2 SCIENTIFIC METHOD 



ticular by what was called by Aristotle syllogism, but 

 more often nowadays deduction. 



These three kinds of inference agree in consisting of 

 judgements or beliefs ; in proceeding from judgements as 

 premisses to another judgement as conclusion ; and in 

 performing this process by a perception of similarity 

 between what is in the premisses and what is in the 

 conclusion. Analogy concludes from one particular being 

 like another, induction from a class being like its in- 

 stances, and deduction from any member being like all 

 the members of a class. Inference in general then is 

 proceeding from judgement to judgement by means of 

 similarity, or what is often called ' parity of reasoning '. 



At the same time, there are minor differences between 

 the kinds of inference. Analogy requires nothing but 

 similar particulars. It is the kind of inference which 

 may be performed by the higher animals, and by young 

 children, without having risen to the universal, without 

 any general apprehension of a whole class, and with- 

 out any general terms or rational language. Mill did 

 an important service in insisting on this inference from 

 particular to particular, by which a dog or an infant will 

 argue from one meal to another, without understanding 

 the whole class of meals, or expressing that class by any 

 general term. On the other hand, we must distinguish 

 more sharply than he did between this particular inference 

 of analogy and the general inferences of induction and 

 deduction, which involve the apprehension of whole 

 classes, and are bound up with the use of general 

 language or rational discourse. Induction and deduc- 

 tion imply generalization. They are reasoning. They 

 are the privileges of man, who alone in the animal world 

 uses the word ' all ', who alone generalizes, or has reason 

 proper. Beasts infer : man reasons. 



