One of the most learned and candid of the Aristotelian commentators — M. Barthélemy St. Hilaire — observes as follows (Logique d’Aristote, p. 435, Paris, 1838) respecting De Sophist. Elenchis:—
“Aristote va donc s’occuper de la marche qu’il faut donner aux discussions sophistiques: et ici il serait difficile quelquefois de décider, à la manière dont les choses sont présentées par lui, si ce sont des conseils qu’il donne aux Sophistes, ou à ceux qui veulent éviter leurs ruses. Tout ce qui précède, prouve, au reste, que c’est en ce dernier sens qu’il faut entendre la pensée du philosophe. Ceci est d’ailleurs la seconde portion du traîté.”
It appears to me that Aristotle intended to teach or to suggest both the two things which are here placed in Antithesis — though I do not agree with M. St. Hilaire’s way of putting the alternative — as if there were one class of persons, professional Sophists, who fenced with poisoned weapons, while every one except them refrained from such weapons. Aristotle intends to teach the art of Dialectic as a whole; he neither intends nor wishes that any learners shall make a bad use of his teaching; but if they do use it badly, the fault does not lie with him. See the observations in the beginning of the Rhetorica, i. p. 1355, a. 26, and the observations put by Plato into the mouth of Gorgias (Gorg. p. 456 E).
Even in the Analytica Priora (ii. 19, a. 34) (independent of the Topica) Aristotle says:—χρὴ δ’ ὅπερ φυλάττεσθαι παραγγέλλομεν ἀποκρινομένους, αὐτοὺς ἐπιχειροῦντας πειρᾶσθται λανθάνειν. Investigations of the double or triple senses of words (he says) are useful — καὶ πρὸς τὸ μὴ παραλογισθῆναι, καὶ πρὸς τὸ παραλογίσασθαι, Topica, i. 18, p. 108, a. 26. See also other passages of the Topica where artifices are indicated for the purpose of concealing your own plan of proceeding and inducing your opponent to make answer in the sense which you wish, Topica, i. 2, p. 101, a. 25; vi. 10, p. 148, a. 37; viii. 1, p. 151, b. 23; viii. 1, p. 153, a. 6; viii. 2, p. 154, a. 5; viii. 11, p. 161, a. 24 seq. You must be provided with the means of meeting every sort and variety of objection — πρὸς γὰρ τὸν πάντως ἐνιστάμενον πάντως ἀντιτακτέον ἐστίν. Topic. v. 4, p. 134, a. 4.
I shall again have to touch on the Topica, in this point of view, as founded upon and illustrating the Megaric logical puzzles ([ch. viii.] of the present volume).
Sophisms propounded by Eubulides. 1. Mentiens. 2. The Veiled Man. 3. Sorites. 4. Cornutus.
The Sophisms ascribed to Eubulidês, looked at from the point of view of logical theory, deserve that attention which they seem to have received. The logician lays down as a rule that no affirmative proposition can be at the same time true and false. Now the first sophism (called Mentiens) exhibits the case of a proposition which is, or appears to be, at the same time true and false.[49] It is for the logician to explain how this proposition can be brought under his rule — or else to admit it as an exception. Again, the second sophism in the list (the Veiled or Hidden Man) is so contrived as to involve the respondent in a contradiction: he is made to say both that he knows his father, and that he does not know his father. Both the one answer and the other follow naturally from the questions and circumstances supposed. The contradiction points to the loose and equivocal way in which the word to know is used in common speech. Such equivocal meaning of words is not only one of the frequent sources of error and fallacy in reasoning, but also one of the least heeded by persons untrained in dialectics; who are apt to presume that the same word bears always the same meaning. To guard against this cause of error, and to determine (or impel others to determine) the accurate meaning or various distinct meanings of each word, is among the duties of the logician: and I will add that the verb to know stands high in the list of words requiring such determination — as the Platonic Theætêtus[50] alone would be sufficient to teach us. Farthermore, when we examine what is called the Soritês of Eubulides, we perceive that it brings to view an inherent indeterminateness of various terms: indeterminateness which cannot be avoided, but which must be pointed out in order that it may not mislead. You cannot say how many grains are much — or how many grains make a heap. When this want of precision, pervading many words in the language, was first brought to notice in a suitable special case, it would naturally appear a striking novelty. Lastly, the sophism called Κερατίνης or Cornutus, is one of great plausibility, which would probably impose upon most persons, if the question were asked for the first time without any forewarning. It serves to administer a lesson, nowise unprofitable or superfluous, that before you answer a question, you should fully weigh its import and its collateral bearings.
[49] Theophrastus wrote a treatise in three books on the solution of the puzzle called Ὁ ψευδόμενος (see the list of his lost works in Diogenes L. v. 49). We find also other treatises entitled Μεγαρικὸς ά (which Diogenes cites, vi. 22), — Ἀγωνιστικὸν τῆς περὶ τοὺς ἐριστικοὺς λόγους θεωρίας — Σοφισμάτων ά, β — besides several more titles relating to dialectics, and bearing upon the solution of syllogistic problems. Chrysippus also, in the ensuing century, wrote a treatise in three books, Περὶ τῆς τοῦ ψευδομένον λύσεως (Diog. vii. 107). Such facts show the importance of these problems in their bearing upon logical theory, as conceived by the ancient world. Epikurus also wrote against the Μεγαρικοί (Diog. x. 27).
The discussion of sophisms, or logical difficulties (λύσεις ἀπορίων), was a favourite occupation at the banquets of philosophers at Athens, on or about 100 B.C. Ἀντίπατρος δ’ ὁ φιλόσοφος, συμπόσιόν ποτε συνάγων, συνέταξε τοῖς ἐρχομένοις ὡς περὶ σοφισμάτων ἑροῦσιν (Athenæus, v. 186 C). Plutarch, Non posse suaviter vivi secundum Epicurum, p. 1096 C; De Sanitate Præcepta, c. 20, p. 133 B.
[50] Various portions of the Theætêtus illustrate this Megaric sophism (pp. 165-188). The situation assumed in the question of Eubulidês — having before your eyes a person veiled — might form a suitable addition to the various contingencies specified in Theætêt. pp. 192-193.