A friend and I were talking about process philosophy the other day, and David Hume came up.

We were talking specifically about causation, and my friend referenced a book whose author noted that Hume undermined the notion that causality has a rational basis.

Hmm...let's have a look:

Hume's strategy dictates that he first show that alternative accounts of our "causal reasonings" are inadequate. This negative project directs his metaphysical microscope toward the intellectualist view that causal connections are made on the basis of the operations of the understanding. Hume proceeds by examining all of the possible ways in which our "causal reasonings" might be based on reason.

Reasoning concerns either relations of ideas or matters of fact. Hume quickly establishes that, whatever assures us that a causal relation obtains, it is not reasoning concerning relations between ideas.

It's not? Well, now I want to read Hume, to see how he "quickly establishes" this. The basis of forming a causal relationship seems very much related to "reasoning concerning relations between ideas."

Effects are distinct events from their causes: we can always conceive of one such event occurring and the other not. So causal reasoning can't be a priori reasoning.

I guess I just don't get this bit. Of course we can conceive of one event occuring and the other not. And of course that causal reasoning isn't a priori. But it is most certainly related to reasoning.

If I see a cue ball hit a nine ball, causing it to move, I can reason that a similar effect will result from a similar cause (e.g. that the cue ball hitting the eight ball will cause it to move as well). This mental mapping onto a new cause and effect relationship that I haven't experienced demonstrates that the expectation is a function of reason, not just of experience.

Or maybe I'm misunderstanding what he's asserting.

Causes and effects are discovered, not by reason but through experience, when we find that particular objects are constantly conjoined with one another.

This implies that we cannot discover cause and effect relationships through reason, but only through experience (which I think I've countered above). If you could only establish cause and effect relationships through direct experience with particular objects, then you wouldn't be able to generalize cause and effect relationships (which we do every day). If I poke a red balloon with a pin, and it pops, I can assume that a yellow balloon will also pop, even though I've never seen a yellow balloon pop. Certainly, this mapping is a function of reason, isn't it? I don't have to see a yellow balloon popped to intuit the causal relationship, do I?

Even after we have experience of causal connections, our conclusions from those experiences aren't based on any reasoning or on any other process of the understanding. They are based on our past experiences of similar cases, without which we could draw no conclusions at all.

I just don't see how this is the case. How are our conclusions not based on "reasoning or any other process of the understanding"? This just seems like nonsense to me. You see water transform into ice when the temperature drops below a certain level. You conclude that the next time water is exposed to temperatures that low it will freeze. How is this not a function of reason?

But this leaves us without any link between the past and the future. How can we justify extending our conclusions from past observation and experience to the future? The connection between a proposition that summarizes past experience and one that predicts what will occur at some future time is surely not an intuitive connection; it needs to be established by reasoning or argument. The reasoning involved must either be demonstrative, concerning relations of ideas, or probable, concerning matters of fact and existence.

Okay.

There is no room for demonstrative reasoning here. We can always conceive of a change in the course of nature. However unlikely it may seem, such a supposition is intelligible and can be distinctly conceived. It therefore implies no contradiction, so it can't be proven false by a priori demonstrative reasoning.

Lost me here. Because we can conceive of "a change in the course of nature" it's unreasonable to draw a conclusion? Because it's possible that the laws of nature will change at any minute, it is not reasonable for me to assume that the next time the temperature drops to freezing that water will become ice?

Probable reasoning can't establish the connection, either, since it is based on the relation of cause and effect. What we understand of that relation is based on experience and any inference from experience is based on the supposition that nature is uniform -- that the future will be like the past.

I don't understand why this supposition has to be absolute. Why can't it be provisional? Why can't you say, nature was uniform yesterday, and the day before that, and so on, so the expectation is that it will be uniform tomorrow. Yes, there's a small chance that it won't be, but every "trial" thus far indicates that it is, so it's reasonable to provisionally assume that it will continue to be the case. It's not a priori, but such an assumption would still be based on reason, wouldn't it?

So Hume says our conclusions about cause and effect aren't based on reason. So what are they based on?

Hume's negative argument showed that our causal expectations aren't formed on the basis of reason. But we do form them, and "if the mind be not engaged by argument...it must be induced by some other principle of equal weight and authority" (EHU, 41).
This principle can't be some "intricate or profound" metaphysical argument Hume overlooked. For all of us -- ordinary people, infants, even animals -- "improve by experience," forming causal expectations and refining them in the light of experience. Hume's "sceptical solution" limits our inquiries to common life, where no sophisticated metaphysical arguments are available and none are required.

When we examine experience to see how expectations are actually produced, we discover that they arise after we have experienced "the constant conjunction of two objects;" only then do we "expect the one from the appearance of the other." But when "repetition of any particular act or operation produces a propensity to renew the same act or operation...we always say, that this propensity is the effect of Custom" (EHU, 43).

It sounds to me like Hume is saying the only way we conceive causal relationships is through conditioning, being exposed to similar relationships over and over again. This just seems wrong to me. This seems to discount the ability to form causal conceptions that one has never directly experienced, which completely contradicts human experience.

I'm also interested in how Hume would distinguish between true causation and false causation, and how this is not a function of reason.

I snap my fingers three times every time I get in my car, believing that I won't have a wreck if I do. If I do this five hundred times without a wreck, according to Hume I have formed this causal link through custom, right? But is there not a causal fallacy at work here? There's no reason why snapping my fingers should lead to a safe trip (unless you want to argue that it creates a greater sense of mental security for me or somesuch, but let's avoid that avenue for now). We could conduct a study with thousands of drivers, asking them to snap their fingers before getting into their cars every day, and compare their rate of accident with a group of non-snappers. What kind of results do you think we'd get?

Is such a process not the application of reason? Hume seems to be saying that the only way to form a causal relationship is through every day experience and habit, that there's not underlying rational basis. If so, then how do you distinguish between sound causal relationships (e.g. water freezing) and unsound ones (snapping increasing car safety)? Because everyday habituation is misleading oftentimes, isn't it? How would we make such a distinction, if not for reason?

Anyway, my head hurts and I haven't read the source material. But the underlying assertions of Hume seem flawed to me in this case. Any philosophers or laypeople out there who want to clarify or correct anything here?