Although there is thus a sphere excluded from physics,[Pg 390] yet physics, together with a "dictionary," gives, apparently, all causal knowledge. One supposes that, given the physical characteristics of the events in my head, the "dictionary" gives the "mental" events in my head. This is by no means a matter of course. The whole of the foregoing theory of physics might be true without entailing this consequence. So far as physics can show, it might be possible for different groups of events having the same structure to have the same part in causal series. That is to say, given the physical causal laws, and given enough knowledge of an initial group of events to determine the purely physical properties of their effects, it might nevertheless be the case that these effects could be qualitatively of different sorts. If that were so, physical determinism would not entail psychological determinism, since, given two percepts of identical structure but diverse quality, we could not tell which would result from a stimulus known only as to its physical, i.e. structural, properties. This is an unavoidable consequence of the abstractness of physics. If physics is concerned only with structure, it cannot, per se, warrant inferences to any but the structural properties of events. Now it may be a fact that (e.g.) the structure of visual percepts is very different from that of tactual percepts; but I do not think such differences could be established with sufficient strictness and generality to enable us to say that such-and-such a stimulus must produce a visual percept, while such another must produce a tactual percept.