AND ON LOGIC IN GENERAL. 201 



4. Terminal precision, metaphysical view. All the schetical words yet used might be 

 called predicables : but it will be historically more convenient to confine that name to the set 

 now before us. I gave them in the Philological Transactions, Vol. vi. No. 129 (1853). The 

 predicables of Aristotle savour strongly of the metaphysical view. I here use the adjectives 

 generic and specific, which, being adjectives, have retained a metaphysical use, and that even 

 with reinforcement. In treating exclusion as well as inclusion, in a definite universe, and with 

 terminal precision, we require three substantive ideas, all attributive, or denoting attributes, 

 though that word is technically applied only to one. There is the attribute, said of all ; 

 the accident, of some and some only (which is therefore non-accident as to what it cannot 

 be said of, the term being equally positive and negative) ; and the evcludent, or excluse, 

 said of none. Each of these exists in a three-fold distinction ; universal, generic, and specific. 



Looking at the notion predicated of, say X, in extension, a universal predicable is one 

 which applies in the same manner both to X and to x, attribute of both, accident (= non- 

 accident) of both, excludent of both. These are of no value in syllogism. The universal 

 attribute and universal excludent are helps to definition, positive or negative, of the universe : 

 the universal accident, which is independent, inessential, irrepugnant, and inalternative, to 

 both X and x, is what we should suppose of an attribute taken at hazard, with reference to 

 a notion taken at hazard. A predicable is generic, when it applies to wider genera of the 

 subject notion, in all the additional extent of those genera : it is specific, when it does not apply 

 to any wider genera, nor gain any new extent of application from the additional extent of the 

 wider genus. Here the subject notion has been treated as a class, or the proposition has been 

 made logico-physical : it will be nearly as easy to treat the subject as an attribute, as well as 

 the predicate. 



XVIII. On the various names which one relation receives, the following remarks may be 

 made. Since every reading in one view has its reading in the other, the mathematical names 

 will so often and so easily pass into the metaphysical, and vice versa, that it is almost as if we 

 gave all the schetical words of either reading to both. To put different, but concomitant, 

 notions under one name is clearness, or at least facility, at the beginning of a subject : but the 

 more progress the more confusion, unless it be prevented by checking development. To put 

 them under different names is confusion, or at least difficulty, at the commencement : but the 

 more progress, the more clearness, and development without confusion. The question which 

 it concerns me to answer, as to the number of names, goodness apart, is but this : — Have 

 I made any divisions in language which are not divisions in thought ? If the answer be in 

 the negative, there is an end of all objection. If logic must take account of n real dicho- 

 tomies, there may be 2" desirable distinctions to draw : in fact, 1, 2, 4, 8, &c. is the logician's 

 progression, rather than 1, 2, S, 4, &c. 



The want of a distinct name for a distinct notion may affect the mode of thought, must 

 affect the mode of expression, of educated men. Shortly after I had first distinguished by name 

 the supercontrary from the contrary, I happened to look at several pieces of controversial 

 theology, written by Oxonians versed in the common logic and distinguished by the name of 

 Tractarians. Their universe was the Anglo-Irish Christian world, consisting of churchmen 

 and heretics, variously named. I was amused by the frequent protection from difficulty which 

 Vol. X. Paet I. 26 



