Saturday, October 20, 2007

Why Philosophers're Crazy

Suppose I believe of some object, for good reasons (e.g. touching part of it), that it's a ball, and that I've also got good reasons (e.g. seeing part of it) for believing that it's red. So, I believe that it's red, and that it's a ball; in short, I believe that it's a red ball. But suppose that my reasons only justify a partial belief of just over 60% that it's a ball, and similarly just over 60% that it's red, so that they only justify a partial belief of less than 40% that it's a red ball (since 62% of 62% is roughly 38%). Then my belief that it's a red ball would seem to be unjustified. So, I might know that it's red, and know that it's a ball, and yet not know that it's a red ball, which is odd. And maybe, if I'm a rational thinker (requiring sufficient justifications for my beliefs), I might not even form the belief that it's a red ball, even as I believe that it's both red and a ball... But no, that would surely be too weird. Perhaps 60% is too low, for the degree of partial belief above which beliefs lie... But many of our ordinary beliefs involve many more than 2 elementary properties; so similarly, anything short of 99% is probably going to be too low too... But that seems unrealistic for, not certainty but belief, so maybe the problem is the forming of logical conjunctions? That is a necessary part of thinking; but maybe, if we want to be rational, we should think less. (PS: maybe this is just a way to make sense of Moore's paradox, e.g. if 70% credence is enough for me to assert a scientific proposition, but less than 50% credence is not enough for me to believe any proposition, and if I have a 70% credence in the axiom of infinity, and a 70% credence in nuclear deterrance, then I could honestly say "the axiom of infinity is true, and nuclear deterrance works, but I don't believe that the axiom of infinity is true and nuclear deterrance works":)


Tanasije Gjorgoski said...

Hi Enigman,

I don't think reasons can be multiplied though, because that would require for them to be independent of each other.

But say, often reasons to see color and form wrongly, would be related to each other. For example, if a reason is bad eyesight which covers both the color and the form, the percentage would still be 60%.

Of course, the whole quantifying of reasons sounds implausible to me.

Enigman said...

Yeah, I think that subjective probability is no kind of probability at all.