210 ETHICAL ASPECTS OP EVOLUTION 



its place a branch of natural science. But success is impos 

 sible, at any rate with our present implements of thought. 

 These considerations are, I think, fatal to the special claim 

 which is sometimes set up for hedonism, that it may be 

 distinguished as a scientific system from other theories 

 of ethics. At the same time they are equally fatal to similar 

 claims on the part of any other system. No theory of human 

 conduct is, or can be made, a science, because the subject- 

 matter resists the application of scientific method. A 

 refutation of hedonism must rest on other grounds than 

 the demonstration that pleasures do not admit of measure 

 ment. That is a feature which pleasure has in common 

 with all other ethical standards. 



If we left the matter here, we should have stated only 

 a half-truth, and that half which, in a stage of thought like 

 the present, when an exaggerated value is attached to scien 

 tific demonstration, is, if not the least important, at any 

 rate the least necessary to insist upon. It is incumbent on 

 us to consider how this want of exactness, which is common 

 to all subjective inquiries, affects their value, whether 

 practical or speculative ; that is to say, the value of their 

 effects, and the certainty with which they impress us. 



Certainty, to begin with the second question, differs not 

 only in degree but in quality. It is, I think, a material 

 oversight to regard knowledge and belief as merely different 

 degrees of the same feeling. Religion, when it distinguishes 

 between faith and conviction, does not assert a mere difference 

 of degree, and still less that articles of faith are less cogent 

 than articles of conviction. It points to a distinction 

 which runs through the whole realm of thought, which is 

 often independent of intensity, and which a careful examina 

 tion of our use of the terms will show us to be intimately con 

 nected with the distinction between objective and subjective 

 series. That our common use always coincides with this 

 division cannot be asserted, nor could invariable consistency 

 be expected in so abstract a matter, but the distinction 

 is usually observed. We do not speak of our belief that 

 the angles of a triangle are equal to two right angles ; that 



