146 HUME vi 



fingers,, it is utterly impossible to conceive that 

 we have not two round objects under them; and, 

 though light is undoubtedly a mere sensation 

 arising in the brain, it is utterly impossible to 

 conceive that it is not outside the retina. In 

 the same way, he who touches anything with a 

 rod, not only is irresistibly led to believe that the 

 sensation of contact is at the end of the rod, but 

 is utterly incapable of conceiving that this sensa- 

 tion is really in his head. Yet that which is 

 inconceivable is manifestly true in all these cases. 

 The beliefs and the unbeliefs are alike necessary, 

 and alike erroneous.' 



It is commonly urged that the axiom of causa- 

 tion cannot be derived from experience, because 

 experience only proves that many things have 

 causes, whereas the axiom declares that all things 

 have causes. The syllogism, " many things which 

 come into existence have causes. A has come into 

 existence: therefore A had a cause," is obviously 

 fallacious, if A is not previously shown to be one 

 of the " many things." And this objection is per- 

 fectly sound so far as it goes. The axiom of causa- 

 tion cannot possibly be deduced from any general 

 proposition which simply embodies experience. 

 But it does not follow that the belief, or expecta- 

 tion, expressed by the axiom, is not a product of 

 experience, generated antecedently to, and alto- 

 gether independently of, the logically unjustifiable 

 language in which we express it. 



