Hume's nominalism and the Copy Principle

Abstract

I consider some ways in which the Copy Principle (CP) and Hume's nominalism impinge on one another, arguing for the following claims. First, Hume's argument against indeterminate ideas isn't cogent even if the CP is accepted. But this does not vindicate Locke: the imagistic conception of ideas, presupposed by the CP, will force Locke to accept something like Hume's view of the way general terms function, the availability of abstract ideas notwithstanding. Second, Hume's discussion of nominalism provides support for the “old Hume” interpretation, that which takes the CP to be a criterion of meaningfulness, as against the “new Hume” reading, according to which it constrains what we can know. Finally, nominalism forces Hume to adopt a more complicated theory of ideas.

Keywords::

• Hume
• Locke
• nominalism
• Copy Principle
• separability

Notes

^1. Hume uses the term “image” broadly, as applying to anything which can be given in, or copied from, experience, not just the visual. The term “impression” applies to “all our sensations, passions and emotions, as they make their first appearance in the soul”, and ideas are “the faint images of these” (T 1.1.1.1; SBN 1, my italics).

^2. According to Pappas (1989, 347), Hume “does not explicitly link [the CP] and abstract ideas”. Rather, the denial of abstract ideas functions “to fend off a quick and easy refutation of [the CP]” (1989, 348). This claim is doubly mistaken. Hume does link the two explicitly: he invokes the CP to argue against abstract ideas. And when he comes to deny abstract ideas, he has already (T 1.1.1; SBN 1–7) established the CP and contended with a counter-example (the missing shade of blue). And he does not say that abstract ideas might constitute a counter-example to the CP. So his view of the order of support is the reverse from the one Pappas suggests. Anticipating this objection, Pappas claims not to be concerned with Hume's intentions (1989, 350). Perhaps he thinks Hume's system would be better supported if the order of justification were reversed, since the denial of abstract ideas can be supported without invoking the CP: Hume cites two additional arguments that do not rely on the CP.

^3. Some commentators impute to him a nominalist “partial consideration” view of abstraction. As to the CP, it is moot whether Locke holds an imagistic conception of ideas, and even if he does, he wouldn't subscribe to Hume's CP as stated, because of terminological differences. Instead of Hume's “impression”, Locke uses the terms “perception”, “experience” and “observation”. And his use of the term “idea” is inconsistent. To facilitate the consideration of the dispute between Hume and (the possibly apocryphal) Locke, I shall adopt Hume's (consistent) terminology.

^4. I am grateful to Amy Schmitter for pointing this out.

^5. All quotations from Locke (1975).

^6. The qualification “that are known” is required because nominal essences change with experience.

^7. Winkler (1989, 39) aptly labels this “the content assumption”.

^8. And not, pace Bennett (1971, 256), if it is definable.

^9. The term “revival set” is Garrett's (1997, 24).

^10. The historical Locke should be impressed by Hume's third argument. He thinks that universals (triangularity, for instance) are impossible objects and that we cannot conceive of the impossible. So he must conclude that there can be no general idea “triangle”, because it is the idea of the universal.

^11. I am grateful to Amy Schmitter for her very painstaking and constructive comments.

Additional information

Notes on contributors

Ruth Weintraub

Ruth Weintraub is Professor in the Department of Philosophy, Tel-Aviv University.