It is natural to suppose that honey tastes the way it does because it is sweet. Democritus, Plato and Aristotle all agree that this explanation is superficial and lacks causal depth; they attempt to explain gustatory phenomena by invoking explanatorily fundamental features of the world. As they work out their causal stories, do they give up on the common-sense explanation of why honey tastes the way it does? In other words, do they deny that sweetness and other sensible qualities are (...) causally efficacious features of the world around us? ;Our sources for Democritus' thinking on this issue present a problematic picture. On the one hand, Aristotle and Theophrastus suppose that Democritus wants to identify colors and flavors with causally efficacious features of the world around us. On the other hand, Theophrastus and others find Democritus defending the view that colors and flavors do not belong to the world around us. This tension, first pointed out by Theophrastus, is merely apparent, if, as certain texts suggest, Democritus distinguishes the colors and flavors which we experience from the colors and flavors which figure in his scientific explanations. ;In the Timaeus, Plato frequently says things like "What is capable of producing such and such effects on the gustatory region has the name 'bitter' thinspace". On Theophrastus' interpretation, Plato is offering definitions of the sensible qualities: flavors are identical with powers to produce certain physiological effects on the sense-organs. Theophrastus reasonably complains that sensible qualities will lack causal efficacy if they are identified with these powers of external objects, but Theophrastus' interpretation is mistaken: Plato is not defining the sensible qualities; he is fixing the reference of our sensible-quality terms. Recent historical work on the distinction between primary and secondary qualities has often emphasized the extent to which the authors of this distinction were motivated by an interest in supplanting prevalent, Peripatetic assumptions about the nature of sensible qualities, but Aristotle's theory does not differ significantly from that of Descartes: both identify colors with causally efficacious properties whose natures are revealed by scientific investigation. (shrink)