68 THE GRAMMAR OF SCIENCE 



— to hold in a region beyond our experience.^ We knozv 

 ourselves, and we know around us an impenetrable wall of 

 sense-impressions. There is no necessity, nay, there is 

 want of logic, in the statement that behind sense-impres- 

 sions there are " things -in -themselves " producing^ sense- 

 impressions. About this supersensuous sphere we may 

 philosophise and dogmatise unprofitably, but we can 

 never know usefully. It is indeed an unjustifiable ex- 

 tension of the term knowledge to apply it to something 

 which cannot be part of the mind's contents. What is 

 behind or beyond sense-impressions may or may not be 

 of the same character as sense-impressions, we cannot 

 say. We feel the siwface of a body to be hard, but its 

 core may be either hard or soft, we cannot say ; we can 

 only legitimately call it a hard-surfaced body. So it is 

 with sense-impressions and what may be behind them ; 

 we can only say sense-impression-stuff, or, as we shall 

 term it, with a somewhat divergent meaning from the 

 customary, sensation. By sensation we shall accordingly 

 understand that of which the only knowable side is sense- 

 impression. Our object in using the word sensation in- 

 stead of sense-impression will be to express our ignorance, 

 our absolute agnosticism, as to whether sense-impressions 

 are " produced " by unknowable " things-in-themselves," or 

 whether behind them may not be something of their own 

 nature.' The outer world is for science a world of sensa- 

 tions, and sensation is known to us only as sense- 

 impression. 



^ This will appear clearer when we have discussed the scientific meaning 

 oi cause and effctt. See Chapter IV. 



^ Herein lies the arid field of metaphysical discussion. Behind sense- 

 impressions, and as their source, the materialists place Matter; Berkeley 

 placed God ; Kant, and after him Schopenhauer, placed Will ; and Clifford 

 placed Alind-stiiff. Professor E. Mach in the paper referred to on p. 65 has 

 reduced the outer world to its known surface, sense-impression, which he terms 

 sensation — leaving no possible unknowable phis which we intend to signify 

 by our use of the word sensation. Such a theory cannot lead to scientific 

 error, but it does not seem a justifiable inference from sense-impression. The 

 variety of inferences cited above shows the quagmire which has to be avoided, 

 especially when the inferences are drawn with a view of influencing judgment 

 in the world of sense. 



