If I dream that Eli can fly, then (usually, and in at least one sense) in the dream I believe that Eli can fly. The question is: is the Eli about whom I have this dream-belief the Eli of the actual world (who exists and who does not fly), or is it of some other thing we might describe as a flying Eli (who does not exist in waking life, and who does fly)? I think that my dream is of both Eli and a flying Eli - they are, in the dream, one thing under two different descriptions. In the dream "Eli" and "flying Eli" co-refer, but in waking life they do not, because "flying Eli" doesn't refer at all. So to say that the belief is about the thing I am dreaming about does not cut any ice.
At any rate, if the object of the dream-belief "Eli can fly" is the Eli bearing all the properties he does in the actual world, then the belief is clearly false.
Also, I can be mistaken in my dreams in this way. While I am dreaming, I believe that I am encountering Murray. Now suppose that I awake to realize that I believed in the dream (as, we might suppose, I sometimes believe in waking life) that "Murray" refers to Merv, not Murray. In that case, my special dream-beliefs about Murray turn out to be false. My first-personal self was wrong.
Was my third-personal self wrong? Was my narrative subconscious, to speak floridly, falsely scripting or falsely interpreting the script of the dream so that Murray (the actual referent of "Murray") did such-and-such? No. There is no question whether the third-personal self is right or wrong. The third-personal self doesn't literally write a script, read it off, and create a visual depiction of how it interprets the text. There is no text, just dreamt experience. And if there is no text to be true or false, there is nothing for the dreamt experience to "truly" represent (or, at any rate, nothing "written" by the third-personal self). We might still be tempted to say that the experience itself is right or wrong, "veridical" or "falsidical". Now, I have always understood "veridical" when predicated of "experience", in something like the following way: "an experience is veridical iff, given a subject's cognitive make-up, the subject acquires only true beliefs from the experience" (there are many good objections to be made here, but this gets the gist of things). Are the dreamt experiences veridical? This, again, hinges on questions like the one I had above about the content of dream-beliefs. But those questions aside, I would point out in favor of dreamt experiences being falsidical that on waking from dreams, I often persist in believing that all of my dream-beliefs are true of the actual world, and it can take a minute or two to convince myself otherwise.
One way or another, I think it is false that the third-personal self is the "ultimate arbiter of what's true in the dream". The third-personal self does not say anything, and so does not issue any arbitrations one way or another. I'd say that the ultimate arbiters are at least two: (1) the phenomenal beliefs of the first-person self and (2) the authorial intuitions of the first-person self. The phenomenal beliefs of the first-person self are all true, at least if we constrict "phenomenal" appropriately. The authorial intuitions are true, given the qualifier below. Still, there remains the question what determines the non-phenomenal, non-first-person-legislated facts in the dream. I'd say that as long as the dreamt events are not events in which the first-person subject is being deceived, then whatever the subject believes is happening in the dream is actually happening in the dream. If it turns out that we can have unattended experiences in our dreams, then there will be further facts internal to the dream that I have not accounted for. Otherwise, I think I've covered all the bases.
Alternatively, non-first-person-legislated facts in the dream might all be phenomenal, in the sense of "phenomenal" in which our phenomenal judgments are indefeasible.
The qualifier about our authorial intuitions: I can believe in my dream that Bill is a cat, and realize on waking that I meant Gary, not Bill. We will have to qualify the indefeasibility of authorial intuition in some way such as
"if I believe in a dream that 'a' refers to b, then if I authorially intuit that F(a), it must be true in the dream that F(b)",
where "a" and "b" might co-refer, and F is an arbitrary property.
Saturday, September 02, 2006
From the Mailbag: Dream-Beliefs
Ian Olasov wrote in with an interesting response to my old post 'Truth in Dreams':