THE FACTS OF SCIENCE 59 



^ 10. — TJie Cations of Legitimate Inference 



We cannot here discuss more fully the limits of belief 

 and legitimate inference. We shall, however, to some 

 extent return to the subject when considering Causation 

 and Probability in Chapter IV. But it may not be with- 

 out service to state certain canons of legitimate inference 

 with a few explanatory remarks, leaving the reader, if he 

 so desire, to pursue the subject further in Stanley Jevons' 

 Principles of Science, or in Clifford's essay on The Ethics 

 of Belief We ought first to notice that the use of the 

 word belief in our language is changing : formerly it 

 denoted something taken as definite and certain on the 

 basis of some external authority ; now it has grown 

 rather to denote credit given to a statement on a more or 

 less sufficient balancing of probabilities.^ 



The change in usage marks the gradual transition of 

 the basis of conviction from uncriticising faith to weighed 

 probability. The canons we have referred to are the 

 following : — 



I. Where it is impossible to apply man's reason, that is 

 to criticise and investigate at all, there it is not only un- 

 profitable but anti-social to believe. 



Belief is thus to be looked upon as an adjunct to 

 knowledge, as a guide to action where decision is needful, 

 but the probability is not so overwhelming as to amount 

 to knowledge. To believe in a sphere where we cannot 

 reason is anti-social, for it is a matter of common ex- 

 perience that such belief prejudices action in spheres 

 where we can reason. 



We cannot demonstrate that a man without a backbone may not exist " out- 

 side " the physical universe, only he would not be a man and would exist 

 " nowhere." The existence of something of which we can postulate nothing 

 at nowhere can never be legitimately inferred from conceptions based on 

 sense-impressions. Such a man would be like Meister Eckehart's deity, who 

 was a non-god, a non-spirit, a non-person, a non-idea, and of whom, he 

 says, any assertion must be more false than true. 



1 Compare the older use in Biblical passages, such as "Jacob's heart 

 fainted for he believed them not," and " Except ye see signs and wonders ye 

 will not believe," or in Locke's definition of belief as adherence to a proposition 

 of which one is persuaded but does not know to be true, with such modern 

 usage as " I believe that you will find a cab on the stand, and that the 

 train starts at half-past eight." 



