Monday, February 8, 2010

"p and I don't believe that p"

A number of folks seem to think that there is some innate "pragmatic contradiction" in assertions of the form: "p and I don't believe that p". Certainly, whenever I've heard these Moorean sentences mentioned, the mentioner assumed this. Yet, there are counterexamples to the "pragmatic contradiction" thesis. And this fact seems to be pretty well-known to people in the relevant field. I mentioned that I had some counterexamples to an ethicist and he found it surprising and interesting. But I then mentioned it to some epistemologists, and they were quite unimpressed. So, here, we have a case where inter-area communication in philosophy has failed: the people in the relevant area know that a thesis is false, while folks in other areas act as if the thesis were uncontroversially true.

For what it's worth, here are some of my counterexamples to the thesis. These counterexamples provide cases where one quite sincerely and unproblematically utters an instance of "p and I don't believe that p". Nobody I've met finds all the examples compelling. In all the examples below, "p" is a sentence and the quotation marks are meant to be right-angle quotes so one can substitute within them.

1. An expert tells me "p" and adds that ordinary people like me don't believe that p. But "p" is a sentence so replete with technical vocabulary that not only do I not know what all the words mean, I cannot even parse its grammar. I sincerely tell someone else: "p and I don't believe that p". In this case, I believe that "p" is true, but I don't believe that p. There are two responses I hear to cases like this. Some people say that the distinction between believing that p and believing that "p" is true is specious, and hence the sentence embodies a pragmatic contradiction. These people have a very low bar for what counts as belief and assertion. They will have to accept the next counterexample. Others say that the sentence is not an assertion if I don't understand it. I worry that this sets the bar for assertions too high. These folks may reject the next example for the same reason, but some of the others might still work for them.

2. An expert tells me: "p and you don't believe that p. Work out the consequences for yourself." I'm not very good at logic, so I have to do this step by step. I thus say: "p and I don't believe that p. By conjunction elimination, p. Hey, that's cool! I didn't know that p, and now I do."

3. Suppose I believe that one has no beliefs when one is in the afterlife, because the afterlife is an undifferentiated beliefless mist of joy. I write you a letter to be opened after my death. In the letter I say: "p. And I don't believe that p. I don't believe it, because right now I am an undifferentiated beliefless mist of joy. Therefore: p and I don't believe that p."

4. I write a paper. I think everything in the paper is true. I present the paper at a conference. When I present it, I am really tired. I am reading the sentences outloud, and sincerely, but I cannot parse all of them, nor do I believe their content. Some of the sentences are intermediate steps in the argument, and I've completely forgotten them (or I never believed them in the first place, though I believed them to be true; sometimes, I write down things in the course of a proof by copying and pasting an existing sentence and transforming it by rules of inference—that's just a matter of syntactic manipulation—without bothering to figure out what the new sentence means). But I still believe that whatever I am saying is true. One of the sentences is "p". So I read the sentence: "p". I then add, surprised at myself: "You know: p and I don't believe that p. I don't believe it because it's too complex to parse, and I remember that this is one of those steps that I've forgotten completely."

5. I program a robot to bring you a drink whenever I say to you: "I don't believe that the robot will bring you the drink, and the robot will bring you the drink" (this example works better with the Moore sentence re-ordered in this way). Now, you keep on interrupting me, never letting me say a whole sentence. I say the sentence sincerely; I don't believe that I will finish the sentence, and hence I don't believe the robot will bring you the drink, but the sentence is so constructed that if I do manage to say it, it'll be true, and that's all that sincerity requires.

6. I'm deaf and have been learning out how to vocalize. I do not believe I can do so yet. I know you're standing somewhere where you can't see my lips and I can't see your reactions. I say to you: "I can speak and I don't believe I can speak." I can say this sincerely, because although it's true that I don't believe I can speak, I also know that if I do succeed in saying it, it is true.

7. I have a mental inertia on which once I form the intention to do an action such as saying a sentence, often I am unable to stop even if I change my mind prior to beginning the action. (Unlike in #6, this is actually the case for me.) So, while the sentence is proceeding from my mouth, it need no longer be true that I have any intention of saying it—though I had to have had that intention. Suppose that I know that as soon as I fully form the intention to speak, Fred will (e.g., by neural manipulation) bring it about that I do not believe that p. So, right now I believe that p, but I know that at the time of utterance I won't. I say: "p but I do not believe that p." This example rests on taking the present tense in a sentence to refer to the time of utterance, not the time of deliberating whether to speak. I think this is correct: think of sentences like "It's 12 o'clock", which you utter while watching the clock—you time yourself to begin to speak so that the clock strikes 12 while you're speaking (or maybe at the very end; Richard Gale has an argument that "now" refers to the time at the end of a sentence, by reference to sports announcers who say things like "He's got the ball, no Jones has now taken it from him, but, wait, no, now he's got it back!")

12 comments:

Heath White said...

I agree that most of these examples work. I would say, however, that this doesn't so much refute the idea of "pragmatic contradiction" as show that we can embrace such things.

Andrew M. Bailey said...

Your third example is similar to this one by John Turri (in the latest issue of _Analysis_): An eliminativist about beliefs asserting some thesis while also denying that she believes it (and that anyone believes anything).

Alexander R Pruss said...

Thanks, Andrew, for that reference. I find Turri's example a bit less convincing than mine, because the following answer is available: if the only counterexample to "p and I don't believe that p" is an eliminativist, then that doesn't show that "p and I don't believe that p" isn't absurd, since eliminativism is thought by many to be absurd.

Martin Cooke said...

Another example might be the Preface paradox? One asserts the conjunction of the propositions in the book, but one also says (in the preface) that one does not believe that conjunction (since one has probably made a mistake somewhere). And that sort of thing could occur with an ordinary proposition, since most of them are composed of simpler assertions. For a simplistic example, suppose a credence of 70% was enough for one to accept as fact, and hence to assert, a scientific proposition, and that at least a 50% credence was required for one to believe any proposition; and suppose that one accepted the axiom of infinity, with a 70% credence, and similarly that current climate change was the result of human activities. (The evidence for one of those facts is largely independent of the evidence for the other, and 70% of 70% is 49%.) Then one could honestly say: "The axiom of infinity is true and mankind is causing the climate to change, but I don't believe that the axiom of infinity is true and mankind is causing the climate to change."

Alexander R Pruss said...

I think the preface paradox is a bit different as it involves more than one assertion.

Alexander R Pruss said...

enigMan: See my post on asserting conjunctions. :-) Though I grant you that your argument is rather nice.

Martin Cooke said...

Thanks, and I agree that the Preface paradox is of a completely different kind. But I think that that is because the point of Moore's paradox is to show that an apparent contradiction is not the same as a contradiction, whereas the Preface is prima facie OK. Incidentally, I wonder if the fact that the intended resolution of Moore's paradox is pragmatic accounts for why your epistemologists were unimpressed by your counter-examples?

Martin Cooke said...

Re your recent post, I agree that asserting a conjunction is not the same as asserting the conjuncts, but since 'p' signifies a sentence I don't think that that fact stops the Preface paradox being another counter-example, but is rather why it is. To be so it relies on "x and y" being equivocal between a conjunction and a short sequence?

Martin Cooke said...

...in other, clearer (hopefully) words, I think that our examples are not so much counterexamples to the pragmatic resolution of Moore's paradox, as examples of what would not give us Moore's paradox, i.e. that Moore's paradox is "p and I don't believe that p" for only some (e.g. most ordinary) p, and that the thesis you refute gets its plausibility from the common use of "p and I don't believe that p" in Moore's paradox.

Alexander R Pruss said...

Note, though, that 2, 3, 4 and 7 can be made to work for any p.

It's pretty rare that you'd be speaking sincerely without believing what you're saying. That's because it's pretty rare that (a) what and whether you speak affects whether what you say is true (cases 5 and 6) or that (b) you expect to change your mind between the forming of the intention to speak and the speech (cases 3 and 7) or that (c) you speak with significantly diminished understanding (cases 1, 2 and 4) and yet self-insight.

Martin Cooke said...

Yeah, I should've said "for only some (e.g. most ordinary) p and contexts".

Alexander R Pruss said...

I've just had a paper using examples 2, 5 and a variant of 6 accepted by the Australasian Journal.