In his A Short History of Modern Philosophy, Roger Scruton defends Locke’s distinction of an object’s “primary” from its “secondary qualities. ” The distinction works like this: “whereas primary qualities resemble the ideas that are produced by them, secondary qualities do not…Primary qualities are really in the objects which possess them, whereas secondary qualities are not” (89). Locke names “colors, sounds, tastes” as secondary qualities (Essay 2.8.10), while his list of “primary qualities” includes “solidity, figure, extension, etc.” (Essay, 2.8.9). Nonetheless, Scruton appears to understand primary qualities more broadly, as “certain scientifically determinable and measurable qualities” (89), perhaps including molecular structure or chemical properties. He then notes Berkeley’s fairly obvious criticism of this view: “since ideas are mental entities, belonging to a wholly different realm…it is prima facie absurd to suppose that ideas can resemble things which are not ideas” (89).
Scruton insists, however, that this objection can be overcome if “we attempt…to free Locke’s insight from the dead theories which enclose it” (89). The real difference between primary and secondary qualities lies in their relations to perceivers: “the secondary qualities seem to stand in need of a perceiver, the primary qualities only in need of an objection” (89). The secondary qualities are indexical — a pan that feels hot to John might feel only warm to Jane, a berry that tastes sweet to Jim might — because of neurological damage or the coffee she just drank — taste bitter to Mary. But the primary qualities not only do not presuppose a perceiver, but also account for the secondary qualities (89), in the sense that a scientific account of an object’s smell will appeal to its molecular or chemical structure, and not vice versa. They are logically prior, ontologically more basic than secondary qualities.
Is this right? I don’t see much evidence that Locke’s discussion is actually aiming at this distinction; Scruton’s account seems more inspired by Locke than a clarification of him. Again, Locke’s examples of primary qualities are “solidity, figure, extension, etc.” But, as Berkeley points out, these are also indexed to perceivers: a ball that appears perfectly spherical to me might, to a mite, appear as cratered as the Moon, and an object that appears perfectly solid actually consists, if I can trust Bohr, of mostly empty space. Here too, esse ist percipi.
What about Scruton’s distinction among qualities in terms of explanatory order? This amounts to noting that a rock is composed of molecules whether anyone encounters it or not; but again, Berkeley would interject that the claim “rocks are composed of molecules” still doesn’t get us perceivers out of the picture. That claim, like all scientific hypotheses, is a strategy for (to borrow Quine’s phrase) “swelling ontology to simplify theory” with respect to the world as it shows up; the claim that rocks are composed of molecules is a way of explaining properties rocks have, properties that perceivers interact with and in terms of which they measure the success of their theories.