reaction against his views were not only certain, they were outright hostile, and later in his life, his colleagues collaborated with his wife to have him committed to an asylum.

Now, as to the question of how we can be relatively certain we cannot be certain, this is a consensus position of epistemologists who almost overwhelmingly adopt fallibilism (IEP) as a position in regards to the certainty of epistemological methods. The short version is that when radical skepticism reared its head thousands of years ago, philosophers have been chipping away at the problem of eking out certainty, mainly by advocating deductive methods and showing the pragmatic nature of induction and abduction, to make claims that some certainty is possible while accepting that doubt is often justified. Today, such a position might be called contemporary skepticism on account that a lot has changed and been learned since the Pyrrhonism. Naturalistic epistemologies of scientific practice also align closely with this view that anything, with some prima facie justification, can and should be doubted. Most scientists agree that belief should be open to justification, which is a far cry from the certainty of a priori reasoning and introspection advocated by Rene Descartes.

If you want an even simpler response, most people are certain that there are reasons to be uncertain. Why? There is a set of very persuasive logical and psychological experiences that undermine certainty: illusion, delusion, confabulation, deception, self-deception, paradox, logical incoherence, lies, paltering, mistakes, errors, fallacy, psychological defenses, rationalization, cognitive distortion, and cognitive bias for starters. Anyone with a modicum of experience in life knows that these experiences are the norm, not the exception. Serious scientists, psychologists, and philosophers spend their lives trying learn about and spot these experiences to gain some certainty about what is going on in the world around them.

  • So from my understanding, that statement is not true? you can be certain using induction and deduction for example. i.e. I don't agree that certainty and truth are mutually exclusive, I think you can be both certain and truthful, for example saying there is a moon is a certain and true statement, in contrast, saying the earth is flat is certain but not true.
    – engma
    Mar 22, 2023 at 13:07
  • @engma The riposte is simple. Yes. You can be certain using induction and deduction, and I never said they were mutually exclusive. Mutual exclusion means there is a logical consequence, and I asserted they are relatively independent. My claim is that the certainty regarding induction and deduction are not guaranteed by the act, because we may do induction and deduction wrong. So, you can say, 'I deduce therefore I'm certain' and the rejoinder: 'You deduced wrongly so your certainty is misplaced'. Thus, you are certain, but not because what you conclude is true.
    – J D
    Mar 22, 2023 at 14:11
  • Let me know if that makes sense; if not, I'll clarify.
    – J D
    Mar 22, 2023 at 14:11
  • I agree that we can deduce and come up with a wrong answer and be certain about it I am fine with that. I am also suggesting that we can also deduce and come with the right answer and be certain, or in other words, I can be certain and accurate at the same time. I would also add that my proposition or question was more about the skeptical view of "You can never be certain", rather than that people can be psychologically be certain.
    – engma
    Mar 23, 2023 at 11:03
  • @engma The circularity you talk about is a form called recursion, and we reason about reason, we are recursing. Thus certainty about uncertainty is a metalinguistic statement. It's a metaepistemological position to be fallibilist which is the belief that we can be relatively certain as a general rule that we must be relatively uncertain about specific claims. IBE and induction accept uncertainty as a premise of their function. Only deduction aspires to certainty. But deduction is subject to error, bias, etc.
    – J D
    Mar 23, 2023 at 14:49

The statement that we cannot be 100% certain about anything is patently false. I am 100% certain I can not lift an ocean liner unaided, that I can not run a mile in thirty seconds, that the mass of the Earth is more than a kilogram, that China is not France, that the Moon is further than London, that the word 'emancipation' is not the same as the word 'banjo', that some politicians are liars, that Shakespeare is famous, etc etc etc.

However, even if that were not the case, you could eliminate any scope for considering there to be a paradox by simply extending the statement to say 'you cannot be 100% certain of anything apart from the truth of this statement'.

  • Eliminating that scope doesn't break the proposition, the statement you mentioned can be easily rephrased to be: "I am certain that you cannot be certain" or "You can only be certain about not being able to be certain", to me the only way out is to just say "You can be certain about things with absolute certainty, but not everything"
    – engma
    Mar 22, 2023 at 13:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .