Liberty and Necessity

Class notes for 10 February

Main points

Hume is a determinist who believes that our values are compatible with the truth of determinism.

In fact, he goes beyond simply arguing for the compatibility of determinism and responsibility since he maintains that the former depends on the latter.

I pointed out an apparent tension between his account of liberty and necessity and his account of causal inference. We do not feel determined in the former case, but we do in the latter.

There is something to Hume’s answer that we feel determined when thinking about other people’s behavior but not our own. But it raises other complications. Since we don’t observe others’ motivations, how do we make the relevant inferences? (Perhaps by looking at their circumstances rather than their motives: And how do we square that with our observations of our own behavior? (Credit where it’s due: Daniel Smythe raised these questions in the Junior Seminar last year.)


Hume points out that we are able to predict and generalize about human behavior. Given my knowledge of how people like you behave, I can predict how you will behave. He takes this to show that we make causal inferences about people’s behavior. Remember what causal inferences involve: all As have been followed by Bs, here is an A, therefore, a B will follow. All people avoid what they believe will be painful, this person believes that touching that flame will be painful, therefore this person will avoid touching that flame.


Hume takes that to mean that we assume that human behavior is caused, just like everything else. Rachel articulated the major competing explanation: people’s behavior is predictable because they are rational. That explains why they make similar decisions even if their behavior is not caused but instead is free: the decisions they make are the rational, sensible ones to make.

As we will see next time, Hume will put pressure on this alternative. He will insist that we explain human behavior by citing motivations and treating them as causes. Reason, by contrast, cannot explain action. Or so he will say.


Adi pointed out that Hume has not proven that all behavior is caused. It may be true that some of our behavior is caused: for example, it is hard to belive that reproduction is simply a matter of choice since if it were you would expect a more even distribution of choices. But even if you grant that our reproductive behavior has causes, it doesn’t follow that all behavior is caused and, in many cases, we cannot accurately predict or make causal inferences about behavior.

Hume’s answer is that the causes are concealed from us, but still there. Adi is right to say that it’s hard to prove or disprove that claim. How can I show that there are, or that there are not, hidden causes? By hypothesis, they’re hidden!

I think that Hume would try to shift the argumentative burden. Given that most of our behavior is caused, why assume that some is not? What could produce action if not a cause? (This is the sort of thing I would say to my friend who grants that our motivations cause our behavior but maintains that there is always a moment of choice between motive and action: why think that the choice is not caused by some motive or other?)


Hume’s argument that responsibility depends on determinism turns on the need to find something enduring to hold responsible for decisions. It’s character, he says.

He makes his case by saying that his opponents can’t find anything persisting to hold responsible. If our actions are not determined by our characters, each decision is determined by a fleeting motive and it doesn’t make any sense to hold a different fleeting motive responsible for what a fleeting motive in the past did.

But why are those the only alternatives? Why couldn’t we hold the characterless chooser responsible, such as the immaterial soul or noumenal self? (The second will be familiar only if you’ve read Kant’s moral philosophy; if you haven’t, don’t worry about it).

I suspect that Hume left this alternative out because he thought it was impossible to understand how anything without character or motives could act at all. So the only alternatives he considered are character, which persists over time, or individual motivations, which do not.