Translated by Thomas Williams
[Question 1: Is logic a science?]
The question is whether logic is a science.
1 It appears that logic is not a science.
For a mode of knowing [sciendi] is not a science; logic is a mode of knowing; therefore, etc. The major premise is evident by analogy: for example, a mode of seeing is not sight. The proof of the minor premise is given by Aristotle in Metaphysics II [3, 995a14-15]: "It is absurd to investigate science and the mode of knowing simultaneously." And 'mode of knowing' is explained as meaning 'logic'.
2 Moreover, in Metaphysics VI [1, 1026a18-19] Aristotle divides the sciences into the mathematical, the natural, and the divine or metaphysical; logic is not included under any of these three.
3 Moreover, science, like demonstration, proceeds from proper principles, according to Aristotle in Posterior Analytics I [9-10, 76a17-b16]. Logic proceeds from common principles. Therefore, etc.
4 Arguments for the opposite view:
Science is an effect of demonstration, and many demonstrations are put forward in logic. Therefore, etc.
5 Moreover, "One who has mastered logic is one who has science" is per se. Therefore, logic is a science, since where there is per se predication, the inference from concrete to abstract is valid.
6 One must say that logic is a science, since the things taught in logic are conclusions on the basis of demonstrations, as is the case in the other sciences. Therefore, they are objects of scientific knowledge [sciuntur], since "demonstration is a syllogism that makes scientific knowledge." (1) Also, logic contains everything necessary for a demonstration: a subject, and a quality that can be demonstrated of the subject through a middle term, which is the definition.
7 But one must realize that logic is considered in two ways. In one way, it is considered insofar as it teaches. Regarded in this way, it proceeds from necessary and proper principles to necessary conclusions and is a science. It is considered in the other way insofar as we make use of logic by applying it in those matters in which it is of use. Regarded in that way, it does not proceed from proper principles but from common principles, and it is not a science. This is evident in the natural treatises, where Aristotle offers logical arguments that proceed from a common middle term and do not cause scientific knowledge in the strict sense. (2)
8 One response (3) to the first argument is that a mode of knowing is not a special science, but it can be a common science.
9 On the contrary: If "A mode of knowing is a science" is true in any way, then, since it is a predication in the abstract, it will be per se in the first mode. The consequent is false; therefore, so is the antecedent. The inference is evident, since every predication in the abstract is essentially true [if it is true at all]. The falsity of the consequent is evident, since the predicate is not included in the intelligible content [ratio] of the subject.
10 Moreover, a mode of knowing is posterior to knowing; therefore, it is posterior to science, since knowing is posterior to science.
11 For that reason another response to the first argument is to say that its minor premise ["Logic is a mode of knowing"] is false, formally speaking.
12 Alternatively, one could say that its major premise ["A mode of knowing is not a science"] is false, since there is genuine scientific knowledge of a mode of knowing. Thus, the intellect understands its object, of which it has scientific knowledge; and in turn, by reflecting on that act of understanding, it can acquire knowledge of itself. So although at first it was, as it were, a mode, it can then have itself as object.
13 As for the proof [of the minor premise taken from Aristotle], one can say that the explanation [of 'mode of knowing' as 'logic'] should be understood materially. And in that way, the predication "A mode of knowing is a science" is true, since logic teaches the mode of knowing insofar as it treats syllogism or argument, by which alone scientific knowledge is acquired.
14 Alternatively, one could say that if 'mode of knowing' is explained as 'logic', the copulation should not be placed between 'science' and 'mode of knowing', but between 'investigate science' and 'investigate the mode of knowing'. For to investigate the one is to investigate the other, since they are concomitant, just as to investigate one thing is to investigate whatever is necessarily conjoined with that thing, even though the one thing is not identical with the other thing.
15 To the second argument I say that in that passage Aristotle is distinguishing the real sciences. Logic, however, is a rational science.
16 As for the third argument, it is evident that logic [as teaching] proves passions on the basis of proper principles, although it is used with regard to common principles.
[Question 2: Is logic a common science?]
The question is whether logic is a common science.
1 It appears that it is not.
For a science is common because its subject is common, but the subject of logic is distinct from the subjects of other sciences.
2 An argument for the opposite is drawn from Boethius (4): "Logic concerns second intentions applied to first intentions." But these second intentions can be applied to all first intentions.
3 One should say that a science is called 'common' on the basis of its subject. Consequently, a science can be understood as common in two ways: either because the subject is predicable of the subjects of other sciences, or because its subject is at the disposal of other sciences.
4 Logic is not common in the first way, except perhaps accidentally, if its subject can be applied to all [subjects of the other sciences]. It is, however, common in the second way.
[Question 3: Is the syllogism the subject of logic?]
The question is whether the subject of logic is the syllogism.
1 It appears that it is not.
For all science comes through syllogism. So if there is a science of syllogism, it will be on the basis of syllogism, which is false, for two reasons:
2 First, a syllogism of which one seeks science is unknown, whereas a syllogism through which one has science is known, since there is no science through something unknown.
3 Second, I pose this dilemma concerning the syllogism through which one has science: either there is science of that syllogism or there is not. If there is not, then by parity of reasoning there is no science of syllogism in general, since whatever is known scientifically regarding the general is known scientifically of each particular. If there is, then that science is through another syllogism, and there will be an infinite regress in syllogisms in order to have science; and that is absurd, since if it were true, nothing would be known scientifically. For the infinite is unknown [incognitum], according to Aristotle in Physics I [4, 187b7-8]; therefore, one and the same thing is both known and unknown [ignotum].
4 Moreover, nothing is the subject both of the whole and of a part. But syllogism is the subject of a part of logic, namely of the Prior Analytics. Therefore, etc.
5 Moreover, there was never a syllogism in sense; therefore, there was also never a syllogism in the intellect. The inference is evident from what Aristotle says in De sensu et sensato. (5)
6 Argument for the opposite view:
The passions of the syllogism are principally demonstrated in logic with respect to the syllogism itself, and with respect to other things on account of the syllogism. Therefore, etc.
7 One view (6) is that the subject of logic is the conception formed by an act of reason, since that is what is common to everything considered in logic. Now the act of reason is threefold: first, the understanding of indivisibles; second, the composition or division of those indivisibles; and third, discursive reasoning from the known to the unknown. Accordingly, the Categories is about the conception formed by the first act of reason; the De interpretatione, which is about statements, is about the conception formed by the second act; and the New Logic, (7) which is about the syllogism and its subjective parts, is about the conception formed by the third act.
8 Another view is that logic is about second intentions applied to first intentions. As Boethius (8) says, these are common to all the things treated in logic.
9 A third view is that the subject of logic is being, since according to Aristotle in Metaphysics IV [2, 1004b22-23], metaphysics, dialectic, and sophistry all deal with the same thing. Now metaphysics deals with being; therefore, etc.
10 A fourth view is that the subject of logic is the statement [oratio], (9) since the old logic treats the statement and its parts, since the statement is a type of utterance [enuntiatio]. Also, statement appears in the definition of the syllogism in Prior Analytics I [1, 24b18-20].
11 A fifth view is that the subject of logic is argumentation, (10) since the whole of logic concerns the species and parts of argumentation.
12 A sixth view is that the subject of logic is the syllogism, since everything else considered in logic is concerned on account of the syllogism.
13 In order to see which view is closest to the truth, one must note that there are three requirements for the subject of a science. First, both what-it-is and that-it-is must be known, since demonstration must presuppose both, as Posterior Analytics I [1, 71a11-16] says. The second is that the passions of the subject be demonstrated of that subject within the science on the basis of what-it-is. The third is that all the other things treated in the science are reduced to (11) that subject and are considered on account of it -- for otherwise the unity of the subject would not make for the unity of the science.
14 The first two conditions are not met by the first three views, since there is in logic no definition according to a general notion [ratio] of any of the subjects identified in those views. Nor is any of them premised as a principle of a science, nor is any passion demonstrated of them in general. Therefore, each of them is too general a subject. Indeed, the three do not differ from each other. That is obvious in the case of the first two. And I will prove it in the case of the third: the subject here is either (a) real being or (b) rational being. If (a), then logic is a real science, which is false. If (b), then either (b1) logic exclusively concerns rational being, i.e., being as considered by the intellect, or else (b2) it concerns rational being as caused by reason. If (b1), then logic could still be a real science, since nothing is the subject of any science except as considered by intellect or reason, since nothing is the subject of any science except as universal. Therefore, one should grant (b2): logic concerns rational being as caused by reason. And that is the same as the first two.
15 As for the argument offered for the first view [in n. 7], it is clear that it argues on the basis of an insufficient premise and commits the fallacy of the consequent, since two other requirements for the subject of a science are lacking.
16 To the authoritative passage from Boethius [in n. 8] I say that logic is about those things, but they are not its first subject, but are common to the subject. It's like the way we say that every science concerns being, since there is no science of non-being.
17 To the argument for the third view [in n. 9] I say that they concern the same thing, not because their subject is the same thing, but because what is considered in logic is the basis for dealing with anything and everything. For many things are dealt with in a science, but it does not follow that every single one of them is the principal subject in the science.
18 The subject identified in the fourth view [in n. 10], namely, the statement, meets neither the second nor the third condition. And as for the claim that statement pertains to syllogism, I say that this is not so unless 'statement' is used equivocally in speaking of the syllogism and of utterance [enuntiatio]. And an equivocal is not the subject [of a science], since it cannot be defined. For in the definition of syllogism 'statement' is used to mean a piece of argumentation.
19 The subject identified in the fifth view [in n. 11], namely, argumentation, fails to meet the first condition. For Aristotle never defines argumentation in general, and he does not offer a definition of it as the principle of a science. Boethius, however, does define it in his Topics. (12) It fails to meet the second condition as well, as is evident from the general argument. (13) And I offer this proof that it also fails to meet the third condition: either 'argumentation' is equivocal, in which case it is evident that there is no unitary science on the basis of the unity of argumentation, since it is neither one knowable thing nor one intelligible thing; or else it is analogous with respect to the things contained under it, in which case the science will not derive its unity from it, but rather from the unity of the first thing to which the others are attributed -- and that is the syllogism -- in the same way that the unity of metaphysics derives from the unity of substance, to which other beings are attributed.
20 So one must say that the first and proper subject of logic is the syllogism. For this meets the first condition, since immediately after determining the parts of the syllogism in the old logic, Aristotle places the definition of the syllogism at the beginning of the Prior Analytics. (14) It also meets the second condition, since in the same work Aristotle shows many passions of the syllogism on the basis of that definition: for example, mode and figure, and then I'm not clear on the rest of this sentence. It also meets the third condition, since it is on account of the syllogism that he considers its parts -- namely, the incomplex and the sentence [enuntiatio] and its integral subjective parts -- in the Prior and Posterior Analytics; as well as the other species of argumentation, since they are reduced to the syllogism as the imperfect to the perfect; as well as the sophistic syllogism as the privation of the syllogism, since it belongs to the same science to know both the positive thing and its privation. Thus the division of logic is evident from the division of the syllogism and the things attributed to it.
21 As for the first argument [in n. 1], I concede that logic is a science of syllogism that is based on some particular syllogism.
22 To the first argument against this response [in n. 2], I say that the syllogism [on which the science of logic is based] is known with respect to both the truth of the premises (since they are immediate propositions) and the inference of the conclusion from those premises (since that is evident through itself, since "a perfect syllogism is one that is lacking nothing to keep it from appearing necessary" (15)). But it is unknown with respect to the passion that is demonstrated of the syllogism in general; with respect to that passion, the syllogism in general is unknown. For one seeks knowledge of the syllogism in general only with respect to its passion, and so it is not the case that one and the same thing is both known and unknown in the same respect.
23 To the other argument against this response [in n. 3], I do affirm that there is science of the syllogism through which there is science. In answer to the question "through which syllogism [is there such science]?" I say: through that syllogism itself. For insofar as through that syllogism some passion is demonstrated of every syllogism, that passion is shown to hold of that syllogism qua syllogism. And this is the only way in which there is scientific knowledge of a particular.
24 To the second preliminary argument [n. 4] I say that the syllogism with respect to the properties that follow formally from it is the subject of the Prior Analytics. But the syllogism is the subject of the whole of logic with respect to all the passions in itself or in its integral or subjective parts or in the things reducible to it. Nor is it required that the subject of a science be predicated of all the things treated in that science, but merely that it be that on account of which the other things are treated. This is clear in the case of natural science. Its subject is the movable body, and yet it deals with motion and nature, which are not movable bodies.
An alternative response is to say that the major premise ["nothing is the subject of both the whole and a part"] is false in the case of sciences in which the subject of the whole requires that determinations be made concerning many points in order for the subject to be known. In such a case, within that science one will have to make determinations concerning those many points and, thereby, concerning the subject itself. And so one part of the science will be about the subject that is the subject of the whole science. That is the case here, as it clearly is for the Physics with respect to the whole of natural science.
25 To the third argument [in n. 5] I say that Aristotle's claim that "nothing is in the intellect [unless it is first in the sense]" is true of that which is the primary intelligible, which is the 'what' of a material thing; but it is not true of everything that is intelligible in itself, since many things are understood in themselves, not because they cause a species in the sense, through the reflection of the intellect. And that is the way it is in the case of the syllogism. As it says in Posterior Analytics I [4, 73b38-39], "'in itself' has a wider application than 'primarily'." For the isosceles triangle has three sides in itself, but not primarily.
1. Aristotle, Posterior Analytics I.2, 71b18-20.
2. See Aristotle, Physics I.2, 184b25-185a4.
3. Given by Robert Kilwardby, Anal. priora (ed. Venetiis 1499 f. 1ra).
4. Actually from Avicenna, Metaph I.2.
5. Aristotle, De sensu et sensato 6, 445b16-17: "Our intellect understand nothing apart from sense."
6. That of Thomas Aquinas in Expos. libri Post. I prooem. The critical Latin edition of the present work cites a passage from Expos. libri. Post.I lect. 20 in connection with the sixth view (see [12]), but that passage in fact rehearses the view that logic concerns all three acts of reason.
7. The New Logic comprised the Prior Analytics, Posterior Analytics, Topics, and Sophistical Refutations of Aristotle, all of which were first reintroduced to the West in the middle of the twelfth century. It was distinguished from the Old Logic, which comprised Aristotle's Categories and De interpretatione, Porphyry's Isagoge, and several commentaries and original works by Boethius. A point of trivia: the De interpretatione was, oddly enough, always known in the medieval Latin West by its Greek title, Peri hermeneias, sometimes written as one word, and often very eccentrically spelled.
8. Actually Avicenna, Metaph. I.2.
9. Boethius, In Categ. Aristot. I.
10. Albert the Great, Liber de praedicabilibus I.4, attributes this view to Avicenna, Alfarabi, and Algazel and then himself asserts the claim that the subject of logic is "argumentation or the syllogism."
11. Not in the modern sense of "shown to be nothing other than it," but in the sense of "traced back to it" or "discussed in terms of its relation to it."
12. Boethius, De differentiis topicis I (PL 64, 1180C).
13. Presumably, the argument given in the second sentence of n. 14.