DIVISION AND CLA SSI PICA TION. 1 2 3 



of division, will yield the best results in any particular depart 

 ment of science ; nor whether when we have made a number of 

 divisions and subdivisions, selecting at each step the basis we 

 have thought best, the results of our process are ex 

 haustive. What, then, ought to guide us in selecting the grounds 

 of our divisions ? Obviously, the end we have in view in making 

 the classification. The^wmz/aim of all classification is, of course, 

 to give us clear ideas, definite, well-ordered knowledge, control 

 over facts, increase of power in retaining and communicating 

 our knowledge about them. But every single department of facts 

 will be found to yield on investigation several widely distinct and 

 very special kinds of knowledge, in addition to what may be de 

 scribed as general knowledge of that department. And hence 

 we may have one or other of two possible purposes in approach 

 ing any sphere of classifiable data : we may wish to classify the 

 contents of the sphere in question with a view to obtaining some 

 special kind of knowledge about them, with this special object in 

 view ; or, without any such particular preoccupation, we may ap 

 proach it with a view to acquiring general knowledge, general 

 information about them. The former process is called Classi 

 fication for a Special Purpose, the latter, Classification for General 

 Purposes. 



When the physician classifies plants on the basis of their 

 medicinal properties, or the agriculturist on the basis of their 

 utility as food for animals, or of their suitability to different soils, 

 we have instances of classifications for special purposes ; while 

 the botanist s classification of plants is for the general purpose 

 of promoting our knowledge of their origin, nature, and general 

 relations and properties. 



These two kinds of classification are also called respectively 

 Artificial and Natural : because in the former we deliberately 

 group the objects (mentally) for a certain definite, arbitrarily 

 selected purpose, by means of a basis determined, if not even in 

 vented, by ourselves ; while in the latter we rather recognize differ 

 entiating attributes already existing in the facts, and so, instead 

 of inventing new mental divisions for practical purposes, we 

 discover existing classes, and so secure a better speculative know 

 ledge of the latter. 



Many modern logicians take exception to the use of the terms artificial 

 and natural in this context : pointing out that all classification is artificial 

 inasmuch as it is a voluntary arrangement of our ideas, not a segregation of 



