Hume makes no secret that he is a determinist, in a Newtonian mold: “It is universally allowed that matter, in all its operations, is actuated by a necessary force” (Inquiry, 91). One billiard ball of matter moves another, and so on; if you know the position of one such particle, you could determine the positions of all others. I’m not sure this holds water in the world as described by Heisenberg, Planck, and Schrodinger (with his famously half-dead cat), but that doesn’t matter as much as what comes next, when Hume annexes human behavior to this causal account, insisting, “The conjunction between motives and voluntary actions is as regular and uniform as that between cause and effect in any part of the universe” (98).
There is an initial plausibility to this line of thought. After all, everyone grants that there is an intimate link between a person’s motives and his choices; asked, “Why did you make a fourth pot of coffee?”, I might say, “Because I was tired, and because I need to be alert to read these next twenty pages, and because I’ve turned myself into something of a caffeine junkie.” It wouldn’t ordinarily occur to me to say, “Because I willed it!”
But there’s a problem here as well, which Hume brushes up against: this account of human action obliterates any distinction between deciding and predicting. (I learned to articulate this distinction, and indeed learned the very example I use below, from Roger Scruton.) Hume writes: “It seems certain that however we may imagine we feel a liberty within ourselves, a spectator can commonly infer our actions from our motives and character; and even where he cannot, he concludes in general that he might, were he perfectly acquainted with every circumstance of our situation and temper” (103n7). An individual might take himself to freely decide for one course of action rather than another, but, Hume insists, that’s simply an illusion generated by our inability to assess the causes (motives) that yield that decision; we’re not necessarily more privileged in assessing that causal web than a spectator. What appear to us to be decisions are really just half-baked predictions.
Let’s apply this Humean analysis to a particular case. At a dinner, someone asks Joe what he’s going to do later that night. Imagine two different responses Joe might give. 1) “Well, I’ll probably go out and get drunk, because that’s just the kind of person I am.” 2) “I’m going to do the right thing tonight, and go home and spend time with my wife and kids.” Option 1 is a predication; in offering it, Joe is merely making an observation about what’s likely to happen, one which he likely isn’t particularly more qualified to give than his therapist, AA sponsor, or brother.
In Option 2, however, Joe isn’t predicting his behavior, but rather making a decision, which differs from predicting in at least two ways. First, a decision involves the offering of reasons, in the sense of final causes — prediction appeals only to efficient causes, but decisions belong to intelligible patterns of action for an end. And second, a decision incorporates the idea of responsibility — in deciding to go home rather than get drunk, Joe is taking ownership of his action as something for which he might be blame- or praiseworthy. He inhabits the action in a way a spectator can’t, and in a way he can’t if he’s merely predicting. On this account of deciding, we’re free, not because our actions are wholly spontaneous or inexplicable, but rather because they incorporate rational ends and moral valuation.
It’s possible, even in a Newtonian universe (again, for the sake of argument), to maintain the existence of this kind of rational freedom. For instance, you might go the way of Leibniz, according to which our actions unfold in a “pre-established harmony,” our bodies enmeshed in the chains of efficient causation that determine material existence, but our soul’s willing according to final causes, the two lines of causation perfectly complementing one another. (Ultimately, this probably the source for the similar distinction in Kant between the physical world and the Kingdom of Ends.) We’d obviously need to say a lot more to determine if that yields a satisfying philosophy of mind or of action, but it at least saves the data of human experience in a way that Hume’s reductionism just doesn’t.