There is a crucial difference between these two cases. All other integrations into a super-science, of sciences dealing with entities and properties apparently qualitatively very distinct, was achieved by saying that really some of the entities and properties were not as they appeared to be; by making a distinction between the underlying (not immediately observable) entities and properties and the phenomenal properties to which they give rise. Thermodynamics was conceived with the laws of temperature exchange; and temperature was supposed to be a property inherent in an object. The felt hotness of a hot body is indeed qualitatively distinct from particle velocities and collisions. The reduction was achieved by distinguishing between the underlying cause of the hotness (the motion of the molecules) and the sensations which the motion of molecules cause in observers. The former falls naturally within the scope of statistical mechanics — for molecules are particles; the entities and properties are not now of distinct kinds. But this reduction has been achieved at the price of separating off the phenomenal from its causes, and only explaining the latter.

All reduction from one science to another dealing with apparently very disparate properties has been achieved by this device of denying that the apparent properties (i.e. the “secondary qualities” of colour, heat, sound, taste, etc.) with which one science dealt belonged to the physical world at all. It siphoned them off to the world of the mental. But then, when you come to face the problem of the sensations themselves, you cannot do this. If you are to explain the sensations themselves, you cannot distinguish between them and their underlying causes and only explain the latter. In fact the enormous success of science in producing an integrated physico-chemistry has been achieved at the expense of separating off from the physical world colours, smells, and tastes, and regarding them as purely private sensory phenomena. The very success of science in achieving its vast integration in physics and chemistry is the very thing which has made apparently impossible any final success in integrating the world of mind into the world of physics.