Forking Hume

Today I find myself wrestling with Hume's fork - a wonderfully cogent skeptical argument that suggests there can be no certainty in our statements about the world. Hume dichotomizes all propositions into two types; first there are statements about ideas, which are analytic in nature, and knowable through logic and deduction. Kant would later dub these statements (such as "1+1 = 2" or "all men are mortal") a priori, or knowledge that can be known in absolute certainty without empirical (worldly) experience.

The second kinds of statements are statements about the world, such as "the sun rises in the east" or "gravity makes rocks fall to the earth." These types of statements are synthetic, a posteriori propositions. A posteriori knowledge comes from empirical data - as we watch the sun continuously rise in the east, or rocks fall to the ground, we come to see the truth in these statements to the extent which they square with our future expectations of the world.

Future expectations, in this case, are the key to understanding why Hume was such a skeptic. In order for us to claim that "every rock dropped will hit the earth" we, of course, need the premise "every rock I've ever dropped has hit the earth." We also need another premise, one Hume is not willing to grant us, and rightfully so.

Hume argues that, implicit in all synthetic a posteriori claims, is a premise that the future always resembles the past. Ironically, we come to realize empirically that this isn't true, and therefore if I can't say with any certainty that the future will resemble the past, I have no good reason for expecting the rock to fall to the ground the next time I drop it. If you were to rebuff me and say "Gravity will always pull the rocks down", I could just as easily say to you "and how do you know that gravity will act the same in the future as it has in the past."

Ultimately, Hume forces us to realize that intrinsic to any of our expectations that we claim to know about the world is a degree of hopeless faith which ostensibly says "I believe that I know X because I have faith that all previous occurrences of X will resemble all future occurrences of X." How ridiculous, in this sense, it is to say that you know that the rock in your hand, when dropped, will fall to the ground, because you believe that this rock, and all rocks that have been dropped and will be dropped, act the same? Surely this is not now we want to go about claiming to "know" things...

So I'm left with a logical argument that makes perfect sense but with consequences that make me very unhappy. According to Hume, statements about ideas (a priori), may be certain, though they cannot truly pertain to the world. This is because we cannot be completely certain about anything in the world, and hence, if we are certain that "a triangle has three sides", we can also be certain that this tells us nothing about the world.

On the flip side, if we cannot be certain of statements about the world; we cannot use our ideas and analytics in matters of fact. Abstract relationships, therefore, cannot be consistently used to express or evaluate material objects/process. What does this mean for physics and applied math? Does this invalidate the very strength of empirical science?

Or does this, once and for all, settle the argument that philosophers are truly useless to the world, just as much as the world is useless to philosophers? I'll leave the conclusions up to you... suffice to say, Hume forked us good.