Note: my question is different from this, as this asks whether only 1 and 2 are necessary, mine is asking why actual belief is needed instead of only awareness.

Why is belief needed for knowledge? I could disbelieve that the Sun is a star, but I may be aware that most people, and most intelligent people believe so, and thus I am aware of the proposition, and so:

  1. [The Sun is a Star] is true.
  2. I know that the proposition tSiaS exists.

By those requisites I should have the knowledge that the sun is a star, even if I don't believe it myself.

I have this gut feeling that awareness is not enough to qualify knowledge, but I can't really come up with a reason why — but at the same time belief seems rather trivial to me, if I am aware and it is true, I could have the knowledge whether I like it or not, belief in this case sounds like an emotional human response. Or maybe this all amounts only to how we define our terms in the end?

  • 1
    Knowledge = justified true belief is challenged for a number of reasons and success of counterproposals is judged based on matching colloquial use of "know" in specific examples as much as possible. "I know that most people justifiably believe that p, but they are mistaken", which is what you describe, most people would not judge consistent with "I know that p". There are proposals for dropping the belief requirement, but on different grounds, see Do machine learning algorithms have knowledge?
    – Conifold
    Commented Dec 3, 2019 at 10:35
  • For the 'mystical' epistemology justified true belief would not be reliable knowledge. It offers a different idea of knowledge and you may like like to look into this. Unfortunately no specialist references spring to mind.
    – user20253
    Commented Dec 3, 2019 at 13:03
  • You aim at determining the conditions of knowledge and you use the term " know(ing)" in condition (2). That seems circular.
    – user39744
    Commented Dec 3, 2019 at 20:25
  • @EleonoreSaintJames I don't aim at determining the conditions of knowledge, I am merely asking a question. And it doesn't seem circular to me, you know what I meant by 'know' in condition 2.
    – user31740
    Commented Dec 4, 2019 at 4:51
  • The combination "I know that p but I don't believe that p" seems contradictory, or at least irrational. In addition, a belief that p, in the present context, is not an emotional attitude. A belief that p is just a readiness to use p as a premise in future inferernces, in order to derive new beliefs, or to decide on actions. Commented Dec 5, 2019 at 12:43

4 Answers 4


The problem with this kind of question is (and always has been) the implicit assertion that there can be knowledge without a knower. These questions begin with the premise that there is some 'fact' (or perhaps 'such-ness' is a better term) out there in the world that is 'true,' and that the epistemological task is to accept only those propositions that express that true such-ness.

But such-ness is neither true nor false; such-ness simply is. When we talk about 'truth' we are talking about our confidence that a proposition expresses that such-ness, and confidence is always — in some way, shape, or form — a matter of belief. One must have a knower before one can entertain the notion of truth, because truth is about the relationship of 'known' propositions to the such-ness of the world.

The real grounds for debate and contention is not over the truth or falsehood of our propositions, but over the applicable standards of confident belief. Ontology and epistemology both boil down to a kind of collective hermeneutic.

  • +1. Interesting angle I hadn't condidered. Hegel and others do disagree with Descartes and Kant on who or what the knowing suibject is. Certainly not the individual person according to Hegel : Geist is a kind of subject-cum-object that comes to create what is known and to know itself in the course of history. Also later Idealist thinkers such as FH Bradley thought in terms of the whole of reality, ultimately and really indivisible into subject and object. Any divisibility was Appearance, not Reality, to adust one of Bradley's book titles. Thanks for stimulating intervention.
    – Geoffrey Thomas
    Commented Dec 3, 2019 at 15:38

Perhaps it isn't. Please excuse me if I go over at least a bit of what is familiar ground to you. By the late 1950s/ early 60s propositional knowledge - knowledge that p - as analysed in terms of justified true belief:

S know that p at time t if and only if p is true at t, S believes at t that p, and S is justified by the evidence (e) avilable to her at t in believing that p.

An internalist view of justification and evidence was generally assumed:

S believes that p at t on the basis of evidence e at t only if S is aware that she has e at t by introspection or immediate sensory awareness. (This neeeds tightening but we can work with it for now.)

Ed Gettier's bombshell in 1963 was to show that the justification condition could be met and the other conditions as well, and S still not know that p by ordinary standards of knowledge. Attention fixed mainly on the justification condition, which was qualified, refined, had sub-conditions added without, so far as I know, a winning candidate emerging. There was no reworking of the justification condition(s) that then or now secured anything like majority consent or clearly passed the test of counterexamples.

Belief as such and on its own never mattered - was never judged remotely sufficient for knowledge - only true belief and belief which we were justified by the evidence available to us in believing.

Jettisoning belief: abandoning internalism

On a different tack externalism abandons the introspection and immediate sensory awareness requirements. No matter what S believes on what evidence at t or any other time. What does matter is how S's acceptance - perhaps only dispositional - of p has come about. If there is a reliable causal connexion between S's mental/ physical state at t and a state of affairs that produced - caused it - then S knows that p at t. The belief condition falls away as otiose.

Imagine an example. If I am asked what the Latin word, mensa, means and I spontaneously reply 'table', I may have no evidence available to me at all that I am right; but I am right, and do know that the mensa means 'table' (in the relevant sense of 'table') because I was taught Latin by a reliable teacher from a reliable textbook; and moreover there is a reliable connexion between teacher, textbook and across the centuries the use of Latin by the ancient Romans. But how can I tell that such a connexion obtains? I can't and don't need to. We have abandoned internalism. The causes of my state of mind/ physical state of my brain are sufficient.

Jettisoning belief: back to Plato & intellectual vision

Some trace the justified true belief analysis of knowledge to Plato's Theaetetus (201c-210d). I hesitate over exactly how we are to take Plato's 'true belief with a logos'. Fortunately the Republic serves my purpose better.

In the central books of the Republic while the Forms are (I'll risk it) ultimate realities and perfect examples of which the corresponding objects of experience on earth are only highly imperfect approximations, Plato's philosophers can at the culmination of a long process of education apprehend the Forms. They do not form beliefs about the Forms. Rather, the Forms (or so I understand) disclose their natures in a moment (for the philosophers) of intuitive self-disclosure lor intellectual vision. The philosophers recognise - penetrate - the nature of the Forms in a sudden encounter of self-evidence.

Now, whatever one may think about this, what is depicted here is knowledge without prior or accompanying belief. Belief has no work to do. Perhaps Spinoza's scientia intuitiva exhibits parallels as might Descartes' self-validating and incorrigble clear and distinct ideas. (I have to pass over here the Form of the Good in Plato and the role of God in Descartes, which provide the metaphysical framework for their respective epistemologies.)

I offer the above discussion not (impossibly) as a full and correct answer but in order to rotate your perspective away, as I think you are looking for, from belief in the analysis of knowledge.

  • I don't know if this is talked about anywhere in philosophical literature, but it seems to me that what you say about externalism also has implications for animal knowledge. It seems obvious to me that animals know things (our cat knows where her food dish is, for example) but on internalism they seem to lack the requisite introspective capacity. I think this is a point in favor of externalism. Commented Dec 3, 2019 at 18:34
  • Why should I deny that your cat has externalist knowlege? The question was (I assumed) about human knowers and I answered on those lines. My answer did not deny this or any other kind of knowledge to cats or any other non-human animals. I agree that an internalist account of non-human animal knowledge is far more problematic. I see no disagreement between us.
    – Geoffrey Thomas
    Commented Dec 3, 2019 at 18:39
  • Oh no, there's no disagreement at all. I was actually agreeing with you wrote. Sorry if it was unclear. Commented Dec 3, 2019 at 19:14
  • 1
    And I apologise for using what plainly appeared an astringent brevity. Glad to have your agreement. You were thinkng aloud in a comment, and I often do the same. All the best - Geoff
    – Geoffrey Thomas
    Commented Dec 3, 2019 at 19:29

All you need to believe is everything is possible, this will allow you to be flooded with knowledge.. What besides a star makes a star a star.. take away the label, take away what you have been told.. look up at them and wonder what you would think they were if you knew nothing about them.. that’s what I’d like to know..


maybe this all amounts only to how we define our terms in the end

I am tempted to read this not just as the last words of the question but perhaps the last word on the question. We should distinguish ordinary language from the spin various arguments tend to introduce here.

You can believe anything, but you can know only true things. If you say "I know that 123 is a prime number" you are mistaken: 123=3x41. Pointing the mistake is usually seen to act reatroactively: you are not using correctly the verb "to know". So, believing depends on the sole believer independantly of content, while knowing depends on the sole content. (This is just like proving: no falsehood is substantially proven, either something is false or the argument is not a proof.) Seen thus, belief and knowledge are on the opposite sides of the subject-object divide. This ties in with the natural attitude which accepts that many truths are established without asking how or why they considered as such; so for many languages an emphatic distinction like "I don't believe this, I know it" is easily grasped. (Establishing the truth status of the content is a different problem.)

JTB was a pious lie with a corresponding "Legend", as recently discussed here.

You must log in to answer this question.