At the beginning of The Groundwork of the Metaphysics of Morals, Immanuel Kant writes:
Ancient Greek philosophy is divided into three sciences, physics, ethics, and logic. This division perfectly conforms to the nature of things and we can hardly improve it, except by articulating its founding principle, so that we can be certain of its completion, on the one hand, and on the other hand, can determine exactly the necessary subdivisions.a
This text of Kant suffices, it seems to me, to show the importance of the role that ancient theories of the division of the parts of philosophy have played in the history of Western thought. I do not intend in the present article to treat all the aspects of this vast theme. It will not be possible for me to speak of all the philosophical schools, nor of all the subdivisions of philosophy that they have proposed. I do, however, think that it could be interesting to characterize the fundamental types of ancient classification by reviewing the conceptual structures, the ‘Denkformen’ which underlie them, and the conceptions of philosophy that they imply.
These types of classification seem to be reducible to three. These three types, in their own ways, each aim to give a complete classification of philosophy. The first is distinguished by the effort to identify the specificity of the object and the methods proper to each discipline. To do this, it uses a method of division, ordering the subdivisions under the form of a conceptual pyramid. It establishes a hierarchy between the parts of philosophy which corresponds to the hierarchy of their objects. The second type of classification is less interested in the specificity of the parts of philosophy than in the solidarity between them. It tries to grasp the correspondences or the linkages which connect them so as to better show the systematic unity of philosophy. The conceptual structure underlying this type of classification is then no longer the pyramid, but the image of a circle or a living organism. Finally, there is a third type of classification. The latter does not exclude the two others and brings in the temporal dimension, that of the succession of stages the student must pass through; or, in a word, the pedagogical dimension. The classification of the parts of philosophy is made here as a function of the stages of paideia: the degrees of spiritual progress of the student. Such a mode of classification will aim at establishing a programme of studies and an order for the reading of certain, specified texts. The underlying metaphor will, this time, be that of the phases of Eleusinian initiation.
These three are the fundamental ancient forms of the divisions of the parts of philosophy that the present essay wants to consider.
The first kind of classification of the parts of philosophy appears in the Platonic milieu. Its appearance is, moreover, tied to a reflection on the scientific method and on the necessity to introduce a classification of the sciences and their parts. The term ‘philosopher’, Aristotle notes, is employed like the term ‘mathematician’.1 Aristotle means that in the same way that the term ‘mathematician’ can designate someone who does arithmetic or geometry or even someone who does astronomy, so the term ‘philosopher’ can designate someone who does natural philosophy, ethics or theology. As Aristotle adds: ‘the science of mathematics also has its divisions: there is a primary and a secondary mathematics, and other kinds which follow, each in their domain’.2 In the Platonic milieu, the division of the parts of philosophy is thus intimately tied to the division of the sciences in general, given that science and philosophy are not yet clearly distinguished, except that Aristotle proposes the principle, according to which a science is more philosophical if it is more theoretical and more universal.3
In Plato, reflection on the scientific method is manifest for example in the Republic,4 which opposes the method of mathematics to that of dialectic. Plato remarks that the proper task of the dialectician – that is, of the philosopher – is to give us an overview of the different mathematical sciences.5 In the Statesman, Plato uses the method of division to define the science of politics, beginning from the basic opposition between the theoretical and practical sciences, in a manner which is, nevertheless, somewhat ironic.6
But it is in Aristotle that one finds the best example of the first type of classification of which we have spoken. The division of sciences proposed in book Epsilon of the Metaphysics7 is presented as a conceptual pyramid obtained by the method of division. One again finds there, from the start, a fundamental opposition between theoretical sciencesc and practical sciences. Theoretical sciences relate to those objects which do not depend upon us, and the practical sciences address those objects which do depend upon us, because the principle of their movement is found in us.8 The practical sciences are then subdivided into ‘practical’ and ‘poetic’ sciences (that is to say, productive sciencesd), based on whether they engender change within the agent or, on the contrary, on works exterior to them. As for the theoretical sciences, these are subdivided into sciences which relate to an unchanging object and sciences which relate to objects capable of motion.e9 The first are distinguished according to the following principle. If this unchanging object subsists in itself, it is a matter for theology. But if it is not unchanging, except when it is separated by a process of abstraction from its matter, then it is a subject for mathematics. The other branch of the division, the science which relates to an object which is capable of motion and which subsists, corresponds to physics.
This classification, which results from applying the method of division, constitutes a hierarchy of the sciences which is founded on the hierarchy of their formal objects:10 that is to say, on the hierarchy of the modes of being that the mind discovers in reality. This hierarchy of objects corresponds to a hierarchy of methods: the most elevated sciences utilize a method more exact than the inferior sciences.11 The practical and political sciences are inferior to the theoretical sciences12 because they relate to contingent objects, human actions; whereas the theoretical sciences relate to being. Amongst the theoretical sciences, the mathematical sciences are superior to the physical sciences, because they abstract from material becoming and retain only an intelligible matter. But the mathematical sciences are inferior to theology because, in the first place, like physics, they relate only to a kind of determinate beings,13 whereas theology has for its subject ‘being qua being’, and second, because mathematics does not totally abstract from matter. Theology appears, therefore, as the first science, supreme and universal.
This classification does not suppose only the Platonic method of division. It is situated within a Platonic problematic and corresponds to conceptual structures which are typically Academic.14 One will firstly recognize the fundamental opposition attested to in the Statesman15 between practical sciences and theoretical sciences. On the other hand, one will easily rediscover, in the hierarchy of objects considered by the theoretical sciences, the Platonic hierarchy which ascends from natural objects towards ta mathêmatika and from ta mathêmatika towards the Ideas, and which appears clearly in the Republic in relation to the ascent from doxa, via dianoia to noêsis.16
Nevertheless, in presenting the classification of the parts of philosophy which we are examining in book Epsilon of the Metaphysics, Aristotle does not purely and simply reproduce a Platonic classification. On the contrary, beginning from the Platonic problematic, he seeks to define the originality of his own doctrine: and, above all, his idea of the supreme science which he elsewhere calls first philosophy and that he here calls theology, which he wants to substitute for Platonic dialectic. One must not represent the classification proposed in book Epsilon as a programme of studies that Aristotle will have defined once and for all, to organize his teaching and to shape the plan of his works. We know that the tradition of Aristotelian commentators understood the text in this manner and that they classified Aristotle’s works conformably to what they considered to be a programme of teaching responding to the natural hierarchy of the objects of science. If we accept this supposition, we are compelled also to state that Aristotle himself, nevertheless, does not respect the distinctions which he establishes in this classification. First of all, the exact place of mathematics, and more specifically of astronomy, remains unclearly defined within his oeuvre.17 Above all, the frontiers between physics and first philosophy are not always distinctly delimited: ontology and physics, theology and physics are often interconnected in Aristotle’s physical treatises as in those on first philosophy, so that one often passes almost imperceptibly from one to the others.18 In addition to that, the scheme of book Epsilon says nothing of the place of dialectic and of the analytics in the ensemble of the sciences; while it is necessary to say that many of the questions concerning first philosophy are equally treated in the treatises dedicated to analytics.19 Finally, the real complexity of first philosophy does not appear in the schema of book Epsilon. This first philosophy is at the same time general ontology, the theory of substance (ousia), the study of principles and theology.20 To put it differently, the teaching of Aristotle in its concrete reality and the content of his books does not correspond to the rigorous classification proposed in Metaphysics Epsilon.
In fact, the goal of the schema of book Epsilon is not to propose a teaching programme. It is to give a definition of the supreme science within the framework of this classification. Through the method of division, Aristotle’s classification eliminates everything that does not belong to this supreme science. The latter is not a practical or productive science, because such sciences are inferior to the theoretical sciences.21 It is not physics, as the pre-Socratics had claimed. In effect, physics has, for its subjects, a determinate genus of being, whereas the supreme science must be universal; and this determinate genus of being is ousia in motion,22 which presupposes an unmovable ousia which precedes it. The supreme science can no longer be identical with mathematics, as certain Platonic philosophers had maintained. As Aristotle says elsewhere: ‘mathematics has become for today’s philosophers the whole of philosophy, even when they say that is necessary to undertake mathematics only in view of another goal’.23 Indeed, mathematics, like physics, is related only to a determinate genus of being, and if the mathematical sciences consider their objects as unmovable, it is by abstracting them from the sensible ousia.24 That which remains after these divisions and exclusions is the first philosophy or supreme science that will be a science of being, as such, taking account, at the same time, of the essence and existence of all things, on the basis of the affirmation of the existence of a single unmovable, immaterial and eternal ousia.25 By defining the supreme science in this way, Aristotle radically distinguishes his theology from Platonic dialectic. For Plato, dialectic, by means of a technique of analysing discourse tied to dialogue, comes to the definition of things and thus of Forms or Ideas which are the foundations of the structure of all reality. But Aristotle refuses to consider the Ideas as ousiai. It follows that Platonic dialogue loses all scientific value in his eyes, because all science must relate either to a genus of being or to being qua being. Aristotle thus considers dialectic to be a simple technique of argumentation by questions and answers. It allows us to speak of everything, but never teaches us anything, because it is content to argue beginning from accepted opinions and common notions, without caring for truth. There is thus for Aristotle a radical opposition between dialectic and philosophy.26 This total change in the content of the definition of dialectic will provoke many confusions in later philosophy.
The first type of classification of the parts of philosophy, which we are exemplifying by reference to Aristotle, is therefore characterized by the attention that it dedicates to defining exactly the specific method of each science. This methodological attention was manifested, as we have seen, in the opposition described by Plato in the Republic,27 between the method of dialectic and the method of mathematics. It is in virtue of the same concern that the Timaeus insists on the necessity of resorting to the method of the ‘likely story,f when it comes to treating of physical things.28 This is also the case in Aristotle, for whom the division of sciences corresponds to a methodological concern that is even more developed than in Plato. In effect, Aristotle thinks that each science must develop its arguments starting from its own distinct principles, and by taking account of those aspects peculiar to the object that it considers. Therefore, ethics has its own distinct method,29 as equally does physics.30 In the context of the classification proposed in book Epsilon of the Metaphysics, Aristotle, for example, insists on the fact that the definition of physics must include a consideration of the matter in which the form is engaged.31 Aristotle reproaches Plato, therefore, for using the same method everywhere, the dialectical method for the analysis of concepts, whether in ethics or in physics, without taking account of the differences between these modes of being.32 This is what Aristotle calls reasoning in a purely formal manner: logikôs; a method which he opposes to that which begins from the nature of things: physikôs, or that which refines the principles proper to a scientific domain: analytikôs.33
Certain evidences allow us to suppose that the ancient Academy’s new classification of the parts of philosophy is simpler than that which we have so far been exposing. In fact, some later testimonies attribute the tripartite classification of philosophy into ethics, physics and dialectics (or logic) to Plato,34 others to Xenocrates.35 If we leave aside for the moment the problem posed by the use of the term ‘logic’, one could think that it is still a matter of a classification obtained by the method of division, opposing first of all practical science (ethics) and theoretical science; then, subdividing the realm of theoretical science into the science of the sensible world (physics) and the science of Forms (dialectic). Effectively, if one considers the catalogue of the works of Xenocrates,36 one could say that we find an entire group of ethical works, another group explicitly consecrated to physics and, finally, an ensemble of works that corresponds roughly to traditional Platonic dialectic. We must suppose then that the mathematical sciences, by virtue of their objects, belong to dialectic (in the Platonic sense),37 which is not impossible. It is thus likely that this threefold division of ethics, physics and dialectics existed in the Ancient Academy. We find here, again, the hierarchical system which ascends from the level of human contingency towards that of divine transcendence.
We are therefore enumerating ethics, physics and dialectic. But did the Ancient Academy employ the term ‘logic’ to designate Platonic dialectic? I think that this is very doubtful, for the following reasons. It seems sure that the Stoics had been the first to employ the word ‘logic’ (to logikon meros) to designate a part of philosophy, and that the presence of the word ‘logic’ in the later testimonies we have cited only reveals the influence of the Stoic vocabulary. This was the opinion of Rudolf Hirzel, exactly one hundred years ago, in his article, ‘De Logica Stoicorum’.38 In effect, there did not exist before the Stoics any text, any book title, whether Platonic or Aristotelian, which could attest to the use of the word ‘logic’ to designate a part of philosophy.
One might object to this position by citing the famous text of Aristotle’s Topics, which distinguishes physical premises from ethical and logical premises. But, despite these appearances, this text could not allude to a true division of the parts of philosophy. First of all, as Aristotle himself says in this context, the division in question is only an approximative method39 to classify received opinions or different contentions that the future dialectician will glean from the course of his lectures, with the goal of making up a collection of premises serviceable in discussion. It is a matter of the ordering of a file or a notebook. But, above all, we must interpret logikos here according to the general sense it acquires with Aristotle.40 We have seen just now41 that Aristotle employs the term logikos in opposition to a ‘physical’ or ‘analytic’ method, to designate a purely formal method, founded on the analysis of a definition (logos)42 and not on the distinct principles of a specific science. The term logikos does not designate a discipline on the same plane as that of ethics or physics, but a purely formal process which can be utilized in ethics as well as in physics. Therefore, I think that one can, with Alexander of Aphrodisias,43 compare this classification of ethical, physical and logical propositions with that of dialectical problems proposed by Aristotle in the Topics,44 some pages earlier. One again finds here, on the one hand, the fundamental opposition between ethical problems and logical problems, and, on the other hand, the indistinct category of problems which are, according to Aristotle, only instruments that allow us to discuss ethical or physical problems. Let us add that it seems unlikely that Aristotle, in a text like the Topics, where he treated of dialectic in the Aristotelian sense (that is to say, as a technique different from philosophy), would employ the term ‘logic’ to designate dialectic in the Platonic sense (that is to say, as philosophy par excellence). Moreover, the word ‘logic’, in these texts of the Topics, is not at all synonymous with ‘dialectic’, since the ‘logical’ problems or premises are only one part of the ‘dialectical’ problems or premises. One could say that Aristotle, the inventor of ‘logic’ in the modern sense of the word, never employed the word ‘logic’, but used the words ‘dialectic’ and ‘analytics’ to designate his invention.
The second type of classification of which we have spoken in the introduction appears with the Stoics. By dividing philosophy into logic, physics and ethics, they perhaps revive a prior classification, but they give it an absolutely new sense, both because of the content of the different parts of philosophy and because of the mutual relations established between them. First, it seems to me that the idea of a ‘logic’, as part of philosophy bearing this name, is something new. This logic encompasses rhetoric and dialectic. But the dialectic here differs at the same time from both Platonic and Aristotelian dialectics. It is no longer Platonic dialectics, since it does not aim any longer at the Forms or Ideas in themselves.45 On the other hand, it is no longer Aristotelian dialectic, since, for the Stoics, dialectic (and rhetoric also) is no longer a simple technique of argumentation belonging always to the domain of the probable. It is a science which, beginning from common notions46 and commonly accepted opinions, elevates itself towards certitude and knowledge of the truth. In any case, this is the position of Chrysippus, for whom dialectic is the science of true judgement and one of the virtues of the sage.47 On the other hand, by suppressing Platonic dialectic as a science of Forms, the Stoics situate all theoretical activity in physics. Stoic physics absorbs theology, which is consistent with an enlargement of the notion of physis. This now no longer designates a particular domain, as it did for Aristotle, but rather the totality of the cosmos and the force which animates it.
One might think that this Stoic tripartition of philosophy, like the Platonic and Aristotelian classifications, has a hierarchical character. One could say that physics represents the highest discipline because it is related to the world and to the gods; that ethics would be a subordinate discipline because it relates only to human action; and, finally, logic could be thought of as the lowest discipline because it is related only to human discourse. Certain ‘pedagogical’ presentations of the Stoic system of the parts of philosophy can give us this impression, and we will return to this thought later.48 But – and this is where we arrive at our examination of the second type of classification – the internal necessity of the Stoic system leads it inevitably to substitute for this hierarchical representation that of a dynamic continuity and reciprocal interpenetration between the parts of philosophy. This unity of the parts of philosophy is founded on the dynamic unity of reality in Stoic philosophy. It is the same Logos which produces the world, which illuminates human beings in their faculty of reasoning and which is expressed in human discourse, all the while staying fundamentally identical to itself in all the degrees of reality. Physics thus has for its object the Logos of universal nature. Ethics, in turn, has as its object the Logos in the reasonable nature of human beings. Finally, logic examines this same Logos as it is expressed in human discourse. From one end to the other, it is therefore the same force and the same reality which is simultaneously creative Nature, the Norm of ethical conduct and the Rule of discourse.49 The method of Platonic dialectic and the Aristotelian method of abstraction allow us to establish a difference of levels between sensible reality and the Forms or Essences. They thereby founded a hierarchy between the parts of philosophy. However, as Émile Bréhier has said, remarkably:
[One finds] no methodological procedure of this kind in the Stoic doctrine;g it is not a matter of eliminating the immediate and sensible given, but rather of seeing Reason embodied in it … It is in the sensible things that reason acquires the plenitude of its reality.50
The difference between levels of reality diminishes and, with it, the difference of level between the parts of philosophy. As Bréhier again has said:
it is one and the same reason which, in dialectic, links consequents to their antecedents, in nature binds all causes, and in human conduct establishes a perfect concord between actions. It is impossible that the good man would not be a physician and a dialectician. In this sense, it is impossible to realize rationality separately in one of these three domains and, for example, to entirely grasp reason in the march of events in the universe, without realizing reason in one’s own conduct at the same time.51
This reciprocal interpenetration corresponds to a model of relations dear to the Stoics, of which one finds other examples in their physics, regarding the enchainment of causes, as well as in the ethics, concerning the relations between the virtues. This is the relation they called antakolouthia.52 The parts bound by such a relationship are mutually implicated and are only to be distinguished by the particular aspect which gives them their name. Thus, each virtue is all the others, and it is only distinguished by the predominance of a particular aspect. Conformably to this model, one can also say that logic implies physics, because dialectic implies the idea of the rationality of the enchainment of events;53 and that logic also implies ethics, since, for the Stoics, dialectic is a virtue which itself encompasses the other virtues: like, for example, the absence of precipitation in judgement or circumspection.54 For, generally speaking, moral good and bad are an affair of judgement.55 Conversely, physics and ethics presuppose logic since, as Diogenes Laertius says: ‘all the themes of physics and ethics can only be examined by resorting to a discursive exposition’.56 Ethics implies physics because, according to Chrysippus: ‘the distinction between goods and evils derives from Zeus and universal Nature’.57 Physics, finally, implies ethics to the extent that the knowledge of the physical world and of the gods is the end of our rational nature,58 and since the perception of the rationality of events implies the rationalization of moral conduct. Evidently, in the Stoic system, there is a reduction of all the parts of philosophy to ethics: the three disciplines of philosophy are defined as virtues59 and it is their reciprocal co-implication which constitutes wisdom.60 Wisdom is indissolubly ethical, physical and logical. The distinction between the three parts only stem from the relations of the sage with the cosmos [i.e. physics], with other men [i.e. ethics] and his own thought [i.e. logic].61 Philosophy as ‘an exercise of wisdom’ consists, therefore, in a constant and simultaneous practice of the three disciplines, as Marcus Aurelius affirms in this meditation: ‘in a constant manner and, if possible, on the occasion of each representation which presents itself in your mind, practice physics, the theory of the passions [that is to say, ethics], and dialectic’.62 This somewhat enigmatic formulation of Marcus Aurelius is explained, I think, by the evolution of the theory of parts of philosophy in the epoch of Epictetus and Marcus Aurelius: the reduction to ethics is here even more accentuated. Epictetus distinguishes, in effect, three domains of askêsis: the discipline of desires, the discipline of impulses,h the discipline of thoughts. In the first domain, one exercises oneself in making one’s desires conformable to the will of universal Nature; in the second domain, one attempts to make one’s actions accord with the will of the rational nature that is common to all men; while in the third domain, one attempts to conform one’s thoughts to the laws of reason.63 We understand, thus, that these three domains correspond, in fact, to the three parts of philosophy: the first to physics, the second to ethics, the third to logic.64 In all three cases, it is a matter of a spiritual exercise: physics as a spiritual exercise makes us aware of our place in the cosmos and makes us accept events with love and complacencyi towards the will of the universal Logos;j ethics as a spiritual exercise aims to make our actions conform to the fundamental tendency of human nature inasmuch as it is rational, that is to say that ethics makes us practice justice and love towards our fellows; finally, logic as a spiritual exercise makes us critique at each moment our representations, so that no unreasonable judgement will be introduced into the chain of our thoughts. In this perspective, the three parts of philosophy are only three aspects of the fundamental spiritual attitude of the Stoic: vigilance. This is the sense of this thought of Marcus Aurelius:
Constantly and everywhere, it depends on you to accept piously the present conjunction of events [this is physics as spiritual exercise], to conduct yourself with justice towards your contemporaries [this is ethics], to apply to your present representations the rules of discernment so that nothing can enter them except that which is objective [this is logic].65
Here again, the simultaneity of the three philosophical activities in the present instant appears clearly.
The Platonic manuals of the Imperial epoch, probably under the distant influence of Antiochus of Ascalon, attempted a synthesis between Aristotelianism, Platonism and Stoicism. They remained faithful to the spirit of Stoicism when they recognized the threefold structure of philosophy as the foundation of its systematic character. One finds this theme in Diogenes Laertius, Apuleius, Atticus and Augustine.66 They attribute to Plato himself the merit of having made philosophy a corpus, a living organism, complete and accomplished,k by binding together dialectic to physics and to ethics. Some of them (Atticus and Apuleius) leave to Plato only the merit of having assembled three pre-existing disciplines: pre-Socratic physics, Socratic ethics and Eleatic dialectic. Others (Diogenes Laertius and Augustine) see in Platonic dialectic a systematizing element invented by Plato in order to synthesize the two elements which Platonism integrates: Pythagorean physics and Socratic ethics. Plato appears, then, as a synthesis of Socrates and of Pythagoras. Indeed, the utilization of the Stoic schema in the presentation of Platonism resulted in a total deformation of the latter. What is most serious is not that the theory of principles (God, the Ideas, Matter) would be placed within physics (after all, the Timaeus, considered, in general, as a ‘physical’ dialogue, could justify this placement).67 Above all, in the Stoic presentation of Platonism, dialectic loses its character of a supreme science aiming at the absolute principle to become only an art of making distinctions and invention in discourse68 or an inquiry into the exactitude of denominations, to quote the expressions of these authors.69 Only Clement of Alexandria, Plotinus and, finally, Augustine70 preserve the distinction between Stoic and Aristotelian dialectic and Platonic dialectic; and they declare that only the latter has for its object true, that is to say divine, realities.
The unitary and systematic schema of the division of philosophy proposed by the Stoics, taken up in the Platonic manuals, finds its highest form and reveals all of its potential in those theories which present God as the common object of the three parts of philosophy. They carry forward the fundamental intuition of Stoicism, according to which the Logos is the common object of the parts of philosophy. According to Clement of Alexandria,71 physics takes God as ousia as its object; ethics has God as the good as its object, while the object of logic is God as intellect. According to Augustine,72 physics has God as the cause of being as its object, while logic considers God as the norm of thought, and ethics takes God as rule of life. This Augustinian order of physics, logic and then ethics corresponds to the order of divine Persons in the Trinity. The Father is the Principle of Being; the Son, Intelligence; and the Holy Spirit, Love.73 The systematic unity of the parts of philosophy reflects here the reciprocal interiority of the divine Persons.
The third type of classification introduces this time a complex datum:l the pedagogical dimension, which implies a method of exposition, a temporal order, a succession of moments, involving an intellectual and spiritual progression. It is a matter of establishing a programme of teaching philosophy which takes into consideration simultaneously the logical order of notions and the capacities of the auditor.
Apparently, this third type of classification adds nothing to the content of the divisions previously established but an extrinsic order of presentation.m One might think that the two first types of classification could be considered either in themselves, in a purely formal manner, or in relation to the auditor, in the pedagogical perspective. It is thus that the commentators on Aristotle at the end of antiquity, in their introductions to the Categories, on the one hand, faithfully present the Aristotelian division of the parts of philosophy, notably the subdivision of theoretical sciences into theology, mathematics and physics; and, on the other hand, expose the cursus which must be traversed by the disciple, a cursus which corresponds in principle74 to the inverse order, since, according to Aristotle,75 there is a radical opposition between the ontological and the pedagogical orders, between what is more knowable in itself and that which is more knowable for us. It can seem logical that an ascending hierarchy of teachings corresponds to the descending hierarchy of the sciences. In the same way, if one considers the Stoic system, one could say that the pedagogical point of view alone serves to introduce some order between the three parts of philosophy, of which we have shown the intimate solidarity and interpenetration.
Nevertheless, it seems that this third type of classification could really be opposed to the two others. The first two types correspond, in effect, to a purely ideal position: they suppose the totality or integrity of knowledge or wisdom as an already realized accomplishment. By contrast, the third type corresponds to the concrete reality of philosophical activity. Philosophy is not given, once and for all. It is realized in communication: that is to say, in the explication, the ‘discourse’ which exposes it and transmits it to the disciple. This philosophical ‘discourse’ introduces the temporal dimension which has two components: the ‘logical’ time of discourse itself (to take up an expression of Victor Goldschmidt76) and the psychological time required by the formation, the paideia, of the disciple. The ‘logical time’ corresponds to the internal exigencies of expression or explication. To be communicated to the disciples, philosophy must be presented in a discursive manner, and thus through a succession of arguments which imposes a certain order. This thing must be said before that other [thing]. This order is that of the ‘logical’ time. But the exposition is addressed to an auditor, and this auditor introduces another component, namely the phases or stages of his spiritual progress. Here it is a matter of a properly psychological or at least of a pedagogical temporality. To take an example drawn from modern philosophy, we see Descartes77 consider this pedagogical aspect when he advises his readers to employ ‘some months or, at least, some weeks’ in the examination of his first and second Meditations. What is at stake clearly appears to be a temporality other than the ‘logical’ time. It is a time for the maturation, the assimilation of the text. As long as the disciple has not assimilated such and such a doctrine, it is useless or impossible to speak to him of other matters. This is why Descartes speaks only of universal doubt in his first Meditation, and of the nature of the spirit in the second. Here the pedagogical requirements influence the doctrinal content of the work.
The two components which we have just described, logical time and psychological time, define the third type of classification of philosophy which we will now discuss. These two components, as we can now glimpse, can profoundly modify the content and the sense of the parts of philosophy. It is equally true that there is a permanent conflict between the two components which we are analysing, as we will observe. It is extremely difficult to safeguard the logical order if one wants to take into account the spiritual state of the auditor.
In Plato and Aristotle, we already find these important pedagogical concerns, which will have a great influence in later tradition. However, we do not find a complete classification of the parts of philosophy inspired by the perspective of the spiritual progress of the student. We have already mentioned the Aristotelian opposition between l’ordo essendi and l’ordo cognoscendi,78 but we have equally remarked that we do not find in his works a concrete programme of philosophical teaching.79 In Plato, we find an outline of such a programme, when the Republic80 recommends that future philosophers devote themselves to the mathematical sciences before turning to dialectic. But it is not certain that the concrete teaching of the Academy81 conformed itself to these theories concerning the organization of education in the ideal State. In any case, Plato is particularly sensitive to the problem of the necessary adaptation between teaching and the spiritual level of the disciple. In the Republic, he thus signals the danger to which young men are exposed who begin to practice dialectic too soon.82 The very idea of spiritual progress is expressed by Plato in two metaphors which will have a considerable posterity: first, the conversion of the prisoners of the cave who are little by little habituated to contemplate the daylight;83 and, second, the initiation into the mysteries at Eleusis (telea, epoptika) of which Diotima in the Symposium enumerates the degrees, when she describes the ascension of the soul towards the Beautiful itself.84
It is with the Stoics that we encounter for the first time an explicit discussion concerning philosophical pedagogy which considers the order and the content of the parts of philosophy. We have mentioned above one important aspect of their doctrine: the mutual implication of the three parts of philosophy whose simultaneous exercise constitutes wisdom. That was only an ideal situation. Let us now turn to another aspect of their doctrine: the distinction and the succession of the parts of philosophy because of the necessities imposed by the requirements of philosophical teaching. It is probably from this perspective that the Stoics claimed that the parts of philosophy are, in fact, not the parts of philosophy itself, but rather the parts of discourse or the logos involved in philosophical teaching.85 It is the discourse of teaching which requires an order and a particular succession. The Stoics are very aware of this pedagogical imperative. Thus, they attempt to fix in a rigorous manner the logical order of the arguments within each part of philosophy,86 as it appears in the summary of Stoic dialectics given by Diogenes Laertius,87 as well as in Cicero’s exposition of the ethics in book III of De Finibus.88 In this last example, it is striking to note that the rigorous progression of thought is at the same time a spiritual progression which, beginning with a physical theory of natural impulsen – that all beings tend to self-preservation – shows how this natural impulse in human beings becomes, at the level of reason, a love of humanity later culminating in wisdom.89 Above all, the Stoics try to determine an order of teaching of the parts of philosophy. Diogenes Laertius and Sextus Empiricus90 report to us the different theories which were current in the school. We would generally agree to place logic at the beginning of the cursus of studies – although Epictetus reserves it to those making progress. Nevertheless, one hesitates about the place of ethics and physics, respectively.91 This hesitation appears clearly in the interpretations that Diogenes Laertius and Sextus Empiricus92 give of the famous comparisons that the Stoics made between the parts of philosophy and organized ensembles such as the egg, the garden or the living being. If logic is always presented as the part which assures the solidity, ethics and physics often exchange their role. Each differently appears as the most important and indispensable part.o
These discussions and hesitations can be explained, as Adolf Bonhöffer has shown, by the differences concerning the viewpoint – logical or pedagogical – according to which the classifications are carried out.93 One could equally well say that physics must precede ethics by founding it logically on the knowledge of the rationality of nature, or that ethics must precede physics so as to prepare the soul for the contemplation of nature.94 Chrysippus himself uses the image of Eleusinian initiation when he places physics last as a revelation (teletê) concerning the gods.95 It is in the same spirit that Sextus Empiricus gives us an account of the order of the parts of philosophy: the spirit must be fortified by logic, then the student’s way of living is ameliorated by ethics, before we can finally approach the most divine objects that physics examines.96
However, it would be too easy to think that the order of the parts of philosophy is simply interchangeable and that one could easily modify it, according to whether one adopts the point of view of logical necessity or that of spiritual progress. In fact, there is a conflict between the logical and the pedagogical aspect. For example, if in teaching one places ethics before physics, one is obliged to present the former without its logical foundations, which are to be found in physics. Ethics here will no longer be a general theory of the ends of human life but will be reduced to a teaching ultimately concerning the social ‘duties’. Bonhöffer97 has insisted on the twofold aspect of ethics in Stoicism. Yet, if, on the contrary, one places physics before ethics, physics will no longer be deployed in its full scope: there will no longer be the supreme teletê of the mysteries of philosophy. Concerning dialectic, it cannot have the same content if it is presented to mere beginners or, on the contrary, if it is exposed to those who have already made progress, as Epictetus wanted. The concrete necessities of the teaching thus conduce finally to important modifications in the content of the parts of philosophy. Unfortunately, we do not know the concrete details of Stoic teaching except from fragments and the doxastic summaries. For this reason, we cannot fully appreciate all the doctrinal variations that this conflict between the logical and pedagogical order was able to produce.
It seems, nevertheless, that the Stoics sought the solution to this conflict in a method of teaching which, while recognizing the necessary distinctions between the parts of philosophy, tried hard to present them all at the same time to the disciple, and always to preserve fidelity to the doctrine as a whole. Indeed, Diogenes Laertius tells us that certain Stoics considered that ‘no part has the first place, that all parts are mixed together, and moreover, that they are necessarily interlaced in their teaching’.98 This is probably what this advice of Chrysippus explains: ‘those who commence by the study of logic must not abstain from the other parts of philosophy, when the occasion presents itself’.99 This method is designed to prevent the danger which not practising logic, ethics or physics during a certain time would represent to the disciple. As has been shown by Ilsetraut Hadot,100 for the Stoics (and for the Epicureans), the method of teaching strives always to be ‘integral’ at each stage of spiritual progress. For the beginner, one proposes sentences or summaries which immediately put them in contact with the whole set of fundamental dogmas and which furnish them with the guiding rules of life.p For the student who is making progress, one offers more detailed developments and more techniques; but by constantly returning to a focus on the fundamental dogmas, this method would never lose sight of what is essential. The disciple thus remains constantly in contact with the three disciplines, even if he studies one amongst them in a more profound manner. There is always a coming and going between concentration and dilation, between simultaneity and the successive order of teachings.
Nevertheless, this method does not totally eliminate the conflict between the logical and pedagogical ordering of the parts of philosophy. In fact, if one is to adapt the teaching to the spiritual capacities of the students, one can be led to two forms of simplification and adaptation which could seem to transform the doctrine itself. When one addresses oneself to beginners, one can even appeal to formulas that are foreign to the school (borrowed from Epicureanism, for example) if one judges them more likely to have an effect on the disciple. This is the principle enunciated by Chrysippus himself in the Therapeutikos.101 This principle of adaptation to the spiritual capacities of the disciple remains alive in all the schools until the end of antiquity and it will often be necessary to take account of it to understand the incoherencies or the apparent contradictions in different works of ancient philosophers.102
In the summaries of Platonic philosophy which we find amongst different authors, from Cicero to Diogenes Laertius, logic is always found in the third place.103 There is probably a tribute there to the Platonic conception of dialectic as the supreme science even though, in these manuals, the content of logic does not correspond to that of Platonic dialectic. Perhaps it is also a matter of the first glimpse of the classification that one can call ‘Neoplatonic’, of which we will now speak.
Effectively, beginning from the first century of the Common Era, one notes the emergence of a classification of the parts of philosophy that is essentially founded on the notion of spiritual progress. The three parts distinguished are, respectively, ethics, physics and ‘epoptic’ (this last word evidently alluding to the supreme initiation in the mysteries of Eleusis). Its first testimony can be found in Plutarch, who, in the Iside,104 affirms that Plato and Aristotle placed, after the physics, a part of philosophy which they call ‘epoptic’ and which has as its object ‘what is first, simple and immaterial’. They think, Plutarch continues, that ‘philosophy finds its end, as in a supreme initiation, thanks to a real touch of pure truth which is found in what is first, simple, and immaterial’. In order to describe philosophical formation, Theon of Smyrna uses the technical vocabulary of the Eleusinian initiation, calling teletê the study of logic, politics (that is to say, ethics) and physics, and calling epopteia the knowledge of true beings.105 Clement of Alexandria also knows this theory. He enumerates the parts of philosophy in this order: ethics, physics (understood as allegorical interpretation), and epoptic, which he explicitly identifies with Platonic dialectic and Aristotelian metaphysics.106 Finally, Origen reveals to us the relationship which exists between this tripartition and the stages of spiritual progress.107 Ethics, according to him, secures a preliminary purification of the soul; physics, by revealing the vanity of the sensible world, invites us to detach ourselves from it; epoptic, finally, opens the purified soul to the contemplation of divine realities. This is why, for him, the three books of Solomon correspond to the three parts of philosophy: the Proverbs assures ethical purification; the Ecclesiastes, which commences with Vanitas Vanitatum, reveals to us the vanity of the physical world; and the Song of Songs introduces us to epoptic. In this succession of texts, Origen specifies that logic, as a science of terms or propositions, is not a separate part of philosophy, but is intertwined with the three other parts, according to certain philosophers – namely, the Aristotelians. After Origin, this tripartition founded on spiritual progress reappears in Porphyry, who edits Plotinus’ Enneads conformably to this schema. The first Ennead corresponds to ethics: that is to say, to a phase of preliminary purification; the second and the third Enneads, to physics; while the fourth, fifth and sixth Enneads correspond to the knowledge of divine realities and thus to epoptic. It is probably through Porphyry that Calcidius knows of the opposition between physics (represented by Plato’s Timaeus) and epoptic (represented by the Parmenides).108 This entire tradition is characterized by certain number of typical traits: the employment of the Eleusinian word epopteia to designate the highest part of philosophy; the identification of Platonic dialectic and Aristotelian theology; the conception of Aristotelian logic as an Organon and, therefore, the integration of Aristotelianism into the cursus of Platonic studies; and finally, and above all, the idea that each part of philosophy does not correspond to a purely intellectual level, nor to the acquisition of a purely abstract knowledge, but that it represents an inner progress which results in a transformation of the individual, his elevation to an ontologically superior sphere. It is evident that, in this schema, ethics and physics assume a meaning that is entirely different to that which they had for example in Aristotelian philosophy. Ethics is now only a phase of preliminary purification, physics is no longer a scientific research, but speculation aiming to make us aware of the fact that sensible reality is but an image. Moreover, as it appears in the Porphyrian ordering of the Enneads, the conflict between the logical and pedagogical order is as acute in Neoplatonism as it is in Stoicism. It is impossible to speak of ethics or physics without considering the whole of the ‘epoptic’; and many of the treatises of the first Ennead, theoretically destined for beginners, are of an extreme difficulty.
This fundamental schema of ethics, physics, epoptic will be the core of the programme of philosophical studies from the end of the first century CE until the end of antiquity. In the Neoplatonic philosophers after Porphyry, this programme contains, on the one hand, the preparatory study of Aristotelian logic and, on the other, the study of mathematics which is the theoretically indispensable propaedeutic to Platonic dialectic, according to the doctrine of the Republic. One thus, finally, has the following scheme: first stage: ethics and logic; second stage: physics and mathematics; third stage: epoptic or theology. We must add that the Neoplatonists interposed into the schema a theory of the stages of spiritual progress which had been outlined by Plotinus and systematized by Porphyry:109 namely, the hierarchy of virtues. The first stage of spiritual progress was, according to this theory, the practice of ‘political’ virtues, that is to say, the accomplishment of the duties of social life according to prudence, justice, courage and temperance. After this indispensable preparation which responded in large measure to the ethical part of philosophy, one could rise to the level of the virtues called ‘cathartic’, which corresponded to a movement of detachment from the body. One passed thus from metriopatheia towards apatheia. This level corresponded to the physical part of philosophy which, as we have seen, had for its goal to detach the soul from the sensible world. Finally, one attains to the theoretical virtues, when the soul is sufficiently detached from the body and can thus turn towards the divine Intellect and contemplate it. This level was evidently analogous to theology or epoptic. All the levels of virtues had their model in the virtues that are specific to the divine Intellect itself and inaccessible to the soul: the paradigmatic virtues.
We have here the broad outlines of a programme of studies which is at the same time an ascensio mentis ad Deum, where one passes by the stages of the purgative way to the illuminative way and the unitive way, respectively.110 But to what does this programme correspond concretely? To form an idea of it, it is necessary to bear in mind that beginning in a certain epoch, which one can probably situate in the first century CE, philosophical teaching to a great extent consisted in a reading and commentary on the works of the founders of the school. This did not exclude discussions with the master on subjects of a general character or, on the contrary, on very personal subjects. Nevertheless, the essential remained the explanation of the texts.111 The Neoplatonic programme of studies corresponds, therefore, to an order of reading Aristotle’s works and Plato’s dialogues. One sees this already outlined by Albinus.112 If we attempt to systematize the testimonies that we possess, it seems that the schema of the parts of philosophy served, above all, to establish the order of reading Aristotle’s work, which represented for the Neoplatonists lesser mysteries: that is to say, a first initiation. It also seems that the scheme of the degrees of the virtues, always intimately tied to that of the parts of philosophy, presided over the establishment of the programme of reading Plato: that is to say, the initiation to the higher mysteries of philosophy.113
One can therefore reconstitute the philosophical cursus in the following way. The first ethical initiation was to be undertaken by the study of simple and striking texts such as Epictetus’ Handbook or the Golden Verses of the Pythagoreans.114 The logical formation of the student was then acquired thanks to the study of Aristotle’s Organon. Then, one studied successively Aristotle’s ethical and political treatises, then his Physics and, finally, his Metaphysics. Marinos tells us that Proclus needed less than two years to be initiated to these lesser mysteries of philosophy.115 After that, one approached the higher mysteries. After the first introduction to the knowledge of self that the Alcibiades represented, the reading of the Gorgias and, eventually, of the Republic led one to accede to the level of the political virtues. One was then elevated to the level of cathartic virtues through reading the Phaedo. Then, at the level of the theoretical virtues, one studied the logical dialogues, namely the Cratylus and the Theaetetus, the physical dialogues, the Sophist and the Politics, and the theological dialogues, the Phaedrus, Symposium and the Philebus. A second cycle would, once again, approach physics with the Timaeus and theology with the Parmenides.116 It is very evident that the dialogues of Plato could not be placed into any such schema, except by means of a forced interpretation.
***
In the Preface to the first edition of his work, Die Welt als Wille und Vorstellung, Schopenhauer writes:
A system of thought must always have the cohesion of an architectural edifice: in other words, a cohesion such that always one of its parts supports another, without being supported by it: the foundational stone must ultimately support all the others, without being itself supported by any other, and the summit must be supported by the rest, without supporting anything else. On the contrary, a unique idea, although all encompassing, must preserve the most perfect unity. Even if one must divide it into parts in order to communicate it, it remains nevertheless that the cohesion of these parts must be organic, that is to say, that each part contributes to maintain the whole and should be maintained in its place by the whole; none being the first, none being the last: the total idea gains in clarity by the exposition of each part, but also the smallest part cannot be completely understood without previously understanding the whole.q
Schopenhauer opposes here ‘systems of thought’ and a ‘unique idea’, but one could say as well that he opposes two types of system: on the one hand, the system of an architectural type, or to quote Leisegang’s phrase in his ‘Denkformen’,117 the ‘pyramid of concepts’; and, on the other hand, the system of the organic type. One recognizes these two types of system in the first two types of classification which we have distinguished. The Platonic-Aristotelian classification supposes a coherence of the architectural type: the inferior degrees of the ontological hierarchy cannot exist without the superior degrees, but these again can exist without the inferior degrees.118 In contrast, the Stoic classification supposes a coherence of the organic type: the parts of philosophy, or rather the parts of philosophical discourse, form a system because they are an ensemble of organized thoughts. But they reveal the three aspects of one and the same idea, which is that of the Logos and they imitate, in their reciprocal implication, the dynamic unity of this unique Logos.119
Each of these two types of classification describe a different ideal type of wisdom. For the first, wisdom is a universal knowledge which embraces the architecture of the system of sciences, their methods and the diversity of their objects; for the second, wisdom is a concentrated attention on the presence of the Logos within all things. The third type of classification does not describe an ideal type, but an itinerary or concrete method which leads to wisdom. It can authorize different classifications: different orders of the parts of philosophy in the Stoic system, or the tripartition ethics–physics–epoptic in the Neoplatonic system. However, these different classifications remain of the same type because, for them, the parts of philosophy are conceived as stages of an inner path which one must traverse. They are the phases of an evolution and of a transformation which one must realize. This third type is properly philo-sophical, in the etymological sense of the term, since it corresponds to an effort, a search, an exercise120 which leads to wisdom. It is this properly philosophical aspect which prevents it from being totally systematic and which explains the hesitations, the incoherencies and the internal contradictions which characterize it.
1Aristotle, Metaphysics, 1004 a8.
2Aristotle, Metaphysics, 1004 a8. We follow Tredennick with the translation of “primary” and “secondary” mathematics here.
3Aristotle, Metaphysics, 982 a5, 982 b10.
4Plato, Republic, 510b.
5Plato, Republic, 537c.
6Plato, Statesman, 258e.
7Aristotle, Metaphysics, 1025b3 ff.
8Aristotle, Metaphysics, 1025b20: nature in itself has the principle of its own movement, whereas the principle of technical productions or of actions is found in the human agent.
9Aristotle, Metaphysics, 1026a10 ff. on the critical problems posed by this passage, cf. Philip Merlan, From Platonism to Neoplatonism (The Hague: Nijhoff, 1960), 62 ff.
10On these problems, cf. Heinz Happ, Hyle: Studien zum Aristotelischen Materie-Begriff (Berlin: Walter de Gruyter, 1971), 565–569.
11Aristotle, Metaphysics, 982a26.
12Aristotle, Metaphysics, 1026a23.
13Aristotle, Metaphysics, 1025b9 & 19 [Translator’s note: ‘kind’ here translates the French genre].
14Cf. Hans-Joachim Krämer, Der Ursprung der Geistmetaphysik (Amsterdam: Grüner, 1967), 146, n. 66.
15Plato, Statesman, 258e.
16Plato, Republic, 510–511; cf. Happ, Hyle, 267. On the association between the soul and the mathêmatika, cf. Merlan, From Platonism to Neoplatonism, 11 & 82.
17Aristotle, Metaphysics, 1073b5 & Physics, 193b22 ff.
18Cf. Happ, Hyle, 477 & 36, n. 149.
19For example, the theory of science or of definition.
20Cf. Happ, Hyle, 311.
21Aristotle, Metaphysics, 1026a23.
22Aristotle, Metaphysics, 1025b19 [Translator’s note: ‘Genus’ here translates the French genre].
23Aristotle, Metaphysics, 992a33.
24Aristotle, Metaphysics, 1026a14. One will remark that the text (1026a8 and a14) refuses to define in a general manner the ontological status of mathematical objects but aims only at certain branches of mathematics.
25Aristotle, Metaphysics, 1025b10–19 & 1026a30–33.
26Aristotle, Topics, 105b30; cf. John D. G. Evans, Aristotle’s Concepts of Dialectic (Cambridge: Cambridge University Press, 1977), 7–55
27Plato, Republic, 510b.
28Plato, Timaeus, 29d.
29Aristotle, Nicomachean Ethics, 1094b11.
30Aristotle, Physics, 193b22; Parts of Animals, 639a1.
31Aristotle, Metaphysics, 1025b28.
32Aristotle, Parts of Animals, 642b5; Generation of Animals, 748a8; Nicomachean Ethics, 1107a26.
33Aristotle, Physics, 204b4 & b10; Nicomachean Ethics, 1147a24 ff.; Generation and Corruption, 316a11; De Cielo, 280a32; Posterior Analytics, 84a8; cf. Mario Mignucci, L’Argomentazione Dimonstativa in Aristotele, tome 1 (Padova: Editrice Antenore, 1975), 484 ff.
34Cf. infra n. 66–69.
35Sextus Empiricus, Adversus Mathematicus, VII, 16.
36Diogenes Laertius, Lives of Eminent Philosophers, IV, 11 ff.
37Cf. Krämer, Der Ursprung der Geistmetaphysik, 146, n. 66.
38Rudolf Hirzel, ‘De Logica Stoicorum’, Satura Philologica, Festschrift Hermann Souppe (Berlin, 1879), 64 ff.
39Aristotle, Topics, 109b19.
40Aristotle, Topics, 105b19: hôs tupôi.
41Cf. note 33 above.
42Aristotle, Topics, 105b19: hôs tupôi.
43Alexandri Aphrodisiensis in Aristotelis Topicorum Libros octo commentaria, ed. Maximilianus Wallies (Berlin; G. Reimer, 1891), 74, 26 & 94, 7.
44Aristotle, Topics, 104b1.
45Cf. Hans-Joachim Krämer, Platonismus und Hellenistische Philosophie, 114, n. 35.
46Cf. Ěmile Bréhier, Chrysippe et l’ancien stoïcisme (Paris: Presses Universitaires de France, 1951), 59 ff., esp. 65.
47Hans von Arnim, Stoicorum Veterum Fragmenta [SVF], tome II, 97–106.
48See Pierre Hadot, ‘Marc-Aurêle: Était-il opiomane?’, in Ětudes de Philosophie Ancienne, 97–106.
49Max Pohlenz, Die Stoa (Göttingen: Vandenhoeck & Ruprecht, 1959), 34; Krämer, Platonismus, 114, n. 35; Andreas Graeser, Zenon von Kition (Berlin: de Gruyter, 1975), 21 ff.
50Ěmile Bréhier, Histoire de la philosophie, tome I, 2 (Paris: Presses Universitaires de France, 1961), 303.
51Ibid.
52Cf. Victor Goldschmidt, Le Système stoïcien et l’idée de temps (Paris: Vrin, 1977), 66; Hans-Jürgen Horn, ‘Antakolouthie der Tugenden und Einheit Gottes’, Jarbuch fur Antike und Christentum, tome 13 (1970), 3–38; Pierre Hadot, Porphyre et Plotin, tome 1 (Paris: Études Augustiniennes, 1968), 239 ff.
53SVF tome II, §952: the determinism of destiny and the principle of contradiction are intimately bound up together.
54SVF tome II, §§130–131; Cicero, De Finibus, III, 21, 72; Diogenes Laertius, Lives of the Eminent Philosophers, VII, 46.
55SVF tome III, §456 ff.
56Diogenes Laertius, Lives of the Eminent Philosophers, VII, 83.
57SVF tome III, §68; Cicero, De Finibus, III, 22, 73.
58Cicero, De Natura Deorum, II, 14, 37.
59Cf. Cicero, De Finibus, III, 21, 72.
60Philo of Alexandria, De Ebrietate, §§ 90–92.
61Cf. Marcus Aurelius, Meditations, VIII, 27.
62Marcus Aurelius, Meditations, VIII, 13.
63Epictetus, Discourses, I, 4, 11; III, 2, 1; II, 8, 29; II, 17, 15 & 31; IV, 4, 16; IV, 10, 13.
64Cf. Pierre Hadot, ‘Une Clé des Pensées de Marc Aurèle: les trois topoi philosophiques selon Marc Aurèle’, Les Études Philosophiques (1978): 65–83.
65Marcus Aurelius, Meditations, VII, 54.
66Diogenes Laertius, Lives of the Eminent Philosophers, III, 56; Apuleius, De Platone, I, 3, 186; Atticus, in Eusebius, Praeparatio Evangelica, XI, 2, 1; Augustine, Against the Academics, III, 17, 37; De Civitate Dei, VIII, 4 ff.
67Diogenes Laertius, Lives of the Eminent Philosophers, III, 76; Apuleius, I, 3, 190.
68Atticus, in Eusebius, Praeparatio Evangelica, XI, 2, 1.
69Diogenes Laertius, Lives of the Eminent Philosophers, III, 79.
70Clement of Alexandria, Stromata, I, 28, 176, 3; Plotinus, Enneads, I, 3, 5, 12; Augustine, Against the Academics, III, 17, 37; De Civitate Dei, VIII, 7.
71Clement of Alexandria, Stromata, IV, 25, 162, 5.
72Augustine, De Civitate Dei, VIII, 4; Epistles, 118, 3, 20.
73Augustine, De Civitate Dei, XI, 25 in liaison with the triad natura, doctrina, usus; cf. Pierre Hadot, ‘Etre, Vie et Pensée chez Plotin et avant Plotin’, Entretiens sur L’Antiquité Classique, tome V (Genève: Fondation Hadt, 1960), 123–125; Olivier de Roy, L’Intelligence de la Foi en la Trinite selon Saint Augustine (Paris: Ětudes Augustiniennes, 1966), 447.
74For example, Simplicii in Aristotelis Categorias Commentarium, ed. Karl Kalbfleisch (Berlin: Reimer, 1907), 4, 23 & 5, 3; Eliae in Porphyrii Isagogen et Aristotelis Categorias Commentaria, Commentaria in Aristotelem Graeca, ed. Adolfus Busse (Berlin: Georgii Reimeri, 1900), 115–119, which enumerate diverse Aristotelian claims concerning the order of teaching, showing that, even in Aristotelianism, the problem was not very (si) simple to resolve.
75Aristotle, Posterior Analytics, 72a; cf. Mignucci, L’Argumenazione dimostrative di Aristotle, tome I, 30.
76Victor Goldschmidt, ‘Temps historique et temps logique dans l’interprétation des systèmes philosophiques”, Acts du X1eme Congrès International de Philosophie, XII (Bruxelles 1953), 7–13.
77I draw this example from Victor Goldschmidt. ‘Temps historique et temps logique’, 11. He cites Descartes, ‘Réponse aux Seconds Objections (contra les […] Meditations)’, in Oeuvres de Descartes, tome IX, edited by Charles Adam et Paul Tannery (Paris: Léopold Cerf, 1904), 103–104.
78Cf. note above on Aristotle, Posterior Analytics. One finds in the work of Aristotle certain pedagogical notations: for example, in Metaphysics 1005b4: one cannot understand certain questions of first philosophy without having previously studied the analytic: or better, in Nicomachean Ethics, 1095a1 ff., young men cannot study ethics and politics if they have not already begun improving their lives.
79Cf. notes 18–21 above.
80Plato, Republic, 521c ff.
81Cf. Harold Chemiss Die Altere Akademie (Heidelberg: C. Winter, 1966), 82–83.
82Plato, Republic, 539b. We note thus in the Phaedrus 271b the idea of a rhetoric which adapts the species of discourse to the different species of souls.
83Plato, Republic, 514a ff.
84Plato, Symposium, 210a.
85Diogenes Laertius, Lives of the Eminent Philosophers, VII, 39: “tripartite, they say, is the discourse of philosophy (ton kata philosophian logon)”, and ibid., VII, 41: Zeno of Tarsis, as opposed to the other Stoics, thought that the parts in question were the parts of philosophy, and are not parts of philosophical discourse. [Translator’s note: see Chapter 3 above.]
86Cf. Goldschmidt, Le Système stoïcien, 61–62. On the subdivisions of ethics, cf. André Mehat, Essai sur les Stromates de Clement of Alexandria (Paris: Éditions du Seuil, 1966), 77 ff.
87Diogenes Laertius, Lives of the Eminent Philosophers, VII, 49, VII, 84, VII, 132.
88Cicero, De Finibus, III, 4, 14 ff.
89Cf. Goldschmidt, Le Système stoïcien, 62.
90Diogenes Laertius, Lives of the Eminent Philosophers, VII, 40–41; Sextus Empiricus, Adversus Mathematicus, VII, 16–19.
91Epitetus, Discourses, III, 2,5; IV, 8, 4.
92Sextus Empiricus, Adversus Mathematicus, VII, 17–19; Diogenes Laertius, Lives of the Eminent Philosophers, VII, 40.
93Adolf Friedrich Bonhöffer, Epiktet und die Stoa (Stuttgart: Verlag von Ferdinand Enke, 1890 [reprint 1968]), 13 ff.
94Cf. Ilsetraut Hadot, Seneca und die Greichisch-Romische Tradition der Seelenleitung (Berlin: Walter de Gruyter, 1969), 115.
95SVF, tome II, §42.
96Sextus Empiricus, Adversus Mathematicus, VII, 23.
97Bonhöffer, Epiktet und die Stoa, 19.
98Diogenes Laertius, Lives of the Eminent Philosophers, VII, 40 (I read pokekrosthai with the manuscripts).
99SVF tome II, 20, 10 (= in Plutarch, De Stoicorum Repugnantiis, 1035a).
100Ilsetraut Hadot, Seneca, 52–56; and her ‘Épicure et l’enseignement philosophique hellénistique et romain”, Actes de VIIIe Congrès de L’Association Guillaume Budé (Paris: Vrin, 1969), 347–355.
101SVF tome III, §474.
102Ilsetraut Hadot, Seneca, 21; and by the same author, Le Problème du Néoplatonisme Alexandrin (Paris: Études Augustiniennes, 1976), 190.
103Cf. Émile Bréhier, Études de philosophie antique (Paris: Presses Universitaires de France, 1955), 215–217. I do not think that Bréhier had good reasons to attribute this third place “to a disdain for logical technique.” On the contrary, this third place is probably a place of honour.
104Plutarch, de Iside, 382d.
105Theon of Smyrna, Expositio Rerum Mathematicarum ad Legendum Platonem Utilium, ed. Eduard Hiller (Leipzig: Tuebner, 1878 [1966]), 14.
106Clement of Alexandria, Stromata, I, 28, 176, 1–3.
107Origen, In Canticum Canticorum, in Origenes Werke, Vol. 8 (GCS 33), ed. W. A. Baehrens (Berlin: C. Hinrichs, 1925), prol., 75, 6. Cf. Évagre le Pontique, Traité pratique ou Le moine, §1 Gaullamont.
108Calcidius, On Plato’s Timaeus, 170, 7 and 277, 5 Waszink.
109Porphyry. Sententia, ed. Erich Lamberz (Leipzig: Teubner, 1975), 32. Cf. Wilhelm Theiler, Gnomon 5 (1929): 307–317; also Ilsetraut Hadot, Le Problème du Néoplatonisme Alexandrin, 152 ff.
110Cf. Henri van Leshout, La Théorie Plotinienne de la Vertu: Essai sur la genèse d’un article de la Somme theologique de Saint Thomas (Freiburg, Switzerland: Broché, 1926).
111Cf. Epictetus, Discourses, I, 4, 7; I, 17, 13; II, 16, 34; II, 21, 10–11; III, 21, 7; Cf. Georg H. I. Bruns, De Schola Epicteti (Kiel, 1897), 14; Aulus-Gellius, Attic Nights, I, 9, 9; Porphyry, Life of Plotinus, §14; Marinus, Vita Procli, ed. Jean François Boissonade (London, Rome, Paris: Lipsiae: Weigel, 1814), 157, 7 ff. See also Eliae in Porphyrii Isagogen et Aristotelis Categorias Commentaria, 115–119, on the order of study of the parts of philosophy.
112Albinus, Eisagôgê, in Karl Friedrich Hermann, Platonis Opera, VI (Leipzig: Novi Eboraci Apud Harperos Phatres, 1853), 147–151; cf. René Le Corre, ‘The Prologue of Albinus’, Revue Philosophique 81 (1965): 28–38. Albinus proposes an order of reading founded on the dispositions of the disciple and on his spiritual progress.
113Compare Marinus, Vita Procli, 157, 41; Simplicii in Aristotelis Physicorum Libros Quattuor Priores Commentaria, ed. Hermann Diels (Berlin: G. Remeiri, 1882), 5, 29 and Simplicii in Aristotelis Categorias Commentarium, 5, 3 ff.; Anonymous Prolegomena to Platonic Philosophy, 10, 26, 49, I.
114Cf. Ilsetraut Hadot, Le Problème du néoplatonisme Alexandrin, 160 ff.
115Marinus, Vita Procli, 157, 41.
116Cf. André-Jean Festugière, “L’Ordre de lecture des dialogues de Platon aux Ve/VIe siècles”, Mus. Helv. 26 (1969): 282–296.
117Hans Leisegang, Denkformen, 2e ed. (Berlin: Walter de Gruyter, 1951), 208, which cites the text of Schopenhauer of which we are speaking.
118The metaphor of Schopenhauer is in effect reversed: it is this time the superior which makes the inferior exist. But the essential component of the metaphor is safeguarded: the relations between the parts of the system are not reciprocal.
119Goldschmidt, Le système stoïcien, 64 makes allusion, concerning Stoicism, to the text of Schopenhauer’s that we have cited.
120Cf. Pierre Hadot, ‘Exercises Spirituels’, Annuaire de la Ve section de l’École des Hautes Études 84 (1977): 25–70. [Translator’s note: i.e. ‘Spiritual exercises’, in Philosophy as a Way of Life, trans. by Michael Chase (London: Wiley-Blackwell, 1995), 79–125]. On philosophy as quest of wisdom, cf. Plato, Symposium, 203 d, where it coincides with the figure of Eros, that of Socrates and of philosophy.
a [Translator’s note] Hadot does not provide the exact reference. Cf. Immanuel Kant, The Groundwork for the Metaphysics of Morals, ed. and trans. by Allen Wood (New Haven, CT, and London: Yale University Press, 2002), 3.
b [Translator’s note] Here, we add explanatory subtitles, based on Hadot’s division of the article.
c [Translator’s note] Here, the adjective is théorétique. See Chapter 3 above on théorétique versus théorique. Here, Hadot uses théorétique not in the specialized sense described there, but as an adjective qualifying a kind of inquiry, as against designating a mode of life.
d [Translator’s note] From here on, we shall translate ‘productive sciences’, given the specific English association of poetry with written verse, whereas the Greek poiesis has a much wider extension, including productive crafts.
e [Translator’s note] The opposition here, in the original language, is between mobile and immobile objects. Aristotle divides between unchangeable or unmovable objects and objects capable of motion: either of being moved by other objects or initiating movements themselves.
f [Translator’s note] Hadot’s expression is fable vraissemblable.
g [Translator’s note] The French noun is dogmatisme, literally ‘dogmatism’.
h [Translator’s note] French: tendances.
i [Translator’s note] French: complaisance.
j [Translator’s note] See Chapter 11 below.
k [Translator’s note] Hadot’s term is achevé.
l [Translator’s note] Hadot’s expression is donnée complexe.
m [Translator’s note] Literally: ordre exterieur de présentation.
n [Translator’s note] Hadot’s term is la tendance; at stake is the Stoic theory of oikeiosis, the ‘adaptation’ of the impulses of natural creatures to their environs.
o [Translator’s note] Hadot’s expression is: la partie la plus intérieure et la plus precieuse.
p [Translator’s note] French: grande regles de vie, an expression whose literal translation in English would be unusual: ‘great/grand rules of life’.
q [Translator’s note] Hadot does not give the citation. See Arthur Schopenhauer, The World as Will and Representation, Volume 1. Translated and edited by Judith Norman, Alistair Welchman and Christopher Janaway (Cambridge: Cambridge University Press, 2010), “Preface to the first edition”, 5. Compare the translation by R. B. Haldane and J. Kemp at: https://en.wikisource.org/wiki/The_World_as_Will_and_Representation/Preface_to_the_First_Edition (accessed 6 June 2019).