io8 THE SCIENCE OF LOGIC. 



hull, masts, sails, rigging, etc.&quot; ; &quot; Man is a being composed of a 

 rational soul and an organized body &quot;. 



The so-called &quot; definitions&quot; to be found in dictionaries belong 

 for the most part to one or other of those classes of substitutes : 

 they are descriptions, or at best distinctive explanations, rather 

 than definitions. And the same is largely true of what passes 

 for definition in literature, prose as well as poetry. It is only 

 in scientific works we may hope to find the precision of thought 

 which it is the aim of logical definition to secure. 



Although it is not the duty of logic, but of each special 

 science in its own department, to construct definitions, it is the 

 duty of logic to control the process, and to secure that all de 

 finition contribute to clear and distinct thought by prescribing 

 certain general canons or rules with which definition should aim 

 at conforming. 



57. RULES OF DEFINITION. The following rules, tradition 

 ally set forth in logical treatises, give the conditions which a good 

 definition should fulfil, and at the same time the characteristics 

 which serve to distinguish an accurate from a defective defini 

 tion : 



/. A definition should contain neither more nor less than the 

 connotation of the term to be defined ; or, should be coextensive with 

 the thing to be defined. 



II. It should be in itself clearer and simpler than the thing to 

 be defined, and, therefore, should not contain ambiguous or figurative 

 language. 



III. It should not contain terms synonymous with the name of 

 the thing to be defined. 



IV. It should as far as possible be expressed by positive rather 

 than by negative terms. 



Briefly, then, a definition should be (i) adequate and precise, 

 (2) clear and simple ; and it should not be (3) tautologous, or (4) 

 negative. 



A very brief word of explanation will suffice for each of these 

 rules. 



Rule I. Should the definition contain, in addition to the 

 differentia, a proprium or an inseparable accident, this will not in 

 any way unduly limit the denotation of the term, but it may 

 mislead by suggesting the existence of a really non-existent class 

 of things. To define &quot; equilateral triangle&quot; as &quot;A triangle 

 having three equal sides and three equal angles&quot; suggests the 



