Why is Hume struggling to reconcile causality with his notion of what is knowable?

Question

This is a follow-up on my earlier question.

Regarding causality as involving "a necessary connection between cause and effect" In what sense was Hume struggling to reconcile the "necessity' with his theory about the origin of knowledge ("perceptions")?

Isn't it the case that human experience is rife with patterns of "necessity" from where experiences about the notion of necessity can be drawn?

Would it be too farfetched to say that this is an ontological problem (his ontology did not contain objects like "necessity") or is it a methodological problem (he didn't think through how his objects interact with each other)?

Answer 1

It was an epistemological problem, Hume's theory of impressions and ideas was a bit too simplistic to describe human cognition realistically, which is understandable given the state of psychology at the time. Kant resolved the problem by postulating "synthetic a priori" of which causality is one. They do not come from "impressions" but rather from understanding and "pure" intuition, see Kant and Hume on Causality. Kant explicitly critiqued Hume's "necessity by association", and used it as inspiration for his position:

"Experience in fact teaches us that something is constituted thus and so, but not that it cannot be otherwise. Hence, if... a proposition is thought together with its necessity, then it is an a priori judgment... The very concept of cause so obviously contains the concept of a necessity of the connection with an effect and a strict universality of the rule, that the concept [of cause] would be entirely lost if one pretended to derive it, as Hume did, from a frequent association of that which happens with that which precedes, and [from] a thereby arising custom (thus a merely subjective necessity) of connecting representations."

Nonetheless, Hume's associationism was quite popular with 19th century philosophers and psychologists, including Mill, Peirce, James, Brentano and even Husserl, and led in part to the Jamesian idea of "stream of consciousness". Of course, empiricists among them rejected Kant's apriorism, but enriched both the theory of perception beyond his "impressions", and the theory of cognition beyond his association of ideas. With some inspiration from Hegel Kantian a priori were reinterpreted as historically accumulated "constitutive principles" based on past experience but not derived from it, either by association or otherwise. Peirce, for example, developed the modern view of hypothetico-deductive genesis of empirical laws in his logic of science, later Reichenbach and Friedman used the term "relativized a priori" for the empiricist replacement to Kant's synthetic a priori, see What are the more complex/interesting examples of synthetic a priori statements?

Answer 2

The assertion in the center is true, but irrelevant.

Humans like necessity, we ascribe it to tons of things. Most of those things are not really necessary. We learn from the exceptions all the time. The first time we drop a helium balloon and it does not fall, one necessity is replaced by another. Were we right or wrong before we encountered the exception?

If things we know are necessary firmly both personally and at a cultural level were not just wrong, science would stop. There would be nothing more to know.

So in what way is our ability to create incorrect assignments of necessity constantly proof of anything?

Hume doesn't have 'a problem'. He is not wrong. Necessity is a logical construct that we do not ever actually observe in reality.

We can observe strong patterns and deduce from them very high likelihoods, but we cannot observe necessity on any finite timeline. The exception might just lie in the future somewhere.

Hume does not rule out the ontological possibility of necessity. Nor does he prescribe a methodology. He just makes the deduction, mathematically obvious if one accepts basic probability theory, that we cannot know more than we can know. Certainty requires faith, and faith can always be either well placed, or misplaced.