320 DR DONALDSON, ON THE ORIGIN AND 



meaning. This logical application of argumentum is by far the most important function of 

 the term, and must be carefully examined. 



If we analyze any process of reasoning, we shall see that it resolves itself into the dis- 

 covery of some connecting link between the premises and the conclusion. The foundation of 

 all reasoning is the common notion that two things agree or disagree with one another accord- 

 ing as they agree or disagree with some third thing. Whether we reason syllogistically or 

 inductively, the test of our reasoning is the middle term, that is, the third idea which helps us 

 to form a judgment. Thus if I wish to assert that Uncle Tom and his fellows are reasonable 

 beings, I find the test of my reasoning in the middle term " man," and my argument runs 

 syllogistically: 



Omnes homines praediti sunt ratione; 

 Afri sunt homines; 

 Ergo, Afri sunt praediti ratione. 

 Or if I wished to prove inductively that all men are reasonable beings, I must find my middle 

 term in an examination of all the varieties of the human race, and my induction will run thus: 

 JEuropcei, Asiatici, Afri, Americani sunt ratione praediti ; 

 Sed Eur. As. Afr. Am. sunt homines ; 

 Ergo, omnes homines sunt praediti ratione. 

 Hence the art of topics is really the discovery of middle terms. And argumentation or the 

 dealing with argumenta, is the application of this test or rule, to prove whether this reasoning 

 rings sound or cracked, namely, to try whether the third term really does agree with the other 

 two. Now this is Cicero's definition of an argumentum. He says, Acad. i. 8 : " Argumenti • 

 conclusio, quae est Graece aVo^ej^ts, ita definitur : ratio, quae ex rebus perceptis ad id, quod 

 non percipiebatur, adducit," that is, by the discovery of the middle term or link of connexion 

 between the subject and predicate of the conclusion sought ; or, as he expresses it elsewhere, 

 Brutus, 32: "habere regulam, qua vera et falsa judicantur, et quae, quibus positis, essent, 

 quaeque non essent consequentia." The argumentum or test, which this regula applies, is 

 obviously unnecessary in those cases which do not require or admit of formal proof; and thus 

 Cicero says, de nat. Dear. m. 4 : " neque ego in causis, si quid est evidens, de quo inter omnes 

 conveniat, argumentari soleo." There can of course be no argumentum or middle term in a 

 truism ; but in all cases of ratiocinatio this rule must be applied, and the topica ars, or art 

 of discovering arguments, is, as Aristotle defines it, /meOo^ov evpelv a(p' ^9 ^vvyjaofieOa auWo- 

 •y'vCeaQai irepl TravTos tov TrporedevTos Trpop\i]ixaTos eg evoo^wv koi avTot Xoyov i/Treyoi/Tey 

 fir)6ti> epov/j-ev vnevavTiov. 



It will be observed that Aristotle says here e^ ei'^o^oji', and he means of course that the 

 purpose of his treatise is " to discover a method by which we shall be able to syllogize about 

 every proposed question from probabilities, and that when we ourselves sustain an argument, 

 we may not advance any thing that is contradictory." Readers of Aristotle do not require to 

 be reminded that the philosopher treats separately of the different kinds of syllogisms. For 

 while the Prior Analytics discuss the syllogism in general, the Posterior Analytics examine the 

 demonstrative syllogism, the Topics enlarge on the probable syllogism, and the Sophistical 

 Elenchi expose the captious or dishonest syllogism. The roTros, however, or " place," is pro- 



