DEFINITION. 97 



for example, defy definition : we must suppose that these terms 

 refer to identical kinds of experience in all men s minds. Similarly, 

 the abstract names of simple attributes, such as redness, sweetness, 

 duty, goodness, cannot be defined, because although we may deal 

 with them after the analogy of concrete objects and so find a 

 genus for them (36) colour, taste, moral feeling or intuition, virtue, 

 in the examples given, we can find no differentiae that will 

 convey any intelligible notion of them to a mind not otherwise 

 acquainted with them. 



The Scholastics regarded classes lower than the species infimae (46) as in 

 capable of definition in the strict sense per genus et differentiam, because 

 they refused to recognize the sub-classes and their distinguishing attributes as 

 species and differentiae. There does not seem, however, to be any sufficient 

 reason for refusing these titles, in their modern logical meaning, to such 

 sub-classes and attributes, or for declining to regard as a definition of negro, 

 for example, the statement that he is a black man. So, also, if we set forth 

 the implication of a significant individual name (28) by assigning its infima 

 species as genus and its group of individualizing properties and accidental 

 attributes (47, 48) as differentia, we may not unreasonably call such a statement 

 a definition. 



(c) A little reflection will convince us that while technical, 

 scientific terms and notions such as triangle, artery, planet, 

 volition, etc., in which the implicational side of the meaning, the 

 reference to attributes, is more prominent lend themselves more 

 easily to definition, the names and concepts of familiar objects 

 such as book, dog, house in which the applicational reference 

 entirely overshadows the connotation are difficult, if not practic 

 ally impossible, to define. 



53. EXEMPLIFICATION OR EXTENSIVE DEFINITION OR DE 

 FINITION BY TYPE. Now, in all the cases just enumerated (52, 

 a, b, c], since the concept or term has an object, be this an 

 attribute or a thing, there ought to be some means of securing 

 between different minds identity of reference to such objects. 

 The process of definition, so far described, is concerned directly 

 and primarily with the connotation of our ideas and terms. It 

 thereby indirectly fixes their reference to things, since denotation 

 follows connotation. What, then, about non-connotative and 

 simple abstract terms ? How is identity of application, and of im 

 plication, respectively, to be secured in these ? And are there not, 

 perhaps, ideas and terms whose application to objects can be more 

 easily and directly fixed than by appealing in the first instance 

 VOL. I. 7 



