79 b 20ληφθέντος τοῦ στερητικοῦ γίγνεται συλλογισμός· ἀμφοτέρων δ᾿ ἀποφατικῶν οὐσῶν οὐκ ἔσται.

Φανερὸν οὖν ὅτι ἐνδέχεταί τε ἄλλο1 ἄλλῳ μὴ ὑπάρχειν ἀτόμως, καὶ πότ᾿ ἐνδέχεται καὶ πῶς εἰρήκαμεν.

XVI. Ἄγνοια δ᾿ ἡ μὴ κατ᾿ ἀπόφασιν ἀλλὰ κατὰ διάθεσιν λεγομένη ἔστι μὲν ἡ διὰ συλλογισμοῦ γιγνομένη 25ἀπάτη, αὕτη δ᾿ ἐν μὲν τοῖς πρώτως ὑπάρχουσιν ἢ μὴ ὑπάρχουσι συμβαίνει διχῶς· ἢ γὰρ ὅταν ἁπλῶς ὑπολάβῃ ὑπάρχειν ἢ μὴ ὑπάρχειν, ἢ ὅταν διὰ συλλογισμοῦ λάβῃ τὴν ὑπόληψιν. τῆς μὲν οὖν ἁπλῆς ὑπολήψεως ἁπλῆ ἡ ἀπάτη, τῆς δὲ διὰ συλλογισμοῦ πλείους. μὴ ὑπαρχέτω γὰρ τὸ Α μηδενὶ 30τῷ2 Β ἀτόμως· οὐκοῦν ἐὰν συλλογίζηται ὑπάρχειν τὸ Α τῷ Β, μέσον λαβὼν τὸ Γ, ἠπατημένος ἔσται διὰ συλλογισμοῦ. ἐνδέχεται μὲν οὖν ἀμφοτέρας τὰς προτάσεις εἶναι ψευδεῖς, ἐνδέχεται δὲ τὴν ἑτέραν μόνον. εἰ γὰρ μήτε τὸ Α μηδενὶ τῶν Γ 35ὑπάρχει μήτε τὸ Γ μηδενὶ τῶν Β, εἴληπται δ᾿ ἑκατέρα ἀνάπαλιν, ἄμφω ψευδεῖς ἔσονται. ἐγχωρεῖ δ᾿ οὕτως ἔχειν τὸ Γ πρὸς τὸ Α καὶ Β ὥστε μήτε ὑπὸ τὸ Α εἶναι μήτε καθόλου τῷ Β. τὸ μὲν γὰρ Β ἀδύνατον εἶναι ἐν ὅλῳ τινί (πρώτως γὰρ ἐλέγετο αὐτῷ τὸ Α μὴ ὑπάρχειν), τὸ δὲ Α οὐκ ἀνάγκη πᾶσι 40τοῖς οὖσιν εἶναι καθόλου, ὥστ᾿ ἀμφότεραι ψευδεῖς. ἀλλὰ καὶ τὴν ἑτέραν ἐνδέχεται ἀληθῆ λαμβάνειν, οὐ 80 aμέντοι ὁποτέραν ἔτυχεν, ἀλλὰ τὴν ΑΓ· ἡ γὰρ ΓΒ πρότασις ἀεὶ ψευδὴς ἔσται διὰ τὸ ἐν μηδενὶ εἶναι τὸ Β, τὴν δὲ ΑΓ ἐγχωρεῖ, οἷον εἰ τὸ Α καὶ τῷ Γ


Posterior Analytics, I

is assumed in connexion with either of them, but when both are negative there will be no syllogism.

Thus it is evident that one term may not-apply atomically to another; and we have explained when and how this is possible.

XVI. Ignorance, considered not in a negative sense Error in respect of terms immediately related. but as a positive disposition of mind, is error reached through inference.a In propositions stating an immediate positive or negative relation it arises in two ways: (a) when we directly supposeb that one term applies or does not apply to another, and (b) when we reach this supposition by inference. The error arising from direct supposition is simple, but that which is based on inference takes more than one form. Let (1) Negative relation. A apply atomically to no B. Then if we infer, taking C as the middle, that A applies to B, our error will be based on inference. It is possible either for both premisses or for one only to be false, (i) For if A (i) Both premisses false. applies to no C and C to no B, and we have assumed the contrary in each case, both premisses will be false (it is possible for C to be so related to A and B that it neither falls under A nor applies universally to B. For B cannot be wholly contained in a genus, since we stated abovec that A is directly inapplicable to it; and A need not necessarily apply universally to everything: hence both premisses are false.) (ii) (ii) Major true, minor false. It is also possible to assume one true premiss: not either premiss indifferently, but AC (the premiss CB will always be false, because B is contained in no genus; but AC may be true); e.g., if A applies

DOI: 10.4159/DLCL.aristotle-posterior_analytics.1960