Question Twenty-Eight
[QUESTION TWENTY-EIGHT
Whether it is Necessary in the Definition of One Relative to Posit Its Correlative]
t is asked whether it is necessary in the definition of
one relative to posit its correlative.
1. It seems that it is not:
Since everything that ought to be posited in the definition is prior and better known, as a genus or differentia with respect to <what is being> defined; but with respect to <a relative> being defined, it is related to its correlative in none of these ways; therefore, etc. The minor is evident with respect to its two primary members through the argument to the opposite in the preceding question.1
2. Second, every definition indicates the essence of the defined; but one relative is not of the essence of the other, since the being of a relative is "to have itself to another"; "another," however, expresses a diversity in the essence; therefore, etc.
3. Third, in the definition of one opposite, the other <opposite> ought not be posited; relatives are opposites; therefore, etc. Proof of the major: first, since contradictories follow all opposites, and in the definition of one contradictory the other <contradictory> ought not be posited; therefore, neither
<ought they be posited> in other opposites.2 The implication is evident, since the consequent does not pertain to the definition of any <thing>, nor does the antecedent.—Second, since any
1. Cf. supra, q. 27, n. 1.
2. Cf. infra, n. 15, where Scotus claims that the form of this argument is not valid.
JOHN DUNS SCOTUS
part of a definition can be predicated of that about which the defined is also predicated, according to Aristotle in the third consideration of Bk. II of the Topics.3 Therefore, if in the definition of one opposite the other <opposite> is posited, two opposites would be predicated of the same <thing>, which is untenable. The minor is evident in that chapter <which begins with>
"as often, however, as it is accustomed to be opposed,"4 where Aristotle enumerates "relative opposition"5 as one species of opposition, and exemplifies <this species>, as "double," "half," giving us to understand that this follows in all relatives.
4. If it is said that the major is true only in opposites other than relatives, against this:
Contraries and contradictories are opposed <to one another> relatively, since a contrary is said contrary to a contrary; therefore, if one contrary is not defined through the other, not every relative will be defined through its correlative.
5. Likewise, "every differentia according to its species is a contrariety," according to Aristotle in Bk. X of the Metaphysics;6 but relative opposites differ in species; therefore, they are contraries. Therefore, if the major is true about contraries, <then> it will also be true about relative opposites.
6. Fourth, a name and its definition signify the same <thing>, according to Aristotle in Bk. IV of the Metaphysics;7 but the name of a relative does not signify its correlative, since then by expressing the one with the other there would be a vain repetition; therefore, neither ought the other be posited in the definition.
7. Fifth, any definition of a relative is good when its correlative is not posited, as this <definition>: "species is that which is predicated of many <things> differing in number, etc.,"8 where
3. Topics, Bk. II, Ch. 2 (110a 5–9).
4. Categories, Ch. 10 (11b 16 [in the apparatus]).
5. Categories, Ch. 10 (11b 25–26).
6. Metaphysics, Bk. X, Ch. 8 (1058a 15–17).
7. Metaphysics, Bk. IV, Ch. 7 (1012a 23–25).
8. Isagoge, Ch. "On Species" (AL I6 9; ed. Busse, 4.11–12).
QUESTION 28
genus is not posited to that with respect to which the species is relatively said; therefore, it is not always necessary for a relative to be defined through its correlative.
8. Sixth, all the reasons that are applied to the preceding question,9 which <insofar as they> prove that one relative can exist without the other, prove that one must not be defined through the other; for nothing is defined through that without which it can exist.
9. To the opposite is Porphyry in the chapter "On Species," who says that a genus is said <with respect> to its species and conversely:10 "Therefore, it is necessary in the accounts of each to use each of the two."
10. Likewise, Aristotle11 says here that it is impossible for one relative to be known definitely unless the other is known; but every non-complex <term> is known definitely when it is known through its own definition; therefore, it is impossible for one relative to be known through its definition unless the other is known; therefore, one must be posited of necessity in the definition of the other.
[I. The Response Of Scotus] 11. This can be conceded, since the essential being of a relative is "to have itself to another"; however, the definition indicates the essential being of the defined; therefore, the definition of a relative ought to indicate that it has itself to another, which cannot be unless the correlative of its defined is posited in it; therefore, etc. This, however, is not only true about relatives depending upon one another, but <it is true> about any
<of those things> which depend essentially on other <things>.
Because those other <things> ought to be posited in their definition, but those <other things> do not define themselves to one another unless they are those which mutually depend essentially (per se) upon each other. This is not true of accidental
9. Cf. supra, q. 27, nn. 1–7.
10. Isagoge, Ch. "On Species" (AL I6 9; ed. Busse, 4.6–9).
11. Categories, Ch. 7 (8b 13–15).
JOHN DUNS SCOTUS
relatives, nor of relatives related in the third mode, as measure and the measurable, nor of relatives according to genus.
[II. To The Principal Arguments] 12. To the first reason:12 the major is true of a definition which is given only through the essentials, of which kind is the definition of substance, or to a less <degree> of an absolute <accident>, not of a relative (respectivi). For since relatives (respectiva) depend essentially on one another, they then mutually define each other, although neither is prior to the other.
13. To the second <argument>:13 the major is only true of a definition that is given without an addition, <but> a definition of a relative or of some accident is not of this kind. Nevertheless, in the definitions of other accidents only the subject is added beyond the essence of the defined. In the definitions of relatives there is a subject and its correlative, and this <is the case> since a relative depends on many more <things> than any accident that is said absolutely.
14. To the third <argument>,14 one can deny the major about relatives and contraries said in a transferred sense and opposed in a private sense, since in all these it is possible for one opposite to be defined through the other.
15. To the first proof of the major:15 the form of arguing is not valid, as is evident in other <arguments>: "a non-quality follows substance; and a non-quality does not pertain to the definition of man; therefore, it is a non-substance" is invalid.16 Nor is that rule, namely, "the consequent does not pertain to the definition of any <thing>, nor does the antecedent," as it is taken, true except about the implication of a superior and inferior in the same genus, such that the implication is not between contradictory opposites and relatives. Since a negation on the part of the contradictories is not superior in the genus to the other relative, as non-father is not superior to son.
12. Cf. supra, n. 1.
13. Cf. supra, n. 2.
14. Cf. supra, n. 3.
15. Cf. ibid.
16. This is an invalid, oblique syllogism. Even if it were to be reduced to a normal, categorical syllogism, it would still contain an illicit middle term.
QUESTION 28
16. To the second proof <of the major>,17 one can say that every part of a definition, in that way in which it is posited in the definition, can be predicated of the defined. And so one relative opposite is not posited in the definition of its correlative in the nominative case—according to which it is opposed to it—but in an oblique case, for example, as "of this" or "to this," and it is predicated of it in this way; for although a father is not a son, nevertheless he is <the father> of a son. The minor can also be denied, since not all relatives are relative opposites, but only if both are taken with respect to the same thing.
17. To the fourth <argument>:18 the major is true about a definition of substance which is only given through its essential
<elements>.
18. Against this: if a relative is at least defined through its correlative, <then> it is necessary that the correlative is at least of the understanding of the relative, since nothing is defined through anything without which it can be understood completely. But the consequent is false, since with equal reason, when it is expressed with it, the correlative would be given to be understood, and then there would be a vain repetition since the same <thing> would be said twice.
19. It is said that a relative, when it is expressed with its correlative, is not understood better through it, although it is given to be understood when it is not expressed. Aristotle seems to imply this when he says19 that the same thing must not be given "to be signified" in relatives, that is, given to be understood through the separated relatives and through the <relatives> joined with their correlatives; and through this he teaches <us> to avoid a vain repetition in the others.
20. Against this: I ask about the name imposed on one relative. If it gives <one> to understand another correlative, this is
<so> either from the imposition or from nature. Either way, it follows that there is not a variation concerning what is taken absolutely and what is joined to another correlative with respect
17. Cf. supra, n. 3.
18. Cf. supra, n. 6.
19. Metaphysics, Bk. V, Ch. 15 (1021a 37– b 5).
JOHN DUNS SCOTUS
to the other <one> of them, nor will there be a variation with respect to that effect, but that it always equally gives the correlative to be understood.
21. Likewise, a correlative is not given to be understood through this utterance unless through the fact that a relative is signified through the <utterance>. But there is not a variation concerning this cause, but that this relative is not uniformly signified, whether another <thing> is expressed with it or not.
Therefore, neither will there be a variation with respect to this, because another relative is given to be understood.
22. To the first of these <objections>,20 it is said that an utterance <that is> imposed to signify one relative is in no way a sign of the other, neither by nature nor by convention. But since a relative, which is signified through that <utterance>, cannot be understood without its correlative, therefore it is necessary— when understanding that relative through this utterance—to understand its correlative. This happens when another correlative is expressed through another utterance and also when it is not expressed, but <this does> not <happen> because this utterance is in some way the sign of that one. Therefore, I concede
<that> on the part of a signifying utterance, no variation happens, whether the correlative is expressed or not; nor on the part of that which is signified through the utterance, but <rather I argue> that it is always necessary for understanding it that its correlative be understood. But there is only a variation from the fact that a correlative, expressed or not expressed through its proper utterance, is given to be understood through this relative; and that accidental diversity definitely belongs to relatives.
23. To the second <of these objections>,21 <a reply> is evident, since through this utterance <that is> imposed for a relative, the correlative is given to be understood in no way; but through that relative for which this utterance is imposed, another correlative is understood, at one time expressed through its proper utterance, at another time not expressed.
20. Cf. supra, n. 20.
21. Cf. supra, n. 21.
QUESTION 28
24. Against this <argument>: "whatever is a sign of a sign, is a sign of the signed";22 but this utterance imposed for this relative is the sign of this relative, and this relative is the sign of the correlative; therefore, this utterance <that is> imposed for this relative is the sign of its correlative, although not immediately.
25. To this <objection> it is said that the major, "whatever is a sign of a sign, is a sign of the signed" is true in this way: so that an intermediate sign is not varied in comparison to the first sign and the last <thing> signed. This does not happen in this proposition, since it is extraneous for this relative, insofar as it gives
<one> to understand its correlative, to be signified through this utterance, and vice versa.
26. To the fifth <argument>,23 it is said that many relatives can be essentially (per se) of the same thing. And according to this, some suitable definition can exist, when one correlative is posited, even though the remaining one is not posited.
22. Cf. Scotus, Super Periherm., Quest. primae, I, q. 2, n. 20.
23. Cf. supra, n. 7, and also infra, q. 29, n. 13.