Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Wittgenstein and doubt

  1. Jan 5, 2010 #1
    Wittgenstein often made the statement that if you cannot doubt that a thing is so, you cannot know it. For example ( the example he gives) , you cannot know that you are in pain because you cannot doubt it. You cannot say," This pain hurts but I doubt that I feel it."
    I can understand how when you are in pain you cannot doubt that you are in pain. However, I do not see the leap from that and saying that you cannot know that you are in pain. Why cannot you know something that you cannot doubt? I have read a lot of Wittgenstein and have never found an answer to this. It seems like he just takes it as a given.
    I thought that maybe it has something to do with the idea that if a statement refers to everything and nothing in particular it is a meaningless statement. Unfortunately, I have failed to make it connect to the idea that you cannot know something that you cannot doubt.
  2. jcsd
  3. Jan 6, 2010 #2
    I think he's using 'know' in a very specific sense. Knowledge is constructed, it has a structure, with dependencies. Sensation does not, sensation is what it is, you can't doubt it because it has no logical, or quantitative form. He talks about his in 'On Certainty'.

    Doubt is part of the structure of knowledge, part of what gives the meaning. You can't understand pain, you can only feel it.
  4. Jan 7, 2010 #3
    You can still experience pain through subtle sensations you are not consciously aware of like internal bleeding right?
  5. Jan 7, 2010 #4
    Not sure what you mean.
  6. Jan 7, 2010 #5
    Frankly the statement makes no sense to me when you're speaking of sensations like pain. The subject is the only one who can know if he or she is in pain. However, if you're talking about tautologies, I would agree. "Either it is raining or it is not raining". You cannot doubt this is true but it conveys no knowledge.

    EDIT: I think we need to define "structured knowledge". Pain has location, intensity, pattern, duration, etc. Eliciting this structure from the subject is invaluable in diagnosis and treatment.
    Last edited: Jan 8, 2010
  7. Jan 8, 2010 #6
    From: On Certainty

    41. "I know where I am feeling pain", "I know that I feel it here" is as wrong as "I know that I am in pain". But "I know where you touched my arm" is right.

    What is the difference?

    You can add 'because' to the end of the last sentence and justify what you say, because it has dependencies. The first three do not.
  8. Jan 8, 2010 #7
    Interesting philosophy. Wittgenstein (W) was very influential. I like his rigorous approach to philosophy. However, he is not infallible. Assuming the pain has an identifiable physical cause, it represents a signal from the site of the dysfunction to the brain just like any sense data, whether it originates from the eye, ear, or internal sensory nerves specific for pain.

    There is a dependency. A blocked artery in the heart will cause tissue death and excite pain receptors and pathways. The pain of a heart attack (MI) is usually very characteristic in terms of location, pattern and qualitative description. One can justifiably say I have pain because.....

    Now the subject may not know the cause (dependency) but will know he has pain just as he knows what he sees or hears even though his eyes and ears. I saw a tree fall. Do I know the tree fell? I may not know the cause of the tree falling but I know the tree fell. Is W saying I can't know the tree fell because I don't know the cause? If that's the case we can't know anything based on sense data.
  9. Jan 9, 2010 #8
    Perhaps W simply meant you must have something to compare with, to judge you are in a certain state - of pain or something. And as soon as you are "comparing" you are "doubting" in some sense. I don't agree with that, but guess he must have meant something like that.

    An other interesting thesis from him is said to be "The problems disappear if you don't think of them". That often works also in practical life, is my experience. Although it must be unwise relying on it.... :rolleyes:
  10. Jan 9, 2010 #9
    Well, people can certainly know pain free states. If they were in pain all the time, it would be "normal" and adaption might occur. There would be nothing to know. But since they experience both, they have a basis of reference.They know they're in pain because they doubt they are pain free.

    It's strange to me that W picked this example. I think his basic idea is correct in that knowledge requires context. Also sense data can be misleading. The Cambridge Encyclopedia of Philosophy (Robert Audi ed, 1995) defines knowledge as a "justified true belief". However the quality of justification is not specified, and we mostly rely on heuristic methods based on context.

    For example, the subject gives a history of pain suggestive of MI. Subsequent tests confirm the diagnosis based on established criteria. This is the justification of MI. The subject might only know it was chest pain, but did go the emergency ward, so he believed it might be significant. The chest pain alone was the justification for a level of concern. He was able to describe the pain, guided by the physician's questions, which lead to the ordering of tests. The subject certainly knew he had pain because it occurred in context. Moreover, he could describe the pain in terms of important details it so as to differentiate from others similar states, giving it structure (and possibly saving his life).
    Last edited: Jan 9, 2010
  11. Jan 9, 2010 #10
    You don't feel a blocked artery, you feel pain.
    You're conflating the two, sensation/explanation, whereas he is being very careful to avoid doing that.
  12. Jan 9, 2010 #11
    As I've said, pain is a sensation. So is vision and hearing. It's true pain is not may not be as discriminating as vision and hearing but one knows they have pain which can be described. They know they have pain because it's contextual. That is they have the experience of being pain free and can doubt they are pain free. That's the way I interpret W's position. If he did mean that we must doubt pain we sense, then we must doubt all sense data. All a posteriori knowledge is based on sense data or inferred from sense data. You're not going to say that W denied all a posteriori knowledge, are you?

    If W was a physician (he did consider going to medical school), would he have told my subject with chest pain that he can't know he has chest pain and sent him home?

    EDIT: I'm not conflating a blocked artery with the sensation of chest pain. The latter is dependent on the former.
    Last edited: Jan 9, 2010
  13. Jan 10, 2010 #12
    No, he's saying pretty much the opposite.

    "I know where I am feeling pain",
    "I know that I feel it here"
    is as wrong as
    "I know that I am in pain".

    Its a similar distinction, I would say, to Kant's phenomenology. The thinginitself is unknowable. That doesn't mean we don't have knowledge of things.
    Wittgenstein is saying knowledge is not about sense experience.
    Knowledge is contextual, pain is not knowledge.
    With regards to doubt, he is stating that you cannot doubt a sensation. You either experience pain or not. Its not open to doubt. Sensation is not, therefore, knowledge, since according to W, knowledge requires doubt.
    I'm going to say, that W distingished between an empirical proposition, and a sensation.

    I am sitting in a chair, is an empirical proposition.
    I am cold, is a sensation.

    W can say, I know I am sitting in a chair.
    W cannot say, I know I am cold.

    There is no direct sense experience in the first.
    I could be sitting on a stool, or the floor.
    Doubt doesn't mean that something is not true.

    But if I'm cold, I can't doubt that I am. I'm either cold or not. So it doesn't involve doubt, and therefore I can't 'know' it.

    W is trying to distinguish between a common understanding of knowing and his very strict understanding of it.
  14. Jan 10, 2010 #13
    In a medical sense, yes the chest pain is a result of the blocked artery, but this is not what W is referring to.

    I can doubt your theory of medicine.
    I can't doubt the pain I'm having.
  15. Jan 10, 2010 #14
    The only sensible thing I can extract from this is that some states can be observed by others (objective) and others states can only be known to the subject (subjective). But in both cases, knowledge is mediated by sense data. How do you know you are sitting in a chair? Because of visual and proprioceptive sense data. How do you know you are cold? Because of sense data from the skin and muscles, How does someone else know you are sitting in a chair. Because of visual sense data. It's true that the other can't know you are cold (unless you're shivering) unless you tell him. But the other gets that information by hearing you.

    I've already said that only the subject can know his own sensations, but they do know them. You're confusing a binary variable with a tautology. In the former, the subject knows he is in one of two possible states. In the latter. if you say I'm either cold or not cold, that's a tautology which cannot be doubted.

    If the subject tells someone else that he is cold, that's subject to doubt by the other. Are you saying that the other knows the subject is cold because he has doubt, but subject himself can't know if he's cold because he has no doubt?

    EDIT: I believe there has to be a distinction between logical tautologies which cannot be doubted but convey no knowledge, and subjective knowledge which only the subject can know without doubt. If you doubt this, please address the apparent contradiction in the previous paragraph.
    Last edited: Jan 10, 2010
  16. Jan 10, 2010 #15
    He's not talking about states, he's talking about knowledge. You're confusing epistemology and ontology.

    W would say 'we have pain', but 'we don't know pain'.
    He sees knowing something as requiring 'doubt'.

    And no, when it comes to words like 'doubt' and 'know', he is not using dictionary definitions. If you haven't already, you should read 'On Certainty'. Otherwise we'll just be going round and round.
    Sigh. No, its not that simple.

    I know the sun will come up tomorrow. How? By inductive reasoning, not sense data.
    If I limited knowledge to sense data, I could never say I know the sun will come up tomorrow. I could only state that 'I know' what is right in front of me.

    And I couldn't say I know there are neurons inside my skull, unless I opened up my skull and looked. And if I sowed it back up, I couldn't say I know they are still there, because I can't see them any more, and they are no longer part of my sense data.

    So you are wrong, what we call 'knowledge' is not mediated by sense data. We use logic and reasoning as well.
    You don't, according to W, you either 'feel' cold or not.
    It has nothing to do with nerve cells. You don't feel nerve cells. You feel cold.
    Fine, but thats not the issue.
    W is saying: I cannot doubt I am in pain, when I am in pain.
    I cannot doubt the sensations that I have.

    The reason this is important, is because I can doubt other things.
    And part of knowing something is being able to doubt it.
    But that is not what W is talking about.
    He is saying: If I am in pain, then I can't doubt that I am in pain.
    And therefore I don't know pain, because knowing something requires that it can be doubted.
    It has nothing to do with other subject vs. object.
    There is no other here. This is about me.

    I feel pain.
    I know I am sitting at a desk.

    No others needed here, you are just complicating the issue.

    *I* can't doubt pain, because, it is part of raw experience.
    *I* can doubt my understanding of a situation, because that understanding is not a raw sensation.

    I can have sensations while sitting at a desk, and certain sensations lead me to believe that I am sitting at a desk, but 'sitting at a desk' is not a sensation. I can doubt 'sitting at a desk'.
  17. Jan 10, 2010 #16
    OK. We can end this now. W is not using the definition of knowledge that I quoted from the Cambridge Dictionary of Philosophy. It's not mainstream philosophy. The standard definition is a "justified true belief". Knowledge is a belief. You believe the sun will rise to tomorrow because it always has. You have observed it consistently. It may not one day due some cosmic catastrophe.

    http://absentpresent.blogspot.com/2008/12/wittgenstein-critique-of-certainty-and.html [Broken]

    Go back and read my last post If you tell someone you're cold, the other can reasonably doubt you are cold. You however cannot. You are saying the other has doubt and therefore can know you are cold, but you can't know you are cold because you don't doubt it. That's simply absurd. You cannot just eliminate the other person. If you were isolated from birth in an empty windowless room, fed by a caretaker who never spoke, how much would you know after ten years? W is elevating some sense data above others without justification.

    Last edited by a moderator: May 4, 2017
  18. Jan 10, 2010 #17
    This is a good example. Are you saying you can doubt sitting at a desk because you doubt your senses or because you could be doing something else at that moment in time? If its the latter, I agree with you because it's one of many possibilities. As long as these possibilities are exhausted and not linked to each other by either/or conjunctions, this is contextual knowledge. If it's the former, you are doubting the validity of all sense data.

    You can doubt the validity of a particular sense datum just as you can doubt the validity of a measurement. Observations, both scientific and common, can be wrong, but we must rely on them, coupled with reason, to have reliable knowledge.

    I don't understand the distinction between the belief that you are sitting at a desk and the statement: "I am sitting at a desk". Is this not a statement of your justified belief?
  19. Jan 10, 2010 #18
    W would probably say all sense data is 'valid'. What might not be valid is the value we give sense data, the way we arrange it.

    I can doubt the meaning I give sense data, but I can't doubt the sensation.
    I can doubt my concept of 'I', 'chair' and 'sitting', but I can't doubt the various sensations that I use to construct those concepts.

    There is nothing about 'hard' or 'soft' that equals 'something to sit on'. Knowledge is constructed, sensation is not.

    'I am sitting on a chair' relies on a huge number of interlinking concepts, all of which I can doubt to one degree or another. This is knowledge. Doubt does not equal denial. This is not an ontological argument about solipsism. Its about knowledge.
    Justification implies the need to justify, it implies that one could doubt.

    In On Certainty, W addresses the difference between belief and knowledge in some length.

    The important part with regards to doubt, is that all knowledge can be doubted. Doubt is an implicit part of knowledge.

    I can doubt my concept of self: that I am distinct from the world, mind is distinct from body
    I can doubt my conception of objects and how the interact: chair and sitting.
    But I can't doubt the sensations involved in sitting.

    That absolutely does not mean I am 'denying' that I am sitting in a chair.

    Compare Descartes 'certain knowledge' (with regards to Cogito Ergo Sum.)
    Wittgenstein would call that a contradiction in terms. Knowledge requires doubt.
  20. Jan 11, 2010 #19
    OK. If you want to say that justification implies doubt, then I have to agree with you and in fact we can doubt our senses. If you read the links I provided (post 16) W's positions in On Certainty are viewed as fairly radical and certainly a departure from mainstream philosophy.

    My understanding has always been that we can have direct knowledge through sense data, and indirect knowledge based on deduction, induction and authority.

    Direct Knowledge: The car is in the garage. Justification: I can see it and I can feel it.
    The car's horn works. Justification: I can hear it.
    The car is cold. Justification: I can feel it.

    Indirect knowledge is usually acquired by way of language which is mediated primarily through the visual and auditory sensory apparatus and requires the a priori ability to learn language. One can debate whether the ability to reason is a priori or learned a posteriori. Ir's probably bit of both.

    For example: They speak Swedish in Sweden. Justification: Authority, unless I've actually been in Sweden. If I have, I would still need to understand Swedish to confirm they are speaking Swedish. They all might in fact be speaking Norwegian.
    Last edited: Jan 11, 2010
  21. Jan 11, 2010 #20
    We can doubt what our senses 'tell us'.
    We can't doubt that we experience sensation.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook