SECTION 26 DEDUCTION. 369 



SECTION XXVI. DEDUCTION. 



CONCLUSION 31. 



Need of Strenuous Mental Application in the Process of Deduc- 

 tion, and need of the Deductions being Graded, Comprehensive, 

 Important, Numerous, Full, Rational and Relevant, Original, 

 Automatically Initiated, and Methodically Developed. 



194. In generalising facts we seek for such similarities 

 as might lead us to formulate a truth larger than the facts we 

 set out with initially. In the descending or deductive method 

 we search also for resemblances, but of a more restricted order 

 than the point of departure of the deduction. In the former 

 process we begin normally with facts; in the latter invariably 

 with a statement. 



Concurrently certain methodological differences between 

 generalisation and deduction require to be elucidated. Roughly 

 speaking, in generalising we mechanically affirm of a whole 

 class what we had observed in a section thereof. E.g., for the 

 word some, we place the word all. If we reversed the process, 

 and for the word all, placed the word some, and called this 

 deduction, we should be trifling in a serious matter. In the 

 first case, we should have a statement of some consequence; 

 in the second, one of no moment. Deduction, therefore, argues 

 a movement which is not methodologically self-evident, as is 

 the movement in generalisation. E.g., "Socrates is mortal", 

 is not a self-evident conclusion from "All men are mortal", 

 for "Plato is mortal", "The Phrygians are mortal", would have 

 been as appropriate. Whilst in generalising there is but one 

 step from many particulars to one general; in deduction there 

 may be innumerable steps from the one general to the many 

 particulars. Once more we see, therefore, that there is a 

 profound distinction between generalising and what we may 

 term particularising. At the same time we should note that in 

 the verifying of certain generalisations, we had to proceed de- 

 ductively, precisely as if our object were to elicit new truth 

 from an established generalisation. In this sense, deduction 

 may be regarded as an auxiliary process in the establishment of 

 a generalisation. Yet we should not exaggerate the difficulties 

 inherent in the deductive process. As in verifying an ordinary 

 generalisation we are greatly assisted by the thorough know- 

 ledge of our subject as a whole, so in deducing we depend to 

 a decisive degree on our intimate acquaintance with the body 

 of truths involved. Total ignorance would spell operating in 

 a mental vacuum. 



There are at least two conditions controlling deductive proce- 

 dure. First, an induction may not be full, that is, only a general 

 statement accompanied by few particulars may have been 

 published ; in which case we may deduce the important statement 



24 



