Consider a thought experiment involving 'something' and three individuals attempting to understand it: one person claims it is a red ball, another asserts it is a simulation, and the third insists it is a cat. Now, an omniscient entity enters and declares it to be X. If we define omniscience as possessing complete and certain knowledge (implying infallible justification), it would imply that X is the true nature of the 'something,' rendering the other possibilities false. (Note: A possibility where something could be subjectively or simultaneously be a red ball, cat, 'X' etc. itself is another possibility but it is still in contrast with the other possibilities mentioned before/not same as the before)

However, the omniscient entity cannot disprove the other possibilities as false since the 'something' could potentially be a red ball, a simulation, or a cat, because they are possible explanations or definitions for that something. Basically, if something has R1, R2, R3, etc. ways to exist/explain/occur, then you cannot prove it to be 'X' because that would imply R1, R2, R3 in the first place are not possible ways at all but by definition they are. The only way one could be certain is if somehow 'X' is the only possible way to explain or exist. Even if the entity were infallible, knowing everything without error or deception, they would need to prove this infallibility. Yet, they cannot establish that their knowledge is exhaustive or devoid of falsehood.

Consider another scenario: a being capable of creating a simulated reality with conscious entities. These entities experience subjective qualia, which are unobservable by anyone other than the experiencing entity. This implies that even the creator of the simulation cannot have absolute certainty about the subjective experiences within it. Applying this reasoning to omniscience, if qualia exists (not saying it does for sure, just seeing the implications down this path), there are aspects of reality unknowable to any entity, including an omniscient one. It is conceptually possible for the being to experience that same qualia, but the problem here is, it cannot be certain if it is actually facing the same qualia.

Furthermore, even if we assume a higher power created the universe, it cannot prove its own act of creation, as it would rely on its own perception and could question its memory as being fallible. In conclusion, complete knowledge of reality is impossible for any entity, human or divine, due to limitations in perception and the existence of subjective experiences that elude direct observation or measurement. Claims of omniscience lack a solid foundation and cannot be substantiated. The arguments surrounding omniscience often lead to circular reasoning or infinite regress, and can lead to rendering the concept self-contradictory. The argument is based on the definition of 'omniscience' stated earlier in this question itself, and not on any other definition.

  • Comments have been moved to chat; please do not continue the discussion here. Before posting a comment below this one, please review the purposes of comments. Comments that do not request clarification or suggest improvements usually belong as an answer, on Philosophy Meta, or in Philosophy Chat. Comments continuing discussion may be removed.
    – Philip Klöcking
    Commented May 7, 2023 at 13:15
  • 3
    Omniscience is meaningful concept in theology, not in human knowledge. Commented May 7, 2023 at 16:18
  • 3
    You confuse knowing, & being able to prove. We all know things that we can't prove. The precise nature & definition of qualia are a subject of philosophical dispute, so you can't simply take their unknowability by any other being as a given.
    – CriglCragl
    Commented May 7, 2023 at 17:55
  • 1
    Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking.
    – Community Bot
    Commented May 9, 2023 at 19:14
  • 1
    @JD My argument isn't just based or rather restricted to physicalism, it spans beyond that. The first thought experiment I presented is not limited to physicalist perspectives but encompasses various possibilities and interpretations. Also some part of the argument goes kind of towards concepts like qualia, which touches upon broader philosophical topics. The arguments goes at an epistemological extreme. Commented Jun 8, 2023 at 17:12

8 Answers 8


The three situations you offer to exhibit that omniscience is incoherent or self-refuting have major problems and need more development to work. The common theme is you are assuming that omniscience would rely on procedural or discursive ways of knowing, and not immediate or intuitive ways of knowing. I’ll highlight the biggest flaw of each:

(1) As others have mentioned, being omniscient and being able to demonstrate what you omnisciently know to finite beings might require omnipotence. It’s conceptually possible for a being to be omniscient and extraordinarily impotent aside from receiving information.

(2) There is a crucial aspect of the qualia debate in philosophy of mind being overlooked regarding ineffability. What does ineffable mean exactly? It means it cannot be rendered in language or propositionally. When each of us feels pain, we clearly understand it at the level of feeling as we rapidly respond to a painful situation. Mystics often find their experiences deeply meaningful although they cannot discuss them with those who have not had them themselves. Now, if physicalism demands that all knowables must be rendered known through propositions, then qualia poses a problem. An omniscient being is not limited to language, but presumably, can know everything directly (whatever that means). So, qualia is not a problem for omniscience. Pantheistic or panexperientialist views of omniscient beings would easily bypass the privacy concern as well.

(3) An omniscient being would not rely on memory to know its own origins. It would presumably have direct and immediate access to information about the past (including its own). So the fallibility of memory simply won’t apply here. Classic phenomenology (Husserl and Scheler) held we could have apodictic insight for some issues (that is, self-verifying intuitions). Now while their claim is probably wrong regarding humans, it’s not logically or conceptually incoherent. An omniscient being would have these regarding everything.

Sure, these considerations expose that omniscience is definitely weird and unphysical (but who proposes that omniscience could be a property of a merely physical being?). But logically inconsistent? I don’t see how.

  • Well omniscience is logically inconsistent, in the sense, that if there are multiple ways or explanations for something to exist, then it would be impossible to know, because as stated earlier, there isn't any one way. So, there would still be a room of doubt, because if such an omniscient being didn't have a question of 'why' for the existence of something, it would imply that something has 1 way to exist, but something can possibly have multiple explanations to exist always. Commented Oct 8, 2023 at 11:27
  • @SiddharthChakravarty Sure, there can always be many explanations, but an omniscient being wouldn’t have to use inferences from the fact of existence to then speculate about the cause, it would know both immediately and intuitively (again I refer to the idea of an apodictic intuition, a self-verifying insight). We have this issue because explanations are retrospective activities for us. Just as an omniscient being won’t rely on memory, it won’t rely on inferences either.
    – Hokon
    Commented Oct 8, 2023 at 17:07
  • the problem here would be then, it would need to know that it's intuition is correct, the possibility of being incorrect arises then. The argument would thus become circular, or one could resort to an infinite regress here too. Commented Oct 9, 2023 at 13:15
  • @SiddharthChakravarty We seem to have at least a few self-verifying intuitions of our own: when we visualize 5+7=12, we know immediately and apodicticly that we are correct. An omniscient being would have the same self-certainty that we have regarding 5+7=12 with everything. It wouldn’t need to make arguments, just as we don’t when we visualize 5+7=12.
    – Hokon
    Commented Oct 9, 2023 at 15:57
  • 5+7=12 is not something uncertain, because it is so only i.e just one way and not multiple explanations, the question 'why' doesn't arise, it isn't the same for asking what 'something' is. Commented Oct 9, 2023 at 17:58

Your question is absurdly vague, as the answer depends so much on your definitions of omniscient, entity and possible. It is also self-contradictory in at least two different ways. At one point you define omniscience as knowing all that it possible to know, but later you cite things that are impossible to know as a reason for denying omniscience, when your definition rules out that objection. Also, you refer to the omniscient entity as possibly having a fallible memory- what???

If we take a literal interpretation of your question as referring to some physical being with up-to-the-minute knowledge of everything within the Universe, the answer is clearly no. To cite just one reason, omniscience in that sense would require the being to know the instantaneous configuration state of every particle in the Universe, and since according to our current physical theories information cannot travel faster than the speed of light, there would be a hard limit on the rate at which the entity would be able to harvest it

  • Even if you go with that definition, you could go on to show that the being cannot be certain of its own beliefs and information or rather question it, thus it wouldn't be able to prove that it knows or has 'knowledge' (by 'to know' or knowledge I mean to have certain information) Commented May 7, 2023 at 16:20
  • Also you brining in physics here doesn't make sense, as that would mean you are asserting physics is true, when we bring up the possibilities of simulation, dream, etc Commented May 7, 2023 at 16:27
  • 1
    You are proving my point about the question being absurdly vague. You ask whether something is possible- what do you mean by possible? If you mean 'possible notwithstanding all the things that we know would make it impossible' then you are asking a meaningless question. Commented May 7, 2023 at 16:58
  • 3
    That itself is nonsense. You cite the impossibility of knowing the qualia experienced by others- why is it not conceptually possible to read minds? Commented May 8, 2023 at 2:45
  • 3
    If the entitity is omniscient then it knows the qualia experienced by you and knows your mind. If it doesn't then it isn't omniscient.
    – gnasher729
    Commented Jun 9, 2023 at 8:02

Here, since the ideas in the comments don't seem to satisfy you, I'll just be putting out some slapdash arguments:

Note: I will try to cut through the vague definitions and their ramifications as much as I can, most of these have already been hashed out in the comments.

  1. Regarding proof: The commentors pulled this up already, a proof constitutes steps of known theorems meant to realize a hitherto unknown theorem. If you know a fact as true already, there is no need to prove it. If you don't immediately agree, ask yourself why axioms don't need to be proved.

  2. Regarding deception of the being: If the being is being deceived, there is someone (or something), doing the deceiving. This being obviously knows more than the previous being, and, if you head down the chain long enough, the amount of information they don't know decreases (approaches 0 if you like :/)

This isn't really necessary because, like the commentors said, by the definition of omniscience, the being would know that it isn't being fooled, otherwise, it's not an omniscient being we're playing with.

  1. Regarding possibilities: This comes up a lot in your arguments. One piece of advice before we start though; It does no good to think about every possibility. If you're considering a million possibilities, it means you're considering 999,999 ones that were never realized. This is true regardless of whether something might or might not be true, i.e., only one fact is true (The car is behind the door you select in a game show, a family's second child is a girl (regardless of the probability that it could have been a boy))

I hope that gives you a new way to think about that particular kind of argument.

  1. Regarding qualia: Only a concept (see disclaimer). Like you said, if we accept the existence of qualia as being something that no-one can know except the entity experiencing the event, then obviously, by definition no-one can know it. (Circular reasoning alert!)

Alternatively, I could say that the existence of an omniscient being renders impossible the existence of qualia.

I think that's all your arguments...

Disclaimer: None of this was created under the delusion that I am up to date with current philosophical theories. I beg pardon for anything that betrays my ignorance of the same.



An entity is something that exists apart from other things, having its own independent existence; dictionary says.

Strictly speaking, by this definition, we cannot say that even each individual (even man) is an entity. Humans have no existence apart form other things, having its own independent existence. Also, existence of each individual is not permanent. A proof must be true if it is a proof. And the truth must be immutable. Then, how or where to store the proof in a mutable about/from the immutable/omniscient? (I believe you can understand what depth this question is from) So, I mean the truth from an omniscient and the truth from an immutable need not be the same as you doubted. If it were the same, there would be no difference in the omniscient from the other three.

And 'the entrance of that omniscient entity to prove something' actually implies that the omniscient you created is an idiot who cannot identify/understand even the other three ‘creatures’ in front of it.

Even if we assume a higher power created the universe, it cannot prove its own act of creation, as it would rely on its own perception and could question its memory as being fallible.

Many of these kinds of ideas are due to our misconceptions of separating the creator from the created.

Complete knowledge of reality is impossible for any entity, human or divine, due to limitations in perception and the existence of subjective experiences that elude direct observation or measurement.

The idea regarding ENTITY and omniscience never goes hand in hand. And omniscience is 'not a an IDEA we knew or experienced before'. So, only a vague 'idea' would always be there until one is escaped from misconceptions.

  • I never said that the omniscient entity has to prove its omniscience to a mortal being (I guess you didn't mean immortal, considering it a typo). The argument explores that the entity would question its own omniscience, and thus would be uncertain, and thus not omniscient. Commented Jun 8, 2023 at 16:00
  • This answer is the most useful.
    – Scott Rowe
    Commented Oct 7, 2023 at 12:55

Yes, they are impossible according to our current theory of quantum mechanics. In particular, the Kochen-Specker theorem shows that for some particles, their squared spin (and thus their spin) is not determined prior to measurement. An omniscient being would have to measure particles in order to know their spin, and such measurements would be detectable in laboratory experiments.

To make this vivid and concrete, I'll explain Albert's sock drawers. This is a thought experiment which presents a discrete version of a laboratory experiment; if the thought experiment can't be explained classically, then neither can the laboratory experiment.

  • There are nine drawers aligned on a three-by-three grid.
  • At the beginning of the experiment, all drawers are prepared so that they either contain no socks or one sock, with half likelihood of either case.
  • The experimenter may open one row or one column, three drawers total, and take all socks. The other six drawers are permanently inaccessible.
  • Here's the punchline: By experiment, each row always has an even number of socks (zero or two), and each column always has an odd number of socks (one or three).

Summing the columns, the total expected number of socks in all drawers is odd, since it can only be one of (3, 5, 7, 9); summing the rows, the expected number is even, one of (0, 2, 4, 6). This is a logical contradiction. Therefore, the drawers don't have predetermined contents; in some fashion, the drawers choose how to respond to the choice of row/column.

So, omniscient beings are impossible, in the sense that they can't tell how the drawers are configured without choosing a row/column, leaving a clear sign of tampering.

  • Thank you for presenting it through the lens of quantum mechanics, however, the argument I presented is not just restricted to our current understanding of physics, as our current understanding of physics itself is one of the possible ways for reality to exist. The argument doesn't explore the existence of such an entity just through one possible way of reality to exist, but rather talks about the difficulty which arises due to multiple possible ways for reality to exist. Commented Jun 8, 2023 at 15:57
  • 1
    "Summing the columns, the total expected number of socks in all drawers is odd; summing the rows, the expected number is even." This seems off. The expected value for any row/column is 3/2 -- not an integer, thus neither odd nor even -- and the sum of the expected values is 9/2. I think the problem is just saying "expected value", instead of something about the observations fitting in the even/odd categories
    – Dave
    Commented Jun 8, 2023 at 17:57
  • @Dave: Rows can only have 0 or 2 socks. Columns can only have 1 or 3 socks. Thus, if we were able to open all rows, we would expect an even number of socks (0, 2, 4, 6) while if we opened all columns, we would expect an odd number (3, 5, 7, 9). The linked article explains in more detail. The point is that classical drawers can't be prepared in this manner.
    – Corbin
    Commented Jun 8, 2023 at 18:03
  • 2
    but that is not a statement about expected values
    – Dave
    Commented Jun 8, 2023 at 18:22
  • 1
    @Corbin I think I get the point, and if I knew a good way to state it, I'd have suggested it. The problem is in the logical statements "The outcome is even(odd)". If you partition the set of trials into rows vs. columns, then for every trial where "rows" is opened the logical statement "the outcome is even" is true. (and odd for columns). This should be provably impossible for any regular probability distribution that is independent of whether rows vs. columns were chosen. I don't know how to express that cleanly and concisely. I'm mainly quibbling with the use of "expected number" here
    – Dave
    Commented Jun 9, 2023 at 15:57

No ,omniscient entity is not self refuting. You are assuming that “something” can be understood in only one way like cat , simulation or ball etc. I can call myself omniscient if I see a universal truth in that something. For example I can say that whatever that something is , is impermanent. I see everything with an eye of impermanence.

Apart from something there are other things as well like feelings , perceptions, choices , consciousness which are subjective. Here also my omniscience works , they are all impermanent.

There multiple ways of understanding something from universals to particulars ,and all can possibly be correct.

  • You haven't just moved the goalposts, you are playing an entirely different game :-)
    – Scott Rowe
    Commented Oct 7, 2023 at 12:49

You seem to have two primary problems. First, you seem to be arbitrarily limiting your "omniscient" being to be unable to know certain things, and then demanding that it know them. That's not self-defeating, it's a tautology. You're essentially asking if there could be any married bachelors. There is no reason to presume that for any practically omniscient being couldn't know the true nature of a thing as well as any subjective and experiential qualities of that thing. All omniscience which people typically consider also includes that ability.

Secondly, you demand that an omniscient being should have need of some external proof that it doesn't know about. An omniscient being doesn't require proof in the first place as it knows everything and knows that it knows everything. The proof, from its perspective, is the fact that it knows it to be true. We require independent rational justification because we are not omniscient and nothing in our thoughts confirms the truth of a thing. None of that applies to an omniscient being.

So, no, an omniscient entity is not self-refuting. It is, however, possible to define a thing and explicitly define contradictions for it, so that it necessarily contradicts, and then you may call it anything you wish, but then we're only talking semantics and tautologies, and not discussing any real or useful concept.

  • The problem lies in the fact that something could possibly exist in some other way too. The burden which would lie on the omniscient being would be to prove that there is only one way for that something to exist, and not multiple ways. It is the same reason why we humans are uncertain too. Commented Oct 8, 2023 at 11:50
  • The only way in which you could not have immediate proof, by definition, is that you unnecessarily restrict your definition of omniscience to the extent that it isn't meaningful. So, even if you conceived of new ways for something to exist, you would then also have to extend your understanding of omniscience to include it for omniscience to be omniscience.
    – DKing
    Commented Oct 8, 2023 at 13:38
  • I hope you understand here something is what the omniscient being is trying to say what he knows it is, if there are multiple ways for that something to exist, it won't make sense for the being to say it knows what that 'something' is. Commented Oct 8, 2023 at 17:44

Being an omniscient entity requires knowing entirety of its own state, plus entirety of the other parts of the universe. A system cannot contain more information than the substance it is made from. So the lowest bound of an omniscient entity is to be the universe itself and utilize every single bit of one's substance to store information.

But it is not enough to just store information. "Knowledge" is a capability to make queries. So there must also be some equipment to access the information.

If the equipment is made from the storage matter, then every time it is used, the state of the storage is also altered; meaning some information has to be sacrificed in order to perform the operation.

To secure the information from being lost, the equipment must be made of non-storage matter. But there's no non-storage matter.

So is it really possible to avoid contradictions when talking about omniscience?

As for any qualia, it is rather experienceable than knowable domain of phenomena, because cannot be converted into bits and used in any deterministic sense (discovering laws of nature, creating tech, etc.).

So here you might already want to discuss omniexperiencing entity. Which is a whole different topic in its own right.

You must log in to answer this question.

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