ARISTOTELIAN METHODOLOGY A Commentary on the Posterior Analytics of Aristotle By J. Athanasius Weisheipl, O.P. (f) S.T.Lr., Ph.D., D.Phil. (Oxon) Pontifical Institute of Philosophy Dominican House of Studies-River Forest, Illinois. ( ) 1958 Edited by John R. Catan Ph.D. (Toronto) Commentary on the Posterior Analytics of Aristotle—Weisheipl 2 INTRODUCTION I. Importance of Methodology Since the time of Francis Bacon (1561-1626) and Rene Descartes (1596-1650) considerable attention has been given to “methodology” by both philosophers and physical scientists (since the 17th Century these have been distinct groups). All at least agree that methodology is concerned with the correct procedure of discovering truth, but not all agree as to what this “method”should be. Forsome the correct method is methodical doubt and justification by clear and distinct ideas (eg. R. Descartes); for others it is the experimental method (e.g., Robert Boyle, Robert Hooker); for others it is the exhaustive classification of individual observations (e.g., Francis Bacon, John Stuart Mill); for still others it is the mathematical method of the physical sciences (e.g., Galileo Galilei, Isaac Newton and most modem physicists). Everyone wants to discover truth. Man has a natural desire to know the truth, not just the facts, but the reasons for their reality. All philosophical investigation begins in curiosity about the unknown. “When we see certain nanifest effects, the causes of which are hidden from us, we are provoked to wonder; and for this reason wonder was responsible for the beginning of philosophy.”(St.Thomas, In I Metaph., lect.3, n.55; cf. Theaetetus 155 D, Clement Alex, strom. VIII, chap. 11, n.60). Illi, qui primo philosophati sunt, et qui nunc philosophantur, incipient philosophari propter admirationem alicuius causae; aliter tamen a principio et modo, quia a principii admirabantur dubitabilia pauciora, quae magis erant in promptu, ut eomm causae cognoscerentur, sed postea ex cognitionem manifestomm ad inquisitionem occultorum paulatim procedentes incoeperunt dubitare de maioribus et occultioribus. {ibid., n.54; cf. St. Albert, I De Praedicab. tr.l, cap.l (Borgnet 1, 2a-b); In Post. Analtx.W, cap. 1, 11 p. 22a-b; J. A.Weisheipl,O.P. “Philosophical Wellsprings," Reality 1 (1950), 8-23) The truth, therefore, which man desires to know is not merely factual truth, i.e., not just a collection of true data, but the explanations of reality. To know the explanation of a thing is to know its causes; this is scientific knowledge. The “scientific method” is the correct procedure to be used by human beings in the discovery and knowledge of scientific truth. Clearly such a method is highly desirable as an instrument of philosophy; and the study of such a method, methodology is important for the acquisition of science and wisdom. II. The Difficulty of Methodology The basic assumption found in modem methodological literature is the conviction that a method that has proved successful in one field of investigation is necessarily applicable to every other field. The mathematical method that has proved successful in mathematical physics is thought to be applicable and proper to the social sciences (moral philosophy); the historical method that is successful for the determination of historical facts is thought to be the method proper to philosophy and theology. This basic assumption cannot be admitted. 1. There is a method proper to each field of investigation, for the proper method of a science depends upon its formal subject and its proper principles. 2. There is a universal method of the speculative sciences in the sense that: Commentary on the Posterior Analytics of Aristotle—Weisheipl 3 (a) there is a general method the human mind must use of necessity to discover scientific truths, i.e., speculative explanations of reality; and that (b) this general method can be used analogically in the various speculative sciences, i.e., the use of this instrument of research depends upon the field of inquiry. This universal method of the speculative science is logic that was perfectly organized (if not invented) by Aristotle in the 4th century B.C. The very name given to the logical works of Aristotle avaA,i)Tixcd indicates its instrumental role as a method of scientific investigation. For this reason Boethius says, “Non tarn est scientia, quam scientiae instrumentum" (Comm, super Porphyium, ed. sec., I, cap.3; cf. St. Thomas, In Boeth. de Trin., q.5, a.l ad 2; St. Albert, Post. Anal., I, tr.I, cap.l, ed. cit., 11. 2b). Although the proper function of logic is to be an instrument and an art, the discovery of the correct rules of logic and the construction of such a universal method is a science because in it we know the proper reasons for the rules. This science is also called methodology, the study of the method of science. 3. From what has been said it is clear: (a) The study of the different method of each science, i.e., the universal method as applied to the different sciences does not belong to logic. “Methodology” is sometimes spoken of as such a study. This kind of examination belongs to metaphysics. b) The study of logic is the study of the universal method of the speculative sciences, which when used in the real sciences (scientiae reales) is used analogically. III. The Order of study If methodology is identical with logic in the sense explained, then we should study the whole of logic. For practical reasons, however, we will concentrate on the Posterior Analytics, which is the apex, the most perfect and only absolutely desirable study among the logical works (St.Albert, I Post. Anal. I. tr. I, ed. cit., I, 2b) Posteriora autem analytica de demonstratione agunt et sunt propterea finis ultimus totius scientiae Logicae (Praef. ed. Leon., I, p.151) Explanation: since all the philosophical investigation originates in wonder over a phenomenon, the explanation or cause of which is hidden from us, the fundamental principle of all method is that we must proceed from the known to the knowledge of the unknown. This, then, is the purpose of logic: .. .docere principia per quae per id quod notum est devenire potest in cognitionem ignoti. (St. Alb. De Praedicab., tr.I cap. 5 ed. cit., 18b). The Posterior Analytics teaches the procedure in discovering necessary principles in order to have certain knowledge through causes, which is scientific demonstration. Hence the Post. Anal., I. has been justifiably considered throughout the ages as a study of Aristotelian methodology. I. The Nature of Logic in General (a) Logic was commonly called Ars artium in the Middle Ages because it is the art by which all other arts and sciences are acquired. It does not consider real being, but “intentions” of Commentary on the Posterior Analytics of Aristotle—Weisheipl 4 real being, i.e., those expressions, mental or verbal, which we use to acquire and express artificially our knowledge of reality. Precisely because logic does not consider reality itself, the Arabians taught logic together with mathematics as a propedeutic to philosophy and not as part of philosophy proper. (Cf. Gauthier, Ibn Rochd, Paris 1948, pp. 47-51; Averroes, In Phys. prooem, Venetiis 1574, IV, fol. r 4; IIPhys., comm.35). (b) In its widest extent logic is concerned with all mental and verbal expressions, i.e., “intentions”, of reality. For this reason it is sometimes called “ars sermocinalis” (cf. ibid., cap. 2, p. 7b; Lib. de Praedicab. tr. 1, cap. 4, ed. cit., 1, 8a; St. Thomas, In Post. Anal. Prooem., n. 6). In other words it is co-extensive with the Trivium. (c) In the strict sense, logic is a method of inquiry, a method of searching for truth, rather than of expressing it, and hence is distinct from Rhetoric, Poetics, and Grammar. As a method it aims at investigating the processes of the mind in proceeding from known to unknown and at establishing certain norms of correct procedure in the attainment of knowledge. St. Thomas defines it: ars directiva ipsius actus rationis, per quam scilicet homo in ipso actu rationis, ordinate, faciliter et sine errore procedat. (In Post. Anal., prooem., n.l) The subject of logic is the method of proceeding from known to unknown, i.e., argumentation. Cum logica sit scientia docens qualiter et per quae devenitur per notum ad ignoti notitiam, oportet necessario quod logica sit de huiusmodi rationis instrumento, per quod acquiritur per notum ignoti scientia in omni eo quod de ignoti notum efficitur. Hoc autem est argumentatio. (St. Albert, De Praedicab., tr. 1, cap. 4, p. 6b) “Argumentatio igiturf says St. Albert following Avicenna and Alfarabi, “logicae docenti proprium subjectum est. ” (ibid.) Argumentation must here be taken in its widest sense to include syllogism, enthymeme, induction and example. Of all these forms syllogistic reasoning is the most perfect. Induction does not give us scire, but intellects. It is scire we desire in science. The enthymeme is merely an abbreviated syllogism; and example is an imperfect induction. Thus syllogistic reasoning is the principal subject of logic. commune = argumentatio subjectum logicae principale = syllogismus 2. The Division of Logic in General (a) The syllogism can be resolved into simpler elements or principles which can be called formal or material because of an analogy to the principles of sensible bodies. (I) Formal principles are the modes and figures which compose the syllogism and which in a sense, make the syllogism to be what it is. Aristotle considers these formal principles of syllogistic reasoning in his Analytica Priora. (II) The matter of the syllogism are the real principles which validate or vitiate the conclusion — “matter”, because they have yet to be determined by the form of the syllogism. Commentary on the Posterior Analytics of Aristotle—Weisheipl 5 (a) If these principles are true, necessary and proper, such a reasoning process yields scientific, demonstrative and certain knowledge. Demonstration is the “pars indicativa logicae”. (b) Probable principles yield opinion and dialectical knowledge. Dialectical principles are common principles and hence do not demonstrate (prove) the conclusion, but indicate that it is very likely demonstrable (provable) if proper principle are found. Dialectics is the “pars inventiva logicae”. (c) False principles, of course, beget sophistical, or pseudo-knowledge, since the syllogistic form is common to all such reasoning, the analysis of formal principles is rightly studied first: the Prior Analytics, (cf. St. Albert, I Post. Anal., tr.l cap. I, p. 3b). The traditional resolution of the syllogism into principles can be represented as follows (cf. St. Thomas, in Post. Anal., prooem., n. 5-6; St. Albert, I Post. Anal., tr. 1, cap. 1, p. 2a; Priora Anal., I, tr.l, cap.l. ed. cit., 1,460a): (consequentiae = forma ... Analytica Priora resolutio seu ANALYSIS necessaria ... Analytica Posterior Incausas (consequentis = principia probabilia ... Topica sophistica ... De sophisticis Elenchis (b) From this it is clear that simple apprehension and judgment are not the principal concern of Logic, for they are not species of its principal subject. Logic discusses the products of the first and second acts of the mind for the sake of the syllogism or argumentation; hence St. Albert says that logic treats simple apprehension and the judgment "per accidens et ratione argumenta¬ tions. " (I) The product of the first act of mind was discussed in antiquity in the following standard texts: (a) Isagoge (seu De Praedicabilibus) by Porphyry. (b) Categoriae (or De Praedicamentis) attributed to Aristotle. (c) De sex Principilis attributed to Gilbert de la Porree. (II) The product of the second act of the mind was discussed by Aristotle in his Peri Hermeneias (or De Interpretatione) Commentary on the Posterior Analytics of Aristotle—Weisheipl 6 Aristotle’'s Posterior Analytics 1. The Text Few have ever doubted that the Posterior Analytics is a very difficult work to understand. Even Themistius, following the Greek text, found many obscurities: Quomodo enim pleraque omnia Aristotelis scripta, quasi de composito, caligine quadam offusa oppletaque habentur, illud potissimum opere hoc (scil. Post. Anal.) intelligitur, cum ob ipsum genus elocutionis, quod, si alicuibi, mixime pressum hie praecisumque est, turn ob capita, quae nullo ordine visunter disposita. (Paraphras. in lib. Post., Praef.) Aristotle’s Posterior Analytics was at least partially translated into Latin by Boethius (470-526). In any case only a partial and corrupt version survived to the 12th century: Translatio Boetti apud nos integra non invenitur et id ipsum quod de ea reperitur vitio corruptionis obfuscatur (From the prologue to an anonymous translation of the Post. Anal., cod. Toledo, Chapter Library 17.14, fol. lr) It was translated anew in the 12th century by James of Venice some time before 1159. James' version became the versio vulgata in the Middle Age; and was often erroneously ascribed to Boethius; it was published with the works of Boethius in PL 64, 711-762. (Cf. L. Minio-Palluelo, "Jacobus Venetus Grecus, Canonist and translator of Aristotle,” Traditio VIII (1952), 265-304). By 1159 Aristotle’s Posterior Analytics was known to the Masters in Paris, but according to John of Salisbury (d. 1182) there was scarcely a Master willing to expound it because of its extreme subtlety and obscurity, fere quot capita, tot obstacula habeat (Metalogicon IV, cap. 6, ed. Webb 171); but John blames this on the bungling mistakes of scribes. John of Salisbury gives us the earliest known exposition of the Post. Anal, in the Latin West. In 1159 an anonymous translator made another version from the Greek without having seen James of Venice’s version, which he condemns for its obscurities. This Versio Toletana (Cambridge, Mass, 1927,223-241) published by Minio-Palluelo (A. L., IV. 2) was little known in the Middle Ages. About the same time, a third translation was made by Gerard of Cremona (d. 1187), this time from Arabic together with the commentary of Themistius. Gerard’s version (A. L., IV. 3) was widely used in the Middle Ages, and it was of great assistance to St. Albert, but it never supplanted the versio vulgata of James of Venice. When St.Thomas wrote his commentary on the Post. Anal, between 1269-1272, he used the version of William of Moerebeke from the Greek (cf. De Rubeis, Diss. XXIII, c. 1-2, ed. Leon. I, cclix-cclxii). Moerebeke’s popular version was a revision of James of Venice’s translation; it is to be found, though contaminated, in most editions of St. Thomas. The medieval schoolmen relied heavily on the commentary of Averroes, translated some time between 1220-1240. The important and influential medieval Latin commentaries on Aristotle’s Post. Anal., are the following: Commentary on the Posterior Analytics of Aristotle—Weisheipl 7 (a) Robert Grosseteste (between 1200-9) Aristotelis Posteriorum Libri, ed. Venetiis 1552, fol. 2r-43v. St. Albert the Great (by 1260) In Posteriorum Analyticorum ed. Borgnet, Paris 1890, t. 11, pp. 1-232. (b) St. Thomas Aquinas (between 1269-72), In Libros Posteriorum Analyticorum, ed. Leonina, Opera Omnia, I, pp. 137-405. Walter Burley (between 1305-1319) Expositio in lib. Post. Anal, not printed. II. The Contents The Post. Anal, is concerned with the manner of acquiring science, i.e., demonstrated knowledge; hence it treats of the demonstrative syllogism and the medium of demonstration. According to Galen (2nd cent.) the original title of the Prior Analytics was ttsqi ovX\oyio\i6v and that of the Posterior Analytics ttsqI drcoSei^efiq. But Alexander of Aphrodisias, a contemporary of Galen, gives their present names as original with Aristotle; and Aristotle himself refers to the Posterior Analytics under its present title in Eth. Nic. VI, c. 5, 1159b26. At present it is impossible to determine the exact title given to the work by Aristotle. At any rate, the First Resolutions, as the Arabs translated it, certainly treats the syllogism, and the second Resolutions are certainly concerned with demonstration. The word "analysis" or "resolution", as Themistius says, is a "redactio uniuscuiusque rei compositae in ea, ex quibus compositio facta est." It is the process of breaking a compositio into its principles, of going from effect to cause. The Post. Anal, show the method of resolving problems to first, immediate, proper and certain causes in order to obtain a scientific explanation, i.e., demonstrative knowledge. III. The Divisions of the Posterior Analytics The divisions of Aristotle’s work are not too difficult in broad outlines. The text is divided into two books and the Arabian commentators generally considered the first a treatment of demonstration, the second of definitionL However, the precise place of definition in Book II is to be a principle of demonstration. Hence it is better to say with St. Thomas that Book I treats of the demonstrative syllogism, and Book II "de medio ex quo syllogismus demonstrative procedit. {In I Post. Anal., lect. 4, n.l) The first chapter of Book I is a propaedeutic to the entire work; it poses the fundamental problem concerning the possibility of learning, i.e., of demonstrative knowledge: either we already know what we are seeking to learn, and thus do not learn, or we do not know what we are seeking, and hence cannot know when we have found it. The last chapter, 19 of Book II indicates the source of all scientific knowledge, the basis of the possibility of demonstrative knowledge: the first principles of human reason, which are acquired from sense experience. The first and last chapters should be considered jointly to appreciate the unity of Aristotle’s work and the force of Aristotle’s answer to Plato’s theory of anamnesis. Commentary on the Posterior Analytics of Aristotle—Weisheipl 8 General Outline of Aristotle’s Posterior Analytics A. The Problem: Is demonstrative knowledge possible? Book I, ch. I B. The solution: I. The Nature of Demonstrative Knowledge. II. 1. Considered in itself: a. What is demonstrative knowledge? Book 1, ch. 2-3b. b. The necessary conditions: i. General conditions required for the principles, ch. 4-5 ii. Conditions required for propter quid demonstration, ch. 6-12 iii. Conditions sufficient for demonstration, ch. 13-15 c. Default of these conditions: i. begets ignorance, ch. 16-18 ii. an infinite regress, which is impossible, ch. 19-21 2. Considered in comparison: a. Comparison of demonstrations one to the other, ch. 24-27 (text 41) b. Comparison of demonstrative sciences, ch. 27-34 (text 42). c. Comparison of science to other ways of knowing II. The Medium of Demonstration. Book II 1. What is the medium of demonstration? Book II, ch. 1-2 a. That there are only four scientific questions, ch. 1 b. That these are all concerned with the medium, si est or quid est. ch .2 2. How is the medium of demonstration to be found? a. How quod quid est and propter quid are related to demonstration. i. Relation of definitio (quod quid est) to the demonstration, ch. 3-10 ii. Relation of cause {propter quid) to the demonstration, ch. 11-12 b. How definitions and causes are to be found. i. Discovery of definitions, ch. 15 ii. Discovery of proper causes, ch. 14-18 C. The Ultimate Basis for the Possibility of Demonstration, ch. 19 * Chapters 9 to 19 follow the divisions of the Greek text (found in Ross and English translation), rather than the Renaissance division found in the editions of Saint Thomas. Commentary on the Posterior Analytics of Aristotle—Weisheipl 9 A BRIEF COMMENTARY ON ARISTOTLE’S POSTERIOR ANALYTICS A. The Problem (or The Necessity of Demonstration) Book I, ch. 1, 71al-71b7 1. The Problem of Learning Aristotle’s opening statement that “all learning and intellectual instruction proceed from pre-existent knowledge” was acceptable to everyone concerned—to Plato, to the Sophist, Nominalists, and to Aristotle himself. Throughout the treatise, however, Aristotle intends to give far more importance to the preposition from than could have been conceded by Plato or the Sophists. Only later in this chapter does Aristotle reveal the precise meaning he sees in this fundamental statement, and the entire treatise is an explanation of his own view. Hence this first chapter sets the theme of the Posterior Analytics by presenting the problem: Is scientific learning possible? In the Meno (80 D-86D) Socrates attempts to inquire into the nature of virtue, a subject about which he admittedly does not have full knowledge. Meno intervenes and objects that all inquiry is impossible, for “a man cannot inquire either about that which he knows, or about that which he does not know; for he does not know the very subject about which he is to inquire.” {Meno 80E) Plato through the voice of Socrates solves the dilemma by means of his doctrine of Recollection: “all inquiry and all learning is but recollection” {ibid. 8ID). Inquiry consists in ordering the questions in such a way as to clear the obstacles to remembering the answer already possessed. In Plato’s view the answer is actually known {simpliciter) and only secundum quid unknown because of imprisonment in this body. All knowledge was acquired in a previous life; present imprisonment impedes facile recalling. Plato admits clearly that all learning proceeds from pre-existent knowledge, but in the sense that all learning is already actually possessed. The Sophists of the Academy attempted to answer Plato’s dilemma by a nominalist solution (cf. Arist., Post. Anal., I, c. 1,71a 34-b3), saying that all learning is simply an aggregation of individual observations. They know only what they have actually experienced, and in a so-called demonstration the minor term is already known to be in the extension of the major term. Learning is possible only in the sense of a completely new addition. This view is very much like John Stuart Mill’s scientific method without the possibility of true demonstration. The formal refutation of both Plato’s idealism and the Sophist’s nominalism belongs to metaphysics, as St. Albert observes, for it is a question of the reality of universals (I Post. Anal., tr. I, cap.6, p. 18a). Robert Grosseteste is correct when he says: ...cum in hoc libro non possit demonstrare quoniam scire sit, quia nullius artificis est stabilire suum subjectum, vel quod est ante suum subjectum, ne artem transgrederetur, vel dubium aut falsum penitus sine ratione supponeret, necesse habuit Arist. ut ante initium huius scientiae saltern modum ostenderet quo convenit scire esse, et aliquid addiscere, et solvere opiniones destruentes scire et addiscere. (Prooem., fol. 2ra) Commentary on the Posterior Analytics of Aristotle—Weisheipl 10 Just as in the Physics, Aristotle establishes the subject of his inquiry by appealing to experience: we learn by proceeding from known to unknown, i.e., by argumentation. The dilemma of Meno is similar to that of Parmenides concerning the possibility of motion (cf. Phys., I, c. 8, 191a 29-32). And Aristotle’s solution to both dilemmas is fundamentally the same. In a certain sense the object of inquiry is known, and in a certain sense it is not known. It is unknown in actu, but it is known in potentia. For Aristotle the end simpliciter incognita, otherwise there is no learning, but only remembering. However, that for which we search is already known in some way, secundum quid, otherwise we would not know what we are looking for and all investigation necessitates going from what we do know to what we do not precisely in virtue of and in the light of what we know, that is, we must proceed from known to unknown. But this is the process of demonstrative reasoning. Therefore all learning is by way of demonstration. 2. The Beginning of Learning: The Praecognita All learning begins with a question, or a problem. Unless one raises a question, or has a problem there is no point in giving him the answer—it is not learning. A question is the expression of man’s curiosity, his desire to know the unknown. Aristotle analyses this situation and sees that there are at least three things that are already known when a question is asked. It is in the light of these three that we pursue our inquiry, which is the process of learning. a. First, we start out with the natural principles of human reason, such as, ‘the whole is greater than any of its parts, and ‘we can really affirm or deny one thing or another’, such principles are natural to every man, and without them learning and demonstration are impossible (cf. Post Anal, II, c. 19, 99b 20-21). That these primary truths are acquired through sense experience is shown in chapter 19 of Book II. Even Plato admits that we have such principles; what he denies is that they are acquired through the senses. Thus we must know—and we do not know by nature—that such common first principles are true. b. Second, the very question we ask implies some knowledge of the subject of inquiry, namely that it exists and something of its nature, i.e., quia est and quid eSt c. Third, before we can ask the question we necessarily know at least the nominal definition of the predicate, i.e., quid nominis. It must be noted with Cajetan (IPost Anal, c.l, Hispraesupositis, Venetiis 1506, fol. 43rd), that Aristotle is here speaking only of questions of fact, quia, not of propter quid questions. For example, if I want to know whether the soul is immortal, I obviously know that there exists such a thing as the soul and I know something of what it is. Likewise, I must have some idea of what ’immortality’ is, or else I could not ask such a question. Cajetan points out that Aristotle is clearly speaking of quia knowledge when he says that we must know quid nominis, for “in ea questione quae terminat questionem propter quid medium est diffinitio quid rei ipsius passionis.” The common first principles, the subject and the predicate of the question are called principles of investigation and of demonstration because they all shed light upon the demon¬ stration: common principles shed a remote light, the subject and predicate illumine by making up the syllogism by being incorporated into the reasoning process. The most important principle of demonstration is obviously the medium of demon¬ stration. But since this is what investigation expects to find, we cannot have pre-existent