This page intentionally left blank

J. D I G G L E N. H O P K I N S O N J. G. F. P O W E L L
M. D. R E E V E D. N. S E D L E Y R. J. T A R R A N T



Professor of Greek Language and Literature,
Vrije Universiteit, Amsterdam


         
The Pitt Building, Trumpington Street, Cambridge, United Kingdom
  
The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa
© Cambridge University Press 2004
First published in printed format 1999
ISBN 0-511-00410-9 eBook (netLibrary)
ISBN 0-521-62368-5 hardback

ou deiÁ v gaÁr hmwÄ n e autwÄ i zhÄ i
kaiÁ oudeiÁ v e autwÄ i apoqnh iskei.
e a n te gaÁr zwÄmen
twÄ i kuri wi zwÄmen´
e a n te a poqnh iskwmen
twÄ i kuri wi a poqnh iskomen´
e an te ou n zwÄ men e aÂn te apoqnhÂiskwmen,
touÄ kuri ou e sme n.


7 Date and authenticity 59 93 127 164 180 209 215 TEXT AND TRANSLATION 235 COMMENTARY 261 APPENDICES I The ending of Aristotle's Protrepticus II Note on the text 336 340 vii 1 1 1 5 10 18 35 58 .1 Philosophical protreptic in the fourth century bce ii.3 Is the Clitophon un®nished? i.2 Summary and analysis of composition i.4 The Clitophon as a Short Dialogue i.1 Introduction i.5 The characters of the dialogue II Meaning and authenticity ii.6 The meaning of the Clitophon ii.3 Protreptic in Plato ii.5 Justice in the Clitophon ii.CONTENTS Preface Abbreviations page ix xiii INTRODUCTION I Prolegomena to the dialogue i.2 Protreptic in the Clitophon ii.4 Elenchos in the Clitophon ii.

CONTENTS BIBLIOGRAPHY 345 INDEXES 1 Literary and philosophical index 2 Grammatical index 3 Index of Greek words 4 Index of passages cited 352 352 353 353 355 viii .

ix . I am very grateful that the board of the Fondation Hardt pour l'e tude de l'antiquite classique allowed me a four-week sojourn in the summer of 1997. the publication of this edition would have taken up much more time and required much more labour. who very kindly suggested to the editors of the series `Cambridge Classical Texts and Commentaries' that they should take a revised version of the book into consideration. Professor Jaap Mansfeld of the University of Utrecht. At the time.PREFACE An earlier version of this book was a privately published doctoral dissertation of the Free University at Amsterdam (Academische Pers. If it had not been for her unequalled competence and precision. It was not until 1995 that I had the opportunity of studying protreptic once again (cf. My aim was to give the Clitophon a more secure position within the development of ancient philosophical protreptic. For this reason. Ms Josselijn Boessenkool. consented to undertake this laborious task. other duties prevented me from realising this project. However. my paper `Protreptic in ancient theories of philosophical literature'). likewise to the editors of CCTC for accepting it. I intended to produce a less provisional edition of it as soon as possible. The board of the Faculty of Letters of the Free University at Amsterdam ®nanced the computerising of the book. I owe a lasting debt of gratitude to my friend and colleague. which permitted me to lay a solid foundation to the book as it presently is. By then. I feel particularly privileged that this is the ®rst commentary on a Platonic text to appear in this distinguished series. 1981). It has been a particular stroke of luck that my pupil. it was clear to me that the most I could hope for was an opportunity for publishing a revised edition of the 1981 thesis.

Moore. I have greatly pro®ted from a number of conversations with Professor Sedley. I no longer have any strong doubts about the authenticity of this work. and I have no compunction in presenting this revised version as x . Susan P.3 of the Introduction. I am grateful to two scholars of the Free University. of course. Professor James Diggle sent me a number of highly salutary notes. During the ®nal stages. Professor D. My former colleague Dr Pauline Allen corrected the English of the ®rst version of this book. which have stimulated me to think some of the problems through once again. van Dijk have been extremely helpful in drawing my attention to various mistakes and inconsistencies. Schenkeveld (emeritus) supervised my thesis. Professor Sedley has suggested numerous improvements for this one. why they want to do it. The late G. The ®rst version of this book appeared under the title A Commentary on the Platonic Clitophon. I am. responsible for such blemishes as remain. As will be clear from section ii. de Vries taught me most of what little I understand about Plato and Platonic Greek.7. and how they do it.PREFACE For the same reason. I thank Professor David Sedley. Pauline Hire. who very graciously enabled me to work at the Cambridge University Library for a couple of days in October 1997. and his constant insistence on his promovendi making clear what they want to do. M. J. it should certainly be dedicated to him. M. But above all. most of the credit should go to his acute observations. If this book had not already been dedicated to the memory of someone else. I'm very proud to be his successor in the chair of Greek in the University which we have both been honoured to serve for most of our lives. G. and if the exposition of my views throughout this book is even passably understandable. even though I still cannot shake o¨ all my misgivings. Caroline Murray and my colleague Dr J. But I now feel that the grounds for my doubts are rather weak.

I have decided to suppress this ± a highly condensed version is here presented as `Appendix ii '. even though only a tiny minority of them have. for which we now have the brilliant monograph of my pupil. I am particularly grateful to Professor Sedley. Boter (The Textual Tradition of Plato's Republic). who was instrumental in bringing about this change of mind. But no matter if this dialogue is authentic or not. unfortunately. which obviously found its way into most of the major academic libraries throughout the world. the Clitophon was being described as a `jewel' ± admittedly in my hearing. I suppose. J. however. My dissertation contained a separate chapter on the MSS of the Clitophon. not from these MSS themselves. the most important question is its literary and philosophical intention ± while rereading what I wrote seventeen years ago. At the time. when I compare the Urtext of this book with the vast amount of work on Plato published during the nineties. A short time ago. I did not have photographs of all MSS at my disposal (see my paper `Supplementary notes'). bothered to read what was after all a minor book on a minor and suspect dialogue. I have felt that my interpretation of it did not require major changes.PREFACE Plato: Clitophon. Dr G. But a rehash of my work as published in 1981 would be pointless in 1999. It is very gratifying that my thesis should have started a spate of fundamental studies on the MS transmission of Platonic dialogues ± nearly all of these studies are based on the ®rst version of the present book. I have. This re-examination has prompted me to change the apparatus in quite a few places. I am just as xi . and by someone who had read the ®rst version of my commentary. I feel sustained by the fact that most of my conclusions are shared by many excellent Platonists today. The textual tradition of the Clitophon is hardly di¨erent from that of the Republic. Indeed. made a fresh collation of ADF ± from photographs.

xii . that it will be taken as a serious and in some aspects unique representative of fourth-century philosophical literature. Yet I hope that this book will contribute to its being considered worthy of being inserted in the crown ± if not. S. R. Amsterdam S.PREFACE aware as anyone that it is not an un¯awed jewel.

con introduzione e commento (Naples 1935) G. Platonis quae exstant opera . D. . in linguam Latinam convertit . ix (Lipsiae 1827). 459±64. . Gonzalez. Zweiter Teil: Satzlehre (Hannover±Leipzig 1898±1904) H. Plato. M. K. Hermann. S. Platone. recensuit. Bury. `Saggio sul Clitofonte'. le Opere tradotte e dixiii . `Clitophon'. Clitofonte. Platonis dialogi iii (Lipsiae 1851). C. M. 465±74 (with Ficinus' translation) G. . Bertini. Plato. Ast. 311±27 see Bekker B. la Magna. A History of Greek Philosophy (Cambridge 1962±78) R. Bekker. Gerth. Modugno. 354±65 I. G. Gildersleeve. Stuart Jones. G. in: J. Denniston. Liddell±R. Platonis opera iv. 966±70 J. Syntax of Classical Greek (New York 1900±11) F. A Greek±English Lexicon (Oxford 1940 9 ) G. Guthrie. Oxonii 1902 R. Cooper±D. RFIC 1 (1873) 457±80 J. KuÈ hner±B. Platonis dialogi Graece et Latine ii 3 (Berolini 1817). Scott±H. AusfuÈhrliche Grammatik der griechischen Sprache. vi (Lipsiae 1853). The Greek Particles (Oxford 19542 ) C. . with an English Translation (London±Cambridge (Mass. 330±1 (scholia) W.ABBREVIATIONS Ast Bekker Bertini Burnet Bury Ficinus Gildersleeve Gonzalez GP 2 Hermann HGPh KG LSJ la Magna Modugno F.) 1929). F. Burnet. Platone.A. Hutchinson. Complete Works (Indianapolis 1997). J. F.

W. Sartori. Platon. Pauly±G. The Roots of Political Philosophy. Wissowa et al. 90±5.±D. Dialoghi v (Bari 1956) 2±10 M. Platone. Platons Werke ii 3 (Berlin 1809). SouilheÂ. ix (Lipsiae 1885). Platonis opera . Orwin.ABBREVIATIONS MT MuÈ ller New Approaches Orwin PW Sartori Schanz Schleiermacher Schw. 102±3 (critical appendix) F. . Proceedings of the Colloquium Held in Amsterdam. edidit M. 111±16 A. Giannantoni. in: T. to Honour C. Stallbaum. L. . Syntax of the Moods and Tenses of the Greek Verb (London 1912 3 ) H. Platonis dialogi.S. RealEncyclopaÈ die der classischen Altertumswissenschaft (Stuttgart 1893±1978) F. Steinhart vii 1 (Leipzig 1859). 47±74 A. Platon's saÈ mmtliche Werke. J. Schleiermacher. Souilhe SSR Stallbaum chiarate ad uso di ogni persona colta xix (Aquila 1930). Goodwin.. Oeuvres compleÁtes xiii 2 (Paris 1930). 163±90 G. Schwyzer±A.. Socratis et Socraticorum reliquiae (Naples 1990±1) G. New Approaches to Greek Particles. 129±49 W. È bersetzt von H. mit Einleitungen begleitet U von K. 453±64 (translation). Schanz. Pangle (ed. textum ad xiv . II Syntax und syntaktische Stilistik (Munich 1950) J. 1996. MuÈller. Rijksbaron (ed). with Interpretive Studies (Ithaca±London 1987). Exhortation]'.. Debrunner. January 4±6. `Cleitophon [or. Griechische Grammatik. Ruijgh on the Occasion of his Retirement (Amsterdam 1997) C. 534 (notes) E.). M. Ten Forgotten Socratic Dialogues. Translated.

. Two Studies in Attic Particle Usage: Lysias and Plato (Leiden 1993) R.S. M. 227±34 (text). O. Republic (Oxford 1993). van Ophuijsen. in: id. . 496±9 (critical apparatus) see MuÈ ller F. M. de Win. 462±8 X. Zuretti. xii (Lipsiae 1825). 507±29 C. Sicking± J. recognovit G. Dialoghi v (Bari 1915). Plato. 495 (notes) C. `Cleitophon'. . Platon. Verzameld werk v (Antwerpen±Baarn 1978 2 ) 469±75 (translation). Susemihl. Water®eld. J.. . Werke in 40 BaÈndchen v 3±6 (Stuttgart 1865).ABBREVIATIONS Steinhart Susemihl Two Studies Water®eld de Win Zuretti ®dem Codicum . 3±21 xv . Plato. . viii (Lipsiae 1825). Platone.


The Clitophon is mainly an attack on Socrates. Is this the Athenian philosopher who inspired a great number of thinkers. what is the relation of these passages to Plato's literary production as a whole? The criticism is uttered by one Clitophon. called ProtreptikoÂv. In Book 1 of Plato's Republic 1 . and so it is. and ®nally to certain passages in Plato where this activity of Socrates' is described or hinted at? As a corollary. by pupils of the Athenian philosopher. was ridiculed by Aristophanes and other comedians and was eventually put to death. The Clitophon has often been dubbed a `riddle'. or is he the literary character who plays the central part in many fourthcentury philosophical texts of a genre called lo gov SwkratikoÂv from Aristotle onwards? This Socrates is said to be an expert in what is called protre pein (I shall translate this throughout the book by `exhort'.1) which is programmatic for the whole of the rest of that work. Its meaning is therefore a problem prior to (and more interesting than) its authorship. who we are told is at the same time rather enthusiastic about the teaching of Thrasymachus. to an interesting passage of Xenophon's Memorabilia (1. its scale as o¨ered here does call for an excuse. Its authorship is dubious ± a decision as to its authenticity would seem to depend mainly on the interpretation of its meaning. In this connection several questions come to one's mind. How does this statement relate to several works.4.1 Introduction Whereas a commentary on the Clitophon requires no justi®cation ± for there is none in either Latin or any of the three major European languages of our time.INTRODUCTION I PROLEGOMENA TO THE DIALOGUE i. for lack of a better equivalent).

occasionally. J. to answer questions for which I found no satisfactory answers in the text of the dialogue. Unless I have gravely deceived myself.I N T R O D U C T I O N I. or. I have therefore left aside the attribution of the dialogue to Plato. such other Socratic texts as I have deemed pro®table to take into account have been used either to test the hypotheses formed on the basis of the Clitophon alone. it is possible to explain the Clitophon from the Clitophon itself. the written work is called an eidoÂtwn upoÂmnhsiv in the Phaedrus. It has become a platitude to say that in cases of disputed authenticity the onus probandi lies with those who want to dispute it ± in fact. D. 149 (G. However. the remarks in Pseudepigrapha 1 (Entretiens Hardt 18 (1971)).1 this character appears as a companion and defender of Thrasymachus. which normally in literary analysis one is obliged to take into account. I do not think that this strictly `ergocentric' method is imperative in Plato ± on the contrary. where an exception is made for texts 2 . the authenticity of the Clitophon has been doubted by many eminent scholars from the early nineteenth century onwards. In general. To what extent is this signi®cant for the interpretation of Republic 1 and Clitophon? What further light is shed on this problem by the similarity of statements in the Clitophon about the result (e rgon) of justice to statements about justice in Republic 1? Clitophon gives an extensive report of his questioning Socrates' companions and refuting them. this is far from being a dogma1 ± but one should not add to the bur1 Cf. Aalders). 12 (R. so that in genuine dialogues a comparative method of interpretation seems to be called for. What is this method of interrogation and refutation. how close does it come to methods observed in other Socratic literature and what are the implications of the similarity for the intention and philosophical provenance of our dialogue? I have tried to answer these questions without any regard to the problem of authorship. Syme).

not its ethical values.4). 3 . Cf. `Zur AÈsthetik der Parodie'. 36±41). My use of the term is therefore much more traditional than that of some recent theoreticians. the arguments for my answers take up a fair amount of space and are scattered throughout the book. 139 speaks of `the disastrous principle ``presumed genuine until proved spurious'' '. Therefore I shall outline here. Pastiche (Munich 1977). W. W. belonging to a genre which as a whole is open to suspicion. to wit philosophical protreptic in its pre-Aristotelian. the questions raised above will be found treated there. The choice of Clitophon. it is the uselessness of protreptic preaching which is the target. but of a speci®c branch of Socratic literature. such sense as I can make of this dialogue. Though I am aware of its de®ciencies. Genres in Dialogue. 2 I have not tried to de®ne this term. without further argument. esp. Travestie. and any rate for Clit.). not of one particular writer. these remarks are relevant for the Clitophon. admirer of Thrasymachus. The Clitophon is essentially a condemnation not of Socrates. 6±8. Die Literatur 30 (1927±8) 439±41. The author is careful not to hit at the core of Socratic philosophy. it is a parody of thoughts. cf. The Bakhtin Reader (London 1994). Parodie. Plato and the Construct of Philosophy (Cambridge 1995). 102±22. Neumann. nor of another philosopher. ethical form. Morris (ed. as the main character suggests how dangerous protreptic can be. but gives no arguments for this somewhat extreme view. I think the following de®nition is satisfactory. ± K. for criticism. 148±9. in which various motifs of this genre are used.: `Parodie ist Nachahmung mit Polemik gegen den Nachgeahmten' (R.1 den by forming hypotheses based on related texts transmitted within the same Corpus. The speech put into Socrates' mouth is a parody2 (as Aspasia's speech in the Menexenus is generally supposed to be). Because I ®nd the questions di½cult and rather involved. If there is indeed such a genre as the Short Dialogue (section i. notably Bakhtin. Nightingale. Dover. Marginal Comment (London 1994). Bakhtin's in¯uence is notable in A. The Introduction and the Commentary together contain my interpretation of the Clitophon.I N T R O D U C T I O N I. P. Karrer.

to whom the Clitophon is ascribed. Moreover. which is. the author implies that he assents to an important aspect of Plato's concept of justice. I have endeavoured not only to elucidate questions connected with structure. In the Commentary. recommended as a better alternative for protreptic. the language of our dialogue is very similar to Plato's. on the whole.I N T R O D U C T I O N I. namely that the true politician is he who renders his fellow-citizens more just. In short. in order to show how these idioms relate to the usage of Plato. I saw no point in increasing the bulk of annotations by referring (more than occasionally) to parallels found in the works of other authors of this period. As I found that. I have adduced many parallels for words. leads to insight (and thereby to knowledge). phrases and constructions which in themselves needed no illustration. intention. The use of elenchos in the Clitophon is typically Platonic. not exclusively protreptic literature. not exhortation. Brandwood's Word Index (but also Ast's Lexicon) has been an invaluable support.1 Clitophon's interrogation of Socrates' companions and ± to a lesser extent ± of Socrates himself serves a double purpose: it proves that mere exhortation towards justice does not lead to knowledge of justice (various discussions of justice are taken over from Socratic literature. It goes without saying that apart from the TLG CD-ROM. 4 . but also to furnish material for settling the questions of authorship. The author's judgement on the respective e¨ectiveness of exhortation and elenchos is identical to Plato's standpoint. the author's intention is to show that his opinion of Socratic literature conforms in every respect to the views found in Plato's literary production. these borrowings are not meant to suggest that Socratic theories about justice are worthless). by implication. expression and textual transmission (in so far as these matters have not been treated systematically in the Introduction). at the same time it is shown that elenchos.

PROLOGUE (406a1±407a4) I. They and their children have followed the traditional curriculum. if such there be. CLITOPHON'S REPORT (407a5±410b3) (1) clitophon's praise (407a5±408c4) (a) Introductory words (407a5±b1) II. 407b1±e2) `Men do not act as they should. A. B. has been reintroduced because on the whole it does justice to the structure of the Clitophon. but in part I did do so. even though a commentary is not necessarily the best place for having one's say on such matters. ± That is not quite right. they do not ®nd them teachers of justice. but neglect to provide their sons. He o¨ers to expound his position. who will inherit it. hoping to bene®t from his words. in part I have indeed not praised you. Clitophon had criticised Socrates' intellectual guidance and praised that of Thrasymachus.I N T R O D U C T I O N I. ± Socrates gives him the opportunity. i. with the knowledge how to use it justly.2 In a limited number of cases. nor have they taken care of themselves similarly in the past. I was unable to refrain from treating questions of grammar and lexicon on a more general scale. because they focus all their attention on amassing wealth.2 Summary and analysis of composition The nineteenth-century division into chapters (Roman numerals). Socrates says someone told him that. in a conversation with Lysias. and 5 . which was abandoned in Burnet's edition. Clitophon answers. Clitophon says that he has been struck whenever Socrates delivered a certain speech like a deus ex machina: (b) Socrates' protreptic (407b1±408b5) ( ®rst part.

one should leave alone. one who cannot handle his own lyre will not be able to handle his neighbour's. Finally. (2) clitophon's criticism (408c4±410b3) (a) Introduction (408c4±409c1) Therefore he asked those companions whom Socrates esteemed most how Socrates' exhortation is to be followed up. they contradict themselves. supposing that exhortation itself is not the goal of life. Discord in the world stems from disharmony. If man is mastered by his pleasures. 407e3±4.2 they are none the less vicious in matters of money ± therefore present education is to be condemned. so with the senses and the whole body. one who does not know how to handle his soul had better leave it alone and cease to live. Interrupting his report.' (third part. not musical but spiritual. likewise. he o¨ers an analogy: one who had exhorted them to the care of the body would reproach them on the grounds that they care only for agrarian 6 . After Socrates' fashion. Clitophon quite agrees with this and similar speeches and considers them very suitable for exhortation and very useful.' These experts are identi®ed by Socrates with those who have learned politics. which is identical to judication and justice. 407e5±8) `Those who care only for their bodies and neglect their souls act likewise: they neglect the ruling part. he is so involuntarily. 407e8±408b5) `What one cannot handle. When men say that injustice is the consequence not of bad education but of a free choice. Consequently each individual and each state ought to care more in this respect than they do now. Clitophon again states his admiration.) (second part.' (III. or at any rate be a slave and hand over the rule of his mind to an expert.I N T R O D U C T I O N I. as they also think that injustice is hateful to the gods. (c) Concluding words (408b5±c4) IV.

justice has still failed to grasp the object of its knowledge. and likewise carpentry can be divided into doctrine and result. Clitophon wished to hear more than a name. Friendship was said by this man to be always a good. `the useful'. Medicine has a double e¨ect. Which art is it that improves the soul? V. who answered that the special result of justice was harming one's 7 . such as carpentry. but unlike the arts. too. as being often harmful. its result is yet unclear. `the ®tting'. (c) Second de®nition of the result of justice (409d2± 410a6) VI. What is the latter? (b) First de®nition of the result of justice (409c1± d2) This pupil answered `the bene®cial'. Finally the most elegant answer given was: to e¨ect friendship in the cities. others. let the result of justice be de®ned similarly. (d) Third de®nition of the result of justice (410a7± b3) VII. Then Clitophon asked Socrates himself. At this point those present were able to accuse him of circular reasoning: medicine.I N T R O D U C T I O N I. but the meaning of these epithets will be de®ned by the arts in question. Being asked whether concord was unanimity in opinion or knowledge he rejected the former. Similarly justice will on one hand produce new just men. the production of new doctors and health (of which the latter is a result of the art. `the pro®table'. not art itself ). Clitophon replies that all these epithets are also valid for the results of each of the arts. The man who seemed best equipped answered that this art is none other than justice.2 products instead of the arts which improve the body. on the other it must have a result of its own. so that the friendships of children and animals (which as a result of a debate he concluded were more often harmful than bene®cial) had to be excluded: real friendship was concord. is concord in this sense.

`II. He thinks that Socrates is still the best in exhorting others to justice but either he can do nothing more. then let it happen. partly blaming him. Was Kleitophon an Sokrates lobenswert ®ndet' (407a±408c). C. Was Kleitophon an Sokrates zu tadeln ®ndet' (408c±410b). 8 . but either Socrates does not know what justice is or he is unwilling to impart his knowledge to Clitophon. divided into `I. for those who have been. For the relation between content (as analysed here) and form. he were going to explain the nature of the body and the treatment pertaining to it. He implores Socrates to do this so that he can stop partly praising. 3±5: `Einleitung' (406a±407a). Kleitophon'.4. That is why Clitophon visits Thrasymachus and others: he is at a loss.2 enemies and bene®ting one's friends. he is almost a stumblingblock in their attainment of the core of virtue and becoming happy.I N T R O D U C T I O N I.2(5). `Hauptteil'. Among other attempts3 at schematisation of the structure 3 By far the most satisfactory is that of Pavlu (`Pseudopl. Clitophon agrees that the care of the soul is all-important and says he has uttered his criticism with this intention. `Schluss' (410b±e)). having exhorted Clitophon to the care of the body. (b) Last appeal (410c8±e5) If Socrates is prepared to stop exhorting him and act just as if. (c) Summing-up (410e5±8) Socrates is invaluable for those who have not been exhorted. cf. Subsequently it turned out that justice never harms anyone. section i. like a layman who can eulogise steersmanship ± this is not Clitophon's view. Clitophon has given up. CLITOPHON'S VERDICT (410b3±e8) (a) Criticism (410b3±c8) Having endured this a long time.

Blits. This disposition is said (437) to correspond exactly to the e rgon touÄ rh torov as de®ned by Theodectes of Phaselis: prooimiaÂsasqai proÁv eunoian.4 The individual traits of the Clitophon are manifest also in Theodectes. 203. constructed so as to make Clitophon appear an equitable critic. rationibus' ± this solution (if it deserves the name) is rightly rejected by Pavlu (5 n. Radermacher. `RaÈ tsel'. dihghÂsasqai proÁ v piqano thta. the ®rst part is about protreptic. 744).4 n. e pilogi sasqai proÁ v orghÁn h e leon. 436) deserves closer investigation because of the conclusions he draws from it. pistwÂsasqai proÁ v peiqwÂ. followed by `eine philosophierende EroÈ rterung. See section ii. and Theodectes was a friend of Aristotle and was in¯uenced by him. ± Ge¨cken's analysis of the dialogue as a judicial accusation was foreshadowed by BruÈnnecke. 272. 452±7). who besides distinguishes three protreptic speeches. 451±2. and is now `more than once sharply criticised'.3). `Socratic teaching'). Ge¨cken. 4) recognises two parts. as Kesters (KeÂrygmes.3. that by Ge¨cken (`RaÈtsel'. Souilhe (163±4) places a dichotomy at 408e2. The return to Socrates at 410a7 is explained `non ex veritatis sed ex artis quasi scaenicae. the ®rst dealing with Socrates. 1). Artium scriptores. which he denies to be ironical (section i. 437 n. the interrogation of Socrates' pupils is the narrative part. to be the prooemium. the Clitophon is a riddle. the second (from 408b5) with his socalled pupils. the Clitophon is unmistakably an Aristotelising text. die den Satz von der Nichtigkeit der blossen Protreptik endguÈltig beweisen soll' ± I am not quite sure whether 408d1±6 or 410b6±c2 is meant. qua in dialogo opus est.2 of the Clitophon. 39±44) after him. the second about justice. This idea was taken over by Orwin: `we might regard this dialogue as a kind of counterApolog y' (`Case against Socrates'. ®nally Socrates is addressed directly for the second time (from 410a7 onwards?). cf. and This schema is taken over by BruÈ nnecke (`Kleitophon wider Sokrates'. Kunert (Necessitudo. on the problems concerning the versions and ascription of this fragment. 1.5. Finally.I N T R O D U C T I O N I. 4 Oratores Attici ii 247 Sauppe. According to Ge¨cken. who makes Socrates the accuser and Clitophon the defendant in a ®ctitious slander suit (`Kleitophon wider Sokrates'. 9 . cf. He considers Clitophon's report of Socrates' speech.

.5 This Platonist Ptolemy. so that Clitophon was in fact one of the persons to whom Socrates reported the Republic). who is mentioned also by Iamblichus.8±9 Diehl. 13±14): the Republic is in given by Yxem (`U fact Socrates' answer (on the premise that Ptolemy must have regarded the eighth tetralogy as a whole. FestugieÁ re. 1.378 W. 48 n. apparently Clitophon was thought to have stayed away through pique (slightly di¨erent A. le gein] pantelwÄ v atopon´ parhÄ n gaÁr ou deÁ thÄi proterai ai Swkra touv dihgoume nou ti na eipen o KleitojwÄn (namely in the Republic. in Tim. I am unable to make sense of Ge¨cken's analysis of the Clitophon.I N T R O D U C T I O N I. iii 218 and n.49. i. 6).20.18±20 Diehl.3. An ingenious distortion of Ptolemy's view is È ber Platon's Kleitophon'. I would suggest A as prooemium. ReÂveÂlation.-J.6 again in connection with 5 Procl.3. ibid. 10 . Commentaire sur le TimeÂe I (Paris 1966). B as a very lengthy narrative. 1. Quite apart from the dubious quality of the remaining arguments. It is attributed by Proclus to PtolemaiÄ ov o PlatwnikoÂv.1 Historical Survey Socrates' silence after Clitophon's plaidoyer does not seem to have caused especial surprise in antiquity. in his dramas. FestugieÁ re. One explanation of it is known to us. e n twÄ i o mwnu mwi dialoÂgwi ktl.3 Is the Clitophon un®nished? i. if the pattern of rhetorical kathgori a is followed at all. is taken to mean `not at any rate in the Clitophon (but in the Republic)'. 340a3±b8).1 Theodectes is. Proclus does not think much of the identi®cation: toÁ deÁ KleitojwÄnta [sc. who identi®ed the missing fourth person of the Timaeus (17a1) with Clitophon: touÄ ton gaÁr e n twÄi omwnu mwi dialo gwi mhd' a pokriÂsewv hxiwÄsqai paraÁ SwkraÂtouv. cf. 2. 7b ˆ 1. This theory is taken over by Susemihl (508).20. Ge¨cken concludes (439) that Theodectes (rather than a pupil of his) is the author of the Clitophon. C (a±b) as roughly equivalent to pi stiv and C (c) ± the closing sentence ± as epilogue (section i.2). Proclus.3.39 ˆ 1. 6 Apud Stob. fond of riddles.

outwv h gaÁr h poÂliv twÄ n A  tlantikoÁn e n polloiÄ v hkaiÁ i Pla twnov soji a toÁ n A kaloiÄ v mo non e rgon a teleÁ v e schken. Scriptorium 20 (1966) 41±54 at 44 n. In Platonis qui vulgo fertur Minoem eiusdemque libros priores De Legibus (Halis Saxonum 1806). I see no reason for Madvig's toÁ A  tlantikoÂn: the Critias is referred to with its regular sub-title.3. P. to wit Plutarch. 32. 12.I N T R O D U C T I O N I. O. 407c6±d2 with express mention of Plato: 439c and 534e. Souilhe .v. As far as I have been able to investigate. `Marsilio Ficino as a beginning student of Plato'. he adduces Ptolemy and Plutarch (cf. the biographer and bibliographer of Aristotle. who was well acquainted with it. 1): Boeckh does not subscribe to this idea himself. PW s. has been identi®ed beyond doubt by A. the beginning of the protreptic speech (407b1±4) was quoted literally by him (4e). supra) as proof that even the most ancient MSS had no more text than ours have.10 An alternative hypothesis explained Socrates' silence as due to a subsequent curtailment in the transmission: the dialogue was not imperfectus (de Serres) but mutilatus. 8 He twice paraphrases Clit. 9 Sol. 10 Cf.9 From the sixteenth century onwards. Hermes 85 (1957) 314±25.8 yet writes about the Critias wv  qhnai wn toÁ  O lumpi eion. There are no ancient readers known to us who explained the absence of an answer as an indication that the Clitophon was left un®nished. some scholars have sought 7 `Der Platoniker Ptolemaios'. Ptolemaios 69). the ®rst to propose this theory was Jean de Serres (in Stephanus' edition). 11 So A. If PeriÁ pai dwn agwghÄv is authentic. Dihle7 with Ptolemy al-gharib. 11 . 1859±60.1 the Timaeus. de Serres probably advanced it to counter the hypothesis found as early as Ficino that the Clitophon is not authentic. 171 n. One reader at any rate says by implication that it was ®nished. 11 (cf. Kristeller. Boeckh.2.11 In the course of the nineteenth century and at times in ours. the notion that Socrates' answer is lacking because the Clitophon is a torso becomes widespread.

in my opinion. 18 So most scholars quoted in the previous notes. cf. but those who spread it around obviously did not think it absurd (cf. Plato (London 1902). 150±1. 1 ˆ Kleine Schriften (Leipzig 1901). Secondly. E. 7. 12 . i 232 n. cf. Gomperz. G. Apolog y. 25. 1: after replacing Clit. `Der lo gov SwkratikoÂv'. enough in12 In itself. Sokrates. 8±11 and (unduly sceptical) van Groningen. K. de Strycker±Slings.16 O. H. Taylor. says that according to some the Laws were transcribed by Philip of Opus o ntav e n khrwÄi (3. by Republic 1. for the problem G. iii 19±26. the Critias is not likely to have been published during Plato's life ± an un®nished Clitophon would provide a parallel for it. 2. 19 De kunst van het gesprek (Antwerpen±Amsterdam 1976).13 Grote.12 To name just a few: Boeckh. there is no objection to this possibility: when D.14 Th. `The case against Sokrates has been made so strong. 13 Index Lectionum der UniversitaÈt Berlin 1840. SAWW 114 (1887) 763. hardly anything can be meant but a publication of a (®nished or nearly ®nished) book found among Plato's `papers' ± it does not matter whether or not we believe the story.I N T R O D U C T I O N I. JoeÈ l.15 A. 10. that I doubt whether Plato himself could have answered it to his own satisfaction' (21). found after Plato's death among his papers. Ritchie.17 Usually. 14 Plato. 16 Plato 12: either un®nished or spurious. de Strycker:19 Plato abandoned the Clitophon because he had expressed the same ideas better in the Euthydemus. 17 Platon. 13). Studien zu den platonischen Nomoi (Munich 1951). Duemmler. Maier. i '. Wichmann. L. 15 `Platonische AufsaÈtze. D. Zur Composition des platonischen Staates (Basel 1895). MuÈller.3. Ideelle Gesamtdarstellung und Studienwerk (Darmstadt 1966). 133 n. also F. this theory is connected with the supposition that the Clitophon was originally intended as a prooemium to the Republic.37). 285±6 n.18 An interesting alternative was put forward recently by E. 64±5. `EKDOSIS'. Plato decided to publish the former as a provoking prelude to the Republic. As the Clitophon itself gives. but that half-way Plato changed his mind and used the alleged dialogue `Thrasymachus' instead.1 to reconcile themselves to Platonic authorship by having recourse to the idea of an un®nished sketch. 5 n. 17.

. scedoÁn kaiÁ e mpo dion).I N T R O D U C T I O N I. in which the key-word eudai mona comes as a sort of shock: although in fact the whole dialogue had been concerned with the way one achieves happiness. 1. taÁ deÁ kaiÁ e phÂinoun ± 410e4± 5 i na mhÁ kaqa per nuÄn taÁ meÁ n e painwÄ se proÁv Lusi an . a6±7 taÁ meÁ n gaÁr e gwge ouk e ph inoun se.). taÁ de ti kaiÁ ye gw). the last sentence is tied up inextricably with the last but one (cf.20 Besides. I shall treat the question without having regard to its authorship.2 dications to decide whether or not it is ®nished. antithetical structure. Comm. Kleitophon'. protetramme nwi deÁ scedoÁn kaiÁ e mpoÂdion touÄ proÁv te lov a rethÄ v e lqoÂnta eudai mona gene sqai (410e5±8) provides everything we should expect from it: it recapitulates the appreciation of Socrates' activities as expounded by Clitophon in the two major sections of the dialogue. `RaÈ tsel'. it ends up in a beautiful climax in the last clause touÄ proÁv te lov arethÄv e lqoÂnta eudai mona gene sqai. in which the prologue is repeated almost word for word (406a2±3 o ti Lusi ai dialegoÂmenov taÁv meÁ n metaÁ Swkra touv diatribaÁ v ye goi . there is besides a clear. 13 . `Pseudopl. yet it does so in slightly stronger language than Clitophon had used before ± this is to be expected in a peroration (axion . . also Comm. touÄ pantoÂv. Ge¨cken. though seemingly artless. .3.2 The problem The Clitophon as we have it certainly does not give the impression of being an un®nished text. ad loc. 5. . even if one does not accept Ge¨cken's analysis 20 Pavlu. Apart from that. 430 n. was not used before (cf. on 410e5 ga r). .3. The closing sentence mhÁ meÁ n gaÁr protetramme nwi se anqrw pwi w Sw kratev axion ei nai touÄ pantoÁv jhÂsw. i. the wordgroup eudai mwn. -moni a etc. . The end of the text clearly looks back to the beginning.

consequently this disposition indicates a ®nished whole. how was it prepared for ± if at all ± in the text of the dialogue that was actually written down? How would the ®gure of Socrates in such an answer correspond to the characterisation in the text? What would Socrates have been able to say in order either to deny the charge or to accept and explain it? These are questions which cannot be answered without giving at the same time an interpretation of the Clitophon. its disposition (exordium ± narrative ± accusation proper ± epilogue). ± We may safely discount the possibility of mutilation of the text posterior to its publication in a more complete form (section i. 5 to that section) does not hold water. 453).3. we should still expect it to be framed in the way it is and to end the way it does. in which the introductory conversation corresponds.3. shows a reasonable similarity to the pattern of a normal law-court accusation. to the exordium of a judicial speech. the chances of its occurring right at the place where Clitophon's requisitoire ends are in®nitesimal. these considerations in themselves do not disprove the possibility that something like a speech for the defence was originally intended by the author. unchecked against other copies.1) even if Boeckh's argument (n. we have to assume that it was precisely this copy or one of its descendants which eventually found its way into the Corpus Platonicum. I shall not waste more words on the possibility. we have to imagine a fanatically anti-Socratic reader cutting away. we shall do better to start with hypothetical questions. if it was intended.I N T R O D U C T I O N I. `Riddle'. in my opinion. half of his copy of the Clitophon so as to provide it with an antiSocratic tenor.2).2 of the Clitophon as a rhetorical kathgori a in all details (section i. Therefore.21 for even if Clitophon's accusation was intended to be answered. 14 . On the other hand such an interpretation is possible only 21 This point is overlooked by BruÈ nnecke (`Kleitophon wider Sokrates'. If an answer by Socrates was intended. If the mutilation was mechanical. 135±6 argues against individual hypotheses based on the assumption). Furthermore. Now. the only scholar who has adduced fundamental arguments against the torso theory (Roochnik. say.

W. Die sokratische Ironie in den platonischen FruÈ hdialogen (Amsterdam 1973). one typical of him (knowledge of what is good leads automatically to doing what is good).4). Hippias etc.I N T R O D U C T I O N I.5. For.22 22 Irony is mainly or exclusively a trait of Plato's Socrates (cf. What Socrates says is in e¨ect this: `In o¨ering to report your praise and blame you have o¨ered to make me a better man [wjeleiÄ n. even if the authorship of this dialogue is dubious. In this interpretation.2 if we are certain that the text as it stands responds to the author's intention.3. if I have learnt about my better and weaker points. Euthyphron. section i.2). Boder. In answering them we are entering a special case of the hermeneutic circle. I therefore feel justi®ed in making this general observation. (1) Socrates' reaction to Clitophon's opinion of him is foreshadowed in what he says after Clitophon has o¨ered to give a detailed account of what he had praised and  ll' ai scroÁn mhÁn souÄ criticised in Socrates' diatribai : A ge wjeleiÄ n me proqumoume nou mhÁ upome nein´ dhÄ lon gaÁr w v gnouÁ v o phi cei rwn ei miÁ kaiÁ belti wn taÁ meÁ n askh sw kaiÁ diw xomai. based on undoubtedly authentic dialogues. cf. 406a5) correspond infallibly to Socrates' weaker and stronger points (cei rwn ± belti wn). With this second assumption Socrates makes it impossible in advance to defend himself: Clitophon is the one who knows in what respect his diatribai deserve praise and blame. 15 . taÁ deÁ jeuÂxomai kataÁ kraÂtov (407a1±4). the other highly ironical: Clitophon's praise and blame (ouk e phÂioun ± kaiÁ ephÂioun. of course. Socrates has no option but to remain silent: he has ± ironically ± placed Clitophon above himself (as he does with Euthydemus and Dionysodorus. I shall strengthen the former and abandon the latter. 23±5) and our author at any rate handles the dialogue in Plato's way (section ii.' Socrates makes two assumptions (cf. ad loc.) and he cannot break the irony (he never does). and all that is left to Socrates is to listen demurely. Comm.].

depicted by Clitophon. kaiÁ nuÄn dhÁ tautoÁ n gigne sqw (410c8±d5) and the clause kai sou deo menov le gw mhdamwÄv a llwv poieiÄ n (e3) seem to me to indicate that Clitophon is not quite serious in stating the dilemma ± he may just have used it to incite Socrates to stop exhorting him and others and get down to business. The second is Socrates the preacher. oion de . obvious that if Socrates did get down to business.5.I N T R O D U C T I O N I. It is. . . section i. Clitophon's attack would have been implicitly justi®ed. however. (2) With these remarks we have already approached the second question. so curiously enough these words. . pampo lloiv kaiÁ pagka lwv legome noiv. Clitophon's praise of the protreptic speech is unmistakably ironic (407a6 pollaÂkiv e xeplhttoÂmhn a kou wn. despite the dilemma stated previously by him h ouk ei de nai se h ouk e qe lein authÄ v e moiÁ koinwneiÄ n (410c5±6). The sentence e peiÁ ei g' e qe leiv suÁ touÂtwn meÁ n h dh pauÂsasqai proÁ v e meÁ twÄ n loÂgwn twÄ n protreptikwÄn. in fact preclude such an answer. The ®rst is the Socrates sketched by the author in the opening conversation: formal in his ®rst. There are in fact two quite di¨erent characters parading under that name (section i.5. It is hard to see why the author should have worked in the irony if he intended to make Socrates wash himself clean of the allegations. ironical in his second r hÄsiv.2). . 408b6±7 loÂgoiv . These two characters can coexist within the framework of one dialogue so long as 16 . In the summing-up and the epilogue Clitophon makes a `last appeal' to Socrates to start telling him all about the care of the soul.3.3).2 Some other features of the text reinforce what may be concluded from the sentence 407a1±4. who moreover states expressly that Socrates had uttered a statement about justice which on closer examination had proved untenable (410a8±b3). which on the surface seem to point towards an answer. namely how an answer by Socrates would square with the character of Socrates as outlined in the text.

an ironical Socrates who admits to knowing nothing and goes on to explain his way of philosophising (like the one in the Theaetetus) would clash with the one who humbly places himself under Clitophon's guidance. 17 . but they cannot occur at the same level without either of them proving false.1). This answer would belong to the direct dialogue. and who has therefore made himself guilty of what is elsewhere called e ponei distov amaqi a (Apolog y 29b1±2). This may have been intended as an opening for a defence. . so it would be up to the ironical Socrates of the prologue. makroÂteron deÁ oude n (410b4±7) . I doubt if even a clumsy writer would fail to realise the impossibility of this task. Clitophon leaves open the possibility that Socrates.23 who has made a false statement about justice. What bene®cial function could have been attributed to an exhorter without knowledge? Plato's Apolog y provides the answer: there Socrates repeatedly testi®es to his lack of knowledge and rather suddenly appears as an exhorter (29d4±e3). to defend the exhorter. he fails to distinguish between the two levels. (3) As to the content of such a defence. True. does not possess knowledge of the subject towards which his exhortations are directed: nomi sav se toÁ meÁ n protre pein ei v a rethÄ v e pime leian kaÂllist' a nqrwÂpwn draÄn. h ouk eide nai se h ouk e qe lein au thÄv e moiÁ koinwneiÄ n (c5±6). even if it were possible for him to explain away the de®ciencies of the pompous preacher who is lacking in knowledge. . But this time the exhortation is inseparably tied 23 When BruÈnnecke says (`Kleitophon wider Sokrates'.3. duoiÄ n deÁ qa teron.2 they remain on separate levels (see section i. 456) that the Socrates of this dialogue could only have answered the charge with a new protreptic.5. Such a confrontation would be bound to take place if Socrates were to answer the charge.I N T R O D U C T I O N I. h tosouÄ ton moÂnon duÂnasqai. though admittedly a good exhorter. who forgoes any claim to knowledge.

Clitophon himself does not possess knowledge and does not pretend to possess it.3).3.4 The Clitophon as a Short Dialogue i . If we take the epic genre as an analogon. But as this promise has already been made (taÁ meÁ n askhÂsw kaiÁ diwÂxomai.4. What is more. 140±3) ± he has been beaten at his own game. Socrates' silence is not a sign of superiority (cf.4). it also presents him with the problem of generic di¨erence. in other words with elenchos (29e3±30a3). `Riddle'.I N T R O D U C T I O N I. There is only one answer left to Socrates once Clitophon has ®nished: an admission of guilt and a promise to mend his ways accordingly. section i.1) or of a fundamental di¨erence between him and Clitophon. section ii. which makes discussion impossible (so Roochnik. The structure of the Clitophon was therefore intended from the beginning. Could this combination of absence of knowledge. we observe that in the course of Greek literature a subdivision develops for which bulk is the criterion.3. there is elenchos in the Clitophon. Socrates had to remain silent: any answer would have been trivial.1 up with the expulsion of conceit. so elenchos is pointless in this case.4.1 The question of genre One of the ®rst things that strike the reader of the Clitophon is that it is so short. and Clitophon is the one who uses it (section ii. The most plausible hypothesis about the length of the oral epic before Homer's time is that it did not last much longer than the average listener could toler18 . but it is directed against Socrates and his companions. Ptolemy. taÁ deÁ jeuÂxomai kataÁ kra tov 407a3±4). exhortation and elenchos have been used in the Clitophon as a defence of Socrates? The answer is no. i. While this has some obvious advantages for the commentator.

27 24 This may have ¯uctuated considerably according to the occasion. and Timaeus and Critias. on the South Slavic parallels.g. Perhaps also the un®nished trilogies Sophist ± Politicus ± `Philosophus' and Timaeus ± Critias ± `Hermogenes' might be fruitfully analysed as belonging to this sub-genre. AJPh 97 (1976) 336±9. Wender. A. more. the composition of the audience and. AJPh 98 (1977) 327±47 at 339.I N T R O D U C T I O N I. 14±17. Lord. One of the most important criteria is the possibility of compression. 27 Among the most prominent features of the monumental dialogue would be: full treatment of the subjects encountered and related ones. for instance a more complicated plot.24 maybe the poems of Hesiod and the data about the number of books of various epics of the Cycle give us an idea. Cf. 25 D. The Singer of Tales (Cambridge (Mass. the Iliad and Odyssey were considerably longer than previous epics used to be.1 ate. `Homer. Avdo Mededovic . lengthier and betterstructured speeches etc. which is now wellknown under the name of Monumental Epic.) 1960). `A note on Plato's un®nished dialogues'. the proof of immortality in Republic 10: mention of immortality is of central importance to the Republic (it is the precondition of one of the rewards 19 . even when this would appear unnecessary ± e. are single dialogues that were split up in the course of the transmission of Plato.26 It is not the place here to enter into details. B. We are told that long South Slavic epics can be compressed into one-sixth of their actual length without great damage. the quality of the singer. and the elephant's child'.4. If this is right.25 the Homeric epics cannot. the former are just long. of course. Therefore the latter are monumental. At this point a decision must be made: is the di¨erence in length an irrelevant factor or does it go hand in hand with a number of structural di¨erences.? If the answer is a½rmative and the di¨erences are signi®cant. W. Haslam is right in claiming that Sophist and Politicus. A similar case can be made for Republic and Laws as Monumental Dialogues rather than abnormally long dialogues. it is useful to assign the new lengthy epics to a special sub-genre. 26 This is true a fortiori if M. more attention to character.

the more so because we shall be tempted to introduce as typical features of the Epyllion what are in fact general characteristics of Hellenistic narrative poetry. 114 and n. 107 (352±48 bce) which we have on the authority of Origenes (ˆ Hieronymus and Syncellus) to `the end of the fourth century or very early in the third' he is biased. besides.28 In fact. after which the main line of thought is resumed. Carl Werner MuÈ ller has of justice). L. virtual absence of arguments ad hominem. the Aspis belongs here.4. and some Homeric Hymns are closely related (one may also think of narrative choral songs like Bacchylides 17). The Epyllion is often considered an invention of Hellenistic poets. ZPE 25 (1977) 95±119. There is no compelling reason for thinking so.I N T R O D U C T I O N I. 283d) behind him. and the creation of the sub-genre typical of that era. but for one aspect. frequent re¯ections about the results that have been achieved so far. As an analogy to the Epyllion. but when he brings down the ¯oruit of Olymp. virtual absence of elenchos as puri®cation. esp. 28 I agree with M. some of them will be typical of other dialogues as well ± especially Phlb. absence of concentric reasoning (section ii.±P.1 As a counterpart of the Monumental Epic we are confronted with the Epyllion.3. frequent digression within the discussion. 36) and commemorated by Asclepiades (xxviii G. the Hesiodic collection of  H oiÄ ai is little else than a string of epyllia. I disagree when he thinks it necessary to assume that `Erinna' was really a pseudonym of a mature poet. It is not easy to ®nd the common denominator for the various representatives of this sub-group (most of them are partly or wholly unknown to us).1 (even in the case of Thrasymachus).5. section ii. is certainly not an `unsophisticated teenage girl'. though I can see his point and he has an unknown ancient authority (cf. 20 . whose  H laka th belongs to the sub-genre. Erinna was imitated by Anyte and Nossis (West. virtual absence of those short-cuts which are created by making a partner willingly grant a highly debatable point. there is no need to pursue the matter further. West (`Erinna'. Ath. Fortunately.) ± all of whom belong to the late fourth or early third centuries. cf. frequent deliberations about questions of method. but a proof is super¯uous after what had been said earlier about the theory of forms.1). Most of these features are closely connected which each other. 116±19) that Erinna.

122 M. Bompaire (Lucien Âe crivain (Paris 1958).30 He gives good reasons for thinking that 25 short dialogues were attributed to Aristippus (320±1 n. Ep. 31 Cf.. this leads to an upper limit of 6. ± J. 5): Diogenes Laertius 3. cf.) but denies the existence of a direct link between these and the Socratic dialogue (312). L. 562±85 and passim) uses the word `petit dialogue' to refer to the various collections of DiaÂlogoi (nekrikoi . [there follow 23. 4). as a genre it will be referred to as `Short Dialogue'. The best explanation seems to be to assume that two titles have been lost. 33 These are also mentioned in the second list given by D. 2).85. e n deÁ e n w i dia logoi pe nte kaiÁ ei kosin. fr.4. titles].29 As a criterion of length he uses the Spuria transmitted under Plato's name minus Axiochus and Eryxias (but including the Alcyon). oi de . Mannebach.32 Some authors in antiquity thought Aristippus' philosophical productions were con®ned to these pieces and to six books of diatribes. . 79 n. Aristippi et Cyrenaicorum fragmenta (Leiden 1961).4. . . that diaÂlogoi is used in a very loose sense (among the titles are ProÁv touÁv e pitimwÄntav o ti ke kthtai oi non palaioÁ n kaiÁ  rh thn thÁn qugate ra31) e tai rav and  E pistolhÁ proÁ v A and that all titles refer to very short pieces. 81.2(7). a lower limit of 1ˆ pages OCT (320 n. Mannebach is aware of the problem.Socr. 32 An average of 3‰ pages OCT according to MuÈller's reckoning (Kurzdialoge. . SSR iv a 144) as well as by Theopompus (FGrHist 115 f 259 ˆ Aristipp.I N T R O D U C T I O N I. 27 KoÈhler and E. not 25. section i. (2.33 It is of course preferable to suppose that Diogenes' list is a sort of contamination of the one volume with its 25 short dialogues with a subsequent list of other 29 I shall translate it for the moment as `short dialogue' until the question whether this is a separate literary genre is treated.1 introduced the notion of Kurzdialog. oi d' oud' olwv graÂyai (there follows a whole di¨erent list).  Enioi deÁ kaiÁ diatribwÄ n au toÂn jasin e x gegraje nai. 21 .83±4 (SSR iv a 144) TouÄ dhÁ KurhnaiÈ kouÄ jiloso jou je retai bibli a tri a meÁ n i stori av . 321 n. SSR iv a 146). 1. e tairikoi etc. 30 But cf.

2. though not necessarily. L. Simon (33) (ascribed also to Phaedo and Aeschines of Sphettos). 2. 199 is. 2. just as there were similar collections of short dialogues attributed to Crito (17). Tarrant's totally misguided views of Thrasyllus as an editor of Plato (Thrasyllan Platonism (Ithaca± London 1993)). Glaucon). Crito] e n e niÁ jerome nouv bibli wi e ptakai deka. to my mind. 7. L. the earliest indication of a tetralogical edition is Hippol. no doubt right in supposing that Thrasyllus wrote an introduction to Plato.I N T R O D U C T I O N I. in one volume and therefore short dialogues). following Usener. one should rather expect a complete list of 25 titles and suppose that this list got somehow mixed up with a list of works of greater bulk.122 (SSR vi b 87) skutikouÁ v au touÄ [sc. the hypothesis that he also produced an edition of Plato is uneconomical. 1.35 Simmias (23) and probably Glaucon. though obviously false. . Questions to be Settled Before the Study of an Author. including dialogues) attributed to Aristippus. 407d4±8 with the words kaiÁ le xiv tou tou [that evil actions are in- 22 .1 philosophical works. He adduces two arguments: All the transmitted Spuria already formed part of the Corpus Platonicum at the time of the tetralogical edition (Kurzdialoge.121 (SSR vi b 42) dialo gouv ge grajen [sc. . touÁ v u pogegramme nouv (there follow 17 titles). 32±41). je rontai kaiÁ a lloi du o kaiÁ tria konta. or a Text (Leiden 1994).37 (43 bce. Simon] touÁ v dialo gouv kalouÄ sin. id. quotes from it). of H.122 (SSR vi b 87.36 MuÈller considers the short dialogue a secondary and later development of Socratic literature.105 (SSR iii a 8) dialo gouv deÁ sune graye [sc. seems to be ineradicable ± it lies at the bottom. ibid.37 the author of that edition put 34 On the analogy of D. 59±60. 75±6). (Simmias). L. 36 D. 35 D. Simon). Mansfeld.124 (SSR vi b 63) GlauÂkwn AqhnaiÄ ov´ kaiÁ touÂtou je rontai e n e niÁ diaÂlogoi e nne a [not short ones. cf. oi noqeu ontai (presumably. for which. pace Tarrant.21.L. The earliest testimony for the tetralogical list is probably Varro.34 We must conclude that there was a collection of 25 short dialogues (or 23 short pieces. 37 The idea that a tetralogical list presupposes an edition ordered that way. L. J.4. apparently] . who quotes Clit. Prolegomena. L.19. This would do more justice to Diogenes' actual words. 2. 124 (SSR vi b 63. Phaedo] gnhsi ouv meÁ n [2 titles] kaiÁ distazoÂmenon [several titles and alternative ascriptions] skutikouÁ v loÂgouv´ kaiÁ tou touv tineÁ v Ai sci nou jasi n. As Thrasyllus must have written a book on Plato in any case (which we know because D. he conjectures the title taÁ proÁ thÄ v anagnwÂsewv twÄ n PlaÂtwnov dogma twn (98). 2. among other things.

39 It is not clear what MuÈller thinks of the authorship of Aristippus' collection.1 the omologoume nwv noqeuo menoi in an Appendix and (since ®ve out of seven dialogues in the Spuria are short dialogues) indicated thereby that short dialogues were not considered Platonic (42±4). one of whom ± Simmias ± is even credited with two epigrams on Sophocles and perhaps one on Plato). was ascribed to Simon the Cobbler. This is hardly the purposeful expression of a view that the Clitophon is part of the Republic. 321±2). the skutikoiÁ diaÂlogoi. 39 AP 7. 23 . 22. Secondly. Hippolytus or his source consulted an edition of Plato in which the dialogues were grouped in tetralogical order (cf. The Clitophon.4. 47 n. See section i. Kurzdialoge. who was a mere literary ®gure. 211±12. 5). The basic premise of this argument is improbable (especially for the Alcyon) and not subject to proof. they cannot have been earlier than Simon's ®rst appearance in Socratic literature. voluntary] e mjanestaÂth e stiÁ n e n thÄ i Politei ai. is only an apparent exception: it was assigned a place within the eighth tetralogy before the Republic because of its subject-matter. Alline. In other words.I N T R O D U C T I O N I. Histoire. but his opinion is not proof. which is certainly a short dialogue (6 pages OCT). Glaucon and Simmias are adespota which were provided with authors who likewise are well-known ®gures from Socratic dialogues (though this time they are also undoubtedly historical persons. the short dialogues ascribed to Crito.2(7). By the same token. best explained if we think of a complete Plato which contained both Clitophon and Republic (or part of it). 60. this proves at most that the author of the tetralogical list did not consider them Platonic. since one group of short dialogues. my review of MuÈ ller. 38 Cf.21. where a similar confusion in Stobaeus is indicated as well). MuÈller argues (Kurzdialoge. the argument about the Clitophon is special pleading. 124.38 if the short dialogues were indeed rejected because of their length. which is in Phaedo's dialogue Simon. It is an interesting slip.4. as Heidel thinks (Pseudo-Platonica.

41 Dodds Grg.124). 5. All this. Hermann. 42 Kurzdialoge. dependent on the one unprovable supposition that Simon the Cobbler was not a real historical ®gure. F. The allegation is obviously false. Dialogue. w n e mnhmoÂneuen upo40 A healthy reaction is displayed by Hirzel. however. Hobein in his PW articles Simmias (4) and Si mwn (6).42 therefore a date later than 350 seems unlikely. It is another matter whether these collections of short dialogues really belong to the authors with whom they are connected. 10 and n. 212.I N T R O D U C T I O N I.40 We have nowadays learned to accept what was highly questionable one century and a half ago. Simon] e rcome nou SwkraÂtouv e piÁ toÁ e rgasthÂrion kaiÁ dialegome nou tinaÂ. I am inclined to agree with him that Aristotle knew this dialogue. The circumstance that there are three di¨erent ascriptions for the Cobbler's Dialogues. does not inspire con®dence any more than Sotion's and Panaetius' silence about the volume with 25 dialogues ascribed to Aristippus. But there is a much more important point. Archaeological claims that Simon's workshop has been identi®ed are discussed by Kahn. Diogenes has a remark about their origin which may be illuminating: 2. 388e2±389a4. MuÈller is unable to prove that short dialogues were not written in the ®rst half of the fourth century. 91: EN 1112a21±3 ± Sis. the Eleatic visitor is the exception which proves the rule. namely that all dramatic ®gures conversing with Socrates in Plato's works correspond to historical ®gures in Socrates' time41 and there is no reason why other Socratics should have resorted to imaginary ®gures. 24 .4.1 We are presented here with a conglomerate of hypotheses started o¨ by Ast and C.122 outov [sc. 19. Let us once again turn to Cobbler's Dialogues. my review. i 102±5. Dialog. 12 and n. For one thing. MuÈller himself dates the PseudoPlatonic Sisyphus in the ®fties of that century (Kurzdialoge. 94±104). cf. is not quite to the point. whereas of the short dialogues ascribed to Glaucon Diogenes bluntly states noqeuÂontai (2.

that the Cobbler's Dialogues and the conversations reported in Xenophon's Memorabilia (and Symposion) belong to the same group of literature.1 shmeiwÂseiv e poieiÄ to. kaiÁ prwÄtov uposhmeiwsaÂmenov taÁ legoÂmena ei v a nqrwÂpouv h gagen. Glaucon and Aristippus were of the same type of apomnhmoneu mata we cannot know. 26±31. That this is indeed ®ction was proved by H.4.I N T R O D U C T I O N I.43 In antiquity legome nou proÁ v A the conclusion drawn from such introductory formulae was that Xenophon made notes of real conversations of the historical Socrates: Diogenes 2.4 PeriÁ qewÄn. nevertheless justly carries the  pomnhmoneu mata. 94. in the case of the Memorabilia. These words are strikingly A similar to the statement about Simon.48 kaiÁ tounteuÄqen akroathÁ v SwkraÂtouv h n [sc. Memorabilien I. Cf. It is not too bold to suppose that if similar conclusions were drawn the material was similar. 1.2 le xw de prwÄ ton a pote autouÄ hkousa periÁ touÄ daimoni ou dia ristoÂdhmon and passim. HGPh. whose major Socratic writing. One is immediately reminded of Xenophon here. namely as reports from memory of what Socrates had allegedly discussed.  pomnhmoneu mata e pigraÂyav. but it gives at least an indication of how the skutikoiÁ diaÂlogoi were presented.1 PeriÁ e gkratei av and so on) arranged them.44 43 Cf. Maier.1 tou twn dhÁ graÂyw o po sa an diamnhmoneuÂsw and Gigon. Xenophon]. Now Xenophon certainly did name A not hear all conversations contained in the Memorabilia personally. Sokrates.4. Simmias. 393± 401. 44 Whether or not the short dialogues ascribed to Crito. 2. in other words. but the ®ction is that he did: 1.3. though ostensibly a defence of Socrates against various attacks. iii 345. Kahn. Dialogue. This may or may not be ®ction (I think it is). within the framework of a larger text with an apologetic character. Xenophon (instead of calling 1. The only di¨erence between them is that whereas the short dialogues `remembered' by Simon bear separate titles. but for the ®rst three collections the names of the 25 . 29±35.

if not always for us. C.2 Short dialogues and the Clitophon (1) Short dialogues are almost exclusively47 duologues. 2±4. W. cf. 262¨.2 Not only. 107. which alone will enable us to reach such a decision.] Just. 129±30. 26 . falls outside the scope of this book.I N T R O D U C T I O N I. 1. MuÈ ller..4. Demod. 192±3). I must add here that a typology of Xenophon's Socratic conversations is an urgent desideratum. were short dialogues written during the ®rst half of the fourth century. 130 n.g.45 It will be understood that a thorough examination of the material. Memorabilien I. If there is consistency.2. one may even go further and draw a direct line from Xenophon's ®ctive presence via Aristotle's dialogues to Cicero. MuÈller and Gigon. Kurzdialoge. Kurzdialoge. Virt. cf. n. X. but quite a number of them have even been preserved. and perhaps [Pl. cf. In [Pl.48 the other is sometimes anonymous. As a deuÂterov plouÄ v. 320). Smp. is in this respect as anomalous as Plato's Smp. I have stressed elsewhere ± review of MuÈller.8±13. therefore. 1.30).3.46 and compare them with the Clitophon. In the main. as is done tentatively by Gigon. the case for the Short Dialogue as a literary genre will have been settled provisionally.] Sis..] Just. 211 ± the insu½ciency of MuÈ ller's treatment of the short Platonic Spuria in this respect. 1 is no dialogue.49 but usually his name is given so that he is identi®able for the contemporary reader. one of the partners is Socrates.4. I shall collect in the next section such general features of the short Platonic Spuria and Xenophon's Socratic conversations as have been observed by other scholars. 1. W. Silent persons are also implied in [Pl. 1. Mem. 94. Demod. In X. MuÈ ller's attitude is ambiguous: `In der Tat kann von einem Genos des Kurzdialogs hoÈ chtens per analogiam die Rede sein' (Kurzdialoge. MuÈ ller. i . More data about Socrates' partner are only provided 45 46 47 48 49 authors certainly suggest it. a silent third is sometimes present (e. and should be used in order to decide whether or not the Short Dialogue is in fact a separate sub-genre of the dialogue. MuÈ ller. 2±4 (for the numbering. Except in Demod.). like the Monumental Dialogue. Kurzdialoge. C. (but cf.

except for the fact that he is apparently not too old for Socrates' and Thrasymachus' sunousi a. Clitophon is called `son of Aristonymus' in order that we may identify him. where it is suggested that the `kleinbuÈrgerliche Milieu' of the Cobbler's Dialogues has a similar e¨ect. is easier `wenn er Alkibiades heisst'. 3). On the other hand. 27 . especially in the prooemium (sections i. the Clitophon is apparently di¨erent inasmuch as there certainly is some degree of characterisation.3 n.5. My personal feeling is that `Selbstidenti®kation' will have been mainly the e¨ect of the questions asked of Socrates' partner. no matter how anonymous he was.5.3).2 when they are functional. cf. C. This. W. as a conscious device to provoke `Selbstidenti®kation' of the reader with Socrates' partner. this characterisation plays a very important part in the author's message. his explanation does not look very probable. Thus in the Sisyphus we learn that Sisyphus is a su mboulov of the archons of Pharsalos (387b7±c3). i.4. whereas in longer 50 Kurzdialoge. there is usually little or no characterisation. he argues.50 Though we must all confess to a complete ignorance about the contemporary reader's reception.5. when Xenophon ascribes to Socrates' interlocutors certain qualities of character which are the topics of the ensuing conversations ± for instance Aristippus' akraÂteia in 2.) In this respect. which seems only a natural one in short dialogues.I N T R O D U C T I O N I. (Of course. MuÈller explains this feature. because the author needed a starting-point for the discussion on (eu) bouleuÂesqai. but the rest of his curriculum vitae is passed over ± even his age is not hinted at (section i. especially n. whether in actual utterances or in stylistic markers. (2) Apart from the scarcity of `external' data about the characters in short dialogues. The Clitophon conforms. 1. 323.2.1 ± this cannot count as characterisation. and to make him concentrate on the subjectmatter.

I N T R O D U C T I O N I.4.2

dialogues personal features are usually much less important for the interpretation of the dialogue. Within the area
of short dialogues the Clitophon is (in this respect) the exception which proves the rule.
(3) Parallel to this lack of individualisation there is normally a virtual absence of situational context, of `scenic
background'.51 Inasmuch as the short dialogue is at the
same time a `dramatic' (two-level) dialogue (section i.5.1),
this is quite normal: among longer dramatic dialogues
Phaedrus and Laws are the only ones to have any scenic
background to speak of.52 The only situational information
in the Clitophon is contained in Socrates' opening words ±
these are the point de deÂpart of the conversation; the words
moÂnw tugcaÂnomen ontev (406a10; cf. Comm.) are necessary
to explain a marker of formal style (section i.5.2). In `indirect' (three-level) short dialogues one would a priori expect
some scenic background, but it is again virtually con®ned
to essential data, usually put together at the beginning of a
conversation, e.g. Lamprocles' un®lial attitude towards
Xanthippe or Pistias' workshop.53 In Clitophon's report
there is no situational context worth mentioning.
Not only are the characters in short dialogues usually
robbed of their individuality, one could say the same of
the conversation. Generally speaking, it is more abstract,
more schematic than in longer dialogues. A number of
51 Cf. MuÈ ller, Kurzdialoge, 322; Gigon, Memorabilien I, 94.
52 This may be explained as due to a desire to bring in a situational
context combined with dissatisfaction at the three-level dialogue (cf.
section i.5.1) as expressed in Tht. ± I discount La. because its situational context is the starting-point for the dialogue. Sis. is the only
dramatic short dialogue which has anything in the way of scenic
background (387b1±5) but it serves as explanation for Sisyphus'
absence, the reason for which is his function of su mboulov explained
above. The Alcyon is too late (cf. my review of MuÈller, Kurzdialoge,
213) to take into account.
53 X. Mem. 2.2.1 and 3.10.9 respectively; 2.8.1 is rather exceptional. X.
Smp. is again ± cf. n. 47 ± anomalous.


I N T R O D U C T I O N I.4.2

peculiarities of short dialogues can be brought under this
(4) Many short dialogues start immediately (or almost
immediately) with a full statement of the problem; often
the ®nal sentence corresponds to this statement, giving a
comprehensive formulation of the result.54 A comparison
of the Meno (the only longer dialogue which begins immediately with the problem) with PeriÁ a rethÄv, which is an
extract of it, is instructive.
Meno (70a1±4):  E ceiv moi ei peiÄ n w Sw kratev, a ra didaktoÁ n h
a reth ; h ou didaktoÁ n a ll' a skhtoÂn; h oute a skhtoÁn oute
maqhtoÂn, a llaÁ juÂsei paragi gnetai toiÄ v a nqrw poiv h a llwi
tiniÁ troÂpwi; 
ra didaktoÂn e stin h a reth ; h ou
PeriÁ a rethÄ v (376a1±2): A
didaktoÁ n a llaÁ juÂsei oi a gaqoiÁ gi gnontai a ndrev h a llwi tiniÁ

Meno's question is here shorn of its personal traits (and
of the askhto n alternative, which does not come up in the
Meno anyway),55 but apart from that the openings are
closely similar. At the end of PeriÁ arethÄv, however, the
conclusion is summed up: outwv e oiken ou te didaktoÁn
einai oute juÂsei areth , allaÁ qei ai moi rai paragi gnetai
ktwme noiv (379d9±10). The Meno ends di¨erently: e k meÁ n
toi nun touÂtou touÄ logismouÄ w Me nwn qei ai moi rai h miÄ n
jai netai paragignome nh h arethÁ oiv paragi gnetai
(100b2±4);56 this statement is followed by a proviso that the
result is not de®nitive until it has been established what
exactly areth is (this question is not touched in PeriÁ
a rethÄv), and the Meno ends with goodbye and a reference
to persuading Anytus.
54 MuÈ ller, Kurzdialoge, 323.
55 Cf. Bluck on 70a2 askhtoÂn; MuÈ ller, Kurzdialoge, 198; section ii.2.3.1.
56 C. W. MuÈller (Kurzdialoge, 217) should not have compared Virt.
379d9±10 to Men. 99e4±100a2.


I N T R O D U C T I O N I.4.2

The di¨erence is clear: whereas Plato is contented with
restating the outcome (qei ai moi rai ) in itself (not at the
absolute end of the Meno), the anonymous author of PeriÁ
arethÄv ®ts the outcome in with a renewed statement of the
alternatives of the opening question, which terminates the
In Xenophon, the correspondence of opening and end is
usually shifted to the general framework into which the
short dialogues are ®tted, e.g. Memorabilia 1.5 begins Ei deÁ
dhÁ kaiÁ e gkraÂteia kalo n te ka gaqoÁn a ndriÁ kthÄ ma e stin,
e piskeywÂmeqa ei ti proubi baze le gwn eiv tauÂthn toiaÂde
(a speech follows). (1.5.6) toiauÄ ta deÁ le gwn e ti e gkrate steron toiÄ v e rgoiv h toiÄ v lo goiv e autoÁn e pedei knuen.57
We have already seen (section i.3.2) that the crucial elements in the prologue of the Clitophon are repeated in the
last sentence but one. Similarly, the interrogation of Socrates' companions after Clitophon's methodical remarks is
encircled by parallel sentences: 409b8±c1 toÁ d' e teron, o
duÂnatai poieiÄ n hmiÄ n e rgon o di kaiov, ti touÄto jamen;
eipe . 410a4±6 thÁ n deÁ upoÁ souÄ legome nhn dikaiosuÂnhn h
omoÂnoian opoi tei nousa e stin diape jeugen, kaiÁ a dhlon
authÄv oti poÂt' e stiÁ n toÁ e rgon. It may be observed that
the conclusion is given in a condensed form; logically
speaking omoÂnoia alone can be subject of the o poi clause
whereas the referent of authÄv is dikaiosuÂnh only; the
condensation has the literary advantage of establishing a
reference to the introductory question.
(5) Not only do beginning and ending correspond in
short dialogues, also half-way one often ®nds markers of
transition, new stages in the argument etc., or alternatively
57 The participle le gwn is concomitant (`while he used to say this'), not
causal, so that it is left to the reader to draw the conclusion that he
actually made people self-controlled by his speeches on the subject;
cf. Gigon, Memorabilien I, 150.


I N T R O D U C T I O N I.4.2

notable absence of such markers, so that a transition is
rendered more obvious by its abruptness. Fluent transitions from one stage of the argument to the next are rare.
This has been noted by C. W. MuÈ ller for the short Spuria
(`Das formale logische GeruÈst der Argumentation tritt in
vielen Kurzdialogen stark hervor' (Kurzdialoge, 323)) as well
as by O. Gigon for the conversations in the Memorabilia
(`genaue bis uÈ berscharfe Disposition'58).
In the Clitophon there are several clear markers of transi kouÂoiv
tion, such as the announcement of the report in A
an . . . gaÁr (407a5; Comm. ad loc.); the interruption in the
report of Socrates' speech (407e3±4) underlines the end of
its ®rst part (section ii.2.2). The sentence 408b5±c4 not
only marks the end of the speech but also terminates Clitophon's praise, as is shown by a comparison with the next
sentence, in which an oblique reference to the Euthydemus
announces disappointment (Comm. on 408c4±7). For the
marking use of irony as framework of the speech, cf. section i.5.3. The second and third de®nitions of the e rgon of
justice are both introduced by clauses containing the participle teleutwÄ n, `®nally' (409d2; 410a7).
On the other hand, we ®nd several abrupt transitions,
one even in the middle of a sentence: the third part of
Socrates' speech is marked o¨ from the second only in
that both begin with a distributive temporal clause: kaiÁ
opo tan au jhÄiv (407e5); kaiÁ otan le ghiv (e8) ± cf. the use
of teleutwÄn as a similar marker ± but there is no main
clause to follow either of them (Comm. on 407e5±8). Likewise, after Clitophon's methodical distinction between
e rgon and di dagma and his request to de®ne the latter in
the ®eld of justice, the transition to the answers given to
him (outov meÁ n wv oimai toÁ sumje ron a pekri nato ktl.,
409c1±2) probably contains both an anacoluthon (Comm.
58 Memorabilien I, 94; cf. HGPh, iii 342 (on X. Smp.).


I N T R O D U C T I O N I.4.2

on 409a7 eipoÂntov de mou) and an asyndeton. Finally there
is to my mind some degree of abruptness in the introduction of the sentence which closes Clitophon's report and
introduces his verdict on Socrates: tauÄta deÁ ouc apax
oudeÁ diÁ v allaÁ poluÁn dhÁ upomei nav croÂnon liparwÄn
apei rhka nomi sav ktl. (410b3±4); the particle de is rather
weak for a major transition. This incongruence of form
and content may be explained by the author's wish to give
a minimum of attention to the ignorance displayed by
Socrates himself (section i.5.3).
There is what may be called a ¯uent transition between
Clitophon's verdict on Socrates and his last appeal to stop
exhorting him and others and get down to business. The
positive side of the balance is stated right away (410b4±6
nomi sav se toÁ meÁ n protre pein ei v a rethÄ v e pime leian kaÂllist' a nqrwÂpwn draÄn); for the negative side Clitophon
hesitates between inability or unwillingness (b6±c6), which
is why he visits Thrasymachus and others, out of sheer
embarrassment (aporwÄn, c8). This last word is explained
by a (causal) epei clause which introduces the last appeal.
Only in the ®nal sentence of the dialogue is the balance
struck de®nitively.
(6) One feature which the Clitophon has in common with
the short Platonic Spuria but which is virtually absent from
Xenophon, is what C. W. MuÈller calls `die radikale Beschneidung der SpontaneitaÈt des Dialogpartners' (Kurzdialoge, 323). In the short Spuria, which are all `direct' dialogues, this is contrived by short answers which almost
always conform to the intention of the questions, and if
not, are still so harmless that they do not constitute real
objections.59 In the Clitophon, this phenomenon is paralleled
59 They are throughout in the type of discourse called `Question and
reply' by Thesle¨ (Styles, 35±41; this is the normal form of elenchos
in all Platonic dialogues).


I N T R O D U C T I O N I.4.2

by the `dummy' status of Socrates' companions in Clitophon's report of his interrogating them: the answers are
given in oratio obliqua and there is no sign of anything in
the way of discussion. See further section i.5.3.
(7) I may be permitted to subjoin an observation of my
own. Short dialogues are rarely what Thesle¨ (Styles, 34)
calls `pedimental'. There is no central60 episode which is
stylistically marked o¨ from the neighbouring parts and
which philosophically speaking constitutes the dialogue's
culmination. Such episodes are Diotima's speech in the
Symposium and the Sun±Line±Cave conglomerate in the
Republic.61 There are no such central culminations in the
short Platonic Spuria; nor have I found anything in Xenophon's `short dialogues' which can be regarded as such.
The two central parts of the Clitophon are each trichotomised (section i.2); as a consequence of this structure a climax is absent. The dialogues62 in the Platonic corpus in
which no pedimental structure is found are Crito, Alcibiades
60 Not necessarily (or even normally) precisely in the middle of the text;
cf. Thesle¨, Styles, 167 and n. 2.
61 See further Thesle¨, Styles, 34 and the literature quoted n. 2; his
analyses of various dialogues, 95±158; 167±8. His conclusion is that
among the authentic dialogues (Hp.Ma. is rejected, as it also is by
C. H. Kahn, `The beautiful and the genuine: a discussion of Paul
Woodru¨ 's Plato, Hippias Major', OSAPh 2 (1985) 261±87) Cri. is the
only dialogue without a central culmination. Pedimental structure
sometimes goes hand in hand with a protreptic concluding episode,
e.g. a myth, as in R., Phd., Grg. (Thesle¨, Styles, 168; Gaiser, Protreptik,
187±96). There is such an episode (the `last appeal') towards the end
of Clit. and of many conversations in X. Mem. (e.g. Prodicus' speech,
2.1.21±34). ± Thesle¨ (Styles, 168 n. 1) suggests that Clit. combines
central culmination with protreptic conclusion but I do not perceive
the former.
62 I discount Ap. (which creates a problem for Thesle¨, cf. Styles, 34 n. 3
and 119) for the simple reason that it is not a dialogue and ought not
to be treated like one. The same holds for the Letters, of which only
Ep. 7 has an obviously central episode, namely the philosophical


I N T R O D U C T I O N I.4.2

2, Erastai, Theages, Hippias Maior, Clitophon, Minos, PeriÁ dikai ou, PeriÁ arethÄv, Demodocus 2±4. With the exception of
Hippias Maior, these are all relatively short dialogues; none
of them is longer than 15 pages (Stephanus). Short dialogues which do have a central culmination of some sort
are Hipparchus, Ion, and Sisyphus. It appears that absence of
pedimental structure is normal in short dialogues; also that
the line which separates authentic works from Dubia and
Spuria cuts across the distinction of short dialogues with
and without pedimental structure.
This last fact is a decisive argument against C. W. MuÈller's
theory that Plato could not have written short dialogues
such as the Clitophon.63 It seems certain to me that the
Short Dialogue is a separate genre and that the Clitophon
belongs to it ± the above observations point that way,
although I would have been glad of more factual con®rmation regarding point (5). Now, it is structure, not
number of pages, which determines whether or not a particular dialogue belongs to this genre. To determine the
precise maximum length seems pointless, though obviously
there must be one (it would be absurd to call the Hippias
Maior a Short Dialogue; the absence of pedimental structure in it must be otherwise explained64). Consequently, if
we can credit Plato with the Crito (which at least by the
criterion of pedimental structure is a Short Dialogue),
there is no reason why we should deny to him such Short
Dialogues as Clitophon, Minos etc. on account of their
shortness, though we may feel compelled to do so on other
63 Kurzdialoge, 43±4; 320±1; 324±5. Cf. section i.4.1 and my review, 211±
64 Either as a sign of inauthenticity (cf. n. 61) or as due to a very early
date of origin.


I N T R O D U C T I O N I.5.1

i.5 The characters of the dialogue

i.5.1 Levels of discourse
If a literary text contains speech,65 there are normally two
levels of discourse.66 First, there is what may be termed the
`lower' level: here the text constitutes the discourse, and
the author and his public are the communicants. Second
comes the `higher' level, where the speech is the discourse,
and the literary ®gures who speak and listen are the communicants. The number of levels is augmented if within
the upper-level discourse there is again speech. This is the
case with Plato's indirect dialogues.67 In the Republic for
example, the lower-level discourse is between Plato and his
readers, the ®rst upper-level discourse between Socrates
65 Throughout this section, `speech' is used as a neutral term, denoting
the uttering of words per se. `Discourse' refers to speech as addresses
by one person to another (sometimes I have, for stylistic reasons,
substituted the vaguer term for the more exact one). `Communication' includes discourse as well as non-linguistic facts and events accompanying discourse. `Public' is meant to be understood as what is
normally called the `ideal reader'. In the absence of a full-blown
theoretical model for the interpretation of the Platonic dialogue, I
have devised this framework ± a synthesis of various and not very
modern theories of narrative text stripped of everything but the barest essentials (cf. esp. W. Schmid, Der Textaufbau in den ErzaÈhlungen
Dostoevskys (Munich 1973), 20±30 and the literature quoted there).
This was done mainly in order to give a satisfactory account of the
role of Socrates in the Clitophon ± it will be seen that level-distinction
yields no very impressive results in the case of Clitophon himself. I
have found this model rather satisfactory in studying the introductory scenes of Smp.
66 Except in the case of various types of monologue.
67 This does not mean that two-level dialogues are necessarily `dramatic dialogues'; Socrates' conversations reported by Xenophon are
often two-level dialogues. Plato consciously avoids this type: rather
than reporting in the two-level style, e.g. ``allaÁ pari hmi,'' e jh o
FaiÄ drov ``all' e rwÂta.'' metaÁ tauÄ ta dhÁ o SwkraÂthv e nqe nde poqeÁ n
h rxato (cf. Smp. 199c1±2), he uses the complicated four-level


We do not study their communication adequately by analysing only what they say. H. 136±64. R. 36 . level. We can now set up a principle for studying the downward relation: two adjacent levels are linked by the identity of the level-content of the higher to the discourse of the lower level. 69 Cf. All discourse presupposes at least two communicants. 231±51. Glaucon etc. At times there is a fourth. one at least at either `end'. In the interpretation of literary texts which have more than one level of discourse. I am not concerned here with the diachronic dimension which the term possesses in modern literary theory.1 and his unknown listener(s). Literaturgeschichte als Provokation (Frankfurt 1970). if any. esp. Other vertical relations between adjacent 68 One level may temporarily have zero realisation. 70 Whether they do or do not belong to the same level as the discourse itself is debatable but does not need to be discussed here.I N T R O D U C T I O N I.69 whereas descriptions of literary ®gures while in discourse should be taken into account as well. The Parmenides has ®ve levels from 127d6 onwards. the second upper-level discourse between Socrates and Thrasymachus.70 All these elements.5. Kunne-Ibsch. only one vertical link between levels is admissible: the downward relationship of one level to the one immediately below. as usually in the second half of the Parmenides. their reactions. for instance Er's report of his journey to the underworld and the proclamation of Lachesis' projhÂthv within that report. the Symposium four from 174a3 onwards. to the communication (at the lower level this is what is usually called `reception'). together with the discourse proper. even a ®fth. In that case one may say in practice that there is a downward relation of one level to the second or even third next. D.68 In order to determine and clarify the nature of this relationship the notion of `levelcontent' will be used here. we must also reconstruct the intention of their words. Fokkema±E. Jauss. Theories of Literature in the Twentieth Century (London 1978). constitute the content of a given level of discourse. W.

37 . along the lines suggested by FriedlaÈnder. e. in other words. let us consider the Republic. 71 A good case is Aristophanes' hiccups in Smp. if a certain passage from the textual discourse is lifted from its level-content and related directly to assumed elements of the author's intention. Here the lower-level discourse (in other words. say the proof of immortality in Book 10. whereas the only acceptable method of analysis is to relate the passages in which the hiccups are mentioned to the whole of the dialogue ± only then can an acceptable explanation be found (e. derision of Aristophanes). iii 16: `Indem Platon den Schluckenanfall er®ndet. Many interpreters connect this directly with an element in Plato's message (e. Assuming partial vertical relations between elements of the higher-level and lower-level communications invariably causes a shortcircuit in literary analysis: for example. by lifting a certain passage.g.g. which consists of Socrates' narrative of the conversation he has had the previous day plus a few accompanying facts.I N T R O D U C T I O N I.g. Now. that he relates it to an anonymous audience which he does not apostrophise and which does not react. the only link between the (higher-level) communication within the text and the (lower-level) communication of author and public is the fact that (higher-level) textual discourse plus accompanying features together constitute the (lower-level) discourse of the author. that it is Socrates who relates it.g.1 levels are inadmissible and invalid. Platon.5. sind wir aufgefordert. ihn und seine Rede fuÈr einen Augenblick dort zu denken. the text. wo er sie eigentlich haÈ tte halten sollen'). If an author communicates a text containing discourse.71 or if one participant in textual discourse is identi®ed with the author. um den Aristophanes zum vierten statt zum dritten Redner zu machen. As an illustration of this principle. out of its context and promoting it to an independent part of the author's message. in this case to the status of Plato's best argument on the subject at the time he wrote the Republic. the communication of Plato and his readers) is identical to the content of the ®rst upper level. short-circuit is caused e. etc.

e. Whether or not the upper-level I is identi®ed with the poet is a problem belonging to the interpretation(s) by the audience of the poet's intention.1 or worse still. an author's indication. If within the text a person calls himself `I'. by simply identifying the literary character called Socrates with Plato. reactions etc. since there can hardly be said to be communication between the narrating Socrates and his silent audience. The Socrates who tells the story is to be kept strictly apart from the Socrates who argues with Thrasymachus and projects a city with Glaucon and Adimantus.) In 72 A classical case is the `poet's I'. The ego and the poet belong to di¨erent levels. his reactions and those of his partners. It stands to reason that the more important the ®rst-upper-level communication of a particular dialogue. which is the only thing which the poet is communicating. there is upper-level discourse. Now this (upper-level) I is part of the level content (i. This is not to ignore the problem by splitting up the poet and the I but to put the poet's I where it belongs: in the (contemporary) reception of his poetry. the scenic background) is identical to the discourse of the ®rst upper level (the report given by Socrates to his anonymous audience). 38 . the text). The communication between a poet and his public consists of a text. (In the Republic the distinction is therefore relatively trivial. a number of concomitant features (such as a title. or when the lower-level communicant refers to a higher-level communicant in the ®rst person singular. There is never a one-to-one relationship between one communicant of the higher-level discourse and one of the lower-level.72 So far we have illustrated the principle of identity of higher-level content to lower-level discourse and that of absence of other vertical relations only for the level of author and public. if recoverable. of the audience. even when they bear the same name or behave in the same way. The content of the second upper level (in the Republic: Socrates' conversations.5. it also applies to distinct levels within a text. the more imperative it is to distinguish between communications of the ®rst and second upper levels.I N T R O D U C T I O N I. a situational context) and the interpretations.

2(2)).1). attention should be paid to possible incongruences of homonymous characters in di¨erent levels. the necessity to distinguish between homonymous characters at di¨erent levels of discourse is probably nowhere more obvious than in the case of the Socrates ®gures in the Clitophon.3. It runs from 407a5 e gwÁ gaÂr to 410b4 a pei rhka and is neatly sandwiched between introductory conversation and Clitophon's summing-up and last appeal. Apelt: `Tritt den Sophisten Euthydemus und Dionysodorus gegenuÈ ber die Ironie des Sokrates mehr als ein Spiel der Klugheit auf. In the Clitophon the second upper level consists of Clitophon's description of Socrates' protreptic speech and his own reactions to it and his questioning of Socrates' companions and of Socrates himself. that the attack is directed at the literary character.). since an attack on the historical Socrates would have been e¨ective only if Socrates were 73 Such as the one contained in the following statement by O. 19.73 In all dialogues with more than two levels of discourse. 1. cf. not the historical person (section i.5. which (together with the content of the second upper level) constitute the ®rst upper level. these ®gures are characterised quite di¨erently even to the point of mutual exclusion (section i. also 21 n.2 Socrates In the Platonic corpus.5. failure to distinguish between the levels in which a character is introduced as a participant in discourse can lead to gross errors of interpretation. As I have indicated above. so zeigt sie sich von ihrer liebenswuÈrdigsten Seite gegen seinen Freund und Gaugenossen Kriton' (Platon Euthydemus (Hamburg 19222). Otherwise Apelt o¨ers a couple of useful remarks on the functioning of level distinction in Euthd. i . notably the Euthydemus.2 some dialogues. 39 .I N T R O D U C T I O N I. I may add here that the double character of Socrates is one of the main arguments for interpreting the Clitophon as a literary rather than a philosophical `pamphlet'. in other words.

ad loc. Usually the ®rst sentence of a dialogue74 contains a vocative form. thÁn QrasumaÂcou deÁ sunousi an uperepainoiÄ are unique in the Corpus Platonicum. 129±30. R. Exceptions are the Cratylus (where the vocative form is postponed for a few lines only). `sozusagen als halb Abwesender praÈsentiert'. 157) says that Socrates `is perhaps deliberately made sti¨ and formal'. It is in a way the most important argument inasmuch as it is wholly textimmanent ± the methodological requirement that the Clitophon should be explained from itself has been satis®ed.2(1). 75 Cf. 422. 76 J. PeriÁ a rethÄv.  riThe opening words (406a1±4) KleitojwÄ nta toÁ n A    stwnumou tiv hmiÄ n dihgeiÄ to e nagcov. PeriÁ dikai ou. Anredeformen (Uppsala 1958). the Symposium (where Apollodorus' audience is too unimportant to become individual ± it serves only to comment brie¯y on Apollodorus' character). which introduce Socrates' partner in the accusative case in a sentence which provides all the `scenic' background needed (the ambiguous hmiÄ n ± cf. Minos. Svennung.4. Ly. section i. 74 Excepting three-level dialogues where the audience of a narrator at the ®rst upper level remains silent: Chrm. 40 . Let us start with the Socrates of the ®rst upper level.75 The Hippias Maior and the Menexenus begin with names in the nominative.76 This is half-way between a normal opening and Socrates' words in our dialogue. ± is made explicit by Clitophon: moÂnw tugca nomen ontev 406a10). MuÈ ller. for his two rhÂseiv this statement is correct. the Ion (where toÁn  I wna cai rein is equivalent to a vocative) and the unauthentic dialogues Hipparchus.2 depicted with at least some consistency. W. oti Lusi ai dialego menov taÁv meÁ n metaÁ SwkraÂtouv diatribaÁv ye goi.. where Socrates starts ®ring questions immediately.5.I N T R O D U C T I O N I. who manifests himself in the opening words. Kurzdialoge. in his willingness to accept Clitophon's spiritual guidance and also in his silence at the end of Clitophon's speech. and which states at the same time the subject of the dialogue. C. Thesle¨ (Styles..

because there is su½cient justi®cation for the third person opening from the text itself.2(4)) and it gives (as several scholars note) a highly formal character to Socrates' words. took it over verbatim) is ingenious but tiv hmiÄ n dihgeiÄ to e nagcov is inexplicable on these lines. among others. . As È ber Platon's Kleitophon'. cf. 7±8: KleitojwÄnta . of course the two are compatible. the disjunction of the proleptic accusative and the oti clause. 14). A (possibly) formal trait is the fact that its two main clauses are nominal.4. It appears from Clitophon's answer that he takes Socrates' formality as dissimulated pique (dhÄ lov ei memjoÂmenov me n moi. Quite apart from the indirect address there is the addition of the father's name.2 The purpose of this singular exordium77 is twofold: it states the subject-matter of the dialogue right at the beginning (as we have seen. prospoiou menov deÁ mhdeÁ n jronti zein 406a8±9).I N T R O D U C T I O N I. . The formality of the opening in itself could point to various states of mind. section i.). I ®nd Socrates' answer to Clitophon's words not so much formal as (ironically) humble. CPh 87 (1992) 95±109. This can be no argument against authenticity (so. . uperepaineiÄ is the theme as set by the teacher. and the ®rst person plural is abandoned. . Clitophon reYxem saw (`U torts the formality towards the end of the dialogue: qeÁ v toÁn KleitojwÄ nta omologouÄ nta (410d5). mh n is certainly not a very formal idiom (cf. 105±8. this is normal practice in the Short Dialogue. a llaÁ . my analysis in `The birth of a written language: An exercise in the pragmatics of the Greek sentence'.5. ad loc. Kleitophon'. 453). Schleiermacher. Clitophon had supposed that they concealed irritation. 78 The order proleptic accusative ± main clause ± subordinate clause is more typical of the written than of the spoken language. Lexiphanes. ± Pavlu's explanation (`Pseudopl. we are now invited to take them as 77 The ®rst parallel in Greek literature is Luc.78 the use of hmiÄ n for e moi and the substitution of metaÁ Swkra touv for met' e mouÄ. the pupil. 41 . obviously being too lazy to frame a beginning of its own. esp. the doublet askhÂsw kaiÁ diw xomai is ironically formal and kataÁ kraÂtov is overt irony.

Compare the analysis of 407a1±4 in section i. Even the ancients do not always read carefully: cf. 22).2(1). who was typical of Plato (section i. i na kaiÁ Melh twi le gwmen mhke q' h maÄ v a dikeiÄ n mhdeÁ a sebei av gra jesqai. . w v i kanwÄv hdh paraÁ souÄ memaqhko tav ta te eusebhÄ kaiÁ osia kaiÁ taÁ mhÁ (12e1±4). a good reason to make Socrates ironical: he had to make clear somehow that Clitophon's attack was not directed at the historical Socrates. on 406a2 Lusi ai dialego menov. 42 . Socrates will no longer be guilty of ase beia.3. His introduction of the ironical Socrates.5. namely the central ®gure of Socratic protreptic writings.170. but at a literary character.12±16 Terzaghi). therefore the irony is slightly out of tune. allaÁ kaiÁ 80 touÄ to KleitojwÄn kakwÄv oi etai (Dio 57d±58a ˆ 2. The same two assumptions which Socrates makes in the Clitophon are present here: (a) Euthyphro is able to teach Socrates. 81 Cf. 26a2±8. nevertheless. also the more explicit passage Ap. Synesius is partly right (though I think he rather missed Socrates' irony) in saying KleitojwÄn deÁ kaiÁ e loidoÂrhsen autoÁ n e n Lusi ou touÄ sojistouÄ 79 kaiÁ thÁn QrasumaÂcou sunousi an prouti mhse´ SwkraÂthv deÁ oudeÁ proÁ v touÄto parwÂxunto. A good parallel to these words of Socrates' is his invitation to Euthyphro: PeirwÄ dhÁ kaiÁ suÁ e meÁ outw dida xai .2 announcing Socrates' familiar deference towards his alleged betters. 406a8±9 prospoiouÂmenov deÁ mhdeÁ n jronti zein. . 80 kai probably determines the complete sentence: `that is precisely where Clitophon understands him wrongly' ± the reference is to Clit.2 n.3. There is a di¨erence in situation: Euthyphro had been pompous and patronising all along.81 I think that the author had.I N T R O D U C T I O N I. and Socrates had used his irony towards him almost from the beginning (compare the sentence just quoted with 5a3±9). at the ®rst upper level shows that he was aware 79 In Lysias' house? No. (b) having learned what things are eusebhÄ. Here Clitophon had said nothing that showed a pretension that he could make Socrates a better man (wjeleiÄ n). Comm.

3. Clitophon's misunderstanding of Socrates' mood is functional: it appears at once that Socrates is not irritated. Socrates' silence at the end of the dialogue has been explained in another connection (section i.2(3)).5. His choice of the Short Dialogue certainly did not hamper him in clarifying the intention of this dialogue through an ably managed.I N T R O D U C T I O N I. After the prooemium.82 Within the limits set by the dialogue as a literary genre. the formality of Socrates' ®rst words can be better understood now: Clitophon's faulty diagnosis dhÄlov ei memjo menov me n moi had to be based on a opening which admitted of this interpretation but which also left room for the subsequent appearance of the Platonic Socrates. this was a desirable side-e¨ect. 43 . characterisation. and 82 It is immaterial whether or not the author thought that the literary character Socrates as found in Plato was a true image of the historical ®gure. this was the best he could do. as he thought him to be. Socrates. better. If this analysis is right. Clitophon's role resembles that of the `poet's I'.2 of the existence of another. The consequence of this silence is that the Platonic Socrates of the ®rst upper level is hinted at in the prooemium and vanishes from sight after 407a4. As the author wanted to criticise a non-Platonic Socrates by means of Platonic methods of argument. Personally I am not even sure Plato thought so. his description of Socrates' speech and his own reactions. if sketchy. It also means that the content of the second upper level has virtually ceased to function as discourse in the ®rst upper level since one of the two communicants of the ®rst upper level is fading away into a `dummy' element. by constructing this misunderstanding the author draws extra attention towards the ironical Socrates and in doing so provides an extra key to the intention of the Clitophon. the author constructed the prooemium very carefully indeed. In its turn.

as for the readers of the dialogue to interpret them ± in other words. One must also bear in mind that the author wishes to point out the danger of protreptic literature ± a Gorgianic Socrates would lose contact with reality and would therefore become less dangerous. nn. for example. Though the opening words poiÄ je resqe. which is quoted in full. The most un-Platonic feature of this Socrates is not his exhorting others instead of questioning them ± exhorting 44 . But are we justi®ed in expecting burlesque? Plato's parodies ran always true to nature (a signal example is. after the prooemium Clitophon becomes a narrator similar to Socrates in. On the second upper level. a genre to which protreptic speeches did not belong (cf. the character called Socrates remains rather sketchy and abstract (like the ®rst upperlevel namesake) for all the space devoted to him. of course. on 407a6 pollaÂkiv) delivers a certain exhortatory speech. 109 and 111). is destined not so much for the ®rst upper-level Socrates to react on them. Stylistically he is left uncharacterised. w nqrwpoi. Thesle¨ (Styles. Comm.2 of his subsequent questioning. even in the protreptic speech. that he gave a de®nition of the result of justice which proved wrong ± it is hinted that he repeatedly failed in that way (410b3±4 tauÄta deÁ ouc apax oudeÁ diÁ v a llaÁ poluÁn dhÁ upomei nav croÂnon). although there certainly are quite a few rhetorical features in the larger sense of the word (cf.5. the alleged speech by Lysias in the Phaedrus) and if. furthermore. no isocola or rhetorical antitheses). say. Socrates' ®rst speech in the Phaedrus is full of burlesque whereas this one is not. we must recall that the former is a parody of an epideictic speech. on 407b1±e2). it is true that there are hardly any Gorgianisms (no poeticisms after poiÄ je resqe.I N T R O D U C T I O N I. We learn that he repeatedly (cf. as well as other similar speeches (408b6 e te roiv toiouÂtoiv). Comm. 157±8) ®nds no trace of `conscious burlesque' in it. refute his statement in its absolute form. the Republic.

151. de Strycker±Slings. 102±3 and n. In the Apolog y.85 We do ®nd it occasionally in other Socratic literature: Xenophon reports two exhortatory speeches addressed to an audience of more than one person: 1.I N T R O D U C T I O N I. Rutherford's interpretation fails to account for the fact that Socrates' exhortatory speeches are addressed to crowds. cf. though Plato rarely depicts him doing it (cf. 100) that the Socrates of the Clitophon is the Platonic Socrates throughout.7 are examples of speeches addressed to his suno ntev.3)83 ± nor this exhortatory speech in itself (as we shall see. 1. Meyer.4 and 1. 1. Plato's comment on the elenctic Socrates (see section ii.6.4.1 e piskeyw meqa ei ti prouÂbibaze le gwn ei v tau thn toiaÂde´ w andrev ktl. 84 The position of te proves that otwi depends from parakeleuoÂmenov as well as from e ndeiknuÂmenov.5. Adam ad loc. 1. 107.84 36c5 e piceirwÄ n e kaston umwÄn pei qein).14). cf..86 but even there. all the elements of the exhortations reported in the Apolog y return in the speech in the Clitophon.2. 1. 85 Cf. in principle ± of the protreptic Socrates in the Clitophon.e. so that the o twi clause is a restrictive apposition to umiÄ n (cf.1 and 2): what is really unPlatonic is his addressing this speech to a crowd. Gigon. Apologie. one might therefore accept Rutherford's claim (Art of Plato. But among other things. 151. the protreptic Socrates is. 1.5.5 (note w andrev) and 1. Plato takes care to stress that the exhortations are directed at individuals (cf.3.7. 83 In principle. Memorabilien I.3. 133. proagageiÄ n: cf.7. cf.19 and 1.5 and 1.1 e piskeywÂmeqa deÁ ei kai alazonei av apotre pwn touÁ v suno ntav a rethÄv e pimeleiÄ sqai proe trepen.1 Xenophon states that Socrates was capable of proagageiÄ n as well as protre yasqai not only through his elenctic questioning (e rwtwÄn h legcen) but also in view of a le gwn sunhme reue toiÄ v sundiatri bousi. 86 That these speeches are exhortatory appears from their position in Book 1: at 1.3). I suggest. section ii. the audience consists of his companions. and the same could be true ± again.5 e moiÁ meÁ n oun e do kei kaiÁ touÄ a lazoneu esqai apotre pein touÁ v suno ntav toia de dialego menov. Apolog y.4.3. 29d5±6 umiÄ n parakeleuoÂmenoÂv te kaiÁ e ndeiknuÂmenov o twi an a eiÁ e ntugcaÂnw umwÄ n.2 others is one of the two aspects of Socrates' jilosojeiÄ n mentioned in the Apolog y (29d5). 1.). In the Apolog y. I disagree with the analysis of Gigon 45 . sections ii.7.6 are examples of e legcov as a means of making better (i. Th.

Maier. and with that of Erbse. to Socrates the historical ®gure. L. section ii. `Architektonik'. Rossetti. QP 3 (1974) 145±57. is from this point of view in complete harmony with it (see sections ii. 87 So especially H. w v d' autwv par' oiv oudeÁ n proshÂkei. and misplaced irritation at that. Therapeia (Chapel Hill 1958). That Socrates is treated ironically by one (Memorabilien I.I N T R O D U C T I O N I. 121. the written lo gov addresses itself to a crowd consisting both of color che sanno and of those who have no business with it (omoi wv paraÁ toiÄ v e paiË ousin. the literary character appearing in the works of Plato. What at ®rst sight appears to be a gross deviation from Plato's literary practice. the Athenian philosopher who died 399 bce. on 407a6 e xeplhttoÂmhn akouÂwn).5.5.3 If my interpretation of the Clitophon as an attack on protreptic writings is correct.3.4±5 as illustrations of `belehren'. 323±7.6).1. Xenophon and others.4±2. Cushman. 46 .4. and o¨ers to talk the matter out. i .6±7 as illustrations of elenchos. 296±305 and passim. on the value of the Clit. Cf. 1. When accosted with what he takes as irritation. see Comm. Clitophon explains his position quietly. I do not intend to go beyond Socrates. `QeraÂpeia in the minor Socratics'. More recently: R.1 as examples of Socrates' (protreptic±didactic) le gein throughout. Phaedrus 275e1± 2). 151). When Socrates reacts with irony.1. for the historical ®gure 286±7.3 Clitophon The ®rst upper-level character of this name is more easily grasped than his opponent Socrates is at this level. The Clitophon is so exclusively dependent on the former that the interesting question whether the latter was indeed at times an exhorter towards virtue87 is irrelevant for its interpretation. 3±29. this remarkable feature is understandable: like Socrates.5. ii. who interprets 1. who takes 1. tells Socrates to his face that his attitude is taken by him as irritation. Sokrates. this young man attacks him immediately with his own weapon (though he is more obviously ironical than Socrates.

(c) some of Clitophon's statements in the report do belong to the ®rst upper level. also in the case of the two anonymous companions of Socrates who volunteer an answer to questions of Clitophon: 409a4 o dhÁ dokwÄ n autwÄ n e rrwmene statov ei nai and 409d3±4 ov dhÁ komyoÂtata e doxen eipeiÄ n (cf.3 who is obviously younger is pretty unique. ad loc. in the second only a couple of stylistic markers indicate it (cf. on 408b5±c4). wnqrwpoi.I N T R O D U C T I O N I. 408c1±4. The summing-up and last appeal are in deadly earnest. 90 The epithet is more or less traditional. in the third. on 408a4±b5). Comm. Bertini (457±8). I detect irony only in the introduction of Socrates' companions (408c5±7 twÄn h likiwtwÄn te kaiÁ sunepiqumhtwÄn h e tai rwn swÄn h opwv deiÄ proÁ v seÁ periÁ autwÄ n toÁ toiouÄton onoma zein).5. 89 This belongs to the second upper level. 407e4. I use it because a young man would be more interested in the relative values of sunousi ai of Socrates and Thrasymachus than an older one. on 407e4). In the rest of Clitophon's report. cf.88 Irony at the same time constitutes the framework of the reported speech. There is no irony elsewhere in the report (with the possible exception of 410b5 kaÂllist' anqrw pwn). Comm. This ¯uctuation goes parallel with the degree of burlesque in the speech itself: rather heavy in the ®rst sentence (poiÄ je resqe. who wrongly consider this a mark of inauthenticity.) practically absent from the last sentence of the oratio recta and the ®rst one of the oratio obliqua. and also because in the 47 . cf. interrupted in 407e3±4 and terminated in 408b5±c4: the ®rst of these sentences is full of irony. Comm. it is clearly present (less so than in the ®rst but decidedly more than in the second sentence. (b) the role of irony at both levels for both characters is crucial for the interpretation. but I treat it here because (a) it would be pedantic to sever Clitophon's reported (second upperlevel) reaction to Socrates from his (®rst upper-level) reaction in direct speech. present to some extent in the closing sentence of the whole speech (cf. Steinhart (51). The consistency of this character (an unabashed young90 88 Cf.).89 The report is announced in 407a5±b1.

91 Note the use of poreuÂomai pro v tina instead of the technical joitaÂw: Clitophon is apparently not a regular pupil. not the villain.91 BruÈ nnecke (`Kleitophon wider Sokrates'. which means that Clitophon is at the moment a pupil or visitor of Thrasymachus.3 man. 463 n. Though one should read poreuÂomai and not poreu somai at 410c7 (cf.). the proposition is of a doubtful value (literary characters have no existence outside a literary text).I N T R O D U C T I O N I. (BruÈ nnecke's second question suggests too much. schon der Nebengedanke ``Thrasymmachos [sic] ist moÈglicherweise unfaÈ hig''? Erscheint nicht damit der Ausblick auf etwas HoÈ heres. 38) asks: `Liegt nicht in dem ``kaiÁ '' usw. The allusion is reinforced by aporwÄn (410c8). auf die Akademie?' The answer to the ®rst question is a½rmative (despite Souilhe 's scepticism. If this is meant to imply that Clitophon's audacity is due to the in¯uence of Thrasymachus. which suggests dissatisfaction with Thrasymachus and the unnamed others. 180 n. who stays condemned. the reader knows that Clitophon will ®nd no knowledge (and a couple of very objectionable ideas) about justice in Thrasymachus' teaching. an allusion that he is not walking towards his doom is necessary.5. The `Ausblick auf etwas HoÈ heres' is in fact provided by Clitophon himself. of the dialogue. The Academy could have been symbolised only by Socrates. hits back hard) is the more amazing since it is fully conditioned by the development of Socrates' character at this level as set out in the previous section. the words kaiÁ allose suggest that he is not going to be an orthodox disciple.5. who.5 ± that the Clitophon was written after Republic 1. 3).) Republic the younger pair Polemarchus±Clitophon matches the older pair Socrates±Thrasymachus. when struck. Some scholars draw a parallel between Clitophon's behaviour here and that of Thrasymachus in Republic 1. As Clitophon is obviously the hero. If one assumes ± reasons for which shall be given below and in section ii. ad loc. 48 .

. Let us now pass to the second upper-level Clitophon. But there is no shyness in Clitophon's report of his discussions with Socrates (410a7±b4 apei rhka) nor in the rest of his appeal and his ®nal conclusion (d5±e8). not to suggest that the statements found in that literature are nonsense. esp.4. We have seen (section i.6. which is rather too humble. 94 It is clear that Clitophon and Lysias had not been alone. Clitophon gets cold feet and starts stammering.2. and when the author keeps irony and earnest wide apart from each other. who appears in two di¨erent settings. cf. 93 For another way of showing agreement with statements ®rst derided. can be replaced by a (more characteristic) command. sections ii.3 In harmony with this dissatisfaction is Clitophon's `last appeal' (410c8±e5). I have for some time entertained the thought that the two anacolutha 410b6±c5 and c8±d5 have quite a di¨erent characterising function: entering the stage of summing up his criticism to Socrates' face. cf. . the appeal is quite sincere. allwn d6) the irony of the report (e xeplhttoÂmhn a kouÂwn. cf. but there is no more irony. 410e4±5 proÁv Lusi an kaiÁ proÁv touÁ v allouv: someone had to report Clitophon's criticism to Socrates. 410d4±5 kaiÁ nuÄn dhÁ tautoÁ n gigne sqw92). 409a3. irony is absent from Clitophon's ®nal appraisal of Socrates' exhortation (410d5±e1): when again he shows his agreement with it (especially with its ®rst two parts.4. The anacoluthon contained in 410c8±d5 brings the advantage that the logical sequel (`I would gladly return to you'). Yet on the level with which we are now concerned. pampoÂlloiv kaiÁ pagkaÂlwv legome noiv) is absent. ii.2(1)) that on the lower level (that of the author framing a text) this appeal is destined to remain fruitless. In a conversation with Lysias94 he is said to have blamed Socrates and 92 For the imperative. The purpose of the Clitophon (to translate this ®rst-upper-level feature in terms of the lower level) is to deride protreptic Socratic literature.3). he does manage to make it clear93 (see also section ii. witness twÄn . It was necessary for the author to insert 49 .5. d2 as discussed below. qaumastwÄv wv e painwÄ . This appraisal of these statements had to be made clear somehow. Likewise. Clitophon's words remain crisp (cf.I N T R O D U C T I O N I.3.

cf.). 95 Strictly speaking this is not the second but the third upper level: tiv h miÄ n dihgeiÄ to represents the second.: his name is mentioned in the ®rst announcement of [a] participants. of these. The correction causes the dialogue to concentrate on Socrates. the ®rst two are remarkable inasmuch as they are used as self-contained sentences following a question (this is rare in Plato's authentic works. but the accents are di¨erent.96 Yet it is signi®cant that he has been mentioned: the reader will interpret Clitophon's philosophical career in connection with Thrasymachus' name (cf. 409a7±c1. Cf. below). 50 . lege sqw. by mentioning him here Plato causes him to be latently present for the reader throughout the text. Clitophon reappears at this level when questioning Socrates' companions and Socrates himself. . Thrasymachus vanishes from sight to reappear only towards the end of the dialogue. His character resembles the ®rst upper-level Clitophon. iii 4.I N T R O D U C T I O N I.3 praised Thrasymachus. FriedlaÈnder. which at the same time echoes Socrates' apostrophe poiÄ je resqe. ad loc. Irony is present in his apostrophe w be ltistoi . ei pe . Platon. the crisp tone is more marked. Though Alcibiades makes his appearance only late in the dialogue. wnqrwpoi (cf. ad loc. .95 This is subsequently corrected: Clitophon had both praise and blame for Socrates. u meiÄ v (408d1±2). On the other hand. All his questions (408d1±409a3.).5. perhaps in the plural a podeco meqa (408d2. Comm. ad 409a3 lege sqw). [b] scene and [c]  ga qwnov sunousi an kaiÁ subject of the dialogue in 172a7±b3 thÁn A  lkibiaÂdou kaiÁ twÄn a llwn [a] twÄn to te e n twÄ i SwkraÂtouv kaiÁ A sundei pnwi paragenome nwn [b] periÁ twÄ n e rwtikwÄn lo gwn ti nev h san [c]. 96 One may compare formally Alcibiades in Smp. c4±d2) end up in an imperative: lege sqw. but hardly anywhere else. There is a note of impatience towards the pupils the detail that the criticism was uttered in a private conversation: without the words Lusi ai dialegoÂmenov (406a2) the meaning could be that Clitophon uttered his evaluation of Socrates and Thrasymachus before a large audience. cf.

Analogy is used as profusely by the second-upper-level Clitophon (apart from the one already mentioned. section ii. At the lower level.) Clitophon makes a curious detour by ®rst putting forward a dilemma: is exhortation to virtue an end in itself or is the pursuit of virtue the logical consequence of a completed exhortation? The phrasing of the ®rst horn of the dilemma may help in explaining why it has been put forward at all: hmiÄ n paraÁ pa nta dhÁ toÁ n bi on e rgon touÄ t ' e stai. The way in which Clitophon manages the questioning is stated by himself to be after Socrates' manner (e panhrwÂtwn. to wit the use of analogies as steps in e pagwgh . ad loc. c7±d1.6). 51 . the reader is supposed to infer from the signal kataÁ se that Clitophon will play the part which in other dialogues is normally assigned to Socrates ± this expectation in the reader (which will be ful®lled completely) is strengthened by the use of a curious device. Styles. `leading on' (cf. as applied to the e rga of carpentry) as by his ®rst-upper-level namesake97 (cf.5. The participle upotei nwn. means `demonstrating'. cf. Only in this way can the words kataÁ se be fully understood. 409b1±6. and refers to the analogy from the care of the body (408e3±409a2). 157.I N T R O D U C T I O N I. It is not just Socratic questioning which Clitophon has in mind. Instead of asking right away what he asks at 408e2±3 (pwÄv arcesqai deiÄ n jamen dikaiosuÂnhv pe ri maqh sewv.).3 which can be expected from someone who is none too reverent towards the master. de on etc. but a very speci®c feature of it. which has not been understood hitherto. 408c9±d1). punqanoÂmenov ti v o metaÁ tauÄt ' ei h loÂgov. accorded by Thesle¨. touÁ v 97 It is probably this feature which earned Clitophon the epithet `verbose'. kaiÁ kataÁ seÁ troÂpon tinaÁ u potei nwn autoiÄ v. the meaning of sumje ron. the distinction of e rgon and di dagma in medicine and carpentry.

ei v pollhÂn ge apori an wv e oiken aji kesqe. therefore the result of soji a is soji a. which is characteristic of that dialogue (as of the Charmides): the only a gaqoÂn is soji a. h e ti le gwmen oti a llouv poih sousin.3. In that place the regress is highly functional as a means of reaching aporia (signalled as in the Euthydemus: 409e10 O te dhÁ e ntauÄqa hmen touÄ loÂgou. oudamouÄ hmiÄ n jai nontai (292d8±e1). 52 .4. This suggests a regressus ad in®nitum (henceforth to be called `circular regress'). and more particularly. Goldschmidt. a circular regress is logically out of place. Here. In the Euthydemus the phrasing is a consequence of the application to men of the circular regress.3. 292e6±7. It may be argued. which is formally recognised as such (nhÁ toÁ n Di a w SwÂkratev. ii.I N T R O D U C T I O N I. the circular regress found in the Euthydemus in Socrates' second conversation with Clinias as carried on with Crito: oi ti e sontai hmiÄ n agaqoiÁ kaiÁ ti crh simoi. the result of soji a must be an a gaqoÂn (in this case agaqouÁv poieiÄ n). o ti de pote agaqoi eisin.4. that this circular regress has a very good artistic function: it introduces Clitophon's questioning of Socrates' pupils. Socrates' protreptikoÂv had been an exhortation to something (stated in so many words: 408d2±3 thÁn SwkraÂtouv protrophÁn h mwÄ n e p' a rethÂn). however.4. kaiÁ e kei nouv au e te rouv (408d5±6). just as the regress terminates it.3). A similar circular regress is found further on in the Clitophon: 410a2 peridedra mhken eiv tautoÁn o loÂgov toiÄ v prw toiv. sections ii. it will be examined in section ii.98 The circular regress results in an aporia.3 mhÂpw protetramme nouv protre pein.5. 79± 81. Dialogues. a complete regress is not reached unless it is proved (quod non) that the end of an exhortation is exhortation itself. Though this may seem a curious way of using a 98 See also R. oi deÁ alloi e keiÄ noi a llouv. a porouÄntev). 505b5±c4 and Adam's note.

iii 337. As we saw. where the more normal questioning form (cf. the use of circular regress is more speci®cally Platonic. This may even be narrowed down to `a typically Platonic way': whereas the use of analogy is generally Socratic. and they dissolve into nothingness as soon as the author has no more use for them. 100 Hence probably the statement (409b1) i atrikh pou tiv le getai te cnh. in the discussion about friendship as the result of justice (which would have required direct speech from Clitophon and indirect speech from the opponent). e.99 Clitophon's respondents at this level are hardly more than an anonymous group. 53 .3 device of Platonic elenchos for what could be called a programmatic purpose. both subject-matter (protreptic) and formulation recall the passage from the Euthydemus. though they are most highly valued by Socrates himself (408c8). it makes clear once more that Clitophon is going to behave in a typically Socratic way. they are introduced with some irony. whereas his opponents are reported in the most formal oratio obliqua with a preference for optatives and in®nitives. HGPh.I N T R O D U C T I O N I. The author underlines this status of Clitophon and his `dummy' partners in the manner of reporting.g. in Oec. and in passing shows that he has a good idea of what the Platonic method is all about. Besides. a rti gaÁr dhÁ e jhn e gwÁ katamanqa nw hi me e phrw thsav e kasta´ a gwn gaÂr me di' w n e gwÁ e pi stamai. o moia tou toiv e pideiknuÁv a ou k e no mizon e pi stasqai a napei qeiv oimai w v kaiÁ tauÄ ta e pi stamai (cf. But compared with him. the confrontation is avoided by leaving unmen99 Xenophon occasionally makes use of Platonic traits.5. The author takes the trouble to make it clear that this Clitophon is the real Socrates.100 Typically. There is no dialogue. who overlooks the fact that the answer to these questions is negative). ad loc.) would smack of dialogue.15 ara e jhn w  Isco mace h e rwÂthsiv didaskali a e stin. Clitophon's own words are in oratio recta throughout. Some of them acquire substance for the sake of the discussion. 19. Guthrie. our author has a far better understanding of Plato's handling of the dialogue.

but (a) it marks the end of a major part of Clitophon's report. among those present at 328b7.I N T R O D U C T I O N I. cf. e5 e rwtwÂmenov.2 and 3). We 54 . 6 h nagka zonto). 9 e panerwtwÂmenov. Of course. Clitophon would have no need to report in full a conversation which Socrates probably remembered as well (though the next sentence implies repeated conversations). In harmony with the general tendency of the report is the curious twist in the sentence eipe n moi . Yet by reporting it as drily as possible he contrives a minimum loss of face on Socrates' part.5. As was the case with the last pupil. sune baine . but here the whole report of the discussion has been cut out. autwÄi.3 tioned who asked the questions (hence 409d6 e rwtwÂmenov. He is named there. so that the discussion can be reported in oratio obliqua throughout. . Having examined the functioning of the literary character Clitophon at both upper levels in this text we may proceed to the only other Platonic text in which he plays a role. . At the author/reader level the shortness of the whole episode provokes a concomitant reaction: attention is focused more on the shortcomings of the companions. ad loc. the Republic.4. it is not stated directly that the questions are asked by Clitophon. with his father's name. and the word order seems to suggest that he was in the company of Thrasymachus and one Charmantides of Paeania. . as here. dikaiosu nhn (409a5±6. 410a1) which is speaking. . only the outcome (and the principal argument) are mentioned in the most impersonal form possible: 410b1±3 usteron deÁ e ja nh blaÂptein ge oude pote o di kaiov oude na´ pa nta gaÁr e p' w jeliÂai pa ntav draÄn.): a oti clause instead of the accusative and in®nitive after eipe n moi would already have bordered on direct speech and broken the pattern of anonymity. The situation is di¨erent in the report of Clitophon's discussion with Socrates. (b) it is the group as a whole (oi paroÂntev. The de®nitive refutation is in direct speech (410a3±6). less on those of Socrates himself (see also sections ii.

(b) it gives Plato the opportunity to o¨er two di¨erent solutions for Thrasymachus without making the situation dramatically improbable (Plato could not very well have made Thrasymachus utter both possibilities himself ). Consequently. (a) It glosses over the argument once again. so that sometimes toÁ touÄ krei ttonov asu mjoron (339e7) is just.I N T R O D U C T I O N I. Clitophon's function is wholly explicable from the needs of the dialogue itself ± his role in the Clitophon does not in any way illuminate his role in the Republic. In the following skirmish the gist of Socrates' remark is repeated and Clitophon tries to save Thrasymachus by suggesting that by `the stronger's interest' Thrasymachus had really meant `that which the stronger believes to be in his own interest' (340a1±b8). The author's intention seems clear enough: Socratic protreptic is the more dangerous as it drives honest and intelligent young people into the arms of false teachers like Thrasymachus. in this 55 . and he expects much from Thrasymachus. towards which Socrates is constantly exhorting others. On the other hand. Thrasymachus rejects this amendment and ®nds another way out: the ruler (ˆ stronger). Clitophon was disappointed when he turned towards Socrates' pupils and Socrates. and for the last time.5. does not err ± when a ruler errs. qua ruler. Wishing to be instructed in justice. Polemarchus breaks in with (unnecessary) assent and Clitophon reacts immediately with an unkind personal remark. It is clear that the Polemarchus± Clitophon episode serves two purposes. The reader who is acquainted with the Republic knows that Thrasymachus will corrupt him beyond healing. Socrates has forced Thrasymachus to admit that leaders sometimes err in perceiving what is for their own bene®t. (The function. the choice of Clitophon as the main character in our dialogue is certainly more illuminating if Republic 1 is taken into consideration. he is strictly speaking not a ruler (340c6±341a4).3 meet him again. after the ®rst phase of the debate of Socrates and Thrasymachus.

Ostwald. Rhodes on Ath. who were radicals (oligarci av e pequ moun). 29. From Popular Sovereignty to the Sovereignty of Law: Law. M. those not united in e taireiÄ ai who strove after the pa triov politei a. From a dramatic point of view. Society. Levi and P. 369±72. if indeed Clitophon. of the words kaiÁ allose.101 A second time Clitophon is mentioned in the constitutional struggle between the surrender of Athens (April 404) and the institution of the Thirty (September 404). As we shall see (section ii.Pol. J. The story is told by Aristotle (Ath. the Pythodorus decree is alluded to by Thucydides (8. A number of politicians of this conviction are mentioned. AC 34 (1965) 464±83. M.3) opposes dhmotikoi and gnwÂrimoi (oligarchs) and among the latter distinguishes on the one side members of e taireiÄ ai and returned fugitives. son of Aristonymus.) If the Clitophon predates Republic 1.5).5. the order Republic 1 ± Clitophon makes much more sense than the reverse one. Finally a few words on the historical Clitophon. Aristotle (34.67. Pol. and Politics in Fifth-Century Athens (Berkeley±Los Angeles 1986). Clitophon thought (or professed to think) that these were not democratic. is identical to the Athenian politician who in 411 (probably early May) proposed that a committee consisting of the ten proÂbouloi and twenty others (such a committee had already been proposed by Pythodorus) should not only bring in proposals for the safety of Athens but should also examine Clisthenes' paÂtrioi no moi. `L'amendement de Clitophon'. among them are Anytus (who was subsequently banished together with Thrasybulus. the character of Clitophon in the latter is not in¯uenced by the former. on the other. a comparison of the de®nitions of the e rgon of justice in the Clitophon with statements in Republic 1 points the same way. 410c7. 56 . A.I N T R O D U C T I O N I.3). Bibauw. if it is the other way round. and 101 Cf. loc. J.3 respect.1). and of Clitophon's last appeal has been treated above. there certainly is an important extra point in the message of the Clitophon. cit.

the Clitophon.5. there is no indication in our dialogue of Socrates' questioning the laws of the city. Remarkably enough. Ostwald. 469±72.' The point should certainly be taken by future commentators of Republic 1. Orwin (`Case against Socrates'. but I do not see that it is relevant to its derivative. Phormisios. That is another way of formulating his obvious enmity toward Socrates. is classed by Aristophanes (with one Megaenetus) as disciple of Aeschylus (965). For the latter. 743±4) raises the interesting point that Clitophon's track-record as an Athenian politician ®ts his role in Republic 1: there. But the vicissitudes of the war may have united in 404 politicians who were diametrically opposed in the spring of 405.I N T R O D U C T I O N I. . the questioner par excellence of the authority of the laws of the city. `Es geht also die Schilderung des Euripides nicht auf den Unterschied in der politischen Richtung. 57 . it is plainly false to speak of Clitophon's `enmity' towards Socrates ± also. .Pol. op. 282). 34. `He asserts what Socrates gets Thrasymachos to deny. who is named (Arist. In never wavering from his interested attachment to legal justice. . he is the sole character in the Republic who stands ®rst and last for the city as it is. Aristophanes' `FroÈ sche' (Vienna 19542). that the will of the rulers is beyond appeal. because the association 102 Cf. he is the representative of legal justice. The question whether or not the historical Clitophon had anything to do with Thrasymachus is immaterial for the interpretation of the Clitophon. (n. cit.3 with the latter restored democracy in 403) and Clitophon.3) as another member of Theramenes' party. Radermacher. where Euripides claims as his disciples Clitophon and `Theramenes the smart' (967). the head of this party (proeisthÂkei deÁ maÂlista) was Theramenes. 26 Scheibe) as involved in a process may or may not be identical to the politician.102 The Clitophon named in a fragment of Lysias' speech  UpeÁ r DexiouÄ a postasi ou (fr. Clitophon's association with Theramenes in this period is also hinted at in the Frogs. sondern vielmehr in Character und in der Lebensanschauung' (L. 101). Ath.

only Souilhe has given the matter any amount of thought (167±9).). especially 954±8. including Theramenes himself. 74. I am forced to sum up the re103 His answer is a½rmative. . line 14). II MEANING AND AUTHENTICITY In this second section of the Introduction I shall try ®rst to draw a picture of philosophical protreptic in the fourth century bce (ii.103 We know too little about the ages of Clitophon (born before 430) and Thrasymachus (dates quite uncertain but he is probably referred to as a teacher of rhetoric in Aristophanes' DaitalhÄv (fr. secondly to confront the results with Socrates' speech reported in the Clitophon (ii. and the speech is probably epideictic. 58 . this seems perfectly possible. HGPh. `De exhortationum . 34. Ostwald. or perhaps. As Thrasymachus is usually associated in our sources with members of the older generation of sophists (Protagoras. 367). Given the lack of an up-to-date comprehensive treatment of philosophical protreptic in antiquity (the only monograph on this subject is Hartlich's dissertation from 1889.3. for two main reasons: (1) A number of Clitophon's associates named at Arist. historia').2).8 K. For Clitophon. Gorgias). which was produced in 427) to pass judgement on the value of the datum (absent from the Republic) that Clitophon was young enough to be the pupil of Thrasymachus. it was written for a client (cf.1). Ath. the context of the Frogs passage. proved himself vehemently opposed to the sophists. iii 296.±A.3 belongs to the data of Republic 1.Pol. 205. as some scholars assume with too much con®dence. makes this plausible. Prodicus. Artium scriptores. Anytus. but it is rather a vague slogan.I N T R O D U C T I O N I. (2) The pa triov politei a is advocated in a fragment of Thrasymachus (78 b 1 ˆ Radermacher. Commentators on the Republic have not bothered to pursue the question. yet it must be borne in mind that at least one of Clitophon's fellow-oligarchs. . are known to have been interested in rhetoric and to have had connections with sophists.5.

4). The conclusions reached in this and the previous sections will be summed up in order to determine the author's intention (ii. considerations of space forbid any kind of extensive argumentation.6). especially the complex of interrogation and aporia which he calls e legcov. It may be thought that in thus concentrating on protreptic literature I am begging the question: one might expect a proof of my thesis that the Socrates who is made the target of our dialogue is not the philosopher himself but more or less a symbol of protreptic. or if within the text one character endeavours to cause such a change in another character or characters.5.1 sults of my studies on this subject.2).2 of this Introduction. ii. As the Clitophon has been transmitted among the works of Plato. The statements about justice in our dialogue will be treated separately (ii. Thus.1.7).1. partly in the comparison of that work with protreptic literature which is contained in section ii. only then can the question of authenticity be settled (ii. we are entitled to call it a protreptic text because of Isocrates' announcement that he is going to send Philip 59 .3).1 De®nitions A text may be called protreptic if its design is to cause a change in the behaviour of those for whom it is destined. It has therefore been necessary to explore the use which is made of elenchos in the Clitophon (ii.1 Philosophical protreptic in the fourth century bce ii . it will be shown that Plato's alternative to protreptic is the dialogue. the attitude of that philosopher to protreptic will be studied next (ii.5). Isocrates' Philippus is an appeal to that king to bene®t the Greeks by uniting them and leading them against the Persians. The answer to this objection may be found partly in the analysis of the characterisation of Socrates in the Clitophon (section i.I N T R O D U C T I O N II.

Isocrates makes some disparaging remarks about philosophers who by exhorting others to virtue (twÄn e piÁ thÁn swjrosu nhn kaiÁ thÁn dikaiosu nhn prospoioume nwn protre pein) try to acquire pupils (e pagage sqai tinaÁ v . Schmeller's stimulating study Paulus und 60 . 282d1±2). . Similarly. eiv thÁn autwÄn omili an.84±5). again the text itself characterises the ®rst of these conversations as an `example of protreptic speech' (282d4±6). cf. or in which a character or characters are impelled to do so. in other words: philosophical protreptic in the wider sense includes all texts written by philosophers or inspired by philosophy which aim at a change of conduct in the readers or characters of these texts (usually in the ®eld of ethics).104 in the stricter sense by what is normally called 104 I am fully aware of the dangers inherent in this term. philosophical protreptic in the stricter sense denotes the texts which incite to the study of philosophy.1. a llaÁ peirasoÂmenoÂn se protre pein (5.3±4). The pseudo-Isocratean pamphlet Ad Demonicum refers to certain `protreptic speeches' which exhort people to (presumably) rhetoric (1.1 loÂgon . again. the distinction between philosophical protreptic in the wider and in the stricter senses corresponds to a well-established di¨erence in genre: in the wider sense. ouk e pi deixin poihso menon oud' e gkwmiasoÂmenon .17. . 15. it is represented mainly by the diatribe. In a stricter sense. I call `philosophical protreptic' all protreptic texts in the two senses de®ned above which belong to philosophical literature. 116). .I N T R O D U C T I O N II. Socrates' two conversations with Clinias reported in Plato's Euthydemus may be called protreptic: their aim is to convince Clinias of the necessity of caring for wisdom and virtue (278d3±5) and thereby to impel him to acquire them (cf. . . which are pointed out at length in Th. From the hellenistic period onwards. . `protreptic' is applied to texts which are intended to impel the readers to pursue a certain study.

330 b . For a recent attempt at delimitation of the term. historia'.28 ˆ 5. Hartlich.1 protreptic tout court. apology. The Euthydemus con®rms this: the wider and stricter senses are often juxtaposed (e. Inasmuch as in Socratic philosophy theory and practice (knowledge and right action) coincide. Schenkeveld. all texts which have a similar intention but in which these aims are achieved indirectly will be called `implicit protreptic'. M. ± a. Yet explicit exhortation to virtue or to speci®c virtues unquestionably existed. fr. E. `De exhortationum .807 W. 310. Eine vergleichende Stilinterpretation (MuÈ nster 1987) 1±54. 11 Schenkl). in: S.-H. in the ®rst conversation of Socrates and Clinias mentioned above.33. as there is of protreptic.. prove or convince by other methods that one must adopt a certain line of behaviour or pursue philosophy. 275a6 jilosojeiÄ n kaiÁ arethÄv e pimeleiÄ sqai ). . 61 .g. cf. the problem is of course that nearly every piece of non-technical philosophical prose is given this label by many modern scholars. . c. Thus.105 But in the fourth century bce this distinction does not obtain. witness the last passage of Isocrates quoted above: here the exhortations are to speci®c virtues (wider sense). 230±47. Another distinction which in the course of my study I have found pro®table is that between `explicit' and `implicit' protreptic (both in the wider and the stricter sense of `protreptic'). the blurring of this distinction is hardly surprising. 105 Though a diatribe by Epictetus is quoted by an author as late as Stobaeus as  Ek twÄn ArrianouÄ ProtreptikwÄn omiliwÄ n (4. Handbook of Classical Rhetoric in the Hellenistic Period. I call `explicit protreptic' all texts which purport to state. myth or other type of philosophical text can be designed to cause its readers or characters to change their moral conduct or to pursue philosophy. 230. Porter. there is no ancient theory of the diatribe. Furthermore. but their aim is stated at the same time to be to induce readers to the pursuit of philosophy (stricter sense). description of behaviour.1. D. Socrates starts from the e ndoxon that all men desire to be die `Diatribe'. bibliographical data in notes 134 and 135 to his p. Any argument. cf. 400 (Leiden 1997).I N T R O D U C T I O N II. d . `Philosophical prose'.

106 It is not always easy to draw the line between explicit and implicit protreptic. philosophise.I N T R O D U C T I O N II.3(6)±(8). the cleft between this group and true implicit protreptic is much deeper. aporia) belong to implicit rather than explicit exhortation. 204±13 analyses various philosophical texts from the post-classical age as (implicitly) protreptic.1.1.1 happy and via various (not always very convincing) steps arrives at the conclusion that in order to be happy one must try and acquire wisdom. n. Protreptiques. Cf. 107 `Aeschines' Miltiades'. 71±99). I consider this group of texts to belong to explicit protreptic: even if some of the means employed (interrogation. 62 . Having set himself o¨ from the writers of 106 Schenkeveld (cf. but the conclusion that philosophy is necessary in order to obtain happiness is here left to the reader. 305±6. 104). refutation. A ®nal distinction must be made between protreptic literature as a general way of in¯uencing the conduct of others and another type of literature which consists of a series of concrete rules of conduct. In this subgenre. or even to lead a reasonably decent life.e. This is explicit protreptic. in which Socrates' partner mends his ways and becomes a follower of Socrates. The dialogue ends with a conversion scene. i. It has been considered `protreptic' both in antiquity (Iamblichus) and in modern times (FestugieÁre. not only because it exempli®es the latter species but also because it makes the distinction itself. the analysis of philosophy as a preparation for death in the Phaedo (64a4± 69e5) is an example of implicit protreptic. a person is convinced by Socrates that his way of life is wrong or that he does not possess enough (philosophical) knowledge to reach the goals set by himself. The pseudo-Isocratean Ad Demonicum is a good case in point. This is especially the case in what I have elsewhere107 called `protreptic dialogue'. because eventually the change in conduct is depicted in the text. On the other hand. section ii.

± Paraenesis is the continuation in prose of the poetic genre known as u poqhÄkai. Kurke. on which cf. he is usually. uses `Protreptik' for what I call `protreptic in the stricter sense' and `ParaÈnese' for `protreptic in the wider sense'. `Pindar's sixth Pythian and the tradition of advice poetry'. I have not adopted this terminology because it causes unnecessary confusion: the distinction between protreptic and paraenesis (as found in Pseudo-Isocrates) is observed throughout antiquity (section ii.4. 85±107. concerned with explicit protreptic. 110 I shall not discuss statements on protroph and protre pein in the orators (notably Isocrates) and the theories of rhetoric (Aristotle. Only in an isolated case (Appendix i ) have I compared specimens of nonphilosophical protreptic in order to settle a point of detail ± even there. I shall henceforward call `protreptic' what is above de®ned as explicit protreptic.2 protreptic speeches in the manner quoted above.2 n.I N T R O D U C T I O N II. Anaximenes of Lampsacus).109 In order to avoid cumbersome circumlocutions. TAPhA 120 (1990). esp. but not always. `De exhortationum . historia'. The distinction is not between euroÂntev and gra yantev. Hartlich. quoting Harpocration paraÂklhsiv´ antiÁ touÄ protroph .108 but between paraÂklhsin (ˆ protreptic) and parai nesin (advice).3 touÁ v protreptikouÁv lo gouv suggra jousi. 63 . historia') the distinction made by Pseudo-Isocrates. I shall call the general type `protreptic'. the outcome is negative. . `De exhortationum . 164) and in many modern studies.110 ii. the author continues: Dio per h meiÄ v ou para klhsin euro ntev a llaÁ parai nesin graÂyantev me llome n soi sumbouleuÂein. Protreptik.2 Protreptic among the sophists Whereas no one can deny that some texts written by sophists are protreptic (for instance the fragments usually 108 Cf. The latter deal with protroph only as a part of forensic oratory and shed no light on Clit. 222. . . the concrete type `paraenesis'. 90±5. 109 Gaiser. L.5). 1. w n crhÁ touÁv newte rouv o re gesqai kaiÁ ti nwn e rgwn a pe cesqai kaiÁ poi oiv tisiÁ n a nqrwÂpoiv omileiÄ n kaiÁ pwÄ v toÁn e autwÄn bi on oikonomeiÄ n ( Following (with Hartlich. unless the context makes it obvious that the word is otherwise used. .

as it was the most current objection to the sophists' trade that areth cannot be taught (section ii. Many scholars acknowledge that we have hardly any data to go by. 35) is wrong in concluding that the distinction made there points to two di¨erent types of speeches (protreptic and instruction): an e pi deixiv is a specimen of the sort of instruction which awaits the prospective pupil (cf. Neue Studien zur Aristotelischen Rhetorik (Halle 1907). Gaiser (Protreptik. Tarrant ad loc.111 On the basis of a passage from the Euthydemus (274d4±275a3) Gaiser (Protreptik. 7) and it may have occurred in protreptic literature far earlier than that.2.1). 64 . Der Protreptikos des Poseidonios (Munich 1912). Dodds on Grg. First. Gaiser's analysis of Protagoras' myth and lo gov (Protagoras 320c8±328d2) as protreptic (Protreptik. the `epideictic' speech. Three possible types of sophistic protreptic deserve mention here.Ma. Prodicus and Protagoras are mentioned (and Hippias implied). But he overlooks the fact that proving this is said in so many words to be the `result of the same art' (274e3±5). (b) the speaker is the best teacher of it. O. 447a6.1. mentioned by Plato as typical of the sophists. GerhaÈusser. where Gorgias. explicit protreptic in the stricter sense. as a separated genre. esp. 2 E. 46±7) arrives at the following ®xed pattern of this type of `Werberede': (a) virtue can be taught. it will remain a question of conjecture whether the sophists inaugurated.3. 68±81).. Kraus. a speech in which the claim that areth can be taught is defended would be a possible type of sophistic protreptic. in other words: persuading (protreptic) is a byproduct of the normal way of demonstrating the sophist's art.2 referred to as the Anonymus Iamblichi). 112 The thesis that virtue can be taught was defended in Posidonius' Protreptici (fr.±K.. 38±40) is plausible enough.I N T R O D U C T I O N II. Secondly. Hp. 282b7±8.112 but there is no proof that such a defence should 111 Cf. yet they tend to regard the assumption that the sophists did indeed create such a genre at least as plausible. W.

1421b18±9 and 1425b36±7.). the author shares the position of Plato. (n. Hartlich. the words kaiÁ ou le gw w v didakto n e stin: this laudable reserve is not in keeping with a protreptic speech. `De exhortationum . GerhaÈ usser. .1. Protreptik. Gaiser. ouk autoÁ dikaiosu nhn e painouÄ ntev a llaÁ taÁv ap' authÄv eudokimhÂseiv (Republic 362e4±363a2. not eulogies. 1358b8±13. we must remember that Clitophon refers to a speech by Socrates which states that virtue can be taught (408b7) and that he includes it among the speeches which he considers to be protreptikwtaÂtouv (408c2). The protreptic speech reported by Clitophon. 115 Arist. dioÂti kalwÄ v authÁn e gkwmiaÂzeiv (410c3± 4). Analogically to one who praises steersmanship without being an expert at it. This shows that our author either ignores or implicitly rejects the distinction made by fourthcentury rhetoric between e painov and protrophÂ. 363d4±5). Dialex. wv crhÁ di kaion einai. 65 . 6 ˆ 2. type of protreptic is the `eulogy of the art'.114 Polus' highly rhetorical encomium of te cnh in general (Gorgias 448c4±9). cit. esp. one may reprove Socrates wv ou maÄ llon onti dikaiosuÂnhv e pisthÂmoni.. Rh. if not identical. who likewise does not distinguish between praising justice and exhorting others to justice: le gousi de pou kaiÁ parakeleuÂontai pate rev te ue sin kaiÁ paÂntev oi tinwn khdoÂmenoi. Gaiser. Cf. 37±8).±K.414.±K. Rh.1±26 D. 284 E. cf. (fr. Protagoras' defence of sophistic (Protagoras 316c5±317c5. 16±31. 114 There is again an interesting parallel from Posidonius' Protr. 113 Cf. historia'. . better attested. 283±91. Another. 112).I N T R O D U C T I O N II.2 be assigned the status of a protreptic genre. as well as the two alluded to by him (408b6±c1 wv didaktoÁ n areth kaiÁ paÂntwn e autouÄ deiÄ ma lista e pimeleiÄ sqai) are exhortations. Al.113 All the same.115 In this respect. Protreptik. An interesting passage in the Clitophon shows that eulogy and protreptic are related. and Gorgias' eulogy of rhetoric (456a7±457c3. op. 40±2) may be considered under this head.

116 Similarly. Gomperz. I conclude that explicit protreptic in the stricter sense did not exist among the sophists as a ®xed genre. This shows not only that protreptic speeches sensu stricto which exhorted people to the arts (like Galen's ProtreptikoÂv) were unknown to the author. though some implicit types (mainly eulogy) may be tentatively considered forerunners of explicit protreptic as found among the Socratics. 66 . Arist. our author could think of nothing better than eulogy. but a person who attached much value to the distinction would not have chosen this analogy. 116 Though it is usually taken as referring to criticism of all arts (HGPh. a eulogy of a reth may be (implicit) protreptic. The apologetic nature of the passages mentioned above may serve as a link between sophistic protreptic and the pseudo-Hippocratic PeriÁ te cnhv. Faced with the question what protre pein would amount to in the ®eld of the te cnai.I N T R O D U C T I O N II. 998a3±4 is no evidence. Met. The analogy from protreptic to bodily care (410d1±4) con®rms this: Clitophon says not protre pein ei v gumnastikhÂn but periÁ gumnastikhÄ v.1. which is an (implicitly protreptic) defence of the te cnh status of medicine. but also that for him the most natural form of exhortation to a te cnh was an e gkwÂmion te cnhv. 232d6±7 a deiÄ proÁv e kaston autoÁn toÁ n dhmiourgoÁn anteipeiÄ n can only mean `what the craftsman himself ought to answer to everyone' (cf. the shift from protre pein to e gkwmia zein was in¯uenced by the choice of steersmanship as an analogy for justice. iii 44 n. Campbell ad loc. A passage from Plato's Sophist (232d9±e1) indicates that Protagoras wrote an apology for all arts. 41). Eine griechische Sophistenrede des fuÈnften vorchristlichen Jahrhunderts (Leipzig 19102 ).. the fragments known as the Anonymus Iamblichi may be analysed as such. as it does not refer explicitly to a separate book. Th. 169±70).2 Of course. and which is full of sophistic topoi. Die Apologie der Heilkunst.

section ii. D. Since it is easy to recognise that Socrates' protreptic speech. F. deal with philosophical protreptic: it must have been more than an isolated phenomenon. (3) Demetrius (section ii. motifs. protreptic.1. that are common to the majority of (but not necessarily to all) representatives of the genre.1. as it did unquestionably in later times. 67 . or whatever we call them. whose agnosticism is also due to his failure to distinguish between implicit and explicit protreptic (cf. we may con®ne this inventory to explicit philosophical protreptic both in the wider (ethical) and the stricter sense. Rhetorica 4 (1986) 309±33).1. we must ®rst draw up an inventory of protreptic texts. not implicit. First.1). I claim that protreptic is to be considered a separate genre of philosophical literature in the fourth century for the following reasons: (1) three philosophers wrote books entitled Protreptiko v.117 a study of all texts which are possibly implicitly protreptic will lead us nowhere. (4) the protreptic texts collected in this section constitute a rather coherent fund of motifs (used extensively in Euthd. In the fourth century at least. Generic Composition in Greek and Roman Poetry (Edinburgh 1972).118 In fourth-century phil117 As against M. is explicit.3 ii. the set of more or less traditional ideas. his examples belong to the fourth century. topoi. protreptic seems to be a genre in content rather than in form (though we know little of the form of the works called ProtreptikoÂv). Besides. Cairns. Jordan's scepticism (`Ancient philosophic protreptic and the problem of persuasive genres'. if one wishes to settle the question whether protreptic existed in the fourth century bce as a ®xed genre.1. In ancient literary composition a genre is characterised by two kinds of features. the general set of elements which together distinguish one genre from another and which enable us to decide whether a given text belongs to one genre or to another. 6. and Clit. reported in that dialogue. 118 Cf. (2) both Euthd.4.3 The protreptic corpus In order to study the relationship of the Clitophon to philosophical protreptic.I N T R O D U C T I O N II. Points (2)±(4) are ignored by Jordan.2) appears to describe the di¨erent species of the genus protreptic. and Clit.): such a coherence is best explained as due to the in¯uence of generic composition. second.

We may also conclude that the speech is protreptic because Socrates 119 This will be done here only in relation to Clit. we may not use them to determine the extent of this corpus. d2±3 thÁn SwkraÂtouv protrophÁn h mwÄ n e p' arethÂn. (7)).119 But as motifs are themselves de®ned by their recurrence within the corpus of protreptic texts.3 osophical protreptic. nor are any of the interlocutors so persuaded.120 We must therefore use other criteria. 121 And because it is treated as protreptic by Chrysippus (section ii. 120 For instance. 283b2±3 does mean that the eristic conversation is not protreptic.. because only one complete protreptic text from that period has survived. there is a fair number of protreptic motifs in Pl. in either case we apply an external criterion. But the irony of Euthd. because it has survived in its entirety. the pseudo-Platonic Alcibiades 1. in a corpus of explicit protreptic. so we can observe that it does not persuade the reader explicitly to care for his soul or to take up philosophy.1(4)).1.121 in drawing this conclusion we have used an external criterion.2). Of course. It does not matter whether the statement that part of a text is protreptic occurs within the same text (outside that part) or outside the text. (section ii. 68 . 410b4±6 nomi sav se toÁ meÁ n protre pein ei v arethÄv e pime leian ka llist ' anqrwÂpwn draÄn). based on the general property of exhortation. We can therefore use our inventory mainly for collecting and studying the motifs. we know little about the general elements (apart from the basic landmark of explicit exhortation as such). but no one would dream of including Smp. because it does not have the generic property of protreptic. statements occurring within the same text should be examined ®rst in the light of the overall intention of that text. We can settle this for Smp. we see that the evaluation of the protreptic is ironical (section ii. in the case of Clit.I N T R O D U C T I O N II. and as they are naturally capable of being used also outside this corpus (cf. Smp. Socrates' speech in the Clitophon is a specimen of protreptic because it is stated to be protreptic (408c2 protreptikwtaÂtouv te hgouÄmai. The problem is that most of the texts which qualify for inclusion in the corpus are mutilated. but that does not detract from the validity of the datum that it is a protreptic speech. These criteria are either external or internal.7.6). below.

`Protreptikos'. (2) Aristippus. we cannot a¨ord to be too particular. (1) Antisthenes. 125 I use DuÈring's section numbers throughout in references to speci®c passages.g. the punctuation of the MSS (of D.2.1. 7. as fr.v.5 and n. Natorp in PW s. Antisthenes der Sokratiker (diss. 18 a±c Caizzi. may be a separate title referring to a di¨erent work (so Hirzel. 2543).1).) is not su½cient evidence for including Aesch. 6. See section ii. 46 [3].2 n. Gigon prints the whole extract from Iambl. Or.123 Still. Antisthenes 10). (3) Aristotle. apart from one uninstructive sentence. ProtreptikoÂv. 320c8±328d2.1. SSR v a 11. kaiÁ andr. Alc. I follow DuÈ ring125 in ad122 For instance. lost (see section ii. SSR iv a 144). Elsewhere the work is called Protreptikoi or ProtreptikoÂv (fr. Aristid. 168. Heidelberg 1970). Pl. that does not necessarily mean that that text was regarded as belonging to the genre protreptic.1.122 if on the other hand a text (or part of it) appears to exhort its readers. Protr. L. 165). 63±4).I N T R O D U C T I O N II.1. Patzer. 123 E. in the corpus (though some of his other comments are). 69 . as the fragments as printed in Ross's edition are too long to be useful. 124 PeriÁ dik. Prt. 272±6) is too haphazard to go by.485. 1. given the scantiness of data in the area. this seems excessively cautious to me. I consider the following texts to constitute a corpus of philosophical protreptic. it may in fact have a wholly di¨erent intention. 302±13) under the heading `to poi protreptikoi '.1. ProtreptikoÂv (fr. this is an internal criterion. 72 n. This work has been regarded as the source of Socrates' speech in the Clitophon (section ii. facsimiles in A.3 accuses mankind of neglecting the real values for human life and states that a better e pime leia is necessary.21±2 Lenz±Behr o d' h n crh simon ei v toÁ protre yai (in Aesch. 73 (Aristotelis Opera iii: Librorum deperditorum fragmenta (Berlin 1987). If a word like protre pein is used in commenting on a text.124 lost. 576 ˆ 1.15±18. cf. PeriÁ dikaiosu nhv kaiÁ andrei av protreptikoÁv prwÄtov deu terov tri tov periÁ QeoÂgnidov d0 e 0 . Alc.4. 121 Mannebach. Both criteria have their disadvantages. section ii.2.

6. Iambl. `De exhortationum .3 mitting as part of this work the extracts contained in the sixth to twelfth chapters (inclusive) of Iamblichus' Protrepticus. Bos. as well as an Oxyrhynchus papyrus (666 ˆ CPG 24. 46. set o¨ from each other and from the following by Iamblichus' favourite formula of transition e ti toi nun. Aristotele. 6) but I will treat them as part of Protr. 38. 93 (cf. 78. Iambl. For further details. Cosmic and Meta-cosmic Theology in Aristotle's Lost Dialogues (Leiden 1989). (cf. 282d1±3). Ierocle (Studi e Testi per il Corpus dei papiri ®loso®ci greci e latini.126 Both the title and the content warrant inclusion in the corpus. 41. 92. 26008 preserves fragments of Arist. as these fragments do not contain protreptic motifs found elsewhere.127 This is typical of protreptic in the stricter sense (compare the end of the ®rst protreptic conversation in the Euthydemus. (Iambl. it would be imprudent to use the Eudemus as a whole for the protreptic corpus. 58. Euthd. Protr. 70 . are the same work (`Aristotle's Eudemus and Protrepticus: are they really two di¨erent works?'. 62.. 96. 54. 104. 247. Dionysius 8 (1984) 19± 51) is partly immaterial to my argument. Florence 1992). cf. Most. other transitional passages contained in b 10. A.6). 90. which is in my opinion a condensation of the two protreptic conversations from Pl. Protr. 126 The following passages I do not consider part of Protr. (so G.1.12±28) rejected also by DuÈ ring. Protr. ± the decision will be justi®ed in Appendix i.I N T R O D U C T I O N II. (25±6). .16 des Places). 31.23 des Places) and an un-Aristotelian piece of dubious origin (66.20±59. 189±216) is not a question that has to be decided here. 63 (e ti toi nun). Hartlich.17±64. 59 (e ti toi nun).: b 52 DuÈring. 43. Vindob. P. I am also dubious about b 29±30 which follows this latter piece and seems to constitute its natural sequel. 22. Protr. 45. on 410e8). Bos' theory that Eudemus and Protr. in Studi su Codici e Papiri Filoso®ci: Platone. Comm. 15. R. Protr. 110 DuÈring). 127 I agree with DuÈ ring that other occurrences of jilosojhte on must be ascribed to Iambl. . Appendix i. which is not from Iambl. especially Aristotle's framing the course of his thought so as to end repeatedly in the conclusion that one must philosophise (b 5. `Some new fragments of Aristotle's Protrepticus?'. 97. not to Arist. b 7±9. Whether or not P. W. which comes between an excerpt of Pl.22±6 (ˆ b 59). cf. with a few exceptions. historia'. b 104±10 are sometimes ascribed to the Eudemus (cf.2 is a doublet of 71. I am highly sceptical about b 23±8. 37. 17 n.. which overlaps with an extract from Stobaeus ascribed to Aristotle (b 2±5 DuÈring).

ˆ 1±11 D.2. 54±6). (6) Aeschines of Sphettos. (SSR vi a 46).2. E¨e. 10±25.. Socrates leaves room for Alcibiades' arrogance because he had said that he was `wedded to ignorance' (118b6) but added that most politicians su¨er from the same evil (118b9±c1). 1.) rightly warns against rashly using X.23 Lenz±Behr) where it is compared with the pseudoPlatonic Alcibiades 1.. 129 Aesch. 172d7±10 ± Aesch. Sokrates.] Alc.6 (71±7. MuÈller. Mem. which may derive from Aesch. 3. more precisely. ˆ 8 D. We learn most about the intention of this dialogue from a passage in Aelius Aristides (  U peÁ r twÄn tettaÂrwn. Mem.3. I use the fragments (1±6 and Vest. [Pl. Vest. Kurzdialoge.1.484. 1 Kr. (SSR vi a 50) and Alc. E¨e gives plausible reasons for the assumption that Chrm. Aischines. Ehlers. B. 164a9±b6. Euthydemus 278e3±282d3. Gaiser. Alc. Ehlers (11 n. Aristides compares the e¨ectiveness of Socrates' handling of Alcibiades in either dialogue.. 1608. See section ii.. this is unsuitable because Alcibiades was so arrogant that he would have criticised even the twelve Olympians. Apolog y 29d7±e3. 71 . 171d8±e5. 306d6±end. SSR vi a 41±54) only. I am prepared to accept his conclusion (especially on account of Pl.16.1.2. I cannot use them here.] Just.3. 572±6 ˆ 1. he should have added the parallel Chrm. Mem. 30b2±4. fr. See Dittmar. 4. but cf. Alc.12±19 goes back to the source of [Pl. ˆ fr.19±20 to Pl.5±6 ± in any case as neither X. MuÈ ller. P. 134±9) and 4. Oxy. In Plato (Aristides did not doubt the authenticity of Alcibiades 1.128 which are protreptic dialogues in their own right: (7) and (8) below. Chrm. cf. 199±203. 5 D. Gaiser uses also X.Mi. below).3 (4) Plato.283. Kurzdialoge.3. B. Alc. C. (5) Plato. W. (Maier. (C. 36c5±d1 (Socrates' protreptic speeches). 164c7±end is intended as a refutation of Aesch.6 is protreptic. But again. 1±3 Kr.8±9 and 4. 137±8 on X. indeed 4. 132c4±6. 86±7) and should have used 1.6. not the derived texts. Protreptik. Alcibiades. 77±95. 1 Kr. 288d5±292e7. Alc. nobody did in antiquity). Chrm. Mem. W.8±9 nor 4. 120±44. See section ii. Hp.129 therefore the thought that he did not stand alone in his igno128 X. 3. Eine vorplatonische Deutung des sokratischen Eros (Munich 1966). is not a protreptic text. 4.4 and n. `Charmides'.I N T R O D U C T I O N II.7± 485. Mem. 1 117c6±e5.

the possibility of an arrogant reaction on Alcibiades' part is suppressed. 46.. ˆ 8 D. the statement seems to be contradicted by P. What Aristid. Though I shall occasionally refer to it. Besides. 132 In `Aeschines' Miltiades'. 1.1. Milt.19. Evidently. inasmuch as Socrates tries to persuade Alcibiades that he falls short of the standards of (moral) education which are necessary for the prospective politician. p.131 for he has said in the eulogy of Themistocles that not even he had acquired suf®cient knowledge. all' e nede hsen are a paraphrasis of a sentence from the eulogy of Themistocles (fr. which they take to refer to what Socrates said in the absence of Alcibiades. Socrates heightens the e¨ect of his words in an appropriate way. 36±48. 1608.I N T R O D U C T I O N II. Even if the verb protre yai had not cropped up in his analysis we would have been justi®ed in adding Aeschines' Alcibiades to the corpus. means is that in Aesch. in which he does not speak evil of that politician. Aischines).1±5. But the words ou deÁ e kei nwi h e pisthÂmh ou sa hrkesen. there is no su½cient ground in the fragments preserved from this dialogue to warrant its inclusion in the protreptic corpus. as a protreptic dialogue. 305±8. Socrates reduces Alcibiades to tears by his eulogy of Themistocles. Krauss (63±4) thinks that the dialogue ended in a climax.3 rance was comforting. Milt.130 so that he may not be spoiled any more than he already is: Alcibiades does not even come near to Themistocles as far as training is concerned. Aristides regards both dialogues as protreptic. 270. In Aeschines. (SSR vi a 50). 1 Kr. Oxy. Compare the de®nition of `protreptic dialogue' (section ii.576. I have tentatively reconstructed Aesch. almost certainly did not have a conversion scene and quite certainly not a ®nal conversion scene. Alcibiades is reduced to tears because of his inferiority to Themistocles. 131 KaiÁ prose ti summe trwv e pe teine toÁ n lo gon. Following Immisch. and that on top of this Socrates shows (e pe teine toÁn lo gon) that even his admired example had not been equal to the vicissitudes of fate.1.485.48±51 Dittmar.1). on account of the following words eipe gaÂr pou ktl.132 130 Aristid. and his despondency ± and at the same time his dependence on Socrates ± is increased. in this way. 72 . And besides.

in this case Socrates (135b7±e5). he thinks. The right e pime leia autouÄ is possible only through self-knowledge (124b7±129a10). ˆ 8 D. Only with this knowledge is true statesmanship possible (133c21±135b6). 128. Euthydemus. cf. especially in 4. The enumeration of so-called agaqa (4.I N T R O D U C T I O N II. Alcibiades states that what is just is rarely useful (113d1±8). the only protreptic text from the pre-Christian era which has been preserved unmutilated (the next one which we encounter is Clement's Protrepticus).. 1 and Aesch. 11) is twice brought to a qumi a (23. `Charmides'). which is identical to knowledge of one's soul or swjrosu nh (129b1±133c20).3 (7) Xenophon. In its entirety. section ii. 31±9).2.1±7 (cf. 73 ..31) which are subsequently proved to be sometimes evil (32±6).2 n. 39) through Socrates' proof of his ignorance (12±23. E¨e. Alc. 1 Kr. (SSR vi a 50) on Themistocles' preparation for politics) and in the adage gnwÄ qi sautoÂn (4. the only true lover of Alcibiades (131c5±e5). until this knowledge has been acquired. will have to show the Athenians what is just and unjust in politics (109b5±c12).24±30. but Socrates proves that the just is always useful (114e7±116d4). he becomes a close follower of Socrates. It is only when this chapter has material in common with other texts from the protreptic corpus that we are on safer ground. as well as from Plato's Hippias Maior). Alc. 133 Cf. Alcibiades 1.2. about to embark on a political career (105a7±b1).133 (8) Pseudo-Plato. n. Alc. as that is.2. As a second line of defence. Alcibiades is ignorant (as most politicians are) and needs an e pime leia (118b4±119b1). this text is de®nitely protreptic.2. cf. yet it contains various motifs not originally of a protreptic nature (it has been established that the author borrowed from an unknown but probably not protreptic dialogue about justice. 1 106e4±112d10 and Aesch. 188. the only way to become an a nhÁr axioÂlogov (40).1. An over-con®dent young man. but he has never learned the di¨erence (112d7±113c7). Alcibiades. fr. Memorabilia 4.2. one has to have oneself led by a better man. Alc. standing at the beginning of a political career (1. recurs in Euthd.

57±8) shows closest a½nity to Men.135 sustains the rejection. 136 I may refer to C.16 ˆ Iambl. Ly.6.I N T R O D U C T I O N II. FestugieÁ re. 70±3 (magic). 1 is not Alcibiades' jilotimi a but human jilotimi a in general (periÁ thÄv e n e kaÂsthi yuchÄ i jiloti lkibia mi av skoÂpon e cei touÄ e le gxai. In Alc. cf. 101±56). Bos.1. `Epinomis'.. 1: 3. the expression is absent from all pre-Republic works but for one occurrence in Ly.3. 127e9) and his despair increases accordingly (124b7±9.4. seems to con®rm this).3. Grg. which cannot be discussed here. 104. 1: 5. In Alc. he means in fact that Alc.217) says that the sko pov of Alc. 1 is a protreptic dialogue. Cf.4. 124b7. Alcibiades maior. Plato's Logic. whereas a comprehensive analysis of all formulas of assent (P. because it 134 When the author of the Anonymous Prolegomena (Hermann. 1: 3. data corrected after H. 135 Some stylistic tests. 1 di¨ers toto caelo from that of Plato's later dialogues. ii (Bonn 1923). We saw that Aristides regarded the dialogue as protreptic. FriedlaÈ nder. once in every 5 pages. the second of which (119a8±124a7) he calls `protreptic'. which inaugurates the third. (my personal impression is that the style of Alc. Lg. as do certain pieces of un-Platonic doctrine. von Arnim. Phlb. 142 (ˆ 92 Westerink) divides the dialogue into three parts.2. the absence of monologue.1..8. 1 74 .16±14. Platonis Dialogi.134 I shall explain elsewhere (section ii.2(7)) 121a3±124b6. `Sprachliche Forschungen zur Chronologie der platonischen Dialoge'. 13. Der grosse Alkibiades. apart from the `central culmination' (section i. Plt.5) my reasons for not assuming that Plato ever wrote a protreptic dialogue. fr. d4. e sti gaÁ r e kaÂstwi h mwÄ n oi on A deiov jilotimi a). 1: 4. notably the occurrence of ti mhÂn. 1: 1. cf. (5 times in 25 pages Didot. R. 1: 2.. passim. Sph. Linguistic evidence. ± Bos rightly claims that if Alc. 55±63 (au toÁ toÁ au to ). This con¯icting evidence is best explained as due to inauthenticity. Protreptiques. Olymp. in Alc. esp. Euthyphr. 88±90 (dogmatism). `maieutic' part). SAWW 169. Lutosøawski. 123d3. this is precisely the part where Alcibiades' need of e pime leia is ®rst stressed and developed (the break at 124a7 is motivated by the ®rst mention of the Delphic maxim. Alcibiades' need of e pime leia is stressed repeatedly (119a9.3 (1912) 31) indicate a late date of composition.5. 6. 1: 5. also Procl. A.136 I do not accept as protreptic the Epinomis (Einarson.. 1: 8. Prm.3 The synopsis will make it clear that this dialogue is explicitly protreptic.2. Phdr. 2 Dillon. 127d6±8). this is my main reason for regarding Alcibiades 1 as unauthentic (I admit it is a subjective one). Tht.

(Einarson. is absent. besides. 218 n. are a number of passages from Plato's dialogues.1. this search is `in the manner of the second discourse in the protreptic of the Euthydemus' (`Epinomis'. including them would blur the line between protreptic and paraenesis. On grounds wholly di¨erent from mine he concludes: `The most important aim of Alc. I think it highly dangerous to reconstruct the plan of Arist.137 Likewise. and it contains a number of protreptic motifs). `Epinomis'.2). There is no exhortation to soji a. my translation).3 fails to meet the criteria stated above (it is certainly implicitly protreptic. but a counterpart to the ®rst discourse. 279). As Einarson rightly remarks. Not belonging to the corpus. 75 . 37 and passim).1). 137 By the same token. he bases this judgement primarily on the doctrines of Alc.4. 1. from that of Epin. which alone would make the Epinomis a protreptic text. has the same structure as an implicitly protreptic one. as Gaiser does). but the subject of the investigation is what soji a consists in. 1 is protreptic to philosophy' (54. jroÂnhsiv or e pisthÂmh. These passages are not part of our inventory of protreptic texts because they are not detachable from their contexts (as are the protreptic conversations in the Euthydemus). I shall call them `protreptic situations' (I do not distinguish between protreptic and paraenetic situations. they are not marked as protreptic is authentic it must be a late work (83±7). even though some passages have the general property of exhortation (in the wider sense). I do not include any material from Xenophon's Memorabilia (apart from 4. 20). Following Gaiser (Protreptik. in which there is explicit exhortation. Protreptik.I N T R O D U C T I O N II. but associated with it. Gaiser.4 onwards to protreptic (see section ii. Protr. Xenophon himself opposes the speeches and conversations from 1.1. it would mean also that the corpus would contain so many heterogeneous texts that it would cease to be useful as a basis for studies in protreptic. It is a priori unlikely that an explicitly protreptic text like Arist. Protr.

132±4). 511±12). they do not justify Gaiser's claim that the structure of the Platonic dialogue is derived from a ®xed pattern of sophistic protreptic (see section ii. (9) Republic 621b8±d3 (cf. Apolog y. section ii. Grg. may have perceived the protreptic character of the Prt. 76 . from this passage.138 (8) Symposium 173c1±d7 (cf. this does not do justice to Clit. de Strycker±Slings.2. but one of his parallels is interesting (cf. I shall use some of these passages occasionally when analysing the protreptic speech in the Clitophon. section ii. The author of Clit. 42±4. Cf. 295a1±c1.1. (5) Protagoras 316c5±317c1 (Gaiser. I reckon: (1) Gorgias 456a7±457c3 (Gaiser. 320c2±328d2 (Gaiser.1. Chrm. The exhortation from beyond the grave in Menexenus 246d1±248d6 belongs to rhetorical. Phd. on 407b5 didaska louv). Euthphr.2).Mi. 217).3 (our external criterion. however. 357e4±8). cf. Comm. La. Hp. Among protreptic situations. cf.. that PoiÄ je resqe. e.g. 138 Rather a piece of argument in the form of a protreptic speech. Pavlu (`Pseudopl. (4) Phaedo 114d1±115a2 (Gaiser. 89c11±91c6.2). 37±8. not philosophical. section ii. w a nqrwpoi. Hp. 526d3±end (Gaiser. also the proof that wrongdoing is involuntary was certainly not borrowed from Prt. (3) Laches 178a1±190b5 (Gaiser.1. 11b6±e4. (6) ibid. Protreptik. on 407b1 PoiÄ je resqe). 166c7±e2.I N T R O D U C T I O N II. (7) ibid.2. who derives the ®rst part of Socrates' speech in Clit.2). 194±6). section ii. Comm. wnqrwpoi was inspired by 353a3 etc. Men. 352e5±357e8 (Gaiser. above). 86b6±c7. 114±18). 193e8±194c6. 486e5±488b1. 79e7±81e2. (2) ibid. protreptic. 139 I have not used passages which are protreptic only inasmuch as a partner of Socrates is encouraged to go on with the discussion.5). 38±40. passage and used a few phrases (esp.1. 372a6±373a8. following Susemihl. 8±9. Kleitophon'. 40±2. I cannot believe.2.139 For the preamble to the laws in Laws 5. 190±2).Ma. 481c5±482c3.

the second from 1. 126. Maier. M. the more so since it has been taken as a reference to the Clitophon:141 Ei de tinev SwkraÂthn nomi zousin. KeÂrygmes. a ii 1). Erbse.4 ii. Though one is considerably later than the fourth century bce. we have to examine two passages which betray re¯ection on philosophical exhortation and are relevant to the Clitophon. w v142 e nioi 140 On the unity of 1. .v.1 Xenophon. 337±40. `Architektonik'. Treu. Polycrates' Kathgori a SwkraÂtouv in 1.I N T R O D U C T I O N II.] as some have actually said': tinev may refer to the readers or listeners of the person(s) called e nioi. 93±4. 1. H. we shall be able to decide later what is relevant and what is not. but Xenophon sees ®t to insert a separate preface concerning Socrates' discourse. .1. 141 Bertini. 77 . As this preface contains statements regarding exhortation made both by Xenophon and by some others.v. The ®rst point is shown in 1.1). taÁ deÁ kaiÁ dialegoÂmenov (1. explicit statements on protroph are not found in his works anyway. PW s. Xenophon goes on to a counterattack and proves that far from being harmful. 43. Erbse. analysis seems to me imperative. 458.4 onwards.3). Memorabilia 1. Socrates was even highly bene®cial to his companions140 taÁ meÁ n e rgwi deiknu wn e autoÁn oiov hn. approved by Bertini. Lit. Maier. 465±9. Sokrates. Gigon.4.1 Having dealt with.4 Early theories of philosophical protreptic Apart from the texts collected in the previous section. 13±17. Erbse. Necessitudo. 42 n. I reserve a separate treatment of Plato till later (section ii. 37 is super¯uous. the distinction between tinev and e nioi makes sense even if w v is retained: `If there are people who think [. 322 n.1±2 and the following chapters. ii. 323±5. Kesters. `Architektonik'. Xenophon (6). cf. 319±22. Memorabilien I.3. 458±60. presumably. Kunert.. Sokrates. `Architektonik'.1.1±2. 142 Jacobs' conjecture oi v. as tekmai romai is often used without a complement (LSJ s. 1777±8.1. A papyrus from the third or fourth century ce (P. I shall treat it rather fully here.3.4.

In this context.1 (Erbse. A. VM 1. 336c4± 6 dialege sqw e rwtwÄn te kaiÁ apokrino menov. 325±6). Prt. cf. Pl. Marchant. On the other hand. In the ®rst interpretation. `Architektonik'.v. dokimazo ntwn ei i kanoÁv h n belti ouv poieiÄ n touÁ v sunoÂntav. Sokrates. we should expect mhÂ.1. Oguse. `They must examine not only A but also B' can be interpreted either as `they already examine A but must also examine B' or as `they examine neither A nor B and must examine both'. Hp. Pl. `Architektonik'. Erbse. Edelstein. mhÁ e j' e kaÂsthi e rwthÂsei makroÁ n lo gon a potei nwn. 323±4). Chantraine. i 457. 246±7. protre yasqai meÁ n anqrwÂpouv e p' arethÁn kraÂtiston gegone nai. 31.4 gra jousi te kaiÁ le gousi143 periÁ au touÄ tekmairoÂmenoi. As the words are intended as an announcement of the following conversations and speeches at any rate up to and including 2. Since the participle skeya menoi has the modal value of the main verb dokimazo ntwn. 162e1. t. these people look only at Socrates' refutations of others and consider these instances of his exhorting others to virtue ± in other words there are two types of discourse: protreptic±elenctic and what may be called didactic (so JoeÈl. only an examination of the whole work and of parallel statements can take us further. Recherches sur le participe circonstanciel en grec ancien (Paris 1962). `correction' and `punishment' are di¨erent things. . even in the ®rst case.145 Lond. there is nothing in this text to prevent us from distinguishing three types of discourse: protreptic. cf. E.I N T R O D U C T I O N II. Gaiser. elenctic and didactic (E. 143 These words seem to be a more or less ®xed combination in polemical writings. Xenophontisches und Platonisches Bild des Sokrates (Heidelberg 1934). proagageiÄ n d' e p' authÁ n ou c ikano n. Protreptik. s. praefatio) reads wv like all MSS. Der echte . 145 Owing to a syntactic ambiguity. 149. Xenophontis Opera Omnia. Memorabilien I. because words in -thÂrion denote either instruments (pothÂrion) or places (desmwth rion) or religious actions (quth rion). kolasthÂriov ii 3. I would rather interpret the phrase kolasthri ou e neka as `as a means of correction' ± the formation would then be analogous to some members of the third group like luth rion. 94±5). La formation des noms en grec ancien (Paris 1933). kaqarth rion. . 119. cf. 30 and n. 2 (Oxonii 19212 ). 78 . there are two possible theories on the number of types of discourse distinguished by Xenophon. allaÁ kaiÁ a le gwn sunhme reue toiÄ v sundiatri bousi. cf. only this place adduced). P. C. 144 It is very improbable that kolasthÂrion means `punishment' (LSJ. not ou . 62±4. Gigon. skeya menoi mhÁ moÂnon a e keiÄ nov kolasthri ou e neka144 touÁ v pa nt ' oiome nouv eide nai e rwtwÄ n hlegcen. Tht.

147 In the recapitulation of his defence of Socrates against the indictment (1.7.147 it becomes clear that the distinction made in 1.. 2) in the eulogy of Socrates which closes the Mem. the second with Antiphon the sophist ± both qualify for the sobriquet touÁv paÂnt ' oi ome nouv eide nai. When we examine Xenophon's use of the word protre pein. with probiba zein and agein acting as intermediaries. 5). 1. (4. 1. proÁv e gkra teian (cf.5. The distinction made initially is gradually given up.1 proetre peto .8. .14 e piÁ kalokagaqi an agein.4.7) addressed to Socrates' sunoÂntev (so named 1.5 and 1.64).6) and a `didactic' speech (1. . but 146 A parallel signalled by H. 7. and this is the sense in which the word is used throughout Mem.11 and a trichotomy at 1. as may be seen from 1. in 1.4. 4.4.1).1 three types of Socratic discourse are distinguished: explicit protreptic (not exempli®ed).1.146 This `didactic' discourse amounts to paraenesis (section ii. ei v tau thn. Obviously.1 proubi baze . though the way in which protre yasqai gradually becomes what in 1. `didactic'.8. Maier (Sokrates.4 betray a curious telescoping of proaÂgein and protre pein: 1.4. this preface is followed by two pairs consisting each of an `elenctic' conversation (1.5. 42 n. 2. protre pw is more than `to exhort' here ± it is perhaps better paraphrased as `by exhorting to cause someone to do'.1. touÁv a kouÂontav e piÁ kaloka gaqi an agein).2. the verb proa gein is not found after 1.4 As I have already shown (section i. . ape cesqai twÄn a nosi wn ktl.1 between protre pesqai and proaÂgein represents a line of thought which Xenophon temporarily adopts. except at 1.19 touÁ v sunoÂntav e do kei poieiÄ n . there is an opposition between pauÂwn and protre pwn: `he made an end to their evil desires' as opposed to `he instilled in them a longing for virtue'. . 79 . Evidently. `elenctic'.I N T R O D U C T I O N II.1).5.7.1 is called proagageiÄ n (cf. The introductory and closing remarks in the chapters following 86).4.11) could be used for both interpretations.1.4.1 protre pein. The ®rst e legcov is a conversation with Aristodemus who does not believe in the gods (1.. .1 arethÄv e pimeleiÄ sqai proe trepe.2 n.14 e do kei . .5.4. .1. praktikwte rouv e poi ei.2). 6. .6.1. at the end of either chapter the usefulness for the listeners is stressed (1. next note) is an argument for a bipartition of Socratic discourse at 4. 1.

has a terminative Aktionsart: it denotes an action which aims at a certain goal and has the attainment of this goal for its natural conclusion (cf. elenchos re148 This becomes obvious as soon as we translate protre pein by `to exhort' ± it makes nonsense of the statements quoted in the previous note.1. If so. 80 . `Architektonik' 323). like a potre pein and pei qein. But there may be other determining factors in the context.2. 149 Cf.4. 1. that does not mean that there were Socratic exhortations ± the words tekmairoÂmenoi may even seem to tell against the assumption that these existed.5. there is an ambiguity in the notion of protreptic which may be perceived more clearly in Stoic theory. Certainly Gigon (Memorabilien I. . carrying out the action by a present or imperfect form (cf. . From 1.7. such as the presence of a plural object. Yet when Xenophon invites these critics to examine Socrates' elenctic and didactic discourse he implies that they had hitherto paid attention only to other types of discourse. which encourage the present/imperfect even when the goal is attained. `protreptic' was mainly accusatory exhortation.4 which is wholly alien to him ± as we shall see (section ii.761 proÁv e tera tina maÄ llon hmaÄv protre yetai h toÁ jilosojeiÄ n.7. .1. ou deÁ a potrepome nou and 4. Erbse.149 Does the passage under discussion prove that Xenophon knew protreptic as a separate genre? It is certainly not stated in so many words: if some people write that Socrates can only exhort others.1 onwards protre pein is precisely what had been called proagageiÄ n at 1. reaching the goal would be expressed by an aorist form. The trichotomy accusation.2). a pe trepe.3). 119±20) is wide of the mark when he translates protre yasqai by `Aufmerksam-Machen'. Intuitively I would say that protre pein.148 Already in Xenophon. .4. Normally.1. SVF 3. and it would seem that protreptic speeches are the only possible ones left.5 i scurwÄv ape trepen). this is not the only time he does so. (30) touÄ deÁ Kriti ou . one may tentatively conclude that for Xenophon.29 Kriti an .I N T R O D U C T I O N II. advice.682 mo non deÁ protetraÂjqai toÁ n sojo n in comparison with 3. his instances of Socrates' `didactic' discourse bear the stamp of advice which Demetrius considers typical of him (section ii.

16 (section ii. iii 23.I N T R O D U C T I O N II. toÁ jilhtikouÁ v meÁ n ei nai twÄn 150 It would ®t in quite well in Aesch. 81 .7. Kleitophon'. protetramme nwi deÁ scedoÁ n kaiÁ e mpoÂdion touÄ proÁv te lov arethÄ v e lqoÂnta eu dai mona gene sqai ). De Oratore 1.4. The impression that Xenophon is actually referring to the Clitophon is rather strong.3 of this introduction). So for instance. Souilhe .7) would appear to us already to border on protreptic speeches. as summarised especially in the ®nal sentence (410e5±8 mhÁ meÁ n gaÁ r protetramme nwi se anqrw pwi w Sw kratev axion ei nai touÄ pantoÁv jhÂsw.4 curs in Demetrius and in Plato (section ii. iii 23. 18±9. We cannot escape from it by pointing to a text in Cicero adduced by Grote (Plato. Ac. between fragments 10 and 11 D.3. perfectum sibi opus esse. eis reliquam facilem esse doctrinam).5. 1. Nor again is it possible to cast doubt on the relationship between our passage and the Clitophon on account of Xenophon's e nioi:151 it is normal in polemic of this time not to name one's opponent but to call him `some people' (cf. `Pseudoplat. when Aristotle says (Politics 1327b38±40) oper ga r jasi tinev deiÄ n upaÂrcein toiÄ v ju laxi. section ii.47. 1. This is quite speci®cally the charge levelled against him in the Clitophon. ± The parallel Cic.4. Alc.1).1. because these examples of `didactic' discourse (1.2) may suggest that these words are Cicero's own invention. but solitum aiunt dicere de®nitely excludes that possibility. ut nihil mallent se esse quam bonos viros. Pavlu. 178. si qui satis esset concitatus cohortatione sua ad studium cognoscendae percipiendaeque virtutis. Plato. 151 Grote. because in this passage (presumably a quotation from a lost Socratic dialogue150) there is not a trace of criticism. Xenophon appears (as the latter do not) to limit protreptic to accusation.204 ut Socratem illum solitum aiunt dicere. Certainly attested is the existence of the criticism that Socrates is successful in exhortation but cannot actually lead people to a reth . quibus enim id persuasum esset.1.

416d3±7 and H. There remains the possibility that the Clitophon is a compilation (`ein Cento aus Texten Platons und anderer Sokratiker'. But it is perhaps possible that by e nioi Xenophon actually means the literary character Clitophon.7. `Dialoghi pseudoplatonici').I N T R O D U C T I O N II. it does not exclude the Clitophon. The issue will be treated in section ii. it might derive directly from Xenophon. ii. not through ®rst-hand knowledge. We shall not be able to judge this possibility until we have compared the parallels which make up the supposed cento and on that basis can form a theory on the intention of the Clitophon (sections ii. who had (to all appearances) been a follower of Socrates rather longer than Xenophon himself. ii. so Hartlich. historia'. Memorabilien I. 152 Cf.4 gnwri mwn.6. alternatively. Whereas it may therefore be regarded as certain that the relationship between the Clitophon and the preface of Xenophon 1. it does not follow that Xenophon has the Clitophon in mind. In the meantime the point may be raised here that if indeed Xenophon was inspired by the Clitophon.152 Besides. The critics of Socrates state their opinion on the basis of an inference (tekmairo menoi ). 1330a1 ± R.4. 82 .1. 229±30. `Pseudopl.2. Pavlu. `De exhortationum . Gigon. .7. he can hardly have regarded that dialogue as Platonic. even if e nioi is taken at face value. which is essentially the position taken up by Carlini. Bonitz. 598±9. Kleitophon'. Index Aristotelicus (Berlin 19612 ). . This is a highly curious statement if it concerns the major Socratic of the day. 19). Republic 375c1±2.4 is more than a coincidence (the minimum assumption being that both texts independently preserve a current opinion about Socrates. proÁv deÁ touÁv agnwÄtav agri ouv he is quoting directly from Plato. 119) and drew among others on the writing which prompted Xenophon's remark (so Gigon.3(7).3(4)).

3 ˆ 3. CR 17 (1903) 57. Innes in S. De Or. 155 One may doubt whether this means `accusing' (stating ± a pojai nomenov (cf. o deÁ i ppon. 279) ± that people do wrong amounts to accusing them) or `a½rming'. n. Mass. Demetrius o¨ers a more general illustration of how one can convey the same messages in di¨erent types of text: (296) KaqoÂlou153 deÁ w sper toÁn autoÁn khroÁ n o me n tiv kuÂna e plasen. Demetrius On Style (Cambridge 1902): `in the way of exposition and asseveration'.9. B. Cic. 7. then Epictetus' identi®cation of protreptikoÁv carakth r and e piplhktikhÁ cw ra (3. 164) is a third argument. Russell±M. A. 10. Innes. if my hypothesis that Demetrius is speaking here about three modes of protreptic is right.10.2 Demetrius. Quintil. Longinus On the Sublime. Russell. The comment on the third manner hqikwÄv (`tactfully') kaiÁ e mmelwÄv kaiÁ ou ciÁ dhÁ toÁ lego menon touÄ to apoÁ SkuqwÄn (297) seems to contrast the third with the ®rst manner. 3. Plin. my discussion in `Aeschines' Miltiades'.1. C. cf. D. that is another argument (cf. `Demetrius periÁ e rmhnei av and Pliny the Younger'. E. Winterbottom (eds. which is not found in classical Greek (cf. If the example of the ®rst manner was really borrowed from the Clitophon. Demetr. cf. Hieron.177. C. Innes in D.4 ii.2 Labourt. Rhys Roberts. D. J.4.5. W.1. Ancient Literary Criticism (Oxford 1972). 156 o ti deÁ a nqrwpoi the MS.45. Ep. Demetrius On Style (Cambridge. 6. C. o deÁ bouÄn. G. A Greek Critic: Demetrius On Style (Toronto 1961): `in the pointed and accusatory manner of Aristippus'. H.154 outwv kaiÁ praÄ gma tau toÁ n o me n tiv a pojaino menov kaiÁ kathgorwÄ n155 jhsin o ti dhÁ ``a nqrwpoi156 crhÂmata meÁ n a polei pousi toiÄ v paisi n. Halliwell. A.33±4. M.). cf. 138. the translators are divided (W. Finally. 303).23. Fyfe±D. Aristotle Poetics. 53 (sic). o ti oi a nqrwpoi most editors.±London 1995)).26±11. which is an argument for `accusing'. 83 . 215. 23. For the construction jhmiÁ o ti. D.9. 3. Grube. Mayor. PeriÁ e rmhnei av 296±8 At the end of his treatment of toÁ e schmatisme non e n loÂgwi (a type of statement or even of whole text in which what we would nowadays call the `illocutionary value' is di¨erent from the denotational value).I N T R O D U C T I O N II. 407a7 e pitimwÄ n). 154 Cf. 153 Demetrius means: apart from the problems of eu pre peia and asjaÂleia discussed in the previous sections.18±9. Ep. ``oi deÁ anqrwpoi Spengel.11.21.

. .v.158 o ma lista dokouÄsin zhlwÄ sai Ai sci nhv kaiÁ Pla twn. leijqeiÄ si. will reformulate the same subject-matter as a question. LSJ. Speaking more generally. metaruqmi sei a n touÄto toÁ praÄ gma toÁ proeirhme non eiv e rw thsin wde pwv. so with regard to the same subject-matter one man will state as an accusation `people leave their wealth to their children. typically Socratic. for example `My boy. (298) eu hme rhsan d' oi toiouÄ toi lo goi toÂte e xeureqe ntev toÁ prwÄ ton. Gesammelte Abhandlungen (Berlin 1885). maÄ llon deÁ e xe plhxan twÄi te mimhtikwÄi kaiÁ twÄ i e nargeiÄ kaiÁ twÄ i metaÁ megalojrosu nhv nouqetikwÄi. poÂsa soi crhÂmata a pe leipen o path r. a llaÁ kaiÁ e pisthÂmhn thÁ n crhsome nhn autoiÄ v. as opposed to the manners of all those who may for one reason or another be included in Socratic literature in general. by a third into a horse.'' (297) toÁ deÁ i di wv kalouÂmenon ei dov Swkratiko n. one which was apparently adopted especially by Aeschines and Plato.'' ama gaÁ r kaiÁ ei v a pori an e balen toÁ n paiÄ da lelhqoÂtwv kaiÁ h ne mnhsen o ti a nepisthÂmwn e sti . i. (`by way of suggestion') seem to overlook this. yet they don't leave with it the knowledge how to use their legacy'. 158 `The Socratic manner in the stricter sense of the word'. as normally in Xenophon: `People should not . .'' touÄto deÁ toÁ ei dov touÄ lo gou A e terov deÁ tautoÁ upoqetikwÄ v157 prooi setai.4 e pisth mhn deÁ ou sunapolei pousin thÁ n crhsome nhn toiÄ v a po risti ppeion le getai.e. kaiÁ ouciÁ dhÁ toÁ lego menon touÄto a poÁ SkuqwÄ n. h polla tina kaiÁ ouk eu ari qmhta. kaqaÂper XenojwÄ ntov taÁ po lla. Bernays. just as the same bit of wax is moulded by one man into a dog. oi on oti ``deiÄ gaÁr ou crh mata moÂnon a polipeiÄ n toiÄ v e autwÄn paisi n.I N T R O D U C T I O N II. ?' In this way the 157 Normal philosophical terminology for `advising'. ± a ra oun kaiÁ e pisth mhn ape lipe n soi thÁ n crhsome nhn autoiÄ v. i 262±71. ± polla w SwÂkratev. Another will express the same thought as an advice. s. . 84 . how much wealth . cf. J. Certainly not `the peculiar manner called Socratic' (Grube). by another into an ox. this type of discourse is called Aristippean.1.' The Socratic type in the stricter sense of the phrase. kaiÁ paideu esqai proetre yato´ tauÄta pa nta hqikwÄ v kaiÁ e mmelwÄ v. oi on´ ``w paiÄ .

216 for a parallel expression. And all this in a personal and harmonious manner. it would seem that the idea as stated in the ®rst form is considered by Demetrius to be of a protreptic nature. These comments closely resemble 159 Cf. and they made a deeper impression than the others through their truth to life. We have to do not so much with three genres of philosophical literature or even of philosophy. lacked nobility. Now. the ®rst type of discourse had less e¨ect and. then.1. This type of speech met with great success when it was ®rst invented. I understand Demetrius to mean that in the ®rst manner there is a harsher sort of protreptic without aporia. 85 . This impression is reinforced by his remark that Socratic writings in the stricter sense became immensely popular and maÄllon e xe plhxan (were more shocking) on account of (among other things) twÄ i metaÁ megalojrosuÂnhv nouqetikwÄi (admonition with nobility.I N T R O D U C T I O N II.4 speaker has at the same time brought the boy into aporia without the latter realising it. from his comment on the `typically Socratic' manner: if this manner is `not after the proverbial Scythian fashion'. If. each connected with a Socratic author (or in the latter case authors). and this can hardly be anything but the ®rst manner.159 then something else must be. as with one idea which can be framed in three di¨erent ways. In the analysis of this passage it is generally overlooked that it is a sort of epilegomenon to the treatment of toÁ e schmatisme non. the speaker in the third manner is said to bring about aporia and to exhort his addressee to education at the same time. and has exhorted him to have himself instructed. while being equally admonishing. has reminded him that he is ignorant. man-of-the-world criticism). not as the saying goes `in the Scythian mode'. So much can be derived. their vividness and their man-of-theworld criticism. apparently. I think.

The related but not identical view that Plato's dialogues had a protreptic e¨ect on many is stated in so many words by Dicaearchus. Gaiser goes far astray when he claims Dicaearchus' support for his own esoteric views (CErc. `La biogra®a di Platone in Filodemo: Nuovi dati dal PHerc.I N T R O D U C T I O N II. cf. Die Berichte uÈber Platon und die Alte Akademie in zwei herkulanensischen Papyri (Stuttgart±Bad Cannstatt 1988). CErc 13 (1983) 53±62. M. The three types distinguished by Demetrius and his comments on the ®rst and third types coincide to a remarkable degree with statements found in Plato's Sophist (section ii. Gaiser. Innes in S. 161 Quoted by Philodemus (Acad.16). D. in the development of Greek literary theories' (D.161 And the conviction that the writings of the Socratics and especially of Plato give us an essentially protreptic picture of Socrates is expressed in Cicero's Academica Posteriora (1.. 313±21. the passage is often taken to say that the intention of Plato's dialogues was protreptic.1. op.11±15). Ind. 1021'. 117).4. Cf.. The Socrates of aporia and elenchos (ut nihil a½rmet ipse. K. Studies in Demetrius On Style (Amsterdam 1964). Halliwell et al. . (n. T. 148. Schenkeveld. 125: proetre yato meÁ g gaÁ r a pei rouv w v ei peiÄ n e p' authÁn diaÁ thÄ v anagrajhÄv twÄ n lo gwn. there seems to be a consensus nowadays that his doctrines re¯ect `an early stage . Yet Demetrius is not our only witness for this concept.1). K.3. 155). 61±2). C. cit.1. More generally. Philodems Academica.4 those contained in 279. It may be thought strange that the interpretation of the Socratic dialogue (at any rate the aporetic dialogue) as having a protreptic intention was already existent in antiquity (quite probably as early as the third century bce160). refellat alios) is there said in so many words to be protreptic 160 Though scholarly opinion about Demetrius' date is very divided. Filodemo. 86 . where Demetrius states that questions sometimes achieve more deinoÂthv than statements (kaqaÂper gaÁr eiv apori an agei toÁn akouÂonta e xelegcome nwi e oikoÂta kaiÁ mhdeÁ n a pokri nasqai e conti). PHerc 1021 col. Cf. Dorandi. the tripartite division is partly akin to the one found in Xenophon. which is obviously false. Storia dei Filoso®: Platone e l'Academia (Naples 1991). Gaiser. .

3.).13 (scripserit ). 71±5.12.18±19) in which philosophical discourse in general is trichotomised. 163 Att. Schmidt. Die philosophische PersoÈnlichkeit des Antiochus von Askalon (GoÈ ttingen 1940). 164 For reasons of space I cannot discuss various texts from Epict. 296sg.43.162 In this case. 12±13. De eloc. quod minime Socrates probabat. G. 13. (esp.1. `Socrates in the Academic books and other Ciceronian works'. esp. `Osservazioni sui tre ei dh touÄ lo gou in Ps.'. it is not implausible that they refer precisely to the Academica.163 We may connect Antiochus' description of the Socratic dialogue as an exhortation to virtue with the Middle Academic view of Plato and Socrates as sceptics: what may have been taken by others as dogmatism in the form of question-and-answer (and therefore may have been a cause of embarrassment for Arcesilaus and his followers) might conceivably have been explained away as protreptic. 406±12. Assent and Argument: Studies in Cicero's Academic Books (Leiden 1997).' should be taken seriously. 419. The main 87 . when Antiochus' expose of the history of dogmatism continues: ita facta est. A.I N T R O D U C T I O N II. in B. Philologus 106 (1962) 14±28.4. 14.-Demetrio.1. Inwood±J.52. 3.23. 12. 12. Antiochus and the Late Academy (GoÈ ttingen 1978). See A. 18. These passages have been related to Demetrius' three ei dh by E.3. id. We happen to know that this part of Cicero's book was derived from Antiochus of Ascalon (most probably from his Sosos). and the emphasis is so clearly on Socratic writings that it is beyond reasonable doubt that we have here a glimpse of the same theory of the protreptic character of Socratic literature that is found in Demetrius. J.4 (omnis eius oratio tantum in virtute laudanda et in hominibus ad virtutis studium cohortandis [ˆ protre pein] consumebatur). Lueder. Glucker.1.33±4. Fam.164 162 Cf. 3. There is a slight indication for this possibility in the same passage from Cicero. ars quaedam philosophiae et rerum ordo et descriptio disciplinae (1. 1. RFIC 96 (1968) 38±46.. 21.3±4. Zur geschichte einer antiken Stil.und Methodenscheidung'. Carlini. Antiochus. 58±88. `Die drei Arten des Philosophierens. Glucker.4. Mansfeld (eds.21.8. 9. 16.4. Att. Cicero's often-quoted words `apo graja sunt etc.17).

1. 1.1 as paraenesis) must have been likewise regarded as part of protreptic. 11..24 ˆ 8. Schol. and is probably identical with. Posid. ± If Demetrius considered elenchos part of protreptic. It is impossible to pass judgement on the connection of the ®rst type with Aristippus. SVF 1. the frequency of this word in Xenophon (as compared to Plato) has often been observed (statistics in JoeÈ l. cf.-H.113.44. fr. cf.1. Clem. 88 . See further my paper `Protreptic in ancient theories of philosophical literature'.1. Grouping together Plato and Aeschines under the heading of interrogation and aporia is only super®cially right: implicit and explicit protreptic have not been su½ciently distinguished in this theory (see further section ii. 2.1. Der echte . Plu.5 and section ii. however. advice.1. be maintained that the passage from Demetrius is the outcome of a process of re¯ection on ethical protreptic.1. though normally protreptic and paraenesis are kept apart in ancient theory: Stob. (Eudorus). Demetrius' example starts with the verb deiÄ .356±7.1. this cannot be identi®ed with any of the three carakthÄrev of Epict. Aristo.14±5 W. ad Demosth. Sokrates. i 467).5). Al.16± 67.1.±K. his second eidov (what I have de®ned in section ii.6 Dindorf. also [Isoc. Paed.] 1. the advising type (which comes close to.3. e pisthÂmhn deÁ ou sunapolei pousin thÁn crhsome nhn toiÄ v a poleijqeiÄ si is strongly reminiscent of the opening of Socrates' speech in the Clitophon. It may. There are some parallels. paraenesis) is rightly connected with Xenophon. the similarity to the theory contained in Plato's Sophist suggests that there existed a theory in which three types of moral exhortation were distinguished: accusation. it must be stressed that it is not a theory concerning philosophical protreptic in its totality. 176 E. 798b. . elenchos.66. . Protr.4 Whatever the origin of the theory found in Demetrius and Antiochus. Exhortation to philosophy (protreptic in the stricter sense) is not touched upon.I N T R O D U C T I O N II. Demetrius' example a nqrwpoi crh mata meÁ n apolei pousi toiÄ v paisi n. especially with what we have called `didactic'. cf. 1. esstumbling-block is Demetrius' ei dov u poqetiko n.

Gomme±Sandbach ad loc. if any (section ii. 13.. An. Clit.1). Or. Halliwell et al. 31.1. Chr. Radermacher. any explanation would do.206.2. Bachm. SSR iv a 148). 278 K.v. contra: Dittmar (Aischines. 121. 3. 2.). Demetrii Phalerei qui dicitur De elocutione libellus (Leipzig 1901). 28.-Plu..39. cf. perhaps Aristid.54 with a reminiscence of Clit. 131.2. Natorp.. P. Hld.320d (depending on Clit. There is a good reason to suppose that (3) was not a widespread usage. Chr. 21 Mannebach. Davies. supposing that it was meant in sense (1). Besides. Cf. C. Ox. cit. Philops.2. loc. Innes in S. Herm. 1.166 therefore. W. 29. 48) and G. (n. the Protrepticus would have been a good guess.3. this is accepted by D. Hamb. Them. 40.2). 166 232: o deÁ gnwmologwÄ n kaiÁ protrepo menov ou di' e pistolhÄv e ti lalouÄ nti e oiken all' a hpoÁ i mhcanhÄv (the insertion of a po (Ruhnkenius±Cobet) is necessary. Rhys Roberts. 2.297..2. as diaÁ mhcanhÄ v laleiÄ n makes no sense). is directly or indirectly dependent on that dialogue. Schol. (cf. Aristid. Eust.2. 165 Most scholars suppose that Demetrius' words. L. if this were true.12±3 (cf.59. 2.I N T R O D U C T I O N II. 905. Paroem. Alexis fr. 4e and Them. cf. fr. (3) people who deliver protreptic speeches. 115 n.1).1.336±7. section ii. are in fact a quotation from Aristippus (fr.. The combination of protreptic and the expression apoÁ mhcanhÄv is striking. 1. This expression is applied metaphorically to (1) people who act from a higher level than their surroundings. 86. (2) unexpected help. 258. op. D. Giannantoni.. 226 K. As (3) was used for the ®rst time in Clit. being presented as an instance of the eidov A risti ppeion. Or. If this were the only verbal resemblance between the two texts. Or. 26.18. 155).d. op. Die Sokratiker ( Jena 1922).165 But it can be shown that elsewhere Demetrius draws upon the Clitophon. misunderstood the expression in Clit. Or.14 (depending on Clit. 10. `Alcman fr.1. 35. 155). 519 n. Valk. 407a8 (wrongly of course). both Ps.210. Demetr. Cf.1). I submit the hypothesis that in this passage he depends on our dialogue as well. Aristippos 8). 656. PW s.. RhM 135 (1992) 94).±A.. Carlini (cf. Clit..4 v. loc. previous note) ®rst proposed the Clitophon as the source. section ii.4 pecially twÄ n d' ue wn oi v tauÄ ta paradwÂsete h* * *i opwv e pisthÂsontai crhÄsqai dikai wv tou toiv (407b3±4. M..14 K. as we shall see. cit. these words are a combination of two protreptic motifs found ®rst in the Clitophon ± section ii. (1958)). W.. 106 again'. cit. quotes the phrase with the express addition of wv e jh tiv.d. An.. cit. cf. P. (n. D. Men. 67±8. Luc. D. not in a source.300. Nestle.5±9 K. Demetr.21. I Cirenaici (Florence n. 89 .

13.2. 111.I N T R O D U C T I O N II. In the Euthydemus.1. toÁ gaÁr zhteiÄ n kaiÁ jilotimeiÄ sqai opwv tiv e stai kaloÁv kaiÁ agaqoÁ v ouk allo ti ei nai h toÁ jilosojeiÄ n).28: kaiÁ outwv dhÁ pareka lei [Socrates] proÁv toÁ e pimeleiÄ sqai kaiÁ prose cein autwÄ i toÁn nouÄn kaiÁ jilosojeiÄ n. of which the outcome is anagkaiÄ on einai jilosojeiÄ n (282d1). If Antisthenes' ®rst three Protreptikoi were indeed entitled PeriÁ dikaiosu nhv kaiÁ a ndrei av.1). Chrysippus (Plu. cf. SSR v a 105 ± to the parallels collected by Caizzi. It is interesting to note that in his PeriÁ touÄ protre pesqai. 1039ef ˆ SVF 3.1. though it must be stressed that Plutarch considers the correction to be Chrysippus' own.5 The development of philosophical protreptic In the fourth century.168 Philosophy is not named 167 kaiÁ jilosojeiÄ n (usually bracketed) is indispensable.5 ii. b 110 DuÈring) together with a correction of two lines of Theognis (175±6). and the sub-title PeriÁ Qeo gnidov certainly points that way. Arist. 90 .167 h idei gaÁr oti touÄto zhtouÄntev ou deÁ n a llo poihÂsousin h jilosojhÂsousi. the accent may have been put by various writers on either.167) mentions Antisthenes' dictum deiÄ n ktaÄ sqai bouÄ n h bro con (fr. Still. 188) coincides with exhortation to philosophy (that these were identical was noticed by Dio Chrysostom. explicit philosophical protreptic appears to have been the domain of Socratic philosophers (if Aristotle may be included among them). add Clit. 67 Caizzi. p. ii.2 n. the most central concept of fourth-century protreptic. As I have already remarked (section ii. he ®rst gives a demonstration of his own. 408a4±7. this explains why exhortation to virtue and to the care of the soul (e pime leia.1. which (in their corrected form) point the same moral. 168 Presumably Antisthenes used some lines from Theognis as a startingpoint for protreptic discussions. Protr. his exhortation was primarily ethical. The ®rst gaÂr is only explicable if in the preceding sentence e pimeleiÄ sqai and jilosojeiÄ n have been identi®ed. Socrates asks the eristics to show by their exhortation that it is necessary soji av kaiÁ arethÄ v e pimelhqhÄ nai (278d2±3).

crhÄ sqai. it appears to me a likely hypothesis that protreptic to philosophy sets in later than. ethical sense. This development is illustrated by the di¨erent use of the phrase `using one's soul' in the Clitophon and in Aristotle's Protrepticus. for that matter. cf. On the other hand. Isocrates' criticism (15. on 408a5 yuchÄi . protreptic in the wider. Aristotle's Protrepticus is a clear specimen of philosophical protreptic in the stricter sense. says nothing. by which Aristotle was strongly in¯uenced (cf. .1.1. that virtue may only be attained through the possession of ethical knowledge (whatever its nature). The protreptic dialogue exhorts to e pime leia. Mem.169 Finally. as Aristotle did not.1) seems to ignore exhortation to philosophy (in the stricter sense). no occurrence in Alc. 1. this specialisation. is natural. exhortation to virtue and exhortation to philosophy become two di¨erent things. only in passing. `Epinomis'. . philosophy is mentioned.84±5. `Epinomis'. This is clear not only from the frequency of the word jilosojhte on but ®rst and foremost from its content. Though the scarcity of our material does not inspire much con®dence. The ®rst protreptic conversation of the Euthydemus. to which protreptic was to be restricted eventually. `The `wisdom' of the Protrepticus is .5 in Socrates' protreptic speeches in the Apolog y (nor. Comm. 264±5).23. Once the Socratic paradox `virtue is knowledge' has been abandoned. 265±6). foreshadows this transition. not so much the science of virtue of the Socratic dialogue as a theoretical science dealing with nature. if at all. The anonymous critic or critics reported by Xenophon criticised Socrates for being able only to exhort others to a reth . and develops out of. 91 . section ii. in the Clitophon). and things divine' (Einarson. 169 X. truth.I N T R O D U C T I O N II. 4. Alc. its absence from the fragments of Aesch. Einarson.2. which we ®nd from Aristotle onwards. . Since the Socratics (including Plato) held. .

See also on 407d8±e2.1. The ®rst protreptic conversation in the Euthydemus and Aristotle's Protrepticus seek to convince by argument. and the facts bear this out.1 n. section ii. that a di¨erence in form runs more or less parallel to this difference in content. 104. had mainly an accusing character. Even the protreptic dialogue seems to support this hypothesis: though it employs elenchos and is opposed as such (together with Plato's dialogues) to accusing protreptic by Demetrius. more will be said on Plato's evaluation of protreptic. 171 The role of Demetrius' advising `character' is di½cult to grasp. The diatribe. not by (negative) statements (apojainoÂmenov) or thundering speeches (kathgorwÄn). the dangers with this term as set out in section ii.2 n. 92 . The fact that the diatribe is not a separate genre in ancient theories of philosophical literature may explain why theories of protreptic always treat protreptic as ethical.I N T R O D U C T I O N II.5 It would appear.L. it is possible to infer that rhetoric and accusation went hand in hand in these works (as they do in the Clitophon).171 Much more could be said about the subject. Pl. 6. This character would seem to be inconsistent with philosophical protreptic in the stricter sense. successor of ethical protreptic.4. 164). Xenophon probably restricts protreptic to accusation and of Socrates' protreptic speeches reported in Plato's Apolog y one at least (29d7±e3) is clearly reprobatory.1. the phil170 Cf. 729a2±b2). because it blurs the distinction between protreptic and paraenesis. When we learn that Antisthenes' Protreptikoi were among his most rhetorical works (D.1. Ethical protreptic. Ap. as do some passages from the general preamble to the laws (Lg. The relationship between Aristotle's Protrepticus and Plato's theory and practice of protreptic (both explicit and implicit). 36c5±d1 seems to belong under this heading.170 quite often has an accusatory character. as I point out in my paper `Protreptic in ancient theories' (cf. perhaps. 726a5±727b4. although we are here on thin ice. there is always a central moment of reproof.1).

I shall now turn to that dialogue (section ii. are promising objects for further research. or various motifs from various protreptic texts are being mixed together into a protreptic pastiche. the place of the Clitophon within the development discussed will be treated in section ii. the di¨erence in using protreptic motifs between Aristotle and his predecessors. But as we are concerned here with the interpretation of the Clitophon. It has been shown already (section i. the structural similarities and di¨erences between the protreptic dialogue and the protreptic conversations in the Euthydemus.3). We must ask at the outset whether one item in the protreptic corpus is being parodied throughout.7. The question cannot be answered with absolute certainty.5. ii.I N T R O D U C T I O N II.2 osophical considerations which prompted Aristotle to take up a genre which his teacher had condemned.3) that Clitophon's report of it is surrounded by clear marks of irony ± these and Clitophon's own evaluation of Socrates' exhortation (410b6±7 tosouÄton mo non duÂnasqai. and the light these may shed on the reconstruction of Aristotle's Protrepticus. as our knowledge of protreptic literature is limited. A de®nite judgement must wait until Plato's attitude towards protreptic has been studied (section ii. so much so that the burden of proof lies with those who claim that the Clitophon derives from one speci®c source. makroÂteron deÁ oude n) constitute the justi®cation for this hypothesis. however.1).2 Protreptic in the Clitophon I use as a working hypothesis the assumption that Socrates' protreptic speech is intended as a parody. subsequently I shall make observations 93 . There are. I shall ®rst discuss two such claims. then study the structure of the protreptic speech.2. strong prima facie indications for the second alternative.

28.2.2. 1. wsper a poÁ mhcanhÄ v qeo v. ii.1.172 von Arnim claimed that both Dio and the author of the Clitophon use one of the three Protreptici of Antisthenes. . it goes on and on. i 118 n. on 407a6 pollaÂkiv] pantacouÄ te kaiÁ proÁv apantav [cf.1 on the various protreptic motifs as we encounter them within the speech. its central theme is a motif found also in the Clitophon. this source must have been 172 Dialog. The speech which follows.14±15 ± these words smack of excessive modesty.I N T R O D U C T I O N II. they could be an excuse for almost any degree of modi®cation). As Dio mentions the battle of Cnidus (13. presented as Dio's own imitation of Socrates' speech. A second speech follows.1 The Clitophon derived from one source ii.5.10±12). . `Der lo gov Swkratiko v'. Dio In the thirteenth Oration. but carefully left out of the ®rst speech: the search for teachers of justice.26). developing the same theme of the insu½ciency of the present curriculum up to 13. wv e jh tiv [see note 174] . section i. allaÁ ple on h e latton ei pw ti. JoeÈl. an ara mhÁ duÂnwmai apomnhmoneu esqai a kribwÄv a pa ntwn twÄn rhmaÂtwn mhdeÁ olhv thÄ v dianoi av. legoÂmenon u po tinov Swkra touv. Therefore he referred e pi tina loÂgon arcaiÄ on.] bowÄ n kaiÁ diateinoÂmenov [`straining his voice'.2 ad ®n. .1 The Clitophon and Dio's thirteenth Oration  qhÂnaiv periÁ jughÄv. reads as a close paraphrase of Clitophon 407b1± c2 memaqhkoÂtav ± after that.] . kaiÁ h xi oun. a misunderstanding of 407a8 umnoiÄ v.17 ameinon oikh sein thÁn poÂlin. Comm. suggnwÂmhn e cein (13. ad loc.  En A relates how during his exile he came to be regarded as a philosopher and many urged him to utter his views on moral problems (13. up to 13. 94 . K. o n oude pote e keiÄ nov e pauÂsato le gwn [cf. it has little in common with our dialogue. cf.2. cf. 64±5. Following a suggestion made by Hirzel. .

220±1. a reference to an utterance of a character within a literary text.19. Phdr. What seems a highly functional element of ridicule in the Clitophon now becomes. as always in this author. 314±15) and P.174 I have indicated other parallels in the critical apparatus. `Kleitophon wider Sokrates'. only H. `Zu Antisthenes'. P. De Dione Chrysostomo Xenophontis sectatore (Gothae 1896). Kleitophon'. J. Sokrates. Von Arnim's hypothesis (adopted by JoeÈl. 256±60. The characterisation of Socrates as a qeoÁv apoÁ mhcanhÄv is the most important and destructive one: this expression must. . too.23±4 de BudeÂ) with Pl. 90 n.347. the words wv e jh tiv in Dio are. 242c3±5 (Socrates). 3). Studies in Platonic Chronolog y (Helsinki 1982). and not long after. a clumsy indication of eulogy. Cf. without arguments. H. 20.g. 171±3. BruÈ nnecke. 25. dismiss it (Pavlu. K. 95 . i 483±5. Cynic Hero and Cynic King (Uppsala 1948).1 written after.315. whereas the Clitophon o¨ers a short recapitulation. 10±11. Dio is thought to follow Antisthenes closely. The similarity of Dio 13 and Clit. `Pseudopl. was ®rst seen by Hartlich (`De exhortationum . Thesle¨. `RaÈ tsel'.2. If all this was already in one source of the Clitophon. not to a comment made by its author. 430±1) or ignore it altogether (e. Euthd. 305c6±7 (Prodicus). Sokrates. Hagen. 18. probably.19±20) with Pl. Two of the results of justice as de®ned in the discussion of Clitophon and Socrates' pupils (toÁ sumje ron and o moÂnoia) are referred to in Dio 13. Maier.3 (ˆ 2. the closing words of the Clitophon are echoed in Dio 13.13 (ˆ 1. Antisthenis fragmenta. Der echte . . Desideri.4 (ˆ 2. R. 394 bce. Dione di Prusa (Messina±Florence 1978). Caizzi.35. 421b3 and 345c5 (Socrates). 86 n. ii 545. Th. Philologus 50 (1891) 381±4. as von Arnim admits. 1) is nowadays usually accepted in treatments of Antisthenes or Dio (R. we are driven to 173 Leben und Werke des Dio von Prusa (Berlin 1898). Exemplum Socratis (Wiesbaden 1979). 174 Besides. 253 n.173 Probably the most decisive argument against von Arnim's thesis is the presence of various reminiscences of the Clitophon throughout the thirteenth Oration. 287 n. DoÈring. Gomperz. HoÈ istad. historia'. 56±64 had argued extensively for direct dependence of Dio on Clit.25±6) with Pl. .I N T R O D U C T I O N II. 24. Wegehaupt. have been used by Antisthenes himself. Souilhe ). . 92±3. 206 accepts it. at best. 1. Griechische Denker.3. Ge¨cken. while authors on Clit.

27±8). 150±6). cf. (n. Diss. philosophy is de®ned as the science of living the life of a kaloÁv kaiÁ agaqoÂv. whereas they are perfectly plausible if regarded as Dio's own variations on a fourth-century theme (to wit.28 Dio says that Socrates eschewed the word jilosojeiÄ n. Ench. 95. 96 .178 175 It would ®t in well with Dio's time. 90. 173).21. the dissimulation of the term jilo sojov is typical of practical philosophy of Dio's own time. 3. esp. 177 Epict. 4. op. 23. 2. Dio has integrated Prodicus' allegory of Heracles at the crossroads.1.1. cit. 3. op. this is predominantly a concept of later Stoicism. If the whole of 13. 4.19.1. (n. HoÈ istad.21±34. (n.16±27 is in essence (apart from rephrasing on Dio's part) a fourth-century text. 470d5±e11. the continuation is evidently Dio's own invention. 173). In that case.176 In 13. NH 7. 46. 17. cit. borrowed from X. the Clitophon). cit. cit. 176 Cf. 3.1 the assumption that the whole of the Clitophon is an extract of a larger but wholly similar work. Mem.24 the Persians are praised for their education because they thought it disgusting to strip and to spit in public ± would any fourth-century Greek have considered the observance of these taboos a mark of paidei a?175 Towards the end of the speech (13. 29±42 purports to be a Socratic dialogue. Cf. (contra) Wegehaupt.23. All the same. Chr.8 passim. At 1.2.113. 64. At 13. Ep. there are some features which seem strange at best. 13.80: Antonia minor is said numquam expuisse. 182±9. It so happens that it is absent from the Clitophon (as are jiloÂsojov. (n. 173).1 is derived from Grg. See too D. Plin. 117.27. 105±6.177 I therefore conclude that Dio used the Clitophon for the nucleus of his Socratic speech.I N T R O D U C T I O N II.12.7.11. op. into a much larger passage of his own making (contra: HoÈistad. von Arnim's thesis cannot be falsi®ed any more and should therefore be abandoned. 178 Similarly. -i a). DoÈring. and embellished it with various motifs culled from everywhere. 173). op.1. cf. Sen.66±83. and Dio was an acute reader.

Aristotle. Themistius gives a slightly condensed and modi®ed version of Socrates' protreptic speech as reported in the Clitophon.2 The Clitophon and Themistius' twenty-sixth Oration In a fair number of publications.180 Plato is supposed to have answered this attack in supplements to Phaedrus and Republic as well as in Protagoras and Clitophon. kaiÁ otan le ghiv. as we now have it. . 180 Consisting of Books 2±4 and maybe 8±9 of our Republic. 97 . Kesters tries to save his case by postulating an impersonal use for the second person singular (AntistheÁne 80). Kesters has stated and defended his conviction. up to 272 of Phdr. 407e5. both Clitophon and Themistius change from direct to indirect speech at the end of the ®rst part of the protreptic lo gov. where Socrates is being confronted with his own words. 179 To wit.g.1 ii.I N T R O D U C T I O N II.1. the second person singular would seem to be the result of Themistius' carelessness in adapting the quotation from the Clitophon: the direct speech is put into the mouth of Philosophy herself. and that he also modi®ed lineends in order to get clausulae which do not violate Meyer's Law. According to Kesters. on the other hand. . intended as an attack on a ®rst version of the Phaedrus179 and on a proto-Republic. who is nowhere apostrophised. the indirect speech naturally depends on the second person singular (kaiÁ o poÂtan au jhÄ iv . e8). In the Clitophon. Now. this version is the original one. Kesters' theories cut across the consensus concerning the chronology of Plato's writings. Epicurus and Carneades in order to `modernise' his fourth-century source.2. what we read in the Clitophon is Plato's quotation of it. H. that what we know as Themistius' twenty-sixth Oration is in fact a work from the fourth century bce. and require the assumptions that Themistius introduced the names of e. In Themistius.2.

Protreptik. cf. that there are several grave infractions of Meyer's Law.2. section ii. Clit. which Themistius has in common with the Clitophon. Die antike Kunstprosa (Leipzig 19234). are to be explained from the semi-dialogical character of the work.2.181 Secondly. for which cf.5 (Downey±Norman) u miÄ n e stai. Leuven 1931. it is precisely in the protreptic speech. 98 . He was. 183 Platoons Phaidros als strijdschrift. inasmuch as it shows a remarkable lack of congruence 181 Cf.±D.3. 45. 3. cf. 64 n. 557e2±558a2. Ath.4 e stiÁ n ara.134. ii. In parts which do not correspond verbatim to the Clitophon there are no infractions. to give one example. 156). E. The instances adduced from ps. 1.21±2 crhÄ sqai jarmaÂkoiv appears to have been preferred to jarmaÂkoiv crhÄ sqai (cf. the ®rst scholar to note that the plan of the third part of Socrates' speech was derived form Republic 1 (discovered independently by Gaiser. KG i 557 Anm. i 109±10. 182 1. Schw. which is elsewhere in the twenty-sixth Oration (as in all of Themistius' works) rigidly observed. but I would like to say that Kesters' interest in the Clitophon has also led to positive results.-X. Wackernagel.3. 407e10 ojqalmoiÄ v crhÄsqai ) in deference to this rule. 135. 143 n. 244. refrained from doing so in this passage because he knew it was present also in the Clitophon: in order to cover up his alleged plagiarism he treated the protreptic speech as if it were a quotation from Plato. Resp.134. and he did so by not applying Meyer's Law precisely there. who allegedly rephrased the whole of a speech by Antisthenes (or another fourth-century author).I N T R O D U C T I O N II. Norden. No further comment is needed. also R.182 Kesters' explanation of this state of a¨airs183 shows how shaky his whole theory is: Themistius. 12±3 nuÄ n paideuÂsewv.2 but in contexts where there is no true dialogue this use seems to be non-existent in Greek. Vorlesungen. ii 922±3.2 The structure of the protreptic speech The composition of Socrates' protreptic speech is curious.2. 21 tou tou crhÄsin.

ad loc. . As far as I know. which has nothing to do with the last part any more than with the ®rst part of the speech. viz. 39±44) are the only ones to observe the tripartite form. It is not very hard to see why the oratio recta is abandoned at the end of the ®rst part.).2 between expression and content. But the part in oratio obliqua is far from being a unity: by far the greater part of it deals with the principle that you must leave alone what you cannot handle (stated 407e8±9. not to the last part beginning with kaiÁ otan le ghiv. cf. the speech is usually stated to be bipartite (so Pavlu.2. which is the subject of the second part (e5± 8).). which refers back to the previous section. An examination of what Socrates is actually saying seems to con®rm this division at ®rst sight: there is hardly any relation between the direct speech and the reported (apart from e teroÂn ti . there are clearly two distinct parts. These are set o¨ from each other not only by the fact that the former is in oratio recta. Yet the author apparently wished to make this accusation within the same speech (e5 toÁ e jexhÄv touÂtwi. constitute an entirely independent whole. the latter in oratio obliqua. `Pseudopl. esp. hence I suspect that o lo gov outov (408a4±5) refers to the whole speech. toiouÄton 407e6±7. . ad loc.2). 451±2) and Kesters (KeÂrygmes. e spoudake nai (407e5±8). worked out in the rest of the speech) ± yet the words which precede the statement of this principle have no connection whatsoever with it: the words kaiÁ o po tan . BruÈ nnecke (`Kleitophon wider Sokrates'. but also by the use of (progressive) au and the phrase toÁ e jexhÄv touÂtwi in 407e5. . Socrates had been vituperating all mankind for being interested only in wealth (407b2±3 crhmaÂtwn meÁ n pe ri thÁ n paÄsan spoudhÁn e cete) ± he could not very well go on to accuse them at the same time of attaching too much importance to the e pime leia touÄ swÂmatov. Kleitophon'. though not even separated from the following by sentence end (cf. 99 . If one looks at its form. 11).I N T R O D U C T I O N II. . 407b1±e2 and e5± 408b5 (section i.

cf. cf. cf. 246d8±e7. d8. swÄma.5±6 des Places). Protr. body is taÁ autouÄ (131a2 and passim). yuch are simply juxtaposed. instead of making Clitophon report three separate speeches by Socrates. Apologie. with the Delphic maxim.I N T R O D U C T I O N II. This scale is worked out extensively in Alcibiades 1. where it is coupled. Our author must have had rather a strong reason for lumping together three unrelated exhortations into one speech. Phlb. incidentally.. b 59). a passage which belongs to the protreptic corpus as de®ned in section ii. 697b2±6 ± where. Cf. where yuch is equated with Man himself (130e8±9 yuchÁ n a ra hmaÄ v keleuÂei gnwri sai o e pitaÂttwn gnwÄ mai e autoÂn). 743e3±6. tri ta (Protr. 92.3. the protreptic character of which has been 184 Cf. 185 Cf. also Arist. This passage. 59. d12). Ep. 233±5.2 407e8).184 a very similar one (crhÂmata ± timh ± yuchÂ) is one of the key doctrines in the Apolog y185 and the Phaedo. Iamblichus found his division into prwÄta. Lg. a11) or taÁ twÄ n e autouÄ (133e1±2. as in Alc. Meyer. there are many more places where crhÂmata. This trichotomy of values is genuinely Platonic. Apolog y. 8 355b2±6. de Strycker±Slings. Gauthier±Jolif on EN 1098b12±14. Apolog y.3 n. 1. 30a7±b4. In these places the order is clearly a hierarchy. It is re¯ected in Aristotle's Protrepticus (b 2 DuÈring ± elsewhere only body and soul are opposed: b 17. 48d8±e10. 98 n. section ii. on Ap. deu tera.2. 137±8. There is in Plato one passage which also is constructed so as to re¯ect this trichotomy. wealth e ti porrwte rw twÄn e autouÄ (131c1. I refer to the ®rst part of the general proem to the laws concerning mankind in Laws 5. which precedes an extract from Clit. 100 . Mx. This reason I take to be his wish to give the speech a structure which itself re¯ects a theme not unknown to protreptic literature: the hierarchic scale of values crh mata ± swÄma ± yuchÂ. b 52 DuÈring (which I reject. so the provisional end at e2 and a few words of appreciation before the speech went on in a reported form were necessary to mask the discontinuity.1. 135±40 and the references given there.1. cf. de Strycker±Slings. 126).

Apologie. Of course.2 stressed by several scholars. who alone gives a correct articulation of this preamble.I N T R O D U C T I O N II. iii 467±70).) Although the trichotomy is usually said to be traditional (or even Pythagorean and `Orphic').2. 553±4. The Laws (Harmondsworth 1970). The similarity in pattern could be a coincidence but for the words already quoted: paisiÁ n deÁ ai dwÄ crhÁ pollh n. strangers. (Cf. the poÂliv. HGPh. Platon.1(2). but the higher status of the body as compared with wealth certainly is (Meyer. 101 . continues with the body (728c9±e5) and wealth (728e5±729b2). though in inverse order. ou crusoÁn katalei pein (729b1±2). ou crusoÁn katalei pein) and Plato continues with other precepts on the behaviour towards the young. and ®nally a long sermon on personal behaviour (730b1±734e2). Though the structure of the general proem is more complex than that of the protreptic speech in the Clitophon. The last passage ends with the value of riches for children (paisiÁ n deÁ ai dwÄ crhÁ pollh n.4 and n. there is no indication that it did not originate with Plato or Socrates. These words stand at the very end of the trichotomic pattern in the general proem whereas nearly the same thought is found at the beginning of Socrates' protreptic speech.2. and appended the rest of what he had to say after he had mentioned the young. cf.186 is announced at the end of Book 4 as follows: taÁ periÁ taÁ v autwÄn yucaÁv kaiÁ taÁ swÂmata kaiÁ taÁ v ousi av (724a7±8). sections ii. The proem itself starts o¨ with re¯ections on the soul (726a1±728c8). Protreptik. 214±17. ii. 187.7. 234. J. 98 rightly says: `Von Bedeutung ist. friends and relations. Plato. T. the identi®cation of yuch with man himself is not novel (cf. it is clear that Plato started this preamble with a pattern identical to that of our protreptic speech (cf. dass KoÈrper und Besitz 186 Wilamowitz uses the term `Predigt'. Saunders. especially the announcing words). i 545±6. Gaiser.

which has a formula e auto n ± taÁ e autouÄ (36c5±6) related to. 631b6±d1 (cf. timai (b2±3).2. to wit plouteiÄ n (279a7). the trichotomy e autoÂn ± taÁ e autouÄ ± taÁ twÄn e autouÄ found in Alc. ta lla kataÁ toÁ swÂma i kanwÄ v pareskeuaÂsqai (a8±b2). eu daimoni a (sources: ka llov. The last part of the speech combines the motif of e pi stasqai crhÄ sqai with several themes of the Alcibiades dialogues. O'Brien. So in the Apolog y. i scuv. education are brought together into a rather coherent argument ± all of these thoughts occur elsewhere in protreptic literature.) The Euthydemus contrasts a whole set of agaqa (which afterwards prove au taÁ kaq' au ta . 93). kaloÁn einai. to be ®lled in with various protreptic motifs.2 nicht als slechthin wertlos dargestellt werden ± das waÈ re kynisch ±'. ugiai nein. sw jrona ei nai kaiÁ di kaion kaiÁ andreiÄ on (b5. but they appear to be kaka as well ± the whole passage is meant as a proof that Euthydemus does not possess self-knowledge (30: o po qen deÁ crhÁ arxasqai e piskopeiÄ n e auto n) and constitutes a parallel to the discussion in Alc. 180±5). plouÄtov. 102 . above. X. soji a (`cleverness'. 140.) as agaqa . cf. The statement is just as valid if one substitutes `Pythagorean' or `Orphic' for `kynisch'). Protr. Daedalus and Palamedes are given as examples of sojoi ).187 Our author used the trichotomy as a formal pattern for Socrates' speech. A highly sophisticated version of this dichotomy `external agaqa ± a gaqa of the soul' is found Pl. ou denoÁv axia. two versions of the trichotomy are also found in the Apolog y. Arist. Protreptik. the external agaqa are lumped together indiscriminately and opposed to the arethÁ thÄv yuchÄ v. 4. du nameiv. kaÂllov depend on the state of one's soul. Aischines. 1. (Of course. justice.31±5 proposes u giai nein. but incompatible with. . Mem. Dittmar. I think we have discovered here an important feature of our author's method of parody: he uses patterns taken over from other Socratic literature in order to furnish them with protreptic motifs unrelated to this patterns.2. Paradoxes. . 281d8±e1). i scuv. eu ge neiai. under the head of crh mata a number of thoughts on wealth. b 4 DuÈ ring makes plouÄ tov. do xa etc. 83±6. tacitly dropped later on because they would upset the conclusion 281d8±e1 quoted above) with soji a. Thus. Gaiser.I N T R O D U C T I O N II. 155. We 187 In other protreptic literature. Lg. 1 (cf. What is said in the short passage on swÄ ma returns in Alcibiades 1 and elsewhere.

87.3).3. jronhÂsewv de . Arist. 408b2 paradoÂnti taÁ phdaÂlia thÄv dianoi av a llwi (the concept of the politician±educator). See HGPh. from 124b7 on the discussion centres on the precise nature of the e pime leia.3 shall see that he does the same thing en deÂtail in the speech (cf. y. 94. The place of the Delphic maxim in this connection (X. iii 467±73. q'. Mem. Comm. fr. (SSR vi a 50).2. Yet there is some kind of unity between the parts: they all end on the keynote of e pime leia thÄ v yuchÄv. cf. Ap.24. Protreptik.2. In fact.52 D. 29e1±3.2. `Charmides'. . It is especially prominent in the Alcibiades dialogues: Aesch. b 34. 6.1. Pl.2. 78 n. 8.2.2 (where it refers to special training in politics: 4.1 and ii. . on 410c8±d5. 124b3). 36c5±8. . sections ii. 306e2±3. 407e7 touÄ meÁ n arxontov ameleiÄ n. ouk e pime lhi oudeÁ jronti zeiv (29d9±e3). 188 This repetition of e pime leia reinforces my opinion that o lo gov ou tov (408a4±5) denotes the whole speech. Alc. . d4 (this is still political. Alc. the central concept of Socratic exhortation: 407d8 kaiÁ deiÄ n e pime leian thÄv nuÄn plei w poieiÄ sqai.3. Clit. 1 119a9. 53 DuÈ ring.1 First part The pattern in which the opening sentence of Socrates' exhortation is cast is the outstanding one of accusing protreptic (section ii. 408b7±8. 4. c9±d1. 82. E¨e. 120b6.2. Euthd. 7). ˆ 1 Kr.188 ii. not the third part. 103 . 202). X. not about the true Values p.I N T R O D U C T I O N II.5): `you care about the pseudo-Values x. ± For e pime leia au touÄ (thÄv yuchÄ v) as protreptic motif. 30a9±b2. 1. Protr. 275a6. the resemblance in wording and construction between these passages is striking: Apolog y: crhmaÂtwn meÁ n (a) ouk ai scuÂnhi e pimeloumenov (b) o pwv soi e stai w v pleiÄ sta (c). 1 129a2) is probably derived from Aeschines (Gaiser. This is the form Plato chose for reporting Socrates' exhortations in the Apolog y: crhmaÂtwn meÁ n ou k aiscuÂnhi e pimelouÂmenov . cf.3 Protreptic motifs in Socrates' speech ii. Clitophon: crhmaÂtwn meÁ n pe ri (a) thÁ n paÄsan spoudhÁ n e cete (b) opwv u miÄ n e stai (c). Mem.3.

the best state of the soul (Apolog y 29e1±2) are left unmentioned (and therefore unharmed) in the parody. It must be stressed that whereas our author parodies the general form of protreptic as found in the Apolog y (in the sentence 407b1±8). . cf.41. our author imposed three di¨erent protreptic 189 Cf. Phaedo ap. On the pattern of accusatory protreptic derived from the Apolog y. kaqeu dontav] e geiroÂmenoi (31a4±5). a8 pei qwn. 104 . 30a1 oneidiwÄ.3 Note especially the anticipation of crhmaÂtwn (pe ri). a lhÂqeia. where verbal resemblances indicate dependence on the famous comparison of Socrates with a gad¯y: a tecnwÄv [!]. on 408e5). which in either place is put into focus by lifting it from the o pwv clause. for Clitophon's subsequent discussion with Socrates' pupils (esp. its content is not aimed at. for a similar manner of saving Socratic principles while attacking them on the surface cf. the author knew very well that the best state of the soul was the object of Socrates' protreptic.I N T R O D U C T I O N II. but the Apolog y itself.189 Conscious imitation is the more probable since in the Apolog y (as here) the simile applies to Socrates' exhortations (the words e gei rwn kaiÁ pei qwn kaiÁ oneidi zwn. Comm. 408e9 o pwv w v be ltiston e stai toÁ swÄ ma. 94. 30e7. proskei menon thÄi poÂlei u poÁ touÄ qeouÄ wsper i ppwi megaÂlwi . However. 408e3±409a3) carefully brings out the point (cf. ei kaiÁ geloioÂteron eipeiÄ n. Sen. 409a3 thÄi thÄv yuchÄv a rethÄi. It is therefore probable that the author of the Clitophon had in mind not just pieces of accusing protreptic as exempli®ed in Apolog y 29d7±e3. . Comm. the Socratic values jroÂnhsiv.2. Imitation of the Apolog y is also found in Clitophon's comment on Socrates' speeches atecnwÄ v w sper kaqeu dontav e pegei rein h maÄv (408c3±4). point backwards to Socrates' report of his protreptic practice. deome nwi e gei resqai [!] upoÁ mu wpoÂv tinov (30e2±5) and wsper oi nustaÂzontev [cf. Ep. cf. on 407d2± e2).

9. 36 D. but also oneself. Mem.2. 191 Cf. which is a `protreptic situation' (306e1±3. The link between motifs (1) and (2) is laid in an inge190 Cf. Aischines.2. 397e3±10 (rather speculative Gaiser. This is the closing remark of the Euthydemus. This is found in the Euthydemus.3) that what one cannot handle should be left alone. but a derived one (the care for one's sons). Erx. Protreptik. Protreptik.190 this is an illustration. Dittmar.3. 1. all of which are present also in the Euthydemus. 357e5 (see on 407b5 didaska louv).4 n. cf. Pl.2. Oec. 283). 74±7). Lg. 105 .4. the thought is also found in the protreptic proem of Laws 5 (729b1±2. These three motifs. Aeschines fr. ii. section ii. section ii. sections ii.4.8±14.I N T R O D U C T I O N II.3. are the following: (1) The amassing of wealth is useless without the knowledge how to use it. (2) It is better to leave wisdom to your children than money.1.6. cf. but a more or less compulsory one. 87e5±88a5.1. ˆ 17 Kr. 4. As has already been noted.3 motifs ± this accounts for the remarkable structure of the opening sentence: opposed to the pseudo-Value (wealth) is not a true Value (such as jroÂnhsiv). section ii. 62±3). This combination of unrelated form and content is typical. This is one of the reasons for the supposition that Demetrius (296.191 The combination of (1) and (2) is absent from the Euthydemus and seems to be an innovation of the Clitophon. also X. cf. (SSR vi a 75).2. 3. of the general principle (to recur in the third part of the protreptic speech in the Clitophon. 203±5.11 (Gaiser.2) derived the combination from the Clitophon itself.2). (3) One should have not only one's children instructed in wisdom. both in the ®rst conversation of Socrates and Clinias (282a7±8) and in the ®nal conversation of Socrates and Crito.3. Men. 307c3±4. Pl. section ii.2 ad ®n. Prt. See Euthydemus 280b8±d7. 804d4±6.

the sons are not mentioned again in the speech. in this way the protreptic speech can resume its accusatory character. 106 . which either alone or in conjunction account for the presence of areth (soji a. whether or not there were indeed teachers of areth was a hotly debated question. Therefore we should pursue wisdom with all our might' ± Ei e sti ge w Kleini a hn d' e gwÁ h soji a didaktoÂn. The normal expression (e pi stasqai) crhÄsqai orqwÄ v is changed to e pi stasqai crhÄ sqai dikai wv in (1). so that Clitophon's interrogation more or less automatically concentrates on justice and its results (409a6 and passim). . . In this way. and wisdom to dikaiosu nh in (2). Now. Comm. on 407b5 didaskaÂlouv. justice becomes the focus of Socrates' exhortation (cf. The same proviso occurs in a parallel development in Socrates' protreptic discussion in the Euthydemus: `We cannot become happy unless by the correct use of things. The teachers of justice are dragged in so as to connect (1)±(2) with (3) ± as we shall see.3 nious way. i kanwÄv was necessary to delude Socrates' audience into accepting their existence for the moment. Apart from 407c1±2 u maÄv te autouÁv kaiÁ touÁ v paiÄ dav umwÄn. The teachers of justice were necessary for the connection of (1)±(2) and (3). twÄn d' u e wn. . dikaiosu nh). 407b2±3) to the addressees of the speech themselves. therefore the parenthesis 407b5±7 ei per maqhtoÂn .I N T R O D U C T I O N II.2. 408b5). already at 407c5 they are left out of account: oude zhteiÄ te oi tinev umaÄ v pauÂsousi ktl. In its fullest form this theory. . they are abandoned as soon as they have done their duty ± and (3) in its turn was necessary in order to divert the attention from the sons (who had been given a prominent place in the opening antithesis crhma twn me n . This parenthesis is consequently a more or less necessary proviso. Another point which the two places have in common is that they both omit one of the three conditions. allaÁ mhÁ apoÁ tautoma tou paragi gnetai toiÄ v anqrwÂpoiv (282c1±3). cf. which can only be provided by wisdom.

O'Brien. TAPhA 40 (1909) 185±201. 132. Paradoxes. again.3 which has been termed `trias paedagogica' (ju siv. apoÁ neoÂthtov deÁ arxame nouv deiÄ manqaÂnein (fr. Kurzdialoge. of course.I N T R O D U C T I O N II. 194 A good synopsis of the history of the problem is given by C. Platon. Gauthier±Jolif ad loc. 144±6. Shorey. (n. 220±48. Gaiser.198 192 Cf.2. 44. h oute askhtoÁ n ou te maqhto n.). e pisthÂmh). p. The Anonymus states expressly that this doctrine of the three conditions applies in every area (eaÂnte sojiÂan ea nte andreiÂan ea nte euglwssi an ea nte arethÁn h thÁ n su mpasan h me rov ti authÄv). 193 It refers. 195 So H.193 Meno confronts Socrates with a problem which had been discussed for at least half a century. 107 . Phaidros (GoÈttingen 1993).). Gunning. MuÈ ller. Prt. This didactic theory was laid down in the Megas Logos (supposedly a protreptikoÂv195): juÂsewv kaiÁ askhÂsewv didaskali a deiÄ tai. an allusion to this programme at Pl. 198 Isoc.196 In a more elaborated form. 163 n. P. E. U (diss.194 Meno's three mutually exclusive possibilities are in fact the three conditions which Protagoras held necessary for anyone who wished to pro®t from sophistic teaching. Sophistik. e pisthÂmh'.. whose fragments. integrated into the theory 1179b20±3. Phdr. Heitsch.. 59 n. EN 1099b9±11.±K. 132±8 are still valuable. 196 Cf. Gunning. Comm. cit. 197 Hp. 269d4±6.. (70a1±4). F. to the qei ai moi rai alternative given at the end of the Meno. Lernen und angrenzende Begri¨e cf. Of the older literature P. on Phdr. mele th.17. cf. MELETH. `FuÂsiv. 1 D. Pl. and indeed it is found in the areas of medicine197 and rhetoric. 175. 192). 323c5±8. W. Lex 2. Hieronymus.338 Littre .192 is found at the beginning of the Meno: a ra didaktoÁ n h a reth . 16 and ad loc. h ou didaktoÁn all' askhto n.4. 3 D. we encounter it in the Anonymus Iamblichi (fr. cf. On melethÂ È bung. mele th. C. allaÁ juÂsei paragi gnetai toiÄ v anqrwÂpoiv h allwi tiniÁ troÂpwi.±K. op. Protreptik. De sophistis Graeciae praeceptoribus (Amsterdam 1915). which announce an original Platonic alternative. Basel 1970). Gomperz. 13. the alternative is added to the list Arist. De Vries. may or may not be part of a protreptic speech. Apart from the last four words. 341.

200 True sophists never held that areth could be acquired by instruction alone (cf. like some modern scholars. La. fr. 108 .I N T R O D U C T I O N II. cf. 202 Loc. by Grg. F. N. Nomos und Physis (Basel 1945). 2. 203 13. areth is acquired by training.21) either misunderstood the issue201 or misrepresented it on purpose. it was possible to hold that you could acquire a reth either by instruction.202 At any rate. ed. cf. 3. its most spectacular aspect was the teachability of the arethÁ politikh . W. 9.199 This opposition called forth a popular but quite un-Protagorean antithesis juÂsei : e x e pimelei av kaiÁ a skh sewv kaiÁ didachÄv. 98±101. Theogn. or by natural ability.3 But of course. 201 He may have been misled.40±2. Democr. Paideia iii 113±14 (Engl. or by training. from which I quoted just now.2.) This is where Euthydemus and 199 O. Prt. 242 D. 429±38. this implied a strong opposition to the traditional views on this point.11 e sti de ti kaiÁ juÂsiv). as Theognis had done before him. also Dissoi Logoi 6. who goes out of his way to deny the existence of a didactic method htiv toiÄ v kakwÄ v pejuko sin proÁ v a rethÁ n swjrosuÂnhn an kaiÁ dikaiosuÂnhn e mpoihÂseien (13. 460b1±7. an argument from vulgaris opinio.203 Such was the status quaestionis when the Meno was written.). 92±3.1. it is clear that by the time Isocrates wrote KataÁ twÄn sojistwÄn. section ii. (Sometimes.21. (n. not by instruction. Heinimann. Paradoxes.100±4. 200 Pl. 323c8±324a1. Pl. cit.86±8. the training is diaÁ thÁn twÄn loÂgwn twÄn politikwÄn e pime leian.2. which had been formulated at their sharpest by Pindar. O'Brien. 179d6±7. Isocrates' own position is intermediate: given natural ability.±K. Now it is self-evident that anyone writing a protreptikoÂv will have to prove or take for granted that virtue (or any other goal of protreptic) comes by teaching or training.. the three necessary conditions of Protagoras had become mutually exclusive in the ®eld of areth . the proof in itself may constitute a protreptikoÂv. Jaeger. and Isocrates. cf. not ju sei. 199).

c4±5. but they must cover up the fact that the existence of instructors was rather questionable in the ®rst place. e xaskh sousin kaiÁ ekmeleth sousin). 6. This willingness is of course indispensable for protreptic. .3.3 Clitophon part company. Dialex. When. Prt. He does so by assuming as self-evident the existence of instructors and then dealing at some length with the alternatives of teaching and training (note the repeated ampli®cation melethtoÂn te kaiÁ askhtoÂn . and Socrates' exaggerated relief (c5±8) shows that Plato was well aware of the fact. the alternatives are as immaterial here as they are in the Euthydemus. as Socrates intends Euthydemus and his brother to take over where he has stopped (e1). 6. cit. d1).7±8. we compare the use either dialogue makes of the `trias paedagogica'. On the other hand.2. at any rate. Theognis. . 282c2) is excluded by a literary trick: Clinias declares that he is willing to accept the teachability of wisdom. having analysed these parallel passages from Euthydemus and Clitophon. In fact. 109 . the author of this passage knew quite as well as Plato did that the Pindaric point of view has to be eliminated in one way or another if you want to write a protreptic speech at all. the author of the Clitophon had to resort to other means in order to dispose of the ju sei alternative. the intrinsical similarities 204 Cf. 319c7±320b3. the possibility that wisdom comes to men only apoÁ tau tomaÂtou (ˆ juÂsei. but the emphasis has to lie on teaching. but after all this is a parody. In the Euthydemus.204 and that with the alternatives maqhtoÂn ± melethtoÂn the possibilities had not been exhausted. 89e4±7 and the discussion which follows till 96b6±9.I N T R O D U C T I O N II. The third possibility (namely that wisdom can be acquired by training) is passed over for artistic reasons: in itself it could have been mentioned here without impeaching the ®nal conclusion (anagkaiÄ on einai jilosojeiÄ n. Not quite an elegant way of getting round an obvious ¯aw in the argument. loc. Men.

In this case Euthydemus 282c1±3 is certainly not the only source for 407b5±7. No argument whatever can be based upon the fact that. It is at any rate easier to suppose that Plato and our author ± if not identical ± had the same way of dealing with the `trias paedagogica' when writing a protreptikoÂv. see also Comm. it is not ruled out by argument and never mentioned again. melethtoÂn. as was argued above. The author of the Clitophon has to conceal the juÂsei alternative. The only possible argument is from similarity of expression. again.3 appear to surpass the di¨erences. Of course. the Euthydemus may have been one of the sources here (as it probably was. they had been at least from 390 onwards. maqhtoÂn.I N T R O D U C T I O N II. he has to take great pains over the meleth alternative. as the Clitophon mentions melethÂ. because he is writing a dialogue. at both places. though. but in that case our author must have had a good insight into the underlying motives which had prompted Plato to write down and leave out what he did. a skhtoÂn Clitophon 407b5±7 ± didaktoÂn 408b7). considering the three shared motifs in this sentence and the third part of the speech. 410d4). Next. didaktoÂn and a skhtoÂn are o¨ered as alternatives. In the Euthydemus. And even here no case can be made from the parallel order didaktoÂn ± ou didaktoÁn all' a skhtoÂn (Meno). askhto n Meno. which is immaterial. The only similarity left is the use of the neuter verbal adjectives (didaktoÂn. maqhtoÂn. in order to do so.2. since naturally the most startling alternative is put ®rst in either case. There remains the question of sources. it is immaterial to the argument and does not come up again. on 408c4±7). Plato can a¨ord to mention the juÂsei alternative once. The choice of these adjectives 110 . The meleth alternative. is left out. the relation of our passage to the opening of the Meno. maqhtoÂn ± melethtoÂn te kaiÁ askhtoÂn (Clitophon). as he is not writing a dialogue but reporting a speech (it is mentioned in the analogy with the body.

1. Chr.2.17±19. 207 didakto n Dialex. on the other hand. ± maÁ Di ' ou k e gwge. 6. both form and content are already formulaic in the Meno as well as in the Clitophon. e jikto n Democr. askhtoÂn. 13.I N T R O D U C T I O N II.2. omwv deÁ oti an apoÁ tau tomaÂtou e pi hi moi sumbouleuÂsw umiÄ n (a reference to kiqari zein h auleiÄ n h i ppeuÂein in 4. ± oudamwÄv.2.±K.. e nqetoÂn. 48 n. paradoth D. maqhto v Virt.4 Par' ou qhnaiÄ oi oudeÁ n e maqon. Under these circumstances.2.g. fr.207 so in fact. ± uses maqhto n. Thus the 205 Arist.. it goes so far. ± ou deÁ mhÁn oudeÁ periÁ palaisma twn ge eiw qasi bouleuÂesqai e n thÄi e kklhsi ai. 111 . 435. dependence of the Clitophon on the Meno (so e.17 etc. cf.208 Xenophon. Prt. maqhto n Dialex. ± all' otan periÁ krouma twn e n luÂrai. 13. paralhptoÂn Men. 93b4±5.3 might seem to indicate dependence. paraskeuasto n Pl. EN 1099b9±10 ± who is certainly dependent on Men. This motif belongs mainly to the protreptic dialogue: Alcibiades 1  qhnaiÄ oi 107a1±9 PoÂteron ou n otan periÁ gramma twn A bouleuÂwntai.1. These two passages probably derive from Aeschines' Alcibiades.205 but does it prove it? The choice of the neuter is just an indication of good fourth-century Greek. 208 This passage was probably used by D. Isoc. 7) is impossible to prove. 6. Memorabilia 4.12 (Xenocrates). In the second sentence (407b8±c6). Pseudo-Platonica. as to ascribe discord and war to insu½cient education. the author uses the motif of inadequacy of present-day education. toÂte anasth shi autoiÄ v sumbouleuÂswn. Heidel. L. 398d7. 206 didakto v only Erx. paradoto n. 59 D. Theogn. .6). . section ii. 376c4. The Clitophon di¨ers from these parallels in that the parody does not especially envisage making one's appearance in the assembly. 4. oud' denoÁ v meÁ n pwÂpote w andrev A akouÂwn tinaÁv einai le gein te kaiÁ praÂttein ikanouÁv ezhÂthsa touÂtoiv e ntuceiÄ n .206 Besides. it had been customary to use verbals in -to n when discussing the issue from the very start. poihto n. 319b2±3. ± ou me ntoi. e qisto n. pwÄv a n orqwÄ v graÂjoien.

407c2±3). 24d9±25a11. Protreptik. 179a4±7. (SSR vi a 77) and my `Aeschines' Miltiades'. e pime leia thÄv yuchÄv ± 209 Other parallels: Pl. Socrates exhorts others to take care mh te twÄ n thÄv poÂlewv priÁ n authÄ v thÄv poÂlewv (36c8 and de Strycker±Slings ad loc. ± The study of music was mentioned (in an unknown connection) in a Protrepticus by Chamaeleon (Ath.I N T R O D U C T I O N II. we encounter an idea occurring elsewhere in protreptic passages. Socr. 184d ˆ 1. cf. Apolog y. a dhÁ paidei an a rethÄ v ei nai tele an hghsqe. cf. ii. This is pure propaganda: no Greek would have admitted that a child's education takes place only. 858d6±9. Similarly. Lg.210 As this theme is being developed in the Clitophon. and is ®nished. within the walls of a class-room. strength. Wol¨. beauty. Ap. 30b4.402 Kaibel). cf. in the individual. riches do not contribute to happiness). Aesch. swÄma ± arcoÂmenon: 129e3±130b4. In the Apolog y. 210 Cf. but only in those from Plato or written under his in¯uence. 28). 326c6±d8. Gaiser. the theme is found in the Alcibiades 1: Ouk ara teicwÄ n ou deÁ trihÂrwn oudeÁ newri wn de ontai ai  lkibiaÂdh.2 Second part The minor importance of the sw matov e pime leia vis aÁ vis the care of the soul is a well-known protreptic theme. oudeÁ poÂleiv w A plh qouv oudeÁ mege qouv aneu arethÄ v (134b7±9. 92e3±6.209 When protreptic literature rejects traditional education (or at least claims its insu½ciency). The Alcibiades 1 provides the nearest parallel (yuch ± arcousa. e1±2). it reduces the notion of education to school-education (very outspokenly here. fr..2. 37 D. Men. Apologie. La. ˆ 11 Kr.3 irony of Xenophon and Alcibiades 1 becomes dark cynicism here. 306. Prt.3. 108±12. EN 1180a1±12. in a protreptic situation (section ii. Arist. 112 .3.1. 325c6±d7.2. 111). just as. ei me llousin eu daimonh sein. Burnet on 24d11 and de Strycker±Slings. The improvement at which protreptic aims concerns not only individuals but also states (407c8 kaiÁ po leiv poÂlesin.

e4.. Aristotle's Protrepticus. E. see section ii. Mem. de Strycker. 132c4±6. 1 117e4±5 anama rthtoi zwÄ si ). 212 `Aeschines' Miltiades'. `Charmides'. 113 . and Alc. 76±104. is signi®cantly frequent in protreptic literature: it recurs in the Euthydemus.401. Chrm. 171d6±7 a namaÂrthtoi . Alc. 213 E. . but I doubt whether this fragment is Aristotelian. 23. Epin. b 108±10. 306. 8.1. . Iambl. 171d8±e5. 128). which are related but not identical: (a) what one cannot handle should be left to others. Memorabilia 4. 204±7. while there is a reference to abstention (a peco menoi. 280e3±281e2. 62±6. 172d7±10 in what is probably a refutation of Aesch. a n diexwÄmen (207 n.3 Third part The general principle of this part of Socrates' exhortation.2.3 touÄ sw matov: 132c1±6). Euthd. 26). `Charmides'. `On Fragment 5a of the Protrepticus' in DuÈ ring±Owen. kreiÄ tton e aÄn thÁn touÂtou crhÄ sin. 34c5. esp.14.16±19 D.1 ˆ 2. Phd. 3. Protr. and it is already found in the Anonymus Iamblichi. 215 Anon. also b 8. 4. Alc. Rabinowitz. 1 117c6±e5. o twi tiv mhÁ e pi statai crhÄsqai. the words anamaÂrthtoi gi gnontai (26) which E¨e compares with Chrm. Ti. X. In these texts. Aesch. X. have another parallel in Alc. R. 88±9.2. ii. Mem. Aeschines' Alcibiades (probably). yet no mention is made of leaving things to others.25±9 belongs closely with Chrm. 79e9±80a5.3.213 See further section ii.2. 353d3±7. and Aristotle's Protrepticus. Xenophon. Pl.212 There is. cf. 896c1±3.3 n.15±6 D.g. b 4 DuÈ ring ± cf. a host of parallels outside strictly protreptic texts.±K. Lg. 34). Arist. E¨e.2. ˆ 1 Kr. Pl. the principle is found stated in two variant forms. Alc.I N T R O D U C T I O N II. 1).211 I may refer also to my reconstruction of Aeschines' Miltiades.2. fr. besides. b 17. others are found in the Apolog y (30a8±b2) and Aristotle's Protrepticus (toÁ meÁ n a rcei toÁ d' a rcetai b 59 DuÈring. 980d6±e2.2.3. (E¨e. cf. (SSR vi a 50) may be an illustration or reminiscence of a general statement of the principle occurring in the lost part of the text ± E¨e should have mentioned this parallel. 204±7. Aristotle and Plato. 1 (cf. 4. 214 Pl.215 Our author's procedure is interesting in that he applies both versions (in the order (b)±(a)) to the human 211 Cf.214 (b) what one cannot handle in the right way should be left alone or it will prove evil.

w ta (c9±11). The relation of e rgon and kaki a: o mmata (b14±c8). . thereby making Socrates exhort his listeners to death and slavery. the same order occurs in the discussion of e rgon and a reth at the end of the ®rst book of the Republic (352d8± 353e11). paÂntwn pe ri twÄ n allwn (b12±3). 78 n. su mpanti twÄi sw mati ) and the sphere of tools and possessions (lu rai crhÄ sqai . And after e rgon and areth have been connected (353a9± b3): o jqalmwÄ n e rgon/a reth (b4±7). . From 353d3 onwards the results are applied to the human soul. allwi twÄn o rga nwn oudeÁ kthmaÂtwn ou deni ). 143 n. wsi n . talla paÂnta (d1±2). . 156). a pote mnein ± drepaÂnwi (353a4). Laws 732a6±b2. As was seen by Kesters (AntistheÁne. w twn e rgon/areth (b8±11). As the conclusion of the whole passage ( justice is the arethÁ yuchÄv 353e7) recurs in the Clitophon 114 . the principle (in form (b)). is applied to the human body (ojqalmoiÄ v crhÄ sqai . 1) and Gaiser (Protreptik. a kouÂein ± wsi n (e7). .I N T R O D U C T I O N II.) Before the interest focuses on the soul. . In the discussion of e rgon: i deiÄ n ± ojqalmoiÄ v (352e5). we ®nd dre panon associated with aspi v and luÂra as examples of tools in an earlier attempt to de®ne justice (333d3±6). (A similar combination is found in the preamble to the laws.3 soul. . Moreover.2.

143 n. apart from that. kreiÄ tton e aÄn thÁn touÂtou crhÄ sin is applied to the ®eld of te cnh (408a1±4) is peculiarly incongruous. 11) compares Alcibiades 1 133e4±5 ostiv deÁ taÁ autouÄ agnoeiÄ . kaiÁ taÁ twÄ n allwn pou an a gnooiÄ kataÁ tau ta (which 216 The similarity in order in Alc. . 209a4±210a8 (where Socrates explains to Lysis that others will leave certain things to him when they will think him better quali®ed than they are themselves) says little or nothing. the inference is inevitable that the plan of this part of Socrates' speech is borrowed ± along with the ®gure of Clitophon and many important features of the second part of the Clitophon ± from the ®rst book of the Republic.2. zu verstehen' (Protreptik. if only because ± as we shall see presently ± the lyre is clearly a corpus alienum in the Clitophon.I N T R O D U C T I O N II. Kleitophon'. 353c) supposes. or alternatively. 350c4±d5 in Republic 1. A common source is not altogether impossible. 217 Gaiser's reference (ibid. . The way in which the general principle otwi tiv mhÁ e pi statai crhÄsqai. as Adam (ad R. 115 . the description of justice as arethÁ yuchÄ v which is common to both Republic 1 and Clitophon would then belong to this source as well.3 (409a2±6). When instead of this Socrates is reported to say that a man who cannot handle his own lyre will clearly be unable to handle his neighbour's. his words are not to the point. should leave it to others.217 Pavlu (`Pseudopl. Gaiser's explanation `von der politischen Abzweckung des Logos her . 156) is more a palliative than a cure. even if the lyre is mentioned at 209b5 and the neighbour (in another context) at c7 ± there is no question of the neighbour's lyre. but this description is foreshadowed by 348c2±6. but I fail to see to what purpose Plato should have taken over the order eyes ± ears ± tools ± soul in Republic 1 from such a source. and vice versa.) to Ly. One expects a statement to the e¨ect that a man who cannot handle his lyre should not handle it. 1 126a5±b10 swÄma ± o mmata ± wta ± po liv is much less striking and probably a coincidence ± if the order was not itself copied from Republic 1.216 The reverse process (borrowing by Plato from the Clitophon) can be excluded.

the next step may have been something like `a man who cannot manage his own household. cf.19±22 adduced by Pavlu.I N T R O D U C T I O N II. At the same time. this parallel may indeed shed some light on our sentence. will be unable to manage the a¨airs of the city'. was. `a man who cannot master his own soul (cf. underestimate the extent to which Plato incorporates material from other Socratics in his dialogues. Chr. (On the other hand. as I think it is. a step towards a de®nition of the requirements of a politician. the second and third parts of Socrates' protreptikoÂv) is a condensation of the end of Alcibiades 1 (from 130a onwards.e.218 218 I do not see the relevance of D. 11 n. Kleitophon'.2. While one should not. Alcibiades 1 133e4±5 quoted above. our conclusion regarding the author's method (cf. It looks probable that the example of the neighbour's lyre was borrowed from some protreptic source and squeezed into this context. this argument loses much of its force: Plato may not have been as careful here as he normally was. This tallies with. in the context from which it was borrowed. the very fact that we can hardly ever locate such foreign material proves that he was very careful in adopting it.2). `Pseudopl. the pattern bodily parts ± tools ± soul (taken over from Republic 1) o¨ered an occasion to drag it in but apparently the author did not realise (or care) that it was out of place. e9).) It is not too bold to suppose that the ®rst part of the argument: ostiv gaÁ r dhÁ mhÁ e pi statai thÄi e autouÄ lu rai crhÄ sqai. 116 .2. if he did so at all. if the protreptic speech in Clitophon is a parody. we may have here a valid argument against Plato's authorship. and considerably reinforces. 10.3 leads up to the statement that such a man can never be a good politician. 3. or possibly. I think. section ii. 408a5). Though I cannot accept his general hypothesis that the whole of Clitophon 407e5±408b5 (i. will be unable to govern those of his fellow-citizens'. see below). dhÄ lon wv oudeÁ thÄi touÄ gei tonov.

the general principle had been so formulated as to lead up naturally to the preferability of death. I agree with Pavlu (`Pseudopl.2. On Pavlu's assumption that this part of the speech was dependent on the end of Alcibiades 1. oudeÁ thÄi e autouÄ) does not make sense in this interpretation.I N T R O D U C T I O N II. the second is added as an (apparent) afterthought. the weight is on the ®rst part. Kleitophon'. Concerning the words kaiÁ teleutaÄ i dhÁ kalwÄ v o lo gov outo v soi (408a4±5). but not. besides. Alcibiades 1. this `beautiful ending' could refer only to the second part of the sentence (from a7 ei de tiv onwards). 12) that they should be taken quite literally: they refer to the ending not just of Socrates' speech as reported by Clitophon. Yet within this sentence. we cannot go beyond that. perhaps it is a (rather feeble) attempt to adjust the argument to this context (it is necessary that the things which people cannot handle belong to themselves). while it would have been equally possible to mention the idea of leaving things to others (version (a)) in the general statement ± this would have laid the emphasis on slavery. The reason why I am con®dent that these words mean more than just the end of the reported speech is the intriguing circumstance that of the only two examples of the fourth-century protreptic literature which are more than an extract (such as Xenophon. The words kaiÁ teleutaÄ i dhÁ kalwÄv o lo gov outo v soi are indeed meant to ridicule the ®nal passage of some protreptic work. or at any rate not in the ®rst place. but of one out of several works which belong to the protreptic corpus and which may have been used throughout the parody. The reversed statement (oudeÁ o stiv mhÁ thÄi twÄn allwn. the ®nal motif of Alcibiades 1.3 On this supposition. Memorabilia 4. the source of the words concerning the neighbour's lyre may have to be looked for somewhere in the sphere of the Alcibiades dialogues. as there is no mention of death being preferable in that dialogue.2) or a hand117 .

219 An allusion to this motif is also present in the Euthydemus: oudeÁ n aiscroÁ n . . Aug. eum nihil hominis esse nec quidquam inter Alcibiadem summo loco natum et quemvis baiulum interesse. SSR vi a 51) klaÂein [sc. having proved that Euthydemus has no knowledge of just and unjust. not in Krauss.2. Mem. 99±103.22): oisqa de tinav andrapodwÂdeiv kaloume nouv.2. In Xenophon. 11c D. 1 and X.77: Quid enim dicimus. It can also be shown that the motif played an important part towards the end in Aeschines' Alcibiades. D. ii Kr. b2) which makes one suppose that Plato had Aeschines' Alcibiades in mind when he wrote these words (cf. 4.2 which was quoted above. c2 pre pei a ra twÄi kakwÄi douleuÂein and 10 ai sqa nhi deÁ nuÄn pwÄ v e ceiv. 1 131a9±b9) makes it probable that Aeschines made Alcibiades concede that his condition was not very di¨erent from that of a slave.2 re¯ect (partly) Aeschines' Alcibiades. Socrates continues (4. the dhmiourgoi are said to be andrapodwÂdeiv (22. 9 D. fr. the other (Aristotle's Protrepticus) with the alternatives `either philosophy or death' (cf. gi gnesqai sojoÂn] uphreteiÄ n kaiÁ douleuÂein kaiÁ e rasthÄ i kaiÁ pantiÁ a nqrwÂpwi (282b3±5).8 ˆ 425. 1. 4. cf.. ˆ 4 Kr. e leuqeroprepwÄ v h ou. 4.39 paÂnu aqu mwv e cwn. 14. In fact.3 ful of tiny fragments. (SSR vi a 53) diaÁ toÁn e rwta on e tuÂgcanon 219 See Cic. this similarity is one of the major arguments for the hypothesis that both Alc. On account of a fragment from Aeschines' Alcibiades (fr. It is especially the reference to lovers (cf. 118 . Civ. Euthydemus takes leave nomi sav twÄ i o nti andraÂpodon einai (39).. 6 D.I N T R O D U C T I O N II. cum Socrates Alcibiadi persuasisset. Alc. 3. cf. ut accepimus. Even though a baiulus is not necessarily a slave. Mem. The slavery motif is found at the end of Alcibiades 1 (135b7±c11.102±5 Dombart±Kalb) this story was regarded by Dittmar as a synopsis of this dialogue (Aischines. cf.. Appendix i ).2. the fact that in the same passage from X.). ut sibi virtutem traderet turpitudinemque depelleret. Tusc. as Euthydemus does in Xenophon and Alcibiades in Alc. e neka tou tou [sc. X. esp. At the end of the conversation. Alcibiades] qe nta thÁn kejalhÁ n e piÁ taÁ go nata aqumhÂsanta (cf. cum se Alcibiades ad¯ictaret lacrimansque Socrati supplex esset.32. Mem. one (Alcibiades 1) ends with an exhortation to slavery. SSR vi a 47). fr. . and speci®es: a r' oun twÄ n taÁ kalaÁ kaiÁ agaqaÁ kaiÁ di kaia mhÁ eidoÂtwn toÁ o noma touÄt ' e sti n.

. touÄ to . Bury has shown221 that all speeches in the Symposium foreshadow Alcibiades' speech ± this particular passage points forward to 218c7±d5. 184a7±8. Alc. p. Alcibiades 1. contrarily to what Bury suggests. The examination of parallels for the slavery motif in the Clitophon shows that this motif has. Alcibiades 1. 219e3.220 Commentators on the Euthydemus usually refer to Symposium 184c4±7 nenoÂmistai gaÁ r dhÁ h miÄ n. . two di¨erent aspects. also contains a well-known protreptic motif: 282a7 kaiÁ paraÁ patro v . Euthydemus. Xenophon. SwkraÂthv o Swjroni skou kaiÁ Fainare thv ktl. section ii. so to speak. where Alcibiades tells Socrates that he is willing to grant him his favours in order that he (Alcibiades) will attain the best possible state of mind. auth au h e qelodoulei a ouk ai scraÁ einai oudeÁ kolakei a (cf.3.222 This is clearly putting into practice the theory about e qelodoulei a from Pausanias' speech. paralamba nein poluÁ maÄllon h crhÂmata. lx±lxiv. lxiii. . . 185b1± 5). 119 . on Smp.). Symposium) ± the presence of 220 The ®rst part of the sentence in Euthd. the other that such a person should have himself led by others (Clitophon. ˆ 4 Kr. but even there Alcibiades is present: R.2. kaiÁ outov agaphto v. (SSR vi a 53) diaÁ toÁ e raÄn belti w poihÄ sai.3  lkibiaÂdou. 221 Comm. does not belong here.I N T R O D U C T I O N II. One is the conclusion that owing to lack of (moral) knowledge a person is no better than a slave (Aeschines' Alcibiades.1. Aeschines. cf. 11c D. 222 Cf. fr. e aÂn tiv e qe lhi tinaÁ qerapeu ein hgouÂmenov di ' e keiÄ non amei nwn e sesqai h kataÁ soji an tinaÁ h kataÁ a llo otiouÄn me rov a rethÄv.223 immediately before the speech about e qelodoulei a there is probably a reference to Alcibiades in the words toÁ upoÁ crhma twn kaiÁ upoÁ politikwÄ n dunaÂmewn alwÄnai ai scro n.2. a possible reminiscence in Aristotle's Protrepticus b 53 DuÈring). Alcibiades 1 131e1±132a2 out ' e ge nesq' e rwÄ n A  lkibiaÂdhi twÄ i Kleini ou e rasthÁv out ' e stin w v e oiken A a ll' h eiv moÂnov. G. 223 katadedoulwme nov.

1 125d10±e6 is a mere coincidence: the art of the captain is used there as an analogical example (along with corodidaskali a. so the Euthydemus in vain tries to make out what e¨ect politikh has on the citizens. there could be little doubt that the `beautiful ending' of Clit. Besides.2. h politikh ] e n thÄi pruÂmnhi kaqhÄsqai thÄv poÂlewv. Comm. e doÂkei ]. on 408b1±2). re¯ects the last set of arguments in Socrates' exhorting of Alcibiades in Aeschines.I N T R O D U C T I O N II. the motif `death is preferable' is combined with the slavery motif: Grg. 120 . if this were true. Comm. (cf.3 an e rasthÂv in Euthydemus and Symposium made me suppose that this second `aspect' also occurred in Aeschines' Alcibiades. Alc. 215e4±216a2. it must be more than a coincidence that both these thoughts are found in close connection also in the Clitophon. 483b2. there is one interesting parallel from the protreptic corpus:225 Euthydemus 291d1±3 atecnwÄ v kataÁ toÁ Ai scu lou i ambeiÄ on moÂnh [sc. Smp. This search ends up in aporia: the `kingly art' makes people sojoi and agaqoi (292c4±5). just as it is not immediately clear in our dialogue to what purpose one should have oneself led by someone who knows the art of governing human beings (cf. 225 Alc. A case might be made for the dependence of the latter passage on Aeschines. But such a case could never be conclusive. but this statement breaks down on the question in what respect they will be all that. If this is true. paÂnta kubernwÄsa kaiÁ pa ntwn a rcousa pa nta crh sima poieiÄ n [sc. This passage provides a counterpart to the identi®cation of kubernhtikh and politikh in the Clitophon. e3). it would seem plausible that Aeschines connected it with the `second aspect' of the slavery motif: `the amaqeiÄ v should have themselves led by sojoi '. on 408b2±5). when Socrates and Alcibiades try to establish the qualities required of a good governor.224 Whereas the comparison of politicians to captains is in itself a commonplace (cf. This supposition strengthens my hypothesis that the general protreptic motif `what one cannot use should be left to others' was stated in that dialogue. 224 In two places in Plato. the parallel discussed above).

cf. but even this deviation from Socratic ignorance as expressed by Aeschines is a long way from h twÄn anqrw pwn kubernhtikhÂ. 99 n. Socrati supplex esset ut sibi virtutem traderet ) but not because Socrates is an expert at h twÄn anqrwÂpwn kubernhtikh . whereas up to 408b2 the Clitophon had been following the Alcibiades-dialogues. cf. explained by fr. and with the more substantial treatment in the Politicus. o mwv wimhn . kaiÁ dhÁ kaiÁ e gwÁ oudeÁ n maÂqhma e pistaÂmenov .2. the subsequent identi®cation of politikh with dikaiosuÂnh and dikastikh in the Clitophon is therefore to be connected with the (rather elusive) discussion of the basilikhÁ te cnh in the Euthydemus.227 Similarly.3 Probably.40). (SSR vi a 53) diaÁ toÁn e rwta on e tuÂgcanon  lkibia dou ou deÁ n di ajoron twÄ n BakcwÄ n e pee rwÄ n A poÂnqein .2. This means that. Socrates does implant some kind of knowledge: e xhgeiÄ to a te e no mizen ei de nai deiÄ n kaiÁ e pithdeuÂein kraÂtista ei nai. . .2. 11c D. 124b10±d5. Gaiser. let alone politikh etc. Memorabilien ii. . paÂnu an pollhÁn e mautouÄ mwri an kategi gnwskon. Socrates expressly says so fr.7. .I N T R O D U C T I O N II.1. Mem. 227 For the in¯uence on the pseudo-Platonic Theages.11 (and 2. ˆ 4 Kr. 135d3±7). . In Aeschines. 4. 49±50). nuÄn deÁ qei ai moi rai wimhn moi touÄ to dedoÂsqai. . 4. e nia deÁ kaiÁ e mimeiÄ to wn e keiÄ nov e pethÂdeuen (4.1). This is typically Xenophontean (cf.226 Let us note at the outset that at this point the Clitophon is fundamentally di¨erent from the Alcibiades dialogues. Gigon. Alcibiades accepts Socrates' guidance (fr. in Alcibiades 1 Socrates' own ignorance is repeatedly stressed (112e1±113c4. In Xenophon Euthydemus ends up becoming a companion of Socrates'. (SSR vi a 52) cum se Alcibiades . (SSR vi a 53): e gwÁ d' ei me n tini te cnhi w imhn du nasqai wjelhÄsai. . 121 . 10 D. 11a D.17. the author now 226 It may be important that the identi®cation of politikh and basilikh recurs in X. ˆ 3 Kr. diaÁ toÁ e raÄn belti w poihÄsai. Protreptik. .

I think.3 suddenly leaves this area. 11c D. rather cleverly managed by dragging in Plato's ideal politician. if the author had made Socrates advertise himself as possessor of such knowledge and teacher of it. Aeschines.2. who is able to make his fellow-citizens better men. This would have to be someone who did know how to `use his soul' and who was able to make others know this as well. In this way one essential aspect of the Alcibiades dialogues (belti w poihÄ sai. who is able to impart the right e pime leia e autouÄ to him because of his love. Therefore the person to whom Socrates' audience is exhorted to hand over the `rudder of their souls' should. this would have crippled the rest of the dialogue. for artistic reasons. On the other hand. there could not have been a direct reference to other teachers of the knowledge how to use one's soul (such as the teachers of justice in 407b5) because then there would have been no point at all in Clitophon's quest among Socrates' disciples. not too clearly identi®able with Socrates or any other person or persons. Alcibiades fr. Alcibiades hands over taÁ phdaÂlia thÄv dianoi av to Socrates.. It is not hard to understand why he should have done so. remain a vague ®gure. This is. In Aeschines and in Alcibiades 1. At the same time the references to Socrates' ignorance had to be suppressed in this speech. the author of the Clitophon had to ®nd another person to have the `rudder of the mind' handed over to. because it is the raison d'eÃtre of our dialogue that Clitophon should ®nd out for himself that Socrates is ignorant. Evidently.I N T R O D U C T I O N II. SSR vi a 122 . Obviously this would not do for a parody of a protreptic Socrates addressing a crowd. Now. quite apart from the question whether or nor he rejected in principle (as Plato and Aeschines did) giving the ®gure of Socrates a teaching role: Clitophon could not very well have proceeded to interrogate Socrates' disciples when Socrates would have proclaimed himself as a teacher. ˆ 4 Kr.

Taylor.I N T R O D U C T I O N II. Plato. 504d5±e3. In the Gorgias it is made quite clear that the ideal politician is no one but Socrates himself : 521d6±8 oimai met' o li gwn  qhnai wn. and as Socrates seems to be portrayed as such at least in the Gorgias. another point to consider. It should be borne in mind. which is essentially the same doctrine in a more elaborated form.5. 253±4 nn. Plato's Euthydemus: Analysis of What Is and Is Not Philosophy (Berkeley±Los Angeles 1992). 515a4±e4.2. ina mhÁ ei pw moÂnov. But as the concept of the politician±educator is typically Platonic. Though Plato creates a circular regress (cf. 230 Cf.3) at this stage. 521d6±e1 and Dodds ad loc. agaqa and their opposites (309c5±d4). 99. 123 .230 To what degree is it probable that here too. however. Chance.228 Yet it may be useful to remember that the politician's art in the Euthydemus is said to make people sojoi and a gaqoi (292c4±5). T. at the bottom of the condemnation of the `Four Men' in the Gorgias. 42±5. 319a3±5 dokeiÄ v gaÂr moi le gein thÁn politikhÁn te cnhn kaiÁ u piscneiÄ sqai poieiÄ n a ndrav a gaqouÁv poli tav.. that the aim of the basilikhÁ sumplokh in the Politicus is to establish in the citizens' souls a lhqhÁv doÂxa metaÁ bebaiw sewv concerning taÁ kalaÁ.. Men. 100a1±2. this is clearly a positive result. Socrates himself is meant by twÄ i maqoÂnti thÁ n twÄ n a nqrwÂpwn kubernhtikh n? For reasons I have expounded above the author is purposely no more explicit. H. inter alia. esp. section i. 503b6±c3. the case for dependence of Clitophon on Euthydemus is strengthened. 228 It lies. Dodds ad loc. There is. a similar claim is made at Plt. cf.229 and the statement as such ®ts in with the function of politikh in the Clitophon. in this connection. cf. di kaia. 229 Cf. 293a2±4. also Prt. This Platonic concept of the politician±educator is of course well-known.3 53) is carried on in a di¨erent form: we are exhorted to turn to the politician for the e pime leia thÄ v yuchÄ v. If we add to this the parallel use of the captain metaphor. epiceireiÄ n thÄi wv a lhqwÄv A politikhÄ i te cnhi kaiÁ praÂttein taÁ politikaÁ moÂnov twÄn nuÄn.

5). 119).I N T R O D U C T I O N II. For this part of our dialogue. to be found in the Gorgias only (see section ii.2. furnished with more or less unrelated thoughts. among the authentic works. As we shall see (section ii. the thoughts are nearly always found in extant protreptic literature ± usually with such a frequency that it is justi®ed to speak of protreptic motifs (only the proof of the involuntariness of wrongdoing is more or less original. cf. which was borrowed from Republic 1. This may be the more probable in that the identi®cation of politics with dikastikh and dikaiosu nh. on 407d2±e2). there would be no further problem in 231 `The ®rst book of the Republic. 124 .4). Comm. Republic 1 is certainly not an explicitly protreptic text.231 If it were. but the fact that almost every single thought does have one or more parallels within the corpus of explicitly protreptic texts is su½cient proof that a cento of protreptic texts was precisely what the author wished to compile. the patterns are derived (with one exception) from other texts of the protreptic corpus.4 Conclusions The protreptic speech in the Clitophon is constructed with the help of a number of patterns. Memorabilien I. at least after Thrasymachus' equation of justice with the advantage of the strongest is proved incorrect. Of course.4 there might be a hint for the bon entendeur. and more precisely. it is correct to speak of a cento (Gigon. seems. As we saw. the de®nitions of the result of justice come from the same source. The only serious objection to this statement on his intention is the pattern of the third part of the speech. at any rate. many thoughts expressed in Socrates' protreptic speech are found also in texts which are not explicitly protreptic (or even not protreptic in any sense).5. ii.2. of a cento of protreptic texts. which immediately follows in the Clitophon.

the readers are referred to Ap. Indeed. there are good reasons to assume that the author of the Clitophon referred his readers to Republic 1 precisely in order to suggest that the only answer to Clitophon's criticism is implicit protreptic as used in the Platonic dialogue (sections ii. Gaiser (Protreptik. however. Gaiser's failure to distinguish between explicit (Gaiser: `sophistic') and implicit protreptic makes his conclusion too facile and impossible to accept. points back to the three works mentioned? See section ii.4 establishing the intention of the Clitophon as a whole: our dialogue could then without more ado be regarded as a demonstration that explicit protreptic gets us nowhere. In fact.. Obviously.I N T R O D U C T I O N II. I wonder where Grube places the conclusion of Thrasymachus' refutation. 125 . I do not venture to suggest.233 two can be identi®ed positively: the protreptic passages from may well be called a protreptikoÁv lo gov' (Grube. the similarity in pattern would escape any but the most acute readers and even they would not know what to do with it.1) or the distinction between e rgon and di dagma (section ii.3. it is rather super¯uous to think of sources. On the other hand.. 232 The choice of the pattern of R.4. If the same man wrote Ap. that by using a pattern derived from Republic 1 in the third part of Socrates' speech he intended to convey the same message. both of which refer the reader to Republic 1. 352d8±353e11 cannot be interpreted as yielding an underlying positive message as can the description of justice as a te cnh for the soul's areth (section ii. eventually. Of the various protreptic texts used in this cento. and Clit. `The Cleitophon of Plato'. 305).2).5. so that no real harm was done.5).7. a good reader would know that Republic 1 was not his target.4.4. and considers that `der Protreptikos im ``Kleitophon'' im besonderen auf den ``Thrasymachus'' hin geschrieben ist' (143)..232 I rather believe that our author took over the pattern for his own convenience. 297 to section ii. Euthd. 233 I have phrased this paragraph with an eye to the most extreme hypothesis: that Plato did not write Clit. 126±8) analyses the whole of Republic 1 as a protreptic loÂgov. this has some bearing on the problem of authorship: does Plato refer to other works of his in the same way as Clit.5 and ii.3(4) and n. Euthd. however.2. as well as to Republic 1.5.

e1±2.3. and that one of them must depend on the other. It is evident that the trichotomy of Socrates' speech is related to the general preamble of Laws 5. it is clear that one or more of the Alcibiades dialogues have been used. 407b2±4. There is a great deal of ridicule in the speech.2.2. ad loc. if the same hand wrote both dialogues. 11±12) that our author copied the Alcibiades 1. Comm.234 Finally. 729a2±4 ± Clit.I N T R O D U C T I O N II. cf. it would not seem improbable.). 126 . itself contains many common protreptic motifs (though its protreptic is far more original than that of the Clitophon was intended to be). there may be a case. which immediately precedes the words paisiÁ n . 729a4±b1. Kleitophon'. and the words kaiÁ teleutaÄi dhÁ kalwÄv o loÂgov outoÂv soi (408a4±5) do not refer to the end of that dialogue. I see no reason for assuming (as Pavlu does. there is no problem. More particularly.4 Plato's Apolog y and Euthydemus. It is futile to raise the question of sources every time we meet a parallel from either of these works ± the Euthydemus. 407e8±9): when one compares the parallels for this particular version of the motif (section ii. given the fact that our author knew the Euthydemus. Euthydemus: 408c4±7. but there is no good reason why the Clitophon could not have been the source here. `Pseudopl. Other parallels: 729a1±2 ± Clit. katalei pein. there are many motifs which the Clitophon and Aristotle's Protrepticus have in common but there is no evidence for one depending on the other. .3.1.3 n.2. that he got it from there. 407c6±7 than the other way round ± again (see previous note). though. . for the general principle of the third part of Socrates' speech (o twi tiv mhÁ e pi statai crhÄsqai. is probably better explained as a reminiscence of Clit. Next. kreiÄ tton e aÄn thÁn tou tou crhÄ sin. 407c8±d1. see section ii. There are obvious reminiscences of either outside the protreptic speech as well (Apolog y: 408c3±4. but that is as far as one can go. but it is 234 The musical metaphor at Lg. in particular. 215).

Comm.). Socrates' claim that he `does nothing else but persuade' must somehow be true of Socrates' role in Plato's dialogues. of explicit protreptic) ± we observed that the author takes care not to hit at the core of Socratic philosophy (section ii.1 ad ®n.1.2. the conclusions agree with the result of the analysis of the characters in the dialogue (sections i. therefore. Also. or even of the Platonic dialogues as such.. If Plato really thought of Socrates primarily as an exhorter to virtue. in this speech. the point of view is Platonic (sections ii. But there is a good case for interpreting the Apolog y as `Besinnung auf das Sokratische GespraÈch' (Gaiser. ii. With this statement. Plato presents us with a Socrates whose main concern appears to be what we have de®ned as explicit protreptic: oudeÁ n gaÁ r allo praÂttwn e gwÁ perie rcomai h pei qwn u mwÄ n kaiÁ newte rouv kaiÁ presbute rouv mhÂte swmaÂtwn e pimeleiÄ sqai mh te crhma twn proÂteron mhÂte outw sjo dra wv thÄv yuchÄv opwv w v ari sth e stai (30a7±b2).3 ridicule of manner (namely.2.3 Protreptic in Plato In the Apolog y.3 ad ®n. the limits of the present investiga127 . 30).3.3.5. we saw that on the two occasions where politics come into Socrates' speech. let alone anti-Platonic. but also for the interpretation of the Platonic dialogue. It will be clear. little explicit protreptic is to be found elsewhere in Plato. So far.2. on 407d2± e2). that Plato's attitude towards ethical and philosophical exhortation is crucial not only for the (relatively minor) problem of the intention of the Clitophon. he had little desire to depict him explicitly as such in the dialogues.3. On the other hand. not enough to warrant such a strong statement.3. i. ii.2. If so. It might even be maintained that the problem of the Clitophon is essentially that of the dialogue in general.): there is nothing un-Platonic. Protreptik.I N T R O D U C T I O N II.5.

what considerations.3.3. or even satisfactorily. ii.5).3.1 Elenchos and maieutikh In the diaeretical process from which the sophist will eventually emerge as `a puri®er of the soul from conceits 128 . that is to say: avoiding the Scylla of subjectivity means getting into the Charybdis of circularity. total subjectivity. Plato nowhere discusses protrophÁ e piÁ jilosoji an or e p' a rethÂn. nevertheless.1).2) and that they harmonise with Socrates' statement in the Apolog y quoted above (ii. Plato's attitude towards protreptic can be reconstructed in relation to his use of the dialogue (ii. that this section will contribute to understanding Plato as much as to grasping the intention of the Clitophon. I have some hope. are obviously too complicated to be treated comprehensively.3. not a discussion. It can be shown that these discussions concern Socrates' own procedure as depicted in Plato's dialogues (ii. I am further forced to refer as much as possible to passages from those very same dialogues whose meaning as dialogues is the core of the problem. in connection with a minor dialogue of dubious authenticity. I must con®ne myself to a more or less super®cial account of my opinion on these problems. and in order to avoid.3.3) as well as with the demonstration of protroph found in the Euthydemus (ii.3. these phrases occur only in the Euthydemus.3. however. I believe. at least to some degree. On this basis. Apart from the Clitophon. in which dialogue he gives merely a demonstration. that elements of a theory of protreptic may be found in two more or less explicit discussions of elenchos: the sixth de®nition of the Sophist and the midwifery episode in the Theaetetus (ii.4).1 tion are set. The questions. philosophical or otherwise. prompted Plato to adopt the Socratic dialogue for saying what he had to say and how he uses this vehicle of his thoughts.I N T R O D U C T I O N II. of exhortation.

it is old-fashioned and `yields little result for much pains'. owing to his amaqi a. showing them to be contradictory (e pideiknuÂousin au taÁv autaiÄ v . taÁ deÁ malqakwte rwv paramuqouÂmenoi. 230b4±8).236 is a process in which various subsequent stages may be distinguished: insistent questioning (dierwtwÄ sin). gentleness towards others ± in short. Vlastos (`The Socratic elenchus'. to be countered by a special form of instruction.235 summing-up of the sixth de®nition). doxosoji a 231b6) consists of believing that one possesses knowledge without really possessing it (229c1±9). 230b5. d2). OSAPh 1 (1983) 27±58. reprinted in Socratic Studies (Cambridge 1994). at 2 and n. it is clear from this passage that Plato himself was the ®rst to use the word in that sense. 230a1±2).I N T R O D U C T I O N II. ± I disagree with G. by far the best commentary is to be found in Goldschmidt. admonition (nouqethtikh ) and elenchos. as identi®ed and described next. points that way. Admonition may take the form either of angry reproof or of more gentle advice (taÁ meÁ n calepai nontev. Dialogues. 236 On the passage 230b4±d4. 237 This is not expressed. .1 that block the way to understanding' (Sophist 231e5±6. the Eleatic visitor concludes that there is a special form of ignorance. but the addition of ate planwme nwn (ˆ having the conceit of wisdom). does not consider himself to be in need of instruction (229e4±230a10). Elen235 Throughout this section I have gratefully used F. e nanti av. 1±37. There are two methods of paidei a. This special form (called amaqi a. The results of elenchos are anger with oneself (and shame. M. testing opinions (e xetaÂzousi ) with the result that they prove to be the result of amaqiÂa.3. a ®rm conviction that the opinions are wrong (c3 bebaioÂtata). 29±31. Elenchos. Cf. 129 . a frame of mind which is the best and the most sw jrwn (d5). because the person subject to it.237 confronting them with each other (tiqe asi par' allhÂlav). To my mind. Cornford's translations: Plato's Theory of Knowledge (London 1935). . 8) that George Grote was the ®rst to apply the term `elenchos' to Socrates' method of investigation in the earlier dialogues. and it is to be countered by paidei a (c11±d4). DieÁ s (Bude text): `ve ri®ant aise ment la vanite d'opinions aussi errantes'.

of which the passage in Demetr. 180±2. As we have already seen (section ii. that is to say. in Demetrius. v 128±9.I N T R O D U C T I O N II. aporetic interrogations met with great success once they had been invented (298) and `shocked more'. Cf. three types of protreptic are distinguished in this text: reprobatory (calepai nontev). op. Meyer. Apolog y.240 238 This phrase. as always in Plato.4. refers to the highest values concerning man. HGPh. Apologie. cf. On account of the Eleatic visitor's hesitation to connect sophistry with elenchos (230e6±231b8. the most prudent course appears to be to assume that the Sph. passage started o¨ a process of re¯ecting and theorising about the Socratic dialogue and Socratic literature. cit.1. 6 (and Index s. 239 Most scholars agree in taking the sixth de®nition as a constructive piece of Platonic doctrine. 27 n. If this identi®cation of expulsion of amaqi a with protreptic is correct. not just as a more or less playful illustration of the diaeretic method. 22d7. Especially the last two statements make it clear that elenchos is intended to make the subject ready for philosophy. 240 In Plato. 229d2.2).v. 231b8) and of the similarity of elenchos to maieutikh (to be studied below). de Strycker±Slings. ± I cannot discuss here the origin of the trichotomies in Demetrius and Plato. as does the favourable judgement on elenchos as opposed to the two types of explicit protreptic. in other words that it is a form of protreptic. esp. and the exhortation to get oneself educated (paideuÂesqai proetre yato) is what the elenchos passage in Plato is all about (cf. 235). this seems to be the best interpretation. 130 . resulting in the compromise h ge nei gennai a sojistikhÂ. 230e2). It is clearly the aporetic dialogue which Demetrius has in mind: the two elements aporia and lack of knowledge are mentioned explicitly (297).). 89. If Plato239 rejects nouqethtikhÂ.3. (n.1 chos is followed by instruction. which cannot be applied until the subject has been rid of all amaqi a (c4±d4). provides an illustrative and instructive example. It is a puri®cation which concerns taÁ me gista (e1)238 and results in true happiness (e3±4). Goldschmidt. Dialogues. advising (paramuqouÂmenoi ) and elenchos. Cornford. on Ap. like the other de®nitions. this trichotomy recurs also in Demetrius. this means that he prefers implicit to explicit protreptic (elenchos versus nouqethtikh ). accomplished more. elenchos is novel and achieves much more than the oldfashioned way.

More than any other Platonic dialogue. 20±1. does Plato also reject nouqethtikh in written form (represented by Aristippus and Xenophon in Demetrius) in favour of the dialogue.1 Here a methodical objection might be raised. not of views. The claim is valid for the Apolog y only on a super®cial reading (cf. Demetrius describes three types of protreptic literature. if it can be shown that what Plato proposes here as an alternative for oral exhortation is a procedure which he elsewhere uses for Socrates' interrogatory method as depicted in his dialogues. or at any rate the aporetic dialogue? I think the answer must be a½rmative.-Pl. we must infer that what is here said about nouqethtikh also re¯ects Plato's attitude towards accusatory and advising protreptic in written form. section ii. 170±1) claims that elenchos was originally a test of persons. If Plato does not indicate this more explicitly. that may be accounted for su½ciently by the plan of the Sophist: he is not giving a comprehensive treatment of paidei a in oral and written form. the Theaetetus shows us Plato as re¯ecting on the principia of his literary 241 Kahn.5).1) stand apart from Platonic dialogues (cf. and for theorising about that method. Dialogue. The Sophist passage deals with oral exhortation in its three forms. 131 .241 In other words. 111. both in theory and in practice. 97±8 (cf. and that only later on did Plato use elenchos to test views. but he uses paidei a as one more illustration of the method of diaeresis (the description of elenchos is proportionally speaking already very extensive as it is).3. section ii.3.I N T R O D U C T I O N II.3). Are we justi®ed in extending what Plato makes the visitor say about admonition and elenchos as paidei a to written loÂgoi. Alc. But it seems clear to me that what is described and identi®ed as elenchos in the Sophist is a testing of persons and of views at the same time. in other words.3. and I would maintain that for Plato it was never anything else. if elenchos as outlined in the Sophist is identical to Socratic questioning. The distinction may help us to understand what makes the Alcibiades dialogues (Aeschines and ps.

I N T R O D U C T I O N II.3. 243 Probably for the simple reason that it would be out of keeping with the midwife metaphor. Theaetetus will henceforth `have the prudence (swjroÂnwv) not to fancy you know what you do not know' (210c3±4). Even more striking is the similarity of the results of maieutikh and elenchos: Theaetetus `will be more gentle and more agreeable (hmerwÂterov) to his companions' (210c2±3). the word e legcov is not used. cf..243 but there are su½cient reasons for identifying elenchos and maieutikhÂ. whereas the dialogue closes with a reference to this episode. In the two passages referred to. the discussions coming between these two passages are an illustration of Socrates' midwifery put into practice. showing that Socrates had been interrogating Theaetetus along these very lines (210b4 h oun e ti kuouÄ me n ti kaiÁ w di nomen ktl. cf. the subjects of elenchos `become more gentle towards others' (Sph. which is what elenchos is all about. Socrates' ability to `prove by every possible test (basani zein) whether the o¨spring of a young man's thought is a false phantom or instinct with life and truth' (150c1±3) is in essence and function identical to `testing opinions' (e xetaÂzein). On account of this reference we are allowed to combine the ending of the Theaetetus (210b4±d4) and the maieutikh episode proper (150a8±151d3): together they constitute the theoretical framework (Socrates as intellectual midwife). 230b9±c1). Besides. Socrates interrupts the interrogation with a lengthy justi®cation of his manner of interrogation. 242 Apart from the passages to be discussed. 132 . 148e7±8). the second stage of elenchos (cf. 210c2 thÁn nuÄ n e xe tasin). this dialogue is quite clearly presented as itself putting into practice what it contains in the way of theory: soon after the start.1 ways and means.242 Besides. the introductory conversation on direct versus reported dialogue (143b5±c6) and the `Apology of Protagoras' on the correct manner of interrogation (167c7±168b6). the interrogative form is of course a second point of agreement.

3. 261 to section ii. 33. KraÈ mer.246 I believe that the maieutikh epi244 If I have read H. Cf.' (Burnyeat claims that the method of the `middle period' is di¨erent ± his implicit claim that Tht. Dialogue.1. 281±3. clearly echoes Socrates' swjroÂnwv. we are justi®ed in analysing as Plato's written implicit protreptic those dialogues which exhibit the same `elenctic' character as the Theaetetus. N. H. `Socratic midwifery.2. Tigerstedt. Maier. 141 n. also H. because the maieutikh episode is much more extensive than the rather schematic description of elenchos given in the Sophist. cf. belongs to the `last period' is at variance with the results of stylometry. the antithesis of doxosoji a and swjrosu nh is not found outside Sph.I N T R O D U C T I O N II. . M. Der Platonische Dialog (Heidelberg 1968). North. anyway). H. Sokrates. Arete bei Platon und Aristoteles (Heidelberg 1959).4. when he says `ce qu'on oppose aÁ cette meÂthode d'admonestation . parallel. 44±5.244 One cannot but agree with DieÁ s. BICS 24 (1977) 7±16. At the same time. . c'est directement la meÂthode dont le TheÂÂeteÁte a donne et le nom et l'illustration continue . 245 Edition of Sph. Gundert. n. Kuhn. and Kahn. and this notion was the starting-point for Gaiser's interpretation of the Platonic dialogue.1 Theaetetus' comment in the Sophist on having this conviction as a result of elenchos. That a major part of Plato's úuvre is protreptic was suggested already by Demetrius and Antiochus of Ascalon (section ii. J.. that it is swjronesta th twÄn e xewn (230d5). 75.3. If I agree with Gaiser that the interpretation of (a series of ) dialogues 133 . Platonic inspiration'. Burnyeat. which she does not mention. 114. . Notice 272. . E. and therefore of maieutikh and implicit protreptic.2). Sokrates. (not so explicitly.) 246 Cf. 183 correctly. 49±50. but for the Tht. since the Theaetetus itself clearly purports to give a demonstration of elenchos/ maieutikh . avec les re sultats qui sont ici deÂcrits de la meÃme manieÁre qu'aÁ la ®n du TheÂÂeteÁte'.245 The identi®cation of maieutikh and elenchos. at 9: `The midwife ®gure signals a return to the aporeutic [sic] style of those early dialogues and to the Socratic method which is the substance of that style. Sophrosyne. 75±9. enables us to understand better what Plato's views on protreptic are. Interpreting Plato (Stockholm 1977). 463. 470±1. F.

It would be foolish. on 410c6 ou k e qe lein au thÄv e moiÁ koinwneiÄ n. however. See section ii. a caution is necessary. we must try to ®nd a criterion by which to separate protreptic from non-protreptic dialogues (not excluding the possibility that in otherwise non-protreptic works. 247 Gaiser (Protreptik. will not be helped by protreptic (151b1±6). 151b2). (though it does not con¯ict with the statements found in Sph. one needs a certain state of mind to start with.1 sode.248 On the other hand. 200c7±d3 and Comm. passage proves that his antithesis `sachliche Untersuchung' : `protreptische Werbung' is too rigid. let alone the Laws.3. to interpret the Parmenides or the Philebus. This state of mind is referred to (within the symbolism of midwifery) as `being pregnant' (148e7.3. and by the same token maieutikh has a wider application than e legas protreptic is a fruitful approach. 276e6 in a passage which uses the same pregnancy metaphor as Tht. in combination with the elenchos episode (neither of which is used by Gaiser247). but it recurs in Phdr. Evidently. 111) denies that the elenctic method has still a protreptic aim in Tht. To draw this line. esoteric doctrine. `pregnancy' is more than the conceit of knowledge (though the latter is included in it).I N T R O D U C T I O N II. Nearly all Platonic dialogues contain a great deal of discussion. we would be forced to stretch the term `protreptic' so far that it would be pointless to use it at all. If implicit protreptic is to be at all successful. At this point. that does not in any way imply assent to the theory of an unwritten. a closer examination of the maieutikh episode in comparison with the description of elenchos is necessary.5 and n. and thereby exhibit traits of elenchos as described in Theaetetus and Sophist. as implicit protreptic: if we did.). Those who do not meet this condition. 248 We do not meet this condition in Sph. is an aporetic dialogue. protreptic passages may occur). See also La. provides a strong support for this interpretation. But the Sph. and so does the fact (not taken into account) that Tht. 134 . 309.: a yuchÁ prosh kousa is the minimum requirement.

Plato needs elenchos as a means of kaÂqarsiv (230c4. d7± 9. that the maqhÂmata are likewise the result of elenchos). we must stop to consider what we have so far learned about protreptic and elenchos. 210c1±2). That does not necessarily mean that elenchos is restricted to kaÂqarsiv. 274d7±e5 should probably be interpreted in this light: the same method is used for protreptic (in this case. This latter aspect of elenchos is only hinted at in the Sophist (230c9±d1 twÄ n prosjerome nwn maqhma twn.3. e1) and therefore concentrates on doxosoji a and its removal: the sophist is de®ned as a species of the genus kaqarth v (231e6). The double function of elenchos explains why Socrates251 employs it throughout Plato's dialogues: he could use the same method for reaching aporia and for conducting philosophical investigations. in other words. and of elenchos in Sph. section ii. 150d2±6 (below. 251 I do not say `Plato' because I have analysed until now only the upperlevel role of Socrates in Tht. elenchos serves a twofold end: it leads Socrates' partners to aporia by delivering them from false opinion (implicit protreptic) and it helps them in a more positive way. 250 Socrates' question at Euthd. it is not said. 150c3. for the moment. Compare also Clinias' surprising progress (Euthd. explicit protreptic) and for teaching a reth . and therefore this divergence in the description of maieutikh in Tht. but he is also able to call up true knowledge (goÂnimoÂn te kaiÁ a lhqe v. 210b4±10). by elenchos. Both processes are brought to completion by the same method of interrogation and testing. 290e1±291a7 and Me ridier's note) with Tht. and of the Eleatic visitor in Sph.4).3.I N T R O D U C T I O N II. I shall illustrate this twofold use of the same method in the next section.250 Apparently. and am going to apply this analysis presently to the upper levels of other dialogues.5. namely in the acquiring of true knowledge. d3.1 cov has in the Sophist:249 not only can the intellectual midwife liberate from doxosoji a (150c2.3. 135 . A translation into terms of the lower level will be attempted in section ii. does not tell against their identity. d7±8. 249 In Sph. but neither is it denied. The latter can be fruitful only through elenchos: questioning and testing the answers.

In the Theaetetus we see Socrates justify himself by means of an extensive description.253 In this sense. 136 . and subsequently an application. he can only pave the way for it. or as the Sophist has it. of elenchos. in other words. Dialogues. 210c4±6).1. the various conversations reported in Plato's dialogues appear to be implicitly protreptic. In other words. The aporetic discussions and the aporetic parts of `constructive' discussions are suggested to be Socrates' alternative for the explicit protreptic condemned in the Sophist. it makes men ready for the care of the soul. ii.1. 253 In Tht. explique e clairement par le texte du Sophiste' (Goldschmidt. cf.3. and reserve the concept of elenchos for the procedure applied by Socrates in the aporetic dialogues. at any rate those conversations or parts of conversations that lead to the ®nal aporia and in applying elenchos remove in Socrates' partners what the Sophist calls `the opinions that obstruct learning' (230d2±3). This dialogue ends up in aporia and Theaetetus emerges from the process more or less252 puri®ed. for maqhÂmata. This is perhaps only `une ruse de l'ironie socratique. and therefore contains an intuitive moment: even if Socrates possesses knowledge. sections ii.5. but the possibility is left open (210b11±c4).I N T R O D U C T I O N II. But it is also true in the sense that philosophical knowledge is based on contemplation of the Forms. Socrates denies throughout that he possesses knowledge himself (150c8±d2. This assumption lies at the bottom of such important works as 252 It is not said in so many words that the puri®cation is complete. ± For the identity of exhortation to the care of the soul and to philosophy. elenchos has for its (®rst) natural conclusion the ®nal state of aporia: only in that state are men ready for the care of the soul.1 Elenchos is said in the Sophist to be a means of puri®cation concerning taÁ me gista. and it achieves the aim better and faster than admonition. he is unable to impart it completely to others. 82). It will be clear from the preceding analyses that I disagree with those scholars who apply a rigid distinction between elenchos and `later' forms of dialogical discourse in Plato.1.

3.256 but the theoretical framework as set out in the Sophist and the Theaetetus justi®es a unitarian approach. which systematically opposes elenchos and dialectic. only re¯ect Plato's later. C. Laborderie.257 ii. and Tht. 256 On later dialogical techniques.I N T R O D U C T I O N II. views of his earlier work (so. and Vlastos' Socrates. I can hardly select even one dialogue and analyse it in its entirety. Fortunately. Ad hominem refutation. and therefore of elenchos. will be tested in the next section. 318±19 on hypothesis in the Protagoras. and R. Robinson's Plato's Earlier Dialectic (Oxford 19532). 257 The result of the test will have to show whether the passages from Sph. for whom the Socrates of elenchos is an entirely di¨erent philosopher from the Socrates of later works. Form and Argument in Late Plato (Oxford 1996). 255 Cf. Kahn. H. Book 1).2 R. cf.2 Elenchos in practice From the two passages discussed in the previous section. 369) or give a true account of elenchos as practised in all dialogues. 137 . 283±311. J. a theory has been reconstructed which may serve as a starting-point for developing a model for the interpretation of the Platonic dialogue. Gill±M. `Afterword: Dialectic and the dialogue form in late Plato'.254 I think that the di¨erences must not be over-emphasised. in C. Kahn. dialectic. M. McCabe. Our ®rst task is now to test the hypothesis that elenchos as analysed above is indeed the pattern of Socrates' interrogation in the Platonic dialogue. for example. one illustration must serve as proof. `Vlastos' Socrates'. Kahn rightly points out (248±50) that elenchos in the stricter sense is not present in all `early' dialogues and appears also in `later' dialogues (the interrogation of Agathon in Smp. diaeresis. Le dialogue platonicien de la maturite (Paris 1978). there is 254 Cf. C. `Did Plato write Socratic dialogues?'.3. Within the limits of this book. hypothesis255 are all di¨erent forms of testing. I do not wish to deny that there is an evolution in philosophical method. Gill... Phronesis 37 (1992) 233±58. modi®ed. Moreover. How valid the approach is.

is too long).. against many modern interpreters. That this is indeed the case is shown by H. `Meno. that ana mnhsiv is the ultimate explanation for maieutikh (cf. 260 If this is right. 86b6±c2. Socrates' interrogation of the slave is in the last resort259 his reply to Meno's complaint that Socrates has numbed his mind (79e7±80b2).. 2). to wit the mathematical passage of the Meno.2 one passage in Plato's works which suits the purpose in an ideal way..3. Socrates shows that being numbed can be bene®cial as it is a sign of progress (84a3±c9). itself. Meno belongs to the large and di¨use group called the early dialogues. cf. Cra.260 Secondly. Benson. Socratic Conversations. 259 The mathematical passage is usually taken for what it is ostensibly: a proof of the theory of anaÂmnhsiv. he does not. the ®rst part of Socrates' interrogation of the slave must also re¯ect the ®rst part of Socrates' conversation with Meno in the Meno itself. v 73 n. 11±25 uses the passage to analyse the type of questions asked by Socrates and to illustrate the importance of the dialogue form as a factor in the interpretation of Plato. Therefore we can expect a priori that the mathematical passage contains an implicit justi®cation of Socrates' method. This is not to belittle the importance of this doctrine for Socrates' behaviour in the dialogues ± I agree. cf. not for other dialogues (and because Tht.258 In the ®rst place. the slave boy and the elenchos'. But even if Plato believed that the passage proves this theory (I neither think it does nor believe Plato thought so. would prove only that the maieutikh episode is the theoretical model for Tht. just as the maieutikh episode contains its explicit justi®cation: Socrates' questioning is exempli®ed here in nuce. (as would perhaps have seemed logical) because Tht. however.261 whereas Theaetetus 258 I do not analyse Tht.. Smp. 84d1. HGPh. 261 These include Phd. no stylistic features have been found which unite these three 138 . ± Stokes. and because an analysis of Tht. the theory itself is brought in only to defeat Meno's pessimism. Bluck on Men. Phronesis 25 (1990) 128±58. H. Meno and Theaetetus are rather wide apart chronologically.I N T R O D U C T I O N II. In fact. of which it is often falsely claimed that stylometry has proved that they belong to the `mature period'. Bluck's Comm. is an aporetic dialogue and therefore cannot illustrate the second function of elenchos. distinguish between the various stages of the interrogation. 11).

The slave o¨ers an alternative opinion: the sides of the square must be larger than two feet. Apolog y. Again. For a possible stylometric link of Phd. If it can be shown that the Meno passage gives a precise illustration of the theory put forward in Theaetetus and Sophist. de Strycker±Slings. n. Prm. 84c11±d2. been delivered from the conceit of knowledge and he recognises the aporia (84a7±b1).3. 540±1.I N T R O D U C T I O N II. 16 n. but not really relevant to the method of elenchos (though very much so to the theory of ana mnhsiv). R. Tht. iv'. after 369. 29. with the second group. therefore. Meno's slave is asked the length of the sides of a square with an area of eight square feet. 405. (previous section. the distance may be ®fteen to twenty years. Chronolog y. 44±5. my review of Brandwood. 253) it is stressed that Socrates does not teach (82e4±5.262 Next. Besides. 85d3±4). this opinion is confronted with his (correct) opinion that the square of three is nine.2 comes late in the middle period. 425a10. so his second answer is that they are three feet long. the theory will become much more acceptable as an overall pattern for the dialogues. this opinion is tested against the slave's (correct) opinion that the square of four is sixteen and thereby he is proved to be wrong (82e14±83c3). `Notes on Politeia. the statement is questionable. Cf. He has. This is the stage called dierwtaÄn and e xetaÂzein in the Sophist. this corresponds to tiqe nai par' allhÂlav and e pideiknuÂnai au taiÄ v e nanti av. Socrates intersperses his interrogation with comments. The `opinions that obstruct learning' have with the works of the second group. and smaller than four. and Cra.. 139 . The slave still thinks he knows what in fact he does not know (82e5±9): he is guilty of doxosoji a. Dialogue. here too.. It is established that he thinks that the sides must be four feet long (82b9± e3). First.. the slave realises that these opinions are mutually exclusive and now confesses his ignorance (83c4±84a2). Kahn. Phdr. my note on R. 262 Here as in Tht. cf. which (as we shall see presently) enable us to determine the various stages of elenchos as applied here.

2 been removed. until he has reached true knowledge. an dessen Ende eine unumstoÈ ssliche Antwort steht. To the slave's conceit of knowledge. P. 245: `Ein Elenchos. because it is a lengthy. 140 . 85b8±9. Socrates goes out of his way to underline that this method has remained the same: he has con®ned himself to questions (84c11±d2.263 The structure. Die fruÈhen und mittleren Dialoge (Berlin±New York 1992). Teilt man diese Sicht der Dinge.. 84b9±c7). dass Platons Methode im Menon von einer destruktiven (sokratischen) in eine konstruktive (platonische) Vorgehensweise umschlaÈgt. liegt es nahe. and he is now ready for maqh mata (cf. in dem die neue konstruktive Dialektik zum ersten Mal vorgefuÈhrt wird. 259. elenchos is applied repeatedly. one who is `thrown into aporia' will automatically `feel a desire 263 Again. Platons Dialektik. 78]. Stemmer. then. 241: `Die Vorstellung. This part of the conversation corresponds to the maqhÂmata of the Sophist. elenchos is again applied.I N T R O D U C T I O N II. in other words. In the third phase the slave arrives (under Socrates' guidance) at the conclusion that the sides must have the length of the diagonal of a square with sides two feet long (84d3±85b7). ist verbreitet [some references in n. Platon konzipiert hier nicht eine neue konstruktive dialektische Methode. of the whole passage may be summed up as follows. repetitive process. even when the slave is about to ®nd the correct answer.264 This last stage is not illustrated completely in the Meno. im zweiten Teil des SklavengespraÈchs den Text zu ®nden. but that is not the point at issue. To aporia.3. that is. Doch diese Deutung des SklavengespraÈchs und des Menon ist unzutre¨end. but it is predicted that the slave will reach it if the process is applied (85c10±d1).' But I disagree with Stemmer's claim that anamnesis equals elenchos. n. d3±4). 264 Cf.' Ibid. Cf. realisiert den anamnetischen Prozess mit seinen beiden Teilen. until he knows that he does not know. Aporia is a necessary and su½cient condition for `attempting to search or learn' (84c4±5). until he has reached the stage of aporia. it is hard to be convinced by these words of Socrates.

Sokrates. and partly in the same words. 384c2. H. Prt. The stages of initial amaqi a and of its ®nal expulsion are contrasted in the same way. ii. 112±17. 93±4 and n. Apologie. Griechische Denker. as in the Sophist. cf. 253. There is a renewed description of his activity267 (now for the ®rst time called 265 I shall not enter here into another characteristic detail of Socratic elenchos: that it is `searching together' (Men. section ii. 84c11.3. With aporia. 506a4. what follows is not protreptic any longer. 84c5±6). 157d1±2.3 Implicit and explicit protreptic in the Apology It has often been remarked266 that in the Apolog y. cit. Wol¨. after the interrogation of Meletus (24c4±28a2). the traits of Socrates' portrait become di¨erent. op. Maier.3. 90b5.3. Grg. 266). The later passage is partly at a higher level than the earlier one: from 29c5 to 30c1 there is (imaginary) discourse between Socrates and his judges. 76. the aim of implicit protreptic has been reached. 266 I give a number of references taken more or less at random: Th. Hackforth. ii 82±4. (n. Meyer. 267 The points of agreement between the two episodes are listed by Hackforth. 80d4. Phdr. 330b6.I N T R O D U C T I O N II. The Composition of Plato's Apolog y (Cambridge 1933). Tht. R. in other words. Even 141 . Apologie. Gomperz. 114±15. 276e7). 25±8. This conclusion is fully backed up by the Meno passage: Plato's aporetic dialogues and parts of dialogues must be interpreted as showing Socrates exhorting his partners implicitly to the care of the soul.3 to know' (cf. that the aporia extends to Socrates as well.265 We had concluded that Theaetetus and Sophist suggest that Socrates' use of elenchos in order to reach aporia and thereby to fertilise the soil for implanting true knowledge is his particular protreptic manner (in Plato). therefore. but (philosophical) investigation. Cra. We see that the elements of the description of elenchos from the Sophist as well as its double function as set out in the Theaetetus recur in this miniature of a constructive dialogue. 151e5.1 n. 114. two of the protreptic speeches are yet one level higher. cf. within this discourse.

initial exhortation and ®nal reproof have the same content: the object of oneidov is oti taÁ plei stou axia periÁ e laci stou poieiÄ tai. cf. poÂlewv thÄv exhortation w a riste andrwÄ n. d5) with the emphasis on the two major moments of examination and conviction (e rhÂsomai autoÁn kaiÁ e xetaÂsw kaiÁ e le gxw. cf. consequently. jronhÂsewv deÁ kaiÁ alhqei av kaiÁ thÄ v yuchÄ v o pwv w v belti sth e stai ouk e pimelhÄ i oudeÁ jronti zeiv. The exhortation is of the reproving type. 30a1±3 ± this is a slightly exaggerated paraphrasis of the  qhnaiÄ ov w n. 28e5. cit. kaiÁ doÂxhv kaiÁ timhÄv. on 408e5).I N T R O D U C T I O N II. it is even suggested that persuasion/exhortation is the only form of Socrates' philosophical activity (30a7±8 ou deÁ n gaÁ r allo praÂttwn e gwÁ perie rcomai h pei qwn ktl.3 jilosojeiÄ n. b4±7). 29c9. A   megi sthv kaiÁ eu dokimwtathv eiv soji an kaiÁ i scuÂn. . Comm.) points out that e ndeiknuÂmenov lacks a complementary clause like it has 23b7) and that parakeleuo menov covers both the initial incitation and the ®nal taunt. 142 . permit us to take these passages side by side. (29d7±e3).. 29e5). so.268 In the rest of the Apolog y. tauÄta parekeleuoÂmhn. . the points of agreement referred to just now. 268 I take it that in 29d4±6 (ou mhÁ pauÂswmai jilosojwÄn kaiÁ u miÄ n parakeleuo menov te kaiÁ e ndeiknu menov o twi an aeiÁ e ntugcaÂnw u mwÄn). e ndeiknuÂmenov refers to e xe tasiv and e legcov (in the terminal sense of conviction of ignorance ± Hackforth (loc. In fact. the emphasis is de®nitely on that part of Socrates' activity which is referred to with the words parakeleuÂesqai (29d5) and pei qein (30a8. taÁ deÁ jaulo tera periÁ plei onov. cf. crhmaÂtwn meÁ n ouk aiscuÂnhi e pimelouÂmenov opwv soi e stai wv pleiÄ sta. and that consequently jilosojwÄ n is coordinated with its two principal aspects. e7 egeiÂrwn kaiÁ peiÂqwn kaiÁ oneidiÂzwn) ± these words are identi®ed in the passage 31b1±7: pei qonta e pimeleiÄ sqai arethÄv . as well as the fact that this extra level is not consistently present. 30a1.3. but this time these are preceded by exhortation and followed by taunt (o neidiwÄ.

36c5±d1) are pieces of explicit philosophical protreptic. 188. but two. occurs in Aristotle's Protrepticus in a very similar context269 and quite often in the Euthydemus. Another. twÄn deÁ kalwÄ n mhdemi an e pime leian poieiÄ sqai toÁ paraÂpan. jilosojouÄnta me deiÄ n zhÄn kaiÁ e xetaÂzonta e mautoÁn kaiÁ touÁ v a llouv). section ii. this serves admirably in conjuring up a picture of a Socrates devoted to e xe tasiv and e legcov in order to check the truth of the oracle. we have to ask ourselves why Plato did not mark this exhorting Socrates more clearly as protreptic (he does so clearly enough in the Euthydemus). especially 275a1±2 kaÂllist' a n protre yaite ei v jilosoji an kaiÁ arethÄv e pime leian. bigger. First. it may be pointed out that e pime leia. In the defence against the accusation of ase beia the description of Socrates examining others (called already thÁn touÄ qeouÄ latrei an at 23c1) is worked out (28e4±6 touÄ deÁ qeouÄ taÂttontov .3.I N T R O D U C T I O N II. 31b4±5. but the more positive aspect of exhortation towards virtue and the best state of the soul is smuggled in at the same time in order that Socrates' practice may appear more positively bene®cial to the Athenians. kaiÁ taÁ meÁ n crh mata zhteiÄ n. problem is the fact that an explicitly exhorting Socrates is hardly ever to be met within Plato's dialogues. kaiÁ taiÄ v twÄ n pollwÄ n au toÁ n akolouqeiÄ n do xaiv.2. For those who hold the Apolog y to be a true 269 b 53 DuÈring h mhÁn andrapodwÄde v ge touÄ zhÄn a llaÁ mhÁ touÄ zhÄ n eu gli cesqai. Plato had made Socrates choose the Delphic oracle for his central theme. b 34. he would have ended up by depicting not one Socrates.3 Socrates' exhortations (apart from the passage quoted in full also 30b2±4. If Plato had over-stressed this protreptic aspect of the Socrates he is creating. . allaÁ mhÁ touÁ v pollouÁ v a xiouÄ n taiÄ v au touÄ . framed so as to refute the old accusations and enmities.270 cf. 143 .2 n. cf. which is the key-word of Socrates' exhortations here. 270 Cf. I think the structure of the Apolog y explains that. If more con®rmation is needed. .

as historical. evidently. Certainly. they can say with Jaeger: `From the Apolog y we know that the real Socrates tried above everything to exhort his fellow-men to practise ``virtue'' and ``the care of the soul'' ' and conclude with him that Plato tried something di¨erent. 272 Paideia.3. my interpretation of the intention of the Platonic dialogue (section ii. as we have already seen. 2.91±2 (Eng. In fact. there is contradiction not just between the explicitly exhorting Socrates of the Apolog y and the implicit protreptic practised by Socrates in the dialogues.5) does not necessarily con¯ict with a historical interpretation of Ap. cf. 1±8. the most e½cient way. Apolog y.3 account of what Socrates actually said before his judges. who uses elenchos in order to reach aporia. For an intelligent defence of the `historical' interpretation.. and in doing so exhorts his partners to the care of the soul in what is. the more so because our analysis of the dialogues so far does not suggest that Plato tries something all that di¨erent. The gradual.3. but at least as strong a contradiction within the Apolog y itself. cleverly managed change of the purely `elenctic' Socrates into a Socrates who uses elenchos side by side with explicit protreptic reminds me too much of Socrates' fallacious behaviour in the dialogues to be able to accept Ap.). Besides.272 But since we have already seen that elenchos and protreptic are tied together in Plato's dialogues. I wholly agree that `Plato wants to push his readers forward to the knowledge of virtue'. I believe that this latter contradiction can be solved only with help of the analysis of elenchos as implicit 271 This is not the place to argue this question.271 there is no problem.I N T R O D U C T I O N II. according to Sph. iv 72±80. After Socrates has ®rst limited his activity to elenchos in order to check the truth of the Delphic oracle. an attempt at harmonisation of the dialogues and the Apolog y is at least worth trying. HGPh. Plato defends Socrates against the charges of Anytus and his colleagues. I can only refer to de Strycker±Slings. 144 . he cannot very well go on saying that he does nothing else but exhort his fellow-citizens to the care of the soul. ed. the Socrates of his dialogues. but which Socrates is he defending? First and foremost.

If Plato had stated that in so many words. which is essentially based on his professed ignorance. between false and true o¨spring. by some divine gift. At the same time he also practises its alternative. but wanted to indicate the e¨ect of Socrates' questioning. Socrates could not have claimed such a divine gift in the Apolog y. as he does in the Sophist and the Theaetetus (and illustrates in the Meno). In the Apolog y. Plato did not want his readers to believe that Socrates actually went around accusing total strangers of not caring for their souls. elenchos. Through practising elenchos Socrates forces his interlocutors to account 145 .3. If this analysis is applied to the Apolog y. if he wanted to show that elenchos is protreptic. Socrates applies the method which is rejected in the Sophist as oldfashioned and as `yielding little result for much pains'. but he at least can distinguish. namely to assume that Socrates' explicit exhortations in the Apolog y are not based on the historical Socrates but on a literary trick of Plato. he could not very well have pointed out. that elenchos. True. truth and the best state of the soul' (29e1±2). to philosophy. Socrates the midwife is ignorant too. the only di¨erence between the Apolog y and the dialogues is that between explicit and implicit protreptic. The only way left for Plato.3 exhortation to the care of the soul. Plato had to make clear somehow that Socrates' elenchos is not just destructive but leads to care about `jroÂnhsiv. by bringing its subjects to aporia. in short. nouqethtikh . he would have destroyed Socrates' case. The second portrait is not intended as a true-to-life account of part of Socrates' philosophical activities. but as a commentary on the ®rst. was to do precisely as he did: to concentrate ®rst on the `elenctic' Socrates and later to combine him with an explicitly protreptic Socrates. that is. is bene®cial for them inasmuch as they will be ready for maqhÂmata. There is only one way to reconcile the two. Since he is writing what purports to be a speech for the defence.I N T R O D U C T I O N II.

206±8. I part company with Benson when he distinguishes between a Socratic and a `newly emerging Platonic understanding of the Socratic method' (50).3 for their opinions and actions. His conclusion is that all three dialogues `display the key features of the method Socrates describes himself as employing in the Apolog y' (65). Just as in the Apolog y the second Socrates explains the ®rst.I N T R O D U C T I O N II.. The Apolog y. poets and craftsmen from the ®rst portrait. 146 . In a subsequent analysis of Euthphr.275 there will be many more people who will force them to an account (39c6±d1 273 In this analysis Orwin's view that the Clitophon is a counter-Apolog y is acceptable. it is elenchos which keeps people awake. After Socrates is sentenced to death. and the more he was acquainted with Plato's dialogues.. Chrm. H. the opposition is between two di¨erent characters. reaches a similar conclusion along roughly similar lines. in the second. which I can only account for if the Clitophon has a polemic intention. and to mend their ways accordingly. The di¨erence is of course that in the Clitophon. `The dissolution of the problem of the elenchus'. de Strycker± Slings. Apolog y. and thus it is god's gift to the city (30e1±a9). OSAPh 13 (1995) 45±112..274 There is an awkward passage near the end of the Apolog y which con®rms the interpretation presented here. with the manner of the former and the results of the latter portrait. La. 29e5 e le gxw. Take away the oracle and the quest among politicians. Benson. e7 o neidi zwn. the concern for the soul (51±2). he sees Socrates' disavowal of knowledge as the central trait (50±1). is essentially a defence of the dialogue. d1 e le gcontev cf. these key features prove to play a major part (53±63) ± some subsidiary features which he perceives in the Apolog y are likewise present. he predicts to those who voted for the death-penalty that far from being rid of having to account for their lives. the easier this task would be. similarly 39d4 oneidi zein ± 30a1 o neidiwÄ. Cf.3. and what you're left with is a consistent description of elenchos as analysed in the Sophist and the Theaetetus. In the ®rst portrait. 275 With 39c7 dido nai e legcon.273 It was left to the reader to combine the two portraits. then. so in the Clitophon Clitophon's own manner of questioning is a tacit correction of Socrates' explicit protreptic. 274 H. and the explicit protreptic of the second.

als seine Testamentsvollstrecker. as illustrated in Plato's dialogues. Apolog y. 278 Sokrates. I disagree with these two scholars when they make the words refer to other Socratics as well: Plato is the only Socratic whose dialogues are written elenchos. `Zur Entstehungszeit von Platons ``Apologie des Sokrates'''. does just that. an die Athener richtet. . Apolog y. esp. . toÁ deÁ u miÄ n poluÁ e nanti on a pobhÂsetai.) 277 So H. in die Arbeit des Meisters einzutreten. The words plei ouv e sontai u maÄv oi e le gcontev refer to Plato's dialogues only: as they were written down and widely read. 147 . Now this prediction makes sense only if it refers to something that had already taken place or was taking place at the time when Plato wrote the Apolog y. die im Begri¨ stehen.277 as such writings do not force their readers to give an account of their conduct in life. Critical invectives against the Athenians (and the condemnation of the four politicians in the Gorgias) do not qualify either. they could reach a wider audience than Socrates could have reached. 210). Er praÈsentiert sich und die Freunde È ¨entlichkeit als die Erben und hier ganz formell vor der O Nachfolger des Sokrates. similarly Burnet on Ap. 106. Erbse. Bibliography). 341±63. As H. 39c8. das Plato . 355±6.'278 I do not subscribe to the suggestion contained in 276 I do not accept de Strycker's answer (which he found unsatisfactory himself ) that `Plato let himself be carried away by his own inspiration so that he is not fully aware of the implications of what Socrates says in 39c1±d9' (de Strycker±Slings.276 Since there is no evidence whatsoever of any of Socrates' pupils ever addressing people in the streets and on the marketplace.3. (I was not at liberty to put my own views forward in this respect in de Strycker±Slings. whereas I have argued above that the second portrait shows the e¨ects of the ®rst. the reference must be to Socratic writings. w v e gw jhmi´ plei ouv e sontai umaÄ v oi e le gcontev).I N T R O D U C T I O N II. Maier puts it: `die Apologie ist ein Manifest. But the written elenchos. in AusgewaÈ hlte Schriften (cf.3 nuÄn gaÁ r touÄ to ei rgasqe oi oÂmenoi meÁ n apallaÂxesqai touÄ dido nai e legcon touÄ bi ou. De Strycker could not accept the interpretation o¨ered here because he regarded the elenctic and the protreptic portraits of Socrates from the Apolog y as simply complementary.

1± 8.279 ii. I can state here only the main features of what I consider to be the meaning of the Euthydemus. Art of Plato. but I do strongly believe. and I hope I have proved. As I have put it elsewhere. as it is by the anonymous interlocutor282). 280 `Aeschines' Miltiades'. Fallacy.280 the Euthydemus is not a protreptic dialogue. considerations of space preclude any argumentation. 148 . inasmuch as the latter practise philosophy and he something of less value (306a1±c5). (HGPh. iv 274±83) and with Sprague. 111±20.4 Protreptic in the Euthydemus The Euthydemus is a three-level dialogue in which there is protreptic discourse on the second upper level. but a dialogue about protreptic. 307±8.3.4 these words that the Apolog y is programmatic in the sense that it announces dialogues that are yet to come. one may compare Isoc. 13. 1±33.3. Plato's epilogue probably intends to show that Isocrates' brand of jilosoji a is inferior even to the eristics. I feel reasonably con®dent that no one can explain the Apolog y satisfactorily if this aspect of its meaning is neglected. Cf. also Rutherford. and even switching from the second to the ®rst upper level (290e1). Obviously. 282 If he is really Isocrates. that the Apolog y is the charter on which the dialogues are founded.I N T R O D U C T I O N II. Its intention can only be grasped if we interpret the messages of all level-contents downwards ± certainly we commit a grave error of method (which in this case is bound to bring about a completely wrong result) if we single out the protreptic parts and identify them with Plato's intention. this was the more urgent because in all his dialogues Socrates employs fallacious arguments not very dissimilar 279 I do not of course pretend that with this analysis a full statement on the intention of the Apolog y has been given.281 Plato wishes to show the di¨erence between Socrates' elenchos and eristic argument (with which no doubt it was often identi®ed. on the other hand. as I think he is. 281 I ®nd myself in agreement with most of what Guthrie says about Euthd.

but Plato is able to show that as a method of implicit protreptic. halfway carried on by Socrates and Crito one level downward. and comes under the head of explicit philosophical protreptic: it ends in the conclusion that in order to reach happiness. It is essentially an attempt to determine what kind of soji a is necessary. HGPh. protre pw 307a2). The second conversation of Socrates and Clinias. but that the representatives of philosophy are repulsive to him (N.283 The ®rst conversation (278e3±282d3) is called `a model of protreptic argument' (282d4±6). the eristic method brings only confusion. This passage is the continuation of Socrates' two protreptic conversations: the search for the speci®c soji a was a ®asco. Since we are concerned here with Socratic elenchos.1). B. 149 . Socrates' elenchos does pave the way for true knowledge. I shall study only the two conversations of Socrates and Clinias. section ii. It would not seem unreasonable to doubt that this passage should be considered protreptic. one must try to acquire it ± in other words. As Socrates is successful where the eristics are not. The search ends up in aporia (292e6±293a1). Socratic elenchos is shown to be superior. Both methods fail to achieve their aim (Socrates' elenchos results in aporia). Crito complains that he would like to give his son a philosophical paidei a. since soji a can be taught (cf.I N T R O D U C T I O N II. but Crito has been exhorted none the less to keep pursuing soji a (jilosojeiÄ n).4 from (only less obvious than) the eristic dilemmas and fallacies (cf. one must practise philosophy (282d1 anagkaiÄ on einai jilosojeiÄ n).3. one needs soji a.3. iv 275±6). Socrates tells him to leave them alone and examine philosophy itself. Plato forces his readers to compare them. is presented as a continuation of the ®rst (288c6±d2). By constantly juxtaposing Socrates' and the eristics' questioning. as the ®rst conversation had already reached the goal of philosophical pro283 There are some features in the ®nal conversation of Socrates and Crito (306d2±end) which justify its interpretation as explicit protreptic.2. The ®eld in which the two types of elenchos compete is explicit philosophical protreptic.

the question whether soji a can be taught (282c1±2) is pointless if soji a means `knowledge'). W. At the same time. 288d8±289c8). Now. on 407a1 w jeleiÄ n). (This holds also for the ®rst conversation: the elements which bring about the aporia at the end of the second had been carefully prepared there.1 and 3). `Platon oder Pythagoras? Zum Ursprung des Wortes Philosophie'. Introducing the concept of virtue as a te cnh is a well-known feature of the early Platonic dialogue. while at the same time soji a is used as `wisdom' (281b6. the concept invariably causes the main aporia. it is clear that jilosojeiÄ n had not been used in its technical sense at the end of the ®rst conversation. d8. Because soji a is a te cnh.I N T R O D U C T I O N II. 150 . e4 and esp. directly at the beginning of the conversation the sojiÂa looked for is assumed implicitly to be a te cnh (with e pisthÂmh serving as trait d'union. I shall comment on two remarkable features which are typical of Plato's use of elenchos.2. it is logical to concentrate on this episode (the most important elements of the ®rst conversation have been studied in sections ii. besides. one must realise that we do not have a pure and simple specimen of explicit protreptic but one bearing Plato's stamp. 285 (1) The statement that soji a is the only agaqo n (281e3±5).). esp. as it does here. in the ®rst conversation. First. Comm. it must have an e rgon (291e1). cf.) and even by te cnh (289c2 etc. (2) The identi®cation of soji a and e pisthÂmh (281a2±b6). A more precise determination of the notion soji a is therefore a necessary addition to the ®rst conversation.3. This identi®cation enables Socrates in the second conversation to replace soji a by e pisthÂmh (288d8 etc. this e rgon must be wje limon (292a8) and therefore agaqoÂn (292a11.)285 As it is precisely these purposes in which we are interested at present. and set up so as to serve his own purposes. it had 284 Cf. but had been re-etymologised as `to strive after soji a' (the allegedly Pythagorean but in point of fact originally Platonic use of the word284). b7 nouÄ n mhÁ e cwn. Burkert.4 treptic.3. 172±4. Hermes 88 (1960) 159±77. On the other hand.

and therefore in aporia (292e6±293a1).5. The more enlightened people (toiÄ v komyote roiv) think the good is wise knowledge (jro nhsiv). J. a passage from the Republic. Kuhn. Hirschberger. O'Brien. section i.3). cf. 79±80). divining that it is something. in which a similar circular regress is signalled.286 In the background lies Plato's idiosyncratic version of the Socratic paradox `Virtue is knowledge': so it is. `Plato wants to make us understand that we had wrongly identi®ed ethical knowledge with the technician's skill' (E. 61±2 and passim. de Strycker). Bijdragen. 289±91 and passim. 151 .3. Tijdschrift voor Filoso®e en Theologie 27 (1966) 214±28 at 217 (my translation).4 been proved that soji a itself is the only agaqoÂn (281e3±5). but knowledge of a very 286 `De eenheid van kennis en liefde in Socrates' opvatting over de deugd'. These statements occur at the beginning of the discussions of the i de a touÄ a gaqouÄ. Erler. 30±5. Sokrates. If Goldschmidt is right in reading the Euthydemus in the light of this passage ± and I have no doubt that he is ± the soji a looked for in the Euthydemus is not. The discussion has resulted in a circular regress (291b8±c1). Paradoxes. Still. every soul pursues the good as ultimate end. 17±18. perhaps. cf. yet perplexed (aporouÄsa) and unable to grasp what it is (505d11±e2). only the assumption that such a Form exists can break through the circularity entailed by identifying jro nhsiv (soji a) and agaqoÂn. points to the way of avoiding it. but in any case knowledge of a higher order than technical knowledge. knowledge of the Forms. but are forced to say knowledge of the good (505b5±10). As Goldschmidt has suggested (Dialogues. Die Phronesis in der Philosophie Platons vor dem Staate (Leipzig 1932). The distinction of levels of knowledge is part of the philosophical message not only of the Euthydemus but of all aporetic dialogues. the only e rgon of soji a is soji a (292d8±e1. Sinn der Aporien. consequently. They cannot tell us what kind of knowledge.I N T R O D U C T I O N II.

3. Dialogue. 541) is completely compatible with what Kahn says about his tentative grouping of the dialogues of the `®rst period' on p. 152 . CPh 71 (1976) 109±12. esp. even if Socrates did. in his early period.I N T R O D U C T I O N II. l'amitie .287 but which I have to state because an important part of my interpretation of the Clitophon and of the dialogue as such depends on it). on ne peut les de®nir que si l'on parvient aÁ de ®nir le Bien avec lequel toutes ces Valeurs paraissent aÁ un moment donne se con287 Instead. Kent Sprague. 38±42. but certainly the theory that virtue is knowledge reposing on a deeper insight of things. `Plato's unitarianism or what Shorey said'. This interpretation of Plato's use of the te cnh analogies is at variance with the one commonly held: that Plato. Kahn. Plato never believed such a thing. is necessary if one wants to understand why he wrote these dialogues at all. Kahn. 280±95. My view that the aporetic dialogues are a separate sub-genre (review of Brandwood. Les parties de la vertu. Lysis. Erler.4 speci®c character. not the knowledge found in the arts and sciences. Hippias Maior. 48. 59±70 and passim. There is no positive evidence that the theory of Forms was already part of Plato's philosophy when he wrote his ®rst dialogues. and that he believed this because he was still under the in¯uence of Socrates' teaching. `Did Plato write Socratic dialogues?'. Euthyphron. Sinn der Aporien. In my opinion (which it would take me too far a®eld to argue at any length here. Only this special knowledge can attain the concept of agaqo n which is essential in de®ning the virtues under discussion. knowledge of another order than technical skill. Euthydemus]. le bonheur. I may perhaps be allowed to refer the reader to R. denies that there is any fundamental break in Plato's views between the aporetic dialogues and the Republic. As Goldschmidt puts it: `LaÁ est la raison deÂcisive de l'e chec des six dialogues [Laches. Charmides. really believed that the knowledge which is (or produces) virtue is not di¨erent from technical knowledge. Chronolog y.

Or ce Bien. 237 and n. nulle part Platon n'en entreprend la de®nition. . Cf. quite unintelligent [cf. there is one element in particular which puts this progress in its proper light. 303±7. 80).. A. all who are favoured by heaven (oisper a n o qeoÁv 288 Cf. 291a4) uttered these words. esp. Hawtrey ad loc. T. some of them. His progress prompts Crito to interrupt Socrates' report: Crito is unwilling to believe that it was Clinias who made these clever remarks. 118. That. In the ensuing conversation of Socrates and Crito.I N T R O D U C T I O N II. AuA 26 (1980) 75±89 at 84. he thinks of Socrates. there is a sentence in the maieutikh episode of the Theaetetus. D. who claims that the te cnh analogies have a twofold dialectical function: exhortation and refutation. s'enteÃter dans cette tentative. misses the irony in Crito's words. Socrates supposes that maybe one of the gods (tiv twÄ n kreittoÂnwn. `Those who frequent my company at ®rst appear. 289 When Crito says (291a6±7) twÄn kreittoÂnwn me ntoi tiv e moiÁ dokeiÄ . .288 A second point of interest is the remarkable e¨ect which Socratic elenchos has on the young Clinias. He has been stimulated by Socrates' questions. does not a¨ect the point. 153 . I would go one step further and identify the two (cf. JHPh 24 (1986) 295±310. next section). Now. `Sokrates' Spott uÈ ber Geheimhaltung (Zum Bild des jiloÂsojov in Platons Euthydem)'. it is in order to mark the more clearly Clinias' astonishing progress. Mais il nous indique qu'aÁ vouloir . however. If Crito does so here. Szleza k. Sinn der Aporien. which provides the best commentary on this passage. `Socrates' use of the techne-analogy'. on ®nit par tourner en cercle' (Dialogues. but as we go further with our discussions.3. kaiÁ polu ge. 279d7±8].4 fondre. so much so that he volunteers good reasons for rejecting the logopoiikhÁ te cnh and the strathgikh as candidates for the art of happiness for which he and Socrates are looking (289c8±290d8). It is unique for a ®rst upper-level character (other than Socrates) to comment on second upper-level discourse in such a way. Roochnik. L. Euthd. and Crito ironically289 agrees. Erler.

What Plato wants to do in the Euthydemus is in some respects similar to his intention in the Apolog y: he destroys the claims of explicit protreptic (by making the conversation end up by turning around). he had recourse in both to the motif of divine intervention. Plato's theory of protreptic must now be reconstructed (in relation to other elements of his philosophy) from the dialogues. implicit protreptic can. Besides. ii.290 Elenchos.3.5 Protreptic and dialogue Until now I have tried to restrict the analysis of the passages as far as possible to the upper levels of the texts in which they occur (where I have deviated from this principle. so signally succeeds where other methods (eristic. as explicit protreptic cannot. from the passages discussed.3.1).5. I do not wish to imply that the readers of Tht. that Socrates can only ascribe its success to divine intervention. Now all that is said or implied about protreptic and elenchos at the upper level or levels must be translated into terms of the lower-level communication between Plato and his readers. then. could understand this sentence only in the light of Euthd.5 parei khi ) make progress at a rate that seems surprising to others as well as to themselves' (150d2±6). yet I feel justi®ed for the following reasons. I did so in order to make some points which for practical reasons were better made in connection with the discussion of particular works). 154 .291 290 See Me ridier on Euthd. but shows at the same time that implicit protreptic is a successful alternative. 291 In doing so I shall inevitably become guilty of causing a `short circuit' (section i. more speci®cally.I N T R O D U C T I O N II. In other words. and ± one may add ± explicit protreptic) fail. but I maintain that as in both passages Plato wanted to make clear how elenchos works. (it explains itself su½ciently). 291a6. suggest a solution for the philosophical problem of virtue and knowledge.

I have tried to do so in a few words. whenever I felt it possible to elucidate the relation of the passages to the whole of the intentions of the works in which they occur. yet they do not bring us very far.3. This consistency of di¨erent texts makes it plausible that they re¯ect a consistent attitude of their author. Not only is it unable to establish philosophical knowledge in the reader (that is not its claim). or philosophy (for Plato. most people like to believe that their behaviour is (most of the time) in accordance with ®xed ethical norms (which is probably true). because they feel it does not apply to themselves. but it also fails to convince him that the care of his soul. In the Sophist.5 Plato rejects explicit philosophical protreptic because it is ine¨ective. a protreptic dialogue like Alcibiades 1 fails to achieve its end. they are plausible enough. these notions are identical) is necessary for his life to be at all worth living. then. and that these norms are mutually consistent and consistent with the rules of conduct prescribed in their society (which is hardly ever true). Both in Plato's day and in our own. For these people. the Eleatic visitor (1) The passages discussed so far yield a coherent picture of protreptic and elenchos.I N T R O D U C T I O N II. so genre-comparison does not help us here. By the same token. that does not necessarily mean that I have lifted them from their contexts regardless of these contexts. as for Socrates. reprobatory admonition like Socrates' speech in the Clitophon is useless. 155 . the dialogue was still in its infancy. when Plato started to write dialogues. (2) If I do not always describe the functions of the passages within their upper-level communication. While we can understand why Plato did not incite his readers explicitly to virtue or philosophy. The cause of its failure is man's doxosoji a: most people feel that they do already possess knowledge and do care about their souls. because most people think they are better than Alcibiades. These statements are the result of a fairly easy transposition of statements found in Plato's works. his reason for writing dialogues is not transparent. though they come from di¨erent texts and different periods in Plato's life. (3) I see no other practicable way.

108±21. and we can apply this to writing and take Plato to reject explicit protreptic. But when the visitor proposes elenchos as a better alternative. o il discorso orale come autoelenchos'. I can only say that the impression is false. 21±37.3. cit. The problem of what writings are preferred by Plato to protreptic merges into the vexed question why Plato wrote at all.5 rejects admonition in oral paidei a. 7 into account because it has already been taken as the basis of Goldschmidt's analysis of all dialogues (Dialogues. if not the only. ± On the relationship between Phdr.) should be compared with Sph. As Socrates points out in the Phaedrus (275d5±9). Proceedings of the II Symposium Platonicum (St Augustin 1992).293 The written loÂgov is the illegitimate brother of the spo292 The words calepoiÁ suneiÄ nai (Phdr. 3±12) and because I doubt its authenticity. All writings lead to doxosoji a (Phaedrus 275b2). op. it does indeed appear that Plato thought oral paidei a (elenchos followed by maqhÂmata) the best. the process of theorising about Plato's work has brought us to the end of the Phaedrus. Sinn der Aporien. the correspondence of the two levels seems to end: a teacher is able to prove a pupil ignorant by `elenctic' questioning. but a written work cannot do the same for the reader. in L. `Knowledge that the mind seeks: the epistemic impact of Plato's form of discourse'. books cannot answer questions. I do not take Ep. Ph & Rh 29 (1996) 225±43. 230b9±c1. `Phdr. 7. 156 . If in the following paragraphs I give the impression that I have picked out a few phrases from Phdr.292 so a written form of elenchos is a self-contradictory thing. I refer to two recent studies: Erler. Now. they certainly cannot ask them. ibid.). Rossetti (ed. 274c ss. Isnardi Parente. M. Predictably. considerations of space here prohibit an extensive analysis of that di½cult passage. and Ep. 210c2±3. Tht. 293 Again. Understanding the Phaedrus.I N T R O D U C T I O N II. to suit my purposes. Schildknecht.. A fruitful attempt to account for the passage as an explanation of the dialogue form in the light of Plato's philosophical views is made by C. For fuller treatment. Gill. 156±72. cf. way to acquire philosophical knowledge. `Dogmatic dialogue in Phaedrus 276±7?' in Rossetti.. 286±92. C.

1654. The spoken word is free from these blemishes (276a5±7). cf. there must be more pro®t for the reader than the idea. 7. But what exactly does the metaphor of illegitimacy entail? Phaedrus echoes Socrates by another metaphor: the written loÂgov is the ei dwlon295 of the spoken loÂgov (276a±9). 296 The words met' e pisthÂmhv (276a5) are opposed to the inability of the written word to answer questions.294 so it is reasonable to assume that the written and the spoken elenchos stand in the same relationship. Plato gives us more than metaphors. 276c9. The dialogue is an imitation of the discussion. cf.297 But at the same time 294 De Vries (ad loc. Interpretationen zu den fruÈ hen und mittleren Dialogen (Berlin±New York 1985). but does Plato write dialogues only because he wanted to imitate the discussions. . This brings us a step further. 1659.3.296 These words are to be taken quite seriously.5 ken lo gov (276a1±2). section ii.I N T R O D U C T I O N II. Pi. A. and therefore to impart knowledge. 276e2±3 (dikaiosuÂnhv te kaiÁ allwn wn le geiv pe ri muqulogouÄ nta) 157 . hence an . The fact that Phdr. Av. 16) ± the references which he gives do not con®rm this sweeping statement. gratifying though it may be.2). O. no qov. There are three serious objections to written works: they cannot answer questions. which cause real knowledge? Surely. whereas discussion is e myucov. and there is no reason whatever to suppose that Plato wanted to except his own writings from this verdict. 18 n.) wrongly says that gnh siov is not used of brothers. le goito dikai wv. cf.3. e7±277a1. and it is a dead thing. . 295 There is a conscious play on two meanings of the word: `image' and `phantom' (cf. of assisting mentally at a Socratic conversation? Fortunately. zwÄnta kaiÁ e myucon). and they are unable to defend themselves (275d4±e5). That the dialogue is Plato's written elenchos is suggested by the analysis of the questioning of the slave in the Meno (cf. SzlezaÂk is right in claiming that `trotzdem will man immer wieder die Dialoge wegen ihres erzieherischen Wertes vom negativen Urteil uÈ ber die Schrift ausnehmen' (Platon und die Schriftlichkeit der Philosophie. they get into the hands of those who have no business with them. 297 I doubt that T. Ar.27.

Erler. in id. 65±98. The reference to R. Ein Beispiel humanistischer Interpretation. 300 While analyses of single fallacies. op. on Phdr. Platonstudien (Amsterdam 1977). Most 158 . 632e5 and 752a1). G. `Zum Spiel bei Platon'. Rowe ad loc.). There is necessarily `a great deal of play'298 (277e6) in the written loÂgov. Kahn.I N T R O D U C T I O N II. Rossetti. .5 we must note that further on Socrates says that a writer who knows the true relation of speaking and writing deserves the name `philosopher' (278c4±d6). 14. 293). at the beginning of the second half ). 18±9). are legion. . H. 138±55.)'. in L. Luther. 31±2 and n. Cf. at 146±7. 501e4 muqologouÄ men ± both in crucial transitional passages) makes it quite clear that the dialogues deserve all the negative quali®cations which are given to the written word in general (the verb is used in the same way in two key passages of the Laws. pollhÂn must denote here lack of seriousness. and Plato's dialogues are full of it. he must have taken trouble to minimise them. at 276d2 (paidiaÄv caÂrin) the word paidia means rather `pastime' (De Vries. `Die SchwaÈ che des geschriebenen Logos. cf.. 299 Cf. 39±40. The end of the Phaedrus itself is marked similarly: 278b7 ou kouÄ n hdh pepai sqw metri wv hmiÄ n taÁ periÁ lo gwn (virtually: `this must be the end of the dialogue'. a short excursus is in order about fallacies in Plato: if anything. Ferber. Comm. 374. they qualify as play. there is surprisingly little literature on the role of fallacies in Plato. Szleza k. As Plato was aware of the disadvantages of writing listed above.300 They fall into three di¨erent classes. R. who points at the quali®cation of the second half of the Parmenides as `play' (137b2. Spel bij Plato (Amsterdam 1949). versucht am sogenannten Schriftmythos in Platons Phaidros (274 b 6¨. This applies especially to what one might is an obvious reference to the Republic (376d9 muqologouÄntev. but that does not follow from 276d1±8. Gundert. real or putative. J. 298 paidiaÂn .. `Warum hat Platon die ``ungeschriebene Lehre'' nicht geschrieben? Einige vorlaÈu®ge Bemerkungen'. de Vries. was ®rst found by W. There is a slight inconcinnity between the two passages: why could not someone devote his leisure to writing a completely serious book? Evidently because books can never be quite serious.3.299 At this point. (1) Conclusions which Plato may well have thought are not fallacious at all. cit. Gymnasium 68 (1961) 526±48 at 536±7. Dialogue. Sinn der Aporien. This is precisely what is accomplished in the dialogue. (n.

302 Cf. or to limit it to instances in dialogues like Prt. Each in turn. has proved rich in confusions. 131±4 (ou di kaion treated as identical to adikon). G. 507a5±6 (cf. Grg. many others blandly assume that Plato is a poor logician. M. S ˆ pleasure. cf. cf. `Persistent fallacies'. K. Klosko. pain is evil). M. Sprague. Top. Teloh. Cat. 132±56. Mind 51 (1942) 97±114. Socrates. though it is in fact fallacious: if P is true of S. Smp. `Criteria of fallacy and sophistry for use in the analysis of Platonic dialogues'. 90±1. 13b36 e nanti on de e stin a gaqwÄi meÁ n e x ana gkhv kako n. 114b6±15. `Plato's sophistry'.3. when considered as a conversation. is rich in fallacious argument. I disagree. Sprague. ProcAristSoc Suppl. so very eloquently Vlastos. Fallacy.. frequently uses fallacy to substantiate his claim that Socrates is depicted as a constantly failing educator. both among the Greeks and nowadays.302 I do not believe that this is particularly frequent. For more fundamental discussions cf. 449: `Each argument contains reasoning which. taken (as it often is) as a straightforward unilinear development. and solve the problem. H. not so 159 . cf.g. `Plato's consciousness of fallacy'. P ˆ good: if so. Arist. who in vain tries to defend the argument). 5) ± this is identical to my own view of the aporetic dialogue as Plato's written protreptic. possible instances include the two refutations of scholars tend either to deny their existence altogether. 331d1±e4. Plato uses it frequently. 201e8±202a3. `Toward a consistent interpretation of the Protagoras'. Socratic Conversations. Sprague. CQ 33 (1983) 363±74. Stewart and R..I N T R O D U C T I O N II. Fallacy and its discussion by M. 1397a7±19. e. it should make the readers want to engage in dialectic' (p. ProcAristSoc 94 (1994) 73±93. Rh. Socratic Education in Plato's Early Dialogues (Notre Dame 1986). There is an instance in the Clitophon at 407d6±7. 51 (1977) 21±44 and 45±61 respectively. Plato was clearly aware that this is a fallacy. For the theory. but I subscribe whole-heartedly to Teloh's further claim that `a Socratic dialogue should stimulate the reader to desire to break the aporia. where Socrates is thought to beat the sophist at his own game. to reach aporia the faster. (2) Plato may have used the fallacy on purpose.301 This is quite a common type of argument.g. McCabe. R. Prt. esp. Stokes. cf.5 call the `logic of opposites'. Klosko calls the `contrarycontradictory fallacy'. A. Robinson. note on 407d2±e2. then the opposite of P is true of the opposite of S (e. 301 Not to be confused with what G.. ± On the fallacies in Euthd. AGPh 61 (1979) 125±42.

Frede. section ii. 300). Klagge±N. 77b5. K.3 as referred to under (2). 206. that no one does evil willingly. not necessarily incompatible with (2): in cases of fallacy through ambiguity. understood as a way to avoid the dangers of the written word.3. esp. prevents readers from deriving knowledge from an authority. Cf. C. `Plato's arguments and the dialogue form'. 160 . 356. `In such cases Plato appears to be asserting by implication.305 Above all. 201±19. the importance of play. Phronesis 29 (1984) 193±236. 70 (1996) 177±93. Oxford 1992). cf. and n. 226±7.5 Polemarchus in Republic 1 (331e1±336a8). 303 Bluck on Men. cf. cit. D. there is the aporia. by M.'303 Cf. Evans. which implies. cf. 368±9.' Cf.3. but especially in the structure of many dialogues. Plato may have intended to suggest to his readers that the ambiguity contains a deeper truth (for example in the case of eu praÂttein `to act well' and `to fare well'). McTighe. passage. Socrates has played on those confusions and used them to elicit discordant replies to his questions. especially because of the fallacies). Methods of Interpreting Plato and his Dialogues (OSAPh Suppl. 211±14. Klosko. cf. `Platonic arguments'. on 407d2±e2. `Socrates on desire of the good and the voluntariness of wrongdoing: Gorgias 466a±468e'. section ii. The role of the questioner. from the Phdr. Smith. who in principle is not committed to the truth of the outcome of the questioning. in J. on grounds quite independent.5. esp.5. op.304 Apart from fallacies. 304 On a special type of fallacies caused by ambiguity. (3).I N T R O D U C T I O N II. esp. D. in which many thoughts and arguments are left un®nished and many loose ends remain (again. (n. but none too clearly. ProcAristSoc Suppl. can be seen in the use of myths. it seems. Comm. It is particularly interesting that this type of fallacy is frequently found when Plato tries to prove the most basic assumption of his ethics. 305 This function of the dialogue is argued for. the solumuch on Socrates' lips (though Socrates is fallible) as on the respondents'.. note on 407d2±e2.

Frede. he thought.' 161 . Sinn der Aporien.307 At the same time. Plato's dialogues could only be completely understood by those who were already fairly conversant with his thoughts (either through contact with Plato himself or through reading a number of his more `constructive' dialogues).I N T R O D U C T I O N II. one can imagine that writing as upoÂmnhsiv is useful. Erler.3. would not be justi®ed. 135a7±b2. Such writings will not cause the conceit of knowledge in those who are not naturally endowed for philosophy. Phdr. 278a7 eu reqei v). 307 Cf.5 tion of the problem of ethical and technical knowledge (section ii.4). But the dialogue form even allowed Plato to present his own views and his own arguments without endorsing them in a way which. Kahn. `Did Plato write Socratic dialogues?'. 276d3±4. 283±6. 308 Cf.308 That does not mean that he will have esoteric doctrines. able to answer questions. cit. if a reader takes the trouble to think through what is suggested in them. not put 306 Cf. to Plato at any rate. 315±19. but not at the two other places mentioned above. the more so as acquiring philosophical knowledge is. At 276d3±4 the writer himself is meant. that may account for the irony. 305). op. 216: `Obviously one can think that certain views and arguments deserve re¯ection even if one does not endorse them. If this is true.306 but that is precisely what is indicated by stamping the written word as a `reminder of those who know' (278a1). 275d1. ± There is some irony at 276d3±4. and by the same token they are capable of defending themselves. not expound. at the other two his spiritual kinsmen. cf. we can understand why Socrates indicates that the philosopher must have `more precious things than those which he has put together in writing' (278d8±9): because the written word is a secondrate medium. M.3. Plato seems to distinguish between those who have `discovered' philosophical knowledge themselves and those who acquire it under spiritual guidance of someone else (Prm. he will only suggest. They are. to some extent at least. (n. a long and gradual process. In the case of the latter.

If the reader does not already know what Plato thinks.I N T R O D U C T I O N II. But those who know will understand what Plato is talking about. Likewise. in that case written elenchos. too. that is to say.1). that is as close as Plato can bring his reader to the healthy state of knowing that one does not know.309 It means rather that he will not expose his doctrines to misunderstanding by stating them in full. the doctrine of a naÂmnhsiv as described in the Meno is a genuine part of Plato's philosophy. and will either turn away (in which case he will belong to those who have no business with philosophy) or he will be stimulated to think about the problem himself. because it does not teach (not explicitly). As I have said above (section ii. but the geometrical passage is not a proof for it. there can be no esoteric doctrines: how can one be reminded of such a secret knowledge by writings which have nothing to do with such knowledge? 162 . or indeed of Forms at all. the Euthydemus suggests the existence of a superior knowledge of the Good. but we do not hear there of the Form of the Good. he will have been exhorted implicitly to the care of his soul. Socrates' method of starting from de®nitions (often demonstrating what is and what is not a de®nition) will show him how to tackle problems. In this sense. It avoids doxosoji a. the dialogue is Plato's written elenchos. he will be ba¿ed. If it ends in aporia. will have achieved its double purpose. 3). he will have acquired some ma qhma. If by thinking through what Plato says the reader will grasp Plato's meaning. this theory does not exhaust the meaning of all of Plato's works. nor does Plato relate this doctrine (at least not in the Meno) to other aspects of his philosophy. the soul's immortality and the theory of Forms.5 down in writing.3. For example. To quote Goldschmidt once again: `Le dialogue veut former plutoÃt qu'informer' (Dialogues. 309 If writing is an ei do twn upo mnhsiv in the sense explained above. Of course.3.

310 I conclude with a comparison of the Platonic dialogue. Dialogues. . as analysed in these sections. in other words. 18: `So ist bei Platon . In both types there is exhortation and both make use of elenchos. with the protreptic dialogue of the Alcibiades type. I take Alcibiades 1 as representative of the protreptic dialogue. In aporetic dialogues. but it is not an aporetic dialogue. but this appeal is made primarily by the implicit protreptic of elenchos (whose strength is. I believe that the passages which contain explicit protreptic have rather an auxiliary function: they may strengthen the appeal to the reader. 57 and n. Yet the tentative line drawn between aporetic and constructive reasoning on the higher level of Socrates and his partners appears to be valid also on the lower level of Plato and his readers: the aporetic dialogues. of course. some less so or not at all. The protreptic dialogue contains aporia (Alcibiades `is wedded to extreme ignorance'. the explicit protreptic at the end of Euthd. Goldschmidt's analysis of the Alcibiades 1 proves that this dialogue has the basic scheme of Plato's dialogues (though he has to admit that in one respect it is quite di¨erent. Styles. also recognised by Gaiser.4 n. 118b6).I N T R O D U C T I O N II. mit der protreptischen Umkehr von Scheinwissen zum EingestaÈndnis des Nichtwissens . (section ii. 283) may be taken as indicating that in spite of eristic.3. For instance. the passage perhaps serves not to discourage the reader but to make him think about it for himself (au toÁ toÁ praÄgma basani sav. Thesle¨. der wesentliche Schritt der philosophischen paidei a selbst getan'). philosophy is necessary in order to reach happiness. . one opinion after another is refuted.5 some dialogues are protreptic. esp. until Socrates' partners have been 310 Though I agree with Gaiser that a group of Plato's dialogues is to be analysed as protreptic. . 323). his method di¨ers considerably from mine. . Gaiser starts from a reconstruction of the structure of the sophistic protreptikoÁv lo gov and recognises elements of this structure in the dialogue (cf. cf. are Plato's alternative for explicit exhortation. they are Plato's protreptic. and the aporetic parts of other dialogues. Protreptik. From the presence of these elements he concludes that Plato's earlier dialogues have a protreptic character. 163 . 1).3. 307b8).

The aporetic dialogue is able to stimulate the reader to think about ethical problems. but it is Alcibiades' ignorance. Plato could never have written anything like the Alcibiades 1. not the ethical problem. and that is su½cient evidence for his total ignorance. inasmuch as it suggests that ethical knowledge is knowledge of a higher order. The Clitophon seems to imply rejection 164 .I N T R O D U C T I O N II. the reader does not pro®t from Socrates' elenchos. ii. Alcibiades is convicted of ignorance. Alcibiades is exhorted and converted.4 Elenchos in the Clitophon We saw that elenchos leading to aporia constitutes Plato's implicit protreptic. and the area of his ignorance happens to be ethical. The protreptic dialogue does nothing of the sort. this is never pointed out in so many words. The process of purging itself constitutes the proof that these partners are lacking in the care of their souls. Alcibiades. but he never did perceive the unbridgeable gulf between explicit and implicit protreptic.4 purged of all conceit of wisdom. Therefore. on the other hand. and Socrates will show him the right way for it (124b9± 135d10). is purged of one false opinion (taÁ di kaia are not identical to taÁ w je lima. whom he wishes to exhort by implication. The protreptic dialogue is a branch of explicit protreptic which uses elenchos. and thereby to take care of his soul. at the same time it contains the germ of a solution of these problems. because it is not really concerned with ethical problems. He needs e pime leia autouÄ (119a8±124b9). Demetrius was partly right in grouping together Aeschines and Plato under the common head of `speci®cally Socratic type' of protreptic in interrogative form. because he never loses sight of the reader. which separates Plato from all the other Socratics we know. 113d5±116d4). which is the focus of interest.

1).1 of explicit protreptic and contains interrogation.3): the use of analogies and the `programmatic circular regress' at 408d5±6. this analogy forces the whole subsequent discussion 165 . This section will deal only with his refutation of Socrates' companions. Therefore we must study now its use of elenchos (if I use the Greek word.4.1 The art of the soul's perfection When Clitophon has been convinced by Socrates' protreptic speeches that e pime leia thÄ v yuchÄ v is necessary.5). on the ®nal regress and the aporia (ii.2). I have thought it more pro®table to study this subject-matter separately (section ii. that does not mean that the method is stamped Platonic a priori. he asks a number of companions how one has to proceed. on the distinction between e rgon and di dagma within the analogy (ii. Clitophon's report of his refutation of Socrates himself is abridged to a degree which makes analysis of that refutation impossible. In his question he uses the words a reth and dikaiosu nh indiscriminately to refer to the goal of the exhortation ± below. The formal features of Clitophon's report of his interrogation have been studied in connection with characterisation (section i. refutation and aporia. there is elenchos also in the protreptic dialogue).4. I shall concentrate here on the content of the analogies (virtue as a te cnh. The question is illustrated by an analogy from bodily care. The manner of refutation is not completely separable from its subject-matter (in this case: the de®nitions of the result of justice). this choice entails in some cases anticipation of the results of that study.I N T R O D U C T I O N II. ii. it can be shown that the end of that refutation is at the same time the end of elenchos as used in the Clitophon. I shall try to show that this identi®cation re¯ects a Platonic point of view. ii.4.

the object of this knowledge remains in the dark. justice and a reth had been treated as identical. When these analogies are used there.I N T R O D U C T I O N II. . So far as the aporia is concerned. it is suggested that justice is not in the full sense of the word a te cnh (cf. on the other hand. on 409a3 thÁ n . which leads to d2±3 h oun dhÁ ti sin ti apodidouÄ sa te cnh dikaiosuÂnh an kaloiÄ to. The e rgon of justice. qua knowledge. especially the aporia of the Euthydemus is closely parallel. when justice is said to produce friendship. an easily de®nable object. The procedure is strictly Platonic. Suggesting by means of analogy that a particular virtue is a te cnh is a procedure so common in Plato that it hardly needs illustration. the Clitophon conforms. and thereby to show the way to better understanding of the nature of ethical knowledge. The circular regress at the ®nal stage of the interrogation of Socrates' friends is caused by the analogy of justice and te cnh.3.3). When the brightest brain of the lot identi®es this te cnh with justice.4. Secondly. The arts have. it is in a 166 .3: 332c5±7 w Simwni dh. In Republic 1 the procedure is applied to Simonides' `de®nition' of justice (to be treated in section ii. No comparable use of the te cnh analogy (in order to reach aporia) is found elsewhere in the Platonic Dubia or Spuria. .5. he is responding in the normal Platonic way to inductive reasoning. h ti sin oun ti apodidouÄ sa ojeiloÂmenon kaiÁ proshÄkon te cnh iatrikhÁ kaleiÄ tai.). and therefore justice itself (note 410a5 dikaiosu nhn h omo noian!) cannot be de®ned (see further section ii. te cnhn). so that the statement that justice is the art of the soul's a reth (409a2±6) introduces a novel element.4) the purpose of these analogies is to prepare the main aporia. Hitherto.4. Comm. As we saw (section ii.1 to start from the assumption that there is a te cnh concerning the best state of the soul. and friendship is equated to `concord in knowledge'. as the discussion develops. but two points deserve attention.

403d2±3. 1361a2. Protr. which is found slightly more often313 it has two distinct uses. R.g. 518d10. Rh. There is one important di¨erence between the use of te cnh analogy in Plato's earlier dialogues and in the Clitophon: in the former. 314 R.315 311 129c5±e7.I N T R O D U C T I O N II. (409a2±3): in the Euthydemus. 499d6±7. but no te cnh is found. Lg.311 Of course.3. 315 403d3±4. 961d5. cf. 504c8±9. Rh.4. Arist. 312 X. Protr. b 59 DuÈ ring. 3. cf. Arist. 7. jronhÄsai. 506d5±6. 1361a4±5. b 46 DuÈring. b21. The object of justice as a te cnh is said to be `the areth of the soul' (409a3). whereas it is in the Clitophon. 6. 167 . namely to denote a particular quality of the soul. 479b3±4. 1369b21. 518e2. Ag.5. Grg. EN 1161a34±5. the di¨erence between swÄma and anqrwpov is inferred from the di¨erence between the tool and its user. e. 313 Grg. Protr. There is only one parallel for the question kaiÁ nuÄn dhÁ ti na jameÁ n einai thÁn e piÁ thÄ i thÄv yuchÄv a rethÄ i te cnhn. 282c8±d2). 1362b13±14. b 46 DuÈring. namely the one needed for attaining happiness (288d9±e2. 517e8±518a1. cf.1 uses a similar one: thÁn e n thÄi yuchÄi au touÄ arethÂn. cf. In the Alcibiades 1. the analogy is not stated explicitly. this more general way of using the te cnh analogies is also found frequently in Plato's genuine works. for instance in the second stage of Socrates' argumentation against Polemarchus to be analysed in section ii. The explicitness of the procedure in the Clitophon makes the misleading character of the te cnh analogy far more obvious than is usual in Plato. This is a rather uncommon phrase. for example. Clinias is asked to name a te cnh. EE 1241b17±18.312 Analogous to `the a reth of the body'. namely to prove or disprove a speci®c argument. which I have not found outside Plato and Aristotle. for instance when it is said in the Republic that the `good' soul (yuchÁ agaqh ) by virtue of its areth makes the body as `good' as possible.1 more general way. Arist.314 or to refer to what may be roughly translated as `the good condition' of the soul.

ruling. in a number of places in Republic 1. and above all. 168 . which had resulted in its contrary (350c10±1 o meÁ n ara di kaiov hmiÄ n anape jantai w n a gaqoÂv te kaiÁ sojo v. not `result' (sight and hearing are the e rga of eyes and ears. in which the just man is said to be h rmosme non (443e2). thinking. in Socrates' report this result is reformulated as thÁn dikaiosuÂnhn a rethÁ n einai kaiÁ soji an (350d4±5). ®rst in Socrates' discussion with Polemarchus (335c4.316 According to Socrates. everything which has an e rgon also has an a reth (353b2±4). life (353d3±10). therefore without justice one cannot live well. 352e5± 10).1 At one place in the Republic the phrase is used for the conception of justice as a harmony of the soul: o monohtikhÄ v deÁ kaiÁ hrmosme nhv thÄv yuchÄv alhqhÁv arethÁ poÂrrw poi e kjeuÂgoi an autoÁn (the oligarchical man. 356). 554e4±5). later in his refutation of Thrasymachus' statement that the unjust are both wise and good (348d3±6). o deÁ adikov a maqhÂv te kaiÁ kakoÂv). Here the attributes of the soul clearly point back to the de®nition of justice at the end of Book 4 (443c9± 444a2). it had already been established that a reth of the soul was justice (353e7±8).I N T R O D U C T I O N II. where the soul's areth is identi®ed with soji a. The e rga of the soul are caring. Only in Republic 1 do we ®nd a normal equation of justice and the areth of the soul. this is not said in so many words. Next.5. justice and a reth had been identi®ed.3 and n. Now. Socrates wishes to refute Thrasymachus' claim that the unjust live better than the just (352d2±4). cf. Soc316 The analogy a reth of the eye ± a reth of the soul is imitated at Alc. Without its proper a reth a soul cannot ful®l these e rga.4. Yet it is only an oblique reference: though the `true areth of the soul' is obviously identical to justice. without its proper areth nothing can ful®l its e rgon well (353b14±c2). 1 133b1±10. section ii. He does so by introducing the notion of e rgon which in this context means `function'.

where justice is the corrective art concerning the soul. In making the soul's areth the object of an art subsequently identi®ed with justice. Obviously. dikastikh and dikaiosuÂnh is understandable from this doctrine only (sections ii.3.4. 169 . the concept of justice as an orderly state of mind is found (504d1±3).3. as well as in the Republic.4): the good politician makes his fellowcitizens better men. the di¨erence between this concept and that of justice as object of a te cnh is mainly that between theory and praxis: the Republic asks what justice is. 317 Socrates could easily have made the point that since human areth is either of the body or of the soul. the soul's a reth as that of other things is said to be the result of (among other things) te cnh (506d5±8). Plato does not take the arguments he puts into Socrates' mouth too seriously. the conception of justice as (knowledge producing) the best state of the soul. In the Gorgias.I N T R O D U C T I O N II. Republic 443c10). and justice is not areth of the body.1 rates is therefore guilty of negligence in adding the genitive yuchÄ v at 353e7±8. the Gorgias how it can be created and furthered. but the similarity is not close enough to make it plausible that our author had this particular passage in mind.317 It is clear that in associating justice and the soul's areth our author follows the Republic ± whether the explicit statement in Book 1 or the doctrine of the later books is hard to decide.5. analogous to medicine in the same way as lawgiving is analogous to gymnastics (464b7±8). not as a speci®c way of behaving towards others (ou periÁ thÁ n e xw praÄxin twÄn autouÄ. At any rate. At one place in the Gorgias. although the author does not make use of it in refuting the positions taken by Socrates' disciples and Socrates himself.2. is clearly present in the Clitophon. therefore it must be a reth of the soul. otherwise he would have made him justify the addition of yuchÄ v along these lines. he diverges from the Republic and moves into the atmosphere of the Gorgias. ii. Clitophon's statement that Socrates often identi®ed politikh .

cf.2 The reason why he does not do so is evident: the discussion was intended to end up in aporia. . but on the one between te cnh and e rgon (for the reason why the latter distinction was introduced.4.4. The latter is clearly and unambiguously distinguished from te cnh as such (e stin deÁ touÂtwn qa teron ouke ti te cnh 409b3±4. is not completely distinguished from te cnh itself. c1 etc. Clitophon proceeds to impose a new restriction on his partners.2 The result of justice Having forced the discussion to start from the assumption that justice is a te cnh. cf. So. thÄv te cnhv deÁ thÄv didaskou shv te kaiÁ didaskome nhv e rgon. a dhÁ ouk e stin te cnh) and is called e rgon (b5.I N T R O D U C T I O N II. . as taught and learned] toÁ meÁ n e rgon. the technical concept of virtue is a well-known Platonic method of creating aporia. 170 . The second analogy makes this quite clear: 409b5±6 kaiÁ tektonikhÄ v deÁ kataÁ tautaÁ oi ki a te kaiÁ tektonikh [viz. why did the author ®nd it at all necessary to make it? He could have restricted himself to stating the obvious. the emphasis is not on the distinction between the two e¨ects (e rgon and di dagma). 6. . The e¨ects (taÁ apotelouÂmena) of a te cnh are said to be twofold: to cause new men to become tecniÄ tai and to execute its speci®c task. toÁ deÁ di dagma. As we saw. skeu h . and that he agrees with it. the other part is]. The former. But the author does more than just adopt a Platonic device: when the object of the te cnh in question is stated to be the soul's areth he indicates (as he had done towards the end of the protreptic speech) that he has completely grasped the implications of Plato's theory of justice. .). d1 taÁ xuÂlina . ii. below). as can be seen already from the wording e stin deÁ tou twn qaÂteron ou ke ti te cnh [obviously. If the distinction between e rgon and di dagma is secondary. though called (with a rare term) di dagma at b6.

3.2 that a te cnh is di¨erent from its object or result. also 408e2±3 pwÄv a rcesqai deiÄ n jameÁ n dikaiosu nhv pe ri maqhÂsewv ± more neutral 410c5±6 h ouk eide nai se [sc. just as ± in general ± he makes no distinction between virtue and te cnh. as far as I can see. dikastikh and dikaiosu nh are identical (408b3±5). (2) the e rgon which is to be de®ned.4). but not to the point.4. that politikhÂ. ii.I N T R O D U C T I O N II. When producing new just men is ®rst called di dagma and then identi®ed with te cnh itself. thÁn dikaiosuÂnhn] h ou k e qe lein au thÄv e moiÁ koinwneiÄ n). this would have made it su½ciently clear that Clitophon is asking not for a de®nition of justice but for a delineation of its e rgon. in other words. That apart from this e rgon it is also a quality of justice that it can impart itself to others (b6±7 thÄv dhÁ dikaiosu nhv w sau twv toÁ meÁ n dikai ouv e stw poieiÄ n) is interesting in itself. is that the opposition e rgon : di dagma serves to prove the thesis which Clitophon had ascribed to Socrates. (2) health. cf. that justice is knowledge. The only justi®cation. Whether he actually thought that justice is knowledge and can be taught is another 171 . By analogy.5. The curious way in which the te cnh analogy is constructed here is to my mind designed for picking up and establishing this point: the two products of medicine are (1) new doctors. The analogy implies that justice can be taught. (408b2±3). This is in keeping with the wording of the end of the protreptic speech: twÄ i maqoÂnti thÁn twÄ n a nqrwÂpwn kubernhtikhÂn ktl.3. It is also implied by Clitophon when he asks Socrates to stop exhorting him and get down to business (410d1±5. This is a simpli®cation of a doctrine elaborated in Gorgias and Politicus and alluded to also in Euthydemus: the just politician makes his fellow-citizens better men (sections ii. this means that justice is its own teaching-matter. It appears that the author does not wish to cast doubt on the statement that virtue can be taught (408b7). the two products of justice are (1) new just men.2. likewise for carpentry and the other arts.

Perhaps the imperative e stw indicates scepticism. we must realise that the introduction of e rgon is possible only because justice had been forced into the framework of te cnh. if not to subscribe to this doctrine. As we saw. section ii. It follows that our author must have had another reason for avoiding the question `What is justice?' and introducing the e rgon. that those who have learned it are the true politicians.2 matter. the author wanted to state at this point that justice can be taught. The reason why the author should have done that must be that he wanted. which. whereas in view of Aristotle's treatment of the relation of justice and friendship (section ii. In order to trace this reason. Of the three de®nitions of the e rgon of justice two appear in Republic 1 as de®nitions of justice itself. When Clitophon at this point of the dialogue constructs his own argument in a way which indirectly supports a motif from the speech which he had reported in a vein of parody. It may seem that the author employed this device in order to work in answers which were suitable material for an easy refutation: yet I do not think this to be true.4. We must now ask why the distinction between te cnh and e rgon was introduced.5. and to underline what he had reported previously as a Socratic doctrine. at any rate to indicate that it was to be taken seriously. this can only mean that the author makes him except this doctrine from the ironical treatment which had been given to other themes in the speech.I N T R O D U C T I O N II. is not relevant to the progress of refutation.) `friendship in the cities' would not have been absurd as a de®nition of justice. as we saw.2. So much for the notion of di dagma.3) the distinction is an impediment rather than a contribution to the ®nal circular regress. the question is the more relevant because (as we shall see. But however that may be. inasmuch as they are able to teach it themselves. our author (whether or not he himself was Plato) shared Plato's view that justice is not so much a series of actions 172 .4.

This is where the introduction of e rgon comes in positively. bene®ting one's friends and harming one's enemies had been presented as a de®nition of justice. the fact that the Clitophon reformulates the outcome of the debate in Republic 1 (oudamouÄ gaÁr di kaion oude na hmiÄ n e ja nh on bla ptein.4. In the Republic. he could not accept `never harming anyone' or `bene®ting everyone' as an adequate de®nition of justice. but he certainly could accept it as a description of the practical manifestation of this state of the soul. To put it more sharply: under the disguise of an aporia reached repeatedly in Clitophon's sessions with Socrates himself. although we shall see that it is in fact a point of view which was defended emphatically by Plato. Given his conviction that justice is a state of the soul. as I have tried to establish. which (as we shall see) was borrowed from Republic 1. let us compare the third de®nition. 335e5) in a positive way. In this light. whereas later on it appeared that the just man never harms anyone.I N T R O D U C T I O N II. This rather 173 . 410b2±3) is not an acceptable de®nition of justice. the author could not accept the result of Socrates' discussion with Polemarchus as a de®nition. the introduction of e rgon in the Clitophon criticises the source of the third de®nition from the Platonic point of view which. Bearing this in mind. by introducing the notion of e rgon. I think that our author had no quarrel with Plato concerning this point of view. becomes highly signi®cant. the author. was also that of the author of the Clitophon: while `bene®ting everyone' (pa nta gaÁr e p' wjeli ai pa ntav draÄn. but. it is a satisfactory description of its e rgon. since (like Plato) he considered justice to be adequately de®ned only as a healthy state of the soul (caused. makes it clear that. but this is not said or implied in the Clitophon. In other words. the result is positive when two and two are put together. seen in its proper light.2 as a state of mind. This is a negative result. by knowledge of the Form of the Good).

ti nov e stiÁ n e pisth mh e kaÂsth tou twn twÄn e pisthmwÄ n. is in fact an important tenet of Plato's philosophy. o tugca nei o n allo authÄv thÄ v e pisth mhv).3(6)).7. this interpretation all but annihilates what is often considered the main argument against authenticity: that it is Socrates himself who tells Clitophon that it is the e rgon of justice to help friends and harm enemies.4. Of course. and that Clitophon refutes this. if the outcome of the discussion leaves Socrates virtually unharmed (section i.2 oblique statement of agreement is in perfect keeping with our author's way of agreeing between the lines with statements which are criticised on the surface. and. they must have understood that Socrates (in other words Plato himself ) was not attacked. the notion that the just man never harms anyone. or for that matter of any Platonic dialogue. the reader is meant to infer that the discussion now turns on a basic item of Plato's thought. In Plato. would never have said such a thing. We might therefore say that when Clitophon turns towards Socrates himself. it is found in the Charmides (166a3±5 a llaÁ toÂde soi e cw deiÄ xai. But those readers of he Clitophon who were acquainted with Republic 1 knew that the de®nition was proposed there by Polemarchus and refuted by Socrates ± when in the Clitophon they encountered a better refutation than the one given in Republic 1. with less emphasis on their 174 . this can only mean that Plato's thought is left unharmed.5.5. the Socrates of Republic 1. If I am right.3). It seems to me super¯uous to trace the origin of the commonplace notion that the e rgon of a te cnh is distinct from the te cnh itself. As we shall see (section ii. though presented as a negative result in Republic 1. It is more than a coincidence that it is precisely the third de®nition which is shielded by the distinction te cnh : e rgon.3). Yet the question remains whether Plato was really capable of exposing the literary character Socrates to such a misunderstanding (see section ii.I N T R O D U C T I O N II.

htiv ouk e sti me rov thÄv oiki av (b 68 DuÈ ring).4. The regress found at the end of Socrates' second conversation with Clinias in the Euthydemus (section ii. as we saw earlier (section ii. e te ra e te rwn e stai. 351d9).  O moÂnoia is shared knowledge. in the Euthydemus (292b4±5. The result of justice as knowledge is knowledge.. The opposition e rgon : di dagma (or anything like it) is not found elsewhere.3 distinctness. The only agaqo n is wisdom. w sper oikodomikhÁ oiki av.) of things (heat) and living beings (horses) which are not a te cnh. esp. As we saw. d1±2) and Republic 1 (332e3 etc.I N T R O D U C T I O N II. 352d9 etc.3. the word e rgon is used there also to denote the speci®c `function' (335d3. `Epinomis'. Its result must be an agaqo n. but there is no direct connection with the Clitophon ± Aristotle is rather correcting the Euthydemus passage here (Einarson. DuÈring. 241).3 Aporia and progress The circular regress at the end of Clitophon's discussion with Socrates' companions is arrived at by the following steps: (1) (2) (3) (4) Justice is knowledge and has a result. 32. h jro nhsiv] poihtikh .1).4.4) is very similar: (1) (2) (3) (4) Wisdom is the only a gaqoÂn and has a result. There is an interesting parallel from Aristotle's Protrepticus: ei gaÁr e stai [sc.4. 272 n. The result of wisdom is wisdom. the circular regress in the Euthydemus may be interpreted as suggesting that the basic principle (wisdom has a de®nable result because it is a te cnh) is wrong: ethical knowledge is knowledge of a higher order than technical 175 . ii. Its result is omo noia.

318 (2)  O moÂnoia is shared knowledge. 176 . sections ii. it is proved at the most that we had wrongly de®ned omoÂnoia (step 3). The Euthydemus therefore contains a fairly clear message as to the nature of ethical knowledge. the circular regress would have shown the way towards a better understanding of ethical knowledge far more clearly: (1) Justice is o moÂnoia.4. The Clitophon may be taken to point the same moral. 505a2±4. (3) Justice is knowledge. the one in the Clitophon does not invalidate it: if it turns out that to de®ne the result as o moÂnoia creates a regress. we saw that the description of justice as a te cnh concerning the soul's areth (section ii. which the author goes out of his way to emphasise (section ii. the distinction between te cnh and e rgon.4. 126±7. In that case we would have had a circular regress similar to the one in Republic 6 (the good is knowledge of the good.2) only complicates matters. or the result of justice (step 2). cf. Evidently the author of the Clitophon did not perceive this possibility. section ii. The regress in the Euthydemus invalidates the basic analysis of wisdom as a te cnh directly.I N T R O D U C T I O N II. C.3 knowledge. Cross±A.4. R.2 and ii.1) implies assent to the conception of justice as an harmonious state of mind. the Clitophon may be taken as conveying the same message.319 When the Clitophon 318 This would have been a plausible de®nition. but this is far from necessary. or he did not intend to suggest the true nature of justice as knowledge more clearly. Now.4.2. D. If Clitophon had asked for de®nitions of justice itself. Plato's Republic.4) and the message would have been the same. Woozley. R. esp. this state of mind depends ultimately on knowledge of the Form of the Good. but there are some di½culties. In the Republic. Besides.5. A Philosophical Commentary (London 1964). 319 Cf.3.

I N T R O D U C T I O N II. to attack the Republic. as knowledge. Euthydemus. As the Clitophon appears. kuÂklwi periio nta poiwÄ n. In any case. 177 . 75. Hp. that is. on the surface.3 ends up by de®ning the result of justice. this de®nition may be taken as a sign that our author subscribed to this view. not in con¯ict. a signal that the reader is to understand the whole discussion of justice as being in harmony. the art of the soul's perfection. In the Clitophon. it is not necessary to assume that he did not completely grasp Plato's theory of justice just because he does not direct his readers' attention to it (Plato does not do this himself in the Euthydemus. compare for example Socrates' words in the Euthyphro h ou k ai sqa nhi o ti o loÂgov h miÄ n perielqwÁ n paÂlin ei v tautoÁn h kei. 15b10). it is stated in so many words that there is an aporia (409e10 aporouÄntev in our dialogue. helps them to understand the idea of justice as `aiming always at bene®ting all' (410b2±3. cf.4. Charmides. which obstructs the readers' view of justice in itself (in relation to the theory of Forms). either). Hippias Maior. the 320 Ly. cf. with that of the Republic is far more urgent than a pointer to the theory of Forms. the circular regress is explicitly marked as such (410a2 peridedraÂmhken ei v tau toÁn o loÂgov toiÄ v prw toiv). 174b11.2). Similar explicit statements are found in Lysis.320 they always refer to the ®nal aporia. (15b10±c1.4. 303e12±3. Dialogues. section ii. Usually the aporia coincides with the end of the dialogue. 222d1±3. Euthd. 291b8±c1. This is normal Platonic practice. the result of elenchos is negative (apart from its purifying aspect). Moreover. where it does not.Ma. The argument has ended where it had started. we have seen that the distinction between justice and its e rgon. but I hope I have shown that there is no contradiction between Republic and Clitophon. Goldschmidt. Chrm. Nobody can expect the whole of the Republic to be present in a Short Dialogue which deals with the inadequacy of explicit protreptic.

We have seen from the Theaetetus and the Euthydemus that Socratic questioning. (410a1).3. the discussion of Clitophon and Socrates is only apparently negative (section ii.4.4). 178 . is at any rate a stimulus for independent thought.4. I infer that with the interrogation of Socrates' companions.3 Euthydemus as quoted in sections i.3. doubtless a conjecture (the hypothesis of a correction of a mechanical error in copying ± i kanoiÁ having been replaced by a dittography of the ®rst part of the following e piplhÂttein ± may be dismissed con®dently). In fact.5. In our analysis of the characterisation of Socrates and 321 Written in the margin as a variant reading (gr. why does Clitophon stress the bystanders' ability to criticise their comrade? The words i kanoiÁ h san are a bit odd anyway. Their ability may be the consequence of Clitophon's elenchos. if our author wanted to indicate the bene®cial function of elenchos. the elenchos in the Clitophon has come to its natural end.2).3) and are treated as such. so the refutation had concerned all of them. so much so. even when the outcome is negative.321 which hardly makes sense.4).3.5. they had all been interrogated and had given the string of de®nitions toÁ sumje ron. though I do so with some di½dence. toÁ lusitelouÄn. toÁ de on. I think an echo of this progress may be perceived in Clitophon's words oi paroÂntev ikanoiÁ hsan ktl. gr. The boundaries of a Short Dialogue are narrow. that in either of these dialogues the motif of divine intervention is used to underline the marked progress in the intellectual capabilities in Socrates' partners (section ii. 1807). One aspect of elenchos as used in the Clitophon has to be mentioned. toÁ wje limon.) by the ®rst scribe of A (Paris.I N T R O D U C T I O N II. he could hardly have done better. they constitute one amorphous group (section i. ii. and they induced an anonymous reader from antiquity (or the early Byzantine era) to change them into epeceiÂrhsan.

283). Both deal with the value of explicit exhortation. the possibility of a negative sidee¨ect. the same must be maintained of the Clitophon. explicitly. Can this be said as well of the Clitophon. the Euthydemus and other works of Plato.3). is obviated by a ®nal passage exhorting again.3 Clitophon (section i. 322 This. the Euthydemus is a dialogue about protreptic. and both show that it does not reach its goal. Plato's aporetic dialogues are implicitly protreptic (section ii.5. are not allowed to walk away with the idea that all this talk of caring about the soul is nonsense. In this respect the Clitophon is again similar. In the Euthydemus. at least. . will have to be recognised even by those who ®nd themselves unable to agree with all details of the interpretations presented in this and the previous sections. makes no sense at all if the present one is not intended seriously). The study of his use of elenchos bears this out completely.I N T R O D U C T I O N II. which employs elenchos as well as explicit protreptic in order to convey a positive message. 410e1±3. the next sentence. 179 . and we saw that here he is completely serious (section i. to philosophy (section ii. Clitophon's last appeal takes up the core of Socrates' exhortation (410d5±e1 qeÁ v toÁn KleitojwÄ nta omologouÄ nta wv e stin katage laston twÄn meÁ n allwn e pime leian poieiÄ sqai.3. which likewise contains aporia? This question can be answered best if we compare Clitophon and Euthydemus.322 and it also shows that elenchos is handled in the same way as it is in Plato. of the reader turning away from exhortation (and thereby from philosophy) for good. . Even those readers who had not understood the implications of our author's frequent allusions to the Republic. One question remains to be discussed.4 and n. therefore. yuchÄ v deÁ .5) we concluded provisionally that the author of the Clitophon has a good understanding of Plato's use of the dialogue. If.4. hmelhke nai ).3.

1) as is also the case with the author's verdict on the third de®nition (section ii. my aim is to determine their provenance and the light this provenance may shed on the intention with which the Clitophon was written.3. Our dialogue is therefore not a contribution to that development (nor does it pretend to be). the series of subsequent de®nitions is concentric: each de®ni180 . the concept of justice as the best state of the soul. likewise.4. or at any rate as a sort of knowledge leading up to that state. Finally. pro®table As is also the case in a fair number of Plato's dialogues.5 Justice in the Clitophon In this section the three subsequent de®nitions of the e rgon of justice in the Clitophon will be compared with similar statements from other Socratic literature.1 Bene®cial.1 ii. ii. at some places a `positive' thought about the role of the just statesman is implied (sections ii. ®tting. It can be shown that the Clitophon is. the role justice plays in it is thoroughly secondary: it serves as a means of proving ignorance.2). Clitophon's statement that Socrates identi®es politics with judication and justice (408b4±5) belongs here.1). While these ideas may give us a clue as to the author's philosophical background.4. Therefore I have left them out of account in the discussions of the three de®nitions. they have no bearing on the explicit statements about justice which are reported and refuted by Clitophon. wholly dependent on other sources which are for the greater part still extant.3.4. I have not attempted to examine the place of these de®nitions within the development of Greek popular or philosophical ethics. can be read between the lines (section ii.5. for these de®nitions.5. ii. useful. To be sure. not to further knowledge.I N T R O D U C T I O N II.2.

I N T R O D U C T I O N II. ®nally as knowing oneself (164d4) ± this de®nition is amended subsequently to `knowledge concerning itself and the other kinds of knowledge' (166e5±6) and the search ends up in aporia.2). the right thing'. however. in the Charmides. and the easiest to refute. then as doing taÁ e autouÄ (161b6). toÁ sumje ron and toÁ de on323 may have been separate de®nitions. the theoretical framework serves to suggest a positive result in the case of the ®nal de®nition. the author needed concentric reasoning. Evidently. Thus. As it seems a priori improbable that toÁ sumje ron. swjrosuÂnh is de®ned ®rst as acting calmly (159b2±6). which may or may not be reached. At most. he would not have made Clitophon address Socrates' companions before asking Socrates himself. next as doing what is good (163e8±11). in that the ®rst de®nition (in this case a string of de®nitions) is the farthest from the object of search. the reasoning is concentric only to a certain extent: the main aporia is reached after the second.1 tion is an improvement upon the former. toÁ lusitelouÄn occurred independently as de®nitions of justice or the e rgon of justice in di¨erent Socratic writings. It conforms. though both in the Clitophon and in the Cratylus these words are used as synonyms (cf. toÁ w je limon. toÁ de on. 323 `the ®tting. and constitutes a closer approximation of the object of the search. 181 . not the ®nal (third) de®nition.5. The progress is ®rst from external behaviour to a mental state (ai dw v).4. Comm. then as ai dw v (160e3±5). if not. on 409c5. the di¨erent pupils of the upper level are probably not to be `translated' into di¨erent Socratic authors in the message of the lower level. then again from external to internal (knowing oneself ) and ®nally from concrete to abstract (knowledge of knowledge). In the Clitophon. next from particular to general (doing what is good). Comm. next from a general formula to a description in terms of relations towards others (doing taÁ e autouÄ ). cf. on 409c5). As we have seen (section ii.

without the terms of the series being divided among separate persons. There is little ground for assuming a connection with the Clitophon ± even if the latter were authentic. answers tauÄta taÁ periÁ toÁ agaqo n te kaiÁ kaloÂn. where Hermogenes. asked what words in the ®eld of ethics are left unetymologised. asks him to de®ne toÁ di kaion: kaiÁ opwv moi mhÁ e reiÄ v oti toÁ de on e stiÁ n mhd' oti toÁ w je limon mhd' o ti toÁ lusitelouÄn mhd' o ti toÁ kerdale on mhd' oti toÁ sumje ron.I N T R O D U C T I O N II. Pavlu's basic premise is false: kerdale ov means `cunning' only if applied to persons. otherwise it is neutral. 15) because its meaning is pejorative (`uÈ bervorteilend. The order of Republic 1 is slightly more logical in that there may be a di¨erence in meaning between the ®rst and the second term of the series.325 and toÁ sumje ron opens the series instead of closing it. allaÁ sajwÄv moi kaiÁ akribwÄ v le ge oti an le ghiv´ w v e gwÁ ou k a pode xomai e aÁ n u qlouv toiouÂtouv le ghiv (336c6±d4). sumje ronta te kaiÁ lusitelouÄnta kaiÁ w je lima kaiÁ kerdale a kaiÁ ta nanti a tou twn (416e2±417a2).5. the parallel would mean little more than it does for Republic 1: we have here nothing but an occasional instance of self-repetition. verschitzt'): Clitophon's refutation uses carpentry and medicine as analogies. in the Clitophon. toÁ kerdale on is absent. Kleitophon'. Thrasymachus.v. in identical order. Yet another parallel is provided by the Cratylus. The absence of justice from the context in the Cratylus passage robs it of its signi®cance for our investigation. in which arts there is no room for `the cunning'. 182 . having accused Socrates of never answering a question himself. 325 According to Pavlu (`Pseudopl. but not between the second and the rest.324 The ®rst three terms occur.1 It is therefore not surprising to ®nd a virtually identical series of adjectives in Republic 1. Leaving aside this third parallel and concentrating on 324 This parallel appears to have been ®rst pointed out by Steinhart. verschlagen. 1). 54±5. words or plans (LSJ s.

Whether or not this is a correct interpretation of the Republic passage (cf. . towards understanding the Republic. and Just. is immaterial. che quelle erano le idee della scuola Socratica' (17). cf. 214.327 The only way to test this supposition is to examine its contribution. Thrasymachus' words would indeed mean that the de®nitions he refuses to accept are Socratic. 183 . W. or we would have to look for a source elsewhere. which one does not expect there. They would also imply assent to the tenor of the Clitophon.5. Sophistik. this meaning could be (and has been326) read into it. C.I N T R O D U C T I O N II. we must therefore choose between dependence of Clitophon on Republic 1. see also Tucker ad loc. 149 and my review of MuÈ ller. Zuretti: `Trasimaco non vuole che Socrate gli risponda che il giusto eÁ il deÂon . 166±7. . H. dependence of Republic 1 on Clitophon. negative. because the string of de®nitions is not defended there. The result is. 327 For a highly fanciful relation between Clit. Alternatively it might be supposed that our author gladly adapted the string of de®nitions which he found in Republic 1 because it suited his purpose ideally: he wished to illustrate that Socratic exhortation does not lead to knowledge. infra (2)). MuÈ ller. in my opinion. Kurzdialoge. we must inevitably conclude that the two passages cannot be independent of each other. si vede manifestamente. the borrowing of the series might imply that Thrasymachus is there taken to mean that the answers precluded by him were in fact those usually given by Socrates. only to uqloi.. Thrasymachus' 326 Cf. or dependence of both on a common source. sapendo. if any. Gomperz. (2) If we suppose that Plato was inspired by the Clitophon passage when writing Republic 1. In neither hypothesis is there real criticism of Republic 1. (1) If the Clitophon depends on Republic 1 (which is the common assumption).1 the passages from Clitophon and Republic 1. Either the string of de®nitions in the Clitophon would become an original invention.

similarly a reference to the accusation contained in the Clitophon is out of place ± Thrasymachus' objections to Socrates' method of question and answer have little or nothing to do with Clitophon's point that Socrates can only exhort others and is unable (or unwilling) to impart knowledge (cf. Comm. Clitophon's refutation of the string of de®nitions gives a strong impression of having been borrowed (cf. this assumption would involve an additional di½culty. The only logical way out is for Thrasymachus to give his own de®nition of justice. on 409c4±d1. sumje ron is countered by Socrates in an analogy: if you ask how much twelve is and add that the answers 2  6. c6±d1). lusitelouÄn. nothing is gained by assuming that the de®nitions were borrowed. you preclude any answer (337a8±b5). that the Clitophon is in this respect dependent on Republic 1. For the Clitophon. 184 . Kleitophon'. Priority of the Clitophon to Republic 1 is therefore unlikely. kerdale on. 3  4. Comm. as one would have to suppose that the characters Clitophon and Thrasymachus were borrowed from Republic 1 even though the string of de®nitions was not. The assumption that the forbidden answers were in fact given in Socratic (or other) literature is irrelevant for the interpretation of Republic 1.1 exclusion of the series de on. (3) The main argument o¨ered sub (2) against the Clitophon as the source of Republic 1 holds good also for the assumption of a common source for both: the episode in Republic 1 can be explained from itself. 346a6±c4. The whole episode has been carefully constructed in such a way that Thrasymachus can plausibly begin by accusing Socrates of never answering (336c2±6) and end up giving an answer himself.I N T R O D U C T I O N II.5. and proceed from hypothesis (1). his source was identi®ed by Pavlu (`Pseudopl. 15) with Republic 1. on 410c6 ou k e qe lein authÄv e moiÁ koinwneiÄ n). w je limon. 6  2 and 4  3 will not be accepted. Therefore I shall discard the assumption.

the Republic passage tries to de®ne the particular e rgon of each art and to discard accidental e rga (such as earning money). Secondly. are inadequate descriptions of the e rgon of justice because they apply to the e rga of every art. passage to be derived from PeriÁ o monoi av by Antiphon the Sophist (509±11): Antiphon is alleged to have said that `the natural base for the good is always to be bene®cial.5±25. `U 4. Sophistik (Darmstadt 1976). Antiphon equates no mov and do xa (87 b 185 . never to harm' (extracted from 87 b 58 D.4. 161. for instance health (medicine). Classen (ed. cf. There is in fact little similarity.330 But there can be no doubt that if the author had intended this. in which conversation there is indeed a passage dealing with o mo noia. 146 n.5. 21±2.I N T R O D U C T I O N II. Mem. The adverb marks a higher stage in the concentric reasoning. 534. 330 Kunert. There are several 328 Schleiermacher. 334. Of course this does not mean that the de®nition of the e rgon of justice as jili an e n taiÄ v poÂlesin poieiÄ n was not borrowed from some Socratic331 source. 493±516 (I quote from the second work) considers the Clit. Steinhart. ii. Protreptik.±K. If the wjeli a of Republic 1 is equated with the e rgon of the Clitophon. 331 E. 505±6) which is compared to Clit. J. The adverb is usually taken as a general description of the style or manner of reasoning of a particular Socratic author. 25. Studi sul pensiero antico (Naples 1938). be it Plato himself. he would have said komyoÂtata e doxe le gein (cf.5. 55 and 72 n. Gaiser. Kesters. whereas the Clitophon tries to prove that general notions like sumje ron. and nothing more. de on etc.328 Xenophon329 or Antisthenes. Comm. who refers to X. 182±3.2 where every art is said to have an i di a w jeli a. 9.. 66±97 ˆ C. n. Bignone. `Authenticiteit'. safety in sailing (steersmanship). Necessitudo. 329 Yxem. 409d6±7 thÁ n jili an agaqo n t' e jh ei nai kaiÁ ou de pote kakoÂn.).2 Friendship in cities This de®nition is ascribed by Clitophon to a companion ov dhÁ komyo tata e doxen eipeiÄ n (409d3±4). È ber Platon's Kleitophon'. cf. on 409d3± 4).

ad loc. just as in Clit.). consequently Antiphon must have thought that justice is e pisthÂmh (510). since the friendships of children and animals cannot possibly be brought under the head of `friendship in the cities'. The latter part of this supposition is reinforced by the exclusion of children and animals (cf. The speci®c result of justice is stated to be not just friendship. cf. As we saw (section ii. Comm. The parallels are not convincing.). o moÂnoia (which Bignone wrongly alleges to be identi®ed with justice) is called e pisthÂmh (409e9). Cf. on 409d9) and the super¯uous kaiÁ dikaiosu nhv e rgon (409e8. but before we examine them. this. The exclusion of these friendships and the resulting equation of jili a and omoÂnoia appears therefore more or less super¯uous.5.2 D. The addition is ignored in the subsequent discussion: when the friendships of children and animals are excluded on the ground that they are more often harmful than bene®cial (409d6±e1). on 409e4 o moÂnoian).2 parallels.  lhÂqeia. the refutation of this de®nition in the Clitophon itself claims our attention. this means that it is friendship tout court which is the target. The simplest explanation of this disagreement is that they do not come from the same source: the de®nition has all the appearances of being derived from a discussion of justice. is a reductio ad absurdum of Protagoras' theory that di kaion ˆ do xa poÂlewv (500±1). whereas its refutation makes more sense if it was borrowed from a treatment of the nature of friendship. HGPh. 44. the same discrepancy is to be observed in the refutation of the ®rst set of de®nitions. because omoÂnoia is ®rst and foremost a political concept. on 409d7±8). they are not real.). iii 150 n.±K. but reconstructed ones. so that the words jili a e n taiÄ v poÂlesin already refer to little else but omoÂnoia (cf. but friendship in the cities. the fragments on which they are based are not from PeriÁ o monoi av. Comm. but from A 186 .5. Comm. and besides. It seems therefore that the de®nition does not match the ®rst part of its refutation. Bignone thinks. incidentally. 1.I N T R O D U C T I O N II.1 ad ®n. the curtailing of part of the argument for this exclusion (cf.

ei adikoiÄ en allhÂlouv. 4. there is nothing gained for the interpretation of the Republic if we assume that the sentence sta seiv gaÂr pou ktl. praÄ xai an ti du nasqai. was borrowed from the Clitophon or any other source. in fact. 9±11). (351c8±d9).5.16 is a eulogy of concord. It is easily seen that the statement that justice causes concord and friendship in cities and other collectives is not given here as a de®nition of justice or its e rgon. 333 o modoxi a is used to refer to swjrosu nh 433c6.1) is applied. if the criterion stated above (section ii. . There is. again. 442d1. it seems. Therefore. cf.333 though there is an oblique reference in the de®nition of injustice as sta siv (444b1). 187 . ± staÂseiv gaÂr pou w Qrasu mace h ge adiki a kaiÁ mi sh kaiÁ maÂcav e n allhÂloiv pare cei. . . 4. this statement is then applied to injustice in two persons (351e3±4) and ®nally to injustice in one person (351e6±7. h deÁ dikaiosu nh o moÂnoian kaiÁ jili an . a parallel in Republic 1. . .4 is essentially a series of unconnected arguments for the statement that toÁ di kaion is toÁ no mimon.I N T R O D U C T I O N II. ara ei touÄto e rgon adiki av. The concepts of friendship and concord are not used in relation to justice in that book. is that it too is found in a defence of a de®nition of justice.16. The only link with the second de®nition of justice in Clit.4.4. justice and its e¨ect are mentioned only as counterpart to a parallel statement (which is taken for granted) about injustice.2 The association of justice and political friendship/ concord is more or less a commonplace in fourth-century philosophical literature. 352a5±8) ± this addition is meant to anticipate the parallelism of justice in the state and in the soul treated in Book 4. miÄ sov e mpoieiÄ n opou a n e nhÄ i ktl.334 332 Discovered.5.332 in Socrates' refutation of the thesis that injustice is stronger than justice (351a2±352d2): dokeiÄ v a n h poÂlin h strato pedon h lhistaÁv h kle ptav h a llo ti e qnov . 334 A more remote parallel is X. ± ou dhÄ ta h d' o v . 4. Mem. . by Kunert (Necessitudo.

and man and wife. more precisely. Alcibiades is forced to confess to his ignorance. The same questions are asked with regard to the arts of measurement and weighing. ± diaÁ ti n' oun te cnhn o monoouÄ sin ai poÂleiv periÁ a riqmouÂv.5. ± ti deÁ oi i diwÄ tai.I N T R O D U C T I O N II. wives are not loved by their husbands. tiÂv e stin kaiÁ periÁ touÄ. ± o moÂnoian. This ability is the e¨ect of eubouli a (125e6). after which Socrates asks: hn deÁ dhÁ suÁ le geiv omo noian. Alcibiades says that he is talking about the friendship and concord which connect parents and sons. 126c1±3). in so far as they do taÁ autwÄ n (127a14±b4). and vice versa. ou diaÁ th n auth n. Republic 1 and Clitophon. kaiÁ tiÂv authÁn te cnh paraskeua zei. ± diaÁ thÁn ariqmhtikh n. (126d8±9). Similarly. Socrates forces Alcibiades to admit that. eu bouli a regarding the welfare of the city (126a4). friendship and concord play an intricate role in a passage in the Alcibiades 1. o tan jili a meÁ n autoiÄ v gi gnhtai proÁv allhÂlouv. ± nai (126c1±12). Socrates establishes that the a reth he as well as Alcibiades strives after is that of the Athenian kaloi ka gaqoi (124e16). this leads up to the paradoxical conclusion: taÁ di kaia oun pratto ntwn e n thÄi poÂlei twÄn politwÄ n jili a ou k e ggi gnetai proÁv allhÂlouv (127c8±9). As doing taÁ autwÄn equals doing di kaia. which consists of being able to rule in the city (125b9). ± ar' oun jili an le geiv o moÂnoian h dicoÂnoian. since concord between man and wife cannot be about typically masculine or feminine arts. toÁ miseiÄ n deÁ kaiÁ stasia zein apogi gnhtai. ± oukouÄn kaiÁ au toÁv au twÄ i e kastov. Some elements in this highly sophisticated argument are common to Alcibiades 1. ± nai .2 The concepts of justice. Socrates then asks which conditions must be present and which ones absent for a city to fare well: e moiÁ meÁ n dokeiÄ w Sw kratev. while on the other hand they can be well ruled only if friendship is present (cf. cities are not well ruled when everyone does taÁ autwÄn. where Socrates makes Alcibiades admit that he does not know the e pime leia proper to him (124e1±127d8). others have 188 . brothers.

while it is quite beside the point in this discussion. 1. The di¨erence between the two is that in the Clitophon the equation is reached by a process of elimination. while in Alcibiades 1 some sort of answer 335 Chrm. The application of o moÂnoia to the state.5.4. Cf. 5. while other authors tend to associate taÁ autouÄ praÂttein and swjroneiÄ n. 72a4±6. iv 165±7. 161b6. 189 . traditionally taÁ autwÄ n pra ttein is given as a de®nition not of justice but of swjrosu nh. there are certain similarities between Alcibiades 1 and Clitophon which one will look for in vain in the Republic. 136±7. 1) do we ®nd a similar statement about justice. as we saw. the private citizens and the individual occurs. Lys. Ti. 433a8±b1) is obviously a hoax: nowhere else in Plato's dialogues (discounting Alc. In either case. the equation of taÁ autwÄn praÂttein with di kaia pra ttein is typical of the Republic exclusively. North. The statement o ti ge toÁ taÁ au touÄ praÂttein kaiÁ mhÁ polupragmoneiÄ n dikaiosu nh e sti . whereas in Alcibiades 1 Socrates makes Alcibiades choose between o moÂnoia and dicoÂnoia as equivalents of jili a. cf. 1006±7 and Van Leeuwen's note. One might say that the Alcibiades 1 takes for granted a conclusion which is the outcome of a debate in the Clitophon. while in the Republic the words are associated without question.I N T R O D U C T I O N II. Sprachliche Deutung als Triebkraft platonischen und sokratischen Philosophierens (Munich 1959).335 On the other hand. 96±8.2 obviously been taken over from the Republic. Secondly. The equation of jili a and omo noia is presented in both as a separate step in the argument. A second point is the epistemological character which in both dialogues is attributed to o moÂnoia (and by implication to jili a). esp. kaiÁ touÄ to allwn te pollwÄn akhko amen kaiÁ autoiÁ pollaÂkiv ei rh kamen (R.3. in Book 1. Sophrosyne. Ar.32. Nub. although in a dissimilar way: in the Clitophon it is concluded that describing omoÂnoia in terms of knowledge does not clarify the concept at all. note eu kaiÁ paÂlai le getai. 26. this intellectualist conception of o moÂnoia is used as a means to reaching aporia. J. 173±4 n. Gomme on Th. concord within the individual was functional there. 49. 99±101. C. Classen. HGPh.

Finally. Heidel (Pseudo-Platonica. quoted by Gaiser (Protreptik. There is even a slight verbal resemblance: Clitophon 410a4±5 thÁ n deÁ upoÁ souÄ legome nhn dikaiosu nhn h omoÂnoian opoi tei nousa e stin diape jeugen ± Alcibiades 1 126d8±9 hn deÁ dhÁ suÁ le geiv omoÂnoian. The concept of omodoxi a is absent from the latter dialogue. with Susemihl (513±14). what is argued in the Clitophon is tacitly assumed in Alcibiades 1: that omoÂnoia consists of shared knowledge. on 409e5 omodoxi an h e pisthÂmhn.336 The Alcibiades 1 would seem to depend either on the Clitophon or on this source (besides undoubtedly using the Republic). cf. 159) in this connection. it is not very logical to assume. 190 . that the former is dependent on the latter. As I have already stated. whereas the source would seem to be a discussion of the `what is x?' type. 2) and others. 145 n. 292b4±7.2 is given as to the object of o moÂnoia ± it is this answer which eventually causes the aporia. ti v e sti kaiÁ periÁ touÄ ktl. The epistemological character of o moÂnoia is wholly absent from Plato's authentic works and expressly denied by Aristotle. See further section ii. seems irrelevant to me: here a number of possible results of statesmanship (plousi ouv touÁv poli tav pare cein kaiÁ e leuqe rouv kaiÁ astasiaÂstouv) are rejected because these results are not morally good but morally neutral. both Alcibiades 1 and Clitophon treat jili a primarily as a political concept.1(3). Three arguments can be put forward in favour of the former hypothesis (the Alcibiades 1 uses the conclusions of this source in their ± curtailed ± form for which the author of the Clitophon appears responsible. the verbal resemblance pointed out above). the argument in the Clitophon shows traces of having been curtailed from a source lost to us. the association in Aristotle of justice and friend336 Euthd.7. Again.I N T R O D U C T I O N II. Obviously these similarities cannot be a coincidence. Comm. 47 n. Since the Clitophon is much more explicit on both points than the Alcibiades 1.5.

5. they are nearly identical (h tautoÁ n ara h e gguÂv ti. (1) Amiability. tauÂthv deÁ ma list ' e ji entai kaiÁ thÁn sta sin e cqran ousan341 maÂlista e xelauÂnousin (1155a22±26). on 409e4 o moÂnoian). (2) Friendship. EE 1241b11±12. Conversely. kaiÁ oi nomoqe tai maÄ llon periÁ authÁ n spoudaÂzein h thÁn dikaiosuÂnhn´ h gaÁr o moÂnoia omoioÂn ti thÄi jili ai e oiken einai.g. `la discorde. periÁ tautaÁ kaiÁ e n toiÄ v autoiÄ v einai h te jili a kaiÁ toÁ di kaion. . though friendship is not mentioned in discussion of justice in Book 5 of the Nicomachean Ethics (or in the corresponding part of the Magna Moralia).340 in short. which are parallel. . the persons possessing which quality are stated to be the intermediate of areskoi and du skoloi (EN 1126b11±20). at EN 1126b19±20 it is said o noma d' ou k apode dotai au thÄ i. MM 1211a7±8. MM 1211a8±9. which may be taken to express an uncertain feeling).2 ship337 deserves a brief mention. at the same time in EE (1234a24) jili a is ¯atly denied to be an areth . cf. Gauthier±Jolif. These and similar statements indicate a line of thought closely related to the association of justice and concord quoted above (and Comm. son ennemi'. 191 . 341 Often translated as if e cqran were an adjective. cf.. The following passage from the Nicomachean Ethics is especially signi®cant: e oike deÁ kaiÁ taÁ v po leiv sune cein h jili a. e oike deÁ ma lista jili ai. justice is rather prominent in the sections on friendship in all three Ethics.339 There are as many kinds of friendship as there are of justice. EN 1161a10±11. however. Its treatment in EN seems to imply that it is an areth (but cf. e. jili a is clearly `friendship'. 338 EN 1159b25±6: e oike de . 1234b30±1). While these persons are called ji lov without reserve at EN 1108a27. 339 EN 1160a7±8. 340 EE 1241b15. 1105b22). 1127a14± 17. 337 Aristotle uses jili a to denote two di¨erent things. In EE (1233b29±34) and MM (1193a20±27). while in MM (1193a37±8) it is doubted. which di¨ers from (1) through the addition of ste rgein (EN 1126b22±3): it is therefore a paÂqov (ibid.I N T R O D U C T I O N II. Whether or not jili a in this sense is an areth is left open in EN (1155a3±4 e sti gaÁr areth tiv h met' arethÄv). as the Eudemian Ethics puts it. They are said to have the same subjects and objects338 and therefore to increase at the same time.

EE 1242a6±7. 11±12. d9±e1) and the relation of o moÂnoia and omodoxi a (cf. In analysing the relation of friendship and justice Aristotle comes close to the utilitarian conception of friendship as found. 192 .I N T R O D U C T I O N II. Gauthier±Jolif ii 2. Comm. the rejection cannot possibly be used for determining the philosophical outlook of the author of the Clitophon.343 This analysis is therefore traditional. for instance. His suspicion that the de®nition of the result of justice as friendship in cities originated from Aristotle ± he quotes (435 n. 1) Pol. cit. 696. if Clitophon's adversary rejects the identity of omo noia and o modoxi a.345 He appears to have overlooked the fact that Aristotle uses o modoxi a in a di¨erent sense from the Clitophon. this word is used by him to denote any group based on the achievement of a koinoÁ n sumje ron.342 Signi®cantly. Cf. the exclusion of the friendship of children and animals (cf. . 342 343 344 345 Cf. Mem.344 Consequently we must not expect it to shed special light on the Clitophon. I fail to see how these slight resemblances warrant Ge¨cken's conclusion that the author of the Clitophon was `Aristotelisch denkend' (434±5). it has even been called `archaic'. h e pisthÂmhn). 334. the notion of koinwni a is very prominent in his discussion of friendship related to justice. Gauthier±Jolif. for example. as does Aristotle. EN 1160a10. on 409e5 o modoxiÂan .2 At the same time. loc. the one which is based on the good. but about friendship based on mutual interest. Some points in Clitophon's refutation of the de®nition can be connected with elements of Aristotle's ethical theories. cf. 1162b21±3.5. nor is there any reason to assume in¯uence of this dialogue on Aristotle. even so.2±10. 2. 657. Comm. it must be borne in mind that Aristotle is here speaking not about the prwÂth jili a. on 409d7±8.. 1262b7± 8 (where justice is not named) and could have quoted with more point EN 1155a22±6 ± is the more unfounded as Aristotle's thoughts on this point are wholly traditional. . in Xenophon. n.

o no mov and toÁ di kaion] ta v te po leiv kaiÁ touÁ v anqrwÂpouv toÁ sunoiki zon kaiÁ toÁ sune con. 7. it could be argued that cumulative evidence makes it plausible (if Plato wrote the Clitophon there is no problem).I N T R O D U C T I O N II. 175. Them. 322c3. where ai dw v and di kh are called poÂlewn koÂsmoi te kaiÁ desmoiÁ jili av sunagwgoi . HGPh. there is no cogent reason to assume that this de®nition was taken over from Republic 1. SSR i c 489. in fact it could very well be sophistic in origin346). 1091 N2 .347 Even in the fourth century this maxim is a solidly embedded rule of conduct: one of the defendants in the Corpus Lysiacum (9. Socrates. See n. At most. 312±13.3 To bene®t friends and harm enemies There can be no reasonable doubt that Plato was the ®rst Greek writer to attack the traditional Greek idea that it is right and just to bene®t one's friends and harm one's enemies. 331 to the previous section. Or. Cf. SSR i c 490). For an analysis of the position of this rule within the wider ®eld of popular morality see M.3 Given the traditional character of the association of justice and political friendship (it is also found in the myth of the Protagoras. 347 Whether he or Socrates was the ®rst thinker to do so will not be discussed here ± the claim that Antiphon preceded Plato has been disposed of by Guthrie. 218a. I have already argued that the refutation was separately borrowed from a work lost to us. 34. iii 112±13. Anon. iii 149. ± Professor Sedley draws my attention to a number of anecdotes in which Socrates amends the maxim (Plu. Or.5. E. 95ab. ii. ii 230.6 touÄ to gaÁ r [sc. they may well have been inspired by Plato. Cf.20) states as his opinion that it is `prescribed' (teta cqai ).6 Downey±Norman.348 346 It is ascribed with no questions asked to Protagoras by Guthrie. Whitlock Blundell. 3. Vlastos. SSR i c 488. 179±99 and 297±300. Iambl. but the ascription to Socrates is probably an interpolation.5. 348 Fr. Supp. much as a speaker in an unknown Euripidean drama had called it a noÂmov. HGPh.10±231. Helping Friends and Harming Enemies: A Study in Sophocles and Greek Ethics 193 . There is of course no telling how old they are.

W.351 although he also says of Socrates di kaiov deÁ wste bla ptein meÁ n mhdeÁ mikroÁn mhde na. who states this rule of life quite often. Greek Popular Morality in the Time of Plato and Aristotle (Oxford 1974). 23. 184). K. cf. Op.I N T R O D U C T I O N II. op. 9.. 180±4. Cf. 1.1. Since MuÈ ller has shown that X. so e gkrath v that he never preferred the agreeable to the good. (n.75). cit. this is re¯ected in a saying. 3.8. Kurzdialoge. add Archil. F. not from Aeschines' Alcibiades (134±9). 1367a20±2 (cf.14 W. 2. An.14. of which an alternative version. Fi lov und jili a im vorhellenistischen Griechentum (Munich 1931).2. 184. From texts like Top. 5. Xenophon momentarily adopts a moral standard which is not his own (in fact Platonic in¯uence seems unmistakable here). Cyr.25. Mem.350 and Xenophon.31. the unfortu- 349 350 351 352 (Cambridge 1989). 4. suggnwÂmh metanoi av krei sswn. cit. 374c4±d3.6±7 V. fr. Dirlmeier.3 At least two Socratics felt no compunction in adopting this rule of life:349 the anonymous author of the dialogue of which PeriÁ dikai ou is an extract. 4. 1363a20±1) it cannot be deduced that Aristotle subscribed to it as well. 113a2±5 and Rhet. 26±59. 28. ascribed to Pittacus (D.3. so just that he did no one any harm (not even a small one!). Sapph. Of course the Greeks recognised as well as we do that in many cases clemency even towards enemies is the best policy (Dover. 6. and for a repertory of places from older Greek literature. Mem. 1. so jroÂnimov that he never made any mistake in judging between good and evil. 1. L. the statement in 4.35. The sentence must be read in its context. 348). puts one at a great advantage over an adversary'.11.. Hiero 2.2. In order to make Socrates as saintly as possible. Dover. 167±8.11). wjeleiÄ n deÁ taÁ me gista touÁv crwme nouv autwÄi (Mem. 194 . J. the ®nal eulogy of Socrates which closes the Memorabilia: Socrates was so pious that he never did anything without the gods' consent.) seems to me to re¯ect a much later attitude.5.12±9 is derived from this dialogue. and to be seen to ignore it. 6. fr. suggnwÂmh timwri av krei sswn (ibid. C.28 (`Prodicus'). MuÈ ller. Dover352 cannot be right in connecting this with the notion that `on occasion to ignore a wrong. can no longer be attributed to Aeschines.

e.5. feels that he stands alone in his conviction. but it is too far removed from Clitophon's formulation to be regarded as its source. touÂtoiv ouk e sti koinhÁ boulhÂ. b2.117±18: duoiÄ n gaÁr pragmaÂtoin proteinome noin mhÁ spoudai oin. A passage from Isocrates probably is a direct attack on Plato's rejection of the rule.g. oli goi d' an tinev twÄn prospoioume nwn einai sojwÄn e rwthqe ntev ouk an jhÂsaien.I N T R O D U C T I O N II. The opposition adikeiÄ n : a dikeiÄ sqai is of course a major theme of the Gorgias.353 There are two passages in Plato in which this traditional rule of conduct is explicitly rejected. cf. a6±7) and that the contrary opinion is the one usually held (d2 oida gaÁ r o ti o li goiv tisiÁ tauÄ ta kaiÁ dokeiÄ kaiÁ doÂxei. 355 The future may be interpreted as an indication that Plato. it is usually considered to re¯ect the Gorgias: 12. 354 In a number of passages the rejection is implicit. oud' a n o tiouÄ n paÂschi up' au twÄ n). 51±2.354 In the introduction to the prosopopoiia of the laws in the Crito. It is repeatedly stressed that this has always been Socrates' opinion (49e1±2 e moiÁ meÁ n gaÁr kaiÁ paÂlai outw kaiÁ nuÄn e ti dokeiÄ . a ll' a naÂgkh tou touv allhÂlwn katajroneiÄ n orwÄ ntav allhÂlwn taÁ bouleuÂmata). Grg. C. 353 Cf. krei ttw thÁn ai resin einai touÄ deinaÁ poieiÄ n e te rouv h paÂscein autouÁv kaiÁ touÄ mhÁ dikai wv twÄn a llwn arcein maÄllon h jeuÂgontev thÁn ai ti an tau thn adi kwv Lakedaimoni oiv douleuÂein. Eucken. The gap between Socrates and normal Greek opinion is unbridgeable (d3±5 oiv oun ou tw de doktai kaiÁ oiv mhÂ. aper apantev meÁ n an oi nouÄ n e contev e lointo kaiÁ boulhqeiÄ en. `Leitende Gedanken im isokrateischen Panathenaikos'. writing after Socrates' death. Socrates states with great emphasis that to do injustice is always evil and that therefore to retaliate against injustice is evil too (49c10±11 oute ara antadikeiÄ n deiÄ oute kakwÄ v poieiÄ n oude na anqrwÂpwn. esp. 195 . MH 39 (1982) 43±70.3 nate addition of mhdeÁ mikroÂn shows that this line of thought is unfamiliar to him. 508d6±e6. b9±10).355 cf.

The de®nition is refuted twice. for the taste of many Platonists. toÁ n d' e cqroÁ n kakoÁn onta blaÂptein) the second refutation sets in: (1) harming a horse or a dog means making them worse in their particular arethÂ. likewise the just cannot make men unjust by justice (c9±d2). likewise (since justice is a (or the) human a reth ) harming a man means making him more unjust (335b6±c8). cf. On the other hand. it is not his function to harm anyone but that of the unjust man (d3±13). 334b8± 9). After a slight modi®cation (335a9±10 toÁn meÁ n ji lon a gaqoÁn o nta eu poieiÄ n. and since the just man is good. `to treat badly'. likewise it is not the function of the good man to harm but of his opposite. 331e3±4).5. I shall not go into the questionable analogies and the ambiguous use of bla ptein (1. it is in fact a long and emotional statement of opinion. blaÂptein deÁ touÁv ecqrou v. Polemarchus is induced to `interpret' this as a de®nition of justice. while at the same time clinging to his de®nition (touÄto me ntoi e moige dokeiÄ e ti. kakwÄv 196 .3 Though formally this passage has the normal dialectical structure of question and answer. there is rather too much argument. Starting from a dictum of Simonides that `giving everyone his due' is just (or the just. ± dokeiÄ moi (332d7±9). The premise on which this opinion is based appears unobtrusively at the end of the opening sentence of the passage: to ge adikeiÄ n twÄi a dikouÄnti kaiÁ kakoÁn kaiÁ ai scroÁn tugca nei on pantiÁ troÂpwi (49b4±5). (2) musicians or skilled horsemen cannot make men unmusical or bad riders by music or the art of riding. in the discussion of Socrates and Polemarchus in Republic 1. which leads up to the absurdity that the just man is a thief (334a10). toÁ touÁv ji louv a ra eu poieiÄ n kaiÁ touÁ v e cqrouÁv kakwÄv dikaiosuÂnhn le gei. (3) it is not the function (e rgon) of heat to chill or of dryness to moisten but of their opposites. because Polemarchus refuses to accept it.I N T R O D U C T I O N II. we need not go into the ®rst refutation. wjeleiÄ n meÁ n touÁv ji louv h dikaiosuÂnh.

that the de®nition was rejected by any other Socratic but Plato.I N T R O D U C T I O N II. 62±8. This question is not answered because Thrasymachus breaks in).14±16 PMG is quoted: Plato would have emphatically denied that this is necessarily true. 197 . 65±6. not as a positive statement as in the Crito. H. ti an allo tiv autoÁ jai h einai.2). 542. Formen der BeweisfuÈ hrung in den platonischen FruÈhdialogen (Bonn 1975). 356 The equivocation is less obvious because in the case of horses and dogs the ®rst meaning goes hand in hand with the second (so the application to humans may seem a correct induction) and also because of the dubious identi®cation of justice and human arethÂ. the two meanings are very e¨ectively employed in the Apolog y (30c8±d6). and it is found stated explicitly in the Clitophon: 410b1±3 usteron deÁ e ja nh bla ptein ge oude pote o diÂkaiov oude na´ pa nta gaÁr ep' w jeli ai paÂntav draÄ n (see section ii. 335c) that men grow worse when they are hurt or made miserable. On the following grounds I consider it proven that the third de®nition of justice in the Clitophon and its rejection were derived from Republic 1: (a) It is unlikely. and besides the ambiguity points a moral which is relevant for the Republic as a whole: true evil for man is to be damaged in his soul (cf.3.5 on this function of fallacies). esp.4. `to make worse'. Cf. 357 I disagree with Tucker's explanation (note on R. That much is suggested by the addition of the super¯uous third stage of the argument (335d3±13). section ii.3 poieiÄ n. for which Simon. cf. 2. then it could be argued that the just man's actions are always bene®cial. as far as our knowledge goes.5. although a positive result might have been reached easily: if it is the function of the unjust man to harm.357 For our purpose it is more relevant to note that for artistic reasons the whole episode is presented as a discussion with a negative result (336a9±10 eien h n d' e gw ´ e peidhÁ deÁ oudeÁ touÄ to e jaÂnh h dikaiosu nh o n oudeÁ toÁ di kaion. Telle. on 407a1 w jeleiÄ n). Comm.356 I am convinced that there is a fallacy and that it was intended.

4 some phrases in the Crito seem to indicate this. this tallies with the situation of Republic 1 but not with the Crito passage. If derived from Plato's works. kakwÄv poieiÄ n anqrwÂpouv). judication.3) omits to report the argument following the de®nition. That Pavlu declared that this 198 . it must come either from Republic 1 or Crito.I N T R O D U C T I O N II. (d) The Crito does not speak of bene®ting friends or harming enemies.5. whereas the opposition is found both in the Clitophon and in Republic 1. whereas in the Crito a di¨erent ground is given (unjust action is evil for the man who commits the action). kakourgeiÄ n. or of either on a common source.5. but for the bare essential statement paÂnta gaÁr e p' wjeli ai pa ntav draÄn (410b2±3).4 Politics. (c) The reason given for the rejection in the Clitophon (the just man always acts to the bene®t of everyone) is implied (and its negative counterpart stated) in Republic 1. The verbs bla ptein and w jeleiÄ n are absent from the Crito passage. (b) The rejection is presented in the Clitophon as the negative outcome of the discussion of a de®nition.5. justice About the words hn dhÁ suÁ politikhÁ n w Sw kratev e ponomaÂzeiv polla kiv. ii. antadikeiÄ n. thÁ n au thÁn dhÁ tauÂthn dikastikh n te kaiÁ dikaiosu nhn wv e stin le gwn (408b4±5) much unnecessary confusion has been created. a ntikakourgeiÄ n. Even though the author of the Clitophon (for reasons which have been discussed in section i. It is frivolous to discuss the possibility of Republic 1 depending here on the Clitophon. this statement su½ces to establish that in the case of the third de®nition the refutation comes from the same source as the de®nition itself. but of acting unjustly or wrongly per se (adikeiÄ n.

). 520b2±3. Kleitophon'. Surely. if not from this passage alone. What is actually said in the Gorgias is that politikh is about the eu exi a of the soul (464a1±b4).358 is due to misunderstanding and lack of interpretative precision. one of them is called dikaiosu nh and dikastikhÂ. politikh is said to be identical to dikastikh and dikaiosuÂnh. in the Gorgias. there are two moÂria of politikhÂ. as BruÈ nnecke says (459 n. this must mean that dikaiosu nh at 464b8 is the corrective part of the politician's art.2. it is in any case the correct interpretation (Dodds on 464b8 could have strengthened his case for dikaiosu nh as the original reading by pointing at 478a5±b1. not mentioned by BruÈ nnecke. Socrates proceeds to distinguish two parts of the care of the body. 465c3.4 identi®cation was not to be found in the authentic works of Plato (`Pseudopl. Nor can I believe that this identi®cation has anything to do with the right of the stronger (BruÈ nnecke and Souilhe . considering his super®cial knowledge of Plato. 199 .e.5.2. kaÂllioÂn e stin) kaiÁ gumnastikhÁ i atrikhÄ v. i. 1). That this identi®cation of judication and justice was denied by Plato in Gorgias 464. section ii. In the Clitophon. loc. cit. gumnastikh and i atrikh (b4±7) and then he says thÄv deÁ politikhÄ v antiÁ meÁ n thÄ v gumnastikhÄ v thÁn nomoqetikhÂn. a nti strojon deÁ thÄi i atrikhÄ i thÁn dikaiosuÂnhn (b7±8). 2) the sequence eyes ± ears ± whole body which is actually taken over from Plato (cf.I N T R O D U C T I O N II. 359 Besides.3). 29) and Souilhe repeats (186 n. which makes him elsewhere condemn as `laÈppische Beispiele' (11 n. 13) is not surprising. from 520b3 o swiper nomoqetikhÁ dikastikhÄ v (sc. It is out of the question that the Clitophon 358 He seems to have been misled by Grg. in the same way as i atrikh is the corrective part of the science of bodily welfare. one branch of the direct tradition and a number of witnesses of the indirect tradition (see Dodds' apparatus) read dikastikhÂ(n) instead of dikaiosu nh(n) at 464b8.359 as appears quite clearly. judication. c2. where both di kh and dikaiosu nh are used for dikastikhÂ). If this is not the correct reading.

(3) A philosopher should be the best man in all separate arts which constitute this complex. Socrates there proves (1) that justice is the same thing as swjrosuÂnh: the art of punishment is identical to the art of improving others (137c6). that the author of the Clitophon took his identi®cation of politics and judication over from the Gorgias. In an attempt at a de®nition of philosophy. 1 (but di¨erences between Xenophon's and Plato's conceptions of dialectic do not disprove dependence). the separation is foreshadowed by Gorgias 520b3 quoted above. or simply knowing oneself. identical to dikaiosuÂnh (137d14.6. therefore philosophy is identical 360 Cf. iii 440 n. Applied to human beings.38. We can go a little further than that. 59±61. Mem. which is swjrosuÂnh (138a1±8). along the lines of Plt. That Xenophon knew the Plt.I N T R O D U C T I O N II. judication is an ancillary (u phre tiv) of the basilikhÁ te cnh in the Politicus (305b1±c8). therefore.5. 231.4 exhibits traces of an un-Platonic doctrine in this respect.360 It would seem. and brought it in as a reference to the educative role of the politician in that dialogue.. cf. X. F. where a clear distinction. no reason is given). the art of discerning between good and bad men implies knowing if one is oneself good or bad. The whole of this interpretation must now be set against a totally di¨erent construction put upon the Clitophon passage ever since C. basilikh and turannikh (138b15±7). Sokrates. the art of punishment is dikastikh (137d12). is made between judges and politicians. Maier. 361 Geschichte und System der platonischen Philosophie (Heidelberg 1838). 200 . 2. In his later dialogues Plato does indeed separate dikastikh from politikhÂ. 138b7±10). this art is also the one which distinguishes between good and evil subjects (137c9±11). (2) This dikaiosu nh/swjrosu nh is next identi®ed with politikh (cities are well governed if there is right punishment. contra HGPh. Hermann:361 it is explained as a borrowing from the pseudo-Platonic Erastai. was argued by H. i 426 n.

(b) when it does occur. 13) that the Clitophon cannot possibly be the source of the Erastai. 362 Guil.363 It follows that either the Clitophon depends on the Erastai or they borrowed the identi®cation of dikastikh and dikaiosuÂnh from the Gorgias. Werner. 58. despotikh and oikonomikh in the Politicus (258e8±11). Now. au th TW) to h authÁ deÁ (B). whose author found the inspiration of his identi®cation of politikhÂ.3) has to be given up: to be sure. SouilheÂ. op.). h e te ra tiv. This would probably imply that my interpretation of h twÄ n anqrwÂpwn kubernhtikh as the educative role of the politician (found in Gorgias. authÁ 364 kaiÁ gignwÂskei touÁv crhstouÁ v kaiÁ touÁv mocqhrouÂv. 51. and the author of the Clitophon may have decided to use this particular representative of the genre to end Socrates' protreptic. the argument is rather confused in detail). Kleitophon'.I N T R O D U C T I O N II. Euthydemus. Politicus. 3. 364 I prefer au th (Schanz. and perhaps also from 476d8±478b1. mention is made of belti stouv poieiÄ n in the Erastai (137c1±d1) but this seems to have no connection with either dikaiosu nh/dikastikh or politikhÂ. it serves only as a link between right punishment and the ability to distinguish between the good and the bad (the argument hinges on 137c9±11 poÂteron h per belti stouv te poieiÄ kaiÁ kolaÂzei orqwÄv. au thÁ deÁ (D). section ii. because (a) apodotic de after a relative clause is rare. 363 Werner. 201 . cf. Pavlu is obviously right when he says (`Pseudopl. 123 n.2. the Erastai is certainly a sort of protreptic dialogue which defends philosophy against the claims of the liberal arts (though it is perhaps better to say that it tries to de®ne the exact place of philosophy with regard to the liberal arts). basilikh . it always follows a demonstrative (GP 2 178). 2.5. 122 n.4 to the complex of arts established in (2) (138d1±139a5.. cit. De Anterastis dialogo Pseudoplatonico (Darmstadt 1912).362 the identi®cation of dikastikh and dikaiosuÂnh probably derived from the above-quoted passages from the Gorgias.3.

in combination with politikh . the latter would not do in the context of the Clitophon. Obviously.5. the other nomoqetikhÂ. 138c9±10).4 Though this consideration does not constitute a real objection. It is a far cry from this to politikh as h twÄn anqrwÂpwn kubernhtikhÂ. Consequently. a second cardinal virtue would have diverted the attention from it). .I N T R O D U C T I O N II. not even in the long series of identical te cnai. all the others were retained and made good sense. would at the end of the whole speech have harmonised quite nicely with the end of the part in oratio recta paÂnt' andra idi ai q' ama kaiÁ dhmosi ai sumpa sav taÁv poÂleiv (407e1±2). perhaps the following observations do. politikhÂ. One can understand why swjrosuÂnh was not named (the whole of the Clitophon is concerned with dikaiosuÂnh. one of them dikaiosu nh/dikastikhÂ. 464b5±6) is said to have two parts. point is the fact that the author of the Clitophon further on goes out of his way to propose the distinction between e rgon and di dagma 202 . the hypothesis that our author selected politikhÂ. Another consideration may be the question why the author of the Clitophon selected precisely the triad politikhÂ. The omission of basilikh and turannikh may perhaps also be accounted for. kaiÁ ai poÂleiv eu oikouÄ ntai otan oi a dikouÄntev di khn didwÄsin ± . when either of them. . Here. On the other hand. dikastikh and dikaiosu nh from Erastai. esp. cf. it is just identi®ed with the moral cluster dikaiosuÂnh/swjrosuÂnh (once dikastikh has been identi®ed with dikaiosuÂnh at 137c14 it does not turn up again. is the art of right punishment in Erastai. the qerapei a thÄv yuchÄv (cf. politikh is not directly connected with dikastikh in Erastai as it is in Gorgias and Clitophon. A third. inasmuch as it is identical to dikaiosuÂnh. and to my mind decisive. dikastikh and dikaiosuÂnh from the Gorgias makes much more sense. In the ®rst place. ± kaiÁ politikhÁ ara au th e stin (138b7±10). politikhÂ. but one wonders why oikonomikh and despotikh were left out.

these are hardly parallels. further section ii. itself can hardly be prior to 330.7. cit. op. which latter is usually dated (by those who do not accept it as Platonic) around 340 ± section ii.4. SouilheÂ. 137e4±138a3. and it could hardly be dated before c. little to connect the Erastai with the protreptic corpus: the gnwÄ qi sauto n motif is the only major similarity.366 the opposition jilogumnasti a : jilosoji a (133e3±5) is only indirectly transposed into the opposition body : soul (134d4) ± no rejection of care for the body here! This parallel between Clitophon and Erastai is evidently of great importance for the origin of the Clitophon. then Amat. 22±3.365 There is. Apart from the parallel under discussion. previous note) in deriving Amat. from the Alc. 366 Werner. and Chrm. apart from the parallels from Plt. discussed. (24). as an Academic attack against Peripatetic polumaqi a. 407c1 ± Amat. He interprets Amat. The date of the Erastai is highly uncertain. This doctrine is absent from the Erastai.1 n. the two dialogues have hardly anything in common. 203 .4 (which complicates his use of the te cnh analogy). This procedure is understandable if we suppose that the doctrine of the politician±physician of the Gorgias and other dialogues was retained by our author as a piece of `constructive' philosophy (cf. 367 Werner detects in¯uence of Alc. If the Clitophon depends on the Erastai.367 the work is almost unanimously rejected (cf. 107±10). (n. it is impossible to maintain the former's authenticity. 1.5.2). 132d1±2. 408a1±4 ± Amat. 330. besides. in Plato's last period). The only purpose of this distinction is that it establishes the point that the just politician makes his fellow-citizens more just (409b6±7 thÄv dhÁ dikaiosuÂnhv wsauÂtwv toÁ meÁ n dikai ouv e stw poieiÄ n).368 As I can make more sense of the Clitophon if I do not take the Erastai into account (especially because of 365 Clit. but precisely that has evidently been taken over from the Charmides. 1 (27) and Phdr. 368 If Werner is right (cf. which ± as is usually the case with Platonic Dubia and Spuria ± is a mere guess. 395 (and by some of those who do accept it as authentic.I N T R O D U C T I O N II. Clit. 362).

Oldenberg370 and adopted among others by Kunert (Necessitudo. 370 De sacris fratrum Arvalium quaestiones (Berlin 1875). Obviously there is a close relationship between these two works. but not published until he decided to incorporate it (whether or not with adaptations) into the whole of the Republic. This hypothesis. it achieves nothing. 18±22).) Wilamowitz. a question must be raised to which the contemporary reader knew the answer whereas we do not: was Republic 1 ever published as a separate dialogue?369 If it was.5 Conclusions One of the three de®nitions in the Clitophon can be proved to derive from Republic 1. I shall disregard the parallel. (e. the Clitophon contains a clear message: when Socratic literature tries to go beyond mere protreptic. witness the `Thrasymachus' (or whatever name may be given to Republic 1 as an isolated dialogue). another is most likely to derive from it. the third has at least a good parallel in a passage from the same book.I N T R O D U C T I O N II. ii. 369 I leave out of account the possibility ± which is far from improbable ± that Republic 1 was written separately by Plato. scriptus est Clitopho dialogus pseudoplatonicus. was ®rst put forward by H. quo edito priusquam ceteri libri emissi sunt. Platon. A polemical text like the Clitophon could not possibly deal with an unpublished sketch. 53: `Platonem primum de republica librum separatim edidisse censeo. When we take into account the choice of the character Clitophon and of Thrasymachus as his teacher. Grube (`The Cleitophon of Plato'. which automatically dates the Clitophon between the publication of Republic 1 and that of Republic 2±10.' 204 . Before we can proceed to reconstruct it ourselves.g.5. 11. cf. we can hardly evade the conclusion that the readers of the Clitophon were meant to grasp the author's intention in the light of the Republic. ii 184.5 the third argument adduced above).5.

147 n. 373 The message is slightly di¨erent if Republic 1 is itself considered to be a `protreptic' dialogue. Before we start doing so. CQ 45 (1995) 58±67.374 we may con®ne ourselves to working out the consequences which this assumption has for the interpretation of the Clitophon. so Grube (`The Cleitophon of Plato'.5. Wilson. but serves admirably as a pre®guration of the just man as described in the later books). section ii. Gaiser (Protreptik. R.5 305). 287. with some reservations. pp. 231. J. two out of three de®nitions which the Clitophon would.2.4 n. 138±9. H. Kahn. on this hypothesis. even if a `Thrasymachus' was 371 Platon. `Proleptic composition in the Republic.371 Julia Annas372 and.3). S. 372 An Introduction to Plato's Republic (Oxford 1981). 17. 126±8.373 It is not necessary for us to decide whether the assumption of a separately published `Thrasymachus' is tenable or not. 143). `Thrasymachus and the thumos: a further case of prolepsis in Republic i '. or why book 1 was never a separate dialogue'. The debate continues.5.2)375 and the argument that no just man harms anyone is constructed in such a way that for the bon entendeur the true meaning of `to harm' (which plays such a prominent part in the proof of the immortality of the soul in Book 10) could not be missed (section ii. the introductory conversation between Socrates and Cephalus makes little sense in a separate dialogue. C. Kahn as quoted in the previous note. CQ 43 (1993) 131±42. ii 45±6.I N T R O D U C T I O N II. 374 Cf. No matter how one thinks about the degree of identity of the hypothetical `Thrasymachus' and Republic 1 as we have it (for example. FriedlaÈ nder.5. 375 Cf. have borrowed from the `Thrasymachus' are derived from contexts where there is a rather clear reference to views expounded in the later books of the Republic: the parallelism of justice and injustice in the state and in the soul is stressed in the passage from which the second de®nition in the Clitophon derives (section ii. Cf. one point must be stressed. 305) and Gaiser (Protreptik. Therefore. 162). 205 .

the Clitophon was written by someone other than Plato. would show himself as critical of this dialogue as he is of explicit protreptic. Of course the hypothesis of a separately published `Thrasymachus' is itself 206 . which would remain extremely speculative in any case. one may well ask why he chose a dialogue of his own. (Of course.2).5 really published separately. one cannot maintain that Plato `felt the need for criticism of his work in general and [Republic 1] in particular' (Grube. the author. in order to make the point that Socratic literature does not teach anything about justice.4) and with his implicit subscription to some Platonic tenets. we had better abandon the possibility. in picking all his de®nitions from the putative `Thrasymachus'. Protreptik.2)? It is clear that dating the Clitophon between Republic 1 and the whole of the Republic brings nothing but confusion. How then. 306. do we reconcile this attitude with his evaluation of Plato's implicit protreptic. 146±7).5. as we saw (section ii.4. it was at any rate a dialogue in which a good deal of what Plato later said about justice in the Republic could already be read between the lines ( justice is a harmonic state of the soul. If it was Plato himself. to commit injustice is to disrupt this harmony). and if the Clitophon was dealing with it. and one which contained pretty much of his theory of justice. the ®rst part of the Clitophon does not criticise Plato's work at all. If.) Certainly. because. one of which is found in the `Thrasymachus' itself (section ii. this question remains open if Plato wrote the Clitophon after the whole of Republic. `The Cleitophon of Plato'. it will be treated below. on the other hand. prior to the publication of the whole of the Republic. Now there are two possibilities: either Plato himself wrote the Clitophon between the publication of the `Thrasymachus' and Republic or someone else did. which he adopts (section ii.I N T R O D U C T I O N II. similarly Gaiser.

Birt.3. cf. a 1 DuÈring). 378 Obviously. Before we start interpreting the relationship of Clitophon and Republic. if a text makes extensive use of (part of ) the Republic and of characters found there. Das antike Buchwesen (Berlin 1882). ± Cicero's statement quoniam in singulis libris utor prohoemiis ut Aristoteles in iis quos e xwterikouÁ v uocat (Att. 472). 207 .3 ± this cannot apply to a separate `Thrasymachus'). L.g. it is out of the question that the second half of the Clitophon is meant as an attack on the Republic. Once this is realised. But there is no indication whatsoever that for any author of the classical period the physical end of a roll had to coincide with a major transition in the text. 4.377 At any rate. 377 Alline. not just because of indications in the text of the Republic (e. the contemporaneous reader is invited to interpret this text in the light of the Republic as a whole. 3. the whole of the Republic is too long to be contained in one roll.2) probably means no more than that Aristotle introduced some of his exoteric works with dedicatory prooemia (we know this for a fact in the case of Protr. not of a particular book of the Republic. it is not until the ®rst century ce that we hear of a division into ten books. each with a separate prooemium (so wrongly Th. Histoire.5. However. because there were none. it is certain that Plato himself did not divide the Republic into books. We are accustomed to think of Book 1 of the Republic as a selfcontained entity. and an anecdote told by Gellius perhaps suggests that the Republic was published in instalments (14.5 not a¨ected by the consequences (or rather absence of them) which it has for the Clitophon. 357a1±2 e gwÁ meÁ n oun tauÄ ta eipwÁn wimhn loÂgou aphllaÂcqai´ toÁ d' hn a ra wv e oike prooi mion) but ®rst and foremost because in our copies of the Republic it is printed as such. unless one resorts with Wilamowitz to the slightly 376 Thrasyllus apud D. one further point must be borne in mind. it does not necessarily mean that Aristotle divided his exoteric works into books..57.378 Consequently.I N T R O D U C T I O N II.376 another one into six books was perhaps used by Aristophanes of Byzantium.16. 15±17.

1) ± we must apparently suppose that he wrote for a public as lazy as himself. i 386 n.4.1) and how could he have acquired that knowledge unless it was (and this is the minimum requirement) by reading the Republic. ``Truth'' on a plate' (Art of Plato. an answer. If there is any dialogue that shows that for Plato there is no such thing as truth on a plate. then. `Socratic teaching'. What. den Staat durchzulesen' (Platon. it is the Republic.I N T R O D U C T I O N II. 208 . he also knows Plato's conception of justice (section ii.5 absurd view that `der Verfasser hat sich die MuÈhe nicht gemacht. This might seem a paradox: the de®nitions which Clitophon turns down are borrowed from the same dialogue which (as I think) for our author constitutes a positive contribution to the question what justice is. Blits.379 As I have stressed repeatedly. our author could borrow the de®nitions. In other words. cf. our author tried to indicate that his criticism is not aimed at Plato. our author understands quite well what Plato's dialogues are all about. Yet the examination of the contexts of these de®nitions in Republic 1 has shown that they are not presented as serious answers there. 100. precisely because the argument in the ®rst episode of the Republic is presented as provisional. did he wish to communicate? The best answer I can think of is that by choosing his de®nitions from the Republic. and the Gorgias besides? If indeed the author wished to communicate the thought that Socratic exhortation to justice does not lead to knowledge of justice.5. the clumsiest thing which he could possibly have done is what he actually does: to direct his readers' attention to the one Socratic dialogue in which there is a very extensive and positive answer to the question what justice is and what its e¨ects are. in which dialogue a well-reasoned de®nition of justice is eventually given. 332 `an answer he can memorize'). since his public was bound to know that 379 A similar objection applies to Rutherford's far more sophisticated interpretation of Clitophon as someone who `wants a solution.

as it were. yet when Clitophon asks them to name the `art of the soul's perfection'. he chose to concentrate on one work (a procedure which entailed the di½cult task of looking for suitable refutations elsewhere). it takes the most intelligent of them (409a4) to answer the question. but although Clitophon says that these speeches `wake us up. Instead. Yet on one point it is unmistakably correct: the ®gure of Socrates in the dialogue is meant to symbolise Plato. and Clitophon goes out of his way to underline this (408e2). many scholars have tried to match the various pupils in Clitophon's report with di¨erent Socratics. The speech reported by Clitophon has been an exhortation to justice.2).6 eventually Plato could and did give an answer to Clitophon's question. In the past. Platonism was the only acceptable form of Socratic philosophy and the Platonic dialogue the only truly Socratic dialogue.I N T R O D U C T I O N II. It would have been easy for the author to prove the inadequacy of explicit protreptic by picking out various definitions of justice together with their refutations from all sorts of Socratic literature. the reader of Socrates' speech will have had no trouble in recognising it as a farrago of patterns and motifs from protreptic literature. I do not believe this to be a fruitful approach.2) and in the third de®nition (section ii. he is meant to take the dumb209 . ii. both in the prelude (section i. from our sleep' (408c3±4).6 The meaning of the Clitophon Socrates' companions have listened to his protreptic speeches.5.4. For the author of the Clitophon. in order to make clear again that he was in no respect critical of Plato. as Clitophon has. Now. the pupils show no sign of having been aroused. Subsequent interrogation proves that their ideas about the result of justice are little more than vague slogans.

Simultaneously. to wit. by Socrates' own inability to give a correct answer. equation of virtue and te cnh. it suggests to the reader that there is another method. Yet the explicit character of protreptic entails its lack of success. such as those of Thrasymachus. making its readers better men by explicitly exhorting them to virtue. namely elenchos as practised in Plato's aporetic dialogues. because he chose to make Socrates a symbol of protreptic and therefore could not make him stand for elenchos at the same time. teach. d4) crumble before a few simple questions. The characteristics of the aporetic dialogue are all present: analogy.3 ± is correct). Clitophon's interrogation of the companions serves a double purpose. as protreptic does not. which can lead to virtue.6 ness of Socrates' companions as a criticism of this genre: it does not achieve its aim. The author of the Clitophon could not contrast protreptic literature and Socratic elenchos as clearly and explicitly as eristic and Socratic paidei a are contrasted in the Euthydemus. The author of the Clitophon clari®es his intention by surrounding the report of the speech by clear markers of irony. and possibly progress (if my interpretation of 410a1 i kanoiÁ h san ± section ii.I N T R O D U C T I O N II. It illustrates the criticism by showing how even the brightest among them (409a4. they run the risk of falling victim to dangerous ethical theories. Yet he takes a great deal of trouble to make it obvious to his 210 . besides. The criticism of the Clitophon is identical to that of the Sophist: explicit protreptic `achieves little for much toil'. aporia caused by circularity. and by Clitophon's last words: those who have been explicitly exhorted cannot become virtuous and happy unless a wholly di¨erent method is applied. concentric reasoning. and cannot.4. I do not wish to imply that the Clitophon contains an absolute rejection of protreptic as such: Socrates is `worth the world' to those who have not yet been exhorted (410e5±6).

The degree to which these characteristics are stressed goes beyond normal practice in Plato's aporetic dialogues. what follows then?' (Pseudo-Platonica.380 I have not seen the second aspect of its 380 I rather agree with Heidel's remark: `Clitopho says in e¨ect: Take the Alc. elenchos as a means of reaching aporia) recommended. Euthd. Platone. b7±c2. 457±60) and Souilhe (177±9). 144±7): aporia is an indispensable element of elenchos and as such is recommended. The intention of the Clitophon therefore has two aspects: explicit protreptic is condemned. may have been among the set of texts which the author of Clit. Heidel's emphasis).4. though Antisthenes' Alcibiades and Protr. condemns. There is. I ®nd myself in partial agreement with JoeÈl (Der echte .1. 211 . I would wish to substitute `protreptic'. d1±4) and there are two circular regresses (408d5± 6. and (certainly) Pl. not rejected.3). I cannot follow Stefanini when he takes Clit.2) and the words kataÁ seÁ . upotei nwn au toiÄ v (408d1. 6. as a criticism made by Plato himself of the aporetic dialogues of his earliest period (similarly Grube. Gaiser. 410a3±4. b5±6.5.1. (2) there are no typical Antisthenean traits in the protreptic speech. 2. but that is su½ciently explained as being due to the author's wish to indicate that Clitophon's technique is elenchos as used by Plato. with the proviso that where these scholars speak of `Antisthenes'. implicit protreptic (more precisely. 410a2±6). but again I would extend `Alc. . Sokrates. Socr. i 483±4).I N T R O D U C T I O N II.1). .: (1) the rhetoric (said to be typical of Antisthenes' Protr. 1 for granted. i 207 o¨ers two sensible arguments against a purely antiAntisthenean tenor of Clit. an abundant use of analogies (408e3±409a2. . L. . Alc. for instance. Protreptik. 1' to `protreptic'. `The Cleitophon of Plato'.. in the Clitophon. In formulating the meaning as I do here. especially Socrates' irony (section i.6 readers that Clitophon's interrogation is a specimen of the aporetic dialogue. Justice is said in so many words to be a te cnh (section ii.5. I may add that the allusions to (perhaps) Aesch. D. section i. BruÈ nnecke (`Kleitophon wider Sokrates'. b1±5. Some features in the text reinforce this message.5) does not cease with Clitophon's report of Socrates' speech.. do not tally with such a tenor. 47 n. section ii. c4±d1. Stefanini. Ap.

too. e¨orts are made to indicate that far from criticising the Republic. on 407b2±e2). he is partly right (cf. He does so by reformulating in a positive way the conclusion that justice is not harming anybody (section ii.3) that the aporetic dialogue is Plato's alternative to explicit protreptic. its protagonist and its de®nitions of justice were borrowed from that Platonic dialogue as every reader who knew it was bound to see.I N T R O D U C T I O N II. section ii. The proof that wrongdoing is involuntary is a parody. Socrates he made the symbol of protreptic literature. he does not hit at the values to which Socrates exhorts his fellow-citizens there (section ii. but as Clitophon is made to attack him. the author recommends this work as providing the ultimate solution of the question what justice is.3. Comm.1). If I have correctly understood the intention of the Clitophon. but not of Socratic ethics (cf. but he overlooks the role of elenchos in the Clitophon.4. I have argued (section ii.5).2).2. the work was written from an unambiguously Platonic point of view. When he parodies Socrates' little protreptic speech from the Apolog y.6 meaning discussed anywhere: when BruÈnnecke (458±9) says that the Republic is the author's alternative for Antisthenes' `ober¯aÈchliche Protreptik'.5. and justi®es the statement from Socrates' speech that justice and politics are identical (section ii. by pointing at the notion of justice as the best state of the 212 . The reasons why explicit protreptic is condemned in the Clitophon are Plato's reasons. our author ran the risk of being seriously misunderstood. The distinction between e rgon and di dagma entails the Socratic paradox that virtue can be taught. the dialogue could easily be interpreted as an attack on Socratic philosophy. Similarly. the Clitophon might have been interpreted as an attack on the Republic.2).4. and the alternative is Plato's alternative. Here. In constructing this pamphlet as he chose to do.

The exaggerated picture of a rhetorical. we can identify the Apolog y and Euthydemus as far as Plato's works are concerned. and indeed of all Socratic literature 381 But cf. Atti dell'Accademia Pontaniana 40 (1991) 327±43. and by taking his de®nitions from the abortive discussions in the ®rst book. Yet. Zwei attische Epitaphien (Berlin 1933). Finally. 382 On the ancient views on Mx..2) about the meaning of the Clitophon: they show that it was not correctly understood.5. In itself. Me ridier (Bude ed.382 But the fact that from Schleiermacher onwards many sensible scholars have considered the Clitophon an attack on Socrates. cf. nobody in antiquity appears to have understood its intention.6 soul (section ii. 76±7. cf. not a philosophical pamphlet. `L'e loquence de Platon dans le Me ne xeÁ ne'. Did the author have no other way of conveying his message than by having Socrates victimised by an admirer of Thrasymachus? How certain could he be of being correctly understood? There are only two testimonies from ± later ± antiquity (Ptolemy and Synesius..5.1).I N T R O D U C T I O N II. one wonders whether the means of attack was well chosen.4. the allusions to literary texts (apart from the Republic.381 the Menexenus was a parody.3. which are said to be unsatisfactory in the Republic itself (354a13±c3.4 ± and there may have been many more) may themselves have served as so many indications that the Clitophon is a literary. when all is said and done. 213 .1 and i. section ii.). sections i. preaching Socrates may have strengthened these indications. Oppenheimer. does indeed show that to a certain extent the form of this dialogue obscures its meaning (though the virtual disappearance of fourthcentury protreptic. as most scholars hold. 357a1±b4. the popularity of Socrates' protreptic speech in later antiquity suggests that this speech was taken quite seriously.2. K. 68±70. 358b2±4 etc. M. Dirat. and even on Plato.5). Likewise. this does not say much: if. one of the Alcibiades dialogues by an author other than Plato or pseudo-Plato ± section ii.

on 408c4±7).3. 19) that Plato did not complete the Clitophon because he had already expressed the same ideas better in the Euthydemus. ii 2 (1874) 369±85.I N T R O D U C T I O N II. I may add that both in the protreptic speech (sections ii. I think the answer to the latter question must be a½rmative. sections ii.7.1 and 3) and in Clitophon's handling of elenchos (section ii.3).4)? An answer to this question is indispensable for solving the problem of authenticity. but rather less clearly. Mnem. I may remind the reader at this point of de Strycker's opinion (section i. The intentions of the two dialogues are not identical. it suggests a solution for the problem of ethical knowledge (and may suggest the theory of Forms). Socrates' conversations with Clinias are specimens of explicit protreptic and of elenchos at the same time: the 383 `Platonica'. The problem is rather whether the intention of the Clitophon is wholly included in that of the Euthydemus. To what extent is the intention of the Clitophon di¨erent from that of the Euthydemus (section ii.3.383 `Plato sua repetere non solet'.4. and that the sentence which connects the two parts of Clitophon's report can be understood only as an allusion to a similar sentence in the Euthydemus (Comm. contributes largely to its meaning being obscured). at 370. or to put it di¨erently. cf.1 and n.1 and 2. The Euthydemus ®rst and foremost contrasts Socrates and eristic (and takes Isocratean instruction in its stride).3. This contrast is to some extent present in the Euthydemus.2. whether the Clitophon brings anything new384 that is not already to be found in the Euthydemus. 214 .6 apart from Plato and Xenophon. The author of the Clitophon contrasts explicit protreptic and elenchos and shows that the result of protreptic is negative and that of elenchos positive. because. as Cobet says. 384 If the Clitophon was written by Plato. he wrote it considerably later than the Euthydemus. the author was rather strongly inspired by the Euthydemus.

Socrates' protreptic speeches are directed to crowds. more generally. because after all Socrates' method is indeed successful there. that is to say. cannot defend themselves.2 ad ®n. as the Euthydemus is not. a message which runs parallel to that of the Euthydemus but is not identical to it. Plato's death is its terminus ante quem. The objections raised against the written lo gov in the Phaedrus (section ii. protreptic literature may be read by anyone (section i.I N T R O D U C T I O N II. so that it would appear to be arti®cial to treat them separately.7 Date and authenticity The problems of date and authorship of the Clitophon are tied up inextricably. The reason why explicit protreptic is rejected in the Sophist.7 subject-matter is protreptic. are illustrated in the Clitophon and cannot be illustrated in the Euthydemus. whether or not his frame of mind makes him ®t for exhortation. concerned with the phenomenon of literary protreptic and. with the problem of the best form of philosophical literature. as explicit protreptic. but he does make a great deal of progress). yet it does have an intention of its own. Socrates must remain silent after Clitophon's requisitoire: writings.5) recur in the Clitophon.3. the Clitophon is. ii. cannot answer questions. the Clitophon is closely related to (and relies on) the Euthydemus. and application of the less 215 . its lack of success and its failure to remove doxosoji a. or from Socratic philosophy (in which case he may fall into the hands of the sophists.). the method elenchos. Secondly. he may turn away from philosophy completely. If the work is authentic. the conversations fail to reach their aim. as elenchos they do not (Clinias does not acquire knowledge.5. as Clitophon does). in this case protreptic literature. including protreptic writings. Socrates fails to give a satisfying answer to Clitophon's question: the written work. Therefore.

385 I take it for granted that the date of its publication virtually coincided with that of its completion. 1).I N T R O D U C T I O N II.4 (and n.7. A. G. 57±61).4. section ii.5.6 appear to have been inspired by one or more Platonic dialogues in which the method of diai resiv was outlined. Priority of the Clitophon is not unlikely but de®nitely unprovable. PW s. most probably Plt. Prm.1 Date I have already indicated (sections ii.4. Sokrates.7.5. 234).11±12 and 4. Ryle.7. Memorabilia 1. Laws 728e5±729b2.1.5.7. cf.3. van Groningen. whatever may be meant by `publication' (for two unsatisfactory answers. but as they are also crucial for settling the problem of authorship they will be treated separately (ii.2. The latest of these is Plato's Republic. I think it very likely that Xenophon refers to the Clitophon. 275d9±e3 that Plato considers it normal that a work. Yet those who doubt its authenticity are entitled to a treatment of all the data relevant for its time of composition (ii. once it has been written. (H. cf.3).2. B. but I see no way of proving it beyond doubt. n.2. cf. 386 3.7. iv 437). is made available for everyone who wishes to get hold of it (but not invariably so. cf. after 370 bce (HGPh. M. 21±54). cf.1. 360 to section ii. the Clitophon must have been published. Treu. Plato's Progress (Cambridge 1966). ii. sections ii.5 presupposes the period of Theban hegemony (cf. `EKDOSIS '.5) various texts on which the Clitophon can be shown to depend.1). 4. Kurzdialoge. As the Memorabilia is at any rate later than the Republic. 8±10 (and contra: C. Maier. 216 .386 there are no chronological problems.4. 24±5 n.1. W. 1776). ii.2) before this problem can be discussed (ii. (2) Plato. Xenophon 6). MuÈ ller. 128d6±e1).385 let us say. The following texts must be considered for settling a terminus ante quem: (1) Xenophon. There is good evidence at Phdr.4.1 unreliable stylometric tests would su½ce to date it in relation to his other works. Language and style are among these data.v.1 and ii.

388 387 As mhde . Cambridge. Gaiser. The hypothesis that Chrysippus quoted Clit.761 und der Dialog Kleitophon'. Some of his grounds are not cogent: hkist' e stiÁ protreptikoÂv ± Clit. but nothing of this is in the text of Euthd. `RaÈtsel'.761 (ˆ Plutarch 1039d±e). on 409e2 ou de . Plutarch's Moralia. An attack on Plato is given by Plutarch as an example of Chrysippus' inconsistency: e n deÁ taiÄ v proÁ v e te rouv a ntilogi aiv h kista jronti zei touÄ mhdeÁ n eipeiÄ n e nanti on e autwÄ i kaiÁ diaÂjwnon. tauÄt' ei rhke kataÁ le xin´ ``o gaÁr toiouÄ tov loÂgov kaiÁ e autwÄ i maÂcetai kaiÁ hkist' e stiÁ protreptikoÂv . 278) proposes Euthd. 19. s. 169. . Cherniss. . which he regards as spurious). . 154.±London 1976) is lectio facilior. Comm.2. . Hartlich (`De exhortationum . . Priority of the Clitophon seems quite likely. so too Ge¨cken. section ii.. but is again not subject to proof. this would seem a clear instance of `emphatic' ou de (mhde ). reported by H. . `Chrysipp iii. proÁv e tera tina maÄllon hmaÄ v protre yetai h toÁ jilosojeiÄ n . cf. SVF 3. An extensive argument for this hypothesis is given by R. nor' (LSJ. but no proof: Chrysippus may very well have taken a statement 217 . Mass. was put forward by Pavlu. (4) Chrysippus. cf. 281b4±e1 in combination with 288e4±289b3. mh (two MSS. ou de a ii 2). qui vita uti nesciat. is stated in Clit. e n gouÄ n toiÄ v PeriÁ touÄ protre pesqai touÄ PlaÂtwnov e pilambanoÂmenov le gontov oti twÄi mhdeÁ 387 maqo nti mhd' e pistame nwi zhÄ n lusiteleiÄ mhÁ zhÄ n. Protreptik. melius non vivat'. Kesters. Souilhe . . 3. `Authenticiteit'. historia'. mhd' cannot mean `neither . 439 n.1 (3) Alcibiades 1 126c1±d10. xiii 2. Kleitophon'. . which he goes on to quote. `qua ex ratione e½ci cogique potest ut. 180±1. 408c2 protreptikwtaÂtouv te h gouÄ mai is an attractive parallel.'' kaiÁ mhÁn ou c e tera deiÄ bibli a dieilhÄ sai touÄ Crusi ppou thÁn proÁv autoÁ n e ndeiknume nouv maÂchn (there follow three quotations in which Chrysippus says or implies that death is preferable to being evil or foolish). 512a2±b2. . cf. `Pseudopl. Eranos 59 (1961) 89± 100.. . but he is silent about the possibility of Chrysippus quoting Clit.I N T R O D U C T I O N II. 388 Von Arnim and the editors of Plutarch before Cherniss refer to Pl. Grg. Westman. (Hartlich saw very well that what `e½ci cogique potest' from Euthd.7.v. 141 n.5.

246d5. (d8±9).. 195d1±2. (b) and (c) are more or less identical.1 and n. Judged from a Platonic point of view. broken down by the question ti nov e pisthmoÂnwv. 862e1±863a2. In later Stoicism.7. Nor do these phrases (or nominalisations of them) occur in the Stoicorum Veterum Fragmenta. Ap.167). (b) `if one lacks justice and virtue in general' (Laws 661c1±5). Mx.1.256 with Chrm. 391 Compare SVF 3. since this work did not con®ne itself to protreptic writings proper: it quoted Tyrtaeus and Theognis (SVF 3. Smp. 410a3±4. As e pi stasqai (manqa nein) zhÄ n is not Platonic. section ii. they are equally far removed from Chrysippus' twÄ i mhdeÁ maqoÂnti mhd' e pistame nwi zhÄn. 164b7±8. 926b4±6. Crito 47e3±7. La.I N T R O D U C T I O N II. 304e2±3. (a). Hp. cf. 390 Chrm. 942a1±4. Chrm. 7 340c4. 218 . 735e3±5. there is only one way left: to ®nd out which particular text from Plato's works induced Chrysippus to use the wording he did use. Republic 445a5±b1).390 it is impossible to maintain that Chrysippus took one of the six texts quoted above in order to reformulate it in yet another Platonic way. or (c) `if one's soul is in a bad state' (Gorgias 505a2±b1. but the statement e pisthmo nwv an praÂttontev eu an pra ttoimen kaiÁ eu daimonoiÄ men is.391 apart from the passage under discussion. 389 Grg.1 In many places throughout his works.Ma. 512a2±b2. 483b2. 38a5±6. 216a1. Lg. 176. Plato states as his opinion that for some people it is better to die than to continue their lives. Ep. typically.2. 854e1±7.389 The only passages deserving serious consideration are those in which it is said that it is better to die (a) `if one cannot handle one's soul' (Clitophon 408a5±6). 62a4±5. Tht. Phd. therefore we may also exclude the possibility that Chrysippus expresses what he found in one of the six passages mentioned in the language of his own philosophy. the phrase e pi stasqai (manqaÂnein) zhÄ n would have been less abnormal. as quoted in the previous note. 176a8±b1. There can be no from Plato and considered it from a protreptic point of view (91±2). R. The fact that the words occur in PeriÁ touÄ protre pesqai leads us nowhere. Consequently. 173d3±4 comes closest.

We know nothing at all about the date of this latter work. 393 The occurrence of e pizhÄn both in Lg. The proof will not be complete unless all other texts are excluded. In (b) there is little to justify jumping from `if one lacks justice and virtue' to `if one does not know how to live'. on 408a5 yuchÄi . Plato does not exactly say in this passage that it is better not to live.393 The remaining four texts. but that life is the greatest evil if one possesses all so-called agaqa without justice and virtue. though it might be argued that `to lack justice and virtue' implies `not to know justice and virtue'. and that he understood it as described. all stating as a premise that life is not worth living if one's bodily constitution is bad. crhÄ sqai. it is possible392 to explain the condition ostiv yuchÄ i mhÁ e pi statai crhÄsqai (408a5) as `whoever does not know how to live'. I therefore conclude that this is the passage he had in mind. start from the analogy of body and soul. it is better not to use it'. 392 But far from necessary.7. is neither Platonic nor Chrysippean. Besides. which is exactly what Chrysippus says. 661c5 and in Chrysippus' refutation must be a coincidence. . since it is the only one in which one ®nds the verb e pi stasqai. a n w v o li giston o toiouÄ tov cro non e pizw hi. (a). 408a6). (c). a formula which. 219 . Comm. this is still not the same thing as `not to know how to live'.1 doubt that the Clitophon passage.I N T R O D U C T I O N II. as shown above. It is hard to see how Chrysippus could have twisted this reasoning into his `not to know how to live'. so we shall have to be content with that of Chrysippus' death as terminus ante quem for the Clitophon. Now this passage contains the application to the soul of the general protreptic proposition `if one does not know how to use a thing. . As in this application `not to use (one's soul)' is explained by `not to live' (toÁ a gein hsuci an thÄ i yuchÄ i kaiÁ mhÁ zhÄ n. e latton de . It is therefore certain that the Clitophon existed and bore Plato's name by the time Chrysippus wrote PeriÁ touÄ protre pesqai. has the best chances. cf.

protreptic of a mainly ethical character continued to be written in the third century. If my interpretation of the Clitophon as a pamphlet attacking Socratic protreptic is right.184.I N T R O D U C T I O N II. L. Kleitophon'. Histoire. `Authenticiteit'. but as this scholar lived somewhat later than Chrysippus. but it 394 D. it would be useless to do so here (see Kesters.1 This philosopher died during the 143rd Olympiad394 (208/ 4 bce).1 n. F. I believe one can make out a good case for priority of the tetralogical list to the trilogies of Aristophanes of Byzantium. ± I shall make no use of the inclusion of Clit. section i. All speculations about tetralogically arranged Academy editions dating from the time of Arcesilaus or even Xenocrates are destined to remain fruitless.4. 201) is refuted if my claim that it is used and ascribed to Plato by Chrysippus is correct. `Dialoghi pseudoplatonici'. 112.7. 7. 19±20) or even to the middle of the third century (Carlini. At what time within this period is it most likely to have been composed? The answer to this question depends necessarily on the view taken of the intention of our dialogue. those who regard it as an attack on the historical Socrates usually tend to assign to it a very early date (often combined with the hypothesis of a separately published `Thrasymachus'). which probably existed already in 45 bce (Alline. Ueberweg's argument that the Clitophon is not attested before Thrasyllus `also uÈ berhaupt auf eine voÈllig unzureichende Weise bezeugt' (Untersuchungen uÈ ber die Echtheit und Zeitfolge platonischer Schriften (Vienna 1861). On the other hand. 220 . Those who think that the meaning of the Clitophon is exhausted when it is stamped as a `cento' or even a `school exercise' will have no problem in dating it to the end of the fourth or the beginning of the third century (Pavlu. `Pseudopl. ± Conversely. quite possibly. This seems to exclude a third-century date. The Clitophon was written between 370 and the end of the third century bce. 55±7). 37). 163±7). it stands to reason that such a pamphlet makes sense only in a period in which Socratic philosophy and the lo gov Swkratiko v do not belong to the past. in the tetralogical list.

7. A. Perhaps we can go a little further than that. Now.I N T R O D U C T I O N II. Bluck. CQ n. as the Socratic schools (apart from the Academy and Peripatos) and Socratic literature had become virtually extinct by then. apart from the accusing tone of the speech (see also on 408e5) and the absence of the word jilosoji a and its cognates. in J. Bos. witness the post-Platonic Alcibiades 1 (usually dated 350±340395).1.1 would have been pointless for the author of the Clitophon to make Socrates the symbol of that type of protreptic if he wrote it in that century. Alcibiades maior. C. ethical type to the quietly arguing philosophical protreptic in the stricter sense as foreshadowed by the Euthydemus and represented by Aristotle's Protrepticus (section ii. 3 (1953) 46±52. . 100±12. 221 . De Alcibiade I qui fertur Platonis (Bonn 1906). de Strycker. R. Comm. interest in ethical protreptic did not cease to exist at once. 101±22 at 121. `Le premier Alcibiade'. accusatory protreptic as parodied in the Clitophon does seem to become less important after the ®rst generation of Socratics: both Plato and Demetrius consider it ine¨ective in comparison with the interrogative type and Plato calls it `old-fashioned' (a criticism echoed by Demetrius when he says that the protreptic dialogues `met with great success once they had been invented'. As we saw. `The origin of the Greater Alcibiades'. 51±2. esp. In fact. Of course. it is possible to trace a line of development in fourthcentury protreptic which runs from the reprobatory. but the Alcibiades 1 is a protreptic dialogue and as such belongs to the interrogative type of protreptic set o¨ by Demetrius (and Plato) from the accusatory type. Bidez. the way it uses certain protreptic motifs indicates that it is closely related to the older type of protreptic and has not much in common with protreptic in the stricter sense (cf.5). 395 H. on 408a5 yuchÄi . crhÄsqai ). Arbs. 64±5. S. Eos ou Platon et l'Orient (Brussels 1945). .s. E. Socrates' speech as reported in the Clitophon is a clear example of the older type of protreptic.

grammars. too. All these data seem to point to not too late a date in the fourth century (discounting for the moment the parallels adduced sub (1)±(3). I would say that our evidence.3. This hypothesis is reinforced by the positive side of our author's message: the aporetic dialogue as used by Plato is recommended as an alternative to explicit protreptic. scanty though it is. ii. the Theaetetus (written after.4. but questions of syntax. 350±330.2.7.1).1. indices. he has only limited access to the corpus (in this case. 369.2.7. There is. it cannot be later than c. If I have interpreted its meaning correctly. 222 . 330).5) seem to have been completely ignored. if indeed Xenophon refers to it. Now.2 sections ii. I submit that the Clitophon was written after the Republic. are as a rule far less easily answered. notably of sentence structure. v 61±2) is Plato's last aporetic dialogue.2 Language and style Two major problems confront everyone who wishes to use linguistic evidence for settling a question of authenticity. and after Plato the philosophical dialogue completely loses its implicitly protreptic function: during the last decade of Plato's life Aristotle was already publishing dialogues in which the injunctions of the Phaedrus (section ii. First. as I think very likely. it seems natural to date the Clitophon not very long after its terminus post quem of 370. ii. specialised studies) normally enable him to settle questions of lexicon satisfactorily. In this respect. does not encourage going beyond. but not long after. commentaries. Plato's undoubtedly authentic works) against which he is checking a given text (the Clitophon): his tools (the TLG CD-ROM. and in which discussion replaces elenchos. Taking all in all.I N T R O D U C T I O N II. and certainly before the end of the third century bce. say. HGPh. it was written towards the end of the seventies or during the sixties of the fourth century bce.

no method. I discuss in the commentary alleged signs of inauthenticity which I cannot accept as such. b6 e xaskeiÄ n. I think it may be accounted for su½ciently by the parallel use of eipe (le ge) and apokri nou as well as by its function of characterising Clitophon. cf.7. as no two sentences are identical ± this is essentially the second problem. b8 katamelethÄ sai.). Kurzdialoge. 398 See also section i.3 for Clitophon's irony. of answering the question whether Plato can be credited with the structure oratio obliqua ± recta ± obliqua observed in the sentence eipe n moi tauÂthn thÁ n te cnhn ei nai. given the fact that elsewhere in the corpus we ®nd jeu gein wv e cei podwÄn e kastov and bi ai oi cesqai jeuÂgwn (cf. supplemented by extensive but not systematic consultation of the whole corpus. 6. 223 . Thus. C. ad loc. If one has discovered a fact of language not recurring in the corpus. ou k allhn h dikaiosuÂnhn (409a5±6. 397 Cf. the use of e wv as a preposition in Alcyon 4). MuÈ ller. ad loc. b6 di dagma. ad loc.5. it must next be decided whether or not it is a sign of inauthenticity. 408c6 sunepiqumhtwÄ n. short of reading the whole corpus. hnper a kouÂeiv suÁ le gontov e jh Swkra touv. they are 406a10 jau lwv. 407a4 kataÁ kra tov. d5 qe v.2 for the formality of Socrates' opening words. 409a3 lege sqw. W. Again.398 It goes without saying 396 And even if one has read the whole corpus. d9 sune baine. but there is no way of proving that this is indeed a su½cient explanation. the parallels collected will have to be weighed carefully. Only in a minority of cases is the decision easy and will it be accepted generally (for example.5.397 but usually matters are more complicated and some degree of subjectivity is inevitable.I N T R O D U C T I O N II.2 for example.). c4 dio ti. but others may think di¨erently. i. I see no problem in the (metaphorical) use of jeuÂgein kataÁ kraÂtov in the Clitophon (407a3±4). 289 and n. the use of lege sqw following a direct question at the end of an utterance (409a3) has no parallel in Plato (cf.). My own method consisted of using the tools referred to above.396 The second problem is how to interpret the data collected.

d2 deiÄ n. a supplement in the form of a study of clausulae is provided by BruÈnnecke. .7.1. 473±7. 408d1 u potei nwn. Phlb. 410b1 (curtailed report). section ii. 224 .. Criti. Ti. it has been treated sometimes explicitly. but as a rule implicitly. 93±4. Both scholars arrive at the same conclusion as I do.400 Various stylometric tests have contributed to establishing a rough chronological classi®cation of the dialogues. gene sqai. about which I cannot make up my mind.... Plt. pronoeiÄ n is common in the fourth century bce. who concludes that Clit. Of the idioms mentioned. had it not been certain that Plato wrote them.. . Untersuchungen.2 that I have raised the question of authenticity far more often myself. comes closest to the dialogues of the latest period (Sph. as does adverbial makro teron. Another possibile way to test the authenticity of the Clitophon is to examine its language with regard to the development of Plato's use of language.399 There are only two idioms which to my mind can be regarded as evidence that Plato did not write the Clitophon (neither of these has been noticed before): 408e4 pronoeiÄ n and 409e1 taÁ plei w.) and to Epin. taÁ plei w `more often' is rare in Greek but occurs in Thucydides. What is quite certain in any case is that there are no possible linguistic objections against the date assigned to the Clitophon on grounds of content in the previous section. . e7±8 e mpo dion touÄ . d4 o n. `Kleitophon wider Sokrates'. we should expect a work ascribed to Plato to exhibit more or less consistently the traits of one period (cf. mh n.3 n. . to these could possibly be added 410b6±7 makro teron (Steinhart).I N T R O D U C T I O N II. 400 This was done by Ritter.. Lg. d5 ou n dh . This is not a very impressive total: six pages (OCT) taken at random from Plato's undoubtedly genuine works (especially his later works) will certainly yield no fewer traits which would have to be considered marks of inauthenticity. 399 See especially on 407a1 all' . 135). 409e8 w mologh kei.

on 408c1). Chronolog y. 58.. the aorist of jaÂnai (409e9) and the accumulation of articles at 409a3 ± none of these are found before. 225 .401 Besides. because ontwv/twÄi o nti is valid). and some only after. Phlb. on 407a2). a lhqwÄv/ w v a lhqwÄv (cf. 540 (on Natorp) and 540±1 (on me ntoi vs. not all stylometric criteria that have been proposed can stand closer scrutiny. in which dialogue two instances of dhÄ lon w v are found against forty-seven of dhÄ lon oti. and the frequency of various clausulae. 65. scedo n/scedoÂn ti (cf. Lutosøawski.402 There remain. it cannot have been written before the Republic (cf. which is avoided in Plt. Young. never dhÄlon w v. The Clitophon is far too short for such tests to be of any value. but the work is so short that this may very well be accidental.7. a majority of those which can pass muster cannot be applied to the Clitophon because of their nature (for example. Comm. the Republic. C.403 Consequently. cf. saje stata qualifying the copula (409e4). table 10. apart from w sper/kaqaÂper. The clausula [ W W [ [ W . 401 Cf.I N T R O D U C T I O N II. answer formulas) or because they are based on proportions (for example.. 403 Ritter. Comm. e terov as a variant for allov (409c2). Lg. cf. the relative frequency of wsper and kaqaÂper. note on 408b1). 168±73). toi nun). until the Republic. Similar phenomena are the use of o ntwv (for twÄ i o nti ) at 409e3.2. Comm. Chronology and my review. Plato's Logic. esp. Chronolog y. 77. Brandwood. M. (cf. For a more general critique. on 409e3±4. Brandwood. 123.. ontwv kaiÁ a lhqwÄv may be valid.2 Now. Untersuchungen. There is nowadays a general consensus about which dialogues were written after the Republic. if the Clitophon (which has two occurrences of dhÄlon w v. Brandwood. none of dhÄ lon oti ) was written by Plato. Besides. Plato invariably writes dhÄ lon oti. however. 2±3. 402 For these reasons. Thus. is not found before full stops in the Clit. in most of these (except Theaetetus and Parmenides) we ®nd dhÄlon wv side by side with dhÄ lon o ti. OSAPh 12 (1994) 227±50. `Plato and computer dating'. phenomena which we do not ®nd at all in Plato before a certain period. the following tests should be rejected.

0 and 31. as these include long stretches of rhetorical prose. `Kleitophon wider Sokrates'.2 the commentary will show that for many peculiarities of syntax. This result would put the Clitophon securely into the last period ± the number is slightly higher than that of most works of that period (the ®gures range from 4. I will try to say as little as possible about the general style of the Clitophon in this connection. if lengthy legal passages are included.9). the results cannot be ascribed to accident: if in a random distribution one ®nds between 46. 77±80).7 for Lg.I N T R O D U C T I O N II. 407 I discount the ®gures for Phdr. 406 In the oratio recta part of Socrates' speech hiatus is avoided altogether. cf. which we already knew.2). 405 But some books of Lg. cf. Raeder `etwa 7'.406 In this case. Chronolog y.404 It is possible to narrow this dating down even further.407 even a text of 3. Both applied the criteria of objectionable hiatus as set out by G..4 for Plt. and Mx. but this can in¯uence the outcome only marginally. respectively). BPhW 30 (1910) 1503±4. There are some markers of what Thesle¨ calls `Onkos style' (Styles.405 but the Clitophon is so small that such a variation should be accepted. BruÈnnecke reports 8. The relative rarity of parallels with Republic shows that the Clitophon is later. Brandwood. the middle-group dialogue with the lowest count. next paragraph. Raeder in his review of Pavlu. to 0. Two scholars have counted instances of hiatus in the Clitophon independently: BruÈnnecke. 153±7 and table 17. has 23.6 Didot pages is a large enough sample. and Cra. Janell (cf.2 instances of hiatus per Didot page (Ly. Kleitophon'.33 instances per (Didot) page. 226 . which is virtually equivalent to what is normally called 404 The relative frequency of such parallels as come from the Laws is adequately explained by the length of the Laws. and should not be taken as an indication that the Clitophon belongs to Plato's latest works. lexicon and idiom parallels can only be given from later works of Plato. `Pseudopl. have a higher rate of hiatus than Clit. ± Phdr.7. 467±8 and H. even though I think it does belong there. note on 407b1±e2.

interlaced word order). as Thesle¨ rightly remarks (79).2.3 `late style'. section i. Moreover. but this is a totally subjective judgement. it is in itself a strong argument in favour of authenticity. Therefore. Besides. This consistency would certainly seem to cancel out the few marks of inauthenticity that my investigation has brought to light. the language of the Clitophon is de®nitely more closely related to that of the Republic and later dialogues than to works dating from before this period. (On the style of the protreptic speech.I N T R O D U C T I O N II. notably a tendency to expand sentences and a tendency to deviate from normal word order.5. and consistently so.7. it rejects explicit protreptic (and recommends the Platonic dialogue) for the same reasons as Plato does. the author shows a deep understanding of what message Plato wished to impart by writing dialogues and what his motives were. (1) The Clitophon is written from a wholly Platonic point of view.7. the conclusion must be that there is no hard linguistic or stylistic evidence against the authenticity of the Clitophon.) In any case. But. and it must be stressed that some other typical features of the late style are absent (anaphora. ii. archaic and poetical words. one wonders how many of his fol227 . Personally. the close relationship between the language of the Clitophon and that of Plato's later dialogues agrees completely with the post-Republic date at which we arrived on grounds of content. the danger of subjectivity in assigning part of a text to this style is greater than in the case of any other of Plato's styles. I ®nd the style of the Clitophon most related to that of the Parmenides (mainly on account of its aridity). cf.3 Authenticity The following arguments may be advanced in favour of the authenticity of the Clitophon.

and stylometry bears this out. (4) The author relies heavily on other Socratic writings. Aristotle. This place. Plato's most intelligent pupil. as I think likely. the Sophist is the ®rst of the six undoubtedly genuine dialogues contained in this group. (2) The language of the Clitophon shows little that can be used as a mark of spuriousness.7.3 lowers had such an understanding. the Clitophon immediately precedes the Sophist. did not know what the dialogue was all about. The Sophist shows a renewed interest in the conceptions which lie at the base of Plato's use of the dialogue. moreover. the latest group. (3) The Clitophon has been transmitted among Plato's works and belonged to the corpus of these works at any rate at the end of the third century bce. In combination with arguments (1) and (2) this argument makes the burden of proof for inauthenticity an extremely heavy one. since the Republic belongs to an earlier group than (Clitophon and) the Sophist. its language is such that it can be placed without hesitation within the development of Plato's style. it is perhaps not an exaggeration to call the Clitophon a 228 . Clitophon's rejection of explicit protreptic and his practising elenchos as an alternative to it re¯ects a point of view which is entirely identical to the Eleatic Stranger's rejection of nouqethtikh and his recommendation of elenchos. in an author whose language is so varied as Plato's this amounts to saying that no linguistic case against authenticity can be made. Besides. The Clitophon must be later than the Republic.I N T R O D U C T I O N II. If. to which it alludes frequently. is in harmony with its content: the Clitophon belongs in the same group as the Sophist. The following arguments may be considered to tell against the authenticity (I list only such arguments as I can take at all seriously myself ). the data on hiatus suggest that within this.

we can be certain that he never refers his readers to other works of his own on such a scale.2. The refutation of the ®rst de®nition (ad 409c4±d1) and that of the second (section ii.5. as well as two other sources for the refutations of the ®rst and second de®nitions of the result of justice). 119). exhibits 229 .3). and in the discussion of justice the same holds for employing dialogues on the nature of justice (Republic 1.I N T R O D U C T I O N II. there are allusions to the Euthydemus at 408c4±7 and 410b4±6.3.1).3 `cento' as Gigon did (Memorabilien I. But apart from these passages. The quotation from the Euthydemus in 408c4±7 makes nonsense of the sentence in which it occurs (though it is better understood if the reader knows the context of the words from the Euthydemus). and it is doubtful whether he can be credited with this one.3.2) do not tally with the distinction made previously between te cnh and e rgon. the example of the neighbour's lyre makes no sense (section ii. on the one hand it is a cento even where we do not expect it to be.7. Even though we do not know to what extent Plato uses material from other authors. the author's style suddenly becomes succinct and even obscure when he is refuting philosophical theses (ad 409c6±d1). on the other. section ii. the only other longer passage in the Platonic corpus for which large-scale use of other published literary material may be assumed. the e pitaÂjiov in the Menexenus. (5) Apart from the fact that foreign material was incorporated. though a cento may be excused by the intention of the Clitophon. Apart from that. the clumsy manner in which it is at times adapted to the context is not in keeping with Plato's manner of writing. In the third part of the protreptic speech. and one to the Apolog y at 408c3±4 (cf. It should be stressed in this connection that `recycling' of Platonic passages is typical of most Platonic Dubia. Extensive use of protreptic literature is only to be expected in the protreptic speech.2. In other words. By contrast. Plato never wrote another cento derived to a large extent from his own works.

7. which I think very likely. only arguments (2) and (3) have been used in favour of the authenticity.6) ± one may well doubt if Plato was capable of doing this.1.I N T R O D U C T I O N II. (4) was used by Gigon (Memorabilien I. (2) especially by BruÈnnecke (476±7). to the best of my knowledge. on the surface. `as some people write and speak about him tekmairoÂmenoi ± on the basis of inferences'. 303) and Guthrie (HGPh. I have not found any myself nor have any been detected.). 119) and (6) has been the main reason for rejecting the Clitophon from Schleiermacher onwards. by others). Grube (`The Cleitophon of Plato'. may indicate that he did not regard our dialogue as Plato's work (see section ii. here an older Socrates is treated with irony and attacked without scruples by a younger Clitophon.4. No matter how much pain the author took to clarify his intention.3 no traces of de®cient adaptation (at any rate. he exposed Socrates to this attack willingly (see further section ii. but there a young Socrates is censured benevolently by a very old Parmenides. (As my interpretation of the meaning of the Clitophon is essentially a new one. In previous discussions of the authenticity. Against the authenticity. 3) point to the attack in the Parmenides. (3) by most scholars. I disre230 . an attack on Socrates (and runs the risk of being taken as an attack on Plato himself ). this strongly suggests that the latter was not written by Plato. his wording. v 388 n. Can Plato really have ascribed to Socrates the view that it is just to bene®t friends and harm enemies.1 ad ®n. a view which Socrates so eloquently rejects in the Crito? (7) If Xenophon does indeed quote from the Clitophon. (6) The Clitophon contains. The partisans of either position have rarely bothered to go into the arguments for the other side. As he wrote the Memorabilia not very much later than the Clitophon itself seems to have been written (ten to ®fteen years at most).

indeed far more plausible. Only the Epinomis and the Letters are forgeries if they are spurious.) Consequently. the defenders of the work's authenticity must accept. that Plato. and that he indulged in borrowing from his own work (explicable for the protreptic speech and Clitophon's interrogation) even beyond what was absolutely necessary. but also the niceties of his use of elenchos. for once.7. Kurzdialoge. 231 . But how strong is the case for the opposition really? The allusions to Plato in the Dubia are of a di¨erent nature 408 It is out of the question that such an author can be considered a forger. could easily have slipped into a collection of Plato's works made after his death. we shall have to examine the arguments given here from both sides. this holds for most of the authors of Platonic Dubia and Spuria. in which he deals with the very same problems that are raised in the Clitophon. W. someone who not only understood completely Plato's aims in writing dialogues and rejecting protreptic.408 On the other hand. who had adapted himself so much to Plato's manner of writing that even his use of language faithfully reproduces that of Plato in his last period. 17±18. chose the combined forms of cento (a dangerously vague term) and Short Dialogue for conveying his message. someone who avoids hiatus to the same extent as Plato did when he wrote the Sophist. Such an author is too much of a hypothetical construct to be at all acceptable: he is so much a lookalike of Plato that it is more economical. MuÈ ller.3 gard here many arguments based on other interpretations that have been pro¨ered. Argument (3) is in itself no problem for someone who believes in an intelligent pupil: a work by such a pupil. to identify him with Plato. moreover. Those who reject the authenticity will be able to deal with arguments (1) and (2) only if they assume that the Clitophon was written by someone very close to Plato. with regard to argument (4).I N T R O D U C T I O N II. cf. written probably in Plato's lifetime. C. someone.

In the Dubia. they are on the whole very obvious. 408c4±7. once the Clitophon had been jotted down. The allusion to Euthd. he and Xenophon may have independently reacted to accusations made against Socrates in the polemics of the time ± each in their own sepa232 . Apolog y. 410b4±6 is even less obtrusive.7. used often in the case of the Seventh Letter).3 from those in the Clitophon. so much so that it is often easy. and that Plato. de Strycker±Slings. if we athetise the Clitophon because of the attack. to identify speci®c passages from the authentic works which they elaborate on. 274e8±275a2 at Clit. even for readers with a super®cial knowledge of Plato. The attack on Socrates ± argument (6) ± may be compared to that of the Parmenides. Some recent scholars treat Plato as if he were the epitome of clumsiness. and we know from the Apolog y that Plato was in¯uenced by the generic conventions of this kind of speeches (cf. It should be noted. it may be argued that as some element in Plato's dialogues has to be the biggest cause of o¨ence. I do have some problems with Clit. Argument (7) could be answered in di¨erent ways. that clumsiness is a highly subjective concept. which I cannot understand unless it is a hidden reference to Euthd. however. Argument (5) will have to be disposed of by assuming that the Clitophon was composed in haste (a convenient route of escape. and so on. but it is a good maxim in classical scholarship that `once is never'. 283a1±7. 235±8). The Apolog y passage is alluded to in a rather unobtrusive way ± and I feel that Plato in his old age must be permitted such a thing.I N T R O D U C T I O N II. The appeal to the Menexenus is not valid: there Plato could draw upon a long oral and literary tradition of e pitaÂjioi. The allusions in the Clitophon which I mentioned are of a di¨erent nature. If the Clitophon was written by Plato. logically. we are forced to athetise the Parmenides next. did not trouble to revise it more thoroughly.

section i. as Professor Sedley suggests to me.1. Besides. I would gladly leave the choice between the two positions to my readers. it seems fair to say that (5) and (6) cancel each other out to a large extent: the more you stress the clumsiness of Clitophon's attack. Alternatively. This is in harmony with what little we know of other interpretations of the Clitophon from antiquity. In other words. The only objections against authenticity that I consider really serious are (5) and (6).1 and n.5. I do not quite understand Xenophon's tekmairoÂmenoi. the argument needed to disprove (1)±(3) ± the intelligent pupil ± is utterly weak and suspect. Xenophon did not know the Clitophon.2). If they value the opinion of someone who has lived with this little work on and o¨ for the past thirty years. the less Socrates is really harmed.3 rate ways. less problematic than the assumption of an intelligent pupil would be.3. 152 of the Introduction). tekmairo menoi in Xenophon) that (the literary character) Socrates is incapable of going beyond protreptic.I N T R O D U C T I O N II. to my mind. just as Aristotle can say tine v when he means Plato (cf. But they are. the hypothesis required to explain points (1)±(3) is very weak and far-fetched when compared with the explanations of (4)±(7) which can be given if the authenticity is accepted. I must say that unless I am entirely mistaken about the intention of the Clitophon (but that still leaves (2) unexplained). Or perhaps Xenophon thought Plato did not endorse Clitophon's criticism.7. although not without 233 .4. section ii.1) and Synesius (section i. those of Ptolemy (cf. but actually means Clitophon when he says e nioi. But if so. Arguments (4) and (7) can be countered without too much trouble. Therefore. which better suits authors of texts whose authority Xenophon tries to undermine than characters within texts. Xenophon may have thought that Plato endorsed Clitophon's criticism of Socrates: it is fair to say that Clitophon infers (cf. If so.

I N T R O D U C T I O N II.7.3

hesitation,409 I accept the Clitophon as a genuine work of
Plato. I repeat what I said at the beginning of the Introduction, that I consider the authenticity problem a minor
issue compared with the problem of the meaning of this
409 Paul Shorey writes (What Plato Said (Chicago 1933), 658) that as a
doctorand in Munich in 1884 he maintained the thesis that the Clitophon is authentic: `I doubt it now.' From what Shorey goes on to
say it becomes quite clear that he regards the Clitophon as spurious.
Over the years my position has become the exact opposite: the balance which I have drawn in this section causes me to claim that the
Clitophon is, after all, authentic. But I am less con®dent a scholar
than Shorey had the right to be.




Cod. Parisinus graecus 1807, s. ix exaratus, cuius
imaginem luce expressam contuli.
eiusdem codicis manus veteris qui dicitur diorthotae, eadem atque manus prima.
eiusdem codicis manus fere aequalis. sunt qui
Arethae attribuant.
eiusdem codicis manus admodum recentior.
manus Constantini Hierapolis metropolitae, s. xii.
Cod. Marcianus graecus 185, s. xii exaratus, cuius
imaginem luce expressam contuli.
eiusdem codicis manus recentiores, s. xiv non
Cod. Vindobonensis Suppl. gr. 39, s. xiii vel xiv
exaratus, cuius imaginem luce expressam contuli.
eiusdem codicis manus recentiores, s. xv non anteriores.


Cod. Parisinus graecus 1809, s. xv exaratus, qui
Clitophontis partem tantum continet (usque ad
pag. 408d3). hic codex quamvis e cod. A derivatus lectiones alterius testis, cum Themistii textu
Platonico cognati, hic illic continet. imaginem
luce expressam contuli.


Cod. Vaticanus graecus 2196, s. xiv exaratus,
cuius imaginem luce expressam contuli. cum Pa
arto vinculo cognatus; eiusdem testis atque cod.
Pa vestigia praebet.


Codicum ADFVa (Pa) consensus.


lectio in rasura scripta.
lectio supra lineam scripta.
lectio in margine scripta.


X pl
X1 , X2

lectio per litteras perscripta.
lectione primae manus punctis deleta
manus prima, altera (etc.).
incertum quae manus






ristwnuÂmou tiv h miÄ n dihgeiÄ to
KleitojwÄnta toÁ n
e nagcov, o ti Lusi ai dialegoÂmenov taÁv meÁ n metaÁ SwkraÂtouv
diatribaÁv ye goi, thÁn QrasumaÂcou deÁ sunousi an uperepainoiÄ .
± O stiv w Sw kratev ou k orqwÄv a pemnhmo neue soi
touÁ v e moiÁ periÁ souÄ genome nouv lo gouv proÁ v Lusi an´ taÁ meÁ n
gaÁr e gwge ouk e ph inoun se, taÁ deÁ kaiÁ e phÂinoun. e peiÁ deÁ
dhÄlov ei memjo menov me n moi, prospoiou menov deÁ mhdeÁ n
jronti zein, hdist' an soi diexe lqoimi au touÁ v auto v, e peidhÁ
kaiÁ moÂnw tugcaÂnomen ontev, i na httoÂn me h ghÄ i proÁv seÁ jauÂlwv e cein. nuÄn gaÁr i swv ouk orqwÄ v a khÂkoav, w ste jai nei proÁv
e meÁ e cein tracute rwv touÄ de ontov´ ei de moi di dwv parrhsi an, hdist' a n dexai mhn kaiÁ e qe lw le gein. 
ll' ai scroÁn mhÁn souÄ ge w jeleiÄ n me proqumou± A
me nou mhÁ upome nein´ dhÄ lon gaÁr wv gnouÁ v ophi cei rwn eimiÁ

406a1±407a4 resp. Synes. Dio 57d±58a ˆ 2.270.12±16 Terzaghi.
407a1 resp. D.Chr. 13.15 ˆ 1.233.3 de Bude .
Tit. PlaÂtwnov (add. D) KleitojwÄn h protreptikoÂv AD: KleitojwÄ n F
A: om. DF
taÁ touÄ dialoÂgou pro swpa SwkraÂthv KleitojwÄn A: om. DF
xun- F
406a1 ti v w
2 meÁ n om. PaVa
3 ye gei F (ut vid.) corr. Fxpl
(sun- et Synes.)
u perepainoiÄ AD: e painoiÄ D 2p : u perepaineiÄ FVa
5 ante o stiv alium interloc. ind. A: ante a7 e peiÁ DF
post o stiv lac. stat.
Schanz: o stiv hhni Hermann: o sho vi tiv (`ille ``quidam'' tuus') Bury: wv tiv
e peidhÁ deÁ PaVa
7 taÁ deÁ kaiÁ e ph inoun bis scrips. D corr. D 2p
10 o nte A4pl , D 2 (ut vid.)
8 mhdeÁ n] mhÁ PaVa
9 jronti zein] ei de nai Axm
proÂv me DF
12 touÄ de - bis scrips. et lineola
11 jai nei ADF: -hi A4r
corr. D
ante ei alium interloc. ind. F
ei dh F
13 h dist ' a n] hdista
407a1 ante a ll ' alium interloc. ind. ADF
me om. PaVa


so it looks as if you're more irritated with me than I deserve. the more so since we are alone. someone told us the other day that in his conversations with Lysias he criticised his sessions with Socrates and was full of praise about his contact with Thrasymachus. although you're pretending you don't care. since you're obviously cross with me. Some things in you I didn't praise. If you allow me to speak freely. socr Why.CLITOPHON socr About Clitophon. the son of Aristonymus. That way you won't be so convinced I'm on bad terms with you. As it is. other things I did. I would certainly be glad to give you my own detailed account of what I said. 241 406 5 10 407 . it would indeed be disgraceful of me not to put up with it when it's you who o¨er to help improve me. I'd be delighted to take the opportunity ± I'm ready to tell you all. you may not have heard the whole truth. clit Someone who wasn't giving you a correct report of what I said about you to Lysias. Now.

Const. Bidez. ind. a8±408b2 cf. Corp. in Grg. 3 ui e wn w Ps. Tim.6 sq. et Plu.: ante voc.Chr. opo te e pitimwÄ n toiÄ v anqrw poiv w sper e piÁ mhcanhÄ v tragikhÄ v qeoÁv umnoiÄ v le gwn´ ``PoiÄ je resqe.Chr.trag.Herm. Paton±Wegehaupt. pouÄ Ps. (poiÄ omnes ceteri exc.21 de Bude .Porph.22. Them. Schanz mhdeÁ n D. DF e do kei D 8 e piÁ w: e k Them. Eloc.-Plu. F 5 ante akou oiv alium interloc. Lex. Cobet: h Ast (cf.) mhc. Them. Luc. Philostr. (a) Him. kaiÁ agnoeiÄ te oudeÁ n twÄ n deoÂntwn pra ttontev. kai moi e do keiv paraÁ touÁ v allouv a nqrwÂpouv kaÂllista le gein. Olymp. Jul. thÁ n paÄsan spoudhÁn e cete AD 2p cf. Demetr.233. (PY). w cf.16 ˆ 1. Plu. tragikhÁ skhnh . Corp. 13. Olymp. Olymp. (PY) Olymp. Tim.21±233. 3.288.D. Const.21 Rabe 4 anaÁ kraÂtov van Herwerden post kraÂtov alium interloc.v. e gwÁ gaÁ r w Sw kratev soiÁ suggigno± A menov polla kiv e xeplhttoÂmhn akouÂwn. resp. Procop. Cyn. taÁ meÁ n a skh sw kaiÁ diw xomai. Or. Ps.2. F -gin.145.28. textum praeb. Westerink.134. Schol.Chr. a8±c2 cf.Porph. ind. twÄ n d' ue wn oi v tauÄ ta a3 sq.8. Iul. 232.Luc. 4a ˆ 1.82. Philostr. A poiÄ ± agnoeiÄ te] anqrwpoi agnoeiÄ te D.11 Bidez. Demetr. Aristid.Luc.232.6 Downey±Norman.3±135. Them.7 de Bude . D. Foerster. Epict. Lib.26 ˆ 267. 18. D. qeoiÄ v F corr.12 sq. b1±c4 cf. 112.  kouÂoiv an. Ps.-Plu. ps. 2 kth sewv ante pe ri add. (disertim): i w 'nqrwpoi Epict. Tim. mhcanhÄ v w Them.: paÄ san poieiÄ sqe spoudhÁn Ps. taÁ deÁ jeu xomai kataÁ kraÂtov.): pwÄv susp. Procop. Iupp. 4a ˆ 1. a7 resp. 4e ˆ 1. Epict.3 Nock.11±12 Macleod. (b)) w nqrwpoi w D.Chr. Schol.Chr. Them. Lib. 36 ˆ 71.5±8 Paton±Wegehaupt. 7.2.4 Flor. Or. b1 cf. ind. AF soiÁ AD: om.-Plu.123 ˆ 2. Or.11 ˆ 1.: thÁn peripaÄsan spoudhÁ n e cete D: paÄsan thÁ n spoudhÁn e cete F: spoudhÁ n e cete apasan Them. s. (UBM) Schol. (ut vid.Chr. 18 ˆ 4.-Plu.: anqrwpoi D. Or. Epict. kaiÁ del. 13.Herm. D. tauÄta w Ps. Plu.: w a nqrwpoi Ps.-Plu.8.Chr.18±20 Kayser.KLEITOFWN 5 b kaiÁ belti wn. a8 resp. Lib. Iul.Herm.D 6 post polla kiv dist.trag. D.: skhnhÄ v Epict.(SY ): au taÁ Them. a8±b1 cf. Schenkl.85.-Plu. Iul. 4e ˆ 1.9 sq. 320d±321c ˆ 2. hab.: transp.-Plu. cf.189 Mal.: a poÁ D. Them.14±17 ˆ 1. Aen. orat.-Luc.4 sq. Iambl. Demetr.220.1 ˆ 1. Ps.26 sq. Luc. (b) Them.6 b1 post le gwn init. Corp. wnqrwpoi.(AL) 242 .Plu. Ps.Gaz. VA 6. F 2s u mnoiÄ v Baumann: u mnoiv DF: u meiÄ v tragikhÄv] tra A1r AD2p F 2s : u mneiv Ven.Chr.-Plu. oi tinev crhma twn meÁ n pe ri thÁn paÄsan spoudhÁ n e cete o pwv umiÄ n e stai. (UBM) poiÄ je resqe post vocativum transp.Chr. Jul.

once I realise what my bad and good points are. and I thought you put things better than any other. every time you disparaged mankind like a god in a tragedy in your lengthy sermons: `Where are you rushing to. to 243 5 b . I'll devote all my energy to the one. clit All right then. while you couldn't care less if your sons.CLITOPHON Obviously. You know. when I used to keep your company I was often stunned by what I heard from you. you human beings? Don't you know that all your actions are beside the point? It's money you do your very best to get. and I'll avoid the other like the plague. Socrates.

(a) 246a ˆ 2. haer. resp. c6±d2 cf. 21 Mannebach. h geiÄ sqe F A1r Them. 13. 13.(AL2 ) kaiÁ dikai wv D.1. 24.) om.32 Holl± Dummer. oute didaskaÂlouv autoiÄ v euri skete thÄ v dikaiosuÂnhv. Ins. Or.-Plu. Theophr.233.7 Keil. id. (a) 439 cd ˆ 3. Or.3 Pohlenz±Sieveking. kapeita ou deÁ n htton kakouÁv gignome nouv periÁ taÁ crhÂmata.6 Kayser. 69. ei deÁ melethtoÂn te kaiÁ askhto n.: cf.27 ˆ 1.55 ˆ 2.93.(a) ge w Them.: o rqwÄv Iambl.235. c7 resp.: katalei yete Ps.237. oi tinev e xaskhÂsousin kai e kmelethÂsousin i kanwÄ v.25.(SY ): kakwÄv Pa Them. a ll' o rwÄ ntev gra mmata kaiÁ mousikhÁn kaiÁ gumnastikhÁ n u maÄv te au touÁv kaiÁ touÁv paiÄ dav u mwÄ n i kanwÄ v memaqhkoÂtav. Plu. b1±3 cf.(AL) thÁn touÄ podoÁv Plu.KLEITOFWN 5 c 5 paradwÂsete *** opwv e pisthÂsontai crhÄ sqai dikai wv touÂtoiv. 20 [66].Chr. Demetr.16±17 NachstaÈ dt±Sieveking±Titchener.(SY ): te Them.5±10 Pohlenz±Sieveking.32 ˆ 1.Chr.18. Procop.239. (b) 534e ˆ 3. D. 33. Aristid. Epiph. kaiÁ adeljoÁ v a deljwÄi kaiÁ po leiv poÂlesin ame trwv kaiÁ (b1) resp.6 ˆ 236.14 Colonna. Or. Pers. c1 te (et Them. tou toiv om.) e Ficino e dikai wv w Them.(AL) 4 pwÄ v] o mwv Hermann 6 kai toi g(e) Plu.Chr. b1 sq. Plu. D.Chr. 13.172. 13.124. 296 ˆ Aristipp. Aen.70.361.18. Fr.900b Migne ˆ 16. Them.17 ˆ 1. a dhÁ paidei an arethÄv einai tele an hghsqe. 24.-Plu. Porph.23. D 2 arethÄv] -thÄv 3 ei nai om. Protr.70.13 ˆ 1.Them.)] kaÂpeit ' D kakouÁv ADFVa Them.(SY ): kaiÁ ou Them.A1r : -twÄn Pa ei deÁ ] ei te F e xaskhÂsousi DF 7 g' e ti ADFVa: ge ti Pa: ge Them.Chr. PaVa Them. c4±6 resp. 1.1 de Bude . ei per maqhto n. Plu.: mikraÁ jronti zete Ps.9 Keil.19 ˆ 1. all' ou diaÁ thÁ n e n twÄ i podiÁ proÁv thÁn lu ran ametri an. Him. resp. VA 1. Philostr.232. Const. pwÄ v ou katajroneiÄ te thÄ v nuÄn paideuÂsewv oudeÁ zhteiÄ te oi tinev umaÄ v pauÂsousi tauÂthv thÄv amousi av.55 ˆ 2. Iambl.: ameleiÄ te kaiÁ post tou toiv Stephanus (ut vid.14 Haury. Eloc. b5±7 resp. Aristid. 4 paradwÂsete w Them. lacunam statui: ou demi an poieiÄ sqe e pime leian Them.9±13 Pohlenz±Sieveking.: orqwÄv pisthÂswntai D2s Them. D. h melhÂkate D. ou de g' e ti proÂteron umaÄ v autouÁv outwv e qerapeuÂsate. kai toi dia ge tauÂthn thÁ n plhmme leian kaiÁ raiqumi an. DF 6 maqhto n] -o.50 Colonna.(AL) gin. D. (b) 439 c ˆ 3. post dikai wv dist. 85.8±10 de BudeÂ.2 ˆ 3. Plu.(AL) 7 all ' ou w Them. telei an Them. Gaz. 59.124.(a) et (b) 244 .Chr. kapeita (et Them.6 de Bude .17 ˆ 1.12±5 des Places.

but that nevertheless you prove none the better in matters of money. Yet you see that in reading and writing. and cities towards cities 245 5 c 5 . Indeed. if it can be acquired by training or exercise. if it can be learned. won't know how to use it in a just way. music and physical exercise you yourselves and your children have had an adequate education ± and this you regard as a complete education in goodness ±. earlier on you never had yourselves taken care of that way. not when the foot doesn't keep in step with the lyre. So how can you not despise the present education system. that brother behaves towards brother. people to exercise or train them adequately.CLITOPHON whom you'll be leaving it. that is ± or. and why is it that you're not looking for people to put an end to this lack of harmonious breeding? It's actually when people are out of tune with this standard and negligent of it. You don't ®nd them teachers of justice.

F 4 paÂlin au PaVa (txt. ind. kaiÁ deiÄ n e pime leian thÄv nuÄn plei w poieiÄ sqai paÂnt' andra idi ai q' ama kaiÁ dhmosi ai xumpaÂsav taÁv poÂleiv. 1. Iambl. (AL) 8 ti v DF 9 ei dh ADVa: ei de F: ei d ' ei Pa: om.4.15±16 des Places. D 3 e gwÁ w Sw kratev e gwÁ F.86±90 Marcovich e5±8 ˆ Stob. o om. Stob. ei per toÁ nikaÄ n e kouÂsion. 7 arxontov ADF2s F2 Stob. periÁ deÁ toÁ arxoÂmenon e spoudake nai.) 5 adiki a] kaki a Hippol. 8 adiei per ± 8 akou sion om.(a) et (b): prosjero menoi Ast 2 kakaÁ post e scata add.12 Wachsmuth±Hense.(a) et (b) 3 touÁ v adi kouv om.23. D ge om. 3. 246 . post adiki a alium interloc. Protr. D 2m keiÄ n hei naii olim conieci. kaiÁ maÂla a gamai kaiÁ qaumastwÄ v wv e painwÄ . Them. F 2s . textum habet Hippol. Schanz post po leiv ®nem orat. kreiÄ tton e aÄn thÁ n touÂtou crhÄ sin´ ei dh tiv mhÁ d4±8 ˆ Hippol. kaiÁ o tan le ghiv w v otwi tiv mhÁ e pi statai crhÄ sqai. pa lin d' au tolmaÄte le gein wv ai scroÁn kaiÁ qeomiseÁ v h adiki a´ pwÄ v ou n dh tiv to ge toiouÄ ton kakoÁn e kwÁn airoiÄ t' an. ind. alium interloc. prius e gwÁ punctis del. 59. thÄv deÁ yuchÄ v hmelhko tav e tero n ti praÂttein toiouÄton. Plu. F2pl : diajeroÂmenai fere Plu. ai F corr.21 ˆ 80. ou kouÄ n kaiÁ touÄ to akouÂsion. Them.53 ˆ 3. wste e k pantoÁ v troÂpou to ge a dikeiÄ n a kou sion o loÂgov aireiÄ . A. kaiÁ opoÂtan au jhÄiv toÁ e jexhÄ v touÂtwi.58 ˆ 3. (SY ): a rxantov FPaVa Them. (katalambaÂnei A2m est scholium) taÁ v poÂleiv delenda susp. e kwÁn om. qamaÁ A: qamaÄ D: qauÄ ma F 5 meÁ n om. D suppl.22±24.19.13±16 Wachsmuth±Hense. touÄ meÁ n arxontov a meleiÄ n. e8±408b1 ˆ Stob. Iambl. d1 prosjero menai AD: prosje romen. Hippol. Hippol. e5±7 resp. Hippol.'' TauÄ t' oun w SwÂkratev e gwÁ otan a kou w sou qamaÁ le gontov.1. resp. Hippol.233. touÁ v a skouÄntav meÁ n taÁ sw mata.  Httwn ov a n hi jate twÄ n hdonwÄ n.16±17 des Places. 7 D2pl tro pou] lo gou Hippol. Protr.KLEITOFWN d 5 e 5 anarmo stwv prosjeroÂmenai stasiaÂzousi kaiÁ polemouÄntev taÁ e scata drwÄ sin kaiÁ paÂscousin. h F 2m ): osa dh Hippol.(a) drwÄsi DF te kaiÁ Plu. ai roiÄ t '] e roiÄ t ' D: ai roiÄ t ' 6 o v an h i ] wv an hÄ F (h in eiÁ mut. ai reiÄ ] e1 q'] d ' F 2 xumpa sav e reiÄ Hippol. u meiÄ v de jate ou di' apaideusi an oudeÁ di' agnoian a ll' e koÂntav touÁ v adi kouv a di kouv einai. 3. Haer. 59. ind.

because he is overcome by desires. feuding and making war and committing and su¨ering the worst outrages. seeing that to overcome them is voluntary? Therefore in either case it stands to reason that injustice is involuntary. Now. when I hear you say them so often. and when you say that what one doesn't know how to use 247 d 5 e 5 . and that each man should take greater private care. than they do at present. then. you say. Isn't that involuntary. And when again you go on to say that those who train their bodies and neglect their souls do something similar: they neglect the part that is going to rule and devote themselves to that which is going to be ruled. So how would anyone choose such an evil willingly? Well.' As for these things.CLITOPHON without measure or harmony. I am full of admiration and I praise them immensely. but that they are so of their own free will ± yet at the same time you have the gall to say that injustice is wicked and an abomination to the gods. you claim that it's not lack of education or ignorance that makes the unjust unjust. Socrates. and likewise all cities greater public care.

ind. b1 diaÂgein twÄi toiou twi w Stob. Them. o v protr. Tou toiv dhÁ toiÄ v loÂgoiv kaiÁ e te roiv toiou toiv pampoÂlloiv kaiÁ pagka lwv legome noiv. Stob. F Them. Stob. thÁn authÁn dhÁ tauÂthn dikastikhÂn te kaiÁ dikaiosuÂnhn wv e stin le gwn. kaqaÂper ploi ou parado nti taÁ phda lia thÄv dianoi av allwi. 1039de ˆ 6. wv au twv A ut semper gaÁr e pisthÄtai A3m et A5s 4 teleudhÁ F Stob.(SY ) post bi on dist. Plu. A 6 kaiÁ e te roiv ± kaiÁ te om.22±17. F xs om. Stob. ou d' a llwi twÄ n orgaÂnwn ou deÁ kthmaÂtwn oudeni .1.16. b5±c2 ˆ Stob. 10 w siÁ DF 11 mhÂt ' A: mhÂte DF Stob. twÄ i maqo nti thÁn twÄn a nqrwÂpwn kubernhtikhÂn. (AL): twÄ i toi. w v ostiv yuchÄ i mhÁ e pi statai crhÄsqai. tou twi mhÂt' akouÂein mhÂq' oraÄ n mh t' allhn crei an mhdemi an crhÄsqai twÄ i swÂmati kreiÄ tton h o phiouÄn crhÄsqai´ kaiÁ dhÁ kaiÁ periÁ te cnhn w sauÂtwv´ ostiv gaÁ r dhÁ mhÁ e pi statai thÄi e autouÄ luÂrai crhÄ sqai. h n dhÁ suÁ politikhÁ n w Sw kratev e ponomaÂzeiv pollaÂkiv. w v didaktoÁ n a rethÁ kaiÁ pa ntwn e autouÄ deiÄ ma lista e pimeleiÄ sqai.240.10±14 Wachsmuth±Hense.9 Pohlenz±Westman ˆ Chrysippus SVF 3.: gaÁr an D: gaÁr an dhÁ AF 2m taÁ F 7 post praÂttonti rasuram duobus signis repletam hab. tou twi toÁ agein hsuci an thÄi yuchÄi kaiÁ mhÁ zhÄn kreiÄ tton h zhÄn pra ttonti kaq' autoÂn´ ei de tiv ana gkh zhÄn ei h. scedoÁn out' a nteiÄ pon pw pote ou t' oimai mhÂpote u steron anteiÂpw. dikaiosu nh Richards 5 post le gwn ®nem orat. dhÄlon w v oudeÁ thÄi touÄ gei tonov. twi Gasda 3 twÄn a nqr. incertum quid antea habuerit: mh pote u peÁ r w n Stob. . 408a1 kaiÁ dhÁ kaiÁ ] kaiÁ dhÁ Stob.60 ˆ 3. 2 paradido nti Them. 12 mhdemi an mhÁ crhÄsqai Stob. om. douÂlwi ameinon h e leuqe rwi diaÂgein twÄi toiouÂtwi toÁn bi on e stiÁ n a ra. A ei h om. b1±5 ˆ Stob.6. . pagkaÂllwv F c2 mh pot ' e v u steron F2pl .13±16 Wachsmuth±Hense.25. kaiÁ teleutaÄi dhÁ kalwÄ v o loÂgov outo v soi.KLEITOFWN 10 408 5 b 5 5 c e pi statai o jqalmoiÄ v crhÄsqai mhdeÁ wsiÁ n mhdeÁ xuÂmpanti twÄi sw mati. Them. Schanz te ADFVa: te gaÁr Pa: deÁ Richards 248 . F add. Pa 4 dhÁ ] deÁ F Stob. h n dhÁ w SwÂkr.d. protreptikwtaÂtouv te hgouÄmai kaiÁ 408a5±7 resp.761.8 ˆ 4.2. et ®nem facit Stob. oudeÁ thÄ i e autouÄ . oudeÁ ostiv mhÁ thÄ i twÄ n a llwn. 4. dikastikh . polit. 3.

Socrates. nor will I in the future. and handing over the rudder of his thinking to somebody else. And it's the same with technical abilities: a man who doesn't know how to use his own lyre will obviously not be able to use his neighbour's lyre either. to leave his soul idle and not to live is better than to live according to his own lights. and which you claim is precisely the same as judication and justice. I suppose. if someone doesn't know how to use his eyes or his ears or his whole body.CLITOPHON would be better left alone ± thus. he is better o¨ spending his life as a slave rather than as a free man. I don't think I've ever said a word against them. I regard them as very suitable for ex- 249 10 408 5 b 5 c . and a man who can't use someone else's lyre won't be able to use his own ± or any other instrument or possession. so numerous and so beautifully formulated. for such a person not to hear or to see or to make any other use of his body is better than to use it no matter how. These speeches and others of the kind. that goodness can be taught and that of all things one should care most for oneself. who has learned the art of steering human beings ± this art which you often call politics. And so this argument brings you to a ®ne conclusion: for a man who doesn't know how to use his soul. and if he must live at all costs.

69. Sa¨rey±Westerink. `` W be ltistoi'' e jhn ``umeiÄ v. omologhÂsantav touÄ t' autoÁ anqrw pwi prakte on einai.Plat. e panerwtwÄ n ou ti seÁ toÁ prwÄton w Sw kratev. e pexelqeiÄ n deÁ ou k on twÄ i praÂgmati kaiÁ labeiÄ n autoÁ tele wv. ka peita w nei dizen le gwn w v ai scroÁn purwÄn meÁ n kaiÁ kriqwÄ n kaiÁ ampe lwn e pime leian paÄsan poieiÄ sqai. wsper an ei tiv hmaÄv prouÂtrepen touÄ sw matov e pime leian poieiÄ sqai. ti tou nteuÄ qen. D: toÁ n Pa 5 ou ti ] o tiÁ v (ut vid.) F2m se ADF 6 7 post onomaÂzein lac.) 2 FVa touÄ t '] to t ' F e perwtaÄn AD e1 ti om. Poll.9.A 1r .a.26 Wachsmuth±Hense. pwÄ v arcesqai deiÄ n jamen dikaiosuÂnhv pe ri maqhÂsewv. Schanz (vel post tou tou lac. kaiÁ atecnwÄv w sper kaqeu dontav e pegei rein hmaÄ v. Theol. suÁ n e piqumhtwÄn D e te rwn F -ai.c. 3. wv o ntov moÂnou touÂtou. (ut vid.83. susp. 4 toÁ AFVa: om. allaÁ twÄ n hlikiwtwÄn te kaiÁ sunepiqumhtwÄ n h e tai rwn swÄn. D 1pl 3 prouÂtounteuÄ qe F dikaiosu nhv ] ti nov Gallavotti peri D -e .17 sq.KLEITOFWN 5 d 5 e 5 wjelimwta touv. Procl. -ai. Hermann 4 deÁ delendum susp.405. h deiÄ toÁ n Swkra th kaiÁ a llh louv hmaÄ v toÁ metaÁ touÄ t' e panerwtaÄn. kaiÁ o sa touÄ swÂmatov e neka c3 sq.23 ˆ 3. kaiÁ  kataÁ seÁ troÂpon tinaÁ u potei nwn au toiÄ v. touÂtwn gaÁ r tou v ti ma lista einai doxazome nouv upoÁ souÄ prwÂtouv e panhrwÂtwn. 3. cf. D 2 om. pwÄv pote nun a podeco meqa thÁn SwkraÂtouv protrophÁ n hmwÄ n e p' arethÂn. Fxpl SwkraÂthn rouv] -e . mhdeÁ n pronoouÄntav o rwÄ n kaqaÂper paiÄ dav w v e stin tiv gumnastikhÁ kaiÁ i atrikhÂ. h o pwv deiÄ proÁ v seÁ periÁ au twÄ n toÁ toiouÄton o nomaÂzein. kaiÁ e kei nouv au e te rouv. all' hmiÄ n paraÁ pa nta dhÁ toÁ n bi on e rgon touÄ t' e stai. c6 cf.64 ˆ 3. touÁv mhÂpw protetramme nouv protre pein. 3. e1±410b3 ˆ Stob. punqanoÂmenov ti v o metaÁ tauÄ t' ei h lo gov.D2 7 kaiÁ o sa ± 9 te cnhn trepe F 5 wnei dize F 6 ampelwÄn D corr.F2s Schanz 9 e panhro mhn PaVa d1 upo tinwn F 2 nun scripsi: nuÄ n w: secl.10±407. D add. proseiÄ con dhÁ toÁn nouÄn toÁ metaÁ tauÄta w v akousoÂmenov. D verbis 4 mhdeÁ n ± 5 i atrikh iterum positis 250 . statuendam putat) deÁ AF d ' D o n scripsi: e n w: e ni Bessarion: e hstiin olim conieci tele wv auto Va 5 par' apanta D 6 e kei noiv Va e te 7 deiÄ ] deÁ F corr.

' I said. and is it impossible to pursue the matter any further and grasp it completely? Is it to be our lifelong duty to exhort those who have not yet been persuaded by exhortation and theirs in turn to exhort others? Isn't this rather the time to ask Socrates and each other. after a fashion. I imitated you. as if we were mere children. and he had noticed that we hadn't the faintest idea that there is such a thing as physical training and medicine. `now. and I asked them what issue was coming next. since we have agreed that goodness is man's very duty. `My excellent friends. barley and vines.CLITOPHON horting people and very useful ± they simply wake us up from our sleep. So I paid close attention in the hope that I would hear what was coming next. what comes next? What do we say is the way to start learning justice? Suppose someone were exhorting us to care for our bodies. in what way do we understand the exhortation to goodness that Socrates is addressing to us? Is it all there is. Those among them who you think are really good I questioned ®rst. in hinting at the answer. If he were then to reproach us and say that it is disgraceful to devote all one's care to wheat. but to some of your contemporaries and your fellow-aspirers or comrades or whatever that sort of relationship to you is to be called. Socrates. and to all other 251 5 d 5 e 5 . I did not put my questions to you ®rst.

) corr.: tauÄtaÁ F 6 tektonikhÄv Stob. hnper a kou hv paginae (del. 2r 3 e panh kein F -ktouÄ to jameÁ n w ou tov] outwv A (ut vid.A 4 : e stiÁ DF 2 ai eiÁ AD (aeiÁ et Stob. D in ®ne nem orat.2 de BudeÂ. ti touÄ to jamen. o du natai poieiÄ n hmiÄ n e rgon o di kaiov. b1 pou ti v AD: pouÄ tiv F e stin A -iÁ .' eipen an i swv o ti gumnastikhÁ kaiÁ i atrikh .KLEITOFWN 10 409 5 b 5 c diaponouÂmeqa te kaiÁ ktw meqa. 3 lege sqw om. toÁ deÁ di dagma. o dhÁ le gomen ugi eian.'' outov meÁ n wv  oimai toÁ sumje ron a pekri nato. 8 e keiÄ touÁv] e kei nouv AF2s toÁ A1r c1 7 toÁ om. F 1pl 409a1 einai ti 2 ei peiÄ n Stob. ei d' e panhro meqa toÁ n tauÄq' hmaÄ v protre ponta `Le geiv deÁ einai ti nav tauÂtav taÁv te cnav. toÁ deÁ ugi eian´ e stin deÁ touÂtwn qaÂteron ouke ti te cnh. thÄ v dhÁ dikaiosuÂnhv wsauÂtwv toÁ meÁ n dikai ouv e stw poieiÄ n. m. D 2 Stob. D2 ) Stob. eipo ntov de mou ``Mh moi toÁ onoma moÂnon ei phiv. D. F 2 : e panhÄ a Schanz 252 . lege sqw. thÄ v te cnhv deÁ thÄv didaskouÂshv te kaiÁ didaskome nhv e rgon. kaiÁ nuÄ n dhÁ ti na jameÁ n einai thÁn e piÁ thÄi thÄv yuchÄ v arethÄ i te cnhn. hnper a kou eiv suÁ le gontov e jh SwkraÂtouv. ind.) post lege sqw ®5 eipen] ei.19 ˆ 1. et D et D 2 ): eipe F a llon F2s 7 d ' e mouÄ D (de mou et Stob. ei pe . kaiÁ tektonikhÄv deÁ kataÁ tau taÁ oiki a te kaiÁ tektonikhÁ toÁ meÁ n e rgon.) 3 u giei an F e sti DF 4 thÄ v deÁ te cnhv Va 5 u giei an F tau taÁ A: tauÄ ta D Stob. A2 extra lin. Va (add. A add. tina Stob.Chr.235. ouk allhn h dikaiosuÂnhn. kaiÁ tauÄta ou san. touÂtou d' au touÄ mhdemi an te cnhn mhdeÁ mhcanhÂn opwv w v be ltiston e stai toÁ swÄ ma e xeuri skein. oblitt. e terov deÁ 409c2 resp. toÁ meÁ n iatrouÁv aeiÁ proÁv toiÄ v ousin e te rouv e xergaÂzesqai. kaqaÂper e keiÄ touÁ v tecni tav e kaÂstouv´ toÁ d' e teron. sec. man. 13. i. dh nav F: ei nai tinav A: einai tinaÁv D (corr. tau thv d' e stiÁ n dittaÁ taÁ apotelouÂmena.) ei phv moÂnon Stob. F Stob. all ' DF w de A: wde F: w deÁ D: w di taÁ om. 9 toÁ swÄ ma del. A 4 au toÁ n F -twÄ F2pl moi om. i atrikh pou tiv le getai te cnh. Stob. A d ' DF: dhÁ A Stob. Baumann 10 e panhrwÂmeqa F corr.'' `O dhÁ dokwÄn autwÄn e rrwmene statov einai proÁv tauÄta apokrinoÂmenov ei pe n moi tauÂthn thÁ n te cnhn ei nai. a llov deÁ toÁ de on. 6 e jh oblitt.bis scrips. allaÁ w de.

' The one who was thought to have the sharpest brain gave an answer to the question. too. as in the other cases the various craftsmen. let one product of justice be to make just men. ``Physical training and medicine. was none other than justice. he said. even though there is one. too. `you hear Socrates talking about'. one to produce always new doctors in addition to those that are already there. a third `the bene®cial'. the other health. He told me that this art `which'. I said: `Don't give me just its name. what do we say that is? Tell me. there are the house and carpentry along the same lines. another `the ®tting'. what do we say is the art which presides over the goodness of the soul? Let's have an answer.CLITOPHON things we try to acquire at great cost for the sake of the body. presumably. but a product generated by the art which teaches and is taught ± that which we call health. in this case. the product which the just man is able to make for us. but do it this way. and another `the pro®table'. not to try to ®nd an art or any means whatever to achieve it. There is of course an art called medicine.' This man. what do you say these arts are. then?'' he would say. But the other thing. 253 10 409 5 b 5 c . Its e¨ects are twofold. the other the teachingmatter. and if we asked the man who was exhorting us to this ``Well. In the case of carpentry. Now the second of these is no longer an art itself. yet when it comes to the best possible condition of the body. the one is the product. replied `the useful'. I think.'' Well. Likewise.

wje lima kaiÁ talla taÁ toiauÄ ta´ allaÁ proÁv oti tauÄta taÂnta tei nei. 4 komyoÂthta Va 5 po lesi DF 6 ou tov] t '] ti F fortasse recte 7 post jili av add. wste taÁ xuÂlina jh sei skeuÂh gi gnesqai. 13. Protr.) nerwtw menov F2m e2 ou deÁ ] ou D 5 a nerwtw menov F le gei Va 7 -wÄ(n) fecit F 2pl o modoxi ai om. 7 h ] ti A2 Va d1 juÂsei A2s (vel A5 ) e sti F 2 wsau twv ± 4 dikaiosuÂnhv om. Stob. taÁ v deÁ twÄ n pai dwn jili av kaiÁ taÁv twÄ n qhri wn.3 de Bude . Iambl. toÁ kalwÄ v.19 ˆ 1. oion h tektonikhÁ toÁ eu.c.'' TeleutwÄn apekri nato tiv w Sw krate v moi twÄn swÄ n e tai rwn. o ti touÄt' ei h toÁ thÄ v dikaiosuÂnhv i dion e rgon. 4 kakeiÄ A: kakeiÄ DF (-a-) D2 : e keiÄ Stob. Va 8 w mol. 59. D 5 talla F om. lege sqw dhÁ kaiÁ toÁ thÄ v dikaiosuÂnhv w sau twv. e4 resp. ov dhÁ komyoÂtata e doxen eipeiÄ n.?) 8 qhri wn] e tai rwn H. e reiÄ toÁ i dion e kaÂsth te cnh. outov d' au e rwtwÂmenov thÁn jili an a gaqoÂn t' e jh einai kaiÁ ou de pote kako n.235. D. taut ' A -uÄ . o deÁ toÁ lusitelouÄn. jili an e n taiÄ v poÂlesin poieiÄ n. MuÈller 9 auton F (ut vid. pa ntwv a gaqhÁn Va 254 . D post teleutwÄ n alium interloc. a dhÁ ouk e stin te cnh. thÁn meÁ n o modoxi an hti mazen´ hnagkaÂzonto gaÁr pollaiÁ kaiÁ blaberaiÁ gi gnesqai omodoxi ai anqrwÂpwn. e pa-ov A1r (outws a. jeuÂgwn dhÁ toÁ toiouÄton oudeÁ jili av e jh taÁv toiauÂtav einai. ouk apede ceto ei nai jili av e panerwtwÂmenov´ sune baine gaÁr autwÄi taÁ plei w taÁv toiauÂtav blaberaÁv h agaqaÁv einai. toÁ deoÂntwv. 6 e ka sth h Stob. yeudwÄ v deÁ onomaÂzein au taÁv touÁv ou twv onomaÂzontav´ thÁn deÁ ontwv kaiÁ a lhqwÄv jili an einai saje stata o moÂnoian. e n om. orqwÄv praÂttein.KLEITOFWN 5 d 5 e 5 toÁ w je limon.Chr. thÁn deÁ omo noian erwtw menov ei omodoxiÂan ei nai le goi h episthÂmhn. o twÄ n allwn ou demiaÄ v. ind. e panhÂiein d' e gwÁ le gwn oti ``KakeiÄ ta ge o noÂmata tauÄt' e stiÁ n e n e ka sthi twÄn tecnwÄ n. D 2m swÄ n Va o v] wv Stob. F 3 twÄ n e tai rwn twÄ n suppl.17±20 des Places. thÁn deÁ jili an agaqoÁn w mologh kei paÂntwv einai kaiÁ dikaiosu nhv e rgon´ d1±e10 resp. lusitelouÄnta.A 2 e stin A corr. av hmeiÄ v touÄto tounoma e ponomaÂzomen.

and that those who call them that do so wrongly.' Finally.CLITOPHON So I retraced my steps and said: `In the other area. proper and appropriate action. he rejected unity of opinion. So let me have a similar answer about the distinctive trait of justice. aiming at. for he said that there must necessarily be many harmful cases of unity of opinion among men. for he was forced to the conclusion that they were more often harmful than good. Socrates. Real and true friendship was in actual fact concord. Upon further questioning he declared that friendship was good. the production of wooden equipment. as its distinctive trait. these names play a part in each of the arts. which of course isn't art itself. bene®cial and so on. For example carpentry will mention right. When he was asked whether by concord he meant unity of opinion or knowledge. What we call the `friendships' of children and animals he didn't admit to be friendships when he was asked about that. He said that the proper product of justice and of no other art was to achieve friendship in cities. In order to avoid that he claimed that they weren't friendships at all. but the aim of all these actions will be stated by each of the arts individually. doing what's pro®table. acting correctly. she will say. never bad. while he had already admitted that friendship was good and the 255 5 d 5 e 5 . too. one of your comrades gave me an answer which was thought the smartest.

: se autoÁn D (deÁ D2 ) kaiosuÂnhn Va b2 ou deÁ n´ apanta Stob. ind. kaiÁ periÁ twÄ n allwn tecnwÄ n w sauÂtwv. kaiÁ periÁ o tou eisiÁ n e cousi le gein´ thÁn deÁ upoÁ souÄ legome nhn dikaiosuÂnhn h omo noian o poi tei nousa e stin diape jeugen. paÂntav] pantoÁv Bury 3 deÁ ] gaÁr F 4 kaiÁ del. kaiÁ a dhlon authÄv oti po t' e stin toÁ e rgon. wv ou maÄllon o nti dikaiosuÂnhv 9 ei nai´ D kaiÁ (et Stob. ou do xan ou san Ge¨cken 10 post do xan ®nem orat. alium interloc. A hmen] h meÁ n D aporouÄ ntev delendum putat Schleiermate A: teÁ DF: om. kaiÁ eipe v moi dikaiosuÂnhv einai touÁv meÁ n e cqrouÁv bla ptein touÁ v deÁ ji louv eu poieiÄ n. F add. A4 ) F Stob. e pecei rhsan A2m 3 h om. wv pollouÄ toiÄ v anqrw poiv axiÂa. A. ind. tau toÁn dhÁ kaiÁ soi tiv e pene gkoi taÂc' an periÁ dikaiosuÂnhv. D2pl c2 soiÁ ti v DF 3 8 oion om. o ge noit' a n kaiÁ periÁ allhn hntinaouÄn te cnhn. a porouÄntev. u steron deÁ e jaÂnh blaÂptein ge oude pote o di kaiov oude na´ pa nta gaÁr e p' w jeli ai pa ntav draÄ n.KLEITOFWN 10 410 5 b 5 c wste tau toÁn e jhsen einai omo noian. ind. ou san Bury: omo n.'' TauÄ ta w SwÂkratev e gwÁ teleutwÄn kaiÁ seÁ autoÁn hrw twn. nomi sav se toÁ meÁ n protre pein eiv arethÄv e pime leian kaÂllist' anqrw pwn draÄ n. diseautoÁ n A (corr. Va cher 410a1 i kanoiÁ h san] gr. TauÄ ta deÁ ou c apax oudeÁ diÁ v a llaÁ poluÁn dhÁ upomei nav croÂnon [kaiÁ ] liparwÄn a pei rhka. D 7 8 ei paÂv Stob. kaiÁ e pist. Bekker: post ei nai transp.) del. w jeli ai A -ei . duoiÄ n deÁ qaÂteron. o te dhÁ e ntauÄqa hmen touÄ loÂgou. h tosouÄ ton moÂnon duÂnasqai. Baumann omonoi ai wv e pist. all' ou doÂxan. kaiÁ e legon oti ``KaiÁ h i atrikhÁ omo noia ti v e stin kaiÁ a pasai ai te cnai. e sti DF -jeuge F 6 pot ' e stin A (e stiÁ n A4 ): pot ' e sti F: pote e sti D post e rgon ®nem orat. [kaiÁ ] e pisthÁmhn ousan. Ast: kaiÁ dikaiosu nhn Hermann: kaiÁ jili an Bertini.A4 DF Stob. oi paroÂntev ikanoiÁ hsan e piplhÂttein te au twÄ i kaiÁ le gein o ti peridedraÂmhken eiv tautoÁ n o loÂgov toiÄ v prw toiv. F2s onti ] o ti D 256 . D (habet Stob.) o moÂnoia w e sti F 5 o poi Bekker: opou w (et F) Stob. Baumann liparwÁ n A se ADF: seÁ A4 : deÁ Va toÁ ] -w Dxs 6 de F nomissase D corr. D 2pl=s katamelehÄsai D corr. makro teron deÁ oude n. oi on mhÁ o nta kubernhÂthn katamelethÄ sai toÁ n e painon periÁ authÄ v.

but one of two things must be true: either you can do only that and nothing that goes any further ± which could also happen in the case of any other art. yet they are able to say what they are all about. being knowledge and not opinion. and it is totally unclear what its product is. for example without being a steersman one might train oneself in making eulogies about how valuable the steersman's trade is for mankind. The very same complaint might perhaps be lodged against you in the ®eld of justice ± people might say that you are 257 10 410 5 b 5 c . I think you are better than anybody else at exhorting people to care about goodness. As we had arrived at this point in the argument and didn't see a way out. I asked you the questions yourself. But this justice or concord of yours hasn't the faintest idea what its aim is. the bystanders were enabled to get at him because the argument had come full circle and got back to where it had started. and likewise for the other arts. and you told me that it was a typical property of justice to harm enemies and bene®t friends. His conclusion was therefore that concord was the same thing. too. They said: `Medicine is a kind of concord. Socrates. it turned out that the just man never harms anyone. Later.CLITOPHON product of justice. This I had to endure not just once or twice but over quite a long period. I have now given up persisting. however.' That's why at long last. as all the arts are. as all he does to everybody is to their bene®t.

F suppl. aporwÄ n´ e peiÁ ei g' e qe leiv suÁ touÂtwn meÁ n hdh pauÂsasqai proÁ v e meÁ twÄ n loÂgwn twà n protreptikwÄ n.7 sq.) F 2 w v scripsi: a ei rhke nai om. i na mhÂ. hv e neka ta lla diaponouÂmeqa. tau thv hmelhke nai´ kaiÁ talla pa nta oi ou me nuÄn outwv eirhke nai taÁ touÂtoiv e xhÄ v. (te lov PlaÂtwnov add. resp.Chr. toÁ e jexhÄv an twÄi protreptikwÄ i loÂgwi e legev oi on toÁ swÄma mou ju sei on oi av qerapei av deiÄ tai.KLEITOFWN 5 d 5 e 5 e pisthÂmoni. Hermann: tautoÁ n gi gnoit ' an H. ind. yuchÄ v de . 13. ou mhÁn to ge e moÁn outwv e cei. taÁ meÁ n e painwÄ se proÁv Lusi an kaiÁ proÁv touÁ v allouv. oion deÁ ei periÁ gumnastikhÄ v protetramme nov h touÄ sw matov deiÄ n mhÁ a meleiÄ n. kai sou deoÂmenov le gw mhdamwÄ v allwv poieiÄ n. dio ti kalwÄv au thÁn e gkwmiaÂzeiv. alium interloc. D2s 6 a nqrw pwi ] anwqen F2s subscr. F 2m w 3 nuÄ n diexhÄlqon Va nuÄ n dhÁ ADF/ kaiÁ souÄ w 4 e painwÄ] -wÄ erasum D 5 d ' e ti F mhÁ meÁ n] mh me D corr. D. D) KleitojwÄn h protreptikoÂv AD: KleitojwÄn F 258 . 7 oi mai A3m poreu somai A2s F2s o poi Bekker o ph(i ) w 8 ge qe leiv e j ' e xhÄv D swÄ ma DF 4 o n] on F d2 h] h D 3 toÁ 1 ] twÄ FVa D tau toÁ n gigne sqw del. MuÈller 5 post o mol. duoiÄ n deÁ qa teron. kaiÁ nuÄn dhÁ tautoÁn gigne sqw.35 ˆ 1. de Bude . h ouk eide nai se h ouk e qe lein au thÄv e moiÁ koinwneiÄ n. qeÁ v toÁ n KleitojwÄnta omologouÄ nta wv e stin katage laston twÄ n meÁ n allwn e pime leian poieiÄ sqai. protetramme nwi deÁ scedoÁn kaiÁ e mpo dion touÄ proÁv te lov arethÄv e lqo nta eudai mona gene sqai. kaqaÂper nuÄn. DiaÁ tauÄ ta dhÁ kaiÁ proÁv Qrasu macon oi omai poreu omai kaiÁ allose opoi du namai. D e sti DF e1 talla1] -a. 410e7 sq. taÁ de ti kaiÁ ye gw.240. w v kaiÁ nundhÁ dihÄlqon.F kaiÁ ± 2 2 ou twv] o ntwv (ut vid. mhÁ meÁ n gaÁr protetramme nwi se anqrwÂpwi w SwÂkratev a xion ei nai touÄ pantoÁv jhÂsw.

I meant this way. I think. 259 5 d 5 e 5 . the soul. Socrates.CLITOPHON none the more an expert in justice just because you make ®ne eulogies about it. as I've illustrated just now. but for someone who is you're actually almost a stumbling-block for reaching complete goodness and so becoming truly happy. that's not what I think. so that I won't have to do as I do now ± partly praise you before Lysias and others. why I go to Thrasymachus and wherever else I can. to neglect that. that for a man who isn't yet persuaded by your exhortations you are worth the world. That's precisely. I beg of you to do just that. but when it comes to the thing which we go to all the trouble for. and just as. but one of two things must be true: either you know nothing about it. For I will maintain. Believe me that all the other things which I've said. Mind you. If you're prepared to stop these speeches of exhortation to me. or you don't wish to share it with me. but partly criticise you as well. beyond agreeing to that. because I'm at a loss. you would go beyond exhortation and tell me what kind of thing my body is by nature and what kind of treatment it therefore needs ± in this case the same thing must happen. if in the area of physical training I had been convinced by your exhortations that I shouldn't neglect my body. You can take it that Clitophon agrees that it is ludicrous to care for other things.




cit. op. Studi sulla tradizione antica e medievale del Fedone (Rome 1972). 53 G. Even if some of them were original (Aristotle quotes from Mx. ii 195 (pro. 12e2±4). section i. L. h miÄ n: some interpreters have a tendency to deny that in Plato hmeiÄ v can refer to one person only. Together with the o ti clause. Hoerber. a MS going back to a popular edition in antiquity. Jowett±Campbell. 257d3±258a2 touÄ d' [sc. though Plt. ii 514±15.5. Carlini. no examples). 263 . M. Anakoluthe. cf. the content of the dialogue is thus stated straight away. 1415b31. `Protreptikos'. as it includes the patently late Alc. A. which cannot belong to the fourth century. 13 want to supply anh r. is more trustworthy than the complex of title. Here the singular is chosen by Schleiermacher and Susemihl. e ristikoÂv for Euthd. MuÈ ller. it must be noted ± as Philip fails to do ± that the Spuria have no such classi®catory sub-titles).) and the author of the spurious Thirteenth Letter (363a7) use the well-known PeriÁ yuchÄ v for the Phaedo). not lo gov. Phoenix 24 (1970) 296± 308. C. On the sub-titles in general.3.. Cf. 130 n. 139 n. Intr.±P. Reinhard. as it reposes on diai resiv (this is an interesting hypothesis. 29. with the words e n twÄi e pitaji wi. ˆ 23 Pf.2(4).  ristwnuÂmou: the main character of 406a1 KleitojwÄ nta toÁ n A the dialogue (`Discourse Topic') is introduced at the very beginning. 62±3 and Hoerber. both Callimachus (Ep. 77. 1. protreptikoÂv: Hirzel. The division according to characters has been claimed by J. Republic.COMMENTARY Title: The simple KleitojwÄn of F. For the formal manner of address. cf.5. 301±4) to be a product of the fourth century bce. But Philip overlooks the possibility that Thrasyllus or Dercyllidas or someone else applied the diaeretical classi®cation to the tetralogical list. section i. shows that these alternative titles can occasionally serve to designate what was regarded as the main characteristic of the dialogue. Kurzdialoge. Pohlenz.4. W. R. it is improbable that all dialogues had double titles right from the beginning. 1 (contra). Philip (`The Platonic corpus'. G. cf.2 and n. SwkraÂtouv touÄ newte rou] hmiÄ n [ˆ Socrates] h klhÄ siv o mw numov ou sa kaiÁ h proÂsrhsiv pare cetai tina oi keioÂthta proves that it can (see also Euthphr. But e pitaÂjiov for Mx. A. `Thrasylus' [sic] Platonic canon and the double titles'. 2. for the historical person cf. This is characteristic of short dialogues. Phronesis 2 (1957) 10±20. Rh. (previous note). section i. Kleine Schriften (Hildesheim 1965). sub-title and classi®cation found in the more learned MSS D and above all A. similarly. esp.

407a5 soiÁ suggigno menov. 201c2. MuÈ ller. See too on 408d2 a podecoÂmeqa. I do not see how la Magna's claim that taÁv . See Intr.±D. ± The omission of me n in PaVa (against ADF) is perhaps defensible in itself (GP 2 165). But even if this supposition is right. Perhaps it was because of Lysias' connection with some participants in the conversation of Republic 1. Why exactly Lysias' name was chosen is hard to say. 7 329b2. a3±4 u perepainoiÄ : it would be out of character for Socrates to say that Thrasymachus gets more praise than he deserves. . h meta tinov diatribh happens to be absent from Plato's works (diatri bw meta is found Phd. . here it might be either.C O M M E N T A R Y: 406a2±406a3 For general discussions cf. In Plato. section i.2 n. Lg. 243. where he is a silent partner. De Win. . The switch to the singular is necessary because sunousi ai (like omili ai ) denotes relationships to di¨erent people.2. The words do not mean that the criticism was uttered in Lysias' house. starting with Synesius (section i. Ap. H. The PaVa reading can be explained as an error caused by homoiarcton. In Euthd. 303b2 Socrates would fall out of his role if he condemned the jubilations explicitly. a3 sunousi an: this refers to basically the same thing as diatriba v. the word means not `praise too much' but `praise very much' (as also Ar. Ep. i 98±101. 227b10±11. 37d1 and Burnet's note). 680). as many readers have supposed. cf. 79). ± Only with mo nw tugcaÂnomen o ntev (a10) does it become clear that h miÄ n is in fact ˆ e moi . wenn er mit Lysias redet: dieser Zug weist auf den Phaidros' (Platon. 335c3 etc. . cf. Prt. Water®eld) opt for `too much'.5.5. . 59d5.. 5 322a2) is too ambiguous. but here the style seems to me to demand me n. sunousi an `l'insegnamento' can be justi®ed. KG i 83±4. metaÁ Swkra touv diatriba v: while the connection of diatribh with an attributive prepositional phrase is frequent enough. who are named in this very sentence. Ep. 94. Eq. a2 Lusi ai dialegoÂmenov: for the importance of this detail. Phdr. 794c5). Wilamowitz' statement `[Kleitophon] wird dem Sokrates nur ein bedingtes Lob erteilen. sunousi a is used for Socratic conversations at La. ± FVa have u perepaineiÄ . because it could also be taken as lo goi (cf. it could only serve to refer the reader of Clit. Most translators (apart from Ficinus. 1) is far-fetched.3 n. section i. but change 264 . a2±3 taÁv . to Republic 1 if this reader knew of the roles played there by Thrasymachus and Clitophon. Vorlesungen. Schw. diatribaÂv `le conversazioni ®loso®che' is opposed to thÁ n . Wackernagel. 629d8 there is no question at all of Tyrtaeus' unduly having praised war and heroes. . cf. i 386 n.5. Intr. in Lg. But the idea could not have been otherwise expressed: taÁv SwkraÂtouv diatriba v (cf..

). I prefer (2). §283. as in every me n/de complex. 585. 1043b25¨. Arist. 38±9. Richards. G. review of GP 2 . that in such contexts Greek can easily do without an emphasising particle. `Notes on Phaedrus'. Ra. lo gov] o stiv. 157. o stiv (Stanford ad loc. . in a sentence tacked on to Socrates'. 556). . h dh ei rhtai. Ar. Met. J. So usually the older translators from Ficinus onwards. note on 408c6 o pwv deiÄ ) as in R. but it may contribute to characterising Clitophon's unabashed attitude towards Socrates in that he comes straight to the point in his very ®rst words. also Grg. Phdr. ou k e sti toÁ ti e stin ori sasqai . . a5  O stiv: there are two di¨erent ways of explaining the construction: (1) assuming an ellipse of hn with o stiv for o stisouÄ n (for the latter cf. w v kentaurikwÄ v e nh laq'. a8 prospoiouÂmenov de : although. there are some cases where in practice the me n-clause carries the weight and the de -clause is hardly 265 . cf. MuÈ ller (reading ti v in the previous sentence). W. 240d6±7. It must be noted. pollaÂkiv . cf. unnecessarily assumes an aposiopesis. 62a1 given by Verdenius. 353c5 htiv hn d' e gwÁ au twÄn h areth (cf. On balance. cf. 508d5 o deÁ dhÁ e moÁv [sc. cf. o stiv hn Hermann. `Digest'. Ti v thÁn qu ran e paÂtaxen. and most twentieth-century translators. la Magna. Platonica. allaÁ poiÄ on me n ti e stin e nde cetai kaiÁ dida xai (`but what you can do is to make clear the quality'). Verdenius. the two clauses are presented as parallel. and recently Orwin and Gonzalez. PCPhS 166±8 (1937) 2 and `Notes sur le texte de Platon'. Mnem. 250±1 belong here (the interpretation of kai at Phdr. So H. Such a lively idiom is not in keeping with the author's style in general. 238d6 and Phd. o sho vi tiv (`your somebody') R. (2) Taking the o stiv clause as an addition to the sentence spoken by Socrates (`Someone who did not give you the right story'): o stiv is the usual relative after tiv. Tucker ad loc. At any rate there is no need to assume corrupted transmission (lacuna Schanz.C O M M E N T A R Y: 406a5±406a8 from optative to indicative is very rare in reported speech in Plato (Riddell. oi d' e n twÄ i astei anqrwpoi. KG ii 363. . 273 (cf. 3. A good parallel is Hdt. Dover compares elliptic ei per `if at all'). a7 taÁ deÁ kaiÁ eph inoun: one of the clearest examples of kai stressing the statement that A is true in some cases preceded by the statement that A is not true in other cases (`partly I did praise you'). w v tiv H. `Notes on Plato's Phaedo'. REG 52 (1939) 23±35 at 33).10 ou gaÁ r dhÁ uetai taÁ a nw thÄ v Ai gu ptou toÁ paraÂpan´ allaÁ kaiÁ to te u sqhsan ai QhÄbai yakaÂdi. 230d4±5 taÁ meÁ n ou n cwri a kaiÁ taÁ de ndra ou de n m' e qe lei dida skein. however. Bury. 11 (1958) 193±243 at 197) supposes that kai combined this value with that of `still' ± I would rather believe that `still' is not expressed in Greek in such cases). add Mx. Some of the examples quoted in GP 2 321±3.

. a9±10 e peidhÁ kai : `the more so because'.g. 454b9±c1 all ' i na mhÁ qauma zhiv e aÁ n kaiÁ o li gon usteron toiouÄ toÂn ti se e teron ane rwmai. Chrm. ei si d3). . `also'. Cf. Tht. e n e moiÁ e stw o ki ndunov´ w v e gwÂ. cf. This use of kai is perhaps to be explained as due to an inversion. Ð KaiÁ paÂnu ge e jh o Kriti av e pei toi kai e stin jilo sojov (second reason). ± mh for mhde n (PaVa) is barely possible: ou de n with jronti zein is very frequent in Plato. as appears from the sentence-®nal position of au to v. 154e6±8 paÂntwv gaÂr pou thlikouÄtov wn hdh (®rst reason) e qe lei diale gesqai. kai is simply the Focus particle (cf. cf. not Socrates' dissimulation. Stallbaum on Euthd. German ja) ± for the distinction between connective. Hp. Cf. 89e2. R. GP 2 370. modal and Focus particles. 153a5 (®rst reason: Homer's authority. Bluck on Men. possibly Smp.). GP 2 370. 303±7.15. If so. This idiom is to be distinguished (as GP 2 297 fails to do) from the far more common use of kai stressing the familiarity or self-evident character of the facts mentioned in the e pei (e peidhÂ) clause. Grg. X. cf. 5. Eng. not rhetorical diction which is aimed at. the e¨ect is well explained by la Magna: `contrapposto all'ignoto.C O M M E N T A R Y: 406a9 more than a concession: the reason why Clitophon is eager to explain his remarks is only Socrates' blaming him. but (e. 3. 288c4 (®rst reason pwÄ v gaÁr an . 464a). Cf. . e pecei rei deÁ periÁ au taÁ maÂcesqai. 289c. 612d7 (®rst reason e peiÁ dhÁ . In most other instances the e¨ect is rhetorical. prospoiouÂmenov de is therefore roughly equivalent to kai per prospoiouÂmenov (so nearly all translators). 91e. which is quite rare in Plato. Headlam±Knox on Herod. Bluck ad loc. At Tht. . Many examples from the Letters and Laws are quoted by Novotny on Ep. 188e4. mhde n: mh negates the in®nitive after prospoieiÄ sqai. e peidhÁ kaiÁ presbu thv ei mi (`since I am. KG ii 232±3. 7 343c5.) R. after all. a9 au touÁv au to v: the juxtaposition is intentional. 266 . . Here we have rather to do with kai as a modal particle (cf. 448c7±8 a lloi allwn allwv is parody. an old man') parakinduneuÂein e toimov. Grg. e gwÁ d' e panerwtwÄ (`which seems obvious in spite of my persistent questions'). especially when the antithesis is preceded by a negative (cf.18. kai qualifying the e pei clause (which it cannot precede). o dokeiÄ meÁ n dhÄ lon ei nai. kai expresses that a second reason is being adduced (the ®rst was Socrates' indignation). R. mhÁ kaloÁn ei nai c2±3). Thompson on Men. Euthd. `too'). a1±3). che li ha riferiti ouk o rqwÄv'.Ma. 197b11 and c5 it is precise. Conditions and Conditionals. 71d7±8 (cf. 603b4 jau lh ara jauÂlwi suggignome nh jauÄ la gennaÄ i h mimhtikh is not. Wakker. . 285b7±c3 ei deÁ umeiÄ v oi ne oi jobeiÄ sqe. 342d2±3 sunwmolo ghse meÁ n kaiÁ tauÄta teleutwÄ n. . The opposite (the de -clause bears the weight) is more frequent. Hipparch. Men. Grg.

. c6 kai . a te etc. But after a gaÂr clause ou n not dh is the normal particle to pick up the inter- 267 . 1 118b5. e peidh ge mo nw e smeÁ n (236c11±d2). 187a6 au toiÁ eu retaiÁ gegono te. consecutive clauses. It is questionable whether the latter can occur outside certain well-de®ned contextual groups (relative. 34c8 auqade steron a n proÂv me scoi h. Ly. Verdenius. I do not see why this use `seems strange at best' (Heidel. also Alc. Bury reads de but translates `so'. 64±5. following demonstratives and certain adverbs such as dio . Pseudo-Platonica. ei de : dh (F) is only apparently better than de (AD): Clitophon repeats what he had said in a9. 8). ei me keleuÂoiv a podu nta o rchÂsasqai. 3±4. review of GP 2 . Vendryes. Adam on R. AusfuÈ hrliche Grammatik der griechischen Sprache. W. Verdenius±J. De Vries applies the word also to adverbial kai (Focus or modal particle) with a similar nuance. W. De Vries on Phdr. e3 kai . a10 mo nw tugca nomen o ntev: both o nte and o ntev are possible (for the plural. 48 n. 303c4 makaÂrioi sjw ). `semi-consecutive' or `conclusive' kai . 36±7. H. a11±12 proÁv e me : the reading of A. cf. 212a2 oi oi t' e ston. Erster Teil: Elementarund Formenlehre (Hanover 1890±2). cf. Traite d'accentuation grecque (Paris 19452 ). Index s. R. o ntev is read by AF and (it seems) D. 270. Blass.. 236d6. i scuro terov d' e gwÁ kaiÁ new terov (Phdr. Ap. a similar phrase has more point: wste kan o li gou. 236c8±d1). Lg. carisai mhn a n. kai Рconsecutive). In a similar situation in the Menexenus. 343e3 mocqhrote rwv. where (in other words) the consecutive link between the two coordinated clauses is not expressed in Greek. 275. J. `Notes on Phaedrus'. 227c7 (and cf. cf.v. touÄ de ontov and what follows. La. a11±12 proÁv e meÁ e cein tracute rwv. J. .C O M M E N T A R Y: 406a10±406a12 This second use is present in most of the Platonic examples quoted by GP 2 (but at Tht. o nte (A 4 D 2 . causal. Cf. Waszink. cf.). Euthd. Cf. ± See on 406a1 hmiÄ n for the justi®cation of the clause. J. Trenker. For other adverbs in this construction. Aristotle on Coming-to-be and Passing-away (Leiden 19662 ). proÂv me (FD) is equally possible. S. i 577. notes on 410b4 kai . and there is no opposition between nuÄ n gaÁ r . Verdenius wisely restricts the term `consecutive' to cases of kai as a consecutive coordinating particle. it is one particular sub-group of what has been called `consecutive'. 250. Le style KAI (Assen 1960 2 ). 922d6±7 o po soi periÁ e meÁ jauÄloi kaiÁ o soi a gaqoiÁ gegoÂnasin (this parallel shows that Orwin's `that I have a low opinion of you' is beside the point). 157a2 and 187b5 kai just means `also'). Phdr. both hands without any authority) is almost certainly a correction rather than the authentic reading. 103. still more poignant is the threat e smeÁ n deÁ mo nw e n e rhmi ai. a12 tracuteÂrwv: cf. id. KuÈ hner± F. a10±11 proÂv seÁ jau lwv e cein: for jauÄlov used for being on bad terms with a person cf.

. L. Socrates radically excludes not listening). i. 240b9 a ll ' e sti ge mhÂn pwv). . `The discourse function of particles: some observations on the use of maÂn/mhÂn in Theocritus'. section i. kataÁ kra tov: for the purport of this 407a1±4 A sentence as well as for its ironical character. G. J. . the only words to separate the combination are ou (ou de . Sicking in Two Studies. the only prose authors who split allaÁ mhÂn at all are Xenophon and Plato ( J. reserves or qualms the partner has raised (cf. cf. . . `Emphasis and a½rmation. Blomqvist. Wakker. occurs in exactly the same context as allaÁ . . 65).. and F is notoriously unreliable so far as particles are concerned. esp. allaÂ' (C. Re examen des emplois de alla aÁ la lumieÁ re de l'eÂnonciation dans Les Grenouilles d'Aristophane'. ou ti ) and the interrogatives except for a few places in the Platonic corpus (this passage. cf. Some aspects of mhÂn in tragedy'.). . a13 h dist' a n dexai mhn kaiÁ eqe lw: for the change from potential to indicative.C O M M E N T A R Y: 406a13±407a1 rupted line of thought. Harder et al. and even in these. Boter. cf. 209±31. readiness to accept a a1 A proposal' (GP 2 342). 481c3±4. 106±7. G. M. which is too small a number for argument. `expressing . `not elsewhere .. mhÂn here in Clit. . Theocritus (Groningen 1996) 247± 63.  ll' . the split form Phd. Tradition of Republic. . the separation of alla and meÁ n dh quoted from Phd. Ion 541a7 a ll ' e keiÄ no mhÁ n dokeiÄ soi (meÁ n TW). `by all means' (Gonzalez).2. 83±9. 960e1 a ll ' e sti mhÁn dunatoÂn. unless' means in fact `three times' (not counting ou etc. in M. and ti v etc. 75±99.5.  ll' ai scroÁn . 252). Gildersleeve i 178.3. J. `rupture discursive'. Grg. Clitophon had given Socrates the choice between listening and not listening to him. In this collocation alla seems to me to have its normal function at the beginning of an answer. GP 2 394±5. Basset. The situation is the same in the case of the more or less synonymous collocations a llaÁ me ntoi (cf. Intr. No argument against authenticity can be made from the fact that Plato does not elsewhere split a llaÁ mhÂn unless it is adversative. Sph. mh n: assentient. 268 . Lg. The basic function of mh n as a modal particle is probably to assert something no matter what the partner in the conversation may think (and thus by implication to preclude possible disbelief. namely to brush aside whatever objections. New Approaches. The split form is somewhat exceptional. 55). C. ou phi. . here. . I think Denniston is wrong in considering allaÁ mh n a combination (with a value of its own) rather than a collocation (each particle keeping its own function): the basic value of mhÂn `explains its a½nity with . . esp. ` A  ll ' e xo loisq' au twÄÎ ko ax. Besides.2(1). Greek Particles in Hellenistic Prose (Lund 1969). 78a10). cf. New Approaches. GP 2 410±2) and allaÁ meÁ n dh (cf. A.

28 (1975) 1±16. Sph.' (Water®eld). a suggestion which is worked out in dhÄ lon gaÁ r ktl. section ii.2(1). [ (. 24e10).'). Socrates suggests that Clitophon's report of his criticism will ameliorate his diatribai . 408c1 scedoÂn. 906e5. 296e7. dhÄ lon oti is absent. cf.. a3 kaiÁ belti wn: kaiÁ (`respectively') is often found in one of two coordinations (or in both).7. (previous note). mhÂn . cf. R. Cf. dhÄ lon wv is not found in those works of Plato which are prior to R. R. 2±10. Pl. Socr. wjeleiÄ n: the passage gains in clarity if one realises that for Plato. cf. note on 408b1 kaqaÂper.. Euthd. 11c D.). (SSR vi a 53). 65. IA 384.C O M M E N T A R Y: 407a2±407a3 Why did the author choose to split allaÁ mhÂn? Certainly not for poetic diction. when a relation between the two ®rst and the second members is implied. cf. to give extra emphasis to Socrates' willingness to listen: `Why. . op. section i. especially in Hp. 255a4). 403). 617e3±4 arethÁ deÁ ade spoton. Mnem.2 and n. (cf. 505c3 ou c u pome nei wjelouÂmenov.) [ Eu ge nhÁ thÁn  H ran le geiv kaiÁ pollhÁn ajqoni an twÄn wjelouÂntwn (see also Hp. note on 409e3±4. though poets do separate these particles. . 408a2).' (the same explanation will hold for the disjunction of alla and meÁ n dh at Phd. GP 2 66.Ma. Rather. Intr. . Ap. at 9). and one of the main functions of ge is to mark focality on a clause-initial constituent. for a lla .Ma.). 23).3. cit. 292a8±11. cf. ge cf. cit. . If proportions are involved the case is entirely different. Grg. as Plato was well aware (Cra. I think. but w jeleiÄ n used absolutely is rare with a personal subject (E. a2 u pomeÂnein: cf. . This is one group out of various unrelated idioms often lumped together as 269 . Cratylus 417c'. . The slightly o¨ended mentor of the opening words now turns humble pupil (wjeleiÄ n is often said of teachers. But the natural explanation of this is that the collocation allaÁ mhÂn attracts words or phrases with focal properties. loc. 441d8. . my paper `The etymology of bou lomai and o jei lw'. . me: omitted in PaVa. fr. In using this word. 47 and n.. in fact. `I call it a favour because . 417c7±8). the verb had replaced o je llein in Ionic-Attic (cf. Ap. 24e4±10 oi de touÁ v ne ouv paideu ein oioi te ei si kaiÁ belti ouv poiouÄsin. to Clitophon: `it would be a disgrace not to accept your o¨er'. h n timwÄn kaiÁ atimaÂzwn ple on kaiÁ e latton au thÄv e kastov e xei. wv: This collocation occurs twice in Clit. Lg. a shame indeed would it be . Aesch. See Intr. . this word implied `to make better'. my paper `Plato. . cf. It may be objected that allaÁ mhÂn is followed by ge more often than not (Blomqvist. . whereas dhÄlon oti is found 131 times (cf. ˆ 4 Kr. ga r: explains w jeleiÄ n (`I say ``help (make better)'' because . and Thg. . . GP 2 119. dhÄlon . . In this case the length of the dialogue is immaterial. ge: ironically restricts ai scroÂn ktl.

w SwÂkratev soi : vocatives do not count as separate clauses. 175±85. a rethÁ n deÁ diw kein (this parallel may have been noticed by the Emperor Julian. In the present case. Prm. Wilamowitz. 772e7±773a1  W paiÄ toi nun jwÄ men 270 . 407a8. and the discourse occasionally reverts to them (`pop'): 407e3±4. which is certainly not a separate value of kai as a connective particle. New Approaches. hn d' e gwÂ. infra 409d3 a pekri nato tiv w SwÂkrate v moi. jeu xomai: not modal. 216a6±7 bi ai . . 16c5±6 qewÄn meÁ n ei v anqrw pouv do siv wv ge katajai netai e moi poqen e k qewÄ n e rri jh (not poqeÁ n). 698d1) is found exclusively. cf. 507d1±2 swjrosu nhn meÁ n diwkte on kaiÁ askhte on. 337e4 PwÄv gaÁ r an e jhn e gwÁ w be ltiste tiv a pokri noito (it is senseless to print ``.g. R. I. a5 A R. ga r: as in Plt. note on a8 mhcanhÄ v). quoted above. . thÄ v gaÁr parelqouÂshv nuktoÁv tauthsi . w be ltiste . and Stallbaum's note. ii 15 (1887) 172±86. diw xomai . cf. van Herwerden. 692d8. 332d5±6. oimai etc. 534). a3±4 jeuÂxomai kataÁ kra tov: `wol schwerlich ein platonischer Ausdruk' (Schleiermacher. taÁ deÁ jeuÂxomai kataÁ kraÂtov: remarkably parallel in wording is Grg. the Socratic paradox `virtue is knowledge' is present. Platon.3. diw xomai . J. 101±29.. . Idem valet de verbis e lauÂnein et diw kein' (H. Tht. 147b2±3 h oi ei ti v ti suni hsi n tinov o noma. 176b4 ponhri an meÁ n jeu gein. cf. Mnem. cf. Smp. 54b2 le g' w PrwÂtarce moi (so rightly DieÁs). my `Adversative relators between push and pop'. jeuÂxomai: cf. who o¨ers a con¯ation of it with Clit. a3 a skh sw . 410b3. at 177).. section i. `GAR introducing embedded narratives'. Cf.C O M M E N T A R Y: 407a3±407a5 `disjunctive kai '. F. Phd. 1). GP 2 59. Phlb. oi comai jeuÂgwn. This is only one manifestation of gaÂr marking a push. Prt. . Plt. quoted in the previous note. esp. kataÁ kraÂtov (Lg. cf. a3±4 taÁ meÁ n a skh sw kaiÁ diwÂxomai. The same holds for parentheses such as e jhn. ii 339 n. .2(1). de Jong. tiÁ v''. 269c4 (introducing the myth of the two eras). In Plato. . Grg. 137b6 ti v ou n ei peiÄ n moi [not moiÁ ] a pokrineiÄ tai. . `Platonica'. 101±4.  kouÂoiv an: cf. 310a7±8 all ' oun a kou ete. But cf. 608d11 (introducing the proof of the soul's immortality). as is proved by e. New Approaches. (to print a comma before an. `In hac iunctura veteribus usitatius est anaÁ kraÂtov.e. akolasi an deÁ jeukte on wv e cei podwÄn e kastov h mwÄn. . R. as most recent editors do. is perverse). Lg. te kai is so used 409b6 oi ki a te kaiÁ tektonikh . Tht. 408b5±c4. 87a8 ti ou n an jai h o lo gov e ti a pisteiÄ v ktl. a transition to a subsidiary stretch of discourse. o mhÁ oi den ti e stin. the narrative is embedded in evaluative statements (Clitophon's praise and blame). Intr. Phlb. i.

1). as it is perfectly possible. Since the exhortation described in the Apolog y is personal. loc. Intr. This word. 1. the verb u mneiÄ n and the (conjectured) iterative optative in this sentence. An explanation of this peculiarity of Socrates' need not bother us here. took it over from Ap. It is another problem whether the pronoun should here be taken as enclitic or not. In the same way. Smp. Phdr.1 a rethÄ v e pimeleiÄ sqai proe trepen´ aeiÁ gaÁ r e legen ktl. 60. section i. also 410d1 twÄn lo gwn twÄn protreptikwÄ n). cf. the repetition is more logical there than it is in the Clitophon. I have therefore omitted the commas which usually surround them in our texts. 13 and n. the imperfect tenses.5. e ponoma zeiv (Verdenius' punctuation) it `interrupts the construction' is clearly circular. 2) that here `the vocative is closely connected with the pronoun'. cf. See also note on 408b3±4. Greek Word Order (Cambridge 1960). X. section ii. 3).. Intr.±D. . soiÁ suggignoÂmenov: the participle denotes the general period of time during which the repeated action of the main verb took place: `when I used to keep you company'. but a comma is probably to be retained after a vocative at the beginning of a sentence (cf. J. that the author of Clit. 505a3) with his insistence in Ap.). This is a strong argument in favour of the reading 271 . Dover. le gwn o ti ktl. 308 n. K. It is also tempting to connect Socrates' statements on the theory of Forms: a qrulouÄ men aei (Phd.). 439c7). as well as o tan (o po tan) ‡ subj.2.. ou deÁ n gaÁ r allo pra ttwn e gwÁ perie rcomai h pei qwn . Schw. even probable. section i.5. 278c6. o e gwge pollaÂkiv o neirw ttw (Cra. Cf.±D. because I have the ± admittedly subjective ± feeling that soiÁ suggigno menov is a separate colon rather than forming one information unit with the preceding e gwÁ gaÁr w SwÂkratev. o po te. exeplhttoÂmhn akou wn: for the ironical value cf.. whereas in cases like 408b3 h n dhÁ suÁ politikh n. J.C O M M E N T A R Y: 407a6 a gaqwÄn pate rwn juÂnti ktl. clearly indicate that the speech to be reported presently was often held by Socrates (cf. Riddell. 1. ± Verdenius' counter-argument (`Notes on Clitopho'. §295.. Schw.3 and n. u mnoiÄ v. The use of the past tense seems to imply that Clitophon no longer frequents Socrates. 198b5 ti v ou k a n e xeplaÂgh a kou wn.3. Intr. Socrates' protreptic maxims in the Apolog y are customary (29d6±7 le gwn oi aÂper ei wqa. w Sw kratev. and qama in 407e3±5. 143 and 146 n. Burnet. Other marks of irony are: wsper e piÁ mhcanhÄv tragikhÄv qeo v. Early Greek Philosophy (London 19304 ). 234d1. 30a7¨. Gi¨ord on Euthd. on his having often uttered the exhortations he quotes. I have followed AD in printing soiÁ . cf.7. 76d7±8. pollaÂkiv akhÂkoav (R.2. (cf. `Digest'. 88.. Mem. . a6 polla kiv: belongs to e xeplhttoÂmhn akou wn. cit.

272 . mhcanhÄv: Timaeus' Lexicon Platonicum has a lemma tragikhÁ skhnh which Ruhnkenius in his edition (ed. Intr. nova cur. where Socrates `enters' the stage in a basket hanging on a mhcanh and behaves (and is treated) like a deity (see Starkie's and Dover's notes on 213±26). cf. be taken with e pitimwÄn toiÄ v a nqrw poiv. The reading of the MSS is backed up by Dio. If this is correct (Billerbeck.26 e piÁ skhnhÁn tragikhÁn anercoÂmenon le gein toÁ touÄ SwkraÂtouv ``i w  n qrwpoi. Koch (Leipzig 1828)) connects with this passage. a6±7 paraÁ touÁv a llouv a nqrw pouv kaÂllista: a more recherche variant of kaÂllist' a nqrw pwn (cf. Kynismus. 189c4.5. Or. jeuÂgein deÁ thÁn ponhri an (cf. Smp. The tertium comparationis is the superior knowledge which delivering admonitory speeches presupposes (so Demetrius).. cf. . a8 wsper epiÁ mhcanhÄv tragikhÄv qeoÂv: cf. apoÁ skhnhÄ v would mean: from the roof of the stage-building (skhnhÂ). poiÄ je resqe ktl. 352e5. not (with H. note on 407a3 diwÂxomai . Webster... for perijeroÂmenon (Ap. 11±12 (this harmonises more or less with the data given by Tim.2. B. 1. op.4. section i. a7±8 e pitimwÄ n toiÄ v a nqrwÂpoiv: on the singularity of Socrates' addressing a crowd. Greek Theatre Production (London 1956).1. cf. in my opinion. . This scene was remembered in later times. 166. MuÈller.v. .C O M M E N T A R Y: 407a6±407a8 poreuÂomai at 410c7 (q. 19c3±4) is an unambiguous reference to it.). jeuÂxomai ). proagoreuÂein toiÄ v e ntugca nousi speuÂdein meÁ n proÁ v thÁn arethÁ n. Themistius and probably Demetrius.2. Suid. The present tenses in 407e4 do not contradict the supposition (cf. cf. there is no reason for printing a comma after anqrwÂpoiv (as all editors do). (t 891) and other lexicographers). 3.. Epict. note on kaiÁ maÂla . cit. apoÁ mhcanhÄv is slightly better (the gods appearing at the end of tragedies may have spoken from either the roof or the crane. 12±13) in that Socrates did speak apoÁ mhcanhÄ v in the Clouds. e painwÄ ). 82 wishes to identify skhnh and mhcanhÂ).2 and n. T. for the exact reference of toiÄ v a nqrw poiv. to wit. The comparison may have been suggested to the author by the famous scene in Aristophanes' Clouds. `praeclarissime omnium' (Ficinus). Intr. He concludes that there was an ancient varia lectio skhnhÄv attested also in some imitations of Clit. G. if indeed toiÄ v anqrwÂpoiv is more or less equivalent to toiÄ v qnhtoiÄ v. . note on 407b1 w nqrwpoi. and others) with umnoiÄ v ± only thus can toiÄ v anqrwÂpoiv be fully understood: it is equivalent. .g. Webster.. not to toiÄ v polloiÄ v as it is e. . Prt. The irony touches on the raw spot: Socrates appears to be lacking in such a knowledge. L. A.'' and Jul. section ii. note on 410b4±6).22. Susemihl. The comparative clause should.11±13 (ˆ 4a) wsper e k tinov tragikhÄ v skhnhÄv . but to toiÄ v qnhtoiÄ v.

As it has been made clear at any rate that Socrates' speech. not to inability (he was writing a parody in any case. . e3 e gwÁ o tan. It can also be used for a long statement not repeated. was delivered repeatedly (cf.. b1±408b5 PoiÄ je resqe .2.2. note on a6 pollaÂkiv). sumpaÂsav taÁv po leiv). . the sons and the teachers are introduced in the ®rst sentence.. . . BruÈnnecke. e1±2 paÂnt' a ndra . KG ii 451. section ii. leÂgwn: for the structure of the speech. d6 httwn o v an h i. b1±e2 PoiÄ je resqe . There is conscious avoidance of hiatus (not counting slighter cases or such as may be eliminated by crasis or elision).) misunderstood this word and took it to mean `shout'. . note on 407e3 tauÄ t' ou n . When so used. Intr. 317a6).. is certainly a conjecture ± they are indirect copies of F ± arising from the confrontation of u mnoiv (DF) and u meiÄ v (A).v. s. c and Ven. Flor.C O M M E N T A R Y: 407b1 u mnoiÄ v: u mneiv. The relative obscurity of the ®rst sentence is probably due to this desire.3. . The only possible readings are umneiv and u mnoiÄ v. 653d6 (and his Index). referring to b9±d2. ± u mneiÄ n used for an often repeated statement: cf. Present education does not provide just use (b8±d2). ps. akou w).2. 3. `Kleitophon wider Sokrates'.. cf. c6±d2 kai toi . paÂscousin. section ii. e kmelethÂsousin.g. Prt. the reading of Mal. 189. ii. d4 ai scroÁ n kaiÁ qeomise v. the style is rhetorical to a limited extent only (cf.22. b8±c6 all ' a mousi av. Since the reference is to a repeated action of Socrates' (cf. c6 a mousi av. umnoiÄ v (Baumann) is a correction of the DF reading rather than a conjecture. Justice can be acquired (d2±8) and therefore should be acquired (d8±e2). LSJ.1). . England on Lg. the optative is decidedly better (the indicative is rare in distributive temporal clauses. The author's desire to allude to as many protreptic themes as he can possibly manage impairs clarity. . Thus. whereas there are two instances of hiatus surrounding the ®rst part of the speech (407a6 moi e do keiv. section ii. cf. which is about to be reported.. notes on b6±7 melethto n .. Chr. D. likewise in a depreciatory manner (e. cf. . section i. cf. . it often has depreciatory overtones (e. Kynismus. Apart from the opening words. Intr. Gi¨ord's explanation is beside the point). ad loc. 273 .g. Euthd. . which is obviously deliberate. taÁ v po leiv: the gist of the argument is clear enough: mankind neglects its duty by focusing all its attention on amassing wealth instead of using it rightly (b1±8). note on 407a6 pollaÂkiv).5. Billerbeck. to be dropped later on (cf. so no great harm was done by a little less clarity). .-Plu. and absent from the Platonic corpus altogether). Clitophon may well mean that this speech is too long. 468. I prefer the latter. 297d3±4 o po te soi tauÄ ta u mnhtai.1. Intr. 4e and Epict.26 (cf. (Intr.2.2.1 and note on 407b5 didaskaÂlouv).

Smp.. FestugieÁre. 138). note on a8 e piÁ mhcanhÄ v tragikhÄ v qeoÂv).123) is no evidence for the contrary ± I suspect that of the two variant readings in Dio Chrysostom 13. it follows automatically that wnqrwpoi and probably poiÄ je resqe are not to be looked for in the source (if any) of which this speech is supposed to be a parody. TGF adesp. Men. in a scene which the author of the Clitophon may well have had in mind. Sam. e3. w anqrwpoi (pseudo-Plutarch 4e. Prt. the crasis adds to the lofty tone. rightly interprets this as i w 'nqrwpoi.1. Ar. 314d6 is rudeness. where Socrates endeavours to pei qein touÁ v anqrwÂpouv kaiÁ didaÂskein (352e5± 6).1 poiÄ metastre jesq' w kakoi . Pseudo- 274 .1).3 n. 408b1 (the nomoqe thv).5. cf. yet it must be noted that wnqrwpoi is not the normal way to address a group (w andrev is. cf. Kleitophon'. 922 ouk oisq' o poi ghÄ v ou d' o poi gnw mhv je rhi. Nub. 223 w 'jhÂmere. Cra. 356c2. section ii.1. On the contrary. Xenophon reports a protreptic speech by Socrates. Mem.16 the shorter version a nqrwpoi agnoeiÄ te is what Dio wrote. 7. e5. Cf. 192d4 (Hephaestus ditto). iv 130 and n.. . the use of w nqrwpoi is an automatic consequence of Socrates being e pitimwÄn toiÄ v anqrwÂpoiv.88 allaÁ gaÁ r ou k oid' o poi tugca nw jeroÂmenov. 23b2 (Apollo to mankind). Kynismus. taÁ twÄ n qewÄn jronwÄ n kaiÁ u perhjanwÄ n taÁ twÄn a nqrw pwn (Schol.1. 354a3. cf. as Socrates is certainly not speaking as a qeoÁ v e piÁ mhcanhÄ v in Prt.2. c5.1). 1638 w daimo ni' anqrwÂpwn Po seidon. If this explanation is accepted. ad Ar. El. but that is not a real argument for adopting Pavlu's interpretation. 8.3) Apollodorus says of himself peritre cwn ophi tu coimi (Smp. Herm. section ii. 357a5). 173a1). Of course. Intr.. ReÂveÂlation. cf. wnqrwpoi: taken by Pavlu (`Pseudopl. Prt. Av. `Dialoghi pseudoplatonici'. poiÄ je rei. section ii. as the words form part of the parody itself (cf. 580). Carlini. beginning w a ndrev. Socrates would then be speaking w v au toÁv . if toiÄ v anqrwÂpoiv above is equivalent to toiÄ v qnhtoiÄ v. Intr. The form wnqrwpoi is con®rmed by the scholiast on Lucian as well as by Epict. which is the archetype of all MSS. in exactly the same way as w a nqrwpoi in Prt. S. . 37) as imitation of Socrates' refutation of oi polloi in Prt. Of course.. 12. wnqrwpoi kaiÁ agnoeiÄ te being the result of contamination from a Plato MS. Elsewhere the persons are at least godlike. 1. like w a nqrwpe. which Pavlu seems to imply (cf.26 iÈ wÂnqrwpoi (so the Bodleianus. 343e6 (Pittacus). Cf. Themistius 320d. Ap. the longer poiÄ je resqe. 5. f126. Libanius 18. ad loc. 3. also Isoc. in a protreptic situation (Intr.. w nqrwpoi indicates the continuation of Clitophon's ironical comparison into the actual report of Socrates' words. In itself this is hardly convincing. cf. in which the apostrophe ± naturally ± is rather frequent (353a3.C O M M E N T A R Y: 407b1 b1 PoiÄ jeÂresqe: this is a tragic diction. Billerbeck.1. cf.22. 352e5±357e8. in fact.

Av. he has w a nqrwpoi. and may be right here. cit. because there is no particular reason for pe mpousin to be stressed (the weight of the clause lies on 'pisko pouv ± it can of course be maintained that kai stresses the verb plus its object). L. see Kamerbeek ad loc. cit. but as ps. This interpretation looks too mechanical to me.27 (how Socrates did and did not address crowds). How unreliable indirect witnesses are in this respect appears from Eus. kai serves merely to connect two statements (so some commentators at S. where the similarity in phrasing may suggest a reminiscence of Clit.?'). È berlieferung des Themistius'. e.Hier. GP 2 321. (b) kai is emphatic and stresses the next verb. D.C O M M E N T A R Y: 407b1 Plutarch.. it may be an old variant (so Schenkl. it is acceptable here: as the ®rst question is a Wortfrage. and a fairly close one Ar.). loc. the same objection holds for Verdenius' suggestion (`Notes on Clitopho'. On the other hand. Epictetus and Olympiodorus. 116d5. and Olympiodorus quotes it e n twÄ i KleitojwÄ nti. immediately after praÂttontev would be the best place for it).. cit. This is probably valid for the passage quoted from OT. u mwÄn kukwme nwn kaiÁ qoruboume nwn periÁ taÁ mhdenoÁv a xia). 112.. loc. The case is di¨erent for kai linking an exclamation and a statement.26 place poiÄ je resqe after the vocative. 42 (ii 600 Conybeare) w a nqrwpoi . At 4. where le lhqav ktl. 6. (c) kai connects two questions. 143±4) that kai is here `motivating' (`and therefore') ± quite apart from the problem whether such a use of kai is not rather a consequence of the semantic properties of the connected clauses. verily we know not etc. OT 415 ar' oisq' a j' wn ei . is certainly a statement. `Yea. adv. loc.. ± where pe mpousin ktl. cf. Several explanations can be thought of: (a) as a rhetorical question is cognitively synonymous with a statement.' (Bury). the 275 . the inversion can be explained as due to an intermediate quotation in some diatribe or other.1 poiÄ je resqe. In Grg. 112) ± it is at any rate a slightly inferior `U one: the apostrophe coming after the question is a greater deviation from word order and therefore more expressive. 7. a rhetorical question remains a question and should be treated like one. cf. poiÄ dhÁ je resqe in a patent imitation of Herm.g. kaiÁ pe mpousin hdh 'piskoÂpouv ktl. can be either.. cit. . w anqrwpoi. kaiÁ le lhqav e cqroÁ v wn ktl. ± Epictetus' i w may re¯ect a deviation common to the Cynic diatribe.8. directly.-Plutarch clearly borrows it from Clit. and unsatisfactory at Av. For Epictetus. (mhÁ a gnoeiÄ te.. A good parallel is S. It is less probable for Ar. kai : between a rhetorical question and a sentence which can be a statement as well as a question (a question-mark somewhere after pra ttontev is suggested GP 2 312.32. both more or less rhetorical. This is excluded at OT. loc. . Phd.. 1033 ou deina . since the former question is in a way explained by the latter (`Isn't it terrible that they are already sending inspectors etc.

44). 274 and n. 164b3) and (as here) to man's moral duties (X.8. . the identity of w je lima poieiÄ n and swjroneiÄ n is established through equivocal use of taÁ de onta praÂttein Chrm. GP 2 321). Aristotle himself gives (Top. the association of wje limon and a gaqoÂn (cf.'' kaiÁ ``a gnoeiÄ te ktl. note on 407b1±e2). but it is too much an idiom of the dialogue to be appropriate here.C O M M E N T A R Y: 407b1±407b2 second one a Satzfrage. h (Ast) is. Of the conjectures that have been o¨ered. b2±8 oi tinev . praÂttein twÄ n deo ntwn ti is found in Arist. Protr. but confusion of h and kai is as unusual in majuscules as it is trivial in minuscules (cf. 3.. That kai is omitted by PY (Dio) is a consequence of their lacking poiÄ je resqe and should not be treated as evidence for È berlieferung des Themistius'. `Ad Themistii orationes'. and. su½ce to justify its adoption here. `Epinomis'. b 2 DuÈring. 113). 164a9±b6). Mem.1. more signi®cantly. . asyndeton (Cobet.. In its turn. EN 1094a24 touÄ de ontov). loc. sec- 276 . praÂttontev: this can refer both to the duties of the tecniko v (the soldier. de on would appear to di¨er hardly. cit. Yet I cannot help feeling that kai is a little odd. Perhaps w nqrwpoi. from wje limon and the other terms in the list of de®nitions of the e rgon of justice at 409c2±3. As zhmiwÄ dev means `harmful' (it is juxtaposed with blabero n at 417d8). what is wanted in given circumstances. de on and zhmiwÄdev are antonyms (418a4± 419b4). 110b10) toÁ de on as an example of an ambiguous word. I suppose. Euripidea (Oxford 1994). I prefer (c) to (b) because kai stressing a verb seems to be rare at the beginning of a sentence (cf. I think this is the right explanation for Av. `U b1±2 a gnoeiÄ te . e qerapeu sate: three di¨erent protreptic motifs have been crammed into this sentence which (even allowing for a probable corruption of the transmitted text) cannot be called lucid and probably was not intended to be (cf. J. Diggle. (no question) should be considered. one may wonder if his readers could understand the sentence if so articulated ± an argument that cuts both ways. 430) and pwÄv (Schanz) are not better. . 3. the doctor. note on 407a1 wjeleiÄ n) explains how Plato could say de on kaiÁ w je limon kaiÁ lusitelouÄ n kaiÁ kerdale on kaiÁ agaqoÁ n kaiÁ sumje ron kaiÁ eu poron toÁ au toÁ jai netai (419a5±8). R. Cobet's deletion (cf. Schenkl. See Intr. asyndeton would be rather harsh. capable of meaning both toÁ sumje ron and toÁ kaloÂn. Chrm. given the fact that the author did not use quotation marks. Einarson.25. In the Cratylus. 469d2. Burnet is too one-sided when he says: ` ``the right thing''.. not ``our duty'' which would rather be touÄ prosh kontov' (note on Arist. though in a di¨erent context (cf. . if at all. The phrase ou deÁ n twÄ n deo ntwn pra ttein recurs at Isoc. (d ) kai serves to introduce a surprised or indignant question (GP 2 311±12). I think. 198) ± the imitations in Procopius and Constantinus Porphyrogennetus do not.

twÄn u e wn but crhmaÂtwn . And. . Themistius' ou demi an poieiÄ sqe e pime leian (from 407d8±e1). 277 . . cf. .2. 706 and Van Leeuwen ad loc. it is not in any MS ( pace Burnet). MT §580. paÄsa h spoudh Ly. e stai: for the construction.39 ou k ai scu nesqe . 1±3. Phrase relative. section ii. .30 ti  rtemisi hv . That amounts to saying that twÄn u e wn is an anticipated part of the opwv-clause. 15a6±7. . backed by Iamblichus. `o stiv apparaõÃt . Thesm. nor has Themistius' spoudhÁ n apasan. . 306d9± e2. son roà le est de manifester en somme la responsabilite de l'ante ce dent dans l'actualisation de ce proceÁs' (Monteil. cf. h duÁ oi sqa ..67. o pwv e pisth sontai crhÄsqai . 248b6. 29d9±e1. J. paÄsan thÁn spoudhÁ n (F) weaker. though it is retained by most editors).5. Nub. 144). 2. . 628e4. . Now. Mem. comme un intensif du simple o v signalant aÁ quel degre e minent l'ante ce dent est concerne par le proceÁ s subordonne . 144 nn. . .1.. See Intr.. the indirect tradition. who gives numerous examples ± but only of ®rst and second person antecedents ± 143 n. Ar. For examples of simple o v so used. 121. cf. twÄ n u e wn . i kanwÄv. 123. .C O M M E N T A R Y: 407b2±407b3 tion ii. ii 399). h tiv ktl. b2±3 crhmaÂtwn . 7. Elsewhere in the Platonic corpus we ®nd h pollhÁ spoudh Phdr. Phlb. shows a bewildering variety as regards its identity. .12. . and is made part of the main clause by anticipation (prolepsis). Thus Dio's hmelh kate (from 407e6?). Lg. 329). . KuÈhner on X. Ps. cf. e stai vs. . 168. X. our MSS do not provide such a verb (the reading ameleiÄ te kai between tou toiv and ou te (b4±5) is a conjecture of uncertain origin (Stephanus?). The sentence is further complicated by the parenthesis 407b5±7 ei per maqhtoÂn .2. 2. An. 2. 219e7±8. also C. The o pwv-clause is on the analogy of speuÂdein. 6. cf. o pwv .1. 1. . cf.. crhma twn pe ri belongs to it. 144.3. Moreover.5. but quite a normal idiom. Autour de te e pique (Amsterdam 1971). . Verdenius. Hdt. pseudo-Plutarch's mikraÁ jronti zete. `utmost'.3. is rather strong (`utter'. . which unanimously does o¨er a verb. warrant only one conclusion: they did not have a verb in their texts any more than we have.1 for the parallel Ap. b3 twÄn d' ueÂwn: the opposition is not crhmaÂtwn vs. b2 oi tinev: after questions and statements expressing indignation. . h tiv ktl. . .1. Euthd. ostiv may be reinforced by ge: Ar. admiration etc.1 A ma lista qwÄma poieuÄmai . `Notes on Clitopho'. no examples in KG (cf.99. oi tinev ktl.-Plutarch's paÄsan spoudh n has no authority. introducing clauses which provide the reason for these a¨ects ± not in LSJ. Equally usual but better known in these contexts is o v ge.3. . thÁn paÄ san spoudhÂn: this. .2.. An. .1. 1. Ruijgh. therefore the presence of a verbum curandi from which the clause is to depend seems imperative.

28d7± e3. thÄ v dikaiosuÂnhv is rather inane.4). To ®ll it in. Likely candidates are e pimeleiÄ sqai. Therefore. pseudo-Plutarch. we shall have to ®nd a verb or verbal phrase opposite in meaning to thÁn paÄsan spoudhÁn e cete and capable of being constructed with both the personal genitive and a o pwv clause. me lein tini R. 345d2. as Schenkl has done without arguments (`U des Themistius'. tou toiv is an adverbial (®nal) clause instead of a complement clause. a meleiÄ n is less likely as it does not appear to govern a opwv-clause. as thÁn paÄsan spoudhÁn e cete : ou te didaskaÂlouv autoiÄ v eu ri skete is a very lame opposition indeed. preferably one which is found with an anticipated genitive and a o pwv clause at the same time within the Platonic corpus. . As a periphrasis has already been used in the sentence. . 278 . Thesle¨. Euthphr. the future e pisthÂsontai becomes normal. memnhÄsqai and periphrastic constructions of the type e pime leian (spoudhÁn) poieiÄ sqai (e cein). whereas it is exceptional (though not impossible. Grg. may strengthen (a)±(e). . the repetition dikai wv . only the ®rst three are found with a personal genitive anticipated from a o pwv clause in Plato (e pimeleiÄ sqai Ap. . . Styles. if. 112±13). jronti zein. which. a simple verb seems better. section ii. the unanimous feeling of Dio.C O M M E N T A R Y: 407b3 Can we follow Burnet and do without a verb at all? The following arguments tell against this.. jronti zein Ap. cf. 69 n. È berlieferung I conclude. as our MSS have it. Themistius and Iamblichus that a verb meaning `to neglect' is missing. it becomes quite unobjectionable if another verb is inserted. KG ii 374. 29d9±e1. . they would be strongly out of character in this pastiche (cf. 384.3. o pwv . e1±3. (a) The absence of a verb creates a false antithesis crhma twn : twÄn ue wn. au toiÄ v. 2). Hence an incidental change to the impossible form e pisth swntai in some secondary MSS of Plato and some MSS of Themistius. as I have argued in Intr.1. The antithesis is further weakened by the anacoluthon twÄn d' ue wn . tou toiv is an adverbial clause depending on eu ri skete. . that we ®nd ourselves before a very old lacuna in the text. (e) If o pwv . 2d1±2. (d ) If twÄn d' u e wn as well as opwv .2. . we should certainly expect a verbum curandi here. 29e1±3). ( f ) Though hardly counting as an argument in itself. tou toiv depend on a verbum curandi. (c) Whereas anacolutha are rather frequent in Clitophon's report. is only half-complete. .1. all of course preceded by a negative. . me lein tini . (b) If indeed the structure of this sentence is modelled on Ap. Of the verbs mentioned. moreover. 520a4.

. and because thÄ v yuchÄv o pwv w v belti sth e stai is basically one phrase. . and it may be backed up by ps. . . section ii. oute .1. its distance from twÄ n u e wn would make the sentence too harsh. Hence Iamblichus writes orqwÄ v (he had been paraphrasing the protreptic parts of Euthd. if the verb came after tou toiv (so Schenkl. .C O M M E N T A R Y: 407b4±407b5 though it would be foolish to exclude the possibility of another verb. . I shall restrict myself to these three. . . ou de ge. Susemihl. . . in the ®fth ± Ap. . cf. . 280e3±281a8. Three places are possible for an ou k e pimeleiÄ sqe (ou me lei u miÄ n. Futur grec. ou de g') but there are a few instances of ou te . . . ou te . Protr. cf. . dikai wv: normal in protreptic contexts is o rqwÄ v: Euthd. . b4 paradw sete: possibly modal (`you will have to bequeath'.-Plu. ou te . b 84 DuÈring. But there is good reason for the deviation. . so ou te is what one expects to have stood here. out' . 88e1.. `the best possible state of the soul'.3. oude .i after tou toiv). shortly before) and Dio orqwÄ v kaiÁ dikai wv. 608b5±6 ou te . and besides. so Schleiermacher.2. on a par with jronh sewv and alhqei av. Of the three possibilities. ou te does not occur in Plato (let alone ou . ou de (R. MuÈller). and it is clear from their gradual disappearance from the speech that they have been needed momentarily but not throughout the protreptic (which is. arxoÂmenon. I do not have the courage to print oute jronti zete: it is euphonically not satisfying. Intr. . . 426b1±2 oute . 429e3 ou te . . paradwÂsete hou te jronti zetei has the advantage of palaeographical probability (the same sort of probability will of course hold for any of the verbs if one reads houte. . . . . . cf. . 29e3 ± the opwv-clause precedes the verb only in virtue of its shortness. . on 406a1 KleitojwÄnta toÁ n A The precise form of the negative can be determined by the following ou te . . ou te . ii 221±2.. . cf. b5 didaskaÂlouv: it is a strange thing to see Socrates advertising on behalf of the professional teachers of virtue (the sophists). ou te . The next problem is where to place the lacuna. Men. 499b2 oute . Tht. . . . note  ristwnuÂmou. . ou deÁ ou n. ou meÁ n dhÁ au ou d'). . note on 407e7 arxontov . 148e3±5 ou t' . . Arist. . By far the best place stylistically is È berafter paradw sete. . ou de g' and Adam ad loc. after 279 . 112±13). ou te . ou de n. . ou de g'. the order genitive ± verb ± opwv-clause is found in four out of the ®ve instances quoted. ou de . it seems that ou . . H. cf. `U lieferung des Themistius'. ou d' au . But I can think of nothing better. . Magnien. . paradwÂsete and tou toiv. mikraÁ jronti zete (which cannot be right because ou te is indispensable). This order is characteristic of formal prose. . ou te . . . . . ou jronti zete): after ue wn. ou te .

Bakker (ed.2. see also Men. t 158). Grammar as Interpretation.1. Verdenius. 70a2. Normally chiasmus is used more or less automatically if the information in the ®rst of a pair of clauses is too compact to be presented in the parallel order. section ii. the sophists.3. I have o¨ered a possible reason in that section. eu ri skete would have meant `you cannot ®nd' (cf. . especially as the parenthesis ei deÁ melethto n . esp. It goes without saying that only forms of the present and imperfect can be so used. 69. 142 (cf. on e1±2 q' a ma kai . Immediately. Lg.3 n. Protreptik.1±7 stresses the futility of looking for teachers of the normal curriculum but not of politics (cf. their existence is questioned (ei per maqhtoÂn) and they are given understudies in the form of anonymous persons oi tinev u maÄv pau sousi tau thv thÄv a mousi av. however. below. one may also think of in¯uence of a protreptic source: X.. `Notes on Epitrepontes'. J. Intr. Cf. Here. the whole Prt. 24). euri skete: rather conspicuous by its durative (here `conative') Aktionsart. . not a sophistic one). Thompson on Men. 357e4±8. Kleitophon'. 4. 184±91. passage has rather a protreptic ring about it. . Epit. the information is not compact at all: melethto n and askhto n are virtually synonymous. my paper `Figures of speech and their lookalikes: two further exercises in the pragmatics of the Greek sentence'. Therefore this instance of chiasmus does have rhetorical overtones. . Intr. Gaiser. Pavlu (`Pseudopl.). .2. Mem. cf. When the presence of didaÂskaloi is so manifestly unasked for in the rest of the parody. b6 a skhtoÂn: cf. 138. 8) drew attention to Prt. Indeed. a Socratic. Greek Literature in its Linguistic Context (Leiden 1997). the same section).1. b6±7 exaskh sousin kaiÁ ekmelethÂsousin: the preverb serves to give 280 . cf. . d8±e1). however. It is. one wonders why they are mentioned at all. 664a4. a genuine Platonic idiom. 42±4.. §304) and Denniston (Greek Prose Style. Styles. Cf. c5±6). cf. in older Greek ou te .C O M M E N T A R Y: 407b6 all. in E. 82). which is not the usual one. b6±7 melethto n . and after the argument about free will they have lost their personality and we are back again in commonplace Socratic terminology (deiÄ n e pime leian thÄv nuÄn plei w poieiÄ sqai. section ii. where Socrates reproaches the common herd for not sending their sons or going themselves to the teachers of the e pisthÂmh hdonhÄv. i kanwÄv b6±7 was inserted with the express purpose of diverting the attention from the question whether there were or were not teachers of justice. ekmelethÂsousin: such examples of chiasmus as are given from Plato by Riddell (`Digest'. 74±7) hardly suggest that the ®gure in itself is a marker of rhetorical style (so Thesle¨. 169±214.

ii 219 (though his parallel Hdt. cf. cf. . .C O M M E N T A R Y: 407b7±407b8 the verbs perfective (`confective') Aktionsart. are probably modal. g' e ti: Themistius' ge (conjectured here by Schanz) and Pa ge ti may indicate a common source (cf. Mnem. Yet g' e ti is clearly lectio di½cilior. my paper `Plato. note on 410b8±9. inasmuch as it does not teach the just use of wealth. implies `the education which you have given to your sons does not teach them to use their money justly'. 325c1. as the preceding reproof. which is unlikely without the article. 403d7). b8 e qerapeuÂsate: the causative component (as usually) not expressed (cf. `RaÈtsel'. `people to train them'. when preceding proÂteron (here and Hdt. Gonzalez. Cratylus 417c'. 43 n. W.64. 5). Chrm. 287a5.Ma. especially the opening (gra mmata kaiÁ mousikhÁn kaiÁ gumnastikhÁn u maÄv te au touÁ v kaiÁ touÁ v paiÄ dav umwÄn. Futur grec. this one contains but one: the present curriculum is inadequate. notes on c5 and on e7 arxontov . 93a6. Orwin. Its form is inversely proportional to its content: the wording is rather ponderous. cf. 266±9. Cf. This topos is manifestly used here to make room for the reductio ad absurdum in the next sentence. cf. e kmeletaÄ n occurs at Hp. . e ti here means `in the past'. esp. in itself it is super¯uous. a mousi av: whereas the ®rst sentence of the speech was loaded with di¨erent protreptic motifs. Verdenius. 312b1±2 281 . Schw. 312c1 with Adam's note.±D. Schleiermacher). Smp 184c4. This verb often refers to instruction of any kind in Plato. Water®eld's `moreover' suggests that he takes ou de . J. `you don't see to it that your sons learn how to use their money justly'. Magnien. further Prt. Bury and Zuretti take e ti proÂteron as apposition to the whole clause (`that which comes ®rst'). so Bertini). more or less. b8±c1 graÂmmata kaiÁ mousikhÁ n kaiÁ gumnastikhÂn: cf.4 is rather di¨erent). . Cra. The origin of the reference is metaphorical: as doctors tend the body. so teachers the mind (cf. `Notes on Plato's Meno'. Bluck on Men. arxo menon. which is the essence of this sentence. It is naõÈve philology to consider e xaskhÂsousin a mark of inauthenticity (Ge¨cken. 3.40. e xaskeiÄ n is new in this meaning. 433 n. 606a4±5 touÄ dakruÄ sai te (`have a good cry' ± a very unusual use of the aorist of this verb) kaiÁ apoduÂrasqai i kanwÄ v kaiÁ apoplhsqhÄnai. Appendix ii ). Cf. e ti as the negative counterpart of e ti de . 3. b8±c6 a ll' . . Prt. ± The futures. b7 i kanwÄ v: underlines the perfective Aktionsart as at R. R. 10 (1957) 289±99 at 296. . being on a par with the o pwv clause of b4. 156b3±157c6 and Prt. but there are no parallels for this in Plato. c2 below.4) it does little more than strengthen it (not `already before'. 8): `you have had yourselves treated' (so. 440c5. carefully embedded between o rwÄntev and i kanwÄv memaqhko tav).

606e3±4 paidei an twÄ n a nqrwpi nwn pragma twn. c1±2 u maÄ v te autouÁv kaiÁ touÁv paiÄ dav umwÄn: the order (the inverse one of b3 twÄn d' ue wn ± 8 u maÄv autou v) re¯ects the decreasing importance of the children. h ghsqe: being lectio di½cilior. 46±7) even though it is backed by Themistius. 575e. 159c3±d2. c3 ei nai: omitted by PaVa Them. cf. Both the present and the perfect occur in Plato. cf. Av. cf. cf. Pax. 9) uses the similarity of Prt. Prt. W. 643e4 thÁ n . c4 pwÄv ou: usually ˆ nonne (cf. Kleitophon'. 278). Mem.v. 496a6. and presupposes one in the adressee. here ˆ qui ®t ut non. The accusative of respect is more normal.3.. cf. Cf. As traditional education always consists of the three subjects in Plato. 312b1±2 and our place to prove the dependence of Socrates' exhortation on Prt. i ppeuÂein. especially in the report of other people's opinions or statements. 1 108c7±8 toÁ kiqari zein kaiÁ toÁ aeidein kaiÁ toÁ e mbai nein orqwÄ v (cf. Men. section ii. J. 408c2.2.. What mousikh amounts to is best illustrated from Alc. . . c2 dhÂ: this particle denotes a high personal commitment to the utterance by the speaker. Dodds on Grg. paidei an. Playing the aulos is mentioned along with the curriculum at Alc. 232c8. Intr. . . There is no need for Hermann's o mwv. 459a. for kakouÁv . . this should be preferred to F's hgeiÄ sqe (on F's tendency to normalise Plato's language cf. very extensively but with a more positive appraisal in Protagoras' protreptic. . . 325d7±326c3. (b) the nominative would be ambiguous as it might be taken as coordinated with o rwÄntev (b8). . 8 (1955) 14±18 at 16. 108a5±6). cf. 4. 269b6. c3±4 kakouÁv gignomeÂnouv periÁ taÁ crhÂmata: understand u maÄv au tou v. `Notes on Hippocrates Airs Waters Places'. Alc. this is not cogent. M. (against ADF). au leiÄ n. cf. Ar. 449a2±4 kakaÂv . LSJ s. 1 106e7. `how can you fail to despise' (Gonzalez. on Grg. van Ophuijsen in Two Studies. R. Van Leeuwen on Ar. 118c8±d4. 94b5. kaiÁ kiqaristouÄ kaiÁ paidotri bou.2. 472 pwÄv ou n ou cwreiÄ tourgon. . De Vries on Phdr. hardly touÁ v paiÄ dav as well. his argument `neque enim ratio inconstantiae ex- 282 . Adam on R. . acc. X. quite possibly rightly. pwÄ v ii 7). paidei an a rethÄ v: Lg.1. dhÂ. cf. Verdenius. 1 106e6 graÂmmata kaiÁ kiqari zein kaiÁ palai ein. ± For gi gnomai `prove'. The personal commitment in the speaker may be ironical (thus in e¨ect a distance is created).6 mentions kiqari zein. peri te poÂlewn dioikhÂseiv kaiÁ periÁ i diwtwÄ n yuchÄv tro pou kataskeuh n. but cf.C O M M E N T A R Y: 407c1±407c4 paraÁ touÄ grammatistouÄ . Dodds' Comm. J. 82. Pavlu (`Pseudopl. . note on 408a4 kai . R. kakoiÁ gignoÂmenoi would be slightly more regular but (a) the re¯exive pronoun (c1) makes the accusative possible here. peri c. Chrm. . proÁv a rethÂn . Mnem.

. .' Plutarch's kai toi ge at 439c is not backed up in his second quotation. 122±4 (slightly di¨erent B. cf. . A conspicuous redundancy in expression makes up for this lack of content. Cf. ii 219±20. which causes. for the changed construction. . c6 kai toi . 4. ge ± in this case one might translate `yet this is the plhmme leia and r aiqumi a . kaiÁ adeljoÁv a deljwÄ i kaiÁ po leiv po lesin (instead of kaiÁ adeljoiÁ kaiÁ poÂleiv . Placed at the end of the second sentence.5). note on d2±5). note on a1 all ' . xxviii) overlooks the fact that this is a rhetorical question. c6 a mousi av: as often. KG ii 152 rightly distinguish between kai toi ge and kai toi . allhÂlaiv). . some poignancy in the choice of this word: for all its mousikhÂ. 131±49). cf. Magnien. . this word announces the sarcasm of the third ± the need of a periphrastic turn for dida skaloi has brought about a rather ®ne by-product. . The split form of this collocation (cf.8 (in a conversation about justice): pauÂsontai d' oi poliÄ tai periÁ twÄ n dikai wn a ntile gonte v te kaiÁ antidikouÄntev kaiÁ stasia zontev. kaiÁ paÂscousin. . . it adds nothing really new to the argument. ge: adversative. Alc. ibid. Prt. however. while serving as well to underline the sarcasm and pessimism: plhmme leian kaiÁ r aiqumi an. . . c6±d2 kai toi .) is directed against the preceding sentence. c6±7 dia ge . For the vagueness of the relative clause cf. New Approaches. more precisely `inverted denial of expectation': the preceding is true even though you would not expect it to be because of the following: he is unhappy kai toi he is rich.4. a ll' ou: dia te . c5 oudeÁ zhteiÄ te oi tinev umaÄ v pau sousi: for the construction cf. mhÂn) is much more frequent than GP2 120 suggests. ame trwv kaiÁ a narmoÂstwv. note on 407b5 didaskaÂlouv. The future is modal. `Sur le roà le pragmatique de KAITOI '. There is. pauÂsontai d' ai po leiv diajero menai periÁ twÄn dikai wn kaiÁ polemouÄsai. pa scousin: the aim of this sentence is twofold: to give room to the sarcastic remark and to work out ± again rather ponderously ± the results of absence of education in justice (already referred to in c3±4 kakouÁ v gignome nouv periÁ taÁ crh mata). . Mem. 348d4±5 zhteiÄ otwi e pidei xhtai kaiÁ meq' o tou bebaiw setai (this would seem the best reading if Burnet's apparatus can be trusted. quasi eam objurgator non intelligat' (iii. the objection raised in the next sentence (u meiÄ v de jate ktl. . polemouÄ ntev. . Obviously. 1 120b3. not this one (cf. KG i 223. . kaiÁ ou (AL of Themistius) is not 283 .C O M M E N T A R Y: 407c5±407c6 quiritur. . There is a slight similarity in wording (probably fortuitous) to X. 564. Futur grec. in fact. my analysis in New Approaches. Jacquinod. cf. present education produces only amousi a. cf. this is not lack of cultural education but of moral knowledge (it refers to kakouÁv gignome nouv periÁ taÁ crhÂmata). 534e. b6±7. . .

as appears from his change of the preverb to dia-. See note on 409d6±7 agaqo n t' .. that Plutarch understood it in this sense.) kaiÁ ou occurs only in polar expressions such as pollaÂkiv te kou c a pax (S.3. The masculine form. A reciprocal middle is out of the question in both meanings discussed: prosje resqai is passive. not with prosjeroÂmenai. ou di' apaideusi an ou deÁ di' a gnoian). war between states: it takes up stasiaÂzousi (the two verbs are coupled in Euthphr. b i 4). 8a1±2). c7 diaÁ thÁn e n twÄ i podiÁ proÁv thÁ n lu ran a metri an: the prepositional phrases are carefully embedded between article and noun. c6 tauÂthn thÁn plhmmeÂleian: of proving kakouÁv periÁ taÁ crh mata. But after the rather extensive criticism of common opinion in b8±d2. Gildersleeve. . . proposed by Ast. Both in the Euthydemus and in the Clitophon the authors grasp this point but carefully circumvent the necessity of proving it (see Intr. . section ii. c8±d1 a meÂtrwv kaiÁ a narmo stwv: used with the same double entendre as a mousi a and plhmme leia (c6). . It must be noted. kaiÁ oude pote kako n. . c8 a deljwÄi .2. cf.v. in other words that vice is ignorance (cf. lo gouv. which is rather odd with stasiaÂzousi. . Water®eld). Schw. as 406a6 touÁ v .C O M M E N T A R Y: 407c6±407d2 Greek: adversative te (. England ad loc. polemouÄntev: includes. repeating tau thv thÄv amousi av in a rather otiose way. The words are found coupled Lg.±D. 691a7. Bury's `without measure or harmony' is perfect. is not only unnecessary. some sort of proof is now necessary. Sartori. . po lesin: connect with stasiaÂzousin. cf. because in the military sense it would mean `attack physically'. there are. The only thing which he contrives to establish is that wrong- 284 . quali®ed by a me trwv kaiÁ a narmoÂstwv. note ad loc. en twÄi podi : instead of touÄ podoÂv (so Plutarch): not the ametri a which is present in the foot but the one (ˆ plhmme leia) which is present in bad handling of wealth. ii 207±8. d2±e2 u meiÄ v de .1). though. but probably ungrammatical (cf.). c6±7 r aiqumi an: of not looking for a didaÂskalov. where (as here) the musical association of plhmme leia is present (cf. OT 1275. not `clash together' (Bury. d1 prosjeroÂmenai: `behave oneself ' (LSJ s. As the preceding sentences have dealt with vice rather than virtue. or sophistic and Socratic exhortation is useless. 409a3. . GP 2 513). . 605). rather more exceptions than it would appear from his collection of examples. Orwin. however. the point which Socrates sets out to prove is that vice disappears by teaching or training. . po leiv: virtue comes by teaching or training. rich more in words than in thought. cf. stasiaÂzousi: here not of internal discord of states. but is not restricted to.

467c5±468c8. 351b3±d7. Fallacy. (c) no one does evil willingly (pwÄv ou n dhÁ . Socrates' question at Grg. section ii. .5 and n. But I now believe that this criticism has more to do with the formulation of the argument than with its validity.. the argument in Prt. (2) The challenge is disproved by a rather less valid argument: (a) people commit injustice either unwillingly (Socrates' view) or when overcome by desire (the view of the masses).g. h a diki a). cf. cf. The proof.3. . (b) injustice is evil (pa lin d' au . 235 (a).3. Yet the fallacies Plato employs in these passages reveal a great deal about his ethical principles (cf. Of course the vulgar reading of (b) contradicts the conclusion of (1). 77b6± 78b2. this commentary is not the place to deal with the gigantic literature on these passages ± I have to con®ne myself to what seem to me the most important objections. O'Brien. e. Socrates' question 77d4±6 implies an equivocation: kako v ± morally evil : kakoÂv ± harmful. Plato frequently sets out to prove the Socratic paradox oudeiÁ v e kwÁn a martaÂnei (or its implication that no one wishes evil). . ai roiÄ t' an. twÄn h donwÄ n). (1) (a) Injustice is committed either through ignorance (ou di' a paideusi an ou deÁ di' agnoian) or willingly (all' e ko ntav). suggesting that neither end nor means is evil. Grg.) (b) itself is arrived at through an argument involving opposites which the author may well have considered valid. which neglects the possibility that vice may be due to lack of training. but evidently the latter is taken here to imply the former. and his arguments are never very cogent. I thought (2) was in fact a sophism.). . This proof is immediately invalidated because (c) is challenged (httwn . is valid only if a gaqo n ˆ h du and kakoÂn ˆ a niaroÂn. (In the ®rst version of this book. This is by no means the same proposition (cf. (b) to be overcome is involuntary (ou kouÄn . but that does not re¯ect on the validity of (2). . Intr. . Cf. p. section ii. Argument (2) could be regarded as a sophism a dicto secundum quid: to do something which is a kou sion in some respect is not the same thing as to act a kwn (cf. 16). which will seem preposterous to many scholars. . Dodds' note.5). Men. conclusion (not stated): therefore injustice is committed through ignorance. in either case (e k pantoÁv tro pou) people commit injustice unwillingly (to ge adikeiÄ n a kou sion o lo gov ai reiÄ ). 301. Sprague. 353c9±355a5. .C O M M E N T A R Y: 407d2 doing is involuntary. consists of two separate arguments. 6). e kou sion). Meno could have answered `Yes' at 78a5 and besides. I am sorry to make such a claim. whereas here the fallacious argument (if it is fallacious) is contrary to the spirit of the Socratic dictum in that this dictum is a pregnant formulation of the conviction 285 . Prt 352e5±357e8. Of course. 468a5±6 is framed as a dilemma. Paradoxes. again the meaning of agaqo n is not quite clear. conclusion: therefore.

di' apaideusi an points back to the didaÂskaloi (b5. Kurzdialoge.C O M M E N T A R Y: 407d2±407d5 (shared by Plato) that lack of self-restraint is in fact lack of knowledge. Another example: [Pl.2. Men. d2±3 ou di' a paideusi an ou deÁ di' a gnoian: not just a rhetorical repetition. ai scroÁn kaiÁ qeomise v: cf.] Demod. Lg. ad loc. once in Min. d3±4 e ko ntav touÁv a di kouv adi kouv ei nai: the second a di kouv might have been missed (cf. Protr. Paradoxes. Lg. 989d2. two inconsistent opinions are contrasted. Protr. either a or not a. 355a5). But the omission (in F) of touÁv as well as of a di kouv makes one suspicious. 984b2. 991b5). O'Brien. M.g.) and does not have here the association `ignobility' (`inopia humanitatis'. amaqi a Prt.v. 863c1. paÂlin and au qiv and esp. 384c1±d2. Again. Hirschig) is necessary. 2 310e4. 357d1. Ep. a gnoia is the more general word. 38±9.6). on the question whether agnoia and amaqi a di¨er. 193±6). Intr. where no correction of the text (wv qaumasi wv gi gnontai oi agaqoiÁ hagaqoiÁ i. 838b10. this coupling of synonyms makes the sentence rather heavy with sarcasm. 100: if a. La. 170e7±171c7 (which is rather more complex) and Arist. 286 . Ancient logicians quote two examples of this argument: Tht. Lg. dhÁ ou n is found at Epin. cf. Cf.. 8 353c4. Particules de liaison.3. APr ˆ Arist. 879c3. des Places. R. 22b7. Aristotle's Protrepticus. Ast. sections ii. note on c6±d2. 89b1±2 ei ju sei oi a gaqoiÁ e gi gnonto ± b9±c1 ou ju sei oi a gaqoiÁ a gaqoiÁ gi gnontai. see Bluck's note on 89a6. MuÈ ller. Rabinowitz. 1 (cf. Lexicon. 325b4. the real Socrates (at least what the author considered to be the real Socrates) is kept out of harm's way (cf. Die stoische Logik (GoÈ ttingen 1974). cf. ei siÁ n a kontev kakoi . where. 183b7 ajrodi sion gaÁr o rkon ou jasin ei nai (o rkon ho rkoni Hertz). C. cf. a 2 DuÈ ring. Ep. cf.1. b. a 2±6 DuÈ ring. . 7 326e5. as here. d5 oun dhÂ: cf. 382b8. 193d6. 860d1 oi kakoiÁ . In combination with tolmaÄte. these combinations cannot be used for questions of authenticity: in the other Dubia. (321d1). cf. 350d5. three times in Epin.) which it usually carries in Plato. ou n dh occurs two or three times in Alc. d4 paÂlin d' au : this `pleonasm' is quite normal in Plato. des Places. 264. . Phlb. Though both ou n dh and dhÁ ou n are practically con®ned to Plato (and Herodotus). s. ii. Neither is found in the Spuria. therefore always b). in Arist. cf. cf. add Prt. four times in the Letters (Ep. 86). (879d2. commonly used in discussing the maxim (e. it is not a basic principle of Socrates' and Plato's ethics which is subjected to Clitophon's ridicule but rather a perversion of it. R. 85±7. It is very interesting to note that argument (2) closely resembles what is once or twice called parasunhmme nov (Schol. b & if not a. Ast. Frede. Lexicon. W. cf. GP 2 468±70. s.vv. likewise Smp.

Ap. Of course. would choose such an evil]. 25e4 touÄ to htoÁ i tosouÄton kako n. KG i 42. which therefore aims at a special. 48±9 who gives examples from Lg. my note on R. Ap. d6  H ttwn ov an h i: for the postponement of the relative. ii 269±70. . focalising e¨ect. 48c7) in three ways: (a) with modal adverb (Cri. Greek Prose Style. 479c8 sumbai nei me giston kakoÁ n h a diki a. e kwÁ n ai roiÄ t' a n (so translated by Bertini. but cf. which I submitted in the ®rst version of this work. 663d7). . a nominal construction is found with the semantically related sumbai nw. or ti v . Lg. where Wakker's alternative paraphrasis `exclusively in the case that. but it is covered up by the hyperbaton httwn o v. there remains a slight inconcinnity between question and answer (diaÁ toÁ h ttwn ei nai or h ttw menov would have been more normal. 607b3. 1) in the preceding question). . d8 to ge adikeiÄ n a kouÂsion o lo gov ai reiÄ : Plato construes the phrase o lo gov ai reiÄ (cf. loc.. (R. cit. the analysis of the argument.' (Bury). Denniston. Cf.C O M M E N T A R Y: 407d6±407d8 to ge toiouÄton kako n: cf. 329d6. The double accusative construction may therefore be defended. For the function of these words (`in either case'). It seems hazardous to explain our construction as an ellipse of ei nai: ei nai can be omitted with dokeiÄ n and nomi zein etc.' applies. where the construction of h gouÄ mai with two predicative accusatives is followed by an accusative and in®nitive. 287 . cf. d7 ei per: `precisely in the case that. 141d6). 408c2±4. kaiÁ touÄ to: as well as ignorance. . 471 n. The relative clause should not be explained as quasi-causal (Water®eld. Phdr. 671c1 (cf. which develops into a copula: Grg. note on d2±e2. Adam on R. who. 238a4. 319±20). cf. Wakker. . cf. (b) with a personal noun as object (R. 19d7 periÁ e mouÄ a oi polloiÁ le gousin... The distinction between o v an and o stiv a n (KG ii 426) proves rather theoretical. 604c7. Burnet on Cri. one expects to ge adikeiÄ n a kou sion ei nai here. but rarely with other verbs. ek pantoÁ v tro pou: ˆ pa ntwv. Jowett±Campbell on R. contrast ei per maqhtoÂn b5±6. However. The clause gives the reason for thinking that h ttwn ei nai is akou sion. who points out that sumbai nw may be followed both by ei nai (Prm. Lg. Gildersleeve. as in the next clause kaiÁ deiÄ n ktl. 390b6±c1. . only. 440b5. Conditions and Conditionals. I withdraw the conjecture to ge adikeiÄ n hei naii akou sion o loÂgov ai reiÄ . Here the postponement splits the phrase httwn twÄn hdonwÄn. 607b3). Prm. . Consequently. `Notes on Politeia iii'. England ad loc. Gonzalez) or quasi-temporal (SouilheÂ) or quasi-conditional (De Win) but as a ± necessarily elliptical ± answer to the preceding question: `Any man [sc. 363a6.' (cf. 345±7. cf. England on Lg. R. Phlb. (c) with accusative and in®nitive. 134b1) and by wn (Euthd. 281e3)). cf. . 35d6. 938c3.

There is no parallel in Plato for the chiastic construction with this collocation. Cf. This is the most frequent cause of chiasmus: in such cases it is wrong to consider it a ®gure of speech. xumpaÂsav taÁv po leiv: rhetorical e¨ect is caused by the climax pa nt' . 408a4±b1 (end of the third part of the speech). e1±2 pa nt' andra . taÁv po leiv: though it has not been proved that injustice is ignorance (only that it is involuntary) these words presuppose that it is.C O M M E N T A R Y: 407d8±407e3 d8±e2 kaiÁ deiÄ n . For the thought cf. cf. as in fact Socrates has done. For e gw preposed. R. section ii.. the moral consequences of the argument (and therefore of the whole speech) are coordinated with the conclusion within the same sentence: the protreptic can thus be framed u poqetikwÄv (Demetrius' words. q' ama kai : this collocation is found only rarely before the latest period (e.g. . Intr.: over twenty times.3. There is hiatus after e gwÂ. 188. next note. For the postposition of te (one would have expected pa nta t' a ndra ktl. note on 407a6 pollaÂkiv. which accounts for the postposition of te). xumpa sav. Intr. . 436d5). .) cf. Chrm. 798a6±8 periÁ taÁ v twÄn anqrwÂpwn dianoi av te ama kaiÁ taÁ v twÄn yucwÄn ju seiv. not by the chiasmus. cf.1. Cf.2. Strictly speaking this word is in- 288 . from i di ai. Ap.1 ad ®n.4. but after all this is parody. qama : cf. For tauÄ ta this is easy to understand. d8 epime leian: the directly reported speech very neatly ends with the signal theme of all Socratic exhortation. section ii. Intr. Of course tauÄ ta is object of a gamai and e painwÄ as much as of le gontov. 54d3±4 tauÄ ta w ji le Kri twn eu i sqi o ti e gwÁ dokwÄ akouÂein (the speech of the Laws). but it is natural enough: both paÂnt' a ndra and idi ai carry emphasis (and perhaps there would have been a slight break in the clause. also Cri. . for which cf. on b6±7 melethto n . . section ii. Rather conspicuously.. It is very frequent in Lg. as it plays an important thematic part in the cohesion of Clitophon's whole report: it refers back not only to the preceding speech but also to Clitophon's praise as last mentioned at a6. One may point out that e pime leia is useless once it is acknowledged that being overcome by desire is an alternative cause of injustice. 155a6. note on 410c8±d5.2) and is no longer a pojainoÂmenov kaiÁ kathgorwÄn. . e3 TauÄ t' ou n w SwÂkratev egwÁ o tan a kouÂw: the left-dislocation (removal from subordinate clause) of both tauÄ ta and e gw is unique in the Platonic corpus.2 n. but then again so would there have been if the pronoun had been placed in the otan clause. the opposite is derived as starting-point for dhmosi ai xumpaÂsav taÁv po leiv. 21b2 tauÄta gaÁr e gwÁ akouÂsav (the oracle) ± here the o tan clause seems to have been used instead of a participle in order to indicate the repetition of the speech.2. above. Lg. e kmeleth sousin.. . Similarly. .

D. e4 kaiÁ ma la a gamai kaiÁ qaumastwÄv w v e painwÄ: kai . As it is printed in our texts. `More colloquialisms. . Thesle¨. cf. e painwÄ . Intr. H. swÂmatov e pime leia is now censured. . e5±8 kaiÁ o po tan . . On the other hand. cf. . . section i. Tarrant. CQ 8 (1958) 158±60 at 159) indicates that this is irony (cf. Smp. 180b1 qaumaÂzousi kaiÁ a gantai. Intr. e spoudake nai: this is the second of the three parts of Socrates' speech (cf. after crh mata. 393b7±8 o tan taÁv rh seiv e kaÂstote le ghi. . We can supply one easily enough: an equivalent of the preceding a gamai kaiÁ .5. . not to misunderstanding. R.C O M M E N T A R Y: 407e4±407e5 compatible with otan akou w: what the author meant is not `any time I hear you saying this often' but `any time I hear you saying this. had it been introduced by o po te au jai hv . we have a temporal protasis without a main clause to follow it. would have lost its lively character. .2. . but extremely forced. It is theoretically possible. o tan akou w and o po tan au jhÄiv (kaiÁ dhÁ kai would have served the purpose). cf. now that he is a pupil of Thrasymachus. section ii. we should better accept it as a contaminated construction. . . 289 . Studies on Intensi®cation in Early and Classical Greek (Helsingfors±Copenhagen 1954). The introduction of the speech was in the imperfect tense with the distributive optative for the subordinate clause ± its conclusion is in the present and the distributive subjunctive (cf. cf. o te le goiv . o po tan' does betray just such a misunderstanding). semi-proverbs and word-play in Plato'.3).. . It is impossible to attach the sentence to these very words by explaining kai (e5) as `so too' (Bury. . la Magna's note `integra: KaiÁ agamai kaiÁ e painwÄ se. note on 407a6 e xeplhtto mhn a kou wn. there is hardly any clear transition between the second part and the third. e teleu ta. as we would then desire a particle to coordinate tauÄ t' . Souilhe and many other translations ± I take it these translations are due to embarrassment.2).. 56) all point to `pathetic style' (Thesle¨. also 408c2 hgouÄ mai ): Clitophon wishes to express that his admiration for Socrates' speeches has not ceased. e painwÄ (e3±4) and by the change from direct to reported speech. e painwÄ does more than just amplify agamai: it also points back to 406a7 taÁ deÁ kaiÁ e ph inoun. Apart from this there is also a technical cause: the `semi-indirect discourse' of 407e5±408b5. qaumastwÄv wv (colloquial. kai coupling semantically related verbs. . which is often'. Cf. Just how ¯uid the transition is can be established only after the structure of this sentence has been elucidated. to detach qamaÁ le gontov from a kou w sou and explain it as a participle of circumstance. the parallel structure. ma la (instead of paÂnu or sjo dra. Styles. 70±1). This part is set o¨ from the ®rst one in a very clear way: by the intervening appraisal tauÄ t' ou n .

the anacoluthon is here less harsh because the speech reported is much longer and it is only natural that from 408a1 kaiÁ dhÁ kai (or at any rate from a4 kaiÁ teleutaÄ i ) Clitophon should have abandoned the formality of the nominal w v clauses. so the best he could do was to lump this (second) part and the third part of the speech together in one allegro sentence. a rxo menon: the futures indicate that Socrates refers to 290 . index s. it is marked only by a renewal (slightly varied) of kaiÁ o po tan au jhÄiv. J. see Intr. .g. section ii. 34d8±9 toÁ d' e jexhÄ v touÂtoiv peirw meqa le gein. for which cf. note on 410d3 toÁ e jexhÄv.. would make the lack of proportion between the second and the third part too obvious. then.C O M M E N T A R Y: 407e5±407e7 Besides. . It would seem. the case is not isolated: the next sentence. the author did not have very much to say on the subject of swÂmatov e pime leia. Intr. the transition from the second part of Socrates' speech to the third is not very noticeable.2. e5 toÁ ejexhÄ v touÂtwi: cf. on the sole authority of Ven. . a rxontov . W. . 189 and W. Phlb. both this full stop and the absence of an apodosis to follow kaiÁ o po tan au jhÄiv ktl. e pegei rein h maÄv is the ± long expected ± equivalent of a gamai kaiÁ . e6±7 e tero n ti praÂttein toiouÄ ton: viz. 17±18. the ®rst one very harsh. that we are dealing with two subsequent anacolutha. There is a good reason for this lack of markers of transition. Ti. both secondary MSS. cf. on Lg. section ii. is essentially parallel: kaiÁ o tan le ghiv ktl.2. Obviously. though more complicated. e painwÄ which was needed. Verdenius. In fact. has no apodosis. `Notes on Epitrepontes'. but this brings in irrelevant information.3. For the question why indirect report was chosen for these two parts. the second tolerably justi®ed. 460d8. The particle is super¯uous (explanatory asyndeton. Cf. `asyndeton. ou deÁ n twÄn deo ntwn (b1±2) ± caring for inessential things.v. e7 touÄ meÁn a rxontov: Souilhe prints touÄ meÁ n gaÁr arxontov against the consensus of ADF Themistius Stobaeus.2. for parallels in thought to the second part. A reduction is desirable and as a matter of fact only the full stop which is printed after e spoudake nai stands in its way. Not adverbial as e. R. explanatory'). to his references add England's Comm. . A colon or a comma would improve the text by totally removing the ®rst anacoluthon. 30c2 taÁ touÂtoiv e jexhÄv hmiÄ n lekte on.2... La Magna makes the phrase refer to the ®nal part of the ®rst speech only. If this solution is accepted. 408b5±c4 touÂtoiv dhÁ toiÄ v lo goiv . . When we assume that the second and third parts were separated from each other by a full stop.

Use of eyes and ears as the ®rst example is slightly awkward (the question of how one can make a bad use of one's eyes and ears at all is not. le gwn: see Intr. . e8±408b5 kaiÁ o tan . Humbert. for the anacoluthon contained in this passage. e8 e spoudakeÂnai: in classical prose.4. 281d1 kaiÁ ambluÁ o rwÄ n kaiÁ akou wn [sc. e9±12 ei dh . and could not easily be. 297). Schenkeveld (Amsterdam 1988). J. if their meaning involves having an opinion of sorts. e8±9 e pi statai crhÄsqai: short for e pi statai orqwÄv crhÄsqai.1 xeiÄ ne A meÁ n arcontav e cein. ii 46±57) tend to suggest that a substantive future participle (type ou d' o kwlu swn pa ra) can occur in this way only when a word denoting presence. or even `that which should rule' (Gonzalez). 430 compares Hdt.). on R. The conditions in which a modal future can occur have never been thoroughly studied. for the general principle and its applications. is either shall or will ') or from Latin (e.C O M M E N T A R Y: 407e8±407e9 physical training as part of the education of the young.g. crhÄsqai: cf. Cobet. in S. ± For the thought. Futur grec. Magnien. Sluiter (ed. cf. `that which is destined to rule' (so la Magna and most translators). Somewhere in this passage the hypotactic force of the ®rst w v ceases to exist. `Ad Themistii  qhnaiÄ e. absence. note on 407e5±8. 3. . 235d. it is not 291 .c (where futures are listed which denote that the subject `vermoÈ ge seiner Bescha¨enheit oder nach Lage der VerhaÈltnisse' can or must execute an action. . touÁ v deÁ a rxome nouv ou k e xein.] maÄ llon h o xu . Magnien's Futur grec is too biased and besides recognises only the modality of volition (though in his translations those of possibility and destination inevitably crop up).g.3. M. Slings±I.4. answered). I prefer this explanation to taking the futures as modal. Mx. section ii. Euthd. 176±7.162. 524e.3. cf. . Ophelos: zes studies voor D.2. Syntaxe grecque (Paris 1960 3 ). 289a2 and passim. ei dh (e9) probably started as a parenthesis giving a few applications of the general principle stated just before. i 115: `The future . verbs expressing states. Cf. It may or may not be true that the perfect has an `intensive' value. . At any rate the examples adduced by KG i 175. u meiÄ v oi kate touÁv orationes'. Euthd. 7. also KuÈhner on X. . need is in the immediate vicinity. the phenomenon has been too often approached from the point of view of a Western European language (e.. cf. id. Mem. the same principle applied to the same ®eld of human ability. Stallbaum on Pl. Gildersleeve. my paper `Het perfectum van Griekse toestandswerkwoorden'. 61±76. R. normally have a perfect with the same meaning. because I doubt that a substantive future participle can be modal in this type of context. e laÂttw an praÂttoi and therefore e laÂttw an e xamarta noi etc.

. Euthd. Simonides. KG i 422) is a related but more stereotyped idiom. which gives a solemn e¨ect (Thesle¨. leÂgwn: stylistically this sentence is a little more lively than the preceding ones: note the (slightly ironical) interaction created by dh (three times) and a ra. kalwÄv. a5 soi: the dative should be compared with Hp. 363b2±3 h  I liaÁ v ka llion [cf. 81). . . . 419b2 does not belong: the whole discussion from 418a5 onwards was meant to prepare the etymology of zhmiwÄdev). 287e1 o ti moi nooiÄ toÁ r hÄma (which disproves Stallbaum's note on R. the dative of the author in quotations (type  O dusseuÁ v le gei  OmhÂrwi. loc. section i. section ii. Lg. . See Intr. a4 kaiÁ . Water®eld) is meant. cf. not marking a climax.3. e11±12 mh t' allhn crei an mhdemi an crhÄsqai: for the construction cf.C O M M E N T A R Y: 407e11±408a5 insigni®cant that it is actually the last example in Euthd. 408a1 kaiÁ dhÁ kai : transitional. is an important one' (GP 2 253). 785b7.. The awkwardness stems from the author's wish to cling to the pattern body ± tools ± soul (cf. kalwÄ v here] ei h poi hma twÄi  OmhÂrwi h h  Odu sseia (cf.. Intr. he must have had something in mind. cf.2.. a4±5 o lo gov outov: the entire speech (not `this argument'. the absence of the article before yuchÄi (a5). so la Magna. dikastikh and dikaio- 292 . . a2 dhÄ lon w v: cf. the rudder/steersman metaphor. 335e2 touÄ to deÁ dhÁ noeiÄ au twÄi (sc. The end of the sentence reveals his intention: the phrase o stiv yuchÄ i mhÁ e pi statai crhÄsqai is picked up by twÄ i toiou twi (b1). cf.3). yuchÄi . note on 407a2. R. as he chose both the principle and the scheme for the sake of parodying the `beautiful ending'. Quite probably the author did not go to much trouble to clarify the thought. . . because the phrase is an automatic result of applying the general principle o twi tiv mhÁ e pi statai crhÄ sqai ktl.2. . dhÂ: this is a collocation of connective kai and dh marking (with some degree of irony) a report or quotation as well-known or even subscribed to by the speaker (GP 2 233±5. 407c2 and note) rather than the combination kai . as it fails to explain kaiÁ mhÁ zhÄ n. consequently the man who does not know how to use his soul should surrender himself twÄ i maqo nti thÁn twÄ n a nqrw pwn kubernhtikhÂn which is equalled with politikhÂ. Jowett± Campbell.5. Republic.cit.3. GP 2 256 (where Cra.). 868b7. ii 186). section ii. On the other hand. to the general scheme which he found in Republic 1. a4±b5 kaiÁ teleutaÄi .. KG i 429).2. .Mi. the postposition of ara. crhÄsqai: we are left in the dark as to what exactly is meant by this phrase ± surely Water®eld's `how to use his mind' is entirely beside the point. Styles. dh which indicates that `the addition . Intr. cf.

356e1).1). Intr. cf. hsuci a in connection with the soul is normally tranquillitas animi (R. For Chrysippus' interpretation. This is a good illustration of the wide gap between Aristotelian protreptic (foreshadowed in the Euthydemus).1. cf. Classen±Steup on Th. ± kata with a re¯exive pronoun (`all by oneself '. i 310±11. Grg. . 653d8. which is ®rst and foremost ethical (cf. C. Intr. Pl. Boyance . but cf. zhÄn by itself would have been su½cient. allwi ). P. sections ii.7.C O M M E N T A R Y: 408a6±408a7 su nh. CQ 20 (1970) 216±20. ± For the dative (of respect) with hsuci an agein. . cf. if only from Phd. twÄ i maqo nti ) justice. 1. The discrepancy is instructive for the history of philosophical protreptic and the place of Clit. 61c9±62c8. The phrase yuchÄv crhÄ siv is used in a completely di¨erent sense by Aristotle (Protr. a6 h suci an a gein thÄ i yuchÄ i: in Plato. `Who did forbid suicide at Phaedo 62b?'. RPh 37 (1963) 7±11. 160a1±2) and mental laziness (Tht. J. The way in which the expression is used here is apparently a novel one (it had to be glossed by kaiÁ [explanatory] mhÁ zhÄ n). 378a4. Prt. . which appeals to the intellect and emphasises theoretical knowledge.79. cf. and protreptic in the traditional Socratic way. 791a3±4. as suggested by Bury on Smp. in it. Lg. as well as with the subsequent discussion on the areth of the soul (cf. section ii. There is a conscious play upon the double function of the soul as the seat of thinking (called dia noia a few lines below) and as the principle of life. A vague reference to rejection of suicide. ii. A similar expression in a similar context Lg. quite probably independently from each other.1) is usually reinforced with au to v in Plato. R. b 90±1 DuÈring): `there are many ways of using one's soul but the truest (kuriwta th) is philosophy (h touÄ jroneiÄ n oti ma lista)'. For other testimonies which are important for the history of this rejection. 505d9 le gwn kataÁ sautoÂn (not `after your own fashion'. . The author of Clit. This is in harmony with the tenor of the ®rst part of the speech. Protr. it may also denote slow intelligence (Chrm. and Aristotle use the same protreptic motif (e pi stasqai crhÄsqai ) and in the same area (yuchÂ). Cf. Lg. R. G. `Note sur la jroura platonicienne'. ± The optative in the ei clause is combined with a present indicative of an expression 293 . 153b11±c1). That is to say `using (ˆ rightly using) one's soul' amounts to learning (cf. 732b2 au toiÁ praÂttontev. a7 praÂttonti kaq' au to n: these words point forward to the second alternative (they are to be contrasted with paradoÂnti .1 (4).. 353d3±10 and Tucker's note. 199b1). especially 409a2±6). Arist.5. Strachan. b 83 DuÈring kaiÁ zhÄ n ara maÄllon jate on . 583c7±8.. KG i 480. which was of course well-known to be Platonic. HGPh. toÁ n e nergouÄnta thÄ i yuchÄi touÄ mo non e contov.7. ei de tiv a naÂgkh zhÄn ei h: cf.

Ma.C O M M E N T A R Y: 408b1±408b2 denoting propriety in the main clause. M. alloiv paradidoÂasin (in Chrm. and at c4.. 82). e3. 41 (1988) 409±14 at 412. is too short for us to rely on criteria based on ratios. The general value of this stylometric criterion (discovered by Dittenberger. 732a7±b2. a specialisation of the general value of ara to denote a distance between the speaker and his utterance (cf. 775b3. cf. Hermes 16 (1881) 321±45 at 337±9. 1 117c9±e6. The Charmides of Plato (Amsterdam 1985). . 410e4. . b1 kaqaÂper: found in this place. Particules de liaison. 171d8±e5 and Lg. . our place and Alc. o ti pra ttousin]. b1±2 kaqaÂper . where the same thought is formulated. a defensible reading. It cannot be a coincidence that in all three cases a ra is preceded by e stin. Lg. 283b2±3 tou tou d' o rov e stiÁ n a ra.. des Places. What is meant in Clit. 409b7. cf. 709e3 ti metaÁ touÄ t' ei peiÄ n o rqwÄ v e stin ara. ii 208. but the optative conveys a nuance of uncertainty which is in harmony with the context. 408c3. no mention of steersmen is made). twÄ i kubernhÂthi e pitre yav an hsuci an agoiv . Yet in this case I am inclined to share SouilheÂ's scepticism (180): the Clit. 555. (e1) o tan de ge pou tinev mhÁ oi wntai ei de nai [sc. All these parallel passages deal with handing over what one does not know how to use in order to avoid a marth mata. it is what is usually called `referential' ara. van der Ben. At the end of Alc. section ii. . Brandwood. b2 taÁ phda lia thÄ v dianoi av: Greek ships of the classical age had 294 . it would seem that what is there called toÁ arcesqai . Its place at the end of a clause is curious but not unparalleled in Plato: Hp.3. 488c2. J.3). Mnem. wsper 407a8. cf.7. my review of N. See also Prt. 355b4±5 e aÁn mhÁ polloiÄ v o no masi crw meqa ara.2. is determined by the exact reference of thÁ n twÄn a nqrw pwn kubernhtikh n (Intr. Ti d' ei e n nhiÊ ple oiv . Chronolog y. esp. MT §§502. ei h is omitted by Stob. .2. KG ii 324). The thought recurs at Alc. R. u poÁ touÄ belti onov (135b7±8) amounts to becoming a close companion of this better man. . a llwi: for the expression cf. cf. 19 and table 4. . b1 a ra: Denniston's classi®cation of this occurrence under a ra expressing surprise (GP 2 36) is not precise enough. 1 Alcibiades promises that he will from now on follow Socrates like a paidagwgoÂv (135d10) and make a start with dikaiosu nhv e pimeleiÄ sqai. 1 135b7±d10 concern persons turning over themselves to others. without it being made very clear what we are to understand by that. section ii. Republic. Lg. 268±72.. where many editors prefer the Byzantine conjecture ama. . `Sprachliche Kriterien fuÈr die Chronologie der Platonischen Dialoge'. Intr.3) is beyond doubt. `used to direct attention to the fact that the speaker is not uttering his own thought' ( Jowett±Campbell. van Ophuijsen in Two Studies.

This 295 .. b3 dhÂ: not. To my mind. S. as at 407c2 and 408a4 (cf. . frequently give to politics' (cf. 155±6. cf.v. to wit in Grg. a mousi av). . van Ophuijsen in Two Studies. b3±4 h n dhÁ . Platonic) doctrine. K. Williams. But these scholars saw the di½culty: why would Clitophon state in so many words that Socrates used to call the leading of men politics? The answer may be found in the interpretation of politikh that I have given in Intr. M. but simply an indication that this is knowledge shared between Clitophon and Socrates (cf. . le gwn: for the comparison of a politician to a steersman (announced b1±2). J. notes). . Orwin. T. (Cambridge 1968). (b) by his inclination to make neat parcels of dependent clauses (cf. 272e4.3. Socrates. . hence the ®nal position of le gwn. 33d7. and Plt. `Notes on Politeia. b2±5 twÄi maqoÂnti . The order as we have it is determined (a) by the author's wish to emphasise the idea `the very same art' (which ± quite normally ± causes anticipation). 434b2. Louis. Les meÂtaphores de Platon (Rennes 1945). Cf. Morrison±R. 291±2 and General Index s. But the article is necessary after a llov. . non vidi ) proposes allwi twi maqo nti. b4±5 thÁn au thÁn .3. Kaiser. Similarly in the next line and at 409b5 (see note). Euthd. note on 407a6 pollaÂkiv. or that it is a particularly important piece of Socratic (in this case. le gwn: a highly unusual word order. cf..3: when the metaphor in thÁ n . for Plato: P.2. `steering oar'. . Pavlu. pollaÂkiv: Bury takes h n as predicative: `which is the name that you. The expression is modelled on taÁ phda lia touÄ ploi ou.c . . Cf.C O M M E N T A R Y: 408b2±408b4 a pair of steering oars. Kleitophon'. my note on R. note on 407b8±c6 all' . a llwi. M. cf. 12: `Socrates nennt also die Politik hauÈ®g so [kubernhtikhÂ]').3. `Pseudopl. on 407c2). the most normal order would be thÁ n au thÁn dhÁ tau thn le gwn w v e stin ktl. I doubt whether this is possible in Greek (o or h i would certainly be more normal). Das Bild des Steuermannes in der antiken Literatur (Erlangen 1954). iv'. . the ironic dh used in referring to other people's opinions. twÄ i maqo nti: A. The consequence is that not only the subject but also the nominal predicate of the subordinate clause is anticipated. to which you often refer with the word politikh ' ± this implies that Socrates used the term in a di¨erent sense from that used by most Greeks. See further Intr. polla kiv can mean that this concept of politics is found in more than one Socratic text. section ii. I think. 85). or perhaps indicating that Clitophon agrees with Socrates (personal commitment of speaker and addressee.2. section ii. . de Strycker±Slings on Ap.. Gasda. we know for a fact that he did so. kubernhtikhÂn is translated into plain language. the statement makes good sense: `the leading (educating) of men. Greek Oared Ships 900±322 b . J. hence the plural as in Plt. Kritische Bemerkungen zu Themistios (Lauban 1886±7. 415.

Richards' opinion `there is no possible construction for the accusative' is mistaken. b5±c4 TouÂtoiv . . c2 te: Richards' de (Platonica. e non senza una tinta d'ironia. E. 639a2±3 ei tiv ai gwÄn trojhÁ n kaiÁ toÁ zwÄion au toÁ kthÄ ma wv e stin kaloÁn e painoiÄ ktl. Earth) oi khthÄ rav a spidhjo rouv pistouÁ v o pwv ge noisqe proÁv cre ov toÂde (M. la Magna). (mh )pote. 157. . I have found one case of prolepsis of subject and part of the nominal predicate in Plato: Lg.5. b7 didaktoÁ n a reth : cf. Bertini. . A. 5. ii. Protreptiques. It is of course the central motif of all Socratic protreptic.) super¯uous. Brandwood. Intr. . Intr. . Hutchinson's discussion ends up in aporia. See Intr.v. . Cri. ± scedoÂn ti sou te qnhken. Plato's Logic. c1 scedoÂn instead of scedoÂn ti (also 410e7) has been considered a mark of late style (Ritter. la recisa a¨ermazione che segue'. M. .2. s. Untersuchungen. `I dare say never' (`serve ad attenuare in certo qual modo.. section i.. 19±20 e qre yat' (sc. .3. Th. cf.. 2. West accepts the transmitted text. MuÈller. Aspis 420±1 a deljoÂv ± w ZeuÄ. (protreptikw)ta touv . 3. b7±8 pa ntwn . scedo n. Anm. Schmidt's conjecture at A.2). pag(kaÂlwv) (cf. cf. is de®nitely unlikely) seems to me a virtually certain example of prolepsis of the nominal predicate: Verrall's `in order that ye might be formed (. 157) and Schanz' hovi pro- 296 .. . cf. (w jelimw)taÂtouv indicate that irony is present.19. In combination with paÂmpoluv: Hp. sections ii. England on Lg. . cf. With ou t' . and his proposal to read dikastikh te kaiÁ dikaiosu nh (Platonica.1.5. pwÄ v jraÂsw. IT 951±2 is textually uncertain and even if the transmitted text is retained. accepted by Page.3.2. c1±2 out' oi mai mh pote usteron antei pw: even now that I have turned my back on you and gone over to Thrasymachus. Chrm. . . table 10.. . .. 5. H. b6 pagkaÂlwv: a favourite vehicle of Platonic irony. Stein on Hdt. De Win) but `probably never'. it admits of another construction. ou t' . Th. LSJ. section ii.1. An extreme example is Men. cf. epimeleiÄ sqai: this had been the recurrent unifying motif of the speech reported by Clitophon. with the usual understatement. Phd. note on 410c8±d5. section ii. even so it is found in Ap. . ..2. Chronolog y.Ma. 188. ii 2. dikastikh n te kaiÁ dikaiosu nhn: cf. Of the two examples given there. On the other hand.) against (with a view to) this occasion' gives more semantic weight to ge noisqe than this verb can bear in Greek. note on b6). .4. pwÂpote: not `hardly' (Bury and Orwin. 28 n. Schmidt's pistoi q'. 286b3±4 pampoÂlla . Grg. Lutosøawski. . KG ii 580. (pw )pote . FestugieÁre. Intr..2 n. 124.2. . 649a2.C O M M E N T A R Y: 408b5±408c2 is rather unusual in Greek. kaiÁ paÂgkala. It would seem that H. . h maÄ v: the parallel duplications in pam(po lloiv) . ou t' . cf.

even though they are not worked in very well (cf. and word-play in Plato'. section ii.3.1. Dionysodorus begins] kaiÁ hmeiÄ v pa ntev e ble pomen proÁv au toÁ n wv au ti ka ma la akousoÂmenoi qaumasi ouv tinaÁ v lo gouv (similarity in wording). Intr. . 59a5. D. the sentence remains of course a confused one. 1. It might seem that the relation between main and concomitant action has been inverted (`I asked my question while being very attentive'). cf. but this goes far beyond the type of inversions listed KG ii 98±9.6. I take it that the author wanted to give a hint that what Clitophon is going to hear is equally disappointing. This means that the author did more than just draw from a general store of protreptic themes (of 297 . Tarrant. cit. semiproverbs. CQ 40 (1946) 109±17 at 114±15). in these contexts. . h maÄ v: there can be no doubt that the author was thinking of Ap.. 923a3. 283a1±7. X. ou te .2.4. it marks the following simile the more sharply (the usual function of the word in Plato: cf. . The sentence could only be understood by those who had read (and were fairly conversant with) the Euthydemus. 2. The only way I can account for this confusion is to explain it as an imitation of Euthd. Now what actually happens is that Dionysodorus' (and his brother's) arguments are utterly disappointing. . 30e2±5. . Cyr. though in view of the status of that MS the conjecture may be an old one. c3 a tecnwÄv: a colloquial word (D. Tarrant. cf. . 115 (d)).3. te gaÁ r of Pa (te Va) is probably conjectural. . There is nothing against ou te .3. . . There Socrates has ®nished his protreptic conversation with Clinias (similarity in situation): twÄi deÁ metaÁ touÄ to e some nwi pa nu sjo dra proseiÄ con toÁn nouÄ n [. c3±4 wsper . Burnet on Phd. Hdt. Intr. . section ii. This interpretation has of course some important consequences. See section ii. c4±7 proseiÄ con .. If indeed the author wanted to allude to the situation of Euthd. 90c4.. .C O M M E N T A R Y: 408c3±408c4 treptikwta touv te are unnecessary. akouso menov) gives the impression that Clitophon is waiting for Socrates to deliver other speeches in which he will go more deeply into the matter of e pime leia thÄ v yuchÄv (which had been the refrain of the speech reported).2. 4. England on Lg.6. . te does not necessarily introduce an opposition. 283a¨.42. But this impression is ¯atly contradicted by the next participle (e panerwtwÄn). An. o nomaÂzein: this is an odd sentence at ®rst sight.3 (neighbour's lyre) and note on 407b2±8).3(5) for the relevance of this habit for the problem of authenticity. but we have noticed before that our author is quite willing to sacri®ce clarity of thought and expression when he has an opportunity for working in more literary allusions. slightly out of character in this formal sentence. op.1. The ®rst clause (proseiÄ con . te (KG ii 292). esp. `Colloquialisms.7.

(SSR vi a 53). but the opposite: Socrates' coevals are opposed to his `pupils' (sunepiqumhtwÄn h . is e rasthÂv. `genitivo partitivo. but cf. e taiÄ rov . . . note on 410b4±6. e k ne ou. . Cri. Ap. . 906d an au toiÄ v twÄn adikhmaÂtwn tiv a pone mhi `provided one gives them a part of their unjust gains'. fr. xcv) possibly so. kai or kai . xxxiii.) kai .. 28 where Apollodorus is called e piqumhth v . Ap. sundiatri bontev (the usual Socratic surrogate for maqhtai ) and e piqumhth v. Meyer. applies it to Socrates' followers: Mem.) kai . but not consistently. dipendente da un tinaÂv sottinteso'. KG ii 251) coupling h likiwtwÄ n and sunepiqumhtwÄ n as against e tai rwn.1. h likiwtwÄn: not many of Socrates' companions were in fact of his age: Crito (Ap. Th. . . thÄv shÄv soji av). c4 toÁ metaÁ tauÄta wv a kouso menov: note the inverted word order. . . if it is indeed an imitation of Euthd. . Socrates). labwÁn oude na pw pote misqoÁ n thÄ v sunousi av e praÂxato. he appears to be co-owner of the jrontisth rion in Aristophanes' Clouds. of course. Euthphr.2. .60 e keiÄ nov [sc. . 14d4 e piqumhth v . c5±6 twÄn h likiwtwÄ n . h : not (as is usually the case with te (. Schw. 298 .C O M M E N T A R Y: 408c4±408c6 course I do not wish to deny that he did): one of his sources was actually the Euthydemus itself. . . Mem. . 11c D. But Xenophon does use it this way. X. ou ti se : see Intr. Both Plato and Xenophon avoid the word maqhthÂv when speaking of Socrates' companions. sunepiqumhtwÄn: a hapax word. . Better known. 195. . onoma zein: an allusion (a rather broad one) to Socrates' denial of having maqhtai (cf. . 102±3. KG i 345. Lg. . Ap. cf. Socrates] gaÁr pollouÁ v e piqumhtaÂv . 49a9) is the only one for certain.3 kai toi ge ou depwÂpote upe sceto didaÂskalov einai tou tou. The participle is probably not ®nal (so la Magna and most translators). i scurwÄv au touÄ (sc. te (.). Cf. c6 te kai . c6±7 sunepiqumhtwÄn . contaminated from suno ntev. cf. when in Tht. Dover's Comm.5. . la Magna ± I doubt that the second half of this explanation is linguistically correct. h .±D. it is di½cult to see the di¨erence (apart from the question of payment). . Ficinus. Alc. 134 n. 33d10. and Sph. Plato theorises about Socrates' work. . swÄn: on the partitive genitive as object instead of the accusative (te mnw thÄ v ghÄv for te mnw thÁn ghÄn). . Yet. 1. ˆ 4Kr. Examples in Plato are rare. cf. . cf. 1.2. 33a4±5 ou v dhÁ diabaÂllontev e me jasin e mouÁ v maqhtaÁv ei nai ). Chaerephon (e mo v . . . which Plato never uses absolutely (unlike e rasthÂv) but always with the genitive of an abstract noun or phrase (e. This is generally Socratic. 20e8±21a1. . cf. 283a6±7: `Ma io stava [sic] attento nella speranza di udire cioÁ che viene in seguito' (Bertini). Cf. Aeschines. Half-way between this absolute use and Plato's practice stands X. what is more. . . Apologie. and.g.. section ii.

loc. Men. . note on 407a5 gaÂr. cit. Hp. Thesle¨. Platonica.. 925e2. 15 n. . 1. Heidel. R. There is no good reason to consider this ad hoc creation a mark of inauthenticity (Steinhart. present. 282d4.3. Cf. toi nun) would perhaps seem more normal. Grg. 516 n. 2). yet it does occur: Cri. u potei nwn: `propounding' (Bury). where Gorgias is meant).2. 674c2.. Richards. quicumque (Monteil. which occurs at Grg. 3.5.Ma. 173b3±4 Swkra touv e rasthÁ v wn e n toiÄ v maÂlista twÄn to te). suggest'. Onom. sometimes also to teachers (cf. A resumptive particle (dh . 157). 79e6. . sunepiqumhth v is quoted from Plato by Pollux. o nomaÂzein. 299 . Intr. 72 n. section i. `RaÈtsel'. 353c5. note on 406a5. Lg. 5. It ought to mean `fellowadmirer' (in the sense of summaqhth v and sumjoiththÂv) but obviously it does not. u pelaÂmbanon proÁv touÁ v ji louv mo non tauÄta e rwtaÄn (`I thought you were putting these questions only in relation to your friends'). c8 touÂv ti: this (or perhaps touÁ v ti ) should be printed. 101±4. ii 2: `lay or put before one.' To some extent ga r mitigates the incongruence we detected in the preceding sentence. `die Sache angreifend' (Schleiermacher). 346c5. Styles. 8. cf. Mem. . u potei nw. LSJ s. 4. 919d7. after the authorial comment h opwv . ou n. Tht. o pwv deiÄ : `we should expect o pwv dhÁ deiÄ or opwv a n de hi' (H. 633a9. etai rwn: referring to pupils. the phrase o poÂqen kaiÁ o phi cai reton o nomazoÂmenoi and its numerous variations (Stallbaum on Euthd. ii 339 n. c7 proÁv seÂ: `in relation to you'. 131±3). not touÁ v tiÁ . cf. X. Cf. Phrase relative.C O M M E N T A R Y: 408c7±408d1 which is used by Plato both for the sophists' pupils and for companions of Socrates (Smp. The sentence provides the justi®cation for proseiÄ con toÁ n nouÄ n: `I paid good attention . my remarks in New Approaches. d1 kataÁ se : cf. 433 n. Xenophon eschews this reference of the word e rasthÂv. Susemihl. 448e8. .. The formation is more allusive than precise. Curiously enough. Pseudo-Platonica. `je leur pre sentais mes di½culte s' (Souilhe ). This is partly right: o stiv (etc. 39. `ita praefatus quaestionem exposui' (Ficinus. It should be connected with u potei nesqai. But gaÂr is used because it marks the return to the narrative level. 503e6 ± esp. `propose a question'.69.v. 288b1).) is not frequently equivalent to Lat. 179e2. In other words. for Grg. it is here used in its typical function of push particle. 49 n. cf.. Wilamowitz.15 e gwÁ se . but it does not iron it out altogether. Ast: `quomodocumque eorum ad te rationem [relationem?] appellari oportet'. For I ®rst questioned your best students. la Magna) ± evidently this word has embarrassed translators. . Platon. 50a6±7 ei te apodidra skein ei q' o pwv deiÄ onomaÂsai touÄto. Ge¨cken. ga r: e panhrwÂtwn resumes the thread of the narrative after the digression twÄ n hlikiwtwÄ n ktl.

C O M M E N T A R Y: 408d1
Let us start with the passage from Tht. `We should examine more
closely the doctrine of universal ¯ux; there seems to have been quite a
battle about it ± In Ionia it is even becoming very popular right now;
Heraclitus' school are strenuous defenders.' ± TwÄi toi w ji le QeoÂdwre
maÄ llon skepte on kaiÁ e x archÄ v, wsper autoiÁ upotei nontai. There is no
question to pose here; nor do I see how the Heracliteans can be said to
present or suggest anything ± the meaning is clearly: `they have demonstrated themselves how this doctrine is built up e x a rchÄv'. Excellently
Schleiermacher: `so wie sie ihn [den Satz] eigentlich vorzeichnen'.
A similar interpretation in Grg., loc. cit., is much more satisfactory
than the one given by LSJ. Chaerephon has asked Polus what one
should call Gorgias if he possesses the same te cnh as Herodicus (doctor),
or Aristophon (painter). What should we rightly call him? Polus starts
o¨ with an encomium on te cnh. Socrates intervenes: Polus should have
said what exactly Gorgias' te cnh was. w sper taÁ e mprosqe n soi upetei nato CairejwÄn kaiÁ au twÄ i kalwÄv kaiÁ diaÁ brace wn apekri nw, kaiÁ nuÄn
ou twv ei peÁ ti v h te cnh kaiÁ ti na Gorgi an kaleiÄ n crhÁ hmaÄv. To my
mind, u petei nato (which refers to the analogies which Chaerephon had
used) is not `submitted' but `shown (by his previous questions how to
answer this one)', `hinted'. Stallbaum tries to connect the two notions:
`interrogando et disputando proponere eoque aliquem sensim ad aliquid
perducere et quasi upotiqe nai s. u poba llein ei quid respondendum sit'
(ad loc.; similarly Dodds), but the parallel quoted from Tht. shows that
the notion of `proponere' is secondary.
The same holds for E. Or. 915: (an anonymous person had proposed
to stone Orestes and Electra) u poÁ d' e teine TundaÂrewv lo gouv twÄi sjwÁ
kataktei nonti toiouÂtouv le gein. On the analogy of Grg., loc. cit. I submit that lo gouv . . . toiou touv is object of u poÁ . . . e teine (and le gein
epexegetic in®nitive): `T. had inspired these words into him'; `T. had
led him on to say these words' (`suggested' is too weak). Whether or not
916 is deleted is immaterial for the meaning.
The way u potei nein is used in Clit. is similar to the situation in Grg.:
Clitophon asks a forthright question, illustrates it by using an analogy
from the care of the body (408e3±409a2) and thereby shows them
how to answer it. The meaning is therefore `demonstrating', `inducing':
`Indem ich nach deiner Weise die Antwort ihnen gewissermassen an die
Hand gab' (H. MuÈ ller); `dando loro, alla tua maniera, una spinta' (Zuretti). In Grg. and Tht. the middle voice is used; in E. Or. and Clit. the
active. It is quite normal for abstract compounds of which the verbum
simplex has a concrete meaning to waver between active and middle; in
Plato (e.g.) u perba llw, -baÂllomai; proti qhmi, -ti qemai; protre pw,
-tre pomai.


C O M M E N T A R Y: 408d1±408d2
d1±2  W beÂltistoi . . . u meiÄ v: as Socrates had started o¨ his speech
with an apostrophe borrowed from tragic diction, so Clitophon opens
his with an imitation of one. su occurs rather often after a vocative in
tragedy (paÂndurte su A. Th. 969 (w ) du sthne su , S. Ph. 759; E. Andr.
68; w kaÂkiste suÂ, E. Andr. 631; etc.) and in comedy (w bdelureÁ ka nai scunte kaiÁ tolmhreÁ su , Ar. Ran. 465; w be ltiste suÂ, Eub. fr. 105.1
K.±A.; Damox. fr. 2.17 K.±A.; Lyn. fr. 1.11 K.±A.), but I have found no
instance of u meiÄ v used that way, nor an instance of either su or umeiÄ v
following a vocative (where the vocative alone would su½ce) in Plato.
Nearest to this comes Euthd. 303c4 w maka rioi sjwÁ thÄv qaumasthÄ v ju sewv (which is di¨erent because of the genitive). ± For the punctuation,
cf. note on 407a5 w SwÂkrate v soi.
d2 nuÄ n: in the MSS, bracketed by Hermann and Modugno, and
indeed there is little point in a temporal adverb; at best `now that we
have been aroused [c3]' ± this may be what Ficinus meant by `deinde' ±
or `in the present circumstances', Verdenius, `Notes on Clitopho', 144 (his
comparison of this passage with 409a2 and 410d4 is not illuminating,
since in those places `now' is completely satisfactory). Probably nun
should be read. Brandwood (Word Index, s.v.) gives ®ve instances of it
(not counting dh nun) but no doubt there are many more (see Dodds on
Grg. 451a3), since Burnet (on whose text the Word Index is based) was extremely reluctant to print nun, accepting it after i qi, deuÄ ro, dh only. I
have found nuÄ n used in questions with (at most) a vague temporal reference at Sph. 253c6; Lg. 630b8; 835d1.
a podeco meqa: `understand', cf. Ast, Lexicon, s.v., ad ®n. If this is a
deliberative question, the present indicative is surprising (though not
excluded, cf. A. W. McWhorter, `A study of the so-called deliberative
type of question (ti poih sw;) as found in Aeschylus, Sophocles, and
Euripides', TAPhA 41 (1910) 157±67 at 165: `even a present indicative
may sometimes appear in a question, approaching the deliberative type;
cf. in Latin, quid ago? So ti le gomen; [cf. Lg. 652a5]'); cf. the very frequent use of jamen in questions in Plato. But it is quite possible that
Clitophon includes himself only for the sake of (ironical) politeness
(cf. Tht. 210b4 h oun e ti kuouÄ me n ti kaiÁ w di nomen w ji le; Jowett±
Campbell, Republic, ii 195±6; Wackernagel, Vorlesungen, i 42±4): in fact,
a pode cesqe is meant. Verdenius, `Notes on Clitopho', 144 suggests that
this is `a parody of Socrates' preference for the plural' ± he may well be
d2±3 thÁn SwkraÂtouv protrophÁn hmwÄ n: for collocation of subjective and objective genitives cf. R. 329b1 taÁv twÄ n oi kei wn prophlaki seiv touÄ gh rwv; 537c2±3 oi keioÂthtov . . . allhÂlwn twÄ n maqhmaÂtwn
(kinship of the studies); KG i 337. ± protroph is a hapax in the Platonic


C O M M E N T A R Y: 408d3±408d4
corpus (it is a worthless Byzantine conjecture at Lg. 920b2), but that
cannot be an argument against authenticity: given the fact that the
author started the sentence the way he did, with pwÄ v . . . a podecoÂmeqa
he had no option but to use the action noun. That it is not elsewhere
used in Plato is due to the fact that Plato seldom discusses exhortation
explicitly. The word is also found in Democritus and Aristotle.
d3±6 w v o ntov mo nou tou tou, epexelqeiÄ n deÁ ou k o n twÄi praÂgmati
. . . , a ll' . . . ete rouv: ABA structure, cf. Dodds on Grg. 452e6 (and
Index s.v. aba structure); de Strycker±Slings on Ap. 40b4±6; J. Th.
Kakridis, Der thukydideische Epitaphios (Munich 1961), 29. For the purpose
of this dilemma, cf. Intr., section i.5.3. This structure is disjunctively
opposed to what follows (d7 h deiÄ ktl.), hence a comma after e te rouv is
better than a question-mark.
d4 o n: a verb meaning `it is possible' is necessary, as was seen by
Bessarion ± I take it that his proposal e ni was supposed to mean this.
The more original meaning `to be present in' is out of place and seems
to require the article with the in®nitive (cf. R. 431e4). As a matter of
fact, e neiÄ nai meaning `to be possible' is found only once in the genuine
works of Plato (Lg. 646d6), though it is well-established already in the
®fth century and occurs frequently in Plato's contemporaries ± it seems
that Plato consciously avoided this usage at least until his old age
(among the Dubia it recurs at Ep. 12 359d2). As long as Plato's authorship
is not excluded beyond doubt, we should be hesitant about excepting
e ni. Ficinus translates the transmitted text: `an quasi solum hoc exstet,
prodire vero ad opus (. . .) minime'. But the coordination of tou tou and
the in®nitive without article seems strange and e n does not elsewhere
follow e pexe rcomai or -eimi.
Therefore, the best solution seems to me to suppose that an original
on was corrupted to e n. I wonder if the corruption to e n has not also
caused a syntactic harmonisation: in other words, if the original text
was e pexelqeiÄ n deÁ ou k on twÄ i pra gmati or toÁ praÄ gma. Both constructions are attested in Plato. With the dative it would have meant `to
pursue the matter' (cf. R. 366e8; Ly. 215e1), with the accusative `to go
through the matter; to investigate it completely' (cf. LSJ s.vv. e pe xeimi
iii 2; e pexe rcomai ii 3). It so happens that the context (especially the
following words kaiÁ labeiÄ n autoÁ tele wv) admits of both possibilities,
though perhaps the dative would be slightly preferable from a stylistic
point of view. Besides, the dative has the advantage of providing an
easier explanation for the corruption.
In the ®rst version of this book, I proposed e hstiin for e n, which was
endorsed by Gonzalez. My new proposal has two advantages over the


C O M M E N T A R Y: 408d6±408e2
former: (1) o n as an accusative absolute (`it [not] being possible') is syntactically on a par with the preceding o ntov, as logically it should be:
both are modi®ed by wv; (2) on better explains the corruption to e n than
e stin does. KG ii 89 quotes on thus used (ˆ e xo n, dunatoÁn o n) at D.
50.22; I have not found a parallel in Plato. For wv modifying ®rst o ntov
then o n (but as a copula, not a substantive noun, as in my proposal), cf.
R. 604b10±c1 wv oute dhÂlou o ntov . . . oute . . . a xion on ktl.
labeiÄ n au toÁ teleÂwv: tele wv auto Va against ADF has weak
d6 ekei nouv: against this reading of ADF, Schanz adopted e kei noiv,
found in Va, possibly an old reading in view of the status of that MS,
but cf. Riddell, `Digest', §185; Jowett±Campbell, Republic, ii 240. After
the in®nitive protre pein, the thought changes to `it will be inevitable'
or the like.
d7 toÁ metaÁ touÄt' e panerwtaÄn: on account of ti tou nteuÄqen, `ought
we to ask . . . the further question' (Bury; cf. H. MuÈ ller; la Magna) is less
likely than `sollen wir nicht . . . nun auch des weitern ausfragen'
(Schleiermacher; so most translators). ± e perwtaÄn (A) is not found elsewhere in Clit. and is therefore probably false.
e1 touÄt' autoÂ: the author is apparently indi¨erent to precise pronominal reference. His meaning is clear enough: d3 tou tou: toÁ protre pein; d4 au to : toÁ praÄ gma ˆ the object of protroph e p' a rethÂn; d5
touÄ t': touÁv mh pw protetramme nouv protre pein; e1 touÄ t' au to : the
object of protrophÂ. This object can be equated with areth ± as
the analogy e3±10 suggests, a reth is to be taken as the best state of the
soul. Cf. next note and note on e8 touÂtou . . . au touÄ .
e2±3 dikaiosu nhv peÂri maqhÂsewv: for pe ri with the genitive instead
of the simple genitive (used here perhaps to avoid two subsequent genitives), cf. England on Lg. 676c6 (and Index s.v. peri c. gen.). dikaiosu nhv takes up e p' areth n (d3); at 408a3 we ®nd again thÄi thÄv yuchÄ v
arethÄ i. This is understandable: Socrates had been exhorting his audience to the e pime leia thÄ v yuchÄ v, a point which Clitophon had grasped
(cf. e7 e pime leian paÄsan poieiÄ sqai; 410b4±6 seÁ toÁ meÁ n protre pein ei v
arethÄ v e pime leian kaÂllist' anqrwÂpwn draÄ n; 410d5±e1), but in the ®rst
part of the protreptic speech the emphasis had been put rather strongly
on dikaiosu nh. There are two possibilities: either the author of Clit.
wanted to imply that dikaiosu nh and areth are identical (cf. 410e7±8
proÁv te lov arethÄ v e lqo nta), that is, that dikaiosuÂnh is the best state of
the soul, or he takes dikaiosu nh as the te cnh which is necessary in order to reach areth (cf. the analogy gumnastikhÁ kaiÁ i atrikh : a rethÁ
touÄ swÂmatov :: dikaiosu nh : a rethÁ thÄv yuchÄv which lies at the bottom


C O M M E N T A R Y: 408e3±408e5
of the next sentence). Cf. Intr., section ii.4.1. Gallavotti's ti nov pe ri
maqhÂsewv (`Miscellanea', RFIC 63 (1935) 508±13 at 511), cf. 409a2±3, is
therefore super¯uous.
e3±409a2 w sper . . . i atrikhÂ: logically, this is one sentence
(Schleiermacher; Bertini), though all editors print a full stop after
ou san (e10). Theoretically d' after ei at e10 is super¯uous (cf. KG ii
487±8), but only theoretically, as the ®rst ei clause (containing many
dependent clauses) is very long. Cf. note on 410d5±e1.
e3 prouÂtrepen: cf. note on e5.
e4 pronoouÄntav: this verb, though quite common in Xenophon and
the orators, is evidently avoided by Plato and Aristotle. It is found only
once elsewhere in the Platonic corpus (Cra. 395c7±8, in the etymology
of Pe loy; it is coupled with proiÈ deiÄ n after the fashion of the Cra.; it
may be important that Plato uses the deponent form pronohqhÄ nai ).
Here it does not have its usual meaning `to foresee', `to provide', but `to
realise beforehand', which in Plato usually is progignw skein (with a
o ti clause, as here, Ti. 70c2±3). The use of pronoeiÄ n instead and, besides, in the active form may be a sign of inauthenticity. Cf. note on
408d1 u potei nwn.
e5 ka peita w nei dizen: kapeita does not mark o¨ two distinct
stages, exhortation (prouÂtrepen) and reproof (wnei dizen): the imaginary
persuader starts his `protreptic' because he ®nds fault with his objects in
the ®rst place (mhdeÁ n pronoouÄ ntav o rwÄ n ktl.). Consequently, exhortation and reproof are the same thing here (cf. 408e3±4 touÄ sw matov
e pime leian poieiÄ sqai ± 6±10 ai scroÁn purwÄn meÁ n . . . e pime leian paÄ san
poieiÄ sqai. . . , tou tou d' au touÄ mhdemi an te cnhn mhdeÁ mhcanhÂn, o pwv
w v be ltiston e stai toÁ swÄma, e xeuri skein). As a matter of fact, what is
given here as reproof is a mirrored condensation of Socrates' protreptic
speech reported by Clitophon, with some verbal similarities (407b2±3
crhmaÂtwn meÁ n pe ri thÁn paÄsan spoudhÁ n e cete ± 408e6±7 purwÄn meÁ n
. . . e pime leian paÄsan poieiÄ sqai; 407b5 ou te didaskaÂlouv . . . euri skete
± 408e8±10 mhdemi an te cnhn . . . e jeuri skein). This is a good indication
that for our author `protreptic' meant `accusatory protreptic'. ± ka peita at times connects two hardly distinguishable aspects of the same
action, as Cra. 411b6±8 oi polloiÁ twÄn sojwÄn u poÁ touÄ puknaÁ peristre jesqai zhtouÄ ntev o phi e cei taÁ onta ei liggiwÄ sin, kapeita autoiÄ v
jai netai perije resqai taÁ pra gmata. prou trepen is apparently meant
to be interpreted as `if someone were going to exhort '.
One is reminded of the structure of Socrates' protreptic method in
the Apolog y (29d5±30a3): ®rst exhortation (parakeleuo menov, d5); if
people contradict, examination (e rhÂsomai au toÁn kaiÁ e xetaÂsw kaiÁ
e le gxw, e5); if they do not meet the standards, reproof (oneidiwÄ, 30a1).


C O M M E N T A R Y: 408e7±409a3
In Ap. too, the content of o neidi zein is not very di¨erent from the
exhortation, cf. Intr., section ii.3.3 and n. 268. It is possible that our
author had this passage in mind when writing these lines; we saw (cf.
section ii.2.3.1) that he probably used the protreptic passage 29d7±e3
for his parody. A verbal reminiscence may be Clit. 408e9 o pwv wv be ltiston e stai toÁ swÄ ma ± Ap. 29e2 thÄ v yuchÄv o pwv wv belti sth e stai.
e7 kaiÁ osa: `and for all other things which', cf. 410a3; Tht. 145a9;
Sph. 219a10; Phlb. 46a8.
e8 diaponouÂmeqa te kaiÁ ktwÂmeqa: `which we labour to acquire'
(Bury; so Sartori, Zuretti, la Magna, Orwin, Gonzalez). For this type of
hendiadys in Plato (denied by Riddell, `Digest', §324) cf. R. 328c1±2;
429e6 with Adam's notes, and the very numerous cases listed by England's Comm. on Lg., Index s.v. hendiadys. For diaponou meqa, cf.
touÂtou . . . au touÄ: Professor Sedley draws my attention to the parallelism of this phrase with e1 touÄ t' au toÂ. As the collocation points forward to the o pwv clause here, the two phrases cannot be co-refential,
but they are strictly parallel. This strongly suggests that touÄ t' au to at
e1 refers to the best state of the soul, in the same way as the phrase here
refers to the best state of the body, and this is borne out by 409a3 thÁn
e piÁ thÄi thÄ v yuchÄv arethÄ i te cnhn. Cf. on e1.
e8±9 mhdemi an teÂcnhn mhdeÁ mhcanhÂn: though of course te cnh and
mhcanh are very often coupled where one would be su½cient (cf. Lg.
831d4±5), here both words retain their original meaning: `to ®nd out
whether there is an art, or, failing that, any means whatever of improving physical condition' (cf. Water®eld and Gonzalez). Perhaps the
caution which caused the addition of mhdeÁ mhcanhÂn was inspired by
407b5±7 ou te didaskaÂlouv . . . eu ri skete . . . ei per maqhto n ± ei deÁ
melethto n ktl.
409a1 deÂ: `denotes that the information he [the speaker] already
possesses is inadequate' (GP 2 173; the instances quoted from Lg. and Plt.
all begin le geiv de ‡ interrogative).
a3 thÁn epiÁ thÄi thÄv yuchÄ v a rethÄ i te cnhn: cf. Intr., section ii.4.1;
justice is here for the ®rst time stated (without argument as in Republic 1
332d2±3) to be a te cnh, cf. a5; d5; 410b7±8. This is a formal expedient,
necessary for the arguments by analogy and the ®nal circular regress
(sections ii.3.4; ii.4.3). Hitherto Clitophon had been more cautious (cf.
note on 408e8±9) and even further on it is clear that he does not regard
justice as a true te cnh (cf. note on 409c4; 410a3; c1±2); for justice as the
a reth of the soul, cf. also on 408e2. ± For the accumulation of articles,
to the two examples quoted by KG i 611 (Sph. 254a10; Plt. 281a8) add
Phdr. 269c9±d1 thÁ n touÄ twÄi o nti r htorikouÄ te kaiÁ piqanouÄ te cnhn; R.


C O M M E N T A R Y: 409a4±409a5
511c5±6 toÁ u poÁ thÄv touÄ diale gesqai e pisthÂmhv . . . qewrou menon; Hp.
Decent. 17 touÁv e v taÁ thÄv te cnhv ei lhmme nouv; De Arte 1 toÁ taÁ twÄn pe lav
e rga (. . .) diaba llein; [Aesch.] 1.12 toiÄ v u peÁ r thÁ n twÄ n pai dwn hliki an
ou sin. Normally Plato takes some care to avoid it, e.g. Smp. 182d3±4
diaÁ thÁn twÄn qeme nwn thÄ v yuchÄv argi an.
lege sqw: rh seiv ending in a direct question followed by an imperative form of le gein (so too 409c1 ei pe ) are rare in Plato. The only instances in the authentic works are Prt. 353a4±6  W Prwtago ra te kaiÁ
SwÂkratev . . . ti u meiÄ v au to jate einai; ei paton h miÄ n; 357c8±d1; Grg.
470a4 le ge; Men. 74a1 ei pe ; Chrm. 165e2 i qi oun ei pe . As Grg. 462d10±11
must be otherwise explained (see Dodds ad loc.; cf. also 463c6 e rwÂta ±
c8 e rwtwÄ dhÂ), jaÂqi is not used by Plato in this way. There are three
parallel cases of apokri nou: Grg. 474c8; 515c3; Hp.Ma. 288e4. There is
no parallel for lege sqw so used, except Phlb. 16d8 ti v au th; lege sqw
mo non, where the addition of moÂnon, to my mind, makes all the difference (cf. the frequency of le ge mo non, `please, go on', in the later
The idiom is relatively frequent in the Dubia and Spuria: Thg. 123c12
(ei pe ); Min. 318a7 (jaÂqi ); and especially Just. 373b4 (ei pe ); c1 (apoÂkrinai ); c2 (jaÂqi ). I discount cases where the imperative is not actually the
end of an utterance, e.g. Ap. 25c6±7 (after a question) w taÂn, apoÂkrinai´ ou deÁ n gaÂr toi calepoÁ n e rwtwÄ; cf. 25d2; Grg. 463a5; Lg. 665b3.
This use of lege sqw is no good argument for inauthenticity (pace Thesle¨, Styles, 15 n. 2), the less so because Plato uses lege sqw as a variant
for le ge in other contexts, e.g. Ly. 204e7 lege sqw . . . ou tino v e stin. It is
in keeping with Clitophon's character (cf. Intr., section i.5.3), and an
incidental preference for an idiom not used elsewhere can be found in
any work of Plato or anybody else (cf. GP 2 lxiii±iv and n. 3).
a4 e rrwmeneÂstatov: this word is used of persons elsewhere only in
Grg. 483c1; e5, where it refers to Callicles' ideal of the `strongest' (cf. R.
564d7). e rrwme nov thÄ i yuchÄi occurs in Xenophon (An. 3.1.42; Hell.
3.4.29), but always denotes bravery (cf. Cyr. 3.3.31). But cf. Ti. 89e8,
where it is used of parts of the soul, and the phrase r wÂmh thÄv yuchÄ v
(Plt. 259c8; X. Mem. 4.8.1); cf. also Prt. 311b1.
proÁv tauÄ ta: with apokrino menov (`a questa mia domanda', la Magna), which is not just a pleonasm of eipe here, as it is at Prt. 314d6;
355d5; Lg. 897d4. See on 410e3 deo menov le gw.
a5±6 ei peÂn moi . . . dikaiosu nhn: the change from oratio obliqua to
oratio recta (h nper akou eiv suÁ le gontov e jh SwkraÂtouv) and back again
(ou k allhn h dikaiosu nhn) can be roughly paralleled from R. 364b5±
c5: aguÂrtai deÁ kaiÁ maÂnteiv . . . pei qousin w v e sti paraÁ sji si du namiv
. . . a keiÄ sqai . . . e aÂn te tina e cqroÁ n phmhÄnai e qe lhi, metaÁ smikrwÄn da-


8. though there are rather fewer examples in Plato than one would suppose (L. though our author does not eschew lengthy sentences in themselves. Hell. On the other hand. Adam ad loc. Alc. 16. there is no grammatical di½culty. 4.17 eipoÂntov deÁ Ku rou o ti tou twn meÁ n ei h  raÂspa´ a liv.3. toiÄ v me ntoi strathgoiÄ v prostaÂxai e jh crhÄnai o pwv kaiÁ Ko rinqov swÂia h i twÄ i  qhnai wn the change to e jh ‡ accusative and in®nitive has dh mwi twÄ n A the same e¨ect. . where the intervening direct speech. tauÄ t' '' e jh ``dihgouÄ . Anakoluthe. it would have been quite normal to use a o ti-clause instead of an accusative and in®nitive after ei pe n moi ± in that case the change of person would have been hardly noticeable.3. causes an unmitigated anacoluthon. as here.''..7. No such softening is attempted here.128).. . Intr. be di½cult indeed to ®nd anything really like our text: in®nitival oratio obliqua.4.2). We must conclude. add Chrm. Cyr. . and 1. Reinhard. Intr.3.8. as the author deliberately chooses to employ the oratio recta for stylistic reasons (cf. 178a7±9). 158c5±d6. similarly Hell. A change from oratio obliqua to oratio recta coinciding with transition from main to dependent clause (as here) is found infra 409c6±d1. jaÂnai Alc.14. The di½culty could not possibly have been circumvented by using a di¨erent word order: the best place for ou k a llh(n) h dikaiosuÂnh(n) is at the end of the sentence. 3.3 ± in the last two examples e jh is inserted as here (cf. . section i. An. ``kaiÁ touÄ to ou n'' e jh ``suÁ toiÄ v meÁ n Aqhnai oiv kecarisme nov e shi . . at Hell. w A ktl. A full stop after 307 . 6. Yet it would. 2 148e5). . .. bla yei . . e lege proÁv auto n.13.3. .9 le gontov deÁ touÄ Ko nwnov hw vi . a7 ei po ntov de mou: if a full stop is printed (as is usually done) after ei pe (c1).28. . . jhÂsei 409d1. [8 lines of oratio recta] . Cyr.''. .4 ei po ntov deÁ Dhmoti wnov e n twÄi dh mwi twÄn Aqhnai wn wv h meÁ n proÁ v touÁ v A  rkaÂdav jili a kalwÄv au twÄi dokoi h praÂttesqai. 87±8. 7. then. a change from oratio obliqua to oratio recta is in itself not very noteworthy (cf. but the style is not improved. Intr. that the author deliberately inserted the clause in oratio recta for a stylistic reason.3). KG ii 565±6).1. 159b1±6. Alternatively. The return to oratio obliqua is understandable: tau thn thÁ n te cnhn ei nai hnper akouÂeiv . 1. Of course. .53. ou k allh h dikaiosuÂnh would be too harsh. Cf. if we print a comma after ei pe . X. . in my opinion. section ii. section i. where the anacoluthon is softened by transition to oratio recta. . 1. with a dependent clause in oratio recta (with change of person) in between (cf. 7.3. Smp. the genitive absolute is pendent. 2 148e3±149a1.5. Less harsh: X. .5. .3±5 apikome nhv deÁ tau thv kaiÁ legou shv´ ``w paiÄ . touÁ v qeouÁv wv jasin pei qonte v sjisin u phreteiÄ n (cf. .C O M M E N T A R Y: 409a7 panwÄn . ``a deÁ kairoÁ v hmiÄ n ei de nai. The closest parallels are Hdt.'' h meÁ n dhÁ taÁ e pagwgo tata .

. 346d1±6). 308 . but there is no opposition between the cleverest pupil and Clitophon. d2. Chrm. jame n ti ei nai di kaion au toÁ h ou deÁ n. Cf. Sicking in Two Studies. There is something to be said for this (though it brings him into trouble at b6±c1): both `new doctors' and `health' are products of medicine. le geiv de tinav . 100. all three de®nitions of the e rgon of justice start with an asyndeton: c1. 61). cf. and also of the roughly parallel anacoluthon opo tan au jhÄiv toÁ e jexhÄv tou twi . MS authority counts for little in these matters. . as against d' e mouÄ (D) as at R.2. pou: cf.1. as does la Magna without arguments. Mainly on the strength of this argument.M.. The most familiar type is e pisthÂmhn pou kaleiÄ v ti. and also has an interactional value ± hence its virtual absence from the orators. cf. For the stating instead of the questioning form. section i. 351b3±4). Burnet on Phd. (Prt.. 399d2±3. Bertini. Bluck on Men. 495c3±4). MuÈ ller. variants are h dikaiosuÂnh praÄgma ti e stin h ou deÁ n praÄgma. 338a4 ei po ntov de mou (AF. note ad loc. but this cannot be an objection (cf. twÄn anqrwÂpwn eu zhÄ n. hence its frequency in Plato for stating the obvious (GP 2 491. e xergaÂzesqai. . . b2 a potelouÂmena: Bury translates both this word and e rgon (b5) as `e¨ect(s)'. . d' e mouÄ D) tauÄ ta. hgouÂmeqa ti toÁn qa naton ei nai. de mou: so AF Stob. b1±6 i atrikh . 65d4±5). b1 i atrikh pou tiv leÂgetai te cnh: the familiar procedure to start an argument by asking if (or stating that) the concept with which the argument deals is more than a name and corresponds to something existent. 2. R.. (Phd.J. Sicking.2. touÁ v deÁ kakwÄ v. kaiÁ o tan le ghiv . . 4. (Prt. Indeed. b3 u gi eian: sc. but that would equate a potelouÂmena and e rgon. kaiÁ teleutaÄ i dhÁ kalwÄv o lo gov ou to v soi ktl.5. C. Stallbaum on Cra. (407e5±408a5. cf. if so. the term e rgon was traditionally reserved for the latter (cf.) I prefer the pendent genitive.3 n. . so another word had to be found to denote both.2. Orwin). Intr. 75e1. Mem. (X. . H. on 410a7).22). 165c10±d2. di dagma: cf. 64c2).4. 410a7. I think that the interaction is better described as an appeal to the addressee to accept the surmise as actual fact for the time being than as a sign of `leaving room for di¨erence of opinion' (Sicking. section ii. (Phd. 330c1). The particle obviously presents a statement as a surmise. Intr. oisqa tinav anqrwÂpouv a cari stouv kaloume nouv.C O M M E N T A R Y: 409b1±409b3 ei pe causes asyndeton in the next sentence. thereby making the clause o dhÁ le gomen u gi eian (b5) super¯uous. 64c2. (Grg. often translated as if it were a nominative (Schleiermacher. 61 ± this does not account for the strong appeal in ou ti pou questions). 57±61.

b5 dh : according to Denniston's treatment of dh in relative clauses. Another possibility is to supply taÁ a potelouÂmena (Bertini. de . Phdr. Starkie on Nub. 9. 246a6 (cf. .±D. toÁ de . when we take it as `of course' (`bekanntlich'. J. 85. cf. while the entities involved are di¨erent' (201±2): this applies in the present case. The construction of the next sentence supports my interpretation of this one. . Ly. The translation `carpentry' is one-sided. 269c2. di dagma: though found in Hippocrates. A. 309 . Hp. 15 n. . 562. b7 e stw: `let us assume that etc. . this should mean `precisely that which we call health' (GP 2 218±19). both used by Plato. Phdr. b6 te kai : `respectively'. cf. Plt.10). Lg. 2). esp. I am willing to take that as a starting-point for the discussion) pollaiÁ meÁ n h donaiÁ kaiÁ ano moioi gigne sqwn. in that a te ktwn also builds houses (cf. 6. de '. De Vries ad loc.13. whereas the word required here is to mean `object of teaching'. which (though the borderline is not always strongly drawn) is properly di dagma. in which. M. unmarked in ma qhma. 304c4±5 (e pisthÂmhv) thÄv manqanome nhv kaiÁ didaskouÂshv. 295c2±3. de : introducing a new instance (GP 2 202) cf. On this combination. `Adverb or connector? The case of kaiÁ . Styles.Ma. toÁ deÁ di dagma would be an appositive sentence ± it seems better to take e rgon and di dagma as predicates. Bury) but then toÁ meÁ n e rgon. 215e7. didaskali a and didachÂ. de ] the states of a¨airs of the two clauses or sentences must be identical or similar. Phlb. ± The concrete products of tektonikh are said to be a house at c6 and wooden equipment at d1. .' Cf. 668) and Xenophon (Eq. as he shows. kaiÁ .C O M M E N T A R Y: 409b4±409b7 b4 thÄv didaskouÂshv te kaiÁ didaskomeÂnhv: cf. Schw. kai is rarely the equivalent of `also'. 187±208. . would mean `the act of teaching'. For the passive of dida skein with a non-personal subject cf. cf. KG ii 126±7. `what's taught' (Gonzalez). Aristophanes (cf. 14a8 (in your opinion there are many heterogeneous kinds of knowledge. The particle can be better understood both here and at d1. who uses maÂqhma throughout. 280c9 oi kodomikhÄ i kaiÁ o lhi tektonikhÄi ). . and should be taken as partitive. van Ophuijsen in Two Studies. tektonikhÄv: the genitive depends on toÁ meÁ n . There is no reason to use this word as an argument against authenticity (pace Thesle¨. `For kai ˆ ``also'' to be acceptable [in the cluster kaiÁ . the word is avoided by Plato. . Plt. 201±3. note on 408b3). . Rijksbaron. its most fundamental semantic value. New Approaches. note on 407a3 kaiÁ belti wn. the tragedians. . although Rijksbaron regards it as an exception to the normal behaviour of kaiÁ . 793c2±5.). The choice of the rarer word was probably inspired by the need to stress the aspect of teaching. 278a2.

Lagas. 283b4 ei pe moi e jh w SwÂkrate v te kaiÁ umeiÄ v oi a lloi. to the te cnh analogies). (c1).). A. `Unterabteilungen innerhalb der zeitlich ersten Gruppe Platonischer Schriften'. after a ®rst attempt has failed (Prm. c2 e terov: continuing a llov as R. au qiv. Prt. xiii). The word is often used in Plato for tackling a problem anew. 1 106c2). c3 e panhÂiein: `I went back' (sc. asynd. 127±39). ei peÂ: cf. i na soi mhÁ diaje rwmai. Instead he transforms them into quali®cations of the performance of the artists. C. Brandwood. Our author is addicted to dh (an argument which could be used either way). (d2). 224±5. 240±6). d' (here). Frogs 39. Chronolog y. 439b10. Platonis Opera. 352a11 etc. 201c7. this is an apposition of toÁ d' e teron. . . -IKOS bei Platon (Freiburg (Schweiz) 1953). Euthphr. tecni tav: tecnikou v would be more in keeping with Platonic usage. J. Maybe (as there) the rarer word was chosen because it is more precise than tecniko v (tecni thv or dhmiourgo v. Cf. lusitelouÄ n: cf. one who has a te cnh for profession.g. c1±3 ou tov . cf. e. elsewhere the MSS have h ia (Schanz. prosei h B. Intr. c4±d1 ka keiÄ . 9d1. d': so DF. teÂcnh: this is a clumsy argument. de on etc. This use of the imperative should be distinguished from cases where the imperative points to surrender (R. where most examples concern appositions following relative clauses. Van Leeuwen on Ar.). (2) Q . cf. . e aÂn ti hmiÄ n e paniouÄsin a lloiÄ on janhÄi.C O M M E N T A R Y: 409b8±409c4 MT §254. apply also to the e rga of the arts. through addition of paÂlin. toÁ de on now be- 310 . d' (409a7). The absence of such a determinant has led most translators astray. . e8. Perhaps d' is better as it creates a parallel between two pairs of question and answer: (1) Q . b8 ekeiÄ : cf. section ii.. an adverbial phrase introduced by e pi . vii p. 142b1±2 (introduction to the `second hypothesis') bou lei ou n e piÁ thÁn upo qesin paÂlin e x a rchÄv e pane lqwmen. this meaning is more clearly determined by the context. one who has the knowledge or the capacity for a profession.. The form hiein lies at the base of the readings at Ly.1. Smp. R. cf. ± Clitophon addresses either the e rrwmene statov of a4 or all the pupils. `ritornava a' miei esempi' (Bertini). tecniko v. 311d6 (and Adam ad loc. 13±16. Sph. have dh . cf. 351d7 e stw . c1 ergon: logically speaking.en Anticipatie-Verschijnselen (Amsterdam 1941). A Stob. `attracted' into the relative clause (KG ii 419±20. Ritter. 206e1 proshÂei T. A. N. note on 409a3 lege sqw. note on 406a12 dhÂ. Clitophon could have carried his case by pointing out that the quali®cations sumje ron. . Syntactische Perseveratie. Euthd. Alc. A. A. cf. asynd. esp. 219a5.5. . but cf. Hermes 70 (1935) 1±30. Smp. 187c3. As a rule. Ammann. after the earliest dialogues e terov is often used as a variant of a llov. etc.

. te cnhn. 379c6±7 twÄ n deÁ kakwÄn all' atta deiÄ zhteiÄ n taÁ ai tia).C O M M E N T A R Y: 409c4±409c6 comes o rqwÄ v praÂttein. The words make far more sense if they are taken as a refutation of sumje ron. c6±d1 a llaÁ .) c4 ka keiÄ : `in the other area'. Kleitophon'.' I conclude that Clitophon's answer was not a free invention. de on etc. therefore ``useful'' is an inadequate description of what carpentry is. adverbs (qualifying pra ttein) are used.1. cf. we see that the adaptation is not a success. c5 o rqwÄv pra ttein: represents toÁ de on of c2. toÁ deo ntwv (sc. teÂcnh: puzzling in syntax and meaning. it is tolerably appropriate to answer: `But the operation of every art can be described as useful. note on 409a3 thÁn . the repetition of the statement is pointless because neither the pupils' answers nor their refutation by Clitophon had at any moment implied that art and object are identical.5. c2±3. . I take proÁv o ti tauÄta paÂnta tei nei as an indirect question. The vague words kakeiÄ ta ge o no mata tauÄ t' e stiÁ n e n e ka sthi twÄn tecnwÄ n cover up the transition. 15) refers to the i di a w jeli a of R. This passage was put in a totally di¨erent framework. namely that of the arts. As before. 346a6±8. The formulation makes it clear that the author thought of the te cnai as one block opposed to aretai (speci®cally justice). w je lima. . not 311 . (Pavlu (`Pseudopl. instead. not of justice as yet another analogue to the several arts which had been named.. section ii. The relative clause a dhÁ ou k e stin te cnh seems to be inappropriate: it had been made clear enough that the products of an art are distinct from the art itself (b3±5). toÁ i dion as a predicative apposition (`as its distinctive trait'. praÂttontev. . So you cannot use the word as a de®nition of justice either. in which one of the virtues (not necessarily justice) had been de®ned by means of a substantivated neuter adjective (not necessarily toÁ sumje ron or toÁ de on or any of the adjectives used here). cf. this sentence has been interpreted by the translators in many di¨erent ways. but a truncated adaptation of a passage from a (probably Socratic) dialogue now lost. e n e kaÂsthi twÄ n tecnwÄn is not a pleonasm. note on 407b1±2 agnoeiÄ te . for the exact meaning of which. he should have said that the xu lina skeu h are themselves sumje ronta. as opposed to justice. which has a di¨erent function and does not explain the distortion in Clit. Cf. Cf. . depending on e reiÄ (so Schleiermacher). Intr. toÁ wje limon is taken up as wje lima (praÂttein) and so on. praÂttein) as object of e reiÄ . as de®nitions not of the e rgon of justice but of justice itself. R. . In the example from the carpenter's art. but a sort of distributive apposition. toÁ kalwÄv. But this result is distinct from carpentry itself (a dhÁ ou k e stin te cnh). the search for the e rgon of justice. toÁ eu. If your opponent says `justice is the useful'. useful action in carpentry results in the production of wooden equipment.

carpentry will mention right. other possibilities are: (1) to supply proÁv touÄ to before e reiÄ (H. gi gnesqai). . There is nothing in the free-will passage to prevent us from assuming a source for the argument (though there is no other reason to assume one). Gonzalez ± if I understand their translations correctly). the article would perhaps be better. R. If we could trust the distinction made by KG i 634 and others (e ka sth te cnh `each art'. as its distinctive trait. Plato's Meno (Indianapolis±New York 1971). toÁ kalwÄ v. even cramped. ad loc. which of course is not art itself. w ste as a sort of equivalent of tei nein proÁ v to (taÁ xu lina . 19). The similarity in wording between this sentence and 410a3±6 does not exclude condensation of a source.' When one keeps toÁ i dion as object. (2) to take toÁ i dion as antecedent of (proÁv) oti (Bertini). 30. 464c5±d3. infra 410a4±5. Bury.) to toÁ de on. e kaÂsth h te cnh `every single art').C O M M E N T A R Y: 409c6±409c7 of jhÂsei. Notes on the Laws of Plato (London 1972). where proÁv e kaston e rgon seems to contradict the canon (`with reference to each separate function'. e kaÂsthi te cnhi (Ven. C. at this place the existence of a source has already been surmised on other grounds (cf.2. which is unnatural word order. cf. But cf. Souilhe . c6±7 e ka sth teÂcnh: for e kastov without the article in Plato. c6 e reiÄ : for personi®cation of the te cnai. given as an example of `every art'. toÁ deoÂntwv: the last adverb goes back (via c5 o rqwÄv praÂttein.3(5)). and in general Dodds on Grg. . One explanation might be that he was not talented enough to retain the easy ¯ow of his expression when the subject-matter became really di½cult. but the subject of jh sei is h tektonikh . As the direct tradition unanimously omits the article (against Stobaeus). sections ii. ii 322. MuÈ ller.). It cannot be a coincidence that at the two places where the author has to disprove a position (the other is the argument against voluntariness of wrongdoing. his reasoning and expression become terse. W. Orwin. cf.7. Men. The other two are introduced because they are in fact more relevant to the carpenter's art than `profitably' or `bene®cially'. 407d2±8). Brown (ed.. but the function of proÂv in proÁv o ti is not the same as in this proÁ v touÄ to. `But the aim of all these actions will be stated by each of the arts individually. 312 . elsewhere his style is rather the contrary (cf. proper and appropriate action. J. Gildersleeve. For instance. further Intr. 184) makes toÁ i dion subject of e reiÄ . it is the best policy not to print it.7. cf. 342a1±d1. T. The change in terminology is immaterial to the argument. c7 toÁ eu . cf. aiming at (she will say) the production of wooden equipment. note on c4±d1). On the other hand we should not overlook the possibility that the falterings arose from curtailing arguments which he found in a source. ii. Saunders. 72a3. K. Guthrie in M.

Modugno. Consequently. d3 twÄn swÄn e tai rwn: ADF Stob. kai ou coupling opposites (cf. note on d9±e1. section ii. ei peiÄ n means that this particular statement was the smartest (so Ficinus. d5±6 jili an e n taiÄ v po lesin poieiÄ n: cf. iv 165 and n. d2±e10 TeleutwÄn .3. d3±4 ov dhÁ komyoÂtata edoxen ei peiÄ n: most translators and interpreters take this clause as a description of a permanent quality of this pupil. d6±7 thÁn jili an a gaqo n t' ejh ei nai kaiÁ oude pote kako n: cf. as here.) had been given. cf. Water®eld. 2. e jhsen is reached. Water®eld). cf. 409a6 and the examples quoted on 409a5±6 and on a7 ei po ntov de mou. `whose answer was considered the cleverest' (see ad loc.) 313 . here split as at Phd. e4±10.. 286±7). cf. Chrm. note on 410a7±8. Phdr. For more parallels and for the syllogistic method which lies behind it. d2 TeleutwÄn: probably we are to infer that Clitophon skips a few de®nitions. . 166c2.C O M M E N T A R Y: 409d1±409d6 d1 xu lina . Intr. Intr.. . . for the latter. i dion rather than e rgon (Ficinus. . Both dh and komyo v indicate irony. `who was reputed to be a most accomplished speaker' (Bury). skeu h: cf. 159c1 ou twÄn kalwÄn me ntoi h swjrosu nh e sti n.2. Note that the reporting tense of the argument is the imperfect throughout (except. Susemihl. .5. MuÈ ller. Cf. section i. kaiÁ ou de pote kako n: te . indicates transition to direct speech.5. GP 2 513).. where the emphasis is on i dion. Gonzalez). Tht. a gaqo n t' . Our author has a strong preference for what Gildersleeve calls `®rst attributive position' (ii 280±2. jh sei: parenthetical.5. 408c6 is a special case. For jhÂsei in a parenthesis cf. in other words. cf. note on 410b1 u steron deÁ e jaÂnh. (But there the intervening word belongs closely to the one preceding te ± maybe here ti (F) is right. Miscellaneous Notes on Plato (Amsterdam 1975). 272b2 etc. . 6). imply that other answers (apart from the series toÁ sumje ron. 165c6. The words o v dhÁ komyo tata e doxen eipeiÄ n. cf. De Vries on Phdr. introducing. HGPh. cf. wmologh kei at e8) until the conclusion w ste . for obvious reasons. note on 407c6±7. cf. .2. . 68b1. twÄn e tai rwn twÄ n swÄ n Va. 146e1±2. 227c7 (and for a restriction. Tht.. . . note on b5 tektonikhÄv. Cf. H. do xan: for the oratio obliqua.. section ii. id. the refutation of a de®nition. d4±5 toÁ thÄ v dikaiosuÂnhv i dion e rgon. . Intr. they fail to remark the aorist in®nitive. komyoÂtata . it is impossible to take these words as an allusion to a speci®c member of the Socratic circle. . cf. where tektonikh is de®ned as e pisthÂmhn thÄv twÄn xuli nwn skeuwÄn e rgasi av. toÁ de on etc. d2 toÁ thÄv dikaiosuÂnhv: if anything is to be supplied at all.).

72 n. Dirlmeier. The parallel from the Lysis (211d6±213a3) which was adduced by Steinhart (55. 32±1117a1 (Gaiser. Cra.2. this is a very natural step to make. Aristotle. because it is based on pleasure instead of the good: EN 1156a31±2 h deÁ twÄn ne wn jili a di' h donhÁ n ei nai dokeiÄ . 523±4 and 529 n. 514±15. ai d' allai kaiÁ e n toiÄ v qhri oiv. Susemihl (513). his treatment of these subjects inevitably contains many strains of thought that go back to discussions in Socratic schools and writings. `pederasty and bestiality' Water®eld ± Modugno's `amicizie dei fanciulli colle bestie' is highly improbable because of the repeated article).5. cf. . Susemihl. See F.. friendship of the young (not necessarily children) is not primary friendship. 11. 146 n. .C O M M E N T A R Y: 409d7±409d8 d7±9 taÁv . EE 1236a38. R. EN 1116b25. a ll' a jobon kaiÁ mwÄron´ h kaiÁ taÁ paidi a paÂnta oi ei me a ndreiÄ a kaleiÄ n. e panerwtwÂmenov: twÄn pai dwn and twÄ n qhri wn are subjective genitives (contra Steinhart. 55. 3) and. a di' anoian oudeÁ n de doiken. . but friendship based on pleasure and on common interest can: EE 1236b5±7 au th [sc. 16)). who wrongly identi®es the o ntwv kaiÁ a lhqwÄv jili a (e3±4) with Aristotle's primary friendship. Nikomachische Ethik (Berlin 1956). Aristoteles. 379. Primary friendship cannot occur in animals. d8 a v hmeiÄ v touÄto tounoma eponomaÂzomen: for the double accusative (of external and internal object). F. 48 n. In Aristotle. Intr. cf. note on 409d9±e1). Children and animals are excluded from friendship as they are from courage in La. Protreptik. Lg. Magna Moralia und aristotelische Ethik (Berlin 1929). Bertini. is certainly wrong in saying that in the Laches the exclusion is ascribed to Prodicus: only the distinction between andreiÄ a and qrase a (197b6±c1) belongs to him. 27). Dirlmeier. h prwÂth] meÁ n oun e n anqrwÂpoiv moÂnon upaÂrcei jili a . 892b7. twÄn pai dwn jili av . For the question whether this exclusion of children and animals was found by our author in the source from which he quoted or made by him on his own initiative. cf. section ii. 314 . is a more or less normal step in Socratic discussions about virtues and virtue-like qualities. 161. Cf. Eudemische Ethik (Berlin 1963). . Pavlu (`Pseudoplat. cf. . Though he does not belong to the Socratics inasmuch as he does not subscribe to the moral paradoxes. too. Heidel (Pseudo-Platonica. 7) is irrelevant (so Kunert (Necessitudo. cf. . Kleitophon'. for once. 197a6±b1 ou gaÂr ti w LaÂchv e gwge andreiÄ a kalwÄ ou te qhri a oute a llo ou deÁ n toÁ taÁ deinaÁ u poÁ anoi av mhÁ jobou menon. Walzer. 406a3±6. 204±5. From all these passages it appears that the exclusion of children's and animals' friendship in Clit. frequently couples children and animals as inferior in his ethical discussions (cf. Arist. For a school of thought which has as its basic tenet that virtue is knowledge (or at any rate reposes on knowledge of some kind). d1±5). Aristoteles.

6)..C O M M E N T A R Y: 409d8±409d9 d8±9 ouk a pede ceto ei nai jili av: for the accusative and in®nitive cf. che . . . Indeed. 1153a28. 315 . .' It is not easy to see why friendship of animals is more often harmful than bene®cial. for the Socratic. The solution was found by Grube. . ma tali amicizie sono le piuÁ frequenti. d9±e1 sune baine . 307 n. note on 409d2. au twÄi. cf. . cf. Hence H. . Phlb. MuÈ ller would read e tai rwn (`Parteigenossen') for qhri wn (74 n. Intr. . for andrei a is always good [cf. Indeed. . hence its subsequent identi®cation with o mo noia in the sense of e pisthÂmh (e4±10). is harmful on principle. d9 sune baine . . The thesis that friendship of children and of animals is harmful more often than not is stated to be the result (sune baine . 35c3. kako n] but karteri a met' a jrosuÂnhv is blaberaÁ kaiÁ kakouÄrgov.4. it can only be a rational thing. . 48 n. agaqaÁv einai represent the argument which he found in his source and for some reason or other chose to pass over (possibly because he did not want the intellectualist conception of friendship to come in before e5. Cf. 39) on the ground that it departs from the meaning it has elsewhere. . . also Heidel. it seems much more likely that the words taÁ v deÁ twÄ n pai dwn . 207c6±7. . where it makes the circularity of the argument ± which he added himself. 410b1.) of a discussion which Clitophon does not report but to which he alludes with e panerwtwÂmenov (`having been subjected to further (e p-) interrogation'). .1 ± more obvious). ei nai: the reason why children's and animals' friendships are excluded. 477 n. Tht. he proposes to read taÁv jili av for taÁv toiauÂtav: `Le amicizie sensuali e brutali sono sempre dannose. . 31. . au twÄi: this phrase was suspected by Steinhart (72 n. 498e10. note on 409d6±7 thÁn jili an . Children and animals are irrational in their feelings. section ii. `The Cleitophon of Plato'. Now another glance at the Laches passage just quoted may give us an idea how this discussion established that friendship of animals and of children is mostly harmful. and the irrational. 1236a2. and comes close to `in den Sinn Kommen'. 192c3±d8 `Not every karteri a is a ndrei a. there is something the matter: `La proposizione che amicizie si¨atte . Aristotle quotes examples of bene®cial friendship among animals (EE 1236b9±10). Cf. Indeed. dunque si puoÁ a¨ermare . friendship being agaqo n. 1: `[it] seems natural as Cleitophon is giving the conclusion of an argument which it is not the place to repeat'. cf. The dative as Grg. Bury). le amicizie siano per la maggior parte dannose'). PseudoPlatonica. Cf. Arist. EN 1152b19±20. when we take sune baine in its usual meaning. ad loc. siano dannose.. 8. non era quella che conseguiva (sune baine) dalle premesse' (Bertini. EE 1224a29. as we are bound to do (`as a result of the argument he was forced to say'. La. I can hardly believe that the author wanted his readers to reconstruct this argument for themselves.

44. Denniston. van Herwerden (Utrecht 1902). `A propos de Platon. which requires an absolute statement: `they are not friendships at all' (most translators from Susemihl onwards).. 61±2.C O M M E N T A R Y: 409e1±409e2 e1 taÁ plei w: comparative and adverbial. `Ou de ``not at all'' '. 316 . Thucydides is the only other author to use taÁ plei w as a comparative adverbial adjunct (1. De Vries on Phdr. 78). a host of examples (most of them to be interpreted di¨erently) are quoted from Homer by J. 583). R. In the genuine works. Verdenius. one would a priori expect adverbial ou de to have the same three values. in Album gratulatorium in hon. ii 193..1 ± in the ®rst passage a temporal interpretation is possible. so Schleiermacher. esp. so Ficinus. it is not comparative (at 144d6 and 146e2 it is opposed to o ligaÂkiv. or `the greater part'. . Schleiermacher). developed from phrases like taÁ plei w le gein (Isoc. a gaqaÂv: practically antonyms since agaqo v and w je limov are practically synonymous (cf. Fraenkel. 153a6. TheÂÂeteÁte 167d6'. Ast.13. e1 blaberaÂv . Here an inclusive or scalar-inclusive value (a climax) is excluded by the context (`they are not even friendships'. whether it should be taken as temporal (`more often'. 5. In the only parallel I have found in the orators. M. .64. . E. Bury) is rather an academic question. 438c1. Jowett±Campbell. Cf. a scalar-inclusive value (`even') and a non-inclusive scalar value (`actually' ± marking a word or phrase as simply the highest point on a scale. H.16 ou k an h nwclouÄmen taÁ plei w. This idiom may be a mark of inauthenticity. but taÁ v twÄn pai dwn kaiÁ taÁv twÄ n qhri wn three times in a row would o¨end even Greek ears. at 146d5 it is even followed by maÄllon h). taÁ plei w is only found as a nominal phrase used as subject or object (`that which is more'. Republic. cf.5. 164. Med. 2). 261a4 gives some carefully chosen instances. in the latter excluded). Prm. This `emphatic' use of oude is often ignored.63). 609). 330a7). . `a ladder of which only the top rung is clearly seen'. R. cf. Denniston admits it for Herodotus (GP 2 197±8. taÁ plei w is only semi-comparative and semi-adverbial (D. `we should not trouble you further'. taÁv toiau tav: our author does not often substitute toiouÄ tov for an attributive phrase (407d5. SouilheÂ) or quantitative (`for the greater part'. In the only other dialogue in the Corpus Platonicum which presents adverbial taÁ plei w (Alc. GP 2 317). `Ou de Homericum'. Greek Prose Style. 55±64. though he normally uses the phrase (like toÁ ple on) as a more recherche alternative for taÁ pollaÁ. `mostly'. e1±2 taÁ v toiauÂtav . id. Mnem. oude : just as kai as a Focus particle can have an inclusive value (`also'). 4. 32 (1979) 163±4. note on 407a1 wjeleiÄ n). e2 toÁ toiouÄton: having to admit that some friendships are harmful and consequently that his statement that friendship is always good was wrong. ibid.

R. cf. e3 outwv: one would have expected touÄto. yeudwÄv: as here. ou de may be the negative of kai as used e. contrarily to what some people think about it.25. 20 (1967) 435±9 at 436) kaiÁ ga r qhn ou d' eidov e cw kako n. Another explanation might be that ou de emphasises that a statement is not true. H. X. e. e3±4 thÁn . 891d2. The phrase is treated as a marker of late style by Ritter.34 (quoted by De Vries. Outside Plato: S. R. Aspis 415. Epin. 1429. At this place. cf. this statement being preceded by the statement that A is true in other cases (so Phdr. the preceding statement that A is true in other cases (speci®cally. Lys. Nub. nowhere else in the corpus. See Intr. 2. Cf. Prm. Among instances not quoted by De Vries but probably to be classi®ed under the general heading of `emphatic ou de ' are R. 962.24. 6. o ntwv kaiÁ a lhqwÄ v jili an: the two adverbs are juxtaposed as Sph. 3. `Theocritea'. ou bd ) might be defended as lectio di½cilior. Plato's Logic. Dover on Ar. 329a8 toÂte meÁ n eu zwÄ ntev. cf. Lg.23. 587c3. El. also ou de . the adverb means `mistakenly' in Plato. section ii. E. Euthd. wv me le gonti (the falseness of what is said about Damoetas' appearance is stressed).. R. .g. 40d2. OC 590. Untersuchungen.g. 328c6 ou deÁ qami zeiv) fall outside the borders of the contextual groups outlined above. on Phdr. but cf. oude pw. Call. 585e1. 17. at 406a7 taÁ deÁ kaiÁ e phÂinoun (cf. Some of De Vries' most convincing examples (e. Comm. Ant. 110. no problem is solved by stating that ou de can `mean' `not at all'. that adults' friendships do deserve the term) may be said to be implied by the context: cf. 53). 263d4. Mem. ad loc. 133d5. emphatic ou de can occur have been de®ned exhaustively. Dysc.. 264a5±6: a good orator starts from the beginning. ou de pote. 175 (I have not been able to locate 317 .1 (Pf. but Lysias ou deÁ a p' a rchÄv all' a poÁ teleuthÄv e x u pti av ana palin dianeiÄ n e piceireiÄ toÁn lo gon).5. 981 and Denniston ad loc. 387.1 n. 8. 4.8. Theoc. 8. Smp. Men. 731. 15c3 ou de (T.ˆ ou .). . in other cases. `Notes on Euthydemus'. GP 2 583 (on Hdt. Phlb. Ep. Mnem. Euthphr. 876b3.g.2). in causal and consecutive clauses. 302c1 (De Vries. Plutarch's work  Oti ou d' hde wv zhÄn e stin kat'  E pi kouron (contrarily to Epicurus' pretensions). 986d2. nuÄ n deÁ ou deÁ zwÄntev belongs to a contextual class described by Denniston under `not even' (GP 2 196) but certainly bordering on `not at all'.106. For some groups of examples ou de can be explained as the negative of emphatic kai .7. I must add that a de®nitive explanation of oude in our passage is not possible until the general conditions in which non-inclusive.): ou de would then stress the statement that A is not true in some cases.C O M M E N T A R Y: 409e3 This idiom deserves closer investigation.. 20. 93 and Lutosøawski.

443±5. this appears very clearly from Aristotle's description of its use (EN 1167a26±b4. onwards (M. 1. but the quali®cation e n taiÄ v poÂlesin (d5) inevitably brings up omo noia. . 259a8. esp. fr. in the Greek phrase jili an e n taiÄ v po lesin poieiÄ n. R. Sph. When friendship of animals and of children has been discarded (and ± as might have been expected ± that of women neglected). Phdr. Cf. cf. Brandwood. cf.. . Plato's Logic. EE 1241a32±3).93. 244a2. 106. e4±10 thÁn deÁ o mo noian . X. op. Lutosøawski. 8. 440±3. 255 D. In fact. saje stata o n. `Zur Entwickelung des Platonischen Stils'. . Brandwood. 305e5.2 n.. o mo noia is above all a political word. cit. . e. Cf. section ii. Schanz. cf. esp. o moÂnoian could have been substituted for jili an without any essential change in meaning (considering that o mo noia was according to Aristotle usually said to be politikhÁ jili a). 55. Page ad loc. it is con®rmed by the ®rst occurrences (Th.2. the word is omitted altogether in the new OCT). cf. This is probably right.4.).g. e4 ei nai saje stata o moÂnoian: cf. is supported by E. 228d7±8 noÂsov .73. 413e1 ontwv is read by d. or ou twv may be `ut dicis' (Ficinus). . 311b9. it is found closely associated with jili a. 402.. 260a1±3. Hermes 21 (1886) 439±59. the parallel from Republic 1 quoted above.3 e kklhsi an poihÄ sai .±K. Chrm. Schanz. men's jili ai are left. Chronolog y. o mo noian: in classical Greek. Intr. For the adjectival use of ontwv and twÄi o nti cf. Med. o ntov by bT. kaqaÂper kaiÁ le getai cf. Chronolog y. perhaps the double construction of e cein. Lutosøawski. 36). J.7. Plato's Logic. as ontwv (instead of twÄi o nti ) does not appear until Republic 5 and becomes frequent only from Sph. periÁ o monoi av.16). At the same time. `slaves' and `free men'). esp. resulting when ou twv is read.5. mentioned 110). In themselves these may cover a wide range of sentiments (quite apart from the fact that `men' can be divided into `Greeks' and `barbarians'. 351d5±6. The whole discussion now appears to be carefully framed so as to end up in a conclusion which had been its starting-point. 76. where (as here) there is an antithesis between a false name and true nature. esp. G. .C O M M E N T A R Y: 409e4 the occurrence in Phlb. 184) for outwv at Euthd. cf. section ii. ± alhqwÄ v for w v alhqwÄ v is found occasionally in earlier dialogues but again starts occurring with some frequency only with Sph. of a lhqwÄv and wv alhqwÄ v Phd. doÂxan: the argument consists of two syllogisms: 318 . 140. At Cra. (M. 162d4. Democr. Mem. 120. Intr. 34±5 ± I disregard the conjecture o ntwv (Ven. `Notes on Euthydemus'. Plt. And.. 109e7±110a1. de Vries. 732 (cf. This adverb is not found reinforcing the copula before Sph. 4. b2±3 politikhÁ dhÁ jili a jai netai h o mo noia. 120.

d9±e1. . jili a (ˆ o mo noia) is always good (e7±8.3. kakoÂn. cf. 409d4) with the easily explained exception of 410a2. cf. . d6±7). 319 . Syllogism (2) is another example of the argument `A is not B. cf. also the optatives of 406a3. section ii. cf. in this context. in the discussion Alc. a `sentiment of friendship'. o modoxi a and o mo noia. The author of Clit. for A is always good and B (sometimes) is not'. therefore (wste) (1c). it would have been too obvious that the argument does not hold water). therefore it is e pisthÂmh (e9±10). section i.. restricts the term o mo noia to what Plato and Aristotle call o modoxi a (had he not done so. therefore jili a is not o modoxi a. EN 1167a28±9) regardless of the distinction do xa : e pisth mh. (2) unity of opinion can be about theoretical things as well as about practical ones (cf. Va has le gei.). i. . Plato's use of the words o modoxi a (R. while friendship cannot (1155b34±1156a3) and o mo noia is jiliko n.2). it is not o modoxi a (ˆ 2). See Intr.5. e5 o modoxi an .C O M M E N T A R Y: 409e5 (1) (2) (a) (b) (c) (a) (b) (c) o mo noia (ˆ jili a) is either o modoxi a or e pisthÂmh (e5). and uses o modoxi a to refer to agreement in opinion as opposed to shared knowledge. but friendship cannot be about theoretical things (EN 1167a24±6. Contrast the report of the protreptic speech. o modoxi a is sometimes evil (e6 hnagkaÂzonto ktl. a logical one. unity of opinion (EN 1167a22±4). 310e9±10) is identical. . (1b) because (gaÂr) (2b) and (2a). h e pisthÂmhn: Aristotle seems to imply that o mo noia had been stated to be identical to omodoxi a.e. Plt. the opposition o modoxi a : e pisth mh is. This intellectualist conception of o moÂnoia is present at our passage.5. the verb o mognwmoneiÄ n covers agreement on theoretical and practical things. MM 1212a19). But elsewhere in this part of the dialogue. the optative is used (408c9. leÂgoi: against this reading of ADF Stob. He uses o modoxi a to refer to agreement about nohta (this much can be inferred from the conclusion periÁ taÁ praktaÁ dhÁ o monoouÄ sin.. 1 126c6±127d5 it is used to prove Alcibiades' ignorance (Intr. The order of the logical steps is (1a).. EE 1241a16±18). Aristotle's terminology is slightly di¨erent from the one used here. note on 409d6±7 thÁn jili an . As two persons may or may not agree on things about which they have doÂxa but must necessarily agree on things of which they possess e pisth mh. Aristotle o¨ers two arguments against o mo noia being taken as omodoxi a: (1) unity of opinion can occur between people who do not know each other (EN 1167a23±4). 433c6.

Point (b) may be argued also against Ge¨cken's o mo noian kaiÁ e pisthÂmhn. ou do xan ou san (`RaÈ tsel'. 432c8±d1). as Euthd. e9±10 wste tautoÁn ejhsen ei nai o moÂnoian kaiÁ e pisthÂmhn ousan. 490c9. See also Adam on R. remains doubtful. gi gnesqai: probably not the copula (so Zuretti. gi gnesqai omodoxi ai: the construction is the passive transformation of a nagka zw with accusative and in®nitive (cf. 4) `to say (think) that necessarily . kaiÁ dikaiosu nhv e rgon: these words are not relevant for the present argument. Some translators (Ficinus. The same goes for Verdenius' proposal (`Notes on Clitopho'. Whether this use of a nagka zein is really di¨erent from the supposed meaning `to prove'. 436 n. (b) this does not explain the addition of kaiÁ dikaiosunhv e rgon (see previous note). Cf. Tht. note on e4±10). 145) to take kai as `and that' (German `und zwar') ± this seems impossible to me when a predicative participle follows. not B'). then. as if ˆ a nagkaiÄ on le gein (nomi zein). a ll' ou do xan: so the MSS and Stobaeus. A few editors assume a lacuna 320 .' But I do not think that kai as a Focus particle can occur with participles unless they are substantive (excepting poetic kai ˆ kai per). SouilheÂ. that concord was the same thing [sc.C O M M E N T A R Y: 409e6±409e9 e6 meÂn: the contrasted idea is expressed by the wste clause in e9. But (a) it is not easy to see how ou san could have crept in. e6±7 h nagka zonto . as can be seen from its formalised arrangement (cf. . I take it that the author wishes to underline that everything which had been said of jili a is true also of o mo noia ± this explanation may help us to reach a decision on the textual problem of the next line.'. it is not surprising that the pluperfect active of this verb is not found elsewhere in the corpus. note on e4±10): o mo noia is e pisthÂmh. LSJ. Gonzalez). I see only one way of making sense of the words as they are transmitted.v. not opinion. being besides knowledge. as Stallbaum says (ad Cra. h ti mazen: of rejecting an alternative. De Win) ignore ou san so that the phrase very neatly renders the conclusion of the ®rst syllogism (cf. as it is unnatural to sever pollai from kaiÁ blaberai . e8 wmologh kei: the pluperfect denotes that the admission made in the past is relevant to the present situation. 1) ± which could be defended palaeographically if one assumed that ou do xan was omitted owing to homoiarcton and inserted afterwards in the wrong place with help of all' (the most normal coordinator in contexts of the type `A. Sartori. Therefore the text must be corrupt. la Magna. . s. a natural anacoluthon. . as friendship]. and that is to take o mo noian as subject of tau toÁ n ei nai and kai as an adverb: `His conclusion was. 196c1±2 i na mhÁ taÁ au taÁ o autoÁ v anagkaÂzoito ei dwÁv mhÁ ei de nai a ma. As Plato seldom compresses the report of an argument.. 292e2. .

Hermann reads o mo noian kaiÁ hdikaiosu nhni. 321 . Zuretti `a¨ermava che cioÁ appunto eÁ la concordia'. CR 32 (1918) 147±9 at 148) very ingeniously changes kai to wv and omo noian to o monoi ai (dative). similarly la Magna). with jili an inserted. Socrates' friend had admitted that friendship is always good and the result of justice. it is the result of justice and it is a sort of knowledge. which had occurred independently to Bertini (477±8 n. non opinio esset. it is possible to make it refer to a gaqoÂn and dikaiosu nhv e rgon rather than to jili an: `therefore he a½rmed that concord was all this. Secondly. it does so indirectly. I prefer the latter: if tau to n is left unspeci®ed. Opposed to them are two other conjectures which make Socrates' comrade say that friendship is identical to concord and (a form of ) knowledge. this makes sense only if it had been stated previously (quod non) that o mo noia is always good and the result of justice (the conclusion would have been a logical error to boot). There are two objections to these two hypotheses. F. If Clitophon goes on to report that the conclusion was (wste) that friendship is identical to o mo noia.C O M M E N T A R Y: 409e9 after kai . not o mo noia.): the discussion is not about justice but about the e rgon (409e9±10) of justice. This brings out more clearly the function of the wste clause: with kai deleted. not about friendship: the reasoning thus ends with all the properties of concord which the discussion has yielded: it is always a good. The latter two conjectures make concord the topic of the clause. (410a4±5 thÁn deÁ upoÁ souÄ legome nhn dikaiosu nhn h o mo noian does not contradict this: see on 410a5 dikaiosuÂnhn h o mo noian.' R. The last statement would have been su½cient for the validity of the syllogism. Ast places kai before omo noian (in a footnote. 1). thÁn deÁ u poÁ souÄ legome nhn dikaiosu nhn h o mo noian ktl. We have to choose then. between Baumann's kaiÁ hjili ani and Bekker's deletion of kai . Therefore I submit that the w ste clause contains statements about concord. This is not quite logical if jili a. . but obviously the author wanted to outline the concept of o mo noia he is dealing with as fully as possible. rightly rejected by Susemihl (524 n. . it refers directly to all that has been said about o mo noia. too' (cf. the argument is criticised by the bystanders on the ground that the concept of o mo noia has not been de®ned clearly enough (410a3±6 kaiÁ h i atrikhÁ o moÂnoia ti v e sti . C. materially the same is Bekker's deletion of kai : both solutions give a text which says that o moÂnoia is identical to friendship and (a form of ) knowledge. G. not agreement in opinion. had been the topic of the last statement made before criticism is raised.).) Better is Baumann's o moÂnoian kaiÁ hjili ani. his text is that of the MSS): `quare idem esse dixit atque consensionem quippe quae scientia. Bury (`Notes on some passages in Plato and Marcus Aurelius'.

51) that interpolation of kai at Cra. van Ophuijsen in Two Studies. 171d7. Lexicon. Cra. . 200c3. 106e1 o po te dhÁ toÁ a qaÂnaton kaiÁ a diaÂjqoro n e stin ktl. followed by a de clause with new information. cf. note on d2±e10. M. la Magna) for which the examples collected at KG i 322 . 372). Besides. The participle must not be taken with oi paro ntev (Ficinus). 588b1. We shall see that the same expedient was used at 410b4 (ad loc. the e¨ect is that of a major transition. s. a pori a is a state common to the interrogator and his target (Intr.v. . o moÂnoian: a considerable number of translators (Ast. Intr. 159e6. entauÄ qa .C O M M E N T A R Y: 409e9±410a4 The interpolation of kai is easily explained: a reader not content with the statement that `concord is the same thing' wanted to add `as knowledge' (which is materially correct but ruins the grammar). Souilhe . . I have argued (`Plato. Water®eld) take this as subject of diape jeugen (with hmaÄv supplied as object).4. Cf. apparently justice is not included. The perfect sundedraÂmhka is found at Plt. as it does at Ap..2 n. G. note on 409a3 thÁn .3 for a possible interpretation of this detail (and n. 417c9 deteriorates an already interpolated passage. See Intr. 398d6. . 1982). For its raison d'eÃtre here. b6. It is true that the aporia is pointed out by the bystanders. cf. Herodotus' mannerism of opening a new section by means of meÁ n dh introducing a resumptive clause. . Der Platontext (GoÈttingen 1942). 26a2. 558a7. When this occurs in a subordinate clause which contains no really new information. a4±5 thÁn deÁ . cf. a4 periÁ otou ei si n: cf. kai is found interpolated at Chrm. section ii. section ii. Cratylus 417c'. 266c5. 61b1.3. touÄ lo gou: cf. The accusative seems to be more normal in Plato. R. `ein aporwÄn ist nicht i kanoÁv e piplh ttein' (third edition (Berlin 1861). cf. 343a1. 642 n. 321 for the variant reading e pecei rhsan). the appositional participle is supported by the similar use of aporwÄn at 410c8. Cf. a2 peridedraÂmhken ei v tautoÁ n o lo gov toiÄ v prw toiv: cf. Sartori. 1. Schleiermacher himself wished to delete the participle as a corrupt dittography of oi paro ntev. `verbal repetition or an anaphoric pronoun helps to make a notion more evident' ( J. a3 kaiÁ a pasai ai te cnai: `and all (other) arts'. peri ad ®n. Phd. cf. te cnhn. because. 1 ˆ Textgeschichtliche Studien (KoÈ nigstein/Ts.. a construction (`in accusativo per anacoluto'. as Schleiermacher remarks. 45. Phd.).3. Zuretti. . Ast. 69a8. but a porouÄ ntev is a piece of comment mixed into the report. Jachmann. 148). R. e9 e jhsen: Plato uses the indicative aorist of jaÂnai sparingly and in post-Republic dialogues only. Phdr. section ii. 410a1 i kanoiÁ h san: cf. 275d1 periÁ w n a n hi. 286 n. since no apori a has become visible yet. (repeating d5±7). De Win. e10 dhÂ: with dh . 265). note on 408e7.. Tht. a porouÄntev: in Plato.4.

tei nousa e stin: for the periphrastic construction. cf. KG i 545). justice is out of place.C O M M E N T A R Y: 410a5±410a7 330 o¨er no justi®cation. 174a. Cf. cf. but almost exclusively if they are accompanied by true verbs of movement (where the end of the movement is anticipated). Greek Prose Style. This is the case at Phdr. Denniston. Plato's Logic. sections i. §18. Tiemann. 204a. 100. note on 410d2 protetramme nov h . klass. note on a5 diape jeugen). cf.. 26b7) or `it is for all the investigation unable to tell' (cf. Cf. For examples in Plato. op. 4. Wochenschr. The asyndetic use of this idiom 323 . There is no need for that (Aerts is mistaken in saying that the other roads to areth are not a priori assumed) and such an analysis is out of the question here: we are dealing with the general tendency of impersonal and intransitive verbs towards periphrasis. Van Leeuwen on Ar.) are constantly interchanged in MSS. 84) add W. R. 319. either `it doesn't recall any more' (cf. 526e2 au to se). Woldinga on X. a phenomenon quite common in fourth-century Greek.. 47c5 poiÄ . a5 dikaiosuÂnhn h o mo noian: strictly speaking. 128). 6 (1889) 248±53. Frequent periphrasis is typical of the later dialogues (Alexander. For the postponement of the interrogative. Aerts.Ma. Cf. ii. I see no problem in dikaiosuÂnhn h o mo noian as object of diape jeugen: medicine and all other arts are able to say what they are all about. Elsewhere in the corpus. De Vries ad loc.)..4.4. 48. cit. True adverbs of place are used in classical Greek where one would expect adverbs of movement (cf. But if the e rgon cannot be de®ned. Riddell. cit. Hp. note on 409e9±10. a6 au thÄv: with e rgon. No such explanation is possible here. not of place. 22±3). Ap. 409c6 (e reiÄ ). J. as the discussion had been about the e rgon of justice.2 (4). `Zum Sprachgebrauch Platos'. 179c6. but this one cannot ®nd and state its subject-matter (cf. W. Plt. further Stallbaum on Smp.. d1 (jhÂsei). substantive) einai ‡ substantival participle.e.± D. KG i 310±11. 294e7) ± preferably the latter.2. Lutosøawski. 308e10 e stiÁ tei nonta. `Participial periphrases in Attic prose'. i. J. Tiemann. Ti. 499a7 mhdamo se a llose. Philol. 23b4. 77±8. add La.. Periphrastica (Amsterdam 1965). Alexander. cit. op. cf. `Digest'. Cf. Nub. To Thesle¨ 's bibliography on the subject (Styles. cf. J. 362±6. neither can justice itself. 305. Smp. o poi: Bekker's correction of MSS o pou: poiÄ and pouÄ (etc. Aerts (op. 228e4 (pouÄ MSS poiÄ pap. Schw. tei nein is found combined with adverbs of movement (Cri. 22) analyses the parallel quoted from Plt. diape jeugen: `it has escaped it'. AJPh 4 (1883) 291± 308. 5± 26 (esp. Intr. a7 TauÄ ta: probably adverbial: `therefore'.28. as `independent' (existential. The postponement is due to the thematic prominence of justice.

The sequence is to be explained as what some modern linguists call `Inzidenzschema' (cf. Chrm. Hdt. 126. R. kaiÁ ei pev: the contrast imperfect : aorist as at 408c9 e panhrwÂtwn. `We might compare the imperfect with a line one of the points of which coincides with another occurrence. the absence of the logical link between this sentence and the previous creates a harsh asyndeton (adverbial tauÄ ta would be only formally asyndetic). cf. section i. cf. the ®rst one is expressed in Greek by the imperfect.3.5. all other translators regard it as object. . 495b5±6 324 . b1 u steron deÁ e ja nh: namely in the course of a discussion not reported by Clitophon. A.. 25±7). 410b1 or 409d9. KG i 157±8. section ii. Stahl. is a `Kurzdialog') may account for that. esp. Of course e rwtaÄn is quite often found in the imperfect without any perceptible di¨erence from the aorist. Smp. pa ntav draÄn: for the double object. Instances include Euthd. Svensson.1. KG i 558. `Historische und despkriptive Linguistik bei der Textinterpretation'. 40±2 (e rwtaÄn). but then so is le gein (KG i 143±4. but cf. Syntax. Schw. Glotta 49 (1971) 191±216. K. 291b1±4. Intr.1 h me rh te e gi neto kaiÁ a ma twÄi hli wi anioÂnti seismoÁv e ge neto. b2±3 pa nta . 8. e pecei rei deÁ periÁ au taÁ ma cesqai´ e peidhÁ deÁ wmolo ghsen ktl. 472b6±7. 280b1±2 sunwmologhsa meqa teleutwÄntev ouk oi d' opwv e n kejalai wi outw touÄ to e cein ktl. 277±8.. eu poieiÄ n: cf. the second one by the aorist.. This occurrence may be at any point of the line.8.5. but di¨erence in genre (Clit. 24. . Pers. note on 409d2±e10). or somewhere in between' (W. An. The order au to v ‡ personal pronoun seems to be more frequent in Plato. 409a5 eipen and in the passage 409d6±e10 (cf. at the beginning. Zum Gebrauch der erzaÈ hlenden Tempora im Griechischen (Lund 1930).3. None of these places is quite comparable to either Clit. 15. 165.. X. cf. note on 409d2. ou c w v e gwÁ nuÄ n r aidi wv le gw. Cf. all' e lkoÂmenov kaiÁ mo giv. . The Greek Imperative (Amsterdam 1966). Plato does not often make use of the reported dialogue to curtail the report of an argument in this way. 201e6±7.21. Stallbaum on Euthd. 50±60 (le gein)). at the end. Theoc. Grg. 201±3): if one action is the framework within which a second action falls. 169c3±d1. a7±8 h rwÂtwn. 4. cf. seÁ au to n: cf. R. 350c12±d1  O dhÁ Qrasu macov wmolo ghse meÁ n paÂnta tauÄta. It would seem that Susemihl takes tauÄ ta in the same way (`Da fragte ich denn'). 105. But what questions are put is clear anyway. F. a8±b1 dikaiosu nhv . 342d2±3 Sunwmolo ghse meÁ n kaiÁ tauÄ ta teleutwÄ n. A. . Bakker. 273b.64. 207a5±6.C O M M E N T A R Y: 410a7±410b2 seems to be unparalleled in Plato. and there seems to be no need for the emphasis on tauÄ ta (if object) created by disjoining it from h rwÂtwn. Strunk. d9 and Intr.±D.

note on 408c4±7). R. ii. Note that the word jilosoji a is absent from this phrase in Clit. Susemihl. Bury. u pomei nav kaiÁ liparwÄ n could only be explained as a zeugma. b3±4 TauÄ ta . ei v a rethÄv e pime leian ± ei v jilosoji an kaiÁ arethÄ v e pime leian. in his edition) is needless.1). (2) `having waited for an answer to these questions'.' Besides. so Sartori and (with many variations) most other translators. kai could be retained as emphasising liparwÄn. . H. Euthd.1. cf. (1) lipareiÄ n is elsewhere intransitive in Plato and all other fourthcentury authors (D. For arethÄv instead of dikaiosuÂnhv or yuchÄv (cf. Bury's panto v (for paÂntav. tauÄ ta is internal object of upomei nav.7. La Magna construes pa ntav as subject. . (2) The phrase ou c apax ou deÁ diÁ v a llaÁ poluÁn dhÁ upomei nav croÂnon does not go well with a pei rhka: `I have grown tired of enduring this not once nor twice etc. twÄ n nuÄ n a nqrw pwn ka llist' a n protre yaite ei v jilosoji an kaiÁ arethÄ v e pime leian. Gonzalez. Souilhe ). In kaÂllist' a nqrw pwn the genitive reinforces the superlative (`le mieux du monde'. Intr. but I doubt whether this construction is consistent with ou c a pax oudeÁ di v. cf. If kai is deleted. Burnet. MuÈ ller. . rather than being fully partitive (contrast 325 . KG i 323±4. nomi sav: `having come to conclusion' (ingressive aorist). 518a4. . (3) The change in tense from upomei nav to liparwÄ n is puzzling. compare kaÂllist' anqrwÂpwn ± twÄ n nuÄn anqrwÂpwn ka llist'. b4±6 toÁ meÁ n protreÂpein ei v arethÄv epime leian ka llist' anqrw pwn draÄn: cf.. 448a7) or as `conclusive' or `consecutive' (cf. note on 406a9±10). 274e8±275a2 u meiÄ v a ra .208 is a hendiadys).1.. but the result (`tutti fanno ogni cosa a ®n di bene') is needlessly vague.2. 408d2±3 thÁn SwkraÂtouv protrophÁ n hmwÄ n e p' arethÂn. but the emphasis is rather on apei rhka (cf. In (2). H. u pomei nav: (1) `having endured this (getting unsatisfactory answers)'. so Schleiermacher. 275a6.C O M M E N T A R Y: 410b3±410b4 etc. . sections ii. carried over from the preceding clause. There can be no doubt that o di kaiov is subject and e jaÂnh predicate. tauÄta . . and its tense self-evident. liparwÄ n becomes a complement of a pei rhka. probably because the author avoided it on purpose (cf. there is room for doubt whether apagoreuÂw can be construed with an aorist participle. 408e3±409a3. 21. but I have seen no parallel for this use of adverbial kai preceding a participle. (cf. At most. 278d3. The phrase looks like a conscious imitation of the sentence quoted from Euthd. b4 kai : there are three reasons which in conjunction justify Baumann's proposal to delete the particle (followed by Schanz. Grg. 410d6±7).. MuÈ ller translates as if kai were not there).

a nqrwpov i 3 b. But the fact that in Plato the comparative always has a complement in the genitive. makroÂteron never equals ple on in Plato. this is an instance of a sentence with two main clause predicates. Anakoluthe. on R. iv'. . parallel: twÄn nuÄ n anqrw pwn kaÂllist').: the second alternative (`you do not want to impart your knowledge'. maxime. Reinhard. 187c1. `Notes on Politeia. 61. 629a6. 40c10 kaiÁ ei te dhÁ mhdemi a ai sqhsi v e stin . c6) is missing: owing to the analogy of praising steersmanship the ®rst alternative has grown so lengthy that the original construction has ceased to govern the sentence. `going further than that in amorous contact'. for which cf. h ktl. my note on R. b8±c1 katamelethÄ sai toÁ n e painon periÁ au thÄv: the verb must mean `to train oneself completely in' not `to deliver' (Susemihl) because (a) meletaÄ n `to declaim' is used intransitively (Phdr. cf. KG i 286 a 10. L.43). v. 39) and Heidel (Pseudo-Platonica. Phd. 403b7±c1 makro tera tou twn suggi gnesqai. 413a8±9 makro tera touÄ proshÂkontov e rwtaÄ n. Tht.v. MuÈller. D. Cra. If so. Steinhart (72 n. 32b5. b6 duoiÄ n deÁ qaÂteron. It seems at least possible that the relative clause o ge noit' a n ktl. Anakoluthe. cf. . 344±5. oudeÂn: usually taken to be synonymous to ou deÁ n ple on. b6±7 makroÂteron .v. . 151±66. optime al. makrote rwv at Sph. 48 n. cf. Schleiermacher). With the pendent h compare Ap. (e2) ei ou n toiouÄton o qa natoÂv e stin ktl. b7±8 periÁ a llhn h ntinaouÄ n te cnhn: cf. (b7±c2) is treated as antecedent of tau toÁ n dhÁ kte (c2±4). cf. or rather as an apposition to a pair of disjunctive clauses (`aber eins von beiden. 363d2) and i 5. which is absent here. Ast. at 326 . cf. Anacoluthon caused by intervention of a comparison or an example is very frequent in Plato. similarly Susemihl). duoiÄ n qa teron is frequently used as a clause apposition. makroÂv i 4 (add R. 258c7. note on 409a3 thÁn . de Strycker±Slings on Ap. LSJ s. LSJ s. Cf. 414. Lexicon. Plt. this is obviously not relevant.'. quam potest. s. 432d7±e3. since the two are not necessarily interchangeable in the Clitophon. cf. . but perhaps `nothing which goes further than that' is preferable (cf. te cnhn. so that duoiÄ n deÁ qaÂteron has to be repeated at c5. 66e5. R. 389a3±7. 8) consider the use of makroÂteron for ple on to be un-Platonic. `nichts Weiteres' H. Ast and England on Lg.c. should be noted. `going further in asking questions than is ®tting'. entweder . `Notes on Politeia. Reinhard.)' with numerous examples. so also at c5.v. . . 228b6. anqrwpov `superlativis apponitur ut eorum vis augeatur (Latin. 33±57.C O M M E N T A R Y: 410b6±410b8 the Euthd. . 283c5. iii'. Although of course makro v and polu v tend to become synonyms. .

39 ± duly repeated by Heidel.1.. 19. Pseudo-Platonica. 687a2±3 etc.3 thÄ i d' u sterai ai e piÁ meÁ n thÁ n po lin ou deÁ n maÄllon e pe pleon. 195a2. 346b3±6 ou de n ti maÄllon. section ii. 1). is only formally similar. cf. rather senseless if the verb refers to the delivery of a eulogy. `The Cleitophon of Plato'.2. . 3. Water®eld. 202b2±4 ou tw deÁ kaiÁ toÁ n  Erwta e peidhÁ au toÁ v o mologeiÄ v mhÁ einai agaqoÁ n mhdeÁ kaloÂn. c4±5 ou mhÁ n to ge emoÁ n outwv ecei: either (1) to ge e moÁ n adverbial 327 . 48 n. 87d1± 2. 244a3±5. `tu sai ben esaltarla ma cioÁ non implica che la conosca meglio' (Sartori). Lg. the opening sentence of the speech). note on 407b6±7 e xaskh sousin kaiÁ e kmelethÂsousin) if training in delivery is meant. I cannot accept the use of this word as a mark of inauthenticity (Heidel. 260c7±8. Intr. De Win). 1).v. Th. R. te cnhn. meleta w ii 1. 233b4±5. the verb means `to train oneself ' also at D. 8) that `das dicht an kai per anstreifende dio ti' is a sign of inauthenticity.1 ama t' eiko v e sti touÁ v e gceirouÄntav taÁ yeudhÄ martureiÄ n kaiÁ thÁ n apologi an euqe wv upeÁ r au twÄ n meletaÄn the adverb eu qe wv proves that training is intended (cf. contra Grube. (b) the pre®x kata. cf. note on 409a3 thÁn .. Sph. cf. 307 n. The aorist in®nitive is chosen because it harmonises with the meaning of the pre®x.255 loga ria du sthna meleth sav kaiÁ jwnaskhÂsav. section ii. kai per is equally possible in contexts like these. H. c1±2 periÁ twÄn allwn tecnwÄn: in view of the following sentence these words imply that dikaiosuÂnh is not among them. mhde n ti maÄ llon oi ou deiÄ n au toÁ n ai scroÁn kaiÁ kakoÁn einai.79. kai per e n pollhÄ i tarachÄ i kaiÁ jo bwi o ntav. cf.1. Not: `because you eulogise it you are not an expert' (so Schleiermacher. 46. c3±4 ou maÄ llon . e neka touÂtou kaleiÄ v maÄ llon au thÁn i atrikhÂn. Intr. There is no reason to assume with Steinhart (72 n. PseudoPlatonica. cf. Plato often uses causal adjuncts and clauses here. Phd. e a n tiv kubernwÄ n u gihÁv gi gnhtai diaÁ toÁ sumje rein autoÁn pleiÄ n e n thÄi qalaÂtthi. c4 e gkwmiaÂzeiv: eulogy and protreptic are evidently felt to be related.2.C O M M E N T A R Y: 410c1±410c4 least in fourth-century Attic (cf. LSJ s. ± e painov peri tinov is frequent in Plato: Smp. the opposition between causal and adversative adjuncts is neutralised: `You are none the more an expert because you can praise it' > `you are not an expert even if you can praise it'. at D. 48 n. For the relationship of e painov and protroph cf. ± In contexts like this one. Phdr. ii 5b). Smp. MuÈ ller. cf. . Phdr. Of course. . dioÂti: `it is not the case that you are more knowledgeable about justice because you praise it beautifully'. but it has its normal function of marking perfective Aktionsart (cf. 307 n. The article probably indicates that e painov is not a particular eulogy but praising in general: `the praise pertaining to it'. adduced by Grube (`The Cleitophon of Plato'. .

More than once. the second is decidedly better. 533a2. note on 408d1 u potei nwn. At a number of places there is a similar ambiguity: Chrm. 176b2. cf.C O M M E N T A R Y: 410c5±410c6 `as far as I am concerned' (here `in my opinion'. Van Leeuwen on Ar. 458d5.g.3.1 n. Orwin. or phrases (such as ou twv e cei ) equivalent to nominal predicates. Clitophon underlines that his position is di¨erent from those who deny that Socrates has any such knowledge.3). 105. Intr. If taken in the ®rst way. at Cra. 109e1±2 (taÁv areta v). ei de nai with a nomen qualitatis as object is found e. Zuretti. Thesm. Lg. R.) or (2) to ge e moÁ n subject of ou twv e cei (`my position'. unless one takes him to imply that Socrates indeed does not want to tell him any more (cf. section ii. turns into that of Tht. But I see no point in introducing the idea of psychological a½nity here: it is the Socrates of explicit protreptic who is being attacked by a Clitophon who himself imitates not unsuccessfully Socrates the intellectual midwife (cf. 188c1. 688a6 (and England ad loc. Grg. Ti. `I'. Lg. 188c4. note on c6 ouk e qe lein au thÄ v e moiÁ koinwneiÄ n). ei rwneuÂsoio deÁ kaiÁ paÂnta maÄ llon poihÂsoiv h a pokrinoiÄ o ei ti v ti se e rwtaÄ i. but are better accounted for as depending on a verbum declarandi implied in ou mhÁ n to ge e moÁn ou twv e cei. Sph. 425c2. section i. so Ficinus. `non eÁ questa la mia a¨ermazione'. 19d3.). 161e4. Gonzalez). In view of Clitophon's immediately re-stating the alternatives h ou k ei de nai se h ouk e qe lein authÄ v e moiÁ koinwneiÄ n. Clitophon denies that Socrates has no `technical' knowledge of justice. Tht. cf. 643a3 (and Ast ad loc. 248. Water®eld). I have seen no good parallel for adverbial to ge e mo n in the construction that one would have to assume if (1) is accepted. e1±3 i na SwkraÂthv toÁ ei wqoÁv diapraÂxhtai´ au toÁ v meÁ n mhÁ 328 . 163c1 (thÁ n oxu thta kaiÁ baru thta). R. 345a2 (and Tucker ad loc. 237b4.5. La. Prt. La. c5±6 h ouk ei de nai se h ouk e qe lein: the in®nitives may be explained as continuing the construction of b6 after the repetition of duoiÄ n deÁ qaÂteron. o ti suÁ [Socrates] apokri nasqai meÁ n ou k e qelhÂsoiv. 426a4±5 (thÁn o rqo thta). c6 ouk eqe lein au thÄv emoiÁ koinwneiÄ n: this possibility is certainly not `slechthin ungereimt' as Steinhart (56) says (at 410e3 Clitophon rather clearly suggests that Socrates does know more than just to exhort others). 337a5±7 prouÂlegon [ Thrasymachus]. c5 ei de nai: object dikaiosuÂnhn (cf.. Tht.). cf. see Intr. Criti. Nor am I much attracted by the parallels R. to suppose that all of a sudden the Socrates of Clit. In the second interpretation.. au thÄv in the next line). SouilheÂ. 338c5. whereas toÁ e mo n is frequently subject of nominal predicates. De Win. Apart from that. 384b6 (thÁ n alh qeian). Plato makes it clear that not everyone is ®t for Socratic `instruction'. (and Plato in general) would make no sense at all of our dialogue.

J. section i. M. 4. which had been left out as object of ei de nai. Tht.C O M M E N T A R Y: 410c7 a pokri nhtai. Mem. Though in reality a vacuum between Clitophon's Socratic and Thrasymachean periods would have been possible. as Denniston says. Not `knowledge' (so Bury and Orwin): even if the phrase `to share justice with me' (or `to let me share justice') is a little odd. though it might be corresponsive (kaiÁ proÁv Qrasu macon . but it remains pointless unless it can be shown that Clitophon. was clear enough. GP 2 307±8. Lg. no such vacuum is hinted at. passage. allou d' a pokrinome nou lambaÂnhi lo gon kaiÁ e le gchi. The present is certainly better. It is less likely to be a correction of oimai than the other way round. alhqeÁ v o neidi zousin.9.4. In fact not only does he understand it. 150c4±7 o per h dh polloi moi wnei disan. (2) The imperfect in polla kiv e xeplhttoÂmhn a kou wn (407a6) is understandable only if Clitophon no longer frequents Socrates' meetings. kaiÁ allose). it is acceptable when the Socratic paradox that virtue is knowledge is taken into account.). 329 . van Ophuijsen in Two Studies. ± The parenthesis is super¯uous. 408c1 scedoÂn (and note ad loc. the consensus of ADF favours the fuller reading (the status of A 3 is uncertain). w v touÁ v meÁ n a llouv e rwtwÄ. cf. koinwneiÄ n: not didaÂskein (cf. cf. so close after dikaiosu nhv e pisthÂmoni. paraÁ deÁ twÄn allwn periioÂnta manqaÂnein kaiÁ tou twn mhdeÁ caÂrin apodidoÂnai. 155b5. Cf. justice. 23d8. Cf. note on 406a9±10. it is. since the meaning. See note ad loc. 338b2 quoted above) ± the author is well aware of Socrates' ways and means. 788d1. but he practises it on an un-Platonic Socrates. kai : probably the adverb with `conclusive' meaning (cf. X.. poreuÂsomai A2 . note on 408c6±7. R. Tht. for the following reasons: (1) It is a fair inference from 406a1±4 that Clitophon has already experienced Thrasymachus' sunousi a. Intr. does not understand the rationale of Socratic questioning. in this literary text it therefore does not exist. like Thrasymachus. The dilemma is probably not intended quite seriously. . cf. autoÁ v deÁ ou deÁ n apojai nomai periÁ ou denoÁv diaÁ toÁ mhdeÁ n e cein sojo n.2(1). c7 oi omai is found in parenthesi at Ap. 798d4. 149). besides. poreuÂomai: so ADF. au thÄv: sc. the ®nal sentence where proÁv te lov arethÄ v e lqo nta can only refer to acquiring complete knowledge of a reth . . 338b1±3 au th dhÁ e jh h SwkraÂtouv soji a´ au toÁ n meÁ n mhÁ e qe lein dida skein. often preceded by dhÂ.3. The fact that the reproof is made with particular insistence by Thrasymachus in Republic 1 might perhaps be thought to favour dependence of the Clit.

5. It would be rash to assume that here e qe leiv. MT §520. This indicates that the author does after all attach some value to it. (to which could have been added D. which is inept.3 n. touÄt' a n eu roite proceiro taton ktl. 22. . i 306 n.. F. 82a6 (poiÄ . 486d5 (allose ‡ poi ).) ei v plou touv a poble yai . the hypothetical force lies rather in the combination e qe leiv ‡ inf. Cf. cf. . 243b5 a lloqi phi. d6 e pime leian poieiÄ sqai. can hardly support the MSS reading. For its aim. touÄto me giston an eu roite ktl. . with no such in®nitive following. as it does at R. MuÈller: gi gnoit' an based on the e qe loiv of inferior MSS and the earliest editions.51 ei gaÁr qe let' e xetaÂsai . c8 e pei : explains aporwÄn. connecting kaiÁ nuÄ n dhÁ qeÁ v ktl. . . de opposition. section i. my note in `Notes on Politeia. e qe loiv Bodl. 360± 1. The dialogue as a whole ends on the same note as the protreptic speech (see Intr.5. which now replaces dikaiosuÂnh de®nitively (announced b5. which is suspect for other reasons (Radermacher's allov ei phi is palmary). The ®nal phrase proÁ v te lov arethÄv e lqo nta eu dai mona gene sqai is neutral. Cri.2.3. like the previous one. .C O M M E N T A R Y: 410c8 (3) The future requires opoi a n du nwmai. cf. 45c1. section i. note on b6). kaiÁ nuÄn dh (d4) may have helped trigger the anacoluthon.55 ei qe lete ske yasqai ti douÄ lon h e leu qeron einai diaje rei.).2 n. 202e7. ii. e1 h melhke nai ). ophi MSS.5. Sph. 414e3. gigne sqw: the anacoluthon is caused. by the intervention of an analogy (cf. . ± See also Intr. . section ii.3. 188). D. In these instances (e )qe lw is followed by a verb meaning `to investigate' or the like. . which gives the impression of an urban periphrasis (ei qe lete ske yasqai ± ei ske yaisqe). sections i. section i.5. iii'. Intr.. allose). cf. 230e1. Phdr. is also 330 . Tht. The anacoluthon is complicated by the meÁ n . 1).6. Hermann deleted tau toÁ n gigne sqw. . because poreuÂsomai (. even the in®nitive which we would expect to parallel pauÂsasqai is missing. R. Given these places. . c8±d5 epeiÁ . ai scunqei hv an e piÁ seauÂtwi. The analogy looks like a repetition of 408e3±409a3.3.). Hermann (Platonis Dialogi. . eqe leiv: C. but it is essential because it diverts attention from the question `What is dikaiosu nh?' to the need of e pime leia.. kaiÁ a llose: cf. d4 qerapei av (cf. Cri. xxviii) attributes a hypothetical force to the indicative on the basis of Alc.. There is no reason to change the text (H. referred to d2 mhÁ ameleiÄ n. Wilamowitz seems to read the present: `Daher bleibt Kleitophon bei Thrasymachos' (Platon.420a5. Phd. C. iii p.) o poi duÂnamai means `I shall go wherever I (now) can'. o poi: so Bekker. F. 97a10 (allose ‡ o poi ). 91. 51d8±e1.. 1 122b8±c3 ei d' au e qe leiv (T Olympiod. 407b8 e qerapeuÂsate). Men. cf. 52b6. . note 92 to Intr. 22.

op. Hdt. 141±7. Gonda. protre pw. Syntax.15. 147 and 148 for the semantic properties of the verb.. Conditions and Conditionals. deiÄ n: the in®nitive is pleonastic. ± Cf. 408e3±4. `A remark on `periphrastic' constructions in Greek'. as LSJ note (s. RoseÂn.C O M M E N T A R Y: 410d1±410d3 hypothetical. cf. J. If in a small majority of cases the distinction appears to hold.2. cit. 239d8±e1 toÁ d' e jexhÄv r hte on.4 hn deÁ toÁ deiÄ pnon poieuÂmenon e n QhÂbhisi `it was at Thebes that the meal took place' (H. d2 protetrammeÂnov h : the periphrasis does not denote a state (cf. ii ). B. 278d2 for a possible parallel for the meaning `persuade that'. neutral as to modal value: `assuming that you are prepared . Cf. note on d2 protetramme nov h . Wakker. cf. Euthd. 12 (1959) 97±112. This distinction was forced upon Greek syntax by grammarians steeped in Latin (where it is obviously present). `Die ``zweiten'' Tempora des Griechischen. protetramme nov h ) is a clear proof of the futility of the distinction between irrealis of the present and irrealis of the past (and past potential). the formulation in Stahl. Zum PraÈ dikatsausdruck beim griechischen Verbum'. it may have been used to emphasise another part of the sentence. d3 toÁ e jexhÄv: object. viz. esp. twÄn lo gwn twÄ n protreptikwÄn: for the plural. elegev: the choice of imperfect for a past unreal condition (cf. section ii.1.4. . 302 331 . cf.1 n. as always.1. ti na h miÄ n w jeli an ktl. aorist and pluperfect. . section ii. cf.. Gonda. MH 14 (1957) 133±54.. when followed by an in®nitive the verb means `to persuade to do a thing' (cf. that it was necessary not to neglect my body'.' Cf. even though. cf. 97±104): `if it had been about gymnastics that I had been exhorted'. though o pwv for w v is a problem). d1±2 periÁ gumnastikhÄv: emphasised. note on 407a6 pollaÂkiv. explained asyndetically by an appositive indirect question as Phdr. that is easily accounted for by verbal aspect. the fact that the focus of the sentence is periÁ gumnastikhÄv (cf. cf. Mnem. Besides. deiÄ n. 125±30.v. not ei v gumnastikhÂn. this facilitated the insertion of deiÄ n: `if it had been about gymnastics that you exhorted me. d1 proÁv e meÂ: depending on twÄ n lo gwn: the construction is a bit loose (pauÂsasqai le gwn proÁv e meÁ touÁ v lo gouv would have been more precise). cf. previous note) gives the in®nitive a more independent status than it would normally have. 111±12). which in fact is the only determining factor for the choice between imperfect. it is best understood if the ei clause is reformulated actively: protre pein tinaÁ touÄ swÂmatov deiÄ n mhÁ ameleiÄ n is hardly o¨ensive. The indicative is. 9. touÂtwn: emphasised by disjunction. note on 407e5. esp. Intr. Intr..

191c8±9 qe v . a colloquialism would be de®nitely out of tune. h melhke nai: see Intr. . The wording of the sentence suggests that of 408e3±409a2 w sper . but not necessarily so: emphasis is not restricted to colloquial style. Thesle¨ 's contention (Styles. h v e neka talla diaponou meqa @ 408e7±8 o sa touÄ sw matov e neka diaponou meqa te kaiÁ ktwÂmeqa. i atrikhÂ: e stin katage laston twÄn meÁ n allwn e pime leian poieiÄ sqai @ 408e6±7 ai scroÁn purwÄ n meÁ n kaiÁ kriqwÄ n kaiÁ ampe lwn e pime leian paÄ san poieiÄ sqai.2. similarly Sartori. and nundh to something in the last part of the conversation (dihÄlqon can both refer to extensive and to concise treatment. . 30d8. it is unlikely that a kaiÁ nundh dihÄlqon refers to the whole dialogue). See the examples collected in KG i 660±1. d5 qe v: cf. . . e1 tauÂthv: the resumption of a word within the same clause by means of an emphatic anaphoric pronoun (ou tov. In Clitophon's last appeal. e keiÄ nov. toÁn KleitojwÄnta: the formality of Socrates' opening words is alluded to. Stallbaum on Chrm. Both sentences give a concise version of Socrates' speech. 163c7. SouilheÂ. .5..C O M M E N T A R Y: 410d5±410e1 and his examples 303. Tht. e1±3 kaiÁ talla . who makes nundh refer to the whole con- 332 . 481c1 with Dodds' note. . la Magna.' The various possible constructions of ei rhke nai. on Phlb. note on 407e7 touÄ meÁ n arxontov. d5±e1 qeÂv . Intr.5. Conditions and Conditionals.. 15 n. Grg. The same holds for the chimerical distinction between unful®lled wishes of the present and those of the past. Water®eld). one through an analogy. cf. section i. e n taiÄ v yucaiÄ v hmwÄn e noÁ n khÂrinon e kmageiÄ on. 27c10. This would mean that something in Clitophon's last few words is the logical antecedent of the whole conversation or at least a greater part of it (contra Modugno. the number of possible referents of tou toiv (cf. note on 408e1) and the number of possible antecedents of a (so MSS) contribute to make this an obscure sentence. dihÄ lqon: `Take it for granted that all the other things which I have now said after those were also said with that intention with which I went through them (speaking about gymnastics) just now. au to v is unemphatic and should not be treated under the same heading) is often colloquial (Thesle¨. ± The asyndeton has perhaps explanatory force: the invitation kaiÁ nuÄ n dhÁ tau toÁn gigne sqw is a realistic one because Clitophon has now been exhorted to the care of the soul. 91). 144±50. Some translators take tou toiv as antecedent of a : `e fa conto che io abbia ora esposto cosõÁ anche tutto il seguito di quanto ho esposto or ora [ˆ just now]' (Zuretti. Wakker. Styles. . cf. . Ap. ta lla paÂnta strongly suggests that nuÄ n refers to a greater part of the conversation than nundh .3. section i. one directly. Cf. 2) that qe v is un-Platonic proves groundless.

). ± I do not understand Verdenius' explanation (`Notes on Clitopho'. 491c6).]. 369b4.] I meant everything following the exhortation [pred. So please do tell me what qerapei a my soul needs given its nature. I can make no sense of what is syntactically speaking the most natural way of construing the sentence: to make ta lla paÂnta . and also with the clause following presently mhdamwÄ v a llwv poieiÄ n (which does not mean `to act in the same way as I did' but `to do what one is asked'.1). `By the rest of my speech [obj.3.' The disadvantages are that both nuÄn and ou twv are di½cult: nuÄ n may perhaps contrast the analogy (nundh ) with what has now actually been intended. Besides. outwv (`likewise') may point back to the resemblance between exhortation towards bodily care and care of the soul. R. or conversely: `By what I said just now about gymnastics [obj.' The latter. dass ich auch uÈ ber Alles daran sich KnuÈpfende. cf. though inane.) is o¨ered by H. Clitophon says that his whole speech was in fact the logical sequel to what is stated in this analogy. Why does Socrates have to be told that Clitophon has told a long story? A variant (moulded on the type pwÄ v touÄ to le geiv. because Socrates never said such a thing.] I meant just what I said just now about gymnastics [pred. 145±6) that nuÄn ˆ `in the present case' and that nu ndh refers to the whole speech. I shall praise you wholesale. if you will do so. Water®eld's identi®cation of a kaiÁ nundhÁ dihÄ lqon with the sequence eyes ± ears ± lyre (407e9±408a4) seems to me to stretch nundh well beyond what can be referred to as `just now'. I can see no candidate but the sentence which contains the last analogy (c8±d5). 328b1. When I spoke of your telling me what qerapei a my body would need given its nature. .]'. namely that he has in fact spoken of the qerapei a of the soul. Intr.. Orwin).C O M M E N T A R Y: 410e1 versation from 409a onwards). a statement that Clitophon had done Socrates' job for him is inconsistent with the ®nal appeal kaiÁ nuÄ n dhÁ tau toÁ n gigne sqw. MuÈller: `und sei des Glaubens. I was referring to what comes next once this has been agreed upon. was ich eben eroÈ rtete. section ii. ears and lyre. mostly coupled with a positive command). dem gemaÈss 333 . . taÁ touÂtoiv e xhÄ v object of ei rhke nai and antecedent of a : `and suppose also that I have made all the other subsequent statements which I rehearsed just now' (Bury. Though of course Clitophon's examination of the pupils has had the purifying e¨ect of Platonic elenchos (cf. if so. This can occur with the perfect (Grg. has at least the advantage that it suits the line of thought: `I agree that e pime leia thÄ v yuchÄv comes ®rst. Clitophon cannot agree (o mologouÄ nta) that it is absurd to care for one's eyes. Another possibility is to construe ei rhke nai with an accusative of the object and one of the predicate (type ti nav le geiv touÁ v belti ouv. cf.

108. e3 kai : probably consecutive. Though with some di½dence. note on 406a9±10. des Places. deiÄ sqai ) wrongly classi®es this under the use of de omai with genitive of person asked expressed and accusative or in®nitive or both implied. but would only be acceptable if in pwÄ v touÄ to le geiv. but it is not in the Greek: sentences like o per e legon (which we translate `as I said' but which have in fact the function of an apposition) do not occur with the plural neuter nor with synonyms of le gw. . ei rhke nai has the resultative sense it ought to have and nuÄ n underlines it: `the whole of my position as it has now been stated'. In fact deoÂmenov is used here with the personal genitive only (half-absolutely. 21. D. this is the more attractive because an adverbial kai in the main clause is needed and the kai before talla must be connective (qe v ± oi ou). cf. the gist of which was given in the previous sentence (referred to by tou toiv). Prm. Perhaps a person who paid attention only to the w v-clause decided that its syntax was inferior inasmuch as dihÄlqon requires an object (not realising that the object is easily supplied from the main clause).g. so to speak) as Ap. kai after wv could be the inverted-place kai which should have followed a main-clause demonstrative (GP 2 295±6). as is shown by the survey of the material in E. A number of translators modify a: `und denke dass ich uÈber alles Andere was sich weiter hieran anschliessen muss gerade so gesprochen habe wie ich es eben hinsichtlich des Turnens that' (Susemihl. this usage should be contrasted with the elliptic construction of e. this suits the context. . If. this squares precisely with the idiomatic mhdamwÄ v allwv poieiÄ n in the next sentence. the pronoun could point forwards to a subject about which no statement has yet been made (`what do you say about this?'). This gives excellent sense. talla paÂnta . If this is correct. outwv . Une formule platonicienne de reÂcurrence (Paris 1929). however. . 35b10±c1 ou deÁ di kaioÂn moi dokeiÄ einai deiÄ sqai touÄ dikastouÄ ou deÁ deo menon apojeuÂgein. I print w v instead of a because it is far superior to any interpretation of the transmitted text that I can think of. we read w v instead of a all problems are solved. w v kaiÁ nundhÁ dihÄ lqon indicates in what way Clitophon's report and the remarks following upon it should be taken. Socrates is invited to take the criticism as an incitement to get down to business about justice. . sou deoÂmenov leÂgw: Brandwood (Word Index. 136d4±5 au touÄ w Sw kratev dew - 334 . 7±16. taÁ tou toiv e xhÄv refers then to everything which Clitophon had said following his report of the protreptic. similarly Schleiermacher). s. it cannot. To the best of my knowledge.v. The weak side of this proposal is that an original wv is not very likely to have been corrupted to a.C O M M E N T A R Y: 410e3 mich ausgesprochen habe'.

kaiÁ proÁv touÁv allouv: cf. Cf. 36d9±10.' absorbs `but I have to add').). The in®nitive depends on le gw. dielqeiÄ n. Intr. section i. .. Intr.2 for the echo of the prologue. ye gw: cf. . e7 teÂlov arethÄ v: cf. . e6 a xion . the latter will have nothing to blame in him) but its extension is easy to account for (`I shall always sustain etc. section i.3. note on 406a2. 280b7±d7 as likely as not). as well as R. . Ast.C O M M E N T A R Y: 410e4±410e8 meqa Parmeni dou (sc. 277a3±4. dass du es doch ja nich anders machen moÈ gest'. jh sw: strictly speaking. 335 .2 for the details of this peroration. . twice with touÄ mh ‡ inf.v. Mx. Phdr. 925e1±2). 407c4). e7±8 e mpo dion touÄ . There is no pleonastic use of le gw here (LSJ s. 280b6 eu daimoneiÄ n an kaiÁ eu praÂttein) is the ®nal note of Clit. eudaimoni a is also the central concept of Aristotle's Protr. The only possibility of making the in®nitive depend on deoÂmenov would be by supplying the previous sentence as object to le gw (`und ich sage dies dich bittend. section i. (eu praÂttwmen). 155d5±6 Kudi an . Euthphr. But in my opinion this would have required tauÄta le gw. (b) iii 7) any more than at 409a4 (see on proÁv tauÄ ta). a llwi u potiqe menov. is an adaptation of Pl. cf. 12e2). 234a5 paideuÂsewv kaiÁ jilosoji av e piÁ te lei. Lexicon. o v ei pen e piÁ kalouÄ le gwn paidoÂv. he would in view of these parallels have written eu pra ttein (cf. (but b 52 DuÈ ring. Intr. . the future is justi®ed only for the ®rst part of the sentence (if Socrates will follow up Clitophon's last appeal. . s.. The premise of the protreptic argument in Euthd. . 494a12 proÁ v te lov e lqeiÄ n. eu labeiÄ sqai. for le gw mh ‡ inf.v. R. gene sqai: e mpoÂdiov in Plato is found once perhaps with the simple in®nitive (R. were the work of a forger. paÄ v. cf. cf. . e5 ga r: explains the whole of taÁ meÁ n e painwÄ . not from Iambl.' Cf. Protr. If Clit. le gw with the in®nitive.3. touÄ pantoÂv: a typically Platonic idiom. .3. Chrm. e4±5 i na mhÁ . toÁ e mpodi zon touÄ i e nai (Cra. (278e3 ara ge paÂntev a nqrwpoi bouloÂmeqa eu pra ttein..2. Euthd. it is said to be the e¨ect of Socrates' exhortation combined with elenchos in Ap. 419c3). 3 315b1±3 and ± rather less obvious ± Ep. 205a1±3 (note te lov a3) and Bury's note. above all. not on deo menov (so Bury. cf. . e5±8 mhÁ meÁn . 8 352b3). being a neutral expression of volition. but cf. 832b1±2. Schleiermacher). is quali®ed by sou deoÂmenov: `I tell you to do just that please. geneÂsqai: cf. supplied from d2±3 ti ou dihÄlqev hmiÄ n. taÁ de ti kaiÁ ye gw.. . further Smp. (Lg. Ep. our construction (not in LSJ) is not found elsewhere in the corpus. e8 eu dai mona geneÂsqai: the word eu dai mwn is here used for the ®rst and last time.

of its most protreptic part (cf. 87. 218 n. 70±4. esp. cf. Rose. . Der Protreptikos des Aristoteles (Munich 1966). one of the two ascribed quotations from Protr. AGPh 47 (1965) 53±79. W. b 104±10 is an indivisible block. Hort. 22 want to place Iambl. Jaeger. 272 and Gaiser. 21. J. de Vogel. G. 37 and `Problems in Aristotle's Protrepticus'. DuÈ ring. `Prolegomena to an edition of the Eudemus' in DuÈ ring±Owen. 139±71 at 171. Aristotle and Plato. . 89. . besides b 110 is closely connected with a 2 ˆ b 6 DuÈ ring. Protr.15 des Places ˆ b 93±6 DuÈ ring after this paragraph. 19±33 at 28) robs the Protr. Gigon. fr. Bignone. L'Aristotele perduto e la formazione ®loso®ca di Epicuro (Florence n. . Protreptik. . 2. (1936)). Protr. `Protreptikos'. inspired. E. Hort. 1). V. . Apart from the fact that b 110 is quite a sweeping statement and makes excellent sense as a peroration. 228. Aristotelis qui ferebantur librorum fragmenta (Leipzig 1886). end on a more exalted note. `De exhortationum .d. h . the consensus about its position rests mainly on the following argument. O. ± Flashar's attempt to attribute b 104±10 to the Eudemus (`Platon und Aristoteles im Protreptikos des Jamblichos'. C. Schneeweiss. . DuÈring. a pite on) and the last sentence of Cic. 102 n.. like Cic. 56±75 at 67 and n. 97±8. 220 n. 253±4. b 110 DuÈ ring (h jilosojhte on . Der Protreptikos des Aristoteles (Frankfurt±Mainz 1969) 110). as DuÈ ring points out (261±2) by the Phaedo (64a4±70b4). `The legend of the Platonizing Aristotle' in DuÈring± Owen. but against this speaks a formal similarity of Arist. historia'. S. Eranos 52 (1954). . Aristotle and Plato. i 90. 115 Grilli (si aut . fr. Hartlich. Mansion. so as to make Arist. opera et cura ponenda est ). Protr. . . aut si . Aristoteles (Berlin 1923). `Contemplation and action in Aristotle's ``Protrepticus'' ' in DuÈ ring±Owen. There are two close parallels 336 .7±90.APPENDIX I: THE ENDING OF ARISTOTLE'S PROTREPTICUS The words h jilosojhte on ou n h cai rein eipouÄ si twÄ i zhÄ n a pite on e nteuÄqen (b 110 DuÈ ring) are regarded by a large majority of scholars as actually the last paragraph of the Protrepticus (Hirzel. 61. 248±56 at 252.

kaqa per jasiÁ n oi taÁ v teletaÁ v le gontev. aliquid dixisse videantur. verumque sit illud quod est apud Aristotelem simili nos adfectos esse supplicio atque eos qui quondam. crudelitate excogitata necebantur. oi prwÄ ton euquÁv juÂsei sune stamen. w sper a n e piÁ timwri ai pa ntev.THE ENDING OF ARISTOTLE'S PROTREPTICUS between this block and the end of Cicero's Hortensius: (1) b 106 . qui nos ob aliqua scelera . sic nostros animos cum corporibus copulatos ut vivos cum mortuis esse coniunctos. fr. ®t ut interdum veteres illi vates sive in sacris initiisque tradendae divinae mentis interpretes. fr. 823 Gigon ex quibus humanae vitae erroribus. cf. . in his studiis nobis omnis opera et cura ponenda est. 115 Grilli (in ®ne dialogi Hortensii. . . fr. poenarum luendarum causa natos esse dixerunt. . ut aliquando terminetur oratio. cum in praedonum Etruscorum manus incidissent. sermonem ®niens). b 107 pa nu gaÁ r h su zeuxiv toiouÂtwi tiniÁ e oike proÁ v toÁ swÄ ma thÄv yuchÄ v´ w sper gaÁ r touÁv e n thÄ i Turrhni ai jasiÁ basani zein polla kiv touÁv a liskome nouv prosdesmeu ontav kat ' a ntikruÁ toiÄ v zwÄ si nekrouÁv a ntiproswÂpouv e kaston proÁv e kaston me rov prosarmo ttontav. . 337 . . quorum corpora viva eum mortuis adversa adversis accommodata quam artissime colligabantur. Arist. aut si ex hac in aliam haud paulo meliorem domum sine mora demigrare. . (With b 107. touÄto gaÁ r qei wv oi a rcaioÂteroi le gousi toÁ ja nai didoÂnai thÁn yuchÁ n timwri an kaiÁ zhÄn h maÄ v e piÁ kola sei megaÂlwn tinwÄ n a marthma twn. 112 Grilli (in extremis partibus Hortensii dialogi). Arist. fr. si aut exstingui tranquille volumus cum in his artibus vixerimus. outwv e oiken h yuchÁ diateta sqai kaiÁ proskekollhÄsqai paÄsi toiÄ v ai sqhtikoiÄ v touÄ sw matov me lesin. (2) b 110 as quoted above. 825 Gigon: Quapropter.

historia'. Rabinowitz. is unwarranted and his method objectionable.97. when viewed in the light of Clitophon's report of Socrates' peroration. Hortensius] philosophandum non esse.APPENDIX I cf. 20. then. quem ad nepotulum meum . but even to answer it in the negative. quem ad exemplum protreptici scripsit (Hist. 17 Grilli and Grilli's note)) means anything but `which he modelled on the protreptic pattern'. fr. however. . 2. `Protreptikos'. libellum. .) Now it can be shown that the Hortensius makes an extensive use of material preserved in Iamblichus' excerpts which are assumed to derive from Aristotle's Protrepticus (for the material. Gall. a passage from the Protrepticus which we know was imitated at the end of the Hortensius may be plausibly located near the end of the Protrepticus itself. 23±7. I may point out (to avoid misunderstanding) that like Hirzel. 81 and n. Ep. we have little to go on here. a long way o¨ from Aristotle's 338 . fr. Hartlich. cf. 8 MuÈ ller (cf. instar protreptici luseram. 54 Grilli were both of them hypothetical syllogisms) which he then proceeds to destroy. which are. Tullius in Hortensio. (Rabinowitz's scepticism. Aristotle's Protrepticus. This supposition.1 ˆ 2. nihilominus philosophari videbatur. It is certain that Cicero imitates a passage ascribed to the Protrepticus: a 2 ˆ b 6 DuÈ ring jilosojeiÄ n le getai kaiÁ toÁ zhteiÄ n autoÁ touÄto ei te crhÁ jilosojeiÄ n ei te mh ± fr.21±2 Hohl ˆ Hort. 38±40. One may point to the eschatological speculations at the end of the Epinomis (992b2±c3). the Hortensius was `modelled on a protreptic pattern' (ad exemplum protreptici ) and if Aristotle's Protrepticus was one of its main examples. If. Aristotle's Protrepticus. fr. also Hort. fr. `De exhortationum . . I do not believe that Trebellius' statement M. is to philosophise. 113 Grilli: adpendicem animi esse corpus nihilque esse in eo magnum. cf. Aug.) It seems evident to me that Cicero tried to improve upon Aristotle: not only raising the question (zhteiÄ n). Aus. 22 praef. 240. automatically raises the question whether Aristotle himself was following a more or less ®xed pattern of the protreptic genre in the peroration of his Protrepticus. DuÈring 152). inasmuch as he sets up an imaginary case (b 6 and Hort. Unfortunately. . 54 Grilli: cum diceret [sc.

The exhortation which Virtue directs to Heracles (Prodicus apud X. 10).7.1. FestugieÁre. 108±15). cf.. neither that part of the Apolog y nor the Apolog y as a whole belongs to the protreptic genre proper. we must hesitantly conclude that there is no evidence for a more or less obligatory reference to death as the ®nal part of a ®xed pattern for protreptic speeches ± not. the properly paraenetic part of the Demonicea ([Isoc. besides.. which is alluded to. 38a5±6 o deÁ a nexe tastov bi ov ou biwtoÁv a nqrw pwi.1. Phd. `Epinomis'. h jilosojhte on oun h cai rein ei pouÄsi twÄ i zhÄ n a pite on e nteuÄ qen ± Clit. Intr. but a work or works unknown. R.THE ENDING OF ARISTOTLE'S PROTREPTICUS drastic dilemma (and inspired by Grg.1. On account of this unconvincing degree of similarity and the general lack of agreement of Clit.27±33) ends with a promise of everlasting fame after death. which says even less. Einarson. section ii. Again. and Protr. The relationship of these works is treated in Intr. (SSR v a 105) deiÄ n ktaÄsqai nouÄ n h broÂcon. o stiv yuchÄ i mhÁ e pi statai crhÄ sqai. One could think of Antisthenes fr. passage and the various parallels in Plato discussed in Intr. though it is implicitly protreptic.]) 1. Protreptiques. there is rather more resemblance between the Clit. with all the provisos necessary owing to the scarcity of our material.13±43) ends with a reference to kalwÄ v a poqaneiÄ n. 339 .36±7).. 67 C. b 104±10 has a rather close parallel in the famous topos of human misery which is placed at the beginning of Epin. a theme which Isocrates places right in the middle of his exhortation to Nicocles (2. it is safer to assume that it is not Aristotle's Protr.. (973b7±974c7.1(4).. section ii. Therefore. 280.7.5 n. 168). Mem. cf. We are therefore compelled to ask whether it is not precisely the peroration of Aristotle's Protrepticus which is parodied in the Clitophon. it may be su½cient here to note that the wording is not particularly close (Protr. quoted by Chrysippus in PeriÁ touÄ protre pesqai (SVF iii 167. Ap. that is. which is found near the end of the a ntiti mhsiv ± however. in the fourth century. By far the closest parallel to b 110 is Pl. section ii. touÂtwi toÁ a gein h suci an thÄ i yuchÄ i kaiÁ mhÁ zhÄ n kreiÄ tton h zhÄ n pra ttonti kaq' au toÂn). 2.

s. suppl. there is no reason to believe that A2 checked the readings against those of another copy. an indirect descendant of A in the Clitophon). 91±9. 409b8 e kei nouv (e keiÄ touÁv DF recte).g. also my paper `Supplementary notes'. it is an independent witness of relatively minor importance. Tradition of Republic. No papyri of the Clitophon have come to light as yet. d3 om. F and its descendants share numerous separative errors. Boter. suppl. of F for Its readings in the Clitophon look like simple conjectures. Tradition of Republic. the theory should almost certainly be rejected for the Clitophon. Whatever its merits for other dialogues. a hand roughly contemporary to A/A 2 . cf. Platonis Opera. Mx. Later hands are. 408b4 dhÁ] deÁ (with Stob. 35±7. s. his theory has recently been revived by B. 7. The hand which is traditionally called A 2 is in fact that of the copyist.Mi. gr.). D (Marc. For a description of A. Schanz. D (and its single descendant) has a separative error at 408a1 (cf. d7 touÄ t '] toÂt '. 36±9). which enters readings it found in W (Vind. For the Clitophon. not a copy. A and its descendants have separative errors at 407a8 umeiÄ v (umnoiv FD. s.APPENDIX II: NOTE ON THE TEXT There are three primary MSS for the Clitophon: A (Paris. D 2 is a late hand. gr. Comm. Hippias Minor (Stuttgart 1996). Here and there in the Clitophon we ®nd A3 . Hippias Maior. Cf. 408a1 gaÁ r a n dhÁ (gaÁ r dhÁ F Stob. As in the Republic. cf. above).).: 406a13 om. ix). while and after he added the accents and other diacritics. 407b6 ei deÁ ] ei te. xiii or xiv) ± they are also the only primary MSS for the 189 has readings that were entered into F by a later 340 . ix pp. gr. A4 and A5 are later hands of no authority. 189 was a gemellus. 80±91. Vancamp (Platon. not to be found in the Clitophon. recte: gaÁ r a n D). e. touÁv a di kouv. Boter. xii) and F (Vind. who made changes and added readings before. 39. where Marc.. cf. 1807. I think. x±xv thought that Marc. Io and Clit. correct but for the accent.. gr. 185. a n. ad loc. thought by some to be Arethas' hand.

cf. Among the secondary This means that the Clitophon was copied into minuscules no less than three times. Tradition of Republic. 44). xv (?). or D and A. Martinelli Tempesta. 99±110. Yet they also exhibit readings that are not found in other Plato MSS but which recur in the MSS of Themistius. La tradizione testuale del Liside di Platone (Florence 1997). Codices Vaticani graeci 2162±2254 (Rome 1985). as can be seen from a number of errors which they have in common against all other MSS. It is important to bear in mind that F is a typical representative of the cheap Plato omnibus as found in later antiquity.6. it has a host of errors. The various later hands will here be called F2 . 410e2 F omits a line. gr. xiv for Pa (106±7). For a description and a characterisation of F. note on 407b1 kai ). who arrives at s. belong together stemmatically is a purely theoretical Whether or not D and F. the hand which is most active in the Clitophon derives most of its readings from Flor. 128±9 claims that it belongs to the second half of s. xiv (earlier authorities assigned it to s. to 408c3 protroph n) and Va (Vat. S. S. a fourteenth-century descendant of A. cf. the D reading (AN for DH ) at 408a1 su½ces to prove this for D. D and F go back to separate copies written in majuscules. contains only the ®rst half. and many untrustworthy variants in word order and particles. including Marc. 341 . and the descendants. have ontwv. 1809. xiv).1 These MSS are descendants of A. which is supplied by F 2 ± in this line outwv looks very much like o ntwv (I am almost certain that this is actually what F 2 wrote). and it is generally accepted for A and F.NOTE ON THE TEXT hand: 407d6 o v a n hi ] w v a n h F 1 : w saneiÁ F2 and the other members of the group. 2196. there are two which merit closer consideration: Pa (Paris. gr. Boter. and they belong closely together. It is certain that A. cf. s. xiv±xv. Martinelli Tempesta. including Marc. Although F has many uniquely true readings. 85. Lilla. Therefore conjectures based on misreading of minuscules should be rejected (cf. who gives an extract of the protreptic speech as an example of how philosophical oratory could be bene®cial 1 For the dates of these two MSS. For Va. 189. s. 189.

so the remaining primary witnesses AL give the best idea of what Themistius' text of the Clitophon looked like. Although some copyists change these things at random. MSS: 407b7 ge ti Pa: ge all MSS of Them. Therefore.3±135. PaVa (and F) agree with the two most trustworthy Themistius MSS in manifest error at 407e7 a rxantov FPaVa Them. 26. SY have obviously been corrected from a Plato MS. As to the apparatus as a whole. closely related to a MS that was available to Themistius.. is manifestly wrong.. AL. the degree in similarity can sometimes be striking: 407e11 mh te a kouÂein mhq' oraÄ n ADF Stob. The myth that ancient prose texts did not have any diacritics or punctuation signs. and supported by Stob. tried to report them exhaustively). cf. I have reported ADF in full. Cf. wherever their readings are not identical to those of A. 114. In one place PaVa agree with all Themistius MSS against all other Plato MSS: 407c3 ei nai om. omnes (very possibly right). which one still ®nds stated as actual fact. I have paid more attention to matters like diacritics. four cases in twenty lines of text are too much of a coincidence. or a further corruption of them... Though each separate case of agreement of Pa and Themistius might plausibly be put down as a coincidence. 320d±321c. and I have consequently reported them in the critical apparatus. AL: kakouÁ v all other MSS of Plato (including Va) and Themistius. once only Pa agrees (in part) with all Them.1. a full stop before 408b1 e stin a ra is shared by F and Them.APPENDIX II (Or. however. even A 4=5 D2 F 2 . spelling and punctuation than is customary (I have not.: g' e ti all other Plato MSS. H. even though I have not adopted any of their readings in the text. there is always a possibility that a diacritic or punctuation sign in a primary Byzantine MS goes back 2 Of the Themistius MSS. ii 134.6 Downey±Norman. section ii. PaVa Them. who quotes 407e8±408b1 toÁ n bi on in Book 3 and 408b1 e stin a ra±b5 in Book 4. Pa has more of such readings than Va. including Va. I conclude that a common ancestor of PaVa was contaminated with readings from a lost primary source. Intr. However slight. È berlieferung des Themistius'. they deserve consideration. `U 342 . Schenkl.2. Pa alone agrees in manifest error with these at 407c3 kakwÄ v Pa Them.2)2.

406a5 10 407a8 b4 5 7 c3 4 6 e3 7 ostiv Bt By So Sl: o stiv hh ni He. but it is super¯uous to report that). n e jelkustikoÂn and minor matters of accentuation and orthography are not listed. But this does not apply to PaVa and the indirect witnesses except Stob. On the other hand. while Plutarch's text cannot be inferred from the distorted ways in which the sentence is twice quoted by him. I do record at d8 that Hippolytus has loÂgov e reiÄ for o loÂgov ai reiÄ on the o¨-chance that this is an ancient variant quoted by Hippolytus himself. h ghsqe By Sl: h geiÄ sqe rell. ontev Sl: onte rell. kai toi Bt So Sl: kai toi rell. if a witness is not cited in the apparatus it agrees with the text. Schanz (Sz). In general the apparatus is negative. that is.NOTE ON THE TEXT to an ancient MS tradition. pwÄ v ou rell: omwv ou He. Thus at 407c7 Themistius is quoted for kaiÁ ou : this means that ADF read a ll ' ou (so do PaVa. 343 . If an indirect witness other than Stob. umnoiÄ v Sl: umneiv rell. though evidently an inferior one. ostiv * * * Sz. Bury (By) and Souilhe (So) Deviations in punctuation. is not named at a relevant place where variant readings are recorded. In matters of elision. but a critical apparatus is not there to give an extra supply of those). I have followed A. it is not useful to know that at 407d4 the MSS of Hippolytus have pa lin d ' au twÄÎ a ma te le gein for pa lin d ' au tolmaÄ te le gein (it is of course an instructive example to illustrate the misreading of majuscule script. paradwÂsete * * Sl: paradwÂsete rell. Thus. elision. crasis. n e jelkustiko n etc. this means that the witness for one reason or another can shed no light on the problem. g' e ti rell: ge Sz. ou te Bt Sl: a meleiÄ te kaiÁ ou te rell. Burnet (Bt). sou Sl (ADF): souÄ rell. touÄ meÁ n rell: touÄ meÁ n gaÁ r So. Concordance of this text (Sl) with those of Hermann (He). I have not recorded obvious errors made in the transmission of secondary sources.

pa ntav rell: pa ntov By. pwÄ v pote nun Sl: pwÄ v poteÁ He. rell: habet He. opoi rell: opou So. e panh iein rell: e panhÄ a Sz. d ' Sl: dhÁ rell. oi omai Sl: oi mai rell. tautoÁn gigne sqw rell: secl. nundhÁ Sz Bt So Sl: nuÄ n dhÁ He By. hovi pr. e panerwtaÄ n rell: e perwtaÄ n Sz. kaiÁ secl. h ntinaouÄn rell: hntinouÄ n Sz (cum D). 344 .APPENDIX II 408c2 8 d2 4 6 7 409a7 c3 6 e9 410a5 6 b2 4 7 c7 d4±5 e2 3 protreptikwta touv te Bt So Sl: pr. omoÂnoian [kaiÁ ] e pisthÂmhn rell: o moÂnoian kaiÁ hdikaiosu nhni e pisth mhn He. poreu omai Sl: poreu somai rell. pwÄ v poteÁ (vel pote) nuÄ n rell. on Sl: e ni rell. te gaÁ r He By. de mou Sl: d ' e mouÄ rell. te Sz. e kei nouv rell: e kei noiv Sz. w v kaiÁ Sl: a kaiÁ rell. poÂt ' e stin Sl: pot ' e sti(n) rell (cum A). He. touÂv ti By So Sl: touÁv tiÁ rell. e ka sth He Sz By Sl: e ka sth h Bt So.

J. Jowett±Campbell. de Strycker±Slings. Works concerning the Clitophon Ast. `Kleitophon wider Sokrates'. J. `Saggio sul Clitofonte'. Plato. M. G.) 1929 311±27 Carlini. see B: 140±7 Ge¨cken. Not listed are well-known works of reference and commentaries on Greek authors (but see Dodds. 465± 74 (with Ficinus' translation) In Platonem a se editum commentaria critica (Berolini 1823). Platonis opera iv. Platonis dialogi Graece et Latine ii 3 (Berolini 1817). recensuit. Oxonii 1902 Bury.A. A. 13±26 Grube. G. I have adhered to these principles less strictly in section A. RFIC 1 (1873) 457±80 Blits. . . in: Cooper±Hutchinson (see B). ix (Lipsiae 1827). J. F. G. R. `Alcuni dialoghi pseudoplatonici e l'Accademia di Arcesilao'. G. A. H. A. W. A. M. . KuÈ hner. `Socratic teaching and justice: Plato's Clitophon'. London±Cambridge (Mass. or more than once in the commentary. Platonis quae exstant opera . i 472±4 (critical apparatus). AGPh 26 (1913) 449± 78 Burnet. P. ASNP 31 (1962) 33±63 Ficinus.BIBLIOGRAPHY Only those works are listed to which reference has been made in more than one section. see B: iii. Tucker in section B). `Das RaÈ tsel des ``Kleitophon'' '. `Clitophon'. H. 354±65 Bekker. `The Cleitophon of Plato'. ii 394±5 (scholia) Bertini. J. see B: 229±32 Heidel. in linguam Latinam convertit . CPh 26 (1931) 302±8 Hartlich. K. I. F. 966±70 Grote. . F. see B: 46±8 345 . with an English Translation. Hermes 68 (1933) 429±39 Gonzalez. Interpretation 13 (1985) 321±34 BruÈ nnecke. M. see Bekker Gaiser.

F. Ten Forgotten Socratic Dialogues. 47± 74 Neumann. Steinhart vii 1 (Leipzig 1859). The Roots of Political Philosophy. D.. U Einleitungen begleitet von K. Naples 1935 Modugno. 330±1 (scholia) HGPh see B: Guthrie. vi (Lipsiae 1853). L. `De authenticiteit van den Kleitophoon'. Platonis dialogi iii (Lipsiae 1851). F. with Interpretive Studies (Ithaca±London 1987). Greifswald 1881 la Magna. in: T. F.). H. Platone. Platons Werke ii 3 (Berlin 1809).S. Platone. Amsterdam 1981 (referred to as `the ®rst version of this book') 346 . Platon's saÈ mmtliche Werke.BIBLIOGRAPHY Hermann.u. A Commentary on the Platonic Clitophon. Translated. `The sophistry of Plato's Protagoras and Cleitophon'. 459±64. Sophia 35 (1967) 46±55 Orwin. con introduzione e commento. G. `The riddle of the Cleitophon'. 6 (1934±5) 161±89 Kunert. G. E. Ch. . Clitofonte. . 3±20 Roochnik. 453±64 (translation). edidit M. B. Dialoghi.k. L. Pangle (ed. R. Platonis opera . v (Bari 1956) 2±10 Schanz. H. M. C. Platone. H. Gymnasiums (Znaim 1909). . L. R. Exhortation]'. 534 (notes) Schneider. Philol. reprinted as `On the Cleitophon'. CanJournPolSc 15 (1982) 741±55. ix (Lipsiae 1885). 592±5 Slings. 102±3 (critical appendix) Schleiermacher. M. Jahresbericht des k. S. op. le Opere tradotte e dichiarate ad uso di ogni persona colta xix (Aquila 1930).). v 387±9 Kesters. in: T. 129±49 È bersetzt von H. Pangle (ed. R. `Der pseudoplatonische Kleitophon'. cit. . Platonis opera . mit MuÈ ller. 111±16 Pavlu. see B: 96±101 Sartori. C. C. Graece et Latine ii (Parisiis 1862). J. `The case against Socrates: Plato's Clitophon'. Stud. Quae inter Clitophontem dialogum et Platonis Rempublicam intercedat necessitudo.. AncPhil 4 (1984) 132±45 Rutherford. 90±5. 117±31 (references are to page numbers of CanJournPolSc) `Cleitophon [or.

J. 163±90 Stallbaum. Mnem. Epiktet. 35 (1982) 143±6 Water®eld. Dialoghi v (Bari 1915). Platonis dialogi. Mnem. 495 (notes) Wolf. Leipzig 1835±8 Billerbeck. MuÈ ller Stumpo. `Annotationes criticae ad Themistii orationes'. recognovit G. General Alline.. A. 469±75 (translation). . J. `Sull' autenticitaÁ del Clitofonte'. textum ad ®dem Codicum . Jahresbericht FriedrichYxem. E. [E. C. The Textual Tradition of Plato's Republic. Leiden 1978 Bos. 1 (1920) 408±19 Susemihl. Verzameld werk v (Antwerpen±Baarn 19782 ). Plato. F. K. Cambridge 1990 Caizzi. 1±35 Zuretti. Republic (Oxford 1993). vaderschap en datering van de Alcibiades maior. Leeds 1976 The Chronology of Plato's Dialogues. . 40 (1987) 35±44 SouilheÂ. 496±9 (critical apparatus) Stefanini. A Word Index to Plato. see H. Mnem. Culemborg 1970 Boter. Interpretatie. L. 507±29 Verdenius. Lexicon Platonicum sive vocum Platonicarum index. G. 227±34 (text). W. O. Giorn. ital. Plato. L. Histoire du texte de Platon. `Cleitophon'. Paris 1915 Ast. G. viii (Lipsiae 1825). Leiden 1989 Brandwood. O. Platon. i 11 (1862) 222±66. see B: 148±51 de Win.S. see B: 289±93 È ber Platon's Kleitophon'. in: id. F. 462±8 Wichmann. crit. Vom Kynismus. Werke in 40 BaÈndchen v 3±6 (Stuttgart 1865). J.] `U Wilhelms-Gymnasium (Berlin 1846). `Notes on the pseudo-Platonic Clitopho'. B. 296±9 Steinhart.BIBLIOGRAPHY `Supplementary notes on manuscripts of the Clitophon'. C. C. Platone. Platon. X. xii (Lipsiae 1825). . . G. M. ®los. Decleva Antisthenis Fragmenta. F. . H. 3±21 B. F. 394±434 347 . R. Milan 1966 Cobet. see B: i 203±10. Oeuvres compleÁtes xiii 2 (Paris 1930).

A. Berlin±New York 1987 FestugieÁ re. K. Basel 1956 Goldschmidt. Stud. M. Stuttgart 1959 Gigon. Gorgias. `Die Architektonik im Aufbau von Xenophons Memorabilien'. Th. I. TAPhA 67 (1936) 261±85 Erbse. Greek Prose Style. Griechische Denker.) Aristotle and Plato in the MidFourth Century. Kommentar zum ersten Buch von Xenophons Memorabilien. Plato. O. A History of Greek Philosophy. S. R. Sophistik und Rhetorik. i±ii Berlin 1964 3 . 308±40 (references are to page numbers of AusgewaÈhlte Schriften) È bungsstuÈ cke Erler. D. C. GoÈteborg 1961 DuÈ ring. GoÈ teborg 1960 E¨e. B. La ReÂveÂlation d'HermeÁs TrismeÂgiste. P. L. Plato and the Other Companions of Sokrates. iii 19753 Gaiser. P. Aischines von Sphettos. Oxford 1952 Dittmar. Der Sinn der Aporien in den Dialogen Platons. Cambridge 1962± 78 Hartlich. M. Basel 1953 Kommentar zum zweiten Buch von Xenophons Memorabilien.±D. J. Indianapolis 1997 Denniston. Papers of the Symposium Aristotelicum Held at Oxford in August. Leipzig 1896±1909 van Groningen. 11 (1889) 209±336 Heidel. Hutchinson Plato. Owen (eds. A. Oxford 1959 DuÈ ring. H. B. Hermes 89 (1961) 257±87 \ AusgewaÈhlte Schriften zur Klassischen Philologie (Berlin±New York 1979). `De exhortationum a Graecis Romanisque scriptarum historia et indole'. `Aristotle's Protrepticus and the structure of the Epinomis'. K. Berlin 1912 Dodds. Platon. W. Complete Works. Leipzig±Berlin 1912 Gomperz. London 1865 Guthrie. Aristotle's Protrepticus. Paris 1973 FriedlaÈnder. G. Pseudo-Platonica. Paris 1949± 54 Les trois `protreptiques' de Platon. E. U zur Anleitung im philosophischen Denken. I. E.BIBLIOGRAPHY Cooper. Protreptik und ParaÈ nese bei Platon.±G. Baltimore 1896 348 . Leipz. `EKDOSIS '. Paris 1947 Gomperz. B. W. A. Les dialogues de Platon. H. J. 16 (1963) 1±17 Grote. V.-J. Hermes 99 (1971) 198±208 Einarson. H. 1957. `Platons ``Charmides'' und der ``Alkibiades'' des Aischines von Sphettos'. Mnem.

Louvain±Paris 1959 KeÂrygmes de Socrate. AntistheÁne. Berlin 1893 `Der lo gov Swkratiko v'. Louvain±Paris 1965 Krauss. Sokrates. H.BIBLIOGRAPHY È ber den Protreptikos des Aristoteles'. Xenophontis de Socrate commentarii. TuÈ bingen 1913 Meyer. R. Essays on the Philosophy of Socrates (Oxford 1992).). CQ 31 (1981) 305±20. Ithaca 1966 O'Brien. H. W. Paris 1929 Rabinowitz. Cambridge 1996 Kesters. `U (1876) 61±100 Der Dialog. Aeschinis Socratici reliquiae. Berkeley±Los Angeles 1957 Radermacher. W. Munich 1959 2 Lutosøawski. Chapel Hill 1967 des Places. Sophrosyne. H. AGPh 9 (1896) 50±66 Jowett. P. Louvain 1935 Plaidoyer d'un Socratique contre le PheÁdre de Platon. L. The Socratic Paradoxes and the Greek Mind. TuÈ bingen 1962 Monteil. i Berlin±Leipzig 1936 2 . Leipzig 1911 KuÈ hner. Artium Scriptores (Reste der voraristotelischen Rhetorik). W. Gotha 1841 (not the Teubner comm. London 1897 Magnien. Th. Der echte und der Xenophontische Sokrates i. H. Benson (ed. H. in German) Kuhn. Le futur grec. De la dialectique. J. Platons Apologie. Sokrates. Leipzig 1895 Jaeger. Die Kurzdialoge der Appendix Platonica. M. K. Aristotle's Protrepticus and the Sources of its Reconstruction i (all published). 35±52 (references are to pages of CQ ) Plato and the Socratic Dialogue. B. Sein Werk und seine geschichtliche Stellung. C. R. Hermes 10 Hirzel. E. Vienna 1951 349 . La phrase relative en grec ancien. Paideia. The Origin and Growth of Plato's Logic. Paris 1963 MuÈ ller. reprinted in H. V. Oxford 1894 Kahn. C. Self-Knowledge and Self-Restraint in Greek Literature. `Did Plato write Socratic dialogues?'. H. W. ii±iii (English edition) Oxford 1944±5 JoeÈl. Campbell The Republic of Plato.±L. Munich 1975 North. G. Paris 1913 Maier. Etudes sur quelques particules de liaison chez Platon.

iv'. 8 (1955) 265±89 [Review of GP 2 ]. Mnem. Helsinki 1967 Tucker. Plato. J. M. London 1929 3 Thesle¨. R. 25 (1972) 42±55 350 . C. `Some remarks on Aeschines' Miltiades'. R. Mnem. 9 (1956) 248±52 `Notes on Menander's Epitrepontes'. E. S. Studies in the Styles of Plato. H. Chronolog y].BIBLIOGRAPHY Reinhard. 29 (1976) 42±51 [Review of MuÈ ller. J. London 1986 de Strycker. R. Ironist and Moral Philosopher. Kritisch-historische Syntax des griechischen Verbums der klassischen Zeit.±S. Berlin 1920 Richards. London 1995 È berlieferung der Reden des Schenkl. Padova 19492 Stokes. J. `Notes on Plato's Phaedrus'. Leiden 1994 Taylor. London 1911 Riddell. London 1900 Verdenius. Untersuchungen uÈber Plato. L. the Man and his Work. E.. H. Socrates. Mnem. C. Kent Plato's Use of Fallacy. Platone. Platonica. W. 118±252 Ritter. Mnem. 43 (1990) 341±63 [Review of Brandwood. B. Slings Plato's Apology of Socrates. Mnem. M. 49 (1996) 403±25 Sprague. The Proem to the Ideal Commonwealth of Plato. Cambridge 1991 de Vries. G. `Die handschriftliche U Themistius'. in J. (eds. Mnem. 27 (1974) 17±43 Vlastos. Mnem. R. Ten Essays in Platonic Interpretation. T. A Literary and Philosophical Study with a Running Commentary. L. `A digest of Platonic idioms' in: id. 31 (1978) 211±14 `Critical notes on Plato's Politeia. WS 21 (1899) 80±115 Slings. 173±92 `Critical notes on Plato's Politeia. `Notes on some passages in the Euthydemus'. G. Kurzdialoge]. 47 (1994) 539±41 `Protreptic in ancient theories of philosophical literature'. H. J. Stuttgart 1888 Rutherford.). J. iii '. London 1962 Stahl. Greek Literary Theory after Aristotle: A Collection of Papers in Honour of D. Mnem. Die Anakoluthe bei Platon. The Apology of Plato (Oxford 1876). Heidelberg 1907 Stefanini. Abbenes et al. A. Schenkeveld (Amsterdam 1995). G. Mnem. The Art of Plato. ZPE 16 (1975) 301±8 `Plato. M. Cratylus 417c'. G. Plato's Socratic Conversations.

J. Amsterdam 1994 Wichmann. Darmstadt 1966 Wilamowitz-Moellendorf. E. Platos Apologie. von Platon. Berlin 1929 351 . iv.BIBLIOGRAPHY Wackernagel. Conditions and Conditionals: An Investigation of Ancient Greek. ii 1962 3 Wolf. O. Basel 1926 Wakker. E. Platon. C. G.1 Frankfurt (Main) 1968 Wol¨. Ideelle Gesamtdarstellung und Studienwerk. i Berlin 1959 5 . Vorlesungen uÈber Syntax. Griechisches Rechtsdenken. U.

123. 155. 132 monumental 19±20 pedimental structure 33±4 protreptic 62. 169±70. 93. 44. 77±89. philosophy as art of 96 paradox 15. 160. 120. 88. 148. 305 apostrophe 40. 116. 47±8. 136. 301 protreptic 59±60. 136. 175±8. 155 ethical 60±1. 293 dialogue 62. 132. 198±204 politeness 54. 304±5 development of 90±3. 151±3. 282. 91. 163±4 short 18±34. 179. 124±7 politician 4. 227±33 fallacy 158±60. 144 implicit 61±2. 28 duologue 26±7 epilogue 49 exordium 40±1 framework 29±32. 229. 221±2. 132±3. 47. 156±63. 314. 50±1. 124. literary 1. 313. 196±7. 142. 4. 23. 140±1. 305 rhetoric 9 conditions 107 352 . 150±3. 155. 220±2 eulogy 65±6. 324 transitions 30±2 diatribe 60±1 edition 12. 88. 140±1 indirect 28. 284±6. 144±6 of individuals 45 philosophical 60±1. 53±4 concentric reasoning 20. 150±3. 140±1. 136±7. 23±4. 324 direct 17. 148. 51±2. 162. 116. 327 explicit 61±2. 180±1 curtailed report 313. 142. 148. 129±31. 133±4 tripartition of 85±6. 210 of historical Socrates 46. 150±3. 139±41. 50 authenticity 2±3. 67±76. 142± 3. 57. 329 paraenesis 62±3. 165±7. 86±7. 271. 46±7. 88 parody 3. 153. 263. 32±3. 164±79. 154±5 discourse 35±9 ergocentric 2 identi®cation 27 levels of discourse 17. 71±4. 41±3. 125. 84. 67. 35±58. 139±40. 130. circular 51±3. 39± 58 interpretation circularity 15. 171±2. 176±7 cento 82. 207. 37±9. 40±3. 154±6 irony 15. 160±3. 34. 315. 322 argument 308. 90±2 rhetorical 76 situations 75±6 among sophists 63±6. 108±9 accusatory 80±1. 142±3. 129±31 regress. 210±11 double purpose of 134±5 in later Plato 136±7. 177±8 epyllion 20 eristic 148±9 genre 67±8 life. 90±2. 122. 71±4. 222±3. 153±4. 43. 296 dialogue aporia 85. 92. 164. 129±35. 319 background 28 characterisation 26±8. 178±9 stages of 129. 140 and maieutikh 132±5 results of 129±30. 284±6 forms 151±3. 216 education 282 elenchos 18. 289.INDEXES 1 LITERARY AND PHILOSOPHICAL INDEX analogy 51. 149±54. 107 theories of 67. 122±4. 78. 231 character. 177±9. 16. 210±11.

309±10 imperfect 313. 228 tetralogies 10. 291. 122±3. 280±1. 307±8. o loÂgov 287±8 alhqwÄv 317±18 alla 268±9 a llaÁ mhÂn 268±9 a llov 295 a maqi a 17. 326 article 284. 305±6. 268. 273 juxtaposition 266 Meyer's Law 98 middle 300 optative 264±5. 289. 315 dual 267 future 278. 313. 295. 304. 310 repetition 316 second person singular 97±8 subjunctive 289 topic 263. 331 pleonasm 331. 129±30 a mousi a 283 353 . 220 2 GRAMMATICAL INDEX ABA structure 302 accusative absolute 302 double 314. 327 anacoluthon 278. 301. 289 colloquialism 289. 311±12. 334 plural 301 `pop' 270 postponement 287. 284 contamination 288±9 dative 292. 273. 330 aorist 185. 306±7 oratio recta 53±4. 288. 322. 223. 306±8 parenthesis 270±1. 293±4. 288 clausula 225 climax 288. 321 imperative 50. 322 periphrastic construction 323. 324. 326. 99. 14 peroration 13 values. 149 stylometry 133. 293. 283. 138±9. 324 indicative 330±1 irrealis 331±2 verbal adjectives 110±11 vocative 270±1 wishes. scale of 100±2 and pseudo±values 105 virtue 151±3 political 108 teachability of 106±10. 306. 299. 327 apposition 310. 22±3. 325±6 hendiadys 305 hiatus 226. 224±6. 312 asyndeton 290. 299 relative 287. 290±1 genitive 298. 288. 297 congruence 267. 99. unful®lled 331±2 3 INDEX OF GREEK WORDS agnoia 286 ai reiÄ . 279. 295±6 `push' 270.3 INDEX OF GREEK WORDS title 263 `trias paedagogica' 107±10 epideictic 64±6 ®gured speech 85 judicial 9±10. 313 participle 78 passive 54 perfect 282. 289±90. 324±5 Aktionsart 80. 319 oratio obliqua 53±4. 320. 322±4. 323 praedicative 295 prolepsis 277±9. 325. 332 causative 281 chiasmus 280.

310 e ti 281 eu praÂttw 160 eu dai mwn 335 eu ri skw 280 e wv 223 hmeiÄ v 263±4 hsuci a 293 qerapeuÂw 281 i kanwÄv 281 kaqa per 225. 296. 155±6. 320±2. de 309 kaiÁ . 142±3 e painov 65 e pane rcomai 310 e panerwtaÂw 303 e peita 304 e pexe rcomai 302 e piqumhth v 298 e pime leia 68±9. 275±6. 305. 320 mhÂn 268±9 mhcanh 89. 168. 95. 269 diatribh 264 di dagma 51. 294 areth 167±70. 180±209. 162 duoiÄ n qa teron 326 ei dwlon 157 ei per 287 e kastov 312 e kmeleta w 280±1 e mo n. 309 didakto n 106. 274. 325. dh 292 kaiÁ dhÁ kai 292 kai toi 283 kata 293 kathgore w 83 katameletaÂw 326±7 kerdale ov 182 kolasthÂrion 78 komyoÂv 313 krei ttonev 153 maqhthÂv 298 maieutikh 132±5. 91. 272. 171± 5. 286. taÁ 130 mele th 107±10 me n 265±6. 139. 181. 313. 122. 56. 305 mousikh 282 muqologe w 157±8 nouqete w 85. .3 INDEX OF GREEK WORDS anagkaÂzw 320 anaÂmnhsiv 138. 319 e rasthÂv 298±9 e rgon 51. 326 me gista. 143±5. 133. 330 e pisthÂmh 107±8. 136. 332±3 oida 328 oi omai 329 354 . 114. 170±2. 110±11. 139. 168±70. 164. 233 e xaske w 280±1 e xeta zw 129. 283 gi gnomai 282 gnh siov 156±7 de 305 deiÄ 88. 325 atecnwÄv 297 autoÂv 324 basilikh 120±1. 288. 170±5. 269. 103. 322 dhÄlon oti/wv 225. 329 kaiÁ . 145 nuÄn 301. 162 anqrwpov 272. 135. 202±3 e rrwme nov 306 e terov 225. 277. 150±1. 138 makro teron 224. 171±2. 166. 325±6 apaideusi a 286 ara 292. 295. 329 dikastikh 121. 150±3. 200±1 blaÂptw 193±8 gaÂr 270. 115. . . 303±4. 202±3. 276 de omai 334±5 dh 282. 198±204 dioÂti 327 doxosoji a 129. 292. 124. 132. 303±4. 171±2 dikaiosu nh 4. 266±7. . 176±7. 299. 176±7. to 327±8 e neimi 302 e ni 302 e nioi 81±2. 129±31. 73±4. 121. 299 ge 269. 294 kai 265. 309. 185. 294.

122±3 72. 295 poreu omai 48 pouÄ 323 pou 308 proaÂgw 79±80 pronoe w 224. 133. 11c Miltiades fr. 42. 233 te cnh 65±6. 317±18 ostiv 265. 111. 112±13 aeschylus Seven against Thebes 19±20 295±6 anaximenes of lampsacus Rhetoric for Alexander 1412b18±19 65 66 107 113. 322 phÄi 330 planaÂomai 129 plei w.4 INDEX OF PASSAGES CITED omodoxi a 187. 318 skhnh 272 sumbai nw 287. 11a fr. 79±80. 322 jili a 185±93. 193 antiochus of ascalon Sosos (?) 87±8 antiphon the sophist fr. 314 jilosoji a 91. 121. 230. 190. 305 saje stata 225. 176±7. 150. 122 118 121 121 118±19. 150±3. taÁ 224. 316 polla kiv 271. 313 tekmai romai 77±8. 185±93. 165± 7. 219. 181. 3 71±2. 82. ) kai 283±4. 221. . 128 pwÄv ou 282±3 yeudwÄv 317 yuch 100. 189. 44 fr. . 81. 148. 169. 10 fr. 226 umne w 273 uperepaine w 264 upoqetikoÂv 84 upotei nw 51. 283 oude 217. 113. 113±16. 192. 96. 67 355 193 185±6 185±6 211±12 69. 319 omoÂnoia 175±6. 9 fr. 277. 304 prosje romai 284 protreptiko v 217. 193±8. 94±5 339 . 331 protroph 65. 325 juÂsiv 107±10 crhÄ siv 113±17. 279. 121. 315±16 sunousi a 264 wv 298 wsper 225 wste 312 wjele w 15. 322±3 tecnikoÂv 310 pa gkalov 296 paidia 158 parakeleuÂomai 142 parasunhmme nov 286 paÄ v 277. 170±5. 90. 142. 305. 309. 318±19 ontwv 225. 1 fr. 202 tauÄta 323±4 te 297 te ( . 225. 316±17 oun 286 oute 279. 299±300 jauÄlov 267 je romai 274 jhmi 83. 297 scedoÂn 296 swjrosuÂnh 129. 335 pei qw 142 peri 303 peritre cw 177. 58 antisthenes Protreptic fr. 292±3 4 INDEX OF PASSAGES CITED aeschines of sphettos anonymus iamblichi Alcibiades fr. 263 protre pw 1. 312. 104.

18±9 3.16.3 56±7 Eudemian Ethics 1224a29 315 1234b30±1 191 1236a2 315 1236a38±b10 314±15 1241a16±18 319 1241a32±3 318 1241b17±18 167 Eudemus 70 Nicomachean Ethics 1099b9±11 107 1116b25±1117a1 314 1155a22±26 191±2 1156a31±2 314 1152b9±10 315 1153a28±31 315 1155b34±1156a3 319 1161a34±5 167 1167a22±b4 318±19 1179b20±3 107 Politics 1327b38±40 81 Protreptic 69±70.19±21 13.48 3. 167 b 68 175 b 83 293 b 90±1 293 b 104±10 70.22.28 108 96 96 116 94±6 274.27 87±8 272±4 87±8 275 erinna Dista¨ euripides Orestes 915 fr. 338 b2 100.23.33±4 4. 336±9 Rhetoric 1358b8±13 65 Topics 110b10 276 chrysippus fr. 1091 356 20 300 193±4 .17±19 13.77 Birds 1033 Clouds 213±26 Frogs 967 Knights 680 275±6 272.8.66±83 3.37 3. 24±5 25 12 21 275 dissoi logoi 6. 115 Tusculanae Disputationes 3.204 Hortensius fr. 274 57 264 aristotle Constitution of the Athenians 29.21. 3. 276 b4 102.32. 221 democritus fr. 75. 17 fr.83±4 6.75 2.4.3 56 34.16 13. 242 dio chrysostom 1. 113 b 23±8 70 b 29±30 70 b 34 143 b 52 70.4 INDEX OF PASSAGES CITED aristippus 21 cicero fr.47.121±2 2. 121 aristophanes 69 Academica Posteriora 1. 276 111 90 diogenes laertius 1. 54 fr.16±27 13. 105. 118.1. 163±4. 100 b 53 118. 29±42 10.16±17 Ad Atticum 4.761 217±19 86±8 87 81 338 338 337±8 336±8 118 demetrius On Style 232 279 296±8 89 86 83±9. 112 fr.2 De Oratore 1. 143 b 59 112±13.26 3.32 194 22.11 108 epictetus 3. 91 a1 207 a 2±6 286. 113 b 7±9 70.

112. 164 115 188±90. 304±5 29d4±6 142 29d6±7 271 29d7±e3 71. 113 174b11 177 Cratylus 408b1 274 416e2±417a2 182 419a5±8 276 419b2 292 439c7 271 Critias 11. 145±6.17 13. 149±50 280b1±2 324 280b8±d7 105 357 .21 22±3 isocrates 1. 91. 126.84±5 63 339 339 276 195 107 108 60 lysias 9. 145 30a1±3 142 30a8±b2 112±13.117±18 13. 294 121 Apology 33. 141±3. 113.53. 143 38a5±6 339 39c6±d1 146±8 Charmides 152±3 154e6±8 266 161b6 189 164a9±b6 71.21 15.20 fr.19. 127.36±7 3. 141±6. 127. 92.p l a t o Alcibiades I 107a1±9 112e1±113c4 113d5±116d4 117c6±e5 124b10±d5 126a5±b10 126c1±127c9 129c5±e7 129e3±130b4 130e8±9 131a2 131a9±b9 131c1 132c1±6 131e1±132a2 133e4±5 134b7±9 135b7±c11 135d3±7 71±2. 155. 1. 148±9. 271 30b2±4 71. 319 167 112±13 100 100 118 100 112±13 119 115±16 112 118. 203 133e3±5 203 137c6±139a5 200±4 Euthydemus 12. 163±4. 39. 221 111±12 121 164 113. 217. 276 164c7±end 71 169c3±d1 324 171d8±e5 71.11±15 86 p l a t o and p s e u d o .3±5 307±8 hippocrates On the Art 66 hippolytus 1.25 12. 147±8 21b2 288 23b2 274 24e4±10 269 28e4±6 141±3 29b1±2 17 29d4±30a3 17±18.3.13±43 2. 73±4.3 207 herodotus 3. 26 193 57 philodemus PHerc 1021.5 1. 142. 142. 294 172d7±10 71. 103±4. 143 30e1±a9 146 30e2±5 104 30e7 142 31a4±5 104 31b4±5 143 33d7 295 36c5±d1 71. 100. 211. 152±3.4 INDEX OF PASSAGES CITED gellius 14. 325 278e3±282d3 71. 102. 214±15 274e8±275a2 143. 294 121. 19 Crito 47e3±7 218±19 49c10±d5 195±8 Epinomis 74±5 973b7±974c7 339 992b2±c3 338±9 Erastai 201.

119 118 106. 149±54. 232 Lysis 152±3 209a4±210a8 115 211d6±213a3 314 222d1±3 177 Menexenus 213. 160±3 275b2 156 275d4±e5 157. 76 108 199±201 285±6 201 308 169 218±19 159 270 195 217±19 199±201 123 76 33±4. 216 729a4±b1 126 729b1±2 101. 107. 105 732a6±b2 113±14. 110±11 77b6±78b2 285±6 79e7±80b2 138 82b9±85d4 137±41 100b2±4 29±30 Parmenides 36. 230 128d6±e1 216 137b2 158 PeriÁ a rethÄv 376a1±2 29 379d9±10 29±30 PeriÁ dikai ou 374c3±d3 194 Phaedo 61c9±62c8 293 64a4±69e5 62 64c2 308 65d4±5 308 76d7±8 271 114d1±115a2 76 Phaedrus 264a5±6 317 269d4±6 107 274b6±278e3 156±8.4 INDEX OF PASSAGES CITED 280e3±281e2 282d1 282a7±8 282b3±5 282c1±3 283a1±7 285b7±c3 288d5±292e7 288d9±e2 290e1±291a7 291b1±4 291b8±c1 291d1±3 292b4±5 292c4±5 292d8±e1 297d3±4 303b2 303c4 305e5 306d6±end 306e1±3 307c3±4 Euthyphro 12e1±4 15b10±c1 Gorgias 448c4±9 448e8 456a7±457c3 460b1±7 464b7±8 467c5±468c8 476d8±478b1 495c3±4 504d1±3 505a2±b1 507a5±6 507d1±2 508d6±e6 512a2±b2 520b3 521d6±8 526d3±end Hippias Maior 281a1 288c4 303e12±3 Ion 530a1 Laches 178a1±190b5 179d6±7 192c3±d8 113 291±2 105. 7th 156. 152±3 40 266 177 40 152±3 76 108 315 197a6±c1 314 200c7±d3 134 Laws 19 629d8 264 631b6±d1 102 632e5 158 639a2±3 296 661c1±5 218±19 691a7 284 697b2±6 100 724a7±8 101 726a1±729b2 101. 126. 109±10 232. 175±6 167 153±4 324 177 120±1 175 120±1. 216 275e1±2 46 276e2±3 157±8 276e6 134 278a1 161 278b7 158 Philebus 48d8±e10 100 358 . 123 52 273 264 301 318 71. 149 105 105 152±3 42 177 65 299±300 65. 294 752a1 158 Letter. 297±8 266±7 71. 232 234a1 40 236d4±249c8 229±30 Meno 70a1±4 29.

193 107 simon Cobbler's Dialogues 23±5 sophocles Oedipus 415 275±6 synesius Dio 57d±58a 359 42 .3 11 posidonius fr. 173±4. 3 66. 176 E. 2 E. 305 333d3±6 114 335c4 168 335e2 292 335e5 173±4 336c6±d4 182±4 337a5±338b3 328±9 340a1±b8 55 340c6±341a4 55 342d2±3 324 346a6±c4 184±5 350c12±d1 324 350d4±5 168 351c8±d9 187. 186. 280 Republic 12.±K. 37±8 Book 1 54±6. 196±8 332c5±d3 166. 76 320c2±328d2 76 322c3 193 323c5±324a1 107±8 330c1 308 343e6 274 351b3±4 308 352e5±357e8 76. 168±9 353d3±10 293 362e4±363a2 65±6 376d9 158 433a8±b1 189 433c6 319 434b2 295 443c9±444a2 168 444b1 187 445a5±b1 218±19 501e4 158 505a3 271 505b5±e2 151 554e4±5 168 608d11 270 612d7 266 621b8±d3 76 Sisyphus 387b7±c3 Sophist 229c1±231e6 232d6±e1 Symposium 173c1±d7 184a7±8 184c4±7 185b1±5 192d4 201e6±7 207a5±6 216a6±7 218c7±d5 219e3 Theaetetus 150a8±151d3 150c4±7 150d2±6 151b1±6 153a5 170e7±171c7 176b4 179e2 210b4±d4 Timaeus 72a4±6 27 19. 124±5.±K. fr. 64 88 65 proclus In Timaeum 7b 10 protagoras fragments/ascriptions fr. 204±9 328c6 317 331e1±336a8 159±60. 284 E. 35±6.4 INDEX OF PASSAGES CITED Politicus 19 258e8±11 201 269c4 270 293a2±4 123 305b1±c8 200 309c5±d4 123 310e9±10 319 311b9 318 Protagoras 316c5±317c5 65. 19. 318 352d8±353e11 114. 285±6 357e4±8 76. 228 129±34 66 36 76 119 119 119 274 324 324 270 119 119 222 132±5 329 153±4 134 266 286 270 299±300 132±5 19 189 plutarch Solon 32. fr.±K.

1±7 4.12±19 4.5±25 4. 308 113 73.4. 4.1 2.1 2.17 2.7 1. 216.27±33 2.38 4. (Oratores Attici ii 247) 9±10 theognis 429±38 thrasyllus 108 22±3.31±6 4.9 4. 102 118 121 271 121 339 308 200 73 280 111±12 121 71.2. 232±3 25 30 360 1.2.1 25. 194 71 118.4 4.4.19±20 4. 320d±321c 97±8 theocritus 6.7.34 317 theodectes fr.4. 187 329 318 121 194±5 . 216 25 45±6.1 1.25±9 4. 79 77±82.39 4.4±1.2 1. 77.11 4.1 4.2.4 INDEX OF PASSAGES CITED themistius 4. 230.5.2 4. 263 xenophon Memorabilia 1.