Ethical Realism

March 26, 2011

Is Knowledge Impossible?

Filed under: epistemology,philosophy — JW Gray @ 12:13 am
Tags: , ,

Some people have thought that knowledge is impossible. It might seem implausible to think knowledge is impossible, but there are important philosophical concerns we can have about knowledge and challenges to the possibility of knowledge can be illuminating. First, I will discuss what ‘knowledge’ means and suggest three different definitions: (a) justified true belief, (b) certainty, and (c) a deep understanding. Second, I will discuss why the belief that knowledge is impossible seems to be self-defeating. Third, I will discuss an argument against the possibility of knowledge known as the Münchhausen Trilemma and explain where it might go wrong. The Trilemma supposedly shows how unsatisfying any proof is in order to show that none of our beliefs are proven—and knowledge is taken to be impossible as a result. I reject the Trilemma because it is too demanding about what counts as a justified belief. Perhaps proof or evidence is not always necessary to have a justified belief.

1. What is knowledge?

We all know what knowledge means to some extent already. We use the word in everyday discourse and we can often spot when someone uses the word wrong. We ask a child if she knows she has hands and she is supposed to say, “Yes” or we worry that she doesn’t know what ‘know’ means. We ask a child if she knows what planets in the universe contain intelligent life and the answer, “No” seems like the correct response here. At some point we have an intuitive understanding of what ‘knowledge’ means. This intuitive understanding is based on how people actually use the word, but it is fallible. There can be mistakes made when a person uses the word.

Philosophers can give technical definitions for the word ‘knowledge’ in an attempt to help us understand and correct our intuitive understanding of the word. The technical definitions must be based on our intuitive use of the word, or we aren’t even talking about knowledge anymore. We would be talking about something else. The intuitive use of the word should be sensitive to what ordinary practice says are examples of knowledge. If the police take me down to the station for questioning and ask me if I know that George is the killer, I’m not supposed to respond saying, “It’s impossible to know if a person is a killer.” The police will not be amused by this statement and are likely to respond, “You know what we mean. Just answer the question.”

I don’t think the word ‘knowledge’ necessarily has only one meaning. It can be used in different ways in different contexts. I suggest the following definitions:

  1. Justified true belief
  2. Certainty
  3. A deep understanding

I will discuss the three suggested different definitions of knowledge.

(a) Justified true belief

Consider an example of a justified true belief. I can see a cat on a mat and believe that a cat is on a mat as a result. The belief is true assuming there is a fact (an actual cat on a mat) that my belief refers to.

Justified true belief is a modest type of knowledge. A belief can be justified without being proven. As long as a belief is reasonable (compatible with proper standards of reasoning), the belief could count as being justified. A justified belief is a belief that is reasonable to have. It seems plausible to think that beliefs can be justified even when no argument is given for them at all because a belief could be reasonable as long as it’s not unreasonable—as long as we don’t keep a belief that we have good reason to reject. For example, I think we all know that “1+1=2” even though we can’t all give an argument for it. This modest use of the word ‘know’ does not necessarily require an argument, certainty, or a deep understanding.

What does it mean for a belief to be true? The word ‘true’ often refers to a correspondence. If a statement corresponds to reality in the appropriate way, then it’s true. We can say that true beliefs correspond to facts, and a fact is the reality that true beliefs correspond to. It might be that some beliefs correspond to reality better than others. For example, the belief that the Earth is a sphere is accurate, even though the Earth is slightly pear-shaped and spheres are ideal mathematical objects that can’t be pear-shaped. We might say that it is true that the Earth is a sphere for most intents and purposes, but it would be more true (or more accurate) to say that the Earth has a shape somewhere between a sphere and a pear.

That the word ‘true’ itself is somewhat unclear, and we should often prefer to use the word to mean something like accurate (that allows for degrees of truth). This is one motivation behind fuzzy logic and multiple truth values, and it seems to capture how we use the word ‘true’ in everyday discourse pretty well. A person can say that they believe that it’s true that the Earth is a sphere without implying that they think it maps reality perfectly.

I don’t want to suggest it really is impossible to say something that turns out to be absolutely true, but that isn’t generally a requirement of our knowledge of the justified true belief variety. General statements seem to be strong contenders for being absolutely true. Perhaps the statement that something exists is so general that we can know that it is absolutely true.

I think our use of the word ‘knowledge’ usually refers to something like “justified true belief” and this is how I usually use the word.

(b) Certainty

We often say that we don’t know something when we are speculating, even though our speculations could be justified true beliefs. We might not be certain (sufficiently know) that someone is a murderer even when we witness that person commit the crime because our mind plays tricks on us, we might worry that the person has an evil twin, and so on. At least some certainty is required before we satisfactorily know that someone is guilty of murder. What we call reasonable doubt is enough to find someone to be not guilty.

When we are certain that a belief is true, it is also a justified true belief. However, the demand for certainty is a more demanding form of knowledge.

Some people equate “certainty” with “infallibility.” However, there can be degrees of certainty. This is what many call degrees of confidence. If something is absolutely certain, then there is no possibility of error. This is the highest degree of justification conceivable. However, not all types of certainty requires such a high level of confidence.

There are different philosophical positions about what a belief we know for certain must be like. For example, one type of certainty only requires that beliefs are justified in the best way we can hope for.)

(c) A deep understanding

Sometimes we think a belief isn’t justified or that we don’t know something if our understanding is too limited. Good examples can be found in what is now called the Gettier Problem. For example, I might see a cow in a meadow and believe that there’s a cow over there. However, I might be looking at a cardboard cutout of a cow while there is really a cow over there hiding behind the cardboard cutout. I have a justification to believe that there’s a cow in a meadow and it’s true that there’s a cow in a meadow, but there’s a sense that I don’t know that there’s a cow there. This sense seems to be that I don’t have the appropriate understanding of what it means for the cow to be in the meadow. The specific cow I thought was in the meadow wasn’t actually there, and the justification I used to support my belief didn’t support my belief in quite the right way.

Some people have said that beliefs must be causally tied to reality in the right way for us to have knowledge. This is a more demanding kind of knowledge than mere justified true belief. Either that kind of justification provides us with more certainty or it provides us with a deeper understanding. Either way, I don’t find it to be a particularly impressive requirement for knowledge in everyday life because we have no way of knowing for certain when our beliefs are causally tied to reality in the right way. We would just beg the question when we say, “My belief is causally tied to reality in the right way.” Even so, some speculation involving how our belief seems to be causally tied to reality could provide us with the deeper understanding that we would like to have.

Similarly, we might prefer to have arguments and justifications for our beliefs, even though it can be perfectly reasonable to have some beliefs even when no argument or justification is attained.

Philosophers are interested in having a deep understanding of our beliefs. We want to know how our beliefs are causally tied to reality and we want arguments for them. However, it seems wrong to say that a deep understanding is always required. Even philosophers commonly rely on axiomatic assumptions. Such assumptions might be defensible, and they might have to be something we could reject on the basis of comparison—we might need to be able to decide one belief is better than an alternative for it to be justified.

2. What if knowledge is impossible?

Let’s say that knowledge is impossible—of the modest justified true belief variety. In that case we can’t know that knowledge is impossible. We would still be forced to have opinions and find some way to explain why some opinions are better than others. In the long run we would still say we have justified beliefs and that we know many things based on the ordinary use of the word ‘know.’

Can we know that knowledge is impossible? No. If knowledge is impossible, then we have no way of knowing that. We would have no way to persuade anyone that it’s true. It wouldn’t even be reasonable to try to do so. It would be strange to expect anyone to believe it or agree with it.

What if it’s impossible to have true beliefs? We can have justified beliefs, even if none of them are true. It might be that no beliefs are absolutely true, but that is not the requirement of the modest sort of knowledge that I’ve discussed. There are different degrees of truth.

What about knowledge as certainty or deep understanding? It might be possible to have knowledge in the sense of having some degree of certainty or deep understanding, but it might be impossible to have absolute certainty or absolutely deep understanding. Those extreme kinds of knowledge could be impossible.

3. The Münchhausen Trilemma

Here is what wikipedia has to say about the Münchhausen Trilemma:

If we ask of any knowledge: “How do I know that it’s true?”, we may provide proof; yet that same question can be asked of the proof, and any subsequent proof. The Münchhausen Trilemma is that we have only three options when providing proof in this situation:

  1. The circular argument, in which theory and proof support each other (i.e. we repeat ourselves at some point)
  2. The regressive argument, in which each proof requires a further proof, ad infinitum (i.e. we just keep giving proofs, presumably forever)
  3. The axiomatic argument, which rests on accepted precepts (i.e. we reach some bedrock assumption or certainty)

The first two methods of reasoning are fundamentally weak, and because the Greek skeptics advocated deep questioning of all accepted values they refused to accept proofs of the third sort. The trilemma, then, is the decision among the three equally unsatisfying options. In contemporary epistemology, advocates of coherentism are supposed to be accepting the “circular” horn of the trilemma; foundationalists are relying on the axiomatic argument. Not as popular, views that accept (perhaps reluctantly) the infinite regress are branded infinitism.

The skeptics argued the following:

  1. All knowledge requires arguments.
  2. All arguments are ultimately circular, regressive, or axiomatic.
  3. Circular, regressive, and axiomatic arguments can’t provide us with the justification that knowledge requires.
  4. Therefore, there is no knowledge.

Let’s assume that ‘knowledge’ here is the modest sort—justified true belief. Why do I reject this argument? First, because the conclusion that “there is no knowledge” seems to be self-defeating. If there is no knowledge, then how can we know that there’s no knowledge?

Second, because the conclusion is so counterintuitive. Our everyday discourse requires that a modest use of the word ‘knowledge’ refers to certain things or we just don’t think the person even knows what knowledge means. In other words, the Trilemma can be taken to be a reductio ad absurdum—the conclusion is absurd to the point that we’ve proven that something is probably wrong with the argument. Either the argument is invalid or at least one premise is false. People who take the argument to be a real proof that knowledge is impossible seem to miss the point.

Third, because the argument is overly ambitious. Extraordinary claims require extraordinary evidence. The conclusion is counterintuitive and the premises are not obvious. The premises could be false and need a great deal of justification, but I have never seen a satisfying justification for the premises. (In fact, the argument requires that the premises couldn’t be satisfactorily justified.) That’s not to say that I know for certain which premise is false. Any of the premises could be false and the argument gives us reason to question them.

Fourth, the premise that knowledge requires arguments seems unjustified and I think it’s probably false. (a) The belief that all justified beliefs must be justified through argumentation could be self-defeating if there is no good argument for it, and I haven’t heard of any good arguments for it. (b) We seem to know some things even though we don’t know how to provide good arguments in support our knowledge. I know that “1+1=2” even though I can’t give an argument for it. This was already discussed in detail above. This implies that beliefs can be justified even when no argument can be given for accepting it.

Perhaps the Trilemma isn’t meant to disprove knowledge of the modest sort. Perhaps it can be used to disprove that we can be absolutely certain that something is true, or that we can have a deep understanding of anything. I am not sure if the argument succeeds against those conceptions of knowledge, but I think such conceptions of knowledge are more ambitious than we should generally claim to attain anyway.


A modest definition for ‘knowledge’ is all that seems needed in most everyday contexts. It tells me if I can wake up in the morning, if I can type out an essay, and if I can think anything worth putting on paper. This modest definition is compatible with our quest for truth, certainty, and deep understanding; but not all of our knowledge is absolutely certain to be true nor is all of our knowledge a deep understanding.

Given the modest definition, we will find the claim that knowledge is impossible self-defeating and the Münchhausen Trilemma to be unconvincing.



  1. I agree with you on much of your post, however, as I look at the subject from a mathematical/ logical perspective, there is a difference between knowledge and “knowledge of.” And this difference may address whether knowledge is possible or not. I agree that the statement “knowledge is impossible” is self-defeating just as saying “truth doesn’t exist” is self-defeating. What we know, however, may be little more than “I am.” The reason we can know this is because another cannot have my thought. It is a Point Of View issue. Even if all is a dream, I would ultimately be the author of the dream. Beyond that we only believe we have knowledge, and the values we are reassigned are reassigned as our perspectives shift.

    So then we get to the matter of knowledge of. And that goes along with most of your post. Knowledge is an internal condition where “knowledge of” is an external condition. That knowledge of falls into one of the three categories you provide but it isn’t knowledge but a perception of knowledge. In other words, it is a belief and is as fallible as our limited, ever shifting perspectives.

    Logically, true “knowledge of” could only be possible for an omniscient, omnipotent, and omnipresent entity. I’m not attempting to make a case for God here but rather to point out that without all three of those characteristics (and all three would apply whether we live in a universe or multiverse) the limitations of our perspectives prevent us from knowledge other than our knowledge of our own existence.

    Comment by themathofgod — January 6, 2012 @ 1:48 am | Reply

    • I don’t understand what you mean by “knowledge of” or why you think only God could have it. Are you talking about certainty? The fact that there’s an “external condition” shouldn’t imply that it’s impossible to know something. Why can’t we know things with external conditions?

      Comment by JW Gray — January 6, 2012 @ 5:30 am | Reply

      • That’s EXACTLY what it implies.We can’t know things with external conditions if our awareness is supposedly internalized — if we believe there to be duality between subject and object. If we were really all subjects in an “objective world”, then in truth we would all be solipsistic. Solipsism disallows any knowledge of an outside world, since any ontologically “internalized” subject is simply making up their own universe.

        In reality, there is no duality — the experience itself IS the reality; and this perspective ALSO disallows any knowledge. Knowledge implies a “knowing” of something beyond any experience; but really you can’t separate the experience from any apparent “knowing”. If I’m thinking that I know something, the reality is that there is an EXPERIENCE of an alleged individual thinking that it knows something. There really aren’t any individuals at all.

        Comment by Elijah Lynn — May 23, 2016 @ 10:27 am

  2. […] Is Knowledge Impossible? […]

    Pingback by What is the Burden of Proof? « Ethical Realism — May 6, 2012 @ 3:53 am | Reply

  3. This is exactly what I was looking for! Thanks for the link 🙂

    So I agree the conclusion is self defeating but only for an academic skeptic. Pyrrhonian skeptics withhold any assent with regard to non-evident propositions and remain in a state of perpetual inquiry.

    Im glad you brought up the Gettier Problem as well, I also JTB is insufficient for absolute certainty or deep understanding as you put it. JTB is necessary but not sufficient for knowledge. The Barn Facades thought experiment shows that the addition of the “no false lemmas” criterion is still insufficient to define knowledge.

    I think certainty in the absolute sense may be unattainable. Truth really boils down to pragmatism and the correspondence theory of truth. However, you pose great problems.

    1. Muchhausen’s trilemma is not a proposition, theory or statement. It is a conjunction of several inconsistent but intuitive propositions. Basically things that you probably hold, but disagree with each other. So it cannot be true or false since it is a set of propositions. The skeptic is not advocating for it’s truth but it’s a logical conclusion which must be grappled by the dogmatist or the internalist. The point is not that the skeptic asserts that Munchhausen’s Trilemma is true, but rather that the dogmatist asserts that all of the premises of the trilemma are true, but denies the conclusion. Since its valid, this means they have to abandon logic or give up one of their premises. Though giving up any of the premises also creates some problems.

    2. You havent addressed the main problem. Is the argument invalid (which doesnt necessarily mean the conclusion is false) or is one of the premises falses. The BoP would be on the dogmatist rather than ignore the problem which you have done. It may seem like a veridical paradox but one that ought to be addressed if we assert that absolute certainty is possible and most of our logic follows reductio ad absurdum.

    3. Your syllogism isnt something generally held by the more esoteric skeptics. See this link

    If beliefs can be justified even when no argument can be given for accepting it then that would lead to alot of problems. There are two basic types of foundtionalism, fallible and infallible.

    Comment by coriechampion — May 16, 2015 @ 2:00 pm | Reply

  4. I have yet to see a convincing argument that doesn’t begin its assumptions by begging the question as the problem of the criterion addresses. Therefore, if this problem can only be answered by begging the question, skeptics and non-skeptics alike have no way of knowing whether it is possible or impossible.

    Comment by coriechampion — May 27, 2016 @ 5:58 pm | Reply

  5. The problem I have with the Gettier Problem is that it seems to overlook a more important flaw in the idea of Justified True Belief. How is it possible to know that a belief is true independently of our justification for the belief? In short, it’s not possible. All of our reasons for thinking that a belief is true is necessarily based upon our justifications for believing it to be true. Thus, the “truth” component of JTB is either useless or redundant. Sure, we want our knowledge to be something true, and not false, but including it in the “formula” for knowledge is begging the question, and doesn’t help us determine if a belief is, in fact, knowledge.

    I wrote about JTB on my own blog back in 2012:

    Comment by macsnafu — May 30, 2016 @ 7:59 pm | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

Blog at

%d bloggers like this: