4. A fourth locus is, where the definiend admits both of a better and a worse construction, and where the definition enunciates only the worse. You may impugn it, on the ground that every cognition and every power must be understood as tending to its best results.[286]
[286] Ibid. p. 143, a. 9.
6. A fifth locus is, where the definiend is enunciated as ranking, not in the lowest and nearest species to which it belongs but, in some higher and more distinct genus. Here the real essence will not be declared, and the definition will thus be incomplete; unless indeed it includes, along with the highest genus, the superadded mention of all the differentiæ descending down to the lowest species. It will then be complete, because it will include, in circumlocutory phrase, all that would be declared by enunciating the specific name.[287]
[287] Ibid. a. 15-28.
6. Assuming the genus to be truly declared in the definition you will examine whether the differentiæ enunciated are differentiæ at all? whether they really belong to the definiend? what is it which they serve to contrast with and exclude, — since, if there be nothing such, they cannot be truly differentiæ? whether the differential term and its counter-differential apply to and cover the whole genus? whether, granting the differentia to be real, it be such, when taken along with the genus, as to constitute a true species, and whether its counter-differentia be such also? This is a locus furnishing many possibilities of impugning the definition.[288]
[288] Topic. VI. vi. p. 143, a. 29-b. 10.
7. Perhaps the definition may enunciate a differentia which is merely negative; e.g., A line is length without breadth. If you are debating with a respondent who holds the (Platonic) doctrine of Ideas, and who considers each Idea or genus to be something numerically one, distinct from all its participants, you will find here a locus for attacking them.[289] He asserts the existence of a Self-long or generical long, a Self-animal or generic animal, each numerically one. Now, upon this hypothesis, since of all long you may predicate either in the affirmative or the negative (i.e., either it is broad or it is not broad), so this alternative may be predicated of the Self-long or generical long; and thus the genus will coincide with, or fall under the definition of, one among its own species. Or, if this be denied, it will follow that the generic long must be both broad and not broad; which is a contradiction still more inadmissible. Accordingly, against one who holds the doctrine of Ideas, declaring the genus to be unum numero, the negative differentia will furnish grounds for attack; but not against any other respondent.[290] For there are various cases in which the negative must be employed as a part of the differentia: e.g., in privative terms, blind is one whose nature it is to see but who does not see. And, even when the differentia enunciated is affirmative, it may have for its condivident member only a negative term, e.g., length having-breadth has for its condivident member only the negative, length not-having-breadth.[291]
[289] Ibid. b. 11-30.
[290] Ibid. b. 29: ὥστε πρὸς ἐκείνους μόνους χρήσιμος ὁ τόπος, ὅσοι τὸ γένος ἓν ἀριθμῷ φασὶν εἶναι. τοῦτο δὲ ποιοῦσιν οἱ τὰς ἰδέας τιθέμενοι· αὐτὸ γὰρ μῆκος καὶ αὐτὸ ζῷον γένος φασὶν εἶναι.
[291] Ibid. b. 33.