444 Mr DE morgan, on THE SYLLOGISM, No. V, AND 



ambiguity, and advanced in opposition to the preceding, I suppose it will be conceded that 

 the particular quantity, if free from the same ambiguity, will settle the matter. 



A logician strong in ancient association naturally tends towards the Latin singular : 

 modern habits tend towards consolidation of plurality of individuals into pieces of extent. 

 Hamilton compromises as follows (VI. 636"'). He wants to translate "Some dogs do not 

 bark" in fully quantified form. He does not say Quidam canes sunt nulla latrantia: this 

 would offend the logical ear. Neither does he say Quidam canis est nullum latrans: this 

 would be purely exemplar. He does say Quoddam caninum est nullum latrans: he speaks 

 singularly of an indefinite section of dog-nature, and so conciliates the ancient exemplarity of 

 phrase and the modern cumularity of thought. 



I take, beginning with Aristotle, a score or more of logicians of all ages, and of every 

 kind of note : I choose them merely because I happen to have access to them at the time of 

 writing. I go direct to the places in which the technical propositional forms are laid down, 

 and to the chapters on conversion. Some writers vary their phrases a little as they get deep 

 into their subjects : but we know that they would all desire that their systems should be 

 described by what they lay down in their fundamental explanations. Hamilton (VI. 626*) 

 has collected a large number of quantifying words both in Greek and Latin ; and might 

 have got more: but it would have been difficult to have found any early writer who heaped 

 his defining chapters with all this variety, or with any noteworthy amount of it. 



Aristotle {Analyt. Pr. cap. 1, &c.) defines the universal as that which belongs to every-one 

 or to no-one, to -jravTi rj nt]Sevi : the particular as that which belongs to some-one, or not 

 to some-one, or not to every-one, to tivI ^ /nrj tiv\ ij [xrj iravrt. Instead of a long quota- 

 tion from cap. 2, on conversion, I pick out all the quantitatives as they stand : they are (i.r]Se- 

 fi'ia, ovoev, wacra, ti, tk, ti, tivI, tivi, fxrioevi, ovoevl, Tivi, ixrioevl, ti, TravTt, tivI, fjit]cevi, 

 ovoevl, iravTt, tivi, tivi, /utjoevi, ovoevi, tivi, tivi, TravTi, ttuvti : not a plural among them. 

 If all this be not exemplar, it must be because Aristotle said one and meant many. But so 

 (by inference) does every person who says 'Any one man is some one animal' he means to 

 speak of all men, and he does it. So that in what sense soever Aristotle is not exemplar, 

 the exemplar system itself is not exemplar. Some will say that Aristotle only distributes: 

 then the exemplar system distributes; and that in modern use does not. 



Again, a person using cumular language would say that a universal negative is upset not 

 only by predication of all, but of some: he would never say that ' none are' is contradicted 

 by 'all are' and also by 'some one is'; he would certainly find intermediate room for the 

 indefinite plural some. Now Aristotle {Anal. Pr. cap. 26) says that the universal negative is 

 destroyed if the predicate be affirmable of Tray or some one, el iravTi koi el tivi : this must 

 be every-one or some one. He had previously said that the universal affirmative is upset if the 

 predicate can be said to belong to no one or not to some one; kuI yap ijv fjLtj^evl sal ^v tivi 

 fjLT] uirap-^ri, avrjprjTai [to kuOoXov KaTriyopiKoi/]. 



Hamilton, in various' places, appends to the word all the parenthesis " [or every]", thus 



' In one place (ix. ii. 303) there is a boldness of assertion 1 sameness of all and every which, but for repeated illustration, 

 which may be quoted as showing that genuine feeling of the I my readers would hardly believe to have existed. Alexander 



