Yu Guo on Cartesian Skepticism

I just received a gracious email from Yu Guo, a reader of Certain Doubts who teaches at Communication University of China, and works on epistemology, philosophy of language, and Kant. Since we ordinarily think of analytic epistemology as being at home most in Anglo areas of the world, it is interesting and encouraging to see interest in it in other areas as well. I’m sure the possibilities of discussion with colleagues that many of us take for granted are not nearly so available in other geographic locations, so I’m posting here a quite interesting paper that Yu sent to me with the email. It’s about closure and the argument for skepticism typically called the argument from ignorance. The paper can be found here.

As always comments and discussion strongly encouraged.


Comments

Yu Guo on Cartesian Skepticism — 7 Comments

  1. It’s very nice of you, Jon, and thank you very much.
    I think I should briefly summarize my position here, so that the reader may get a better idea of what I am arguing for. Here’s the abstract: I argue that the reason we find the BIV argument (and Cartesian skeptical arguments in general) puzzling is because our intuitive construal of the argument equivocates between two different analyses of ‘S is not a BIV’, a disjunctive analysis (into ‘S is not a handless brain or S is not deceived into falsely believing that S has hands’) and a conjunctive analysis (into ‘S is not a handless brain and S is not deceived into falsely believing that S has hands’). Once disambiguated, the BIV argument turns out to be either uncontroversial and epistemologically uninteresting, in which case it serves the ends of neither the skeptic nor the Moorean nor the contextualist, or controversial indeed, but in which case it is no longer based on the epistemic closure principle.

    (I want to apologize in advance for any lingusitic flaws in my wording. Also, I was just told this morning by an English Prof. that it’s considered rude to make rampant use of phrases like ‘Note that..’ and ‘Recall that..’ in a paper written in English. I must apologize if that sounds impolite and I assure the reader that I didn’t mean to.)

  2. I am, of course, not the best person to have strong opinions about this (not being a native English speaker), but isn’t what the English Prof. told Yu Guo false?

  3. I was puzzled by this as well Juan. It’s certainly not considered rude where I’m from, and that’s England! It sounds to me Yu Guo that you’ve a better grasp of the English language than your English Prof.

  4. Juan, Duncan and Jon, thank you all for the very kind and instructive comments. (By the way, English Profs here have been quite famous for sometimes issuing verdicts about English usage that strike native speakers as counterintuitive. There is an interesting post at Language Log:
    http://itre.cis.upenn.edu/~myl/languagelog/archives/003528.html)

    I think I should give a summary here of the main claims of my paper, so as to make my position clearer (though the specific arguments for these claims would have to be found in the paper itself). So, here is how it goes:

    Some symbols:
    B = ‘John is a BIV’.
    h = ‘John has hands’
    b = ‘John is a handless brain’
    ~d = ‘John is not undetectably deceived into falsely believing that h’
    Kp = ‘John knows that p’

    The BIV argument is the following:
    ~K~B
    ~K~B -> ~Kh
    Therefore, ~Kh

    It depends on the following closure:
    (1) (Kh & K(h -> ~B)) -> K(~B)

    I claim that we intuitively analyze (1) into:
    (2) (Kh & K(h -> (~b V ~d))) -> K(~b & ~d)

    I support this claim with two arguments.
    First, it explains two things:
    (a) why we are willing to grant that h -> ~B ( Explanation: because (h -> ~b) -> (h -> (~b V ~d)) ) (I cite Pritchard and Feldman as evidence that we do reason this way);
    (b) why we find K(~B) implausible ( Explanation: because we find K~d implausible )
    Second, it explains why we consider only the spatial dimension of a BIV (i.e. not having hands) when we judge
    (c ) John knows that Mary is not a BIV,
    to be true, but consider also the epistemic dimension (being something of which one cannot know oneself not to be a instance) of a BIV when we judge
    (d) John knows that John is not a BIV
    to be false.
    (Very roughly, the explanation is that we intuitively analyze ‘Mary is not a BIV’ as it occurs in (c ) into a disjunction, but analyze ‘John is not a BIV’ as it occurs in (d) into a conjunction. For details, please see section II and III of my paper.)

    (Besides arguing for the truth of the claim that we intuitively analyze (1) into (2), I will also offer an explanation of why this claim is true. This explanation is found in section III of my paper.)

    But (2) is not an case of closure, because ~B equivocates between (~b V ~d) and (~b & ~d).

    There are two options for turning (2) into a case of closure.
    The first option is to turn (2) into:
    (3) (Kh & K(h -> (~b V ~d))) -> K(~b V ~d).
    The second option is to turn (2) into:
    (4) (Kh & K(h -> (~b & ~d))) -> K(~b & ~d).

    But both options face problems. Very briefly, the problem with (3) is that it is not controversial, in the sense that it is just like any other ordinary closure that we would intuitively endorse (such as ‘If John knows that he has hands, and knows that if he has hands then he is not a handless ball, then John knows that he is not a handless ball’). Thus (3) cannot be exploited by the skeptic.

    The problem with (4) is that ~d itself stands in need of analysis, and again there are two options for analyzing ~d. The first option turns (4) into a uncontroversial closure, hence unserviceable for the skeptic. The second option turns (4) into a conditional that is not a case of closure at all. We have no reason to be bothered by a skeptic who appeals to a pseudo-closure. (For details on this, please see section II of my paper)

    The above is a brief summary of my main claims in the paper. I am very, very much looking forward to criticisms and comments from you all 🙂

  5. (Here’s an additional argument for my position. In what follows I show how the law of non-contradiction is apparently threatened by Cartesian skepticism. And I offer a explanation of this threat using the notion of conversational implicature)

    Consider the disjunctive closure (DC):
    [Kp & K(p -> (p V q)] -> K(p V q).

    DC cannot be exploited by the skeptic, because a subject’s epistemic position (intuitively) with respect to p necessarily equals her epistemic position (intuitively) with respect to (p V q). (By contrast, in the skeptic’s closure, the subject’s epistemic position (intuitively) with respect to ‘I have hands’ does not equal her position (intuitively) with respect to the known-to-be-entailed ‘I am not a BIV’.)

    Further, if our intuition judge ‘John knows that John has hands’ to be true, then our intuition MUST, on pain of irrationality, judge whatever sentence of the form
    ‘John knows that: John has hands or p’
    to be true, whatever that p may be. (I’m setting aside irrelevant complications of the closure principle)

    It follows that we MUST have the intuition that, if John knows that John has hands, then he must know the following sentence to be true
    ‘John has hands or 3+2=1’;
    ‘John has hands or Betrand Russell is a Martian’;
    ‘John has hands or he is not Betrand Russell’;
    (1) ‘John has hands or he is not undetectably deceived into falsely believing that John has hands’.

    Let’s see what (1) amounts to. Assuming that ‘John has hands’ is equivalent to ‘He is not handless’, (1) amounts to:
    (2) ‘John is not handless or John is not undetectably deceived into falsely believing that John has hands’.

    This is equivalent to
    (3) ‘It is not the case that (John is handless and John is undetectably deceived into falsely believing that John has hands)’

    But we have the strong intuition that John does not know (3) to be true. So here’s a very paradoxical situation indeed: (1) and (3) express the same proposition, and yet our intuition says John knows (1) but not (3). So it seems not only the closure principle, but also the law of non-contradiction is at risk. Are we to say that John knows that p and does not know that p?

    (Of course there is some delicate issues about sentences and propositions here, but I think the assumptions I’ve made is harmless)

    My way to preserve the law of non-contradiction is to deny the legitimacy of our intuition with regard to (3). Although (1) and (3) express the same proposition logically, we intuitively but wrongly construe (3) to be the following proposition:

    (4) It is not the case that John is handless and It is not the case that John is undetectably deceived into falsely believing that John has hands.

    The explanation I offered in my paper is semantical proximity between (3) and (4). In fact, there can be a further explanation, in terms of conversational implicature. Consider the following dialogue:
    Dad: What knowledge have you gained today?
    Son: Well, I now know that a lot of things you taught me is wrong! For example, I now know that it’s simply false that Napoleon is a girl and the earth is flat and 6+2 equals 9……..

    Now, the son’s self-knowledge-attribution is logically true provided only that he knows that Napoleon is not a girl. But the knowledge attribution conversationally implicates that he knows the conjunction that Napoleon is not a girl and the earth is not flat and 6+2 does not equal 9.

    Hence the following principle:
    K~(p & q & r) conversationally implies K(~p & ~q & ~r).

    And this is why we intuitively deem (3) to be false: because (3) implicates (4), and we intuitively deem (4) false.

    Criticisms and comments strongly welcomed.

  6. A: Do you know that you have hands?
    B: Of course.
    A: Now, for any q, do you know that you have hands or q?
    B: Yes.
    A: Is that what your intuition tells you?
    B: Sure.
    A: So, do you know that you have hands or 3+2=9?
    B: Yes I know.
    A: Do you know that you have hands or up to 5% carbon dioxide in medicine is added to pure oxygen for stimulation of breathing after apnea and to stabilize the O2/CO2 balance in blood?
    B: Yes I know. See, I’m not really interested in whatever follows the word ‘or’. If I know p, I know p or q, whatever that q is.
    A: So, do you know that you have hands or you are not undetectably deceived into falsely believing that you have hands?
    B: Of course I know, given what I just said!
    A: Is that what your intuition tells you?
    B: Yes!!
    A: Now, let me ask you this question: Do you know that you are not handless or you are not undetectably deceived into falsely believing that you have hands?
    B: Surely I know, because ‘you are not handless’ is the same thing as ‘you have hands’!
    A: Do you agree that (~a V ~b) is logically equivalent to ~(a & b)?
    B: Yes, that’s a logical truth.
    A: Now, answer this: You think you have hands, but do you really know that it is not the case that you are in fact handless and you are just undetectably deceived into falsely believing that you have hands?
    B: Well.. I guess I can’t know that.
    A: Is that what your intuition tells you?
    B: Yes. My intuition says that I cannot know myself not to be the victim of an undetectable deception.
    A: So is your intuition saying that you know (~a V ~b), but do not know ~(a & b)?
    B: ……

Leave a Reply

Your email address will not be published. Required fields are marked *