300 (Years of Hume)


From Philosophy Now:

Like most philosophers of his time, Hume conceived of thought as a flow of mental images. Seeing a tree, imagining a tree, or remembering a tree, were all thought to consist of our having a mental image, more vivid for the seen tree, less vivid for the imagined or remembered tree. A sentence like ‘The Earth is round’ would have a certain type of mental image as its meaning, and believing that the Earth is round necessarily involved a vivid mental image of that type. This theory also explained why certain beliefs were logically impossible. For example, a four-sided triangle was logically impossible (and a three-sided triangle logically necessary) because we could not form a mental image of a triangle that did not have three sides. (Try it.) Hume’s disturbing insight from this way of thinking about thinking, was that all our factual and moral beliefs can therefore only be justified in terms of the psychological laws that govern the succession of images in our minds.

Consider perhaps Hume’s most famous argument, which begins with the question, ‘What justification do we have for our factual beliefs?’ By ‘factual beliefs’, Hume meant those beliefs that we can imagine (that is, form a mental image of) and as a result of these images say are either true or false, occuring or not occuring. For example, when we see two billiard balls collide on a table, we believe that the impact of the first ball will cause the second to move in a particular direction. This belief is ‘factual’ because we can also imagine the second ball not moving at all, or returning in the direction from which the first ball came, or vanishing in a puff of smoke. Since we can imagine any of these things, they are all logically possible. Therefore, Hume concluded, there is nothing in the motion of the first ball from which we can logically infer the motion of the second. That we have an accurate belief as to how the second ball will move is not based on any logical deduction from the movement of the first, but from our past experience of seeing billiard balls collide.

But, Hume persists, what is our justification for drawing conclusions from experience? Only our belief that the future will be like the past. But this too is a factual belief. We can imagine that the future will not be like the past – for example, that tomorrow billiard balls will vanish upon being hit by other billiard balls. So our belief that the future will resemble the past is itself not based on any process of deductive reasoning, but solely on experience. So how is experience itself justified?

To justify anything, you give reasons. And you justify those reasons by giving still other reasons. This implies three possible structures for any chain of justification:

(1) Reasons go on forever, without repeating.

(2) Reasons go in a circle – that is, eventually a reason is repeated.

(3) The chain stops, with a final reason.

Structures (1) and (2) would plainly provide unsatisfactory justifications, which leaves structure (3). But if a chain of justification is to stop in a satisfying way, the last reason given must not require further justification. And since we can imagine the contrary of a factual belief, a factual belief cannot be a final reason. So a factual belief that the present is like the past cannot be the final justifying reason for any conclusions about the world.

“David Hume at 300”, Howard Darmstadter, Philosophy Now