The following study can only be a very general sketch:1 one would need a whole work to treat this topic in all its details. Our exposition will especially focus on the relations between philosophy and dialectic; rhetoric will only appear to the extent to which it is intimately related to dialectic. On the other hand, we will not analyse all philosophical schools: to be more specific, we will leave the Epicureans aside.2 After having presented the relations between philosophy, dialectic and rhetoric in Aristotle, which will allow us to evoke the Platonic background [of Aristotle’s position], we will then study the evolution of these relations in Hellenistic and Neoplatonist philosophy. We will thus be led to a set of methodological conclusions.
It is in the work of Aristotle that we find the most elaborated ancient theory on the relationship between the three disciplines.3
First of all, what are the common points between rhetoric and dialectic? They both suppose an initial situation marked by conflict: that is, a situation where two contradictory responses can be given to a theoretical, juridical or political question (or problem).4 However, rhetoric and dialectic are not interested in the response as such. They are only concerned with the means which will allow an adversary, a judge or the people to accept one of the two. What counts, therefore, is not so much the theme of the discussion, but rather the interlocutora himself, who must be persuaded.5 In order to conduct the listener to a given conviction, one must start from those things of which the listener is already convinced; that is to say, one must start either from his own opinions or, more generally, from opinions that are universally accepted, common and natural notions, admitted by everyone.6 This point, for Aristotle, is what radically distinguishes rhetoric and dialectic from the sciences such as mathematics, for example.7 The latter have, indeed, their own principles and techniques which are only understood by specialists, and they are applied to one specific domain of reality. Rhetoric and dialectic, on the contrary, are ‘common’ and universal. They do not have any specific and determinate domain of reality on which they focus; rather, they can be applied to any subject of discussion, and they start from principles which are accessible to all. They do not bring knowledge concerning any specific issue. They seek only to persuade the interlocutor, with the help of what he already knows. This is why they are both able to prove the positions ‘pro’ and ‘contra’:b8 this is what Cicero will call argument in utramque partem. This does not deprive them in any way of logical rigour. From the moment that one admits a certain point of departure, the necessary chain of syllogisms will inexorably lead to the conclusion. The syllogistic reasoning is the same in science and dialectic.9 It is solely the principles which differ between them: in each science, its own specific principles; in rhetoric and dialectic, common notions. Furthermore, rhetoric and dialectic, as we have already said, are not concerned with the conclusion in itself, but rather with the means that lead to the conclusion.
Despite this fundamental kinship, rhetoric and dialectic are different from one another. Let us first speak about the characteristics that are specific to dialectic. Dialectic is the art of discussing; that is to say, it formulates the rules of the dialectical joust,10 in which an attackerc argues against the thesis of an adversary in order to oblige the latter, through skilful questions, to necessarily admit the thesis which contradicts his own. ‘To interrogate’ (erôtan), this is the task of the attacker; to construct a syllogism, starting from the premises (questions posed to the adversary) which will lead (if he concedes the truth of these premises) to a conclusion which involves the contradictory proposition to the adversary’s thesis. The attacker himself does not have a thesis. However, he is capable of arguing in such a way that he who defends a thesis will be inexorably constrained to contradict his own thesis. One recognizes here the Socratic situation. As Aristotle says: ‘Socrates asked questions, he did not provide answers, since he confessed to know nothing.’11 Not only does dialectic teach one how to ask questions. It also teaches how to respond; that is to say, it teaches one how to defend a thesis against the trapsd of the questioner.e12 The respondent defends a thesis, but only for the sake of the exercise: he must be able to defend the thesis and the thesis that contradicts it. He thus only appears as if he knew.
This game of questions and answers and its tortuous itineraries, which often puts off the readers of Plato’s dialogues, cannot be found in rhetoric. The latter uses continuous discourse;13 that is, an exposition which is not interrupted and structured in questions and answers. This was the method of the sophists. The listener who must be persuaded does not participate in the argumentation, as was the case in dialectic. In dialectic, the questioner and the respondent specify, at each step of the discussion, the points on which they agree. And this is exactly the danger for the respondent, who can let himself be led to imprudently agree to a premise (that is, to say ‘yes’ to a question) in such a way that the logical consequences of the premise thus conceded will constrain him to admit as true the contradiction to his own initial thesis. In rhetoric, the discourse is developed with no obstacles, starting from the positions that the listener is supposed to accept.
Moreover, dialectic is more universal than rhetoric. The former treats all possible subjects, and does so in the most universal way, without the intervention of concrete data.f On the contrary, rhetoric takes as subjects, above all, concrete problems14 (Did this man commit murder? Should we go to war?). What is at stake for rhetoric is the kind of problems that the ancients called ‘civil’ or ‘political’15 which are, by definition, contingent. One should not be surprised to find the phrase ‘civil questions’ in the Latin rhetoricians, employed to designate the problems of rhetoric.16 There will always be a close link between rhetoric and ethical and political problems. This opposition between the generality of dialectic and the particularity of rhetoric corresponds to the opposition between thesis and hypothesis. As Boethius will say:
Dialectic is only concerned with the ‘thesis’. The thesis is a problem (questio) without any concrete or particular circumstances (circumstantiae). Rhetoric concerns itself with discussing hypotheses, which is to say, discussing problems which presuppose a multiplicity of concrete circumstances (Who? What? Where? When? Why? How? By what means?)17
For Aristotle, there is a fundamental opposition between philosophy and these two techniques of rhetoric and dialectic. In effect, this position reflects the novelty of his conception of philosophy, compared to that of Plato. One knows that in the Phaedrus, after having criticized the rhetoric of the rhetors, Plato proposed a philosophical rhetoric.18 This would not be a rhetoric of mere appearance, but one which would be founded on the knowledge of truth19 and on the differentiation of possible relations between different kinds of souls and the kinds of discourses capable of affecting each of these types of souls.20 In Plato’s mind, this philosophical rhetoric presupposes dialectic, which for him means philosophy, and which is what will allow rhetoric to be founded on the knowledge of truth. Dialectic is defined in the Phaedrus as a twofold movement of thought. On the one hand, it reduces the multiplicity of notions to the unity of a Form, that is, of an Idea in the Platonic sense. On the other hand, dialectic re-descends from this Form, distinguishing and organizing the subordinated forms which it implies.21 In Plato, this method is closely linked to the practice of dialogue. Indeed, as one notes in Plato’s dialogues, it is in the game of questions and answers, through the slow progression of the interlocutors’ reflection on their own discourse, that the ideal Form gradually emerges, together with the division of particular forms which are enveloped in this first Form.g The logical architecture of reality is discovered by discourse and by the discourse about discourse. As Léon Robin excellently writes: ‘The reflective and free adhesion to a thesis subject to examination is essential to dialectic, i.e., the dialogical method of research in common by questions and answers conducted according to an order.’22
This is precisely what Aristotle refuses. For him, Platonic dialectic is a purely formal method which is incapable of leading us to a truly scientific form of knowledge. Indeed, Platonic dialectic starts from common notions, whereas for Aristotle each science must start from principles specific to the domain of reality it studies. In addition to that, Platonic dialectic moves within the homogenous and universal domain of discourse, whereas each science has as its object a determined ousia, which it must attain through specific methods, ultimately relying on observation from the senses. Platonic dialectic cannot exit the domain of discourse. It only attains abstract notions rather than ousias, which, for Plato, would be the Ideas subsisting in themselves.23 The refusal of the theory of Ideas and the refusal to consider dialectic as a philosophical science go hand in hand in Aristotle. Generally, Aristotle employs the term logikôs either to designate this purely formal way of arguing which relies on the ordinary meaning of the words, or to designate the analysis of what is logically implied in a definition.24
In Aristotelian philosophy, there should then be a radical separation between rhetoric and dialectic, on the one hand, and the philosophical sciences, on the other.25 Since the status of science is denied to dialectic in the Platonic sense, and since dialectic is preserved only as a technique of persuasion, indifferent to the pursuit of the truth, there should be no room for it in philosophical science. In fact, Aristotle often distinguishes with great clarity dialectical from demonstrative premises: the former have an interrogative form, to which one could respond with a ‘yes’ or ‘no’. One will ask the interlocutor: ‘Is pleasure a good? Yes or no?’ The rest of the discussion will depend on the answer that the interlocutor gives. Demonstrative premises, on the other hand, are not interrogative; they are already choices between the contradictory alternatives. For example: ‘pleasure is a good’. One can, then, extract the consequences from this affirmation. Therefore, they allow us to build up a science. Since the consequences are themselves a result of the principles proper to that science, they lead us to determinate conclusions; whereas if dialectical premises are taken as [the] starting point by the questioner, they can only lead to a refutation of the respondent’s position.26 Aristotle thus distinguishes didactic from dialectical arguments:
we call didactic those arguments which reach their conclusions starting from principles that are specific to each discipline, instead of beginning from the opinions of the person who responds. We call dialectical those arguments that reach a conclusion which contradicts the given thesis starting from premises that are probable.27
With this fundamental opposition posed, Aristotle still recognizes that dialectic has a role to play in philosophy.28 First of all, it is a sort of intellectual gymnastics, which one should practice in the scholarlyh exercise of the dialectical joust. Aristotle formulated the rules of this kind of joust in the eighth book of the Topics. And the practice of dialectic will remain prestigious for a long time in the philosophical schools, as we will see in what follows.
Dialectic (and eventually rhetoric) represents a means to adapt oneself to one’s interlocutors. Moreover, dialectic and rhetoric can play a protreptic role in philosophy: inviting someone to a philosophical conversion, starting from common notions that every man can understand and accept.29 They can also have a polemical mission. We can, for instance, surmise that the Aristotelian dialogues, today lost, which according to Cicero presented successive arguments for and against a thesis, were precisely aimed at refuting a doctrine after having clearly exposed it.30 One can often find this procedure in Cicero. One can also note that, in the Metaphysics, Aristotle – who reproached Platonists for developing non-scientific and merely formal arguments – claims that he can refute Platonic doctrines by using arguments even more formal than those of his rivals, and therefore beating them on their own terrain, thanks to dialectic.31 In general, the knowledge of common notions, accepted by everyone, or the knowledge of the theories of one’s adversary, allow one to take as one’s starting point in the discussion the adversary’s prejudices and presuppositions, in order to lead him to detach himself from them. The technique of refutation which characterizes dialectic will make it possible to persuade one’s interlocutors to renounce their own untenable positions.
Finally, within philosophical science itself, when a particular science such as physics is led to speak about absolutely universal principles, it will be constrained to call upon dialectic, to the extent to which the ‘axioms’ that it supposes are founded on universally accepted opinions. As Aristotle said:
It is impossible to speak of first notions in each science while at the same time relying on the specific principles of this science, because principles are precisely that which is first in relation to all the rest. It is thus necessary, if one wants to examine them, to make recourse to what exists in terms of generally accepted ideas concerning each of these notions. This task is proper to dialectic.32
Dialectic, then, comes onto the scene when a science reaches its original limit and cannot reflect on its own principles. This is why first philosophy cannot explain the use of the identity principle, which is the fundamental axiom of all science, except by resorting to a dialectical method.33 Another boundary of science appears when one finds affirmations of existence. For example, physics poses the problem of the existence of time.34 Now, science has as its object the essence and not the existence of things: one does not know, properly speaking, an existence; rather, one can only be convinced of it (or not). On the other hand, as Wolfgang Wieland35 correctly remarked, rhetoric, and especially judicial rhetoric, is prepared to face this sort of issue, since it must often convince people of the existence of a certain fact. When treating of problems of existence, Aristotle uses, then, a mode of rhetorical argumentation, which he calls ‘exoteric’. Simplicius, when commenting on this phrase, explains: ‘That which is exoteric is not specialised (koinon), and that which reaches a conclusion starting from reasons accepted by everyone (endoxon).’36
In principle, in his scientific works, Aristotle should have used only the apodictic methods he had proposed in his Analytics. However, like many theorists of methodology, he is often disloyal to his own principles. His procedures of argumentation sometimes happen to be dialectical, and even rhetorical. Like every science, philosophical science should go from the premises to the conclusion; that is to say, it should start from general principles, specific to the science in question, and advance towards consequences which would be more and more distanced from these premises. On the contrary, rhetoric and dialectic correspond to an inverse movement of thought: they lead us from the conclusion to the premises. Indeed, in rhetorical or dialectical argumentation, one knows the conclusion in advance. In judicial rhetoric, the lawyer knows from the start whether he will answer ‘yes’ or ‘no’ to the question: ‘did this man commit murder?’ In dialectic, the attacker and the respondent have defined positions from the start. To the question: ‘is pleasure a good?’, one answers ‘yes’, the other answers ‘no’. What one must find are not the conclusions, but rather the premises; that is, the propositions from which one will be able to ground a conclusion which is already known. Aristotle’s Topics and his Rhetoric precisely provide the ‘places’ (topoi), that is, the typical schemasi of argumentation that will allow one to find, concerning any given issue, the premises that will necessarily lead to some conclusion. The dialectical and the rhetorical method are therefore hypothetical; they lead us back to the necessary conditions which make the conclusion possible. Now, as Wieland has shown, it is precisely the rhetorico-dialectical method that one finds applied in many points of Aristotelian philosophy, rather than the deductive method defined in the Analytics.37 That which Aristotelian science seeks to discover are the principles on the basis of which syllogistic deduction can take place. These principles often constitute general schemas of argumentation that are, indeed, analogous to the topoi of dialectic.
Must we go further and say, with Pierre Aubenque, that the first philosophy that Aristotle wanted to substitute for the Platonic dialectic remained at a stage of a purely aporetic dialectic?38 Given the limited scope of this article, we can only raise this problem which shows the importance of dialectic in Aristotle’s philosophy.
To describe the relations between philosophy, dialectic and rhetoric in post-Aristotelian ancient philosophy, it would be necessary to clearly distinguish three main domains of application of dialectic and eventually of rhetoric.39 Dialectic can be a pedagogical or school exercise, and Aristotle formulated the rules of this exercise in the eighth book of the Topics. Dialectic and rhetoric can also be a method for the teaching of philosophy. And finally, both dialectic and rhetoric can be a constitutive part of philosophy, as subdivisions of logic.
In what concerns pedagogical dialectic, we have already seen that Aristotle formulated in the eighth book of the Topics the general rules of dialectical discussion, opposing for the sake of intellectual gymnastics a questioner and a respondent. It is probably to the pedagogical exercise that Polemon,40 Xenocrates’ successor at the Academy in the third century BCE, referred when he said that it was better to exercise oneself in the difficulties of life than in dialectical questions, since he criticized those whom one admires for their argumentative performances (erôtêseis), but who contradict themselves in their inner dispositions.41
Moreover, it is to the dialectical joust that we can link the habit of using the word ‘interrogate’ (erôtan in Greek, interrogare in Latin) to designate all forms of syllogistic argumentation until the end of antiquity. This is because, in dialectical exercises, it is he who interrogatesk who puts forward syllogisms.42 Approximately four centuries after Polemon, we find a parallel between the dialectical exercise and the exercise of virtue in Epictetus:
In the same way that we exercise ourselves in responding to sophistical interrogations, we have equally to exercise ourselves daily to confront representations, for they also pose interrogations to us.43
Aulus Gellius, who writes in the first half of the second century CE and who had been the disciple of the Platonist Taurus, mentions the pedagogical exercises in dialectic when he evokes one of the fundamental rules of these exercises: that which established that the respondent could only reply by saying ‘yes’ or ‘no’. But he remarks that one had the right to refuse to answer sophistical questions such as: ‘have you stopped committing adultery, yes or no?’44 In the fifth century, we find Proclus, in his Commentary on the Alcibiades, perfectly informed about the fundamental rules of the dialectical exercises.45
Throughout the Middle Ages and into the twentieth century,46 the dialectical exercise of argumentation will, in fact, continue to be an important element in the formation of future philosophers.
We have less details concerning rhetorical exercises. However, we know from Cicero that both Academics and Peripatetics attributed great importance to the training of speech.47 For this reason, they practiced exercises in which arguments for and against a thesis were successively developed. Generally speaking, it seems that philosophical education included as one of its constitutive parts the composition of ‘theses’, in the rhetorical sense of the word: that is to say, the composition of continuous speeches (in contrast with dialectic) and expositions on general topics, such as: ‘should the sage get married?’ Such exercises seem to have formed part of the exercises proper to philosophical formation in the Platonic, Aristotelian and Stoic schools.48
The problem we now pose is effectively different [from the preceding]. Here, we are not concerned with knowing if one practiced dialectical or rhetorical exercises during philosophy courses, but rather if the master used a method which would involve dialectical and rhetorical traits in order to communicate the content of his philosophy to his students. From this perspective, one can say that the period between the third century BCE and the fourth century CE corresponds to a triumph of dialectic and rhetoric in philosophical teaching. Evidently, the use of the two disciplines acquires multiple forms and meanings; the phenomenon is, nevertheless, a general fact.
On this topic, we possess a very interesting text from Cicero specifically on the teaching methods of the Academy. There, we can clearly see the fundamental importance of the question–answer schema, but also the more dialectical or rhetorical orientations of different pedagogical methods.l It will be very useful to analyse this text with precision. At the very moment when he begins speaking in the dialogue of De Finibus in order to refute the Epicurean position,49 Cicero warns his interlocutors: ‘First of all, I ask you to understand that I will not develop a schola as a philosopher,m which is something that I have never appreciated in a philosopher.’ By schola, Cicero understands a continuous discourse, and thus a rhetorical type of discourse.50 Indeed, he continues:
When has Socrates, who we could rightly call the father of philosophy, done such a thing? That is rather the custom of those who one used to call ‘sophists’. Among the latter was Gorgias of Leontium, who challenged Socrates to propose to him any topic on which he wanted to hear him speak: a daring enterprise, I would say even a temerity, if this practice was not later transmitted to our philosophies.
Cicero specifies what could be understood as a rhetorical form of teaching: a listener poses the question and the master replies by means of a continuous discourse. He suggests that the philosophers of his time also used this form of teaching, before coming back to this detail later in the text. To this rhetorical method, the rest of Cicero’s text opposes the dialectical method:
However, this Gorgias I have just named as well as the other sophists, as we can learn from Plato, we see them played with by Socrates. Because the latter, asking questions and interrogating, had the habit of leading those with whom he discussed to reveal their own opinions, so that he could himself argue, if it seemed interesting for him, against that which they had proposed.
One is clearly speaking here of the dialectical method, since a dialogue takes place between a questioner and a respondent. ‘This method,’ Cicero continues:
was not preserved by Socrates’ successors, but Arcesilaus revived it, establishing the following rule: those who wanted to hear him speak should not ask him what he thought, but should themselves speak what they thought; when they said this, he would himself argue against them; his listeners, however, could defend their own opinion whenever possible for them.
According to Cicero, then, the Ancient Academy (‘that which comes after Socrates’) had renounced dialectical method as its means of teaching until Arcesilaus reintroduced it. In Arcesilaus’ school, the master, who is the questioner and the attacker, has no thesis. It is rather the student, the respondent, who defends a thesis. Cicero stresses the difference between this dialectical method, in which the respondent engages in a dialogue with the master, and the rhetorical method of teaching, in which there is no dialogue:
For other philosophers, he who posed a question then holds one’s tongue: this is the current practice nowadays in the Academy. He who wants to hear the master speak, tells him, for instance: ‘pleasure, it seems to me, is the sovereign good’. The master, then, argues against this opinion by giving a continuous speech, in such a way that one can conclude that those who say ‘it seems to me …’ are not completely convinced of what they say, of the opinion they express, and want to hear the contrary.
This is the rhetorical method, therefore, because the master’s discourse is continuous. Nevertheless, one can also note the point in common between the dialectical and rhetorical methods of teaching: namely, the teaching is always realised contra thesim: that is, against a thesis that one refutes.
With the support of this text, and of others that complement it, one could, in effect, present the different aspects of the teaching of philosophy in that period in the following way:
1The teaching is always done contra theism:51 which is to say, it proceeds against a determinate position interrogatively or affirmatively presented by a listener: ‘in my view, death is an evil’ or, again: ‘can the sage be angry?’n
2Regarding this thesis, the master can then develop his teaching either in a dialectical or in a rhetorical way.
3If the teaching is done in a dialectical way, then a dialogue takes place through questions and answers between the master and his listener. Against the thesis presented by the listener, the master argues syllogistically, by posing to the listener questions whose responses will necessarily lead him to accept the contradictory thesis to the one he had earlier proposed to the master. This is Arcesilaus’ and Polemon’s method.52
4If the teaching is done following the rhetorical method, no dialogue follows the initial question, the thesis proposed by a listener. But regarding this thesis, the master develops either only one continuous discourse of refutation, or two antithetical discourses, one for and the other against.53 In this latter case, the method of argumentation employed is that of in utramque partem. These two types of rhetorical argumentation were used by Xenocrates,54 Arcesilaus,55 Carneades56 and Cicero57 himself.
All this concerns only the method. However, it is evident that common methods can acquire different meanings, depending on the philosophical orientation of the masters who use them. One could, in fact, distinguish three fundamental orientations. The Ancient Academy (for instance, Xenocrates or Polemon) uses dialectic or rhetoric in a still Platonic spirit. This means that, whether rhetorical or dialectical, the argumentation is always already pre-directed. As Cicero says, the listener who poses the questions knows in advance what the master will say. Additionally, he does not maintain the thesis he proposes to the master: in truth, he shares the master’s position. The question proposed is then purely formal. It only serves as a pretext for the master to expose his own doctrine, either by means of dialectical refutation or rhetorical abundance. Arcesilaus, on the other hand, has very different intentions. He claimed to not uphold any determinate thesis himself. Arcesilaus considered himself an inheritor of Socrates, who knew only that he knew nothing. Both when he argues dialectically against a thesis in order to destroy it, and when he rhetorically develops two antithetical discourses, his aim is not that of teaching any particular content. His teaching has no doctrinal content, he only aims at provoking an epochê, the suspension of judgement. ‘He suspends the assertion,’ Diogenes Laertius says, ‘because of the contradictions between discourses.’58 Finally, the third orientation is that of the probabilistic academic sceptics. For them, the rhetorical method is that which is most important, especially under the form of in utramque partem discourse. Indeed, while remaining loyal to Arcesilaus’ ideas concerning the impossibility for the human spirit to reach truth, they also think it is possible for people to choose among opposed opinions that which will seem to be closer to the truth: the most probable one.59 From this perspective, the use of rhetorical means will allow one to be more persuasive. But one should seek equally to present the ‘pro’ and ‘contra’ so as to leave the listener the free choice.
There evidently was a sort of pre-established harmony between the Ciceronian ideal of the philosopher-orator and this tendency in the Academy of his time to grant a preponderant place to rhetoric in the teaching of philosophy. As Cicero himself puts it: ‘we use this philosophy that has engendered oratorical abundance and which expresses ideas which do not differ that much from common opinion’.60 Rhetoric, dialectic and philosophy here come together in order to allow the listener or interlocutor to discover by himself that which seems more probable to him: that is to say, that which is more likely to correspond to common notions admitted by humankind.
In what concerns the ancient Stoics, we do not know much about their method of teaching. We do know, however, that Chrysippus firmly opposed Arcesilaus’ method, and that he refused to drag the listener, especially the beginner, into the dangers of the in utramque partem argumentation. He wanted philosophers such as the Stoics who developed a science which should lead one to live a life in accordance with itself, to begin by acquiring the basic principles and by studying the doctrine as a whole.o61 The polemic with adversaries which played a key role in teaching for the Academics was for the Stoics reserved for a later stage of the philosophical education.62 However, dialectical interrogation was a procedure of exposition dear to the Stoics. Cicero tells us, indeed:
They sting you as with darts, with small sharp interrogations. But those who respond ‘yes’ to them are not transformed in their souls, and they leave in the same way they came. This is due to the fact that, even if the thoughts they express can be true and surely sublime, they are not handled as they should be, but rather, narrowly.63
For Cicero, the continuous and ornate discourse of rhetoric has more psychagogic power than the brevity of dialectical interrogations. Elsewhere, Cicero tells us that Cato, as a good Stoic, does not develop the subject matter which he treats, but that he proves his thesis by short interrogations.64 Effectively, for the Stoics, dialectic was not only used to ‘sting’p the interlocutor’s attention. Through its sharp questions, it should above all allow one to address the interlocutor beginning from that which he already accepts. Émile Bréhier had correctly shown the importance of this key aspect in Stoic dialectic:
The goal to be pursued by the Stoics is, first of all, to create a steadfast belief within the disciple’s mind; by a peculiar postulate, they superimpose the objective conditions for persuasion and a strong subjective conviction. However, the aim of the dialectician is not exactly the invention [or] the discovery of new theses; all his effort centres upon the discussion of theses which naturally present themselves to the human mind. Once they pass the testq of discussion, from uncertain and unstable opinions, they become firm and systematic beliefs. Thus, the Stoics only seek to prove theses which they consider common opinions or those which they link, by a kind of artifice, to the general beliefs of humanity: the existence of the gods, the truth of divination. Moreover, they seek to prove, less to establish the validity of a thesis than to establish a conviction capable of resisting all opposing argumentation. For this reason, dialectic is here a defensive weapon which allows one to escape one’s adversaries. This is why it contains a theory of sophisms.65
After the first century CE, the method of philosophical teaching is modified in all schools, to the extent that the master’s teaching in large measure involves an exposition of the main texts of the school founders. One can note this phenomenon equally in the Platonists, the Aristotelians and the Stoics.66 In truth, as Hans-Georg Gadamer remarked,67 one can detect in the hermeneutical relationship of the exegete with the text that he explains the same fundamental situation of dialogue that exists between an interrogator and a respondent which characterizes dialectic. This is true to the extent that the commentary generally presents itself in the form of zêtêmata;68 that is to say, under the form of questions that are put to the text, so to speak.
However, the game of questions and answers in philosophical teaching does not completely disappear. After the explanation of the text made by the master or by a student in front of the master, the students had the chance to ask questions. Here we find, once more, the ancient traditions. We know from Aulus Gellius that, when attending a course of the Platonist Taurus, he asked the master: ‘does the sage ever get angry?’ One can here recognize the situation we have already seen described by Cicero. A student poses a question with the intention of hearing the master responding ‘no’. Aulus Gellius wanted to give Taurus the occasion of demonstrating that the sage does not get angry. Taurus, however, will propose a distinction between impassibility and insensibility, giving a nuanced response. In any case, Aulus Gellius specifies that Taurus’ answer was graviter et copiose: that is, abundant and noble in style. This means that his exposition was a putting to work of the rhetorical method.69
We find the same kind of practicesr in Neoplatonism. Porphyry tells us that Plotinus would seek to stimulate the listeners to pose questions, after having explained Platonic and Aristotelian texts as he would usually do in his courses. That created a situation of disorder, according to Porphyry, opening the space for pointless questions.70 Porphyry kindly underscores how, after his arrival at Plotinus’ school, the latter chose him as a privileged interlocutor, which did not please everyone. For three consecutive days, he was the one interrogating Plotinus about the union of body and soul. A certain Thaumasius complained about it, saying he could not stand this game of questions and answers with Porphyry, but that he rather wanted to hear Plotinus himself developing a general thesis, speaking in a way which would allow one to take notes.71 Plotinus would respond that: ‘if we did not have to respond to the aporias that Porphyry’s questions present, we would not have anything to say that could be worthy of noting’.72 We find here, once again, the opposition between dialectical and rhetorical methods. The dialectical method is that which Plotinus employs: it consists in the ‘game of questions and answers’ mentioned by Thaumasius. The rhetorical method is that which Thaumasius wanted to see Plotinus practicing. The epithet employed by Thaumasius to designate the kind of discourse he wished to hear is katholou. This is a technical term to designate the ‘theses’ in the rhetorical sense of the term; i.e. the general questions developed in a continuous discourse.73
The Stoic Epictetus also divided his classes between the explanation of texts by Chrysippus or Antipater and discussions with his listeners. The Discourses published by Arrian are nothing other than these discussions with the students.74 In them, we find the trace of both continuous speeches, and thus of the rhetorical method as well as of dialogues, and thus of the dialectical method, more frequently. Epictetus knows the Socratic technique of dialogue very well, and he makes explicit reference to it.
The dialectical method still occupies an important space in the teaching of the Christian, Origen. For him, dialectical discussion between master and disciple plays an essentially critical role. According to his disciple Gregorius, Origen asked questions and listened to his students’ answers so as to test them:s
In a way that was effectively Socratic, he made us stumble by his discourse, when he saw us completely indolentt … In the beginning, it was not without pain and difficulty that we subjected ourselves to these conversations, for we were not yet accustomed or trained to follow reason, and therefore he purified us.75
These different texts confirm, then, the importance of dialectical and rhetorical methods in the philosophical teaching of antiquity.
We have presented above the radical opposition between Plato and Aristotle regarding the relations between rhetoric and dialectic, on the one hand, and philosophy, on the other. For Plato, philosophy was essentially dialectic: that is to say, communal research in dialogue, of the Truth of the forms which founds the very possibility of dialogue (and a philosophical rhetoric can only be grounded on this dialectic). For Aristotle, on the contrary, dialectic and particularly rhetoric are absolutely exterior to philosophy, given that they are but techniques of persuasion, and not sciences that would investigate a specific realm of reality. They start from common opinions rather than scientific principles. Instead of dialectic, philosophy should normally use the analytic,u a special method for scientific demonstration. Is it necessary to recall en passant that the creator of logic, in the modern sense of the word, never uses in his work the term ‘logic’ to designate a particular discipline, either dialectic or analytic?
The use of the word ‘logic’ to designate a specific part of philosophy only appears, for the first time, with the Stoics, who put it on the same plane as ethics and physics.v This Stoic logic, the science of human discourse, comprises two parts. On the one hand, there is rhetoric, defined by the Stoics as the science of speaking well; that is, of continuous discourse. On the other hand, there is dialectic, defined as the science of ‘discussing with rectitude’; which is to say, the science of dialogue or of discourse, insofar as it is broken down into questions and answers.76 The theory of this dialectic contains the following chapters, organized in a rigorous way:w
[1]one begins by establishing the cognitive relationship between man and the events of the world, a cognitive relationship designated by the name of phantasia, which we could translate as ‘representation’;
[2]next, one examines the criteria of truth of this representation;
[3]then, one proceeds to the expression of this representation in discourse, initially considering this expression as sound, as voice, and then examining its value in terms of meaningx or as ‘signified’;
[4]finally, one studies the propositions that express events and the reasonings which link these propositions among themselves.77
The theory of dialectic thus enunciates the rules which allow us to speak of reality in an exact way. True dialectic, as part of philosophy, is not this abstract theory, however. It is, rather, the practice of this theory, the lived philosophy which is a permanently vigilant attention to retain in thought and discourse an exact representation of reality. From this perspective, Stoic dialectic retains essential characteristics of the Socratic–Platonic dialectic. It is fundamentally a critique which is exercised through interrogation, addressed to oneself and others, and which aims at detecting errors in our representations and in those of others. It is a critique, but it is also a maieutic which helps us to develop the natural notions that we already have within ourselves which, through a suitable approach, can be organized into a coherent and systematic vision of reality: the human logos asks only to be in harmony with the logos of nature. Understood in this way, dialectic is then a spiritual exercise and only the sage is truly a dialectician, because he is the only one capable of avoiding errors in judgement.78
In the Academy, from Arcesilaus to Cicero, both rhetoric and dialectic, probably joined together under the name of ‘logic’ (as in the Stoics), are constitutive parts of philosophy. More precisely, they constitute the totality of philosophy, to a certain extent. This is particularly true for Arcesilaus, but also for the Academics of Cicero’s time and for Cicero himself.
We know that, for Arcesilaus, dialectical or rhetorical argumentation was exercised against every dogmatic thesis, and that it should lead to the total suspension of judgement. The other traditional parts of philosophy, namely physics and ethics, would disappear to the extent that they would not have specific contents anymore, since every thesis is refutable, as is any physical or ethical doctrine. Nevertheless, logic would assume a sort of ethical value to the extent that it would operate an epochê, a suspension of judgement leading to ataraxia,79 peace of mind. From this perspective, the practice of this rhetorico-dialectical logic had an existential significance.
Equally, in Cicero’s time, there is a preponderance of rhetorico-dialectical logic among the Academics. This time, however, the other parts of philosophy are not completely eliminated. Certain physical or ethical theses are admitted as being more probable than others. However, the method which permits recognizing them as probable, or of being convinced of them, is an essentially rhetorico-dialectical one. It is no longer a matter of completely suspending judgement, but of seeking that which is closest to the truth, as against the truth itself, and of thus producing a subjective conviction. It is a matter of finding out what one can reasonably admit. Therefore, one will seek to dialectically criticize theses that are doubtful and obscure, and to expose rhetorically – that is to say, with persuasive force – the possible solutions of a problem, with the aim of choosing that which is more probable. In this way, the disciplines of rhetoric and dialectic, which Aristotle considered foreign to philosophy, become essential to it. As for Plato, dialectic is philosophy. But the dialectic of the Academics is no longer the dialectic of Plato, the way of accessing the universe of Forms. Rather, paradoxically, it corresponds to the persuasive techniques that Aristotle rejected from the domain of philosophy, because they would be situated uniquely in the domain of the probable and of common opinions.y
It is this rhetoric–dialectic logic that we will find in Cicero. For him, logic (ratio disserendi)80 implies a dialectical part (ratio disputandi) and a rhetorical part (ratio dicendi).81 As a technique of discourse by questions and answers, dialectic implies a theory of the criteria of distinction between truth and falsity, as well as of the relations of opposition and consequence, a science of definition and division, and a teaching concerning the figures of reasoning.82 Rhetoric, on the contrary, is the technique of continuous discourse and implies a theory of invention, of disposition and of style. What is at stake in rhetoric is finding the arguments, then disposing and presenting them in such a way that the listener would be moved and persuaded.83 It is evidently this latter rhetorical part which interests Cicero, the orator. Dialectic is only valuable for him to the extent to which some of its precepts can be used by rhetoric. For that matter, Cicero reproaches the Stoics for having almost wholly neglected rhetoric.84 And by rhetoric, Cicero understands especially the theory of invention: the ‘topics’.85 He even divides logic no longer into dialectic and rhetoric, but into dialectic (ratio iudicandi) and topics (ratio inueniendi).86 For him, the Peripatetics are the masters of this art of invention. Cicero seems to believe that his Topics are a summary of Aristotle’s Topics.87 As a matter of fact, Cicero had to use a post-Aristotelian handbook which marks a clear evolution of the Aristotelian doctrine of the Topics, originally dialectical, in the direction of a rhetorical use.88 In particular, one notes that, in the second part of Cicero’s short treatise on the Topics, referring to the ‘theses’ (proposita) which one will be able to deal with thanks to the theory of the places, Cicero gives as examples theses which are surely not dialectical, since they do not admit only ‘yes’ or ‘no’ as an answer, which is characteristic of that discipline.89 Cicero’s Topics are, therefore, a characteristic example of the evolution that leads to a predominance of rhetoric over dialectic in philosophy. The ‘theses’ of which Cicero speaks correspond to Platonico-Aristotelian dialectical argumentation, in which a thesis having been proposed by a defendant,z a questioner would then seek to make the defendant accept the contradiction within it. However, Ciceronian theses are no longer dialectical arguments. The initial theme remains of the same kind: ‘should the sage get involved in public affairs?’, or again: ‘can virtue be taught?’ But the way in which the issue is handled is completely different. This will now be dealt with through a rhetorical exercise, of an exposition which takes the form of continuous discourse. This is the meaning of Cicero’s work, the Stoic Paradoxes. Cicero explicitly tells us, in the Preface, that his aim was to present the Stoic theses using the rhetorical method: which is to say, to rhetorically present these positions that were treated dialectically within this school.90
The ambiguous position of Cicero’s Topics sheds light on the hesitations of the encyclopaedists of late antiquity concerning this work. Martianus Capella places the Ciceronian theory of places in his Rhetoric.91 Cassiodorus, on the contrary, places Cicero’s Topics in the corpus of dialectical authorities that he establishes for the library of Vivarium.92
In the work of one of Cicero’s contemporaries, Antiochus of Ascalon,93 who claims to return to the doctrines of the ancient Academy, logic always entails the two parts that we referred to above: dialectic and rhetoric. However, as in Stoicism, rhetorico-dialectical logic becomes again a means to achieve and express the true and the necessary.94 The same tradition is found in Alcinous,95 a Platonist from the second century CE. Antiochus and Alcinous divided dialectic into the theory of division, the theory of definition and the theory of demonstration:96 a subdivision that will continue to take pride of place in Neoplatonism.97
With Neoplatonism, a new turning point in the evolution of the concept of dialectic takes place. This phenomenon occurs due to the integration of Aristotelian logic into the Neoplatonic system, and because of a new awareness regarding the originality of Plato’s dialectic. Dialectic in the Aristotelian sense, understood as an art of discussion which starts from accepted ideas, regains its place in the Aristotelian system of logic, alongside analytic, which represents a theory of specifically philosophical reasoning. This Aristotelian logic, which is but a tool (organon) of philosophy and a propaedeutic to it,98 is evidently distinct from Platonic dialectic.99 The latter is considered as the supreme part of philosophy which fixes our gaze on the ‘intelligible’, that nourishes our soul on the plane of truth,100 and through which we exercise contemplation of the intelligible. In the Neoplatonists, dialectic is practiced according to Plato’s own methods: division, ascension to first kinds,aa synthesis and analysis.101 However, an aspect of Plato’s dialectic disappears in the Neoplatonists. Indeed, for Plato, it was through dialogue that philosophers could experience the reality of Forms, which, in their turn, made dialogue possible and gave it its validity. On the contrary, for the Neoplatonists, dialectic becomes a monologue.102 It preserves only a very vague trace from the dialogical structure: it is perceived as a path (diexodos)103 and even an errancy which is fulfilled through the Forms. It follows the movements of procession and conversion, the phases of distinction and reunion, and the stages of internal multiplication in which the intelligible world is constituted. These procedures of analysis and synthesis, of affirmation and negation, aim at making us catch a glimpse of the reflection of the transcendent One in the intelligible multiplicity. And, for the Neoplatonists, this Platonic dialectic is but one of the methods of theology amongst others.104
Let us then conclude this too-brief sketch by making some general remarks. First of all, we have seen that the term ‘dialectic’ in antiquity can have many different meanings which correspond to the different stages of this lively history. After having been identified by Plato himself with philosophy, that is, with science, dialectic is then excluded from philosophy and reduced to a technique of persuasion by Aristotle. This technique of persuasion will be, once again, identified with philosophy by the probabilist Academics, whereas the Stoics would come to elevate the same technique of persuasion to the dignity of a scientific method, as a constitutive part of philosophy. In Neoplatonism, this Aristotelian–Stoic dialectic will again be excluded from philosophy strictly speaking; whereas Platonic dialectic is reduced to a monologue, becoming a theological method. In the Latin West, on the contrary, it is dialectic as a method of persuasion, drawn from Aristotelian and Ciceronian sources, which will be bequeathed to the high Middle Ages.105
This dialectic, as we have also seen, tends more and more to mingle with rhetoric after Aristotle. One renounces, in teaching, the technicalitybb of dialogue, which is extremely hard to operate, becoming satisfied with the simpler means offered by continuous discourse. One contents oneself with integrating into rhetoric some elements of dialectic which can be useful to the former.106 From this perspective, one should evoke the problem of this so-called literary genre that modern historians call the ‘diatribe’. As Hermann Throm had shown,107 the ‘diatribe’ is just a thesis. And a thesis is originally part of a dialectical exercise, in which one poses a dialectical problem (that is to say, a question), a problem with a broad scope to which the dialogical argumentation will have to answer. The ‘diatribe’ is, therefore, a thesis that is treated in a rhetorico-dialectical way. The only thing that survives from dialectic are the outlines of fictional dialogues (moreover, this is the reason why the philosophical works of Cicero and Seneca are called dialogi).108 It is rhetoric, however, that provides the main means of amplification and persuasion. Additionally, the themes of the ‘diatribe’ correspond to the traditional set of issues which were the object of theses.109
Even when it was combined with rhetoric, dialectic was no less dominant in the whole of ancient philosophy than rhetoric. We were able to note the importance of the ‘question–answer’ schema in philosophical teaching, involving either replying to a question by means of a continuous discourse or, on the contrary, by means of questioning dialogically. This fundamental structure, inherited from Socrates and Plato, is of capital importance for Aristotle. Concerning the latter, Ingemar Düring correctly underlined:
What characterizes Aristotle’s method is the fact that he is always in the process of discussing an issue. Every important outcome of his investigations is almost always the answer to a question posed in a clearly determined way, and it is only valid as a response to this specific question.110
Similarly, Wieland notes that ‘it is so essential for the Aristotelian structure of thought to answer to specific questions that the detailed explanation of the meaning of the question becomes itself a philosophical task’.111 In fact, one must recognize that, in this characteristic, Aristotle does not differ from the whole of the ancient philosophical tradition. As we have seen, for the Academics, teaching consisted in replying to a question, and we know that Plotinus’ treatises, for example, were written in order to respond to the questions that were posed to him in the course of his teaching.112 The Stoic doctrine could seem to present itself in a more systematic way. However, one must not forget that it is known to us by means of doxographic summaries, and that we are completely ignorant of what the works of the school founders could have been. In any case, in the Stoic works that have been preserved, as, for instance, those of Epictetus and Seneca, the question–answer structure plays a fundamental role. In general, the philosophical works of antiquity, the different treatises written by an author never present themselves as the parts of a system that would aim at being perfectly coherent or complete. Each work seeks to respond coherently to a given question, to a very precise problem. This does not exclude the incoherencies that can emerge between the different responses to different issues given by the same author. The coherence is situated within the limits of the dialogue between ‘question’ and ‘answer’. To put it in a more profound way, this ‘question–answer’ structure imposes upon thought ethical exigencies and a logic that Gadamer admirably analysed.113 Let us add that this centrality of the ‘question–answer’ schema could invite us to investigate what we might term the ‘problematics’ in ancient philosophy:cc that is to say, a reflective inventory of the types of questions to which it sought to answer.
In the ‘question–answer’ schema, the reference to the listener or interlocutor plays a central role. The philosopher always addressed someone. In the Phaedrus, Plato already wanted to constitute a philosophical rhetoric that would be able to adapt the different types of logoi to the different types of souls. In ancient philosophy, a doctrine is never completely separated from pedagogical care and philosophy, to a large extent, is identified with its teaching, which means to say that it is fundamentally formative. Teaching must always adapt itself to the exigencies of the disciples’ spiritual progress and spiritual level. Cicero, as we have seen, considered rhetoric as more efficient than dialectic to produce the conversion of his listeners. By contrast, Proclus, in his Commentary on the Alcibiades,114 affirms precisely the opposite. Either way, one must take into consideration this ‘pedagogy’ of the master if one wants to understand the precise meaning of this or that work. One would need, for instance, to determine whether it was addressed to beginners or to progressing students,dd to adversaries or to laymen.ee
This reference to the listener which corresponds to the dialectical and rhetorical aspects of ancient philosophy, invites us also to recognize the capital role played by the recourse to common notions admitted by all people. This is, even still, an important topic for reflection. Given that philosophy always implies a radical conversion from the natural attitude,ff a break with the ‘quotidian’ view of things, it is, nonetheless, concerned with addressing as its starting point what is evident and natural to man. The total transformation of one’s worldview will only be accomplished by a deepening and a systematization of these common notions.
Finally, the reflections we have developed here concerning the relationship of rhetoric, dialectic and philosophy in antiquity remind us once more that philosophy has never been pure theory; or, to put it differently, even when philosophy sought to be pure theoria, as in Aristotle, it was still the concrete choice of a life dedicated to theoria.gg In other words, philosophy, at least since Socrates, has always implied a conversion, a total transformation of one’s way of living and thinking.115 From this perspective, dialectic and rhetoric correspond to the psychagogic power of language.
1This chapter appeared in Studia philosophica (1980), 139–166. It formed the object of an exposition before the Groupe vaudois de philosophie on the 18th of June, 1979.
2One will find valuable indications on this subject in Petrus Hermanus Schrijvers, Horror ac divina voluptas. études sur la poétique et la poésie de Lucrèce (Amsterdam: A. M. Hakkert, 1970).
3Cf. Hermann Throm, Die Thesis (Paderborn: F. Schöningh, 1932); Antje Hellwig, Unterschungen zur Theorie der Rhetorikbei Platon und Aristoteles (Göttingen: Vandenhoeck & Ruprecht, 1973); John D. G. Evans, Aristotle’s Concepts of Dialectic (Cambridge: Cambridge University Press, 1977). See also the important systematization given by Alexandri Aphrodisiensis in Aristotelis Topicorum Libros octo commentaria, ed. Maximilian Wallies (Berlin; G. Reimer, 1891), 4, 6.
4Cf. Aristotle, Topics, 158a16; Prior Analytics 24a, 22ff. In rhetoric this situation is that of the amphibêtêsis, for example Aristotle, Rhetoric, 1416a9. See also ibid. 1357a4.
5Cf. Theophrastus, in Andreas Graeser, Die Logischen Fragmente des Theophrast (Berlin: de Gruyter, 1973), 4.
6Aristotle, Rhetoric, 1354a3, 1355b27; Topics, 100a25–30; Refutation of Sophisms, 183a36.
7Aristotle, Rhetoric, 1355b8 and 25-34; 1358a10–35 and 1359b8–10; Posterior Analytics, 77a 26 ff.
8Aristotle, Rhetoric, 1355a29 ff.
9Aristotle, Prior Analytics, 24a25.
10Cf. Paul Moraux, ‘La joute dialectique d’après la huitième livre des Topiques’, in Aristotle on Dialectic: The Topics, ed. Gwilym E. L. Owen (Oxford: Oxford University Press, 1968). Cf. Aristotle, Topics, 104a8.
11Aristotle, Sophistical Refutations, 183b8.
12Ibid., 183b6.
13Aristotle, Rhetoric, 1357a1–4.
14Ibid., 1354b4; 1355b25.
15Cf. Hermann Throm, Die Thesis: Ein Beitrag zu ihrer Entstehung und Geschichte (Rhetorische Studien) (Université de Lyon: Persée, 1938), 92–94.
16Cf. Heinrich Lausberg, Handbuch der literischen Rhetorik (Munich: Max Hueber Verlag München, 1960), p. 41.
17Anicius Manlius Severinus Boethius, De Topicis Differentiis, trans. Eleanore Stump (Ithaca: Cornell University Press, 1978), IV, cols. 1173–222, 1205c. We will see below, notes 89–90, that the theses do not always remain dialectical in the strict sense, but that they take on a rhetorical aspect to the extent that they have renounced argumentation by question and response.
18Plato, Phaedrus, 257b; 276e.
19Plato, Phaedrus, 262c.
20Plato, Phaedrus, 271b.
21Plato, Phaedrus, 265c–266c; 277b.
22Léon Robin, Platon, Phédon (Paris: Belles Lettres, 1952), 12, n. 2. It seems difficult to admit the theory of Gilbert Ryle, ‘Dialectic in the Academy’, in New Essays on Plato and Aristotle, ed. Renford Brambough (London: Routledge & Kegan Paul, 1965), 55 and Hans-Joachim Krämer, Platonismus und Hellenistische Philosophie (Berlin: De Gruyter, 1971), 21, according to which the word ‘dialectic’ will have had two senses in Plato’s Academy: the first will have designated Ideendialektik, that is to say the science of ideas; the second will have related to the technique of dialectical jousting of which the Topics of Aristotle had fixed the rules. One will rediscover the traces of this dialectical gymnastics in the divisions which play a great role in the last dialogues of Plato. But in this hypothesis the vocabulary of Aristotle becomes incomprehensible. We see indeed at the beginning of the Topics the distinction between dialectic (which operates conformably to opinion) and philosophy (which seeks after the truth); so why does Aristotle not distinguish between these two senses of the word ‘dialectic’ which, according to the hypothesis of Ryle and Krämer, existed in the Academy? If the word ‘dialectic’ had two senses, he would have had to say: I admit dialectic as a technique of discussion, but I do not admit dialectic as the science of Ideas. In fact, all the work of Aristotle shows clearly that what he refuses in Platonism is precisely a dialectic which will be at the same time ontological and ‘dialogical’: it is precisely the confusion between two methods that he denounces: the scientific method and the method of discussion. It seems very clear that the allusions which Plato makes to dialectic in his dialogues imply an identification between science and the technique of discussion.
23Aristotle, Metaphysics, 992b18 ff.; 1004b18; 1025b5 ff.; cf. Evans, Aristotle’s concept of dialectic, 7–52. See also Aristotle, Sophistical refutations 172a12–40.
24Aristotle, Physics, 204b4 and 10; De Caelo, 280a32; Posterior Analytics, 84a8; Metaphysics, 1029b13; Generation of Animals, 747b28. For allusions to the danger of the dialectical method, see Parts of Animals, 642b5; Generation of Animals, 748a8; Nicomachean Ethics, 1107a28.
25Aristotle, Topics 105b30; a neat distinction between philosophy, which treats of things according to truth, and dialectic, which treats them conformably to opinion (pros doxan).
26Aristotle, Prior Analytics, 24 a22; Sophistical Refutations, 172 a16.
27Aristotle, Sophistical Refutations, 165a38.
28Aristotle, Topics, 101a25 b4.
29Aristotle, Topics, 101a30.
30Cicero, De oratore, III, 80; De Finibus, V, 10; cf. Ingemar Düring, Aristoteles (Heidelberg: Bibliothek der Klassischen Altertumswissenschaften, 1966), 134, n. 59 and 155.
31Aristotle, Metaphysics, 1080a10.
32Topics, 101a37.
33Metaphysics, 1006a12; Prior Analytics, 43a38.
34Physics, 217b30.
35Wolfgang Wieland, ‘Aristoteles als Rhetoriker und di Exoterischen Schriften’, Hermes 86 (1958), 332-346.
36Simplicius, On Aristotle’s Physics, 695, 33 Diels.
37Wolfgang Wieland, Die Aristotelische Physik (Gôttingen: Vandenhoeck & Ruprecht, 1962), 216. Cf. also Jean Marie Le Blond, Loqique et méthode chez Aristote (Paris: Vrin, 1939).
38Pierre Aubenque, Le problème de l’être chez Aristote, 2e édn (Paris: Presses Universitaires de France, 1966).
39On the ‘continuity of academic dialectic’, cf. Hans-Joachim Krämer, Platonismus und Hellenistische Philosophie, 14–58.
40Diogenes Laertius, Lives of the Eminent Philosophers, IV, 18. The expression dialektika theôrêmata which we have translated by ‘dialectical questions’ recalls Aristotle, Topics, 104b1 where the ‘dialectical problem’ is defined as a theôrêma.
41Diogenes Laertius, Lives of the Eminent Philosophers, IV, 19.
42Some examples: Teles, Teletis reliquiae, 2nd ed., ed. Otto Hense (Tübingen: G. Olms, 1909), 35, 9 (and the note); Epictetus, Discourses, II, 18, 18; II, 19, 1; 111, 21, 10; 1, 7, 1; Diogenes Laertius, Lives of the Eminent Philosophers, II, 108; Sextus Empiricus, Hypotyposis, I, 20; II, 185; Cicero, De Fato, 12 (interrogare = to argue).
43Epictetus, Discourses, III, 8, 1.
44Aulus Gellius, Attic Nights, XVI, 2.
45Proclus Diadochus, Commentary on the First Alcibiades of Plato, ed. Leendert G. Westerink (Amsterdam: North-Holland Publishing, 1964), 283: in dialectical interrogations, the questions must be formulated in such a way that one can respond to them by ‘yes’ or ‘no’, and it is the respondent who ‘says’ something, that is to say who has a thesis.
46Cf. Marie-Dominique Chenu, Introduction à l’étude de Saint Thomas d’Aquin (Paris: Institut d’études Medievales, 1954), 73–77 (the disputatio). The exercise of dialectical argumentation was still practiced recently in the teaching of Thomist philosophy.
47Cicero, Orator, 46; De Finibus, V, 10; Tusculan Disputations, II, 9. Cf. Diogenes Laertius, Lives of the Eminent Philosophers, V, 3; Quintilian, Institutes of Oratory, XII, 2, 25.
48For the academic and peripatetic schools, see the preceding note; for Stoicism, one finds indications which give us glimpses of the existence of this practice in Epictetus, Discourses, II, 1, 30 & 34; II,17, 35; II, 6, 23.
49Cicero, De Finibus II, 1, 1–3. [Translator’s note: The ensuing quotes from Cicero all come from this textual location.]
50As the ensuing text shows, the schola is exactly a continuous discourse improvised in order to respond to a question posed to the improviser by an auditor; cf. Cicero, Tusculan Disputations, I, 7 and Laelius, On Friendship, 17. One finds an analogous process in the disputatio quodlibet of the middle ages.
51Cf. Diogenes Laertius, Lives of the Eminent Philosophers, IV, 19; IV, 40; V, 3; Philodemus, On Rhetoric, II, 173, 5 Sudhaus.
52Diogenes Laertius, Lives of the Eminent Philosophers, IV, 19; Polemon taught while on his feet, making a continuous discourse in response to a thesis. He argued while walking.
53Cicero, Lucullus, 7; Tusculan Disputations, II,9; De Fato, 1.
54Plutarch, Apothegmata laconica, 220e, where we see Xenocrates develop a thesis—only the sage is a good general (which assumes the question: is the sage a good general?) In what concerns the arguments in utramque partem, one can suppose in light of the testimony of Cicero that it was a current usage amongst the Academicians and Peripatetics.
55Cf. Diogenes Laertius, Lives of the Eminent Philosophers, IV, 40; Arcesilaus makes continuous discourses (legein) contra the theses.
56Lactantius, Institutes of Oratory, V, 14; this is the story of the two discourses for and against justice upheld in Rome by Carneades; cf. equally Cicero, De oratore, II, 161.
57For discourse against a thesis, see the five scholae which constitute the five books of the Tusculans; for discourse in utramque partem, see the De divinatione, according to the De Fato, 1.
58Diogenes Laertius, Lives of the Eminent Philosophers, IV, 28.
59Cf. Cicero, Lucullus, 7: ‘Ut … exprimant aliquid quod aut verum sit aut ad id quam proxime acedat.’
60Cicero, Parodoxa Stoicorum, Proem, 2: ‘Nos ea philosophia plus utimar quae peperit dicendi copiam et in qua dicuntur ea quae non multum discrepent ab opinione populari.’
61SVF, vol. II, §127 (Plutarch, De Repugnantiis Stoicorum, 1036 a).
62Cf. Ilsetraut Hadot, Seneca und die Griechisch-Römische Tradition der Seelenleitung (Berlin: De Gruyter, 1969), 55: the exposition and the discussion of adverse theses only comes in the third and last stage of the teaching, if one can judge from the order of the letters from Seneca to Lucilius.
63Cicero, De Finibus, IV, 7.
64Cicero, Parodoxa Stoicorum, Proem, 2.
65Émile Bréhier, Chrysippe et l’ancien Stoicisme (Paris: Presses Universitaires de France, 1951), 63.
66For example, Epictetus, Discourses, I,4, 7; I, 17, 13; II, 16, 34, etc.; Aulus Gellius, Attic Nights, I, 9, 9.
67Hans-Georg Gadamer, Wahrheit und Methode, 2nd edn. (Berlin: Akademie Verlag, 2011), 351 ff.
68On the history of the word zêtêma, cf. Heinrich Dörrie, Porphyrios ‘Symmikta Zetemata’ (Munich: Beck, 1959), 1–60. A great part of the treatises of Porphyry and Plotinus are zêtêmata concerning the texts of Plato.
69Aulus Gelius, Attic Nights, I, 26, 1–11; Copiose marks well the rhetorical character of the exposition, for example at Cicero, De oratore, II, 151.
70Porphyry, Life of Plotinus, 3, 36.
71Or more probably: ‘to speak on the texts’, cf. Porphyry, La vie de Plotin, vol. II, ed. Luc Brisson et al. (Paris: Vrin, 1992), 155. [Addition of 1998].)
72Porphyry, Life of Plotinus, 13, 10-17, cf. the preceding note.
73Cf. Throm, Die Thesis, 28, 87, 94, 130. The thesis, as exposed on a general theme (without consideration of the particular circumstances) is an exercise theoretically reserved to the philosophers, even though many rhetoricians practiced it.
74Cf. Epictète, entretiens, livre I, ed. and trans. Joseph Souilhé (with the collaboration of Amand Jagu) (Paris: Les Belles Lettres, 1943), préface, xxix.
75Gregory the Thaumaturge, Prosphônêtikos (Address of Thanks to Origen), VII, 97–98.
76SVF, vol. II, §48 (= Diogenes Laertius, Lives of Eminent Philosophers, VII, 41).
77Cf. Diogenes Laertius, Lives of Eminent Philosophers, VII, 43–44.
78SVF, vol .III, §548 = Ioannis Stobaei Anthologium, Vol. 2, Curtius Wachsmuth & Otto Hense eds. (Berlin: Weidmannsche Buchhandlung, 1909), 111, 18. On the development of natural notions, cf. Victor Goldschmidt, Le système Stoïcien et l’idée de temps, 2e édn (Paris: Vrin, 1977), 159. On the critique of representation, cf. Goldschmidt, ibid., 118.
79SVF, vol. II, § 48 = Diogenes Laertius, Lives of the Eminent Philosophers, VII, 41.
80Cicero, De Fato, 1; De Finibus, 1,7, 22 and IV, 8, 7; Orator, 113: dissenere embraces disputare (dialectic) and dicere (rhetoric); Prior academics, 30.
81Cicero, Orator, 113.
82Cicero, Orator, 114–117; Brutus, 152; Lucullus, 91.
83Cicero, Orator, 43 ff.; De oratore, II, 115; Brutus, 185.
84Cicero, De Finibus, IV, 3, 7 ff.
85Cicero, De Finibus, IV, 4, 8. The theory of invention is one of the five Roman rhetorical canons. See Orator, 44; Parts of Oratory, 5. See also De oratore, II, 157.
86Cicero, Topics, 6.
87Cicero, Topics, 1–5.
88The Topics of Cicero (53) nevertheless integrates the Stoic theory of reasoning, which underscores their post-Aristotelian character.
89Cicero, Topics, 83; the theses related to the descriptio are of the type: what is a miser? What is a flatterer? On the other hand, the ‘practical’ theses (86) involve exhortations or consolations, both psychagogic techniques which are evidently not dialectical.
90Cicero, Paradoxa Stoicorum, Proem, 5: “I transpose in our oratorical style what one calls thetica in the schools.” On the relations between rhetoric and philosophy in Cicero, see the fundamental work of Alain Michel, Rhétorique et philosophie chez Cicéron (Paris: Presses Universitaires de France, 1960).
91Martianus Capella, De nuptiis philologiae et mercurii, V, 474–501.
92Cassiodorus, Institutions of Divine and Human Learning, II, 15–17.
93Cf. Cicero, Prior Academics, 32 (a summary of Platonic philosophy inspired by Antiochus of Ascalon).
94Cicero, Prior Academics, 30–32.
95Albinus (Alcinous), Didaskalikos, in Platonis dialogi decundum thrasylli tetralogias dispositi, ed. Karl F. Hermann (Leipzig: Teubner, 1880), Vol. 6, 154, 7–156, 20.
96Cf. Reginald E. Witt, Albinus and the History of Middle Platonism (Cambridge: Cambridge University Press, 1937), 36 and 61. Cf. Albinus (Alcinous), Didaskalikos, 153, 26.
97Cf. Porphyry, Porphyrii Isagoge et in Aristotelis categoris commentarium ed. Adolfus Basse (Berlin: G. Reimer, 1887), 1, 5–6; Proclus Diadochus, Commentarium in Platonis Parmenidem, In Procli Opera Inedita, ed. Victor Cousin (Paris: Augustus Durand, 1864), 982, 11.
98Cf. Proclus, Theologie Platonicienne, ed. and trans. Henri-Dominique Saffrey & Leenert G. Westerink (Paris: Les Belles Lettres, 1968), 1, 10, 18.
99The opposition appears clearly in Plotinus, Enneads, I, 3, 4, 18, but already in Origen, In Canticum Canticorum, in Origenes werke, Vol. 8 (GCS 33), ed. Wilhelm A. Baehrens (Berlin: C. Hinrichs, 1925), 79.
100Plotinus, Enneads, I, 3, 4, 10 and 5, 9.
101Plotinus, Enneads, I, 3, 4, 12–16; cf. Werner Beierwaltes, Proklos, 2nd ed. (Frankfurt-am-Main: Klostermann Vittorio GmbH; 1979), 248 ff.
102Cf. Beierwaltes, Proklos, 240, n. 1.
103Proclus Diadochus, Commentarium in Platonis Parmenidem, 993, 9.
104Proclus, Theologie Platonicienne, 1, 20, 6 ff.; the modes of theological exposition are the Orphic mode, which utilises symbols, the Pythagorean mode, which utilizes numerical symbols, the Chaldean mode, which is the direct revelation under divine inspiration, and the dialectical mode proper to Plato.
105Abelard for example is still very close to the Ciceronian tradition of the Topics, as I have indicated in Marius Victorinus (Paris: Collection des Études Augustiniennes, 1967), 197, n. 36.
106Cicero says that dialectic is a contracted rhetoric and that rhetoric is a dilated dialectic, at Brutus, 309.
107Throm, Die Thesis, 77 and 149. Throm is perhaps wrong to tie too exclusively dialectic in ‘general’ to rhetoric in ‘particular’.
108Cf. Hadot, Marius Victorinus, 211–214.
109Throm, Die Thesis, p. 78–79.
110Ingemar Düring, ‘Aristoteles und die Platonische Erbe’, in Aristoteles in der Neueren Forschung, Paul Moraux ed. (Darmstadt: Wissenschaftliche Buchgesellschaft, 1968), 247.
111Wieland, Die Aristotelische Physik, p. 325.
112Porphyry, Life of Plotinus, 4, 11 and 5, 60.
113Gadamer, Wahrheit und Methode, 345: ‘Fragen heist ins Offene stellen. Die Offenheit des Gefragen besteht in dem Nichtfestgelegtsein der Antwort … Nun ist die Offenheit der Frage keine uferlose. Sie schliesst vielmehr die bestimmte Ungrenzung durch des Fragenhorizont ein.’
114Proclus Diadochus, Commentary on the First Alcibiades of Plato, 172, 6: the dialectical genre obliges the audience to pay more attention than continuous discourse; it obliges the interlocutor to seek and to find for themselves.
115Cf. Pierre Hadot, ‘Spiritual Exercises’, in Philosophy as a way of life, 79–125 [and Chapter 4 above].
a [Translator’s note] French: auditeur. Sometimes we will translate this more literally as ‘auditor’ or ‘listener’, in the context of a dialogue or discussion, the addressee of a philosophical discourse participates in discourse and is, therefore, an ‘interlocutor’.
b [Translator’s note] French: le pour et le contre; more literally: the positions ‘for’ and ‘against’ a certain thesis.
c [Translator’s note] Literally, from Hadot’s attaquant.
d [Translator’s note] French: pièges.
e [Translator’s note] The French is interrogateur, which we render as ‘questioner’, given the historical and legal associations in the English ‘interrogator’. ‘Inquirer’ might also have been used but it is specifically someone charged with asking questions according to certain rules that is at stake here. Inquiry may involve asking questions, but it also has a wider denotation.
f [Translator’s note] French: données.
g [Translator’s note] Original: cette Forme première.
h [Translator’s note] Hadot’s adjective is scolaire.
i [Translator’s note] Hadot’s phrase is schémas-types.
j [Translator’s note] Here, exercise scolaire: we translate the adjective as ‘pedagogical’ to avoid the association of the term ‘scholastic’ in English with specifically medieval forms of university teaching. ‘Scholarly’ in English also carries a different set of connotations than those in play here; Hadot’s term designates something closer to a ‘school exercise’, which we will accordingly sometimes use.
k [Translator’s note] Literally, celui qui interroge.
l [Translator’s note] Original: méthodes d’enseignement.
m [Translator’s note] Rackham in the Loeb edition of the text translates schola as ‘formal lecture’, at Marcus Tullius Cicero, On Ends, trans. H. Rackham (Cambridge, MA, and London: Harvard University Press, 1931), 79.
n [Translator’s note] The French expression is se mettre en colère…
o [Translator’s note] Hadot’s expression is, more literally, ‘from one end to the other’ (d’un bout à l’autre).
p [Translator’s note] Original: ‘piquer’.
q [Translator’s note] Here: épreuve.
r [Translator’s note] Hadot’s term is usages here: viz. customs, usages, manners, fashions, customary or habitual practices.
s [Translator’s note] Again: les éprouver.
t [Translator’s note] Original: réstifs.
u [Translator’s note] French: l’analytique.
v [Translator’s note] See Chapter 6 above.
w [Translator’s note] We have added the indentation and enumeration of Hadot’s inventory here, for purposes of clarification. The original is continuous text.
x [Translator’s note] The French is: valeur de sense.
y [Translator’s note] See section 1 above.
z [Translator’s note] Original: défenseur.
aa [Translator’s note] Original: la remontée aux genres premiers.
bb [Translator’s note] Original: technicité.
cc [Translator’s note] Original: la problématique dans la philosophie antique. See Chapter 2 above.
dd [Translator’s note] Hadot’s expression is progressants, and it is possible to surmise that he had in mind here the Greek prokoptonta, ‘those making progress’ in the Stoic texts.
ee [Translator’s note] French: les profanes.
ff [Translator’s note] Hadot’s expression is, exactly, attitude naturelle. Given his awareness of Husserl’s work (see Chapter 1 above), it is probable that Hadot wanted to evoke the Husserlian connotations of this term here.
gg [Translator’s note] See Chapter 3 above, where Hadot develops this reading of Aristotle’s conception of theoria.