Direct Realism is worth defending. Its common sense appeal and epistemic optimism affirm humanity's moral responsibility, preserve its confidence in scientific investigation, and satisfy its intuitions about phenomenology and the objective physical world. Although the Problem of Secondary Qualities presents a formidable obstacle to accepting Direct Realism, there are Direct Realist accounts of perception and theories of secondary qualities, like Thomas Reid’s, that can overcome it.

The Problem of Secondary Qualities infers the falsity of Direct Realism from three premises: a spreading principle, that a good analysis of secondary qualities should extend to perception generally (SP); an empirical claim, that secondary qualities are among the objects of human perception (OC); and a supposed scientific discovery, that physical objects do not possess secondary qualities (NPT). From these, it concludes that no perceivable qualities are possessed by physical objects, which would make Direct Realism false.

All three premises from the Problem of Secondary Qualities are prima facie plausible. But while SP and OC hold up against a wide range of criticisms, NPT is somewhat weaker. It rests on a poorly defended hypothesis about whether secondary qualities are physically causal. They are not, it is supposed, because scientific properties account for the whole of physical causation, particularly between the objects of perceptions and human sense organs, making additional properties superfluous. But scientists and philosophers who say that physical objects do not possess secondary qualities because they are non-causal have spoken too soon. Even if causal-scientific properties formed the entire causal story about stimulating sense organs, this does not bar secondary qualities from the world of science. They can be objective and causal if they are identical to known scientific properties.

One way to make these a posteriori identifications is on the basis of Thomas Reid's theory of secondary qualities. According to Reid, secondary qualities are physical properties and objects of human perceptions. But original conceptions of them are relative and obscure. You do not understand the natures of secondary qualities merely by perceiving them. Rather, you grasp them in terms of their relations to your sensations, which are natural signs for the qualities.