In Part I, the most vulnerable premise of the Problem of Secondary Qualities seemed to be the Secondary Quality Non-Physicality Thesis (NPT), that if an object is physical, then it does not possess secondary qualities. Robinson characterised this claim as a deliverance of science, citing Galileo, Descartes and Locke as authorities, and commitments to NPT appeared in the writings of contemporary scientists. In spite of philosophy's high regard for science and strong desire to accommodate its best theories, acceptance of NPT needed a more detailed argument. Frank Jackson offered such an argument, one that rests on a hypothesis regarding the causal potency of secondary qualities. Scientific properties, Jackson said, fully account for the causal connections between physical properties and sense organs. So sensations offer no reason to posit non-scientific properties in physical objects, anything in addition to scientific properties would be causally superfluous. This creates a problem for secondary qualities, said Jackson, because secondary qualities, being conspicuously absent from current scientific theories, are not scientific. Sensations do not count as evidence for the possession of secondary qualities by physical objects. Without the evidence of sense, why think that physical objects possess secondary qualities at all?

Rejecting the Non-Physicality Thesis

Rejecting NPT means challenging Jackson's claim that secondary qualities are not scientific and therefore not causally potent. How can secondary qualities be causally potent and therefore scientific? They seem not to be because they are not readily understood in scientific terms. Jackson is right that the language of science does not mention secondary qualities by their traditional names, and the conceptions by which perceivers originally grasp secondary qualities are not scientific notions. But these facts about secondary quality names and notions do not demonstrate anything about the natures of secondary qualities themselves, only the way that English speakers tend to think and talk about them. Even if traditional secondary quality vocabulary words do not appear in scientific explanations, it may be that scientific theories make reference to secondary qualities. This would be the case, for example, if secondary qualities were identical to known scientific properties or combinations of scientific properties.