Traditionally, Knowledge is defined as a True Justified Belief (Let us ignore epistemic caveats and objections to this definition).

According to Wittgenstein, there is no place for knowledge where there is no place for doubt (or at least this is what I understood from references about his philosophy).

If so, then how can we define doubt in terms of Justification, belief and Truth.

I know we cannot define doubt in terms of truth, because we can doubt something that is True.

I also know we should use Belief in our definition, as doubt is a non-belief (or at least, incomplete belief).

And I am very hesitant to whether the lack of justification plays any role in Doubt, because even the most justified truths (like the sun) can also be doubted.

So, how can we define doubt? are there any epistemic or analytical theories defining "doubt" in less ambiguous terms?

References : Wittgenstein's Epistemology


  • Comments are not for extended discussion; this conversation has been moved to chat.
    – Geoffrey Thomas
    Commented Jan 28, 2019 at 14:47

6 Answers 6

  1. We can focus in on 'doubt' in the first place by drawing a distinction, which I think few would deny, between (1) 'I doubt that p' and (2) 'It is doubtful that p'. (1) is a doxastic state, a cognitive state of the person, while (2) may be merely a mathematical calculation of probability produced by (say) a computer program. The computer is not in a doxastic state - as of now (2019).

  2. We can further close in on 'doubt' by making some contrasts. I do not doubt that p if I strongly believe that p or (think I) know that p.

  3. In a curious locution I can 'half-believe' that p. I am not quite sure how to characterise 'half-belief'. We might try : I half-believe that p if I intellectually reject p (often or usually a superstition) yet nevertheless allow it to influence my thinking and actions. Thus I may intellectual reject the belief that 'touching wood' has any effect on my fortunes, yet still allow this belief to influence my actions. Rationally I know that touching wood has nothing to do with what will happen, but still say, 'My cold will be gone by tomorrow - touch wood' - and make certain that I do touch wood. (Peter McKellar, A Textbook of Human Psychology, London: Cohen and West, 1952: 320.) I might here be said practically to doubt what intellectually I reject. (A form of cognitive dissonance ?)

  4. Belief and doubt can co-exist if indexed to different considerations. In a law court, for instance, I may grant that X's guilt is beyond reasonable doubt, given the evidence available, and, having no specific reason to believe that the evidence is incomplete and relialble, believe that X is guilty. Yet I may retain a degree of scepticism about X's guilt from purely general considerations about the permanent possibility of miscarriages of justice.

  5. If (Platonically) we draw a distinction between knowledge and belief, the mere fact that my state of mind is recognisably one of belief and not knowledge, necessarily induces a withdrawal of unqualified assent, no matter what the empirical evidence.

  6. Wittgenstein's views on knowledge and doubt are not entirely clear. A standard view, expressed by Kenny, is that there is no place for knowledge where there is no place for doubt :

'We can now see why Wittgenstein rejected the idea that when I am in pain I know that I am in pain. Throughout his life he thought of knowledge as involving the possession of a true description of a state of affairs.... That is why, since "I am in pain" is not, when I am in pain, a true description in the normal sense, "I know that I am in pain" cannot be in order if "know" is being used in the normal sense. Since the truth of "I am in pain" coincides with its truthfulness, and there is no such thing as making a mistake here [GT : no place for doubt], Wittgenstein does not want to call it an assertion, or the expression of a piece of knowledge' (A.J.P. Kenny, Wittgenstein, Harvard, 1973: 201.)

  1. We should pause over this reading of Wittgenstein - I put it no stronger than that - in light of the following:

The shortest way with this position is to undercut the contention that Wittgenstein thought 'I am in pain' is not a true description in the normal sense. I would grant at once that Wittgenstein hints in a number of places that what I have called an avowal need not be a description under all circumstances. The question before us, however, is whether 'I am in pain' could under any circumstances be a description. Kenny has adduced no firm evidence that Wittgenstein gives a negative answer to this question. And indeed there is evidence that Wittgenstein would give a positive answer. Wittgenstein himself alludes to the locution 'I describe my state of mind' as if it is a permissible way of talking (Investigations, §291). Besides, Wittgenstein hints that 'I am afraid' is sometimes a description of a state of mind and sometimes is not.

  • Thank you for this detailed account +1, as it occurs to me, if we suppose that for Wittgenstein 'I am in pain' is not a description. Then it follows that we cannot 'doubt' it, since it is difficult to say that we can 'know' something that is not a description. Which does perfectly match the fact that I do not doubt that I am in pain.
    – SmootQ
    Commented Jan 30, 2019 at 17:30
  • There is yet another question as for the 2nd point: "I do not doubt that p if I strongly believe that p ", does 'strongly' here make 'doubt' more like a spectrum than discrete values "doubt and no-doubt" ?
    – SmootQ
    Commented Jan 30, 2019 at 17:35
  • 1
    Yes, you are absolutely right. If there are degrees of belief, there are degrees of doubt. I'll try - with acknowledgement - to work this into the answer. One can have a slight suspicion, consider something to be very likely, or take it to be almost certain. Different degrees of doubt match these - and probably many other states of mind. I think I kept too closely to the word' doubt' and omitted concepts, such as slight suspicion, which plainly involve the concept but not the word. Much appreciated. Best - Geoffrey
    – Geoffrey Thomas
    Commented Jan 30, 2019 at 18:51
  • Thank you so much Geoffrey, I think this is the closest to a best answer I can get, since it is already difficult to define knowledge , it is even more difficult to define terms that may (or may not) rely on knowledge itself. Even knowledge remains somehow ambiguous (i.e Gettier's problem). Thanks again
    – SmootQ
    Commented Feb 2, 2019 at 11:27

According to Wittgenstein, there is no place for knowledge where there is no place for doubt.

Likewise there is no place for knowledge where all there is, is doubt. For doubt to make any headway it must doubt something that is worth doubting, and this will depend upon circumstance and context.

Doubt was made famous by Descartes as a possible foundation of his system of thought and has been fetishised ever since; but it’s worth noting that Descartes did not make much use of it. As a foundation it does not go very far, and one might even question whether foundations are even neccessary after all thought, as life, begins in media res.

As for a definition, why not merely invert each term in your definition of knowledge? That is, you doubt whether a proposition is true, is justified or is a belief. For example, if someone believes the sun to be a sphere but actually says that the sun is a cube and insists that this is his belief. Well, we know from this account that what he says is his belief is a lie, he believes otherwise; and so his stated belief, is in fact not a belief. It’s a lie.

Now of course you might say that this is not a definition of doubt, it is only showing or demonstrating how doubt is used; well, if I recall rightly, the later philosophy of Wittgenstein on language insisted that meaning is use or maybe even, meaning is in use.

So by showing how doubt can be used to begin to answer your question should help you to understand how doubt can be understood and it’s meaning elucidated.

  • I am agree with the fact that language a tool, and that meanings (and definitions) are in the use of that tool we call language. But what makes it all look difficult is this : To know X in the TJB sense, means 'to eliminate or decrease' doubt about X , let's suppose that it means 'to eliminate all doubt' , then : Knowledge would be the opposite of doubt, so D ↔not-TJB , right?! but we know that D does not depend on Truth or Falsehood, which makes us reduce D to D ↔not-JB. But we already know that Doubt does not mean "not justified belief' , since doubt is not a belief, therefore not-B?!
    – SmootQ
    Commented Jan 30, 2019 at 15:19
  • But the concluded definition in my analysis of Doubt seems very weak, since 'doubt' is not 'not-belief'. You can believe something while doubting it. Thank you so much, I finally accept that doubt maybe is just in our use of the word.
    – SmootQ
    Commented Jan 30, 2019 at 15:19

[I believe, not only ideas but the correct usage of language plays an important role in forming any definition. Since I am not a native speaker or a linguist what I can do is to give an idea regarding the definition.]

One's all doubts end only by the unification of one's self with the Ultimate Truth (In other words, by the realization of one's real nature). Until then there must be doubt to all who haven't reached that level...at least a little. ("O Truth/God, who are you?" This is also a doubt.) Also, some people withdraw themselves and live a solitary life after their doubts completely burnt out. (I mean, after their Truth realization...) This shows they have no desires at all.

Desires and wishes keep our life going on...I didn't forget it. Keeping the aforesaid idea in mind I shall try to define 'doubt' (in terms of justification, belief and Truth) if you are badly in need of it. Since we want more precisely, I didn't try to compare the meaning given in dictionaries or in any other references including Wittgenstein's Epistemology. So I didn't use give importance to the word, 'feeling'.

Doubt is the driving force that emerges from the mind, which subsides only when one's all beliefs are burnt out by getting justification without a second thing and that happens by the realization of the Truth.

Here, I can't substitute the words desire or wish for the word, 'doubt' because the one who has desire/wish must still have doubts. But if one's doubts are completely vanished, he will have no desire/wish at all.

The following link might be useful to understand two words related to the mind:


Vikalpa literally means ‘doubt,’ ‘choice’.

The word Vikalpa has several senses. It's different inference as per varied theories is as follows:

When manas or the mind is defined as that part of the antahkaraṇa[1] which is responsible for saṇkalpa and vikalpa, the word vikalpa stands for doubt.

  • @SonOfThought If panpsychism is true, then what makes human conscious experience more reliable than the limited bits of consciousness a cup of coffee has, if the difference between our consciousness and knowledge and that of a cup, is a matter of quantity not quality? If one does not know the answer, then it follows that introspection, at best, is unknown to have any authenticity at all, and at worst, not reliable.
    – SmootQ
    Commented Jan 28, 2019 at 11:38
  • Comments are not for extended discussion; this conversation has been moved to chat.
    – Geoffrey Thomas
    Commented Jan 28, 2019 at 14:48
  • 1
    @SmootQ:Thank you for sharing the latest trends of materialists. See, the last part of your comment "...matter is even more mysterious, we only know the forms matter can take, but we can, and will, never know what matter actually is." ~ At last you reached your main question about 'doubt'. In my answer I had mentioned -- "O Truth/God, who are you?" This is also a doubt. You may alter this as "O matter, what are 'you'?" To this, you won't get the answer until the unification with it (or what you call matter). It will be nothing, but consciousness. Commented Jan 29, 2019 at 15:02
  • Again, when you reach the minutest/subtlest, you must reach an unnamed somethingness; not a nothingness. But that somethingness, if it is a matter in a materialist's point of view, it must have a property (or properties). For the a materialist, let's assume it has no property at all; but at the same time it must have property as well. Otherwise a materialist couldn't perceive this world. So it must be as Sri Ramakrishna mentioned (as a wonder). Commented Jan 29, 2019 at 15:04
  • "...ultimate reality transcends both consciousness and the cosmos, where consciousness and the cosmos can be reduced to something that is not objective and not subjective,..." ~ What if that consciousness could make all the knowledge perceived with our senses (including the cosmos as perceived with your eyes) made meaningless? Then, would it be as mentioned by Sri Ramakrishna? SmootQ, Thanks for your great responses. Thanks to you also, Peter. Commented Jan 29, 2019 at 15:07

Wittgenstein was trying to limit skepticism

From the times of René Descartes (and possibly even before that) and his Meditations on First Philosophy, we know that is possible to doubt almost every knowledge, and certainly those made empirically. If we define doubt as probability that some statement may not be true, this probability would always be greater then 0 for statements based on perception . For example, Sun rises in the East . But, is there a Sun and is there an East ? Therefore, there is a reason for doubt.

Wittgenstein and contemporary philosophers were aware of that, but they were also aware that philosophy that rejects empirical knowledge remains barren and fruitless, because modern science and life in general are based on empirical data. Therefore, lot of his work was about conditional knowledge: if we take that Sun rises in the East as empirical fact, how does it fit into our complete system of knowledge? What this sentence really means in the language it is spoken ?

In the context of Wittgenstein's solution to the problem of knowledge (which is not perfect, but it is "useful" for the purpose of qualifying knowledge) doubt is necessary ingredient of it. To use our example with the Sun, if we decided to believe our perception (true justified belief ) , there is little doubt that Sun indeed rises in the East. What doubt remains is purely based on skepticism that rejects our perception. For Wittgenstein such doubt is insignificant. But doubt that all other suns on all other planets therefore must rise in the East (hasty generalization) would be considered valid. Amount of doubt that accompanies the knowledge would therefore serve as indicator of quality for that knowledge.

  • I agree with Wittgenstein, suppose that we have a statement A that we do NOT doubt, which, according to your definition : probability of the statement being false is 0. As the tradition goes, we only 'know' something by knowing its opposite. To actually know 'A' is to reach the understanding of 'not-A'. So, ,according to your definition : if there is no doubt, then not-A is ungraspable in essence. And if we do not know what not-A might be, then it would follow that we do not know A itself. Which means that doubt, or the possibility of not-A, maybe is necessary for knowledge.
    – SmootQ
    Commented Jan 30, 2019 at 11:31
  • Thank you so much for your answer +1
    – SmootQ
    Commented Jan 30, 2019 at 11:32
  • @SmootQ Well, most of the statements that leave no doubt are tautological, and therefore almost useless as knowledge. You are right that if we cannot even grasp opposite statement, then original statement is all encompassing for us, therefore not really understood.
    – rs.29
    Commented Jan 31, 2019 at 6:56

A comparative etymological attempt to answer the question where the word ‘doubt’ came from.

If we assume ‘doubt’ is a compound of ‘δύο πόδια’ (two feet) you can underSTAND (see the standing metaphore for righteous straight upright thinking) that it gives an image of standing on two feet regarding two options for the truth. Doubt could etymologically be defined as being ‘on two feet’ or considering two (or more) foundations for the truth and expecting one of them to be true. Left is what os left and right is correct but they can also be interpreted as the left and right sides.

  • Thank you friend, but the question is not about the etymology of the word 'doubt', it is about the meaning behind the word, in other words : "what propositions like I doubt that X" actually mean in terms of Justifications, Belief and Truth, be it in English, or Arabic or French or Chinese, it doesn't matter.
    – SmootQ
    Commented Jan 31, 2019 at 20:14
  • A choice between justifications, beliefs and truths. That is at least the intention of my answer. To show ‘doubt’ defines itself by ‘choices creating uncertainty about a single X’.
    – Ajagar
    Commented Feb 1, 2019 at 7:11
  • Thanks, and how do you define 'Standing on two options' , in terms of Justification and Belief? how can you further define this word 'standing', is it non-justification AND non-belief or justification and non-belief or something else ? if you look in the tags I added , it is analytic philosophy. Thanks again
    – SmootQ
    Commented Feb 1, 2019 at 8:41
  • By the way, the statement 'two options of truth' is a tautology of the form (P v ~P) . Which means that Truth cannot contribute to the definition of doubt, because if D is the definition of doubt, and (P v ~P) is your tautology where P and ~P are two options for truth, then we can rule it out of the definition, so D & (P v ~P) is equivalent to D. So if P is an option, then the statement P or not-P cannot be a part of the definition, because it is already a tautology, and does not contribute anything.
    – SmootQ
    Commented Feb 1, 2019 at 8:46
  • So if you say : Doubt is standing on two options for truth, then the definition of doubt lies in the word 'standing' , not in the word 'truth' or 'two option', because if you list the two options like this : P or not-P, then this does not add anything, which means that Truth is probably not a part of the definition of 'standing' or 'doubt'.
    – SmootQ
    Commented Feb 1, 2019 at 8:49

Before: doubt is related to truth. Not the absolute truth or a philosophical truth, but personal truth. A doubt exist when a concept can be related either with truth of falsehood. For example, my son has my keys can be true since he usually hides your keys, but can be false since he says he is not hiding them.

So, doubt is the inability to relate a concept with truth or falsehood.

You must log in to answer this question.

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