Introduction*
Concerning dialectic, Plato and Aristotle might be thought to stand on each side of a very wide gap. To Plato, dialectic is the best means available to philosophy for reaching truth, whereas Aristotle seems to grant dialectic little more than the function of testing propositions and thus denies a direct access to philosophical insight through dialectic. However, even if this were an adequate description of Platonic and Aristotelian dialectic (and it hardly is), one question would remain: what happened in between, or in other words, how did the concept of dialectic develop from Plato to Aristotle? The present volume aims at giving some answers to this question.
The last four decades of scholarship in ancient philosophy have produced numerous investigations of dialectic concerned with Plato or Aristotle separately, but there has been virtually nothing on the development of dialectic from Plato to Aristotle. This is not only true in the English-speaking world, but holds for the last forty years of French and German scholarship as well. The remarkable essays in G.E.L. Owen’sAristotle on Dialectic – the Topics (Reference Owen1968), W.A. de Pater’sLes Topiques d’Aristote et la dialectique platonicienne (Reference Pater1965), and the investigations of J. Stenzel, Studien zur Entwicklung der platonischen Dialektik von Sokrates zu Aristoteles (Reference Stenzel1931), are the most recent works which may be said to be concerned with the development of dialectic from Plato to Aristotle.1 But even these studies do not cover all the aspects of the issue (nor do they claim to do so); they operate, rather, within a somewhat narrow conception of dialectic that is clearly reflected in the topics dealt with. The focus is primarily on methodology (dialectic and definition) and ontology (dialectic as concerned with forms, ideas or principles). Quite generally, one might say that these previous investigations place their emphasis on dialectic as a theoretical issue and tone down the fact that ancient dialectic is also intended for use in an actual debate with a real interlocutor.
There are probably two main reasons for this. First, these studies all flow from the vein of developmentalism, which focuses on theory or doctrine as a natural starting point. Second, this neglect of the practical aspect of dialectic might stem from a predominantly modern concept of epistemology as concerned ‘monologically’ with the relation between knower and object (and less with the epistemology of two opposed claims to knowledge and their proponents, confronted in argument). Be this as it may, the focus on method and ontology has also left its mark on some of the most influential studies of Platonic and Aristotelian dialectic of the last century. In his Plato’s Earlier Dialectic, Robinson is well aware of the problems posed by reading Plato’s dialogues as evidence for a theory of dialectic; but, nevertheless, this is what he does in the main part of this seminal work.2 Likewise, Vlastos in ‘The Socratic elenchus’ (Reference Vlastos1983) treats the logic of the elenchus and its methodological status largely in abstraction from the dialectical setting of the elenchus. Among Aristotelian scholars, Owen’s ‘Τιθέναι τὰ φαινόμενα’ (Reference Owen and Mansion1961) started the still-unsettled debate about the epistemic status of the premises in dialectical argument (the so-called ἔνδοξα), which in time turned into a problem about the role of dialectic in establishing the foundations of knowledge or science.
The contributors to the present volume do not abandon this interest in dialectical method or ontology. In the present context, however, dialectic means primarily argumentation directed at an interlocutor, or in the words of Aristotle: dialectic is argumentation πρὸς ἕτερον (Top. 8.1.155b7). The practice of dialectical argumentation and its extension into the literary form of the dialogue makes up the core of the present volume. The main part of this introduction is devoted to an outline of dialectic conceived primarily as question-and-answer argument.
Dialectic
The contributions are not concerned with questions about origin. Whether there was dialectic before Socrates (as the Ancients themselves seem to have believed), what it was like and how it took shape and evolved must be addressed at some other occasion.3 In the context of the present volume, ‘dialectic’ is a form of argument closely associated with the enigmatic figure of Socrates as depicted by Plato. The Apology contains the following outline of Socrates’ manner of arguing (20c–23c): acting as questioner, Socrates enters into dialogue (διαλεγόμενος) for the sake of examining (ἐξέτασις) claims to knowledge elicited from a respondent, whom he scrutinizes in the elenchus (ἐλέγχειν). However, to Socrates, question-and-answer dialectic is as much a certain form of conducting one’s life as it is a certain form of conducting an argument (Ap. 28e); and, according to the man himself, it is this dialectical ‘business’ that has brought him before the court to defend his life (Ap. 20c–d). The activity here described came to be designated διαλέγεσθαι (conducting a dialectical argument); and it is precisely this Socratic way of having an argument that might be said to bridge the apparent gap between the Platonic and Aristotelian concepts of dialectic, since this ‘business’ of question and answer provides one basic, common feature in the dialectic of both.To Plato, as to Aristotle, the fundamental meaning of conducting a dialectical debate is captured by the Socratic notion of giving an account or taking one up for examination in question-and-answer form, i.e. λόγον δοῦναι / λαβεῖν (Prt. 336c–d, Rep. 7.531d–e; Top. 1.1.100a18–20, SE 1.165a24–8). It is as the heirs of Socratic dialectic that the Platonic and Aristotelian conceptions of dialectic become comparable.
The illustration overleaf represents a general model of Socratic dialectic (πρὸς ἕτερον-argumentation). It is intended as a framework suitable to describe the development of dialectic from Plato to Aristotle; thus, it represents the common ground shared by both. The main characters are the questioner and the respondent (1–2). Every other element in the illustration (3–7) refers back to these in some way.
Each of the seven elements in this illustration appears in some form in Plato’s depiction of dialectical argument (references will be given as we proceed). But the description of each element is primarily taken from Aristotle’s Topics and Sophistical Refutations. In these writings, Aristotle seems to describe and refine a method of argumentation which he found, at least partly, in Plato’s depiction of dialectic. The dialectical discussions in the Academy must, of course, also have been a source of inspiration to Aristotle; but presently we shall concentrate attention on the relation to Plato’s dialogues. Obviously, the use of Aristotle’s terminology to describe a common basis of dialectic involves the risk of misrepresenting Plato’s dialectic or distorting the picture of development by viewing dialectic from the point of view of its final stage rather than from its beginning. Aristotle’s categorizations of dialectical argument do indeed often seem to make explicit what is only implicit in Plato’s depiction of dialectic; but even so, we shall try to steer clear of misrepresenting Plato by using Aristotle’s categorizations mainly as a heuristic tool. The strength of the present procedure is that it provides the present investigation with a set of precise points of comparison. It cannot be completely neutral, but, as will become clear, Aristotle’s terminology is surprisingly well suited to describe even the points on which Plato and Aristotle dissent considerably.
Graphics by Martin Emborg
It might seem objectionable to draw on the Sophistical Refutations in a description of Socratic dialectic. Is dialectical argument not fair play, as opposed to the foul play of the sophists? Maybe so, but in the practice of question-and-answer argument, some interlocutors will inevitably use foul play. Thus, Plato tacitly and Aristotle openly insist that it belongs to the dialectician to know and be able to counter sophistical arguments (Euthydemus, Sophist, SE 11.172b5–8, Rhet. 1.1.1355b15–17).4
(1) Questioner
The questioner must extract a thesis from the respondent and test or refute it by asking questions that will lead the respondent to grant premises from which the questioner may infer a contradiction of the thesis (see Prt. 349a7–d1). This argumentative activity of the questioner is called ‘testing or refuting’ (ἐλέγχειν), ‘examining’ (ἐξετάζειν), ‘attacking’ (ἐπιχειρεῖν), ‘destroying’ (ἀνασκευάζειν) the thesis of the respondent, or ‘constructing’ (κατασκευάζειν) a claim in case the respondent defends a negative thesis. All this corresponds to the Socratic notion of taking up an argument for examination, i.e. λόγον λαβεῖν (Men. 75d1–2). It is unclear how Plato would describe the logic of the arguments at the disposal of the questioner. He rarely mentions ‘deductive argument’ (συλλογισμός, Tht. 186d, Ti. 87c) and never uses ‘induction’ (ἐπαγωγή) in a logical sense.5 In Aristotle’s terminology, the questioner has these two types of argument at his disposal (Top. 1.12.105a10–12), and some of the forms of dialectical argument found in Plato; division (διαίρεσις) and the Socratic inference by analogy count, according to Aristotle, as special instances of συλλογισμός and ἐπαγωγή respectively (An.Pr. 1.31.46a32–3, Rhet. 2.20.1393a22–b8). Aristotle’s broad definition of ‘συλλογισμός’ suggests that he thought of this as any valid argument (Top. 1.1.100a25–7). The logic of the dialectical συλλογισμός (Top. 1.1.100a29–30) and the epistemic achievement of the elenchus have been major matters of controversy for many years and there is still no general agreement among scholars as to how these problems should be solved.6
The question is the prevalent mode of speech in dialectic. It is the engine that drives the debate forward, with the answers acting like fuel, as it were.7 There is, however, disagreement between Plato and Aristotle as to what sorts of question should count as genuine or legitimate dialectical questions. According to Aristotle, the question, ‘What is X?’ (which is typical of Socrates, as depicted in some of Plato’s dialogues) is illegitimate in dialectic. One should not ask, ‘What is virtue?’, but rather offer a proposition, which may be answered by a yes or a no (Top. 8.2.158a14–22); for example, ‘Can virtue be taught?’ This dissent obviously indicates a change in the concept of dialectic, but it does nothing to impede or challenge the common basis of dialectic in Plato and Aristotle.
The questioner has a set of different options at hand in posing his questions. Aristotle offers the following list: a questioner may introduce distinctions between the different meanings of the same term (Top. 8.2.157b2–7, Prt. 329c–d, Men. 74c); he may ask the respondent to provide a counterexample to a universal claim, which the respondent refuses to grant (Top. 8.2.157a34–5); he may censure the respondent if he does not abide by the rules of the debate, or he may even break off the argument (Top. 8.2.158a25–30, Prt. 335a–d, Grg. 461e–462a); finally, he may obtain premises which go beyond what is necessary for a deduction of the contradictory of the respondent’s thesis (Top. 8.1.155b20–4). Such premises might serve to support one of the questioner’s necessary premises by ensuring an induction of the universal, which might make the respondent accept the necessary premises more easily (Men. 72d–73c); or they might serve to prolong the argument in order to either confuse the respondent (e.g. by introducing irrelevant premises) or to conceal the conclusion the questioner is aiming at. Finally, these ‘extra’ premises may simply aim to make the argument clearer, e.g. by referring to examples, poetry, or analogies (Prt. 330a–b, Lys. 215c–d, Rep. 1.334a–b). It seems quite fair to think that a majority of these Aristotelian manoeuvres reflect the practice of dialectic as depicted in Plato’s dialogues.
The premises offered by the questioner must be what Aristotle calls ‘endoxic’ premises (ἔνδοξα), which means that they must express claims that are acceptable either to all men, or to the majority, or to the wise (the majority of them or the most renowned) (Top. 1.1.100b21–3). This assertion might seem to conflict with the Platonic Socrates and his dismissal of ‘what other people say’ as irrelevant to dialectical argument (Chrm. 161b–c). On consideration, however, the arguments in the Charmides or Laches – as in most Platonic dialogues – are actually based on common assumptions about temperance or courage. Socrates’ demand that his interlocutor should say what he means, then, might do no more than express the general point that a dialectical premise should be acceptable to the respondent. In fact, Aristotle’s different types of ἔνδοξα probably correspond to different types of respondents. The dialectical problem, ‘Is the earth made up of squares or triangles?’, would not yield an endoxic thesis if offered to an ordinary man, since neither answer would be acceptable to either all or the majority of men (I presume). If posed, however, to a member of the Academy armed with the Timaeus, this problem might actually yield an endoxic thesis, at least insofar as members of the Academy might want to affirm that ‘the world is made up of triangles’; i.e. this claim would be acceptable to a renowned group of philosophers. The point of demanding endoxic premises as a constitutive element in a dialectical debate seems to be, then, that such a debate must take place on some shared ground of rationality. Dialectic will not debate claims that are in no need of argumentative support, or claims which no-one would find credible in some sense. The problem, ‘Is arithmetic ambition or distraction?’, only makes sense in the somewhat private rationality of Wonderland. Dialectic, however, is based on a common rationality in the world of men (see Top. 1.11.105a3–9). Even if it is confined to the realm of opinions, appearances, and controversy, dialectic is still based on minimum norms of rationality, such as a shared language and men’s willingness to accept inferences or apparent inferences.
This requirement in regard to dialectical claims and premises gives rise to a delicate problem: must the questioner be committed to the claim he is striving to render victorious? And subsequent upon this, must the questioner have some positive knowledge to conduct the debate – and if so, what kind of knowledge? Whereas it seems clear in regard to Aristotle that both questioner and respondent might, depending on the type of dialectic pursued, dissociate themselves from the claims they are representing (Top. 8.5.159b27–35), it is disputed in regard to Plato whether the questioner (normally Socrates) must endorse the claims he brings to victory (if any) when subjecting a respondent to his elenchus.8 The difficulties connected to this, obviously, have to do with the fact that Socrates persistently claims not to know anything or just small and trivial matters (for the inevitable exception, see e.g. Smp. 177d). Here is not the place to go any further into the much-debated topic of Socratic ignorance. Aristotle seems to allow for both a knowing and an ignorant questioner depending on the form of dialectic being conducted. In so-called peirastic dialectic, a questioner might be a non-specialist or even an ignorant (SE 11.172a21–7), and Aristotle may very well have Socrates in mind here (see SE 34.183a37-b8). But as a rule, the questioner in Aristotle’s dialectic has quite a lot of knowledge, even if this is confined to the realm of δόξα (Top. 1.14.105b30–1). First of all, the dialectician knows how to argue, i.e. he knows a method (Top. 1.1.100a18–21); and he knows τόποι (see below), definitions, arguments (Top. 8.14.163b17–23), and ἔνδοξα (Top. 1.14.105a34–b18). Obviously, these problems greatly influence how one understands the epistemic status and achievement of not only dialectical arguments but also the proponents of dialectical arguments.
The idea that Aristotle describes a method which he found depicted in Plato’s dialogues faces a problem in regard to a very central part of Aristotle’s dialectic, the τόπος. It is unclear, and not terribly well researched, whether there is anything in Plato’ depiction of Socrates that could count as argument by means of a τόπος.9 The four predicables, which formally structure the classification of τόποι, might partly have been anticipated by the treatment of division (διαίρεσις) in Plato’s dialogues; but the only other clear relation to Socrates seems to be that a τόπος is primarily useful for destroying or establishing definitions. Since, however, no-one has been able to show any clear line of development from Plato to Aristotle in regard to the τόπος, this will not be considered any further here.10
(2) Respondent
If the questioner must strive to deduce a contradiction of the respondent’s thesis (or lead the respondent to hold a paradox or maintain a clearly false conclusion), the task of the respondent, on the other hand, is to solve (λύειν) the argument put forward by the questioner, or, if he cannot do so, at least hinder or block (κωλύειν) any conclusion so as to stand un-refuted (Top. 8.10.161a1–15). The most efficient way of solving or freeing oneself from a refutation is to point out why a premise of the argument is false or otherwise problematic (Top. 8.10.160b33–9,Euthd. 277e–278a). If refutation is inevitable, the respondent must show that the refutation resulted from the thesis itself, not from his way of handling it (Top. 8.4.159a18–22). This activity corresponds to the Socratic notion of giving an account by answering questions, i.e. λόγον δοῦναι (with slight terminological variation, Grg. 474b, Prt. 339a). Aristotle makes a noticeable observation about the respondent. He claims that there has been no former clarification of the dialectical manoeuvres with which the respondent may conduct his part of the argument (Top. 8.5.159a32–7). Subsequently, he lays down the rules for the respondent himself. The respondent’s options are not confined to confirming or denying the premise offered by the questioner by simply answering yes or no (this was apparently the state of affairs among Aristotle’s predecessors; see SE 17.175b7–10). A respondent may make distinctions so as to avoid ambiguity in the terms he concedes (Top. 8.7.160a22–34, Euthd. 295b–d, Men. 75d5–7), and he may simply say, ‘I don’t understand’, if offered an ambiguous premise (Top. 8.7.160a17–22, Tht. 164d, Sph. 222d). These simple moves are in fact formidable defensive tools in the hand of the respondent, since one of the most effective and most widespread sources of elenchus is the ambiguity of words (SE 1.165a3–13). Failure to introduce a distinction when ambiguous terms appear in the premises is, therefore, to count as a grave fault on the part of the respondent (SE 17.175b28–33). It is interesting to note that these moves allowed to the respondent do in fact reflect how the dialecticians (not the sophists or the ordinary interlocutors) in Plato’s dialogues argue – as did most of the moves allowed to the questioner.
In case the questioner forces the respondent to assert two incompatible claims, and the respondent can offer no solution (λύσις) to the argument, the respondent is reduced to an ‘impasse’ (ἀπορία). This receives different interpretations by Plato and Aristotle in regard to its meaning and place in dialectical argument, but the minimal basic sense of ἀπορία seems to be the intellectual loss resulting from having all one’s argumentative means exhausted. It can either occur as a result of an elenchus (Men. 84a–b) or because the respondent is faced with opposed but equally powerful inferences (Top. 6.6.145b16–20). The philosophical purpose of ἀπορία will be touched upon below.
In Plato the respondent seems to be expected to endorse the claims he makes (but see note 8 above), whereas this is not necessarily the case according to Aristotle. The importance of this observation and its impact on the development of dialectic is an issue which needs to be further investigated. It is, in fact, generally the case that the role of the respondent in both Plato and Aristotle has received remarkably little direct scholarly attention.11
(3) Relation of questioner to respondent
The questioner and respondent might stand in an equal or an unequal epistemic relation to one another, since the interlocutors might both be dialecticians, both ordinary men, or one dialectician and one an ordinary man (Top. 1.12.105a16–19, Top. 8.2.157a18–20, SE 11.172a23–7). In Plato’s writings we normally find an unequal relation between the questioner (old, experienced, superior) and the respondent (young, inexperienced, inferior). This is the case with Socrates in the role of the superior and experienced questioner, e.g. in Lysis (223a–b) and Meno (76a–c); whereas the tables are turned in Parmenides, with Parmenides as the old, experienced and superior questioner engaged in discussion (in the first part) with Socrates as the young, inexperienced respondent (Prm. 127b–c). Aristotle, on the other hand, seems to envisage the relation between questioner and respondent in more equal terms. An equal relation obtains when two fully trained and fully experienced dialecticians enter into debate. Even though this situation is never expressly mentioned in the Topics, it seems, nevertheless, to be the situation for which the rules of the Topics were primarily designed; i.e. a situation of arguing with or against other philosophical schools, i.e. mainly the Academy.12 Whereas a dialectical debate on the basis of an unequal relation seems to imply the possibility of some sort of progress in knowledge, the purpose of a debate among equals might be intellectual gymnastics, testing the respondent if he makes a claim to knowledge or, more ambitiously, to settle a philosophical dispute among competing schools.
(4) Subject in question
It is a danger in any debate that questioner and respondent do not direct their argument at the same subject matter. In fact, as is clear from the Euthydemus, it seems to be a sophistic strategy to ‘go for the name’ (τὸ πρὸς τοὔνομα διαλέγεσθαι, Top. 1.18.108a35) instead of the thing in question, i.e. to direct one’s argument against a specific sense of some term knowing well that the respondent takes it in some other sense. Socratic dialectic, on the other hand, seems to be based on the demand that both interlocutors agree on the meaning of the terms under debate. According to Meno, doing so must count as ‘more dialectical’ (Men. 75d) than striving to refute mere words; and in the Topics, Aristotle points out that a questioner acts ridiculously if he questions with one specific meaning of a term in mind, although the meaning of the term has been agreed by both interlocutors to be something else (Top. 1.18.108a18–37). The real danger here is that of dialectic turning into eristic or mere word play. The importance of the distinction between arguing about a thing or arguing about a word lies in pointing out that dialectic proper is a serious business and is philosophically more ambitious than simple word-play (see e.g. Euthd. 285a, Rep. 5.454a, SE 5.167a23–5 and SE 10).
(5) Character of questioner and respondent
As mentioned already, question-and-answer dialectic is not confined to valid (or apparently valid) rational argument. The dialectician may employ argumentative strategies which involve his own or his interlocutor’s character. Consequently, the character of the proponents of the arguments is a feature which must be taken into account in a dialectical debate. The importance of character is neatly brought out in the literary settings of Plato’s dialogues. It is obvious from these that being ill-tempered (Thrasymachus, Rep. 1.336b–337a), arrogant (Euthydemus, Euthd. 293a), or, on the positive side, courageous and mildly disposed (Theaetetus, Tht. 144b) interferes heavily with a respondent’s ability to argue dialectically. In general terms, Plato points out that it is no small part of dialectic to know different kinds of interlocutors (their souls) and address them in different ways according to their character (Phdr. 276e–277a, 277b–c). Aristotle seems to have transformed this into a technical use of character, which is quite different from Plato’s Phaedrus. The character of the interlocutors may indeed form part of the argumentative strategy of the questioner. If he forces the respondent to accept a nasty moral claim, this will make the respondent’s argumentative position very difficult, because in the eyes of the audience he will appear to have a disgusting character, and they will hate him for it (Top. 8.9.160b17–22). Correspondingly, the respondent should generally be on guard against conducting his part of the debate in a way that would make him appear to be an obstructive interlocutor (δυσκολαίνειν), which is both a technical fault and a disadvantage in regard to character (Top. 8.8.160b2–13, SE 17.175b33–6). Such uses of character surely bring dialectic close to sophistry, but since they do appear in the arsenal of some interlocutors, the dialectician must know them as well. The psychology of dialectical argument is a very real feature of Socratic dialectic.
(6) The audience
From dialogues such as Lysis or Euthydemus, it is clear that dialectical debates often attracted an audience. From the point of view of dialectical argument, the audience is quite naturally not as important as the interlocutors themselves. Nevertheless, the presence of an audience interferes with the argumentative options of the interlocutors. There seems to be three ways in which an audience can alter the conditions under which the dialectician operates. The first consists in having the audience form a negative opinion about the character of one’s opponent (see above). The second is slightly more sophisticated: the questioner might hope to convince the audience that his argument, though actually flawed, has brought about a successful refutation of the respondent. If this manoeuvre succeeds, the confusion as to whether the respondent has been refuted or not will be to the advantage of the questioner, or, even worse, in the eyes of the audience the respondent will appear to have been refuted, since the members of the audience will themselves jump to conclusions (SE 22.178a20–7, An.Post. 1.24.85b18–22; see Ap. 27a, where Socrates accuses Meletus of answering in such a manner). This, obviously, makes the position of the respondent difficult, so much so that Aristotle warns not against actually being refuted but against appearing to have been refuted (SE 17.175a40–41).13 That this is an important issue for both Plato and Aristotle is clear from the fact that both take some trouble over explaining how it is possible that a respondent in possession of truth might appear to have been refuted in the eyes of an audience (Epist. 7.343c–e, SE 16.175a20–30). Considering that the normal reaction of the audience seems to have been applause for the winner, apparently with an implied ridicule of the loser (Euthd. 276b–c), this concern is quite natural.
The third and last role of the audience is to act as a jury of the debate. This is not found in anything but a metaphorical way in Plato (Phd. 63b, e) and seems elsewhere to be rejected as irrelevant to dialectic (Rep. 1.348a-b), though the audience in the ‘non-dialogical’ Timaeus is given the role of ‘judges’ (Ti. 29c–d). In Aristotle, however, some debates apparently had a fixed duration of time within which the argument should be concluded (Top. 8.10.161a10–12). Afterwards, some evaluation seems to have taken place. Aristotle distinguishes between evaluating the argument in itself and evaluating the manner in which it was handled by the questioner (Top. 8.11.161a16–b10). An argument might be intrinsically bad, but nevertheless the questioner might have argued his case in the best possible way. So clearly, the evaluation should focus on the dialectical skill of the questioner and on the intrinsic qualities of the argument as separate issues (Aristotle offers five points of critique relevant to the argument in itself).14 The main concern of the critique seems to be to determine whether the debate was conducted in the right way (καλῶς); and if not, to whom the failure should be ascribed. Arguing under these constraints ought to have a considerable impact on the argumentative behaviour of questioner and respondent.
(7) Use of dialectical argument
The main difference between Plato and Aristotle in regard to dialectic occurs in their assertions of the use (χρῆσις) or uses of dialectic. According to Topics, dialectic is useful for exercise (γυμνασία), for casual encounters (ἐντεύξεις), and finally with regard to the philosophical forms of knowledge (πρὸς τὰς κατὰ φιλοσοφίαν ἐπιστήμας), i.e. with regard to truth (Top. 1.2.101a25–8). This list of uses is apparently intended by Aristotle to be exhaustive, and in fact, it seems to cover all the main aspects not only of his own but also of Plato’s conception of dialectic. Not surprisingly, however, there is considerable divergence as to how these uses are interpreted in Plato and Aristotle respectively.
The use of dialectic for exercise or training is a prominent feature of the dialectic of the Parmenides and the point of the exercise is quite serious, i.e. to reach truth for which the exercise seems to act as an enabling condition (Prm. 135d–e). However, contrary to what we find in Aristotle, the exercise of the philosopher’s rational faculties is usually a task belonging to mathematics in Plato (Rep. 7.525d–526b). This probably reflects the fact that, to Plato, dialectic is primarily concerned with truth (corresponding in some way to Aristotle’s third use – πρὸς τὰς κατὰ φιλοσοφίαν ἐπιστήμας). To Aristotle, exercise might be the primary function of dialectic.15 The point of the exercise is to acquire or maintain dialectical skill (Top. 8.14.164a16–b2, Top. 1.2.101a28–30). Further, the experience, the swiftness of mind, and the sense for the right moment, all of which result from dialectical exercise, are effective safeguards against being refuted – especially in regard to sophistic refutations (SE 16.175a12–26). Finally, the exercise also seems to serve one of the philosophical uses of dialectic: going through ἀπορίαι, as will become clear below.
The second use of dialectic is in regard to so-called casual encounters (ἐντεύξεις). Dialectic is useful in regard to these because, having acquainted himself with the δόξα of ordinary people, a dialectician may communicate on this basis, i.e. from assertions which are familiar and acceptable to his ordinary interlocutor, and use this as a vantage point for changing the interlocutor’s mind, persuading him as it were (Top. 1.2.101a30–4). Aristotle offers no explanation of the term ‘casual encounter’, but from the Rhetoric it is clear that this use of dialectic is associated with persuasion (Rhet. 1.1.1355a24–9). As such, the use of dialectic in casual encounters may reflect Plato’s Phaedrus in some sense. On the surface, this dialogue treats rhetoric; but persuasion, rightly understood, turns out to be a matter for dialectic (Phdr. 271cff.). But what exactly is ἔντευξις-dialectic? The term suggests two distinct but related ideas. Evidently, Aristotle means ἔντευξις to cover a dialectical situation which is partly a result of chance. This corresponds neatly to the situation depicted in a number of Plato’s dialogues, which are exactly casual dialectical encounters.16 It may, however, also refer to the formal character of this sort of dialectic as a less strict way of arguing. This should probably be taken to indicate a less regulated but not necessarily a less valid form of argumentation, maybe adjusted to the un-technical manner of arguing of ordinary men (Rhet. 1.1.1354a1–7). This does not imply that dialectic used in casual encounters should be neglected by philosophers. To the contrary, the reference to persuasion suggests that this use of dialectic aims primarily to change the mind of the respondent (cf. δόξα), rather than to test or refute his thesis formally. As such it presupposes at least that the respondent actually believes the claims he is defending, and further, that there is such a thing as philosophical persuasion (see Prt. 352d–353a, Euthd. 275a).
The third use of dialectical argument concerns the philosophical forms of knowledge. The interpretation of this has been among the most controversial issues in Aristotelian scholarship for some decades. Whereas Plato seems to have no problems with dialectic as a means to reach truth, Aristotle is more hesitant about the role of dialectic within philosophy. Let me be very brief here. Dialectic is useful to philosophy in two respects: (1) with a view to seeing whether a claim or its contradictory is true (or false); and (2) with a view to establishing or critically testing the principles of the philosophical forms of knowledge (Top. 1.2.101a34–b4).
Of these, the first is perhaps the least controversial. Using dialectic to gain an overview of some disputed claim consists in going through the ἀπορίαι on both sides (πρὸς ἀμφότερα διαπορῆσαι), i.e. debating a claim and its contradictory.17 To do so, the interlocutors may have to use false premises, since a claim and its contradictory cannot both be true (to establish a false claim, the interlocutors must use at least one false premise). In dialectical exercise and testing, however, false premises are allowed in order not to restrict the argumentation (exercise) or in order to establish a true premise through a false (testing) (Top. 8.11.161a24–33). Thus, dialectical exercise may also serve the purpose of making the interlocutors more apt at going through ἀπορίαι, since it produces skill at handling false premises. The philosophical point of bringing a respondent to ἀπορία seems to be that ἀπορία forms a necessary condition for searching after truth.
As to the second philosophical use of dialectic, the main problem has been to understand the role of dialectic in regard to the first principles. It is neither possible here, nor relevant to this introduction, to go into this complicated problem at length. Because dialectic concerns itself with examination (it is ἐξεταστική), it has in some way access to first principles (Top. 1.2.101b2–4). One current way to interpret this is to take ‘examination’ (ἐξεταστική) as synonymous with ‘testing’ (πειραστική). This would mean that the access to first principles is somehow connected to peirastic dialectic (though it is far from clear how exactly this is so).18 At any rate, the presence in Aristotle’s dialectic of these two philosophical uses seems to indicate that the dialectical examination for which Socrates was brought on trial may actually lead from the realm of mere belief (δόξα) to real knowledge (ἐπιστήμη).
Dialogue
According to Rudolf Hirzel, whose study of the development of the dialogue form is still the most thorough, Socrates caused a revolution in language as much as in philosophy.19 What Hirzel meant by this was that the Socratic writers made everyday Attic into the leading medium for cultivated communication. But it seems to be true in another sense as well. Socrates appears to have caused a new literary form to be generated: the Socratic dialogues (οἱ σωκρατικοὶ λόγοι), literally ‘the Socratic writings’.20 The dialogue form, in other words, is intrinsically connected to Socratic dialectic and for this reason studies into the form and content of the dialogue should be included in a volume on the development of dialectic.
Whereas the scholarly literature concerning Plato’s use of the dialogue form is too extensive to be treated here, it is striking how relatively little scholarly attention Aristotle’s dialogues have attracted.21 Neglect is not responsible for this, but rather the fact that desperately little of these works remains extant. Recently, however, a translation of the fragments with introductions and commentaries has given a more accessible basis for the study of the Aristotelian dialogue than the somewhat bulky Greek editions offer.22 It is, unfortunately, difficult to establish a clear criterion for what may count as an influence from spoken dialectic on written philosophical texts. A writer of philosophy may be taken to be ‘dialogical’ if he or she in some sense enters into dialogue with some philosophical predecessor, i.e. either ‘answers’ some question or comments on a thesis of a predecessor. On this broad definition of ‘dialogical’, presumably every philosopher could be called ‘dialogical’.23 In the present investigations, however, ‘dialogical’ refers more narrowly to one of the above mentioned aspects of Socratic dialectic, (1–7) above. This provides a useful guideline for the investigation of dialectical elements in written philosophical texts of Plato and Aristotle.
Development
‘Development’ is a term that will trigger much suspicion among a considerable number of scholars, Platonic as well as Aristotelian. In the present volume, it means nothing more than a diachronic or comparative perspective on the dialectic of Plato and Aristotle. By taking ‘development’ in this non-evaluative sense, the present volume operates with a rather flexible or non-committed concept of development. Contrary to the previous investigations of Owen, De Pater and Stenzel, the present volume does not subscribe to one definite scheme of development in either Plato or Aristotle. The danger of developmentalism in the manner of Jaeger is that it easily imposes a very definite doctrinal scheme that tolerates no opposed directions or contradictions in a given stage of development (developmentalism is supposed to explain why there are contradictions at different stages, but cannot accept contradictions in the same stages, which is why there is a tendency in developmental studies to introduce ever more stages). A purely diachronic concept of development does not face the problem of opposed directions or contradictions, since on some issues development may show considerable agreement in the transition from one thinker to the other, and on different issues it may reveal severe disagreement.
Recently, scholars have presented a third way of interpreting the development, or lack of development, from Plato to Aristotle. More specifically, Neo-Platonist interpretations have been advanced in harmonizing or anti-developmental efforts.24 Whatever the merits of ‘Neo-Platonizing’ Plato and Aristotle might be, such an interpretive framework cannot help the present investigations for the simple reason that the Neo-Platonist commentators seem to take no interest in question-and-answer dialectic and apparently read Plato without taking notice of Socrates.25 Such an interpretive frame would make The Development of Dialectic from Plato to Aristotle an impossible task. In order to trace aspects of the development of dialectic, it is advisable to keep Socrates in mind (Remembering Socrates, as a recent volume encourages us to do)26 instead of pretending that he left no mark on the philosophical thought of Plato and Aristotle. Nevertheless, it should be noted that the Socrates kept in mind here is not the historical one. A regrettable consequence of this is that the so-called minor Socratics and their portrayals of Socrates receive almost no attention in the present volume.27
Structure and content of the volume
The chapters are ordered thematically into three groups that reflect the concept of dialectic outlined above: (1) Dialectic as an inter-personal activity; (2) Form and content in the philosophical dialogue; and (3) Dialectical methodology. Complete consistency in each group is impossible and some overlaps are inevitable.
To cover all aspects of dialectic and its development would be a very difficult task and to do so has not been the aim of this volume. What readers will probably miss most is a chapter on Plato’s dialectic in the middle dialogues (Republic and Phaedo in particular). This has not been left out due to a dislike of the method of hypothesis or of the theory of ideas, but reflects, rather, that this issue – like that of the Socratic elenchus – has received its due share of attention elsewhere.28 The chapters cover themes which are less well investigated (but no less central) or such that have not been considered at all previously. It will, further, be noted that the individual chapters may not be in complete agreement on some issues of interpretation. It would indeed be remarkable to find two scholars of ancient philosophy in complete agreement, let alone twelve as in the present volume. I have not sought to impose one and the same interpretation unto all the authors, but have instead welcomed different views as an advantage for the treatment of the subject of this book. Thus, the individual chapters should not be read as a continuous narrative. Each chapter sheds light on one or more aspects of the development of dialectic either by directly comparing Plato and Aristotle or by treating a central issue of dialectic in the writings of one or the other. The picture which emerges from the book gives, I hope, a fair representation of dialectic and its development from one great philosopher to another. The summaries below offer an outline of the content of each chapter.
Part I: Dialectic as inter-personal activity
Luca Castagnoli: Self-refutation and dialectic in Plato and Aristotle
It is a common assumption among present-day scholars that arguments put forward in a dialogical or dialectical context may be stripped of their dialectical clothing and formalized in accordance with twentieth-century logic without any logical loss. Castagnoli treats four arguments of self-refutation (two from Plato and two from Aristotle) and shows that none of them may be stripped of their dialectical context without ruinous logical results. An account of the logic of self-refutation arguments must take into consideration that these arguments depend upon the presence of both a questioner and a respondent (they are thus eminent examples of πρὸς ἕτερον-arguments). To strip them of this dialectical dimension and interpret them ‘monologically’ by use of the so-called Consequentia Mirabilis is not only historically incorrect but also a misrepresentation of the logic of these very self-refutation arguments. As to the development of such arguments, there is a remarkable continuity between Plato and Aristotle here, despite what seems to be an Aristotelian depreciation of Plato in this regard.
Marja-Liisa Kakkuri-Knuuttila: The role of the respondent in Plato and Aristotle
In his Topics, Aristotle claims to be the first to try to establish the rules for the respondent in a dialectical debate. This chapter investigates the relevance of this in regard to the problem of the epistemic status and achievement of the Socratic elenchus, and argues that the sophistication of Aristotle’s concept of elenchus primarily hinges on the rules for the respondent established in the Topics. This does not mean that Plato was unaware of the importance of the part played by the respondent. By investigating the Gorgias and Euthydemus, Kakkuri-Knuuttila shows that Plato clearly saw the limitations of having a respondent who may only answer ‘yes’ or ‘no’ to a dialectical question. However, Plato leaves the task of formulating new norms for the respondent to Aristotle. The rules laid down in Topics 8 endow the respondent with the primary responsibility of ensuring the quality of the dialectical argument. As such, the respondent attains an unprecedented importance for the epistemic achievement of the elenchus. But, as Kakkuri-Knuuttila demonstrates, the two most effective rules for the respondent were anticipated by Plato.
Hallvard Fossheim: Division as a method in Plato
Division (διαίρεσις), as depicted in Plato’s later dialogues, has traditionally been understood as a dialectical method designed to yield new knowledge. Against this reading, Fossheim argues that διαίρεσις is a method for presenting and organizing knowledge that has already been obtained. Thus, division belongs to the inter-personal dialectic of cooperative instruction. Fossheim argues that διαίρεσις depicts a new, cooperative ideal of dialectic and points out a tendency in Plato to move steadily away from competitive dialectic towards cooperative dialectic (this is a route from elenchus, over hypothesis, to διαίρεσις). Further, it is argued that this ideal of dialectic may be applied to the state of the soul of the interlocutors. The stages of dialectical argument (elenchus, hypothesis, διαίρεσις) come to represent stages in the development of the human soul.
Part II: Form and content in the philosophical dialogue
Morten S. Thaning: Dialectic and dialogue in the Lysis
In the Phaedrus, Plato envisages dialectic under two aspects as the ability to address different souls in different manners, and as performing divisions and collections. This chapter interprets these two aspects of dialectic as directed at two different kinds of audience and takes Lysis as a paradigm dialogue in this regard. The aspect of how to address different souls in different ways is brought out by the confrontation of three forms of protreptic speech in the Lysis: the encomium, the sophistic ἐπίδειξις, and the dialectic of Socrates. Dialectic as the performance of division and collection is illustrated by the investigations into friendship in the Lysis. Thaning argues that the five attempts at defining friendship constitute five pieces, as it were, in a dialectical puzzle (ἀπορία). The ‘solution’, i.e. λύσις, to this ἀπορία lies in the ability to perform divisions (of friendship) and a synoptical collection of them into a single definition. However, it is impossible to do so from the Lysis alone. To solve the ἀπορία, one needs further Platonic notions (mainly from the Symposium and Republic). These two observations about the Lysis make Thaning conclude that the dialogue was intended for two distinct audiences: (1) the novice, who should be led to engage in philosophy, and (2) the fully trained dialectician, who should use the dialogue as an exercise in division and collection.
Holger Thesleff: The Laches and ‘joint search’ dialectic
Many of the anomalies in the dialogue Laches have been observed and described, but until now there has been given no satisfactory explanation of them. Thesleff undertakes to explain these by focusing on the concept of dialectic implicit in the Laches. The author points out that the Socratic elenchus, or destructive elenchus, is not the predominant form of dialectical argumentation in this dialogue. Rather, Socrates applies constructive argument in large parts of the dialogue, emphasizing repeatedly that dialectical argument is a joint search or cooperative effort. By combining philosophical, literary, and chronological analysis, Thesleff demonstrates that the Laches forms a turning point in Plato’s concept of dialectic from Socratic elenchus to ‘joint search’ dialectic.
Charles H. Kahn: The philosophical importance of the dialogue form for Plato
Kahn deals with a classic issue in Plato scholarship, i.e. Plato’s use of the dialogue form for philosophical communication. Kahn’s argument is based on the introduction of two notions: (1) Plato’s ‘perspectivism’, i.e. the observation that Plato seems to avoid absolute formulations of philosophical theories and tends to approach and describe the same philosophical point from various perspectives in various contexts, and (2) the presence of ‘schemata’ in Plato’s philosophy. A schema is not an explicit philosophical theory but a core of philosophical insight – a focal point – which can be described in a great variety of ways but which does not allow for a final formulation. In combination, 1 and 2 explain why Plato uses the dialogue form: it is ideally suited to describe, from various angles (perspectivism), the same philosophical insights (schemata) in different contexts.
Jakob L. Fink: How did Aristotle read a Platonic dialogue?
Aristotle’s approach to Plato seems to be completely indifferent to the dialogue form. This has led some to claim that he took this literary form as little more than a dispensable vehicle for the conveyance of philosophical theories. This chapter undertakes an investigation of what Aristotle did say about the dialogue form (οἱ σωκρατικοὶ λόγοι) and its relevance for the transmission of philosophical knowledge. Despite the current assumption to the contrary, Aristotle does in fact make a number of significant claims, which are collected and used for a reconstruction of his interpretation of Plato’s dialogue form. The main points are his recognition of the poetical nature of the dialogues and his focus on the depiction of character (ἦθος) in these. It is argued that Aristotle interprets the dialogues as ‘paradigms’ for imitation and, as such, as protreptic texts. The chapter concludes by investigating Aristotle’s own fragmentary dialogues through a comparison of the dialectical elements in these and in the Aristotelian lecture. It is pointed out that both use a predominantly ‘monological’ type of dialectic but that the lecture is by far the more Socratic of the two.
Part III: Dialectical methodology
Vasilis Politis: What is behind the ti esti question?
Socrates is famous for his search after definitions and his focus on the ‘What is Φ’ question (τί ἐστι). The demand for a definition and the subsequent test of it stand at the root of dialectical argumentation. This definitional procedure has earned Plato both the honorific title of being the first to point out the priority of definition in philosophical inquiry as well as the accusation of committing or depicting the ‘Socratic Fallacy’ (levelled against him by Geach). What has, until now, never been asked is what motivates and justifies the demand for definition itself. This is what Politis takes up in this chapter. As can be seen from some of Plato’s dialogues, a definition must satisfy a generality requirement, a unity requirement and a requirement to be explanatory. The justification for these demands on definition can be found in a certain type of ἀπορία, as Politis demonstrates by focusing on the generality requirement. The upshot is that, in the order of inquiry, definition loses its position of priority to ἀπορία. It is a certain type of intellectual bewilderment (ἀπορία) that motivates and justifies the τί ἐστι question.
Hayden W. Ausland: Socratic induction in Plato and Aristotle
Whereas the logic of deduction (συλλογισμός) received a fully worked-out theoretical interpretation in Antiquity (in Aristotle’s Prior Analytics), the same is not true of its counterpart, induction (ἐπαγωγή). In both Plato and Aristotle, ἐπαγωγή is associated with Socrates and Socratic dialectic, but in two different ways. Through Aristotle’s ‘history’ of philosophy, Socratic ἐπαγωγή was given a narrow logical interpretation, whereas Plato and the so-called minor Socratics testify to a more rhetorical use of ἐπαγωγή. The chapter traces the development of inductive argumentation in Xenophon, Plato, and Aristotle and considers the later Peripatetic and Neo-Platonist interpretations of ἐπαγωγή as a striking contrast. Ausland argues that the Peripatetic approach is based on a too narrow ‘logical’ reading of Aristotle’s remarks about Socrates. Not only does Aristotle himself speak about ἐπαγωγή in a variety of senses, he also employs ‘inductive’ arguments which preserve a much broader understanding of this phenomenon than his account of Socratic ἐπαγωγή would suggest. This tension in the interpretation of Socratic induction challenges the claim that induction is exclusively or primarily a logical matter. There are distinct traces of rhetorical influence in the development of this particular dialectical form of argument.
Louis-André Dorion: Aristotle’s definition of elenchus in the light of Plato’s Sophist
The elenchus has formed the centre of attention in a large number of studies into dialectic, but a meticulous comparison of elenchus in Plato and Aristotle has not been undertaken yet. This chapter is devoted to such a comparison. The Sophist provides the only explicit account of the constituents of an elenchus in Plato. The relevant passage makes clear that Plato knew very well how to formulate the validity requirements operative within a genuine elenchus. But more importantly, the same passage shows that the elenchus serves primarily a moral purpose to which logic is a subordinate matter. Dorion points out that Aristotle’s definition of elenchus reflects (to a much higher degree than has previously been recognized) the logical part of Plato’s account. However, the moral dimension is completely absent from the Aristotelian elenchus. This absence should be explained by the fact that Aristotle rejects the Socratic doctrine of virtue as knowledge. The outcome in terms of the development of dialectic is a de-personalized and ‘de-moralized’ concept of elenchus in Aristotle’s dialectic.
Robert Bolton: the Aristotelian elenchus
Aristotle’s Sophistical Refutations is a rather neglected member of the Corpus Aristotelicum. The sophistic or apparent elenchus, nevertheless, is a vital source for understanding Aristotle’s interpretation of a genuine elenchus, since the sophistic elenchus is defined by its aberration from the genuine. This chapter, thus, takes the Sophistical Refutations as the key to the Aristotelian elenchus. Bolton challenges the interpretation of the elenchus as an argument directed exclusively at the exposure of inconsistencies. The Aristotelian elenchus aims to establish a proposition as false. This is clear from Aristotle’s definition of it as a demanding form of inference, i.e. a deduction (συλλογισμός) of the contradictory of a respondent’s thesis. By such a deduction of a contradictory of a false claim, the elenchus may in fact establish other propositions as true. This genuine elenchus is the task of peirastic dialectic. The upshot is that the peirastic dialectician may establish truth through the ignorance of the respondent concerning a specific subject and through the so-called ‘common things’ (τὰ κοινά), here interpreted as things that work in argument with many people.
Wolfgang Kullmann: Aristotle’s gradual turn from dialectic
This chapter investigates Aristotle’s conception and use of dialectic through the whole of the Corpus Aristotelicum. As early as the Topics, Aristotle limits the role of dialectic in the philosophical sciences. The Platonic conception of dialectic as a purely argumentative procedure starting at the level of opinion and leading from here to an ultimate explanation of the world is, to a certain extent, still present in the Topics; but the primary role of dialectic is no longer to provide or apprehend principles. Rather, dialectic serves the purpose of intellectual training and public encounters (and only in the third place the apprehension of principles). Kullmann interprets the subsequent development in Aristotle’s dialectic as a gradual turn away from this third use of dialectic in philosophy. A rival to the method of dialectic is present in Prior and Posterior Analytics, which offer a new procedure for apprehending principles by means of sense-perception, induction and experience. The turn from dialectic, however, is gradual and finds its fullest expression in the biological writings which conform to the new methodology of perception and experience presented in the Analytics. Kullmann interprets this development in Aristotle’s conception of dialectic to the effect that the true heirs of Plato’s dialectic, conceived as a science directed at explaining the world, are the individual philosophical sciences as found in Aristotle’s writings.
* I wish to thank Luca Castagnoli, Sten Ebbesen and the readers of the Press for their comments on drafts of the introduction.
1 HambruchReference Hambruch1904, KappReference Kapp1942, SichirolloReference Sichirollo1966, and the relevant articles in BertiReference Berti2008 should also be mentioned. NarcyReference Narcy and Cordero2000 and DixsautReference Dixsaut and Tsimbidaros2004 are both firmly rooted within the horizon of Reference OwenOwen, Reference PaterDe Pater and Reference StenzelStenzel.
2 Robinson Reference Robinson1953: 62.
3 For the Eleatic Zeno as inventor (or discoverer) of dialectic, see Aristotle’s Sophist (Fr. 65 R3 = DL 8.57).For the rival candidate Protagoras, see DL 9.53. WilpertReference Wilpert1956/57 has tried to sort out Aristotle’s account of the origin and development of dialectic. Roughly speaking, the picture is this: Zeno of Elea ‘invents’ dialectic; Socrates, Plato and the Academy bring it forward; and Aristotle completes it. For reservations as to Zeno’s role, see Dorion Reference Dorion, Canto-Sperber and Pellegrin2002: 200–8.
4 What follows is largely inspired by the accounts given by Brunschwig Reference Brunschwig1967: xxix–xlv, MorauxReference Moraux and Owen1968: 277–90, SlomkowskiReference Slomkowski1997: 9–42, Rapp & WagnerReference Rapp and Wagner2004: 11–18. The idea that Aristotle’s dialectic is a useful guide to Plato’s dialectic has been elaborated among others by FredeReference Frede, Klagge and Smith1992 and, in particular, BoltonReference Bolton1993.
5 The absence of words does not mean that these dialectical manoeuvres were unfamiliar to Plato. For inductive arguments in particular, see RobinsonReference Robinson1953: 45–6.
6 The logic of the dialectical συλλογισμός is interpreted (in connection with the interpretation of the τόπος) either as a modus ponendo ponens or tollendo tollens, BrunschwigReference Brunschwig1967: xli, i.e. as a hypothetical syllogism, or as a categorical syllogism, SchrammReference Schramm2004: 133. For various positions on the Socratic elenchus, see the articles in Scott Reference Scott2002.
7 This metaphor is taken from a Danish commentary on Plato’s Smp. See OlsenReference Olsen2003: 10.
8 MannReference Mann, Rapp and Wagner2006: 116–17 with note 18 argues (summarily) against the claim that the ‘say-what-you-mean’ requirement is constitutive in Socrates’ way of arguing. For the more traditional view, see VlastosReference Vlastos1983: 38.
9 There are connections between the Socratic elenchus and the rules for definition operative in Aristotle’s τόποι (De PaterReference Pater1965: 19–20) and instances of τόποι in Plato’s dialogues (De Pater Reference Pater1965: 67 and Hambruch Reference Hambruch1904).
10 It is very difficult even to state clearly what a τόπος is. For a short account with references, see PrimavesiReference Primavesi, Ritter and Gründer1998.
11 The exceptions being Beversluis Reference Beversluis2000 for Plato and Slomkowski Reference Slomkowski1997: 36–42 for Aristotle.
12 The Topics has a large number of τόποι to be used against ‘those who posit the ideas’ (e.g. 2.7.113a24–32, 3.6.120b3–6, 4.2.122b25–35, 6.3.141a5–9, 6.6.143b23–4, 6.8.147a5–11) and at least one τόπος useful for establishing ideas (5.7.137b8–13). See the comments ad loc. of BrunschwigReference Brunschwig2007: 192 for the textual difficulties of this passage. See SE 12.172b29–31 for a reference to debates with other ‘schools’.
13 FaitReference Fait2007: xxv has a good discussion with further references and examples of how the audience might supply premises which were not actually conceded by the respondent.
14 See Kakkuri-KnuuttilaReference Kakkuri-Knuuttila2005: 65–9 and SmithReference Smith1997: 141–6 for further details.
15 The ‘gymnastical’ aspect of Aristotle’s dialectic was somewhat neglected in the scholarly literature until PrimavesiReference Primavesi1996 argued that it is the central purpose of Aristotelian dialectic. For an even sharper formulation of his position, see Primavesi Reference Primavesi, Hempfer and Traninger2010.
16 ἔντευξις derives from ἐντυγχάνειν ‘to chance upon’. In Plato, dialectical arguments very often result from some chance meeting (e.g. Euthd. 272c, Lys. 203a, Tht. 142c), or a casual encounter leads to the narration of a dialogue (e.g. Prm. 126a, Smp. 172a, Tht. 142a). Other dialogues seem to be casual even if τυγχάνειν (in some form) is not used expressis verbis, e.g. Chrm., Euphr., Ion, Menex., Phdr. Even Rep. seems to result from a casual encounter. Aristotle may have had this aspect in mind in coining the term ἔντευξις. See also De Pater Reference Pater1965: 81–2.
17 This use of dialectic seems to be inspired to some extent by Plato’s Parmenides. See Brunschwig ad. loc. Reference Brunschwig1967: 116 and Reference Brunschwig2007: 302 (ad 163b9–16).
18 BoltonReference Bolton, Devereux and Pellegrin1990 is the main proponent of this interpretation. Opponents of this take dialectic as offering a critical test of first principles. See SimReference Sim and Sim1999 for more details on both positions and the relevant Aristotelian texts.
19 HirzelReference Hirzel1895: 93–4.
20 Since Schleiermacher, the dialogue has been interpreted as the literary form which most truly imitates the oral philosophical activity of Socrates; see Schleiermacher Reference Schleiermacher1804: 18–19. Aristotle (Poet. 1.1447b11) is the first known reference to the dialogue form. It is disputed what the literary ancestor of the dialogue is. There are three candidates: tragedy, PlegerReference Pleger and Meyer2006; comedy, ClayReference Clay and Vander Waerdt1994; and finally the sophistic protreptical speech, Gaiser Reference Gaiser1959.
21 GillReference Gill, Annas and Rowe2002 surveys the main modern approaches to Plato and the dialogue form with a useful bibliography. SzlezákReference Szlezák2004 and RoweReference Rowe2007 are some important recent contributions.
22 Flasharet al. Reference Flashar, Dubielzig and Breitenbach2006. The edition of GigonReference Gigon1987 is the most complete, but also the most inaccessible.
23 One of the most recent investigations into the development of the dialogue seems to me to suffer from this broad definition of ‘dialogical’; see MeyerReference Meyer and Meyer2006: 10.
24 GersonReference Gerson2005, Karamanolis Reference Karamanolis2006.
25 Gerson Reference Gerson2005: 29. But see SedleyReference Sedley, Annas and Rowe2002, another plaidoyer for an ancient approach, for some modification.
26 Judson & Karasmanis Reference Judson and Karasmanis2006.
27 An outline of these portraits and their interrelation with one another and Plato has been given by KahnReference Kahn1996: 1–35. VlastosReference Vlastos1991 is a brilliant account of the historical Socrates.
28 See BaileyReference Bailey2006, DixsautReference Dixsaut2001chap. 2, KarasmanisReference Karasmanis and Tsimbidaros2004, MittelstrassReference Mittelstrass and Höffe1996 to mention just a few.
