What is a cause? Notes for Ocober 18

Main points

Hobbes made fun of the Romans because they couldn’t accurately distinguish between causation and correlation. They wanted to have a general named Scipio lead them because generals with that name had won in Africa in the past. But the particular Scipio they picked wasn’t up to the task. (I did a little more digging in Plutarch. The second passage makes the point that Barrett did. Cato, at least, was pretty clever. Caesar was too.)

In the sections we’re reading today, Hume offers his own definitions of cause. But can Hume do any better than the Romans did? Can he explain what the difference between correlation and cause is?

Probably not. But that’s what we should expect. If there were a clear difference between them, we would have reason to identify causes, as distinct from mere correlations, and that would give us reason to make causal inferences.

What is a cause?

John and Mike pointed out the distinction between cause and correlation. But what does this distinction consist in? That is, what is different about causes and correlations?

Eleanor pointed out one feature. Causes make their effects happen. This isn’t so with two correlated events. The one happens independently of the other.

Martin identified another feature. If an A caused a B, then every A should cause a B, provided the circumstances are similar enough. But if C and D are merely correlated with one another, their relationship won’t be uniform. Causes always produce their effects. But there is no reason to think that the correlation between C and D will always hold.

(Well, it might hold in some cases. Suppose C and D are both effects of E. Then every time an E occurs, a C and a D will. C doesn’t cause D, but the relationship between C and D will be uniform.)

But must a type of cause always produce the same effect, John asked?

If the question means “is it logically necessary that every A will cause B as its effect?” then the answer is no. That’s the lesson of Hume’s arguments that we do not draw causal inferences using demonstrative reasoning. The billiard ball could go up, the sun might not rise, it isn’t contradictory to imagine any of those things, and so on.

If it means “do we think that an individual cause made it necessary that its effect happen?” then the answer, according to Hume, is yes. We can’t make any sense of this ‘necessary connection’ between cause and effect, but believing it is there is part of believing that A caused B. This is Eleanor’s observation: we think causes make their effects happen.

If the question means “do we think that causes are always uniform?” then the answer is a matter of debate. Hume thinks we do. He’s with Martin.

Anscombe disagrees. She notes that we can sensibly say that I spilled the milk (i.e. I caused the milk to leave its glass and spread across the table) even though I am often in the vicinity of milk without spilling it.


I mentioned occasionalism, the view that so-called ‘secondary causes’, that is, the cause and effect relations that we finite creatures are familiar with like the examples we have been discussing, are just occasions for God to make the resulting effects happen. There are roughly two paths to occasionalism:

  1. The philosophical path. We can’t understand how limited, finite bodies could make it necessary that effects happen. But there is such a thing as cause and effect. Therefore, there must be something with infinite power that makes it necessary that the effect happens.
  2. The theological path. God is omnipotent, meaning he has infinite power. If a finite cause could make its effect necessary, then it could resist God’s power. ‘Necessary’ means necessary, no matter what else happens, after all. Therefore, finite causes must be incapable of making their effects necessary without God’s intervention.

Warning: I said those were rough. You really don’t want to read the actual arguments right now. Trust me on that.

Of course, Hume points out that if we can’t understand how a finite thing has the power to make its effect happen, we can’t understand how something infinite can. The problem is that we don’t understand power, not that we need more of it, whatever it is, in order to understand causation. As Barrett said, throwing in “omnipotent” is just adding a word, not explaining anything.

Hume took a shot at occasionalism because his big-P Philosophy, his understanding of our place in the universe, is a naturalistic one. While his contemporaries thought that we were essentially rational creatures, made in the Image of God, Hume did not. He thought that we are natural beings, like other animals. We don’t have souls or minds that are different in kind from the rest of the animal world. He didn’t think there was anything supernatural and he thought that religious belief was often pernicious.

This will all be clear in his arguments about miracles and divine providence. But I wanted to point out that it’s present here too.


We will have to discuss Anscombe’s views on Friday. There’s just too much interesting stuff!

This page was written by Michael Green for Problems of Philosophy, Philosophy 1, Fall 2006.
Name of website