The Necessity of “Necessity”: Hume’s Psychology of Sophisticated Causal Inference

Much of what Hume calls probable reasoning is deliberate and reflective. Since there are aspects to Hume’s psychology that tempt some commentators to think, on the contrary, that for Hume all such reasoning is simple and immediate, I will be concerned to emphasize Hume’s recognition of the sophisticated sort of probable reasoning (section I). Though some of the details of my case may be new, the overall point of this section should not be news to recent scholarship. But once we recognize that this reflective and deliberate reasoning constitutes a significant portion of all probable reasoning, it becomes legitimate to ask how Hume accommodates this reasoning in his psychology, his ‘science of man.’ I believe that Hume has an answer to this question. I will explain in what way Hume could have thought that probable reasoning can be sophisticated: in short, sophisticated probable reasoning involves the use of the concept of evidence or epistemic support (section II). Hume’s psychology, constrained by his empiricism, must therefore explain how we come to have this idea. This will allow us to see his discussion of necessary connection in a new light. Hume does not reject the idea often referred to as ‘the idea of necessary connection’; indeed, it has precisely the role that I’m suggesting the concept of evidence has in Hume’s picture of sophisticated causal inference. However, the metaphysical accretion that gives the idea of necessary connection its name is not what’s doing the work in Hume’s psychology of this sort of reasoning (section III).¹ So the idea of evidence or epistemic reason that figures in Hume’s psychological theory is not really that of necessary connection at all, though it is often referred to in this way, even by Hume. I then address a challenge to my claim about the role of this idea in sophisticated probable reasoning.

I Simple vs. sophisticated probable reasoning

There are passages in Hume’s writings that suggest that many instances of probable reasoning — which for Hume is causal reasoning² — are immediate and unreflective (e.g. Treatise 1.3.8.13; SBN103). But did he think of all probable or causal reasoning in this way? Hume’s explanation of our inferences in terms of custom and habit might tempt us to answer yes. Suppose we experience an object/event A followed by an object/event B, and furthermore, we experience that every object resembling A is followed by an object resembling B. Hume’s claim is that as a matter of habit, when presented with one such constantly conjoined event, the other will come to mind immediately (1.3.6.15, SBN93), ‘without any new reasoning or conclusion’ (1.3.8.10, SBN102).³

A further consideration suggestive of the immediacy and unreflectiveness of causal inference might be taken from what Hume says about reasoning in animals. Hume thinks that his view about the role of custom in probable reasoning is strengthened if the same principles are at work in animals (ECHU 9.1, SBN 104).⁴ He goes on to argue that probable inference in animals is indeed ‘not founded on any process of argument or reasoning’ (ECHU 9.5, SBN 106), and that this supports the view that the circumstance for the generality of mankind is the same (ECHU 9.5, SBN 106-7). Probable reasoning in animals is immediate and unreflective, and Hume thinks that the situation with us is no different. No wonder, then, that many commentators appear to take Hume to think that all causal inference is immediate and unreflective.⁵

Hume no doubt thinks that some instances of causal reasoning are properly described this way. But there are others that are more deliberate and reflective. Some of these are undertaken in the various sciences, as Hume makes clear in the Enquiry.⁶ Closer inspection of several cases discussed in the Treatise will clarify what Hume thought was going on in these sophisticated inferences.

Case 1: Separating the superfluous from the essential. The first sort of case is addressed in Hume...