SCIENCE AND THE UNOBSERVABLE DINGLE 217 



physically and not only practically unobservable. And, on the other 

 hand, when we say that absolute motion is physically unobservable, 

 we are again assuming omniscience. We cannot observe such motion 

 by optical, acoustical, electromagnetic or any other means within our 

 present knowledge, but, without the assumption, who can say that 

 there is not some undiscovered physical medium through which it 

 may be detected? If there is, absolute motion is merely practically, 

 and not physically, unobservable. It is clear, I think, that unless we 

 are omniscient the two classes are indistinguishable. 



Our analysis of unobservability, then, finally brings us to this. If 

 we assume that we are omniscient we can distinguish three classes — the 

 practically, the physically, and the logically unobservable. If we do 

 not assume that we are omniscient we can distinguish only two 

 classes — the actually and the logically unobservable, let us call them. 

 The importance of this conclusion for our purpose is this. We are 

 going to look at the actual practice of physics, to see what kinds of 

 unobservable are excluded and what kinds are not. If we find that a 

 distinction is made between the practically and the physically unob- 

 servable, then we know that physics is assuming omniscience; but if 

 no distinction is made, then there is no such assumption. 



Let me state the result at once, afterward giving examples to 

 justify the statement before proceeding to consider the validity of the 

 principle we are considering in its definite form. The practice of 

 physics is to recognize three classes — the practically, the physically, 

 and the logically unobservable. Of these it excludes the physically 

 and the logically unobservable from its considerations, and aims at 

 describing the universe in terms of the observable and the merely 

 practically unobservable only. It thus assumes omniscience, in the 

 sense in which I have defined the word. 



It will not take us long to see that physics includes the practically 

 and excludes the logically unobservable. No physicist denies that the 

 moon has a far side in the same sense as it has a near side. We 

 assume without question that the earth has an interior, that there are 

 stars outside the range of our telescopes, and regions beyond the 

 obscuring clouds of the Milky Way. All these things could be 

 observed if known means of observation have precisely the properties 

 we believe them to have and we had the skill to make full use of them. 

 Hence the practically unobservable is admitted to physical theory. 



We may deal equally summarily with the logically unobservable. 

 Nasty things have been said about the reasoning of some modern 

 physicists when they step outside the bounds of their equations, but 

 I do not think the bare, unadorned physical theories themselves have 

 been called illogical, either with pride or with shame. If, then, the 

 structure of physical theory allows ontological significance to any- 

 thing which is logically unobservable, it does so through an oversight, 



