PREFACE

I owe a special debt of gratitude for assistance and encouragement in the preparation of this translation to Professor Charles De Koninck of Laval University, Quebec, Canada; and to my husband who, in using a part of it in a graduate course at Notre Dame University, discovered that its completion would contribute to knowledge of the thought of Aristotle and St. Thomas. Special acknowledgment must also be made to Professor Richard Connell of Marquette University for his painstaking examination of the manuscript and his many helpful suggestions; and to Fr. Henri DuLac, Chairman of the Department of Philosophy at the College of St. Thomas, St. Paul, Minnesota, whose doctoral dissertation "On the Modal Proposition" was very helpful in relation to Cajetan's Commentary on that subject.

South Bend, Indiana  Jean T. Oesterle

January, 1961

Translator's Introduction

I. THE TEXT OF ARISTOTLE

Aristotle's most fruitful work began about the forty-ninth year of his life (335 B.C.) when he founded his school in Athens, and lasted until he withdrew to Chalcis in 323 B.C. His departure to the seclusion of Chalcis was generated by a charge of impiety apparently brought about by a mixture of political feeling and doctrinal hostility. It goes without saying that Aristotle has undoubtedly contributed more than anyone else to the science of logic. Certainly he had no predecessors in this field and there was not much by way of additional original contribution until centuries later. His logical works, which have been known since at least the sixth century as the Organon, or tool, comprise eight treatises, if one includes the Rhetoric and the Poetics: the Categories, the Peri Hermeneias (On the Enunciation), Prior Analytics, Posterior Analytics, Topics, and On Sophistical Refutations. It will be noted in this doctrinal, i.e., scientific, ordering of the logical works, that the Peri Hermeneias is second, but it seems to be later in date than at least three of the other logical treatises since he mentions these in this work: On Sophistical Refutations (at 17a 36), Prior Analytics (at 19b 31), and the Topics (at 20b 26). In fact, Father J. Isaac in his work on the Peri Hermeneias thinks that it may even have been a last work. His argument is that the reference to the Prior Analytics, which presupposes the doctrine elaborated in the Posterior Analytics on the demonstrations of definitions also establishes the Peri Hermeneias as later than the Posterior Analytics. In addition, the elaborate doctrine on enunciations about future contingent events in Chapter 9 and on the consequents of modals in Chapter 13, and the absence of any reference to the Peri Hermeneias in any other work of Aristotle may indicate that it was composed at the end of his life.

             There has been no serious question as to the authenticity of the Peri Hermeneias. Andronicus (early first century, B.C.) suspected its genuineness because of a reference to the De Anima to which he found nothing which corresponded in that treatise. Aristotle makes this reference at 16a 8 and St. Thomas in his Commentary (Lesson II, n. 6) answers Andronicus' objection, making reference to Book I of De Anima. Professor W. D. Ross, on the other hand, in his book Aristotle, suggests that the reference at 16a 8 should be transferred to 16a 13, and relates it to De Anima, Book III, Chapter 6, following H. Maier, who wrote a defense of this work.

             There is, in fact, both external and internal evidence supporting the authenticity of the Peri Hermeneias. For example, Theophrastus, a fellow-academician of Aristotle's, and Aristotle's friend Eudemus, both wrote books that apparently presuppose it. Ammonius Hermiae (fifth century, A.D.), who wrote a commentary on the Peri Hermeneias and who doubted the genuineness of its last chapter (from 23a 27 on), mentions that Andronicus was the only critic of its genuineness as a whole. As to internal evidence, the style and grammar, though less dialectical and more didactic than in his treatises on other subjects, is typical of Aristotle's logical works. This difference can perhaps be ascribed to the fact that Aristotle was the first to write on logic and hence had no predecessors to take into account.

             The Oxford edition of the Greek text, edited by L. Minio-Paluello, Praelector of Mediaeval Philosophy at Oxford, (Oxford: The University Press, 1949) has been used for the translation of Aristotle's work. Reference can be made to the Preface of that edition (p. xiv) for the codices and ancient translations used for the preparation of the Greek text, and to page 48 for the identification of the symbols used to refer to them.

II. THE COMMENTARY OF ST. THOMAS

St. Thomas, who lived barely forty-nine years (1225?-1274) and who besides many other writings commented on twelve works of Aristotle between the years 1260 and 1273, wrote the present commentary, which is dedicated to the Provost of Louvain, between 1269 and 1272. While much learned work and research has been done to try and establish the exact dates of his commentaries by such men as Grabmann, Mandonnet, and Walz, to mention only a few, no certain and definitive conclusions have been reached. It is known, however, that during the time of the writing of this commentary he also wrote the commentaries On the Ethics, On the Meteorology (up to Book II, Lesson 10), On the Posterior Analytics, the only other logical work he commented on, On the Politics (up to Book III, Lesson 6), and perhaps had begun the commentaries On the Heaven and the Earth and On Generation and Corruption. In addition, he wrote on many theological subjects. It is interesting to note that St. Thomas began writing his commentaries on Aristotle's works after attending a general chapter of his Order in 1259 where along with St. Albert the Great among others, he helped to organize the curriculum of studies for Dominican schools. Perhaps his experience of teaching theology had made evident its need of philosophy.

             There appears to be no doubt at all about the authenticity of St. Thomas' Commentary. The official catalogues of his works, particularly the one by Reginald of Piperno who was St. Thomas' assistant and secretary from 1259 until his death in 1274, all list the Commentary on the Peri Hermeneias.

             However, St. Thomas did not write a commentary on the entire work, although it seems to have been his intention to finish it. That intention is evident in the first two lessons of the second book where he states the ordering of the remainder of the text of Aristotle and in no way indicates that he does not intend to go on with his exposition of it. There seem to be differing opinions as to the reason for its being unfinished. The Piana editors (edition of St. Pius V) said that it was because of his death. This well may be the most accurate theory since this commentary was among the last writings of St. Thomas. Another view held that it was not completed because he was either too much occupied with other things or that the Provost for whom he wrote it, being a young but wise student, intended to spend more time on study before asking more of St. Thomas.

             The Leonine edition of the Commentary has been used for the translation of this work. It is based largely on the Piana edition with the variants in texts noted. In the Preface of the Leonine edition will be found the codices and editions from which it was prepared together with their identifying symbols.

III. THE COMMENTARY OF CARDINAL CAJETAN

Soon after the death of St. Thomas--in fact, before the end of the thirteenth century--his work on the Peri Hermeneias was completed by others, and later copies of the work usually had such a supplement. All of these were made by Dominicans or were extracts from their own commentaries. The statement made in the Venetian edition of 1496 testifies to this: "Many of the most renowned doctors of the same Order have furnished a supplement, i.e., have completed what was lacking in St. Thomas' Commentary."

             The best known of these supplements is that of Cajetan (1469-1534) which was explicitly composed with the intention of completing St. Thomas' work. Cajetan finished it in 1496 and it appeared the same year in the third Venetian edition. It was published so repeatedly after this along with the commentary of St. Thomas that it has become inseparable from it.

             The text of Cajetan's commentary as it appears in the Leonine edition has been used for this translation. It is not a critical edition of his work (nor is one available at the present time). Hence, no corrections have been made nor have variants in texts been noted by the Leonine editors, but obvious difficulties with the text have been noted by the translator.

IV. THE SIGNIFICANCE OF THE DOCTRINE

The comment often made about the Peri Hermeneias is that it is somewhat elementary, treating as it does such simple things as names, verbs, speech, (i.e., the verbal explicitation of reason) and the enunciation. It is indeed true that in contrast to the Categories, the mode of exposition in the Peri Hermeneias is both more elementary and elaborate. It is elementary in the sense that it is confined to the essentials of the enunciation, also called the proposition; its elaboration is made manifest in the extensive analysis of the parts of the enunciation and the enunciation itself. It should be noted in this connection that Aristotle does not expressly treat of syncategorematic terms--such as the quantifiers "all," "some," etc., and the connectives "or," "if," etc.--which in modern logic are taken to be the most basic. In fact, what Aristotle has to say about the essential parts of an enunciation, the name and the verb, most modern logicians would consider foreign to pure logic. But whatever may be said of Aristotle's development of logic, his definitions of such simple things are important for setting order in our thinking as expressed in ordinary language. Indeed, they are particularly relevant for grasping the difference between the words of common speech and the symbols of formal logic (which Albertus Magnus called termini transcendentes), mathematics, and theoretical physics, especially since little attempt has been made in modern times to consider these elementary and basic matters carefully.

             Hermann Weyl, for instance, in an essay entitled "The Mathematical Way of Thinking," remarks: "The mathematical game is played in silence, without words, like a game of chess." He then significantly adds: "Only the rules have to be explained and communicated in words, and of course any arguing about the possibilities of the game, for instance about its consistency, goes on in the medium of words and appeals to evidence." This passage, as well as his whole influential essay, raises an important question: Just what is the difference between a word and a symbol? It is with respect to an answer to such a question that the Peri Hermeneias remains an important work and one that has special interest for the modern reader. By means of it we can see that in the phrase, "the time t," for instance, the word "time" and the symbol t do not stand for the same thing nor in the same way. The symbol has an operational value which is not true of the word. And it may be worth noting, further, that the operational value in question is a mechanical one, which is the same as to say, in modern terminology, that it is utterly formal. By this we mean that although a symbol, such as t, cannot be taken out of context, one does not have to think of what the word "time" stands for to use the symbol t operationally.

             An interesting point of difference between Aristotelian and modern logic that will be found in this treatise concerns the infinite name, e.g., non-man or non-tree. Aristotle states expressly that infinite names are not names, and the reason is that while they deny a given meaning, such as man, they do not impose a new meaning. Name and infinite name are in fact contradictories, excluding a middle. August de Morgan, on the other hand, in his Formal Logic, considers both as names. In doing so, he pointedly observes that he intends to draw no distinction between contrary and contradictory names. Now this distinction can be overlooked in some instances. For example, the genus "color" can be divided into "white" and "nonwhite." The mode of signification can be interpreted here as one of contrariety, for we understand "nonwhite" of any other color. But the mode of signification could also be understood as an opposition of contradiction, since "nonwhite" can be said of what is not in the genus of color at all, such as an integer, a square root, and even of what in no sense can be. The point is that if "nonwhite," understood as a contradictory term, were a name, its interpretation should be a possible one, i.e., it would have a determinate meaning, but the fact is that it has an infinity of possible meanings and therefore no determinate meaning at all.

             A modern reader will also note that in this treatise all enunciations are predicational. He may well wonder why Aristotle did not consider what are today called relational propositions, e.g., "Peter is taller than John." The fact, however, is that relational propositions are included in predicational propositions, for a closer inspection of such a proposition will reveal that "to be taller than" expresses a relation of inequality, and it is this relation that is being predicated.

             It will also be noticed that Aristotle makes no case here for conditional enunciations, e.g., "If p, then q." St. Thomas suggests the reason for this in his exposition of the text, namely, that such an enunciation signifies something that is one by hypothesis and this, unless confirmed by the truth of the simple enunciation, will not fulfill the requirements for demonstrative science to which the whole of Aristotelian logic is ordered.

             A point of special significance for both Aristotelian and modern logic is that of supposition, "supposition" being taken here as the verification of the term in a proposition in relation to the copula. It has been asserted that this doctrine is a scholastic invention, now happily fallen into oblivion, as H. W. B. Joseph says in his Introduction to Logic. But this remark seems to ignore the import of Aristotle's words in 21a 25, where he says: "Homer is something, say, a poet. Is it therefore true to say also that Homer is, or not?" And he goes on to say: "The 'is' here is predicated accidentally of Homer, for it is predicated of him with regard to 'poet,' not in itself." What is meant here should be distinguished from what Aristotle intends in the Sophistical Refutations where he says (165a 5): "It is impossible in a discussion to bring in the actual things discussed; we use their names as signs instead of them." The passage from the Peri Hermeneias is about the name in relation to the verb in a proposition, which is exactly the point at issue in the mediaeval doctrine of supposition, whereas in the Sophistical Refutations he is talking about names taking the place of things, i.e., that every name stands for the thing named, which refers to signification. Supposition, then, is not a later invention, nor an unimportant one, since it shows why such a statement as "Homer exists" is false whereas "Homer is a poet" is true. For to signify with time is essential to the verb as distinguished from the name, as Aristotle points out in defining the verb, but in the case of supposition, time has a new relevance: "Homer exists" is false because "to exist" is in the present tense. That time is relevant to the verb as such will appear irrelevant to some modern logicians. But again we must bear in mind what Aristotelian logic intends. It is not formal in the modern (and quite legitimate) sense of this term "formal"--not even in the Prior Analytics, for it too is about second intentions.

             The expressions "first intention" and "second intention" are not in Aristotle's text, but they are in no way foreign to the intent of Aristotle. The expressions are simply a convenient way of distinguishing, for example, what the predicable genus "animal" is from what an animal is. We first observe that there are animals of different kinds and that "animal" can be said of them all, and that while a horse is an animal we do not say that an animal is a horse--rather that some animal is a horse. Nor do we say that "horse" is this one. And so we notice that our mind has related one thing to many in a special way. This is how we come to know in a reflexive way what a predicable genus is. What animal is--however vaguely known--is a first intention, but the particular type of relation our mind conceives as we predicate it of different kinds is a second intention.

             When we know an animal, whether it be a man, a horse, or an elephant, we know by means of some similitude of the known in our mind. Such is the case of first intentions. But in conceiving a second intention there is no similitude between what we conceive--the relation of reason--and that apropos of which the mental relation was formed. We can draw something like a horse, but we can sketch nothing like a predicable species. In other words, there is no resemblance between what we conceive as a man, for instance, and the relation of reason we form about man as predicable of many individuals. There is, in this respect, an analogy between words and second intentions. The word "square," for instance, is not square, and the name "house" does not look like a house. But it is only an analogy, for second intentions are not arbitrary in signification, as words are, although a sophist can use second intentions arbitrarily.

             Now the verb raises a problem of its own in this connection because of the time factor. What is time as a defining element of the verb in logic? What does the word "time" stand for in the definition? Real time, of course (and not just any kind of succession): time past, present, and future. Let us return to our example of supposition. "Homer is a poet" or "Homer no longer exists" are not statements about something logical, but they do have a logical structure: they comprise name and verb, subject and predicate; they are also instances of true enunciations. The reason why "Homer exists" is false is not a logical one; it is false because Homer is dead. Logic is not concerned with Homer, nor with his kind of existence or nonexistence. But the point is that when we do talk about him, expressing our mind on the subject, we do not simply talk about him, we also express our mode of thinking about him, and to examine the order among the conceptions in our mind as we think and talk about him is the business of logic. Hence we observe in this particular case that if the name "Homer" is to stand for the historical person, and "to be" is a predicate, we cannot form a true enunciation without regard to the tense of the copula. This is a rule of logic. To imply, on the other hand, that time has nothing to do with verbs, that such a consideration is purely grammatical (since different languages have different tenses and some skip the copula) would mean that logic has nothing to do with these, and therefore that the copula itself is foreign to logic, since it too signifies with time. If this were true, the Peri Hermeneias would not be a logical treatise.

             In this work Aristotle also takes up the question of enunciations about future contingent events, and this matter deserves extended comment. This is a recurrent problem in our time because the symbolic reconstruction of time is sometimes believed to be a substitute for the time we experience and express in ordinary language. The symbolic reconstruction of time in relativity theory, for instance, when interpreted as an adequate substitute for time as we name it, can only lead to determinism--a determinism of a baffling kind, since, from this point of view, time has no direction. Thus, according to this interpretation we would have to say not only that what happened later was entirely determined in what happened before, but also that what happened before was equally determined in what happened thereafter. The determinism of Laplace, often equated with the principle of causality, left no room for contingency, nor would the symbolic reconstruction of time as interpreted by some philosophers of science. If 'the principle of causality' so understood were true, a considerable part of the Peri Hermeneias, including the treatise on modal propositions, would be irrelevant. But if some measure of indeterminism (not necessarily of the kind called fortune or chance) is to be considered in physics, as in thermodynamics and quantum theory, it will require an importation foreign to this principle of causality, such as Aristotle's potentia, which is Heisenberg's expressly stated view, one not unlike that of Max Born. But this potentia implies time as we know it before its symbolic reconstruction, which shows that 'time' is not quite the same as t after all.

             There are two questions that arise regarding Aristotle's treatment of enunciations about future contingent events. First, why should he mention contingency at all in logic? The purpose of logic, the ordering of second intentions, which are nothing outside of the mind, seems to rule out such a consideration; and again, as Aristotle himself says later on in the Posterior Analytics, there can be no science of the contingent; hence why consider enunciations about future contingent events? The second question concerns the adequacy of his treatment of such enunciations, for it seems to ignore what lies between what shall be or what shall not be, i.e., what might be.

             Aristotle's discussion of contingency in the Peri Hermeneias brings out forcefully the idea that although the primary concern of his logic is to set order in our mind, it is nonetheless remotely based upon reality, and in the end is no more than a tool (organon) in the service of the other sciences. Now it should be pointed out first of all that contingency in one way or another enters into all of the logical works of Aristotle. Even prior to them--prior in a doctrinal sense--Porphyry's Isagoge introduces contingency in the definition of predicable accident--"that which can be present or absent without destroying the subject." In the Categories, accident means something quite different: that which is present in and is predicable of a subject; and while some such accidents are necessarily present in a subject, others can be present or absent, and in this respect they are also predicable accidents, e.g., to blush.

             In the Prior Analytics (Book I, Chapter 13) contingency is discussed as affecting the form of the syllogism. It is here that Aristotle first points out the distinction between the possible as opposed to the impossible and the possible as opposed to the necessary. In the Posterior Analytics he shows that there can be demonstration only in necessary matter, excluding thereby the extreme cases of possibility opposed to the necessary, namely, fortune and chance. On the other hand, he notes that demonstration in necessary matter does not exclude natural probability, i.e., what happens for the most part. The Topics introduces a new type of contingency, i.e., one that in argumentation does not lead to a necessary conclusion. This possibility is the same as likelihood or verisimilitude, as distinguished from truth. In the Rhetoric the enthymeme persuades in contingent matters, largely from singulars, by an appeal to appetite. In the Sophistical Refutations Aristotle shows that it is the infinity of ens per accidens that is exploited by the sophist, such exploitation being rooted in his own desires. Finally, in the Poetics, contingency is shown to be necessary for tragedy.

             We must now bring out the proper reason why contingency is discussed in the present work, and how it shows that logic, as Aristotle conceived it, is necessarily, though only remotely, based upon reality. We have already pointed out the time factor in the definition of the verb. But time appears in a new guise in the discussion of enunciations about the future. If a statement such as "Socrates will be in the forum tonight" is true, it implies that it is impossible for him not to be there tonight. But common experience tells us that though this may be his firm resolve, he may not get there after all, for any number of reasons, e.g., his wife, lack of transportation, being dissuaded by friends, threatened by enemies, etc. Now if it is true that such a statement cannot be determinately true, it means that we are here faced with a possibility to be and not to be: Socrates now at home, or on his way to the forum, can reach the forum and not reach it. The fact of being at home, or on his way, coincides exactly, in time, with ability to be in the forum thereafter.

             The foregoing is an instance of what Aristotle calls "simultaneous potency of contradiction." Potency, so understood, implies temporal succession, yet a temporal succession of a special kind. Even while at home, Socrates' remaining there for as long as he does implies time, but not quite the time this kind of potency demands. If he could not possibly leave home, time would go on all the same, but there would, for that time, be no simultaneous potency of contradiction. Where there is no such potency, every proposition will be either determinately true or determinately false. Now there is no such potency in the past, i.e., what has been can no longer not have been; and what is present cannot fail to be present so long as it is present. Any proposition stating where Socrates was or is now will be either determinately true or determinately false. Let us notice that all this is explained by merely pointing out particular instances which are familiar enough. All of which goes to show that the division between propositions about past and present, and propositions about non-necessary future things or events cannot be made without reference to the uncertain temporal succession that simultaneous potency of contradiction implies--whether this potency be active, as Socrates' will and resolve to leave his home or not to leave it, or passive, as when he is prevented from leaving his home. We must note, then, that the time factor in the definition of the verb does not take care of the time factor as it appears in propositions about future contingents.

             Now if there is this kind of potency in the real world, in the world of first intentions, does this entail that we will find its counterpart in logic as well? The matter is not that simple. The potency we have been talking about is real. But we have also been talking about the way we talk about it, and this reflects our way of thinking. All we aim to do in logic is to set order in our thinking, and one way of achieving it at this juncture is to notice in an explicit way that not all propositions are either determinately true or determinately false; if they were, this would assume, contrary to common experience, that everything is or occurs of necessity.

             Further, it may be opportune to recall Aristotle's warning that when we use the word dynamis--possible or potency--apropos of mathematical subjects, it stands as a metaphor, which simply means that it is applied, without change or extension of meaning, to something that the word does not properly signify. His specific example is geometry; but the same will be true of logic where some conversions, for instance, are "possible" and others not. The "power" of a number or the very name "predicable" are instances of possibility or potency taken as a metaphor. The reason is that the denomination is taken from something extrinsic to that of which the word "potency" is said. A number, for instance, has a power because of what our mind can do with it operationally; the potency of a predicable genus will be greater than that of a species because of its wider predicability--predication being an act of the mind. But even an apt metaphor requires some foundation. In the present case, the basis is none other than reason itself as applied to numbers or second intentions, but which are not the power that reason is.

             By way of fully answering the question as to why Aristotle considers contingency in logic, then, we should say first of all that his purpose in considering contingency in the Peri Hermeneias as well as in the rest of the Organon is, of course, strictly logical. But a logic of second intentions cannot treat of these in vacuo; there must always be some reference to first intentions, as was clear from the outset in the Categories, which points to instances in nature or in mathematics. Hence in the Peri Hermeneias, Aristotle cannot establish or analyze the special type of proposition which is neither determinately true nor determinately false without reference to the extra-logical. Or to put the matter in another way, if there were no contingency in human actions or in nature, the question of propositions that are neither determinately true nor determinately false would never arise.

             It should perhaps also be pointed out that the whole question of modal propositions, which is taken up later in this work, would have to be ruled out of logic if we prescinded from contingency, for we manifest necessity by opposing it to contingency and vice versa; otherwise contingency would be opposed only to the impossible and therefore might be reducible to the necessary since the necessary is not impossible, else it would not be.

             With respect to the adequacy of Aristotle's treatment of propositions about future contingent events in view of a possible middle ground between what will be or will not be (the second question we raised above), we should point out that from the viewpoint of the opposition of contradiction there is no intermediary: there is going to be a naval battle tomorrow or there is not going to be; hence, someone who reasons in such a way as to think the event is likely would have to make a distinction in his own mind if he is to be in conformity with reality, i.e., if his knowledge is to be true. For while he may have every reason to believe that it is likely that there will be a naval battle, and in this he may be quite right, he must know that it is only likely; and whether likely or not, there will in fact be or not be a naval battle tomorrow.

             The distinctive way in which contingency is treated in a logical context, namely, that it is a consideration of second intentions as relative to first, can be pointed up by considering the treatment of contingency in other works of Aristotle. In the Physics, for example, he is concerned with the ancient philosophical opinions that the whole universe was formed by chance (Democritus) and that even all the natural species are originally the work of chance mutations (Empedocles), which is another way of saying that contingent causality lies at the root of all things. Aristotle approaches this problem in an entirely different way. He begins by analyzing fortune, the extreme kind of contingency, which is well known because it occurs in human action, although his aim is to manifest chance in nature, more obscure to us, and recognizable only in familiar living things. In contrast to this, in the Peri Hermeneias he merely mentions our deliberation as a principle of future events by way of manifesting the fact of contingency.

             In On the Heavens (Book II, Chapter 12) he shows that whatever can cease to exist will at sometime actually cease to be (a conclusion that becomes a basic proposition in the Tertia Via of St. Thomas). This contingency is quite different from the kind treated in the Physics; it is not opposed to the necessary in the way chance and fortune are, as can be seen in On Generation and Corruption (Book II, Chapter 11).

             In the Ethics, which deals with human actions, contingency is treated in still a different way. Contingency is the very reason why it is so difficult to reach the proper principles of moral science, all of which are uncertain when it comes to applying them here and now--whereas the general principles hold in all circumstances, e.g., good must be done and evil avoided.

             In the Metaphysics, as in the Posterior Analytics, Aristotle treats accidental being only in order to exclude it (except as to what it is in general) from the subject of science. In this connection he shows in the Metaphysics that the true and the accidental (in the sense of contingent) are comparable: both are extrinsic to what is primarily called "being," namely, to what is per se and outside the mind. The consideration of being as true is logical. The Posterior Analytics showed that there can be no necessary discourse about any particular contingent thing or event, whereas there can be true statements about these, e.g., "Socrates is walking." But the subject of Metaphysics is "being," and "being" is divided into necessary and contingent. It seems, therefore, that this science should treat of the contingent as well as of the necessary. Actually, although metaphysics does define what the contingent is and what its cause is, being a science, it cannot extend to what is purely accidental, as was shown in the Posterior Analytics. But the Metaphysics makes the same point. Is this process not circular? It would be if we divorced the logical entirely from the real--second intentions from first intentions. The logician shows that there is contingency on the basis of common experience, but he is not concerned to analyze the nature of contingency; the metaphysician proves that not everything that happens in the world has a per se cause, and that a sufficient cause, once posited, does not necessarily entail its effect. Both the logician and the metaphysician start from common knowledge, but the one examines the logical conditions of demonstration whereas the other examines directly the very nature of real contingency, to show in a distinct and definitive way the reason why there can be no science of the contingent.

             We have dwelt at some length on the role of time and contingency because it is so basic to a large part of the Peri Hermeneias and needs to be particularly stressed nowadays for three reasons: (a) traditional logic takes it all too much for granted; (b) it provides a fine occasion to point out, in a particular instance, how Aristotelian logic is founded, however remotely, on reality without presupposing any distinct knowledge of the subjects of other sciences, and (c) how it differs from without replacing purely symbolic logic.

             What has just been said conveys that we would distinguish between Aristotelian and so-called traditional logic. Indeed, the difference between these two is perhaps greater than that which can be seen between the Organon and logic in the modern sense. For traditional logic, at least as it is usually taught, presupposes distinctly articulated knowledge of reality and, as formulated in textbooks, is seen to be in the end utterly dependent upon metaphysics, which entails that to be understood logic should be the last course in philosophy--a position which St. Thomas declares absurd. Aristotle's logic, on the other hand, requires no such distinct knowledge of real things. All we need to start with is things as commonly named and nominally defined. In fact, if we do not begin in this way the student will never see that in the first operation of the intellect, for instance, to go from "man" to "rational animal" is to go from confused to more distinct knowledge, or to put it another way, from what is known to what was unknown apropos of the same thing. Because it supposes no more than a common confused knowledge, logic, however difficult it may be since it is about second intentions, can and should nevertheless be the first course in philosophy and taught preferably to freshmen, for the development of reasoned knowledge about reality depends upon the student's possession of logical processes of thought.