Among Plato's works, the Statesman is usually seen as transitional between the Republic and the Laws. This book argues t...
41 downloads
563 Views
6MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Among Plato's works, the Statesman is usually seen as transitional between the Republic and the Laws. This book argues that the dialogue deserves a special place of its own. Whereas Plato is usually thought of as defending unchanging knowledge, Dr Lane demonstrates for the first time how, by placing change at the heart of political affairs, Plato reconceives the link between knowledge and authority. The statesman is shown to master the timing of affairs of state, and to use this expertise in managing the conflict of opposed civic factions. To this political argument corresponds a methodological approach which is seen to rely not only on the familiar method of 'division', but equally on the unfamiliar centrality of the use of 'example'. The demonstration that method and politics are interrelated transforms our understanding of the Statesman and its fellow dialogues.
CAMBRIDGE CLASSICAL STUDIES General Editors
M. F. BURNYEAT, P. E. EASTERLING, M. K. HOPKINS, M. D. REEVE, A. M. SNODGRASS
METHOD AND POLITICS IN PLATO'S STATESMAN
METHOD AND POLITICS IN PLATO'S STATESMAN
M. S. LANE University Assistant Lecturer in History in the University of Cambridge, and Fellow of King's College
I CAMBRIDGE UNIVERSITY PRESS
PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE
The Pitt Building, Trumpington Street, Cambridge CB2 I R P , United Kingdom CAMBRIDGE UNIVERSITY PRESS
The Edinburgh Building, Cambridge CB2 2RU, United Kingdom 40 West 20th Street, New York, NY 10011-4211, USA 10 Stamford Road, Oakleigh, Melbourne 3166, Australia © Faculty of Classics, University of Cambridge 1998 This book is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1998 Typeset in 11/13 pt Times New Roman and Greek New Hellenic A catalogue record for this book is available from the British Library Library of Congress cataloguing in publication data Lane, M. S. Method and politics in Plato's Statesman / M. S. Lane. p. cm. - (Cambridge classical studies) ISBN 0 521 58229 6 (hardback) 1. Plato. Statesman. I. Plato. Statesman. II. Title. III. Series. JC71.P314L36 1997 / 32o .oi - dc2i 97-6077 C I P ISBN 0 521 58229 6 hardback Transferred to digital printing 2005
AO
To my parents, elders, and teachers
CONTENTS
Acknowledgements
page xi
Abbreviations and note on text
xii
Introduction. Method and Politics in Plato's Statesman Part I
i
Method Introduction The Sophisfs use of example and division The Statesman's use of example and division
13 21 33
The analysis of example Conclusions
61 75
Kinship and method Shepherding Weaving
Revising common sense: the delusions of the given Rhetoric and method
Part II The story as a fulcrum of the dialogue Introduction: story-telling and self-criticism Telling the story Zeus versus Kronos: autonomy versus authority The imperative of invention The necessity of narrative
Criticising the story
The insufficiency of narrative: remaining rivals Delusions of grandeur: Kronos as a 'great example' Delusions of grandeur: measured by the mean On the temporality of the kairos
Part i n Politics Introduction Rivalry revisited: subordination of the arts to the mastery of the kairos ix
33 40 46 75 89
99 101 101 III 114
117
117 119 125 132
137 139
CONTENTS
Rivalry renewed: the challenge and subordination of law 146 Two concepts of law: as mortmain and as memorandum 148 A third view: law as makeshift in the second-best regimes 155 Political knowledge as weaving 163 Weaving as political example: gender and simple unity in the Lysistrata 164 Weaving as political example: complex unity in the Statesman 171 Conclusions 182 Delusions of righteousness: the political significance of evaluative conflict 18 2 Delusions of righteousness: the political significance of authority and time 193 Select bibliography General index Index Locorum
203 218 224
ACKNOWLEDGEMENTS
This book has emerged from a doctoral dissertation at the University of Cambridge which was supervised by Myles Burnyeat and Malcolm Schofield and examined, most constructively, by David Sedley and Christopher Gill. Many debts are recorded in the dissertation, including the financial support of the Marshall Aid Commemmoration Commission, the Harry S. Truman Foundation, and the Mary Isabel Sibley Fellowship of Phi Beta Kappa, and the intellectual and moral support of valued friends and colleagues. My thanks once again to them all. Here I mention only those debts to individuals incurred in the writing of the present book. Foremost among these is that owed to Myles Burnyeat, who championed the book for Cambridge Classical Studies, and to the other editors of the series. Malcolm Schofield and M. M. McCabe very generously commented on a penultimate draft. Serving with Malcolm and Christopher Rowe on the editorial team of the forthcoming Cambridge History of Greek and Roman Political Thought has kept me Platonically engaged while writing this book, as has the continuing camaraderie of Verity Harte. Alonso Tordesillas kindly invited me to participate in the work of the Centre d'etudes sur la pensee antique 'Kairos kai Logos'. Between completing the dissertation and completing the book I became a University Assistant Lecturer in the Faculty of History and a Fellow of King's College, Cambridge. To Quentin Skinner and Jonathan Steinberg, who have chaired the Faculty during my time there, and to Patrick Bateson as Provost of King's, go my thanks for the support of their respective institutions, as to the colleagues and administrative staff thereof. Finally, warm gratitude to farflung family and friends and above all to A. L., who embodies loving care in the face of all exigencies. XI
ABBREVIATIONS AND NOTE ON TEXT
Works of Plato are cited from the edition of J. Burnet, Oxford Classical Texts series (Oxford, 1900-1903), now superseded for volume I by the revised edition of E. A. Duke, W. F. Hicken, W. S. M. Nicoll, D. B. Robinson, and J. C. G. Strachan (Oxford, 1995). Unless otherwise stated, I quote from the excellent recent translation of the Statesman by C. J. Rowe, an edition and translation to which I and other students of the dialogue are much indebted. Reprinted by permission of the publishers from Plato: Statesman, translated and edited by C. J. Rowe, Warminster: Aris and Phillips, 1995. I have also quoted from the Loeb translation of the Sophist, reprinted by permission of the publishers and the Loeb Classical Library from Plato: Theaetetus and Sophist, volume vn, translated by Harold N. Fowler, Cambridge, Mass.: Harvard University Press, 1921.
The abbreviations for Platonic works are made clear in the text; the most often used are St. Statesman (Politicus in OCT) So. Sophist (Sophista in OCT) Works of Aristotle are cited from the edition of I. Bywater, W. D. Ross, W. Jaeger, Oxford Classical Text Series (Oxford, I897-I957)-
ABBREVIATIONS
The abbreviations for Aristotelian works used are as follows: Met. Metaphysics Pol. Politics N.E. Nicomachean Ethics LSJ DK
Liddell, H. G., R. Scott, H. S. Jones, eds. A. Greek-English Lexicon (Oxford, revised 1968) Diels, H. and W. Kranz, Vorsokratiker (Berlin, 10th edn, i960)
Xlll
INTRODUCTION
METHOD AND POLITICS IN PLATO'S STATESMAN As the discovery of truth and the direction of life are the twofold function of philosophy, so Plato saw a twofold counterfeit of his ideal educator and governor in the professors of wisdom and the public men of his time. The one corrupted inquiry with controversy, the other spoiled politics with faction. Lewis Campbell1
A colleague once remarked to me that the Statesman is a 'very lonely' dialogue. Interpreters as different as the dean of Anglo-American analytical scholarship, Gilbert Ryle, and a leading Straussian have found it wearying and rebarbative to read.2 It has won little reflected glory from the analytical attention paid to its companion the Sophist in the last thirty years.3 Some have taken it to be mainly a discourse on the method of division, itself a procedure of dubious import, and in any case presented more fully in the Sophist, Philebus and Phaedrus*. Others consider it essentially a discourse on 1 2
3
4
Campbell (1973) i, referring to the Sophist and the Statesman together in introducing his edition of both dialogues. Ryle (1966) 285 calls it 'this weary dialogue'. Benardete (1984) Vol ni.73 asks, 'Does the Statesman demand a special effort on our part not to grow tired?', though it should be said that unlike Ryle he finds this wearying quality to be part of the dialogue's message about the application of knowledge to politics. See also Grene (1950) 181: 'The Sophist and the Statesman are among Plato's work unique in that they are dull.' An exception is McCabe (1994). Studies in the Straussian tradition have tended to treat Theaetetus, Sophist and Statesman together: Benardete (1984), Klein (1977); see also Dorter (1994), though his discussion of the Statesman is mainly a summary of its text. A recent and welcome wave of studies of the dialogue includes the papers of the Third International Symposium Platonicum held in Bristol, 1992, selected in Rowe (ed.) (1995), with others in Nicholson and Rowe (1993); the translation introduced by Annas (1995), the masterful edition with commentary of Rowe (1995), and Rosen (1995). Miller (1980) is also thoughtful. Taylor (1961) 9 says that 'the really serious business of both dialogues' is with method rather than the identification of sophist, statesman, or philosopher. Ryle
INTRODUCTION
political theory, 5 though pallid beside the poignancy of the Apology and Crito, the vitriol of the Gorgias, the grandeur of the Republic, the monumentality of the Laws. Seldom have studies of the method and politics of the dialogue been combined in more than a consecutive way. 6 This book explores their intimate connection. Both the method and the politics of the Statesman hinge on the question of the authority of political expertise and how it is to be distinguished from rival forms of expertise rampant within the city. The statesman cannot be defined without distinguishing him from similar rivals, nor can he rule without using his knowledge to command these rivals within the city. In both cases a key rival is the rhetor. He is not banished from the city, but allotted a strictly subordinate role within it, while his concepts for argument and skills in politics are appropriated and revised in defining the statesman. To resolve the challenge of rivals requires a method of definition which appropriates the rhetorical notion of 'example', and a politics of the authority to command which appropriates the rhetorical notion of the 'kairos* or the appropriate action at a given moment. Both method and politics, moreover, strive to reform assumptions and puncture delusions about the evident worth of certain forms of expertise or political action. In its fundamental concern with the nature and authority of political expertise, the Statesman provides one answer to the problem posed in the second protreptic of the Euthydemus (288-92): what could be the subject matter of political expertise? After all, Socrates and Crito muse inconclusively
5
6
(1966) 285-6 sees whatever value the divisions have as the value of the dialogues: he considers division in the Statesman a demonstration 'for beginners only', compared to the more advanced Sophist. Contrast Rosen (1995) 2: the dialogue is 'a demonstration of the inappropriateness of diaeresis to the study of human affairs'. Skemp (1961) 67 finds the interest of the dialogue in the politics and the 'myth': he considers the divisions in the Statesman as at best 'a gentle satire' on the clumsy arrogance of the Academy youth. Taylor (1961) 192 is representative: '[T]he Politicus [the Latin name for the dialogue] has a double purpose. It is meant to provide a second and still more elaborate demonstration of the method of "division" . . . and also by leading us to a sound definition of the statesman, to enforce certain fundamental material points of constitutional theory.'
METHOD AND POLITICS IN PLATO'S STATESMAN
there, the 'kingly art' can't possibly know how to carry out the specialties of other experts, be they doctors or carpenters, as well as they can. Yet the kingly art, which is wisdom, must know how to use the goods these other specialists provide if they are to be beneficial rather than harmful. How can the kingly art know how to use what it does not know how to produce? And in virtue of what kind of knowledge, then, can the political expert (or possessor of the 'kingly art') 7 claim to rule? In the Republic, the philosophers are to rule not in virtue of any peculiarly political knowledge they possess, but rather in virtue of their synoptic and pervasive understanding of the Good. If the question were to be pressed, however, what specifically counts as political knowledge in the Republic, it is not clear what the answer would be. 8 This is the question which is pressed in the Statesman. Given how many forms of expertise there are in the city - cobblers, generals, navigators and so on - two questions must be asked about the postulate of a purely political expertise. First, what does it know? Second, how does it rule? The Statesman answers these questions (and so circumvents the problem of possession and use) by distinguishing between knowing what to do and knowing when to do it. In assigning to statecraft the unique role of commanding when each expert should perform his work and so coordinating the work of different experts, the dialogue emphasises time as the dimension of political action. Political expertise is neither meta-knowledge nor another species of knowledge, but rather knowledge of the relation 7
8
The Euthydemus talks only of 'wisdom' and the 'kingly art' (basilike techne) in this context. The Statesman alternates indifferently between 'politike episteme\ 'politike techne', and 'basilike technt, though it should be noted that the Sophist uses techne and never the sometimes more prestigious episteme for the dubious art of the sophist. I follow Rowe (1995) 1-2 in usually translating all the Statesman's terms by 'political expertise' while using 'statecraft' or 'the art of . . . ' in contexts where 'expertise' would be awkward. Note also Rowe's discussion ad loc. of the contrived technical formation of politikos (statesman), which he suggests may be a neologism. Myles Burnyeat has argued in lectures at Cambridge that the Euthydemus can be read as postdating and critically commenting on the Republic. His lectures stimulated my interest in the problem about the kingly art in the Euthydemus. However, the present argument does not depend on accepting any particular view about dating the dialogues.
INTRODUCTION
between other forms of knowledge and the temporal demands of the moment of action, or the kairos. Based on mastery of the kairos and command of the other experts, the political expert's authority is exalted against the static and inflexible authority of the written and unwritten laws. But at the same time, legal authority is defended against the vagaries of personal rule, whether of tyrant or faction, in the absence of a true political expert. It is this double move, first attacking and then offering a limited vindication to the law, which has led some scholars to identify a newly realistic attitude in the Statesman,9 one which is more favourable to democracy 10 (a law-abiding democracy being classed as better than any lawless regime), and which marks a transition toward the insistence on the rule of law rather than men which characterises the Laws. The eagerness to find evidence of softening and transition, 11 however, risks obscuring the uncompromising vindication of the nature, possibility, and authority of political expertise (in relation to its rivals) which I claim to be the central concern of the dialogue. Indeed, one advantage of my interpretation is that it rescues the Statesman from the halfway house of transition, dependent on a dating of the works impossible to establish with certainty, and finds in it a philosophical identity which is based rather on its pressing of certain questions which are not pressed so far or so hard - for whatever reasons - elsewhere in Plato. To say that the Statesman presses its own philosophical questions is not to say that its answers are wholly stable, or satisfactory from a modern perspective. The idea of certain 9
10
11
Annas (1995) exemplifies this view of the dialogue as, on the one hand, taking a new interest in the 'real world' (x) but, on the other hand, 'unstable because Plato has not yet thought through the degree of compromise that these new ideas demand' (xvi). Cf. also Barker (1918) 330: 'Plato makes his peace with reality . . . ' The question of whether the Statesman is really kinder toward democracy than other Platonic works will be looked at in the context of fourth-century debates in Part m. Annas (1995) xxii is again representative in viewing the dialogue as transitional: 'The Statesman is in some ways a record of complication and even confusion. But not only does it help us to see how we get from the Republic to the Laws, it is a record of the entanglements that only a very great and original thinker, defending and qualifying his boldest work at the same time, could get into.'
METHOD AND POLITICS IN PLATO S
STATESMAN
knowledge of the kairos founders on two difficulties: the multiple variables involved in having such knowledge, and the touchy problem of knowing how to recognise who does. The statesman must know precisely what needs to be done at every moment and by whom, surmounting the stunning difficulties of calculation of the future in a prospective judgment without the benefits of hindsight. Although concentrating on the practical and temporal knowledge of the statesman, the dialogue does not separate the unity of reason. 12 The Statesman recognises the need for reason to be practical but considers this still a purely intellectual achievement and one which can in theory be perfect. There is no place in this conception of practical reason for the lesser precision allowed it by Aristotle, nor for the latter's emphasis on the contribution of habituated perception and judgment to its success.13 Later political thought tended either to accept the Aristotelian version of practical reason or even to strip away all claims to reason from the practical effects of will or virtue. 14 The Statesman can claim, in this regard, very little influence on the subsequent history of political thought; even where its themes of temporal knowledge and decision are taken up, they are associated with the mild counsels of prudence rather than the knowing commands of the statesman. In insisting on an objective kairos knowable by the political expert, the Statesman even more than the Republic remains within the ambit of the Platonic agenda identified by Karl Popper: its question is not, 'How can we so organise political 12
13 14
Phronesis is occasionally mentioned in the Statesman but never denned or explicitly linked to knowledge of the kairos, while the claim that the statesman has episteme is never renounced: for these reasons I find no grounds for distinguishing the statesman's knowledge as a special form of practical wisdom, pace Rosen (1995) vii: '[t]he central theme of the Statesman is the relation between phronesis ... and techne\ Griswold (1989) 152-3 notes the cosmos' possession of phronesis in the Statesman's story (269di) but observes that the dialogue's focus on episteme and techne renders this 'rather far' from Aristotelian phronesis. Lear (1988) 170-4 gives a clear account of Aristotle's distinction. One version of seizing the moment was offered by Machiavelli's virtu, on which see Skinner (1978) Vol. 1, 128-38; other Renaissance conceptions emphasised the importance of virtue, in particular the virtue of temperance, in being able to act prudentially in line with the kairos; see the suggestive remarks in Hutson (1996).
INTRODUCTION
institutions that bad or incompetent rulers can be prevented from doing too much damage?' but rather 'Who should rule?' 15 The Republic had only asked this after identifying the needs of the soul, the city, and the relation of the virtues to knowledge. The Statesman takes for granted that the person with knowledge should rule, and goes on to ask, 'what does rule consist in, and how is knowledge related to rule?' and the theoretical interest of this question will concern us in this book. Still, it is significant that neither the Statesman nor any other Platonic dialogue asks a further question which might be thought to arise for a commitment to wise rulers: 'is there a way for non-experts to know who should rule?' We will see in Part in that the Statesman discusses what should happen when people meet with a true political expert, and even discusses why what should happen (recognition and obedience) might not, but offers no instruction beyond this as to how the people should discriminate between, as it were, a true and a false prophet. The comparison with prophecy is especially telling because at stake in the claim to know the kairos is the claim to know the future. To be told that one ought to submit to a true prophet, and even be told what such a true prophet would know and be able to do, is of little use if false prophets abound on every side. In this light it is not surprising that the Statesman has proved so infertile in later political thought. Though I hope to show that it is rewarding to think through, it has evidently been a difficult text to think with in the development of the tradition. Though lonely in its recalcitrance to readers, the Statesman is less genuinely singular than most other Platonic dialogues. It continues the conversation recorded in the Sophist begun when a 'stranger from Elea' is invited by Socrates and Theodoras to ascertain the relative worth of sophist, statesman, and philosopher (So. 216C-217C). This Stranger appears only here and conducts only these two discussions exploring two 15
Popper (1995) 120, a work written during the Second World War andfirstpublished in 1945. The question of whether Popper read Plato fairly notwithstanding, this is a ringing appeal to the liberal conscience.
METHOD AND POLITICS IN PLATO'S
STATESMAN
parts of the same question. Some have claimed that because the Theaetetus is also mentioned implicitly at the beginning of the Sophist, any interpretation of the pair must encompass it as well. 16 This claim requires examination. Theodorus opens the Sophist (216a) saying: 'Here we are once more, Socrates, as yesterday's engagement requires of us, and we have brought a visitor too.' 'Yesterday's engagement' almost certainly refers to the close of the Theaetetus describing a meeting between Socrates, Theodorus and Theaetetus, ended when Socrates leaves to meet his indictment and instructs Theodorus to meet him again the following dawn (2iod). But to insist too much on this back-reference creates some difficulties: are we then to assume that the framing claim of Euclides to be narrating the Theaetetus (i43a-c) encompasses the two unnarrated dialogues as well? 17 If not, then we have in the Sophist and Statesman a performed version of conversation, 18 while the Theaetetus narrates the previous day's events with the hindsight of several decades. Nor is the Theaetetus' topic explicitly linked to the next day's reunion. In this, as in the presence and leadership of the Stranger from Elea, Sophist and Statesman are tightly bound together to an extent much beyond the claimed chronological proximity of the conversation narrated in the Theaetetus. The extent of the textual connection between them justifies a common reading of the pair without prejudice to the possible connections of the Theaetetus or of other dialogues. Such a common reading is offered here only in part. I have not attempted a global interpretation of the two dialogues the epigraph is as stimulating a remark as I know on the subject - or the vexed question of the 'missing' philosopher. 19 16
17 18 19
Miller (1980) 1-2, Klein (1977) 3 and Benardete (1984) xvi, though the last acknowledges that the relation between the pair is closer. Dorter (1994) tries to argue for a quartet of 'Eleatic' dialogues, including the Parmenides on the basis of Socrates' remark to Theaetetus that he once met Parmenides as a young boy (Tht. 183c). Myles Burnyeat once asked me this. Perhaps conducted without a break: Dies (1935) vii. Sprague (1976) 100 and Klein (1977) 177 identify the statesman with the philosopher; Griswold (1989) 163 n. 13 seems to me right in observing, against them, that
INTRODUCTION
I have, however, attempted here to lay out the methodological relationship between the Stranger's two inquiries, since the methods of division and example are introduced in the Sophist, and more fully explored in the Statesman. Perhaps these comments will ease the way for further attention to the oddly neglected task of interpreting the two dialogues together. On this note it must be said that I have taken one liberty and refused another. I take the liberty of identifying the Eleatic Stranger's arguments with Plato's, in the spirit of identifying what the arguments are on their own terms rather than seeking clues that they are to be disregarded or minimised. I refuse the liberty of speculating here on the relationship between Socrates and the Eleatic Stranger, or Socrates and Plato. Such speculation would require at the very least that fuller interpretation of the Sophist and the Statesman as a pair suggested above. The discussion tries not to assume knowledge of the dialogue, but inevitably those able to follow along with a text either in English or Greek will gain more from the proceedings (and find them more interesting).20 The introduction to each Part, and the conclusions to Parts i and m, are designed to be so far as possible free-standing, summarising the main lines of interpretation and linking them to wider issues in philosophy and the history of political thought. Those who turn to them first, or solely, will find the principal results of the book on the trust that the sections in between have established the textual evidence. To conclude this introduction I sketch the progress of the argument through the three parts. The fundamental contention of Part I, dealing with method, is that the Sophist adumbrates and the Statesman develops
20
the Stranger says explicitly that sophist, statesman and philosopher are three (So. 217b). But see the persuasive arguments for finding the philosopher rather in the practice of philosophy in the dialogues in Frede (1996) 146, 149-51, and McCabe (forthcoming). Rowe (1995) is an invaluable critical edition, with introduction, translation and commentary. Translations by Waterfield in Annas (1995) and Ostwald/Skemp (1992) are also widely available.
METHOD AND POLITICS IN PLATO'S STATESMAN
a combined method consisting of two elements: division (diaeresis) and example (paradeigma). Whatever may be true of division in other dialogues, we fail to understand its role in the Sophist and the Statesman if we fail to recognise its intimate links, in both dialogues, with the use of an example of a special kind. The characteristic features of such examples include their relative insignificance compared to matters of political import, their structured definitions, and their relevance to the subject of the inquiry in hand. Division works adequately to define the main objects of inquiry only when it elucidates the structure of definition either of the example or of the item exemplified.21 Moreover, I shall show that the Eleatic Stranger's handling of examples involves attention to their implicit logic of similarity and to their role in inquiry, developing a theoretical construction of example which compares with work in contemporary philosophy of science and contrasts with the use of example in the ancient practice of rhetoric. The link between method and politics, in the centre of the dialogue and so in Part n of the present book, is provided by the outlandish muthos ('story', in my translation, rather than the usual 'myth'). The story has tended to monopolise a literature of its own, 22 often with some reference to the politics preceding it in the dialogue, but with none to the politics which follow or to methods other than 'myth' itself. I interpret the story as the fulcrum of the dialogue, which in hindsight makes sense of the preceding method and politics, and in foresight introduces the methodological and political themes still to come. In the aftermath of the story, method and politics are themselves unified. Reflecting on his own story after telling it, the Eleatic Stranger suggests that it was flawed by a political prejudice (a traditional belief in kings as shepherds, 21
22
Contrast D o r t e r (1994) 227 and passim, w h o seeks a 'method of hypothesis' (like that of the Republic) t o complement division in the Statesman, b u t is reduced t o gesturing a t the general 'overcoming of incompleteness' in lieu of textual evidence for such a method. M o r e papers were given o n it t h a n o n a n y other topic at the 1992 Third International Symposium Platonicum which was dedicated t o the dialogue and held in Bristol.
INTRODUCTION
ruling a tame and unthinking population) which turned into a methodological prejudice (the telling of an overly long story which cannot be properly used as an example) (St. ijjbi-6). The puncturing of prejudices, the purportedly common-sense assumptions which often distort both intellectual inquiry and political choices, is signalled in the headings, in all three parts of the book, about the identification and curing of 'delusions'. According to the Stranger's self-criticism of his story, the method necessary in its stead is an adequate, and appropriate, example. The example he now chooses is the art of weaving, and this example eventually guides not only the definition of statecraft as an intellectual competence, but also an analysis of its characteristic activity. As argued in Part m, statecraft emerges in a special relation to the other arts, able to judge and coordinate their opportunities for action, which results inter alia in the subordination of rhetoric. The political expert is also to carry out the task of weaving together two conflicting factions in the city. Each of these factions is conceived as characteristically disposed to err on one side or the other of the mean, in making evaluative judgments. I shall show that it is essential to recognise the dynamic temporal background assumed in this analysis in order to make sense of both the problem and its proposed resolution; further, that such a conception of evaluative conflict speaks to an important issue in the prior, and subsequent, history of political thought. The ideal of statecraft is, however, only one half of the political theory developed in the Statesman. A wise ruler is, as the story somewhat fancifully suggests, not always available to guide human affairs. Indeed, according to the story, in the present cosmic epoch the god has retired from active guidance, leaving humans to imitate the cosmos in forced self-rule. If the story's cosmology reassuringly insists that the god will someday return to manage human affairs, the political parallel drawn is less comforting. Arrival of a genuine statesman is a theoretical possibility, and must be kept open as a practical one. But in his absence, without the guarantee that the true statesman has ever come or ever will, humans must contrive to rule themselves. And, making use of the idea of imitation 10
METHOD AND POLITICS IN PLATO'S STATESMAN
in the story, the dialogue develops a paradoxical account of just how such imperfect self-rule can, by strict adherence to law, achieve some feature of the absent political ideal while remaining open to its eventual realisation. The Statesman provides very little of the meditation on institutions, and none of the meditation on human weaknesses and their possibly calamitous effects, that we have come to expect of a political theory (and which the Laws does amply provide). In this respect the Statesman is austere. The dialogue invites us to consider only the absence of natural authority (in the story), the plethora of rivals for political authority, and the concomitant need to define and establish the rule of the true political expert. But whereas a sceptical thinker like Hobbes, for example, would pursue the establishment of such authority as a matter of artifice, the Statesman pursues the objective 'art' in it: the political competence which can establish a state most successfully. Construing political knowledge dynamically, accepting a certain inevitability of conflict, and also considering options for politics in the absence of knowledge, the Statesman tests the limits of an objectivist approach to political value and is limited in turn by the possibilities of such an approach.
ii
PART I
METHOD
Introduction
Socrates has asked the Eleatic Stranger whether people in his country consider sophist, statesman and philosopher one, two, or - there being three names - 'divide' (diairoumenoi) them into three 'classes' (gene)1 and ascribe a class to each name (So. 217a). The Stranger answers, 'three', but says that it is not a small or easy thing to 'define' each 'clearly' (So. 217b): he undertakes first to search for and 'make plain by argument' what the sophist is. The implication is that the inquiry set by Socrates will require the establishment of definitions of sophist, statesman, and philosopher. This task is undertaken and achieved for sophist and statesman in the eponymous dialogues: definition is what they set out to give and claim to have achieved. Both inquiries contrast the definition sought with the mere assertion of a name, though the having of the name in common (So. 217c) is the starting point for the argument. Both involve separating the expertise in question from a host of rival arts, or forms of expertise, claiming the name for themselves. (Some of these are assigned roles as subordinate or collaborative kinds of expertise, while others are unmasked as impostors). And both define a paradigmatic example in order to originate, or complete, their search. Definition is achieved by dividing kinds of knowledge in order to connect them in an analytical account (logos) giving common meaning to the name. The argument of Part 1 is that the Sophist introduces the interaction of division and example; the Statesman shows that 1
See the discussion of genos and eidos below (pp. 17-18). 13
METHOD
their interaction is required successfully to define the statesman, and analyses the crucial importance of example in enabling division to succeed by picking out which similarities between arts are relevant and which are to be disregarded. The Stranger sometimes employs division without example, though never example without division. Part of the burden of the Statesman is to show why, for important matters, example and division must go hand in hand, though this is not an inevitable feature of any inquiry. Linking division with example is rather an achievement, to be understood and recognised as such, and so here analysed both in its parts (division and example separately) and whole (their interaction). In identifying a method of 'example and division' in these two dialogues' quests for definition, I mean to challenge the widespread assumption2 that the method they use is the 'collection and division' which (so the assumption goes) is identical with that prescribed (by Socrates) in the Phaedrus (265d-266b), in the Philebus' 'gift of the gods' (i6b-i8e), and (by the Stranger) in the 'philosopher's science' of the Sophist (253c-d; cf. 25ia-259e) and the instruction on dealing with likenesses and unlikenesses in the Statesman (285a-b). These passages, which are severally difficult and collectively diverse, 1 label the 'prescriptive passages' in contrast with the 'actual divisions' made in the course of defining sophist and statesman. The 'actual divisions' are strictly confined to forms of expertise {technai or epistemai), the persons who exercise them or the objects on which they do so. As we shall see, examples are explicitly introduced in connection with these actual divisions; no explicit mention is ever made of anything called 'collection' (and interpreters disagree as to what, in the actual divisions, does illustrate it). None of the prescriptive passages from the Stranger's dialogues or Socrates' say anything about example. I seek therefore to clarify the actual workings of 2
So Bluck (1975); Hackforth (1972a) 134-6 and (1972b) 142-3; Stenzel (1940) introduction (by Allan) xliii and ch.9; and Nehamas (1984). Healthy scepticism however is expressed by Gosling (1975) 82, 87, 202-3. Wedin (1987) attempts to find a middle ground, arguing that the Statesman corrects some aspects of the Phaedrus' method of division. 14
INTRODUCTION
division, and its interaction with example, without the prejudice involved in importing prescriptive assumptions about what division must be. To adopt this strategy of attending to the 'actual divisions' given by the Stranger is not to deny that there may well be affinities with the prescriptive passages, or that those passages do establish some notion of division as a logical method applicable to more than just the technai. One of the best and most recent defences of the assumption of a single method lying behind both prescriptions and actual divisions may illustrate both the affinities and the risks. M.M. McCabe has recently offered a stimulating account of 'collection and division' as signifying a new Platonic answer to the problem of what makes an individual one. 3 Unlike Aristotle, who would answer that what makes an individual one is 'being one of a kind', she claims that Plato wrestles his way to the answer that what makes an individual one is its relation to other individuals, its place in a context. When she speaks of 'collection and division', however, she appeals almost exclusively to what I called above the 'prescriptive passages', particularly of the Sophist and Philebus. Her only reference to the 'actual collections and divisions Plato gives' is as 'comical analyses of absurd skills (weaving or angling)', 4 and though this remark seems tongue-in-cheek, the 'actual divisions' are never accorded further analysis (indeed the Statesman is not discussed in much detail at all). The very fact that her account can be so persuasive for the 'prescriptive' passages, while paying virtually no attention to the 'actual divisions' which are of technai and using examples, suggests that distinguishing between these discussions is both possible and advisable. Discussion of this general question about division affords the opportunity to comment on three more standard issues in the existing literature on division. These are the distinction between meros and eidos; the distinction between eidos and genos; and the issue of whether division of eide is dividing
3
McCabe (1994). 15
*Ibid., 222.
METHOD
'Forms' (eide) in the sense in which such 'Forms' are construed in the Phaedo or Republic. Consider the last and most controversial question first. In Platonic scholarship, whether a dialogue alludes to 'Forms' has become a touchstone for the debate as to whether Plato's dialogues are universally consistent, or if they are not, whether he changed his mind ('developed' his thought) in identifiable directions which afford or concur with a chronological ordering of the dialogues as composed.5 To this structural issue corresponds a highly charged emotional and intellectual one: was there a core to Plato's teaching, and was that core the so-called 'doctrine of Forms'? G. E. L. Owen enlisted the Statesman in his well-known argument that Plato did change his mind about Forms, and that the Timaeus (in which Forms appear) must therefore be dated to the 'middle period', after which a 'late' Plato critical of Forms and concerned with the analytical philosophy of language produced the Parmenides, Theaetetus, Sophist, Statesman, Philebus, and Laws.6 His polemic, though unconvincing in some details, helped to establish a robust (though not uncontroversial) tradition of reading the 'late' dialogues without reference to Forms and in the spirit of philosophical analysis. This book is within the spirit of that tradition in finding no arguments of the kinds appealing to 'Forms' in the Phaedo or Republic in the Eleatic Stranger's dialogues, and so in not assuming that such 'Forms' are what he means when he uses the ubiquitous Greek word eidos.1 However, this need not indicate a dramatic rupture on all fronts between 'late' Plato and all other Plato. That the 5 6
7
For a brief discussion, see Lane (forthcoming b). Owen (1965) (given a general rebuttal by Cherniss (1965)) claimed that part of the Statesman's 'myth' about the divine shepherd contradicted accounts of political shepherding in the Critias, the Timaeus' incomplete sister-dialogue; this use of the Statesman was decisively refuted by Gill (1979). A more successful assault by Owen (1973) on evidence for late-period Forms in Statesman 285d-286b (without reference to the dating of the Timaeus) is considered below, pp. 70-75. This is further supported by the fact that in the Eleatic Stranger's hands, eidos is used to categorise the technai rather than the virtues or standards for comparison suggested by middle-period arguments involving Forms. 16
INTRODUCTION
Statesman approaches politics differently from the Republic may mean that they are simply exploring different aspects of common questions. One finds in each dialogue glimmers of themes which appear and reappear throughout others, being reworked, explored, and treated from different perspectives in the individual world of each dialogue. As on McCabe's account of individuation, the dialogues constitute contexts for one another but each must be understood, within those contexts and the broader contexts of intellectual and political history, as distinct. If eidos in the Sophist and Statesman is not used in contexts which suggest Phaedo- or Republic-style explanatory Forms, how is it to be translated? The question is especially acute with regard to the interplay between eidos and genos common to both dialogues, terms which are tempting to identify with the 'species' and 'genus' of Aristotelian biological science.8 This temptation must be resisted.9 Though the root sense of 'genos' is extensional ('race', 'stock', 'kind'), it can also be used intensionally ('sort', 'type'); conversely, though the root sense of 'eidos' is intensional ('shape', 'outward appearance'), it can also be used extensionally ('domain', 'class'). 10 And indeed in the actual divisions of our dialogues the terms are used interchangeably in many contexts. Such flexibility within each term and between them thwarts attempts to establish a strict skeletal logic of division as an 8
9
10
E.g. Taylor (1961) 9: the serious business of both dialogues is the value 'of precise and careful subdivision of genera into their constituent species'. Perhaps it is truer to say he is tempted by Mendelian, not Aristotelian, influence: Pellegrin (1986) 69 (cf. 50-112) argues that in Aristotelian biology 'the pair genosjeidos . . . constitutes a diaeretical tool functioning at any level of generality at all'. They do not connote only the proximate genus and infimae species, but are used flexibly relative to each stage in the inquiry. Cornford (1935) 268-72 rejected the Aristotelian interpretation of Platonic division but still uses 'genus' and 'species' language for division's analysis of 'Forms'. Aristotle himself notoriously attacked division as a method of proof (An.Pr. 1.31, though Rose (1968) argues that it influenced his development of the syllogism), but reformulated an acceptable version (An.Po. 11.5 9^28-31) and used this in his biological works as a method of denning natural kinds. I owe this schematisation to De Rijk (1986) 32-3. It is meant to illustrate the force and point of these two contrasting approaches to division, though a given word may be pressed into the service of either approach on suitable ontological assumptions. 17
METHOD
exercise in extensional set theory or intensional analysis.11 I argue below that division needs to be understood as an exercise in the clarification of true beliefs which can establish knowledge. 12 It is inquiry conceived as the restructuring and clarifying of understanding. 13 That, as has been observed,14 its 'logic' raises new questions at each step (the interlocutors having to decide in which branch of the division their quarry is lurking), goes to show that the logic of proof is not what division provides. It relies not only on the prior grasp of resemblances, but also on a grasp of what are 'real' distinctions between kinds or individuals as opposed to specious ones. This latter point is thematised by the Stranger as the distinction between eidos and meros (262a-263b): dividers must take care not to hack off mere bits. 15 The advice rests on the fundamental ontological realism which is never relinquished by Plato, and reminds us of the beginning of the Sophist where the Stranger is instructed by Socrates (and instructs Theaetetus) that they must test which names really correspond to kinds or individuals. The Stranger's use of division has been all too often, and all too quickly, assimilated to mentions of 'division' elsewhere in Plato. Yet the bearing of his use of example on the images, analogies, similarities, comparisons of all kinds so prevalent 1
* This, to my mind, vitiates the subtle debate on the deep structure of division between Cohen and Moravcsik in Moravcsik (1973a). More useful in laying out the surface logic of the divisions is Cavini (1995). 12 Sayre (1969) 187-91 notes that Aristotelian genus-species analysis does not apply to Platonic dialectic. Gulley (1962) 114 claims that division is 'a method of classification, not of definition in terms of genus and specific differences', though definition in other terms is announced in both Sophist and Statesman as the goal. 13 I deliberately assert both sides of a contrast made in McCabe (1994) 258 n.79, that division is 'not a method of inquiry, but a system of understanding, a way of structuring our explanations'. That 'inquiry' and 'system of understanding' are not incompatible is argued below, pp. 68-70. 14 E.g. Gulley (1962) 112 on 'the additional power' required at every stage 'of recognising and selecting "real" resemblances and differences' [for him, between Forms]. The role of example, it will be argued here, is precisely to facilitate such recognition and selection in the working of division. 15 It is impossible not to acknowledge the resonance here of Phdr. 265c, which prescribes 'dividing things by kinds (kat'eide) where the joints are, and not trying to break off any part (meros) in the way of a bad carver'. This procedural advice applies equally well in prescription and practice. 18
INTRODUCTION
throughout the dialogues has gone relatively unexplored. Plato's most famous protagonist, Socrates, is famous for comparing familiar arts - horsemanship, medicine, shoemaking with the disputed knowledge of virtue he presses interlocutors to define. In the architecture of many dialogues, as in the give and take of Socratic argument, comparisons are often fundamental. Essential to the Republic is the comparison between the structure of the city and the structure of the soul; central to the Theaetetus is the comparison between the structure of belief and first a wax block then an aviary; the Gorgias makes unflattering analogies between cookery and rhetoric, and so on. Comparisons are indeed so prevalent in the dialogues in which Socrates is the central figure that they seem largely to have escaped philosophical scrutiny. Certainly Socrates himself (as portrayed by Plato) does not typically stop to inquire into the validity of comparison as a form of argument. His mode of argument is to fire off analogies and examples so as to press conclusions on the basis of an apparently obvious similarity. Occasionally Socrates himself or another figure does briefly worry about the notion of resemblance, which after all (as we shall see) is at the heart of any comparison. Resemblance is called into question in the Protagoras' doubts about similarity (33idi-e4), in the Parmenides' regress of likeness (I32di-i33a7; cf. 147-8), in the Philebus' debate over whether all pleasures, and all forms of knowledge, must be like one another (12c-14a), in the Sophist's warning about the slipperiness of resemblance (23ia4-bi). In the Statesman example (paradeigma) itself is used to secure relations of resemblance. The process of establishing genuine resemblance by means of example is laborious and slow. It contrasts, I shall argue, with the rhetors' habit of appealing to paradeigmata whose practice in this regard appears very close to the rapid-fire reliance on resemblances characteristic of Plato's Socrates. 16 In this respect the Statesman works out carefully 16
The classic study of 'paradigme' in Plato, Goldschmidt (1947), casually subsumes a variety of Greek terms under that French one, and treats the Sophist's 19
METHOD
what is characteristically glossed over or exploited in the Socratic dialogues.17 One apparent obstacle to the present argument is that the Statesman, which makes much of examples both in analysis and in use, begins its inquiry without one. Claiming to follow the same method used in the successful inquiry about the sophist, the Stranger plunges into a series of divisions at the outset of the Statesman. But in doing so he does not - as he had done in the Sophist - choose an example for practice first. These early example-less sallies in the Statesman end largely in failure, and it is in order to rescue the inquiry that example is introduced in the middle of the dialogue, heralded as important and explained at length. So the delay of example in the Statesman serves to highlight its importance. In contrast, the Sophist starts off with an example straight away, but does so almost casually, even slyly, without ever reflecting on what examples are or why they are important and reliable.18 Contrasting tactics - in the Sophist, the quiet pervasiveness of the example; in the Statesman, its belated but ballyhooed introduction - both serve to emphasise the significance of example in inquiry. The discussion falls into four further parts, two on use and two on analysis. The first considers briefly the use of example and division in the Sophist, in order to set the stage for the second, their uses in the Statesman. The third offers an analysis of the Statesman's own analysis of example, and the fourth part concludes by analysing the workings of division and example in relation to philosophical issues of paradigms and rhetoric. Though separate analyses of example and divi-
17
18
and Statesman's use of example as no different in principle from other Platonic comparisons. If this could be ammunition for those wanting the Stranger to 'correct' or advance beyond Socrates, it could also be read as providing support for (through analysis of) his characteristic practice. Cf. Benardete (1984) Vol. in. 103: 'it is characteristic of the Sophist that the connection between its action and its argument is at the most allusive. It never becomes a thematic part of the dialogue. It seems to be just a coincidence ... that the angler, chosen as an example to illustrate the Stranger's way, turns out to be the first model for the understanding of the sophist. In the Statesman, on the other hand, the Stranger almost obsessively harps on their [sic] own doing . . . ' 20
THE SOPHIST S USE OF EXAMPLE AND DIVISION
sion are made, and they do sometimes appear separately in the dialogues, the fundamental thesis of Part i remains that example and division in Sophist and Statesman are, when used most successfully and for the most difficult tasks, harnessed together in intimate tandem. The Sophist's use of example and division In light of the general task set him by Socrates and discussed above, the Stranger embarks on a question-and-answer inquiry about sophistry with young Theaetetus. He begins by remarking to the boy: ES:
. . . as yet you and I have in common, about [the sophist], only the name (onoma); but as to the thing (ergon) to which we give the name, we may each have a conception of it in our own minds. Yet we ought always in every instance to come to agreement about the thing (pragma) itself by definition/argument (logos), rather than to agree about the name alone without definition (So. 218CI-5). 19
Three elements are in play in this statement: the name, the thing, and the definition or true conception of the thing (logos). Having agreed to investigate the sophist the interlocutors both know this name, but they have no guarantee that they are each privately defining it in the same way, nor do they have any guarantee that their individual definitions are sound. The goal of the inquiry is to attain agreement between the two interlocutors about the definition of the thing, to replace the literally nominal agreement from which they begin. Notice however that names are not wholly disparaged. This point will be explored further below. Having defined the goal of common agreement by argument about the 'thing', sophistry, the next question is how to proceed. The Stranger immediately suggests that they begin by choosing an example. He asks, 'shall we take some lesser thing and try to use it as an example (paradeigma) for the 19
Campbell (1973) 13 n.2 succinctly observes: 'ovoiaoc = the name, is distinguished on the one hand from epyov or irpayija, the thing, and on the other from Aoyos, the definition or true conception of the thing . . . The union, TO irpayna OCUTO 5ia Xoycov, is opposed to the mere name, TO ovopa novov x w pis ^oyou.' 21
METHOD
greater?' (So. 2i8d8-9). He justifies the suggestion by observing that a 'minor' 20 and 'trivial' example will be easier to work with than the greatest subjects (So. 2i8di, d8). Example so far implies mere practice on something trivial before the real show begins. But the constraints which the Stranger quickly places on a suitable example suggest that something more is at stake. Any example chosen must be (i) 'familiar', allowing the inquiry to get off the ground; (ii) 'minor', reinforcing the opening idea that nothing too serious is at stake; and (iii) 'having no less a definition21 (logon) than any of the greater things' (So. 2i8e2-3). Though 'minor' plays down the importance of the example, the requirement that it be definable just as the 'greater things' exemplified are, begins to indicate the nature of its significance. The indication is strengthened when the Stranger suggests an example - the art of angling and its practitioner. Having checked that the angler indeed meets the three constraints (So. 2i8e4~5), the Stranger announces his twin aims in using this example: I hope he offers us a method (methodon)22 and a definition (logon) not unsuitable to our purpose. (So. 219a 1-2)
Notice the careful articulation of this hope. The definition of the angler is linked in expectation to the purpose of the inquiry (showing that to have read the opening words about example as implying an isolated practice-run was to have misread). And the hopes placed on the example of angling include not only a relevant definition but also a useful 'method'. The example is not inert terrain for a quick and easy trial of an 20
21 22
I translate smikron as 'minor' rather than the usual 'small': what is meant is not that angling and sophistry differ in size, whatever that would mean, but in importance and difficulty for purposes of definition. Smikron is glossed by phaulon at d8 (here, 'trivial'). The latter term is somewhat provocative; compare the Hippias Major, where Hippias is affronted by Socrates' 'trivial' examples of pots (288CI1-3) and ladles ( 2 9 ^ 3 - 4 ) . Logos as 'definition' is supported by the context of ngsn. Vlastos (1994) 1 n.5 notes that Plato coined methodos in the Phaedo to flag the distinctive 'method' of the elenchus (Phd. 79e3, 97b6). Its appearance here flags the introduction of an equally distinctive method (of division, introduced in conjunction with example). 22
THE SOPHIST S USE OF EXAMPLE AND DIVISION
independent method. If the Stranger's hopes are fulfilled, the angler-example will provide23 the method for the main inquiry, and the definition of the angler will also somehow prove relevant to the definition of the sophist which is the main task in hand. Let us see how the search for angling works. The Stranger begins by ascertaining with Theaetetus that the angler as such possesses a characteristic expertise - the assumption is that it is this expertise which defines him and even lends him its name. In the Sophist, the word techne is used to denote the expertise of the angler, as well as the many kinds of expertise evaluated in the course of the divisions. Possession of such an expertise defines its practitioners, particularly in Plato. 24 Techne in Greek describes a cognitive grasp giving the systematic and reliable ability to act on some part of the world. Its connotations are strongly positive,25 the technai (plural) considered so crucial to human cultural achievement that they are often imagined in myth and drama 26 as gifts from the gods. The term is used consistently in the Sophist to describe the kinds of expertise encountered in the divisions, and intermittently in the Statesman for the same purpose. Even though the two terms seem to be used interchangeably in the Statesman,21 the fact that the more wholly intellectualist episteme is missing from the Sophist may cast a faintly pejorative shadow back on the status of sophistry. 23
24
25
26 27
D a v i d Sedley has objected that the example m a y simply introduce here a m e t h o d which has been independently elaborated elsewhere (perhaps in the 'prescriptive' discussions of division by Socrates). But the embedding of division in example here is n o t arbitrary; it proves essential to the progress of the inquiry in a way which n o prescriptive discussion of division suggests. The context of example is constitutive of w h a t division is for this inquiry, a n d in the context of the Stranger's conversations originates the mention of division. R o w e (1995) 2 - 3 observes that Plato makes characteristic use of adjectives (also m a d e into substantives) ending in -IKOS for the particular kind of specialism (e.g. TTOAITIKOS), a n d substantives ending in -IKT) (e.g. &pi0|jr|TiKT)) for the activity. T h e association of techne with universality, teachability, precision, and concern with explanation is surveyed in N u s s b a u m (1986) 9 5 - 7 . A n d in the story told in the Statesman itself: see Part 11. R o w e (1995) 178, note to 258d5, observes that in the Statesman 'the two terms will be used interchangeably throughout'; he cites Gill (1995) 294 n.15. Neither remarks the discrepancy on this point between the two dialogues.
23
METHOD
Because both sophistry and statecraft are treated as forms of expertise, the examples chosen to discuss each one are themselves forms of expertise, and they are defined in terms of their similarity to and difference from other such forms, progressing in each case toward the isolation of a single expertise. These divisions of forms of expertise are described indiscriminately in terms of the forms of knowledge themselves and the persons who possess them. 28 Assured that the angler does have an expertise, the Stranger asserts that there are two kinds (eide) of all forms of expertise (2i9a8). Each kind must be named, and this is done by enumerating various arts belonging to each kind, until a sufficient sense of identity emerges for the kind to be properly named. First are listed a number of kinds of expertise 'all of which might properly (dikaiotata) be called by one name [productive-art 29 ]' (2i9bi-6); there follows on the other side of the division another list of forms of expertise, concluding 'all of this part of expertise might properly (diaprepseieri) be called acquisitive-art' (219C6-7). In both cases use of a normative term signals that these names are not just convenient or familiar, but correct. Theaetetus is then asked to judge in which of these two named kinds angling belongs. His choice of acquisitive-art (2i9d3) establishes it as the candidate for the next division-cut in which the same procedure will be repeated. This first completed step of division models all those which, though compressed, will follow. A single kind of expertise (in the first case, expertise as such) is split into two or more 30 28
29
30
T h e original topics are the personages (So. 2i7a3) and in the Sophist in particular the person is mentioned m u c h m o r e often t h a n the expertise (e.g. So. 2 i 8 b 7 , C], Q4), but the divisions are clearly conceived as divisions of the technai which certain persons m a y possess (e.g. So. 2i9a8). N o translation of the Sophist has been done on Rowe's principle of rendering techne as 'expertise', although its shadier connotations m a y sometimes be m o r e appropriate there (in the absence of episteme) t h a n in the Statesman. I have modified the L o e b translation of the Sophist (Fowler (i 921)) to suggest 'expertise' in some contexts, in a n attempt to preserve b o t h the commonality and the difference in this usage between the two dialogues. N o w h e r e is division required to be dichotomous, though it is so practised in the Sophist and in the first half of the Statesman (the 'shepherd-divisions', below); any
24
THE SOPHIST'S
USE OF EXAMPLE AND DIVISION
kinds. Sometimes these are named immediately; more often a few of each kind are enumerated until an appropriate name can be decided upon. Then one of these kinds is identified as including the target, and made the subject of the next division. Though we tend to speak of the 'division of angling', the division is rather of all the kinds of expertise which include angling until the latter is sufficiently isolated to be clear. The names of those divided along the way themselves are linked into a definition of the target, as will be seen below. One qualitative feature of division which the Stranger highlights, if discreetly, is the giving of names.31 At many points the normative dimension of names is reiterated, for example: ' . . . is not that genos worthy (axion) of being called by another name?'. (So.
' . . . since there is . . . one art involved in all of these operations, let us honour it (axiosomeri) it with one name'. (So. 226C5-6) ' . . . we must examine further and see whether [education] is all already indivisible, or still admits of division worthy (axian) of having a name'. (So.
The last quotation reveals most clearly a fundamental link between division, names,32 and value. A division is most characteristically accomplished when the separation is accompanied by the awarding of just those names which are deserved and appropriate. Conversely, some things which turn up in the divisions do not deserve names. The Stranger remarks in one case that 'all that must be considered one eidos,
31
32
such restriction is even disavowed later at St. 287C2-5, though it is conceivable that the failure of the early divisions is linked to their rigid bifurcations. Cherniss ((1962) 46-7) insisted on the inapplicability of Aristotle's criticism of dichotomous division to these two dialogues. A number of other passages indicate the giving of names as a crucial step in a division, although not drawing an explicit connection with worth: 'by what name shall we say this ought to be called?' (So. 22ia4); 'some such name as violent... what other name . . . ' (So. 225a9~bi); 'let us try to tell the name by which we must call each of these' (So. 225d4~5); 'now what name is to be given to that part of instruction which gets rid of this [stupidity]?' (So. 229CH-12, with C8-9). AeT in some of these examples plays a normative role as well. Moravcsik (1973c) recognised the prevalent attention to names in the workings of division; cf. Loriaux (1955) 158 who cites fifty-four passages in the Sophist and Statesman using terms such as prosagoreuein, eponomazein and their cognates. 25
METHOD
but it received no name from our predecessors, nor is it now worthy (axion) of receiving one from us' (225C2-4). Just how this line is drawn - as to which names are worthy and appropriate, which unworthy and to be neglected - will be discussed below. The process of division requires that once sufficient names have been awarded to the two or more new branches, the interlocutors must then agree as to which kind the target belongs. That chosen kind is then itself divided and the process repeated until the target is finally isolated. A definition of the target is then woven 33 out of all the names of the arts which have turned up in the relevant branches of the division (268C5-6), names which have been clarified by the process of argument. The result in the case of angling is recapitulated by the Stranger as follows: Now, then, you and I are not only agreed about the name of angling, but we have acquired a sufficient (hikanos) definition (logon) of the thing itself. For of expertise as a whole, half was acquisitive, and of the acquisitive, half was coercive, and of the coercive, half was hunting, and of hunting, half was of living things, and of living things, half was water hunting, and, taken as a whole, of water hunting the lower part was fishing, and of fishing, half was striking, and of striking, half was barb-hunting, and of this the part where a blow is struck by pulling from below upwards at an angle has a name in the very likeness of the act and is called angling, which was the object of our present search. (22ia7~b3)
Notice the double-edged emphasis on naming in this summary. On the one hand, the Stranger is consistent with his opening remarks in unfavourably contrasting (mere) agreement about a name with (real) agreement about a definition (logos) of something. On the other hand, the final identification of the art of angling involves naming it - even though the name, descriptive of the world as it is, would not have sufficed without the full definition. The definition establishes the identity of the expertise in question by relating it to the 33
'Woven' here is my own term, but it echoes the Stranger's call to 'weave' the logos of the final definition of sophistry later in the dialogue (So. 268C5-6). See below, pp. 38-9, for further remarks on this term which is, in the light of the Statesman's use of weaving, pregnant with meaning. 26
THE SOPHIST S USE OF EXAMPLE AND DIVISION
context of other forms of expertise; both similarities and differences are acknowledged in the course of the definition. Angling having been chosen as the example for sophistry, no recursive choice of an example for the example seems necessary. The division just summarised works - without an example - to define angling successfully. The early divisions of the Statesman also proceed without an explicit example, but less successfully, and example has to be introduced to save the inquiry. Perhaps one effect of the angler being 'minor' and 'trivial' is that its division - itself an example - can dispense with example in reaching an adequate definition. In any case the combined method of example-division is a condition for success only in the case of the 'greater things' which are the main targets of inquiry in both Sophist and Statesman. Example and division can certainly be employed separately, and division is used alone when the target-art is itself an example for something more important. But only the combined method proves successful in the case of the 'greater things'. Let us now effect the comparison between angling and sophistry as is done in the dialogue. Having completed his definition of the example, the Stranger begins again by asking whether the sophist, like the angler, possesses an art. He answers himself that the sophist does indeed have 'some art' (22id). This reply allows division to proceed according to the method introduced by the example, and we will consider the way this works in a moment. First the oddity in the reply must be explored. In the Gorgias, sophistry is decidedly not a genuine art, being only a counterfeit of statecraft (G. 465C2-3), and this attitude is wholly characteristic of Plato's many discussions of sophistry. Yet the entire architecture of the Sophist's divisions depends on treating sophistry as an art, which is the literal if not performative impact of the avowal 'some kind of art'. It may be that the instability and lightness of the Sophist's divisions reflect, indeed are designed to highlight, the hovering of sophistry on the frontier between art and not-art. That division produces six slippery and overlapping definitions of sophistry (though concluding with a seventh which claims to capture the 'being' of the sophist 'most 27
METHOD
truly'34), on this reading implies more aspersion on sophistry than on division.35 Despite these possible shadows, Young Socrates is willing to agree that the sophist does indeed possess 'some kind of art', perhaps lulled by the very word sophistes which is linked to the word for 'wise' and, as Rowe36 points out, before Plato could mean simply 'expert'. The Stranger now exclaims: ES: THT: ES:
Good gracious! Have we failed to notice that the man is akin (suggene)31 to the other man? Who is akin to whom? The angler to the sophist.
THT:
HOW SO?
ES:
They both seem clearly to me to be some sort of hunter. (22id8-e3)
Kinship between the angler and the sophist takes us very far indeed from any early (mis-)reading of the example as a trivial practice-run unrelated to the real task in hand. The example cannot be a mere arbitrary illustration if it is indeed akin to the target of sophistry. And this says something important about the way 'expertise' is here being understood. To say that the two men are both hunters of some sort does not mean that the sophist puts on Wellington boots and seizes a fly-rod in order to hunt like the angler. Rather the kinship inheres in certain abstract patterns of orientation and confrontation which may be acted out with very diverse, as it were, props, costumes and scenarios. A hunter offishand a hunter of souls are both hunters not merely metaphorically. As kin, both are hunters under the skin. The example and its 'kinship' hooks the inquiry about sophistry onto the third division-cut made in the course of 34
35
36 37
The final definition at 268c8-d4 is described by the Stranger as what they must say to be, literally 'the being of the sophist, most truly, as it seems' (TOV OVTCOS j n . . . TaAr|06crraTa, cos eoiKsv, epeT). Further aspersion may be implied in the choice of angling as the example for sophistry; as Mary Hannah Jones once remarked to me, angling is outlawed as bad for moral education in the Laws (823d-824a). Contrast the contemporary cultural status of weaving (the example for statecraft), discussed below, pp. 164-5. Rowe (1995) 2 - 3 . The composition of ovyyevsia - 'together' + yevos - already hints at the fact that six of the definitions of sophistry will derive from kinds identified in the definition of angling. 28
THE SOPHIST S USE OF EXAMPLE AND DIVISION
defining angling. Recall from the recapitulation quoted above: T o r of expertise as a whole, half was acquisitive, and of the acquisitive, half was coercive, and of the coercive, half was hunting . . . ' Both anglers and sophists belong here as kinds of hunters, and both are, according to the fourth division-cut, hunters of living things. To this point the two kinds of expertise (qua their experts) share a common logos (222a2-3). But the sophist is now sought in the previously neglected branch of the fifth division-cut. If the angler hunts living things in water, the sophist hunts those on land (222a9-b3). From this point the divisions diverge, the one path leading to a definition of angling, the other to a definition of sophistry. Indeed, six of the seven divisions by which sophistry is eventually seven times defined begin, like this first one, from some nodal art shared with the division defining the kindred art of angling. None of the six divisions 'kin' to angling challenges the original distinction between productive and acquisitive arts, and indeed all of them remain like angling within the 'acquisitive' kinds of art only. The following table displays the starting points of the six kindred divisions in relation to the definitional division of angling:38 Beginning subkind From Land222a2ff. hunting Exchange 223c6ff.
Div. ist
Kind(s) shared with angling Acq.-Coerc.-Hunt.-[not Water but-]
2nd 3rd 4th 5th 7 th
Acq.-[not Coerc.but-] 39 [as second - adds settled retailers] [as third - adds retailers of own wares] Acq.-Coerc.-[not Hunting but-] Fighting Acq.-Coerc.-Fighting-ControversyArgumentation-Disputation 40
38
39
40
22461-4
T h e analysis here is restricted to the procedural features of the analysis of angling. F o r one reading of its political significance, see Benardete (1984) Vol. 11.77-99. The Stranger misremembers his own example here: he actually contrasts exchange with hunting, forgetting that in the case of angling, exchange h a d been contrasted with coercion (of which hunting was one part; 2 ^ 4 - 7 with e2). N o t h i n g hangs o n this alteration, however, a n d in light of the five other careful back-references to angling, it seems prudent to judge it merely a n imprecision. Cf. 5th division, 225C9. T h e subject m a t t e r taught by teachers of disputation is said to be 'images', b u t the Stranger imagines that the sophist might deny this
29
METHOD
Some comment on the mysterious sixth division is in order. The sixth division, which is the only one defining a 'noble' and admirable sophist is also the only one without any roots in the dubious art of angling. It begins from another 'track' (ichnos, 226b2). As an exception, this one would seem to prove the 'rule' discussed above, that multiple definitions of sophistry derive from its ontological instability. Only this 'noble sophist' 41 escapes kinship with the angler and his multiplicitous relations, and thereby is also kept aloof from too close an affinity with the six other depictions of sophistry. Yet this isolated division, rejecting the example of angling, nevertheless forges its own use of examples marked by the same word as in the overall inquiries of Sophist and Statesman (paradeigmata, 226ci). The Stranger mentions expressions connected with household arts, citing 'sift', 'strain', 'winnow', 'separate', 'card', 'comb', and 'beat the web' (226b2-io). All these, one might observe, are the kind of 'menial' and 'minor' activities which a paradeigma was required to be (see above). 42 The exchange goes like this: THT: ES: THT:
ES:
Why do you use these as examples (paradeigmata) and ask about them all, and what do you wish to show in regard to them? All those that I have mentioned imply a notion of division (diairetika). Yes. Then since there is, according to my argument, one art involved in all of these others, let us give it one name . . . the art of separation (diakritikeri). (226c 1-8)
Notice that it is not the Stranger, but Theaetetus, who calls the assembled activities paradeigmata. Clearly, the assembled list does not bear the internal, articulable structure of the
41
42
definition of himself by denying that images are possible. Thus this seventh division occasions the great investigation into Being and Not-Being, which enables it to be resumed and completed successfully at the very end of the dialogue. Sayre (1969) 151-4 takes the 'noble sophist' to be wholly admirable and identifies him with Socrates; contrast Cornford (1935) 182 for whom the sixth sophist's evocation of Socrates highlights the negative features they share. They are also vocabulary used for weaving, a choice which seems to allude to the Statesman's central example of weaving and so to reinforce this sixth division's links to example. 30
THE SOPHIST S USE OF EXAMPLE AND DIVISION
paradeigma of angling, as worked out and worked into the analysis of sophistry. These are garden-variety examples, instances cited to make an immediate point, in a mini-method perhaps more like Socratic analogy than the carefully divisible example which the Stranger has been articulating. Yet Theaetetus seems to have picked up the connection between example and division in the conversation so far, if not yet appreciating the full import of the analytical use of example, and he underscores the link by introducing it into just the section of the dialogue from which it would otherwise be missing. And the Stranger responds to his observation by making a provocative connection between division and example in the content, as well as the method, of this sixth division. The examples make manifest the core activity of separation, a common thread tying household sifting to the Stranger's own method of division in aid of definitional inquiry. The moral that division's root act is separation, as well as the playfully doubled link between division and example, will both be emphasized in the Statesman. In the course of the seven divisions of sophistry as in the exemplary division of angling, names are invoked many times - sometimes positively as when the proper name for an art is sought, other times negatively when the Stranger urges Theaetetus not to be overly distracted by names, or advises that a certain art is unworthy of being named. The clue to reconciling these seemingly opposite attitudes toward names is a remark made in the course of the fifth division. When Theaetetus requests a name for all the arts which purify the body, the Stranger - knowing that their quarry lies on the other side of the division, in the arts which purify the soul - replies: ... it will not matter at all to our method what name soundsfinest;it cares only to unite under one name all purifications of everything else and to keep them separate from the purification of the soul. (So. 227CI-5)
The name for the purifications of the body is instrumental, designed to achieve a single task: to group these purifications together and to keep them separate from the purifications of the soul. So this name need not be precise or elegant, only 31
METHOD
serviceable; though a misleadingly glamorous43 name would not do, there may be a number of more modest names which would. The criterion for choosing a given name is simply its adequacy to the task at hand, and this is defined teleologically in terms of the inquiry's overall goal. Sometimes, the direction of the inquiry requires careful and discriminate naming; sometimes, any serviceable name or even none at all will do. The seemingly opposed attitudes to names are explicable as two sides of the teleological project of inquiry.44 There is a further refinement to this Janus-faced account of names. Sometimes names are taken as available in the language; but at other moments there is a dearth of names, which have then to be invented. But if division can so spur new names, this implies that an insufficiency of names may be due to an earlier inattention to division. The Stranger makes this very point in the last part of the dialogue, in his return to the seventh division defining sophistry: Where, then, can the fitting (prepon) name for each of the two [kinds of imitator] be found? Clearly it is difficult, because there was, it seems, among our ancestors a long established and careless indolence in respect to the division of genon according to their eide*5 so that nobody even tried to make such divisions; therefore there is not a great abundance of names. However, even though the innovation in language be a trifle bold, let us, for the sake of making a distinction, call the imitation which is based on opinion, opinion-imitation, and that which is founded on knowledge, a sort of scientific-imitation. (26yd4-e2)
43
44
45
The rejection of what sounds 'finest' suggests that the method of division, like the story's diagnosis of example, is intended to puncture delusions of grandeur which are liable to interfere with accurate political understanding. See Part II below. Despite her focus on prescription, again, McCabe (1994) 258 perceives this aspect of division's procedure: T h e trouble with words ['word' also translates onoma] . . . is that they allow us to see divisions where none really exist - names mislead us into thinking we have properly described the nature of things, where we have not done so . . . The good thing about words, on the other hand, is that they allow us to see how the nature of things is structured . . . ' She also (forthcoming) emphasises the Statesman's concern with teleology generally, a point congenial to and supported by my argument here. Although genos/eidos here can be roughly equated with genus/species, I leave them untranslated in order to preserve the flexibility with which they are used to signal relative classifications throughout the dialogue; see above, pp. 17-18. 32
THE STATESMAN
S USE OF EXAMPLE AND DIVISION
The Stranger diagnoses the lack of names as due to lack of division. It seems that division, by laying out the arts clearly and adequately, acts as a spur for the invention or application of just those names which are dialectically required. Without divisions, names may happen to be sparsely or profusely provided, but there would be no reason to expect their provision to correspond to the nature of things. So if names are useful for division, they are useful only in certain contexts regulated by the goal of the inquiry; and they may also in these contexts be generated by division. It follows that names are valuable but their value cannot be taken on its face. They are, as it were, not evidence but tools. The name 'sophist' shared by the Stranger and Theaetetus at the outset of the inquiry was a launch-pad, but the goal of the inquiry is to use such shared names to lever the discussion to a definition of the thing in question. Thus, just as the Stranger had originally hoped, both the method and the definition provided by the example of angling prove useful for the definition of sophistry. Definition is exemplified by defining (which is done by dividing) the example. And though division is used without an example in the case of the example itself, for the more difficult and more significant inquiry into sophistry example proves to be an indispensable partner of division. Yet this close and persistent connection between example and division is never remarked upon in the Sophist after the Stranger's first quick expression of hope. The dialogue makes little fuss about its use of division, getting under way rapidly and with little reflection. It is left to the Statesman to celebrate and highlight example and the divisions which articulate it. The Statesman's use of example and division46 Kinship and method The Sophist, as we have seen, begins with an injunction to reach agreement by argument (logos) about the thing underlying a 46
Line numbers in this section and the remainder of Part I refer to the Statesman unless otherwise marked. 33
METHOD
shared name, and goes on to explore the kinship (suggeneia) between two arts which though different in name turn out to have a certain common definition (logos). The Statesman begins, even before any inquiry is launched, with some banter on these same themes. A boy bearing the name of 'Socrates' has replaced Theaetetus, his companion in mathematics and gymnastics, as the Stranger's interlocutor. The passage is paraphrased and extended by Alexander Nehamas as follows: Socrates says that while he and the younger Socrates have their name in common, it is much more important to see whether they are akin (suggeneia) to one another through logov. through conversation, we get to know others better; through definition, we get to know what is akin to what and so, strictly speaking, what each thing is. 47
The 'kinship' of angler and sophist has been revealed through argument/definition; now the 'kinship' of Socrates and the youth who shares his name is to be tested through argument/ conversation. In the former case two things with different names turned out to share a substantive similarity. In the latter case, two people with the same name may or may not turn out to share any substantive character traits in common; as it happens, Young Socrates proves himself at various points cagier, brasher, and less astute 48 than Theaetetus, and his mathematical abilities do not seem to be matched by any ethical inclinations which would liken him to his namesake. At stake, as we have seen already in the Sophisfs emphasis on names, is the fundamental question of whether the likeness or difference indicated by shared or different names is real. Young Socrates' first task is to agree that the statesman, like the sophist, must be defined as one of those possessing a form of expert knowledge 49 (258b3-4). This is structurally 47 48
49
Nehamas (1984) 31, summarising St. 257di-258a3. Cagier: he gives only cautious approval to the need for a new division, and tells the Stranger that the search is really his (the elder's) alone (258C2, C9-10). Brasher: he makes the unscientifically reckless division between men and beasts, for which he is chastised (262a3~4). Less astute: although described as a promising mathematician like his friend Theaetetus (Tht. i47C7-i48b4), he has so little understanding of politics as to have to inquire as to the most basic moral of the story (271C4-7; cf. Part 11, p. 107). Episteme is used here, whereas techne was the word used at the corresponding point in the Sophist; see n.7 above. 34
THE STATESMAN'S
USE OF EXAMPLE AND DIVISION
identical to the question which was twice used in the Sophist to begin a division of technai, and it plays the same role here: 'in that case we must divide the kinds of knowledge (epistemas) as we did when we were considering [the sophist]?' (258b6-7). Besides the substitution of episteme for techne discussed above, this adoption of the Sophist's method involves one crucial adaptation. As we have seen, the search for the sophist began with the choice of an example, which then provided the method of division and a definition relevant to the main inquiry. Here at the outset of the Statesman, no example is chosen. With the method of the Sophist fresh in mind, no example appears to be needed to set the divisions going. But this appearance will prove deceptive. In this silent alteration a crucial aspect of the method in the Sophist goes missing and, for the moment, unremarked. The Stranger's instruction to divide 'as we did' in the Sophist, however, has an unexpected consequence. Though they engage in division in both cases, they begin by dividing differently. In the Sophist the first cut divided productive-arts from acquisitive arts. In the Statesman the first cut divides practical arts from theoretical arts. No argument is given in either dialogue for any extensional identity between these two basic divisions. Indeed, a later division of theoretical-art yields a group of arts which issue their own commands, and only this sub-division of theoretical art is said to give commands for the sake of production, that is, producing things ( 2 6 i a n b2). So basic, and unremarked, a difference in the two opening divisions of expertise must lead us to inquire into just how the divisions map onto reality. The idea that the divisions map reality was introduced by Skemp thus: The notion of a 'world' of Forms, i.e. of a Reality which is a kind of area occupied by Forms and susceptible of a dialectical Ordnance Survey, is fundamental [to the dialogue]. Skemp (1942) 74 n.3
Skemp's reading of course bears on something much more significant than merely the idea of a map: the notion that the Statesman's divisions divide or otherwise involve Platonic 35
METHOD
Forms. This question has been commented on above (pp. 1618). But the image of a map also requires consideration. Ordnance Surveys are normally thought of as unique and definitive, the complete account of a topographical area. So to think of them, however, is already to assume that one knows and shares the purpose for which they are made. They appear unique and definitive insofar as there is tacit agreement on the features which a comprehensive map should have (marking elevation, declination, and so on). But an oil surveyor might well have inquiries in mind which an Ordnance Survey map would not meet. The point is that maps are relative to the purposes of inquiry, and no map is even in principle definitive for every inquiry: a map suitable to the oil surveyor would not serve a motorist in haste. Skemp's appeal to mapping captures one important feature of the Statesman: the way that names have there to be tempered and even invented to correspond to the relevant features of a reality not otherwise defined, but everywhere assumed. But it is equally important to observe that the Sophist and the Statesman, from the outset, present radically different maps, which like all maps are shaped by the purposes they are made to serve. If division maps reality, it does so precisely in being neither exhaustive nor fixed in the distinctions it makes, names it gives, and definitions it constructs. And so long as a sufficient definition of the target is achieved, the divisions made along the way may be as sketchy and provisional as the interlocutors like. The teleological aim of the inquiry governs the degree of completeness it is compelled to attain as well as the very content of the definitions reached. By 'teleological' is meant the literal fact that each inquiry aims at a telos: definition of sophist and statesman, and thereby (in part) resolution of Socrates' initial query of the Stranger. Each such telos puts a 'teleological' constraint on the unfolding development of the discussion. Certain words characteristically reflect this constraint in the text of the Statesman in particular. The most notable are saphes for clarity and teleos for completeness; the latter, of course, implies completion precisely as the realisation of the telos. 36
THE STATESMAN'S
USE OF EXAMPLE AND DIVISION
Repeatedly the Stranger checks himself or the boy, or urges them both on, by observing that their clarifying task is not yet complete. 50 What must be completed is not a map of the entire set of the arts in their interrelation, but rather the definition of the target (and even that itself must be complete only in relation to answering Socrates' original demand as to whether sophist, statesman, and philosopher are one, two or three). The goal is a portrait of a single art, sufficiently clear and detailed for these purposes, 51 not a panorama of them all. It follows that such clarity about the other arts which the divisions may achieve is either necessary for the inquiry or an unintended bonus. Consequently, no special effort is made to ensure that these other arts are completely characterised, or adequately distinguished from one another. And this teleological framework engenders the double-sided practice of naming which has been noted above. This feature of inquiry is most broadly formulated early in the Statesman: ES:
So in what direction will one discover the path that leads to the statesman? For we must discover it, and after having separated it from the rest we must impress one character (idean) on it; and having stamped a single different form (eidos) on the other turnings we must make our minds think of all kinds of knowledge as being two forms (eide). (258C3-8)
This proposal echoes one made in the Sophist for the division of purifications: put the one of interest (the target) by itself, and all others in another class. The teleological character of the inquiry is evident here. Since the aim is to define statecraft, all other knowledge must be understood as in some way other-than-statecraft.52 There follows a certain 50
51
52
ou . . . TravTairaai . . . TeAecos, 267CI1; not yet 6Aov or crocks, 275a5; TEAEOV, 28oe6; n o t yet craves or TeAsov, 2 8 i d 2 . Contra Sayre (1969) 216 w h o insists that '[a] definition of the sort which the Sophist exhibits is adequate if it formulates b o t h necessary and sufficient conditions for being the kind of thing defined'. The divisions offer a notion of 'sufficiency' relative to the purposes of the inquiry instead. C o m p a r e Ackrill (1970) 384: '[Plato] is not here seeking to bring to light the structure of a whole genus, b u t to achieve a definition of a particular species. F o r
37
METHOD
53
provisionally in the discussion and divisions of all those other forms of knowledge. If we were defining carpentry we would presumably have to establish a kind of 'knowledgeother-than-carpentry' in which statecraft would now be included along with all other forms of knowledge; a changed telos would change the very classifications employed. It is important to observe that such lopsided dichotomies are not presented by the Stranger as mere way-stations, a first step in what will be a fuller and more complete exposition. Rather, the formation of these two complementary but so different kinds is demanded as an onerous intellectual discipline. To define something requires 'bringing one's mind' to consider it as something distinctive and set apart from all other things of its erstwhile kind. To recognise an eidos is a constructive intellectual achievement, one from which the purpose of the inquiry - that is to say, the purpose of the inquirers - is ineliminable. If the results achieved by division are provisional and relative to the aim of the inquiry, in what sense can they be said to be definitions? Both dialogues 54 claim to reach definitions by linking, or 'weaving' together the names of the forms of expertise identified in the path of the division leading to the target. Yet it may be objected that if the division paths are provisional and multiple, surely they cannot constitute a proper (presumptively unique) definition.55 The point is well
53
54
55
this purpose the important thing is at each stage to hit on the relevant sub-genus of the superior genus; the irrelevant sub-genus can be thrown away - and it doesn't matter if [tjhere are some other (irrelevant) sub-genera we have not mentioned.' But Ackrill takes this specially to justify dichotomous division, which it need not do. T o use this word t o express this thought is t o gloss a remark by Campbell (1973) xi: ' . . . Plato employs division by exclusions precisely as a provisional expedient [emphasis original]. His object is n o t the classification of m a n y objects b u t the definition of one.' So. 2 2 i b - c defines the angler ('not only the n a m e . . . b u t the definition (logon)') in this way; So. 2 6 8 c - d does the same for the sophist, claiming to 'weave' (sumplexantes) the n a m e by so doing. St. 2 6 y a - c does the same for the shepherddivisions of the statesman, though this is soon criticised as misleading a n d incomplete (see below, p p . 4 3 - 6 a n d Part 11) a n d the final definition of the statesman is n o t given by such reverse summation. Cherniss (1962) 47 rightly notes the provisionality: the stages of division are imp o r t a n t 'rather as a safeguard t o insure the right direction of the search (Politicus
38
THE STATESMAN
S USE OF EXAMPLE AND DIVISION
taken only in so far as it highlights the fact that definition conceived as a unique verbal formula is not the dialogues' aim. The call for 'definition' is better understood as signalling the need for, as Annas puts it, 'expert knowledge of what one is talking about'. 56 And this is entirely consistent with a focused and limited notion of what one is, indeed, talking about: a notion focused and limited by the goal of the inquiry and so provisional in relation to a comprehensive understanding of the general field. Definitions are, in these dialogues, presented as provisional insofar as they are relative to a given aim of the inquiry; they are demands thrown up by the course and direction of the argument. As remarked above, the Statesman embarks on this inquiry without choosing any example to govern it. We should not be surprised that the subsequent attempt to practise division in the absence of an example is not really successful. Division of theoretical-art continues until the Stranger invites a conclusion: 'let's go back to the beginning, and gather together (suneirdmen) from there to the end our account (logon) of the name of the expertise of the statesman' (267a4~6). The definition reached is, roughly, the statesman as giver of his own orders to the human herd (267a7-c4). The mere 'gathering together' of this first definition may 57 warn that the definition has failed to attain the requisite unity, indicated elsewhere by the language of weaving (sumplekein). Whether or not the semantics suggest failure, failure is confirmed when the Stranger rejects their definition as flawed, not in all ways 'sufficient' (hikanos) or 'complete' (teleos) (267c8-di). This is the signal for introduction of the story of the divine helmsman. But as that story is the topic of Part n, the task here is to examine the failure of this first definition more closely.
56 57
262B) t h a n as representative of the necessary ingredients of the idea, for "longer" and " s h o r t e r " roads m a y lead to the same conclusion {Politicus 265A, 266E)'. But he concludes, wrongly in m y view, that division is therefore unable to attain definitions a n d is rather a mere h a n d m a i d e n of recollection (pp. 4 6 - 7 ) . A n n a s (1995) xi. T h e possibility is in the contrast of suneirein with sumplekein at So. 268C5-6; but see the cautionary remarks on the former verb in Aristotle, N.E. H 4 7 a 2 i - 2 2 , in Burnyeat (1980) 89 n.6.
39
METHOD
Shepherding No formal example (paradeigma) is adopted at the outset of the Statesman, as was done in the Sophist. Nevertheless the Statesman's first and ultimately unsatisfying foray into division is guided by the tacit introduction of the language of shepherding. 58 The role of this language is difficult to describe precisely. Certainly the language of shepherding as used for statecraft goes far beyond a simple analogy. The statesman is discussed as shepherder of the human herd, distinguished from other herders only by the identity of his flock (people, not oxen, cows or pigs). The discussion proceeds as if statecraft were straightforwardly a kind of shepherding and the various kinds of shepherding did not differ in the nature of their activity. However, this more-than-analogy still fails to become an explicit 'example'. The shepherd is never announced as an 'example', and more importantly, there is no division of the arts involved in shepherding as there was for the explicit 'example' of angling in the Sophist. Shepherding is presented as a pursuit without internal differentiation, being divided not into kinds of expertise but only into the kinds of herds it tends. And so in contrast with the Sophist, where clear-cut and explicit adoption of an 'example' enabled the divisions to proceed successfully, in the Statesman a tacit adoption of something like an example but never clearly recognised, announced or defined leads ultimately to failure and frustration. Shepherding language first appears in the course of the fifth division-cut. To recapitulate the steps of the division to that point: theoretical art is divided into calculating arts, which only judge but do not use their judgments, versus commanding arts which issue orders to others based on their judgments. Commanding arts are then divided into those which pass on others' commands, like heralds or more fancifully retail mer58
The literal term is 'herding', but 'shepherding' is more colloquial in English so long as it is understood without special reference to sheep (remembering that shepherd's pie is often made with beef). 40
THE STATESMAN'S
USE OF EXAMPLE AND DIVISION
chants, versus those which issue their own commands. A name has to be fashioned for the latter class, since this genos 'happens pretty much to be without a name of its own', and the name chosen is 'self-directing' (26oe3-6). Only this side of the division is named since it is only in this class that the 'class of kings' is to be placed (26oe5~6). Once again, the purpose of the inquiry determines a doubleedged practice with regard to names. The 'self-directing' arts are given a name even though it has to be invented. But of the other half of the division the Stranger says, '[can we not continue] taking no notice of all the rest, leaving someone else to propose another name for them? For we set up our investigation in order to find the person who rules, not his opposite.' (26oe6-26iai) Since the ruler is not to be found among the heraldric-retailing-commanding arts, that awkward conjoined description is enough and no name need be coined for them. These self-directing arts are the ones said to be productive, to command for the sake of producing things, and they are divided into those which produce lifeless things versus those which produce living things. This division pits architect (previously a guiding model, now split off from the search for statecraft) against herder - at which point the notion of shepherding enters at last. Observe the vocabulary in which it is couched. Numerous words are put into play, with several variations of 'rearing' (trophein) but accompanied by one appearance of'caring for' (epimeleian, 26id5). 59 Young Socrates is indifferent to this plethora of possible names for the shepherd's activity: he wants simply to use 'whichever of these names crops up as we talk' 60 (26ie4). The Stranger praises him for 'not paying serious attention to names', predicting a rich future in wisdom because of it (26165-7). But his praise will soon be proved ironic. For the
59
60
These words will reappear a n d be m o r e carefully distinguished in the exaggerated portrait of K r o n o s as the shepherd of h u m a n s , in the story. Here I translate with Waterfield, in line with the ironic reading which is m a d e appropriate in retrospect (see below). 41
METHOD
boy immediately makes a hasty division between nurture of humans and nurture of beasts, an error which the Stranger diagnoses as in part a confusion about names: And to me you appeared then to think that in taking away a part you had left behind the rest as in its turn a single kind (genos), consisting of all of them, because you had the same name, 'beasts', to apply to them all. 61
Young Socrates has drawn precisely the wrong inferences from the Stranger's double-edged practice with names. 62 On the one hand, he professes indifference to names like 'herdrearing' which really do matter, names which (at least as the division proceeds) will feature in the eventual definition of the statesman. On the other hand, he is so bedazzled by his own dexterity in finding the common name 'beasts' for living creatures other than humans, that he wrongly concludes that a common name in common parlance must correspond to a real and single kind. In short, the boy makes light of names when they matter and makes them matter when they don't. He ignores the need to invent a name as a tool, and misreads another name as solid evidence. In taking his cue, perhaps, from occasions when Theaetetus and the Stranger both enjoin an appropriate neglect for names (e.g. So. 22od4, 227c 1-5), he misses entirely the way that the argument's context in the overall inquiry determines the appropriate double-edged treatment of names. In other words, the boy's error is in taking the mere existence of a name as licensing a division, forgetting that division must be employed precisely to scrutinise and revise names. A further run of divisions follows, taking the interlocutors from tame animals through terrestrial to walking ones, and then down two paths to identify humans as the kind of herd tended by the shepherd-statesman. The longer path involves
61
62
Genos as 'kind' here brings out the error better than Rowe's neutral 'class'. A n d 'beasts' brings out the fact that h u m a n s have been excluded better t h a n Rowe's 'animals'. R o w e ' s note (1995) on the passage picks u p half of this point: ' W h a t this is supposed to illustrate is perhaps primarily the danger of relying on names . . . '
42
THE STATESMAN
S USE OF EXAMPLE AND DIVISION
horns, interbreeding abilities, and a pedantic joke about the geometrical relation between the two-footedness of humans and the four-footedness of pigs; here the Stranger reinforces the equal attention which division gives to what is undignified and minor (smikroteron), a lesson which careful readers of the Sophist's introduction of the method would already have learned. The shorter path has two steps: distinguish quadrupeds from bipeds, then feathered from featherless bipeds the latter are (all and only) humans. These two different paths 63 - one starting with footedness, the other arriving there - reinforce the provisional and teleological character of the maps which division constructs. When both paths have isolated humans as the herd of the statesman-shepherd, the Stranger as seen above 'gathers together' the definition of the name of the statesman's art. He retraces the steps of the divisions leading to statecraft from the very first division into theoretical art, via the 'longer path,' and concludes somewhat awkwardly: the segment from this, a part relating to a two-footed flock, concerned with rearing of human beings, still left on its own - this very part is now what we were looking for, the same thing we call both kingly and statesmanlike. (267CI-3)
But this account, while stated 'in a certain way,' is - the Stranger fears - not complete. His main concern is that, whereas shepherds of other animals have no rivals for the multifaceted exercise of their art, the statesman-shepherd will find himself besieged by rivals arguing that they, not he, are responsible for some aspect of the art of caring for humans (267e4~268di). This criticism and its significance will be developed further in the story and the corresponding Part II here. Here, however, we may consider the question of what has gone wrong. Why did the shepherd-divisions (as I will refer to the whole set of divisions down to 267C3) fail? 63
Dorter (1994) is much concerned with the fact of two paths, and the nature of the difference between them. His desire to find a substantive point here contrasts with my emphasis on the floundering of these early divisions before the example of weaving is introduced. 43
METHOD
Their failings are interrelated and twofold: in method and in politics. Consider first the failings in method. We have already noted that this entire series of divisions, heavy with plodding humour as they are, divide only (the arts of caring for) different kinds of animals, without reference to any internal structure or differentiation of the activity of herding. The divisions fail, then, to distinguish factors internal to the art of rule. They also fail to distinguish between humans and other animals in any qualitative terms whatsoever. Only evident biological facts such as habitat, hornedness, footedness and so on are used to distinguish among different kinds of animals, and these differentiae are applied as directly to humans as to pigs and cows. Detached from the logos of an example, from the definition and kinship which shape inquiry, these dichotomies furnish no resources with which to make any meaningful comparison. The method of division-saws-example, applied to an important telos of inquiry, creates an evaluative vacuum with political consequences. To include a division of land- and wateranimals in the definition of statecraft, discussing humans as if one had first to make sure what kind of creatures they are, is to discount any internal knowledge of human goals. Instead of raising questions about human judgment, values and desires, we are subjected to a pedantic discussion which objectifies humans as just another kind of herd. The story will reject this estranging objectification, offering instead a narrative which - while still putting humans into a cosmic framework reveals the mode of self-government which is distinctive to them. In light of this coming rejection it is plausible to read the point of the alienating animal-divisions as, precisely, to make the reader uncomfortable with so 'inhuman' a discussion of politics. The very estranging64 quality of the divisions 64
Aristotle, Pol. I 3 i i b i 5 , uses allotriotes to refer to 'estrangement', a reference which - though in my view inapplicable to division generally - neatly illustrates the link between its example-less and so estranging practice and an estranging politics here (I owe this reference to Rowe (1995) 181, note to 26ia3-6). Cf. also Gill (1995) Appendix on the dialogue's method of 'defamiliarisation and theorised reconstitution'. 44
THE STATESMAN
S USE OF EXAMPLE AND DIVISION
insinuates that humans are not in fact a herd like any other, and their governance not a matter of ordinary shepherding.65 One final political implication is suggested by the shepherddivisions, this time with reference to the language of shepherding itself. The model of the king as shepherd was ingrained in Greek culture by the Iliad. Its deep resonance may account for the Statesman's success in sneaking in shepherding without having to class or employ it as an explicit 'example'. Indeed most of Plato's political discussions take up the model of shepherding at some point, though the polemical force of the model varies with each context. Republic I dissects an inherent tension (Rep. 345a-347a): do shepherds fatten their sheep merely in order to eat them (as Thrasymachus insists), or is Socrates right to believe in a genuine art of shepherding guided solely by the best interests of the herd? The Statesman's general account of ruling sustains Socrates' view that true rulers will have the true interests of the ruled in mind. But this view of ruling cannot be satisfactorily modelled in terms of shepherding. That widely accepted model of rule will be shown to lack both the internal differentiation as an art, and the special (and internal) applicability to humans, that an adequate example for politics requires. 66 Division without paradeigma, as displayed in the shepherddivisions, ends in stalemate. The definition thereby reached, of the statesman as herder of the human herd, is unsatisfying (267C5-268d4), yet the method so far offers no resources for its correction. This failure is a key juncture in the dialogue. From this point onward, new methods and new political examples will be introduced, with a rising level of self-awareness about what a satisfactory definition of statecraft would require. 65
66
Contrast the discussion in B u m y e a t (1992) 183 of the deliberate alienation achieved in the Republic by such devices as the guard-dog imagery and the stockbreeding terminology. It must be noted that the Critias a n d the Laws each appropriate the model of shepherding in ways beyond the scope of the discussion here. See Gill (1979), rebutting Owen's claim (in Owen (1965)) that the Statesman offers 'second thoughts o n politics' correcting the Republic, Timaeus and Critias. Gill argues inter alia that K r o n o s is a m o r e psychologically subtle shepherd in the Critias t h a n in
the Statesman. 45
METHOD
The story which is introduced, as Part n will show, serves as a kind of halfway-house between the shepherd model and the genuine paradeigma of weaving which will eventually guide the inquiry to an adequate definition of statecraft. Awareness of the story as an inadequate paradeigma finally puts the topic of paradeigma onto the explicit agenda of the dialogue (in contrast to the seemingly casual ubiquity of paradeigma in the Sophist from the start). Having seen how shepherding fails to function, being neither a technically nor politically appropriate example, it is best to proceed by examining how weaving - the example which proves both genuine and appropriate - succeeds in functioning. The deliberate choice of weaving as a paradeigma is preceded by a lengthy and systematic account, by the Stranger, of the nature, value and importance of paradeigmata. Consideration of this systematic account of paradeigma is postponed to the next section, so that we may first describe the way that, once introduced, an adequate paradeigma is actually used in the Statesman. Weaving ES:
. . . what quite minor (smikrotatori) example is there which we can set beside statecraft, and which, having the same activity (pragmateian) as statecraft, will help us sufficiently to discover what we seek?67 (279a7~ bi)
This question recalls almost precisely the Stranger's selfinterrogation at the beginning of the Sophist as to what 67
My translation, to highlight the references to the themes of being 'minor', 'example', and 'sufficient'. Contra Jowett and Skemp, Rowe employs 'model' to translate irapaSeiyiJia; Rosen concurs, arguing that 'model' embodies a normative sense of standard or rule in contexts where 'example' may be simply any one of a kind (Rosen (1995) 80-3). Certainly the two central TrapaSsiyiiaTa of angling and weaving are normative (cf. Rowe (1995) 201 n.to 2^jjdi-2). Yet the Sophist introduces the term in an everyday illustrative sense, and the Statesman's irapaSeiyiaa of irapdSEiyna shows children learning letters for which no word could be more normative than any other. To distinguish too sharply between these semantic notions is to obscure the fact that Greek combines them in a single term (LSJ s. v.). The main point is, as Rosen (1995) 83 goes on to say, 'we require not merely an example of a paradigm but a paradigmatic example'; I prefer the deceptively ordinary connotations of 'example' to suggest this. 46
THE STATESMAN
S USE OF EXAMPLE AND DIVISION
example to choose for sophistry. The two dialogues place parallel constraints on the choice of the example. These are summarised in the following table. Each dialogue enjoins that the example chosen must be: Sophist (2i8e2~3) (a) minor (smikron) (b) familiar (eugnoston) (c) have logos (d) share kinship (be suggene) with telos of inquiry (22id9)
Statesman (279a7~bi) (a) minor (smikrotaton) (b) ready-to-hand (prokheiron)68 (c) [take part (mews) of it] 69 (d) share activity (pragmateian) with telos of inquiry (279*7-8)
Kinship and shared activity are, I take it, parallel ways of establishing a meaningful affinity between example- and targetart. Yet there is a subtle difference between them manifest in the workings of the two inquiries. Recall that in the Sophist, all but one division of sophistry began from a 'nodal' art identified in the division of angling. Individual arts common to the logoi of example-art and target-art underpinned the comparisons. The Statesman uses no such single common art between weaving and statecraft as a launch pad. Instead, a second-order distinction made in the division of weaving (28idio) - between contributory arts (sunaitia) and cooperative arts (aitiai) - will be taken over and applied to the division of statecraft in order to relaunch it successfully. Weaving and statecraft do not, we will find, have the same contributory arts; each has a proper set of its own which parallel one another. 68
69
The two constraints of (b) m a k e the same general point: we must have some kind of cognitive grasp of the topic in order to begin the process of clarification through inquiry which will result in knowledge. In the later technical analysis of example, this initial cognitive grasp will be referred to as 'true belief ( 2 7 ^ 9 , 278b3); see p. 64 below. The specific suggestion is that they take only 'part' (mews) of weaving, i.e. woolweaving (279b 1-5); this injunction could be read as recommending a kind of weaving rather t h a n a part of its logos. But the next sentence enjoins the same procedure of division 'just as [when] we divided each subject before by cutting off parts from parts (mere merony (iq
I2
3
62 61
70-71 63,70 62 62
66,67 47 n.68, 63
227
47 n.68, 63 69, 120 64, 119, 178 n.73 70
70 n.106 70 n.106 63 62
47, 1 2 2 61,74 47 n.69 47 n.69 50 169
49 49 51
50 50 50
37 n.50 173
5i,55 51 51
5i, 54 52 52 37 n.50, 54 47 54 54 54 54 54 55 55 47 n.69 72, 126 126 125 128 128 126 129 128 129 143 127 172 131 14
16 n.6, 62, 70, 71
INDEX LOCORUM
Statesman (cont.) 285d8-9 285d9-86a4 28502-4 286a2-4 286a4~7 286a7-b2 28607-10 286CI 286d4-87a7 287a7-b2 28704-8 287C2-5 288b4-8 288C4 289ai-2, 3 289a9~b2 289C4 289c-d 29ia8-b3 29ib-c 29id-92a 292e~93a 293a6—b4 293a9~C3 29366-7
74 73
2QASL2
148, 161
2Q4a6—8 294aio-b6 294a-95a 2O4b2—6 294b8—C4 294CI 294d—e 295a2 295a9~-b2 295b—d 296c-d 296e4~97a2 297dio-e5 29763, e4 298b2 298c-e 299b 299C4-6
149 149
300b1-6 3ood6 3ooei 3ooeii~3oia3 3Oia7 301CI 3Oic6-d4
3Oid4-6 3Oiei 3O2a2-b2 3O2d7~3O3e2 3O2e6 3O3a-b 303D9-C5 30404-62 30469-11 3O5cio-d4 3O5b4-7 3O5d2~4 30562-4 3O5e8-io 30606-10 3o6e9-io 307b! 3O7b2
73,74
73 73 73
75, 161
75 75 56 57
25 n.30, 59
59 59 n.8o 59 n.8o 59 59
3O7b5
140
183 148 148 148 148 162 n.52
148
199 152 153, 154 162 n.52
153 156 155
162 n.53
154 154 158 156 156
20ieiff.
2O7d 2iod Timaeus
in
157 160
29b-c 38b3-4
159
161, 162
228
160 160 135, 142
143
142 143 n.20 135, 196 53,142^9,164,177
164 171, 172 n.70
175 171
175 171
179
Theaetetus I43a-c
157 n.43
159
3o8a6 3o8b2-8 3O9a9-io 3O9b7 30905-8 3O9dio-e3 30905-8 3ioa3 3iob2-3 310C4-7 310C7 3iodi 3iod6 3iiai-c6
150 152, 153 152
160
175
3o8d
198
139 157
3O7b9~c2 307c1-2 307C2-5 3O7d2~5 3O7d7 3O7ei-3o8d9
308C5
151 n.34
161, 162
171 172 172, 174
176 172, 176 172 n.70, 176 172 n.70
173 143 n.io
174 174, 178
179 179 180
174 181 161 181 182
178 16, 19
7 34 n.48
70 n.105
67 n.97 7
16 n.6, 45 n.66, 91, 93 n.156, 107, 109 n.18, 138-9, 157, 163^55
138 145
INDEX LOCORUM POLYBIUS
The Histories vi.9.2-3
1288
178 n.74
Electra 31
THUCYDIDES
Peloponnesian War 8.4 3.82.4
SOPHOCLES
133
132
229
170
184-5