The Theory of Ideas

Class notes for 6 January

Main points

The theory of ideas provides Hume with the basic elements of his science of man. He will use these elements in his explanations of our thoughts and feelings. It’s a little misleading to call it a theory in Hume’s hands since he just lays it out there without much argument or, apparently, care. It’s more a set of assumptions that seemed obvious to Hume and his contemporaries.

According to Hume’s version of the theory, all of our thoughts and feelings are perceptions. There are two distinctions that cut across one another: impressions vs. ideas and simple vs. complex. Impressions include our sensations, passions, and emotions: everything we feel. Thinking and reasoning is done with ideas. (For feeling vs. thinking, see Abstract, par. 5).

The first three sections concern impressions and ideas. They are about thinking in the sense of having a feeling or thought.

The next two sections (§§ 4-5) are about connections among ideas. They concern thinking in the sense of moving from some feelings or thoughts to others. Section 5 gives philosophical relations, the seven different ways that perceptions can be compared with one another. Section 4 concerns the movements from one perception to another that we actually tend to make: these are called principles of association or, later, natural relations.

The last two sections (§§ 6-7) concern questions about our ability to think about things apart from their qualities. Section 6 concerns our thought that a thing can remain the same despite changes in its qualities. Section 7 concerns our thoughts about general categories: how do we have a thought about human beings in general, as opposed to thoughts about particular human beings?

Our discussion of the last part (§§ 6-7) was brief because we will encounter these topics again, in 1.4.2.


The most important point I wanted to establish concerns the problems with Hume’s claim that every simply idea is copied from a simple impression.

They are important because Hume will argue that several apparent ideas are questionable on the grounds that there is no corresponding impression from which they might be derived. You have already seen an example of this kind of argument in the Abstract (see paragraph 7).

A question we will want to keep in mind is: “Given the apparent concessions in Book 1, part 1, how can he prove that?” (Of course, even if he hadn’t conceded it, we could still ask how he can prove it: asserting the theory of ideas doesn’t amount to proving anything).

How can he refute an opponent who says that he did too have a simple impression of causal necessity? How can he show that someone who claims to have reasoned his way to moral distinctions has not done something similar to the person who filled in the missing shade of blue?

Corrections concering substance

Section 6 is about how we can think of things as independent of their qualities. We do, in fact, think that the book will remain the same even if its qualities change: I will have the same book even if I shave off the edges or bleach the cover. Since Hume is trying to explain all of our thoughts, how does he explain this one?

I put the point in negative terms in my lecture. I said that he denied one description of this kind of thought, namely, that what we think is that the substance of the book remains the same while its modes change. (“Modes”, “accidents”, and “qualities” are used more or less synonymously, though any particular author may mean something specific by any one of them, so you always have to be on your toes).

That’s true: Hume denies that we have any idea of a substance. Bear in mind that, as Katie’s question about putting the book in a wood chipper (heavens forbid!) showed, “substance” is not ‘matter’ or ‘stuff.’ Rather, it’s supposed to be matter conjoined with a ‘substantial form.’ What’s a substantial form? It’s supposed to be different for every kind of thing: for a knife, it has something to do with having the ability to cut; for a book, it probably concerns the words; and so on. Hume, in any event, isn’t having any of it: he denies we have any idea of what a substance would be.

But what I should have added is that Hume has something positive to say about how we can think of a thing’s persisting through changes in its modes. His answer is linguistic, very much like the answer in section 7. We associate a word with a collection of qualities. When we learn that a thing has a quality that we had not previously associated with the word, we add that to the collection. In section 7, we are said to associate words with different things that resemble one another, such that we can think about kinds of things apart from their qualities even though, strictly speaking, there is no single idea corresponding to that word.


I have two additional things to throw in that were not in the lecture. These are minor points that I am including mainly for my own edification (writing them out in public helps me to think them through).

If you don’t know what innate or abstract ideas are supposed to be, you need not worry about this.

Innate ideas

Hume used “perception” where Locke used “idea” (see Why? Here is a speculative answer.

Hume wanted to agree with Locke’s contention that there are no innate ideas, that is, no ideas in our minds prior to our having experiences.

But, as he noted, the most that Locke had shown is that all ideas are derived from sensations or passions. Since we don’t know where either comes from, we can’t conclude that they are not innate.

Moreover, in the case of some of the passions, it seems pretty clear to Hume that they are part of human nature and not learned from sensory experience. Recall Katie’s unfortunate encounter with a diaper: there are some things that we just find disgusting, without learning to do so.

For references in Hume, see the Abstract, par. 6 (“natural affection, love of virtue, resentment, and all the other passions”); see also (passions that “arise from natural impulse or instrinct,” such as the desires to punish enemies and bring happiness to our friends, hunger, lust, and other “bodily appetites”); and 2.2.11 (the “bodily appetite for generation”).

So, “idea” is reserved for the parts of our mind that fit Locke’s claim that there are no innate ideas, even if that means revising Locke’s own definition of “idea.” “Impressions,” by contrast, are parts of the mind that may be innate (or, at a minimum, we cannot show that they are not innate).

Maybe. I should review Locke more carefully than I have done before I can say this with confidence. But it seems to make sense.

Simple and abstract ideas

How can Hume maintain both that we can separate every simple idea from every other idea and also that we cannot have abstract ideas?

If I can separate the (putatively) simple idea of the apple’s red color from the complex idea of the apple, why can’t I have an abstract idea of the color red? Isn’t that just the idea of what all red things have in common?

There’s some fancy footwork going on concerning “distinctions of reason” that, I think, is supposed to answer this question ( But I have a hard time following it. I suspect this is going to be an issue again in 1.4.2. So, for now, I’m going to leave it as a question.