A Problem With The Lockean Thesis (Part II)
[I planned to write only two blog posts on the Lockean Thesis, but now, it seems that there'll be a Part III. No more after that, or I might as well write a paper. In this post, I consider, and then reject an attempt to resolve the problem I raised for the Lockean Thesis in my previous post. In my next post, instead of trying to resolve the problem, as I said I would, I'll try to locate the source of our intuitions for thinking that (4) to (6) sound better than (1) to (3).]
It seems that I can be rational in believing that Mark Twain is Samuel Clemens with less than complete confidence. But suppose I’m offered a bet that will win me a dollar if Mark Twain is Samuel Clemens, and condemn me to eternal torture if not. I’ll decline the bet. Question: in doing so, do I still believe that Mark Twain is Samuel Clemens?
I’m inclined to say `No’. Some philosophers suggest that the threshold that one’s degree of belief in p has to meet for one to have a belief that p varies with context. (Foley (1993) does not make such a suggestion.) This seems plausible to me, and if it’s true, we’re allowed to say that in some context in which I’m not considering the bet, I believe that Mark Twain is Samuel Clemens, but in a context in which I’m, I don’t (although I might believe there’s a high chance that Mark Twain is Samuel Clemens, or something like that). In both contexts, my degree of belief in `Mark Twain is Samuel Clemens’ is the same, although the threshold for belief is higher in the second context than in the first.
Can shifting thresholds help with the problem I raised for the Lockean Thesis? Consider a context in which I believe that Mark Twain is Samuel Clemens. If, in the context, one is rational in having such a belief just in case one is rational in being sufficiently confident (say, at least 90% confident) that Mark Twain is Samuel Clemens, then both statements below should sound equally felicitous (or infelicitous):
(2) I’m slightly more confident that Mark Twain wrote Huckleberry Finn than I’m that Samuel Clemens wrote it, and I believe that Mark Twain is Samuel Clemens.
(5) I’m slightly more confident that Mark Twain wrote Huckleberry Finn than I’m that Samuel Clemens wrote it, and I’m 95% confident that Mark Twain is Samuel Clemens.
But they don’t. (2) sounds worse than (5). How do we explain this?
Here’s a possible explanation. It might well be rational to believe that Mark Twain is Samuel Clemens, while being slightly more confident that Mark Twain wrote Huckleberry Finn than that Samuel Clemens wrote it. It’s just that when one utters (2), and considers that one is more confident that Mark Twain wrote Huckleberry Finn than that Samuel Clemens wrote it, the threshold for belief changes (goes up to 1, in fact). In considering such a thing, one does not believe that Mark Twain is Samuel Clemens. This explains why (2) sounds infelicitous, but its sounding infelicitous does not count against the Lockean Thesis.
There’s a question why merely considering that one is more confident that Mark Twain wrote Huckleberry Finn than that Samuel Clemens wrote it would raise the threshold for believing that Mark Twain is Samuel Clemens. (Would it?) But in any case, the explanation above doesn’t seem to work in general, for the problem remains even if we change (2) and (5) to read:
(2′) Jack is slightly more confident that Mark Twain wrote Huckleberry Finn than he is that Samuel Clemens wrote it, and he believes that Mark Twain is Samuel Clemens.
(5′) Jack is slightly more confident that Mark Twain wrote Huckleberry Finn than he is that Samuel Clemens wrote it, and he’s 95% confident that Mark Twain is Samuel Clemens.
(2′) sounds worse than (5′), and the infelicity of (2′) does not stem from Jack’s considering that he’s slightly more confident that Mark Twain wrote Huckleberry Finn than he is that Samuel Clemens wrote it. We can imagine that he’s considering no such thing.