This dialectic debate, which Aristotle found current at Athens, he tries in the Topica to define and reduce to system. The dialectician must employ Syllogism; and we are first taught to distinguish the Syllogism that he employs from others. The Dialectic syllogism is discriminated on one side from the Demonstrative, on the other from the Eristic (or litigious); also from the scientific Paralogism or Pseudographeme. This discrimination is founded on the nature of the evidence belonging to the premisses. The Demonstrative syllogism (which we have already gone through in the Analytica Posteriora) has premisses noway dependent upon opinion: it deduces conclusions from true first principles, obtained by Induction in each science, and different in each different science. The Dialectic syllogism does not aspire to any such evidence, but borrows its premisses from Opinion of some sort; accredited either by numbers, or by wise individuals, or by some other authoritative holding. As this evidence is very inferior to that of the demonstrative syllogism, so again it is superior to that of the third variety — the Eristic syllogism. In this third variety,[8] the premisses do not rest upon any real opinion, but only on a fallacious appearance or simulation of opinion; insomuch that they are at once detected as false, by any person even of moderate understanding; whereas (according to Aristotle) no real opinion ever carries with it such a merely superficial semblance, or is ever so obviously and palpably false. A syllogism is called Eristic also when it is faulty in form, though its premisses may be borrowed from real opinion, or when it is both faulty in form and false in the matter of the premisses. Still a fourth variety of syllogism is the scientific Paralogism: where the premisses are not borrowed from any opinion, real or simulated, but belong properly to the particular science in which they are employed, yet nevertheless are false or erroneous.[9]
[8] Topic. I. p. 100, b. 23: ἐριστικὸς δ’ ἔστι συλλογισμὸς ὁ ἐκ φαινομένων ἐνδόξων, μὴ ὄντων δέ, καὶ ὁ ἐξ ἐνδόξων ἢ φαινομένων ἐνδόξων φαινόμενος. οὐ γὰρ πᾶν τὸ φαινόμενον ἔνδοξον καὶ ἔστιν ἔνδοξον. οὐθὲν γὰρ τῶν λεγομένων ἐνδόξων ἐπιπόλαιον ἔχει παντελῶς τὴν φαντασίαν, καθάπερ περὶ τὰς τῶν ἐριστικῶν λόγων ἀρχὰς συμβέβηκεν ἔχειν· παραχρῆμα γὰρ καὶ ὡς ἐπὶ τὸ πολὺ τοῖς καὶ μικρὰ συνορᾶν δυναμένοις κατάδηλος ἐν αὐτοῖς ἡ τοῦ ψεύδους ἐστὶ φύσις.
[9] Ibid. i. p. 101, a. 5-17.
Upon the classification of syllogisms here set forth by Aristotle, we may remark that the distinction between the Demonstrative and the Dialectic is true and important; but that between the Dialectic and the Eristic is faint and unimportant; the class called Eristic syllogisms being apparently introduced merely to create a difference, real or supposed, between the Dialectician and the Sophist, and thus to serve as a prelude to the last book of this treatise, entitled Sophistici Elenchi. The class-title Eristic (or litigious) is founded upon a supposition of dishonest intentions on the part of the disputant; but it is unphilosophical to make this the foundation of a class, and to rank the same syllogism in the class, or out of it, according as the intentions of the disputant who employs it are honest or dishonest. Besides, a portion of Aristotle’s definition tells us that the Eristic syllogism is one of which the premisses can impose upon no one; being such that a very ordinary man can at once detect their falsity. The dishonest disputant, surely, would argue to little purpose, if he intentionally employed such premisses as these. Lastly, according to another portion of Aristotle’s definition, every syllogism faulty in form, or yielding no legitimate conclusion at all, will fall under the class Eristic, and this he himself in another place explicitly states;[10] which would imply that the bad syllogism must always emanate from litigious or dishonest intentions. But in defining the Pseudographeme, immediately afterwards, Aristotle does not imply that the false scientific premiss affords presumption of litigious disposition on the part of those who advance it; nor does there seem any greater propriety in throwing all bad dialectic syllogisms under the general head of Eristic.
[10] Topic. VIII. xii. p. 162, b. 4.
The dialectician, then, will carry on debate only by means of premisses sustained by real opinion; which not only always carry some authority, but are assumed as being never obviously fallacious; though often inconsistent with each other, and admitting of argumentation pro and con. These are what Aristotle calls Endoxa; opposed to Adoxa, or propositions which are discountenanced, or at least not countenanced, by opinion, and to Paradoxa (a peculiar variety of Adoxa),[11] or propositions which, though having ingenious arguments in their favour, yet are adverse to some proclaimed and wide-spread opinions, and thus have the predominant authority of opinion against them.
[11] Ibid. I. xi. p. 104, b. 24: περὶ ὧν λόγον ἔχομεν ἐναντίον ταῖς δόξαις.
Of these three words, Paradox is the only one that has obtained a footing in modern languages, thanks to Cicero and the Latin authors. If the word Endox had obtained the like footing, we should be able to keep more closely to the thought and views of Aristotle. As it is, we are obliged to translate the Greek Endoxon as Probable, and Adoxon as Improbable:[12] which, though not incorrect, is neither suitable nor exactly coincident. Probable corresponds more nearly to what Aristotle (both in this treatise and in the Analytica) announces sometimes as τὸ ὡς ἐπὶ τὸ πολύ — that which happens in most cases but not in all, as distinguished from the universal and necessary on one side, and from the purely casual on the other;[13] sometimes, also, as τὸ εἰκός or τὸ σημεῖον. Now this is a different idea from (though it has a point of analogy with) the Endoxon: which is not necessarily true even in part, but may be wholly untrue; which always has some considerations against it, though there may be more in its favour; and which, lastly, may be different, or even opposite, in different ages and different states of society. When Josephus distinguished himself as a disputant in the schools of Jerusalem on points of law and custom,[14] his arguments must have been chiefly borrowed from the Endoxa or prevalent opinions of the time and place; but these must have differed widely from the Endoxa found and argued upon by the contemporaries of Aristotle at Athens. The Endoxon may indeed be rightly called probable, because, whenever a proposition is fortified by a certain body of opinion, Aristotle admits a certain presumption (greater or less) that it is true. But such probability is not essential to the Endoxon: it is only an accident or accompaniment (to use the Aristotelian phrase), and by no means an universal accompaniment. The essential feature of the Endoxon is, that it has acquired a certain amount of recognition among the mass of opinions and beliefs floating and carrying authority at the actual time and place. The English word whereby it is translated ought to express this idea, and nothing more; just as the correlative word Paradox does express its implication, approached from the other side. Unfortunately, in the absence of Endox, we have no good word for the purpose.
[12] Aristotle gives a double meaning of ἄδοξον (Topic. VIII. ix. ix. 160, b. 17):— 1. That which involves absurd or strange consequences (ἄτοπα). 2. That which affords presumption of a bad disposition, such as others will disapprove — οἷον ὅτι ἡδονὴ τἀγαθὸν καὶ τὸ ἀδικεῖν βέλτιον τοῦ ἀδικεῖσθαι.
[13] Topic. II. vi. p. 112, b. 1: ἐπεὶ δὲ τῶν πραγμάτων τὰ μὲν ἐξ ἀνάγκης ἐστί, τὰ δ’ ὡς ἐπὶ τὸ πολύ, τὰ δ’ ὁπότερ’ ἔτυχεν, &c. Compare also Analyt. Post. I. xxx., et alib.