Substance, Body, and Mind

Class notes for 27 January

Main points

Sections 3-6 discuss different attempts to understand why we regard bodies and minds as distinct, coherent things. Sections 3-4 are about bodies; 5-6 are about minds.

In both cases, Hume believed that the attempts to understand bodies and minds fail.

The ancient philosophy’s attempts to explain the identity of objects despite change and the relationship between a thing and its parts both rely on substance but we have no idea of what a substance is and so they fail.

The modern philosophy uses causal reasoning to distinguish primary and secondary qualities. This reasoning conflicts with the senses, which do not distinguish between qualities in this way. The former leads us to believe that bodies do not really have colors, the latter leaves us unable to think of bodies without colors. Hume’s discussion leaves out most of the material you need to make sense of the modern philosophy, so I have written up a quick background.

I singled out and as especially interesting paragraphs. They identify a position Hume calls “moderate skepticism” that seems to be his own attitude towards these problems. This position also seems to have some bearing on his beliefs about the justification for making causal inferences.

1.4.5 concerns two different attempts to understand the fundamental constitution of the mind: is it material (e.g. a brain) or immaterial (something else, perhaps a soul)? Hume’s position is that each side has effective arguments against the other. We discussed the immaterialist’s argument about local conjunction. According to Hume, the alleged impossibility of conjoining something with matter’s properties (e.g. extension, spatial dimensions and location) and something like an immaterial soul brings down both sides.

I will first add a few remarks about local conjunction and then I will try to explain two things in 1.4.5 that we did not discuss in class: Hume’s references to Spinoza ( and the argument about causes ( NB: Hume’s conclusions are neatly summarized in

Local conjunction

To make sense of the arguments about local conjunction, you need to know what “substance” meant. A substance is something that does not depend on anything else to exist. Qualities (aka “modes” or “accidents”), by contrast, depend on something else to exist: substances! They “inhere in” or are “modifications of” substances.

An example may illustrate the distinction. The fist I am making depends on my hand’s existing, but my hand does not depend on my fist’s existing. (While I don’t think anyone regarded a hand as a substance, you get the idea of the asymmetrical relation between substance and quality, I hope).

Immaterialists held that the mind cannot be a material substance. An argument for their position was that some of our thoughts cannot be conjoined with matter: some of our thoughts cannot have exact spatial locations, but all material things can. That meant, in their opinion, that those thoughts could not be modifications of a material substance: you could not have those thoughts if your mind were material.

An immaterial substance, according to the immaterialists, does not have the features of matter. It lacks extension (spatial dimension) and it is simple (it cannot be divided or broken down into parts).

But, according to Hume, if we grant the immaterialists’ argument that things with extension (like matter or material brains) cannot be conjoined with things that lack it (like desires or thoughts about moral distinctions), materialists can pull the same trick. Some of our thoughts do have spatial dimensions and can be given a spatial location. It follows that these could not be conjoined with an immaterial substance since, by the immaterialists’ argument, you can’t conjoin something that lacks extension with something that has it.

Most of you didn’t like that last part, although it’s hard to formulate an objection given that we aren’t terribly familiar with these terms. Talking with Adi helped me see how to make Hume’s point a little easier to swallow.

I don’t think he’s committed to saying that we can give an exact spatial location to all of our visual or tactile perceptions. Instead, I think he’s only committed to saying that we can given them a spatial location relative to some others. My visual perception of the top of the window is located above the visual impression of the lamp. I think that’s enough to embarrass the immaterialists: any perception with any extension is trouble for them.

If he had to make the stronger claim that we can give spatial locations to all of our visual impressions, his position would be very hard to understand. My present impression of the window is … where relative to my memory of seeing Half Dome? (Other than “in your head, silly” — where are they relative to one another in my head?)

Spinoza and the immateriality of the soul

Here is a crude version of Spinoza. Substance must be infinite: this is supposed to follow from a very complex argument that, happily, we can ignore. Given that argument, there can only be one substance and everything we see is a modification of this substance. Not only that, but God is identical with this substance.

This picture of God is quite a way away from the personal God, the one that cares about and intervenes in human affairs. It was thus considered unorthodox at best and atheistic at worst.

Hume’s discussion of Spinoza is meant to embarrass immaterialists. They believe that all of our ideas are modifications of a simple immaterial substance: a soul. But, Hume reasoned, their picture of the mind is parallel to Spinoza’s description of the universe. If Spinoza’s system was impossible, as many immaterialists believed, then so is immaterialism about the mind.

That’s what Hume was trying to show, at any rate.

Incidentally, if you want the real version of Spinoza, we are lucky enough to have a top flight expert teaching a course on Spinoza in the Spring term: Steven Nadler. Thumbs up!

Material and immaterial causes

Many of Hume’s contemporaries (and our own as well) assume that there cannot be causal relations among things of a different kind. Matter cannot cause immaterial effects and immaterial substances cannot cause material effects.

Immaterialists, for example argued that matter alone cannot cause our thoughts since our thoughts are of things that go beyond the qualities of matter (; see the annotation to for references).

Hume used his theory of causation to undermine this argument. As cause Bs if As are constantly conjoined with Bs. There is no requirement that As and Bs are the same kind of stuff, material or immaterial.

A similar point could be made against materialists who assert that the mind must be material (e.g. states of a brain) because bodies are material. That follows only if matter causes our perceptions and matter can only cause material effects.

In any event, the alternative to admitting that matter has causal efficacy (in this sense) is occasionalism, the view that God is the real cause of all effects. Hume points out that this would mean that God causes all of our thoughts as well, since we have no more idea of how one thought can cause another than we do of how matter can cause effects. But that means God would be responsible for all of our actions as well, including the evil ones. According to Hume, Malebranche tried to dodge that conclusion by excepting volitions (= will, voluntary causes of action) but, Hume argued, that was an ad hoc assumption, motivated only by a desire to square his position on causation with his religious convictions. (

In addition to being great fun (well, for me), that argument exposes a sense in which Hume claimed there really are causes: constant conjunctions. That isn’t enough to get our idea of causation, which includes the idea of necessary connections between cause and effect, but it does suggest he wasn’t what I called a causal nihilist.