The more i study science and maths these days, the more i continuously question myself whether all the knowledge we have built up until now is in fact true or correct.

But before being picky with what true is, like i know in science that knowledge will never be perfected, as scientific theories are just theories which claim to understand the natural phenomena in a better way than before, justified by experimental verification. But the scientific theory or knowledge built may always be superseded with a better scientific theory, which questions whether at any point in time what we know from science is really true at all?

This is just one aspect of the confusion which i am having; however, more generally, if i were to say everything we know up til now is wrong, does this claim have any validity?

In other words, if i were to say that everything we know is wrong, is this considered true?

  • 1
    The ultimate test of truth is generally whether the entire body of human knowledge uses an economical number of concepts, whether all concepts integrate with one another (i.e. there are no contradictions between different areas of knowledge within the body), and the entire body of knowledge proves to be useful for human purpose in practice. It cannot currently be said that all we know is truth, but a measure of the quality can still be how large the body of knowledge is, the scope of circumstances covered, and the number of contradictions it has produced relative to its overall size.
    – Steve
    Commented May 28, 2019 at 17:03
  • 1
    You might be interested in the KK-Principle (iep.utm.edu/kk-princ). It's the statement that if you know p, then you know that you know p. Externalist epistemologies often reject this principle, saying that all you need to have knowledge is certain conditions in place, whether or not you know these conditions are in place or not (and therefore, whether or not you know that you know). Commented May 28, 2019 at 18:38
  • 2
    By the usual meaning of "know" everything we know is right, if it is wrong we do not know it, we only think we know it. The usual position is that most of things we think we know are right, and hence we do know them, but some are wrong, and we do not know which. That is why knowledge can always be perfected.
    – Conifold
    Commented May 28, 2019 at 19:35

4 Answers 4


This is a very common issue when dealing with science. Much of science's approaches to Truth (with a capital letter) is through abduction, an approach which assumes the most likely hypothesis is true. If you read the linked SEP article, this is fraught with nuances, as you suggest.

Personally, I am a fan of radical skepticism, and the Aggripan Trilemma. It's not popular these days to declare that one cannot truly know anything, but if nothing else it's an interesting foil to challenge the current tendency for other systems of thought to declare they have the Truth.

As an example, I find it is very common to assume that we know the true semantic meaning of a word or sentence. While we tend to permit axioms to be challenged, tested, and sometimes proven, the behavior of the language we are using is typically assumed to be true (for good reason: there's major issues that arise with a language defining its own semantics). What if the answer is simply that we are not using the same meaning of "knowledge?" I might be communicating about what I believe knowledge is, and you might be communicating about what you believe knowledge is!

An interesting follow on might be to question how one might live life if nothing can be known, using whatever concept of knowledge suits you. I believe there are some related things targetted by Zen Koans, which many state are designed to try to break you out of the way you think. Do we really need to know Truth? Myself, I find the answer I get about that question depends greatly on the individual doing the answering.

  • 1
    Actually, it is almost universal today to declare that all knowledge is fallible, but that does not amount to radical skepticism. There is a difference between fallible/imperfect knowledge and no knowledge at all. And I am not sure how "behavior of the language", or any behavior, can be "true", that does not seem to make sense.
    – Conifold
    Commented May 29, 2019 at 19:37

From a science perspective, we don't know that what we know is right. You could argue that at the most fundamental level the only thing that an individual knows is that they exist, that they are conscious.

It is also a mistake to say that science can "understand" natural phenomena. Scientific theories can really only describe natural phenomena. They do this by creating a model, often a mathematical model. For example, it is not correct to say that that anyone "understands" quantum theory. We have equations to describe the behaviour of very small particles that are moving at a large fraction of the speed of light. It turns out that compared to objects in our everyday experience, the behaviour of these particles is bonkers. This means that it essentially impossible to "understand" the behaviour of say, an electron in the same way that I "understand" the colour green, or the word "dog".

Scientific theories are models that describe natural phenomena. Our current theories do a better job at describing these phenomena than previous theories, and it is plausible that better theories will supplant (more likely, improve) our current ones.


It's impossible for everything we know to be wrong.

There is certainty in our thought.

Consider the following: (i) not everything is what it seems. This means that the objects of our thinking can differ from what we think they are. This claim cannot be disputed without assuming its validity first. Thus, trying to dispute this claim necessarily leads to a contradiction.

By the Principle of Non-Contradiction demonstrated by Aristotle in his Metaphysics, we know contradictions are not true.

From (i) [the possibility of error] it follows that the external world is not created by our mind, but rather, is independent of it given the possibility of inadequacy between our thoughts and the external world.

Thus, we know something besides our conscience exists.

This is an example of something you could know for sure, but by studying logic and a priori claims there is more you can discover.

A priori knowledge is the only we can have certainty about. Logic is a priori and every empirical scientist assumes the validity of logic in order to do science. Thus, any science has an a priori foundation. This foundation is what we can know for sure, the rest could be the case or not. But don't discard empirical knowledge for a priori reasoning is limited.


One of the most influential definitions of knowledge dates back to Plato, and describes knowledge as "justified true belief." We cannot "know" anything that is not true, we cannot be said to "know" something that we do not believe, and we don't actually "know" something if we don't have any legitimate reason for believing it (even if it's true).

From the point of view of radical skepticism, we have no access to whether things are actually true or not. So, a lot of modern thinkers tend to focus on what justifies our beliefs. From this point of view, science is essentially a standardized process for generating linked networks of mutually justified beliefs. Disciplinary hygienes such as empirical experiment, peer review, randomized trials, and so forth, are designed to minimize the admission of unjustified beliefs to the established canon. In theory, every scientific "fact" accords with all the others.

You are right if you say that we don't know if any given scientific belief is true --and in fact, wrong beliefs go out under the scientific imprimatur all the time. However, most people consider scientific beliefs to be largely reliable in the aggregate. There aren't many other systems that do as well at conserving and extending justification. Systems of religious belief have a putative advantage over science in as much as they are typically founded on claims of direct access to truth (rather than inferred truth). But many of their elaborated dogmas are only loosely justified by (or even connected to) their core truths/beliefs.

You must log in to answer this question.

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