Virtue, Happiness, Knowledge: Themes from the Work of Gail Fine and Terence Irwin 9780198817277, 0198817274

Table of contents :
David O. Brink, Susan Sauve Meyer, and Christopher Shields: Introduction1: Lesley Brown: Rethinking Agreement in Plato2: Ralph Wedgwood: Plato's Theory of Knowledge3: Dominic Scott: Justice and Persuasion in the Republic4: Richard Kraut: Plato Against Democracy: A Defense5: Susan Sauve Meyer: Self-Mastery and Self-Rule in Plato's Laws6: Verity Harte: Plato's Philebus and the Value of Idle Pleasure7: Christopher Shields: A Series of Goods8: David Charles: Practical Truth: An Interpretation of Parts of NE VI9: Paula Gottlieb: Aristotelian Feelings in the Rhetoric10: Julia Annas: 'Ought' in Aristotle's Nicomachean Ethics11: Karen Nielsen: Deliberation and Decision in the Magna Moralia and Eudemian Ethics12: John Martin Fischer: The Freedom Required for Moral Responsibility13: Allen Wood: Virtue: Aristotle and Kant14: Roger Crisp: Richard Price on Virtue15: David O. Brink: Eudaimonism and Cosmopolitan ConcernBibliographies of Gail Fine and Terence Irwin

Citation preview

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Virtue, Happiness, Knowledge

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Virtue, Happiness, Knowledge Themes from the Work of Gail Fine and Terence Irwin

edited by

David O. Brink, Susan Sauvé Meyer, and Christopher Shields

1

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

3 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © the several contributors 2018 The moral rights of the authors have been asserted First Edition published in 2018 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2017964292 ISBN 978–0–19–881727–7 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 06/07/2018, SPi

Contents Notes on Contributors

vii

Introduction1 1. Rethinking Agreement in Plato Lesley Brown

18

2. Plato’s Theory of Knowledge Ralph Wedgwood

33

3. Justice and Persuasion in the Republic57 Dominic Scott 4. Plato against Democracy: A Defense Richard Kraut

77

5. Self-Mastery and Self-Rule in Plato’s Laws97 Susan Sauvé Meyer 6. Plato’s Philebus and the Value of Idle Pleasure Verity Harte

110

7. A Series of Goods Christopher Shields

129

8. Practical Truth: An Interpretation of Parts of NE VI David Charles

149

9. Aristotelian Feelings in the Rhetoric169 Paula Gottlieb 10. ‘Ought’ in Aristotle’s Nicomachean Ethics184 Julia Annas 11. Deliberation and Decision in the Magna Moralia and Eudemian Ethics197 Karen Margrethe Nielsen 12. The Freedom Required for Moral Responsibility John Martin Fischer

216

13. Virtue: Aristotle and Kant Allen W. Wood

234

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

vi  Contents 14. Richard Price on Virtue Roger Crisp

253

15. Eudaimonism and Cosmopolitan Concern David O. Brink

270

Bibliographies of Gail Fine and Terence Irwin Index Locorum General Index

293 305 315

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Notes on Contributors Julia Annas  is Regents Professor of Philosophy at the University of Arizona. She holds a B.A. (Hons) from Oxford, a Ph.D. from Harvard (1972), and an Honorary Doctorate from the University of Uppsala. Her research interests have ranged over a wide field of ancient philosophy, but for some years have focused on ancient ethics, and she has also worked in the field of contemporary virtue ethics. Her publications in ancient philosophy include An Introduction to Plato’s Republic (1981), The Morality of Happiness (1993), Platonic Ethics Old and New (1999), and Virtue and Law in Plato and Beyond (2018). She has also published Intelligent Virtue (2011) and is working on a book on virtue ethics. David O. Brink  is Distinguished Professor of Philosophy at the University of California, San Diego. He received his Ph.D. from Cornell University in 1985, where he studied with Gail Fine and was supervised by Terry Irwin. His research interests are in ethical theory, history of ethics, moral psychology, and jurisprudence. He is the author of Moral Realism and the Foundations of Ethics (Cambridge University Press, 1989), Perfectionism and the Common Good: Themes in the Philosophy of T.H. Green (Clarendon Press, 2003), and Mill’s Progressive Principles (Clarendon Press, 2013). Lesley Brown is Fellow in Philosophy emeritus at Somerville College, Oxford and a member of the Philosophy Faculty of the University of Oxford. She has published extensively on Plato, particularly on Plato’s Sophist, and on Aristotle, including ‘Why is Aristotle’s Virtue of Character a Mean?’ in the Cambridge Companion to the Nicomachean Ethics (2014). David Charles  is Professor of Philosophy at Yale University. He is the author of Aristotle’s Philosophy of Action (Duckworth, 1984) and Aristotle on Meaning and Essence (Clarendon Press, 2000) and edited Definition in Greek Philosophy (Clarendon Press, 2010). Roger Crisp is Uehiro Fellow and Tutor in Philosophy at St Anne’s College, Oxford, and Professor of Moral Philosophy at the University of Oxford. He is author of Mill on Utilitarianism (Routledge, 1997), Reasons and the Good (Clarendon Press, 2006), and The Cosmos of Duty: Henry Sidgwick’s Methods of Ethics (Clarendon Press, 2015). He has translated Aristotle’s Nicomachean Ethics for Cambridge University Press (2000) and edited the Oxford Handbook of the History of Ethics (2013). John Martin Fischer received his Ph.D. in philosophy from Cornell University in 1982. He was fortunate to study with both Gail Fine and Terry Irwin during his time at Cornell. His primary research interests are in free will and moral responsibility.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

viii  notes on contributors He is the author of The Metaphysics of Free Will: An Essay on Control and (with Mark Ravizza) Responsibility and Control: A Theory of Moral Responsibility. Oxford University Press has published four collections of his essays: My Way: Essays on Moral Responsibility; Our Stories: Essays on Life, Death, and Free Will; Deep Control: Essays on Free Will and Value; and Our Fate: Essays on God and Free Will. He is Distinguished Professor of Philosophy at the University of California, Riverside. Paula Gottlieb is Professor of Philosophy and Affiliate Professor of Classical and Ancient Near Eastern Studies at the University of Wisconsin-Madison. She received her B.Phil. at Oxford and studied for the Ph.D. with Terry Irwin and Gail Fine at Cornell. She specializes in Aristotle’s ethics and metaphysics. She is the author of The Virtue of Aristotle’s Ethics (Cambridge University Press, 2009) and ‘Aristotle on NonContradiction’ for the Stanford Encyclopedia of Philosophy. She is presently writing a book, tentatively titled Aristotle on Reason and Feeling, for Cambridge University Press. Verity Harte  is Professor of Philosophy and Classics at Yale University. She received her Ph.D. from Cambridge University in 1994. Her research interests are in ancient philosophy, especially the philosophy of Plato and Aristotle, with a particular focus on topics in metaphysics, epistemology, and philosophical and moral psychology. She is the author of Plato on Parts and Wholes: The Metaphysics of Structure (Oxford, 2002) and of numerous articles on Greek philosophy. Richard Kraut  is Charles and Emma Morrison Professor in the Humanities, in the Department of Philosophy, Northwestern University. His interests include contemporary moral and political philosophy and the ethics and political thought of Socrates, Plato, and Aristotle. Among his publications are Against Absolute Goodness (Oxford, 2011), What is Good and Why: The Ethics of Well-Being (Harvard, 2007), Aristotle: Political Philosophy (Oxford, 2002), Aristotle, Politics Books VII and VIII: Translated with a Commentary (Clarendon Aristotle Series, 1997), Aristotle on the Human Good (Princeton, 1989), and Socrates and the State (Princeton, 1984). He is a member of the American Academy of Arts and Sciences. Susan Sauvé Meyer is Professor of Philosophy at the University of Pennsylvania. Trained at the University of Toronto (B.A. 1982) and Cornell University (Ph.D. 1987), she taught at Harvard University before joining the faculty at the University of Pennsylvania in 1994. A specialist in Classical Greek and Hellenistic philosophy with special interest in the ethical tradition, her publications include Aristotle on Moral Responsibility (1993; reissued in 2011), Ancient Ethics (2008), and Plato: Laws 1 and 2 in the Clarendon Plato Series (2015). She is currently an editor of the journal Archiv für Geschichte der Philosophie. Karen Margrethe Nielsen is Tutorial Fellow at Somerville College and Associate Professor in the Faculty of Philosophy, University of Oxford. Her publications include ‘Spicy Food as Cause of Death: Coincidence and Necessity in Metaphysics E 3’, Oxford

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

notes on contributors  ix Studies in Ancient Philosophy (2017), ‘Vice in the Nicomachean Ethics’, Phronesis (2017), ‘The Will: Origins of the Notion in Aristotle’s Thought’, Antiquorum Philosophia (2013), and ‘Deliberation as Inquiry’, Philosophical Review (2011). She received her Ph.D. from Cornell University in 2006, on a dissertation supervised by Terry Irwin and co-supervised by Gail Fine, on Aristotle’s theory of decision (prohairesis). The topic has held her attention ever since. Dominic Scott is a professor of philosophy at the University of Oxford and a fellow of Lady Margaret Hall. He received his Ph.D. from the University of Cambridge and taught in the Philosophy Department there from 1989 to 2007. He was professor of philosophy at the University of Virginia from 2007 to 2014. He is the author of Recollection and Experience (Cambridge University Press, 1995), Plato’s Meno (Cambridge University Press, 2006), and Levels of Argument: A Comparative Study of Plato’s Republic and Aristotle’s Nicomachean Ethics (Oxford University Press, 2015). He co-authored The Humanities World Report 2015 (Palgrave Macmillan, 2014), and edited Maieusis: Studies in Honour of M. F. Burnyeat (Oxford University Press, 2007) as well as The Pseudo-Platonic Seventh Letter: A Seminar by Myles Burnyeat and Michael Frede (Oxford University Press, 2015). Christopher Shields is Shuster Professor of Philosophy at the University of Notre Dame. Previously he taught at the University of Oxford and the University of Colorado at Boulder. He has held visiting posts at the Chinese University of Hong Kong, the University of St. Louis, the Humboldt University of Berlin, Cornell University, Stanford University, Yale University, and the University of Arizona. He is the author of Order in Multiplicity: Homonymy in the Philosophy of Aristotle (Oxford University Press, 1999), Classical Philosophy: A Contemporary Introduction (Routledge, 2003), Aristotle (Routledge, 2007), Ancient Philosophy: A Contemporary Introduction (Routledge, 2011), with Robert Pasnau, The Philosophy of Thomas Aquinas (Westview, 2003; 2nd rev. ed. Oxford University Press, 2015), and Aristotle’s De Anima, Translated with Introduction and Commentary (Oxford University Press, 2016). He is the editor of The Blackwell Guide to Ancient Philosophy (Blackwell, 2002) and The Oxford Handbook of Aristotle (Oxford University Press, 2012). Ralph Wedgwood  is Professor of Philosophy at the University of Southern California. He received his Ph.D. from Cornell University in 1994, where he studied with Gail Fine and was supervised by Terry Irwin. Later, he was their colleague at the University of Oxford, where until 2012 he was a member of the Faculty of Philosophy and a Fellow of Merton College. His principal research interests are in ethics, epistemology, and the theory of practical reason and rational choice, with a subsidiary interest in the history of those subjects. He is the author of The Nature of Normativity (Clarendon Press, 2007) and of The Normativity of Rationality (Clarendon Press, forthcoming).

OUP CORRECTED PROOF – FINAL, 06/07/2018, SPi

x  notes on contributors Allen W. Wood is Ruth Norman Halls Professor at Indiana University and Ward W. and Priscilla B. Woods Professor Emeritus at Stanford University. He was born in Seattle, received his B.A. from Reed College in 1964 and his Ph.D. from Yale University in 1968. His interests are in the history of philosophy, especially German philosophy in the eighteenth and nineteenth centuries, and in ethics, social and political philosophy, and philosophy of religion. He is the author of a dozen books and editor or translator of about a dozen others. His most recent books are Fichte’s Ethical Thought (Oxford University Press, 2016),  Formulas of the Moral Law (Cambridge University Press, 2017), and a second (revised) edition of his translation of Kant’s Groundwork for the Meta­ physics of Morals  (Yale University Press, 2018). Forthcoming in Spring, 2018 is a second (revised) edition of his translation of Kant’s Groundwork for the Metaphysics of Morals (Yale University Press).

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

Introduction Through their writing, their teaching, their mentoring, and their broader scholarly output, Gail Fine and Terry Irwin have reshaped the character of ancient philosophy as an academic discipline. Their contributions to the discipline do not, however, end there. On the contrary, their wide-ranging achievements extend into all periods of the history of philosophy and indeed into several areas more systematic than historical. Or perhaps one should say, rather, that their work defies any ready classification as being either historical or systematic, because whatever its primary focus on a given occasion, what they write cannot be pigeonholed as either exclusively scholarly or thematic; for they practice an unremittingly philosophical form of history of philosophy, or, judged from another angle, a historically enriched form of systematic philosophy. That is, as they pursue it, philosophy engages the discipline’s history in a manner animated by its current and perennial concerns, but it does so while remaining fully sensitive to the original context of its production. Their work combines the highest level of scholarly rigor and rich philosophical insight. Animated by a purely philosophical spirit, it is never narrowly antiquarian in orientation. Although alert to matters of text and transmission reflecting painstaking philological care and exceptionally broad scholarly erudition, their work never loses sight of a simple question: should we too believe this? Their students, their colleagues, and the broader philosophical public have been the beneficiaries of their sustained and remarkable activity. In an effort to express their admiration and gratitude to Terry and Gail, the contributors to this Festschrift have offered these essays to mark the occasion of their retirements from the University of Oxford and Cornell University, where both have held permanent posts in their long and distinguished careers. They have between them educated several generations of philosophers, many of whom have in their turn begun the process of passing along to their own students the legacy of excellence originating in the careers of Terry and Gail. Most of the essays in the present volume made their first appearance at a conference held at Cornell University in September of 2013, dedicated to Terry and Gail, who kindly presided over the proceedings. The speakers at the conference found themselves, in typical fashion, challenged, encouraged, and schooled by Terry and Gail, and indeed by the assembled audience of their past and current students, their colleagues, and those whose work they have influenced. Since then, the editors have

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

2  Virtue, Happiness, Knowledge carried the spirit of the conference forward, commenting upon each of the chapters, engaging in mutually beneficial exchanges with their authors, and working to produce a volume worthy of its two honorees. We offer it to Gail and Terry with admiration and continuing gratitude. Prefatory to the chapters which follow, we have undertaken to offer brief overviews of the careers and contributions of Gail and Terry. Although they share a common method and a dominant period of focus, their individual contributions head in distinctively different directions, often arriving at instructively divergent points of view, at times complementary and other times at variance with one another. Accordingly, we begin by recounting their philosophical careers individually, before returning briefly to their shared influence and legacy.

Gail Fine Gail’s contributions to ancient philosophy center on Plato and Aristotle, but extend to Hellenistic philosophy as well, where she has done seminal work on Academic Scepticism. Her work is characterized by a lively, minute form of textual engagement motivated by broader philosophical considerations: she wants to know what the philosophers she studies maintain, to ascertain why they maintain what they do, and then to determine whether we ourselves should agree with them, and, if we do, whether we should do so on the basis they advance for holding the views they espouse. We find, then, careful exegesis and philosophical assessment in equal measure. It should not be inferred, however, that these activities parcel into discrete components in her work, beginning with neutral exegesis where positions are discerned and dispassionately characterized, followed by argument reconstruction, and then rounded off by critical appraisal. On the contrary, it is a hallmark of Gail’s methodology that each of these activities informs the others in a symbiotic, mutually enhancing way. Generally speaking, on her approach it counts as a good reason to discount an interpretation of Plato, or Aristotle, or Sextus that it ascribes without compulsion a view that is transparently implausible or simply false. Accordingly, Gail’s approach to ancient texts is guided by the thought that we are better served by reading a supporting argument as enthymematic than by concluding that it is transparently invalid or unsound, unless, again, we are faced with an unanswerable reason for doing so. This is not because ancient authors never say false things or give bad arguments for their views, whether true or false. Rather, Gail’s approach commends the thought that if we read a text understanding our first hermeneutical impressions to be unassailable, then we do our authors a disservice: we are apt to miss their deeper meanings and motivations. By the same token, if on a first reading we find a philosophical position alien to the point of being unintelligible, that, Gail thinks, is as likely our fault as it is the fault of the author being studied. In consequence, we should in every instance strive to understand and assess the philosophers of antiquity in terms we ourselves can readily understand and articulate in our own vocabulary.

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  3 This last point has induced some of her readers to suspect Gail of courting anachronism: we should, such scholars advise, grapple with philosophers of earlier periods in their own terms, not in the terms we happen to find congenial, and we should avoid allowing our independent assessments of the plausibility of their positions to colour— or, or as they would have it, to discolour—our judgment as to the accuracy of an ascription. Gail, they may rightly point out, is happy to recruit both Chomsky and Damascius when elucidating a single Platonic text (2014, 154 n. 51, 165); yet Chomsky writes in a place and time far removed from Plato and in an idiom utterly alien to that of Damascius. Put in its most unsympathetic terms, this sort of reaction calls into question the unflinchingly logocentric method Gail characteristically employs. Does this charge have any traction? Each time she encounters an ancient text Gail’s method involves posing a pair of questions: (i) what are the possible meanings of this text? and (ii) which among them is most plausibly supported by the argument it offers? When we know the argument, we have a handle on the position, but not before. It is rarely if ever the case that we can simply light upon the correct interpretation of an ancient philosophical text as a matter beyond question or controversy from a reading which avoids assessing its author’s philosophical motivations and objectives, no matter how thorough our reading may otherwise be. Competing interpretations of varying degrees of plausibility will invariably present themselves; we are then asked to choose among them. Gail’s way of choosing begins by determining, with all due charity and intellectual humility, which of the alternatives is best supported by the argument the author of the text promulgates. Where no immediate argument is given, one may equally determine, as her approach suggests, which among the positions most readily comports with claims motivated by argument elsewhere. We may then, adapting a metaphor deployed in Plato’s Republic (434e), rub the passages together to see which interpretation emerges from the process, as fire emerges from the sparks of fire sticks when rubbed together. If we proceed in this way, our governing impulse will be primarily logocentric, in the sense that it will enjoin us to ferret out the argumentative underpinnings of a claim as our first and most secure—if not our sole—guide to its likely meaning. At any rate, Gail’s governing practice seems to reflect some such approach; she does not spend a great deal of time overtly defending or even describing her philosophical method. Still, the same method structures her work in virtually every period of her long and productive career. She deals primarily with non-value areas of philosophy, concentrating especially on issues in metaphysics and epistemology, and their intersection, as they crop up in the philosophers of special concern to her, taken both individually and in concert. This last point bears emphasis because one core area of Gail’s research has concerned the philosophical interaction of Plato and Aristotle. In the early twentieth century, the pioneering German scholar Werner Jaeger advanced a striking set of views about Aristotle’s development, focussed centrally upon his evolving attitudes towards Plato. Jaeger thus kicked off a long and fruitful scholarly dialectic to which Gail has made lasting contributions.

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

4  Virtue, Happiness, Knowledge In general terms, Jaeger (1923) took the view that Aristotle began life as a dutiful Platonist who gradually grew critical of his teacher, developing into his own master as he matured and forged a system of philosophy markedly incompatible with Plato’s, rejecting most conspicuously the cornerstone of Plato’s thought, his theory of Forms. Jaeger’s view won many adherents but also some partial detractors as well; it was not, however, frontally assaulted with any success until 1965 with the publication by G. E. L. Owen (Gail’s doctoral advisor) of a British Academy lecture entitled ‘The Platonism of Aristotle’. In this work, Owen attempted to turn the tables completely by arguing that only as he matured did Aristotle come to appreciate the profounder dimensions of Plato’s thought, with the result that, far from beginning as a dutiful Platonist who emerged incrementally as an autonomous thinker and harsh critic, Aristotle actually began life as an impetuous critic who came to adjust his own thinking with an increasingly appreciative eye on delicate problems acknowledged by Plato himself in some admirably self-critical moments. One such moment, a crux of sorts, is the so-called Third Man Argument directed against Plato’s theory of Forms, introduced by Plato in his Parmenides, according to some scholars merely maieutically and heuristically, in an effort to clarify and defend his conception of Forms, but according to others as a candid negative assessment of his own system intended to present an unanswerable criticism of his signal contribution. This Third Man Argument came to play a significant role in Gail’s intellectual development. She investigates the argument in several articles, and then returns to it along with various other Aristotelian assaults on the theory of Forms in her magisterial On Ideas: Aristotle’s Criticisms of Plato’s Theory of Forms (1995). This work is a study of an uncommonly rich text of Aristotle’s, the Peri Ideôn (On Ideas), sections of which were preserved by Alexander of Aphrodisias. In the fragment recorded by Alexander, Aristotle assails Plato’s theory of Forms by means of a series of complex arguments, including a version of the Third Man Argument. Gail’s treatment of these arguments represents the pinnacle of an argument-focused assessment of Aristotle’s relationship to Plato and his theory of Forms. The picture that emerges is—as a philosophical as opposed to a doxagraphical matter—far more detailed and philosophically nuanced than anything produced by either Jaeger or Owen. It is, and will remain for many years to come, an indispensable resource for philosophical scholars investigating Aristotle’s relationship to Plato in metaphysical matters. This work is, however, but one of Gail’s signal achievements in ancient philosophy. Judged by their impact and the amount of discussion they have generated, Gail’s contributions in four areas merit special note. First is the issue already introduced, Aristotle’s criticisms of Plato, including, but not limited to, those pertaining to the theory of Forms; second is her widely influential, early account of knowledge and belief in the middle books of Plato’s Republic; third is her career-spanning interest in the paradox of inquiry, as it was formulated originally in Plato, but then also as it appears subsequently in various guises in post-Platonic ancient philosophers; and fourth is her engagement with Plato’s epistemology more broadly, which already occupied her

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  5 in her Harvard doctoral thesis, ‘Plato and Acquaintance’ (1975). Her concern with Plato’s epistemology surfaces over and over again throughout her more than 50 scholarly articles, critical discussions, and scholarly monographs. Her great body of work comprises many publications not mentioned in this short discussion, including several important contributions on the nature of substances and universals, most but not all of which take Plato and Aristotle as their focus, as well as assays into scepticism, subjectivity, perception, causation, and determinism. (A full bibliography of Gail’s scholarly publications can be found at the end of the volume.) Without any attempt at being comprehensive, then, we may characterize some of Gail’s most prominent publications, in an effort to provide an indication of her lasting contributions to the field and to reflect at least briefly on her distinctive philosophical methodology. Beginning with the last recurrent theme mentioned, we may focus first on Gail’s approach to Plato’s epistemology. In an earlier, much discussed work, ‘Knowledge and Belief in Republic V’ (1978), later reprised, refined, and expanded as ‘Knowledge and Belief in Republic V–VII’ (1989b), Gail articulates a deep problem in Plato’s epistemology and proceeds to offer a startling solution to it—startling, at any rate, to a certain sensibility, one characteristic of an older generation of scholars who had understood Plato’s metaphysical epistemology in such a way that the problem Gail articulates with such clarity and force barely comes into focus. The problem is this. In the middle books of the Republic, Plato offers an account of knowledge which seems first to bifurcate the world into the knowable and the unknowable and then to suggest that knowledge (epistēmē) is, reasonably enough, restricted to those sorts of objects which are knowable, namely Forms. Forms are, on this picture, suitable objects of knowledge because they are stable, precise, context-invariant, and incapable of slipping away, as Plato puts it in the Meno, in the manner of the statues of Daedalus. These seem at first to be fixed in place, like other statues, but then prove so lifelike that when left unshackled, they scamper away, leaving those who possessed them empty-handed. What value they have to those whose they are, then, lasts only so long as they are secured. Evidently, beliefs are like that: they are fine as far as they go, but unless tied down, they slip away when unobserved, and so prove of no lasting value. To be secured, however, beliefs must be bound with the chains of reason, tethered by an account (logos) or reasoned explanation (aitios logismos); then, and only then, do beliefs give way to knowledge. In this way, an item of genuine knowledge is superior to any given belief, because in addition to being true, as a belief may (or may not) be, it is supported in the right way by an anchoring account. Now, if all knowledge qualifies as knowledge only if it is accompanied by an account, then—on the assumption that an account is also something that has to be known by the knower whose belief it anchors—knowledge will require prior knowledge if it is to qualify as knowledge at all. If that is so, however, knowledge will be impossible. For each attempt at a knowledge claim will require a prior, more fundamental justifying knowledge claim. One may counter, as many scholars take Plato to counter, that some

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

6  Virtue, Happiness, Knowledge kinds of knowledge escape this regress: some knowledge is privileged, qualifying as a first principle, as something known simply and immediately, without any need for further justification. An older generation of scholars accepted this second outcome, taking a cue from Plato himself, who introduces the Form of the Good as the ‘unhypothethical first principle of all’ (Rep. 510b7, 511b6). The suggestion thus lies near that knowledge of this Form, the Form of the Good, is foundational for all other knowledge, and that Plato’s solution to the regress problem is thus a version of epistemic foundationalism. This near-lying suggestion is also, however, as Gail presses, deeply problematic. Indeed, its manner of being problematic nicely illustrates Gail’s general approach to interpreting Plato. To begin, the notion of a first epistemic principle, an ungrounded ground accessed directly by an unnamed faculty of mind via some manner of immediate apprehension, strikes many as inherently, intractably mysterious. Still, one may aver, this is Plato’s view and we are left to make such sense of it as we can. Is it, though, Plato’s view? To many it seems so. Is this not, after all, the immediate purport of Plato’s commitment to the Form of the Good as an unhypothetical first principle? Indeed, the entire analogy of the sun in Republic VI supports this view. After all, as the Form of the Good is to the intelligible realm, so the Sun is to the visible realm. As the Sun illuminates other things but is visible itself by its own nature, so the Form of the Good renders other things intelligible but is itself intelligible in virtue of itself, by its own nature, and so needs nothing beyond itself to vouchsafe its grounding role in Plato’s foundationalist epistemology. This seems very close to a straightforward statement of epistemic foundationalism on Plato’s part. As Gail is quick to point out, however, and as we have already seen, Plato maintains precisely the opposite regarding objects of knowledge as a class: he distinguishes knowledge (epistēmē) from mere belief (doxa) by demanding that knowledge be accompanied by an account or reasoned explanation; the foundationalist model mooted precisely abjures that requirement. When a cognizer becomes acquainted with a Form, or at least the Form of the Good, there is no further justification needed or wanted; indeed, the very possibility of providing a justifying account seems ruled out by the unhypothetical character of the experience. How, if this is so, is knowledge (epistēmē) to be distinguished from mere belief (doxa)? There seems to be little wriggle room here, since Plato’s contention that knowledge (epistēmē) requires something more than mere belief (doxa) is not a singular or even rare in his writings. On the contrary, he asserts it repeatedly (e.g. Rep. 510c, 531e, 533b–c; Phaedo 76b; Symp. 202a; Tht. 202c; Laws 966b, 967e). More to the point, it figures centrally in his brief in Republic V for the claim that only philosophers can rule: they, and they alone, can differentiate Forms from one another by reason and by giving an account of what each is (by providing a logos), including, as Plato acknowledges, the Form of the Good (Rep. 534b–c). Are we, then, left with a simple contradiction? Gail thinks not. All knowledge, including knowledge of the Form of the Good, requires, just as Plato says, some accompanying justification or account (logos) to

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  7 secure it. It does not follow, however, that the justificatory chains of our claims to knowledge must be linearly ordered, with each justification appealing to a justifying principle prior to it in a justificatory order. Indeed, it does not even follow that the ‘chains’ of justification be chains. Individual justifications might rather fit into a web, deriving justification one from the other and all in concert from the totality which comprises them, conferring justification holistically rather than atomistically. Plato might, after all, be an epistemic coherentist rather than a foundationalist. Such, at any rate, is Gail’s contention. Gail’s contribution to this debate has occasioned a great deal of discussion, including a good deal of dissent. This is, however, only because it is novel, provocative, philosophically alert, and uncommonly creative. It also turns out to be motivated not by some anachronism, but by a looming contradiction internal to Plato’s writings which others have failed to acknowledge. Philosophically minded readers of Plato have rightly taken note. The same again holds of her career-spanning interest in the paradox of inquiry (1992b, 2004a, 2007, 2010a, 2010b, 2013, 2014). Like her work on the middle books of the Republic, Gail’s engagement with the paradox of inquiry reflects her deep and abiding interest in both epistemology and metaphysics—in this case beginning already in her undergraduate days, when, as she reports, she read the Meno and was ‘immediately enchanted’ (2014, vii). It is easy to see why she should find this dialogue so captivating. Its core paradox raises deep questions about Plato’s aims and objectives, and thereby also about the aims and objectives of a broader sort of philosophical mission. Readers of the Meno will recall that a problem about the goals of Socratic inquiry is first interjected into the dialogue by a frustrated Meno on the occasion of his having been refuted by Socrates. Meno’s frustration results in part from his failure to answer a general sort of question posed by Socrates, a so-called ‘What is F-ness?’, where ‘F’ stands in for some virtue term: ‘What is courage?’, ‘What is piety?’ and so forth. Meno grows indignant, asserting that the Socratic ‘What is F-ness?’ question is unanswerable, insinuating that in posing it Socrates is setting up his interlocutors for failure. Socrates rejoins with what he purports to be a precisification of Meno’s complaint. In so doing, Socrates presents what has come to be known as ‘Meno’s Paradox’: for all x, inquiry into x is impossible, since either one knows x, in which case there can be no occasion for coming to know it by inquiry, or one does not know x, in which case one will fail to recognize it even if one stumbles upon it in the course of inquiring after it, rendering any attempt to inquire into what one does not know futile. Socrates’ response to the paradox as he himself formulates it is perplexing: he first dismisses it as a debater’s trick (Meno 80e), yet then proceeds to counter it by means of an elaborate, even extravagant response involving his remarkable doctrine of recollection, according to which all learning is actually recollection, coupled with a commitment to the prenatal existence of the human soul. As Socrates shows by cross-questioning a slave about a simple geometrical problem, the slave already has the answer he professes not to know somehow, so to speak, lurking unrecognized in his soul. If this is correct, the process of

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

8  Virtue, Happiness, Knowledge ‘learning’ is shown to be not the acquisition of something from without, but really rather a dredging up of what one already knows. This in turn implies, Socrates contends, that the soul is immortal, since it shows itself able upon examination to recollect things never learned in this lifetime. Looked at from one angle, then, Socrates’ response may appear concessive to a fault, implicitly allowing that the debater’s argument, no matter how slippery, actually has a true conclusion: inquiry into what one does not already know is impossible. This appearance is, however, misleading at best. The situation proves far more complicated, in ways Gail, more than any other scholar, has painstakingly made clear. Indeed, all of the ingredients for the sort of philosophical problem which excites Gail’s interest are present in this exchange. There is an epistemic dimension: what can be known and how? What standards are appropriately employed in determining that someone has and can demonstrate having genuine knowledge, as opposed to conviction or earnest, even true belief? There is a metaphysical component as well: what kinds of things can be known? Propositions? Abstract entities? Only what is assertoric and truth-evaluable? Sensible individuals, in addition to the abstract objects of knowledge? These questions play out against the backdrop of the worry Meno rightly presses in his initial outburst: what does Socrates want to achieve in posing his ‘What is F-ness?’ question? What form will a successful response take? Are there perhaps different kinds of knowledge for different kinds of objects? That is, does knowledge (or virtue, or piety, and so on for any of the various values of ‘F’ in the ‘What is F-ness?’ question) admit of a univocal analysis, as Socrates so often seems to assume? Finally, there is a multi-tiered set of textual questions about Plato’s presentation of these issues. Is he speaking in propria persona using Socrates as a dramatic character or representing the historical Socrates he knew? Are we to take Socrates’ presentation of the doctrine of recollection at face value, or as going proxy for some milder doctrine, about the possibility of innate or perhaps a priori knowledge? In sum, in Meno’s Paradox we have a complex and multitiered nexus of textual, exegetical, epistemological, metaphysical, and methodological questions, each of which impacts the other in appreciable ways: plainly grist for Gail’s mill. Scholars have responded in all manner of ways to this situation, some even maintaining that Socrates cannot be in earnest in offering his doctrine of recollection in response to Meno’s Paradox, that he is sparring, sporting, or otherwise spoofing. Gail, by contrast, sees a genuine problem met with an earnestly proposed solution. According to Gail, Plato responds by showing that though perfectly valid, Meno’s paradox is unsound. Its premise is false: the claim that if one does not know x, then one cannot inquire into x. On the contrary, one can, just as Socrates’ exchange with the slave illustrates, begin in belief, even false belief, and steer one’s way to knowledge. Since we can in fact move from belief (doxa) to knowledge (epistēmē), not least through the instrument of a shrewdly deployed Socratic elenchus, it is simply false that inquiry is impossible for those lacking knowledge. When we confront Plato’s Meno, we are asked to determine how deep the paradox of inquiry runs. The discussions inaugurated by Gail provide ample reason to believe that

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  9 no simple quick fix suffices. We see this same result in another way as well, a way made clear in Gail’s latest work on this topic, a book-length study of the paradox of inquiry as it appears not only in Plato but also in the philosophers influenced by him. For, as she shows, the paradox of inquiry has a long afterlife in antiquity, captivating a full spectrum of top-tier philosophical minds, including Aristotle, Epicurus, the Stoics, and the Sceptics. Their nuanced attention to aspects of the problems it poses gives us further reason to conclude that Meno’s paradox cannot be dissolved simply by disarming it as a debater’s equivocation. Gail’s career-long interest in this topic thus eventuates in another lasting contribution to the field, The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus (2014). Yet another agenda-setting thesis emerges in a different early work of Gail’s (1978), one offering an especially clear illustration of her hermeneutical methodology. It also shows how Gail’s work on Plato crosses readily over from epistemology into metaphysics. On one natural reading of the middle books of Plato’s Republic, we find him committed to an epistemically driven bifurcation of reality. Given their invariability and context-insensitivity, Plato suggests, Forms are uniquely suited to serve as objects of knowledge (epistēmē); by contrast, sense particulars, given their ceaseless variability and context-dependence, are suitable objects of belief (doxa) but never of knowledge. Can we know, for instance, that an elephant is large? Well, fairly plainly, in the class of mammals, elephants are indeed large. Still, in comparison with Mt. Everest or the Milky Way, an elephant is puny. So, is an elephant large? Well, suggests Plato: yes and no. By contrast, largeness itself, taken by itself, is never anything but large: it does not rely on a context of appraisal for being what it is. One Platonic way of thinking, then, differentiates Largeness, the Form, a transcendent, abstract entity, from sensible large things, which can never escape their context sensitivity. As a result, the Form Largeness itself is not reducible to large sensibles, taken either individually or corporately. The requisites of knowledge thus imply that the objects of knowledge are supra-sensible, demarcated by their very natures from anything we might perceive by the senses. This doctrine, again congenial to an older generation of scholars, is sometimes dubbed the ‘Two-Worlds Theory’. According to the Two-Worlds Theory, Plato divides reality into two mutually exclusive domains, a domain of sense perceptibles which serve as objects of belief and a domain of abstract entities which serve as objects of knowledge. No object of belief can be known; no object of knowledge can serve as an object of belief. To revert to the earlier illustration of the Meno, if we think that we know what largeness is because we know that Jumbo the elephant is large, then we have truncated vision: we will find the predicate ‘. . . is large’ scurrying away as we shift Jumbo from one context of appraisal to another. When we know something, we know it stably and securely; so, our attitudes towards largeness vis-à-vis Jumbo do not amount to knowledge of largeness. Large things are perceived and not known; known things are grasped by the mind and not perceived. It must be said that Plato does much to encourage the Two-Worlds Theory. In Republic V, for instance, he individuates our mental capacities by what they are ‘over’ or

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

10  Virtue, Happiness, Knowledge ‘set over’ (epi), thus: (i) knowledge (gnōsis, epistēmē) is set over (epi) what is; (ii) belief (doxa) is set over (epi) what is and is not; and (iii) ignorance (agnōsia) is set over (epi) what is not (Rep. 477a–b). One natural way of taking this passage would be that knowledge ranges over Forms (over what is, what exists fixedly and permanently), belief ranges over sensibles (which are ever shifting, and never completely or permanently what they are, because of the point about context sensitivity already mentioned), while ignorance ranges over what is not (because it takes no object at all). Bracketing the point about ignorance, which raises special difficulties for all approaches, the TwoWorlds approach takes Plato here to be more or less stating a straightforward commitment to the mutually exclusive and exhaustive bifurcation of reality, into the knowable and the sensible, the realm of Forms and the material realm. Gail begins with a simple, deflating pair of observations. First, if we think that knowledge is the same as justified true belief (or that it at least requires justified true belief), then knowledge must involve belief, rather than exclude it. Since we do in fact think this, we cannot easily understand Plato as overlooking it so blithely. This is another occasion, however, where some may suspect Gail’s method of courting anachronism. What matters, after all, is not what we think about knowledge and belief, but what Plato thinks; and has he not just effectively asserted that they are disjoint? Not exactly, it turns out. Gail too has textual evidence on her side. As we have already seen, it is not us, but rather Plato himself who urges, in the Meno and then also in the Theaetetus, that knowledge (epistēmē) precisely is true belief together with an account (logos) or reasoned explanation (aitios logismos). So, on his approach, far from being belief-excluding, knowledge (epistēmē) is belief-entailing. Hence, not only can Forms be objects of belief, they evidently must be. Moreover, Plato repeatedly speaks of his beliefs about Forms, including even the Form of the Good, about which he expressly claims to have no knowledge. So, the Two-Worlds Theory, however recommended by Plato’s manner of speaking, seems incompatible with his practice; it also seems at variance with some deeply intuitive convictions, such that one can know one’s neighbour, though she is not a Form, and one can believe, but not know for sure, that Plato’s Forms are meant to be paradigms after which the sensibles which participate in them are named. Gail’s first observation recommends, then, a second. As she rightly maintains, when he asserts that knowledge and belief are set over (epi) specifiable objects, Plato may be taken in a variety of non-equivalent ways, depending on how one understands the word ‘is’. The English word ‘is’, like the Greek word it translates (esti), can be used in at least three different ways: (a) existentially (‘There is a God’ = ‘God exists’); (b) predicatively (‘Miss Stanbury is prim’, an instance of the general predicative schema ‘x is F’); and, if somewhat less frequently in English than in Greek, (c) veridically (‘Tell it like it is’ = ‘Tell the truth’) In principle, then, to take just his first sentence as an illustration, when Plato says that knowledge is set over (epi) what is, he might mean that knowledge takes as its objects: (a) what exists; (b) what is F, for some predicate or other; or (c) what is true. As Gail rightly observes, proponents of the Two-Worlds Theory move too

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  11 readily and without argument to (a), the existential use of ‘is’. Beyond the problems already noted, this interpretation has some immediately perplexing consequences even taken in its own terms: it is not at all clear what it means to assert that knowledge concerns what exists, whereas belief concerns what both exists and does not exist. What can both exist and not exist? Taken distributively, that one has beliefs both about what exists and about what does not exist, the suggestion seems more promising. Yet then we have a different worry. Surely if we can have beliefs about what does not exist, so too can we know things about what does not exist. One can, after all, know that Santa Claus does not exist; and one can believe falsely of the tree in the front garden that it is a poplar, when it is in fact a cyprus. In the first case we have knowledge pertaining to a non-existent and in the second false belief pertaining to an existent. Perhaps, then, Plato has some one of the other, non-existential senses of ‘is’ in mind? Gail argues for (c), the veridical sense, such that Plato’s meaning is rather something less extravagant, that knowledge is in every instance of something which is true, namely a true proposition, whereas belief can be set over true or false propositions indifferently. Put another way, knowledge is truth-entailing, while belief is not. That much seems neither alien nor even problematic. How, though, are we to adjudicate these differing interpretations? One can generate them, just as Gail does, only to find that there is no way forward. Even in that case, however, one will already have shown that Plato’s text does not demand the existential reading, and so does not simply state or even imply the Two-Worlds Theory. Still, with no further argumentation from one side or the other, the debate will stalemate. Fortunately, there is a way forward. For, happily, the second phase of Gail’s logocentric method is still to be deployed. One can proceed just as Gail is disposed to proceed, by looking to the text of the Republic to see which, if any, of these alternatives its arguments require, or, failing that, which they recommend. Gail argues that upon close inspection, the arguments of Republic V–VII in fact recommend (c), the veridical reading, which in turn carries with it various exegetical and philosophical advantages. Here too Gail’s solution has garnered a great deal of scholarly attention, once again a fair bit dissenting. All this discussion, including the dissent, is welcome: Gail’s work has inaugurated a healthy, productive scholarly debate, and has thus advanced our understanding of Plato and the issues he brings to our attention. That she should have this effect is altogether unsurprising. In assessing the so-called Two-Worlds Theory, Gail advances a textually informed, philosophically nuanced, bold and attractive thesis, one which has proven agenda-setting for all the best reasons. We have focussed mainly on Gail’s influential discussions of Plato’s metaphysics and epistemology. Although our doing so provides at least a faint portrait of her approach to the study of ancient philosophy, we would be remiss not at least to mention her many other accomplishments. She has additionally done pioneering work on Aristotle’s metaphysics, on issues in ancient Scepticism, and on an intricate set of questions regarding the relationship between Plato and Aristotle. In particular, and not only in the monumental work already mentioned, On Ideas: Aristotle’s Criticisms of

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

12  Virtue, Happiness, Knowledge Plato’s Theory of Forms (1993), Gail has done more than any other living scholar to investigate Aristotle’s criticisms of Plato’s theory of Forms. Gail’s work in all these areas proceeds with an unsurpassed level of attention to argumentative detail and displays an unremitting commitment to philosophical precision, all delivered with her trademark clarity and candour. The beneficiaries—her students, her colleagues, her readers—come away with renewed respect for the ancient authors whose philosophy it has been her life’s work to make vivid. When we encounter her engagements with Plato, or Aristotle, or Sextus, we are treated to philosophical dialectic at its highest level, and we are reminded that the ancient philosophers are, well, philosophers, as we ourselves wish to be. Gail shows that by treating the ancients as the philosophers they claim to be we ourselves stand to become, as she is, philosophers worthy of the tradition they inspire.

Terence Irwin Through his research and graduate student supervision, Terry has had a profound influence on ancient philosophy and the history of ethics. In over four decades in the professorate, Terry had published five monographs, five translations, and well over 100 scholarly articles and helped shape the philosophical communities at Cornell and Oxford universities. He has trained many students who have gone on to successful academic careers, including several who are represented in this Festschrift. The depth and breadth of his work in Greek philosophy and the history of ethics have reshaped the agenda in those fields, for which we are all in his debt. Terry made a dramatic entry into scholarship about Greek philosophy with the publication of Plato’s Moral Theory (1977a). It is common to read Plato’s earlier dialogues as representing the philosophical thought of the historical figure Socrates and later dialogues as representing the thought of Plato himself. Terry defends a striking version of this developmental thesis by contending that there is significant change in philosophical commitments from Socrates to Plato. Focusing primarily on ethical issues in Plato’s early and middle dialogues, he argued that, despite agreement on some claims about the nature and importance of the virtues, Socrates and Plato disagreed on several important methodological and substantive issues. On Terry’s reading, whereas Socrates seeks reductive real definitions of the virtues that eliminate disputed terms in the definiens, Plato defends real definitions of the virtues that are non-reductive. Further, Socrates accepts the unity of the virtues, believes that the virtues are forms of ethical knowledge, and denies the possibility of akrasia or weakness of will. By contrast, Plato conceives of virtue as an affective and conative state that is regulated by ethical knowledge; though his position about the unity of the virtues is not certain, he emphatically embraces the reality of akrasia. Moreover, though Socrates believes that virtue is both necessary and sufficient for happiness, he nonetheless thinks that virtue has only instrumental value. Like Socrates, Plato thinks that the demands of virtue are always

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  13 dispositive. However, Plato rejects the Socratic assumption about the instrumental value of virtue, claiming that virtue is valuable for its own sake as a proper and controlling part of happiness. Despite claiming that virtue dominates happiness, Plato does not think that it exhausts happiness. For Plato, unlike Socrates, virtue can have a price, but it is one that is always worth paying. Terry’s book was arresting, not just for its provocative interpretive claims but also for the resourcefulness of its arguments, both exegetical and philosophical. In addition to a densely argued text, covering Plato’s early and middle dialogues, the book contained an even denser and extremely rewarding set of notes, situating his claims in a rich secondary literature and offering reasons for agreeing or disagreeing with various received interpretations of Plato. The book, especially Terry’s reading of Socrates as an instrumentalist about virtue, proved extremely controversial, but critics recognized its value. The great classical scholar and philosopher Gregory Vlastos, Terry’s supervisor at Princeton, reviewed the book for the Times Literary Supplement, expressing reservations about Terry’s dense analytic prose and his instrumentalist reading of Socrates but also unequivocal praise for it as being ‘rich in philosophical erudition, arresting in its claims, and powerful in its reasoning.’ While taking issue with various aspects of Terry’s reading of Socrates, Vlastos also took care to record his admiration for Terry’s book. Those who might think I am doing so [depreciating the value of Plato’s Moral Theory] would only reveal how sadly they misunderstand the standards by which a work of pure scholarship is appraised in its own domain. If an interpretation is as clear, bold, imaginative, yet painstakingly attentive to its texts and philosophically as rigorous as is Irwin’s, it has more than proved its worth. Its effect on those who work with it, testing it out, text by text and argument by argument, will be bracing, compelling them to bring into sharper focus things which had been heretofore fuzzy in their perception of what is being said in the dialogues and to explore connexions they had never troubled to track down before . . . . For this I am greatly in his debt, which is all the heavier because of the many pages of level-headed and level-tempered criticism his notes have devoted to my own work. I have never had more valuable help from a critic.  (231)

This judicious assessment of Terry’s first book might equally serve as an assessment of his scholarly oeuvre. Shortly after the publication of Plato’s Moral Theory, Terry contributed a translation, with extended notes, of Plato’s Gorgias for the Clarendon Plato series (1979a). The translation quickly became the best English translation of the dialogue, and the notes raise important scholarly questions about the text, not just concerning individual passages but also the role of the Gorgias—especially on topics about happiness, pleasure, virtue, and akrasia—in relation to the Protagoras and the Republic. In Plato’s Ethics (1995a), Terry returned to the themes of Plato’s Moral Theory. Though it endorses the main developmental picture of the relation between Socrates and Plato in the earlier book, it is not a second edition of the first book. Instead, it broadens the scope of the first book by including examination of ethical doctrines in

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

14  Virtue, Happiness, Knowledge Plato’s later dialogues, especially the Philebus, and it aims to provide a more balanced and accessible discussion of Plato’s ethics and rival ways in which it has been interpreted. Terry’s examination of methodological and ethical themes and commitments in various dialogues yields important insights about individual dialogues and a striking developmental narrative. The extended engagement with the ethical argument of the Republic (eight chapters) is masterful. This alone would make Plato’s Ethics the best monograph on Plato’s ethical theory, though its scope, depth, and systematicity make an even greater scholarly contribution. Building on several important articles on Aristotle’s metaphysics and ethics and a translation with notes and glossary of the Nicomachean Ethics (1985a, 1999), Terry’s monograph Aristotle’s First Principles (1988a) was a tour de force. It provides a systematic treatment of Aristotle’s methodological and substantive commitments, ranging from his dialectical methods to his conception of explanation and science, the metaphysics of matter, form, substance, and essence, the soul and action, eudaimonia, the virtues, and political science. Terry begins with an apparent tension between Aristotle’s dialectical methods and his naturalistic realism. How can dialectical methods, in which inquiry is driven by common and respected beliefs, be a guide to objective facts and truths? He argues that the resolution of this tension consists in the pursuit of dialectical methods involving privileged beliefs (which he terms ‘strong dialectic’). Strong dialectic appeals to first philosophy, which involves a particular kind of dialectical study of substances with an essence. Terry reconstructs Aristotle’s application of this conception of first philosophy to substance itself and then to the human soul, the human good, and the ethical virtues, and finally to political community. Though the book is chock-full of resourceful interpretive proposals and philosophical assessments of individual topics in Aristotle, which move the state of scholarship on those topics forward, perhaps its greatest contribution is to provide a sustained and systematic assessment of how Aristotle’s various philosophical commitments hang together in a system that is informed by his commitments to dialectical method and philosophical naturalism. Its combination of interpretive and philosophical scope and detail is unrivaled, and it remains the most impressive single monograph on Aristotle’s philos­ophy as a whole. Though much of Terry’s scholarship concerns various figures and issues in Greek and Roman philosophy, he has always had diverse interests in the history of philosophy, especially the history of ethics. He taught the British idealists at the beginning of his career at Harvard, he regularly taught seminars on Kant at Cornell, and in the early 1980s he began teaching a two-semester sequence in the history of ethics. At Oxford, he has continued this engagement with medieval and modern, as well as ancient, figures and traditions. Terry’s interest in the history of ethics culminated in the publication of his monumental three-volume work The Development of Ethics (2007–9), a masterful reconstruction and assessment of figures, traditions, and ideas in the history of ethics in the Western tradition from Socrates through John Rawls. The three volumes weigh in at over 11 pounds and span 96 substantial chapters and over 2700 densely

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  15 formatted pages. The Development of Ethics covers not only familiar figures, such as Plato, Aristotle, the Stoics, Aquinas, Hobbes, Locke, Hutcheson, Butler, Hume, Smith, Reid, Kant, Hegel, Mill, Green, and Sidgwick, but also a rich variety of ancient sources (including the Cynics, Cyrenaics, Sceptics, and Church Fathers, such as Augustine), medieval, renaissance, and reformation sources (including Scotus, Ockham, and Machiavelli), sources for natural law (including Hooker, Vasquez, Suárez, Grotius, Pufendorf, and Barbeyrac), continental rationalists (including Spinoza and Leibniz), British moralists (including Cumberland, Cudworth, Shaftesbury, Clarke, Balguy, and Price), post-Kantians (including Marx, Schopenhauer, Kierkegaard, Nietzsche, and Heidegger), and twentieth-century Anglo-American sources (including Moore, Ross, Stevenson, Ayer, Lewis, and Hare). This is just a sampling of the more familiar historical sources that Terry discusses. He also discusses a wealth of less familiar philosophical and theological figures. Most chapters are devoted to individual figures, and several figures get multiple chapters (Aristotle gets four, Aquinas nine, Scotus two, Suárez two, Hobbes three, Hutcheson two, Balguy two, Butler four, Hume five, Reid two, Kant seven, Hegel two, Mill two, Sidgwick three, and Rawls two). A few chapters discuss traditions and themes. The combination of scope and depth in The Development of Ethics is without precedent. The broad scope of Terry’s inquiry pays various dividends, not the least of which is that he is able to show that some ideas and themes often taken to be distinctive of modern ethics have their origins and antecedents in antiquity. It is hard to think of anyone else as well qualified historically, philologically, and philosophically to undertake such an ambitious interpretive and philosophical task and carry it out with such authority. Any reader of his three volumes should find the process immensely illuminating though also humbling. Though Terry has a keen sense of the context of the figures and ideas he discusses, he is relentless about understanding and assessing the philosophical content and implications of the texts. Though he provides self-contained reconstructions and assessments of various figures and traditions that make good sense of those figures and traditions on their own terms, two general principles emerge from and guide his discussion. The first principle is a methodological commitment to Socratic dialectic, as refined and practised by Aristotle. This makes Terry’s approach to the history of ethics essentially comparative. In understanding and assessing the philosophical claims of a particular figure or tradition, he finds it fruitful to compare the philosophical commitments and resources of that figure or tradition with the commitments and resources of other figures and traditions. This is valuable, Terry believes, whether those different figures and traditions were in actual and conscious conversation or not. The second principle is a substantive set of commitments that are perhaps clearest in Aristotle and Aquinas but that Terry thinks influence, in various ways, much of the history of ethics. This principle Terry calls ‘Aristotelian naturalism’, which takes the central ethical concept to be a teleological conception of a final good, which should be identified with the agent’s happiness or eudaimonia, where conceptions of eudaimonia

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

16  Virtue, Happiness, Knowledge should be constrained by human nature. Aristotelian naturalism implies that the principal ingredient of the human good must be the realization of our rational nature. The virtues, including the moral virtues, must be understood as essential expressions of this rational nature. If we think of Aristotelian naturalism as receiving its fullest expression in Aquinas, then we can understand a good bit of how Terry approaches the history of ethics. Though few figures accept every element of Aristotelian naturalism, many figures accept one or more elements. Those who accept several elements and who, therefore, figure as central characters in Terry’s narrative are Aristotle, the Stoics, Aquinas, Suárez, Butler, Kant, and Green. Even those who cannot plausibly be viewed as Aristotelian naturalists can still be usefully understood, Terry argues, as contesting one or more elements of Aristotelian naturalism. One may wonder whether Aristotelian naturalism is the best lens through which to appreciate the history of ethics. But there can be little doubt that it provides one especially fruitful framework with which to understand and assess the history of ethics and that Terry has made a powerful case for the plausibility of the essentials of Aristotelian naturalism that every thoughtful reader should take seriously. The Development of Ethics is the most significant contribution to the study and appreciation of the history of ethics to date, and it is unlikely we will see its equal anytime soon. The Development of Ethics is a tough act to follow, and it might seem like a suitable capstone to a distinguished scholarly career, as Terry reaches his mandatory retirement at Oxford. But, as will come as no surprise to those who know Terry, his scholarly activity is not slowing down. He has been working on a series of essays about the relation between ancient and modern conceptions of virtue, value, and duty, and he is now embarked on a critical edition of Aristotle’s three ethical works—the Nicomachean Ethics, Eudemian Ethics, and Magna Moralia—that will combine the Greek texts, translations, and commentaries. Those interested in Greek philosophy and the history of ethics eagerly await the completion of these projects. Terry will leave an unsurpassed legacy of scholarship in Greek and Roman philos­ophy and the history of ethics, one which alters the way in which familiar figures, texts, and traditions can be understood and assessed. Part of this scholarly legacy is reflected in his engagement with and influence on his peers and colleagues, several of whom are represented in this Festschrift. But he will also leave a pedagogical legacy, perhaps equally important. Terry has trained many students in ancient philosophy, the history of ethics, and ethical theory. This influence is reflected in this Festschrift by its three editors and by four other contributors, but this is just the tip of a larger pedagogical iceberg. Among Terry’s many virtues as a supervisor are his willingness to take the views of his students seriously, even when he disagrees with their substance, his prompt and copious feedback, his ability to distinguish the forest from the trees in this feedback, and his insistence that we take seriously alternative readings of the texts and arguments under discussion and not settle for easy victories. Though each of us reflects Terry’s influence in different ways, we have all learned from him the pleasures of doing the history

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

introduction  17 of ­philosophy in a way that combines careful and systematic reading of the texts, philosophical imagination and charity, and respectful and constructive engagement with the secondary literature. For this inheritance, we are greatly in his debt.

The Influence of Irwin and Fine In discussing them individually, we have had occasion as students of Terry and Gail to note with gratitude their dedication to graduate education. They have given us gifts we can never repay, and they have served as models for us as we have turned to our own professional careers in higher education. We are doubtful that we will ever attain the standards they have set. Even so, we may still regard them as Plato regards his Forms, as regulative ideals to emulate. The closer we come, they greater the benefit to our own students, and, more remotely, the greater the benefit to their own students in turn. Such is the length of their legacy. When we pull back from their individual contributions, however, whether pedagogical or professional, we may note in conclusion one final, unique feature of their legacy: though individually unsurpassed, they are corporately astonishing in their reach and influence. With the exception of their joint translation of selected texts of Aristotle, which has become the standard in the English-speaking world, Terry and Gail have gone down individual pathways in their publications. Still, one has the sense that they have often walked together on the path to production, that each has been the first and best critic of the other, that each has just cause to thank the other warmly, as one sees them doing over and over again in their individual acknowledgements. This gives us reason to appreciate them as a philosophical couple, as well as philosophical individuals. To highlight their already entrenched and growing legacy, perhaps it will serve to close by mentioning one small piece of evidence, which gives some faint indication of the reach of their works. When one of the editors of this volume served as a referee for the ancient philosophy section on an annual meeting for the American Philosophical Association, the following astonishing statistic came to light: of the 37 papers received for blind reviewing, fully 36 of them began by citing the work of either Terry or Gail, at times to take issue, at times to augment, at times simply and more neutrally to record the state of play in the issue to be engaged. These submissions jointly give ample testimony to the remarkable and wholly justified influence of these two figures. Their work has been, and will remain, agenda-setting. Indeed, Gail and Terry have set the agenda because they have set the standard. We offer the essays in this volume as testament to the standard they have set.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

1 Rethinking Agreement in Plato Lesley Brown

1.  Introduction: Varieties of ‘Agreeing’ (Homologein, Homologia) Readers of Plato1 are aware how frequently agreeing and agreement feature in his dialogues, in various guises. There are substantive discussions of agreement, such as the famous argument mounted by the personified Laws in Crito to the effect that Socrates by remaining in Athens has agreed to obey the laws,2 as well as critical mentions (in Republic and in Laws) of the thesis that justice is based on an agreement (sunthēkē/homologia). In addition to these explicit discussions of agreeing and agreement as making a promise, the term homologein (usually translated: to agree) and its cognates occur with great frequency, but often almost unnoticed, in everyday philosophical exchanges. Examining them and their implications is the task of this essay. First, I draw some distinctions to identify and describe the use in which I am chiefly interested, which I have labelled declarative homologein. English has both ‘to agree that such and such’ and ‘to agree to do something’ (where ‘to agree’ is roughly to promise or consent). In Greek we find homologein when used of persons with the same two meanings: (1) homologein as to agree that . . . or declare that . . . or concede that . . . and (2) homologein as to agree or promise to do something. The difference between uses (1) and (2) is often marked syntactically, but sometimes only the sense given by the surrounding context makes clear which use is in question.3 I have chosen the label declarative homologein for use (1).4 This label marks the fact 1   This essay is a token of my immense gratitude to Gail Fine and Terry Irwin for the benefit I have derived from their penetrating scholarly work, and for friendship and fruitful discussions over many years. 2   For discussion of the Crito argument, see Kraut (1984) and Brown (2006). 3   Homologein as to promise to . . . usually takes a future infinitive; homologein as to declare that . . . almost always takes an accusative and infinitive construction (or simply an infinitive as at Prot 317b4 homologō te sophistēs einai kai paideuein anthrōpous; I declare/admit that I am a sophist and that I educate people). But no syntactic marker can distinguish whether ta hōmologēmena are things that have been promised or things that have been declared or conceded; usually the surrounding context will make this clear. 4   In Brown (2006) I used Raz’s labels cognitive agreement and performative agreement for (1) and (2) respectively (Raz 1986: 80–2). I now prefer ‘declarative’ to ‘cognitive’ for Greek homologein (1), since

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  19 that the use in question takes a that . . . clause (either expressed or understood, when a respondent is said to homologein in response to a question or thesis put to him). From my study of Plato’s usage, I have concluded that in use (1) homologein is always a public act of asserting (or of indicating assent). This is one respect in which it differs from the English ‘agree that . . . ’, since we can say ‘Mary agreed with Jane that the decision was unfair, but she kept her views to herself.’ Here ‘Mary agreed that . . . ’ reports a matching belief that isn’t declared or made manifest. By contrast, when a person is said to homologein that such and such, this always records a public assenting or agreeing or declaring, not a private believing.5 Before pursuing the declarative use, I mention and set aside a third group of uses. Homologein in use (3) means roughly to be consistent with, or consonant with, or to correspond to. As such it can be used of persons, theses, words, and so on; its opposite is diaphōnein (literally, to be out of tune with). In this use homologein does not designate a speech act.6 Plato uses what I have labelled the declarative homologein with great frequency in philosophical conversations, both in reported dialogue, where the narrator reports a speaker’s assent or his declarations, and in the question and answer exchanges of direct speech, where Socrates (or another questioner) asks ‘do you agree that p?’ or reminds an interlocutor ‘earlier you agreed that q’ or speaks of ‘what was agreed’. The pages of Plato abound in such uses of homologein, or (in English) of ‘he agreed’, ‘do you agree . . . ?’ ‘from what you agreed it follows that . . .’ and so on. At any rate, that is how translations render the plentiful uses of homologein in these types of exchange; they are pretty much uniformly rendered with ‘agree’. But ‘agree’ can, I believe, be misleading as a translation of homologein in its declarative use (homologein that such and such). Here’s why. In English ‘agree that p’ is used in such a way that it typically conveys ‘agree with the speaker that . . . ’, in its second- and third-person uses (you agree(d)/he agrees/agreed that . . .). Sometimes the agreeing described is with a third party. But in English usage, absent any indication of a third party, if I say ‘you/she agreed that p’ or ask, ‘do you/does she agree that p?’ I imply that I too hold the belief that p. In the standard case then, the person who reports another as agreeing that p indicates her own belief that p, and if A asks B ‘Do you agree that p?’ she indicates that she, A, holds that view.7 So, if we read exchanges between Socrates and his interlocutors in terms of agreeing, then we are likely to understand Socrates endorsing

Raz’s ‘cognitive agreeing’ covers believing as well as expressing a belief, while homologein (1) is always expressing a belief. 5   This is something observed from my study of these uses, not a stipulation. Thanks to Christopher Shields for asking for clarification here. 6   Des Places, Lexique (1970) s.v. homologein offers a triage of uses very similar to that I have offered: (1) concéder (que); (2) s’engager (à); and (3) s’accorder (avec). 7   For that reason, the originally proposed question for the 2014 Scottish referendum: ‘Do you agree that Scotland should be an independent country?’ was thought to be biased, and was replaced with the more neutral formulation ‘Should Scotland be an independent country?’

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

20  lesley brown the theses that he extracts from his interlocutors.8 But, from investigating the uses of declarative homologein, I have concluded that it does not convey speaker endorsement. I present the evidence for this in Section 3. Why does this matter? It may make a difference to how we read exchanges between Socrates and his interlocutors, to how we understand the dialectical situation. And it may have some bearing on the debate over the nature of the Socratic cross-examination, the debate between a constructive and a destructive reading. I return to this issue in Section 4.

2.  Two Texts Where Much Hangs on the Implications of Homologein 2.1 In Plato’s Meno, Socrates responds to a complaint from Meno and introduces what has come to be known as ‘the dialectical requirement’, following Irwin’s influential discussion.9 Socrates has offered a definition of shape as ‘the only thing that always accompanies colour’, to which Meno objects: ‘what if someone says he doesn’t know what colour is?’ In reply Socrates distinguishes two ways of handling such an objection, that used by controversialists (eristikoi) and the more dialectical way of proceeding. This is ‘to answer not just by speaking the truth but also through the things that the person questioned first agrees that he knows (prohomologē(i) eidenai)’.10 Some influential critics understand this form of words, with the verb prohomologein to show that Socrates requires, not merely that the respondent say that he knows the things in question, but also that the respondent does know them, if the dialectical requirement is to be satisfied. For example, Irwin writes, ‘Socrates’ dialectical requirement (DR) implies that if A defines x for B as yz, and B does not know about y and z, A’s definition is a bad one.’11 Presuming that the dialectical requirement does require knowledge in the interlocutor (and not merely that the interlocutor say that he knows), critics proceed to discuss what sort or level of knowledge Socrates credits to his interlocutor. Here is Scott’s comment: When he requires that the interlocutor agree that he knows such things, what does he (Socrates) mean by ‘know’? On the evidence of the lines just quoted, we might think that the level of knowledge required is quite informal.12

8   Not all scholars understand ‘agree that . . .’ to have these connotations; cf. n13. My conclusions relate to homologein, not to English ‘agree that’, which I cite only for comparative purposes. 9   Irwin 1977: 136–8. 10   Meno 75d5–7, as translated in Scott 2006: 35. Scott reads prohomologē(i), the MSS read proshomologē(i) (agree in addition), but the choice between them does not affect the issue that concerns me, regarding the import of homologein. I accept Scott’s choice, since the sense of the passage calls for the prefix pro (in advance); cf. Ar Top 110b3 for its use as a technical term. 11 12   Irwin 1977: 136.   Scott 2006: 36.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  21 But rather than ask that question, we should ask instead whether the verb prohomologein does or does not carry the implication that Socrates requires some actual knowledge in the interlocutor. The lines quoted from Scott show that he takes Socrates’ remark, with the verb ‘agree in advance’ to imply that Socrates requires actual knowledge in the respondent. If (as I argue) it doesn’t, then the ‘more dialectical’ way is to proceed by getting the interlocutor first to say he knows the things which will feature in the definition to be offered. If that is the right understanding of prohomologein, then Socrates is not making it essential that the interlocutor actually have any knowledge. On this reading, the dialectical requirement is a rule in a certain kind of philosophical exchange: one which requires getting a certain statement from the other party, but not one that requires actual knowledge of that other person. This approach, embraced by Gail Fine, has the advantage of avoiding the implication that Meno (the respondent at the time) has actual knowledge.13

2.2 Aristotle, Soph El 183b7–8 is another passage where much may hang on the implications of homologein. This is Aristotle’s famous remark about Socrates disclaiming knowledge. Socrates, says Aristotle, used to ask questions but not answer them, hōmologei gar ouk eidenai ‘for he used to agree/acknowledge/maintain that he didn’t know’. Can we infer anything from the choice of verb—hōmologei—about Aristotle’s attitude to the denial of knowledge? I believe we cannot, but that some have done so. For instance, those who translate ‘agreed that he did not know’ or ‘acknowledged that he did not know’ may thereby give the impression that Aristotle is endorsing Socrates’ denial of knowledge, is regarding it as true.14 A different inference seems to have been drawn by Irwin, who writes, ‘[Socrates’] repeated disclaimers of knowledge are too frequent and too emphatic to be dismissed as ironical without strong reason; Aristotle takes them seriously (SE 183b6–8) and so should we.’15 That is, Irwin seems to take the force of Aristotle’s chosen verb of reported speech—homologein—to be that Aristotle took the disclaimers to be sincere. But as I shall argue, the use of homologein shows neither that Aristotle took the remarks to be true—that is, he is not endorsing Socrates’ claim not to know (Sections 3, 5)—nor that he took them to be sincere and non-ironical (Section 5).

13   Fine  2014: 52–3 ‘A second way of defending the view that, as stated in 75d, DR [the dialectical requirement] doesn’t require knowledge is to note that Socrates doesn’t actually say that it does. He says that it requires one to agree that one knows. Perhaps the force of “agree that one knows” is that one must think one knows.’ At n66 ad loc. Fine references arguments by Brown from an unpublished draft of this paper. On the main point we concur: Socrates can say ‘X homologei that p’ without endorsing p, though I am less comfortable than Fine with using ‘agree that’ as a translation without the implication of speaker endorsement. 14   We would be puzzled by the use of ‘acknowledged’ in the following: ‘She acknowledged that she was driving the car when it was in the collision, but this was proved false, as her husband was the driver at the time.’ 15   Irwin 1977: 39–40. Cf. idem 1995: 17 ‘According to Aristotle Socrates made no pretence of knowledge but ‘acknowledged that he didn’t know’ (Top [i.e. S.E.] 183b7–8).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

22  lesley brown Aristotle writes merely that Socrates used to assert that he didn’t know, and we can’t tell (alas) what attitude he took to the famous disclaimer of knowledge.

3.  Texts Where Homologein Does Not Report Any Matching Belief The following texts use declarative homologein to describe an expression of belief (a saying) that does not match the speaker’s (X is not agreeing with the speaker that . . .) and that is not an agreeing with anyone else. The first is from a play of Menander, The Samian Woman, c. 320 bce.

3.1 Menander, Samia at line 524 (cf. 651). Here the context makes crystal clear that the use of homologein does not signify a shared belief, or one that the speaker endorses. The Samian woman, Chrysis, wants to help a pair of lovers, and pretends that the girl’s baby is her own. When the young man eventually explains the situation to his father, he says ‘Chrysis is not the child’s mother, she’s doing this as a favour to me, “homologous’ hautēs” declaring that the baby is hers’. The speaker plainly does not endorse what Chrysis is declaring, nor is he saying that Chrysis is concurring with anyone else in making the claim. In this use of homologein to report Chrysis’ utterances, the usual force of ‘homo’—together with—has been lost; a point I return to in Section 5. It just means: she’s declaring (or claiming or asserting) that the child is hers (a well-meant but false claim, as the speaker has indicated). Next, some passages in Plato where the speaker uses homologein to report the other party’s claim but without endorsing it, and where no other party is indicated as sharing the belief in question.

3.2 Theaetetus 171a9. Socrates is trying to show Theodorus how Protagoras’ thesis (that whatever X believes is true for X) is self-refuting.16 Socrates: Secondly, it has this most exquisite feature: as regards the opinion of those who hold a belief contrary to his opinion (namely the belief that his is false) Protagoras—I presume— concedes (sunchōrei) that theirs is true, seeing that he claims that all men judge what is (homologōn ta onta doxazein hapantas). Theod: Undoubtedly. Soc: And if he agrees (homologei) that the opinion of those who think him wrong is true, then wouldn’t he be conceding (sunchōroi) that his own opinion is false? Theod: Necessarily. Soc: But the others don’t concede (sunchōrousin) that theirs is false? Theod: Indeed not. Soc: But Protagoras, for his part, admits (homologei) this judgement to be true, given what he’s written.  (Theaetetus 171a6–b8)

  I pass over issues of text and interpretation of this argument. Cf. Castagnoli 2010.

16

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  23 It is the first use of homologein, in bold, which provides my evidence. Socrates describes Protagoras as homologōn ta onta doxazein hapantas, ‘seeing as he claims that all men judge what is’ (that is, he claims that all beliefs are true). Now Socrates has just suggested that no one apart from Protagoras holds the thesis that all beliefs are true (170e9–171a2). Hence what he indicates by the phrase in bold is that Protagoras claims or maintains that all beliefs are true. Given what preceded, there can be no implication that anyone shares the belief, despite the use of homologōn. And, of course, Socrates himself does not share the belief; he is not endorsing it.

3.3 Symposium 201b9, part of the so-called ‘elenchus of Agathon’. Socrates is challenging Agathon’s description of Love as beautiful. He has got Agathon to admit that Love desires what is beautiful and that one loves what one lacks and does not possess; so Agathon will have to admit that Love is not beautiful. Next Socrates asks: So do you still maintain (homologeis) that Eros is beautiful, given this is so? (201b9)

To translate ‘agree’ here could well be misleading, if it is thought to imply a shared belief. K.J. Dover in his commentary has noticed a problem, but his comment shows that he clings to the assumption that homologeis marks a shared belief.17 Discussing the line quoted, and specifically Socrates’ question framed with homologeis he writes: ‘“agree (sc. with popular belief)”; Socrates himself does not believe that Eros is kalos’. Troubled by the usage of homologeis, when Socrates evidently does not share the view, Dover takes Socrates’ use of homologein to require that there is some other party with whom the question implies Agathon shares the view. But this is both implausible and unnecessary. This passage shows that Socrates uses ‘homologeis’ simply to ask if Agathon still gives it as his opinion that Eros is beautiful, not if he agrees with Socrates, or anyone else.

4.  Implications for Interpreting Socratic Questioning My reason for this investigation is the belief that readers may make a wrong inference from Plato’s uses of homologein. Readers are prone to infer from such uses that the speaker or questioner endorses the theses in question. One reason for this may be that the verb is standardly translated ‘agree’, which (to many people’s ears) conveys that the one who is said to agree . . . is thereby said to share either the speaker’s belief that . . . , or to share the belief with a third party. Another reason is the prefix homo typically does indicate something shared or common; see Section 5 for further discussion. Now it is a matter of considerable importance for the understanding of Plato’s presentations of Socratic questioning which theses he presents Socrates himself accepting, as I discuss presently.  Dover 1980.

17

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

24  lesley brown Sometimes it is clear from the context that Socrates shares the views on which he is asking his interlocutors questions. The exchanges early on in the Crito (48–9) are a good example. Socrates reminds Crito of what was ‘agreed by us often in the past’ (hōmologēthē) and asks if all those earlier agreements/homologiai have been washed away in the last few days. Here, whatever the force of the homolog terms, the surrounding text makes it clear that Socrates is referring to theses both parties accepted in earlier conversations. Note how Socrates goes on to ask if they share the views, using terms which make it clear that beliefs of both of them are in question.18 Laches 198c is a further passage where the surrounding text makes clear what Socrates’ own view is.19 But often it is unclear where, in Plato’s presentation, Socrates stands. There is a wellknown controversy about the nature of many of Socrates’ inquiries, especially in the so-called elenctic dialogues. Does Socrates believe he can derive positive conclusions from his cross-examinations? On the constructive view (defended in different ways by Vlastos and by Irwin), the answer is yes. Constructivists hold that Plato indicates which premises in a Socratic examination are accepted by both parties, and indicates that they ‘deserve to be accepted’, to use Irwin’s apt phrase.20 Here is Vlastos’ four-step procedure for the constructive elenchus:21 (1) The interlocutor asserts a thesis, p, which Socrates considers false and targets for refutation. (2) Socrates secures agreement to further premisses, say q and r. The agreement is ad hoc; Socrates argues from q and r, but not to them. (Vlastos refers to these as ‘the agreed-upon premisses’ at n30.) (3) Socrates argues, and the interlocutor agrees, that q and r entail not-p. (4) Thereupon Socrates claims that not-p has been proved true, p false. The well-known problem is this: what entitles Socrates to single out the interlocutor’s original thesis, p, as the one to be rejected, on the basis of the interlocutor’s acceptance of the further premises q and r? This is not the place to discuss the different solutions offered by Vlastos and by Irwin. Instead I want to suggest that in Vlastos’ presentation, illicit support for the constructive view may come from what I regard as a mistaken inference from the very common use of homologein when Socrates invites or records submissions from his interlocutor. At Step 2 Vlastos writes about Socrates securing agreement to the extra premises. And on p. 40 he writes: ‘Socrates does of course have reasons for q and r, but he doesn’t bring them into the argument. He asks the interlocutor if he agrees, and if he gets agreement he goes on from there.’ Later in the article, following 18   Crito 49d: So, consider very carefully whether we have this view in common and whether you share my belief (koinōneis kai sundokei). 19   Laches 198b, Socrates asks Nicias if he speaks of the same virtues as he, Socrates. (Yes, Nicias does) ‘So we admit these (tauta men homologoumen), but now see if you hold a different view from me on the following’. It is not the mere use of ‘we agree/admit’ or ‘our admissions’ (because Socrates is prone to say, ‘we claim . . .’ when what he really means is ‘you claim . . .’) but the surrounding context that establishes what Socrates is endorsing. 20 21   Irwin 1995: 20.   Vlastos 1983: 39.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  25 his ambitious attempt to argue that ‘The set of moral beliefs held by Socrates at any given time is true’, he writes, ‘from which it follows that q and r are true, since Socrates has agreed to them’ (p. 55, my emphasis). At this point Vlastos makes clear that that what he called ‘the agreed-upon premisses’, the ones Socrates ‘secures (the interlocutor’s) agreement to’ (Step 2), are ones that—on his interpretation—Socrates as well as the interlocutor has agreed to. On the destructive or negative view, once the prevailing interpretation and defended in recent years by Benson, Frede, and others, the aim of a Socratic examination is simply to show the interlocutor that the theses he has admitted form an inconsistent set, to show him that he has contradicted himself and holds inconsistent beliefs.22 On this destructive reading, when Socrates introduces extra premises into an argument, it doesn’t matter whether or not Socrates himself believes them. All that matters is that the interlocutor professes belief in the premises. They do not need to be selfevident, or generally believed, or believed by Socrates.23 This is Benson’s so-called doxastic constraint.24 My aim is not to adjudicate between these interpretations, but to make a small contribution to the discussion through my reflections on the implications of homologein. As already noted, there are indications that Vlastos may have assumed that, when Socrates asks an interlocutor if he homologei a certain thesis, Plato presents Socrates as seeking agreement for his own view. Or so the passages previously quoted show (Vlastos pp. 40, 55). Now I cannot be sure that Vlastos is relying on a mistaken inference from the frequent use of homologein in Socratic questioning—relying on the idea that the term is used to mark a shared belief—but on the basis of the remarks quoted I suspect it to be so. But there is a small piece of evidence that critics can be misled (as I believe) by the use of homologein even in Benson, who defends a destructive reading. He considers a problem for his view, the worry that by using ‘we agree (homologoumen) that q’ or ‘it is agreed that q’ Socrates ‘is thereby making available to the interlocutor his beliefs’.25 He evidently thinks he has to explain away Socrates’ use of the language of homologein when introducing or referring to extra premises, since it is crucial to his interpretation that in the elenchus the extra premises introduced need not be believed by Socrates. But if my line is correct, Benson need not be troubled by Socrates’ use of homologein to ask for or record an interlocutor’s views, since it carries no implications whatever for what Socrates believes.   Benson 2000; M. Frede 1987: 203–4; cf. idem 1992: 209–14.    23  Cf. Frede 1992: 212.   Benson 2000 introduces the ‘doxastic constraint’ at 37. Irwin 1993 has a valuable discussion of some rare occasions (chiefly in Prot and Gorg) where Socrates relaxes the ‘say what you believe’ rule to explore a thesis believed by others but not the interlocutor. 25   Benson 2000: 47n52: ‘It might be thought since Socrates frequently introduces the premises of the elenchus with such phrases as “we agree (homologoumen) that q, don’t we” or “it is agreed that q, isn’t it?” (see for example Laches 198b2) that he is thereby making available to the interlocutor his beliefs’. Benson goes on to reply to this worry. Here’s my reply. We can explain the use by Socrates of ‘we agree’ by noting it as a polite use, meaning simply you agree (i.e. you declare). As to ‘it was agreed’, if this is a passive of homologein, then Socrates does not thereby make available to the interlocutor his own belief, but records what the interlocutor has earlier maintained. 22 24

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

26  lesley brown I do not claim that my reflections damage the case for a constructive reading, especially not one so carefully argued as Irwin’s.26 I am simply cautioning against taking Socrates’ uses of homologein as asking for or recording agreement with his, Socrates’, views.

5.  Confrontation Rather Than Consensus? Many contexts show that the practice of asking for or recording an act of homologein from another is not invariably the friendly pursuit of consensus. First, it is clear that the term often has a forensic ring, as if the setting is like a law court, and indeed the term is often found in orators’ speeches.27 We find many texts where ‘deny’ (exarnon einai) is the alternative to homologein.28 In such exchanges, the purpose is to pin the interlocutor down, to extract some affirmations or denials, not to reach a consensus. The orator who asks for or records a homologein from the accused is not revealing anything about his own beliefs; he is not endorsing what he records the accused as having said. Second, the plentiful uses of homologein terms in Euthydemus provide evidence for my thesis. The bulk of the dialogue consists of Socrates’ narration of the eristic tricks played by the brothers Euthydemus and Dionysodorus on Socrates and his companions, while in the protreptic interludes Socrates discusses serious matters in a serious and non-eristic manner. The verb homologein is used with great frequency in both the eristic and the protreptic stretches, but it is the uses in the eristic parts that provide my evidence. There are, as we expect in a narrated dialogue, many uses of the verb (usually in the imperfect, hōmologei) to report a subject’s saying yes to a question.29 But we also find it used in direct speech; the remarks of the brothers bristle with demands that their interlocutor homologei—make a claim—one way or another, or triumphant remarks that the interlocutor made a certain claim, using homologein.30 Now it is perfectly plain in Euthydemus that the brothers’ aim is to get their victim to contradict himself, or to give voice to a shocking or absurd claim. There is no attempt to reach a consensus, and the brothers are not even interested in finding out what the interlocutor really believes. To illustrate the point, I turn first to 295a, where we find the ‘Know everything always’ fallacy. Socrates has already reported how the brothers haven’t shied away from declaring that they know (homologountes eidenai, 294d7) all sorts of shocking things.   For example, in Irwin 1995, ch. 2.   For frequent use of homologein, with its antonyms exarneisthai and antilegesthai, in forensic speeches see Demosthenes Or 27: 30. 28   Protagoras 317b6, Protagoras speaking: I admit that I am a sophist and that I educate people; I think that’s a better way of protecting myself than that other one—admitting it rather than denying it. Charmides 178c Socrates speaking: Charmides said that it wasn’t easy either to admit (homologein) or to deny (exarnon einai) what I’d asked him (viz. whether he is self-controlled). 29   E.g. 276a–277c has six occurrences of Hōmologei reporting that Cleinias said yes to the brothers. 30   Between 295a and 302e there are 15 occurrences of a form of homologein in direct speech, in questions and answers, as the brothers question Socrates and Ctesippus. 26 27

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  27 Next the tables are turned; and Euthydemus tries to get Socrates to homologein that he knows everything always. The questioner does not allow him to enter the appropriate qualifications (as when Socrates wants to say: I always know everything I know with the same thing—the soul). ‘If you’ll answer me, I’ll display you too homologounta these amazing things’ (295a5): Euthydemus thus boasts that he will make Socrates utter words which commit him to the absurd claim. No one actually thinks Socrates knows everything, or that he believes he does. And for sure Euthydemus, in trying to get Socrates to homologein, to give voice to this claim, is not trying to get him to agree with any belief he (or anyone else) holds, for the obvious reason that no one believes that Socrates knows everything always. Here’s one more example. At 301c–d Dionysodorus tricks Socrates by misconstruing his answer—that a butcher is the person to cut up and skin—as saying it is appropriate to cut up and skin a butcher (the trick involves reversing subject and object).31 ‘Did you admit/declare this (hōmologēsas) or not?’ Socrates: ‘yes, I admitted it (hōmologēsa); forgive me’. Of course, neither party believes that one should cut up a butcher, or that Socrates was sincerely expressing that belief. Getting Socrates to homologein is getting him to commit himself to something that can be construed as an assertion to that effect. It may be objected that one should not liken the sophistical refutations that the brothers in Euthydemus delight in to Socratic questioning. One important difference is that Socrates standardly demands that his associates are sincere when giving their answers.32 Without denying that difference, I appeal to the frequent use of homologein by all the questioners (as well as by the narrator) in Euthydemus to back up two claims. First, the language of homologein should not be taken as indicating or inviting consensus or agreement with the questioner (or with anyone else). Second, the verb can be used of a statement that all present know is not the sincere opinion of the one who is said to homologein such and such. Indeed, it is clear also from Crito 49d1 that to describe someone’s avowal in terms of homologein is not to treat it as sincere; there Socrates warns Crito ‘be careful not to homologein these things against your own belief ’.33 Objection: what about the homo- prefix of homologein? Does that not indicate that the belief whose avowal is reported is shared by someone else? Or that the one who homologei says the same as someone else? Standardly the force of the prefix homo- is indeed that of ‘together’, or ‘alike’. No doubt in origin that was the effect of the prefix in all uses of homologein, and the homo- prefix certainly retains its force in other uses of the verb, notably in what in Section 1 I labelled use (3), when a person or thing is said to correspond to or be in harmony or consistent with another. And there are many 31  Dionysodorus: prosēkei de ge, hōs phēis, ton mageiron katakoptein kai ekderein; hōmologēsas tauta ē ou; Soc: hōmologēsa, ephēn, alla suggnōmēn moi eche (301d2–4). 32   Exceptions to the sincerity requirement are discussed by Irwin  1993 and by Kahn  1992: 255–6. I return to the issue in Section 6. 33   kai hora, ō Kritōn, tauta kathomologōn, hopōs mē para doxan homologēis. (The verb kathomologein occurs twice in Plato, here and at Gorgias 499b4. It evidently means: say yes again and again.)

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

28  lesley brown other verbs where the prefix homo-retains the force of jointly or together with, such as homodoxein and homonoein, both meaning roughly to believe the same as someone else. But I take the passages cited in Section 3 to show that when A says that B ‘homologei’ in what I have labelled the declarative use, A does not imply that the view expressed by B is shared by A or by anyone else. It seems that, in the declarative use of homologein, the homo- prefix has entirely lost its force. This is undoubtedly the case in later writers, and, if I am right, it is so already in Plato.34 Alternatively, in the uses I am exploring, the only force the prefix retains is that one who homologei says/expresses belief in/the same as the logos says (but not necessarily the same as someone else says or believes).35 How then, it may be asked, can we tell whether Plato is indicating mutual agreement between speakers, if (as I claim) the mere use of homologein does not do so? As already noted, Plato has a variety of ways of making it clear when he is representing his speakers as sharing a belief.36 And we find mutual agreement indicated by middle voice forms of anomologeisthai or of diomologeisthai; the latter is prevalent in Republic and Sophist, where many of the conversations are explicitly represented as shared searches for the truth.37

6.  Homologein as Declaring, and as Promising: Similarities and Differences Many scholars have seen analogies between agreeing as declaring or asserting, and agreeing as promising. They compare what Socrates expects of his interlocutors when in answer to a question they declare that such and such with what is expected of one who makes a promise to do something. Indeed, Plato’s text seems to encourage such reflections. In Crito 49aff. Socrates starts by recalling the declarative agreements Crito has made in the past, when he has accepted certain ethical principles, and he then turns to agreeing as promising (49e6–7, recalled at 50a2–3).38 At 51e Plato makes the 34   In the New Testament homologein standardly means to profess or profess belief in, with no implication of agreement in the sense of a shared belief. 35   For a later case where a similar prefix has lost its original force, cf. sugkatatithesthai, whose original meaning was to deposit (e.g. an opinion) together with someone else, as at Gorg 501c. By the time sugkatathesis is used for ‘assent’ sc. to impressions in Stoicism, the sun prefix has lost that original meaning: perceiving subjects accept the content of an impression, but don’t cast their lot or vote together with someone. (Thanks to Tobias Reinhardt for the point.) 36   Sec. 4, with n18 and n19. 37   The verb sugchōrein might be thought suitable to indicate mutual agreement, and some uses of sunechōrei to indicate assent in Euthydemus seem to bear this out, as when it is used to record Cleinias’ assent to theses Socrates puts to him (280a5, 281a1, 281c8). However, at 286c9 it indicates Ctesippus’ conceding an outrageous thesis (that in all cases either one says the truth or says nothing). Thus, like homologein, it is used to indicate an interlocutor’s conceding a thesis. Adorno 1996 argued that homologein indicates rational assent while sugchōrein has emotional or affective overtones, but the evidence does not bear this out. 38  49e6–7 poteron ha an tis omologēsēi tōi dikaia onta poiēteon ē apatēteon; should one do the things he has promised to another, provided they are just, or should one play false? See Brown 2006: 76 n6 for a defence of this construal.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  29 Laws argue that Socrates has agreed to obey the laws of Athens by remaining in his birthplace. And Plato uses the same verb, emmenein, for sticking to what one has agreed, both for continuing to assert what one has previously asserted (e.g. at Crito 49e3–4) and for keeping a promise (at Crito 50a2). Here (from Irwin) is one example of a critic using an analogy between agreeing as promising and agreeing as declaring one’s beliefs. Discussing the Crito, he writes: ‘The citizen is free to make or withhold agreement to the laws, and free to cancel agreement by emigration, whenever he likes . . . . Socrates imposes only this sort of obligation on Crito: Crito must accept conclusions only if he thinks they follow from premises he accepts, but he is free to reject the premises or inferences as he pleases. Argument requires only the sort of good faith required of a good citizen . . . The interlocutor is not coerced, but simply required to abide by his agreements.’39 The point that interests me is that of the obligation to ‘abide by one’s agreements’ in the two different cases. We find Vlastos too relying on an analogy between giving up a premise one had asserted (‘agreed’) and failing to keep a promise, in his discussion of the elenchus. He writes: ‘Whenever Socrates proved to his interlocutors that the premisses they had conceded entailed the negation of their thesis, why couldn’t they hang on to their thesis by welshing on one of the conceded premisses?’40 Thus Vlastos, using a now discredited term for refusing to pay money one owes, likens changing one’s mind in argument to failing to keep a contract or promise.41 Kraut (restating Vlastos’ query) uses a less colourful expression to make the same point: ‘Why can’t the interlocutor simply renege on his agreement to accept this or that premise? Or why can’t another interlocutor . . .  refuse to make the same agreements?’42 All these discuss the interlocutors’ ‘agreements’ (that is, their declarations) in terms drawn from the sort of commitments made by one who promises. However, I want to argue that there is an important disanalogy between the two situations—changing one’s mind about what one believes, and failing to keep a ­promise—and that Plato is fully alive to it. We noted that he uses the language of sticking to or abiding by what one has agreed (emmenein) in both cases. But the difference Plato rightly signals is this. The mere fact of not keeping your promise is prima facie wrong, and to do so is to play false, exapatan (Crito 49e7). But failing to abide by what you have declared you believe is not in itself wrong or deceitful. Indeed, Socrates often makes it clear to the interlocutor that he is free to change his mind if things now appear differently to him.43 What Socrates warns his interlocutor against is an undeclared change of mind, as in Republic 345b7–9, where Socrates tells Thrasymachus: Stick to (emmene)

40   Irwin 1986: 57.   Vlastos 1983: 49–50.   OED Welsher: a bookmaker at a race meeting who takes money for a bet and absconds or refuses to pay if he loses. 42   Kraut 1983: 61. 43   Crito 49de; Prot 349c7–d2. This is another respect in which Socrates’ practice differs from eristic confrontations, where a respondent may not withdraw a concession. 39 41

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

30  lesley brown the things you said, or, if you change position (the verb is metathesthai), do so openly and don’t deceive us (mē exapata). The reason why there is typically a moral difference between changing one’s mind (failing to stick to one’s beliefs) and not keeping a promise is related to the different ‘direction of fit’ with the truth in each case. To promise is to undertake to make something true in future, while to affirm (‘agree’ in the declarative use) is to express a belief that something is true. Here’s the difference. A belief (or expression of belief) is at fault if it doesn’t match the truth. But a promise does not have to reflect reality as it is now; rather, the job of the one who agrees/promises is to make things come out as promised, to do what they promised to do. When someone has made a promise, they cannot just say: I have changed my mind and it no longer seems a good idea to do what I promised. But it’s different with expressions of belief. To change one’s belief is not in itself reprehensible, quite the reverse if one has seen good reasons to do so. Pace Vlastos, to change one’s mind about what to believe should not be likened to reneging on an agreement (in the sense of a promise or contract), and Plato is quite alive to the important difference, as the citation from Republic 345b shows.

7. Conclusion My focus has been on how Plato uses homologein and its cognates when he represents (whether in a narration or in direct speech) those situations when a questioner seeks the opinions of the respondents, asking for answers, recording them, harking back to answers already given, and so on. Sometimes Plato uses the plain legein or phanai to phrase the questions or reports: do you say/did you say/earlier you said and so on. But very often his verb of choice is homologein, almost always translated ‘agree’. It is natural to expect that the prefix homo-adds something to the plain legein: not just ‘do you say . . . ’, but ‘do you say together with . . . ’, or ‘do you say the same as . . . ’? But in this essay, I have marshalled evidence that speaks against that expectation. Section 3 presented some texts which indicate that the effect of homologein is (in that respect) no more than plain legein, to say.44 That is, we find it used in contexts where the speaker, in saying that X homologei that such and such, is not implying that the belief X expresses thereby is shared by the speaker or by anyone else. It might be proposed that the uses cited in Section 3 are exceptions, and that standardly homologein is used to mark someone’s expressing agreement with the speaker or with someone else. Examination of the plentiful uses of homologein in Euthydemus has suggested otherwise. In Section 5, I showed how when the brothers invite their opponent to homologein one way or another, or predict that he will homologein that such and 44   The choice of homologein, in place of the plain legein, may be to add weight, to convey that more is at stake than mere saying. The verb is used to mark declarations that the speaker (whether in a law court, or in a philosophical discussion) is supposed to stand by. The perfect tense hōmologēkas (frequent in Euthyd) is used to remind an interlocutor of what he is committed to, given what he has already maintained.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Rethinking Agreement in Plato  31 such, they are simply requiring the opponent to make a claim or predicting that he will claim that . . . . They are asking the other person to say something and to stand by it, not to say something in agreement with themselves or anyone else. The interlocutor is required to agree to one thesis or another, but not thereby to agree with anyone.45 And there is no warrant for assuming that anything different is implied when Socrates uses the term to record or to request answers from his interlocutors. This point is unaffected by the well-known and important differences between Socratic questioning and eristic practice, already touched on: viz., that unlike the eristic brothers, Socrates generally demands sincere answers, and he allows his interlocutors to withdraw their answers if their beliefs have changed (cf. Section 6). For all those differences, the implications of (declarative) homologein are identical in any context of use, or so my investigations suggest. When an interlocutor is asked to homologein, or when he declares that he has agreed to a thesis (hōmologēsa), no one else’s beliefs are indicated thereby. Nothing can be inferred, from the use of homologein, about the beliefs of anyone other than the person who has made the claims or concessions; the use of the term does not imply any other person who shares the belief in question.46 There are some indications that readers of Plato have thought otherwise (as indeed I once did myself), and have used what I take to be unwarranted assumptions in their interpretations of Socratic questioning. Hence this exercise in rethinking agreement—or to be precise, in rethinking the implications of declarative homologein—in Plato.47

References Adorno, F. (1996) ‘Appunti su omologein e omologia nel vocabolario di Platone’. In: Adorno, F., Pensare Storicamente, Firenze: L. S. Olschki, pp. 49–65. Benson, H. (2000) Socratic Wisdom. New York: Oxford University Press. Brown, L. (2006) ‘Did Socrates agree to obey the Laws of Athens?’ In: Remembering Socrates (eds L. Judson and V. Karasmanis) Oxford: Oxford University Press. Bury, R.G. (1932) The Symposion of Plato. Cambridge: Cambridge University Press. Castagnoli, L. (2010) Ancient Self-Refutation. Cambridge: Cambridge University Press. 45   Section 5, last paragraph, notes that middle forms of the verb are used to express mutual agreement. R.Geiger, Dialektische Tugenden (2006), pp. 80–1 notes that one should observe the distinction between ‘einer Aussage zustimmen’ and ‘einer Person zustimmen’, and again between Zustimmung and Übereinstimmung. He does not, however, tie these to specific Greek expressions. 46   One more instance: at Symp 223d2–6 it is related that, at the end of a long night’s drinking, the sober Socrates forced Agathon and Aristophanes to homologein that to know how to write comedies and to write tragedies belongs to one and the same man. The narrator’s term homologein indicates that the drunken poets were forced to assent to that thesis, but not that it was one Socrates himself accepts (pace many commentators on the passage, for example Bury 1932: 171 ‘the thesis here maintained by Socrates’). 47   This paper offers a segment of my research on agreement terms in Plato, versions of which have been presented at University College, London (2012), at symposia in Bordeaux and in Pisa (International Plato Symposium 2013) and at Cornell’s colloquium to honour Gail Fine and Terry Irwin (2013). I am very grateful to audiences at all those places for their questions, to Gail Fine for her unfailingly helpful comments on an earlier version, and to Christopher Shields whose comments incited me to clarify many points.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

32  lesley brown Des Places, E. (1970) Platon, Oeuvres Complètes, Tome XIV, Lexique. Paris: Les Belles Lettres, 1970. Dover, K.J. (1980) Plato, Symposium. Cambridge: Cambridge University Press. Fine, G. (2014) The Possibility of Inquiry. Oxford: Oxford University Press. Frede, M. (1987) ‘The skeptic’s two kinds of assent’. In: Essays in Ancient Philosophy. Oxford: Oxford University Press, pp. 203–4. Frede, M. (1992) ‘Plato’s arguments and the dialogue form’. In: J. Klagge and N. Smith (eds) Methods of Interpreting Plato and his Dialogues. Oxford: Oxford University Press, pp. 201–19. Geiger, R. (2006) Dialektische Tugenden. Untersuchungen zur Gesprächsform in den Platonischen Dialogen. Paderborn: mentis-Verlag. Irwin, T.H. (1977) Plato’s Moral Theory. Oxford: Oxford University Press. Irwin, T.H. (1986) ‘Coercion and Objectivity in Plato’s Dialectic’. Revue Internationale de Philosophie 40 (56/7), 47–74. Irwin, T.H. (1993) ‘Say what you believe’. Apeiron 26 (3), 1–16. Irwin, T.H. (1995) Plato’s Ethics. Oxford: Oxford University Press. Kahn, C. (1992) ‘Vlastos’ Socrates’. Phronesis 37, 233–58. Kraut, R. (1983) ‘Comments on Vlastos’. Oxford Studies in Ancient Philosophy I, 59–70. Kraut, R. (1984) Socrates and the State. Princeton: Princeton University Press. Raz, J. (1986) The Morality of Freedom. Oxford: Oxford University Press. Scott, D. (2006) Plato’s Meno. Cambridge: Cambridge University Press. Vlastos, G. (1983) ‘The Socratic Elenchus’. Oxford Studies in Ancient Philosophy I, 27–58.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

2 Plato’s Theory of Knowledge Ralph Wedgwood

0. Introduction In his middle-period dialogues, Plato worked out at least the rough outlines of a ­distinctive theory of knowledge. According to the chronology that seems most ­plausible, he first sketched the outlines of this theory in Meno; then he revised and developed the theory further in Phaedo and the Republic; and finally, in Theaetetus, he probed the theory in depth, by subjecting it to a series of searching questions. In this essay, I shall propose an interpretation of the version of the theory that is presented in Phaedo and the Republic, although I shall also consider what seems to be the slightly earlier version of the theory that we find in Meno. Unfortunately, I shall not be able to pay more than the most glancing attention to Theaetetus. My interpretation will focus, as narrowly as possible, on Plato’s view on the ­question, ‘What is knowledge?’. I shall touch only briefly on Plato’s views on related questions, such as ‘What is belief?’ and the like. My interpretation of Plato’s view on this question will be at least in part conjectural: although I shall show that it fits much of the textual evidence, I shall not be able in the available space to check this interpretation against all the relevant textual evidence, nor shall I be able to argue in detail that this interpretation is preferable to all alternatives. The primary goal of this essay is just to put this interpretation forward for consideration. No contemporary scholar of ancient philosophy has studied Plato’s epistemology in greater depth than Gail Fine.1 The debt that my interpretation owes to her work is both obvious and immense. I shall admittedly disagree with her on one crucial point: I shall accept a version of the traditional interpretation of Republic 476e–480a, according to which what in that passage is called ‘knowledge’ (epistēmē or gnōsis) consists in a kind of grasp of the Forms, whereas what in that passage is called ‘opinion’ (doxa) consists— at least in a sense—in a kind of grasp of perceptible concrete things. On many other points, however, I shall accept Fine’s interpretation. Moreover, I shall also follow her methodology—which involves not just the most rigorous philosophical analysis, based on a painstakingly close reading of the primary texts, but also an attempt to   See the papers collected in Fine (2003), and also Fine (2004, 2008).

1

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

34  Ralph Wedgwood bring Plato into a dialogue with contemporary epistemologists, in a way that looks more for continuities between Plato’s thinking and that of our contemporaries than for contrasts or dissimilarities. While it is agreed on all sides that it is important to avoid anachronism, I strongly agree with Fine’s belief that a consideration of such contem­ porary ideas is often helpful in understanding Plato’s thought. Briefly, my interpretation of Plato’s theory of knowledge is the following. 1. Plato is a kind of contextualist about words like ‘knowledge’. The heart of Plato’s theory is an account of four different levels of cognitive mental states, which he illustrates with the image of the four segments of the Divided Line (Republic 509d–511e). But as Plato explicitly admits, sometimes he uses his principal term for ‘knowledge’ (epistēmē) to cover both of the upper two levels, and sometimes just for the highest level. There are also indications that his usage of other Greek words for ‘know’ (such as eidenai and gignōskein) and of the term doxa (‘belief ’ or ‘opinion’) varies with context as well. I shall explain the sense in which this makes Plato a contextualist in the next section. 2. Plato assumes that knowledge is a factive mental state, belonging to the inclusive genus of cognitive states that also includes all kinds of belief or opinion. As I  shall argue, one distinguishing feature of knowledge, in his view, is that it must satisfy the condition that contemporary epistemologists call ‘adherence’. Indeed, he may think that genuine knowledge must satisfy adherence to the highest degree—that is, in effect (as I shall explain), genuine knowledge must be utterly indefeasible. If he does think this, it would explain why he also holds that all genuine knowledge is a priori—which, he speculates, is best explained by the Theory of Recollection. 3. It is also plausible that Plato thinks that every truth that can be known is necessary. Together with the explanation of knowledge that is based on the Theory of Recollection, this guarantees that genuine knowledge also satisfies the condition that contemporary epistemologists call ‘safety’. Indeed, knowledge is safe to the highest degree—that is, infallible. If, as seems plausible, Plato assumes that all truths that are necessary in this way are in some sense aspects of the Forms, this vindicates the traditional interpretation that he holds that all genuine knowledge consists in a grasp of some aspect of the Forms. 4. For Plato, knowledge always requires at least some grasp of the explanation of the truth that is known. The truth about the Forms constitutes an intelligible explanatory system of necessary truths; and the different levels of knowledge correspond to the different degrees to which the thinker grasps this explanatory system of necessary truths. It may be that all adult human beings have at least a  rudimentary grasp of some fragments of this system of these necessary truths. But no human being has yet achieved the highest level of knowledge, which  would consist in a coherent synoptic grasp of the whole system of ­necessary truths.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  35 5. We can use our understanding of Plato’s terminology to show how the texts ­support this interpretation. As Plato admits, he switches between using epistēmē strictly, to refer to the highest level of cognition, and using it more loosely, to refer to both of the two upper levels together. In some other dialogues, he uses doxa as a generic term for the genus of which all four levels are species; but in the Republic, he mostly uses doxa for mere belief—a cognitive mental state that falls short of counting as epistēmē—though what this means depends on how in  turn the term epistēmē is being understood. Moreover, Plato sometimes allows himself to use some of the Greek words for ‘know’ (such as eidenai and gignōskein and their cognates) in an idiomatic sense, to stand for what in fact, according to his official theory, is just a true belief of a relatively reliable kind (in his terminology, confidence or pistis). In what follows, I shall take these five points in turn.

1.  Plato’s Contextualism There are two famous passages where Plato contrasts knowledge and opinion: Meno 98a, and Republic 476e–480a. But later in the Republic, this duo of contrasting states is elaborated into a quartet, when Plato gives an account of four different levels of ­cognitive state, illustrating these states by the image of the four segments of the Divided Line (509d–511e). The names that he gives to these four states there are: intellection (noēsis), thought (dianoia), confidence (pistis), and imagination (eikasia). It is natural to wonder how the pair discussed earlier (knowledge and opinion) are related to the quartet that is discussed later (intellection, thought, confidence, and imagination). Plato answers this question explicitly, in a slightly surprising way, in a passage in the middle of the description of the education of the guardians in Book VII. In this passage, Socrates contrasts ‘dialectic’—which is his name for the highest form of intellectual inquiry—with five other branches of learning (namely: arithmetic, the two-dimensional geometry of planes, the three-dimensional geometry of solids, astronomy or the four-dimensional geometry of motion, and harmonic theory), and makes the following comment about these other branches of learning (533d–534a): From force of habit, we have often called these branches of learning kinds of ‘knowledge’. But they need another name, clearer than ‘opinion’ and darker than ‘knowledge’. We defined it as ‘thought’ somewhere before. But I don’t suppose we will dispute about names, with matters as important as these before us to investigate . . . . It will be satisfactory, then, I said, to do as we did earlier and call the first portion ‘knowledge,’ the second ‘thought,’ the third ‘confidence,’ and the fourth ‘imagination’. The last two together we call ‘opinion’, while the first two we call ‘intellection’. Opinion is concerned with becoming; intellection with being. And as being is to becoming, so intellection is to opinion; and as intellection is to opinion, knowledge is to confidence and thought to imagination.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

36  Ralph Wedgwood Here Plato clearly characterizes both the two lower levels (confidence and ­imagin­ation) as species of ‘opinion’ (doxa). This characterization would be pointless if doxa were in fact a wide genus that included all four levels of cognition. So this is one of the passages in the Republic where he uses the term doxa for mere belief—that is, for a kind of cognition that falls short of knowledge. However, he also explicitly admits that his terminology is not invariant between ­different contexts: he says that earlier he called both of the two upper-level kinds of ‘knowledge’ (epistēmē), though here he now prefers to use this term more strictly, so that it refers only to the uppermost level. Moreover, without explicitly taking note of the fact, he also switches around his use of the term ‘intellection’ (noēsis) as well. When he gave the image of the Divided Line at the end of Book VI (509d–511e), he used the term ‘intellection’, not to refer to the upper two levels together, but just to refer to the highest level alone. Here, however, in Book VII, he explicitly uses the term ‘intellection’ for both of the two upper levels together. So, in effect, Plato has switched around the terms ‘intellection’ (noēsis) and ‘knowledge’ (epistēmē) between the ­discussion of the Divided Line in Book VI and this later passage at 534a. It is clearly more charitable for us to interpret this switch as deliberate, rather than as a mere slip on Plato’s part. By switching his terminology around in this way, Plato seems to be warning us against assuming that the same word always expresses the same concept and refers to the same entity. On the contrary, we need to look at the underlying structure of the concepts that are being expressed, and not just at the terminology that is being used. Fundamentally, there are these four levels of cognition, which following Fine (1990), I shall refer to as L1, L2, L3, and L4; and there are also the various genera to which these four levels of cognition belong. Our fundamental task as epistemologists is to understand these different species and genera of cognitive states, not to worry about the ­terminology that we use to refer to them. With respect to the terminology itself, however, Plato is in a sense a contextualist. In different contexts, he uses cognitive terms like epistēmē to refer to different mental conditions—sometimes using the terms to refer a more restricted condition, which includes only L4 (the highest level of cognition), and sometimes to refer to a more inclusive condition, which includes both L3 and L4 (the upper two levels) together. We shall later see evidence that he also uses the other Greek words for ‘know’ (such as eidenai and gignōskein) in a similarly flexible way as well. Admittedly, the text does not definitively establish whether Plato is a full-fledged contextualist, like such contemporary philosophers as Keith DeRose (2009), among others. To be such a full-fledged contextualist, he would have to think that cognitive terms like epistēmē are not ambiguous, but have a single univocal meaning, and that in using the terms so that their extension shifts between contexts in this way, he is using the terms strictly in accordance with this univocal meaning. In other words, according to contextualism, it is part of the univocal meaning that cognitive terms like epistēmē have that their extension shifts between contexts in this way. Nonetheless,

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  37 the  text is at least compatible with Plato’s being a contextualist in this full-fledged sense; and so I have taken the liberty of referring to Plato’s flexible use of this cognitive vocabulary as “contextualist.” Finally, it is also fairly clear that Plato’s use of the term for ‘belief ’ or ‘opinion’ (doxa) varies with context as well. Sometimes, as at this point in the Republic, he uses it for mere belief—that is, for the kind of belief that falls short of counting as ­knowledge. But in other contexts, he uses it more broadly for the genus of which all four l­ evels are species. This seems to be the usage that we find in the second half of Theaetetus, where both of the two definitions that Plato considers—that knowledge is true belief (187b), and that it is true belief accompanied with an account (201c–d)—clearly imply that knowledge is a species of belief. On the relation between knowledge and belief, it is not Plato’s view but only his ­terminology that changes between Republic and Theaetetus. In both works it is assumed that there is a wider genus of cognitive states, and that both knowledge and mere belief are species of this wider genus. The only difference in the terminology is that sometimes (as in Theaetetus) Plato uses the word ‘doxa’ for the wider genus, and sometimes (as in the Republic) Plato restricts the word for the species of belief that falls short of knowledge. We shall return to this understanding of Plato’s terminology in the last section, when we survey some of the evidence for and against the interpretation that will be defended here. As we shall see, this understanding of Plato’s terminology will help us to evaluate this evidence in a more precise and discriminating way.

2.  Adherence, Indefeasibility, and the a Priori Towards the end of Meno, Socrates makes the following suggestion about the differ­ ence between knowledge and mere true belief (97e–98a): True opinions are also a fine thing and altogether good in their effects so long as they stay with one, but they won’t willingly stay long, and instead run away from a person’s soul, so that they are not worth much until one ties them down by reasoning out the explanation. And that is recollection, Meno my friend, as we agreed earlier. And when they’ve been tied down, then for one thing they become items of knowledge, and for another, permanent. And that’s what makes knowledge more valuable than correct opinion, and knowledge differs from right opinion by being tied down.

Here, Plato suggests that knowledge differs from mere true opinion because true opinions ‘run away’, while knowledge is ‘permanent’. According to Bernard Williams (1978, 38), this suggestion should not be interpreted as the ‘the blankly psychological proposition (in any case, surely, very dubious) that one is more disposed to forget what one merely believes than what one knows’. Instead, it should be interpreted as the ‘point . . . that knowledge cannot rationally be rendered doubtful’—whereas mere beliefs can ‘rationally be rendered doubtful’. For a belief to ‘rationally be rendered

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

38  Ralph Wedgwood doubtful’, I assume, is for the belief to be given up because it is rationally undermined. Let us also assume that in all the cases that concern us, the belief in q ­ uestion will be given up if and only if (and because) it is rationally undermined. In effect, then, the suggestion is that knowledge (unlike mere belief) cannot be ­rationally undermined. Williams interprets this suggestion as implying that knowledge cannot be rationally undermined. This interpretation is plausible—at least on some understanding of ‘can’. Plato is presumably not suggesting merely that knowledge is ‘permanent’ in the same way as a belief that by a strange fluke just happens never to be undermined. Instead, the suggestion is that whereas with mere beliefs, it could easily happen that these beliefs are rationally undermined, it could not easily happen that any genuine knowledge is rationally undermined. To say that it could ‘easily happen’ that a belief is undermined is to say that there are possible worlds that are sufficiently similar to the actual world (in the relevant respects) in which the belief is undermined. Presumably, when a belief is rationally undermined, it is undermined by new information or new reflections. There seem to be two ways in which this could happen: 1. The true belief might have been unjustified all along, and the new information or reflections might prompt the believer to think more rationally, and to give up the belief. 2. The true belief might originally have been rational and justified, but the new information might defeat that original justification. Cases of the first kind (1) are true beliefs that were never justified at all; such cases were never cases of justified true belief (JTB). Cases of the second kind (2), however, involve a JTB whose justification is later defeated. According to Plato, the fact that the belief ‘ran away’ or was rationally undermined in this way shows that it was never a piece of knowledge in the first place. Thus, if we agree with Plato’s judgment on cases of this second kind, we must conclude that they are cases in which a JTB fails to count as knowledge—or in the terminology that has become common after the work of Edmund Gettier (1963), they are ‘Gettier cases’. However, these cases are importantly different from the cases that were originally discussed by Gettier (1963). In each of the original Gettier cases, the believer might easily have had a belief, in a very similar way to the way in which the believer actually believes the particular proposition in question, even if the proposition that the believer would then have believed were false. As many contemporary epistemologists would say, the belief in question fails to be safe.2 The problem with cases of type (2) is different: it is not that the believer might too easily have believed something false; it is that the justification of the belief in question might too easily have been defeated. Thus, the closest parallel to the cases that Plato has in mind are cases like the ‘assassination case’ that was first presented by 2   For seminal discussions of the idea that knowledge involves safety, see Sainsbury (1997), Sosa (1999), and Williamson (2000).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  39 Gilbert Harman (1973, 143f.). In this case, there is a JTB that fails to count as knowledge, because there is a mass of (misleadingly) defeating evidence in the believer’s environment, and it is simply a fluke that the believer does not encounter this defeating ­evidence. But this defeating evidence—we may suppose—consists entirely of undercutting (rather than ‘rebutting’) defeaters. So, if the thinker had encountered this defeating evidence, she would simply have given up on having any beliefs about the topic in question: she would not have come to believe the proposition’s negation. In other words, this belief fails to count as knowledge, not because the believer could too easily have come to believe something false, but because the belief could too easily have been rationally defeated even if it were true. That is, on this approach, for you to know p, there must be no possible world that is sufficiently similar to the actual world, in which p is true, but your belief in p is rationally defeated or undermined. This condition on knowledge is broadly akin to the fourth of the four conditions that were imposed on knowledge by Robert Nozick (1981, 176–8)—the condition which has more recently come to be known as adherence.3 For a thinker’s belief in a true proposition p to ‘adhere’ to the truth is for the belief to be such that, in all the relevantly similar possible worlds in which the proposition p is still true, the thinker would still believe p. In effect, this idea of a belief ’s ‘adhering’ to the truth across the relevantly similar possible worlds seems equivalent to giving an explicitly modal gloss on Plato’s talk of the belief ’s being ‘permanent’. It is clear that a belief in a proposition p can be safe without adhering to the truth in this way. This will happen whenever there is no relevantly similar world in which the corresponding belief that the believer has is false, but there is a relevantly similar world in which although p is true, the believer encounters misleading defeating evidence, and so does not believe p in that world. What adherence requires, then, is that there must be no relevantly similar worlds in which the thinker encounters misleading defeating evidence in this way. Now, the term ‘relevantly similar’ as it appears in this statement of adherence is presumably a context-sensitive term: in some contexts, the term might have an inclusive extension, so that many worlds count as ‘relevantly similar’; in other contexts, the term might have a more restricted extension, so that many fewer cases count as ‘relevantly similar’. Depending on how inclusive the term’s extension is, the adherence condition will vary in strength. At one extreme, it might in effect be the condition that there is no possible world whatsoever in which the thinker encounters such misleading defeating evidence. At the other extreme, it might merely be the condition that the thinker does not in fact encounter such misleading defeating evidence in the actual world. In between these extremes, adherence requires that the thinker should encounter no such misleading defeating evidence in a range of possible worlds besides the actual world, but not that the thinker should encounter no such defeating evidence in any possible world whatsoever.   For a systematic discussion of adherence, see Bird (2003).

3

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

40  Ralph Wedgwood There is no sign in the texts that Plato is aware of this issue with the interpretation of a­ dherence. However, it may be that, at least implicitly, he interprets adherence as the ­strongest of these conditions—that is, as the condition that there is no possible world in which the thinker’s knowledge would be rationally undermined, or in other words, as the condition that knowledge is indefeasible. If all genuine knowledge is indefeasible in this way, then this would guarantee that all genuine knowledge must satisfy adherence to the highest degree. If Plato does assume that all genuine knowledge is indefeasible in this way, then this could explain one of the most striking claims that he does seem to make—the claim that all genuine knowledge is a priori. Most philosophers would assume that all of our empirical justification for beliefs about mind-independent reality is defeasible; there is always the possibility of new experiences that would defeat any such empirical justification. So, on this assumption, if any knowledge about mind-independent reality is indefeasible, that knowledge must be a priori. There is much textual evidence in favour of interpreting Plato as regarding all genuine knowledge as a priori. First, in Meno (81d) he declares that ‘seeking and learning are in general recollection’; and at 86b, he seems to treat ‘what we happen not to know at present’ as the same as ‘what we do not remember.’ There are many controversial issues surrounding the interpretation of the Theory of Recollection, but it is clear that when we ‘recollect’ something, we in some way ‘retrieve’ a truth that is already ­contained within the soul. In this sense, recollected knowledge is not derived from our sensory experience of the external world; it is knowledge that we possess purely by exploiting the resources that are already, antecedently to sensory experience, contained within the soul. In other words, it is a priori. Another passage where Plato insists on the a priori character of genuine knowledge is in Phaedo (65a–66e). For the attainment of wisdom (phronēsis), the bodily senses are said to be a positive ‘hindrance’ (65a). Sight and hearing and the other bodily senses ‘hold no truth’; and when the soul attempts to inquire into anything with the body, it is ‘utterly deceived’ (65b). Instead, it is ‘through reasoning that any of the things that are become manifest’ to the soul (65c). He seems to equate this kind of reasoning with thinking (dianoeisthai, 65e); and he says that the truth of these things is not seen through the body, but it is ‘whoever of us who has prepared himself for thinking most fully and precisely of each object of inquiry itself who would come closest to knowing (gnōnai) each thing’ (65e). Finally, the prominence of mathematics and philosophy in Plato’s conception of knowledge also encourages the conclusion that he thought that all genuine knowledge is a priori. Admittedly, there are some striking passages where he seems to allow for  knowledge of truths—such as the road to Larissa (Meno 97a)—that would be knowable only empirically if at all; we shall return to the issue of how to understand these ­passages in Section 5. But the general picture that seems to emerge is that the paradigmatic examples of knowledge are mathematics and philosophy—which are plausibly taken to be a priori.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  41 To sum up, according to Plato, knowledge is a factive cognitive state, which (a) adheres to the truth, and (b) is a priori—perhaps because it must adhere to the truth to the highest degree, or in other words, must be completely indefeasible.

3.  Necessity, Safety, and the Forms How could it be that all the truths that we ever come to know are already ­contained within the soul itself waiting to be recollected? This is never fully explained in Meno, but it seems to have something to do with the fact that the soul is imperceptible, immaterial, and eternal (existing for all of time—past, present, and future), and so in some way ‘akin’ to the Forms, as Plato argues at length in Phaedo (78b–80c). It also seems that the Forms collectively are the source of an intelligible system of necessary truths, which somehow articulates the nature of the Forms. So, presumably, one of the main ways in which the soul is ‘akin’ to the Forms is in somehow ­reflecting this system of necessary truths. This explains why all knowable necessary truths about the nature of the Forms are somehow built into the nature of the soul itself—even though, for most of the time, at least while we live an embodied human life on earth, these truths are inaccessible to consciousness (or as Plato suggests, have been forgotten). How exactly can we retrieve these truths that are built into the soul itself? Some ­philosophers might propose that we can sometimes retrieve such necessary truths though a kind of inference: perhaps one can sometimes ‘recollect’ a truth p by means of a process of inference whose conclusion is p. But this proposal would suffer from the following defect. The capacity for knowledge-yielding inferences of a certain sort is itself a mental state that has a certain kinship with the knowledge that those inferences are valid. So, this capacity would itself demand explanation—and the explanation of how we possess this capacity will itself have an affinity to the explanation of how we know the very necessary truths that are in question. So, any explanation based only on the appeal to such a capacity for inference is superficial, and fails to get to the heart of the matter. At all events, Plato’s account of how recollection works is not just based on such an appeal to inference. When Socrates interrogates the slave in Meno (82b–85b), what happens seems just to be that Socrates asks the slave a question, and then the correct answer to the question pops up in the slave’s head. By asking these questions, Socrates gets the slave consciously to focus on certain propositions. So, Plato’s picture seems to  be this: for at least some of these necessary truths, when we consciously focus on these truths, we immediately come to understand and know them. This seems to be  ­suggested by his remark (Republic 508d) that ‘when the soul focuses on something  illuminated by truth and what is, it understands, knows, and apparently ­possesses understanding’. In other words, given that these truths are already built into the soul, sometimes simply focusing consciously on these truths is sufficient for ‘recollecting’ them.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

42  Ralph Wedgwood As I understand it, this account of what it is to ‘recollect’ a truth does not appeal to any kind of ‘direct acquaintance’ with the truth in question. Instead, the account appeals to truths that are buried deep in unconscious propositional memory, and then awoken into conscious knowledge. A perceptual reading of Platonic recollection would be quite misguided. (Indeed, it may be that one of the chief functions of recollection is to explain our capacity for inferring necessary truths. As I shall tentatively suggest in the next section, Plato may think that whenever recollection provides us with any kind of knowledge, what we recollect is never a single truth all by itself, but rather an argument, which involves both certain premises and a conclusion that ­follows from them.) It is part of this account of how a priori knowledge is possible that it is knowledge of aspects of the Forms—that is, of necessary truths. There is no hint anywhere else in Plato’s middle-period writings of any other account of how a priori knowledge might be possible. So, it seems plausible that Plato held not only that all genuine knowledge is a priori, but also that it is all knowledge of necessary truths. This necessitarianism about knowable truths is inherited by Aristotle, who argues in the Posterior Analytics I.33 (88b30–89b9) that the knowable truths are always necessary truths, while the objects of ‘opinion’ (by which Aristotle means mere opinion, the kind of belief that falls short of being knowledge) are contingent propositions—or as Aristotle puts it, ‘what is true or false but can also be otherwise’ (89a3).4 I have already suggested that Plato may think that genuine knowledge must be ­indefeasible, and so satisfy the ‘adherence’ condition to the highest possible degree; at least, this would provide one possible explanation of why he thinks that all genuine knowledge is a priori. This raises the question of whether he also thinks that ­knowledge must satisfy safety to the highest degree. For a belief to satisfy safety to the highest degree would be for it to be impossible to have a belief in a false proposition on any basis that is similar to the basis of the belief in question. In that sense, for a belief to satisfy safety to the highest degree is for it to be based on an infallible basis for the belief in  question. (It is striking that at 477e, Plato says that knowledge is ‘infallible’— anamartēton—although there are admittedly several possible interpretations of that term in that passage.) As we have seen in considering the Theory of Recollection, whenever anyone has genuine knowledge of a truth, Plato would say that the basis for this knowledge is precisely that the truth in question was already built into the ­knower’s soul. So, for a belief that has a basis of this sort to be maximally safe, it would have to be impossible for any false proposition to be built into the soul in this way. If the only truths that can be built into the soul in this way are necessary truths that constitute aspects of the Forms, then it will be impossible for any false proposition to be built into the soul in this way. So, if Plato accepts the sort of necessitarianism about knowable truths that I have described, then he would in effect be committed to the conclusion that all genuine knowledge satisfies safety in the highest degree. 4   See the interpretation of Aristotle’s account of the distinction between knowledge and belief that is given by Fine (2010).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  43 The view that Plato accepts this sort of necessitarianism about knowable truths is also supported by the traditional interpretation of the Republic according to which it argues that all knowledge concerns the Forms. As Socrates says at 529b–c: I can’t conceive of any branch of learning that makes the soul look upward except that which concerns what is and what is invisible, and if anyone attempts to learn any of the sensible things, whether by gaping upward or squinting downward, I’d claim—since there’s no knowledge to be had of such things—that he never learns anything . . . .

In this passage, the reference to ‘what is and what is invisible’ is clearly a reference to the Forms, while ‘sensible things’—that is, perceptible concrete things—are explicitly said to be unknowable. As I shall argue later in this section, the same point—that the ‘objects’ of knowledge are the Forms—is also made in the famous passage at the end of Book V (476d–480a). It is not immediately clear what it might mean to say that the ‘objects’ of knowledge are Forms. Plato uses a variety of different expressions to capture this idea of the ‘object’ of a cognitive state. Often, he uses the preposition epi, apparently in the sense of towards or in reference to. For example, he says that ‘knowledge’ (gnōsis) is ‘towards what is’ (477a), that opinion is ‘positioned towards something else’ (477b), and that the person who has an opinion ‘bears the opinion towards something’ (478b)—although he also uses phrases without epi, as when he says, ‘We had to assign ­ignorance to what is not, and knowledge to what is’ (478c). It would be a mistake, I think, to identify the ‘object’ of these cognitive states, which Plato is trying to indicate with these phrases, with the content of the state that is indicated by the noun clause that is the grammatical direct object of the cognitive verb. Instead, I propose Plato is trying to offer an explanatory analysis of the cognitive state: he is suggesting that the state is ultimately or fundamentally explained by a kind of cognitive connection with or grasp of this ‘object’. However, just because the ‘object’ of these cognitive states is an entity like a Form or a perceptible concrete thing, it does not follow that these states do not also have ­propositions as their contents: the ‘object’ of the state may be distinct from its propositional content. In fact, Plato is perfectly happy to fill in his specifications of pieces of knowledge by noun clauses that indicate propositions, as when he says that ‘­knowledge (epistēmē) is by nature towards what is, to know how it is (gnōnai hōs esti)’ (477b).5 So, it seems plausible that Fine (1990) was right to argue that the kind of knowledge under discussion here is ordinary propositional knowledge. Thus, on the interpretation that I am proposing, according to Plato all the propositional knowledge that one has is fundamentally explained by one’s having a kind of cognitive connection with or grasp of the Forms. Since it also seems that both the 5   Admittedly, the exact interpretation of this sentence is disputed. But on any plausible interpretation, the clause that I have translated ‘how it is’ (hōs esti) indicates that it is part of the knowledge in ­question that it in some way involves predicating or describing its object—that is, in effect, it involves a proposition.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

44  Ralph Wedgwood existence and the intrinsic character of the Forms are metaphysically necessary, if one knows a proposition solely because of having an appropriate kind of cognitive connection with the Forms, that proposition would surely have to be a necessary truth. Thus, if I am right to accept a version of the traditional interpretation of the end of Book V, according to which the ‘objects’ of knowledge are the Forms, and the ‘objects’ of opinion are perceptible concrete things, this passage supports the conclusion that all genuine knowledge for Plato is knowledge of necessary truths. By contrast, on my interpretation, according to Plato all mere belief or opinion is explained by some kind of cognitive connection with or grasp of concrete perceptible things. The idea here may be that all opinion is ultimately explained by some kind of  sensory experience, which Plato may view as essentially consisting in a kind of ­cognitive connection with concrete perceptible things. It is clear why all the beliefs that  we would regard as empirically justified would in his view be explained by ­sensory  experience. But he may also have thought that even those beliefs that we would regard as irrational or unjustified are best explained by the thinkers somehow misreading or being misled by their sensory experiences. (Perhaps even beliefs in necessary falsehoods—like Meno’s slave’s initial belief that the square whose sides are double in length is also double in area (82e)—are also in Plato’s view explained by ­sensory experience.) What reason is there for accepting this interpretation of the end of Book V? Clearly, one of the passage’s main goals is to explain the difference between three cognitive conditions—knowledge, ignorance, and opinion—in terms of a corresponding difference in their ‘objects’.6 It seems clear that these three cognitive conditions are understood to be disjoint: no particular token state can be an instance of more than one of these three types. Ignorance is clearly disjoint from both knowledge and opinion: to be ignorant of a certain domain is to be cognitively connected to nothing in that domain (478b–c)—this is the sense in which the ‘object’ of ignorance is ‘what is not’. In  this way, being ignorant of a domain excludes both having opinion and having knowledge within that domain. Since the term ‘opinion’ in this passage seems to refer to ‘mere opinion’, no token state can be both a state of knowledge and a state of ‘opinion’— that is, a cognitive state that falls short of knowledge. It follows that these three kinds of cognitive condition are disjoint. Since Plato seems to be aiming to explain the differences between these three types of states in terms of corresponding differences between their objects, the sets of the objects of these three types of state would also have to be equally disjoint. The set of the objects of ignorance would have to be the empty set, since as we have seen, being ­ignorant about a domain involves being cognitively connected to nothing within that domain; and every pair consisting of the empty set and any other set is disjoint. Since 6   In this way, Plato’s approach is parallel to the way in which Aristotle in De Anima II.6 (418a6–25) explains the differences between the different senses by identifying the different ‘special objects’ of the senses.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  45 the Forms and the perceptible concrete things are also disjoint non-overlapping sets, one way to interpret the passage as identifying three disjoint sets for each of these three cognitive conditions would read it as identifying the empty set for ignorance, the set of Forms for knowledge, and the set of perceptible concrete things for opinion. Ultimately, however, there is a simpler and more powerful reason for accepting a version of the traditional interpretation of this passage. As we have seen, Plato claims that knowledge has as its object, or is ‘towards’ ‘what is’ (477a); in a parallel way, he concludes, opinion has as its object, or is ‘towards’, what is ‘intermediate between what purely is and what in every way is not’ (478e). The text (479a–c) also seems to make it clear that Plato identifies (a) what is with the Forms (for example, as Plato says, ‘the Beautiful itself ’), which are delighted in and loved by the philosophers, and (b) what is intermediate between what is and what is not with the concrete perceptible things (as he says, ‘the many beautiful things’), which are delighted in by the ‘lover of sights and sounds’. The conclusion that Plato draws at 480a, apparently making the very same use of the preposition epi as earlier at 477a–b is that the ‘objects’ of knowledge are the Forms (like ‘the Beautiful itself ’), while the ‘objects’ of opinion are the perceptible concrete things (like ‘the many beautiful things’). So shall we say that the latter (i.e. the philosophers) delight in and love the things that k­ nowledge is towards, while the former (i.e. the lovers of sights and sounds) delight in and love the things that opinion is towards?

It is admittedly far from clear exactly what Plato means by saying that perceptible things are ‘intermediate between what is and what is not’ (478d–e). It may have to do with the fact that the existence and character of perceptible things are contingent, and so every such thing exists and has a certain intrinsic character in one possible world, and either does not exist or does not have that intrinsic character in another world. But at all events, the important point for our purposes is that in this passage he endorses the view that the ‘objects’ of knowledge are the Forms. As I have explained, what this seems to mean is that all genuine knowledge is ultimately explained by our cognitive connection with or grasp of the Forms, and so must be knowledge of metaphysically necessary truths. In this way, then, it seems to me that a version of the traditional ‘Two-Worlds’ (TW) interpretation of Plato’s theory of knowledge is correct. Several objections have been raised against this TW interpretation—and especially by Fine (2003). I shall consider these objections to the TW interpretation in Section 5.

4.  Explanation and the Levels of Knowledge So far, it seems as if Plato is defending an extreme conception of knowledge, according to which all knowledge must consist in a priori knowledge of necessary truths—perhaps because all knowledge must be both indefeasible and infallible. So how can it also be that—as I have claimed—Plato is also a kind of contextualist about knowledge?

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

46  Ralph Wedgwood The answer is that Plato also thinks that knowledge must meet a further condition. This condition is in a way analogous to the ‘justification’ condition that later ­epistemologists imposed on knowledge. It is the condition that knowledge requires a grasp of the explanation of the truth that is known.7 The point that knowledge must satisfy this condition is made at several points. In Meno (98a), it is said that knowledge is ‘tied down by reasoning out the explanation’ (aitias logismōi); in Phaedo (76b), it is said that ‘a man who has knowledge would be able to give an account of what he knows’; in the Republic (534b), it is said that the person who cannot give an account ‘to that extent’ (tosouton) lacks understanding (nous) concerning the thing in question; and in Theaetetus (201c–d), the suggestion that knowledge is ‘true belief accompanied by an account’ seems to be treated as the most promising of the accounts of knowledge under consideration. Some contemporary contextualists like Stewart Cohen (1999) hold that justification comes in degrees, so that it is only the context in which the term ‘know’ is used that determines how much justification a person needs for a proposition for it to be true to say in that context that the person ‘knows’ the proposition. In an analogous way, as I shall argue, Plato thinks that there are levels of knowledge, where each of these levels corresponds to the degree to which the person grasps the explanation of the truth that is known; and in different contexts words like ‘know’ can be used more or less strictly. When the term is used less strictly, it refers to all levels that involve any degree of grasping the explanation of the truth that is known; but when the term is used more strictly, it refers only to the highest degree of grasping this explanation. Plato’s account of the different levels of knowledge is made most explicit in his ­discussion of the Divided Line (Republic 509d–511e). Strictly speaking, this analogy is only part of an extended discussion, the overall goal of which is to convey the central role of the Form of the Good both in knowledge and in reality. Unfortunately, I will not be able to explore the role of the Form of the Good in any detail; I shall simply focus on what this passage tells us about the nature of knowledge. As we have seen, the four segments of the line correspond to two different kinds of  opinion—imagination (L1) and confidence (L2)—and two different kinds of ­knowledge—thought (L3) and intellection (L4). In interpreting the distinction that is drawn here between these two levels of knowledge (L3 and L4), it seems best to assume that both levels of knowledge satisfy all the conditions of knowledge that I have identified so far: that is, both levels of knowledge involve a priori knowledge of necessary truths. Similarly, both levels of opinion (L1 and L2) involve beliefs, which are fundamentally explained by sensory experience—which Plato seems to regard as consisting in a kind of cognitive connection with concrete perceptible things. Plato’s account of the analogy at 509d implies that the four segments of the Divided Line have certain proportions to each other. Let the lengths of the four line-segments be, respectively, l1, l2, l3, and l4. Plato tells us explicitly that the following three ratios are   For a revealing exploration of this point, see Burnyeat (1980).

7

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  47 all the same: l1/ l2 = l3/ l4 = (l1 + l2)/(l3 + l4).8 It follows by elementary algebra that l2 = l3. It is in my judgment preposterously implausible to suggest that Plato might somehow have failed to notice this simple mathematical fact. Moreover, even if—quite ­unaccountably—Plato himself failed to notice this fact, there were plenty of brilliant mathematicians at the Academy who would have pointed it out to him. So, it seems to  be an important feature of this image that the length of the line-segment that ­represents confidence (L2) is equal to the length of the line-segment that represents thought (L3). Most scholars hold that it is also part of Plato’s image that the divisions of the line are unequal. In other words, the ratio in accordance with which the line is divided (l1 + l2)/ (l3 + l4) ≠ 1, and in consequence, l1 ≠ l2, l3 ≠ l4 and (l1 + l2) ≠ (l3 + l4). It is not totally clear what these unequal divisions are meant to convey. But one reasonable speculation is that it is meant to indicate that opinion (that is, L1 and L2 together)—in the sense of mere opinion, the kind of belief that falls short of knowledge—is inferior to knowledge (L3 and L4 together), and that in a similar way, imagination (L1) is inferior to confidence (L2), and thought (L3) is inferior to intellection (L4). Since as we have seen, the line-segment that represents confidence (L2) is equal in length to the line-segment that represents thought (L3), what seems to be suggested is that confidence (L2) is not inferior to thought (L3) in the way in which opinion in general is inferior to knowledge in general. In short, there is a ‘good kind of belief ’ and a ‘second-rate kind of ­knowledge,’ and neither of the two—at least as Plato now thinks of things in the Republic—is inferior to the other. The distinction between imagination and confidence is sketched extremely briefly in the discussion of the Divided Line. On the whole, however, the view of many commentators seems plausible: imagination (L1) is a state that involves uncritically taking sensory experiences at face value, without any deployment of an ability to discriminate between reliable and misleading appearances; confidence (L2), on the other hand, is a state that involves some deployment of a reliable discriminatory ability of this sort.9 In fact, it seems clear that the same distinction is explained again much later, in Book X (602c–603b), when Plato distinguishes between two kinds of belief that are involved when we see something that ‘looks crooked when seen in water and straight when seen out of it’ (602c). Cases of this kind, he argues, reveal that there are two kinds of belief—clearly implying that both kinds involve a part of the soul’s ‘believing’ something (doxazon). It seems plausible to me that Plato’s reason for implying that the superior kind of belief in these cases counts as belief, rather than as knowledge, is because its content is a contingent fact about concrete perceptible items, rather than a necessary truth about the Forms. 8   This interpretation is clearly supported by Plato’s use of the phrase ton auton logon at 509d5; it does not matter whether we follow the Oxford Classical Text in reading anisa at 509d4 or the various ancient and modern critics who have suggested amending the text to an’ isa. 9   This is in effect how Fine (1990) interprets the distinction.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

48  Ralph Wedgwood The inferior kind of belief is the one that uncritically takes sensory appearances at face value; this kind of belief is attributed to an ‘inferior’ part of the soul (603a). The other kind of belief somehow ‘places confidence in measurement and calculation’ (603a); because it ‘places confidence’ in such measurement and calculation it is attributed to the ‘best part of the soul’. It is striking that Plato uses the term pisteuon for ‘places confidence’—a term that is obviously cognate with the term that he uses for the L2 kind of belief, pistis. This makes it plausible that this superior kind of belief is what was earlier called confidence (pistis): that is, it is a belief about contingent concrete things that involves a reliable ability to discriminate veridical from misleading appearances. Presumably, exercising this ability will often involve a sort of ‘measurement’ or ‘calculation’. Such measurement or calculation presumably involves some application of mathematical reasoning to concrete perceptible things; this sort of reasoning would presumably only be possible for those who possess at least some degree of knowledge (if only the lower L3-kind of knowledge, thought). The distinction between the two levels of knowledge is explained in somewhat greater detail. The lower kind of knowledge is illustrated in the Republic (510c–d) by the examples of the kinds of mathematics that had been developed in Plato’s time. We know from Meno that this sort of mathematics (even the very rudimentary grasp of mathematics that Meno’s slave achieves in the course of his discussion with Socrates) involves at least the beginnings of recollection: the soul consciously focuses on certain necessary truths, and because these necessary truths about the Forms are already built into the soul, the soul comes to have a priori knowledge of those truths. What Plato says here is that L3-level knowledge (thought) has two limitations compared to L4-level knowledge (intellection). First, even though it is concerned with abstract m ­ athematical objects, it still makes use of concrete visible objects as images (this point is made three times, at 510b, 510d and 511a). Second, it cannot avoid relying on ‘hypotheses’ instead of on the ultimate ‘unhypothetical’ explanatory first principle (510b–c and 511a–b). The first point might seem puzzling: if L3-level mathematical knowledge really is a priori knowledge of necessary truths, why does it still use perceptible things as ‘images’? It seems that Plato must be assuming that while we are only at this level of knowledge, we still need images of this sort—such as figures and diagrams and the like—to focus clearly enough on the relevant truths in order to recollect them. (It seems clear that in his discussion with Meno’s slave, Socrates draws diagrams in order to help the slave to focus on the relevant questions.) By contrast, if we ever acquired complete L4-level knowledge of the whole intelligible system of truths, then we would somehow be able to focus on each of these truths without any such reliance on images. It is because of this feature of L3-level knowledge that it still has some kinship with opinion: although it is a state that is fundamentally explained by the soul’s cognitive connection to the Forms, it still has a kind of dependence on the soul’s connection to perceptible concrete particulars.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  49 The second difference between L3-level and L4-level knowledge concerns the degree to which the thinker grasps the explanatory system of necessary truths. L3-level knowledge involves relying on ‘hypotheses’ and deriving ‘conclusions’ from such hypotheses, but failing to provide any ‘further’ explanation of these hypotheses. Presumably, these ‘hypotheses’ and ‘conclusions’ are themselves necessary truths that belong to the whole explanatory system of necessary truths that flow from the nature of the Forms. So, whenever we have any level of a priori knowledge of these necessary truths, we must grasp at least a small argument in which some of these truths are ‘hypotheses’ and others are ‘conclusions’ that logically follow from them. If you grasp an argument in which a true conclusion follows from some true ­hypotheses, you can be said to have at least some grasp of an explanation of this ­conclusion. But if you cannot derive these hypotheses from any ‘further’ principles, you have no ‘further’ explanation of these hypotheses. This raises a problem about how you can be said to ‘know’ these hypotheses, given that (as we have seen) Plato insists that to know a truth, one must have some grasp of its explanation. The solution to this problem proposed by Fine (1990) seems plausible to me: there is still a sense in which I grasp a kind of explanation of these hypotheses, since these hypotheses are  in a way explained by their explanatory power in relation to the truths that follow from them. So L3-level knowledge involves grasping at least a small argument involving these necessary truths, and so having a limited though genuine insight into the explanation of the truths in question. L4-level knowledge, by contrast, involves a complete grasp of the whole explanatory system of necessary truths, including the ultimate ‘unhypothetical first principle of everything’, which Plato seems to identify with the Form of the Good.10 To achieve L4-level knowledge, a few one-off bits of recollection would not be sufficient. One would have to use what Plato calls ‘dialectic’ (511b) in order to fit all of these recollected truths together in the right way, and thereby to achieve a grasp of the whole explanatory system of necessary truths. It seems that Plato thinks that anyone who has any grasp of mathematics at all has at  least L3-level knowledge, while L4-level knowledge has perhaps never yet been achieved by any actual human being. This reveals that it is in a way an oversimplification on Plato’s part to write as though there were exactly two levels of knowledge. If one’s grasp of this explanatory system of necessary truths comes in degrees, there is a huge ­indeterminate number of levels of knowledge, extending all the way from the most rudimentary grasp of tiny fragments of the explanatory system of truths through intermediate levels to a comprehensive grasp of the whole system. Indeed, there is some—admittedly somewhat equivocal—evidence that Plato thinks that virtually everyone possesses at least certain very elementary kinds of L3-level knowledge. When the argument for the Theory of Recollection is presented in Phaedo   For an illuminating discussion of Plato’s conception of the ‘unhypothetical’, see Bailey (2006).

10

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

50  Ralph Wedgwood (72e–77a), it starts out from the assumption, which is enthusiastically accepted by Simmias, that ‘we know what the Equal itself is’ (74b)—where the phrase ‘the Equal itself ’ clearly refers to the Form of the Equal. However, Simmias explicitly denies ­having the kind of knowledge of the Equal itself that would involve being able to give an account of it ‘properly’ (axiōs, 76b); we may take this as the claim that Simmias does not have L4-level knowledge of the Form of the Equal.11 Nonetheless, with respect to the L3-level knowledge of the Equal itself that Simmias does have, Socrates and Simmias seem to agree that he got that knowledge from ­perceiving concrete perceptible objects like ‘equal sticks and stones’ (74b), and that we have been having the relevant kind of perceptions ‘from the very moment we were born’ (75b). The implication seems to be that Simmias started to acquire the kind of knowledge of what the Equal itself is that he actually has not long after he started having perceptions of this kind. This kind of knowledge may be plausibly identified with a possession of the concept of equality. So, it may be that for Plato, certain particularly elementary pieces of L3-level knowledge are necessary for even possessing the basic mathematical concepts that we have; and since the other examples of Forms that he cites in Phaedo and the Republic are the Forms of ethical qualities like the Good and the Just and the like, Plato may also think that certain rudimentary pieces of L3-level knowledge are also necessary for possessing the fundamental ethical concepts too.12 At all events, we know that even Meno’s untutored slave starts to recollect certain necessary mathematical truths; in so doing, according to the theory that Plato develops in Phaedo and the Republic, the slave recollects certain aspects of the Forms—certain small ­fragments of the complete intelligible system of necessary truths. According to the interpretation of this theory that I am proposing here, in recollecting these truths, the slave acquires some rudimentary pieces of L3-level a priori knowledge of necessary truths. But the kind of grasp of mathematical truth that the slave achieves in this ­passage is, of course, as Plato presumably knew, fairly widespread. So, on my interpretation, at least some basic pieces of L3-level knowledge are widespread. Some readers may think that this implication of my interpretation—that according to Plato, some rudimentary forms of a priori knowledge of necessary truths are   For a seminal discussion of this part of Phaedo, see Ackrill (1997, essay 1).   This interpretation of this passage in Phaedo is emphatically denounced by Dominic Scott (1999), who claims that it involves Plato in ‘appalling difficulties’ (114); according to Scott, ‘recollection only starts with the process of philosophizing, and thus only a rather limited number of people recollect’ (96). However, Scott’s arguments are not persuasive: his suggestion that only philosophers recollect seems incompatible with the indisputable point that according to Plato, Meno’s untutored slave begins to recollect certain truths of mathematics before Socrates’ and Meno’s eyes; the kind of grasp of elementary mathematics that the slave achieves is certainly not confined to philosophers. Scott regards it as an ‘absurdity’ (106) to read Plato as claiming that any ordinary thinkers are aware of the way in which ordinary concrete perceptible objects ‘fall short’ of the Forms. But in fact, it is not implausible to suggest that, at least at the time when he wrote Phaedo, Plato believed that even Meno’s slave is implicitly aware of the difference between a necessary truth (like the theorem that the square on the diagonal is double the area of the original square) and any contingent truth about the perceptible diagram before him; and this implicit awareness of this difference is, I suggest, precisely what Plato means by speaking of our awareness of how concrete perceptible objects ‘fall short’ of the Forms. 11 12

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  51 widely shared—is incompatible with the view, which Plato seems to endorse, that non-­philosophers are ‘in reality deprived of knowledge of each thing that is’ (Republic 484c). However, it is open to me to interpret this occurrence of the word “gnōsis” (‘knowledge’) as referring to something like L4-level knowledge. This is one example of a p ­ attern that I shall explore in more detail in the following section: once we understand the flexible way in which Plato uses his terminology, the apparent obstacles to my interpretation will disappear.

5.  Cognitive Terms in Context In this final section, I shall try to show how my interpretation of Plato’s theory and of his use of terminology can explain some puzzling features of the relevant texts, and also answer some of the key objections that may be raised against my interpretation. First, we can now see why in Phaedo, Simmias seems to contradict himself on the issue of whether people like him have knowledge of Forms like the Equal itself—with Simmias saying at 74b that ‘we know what the Equal itself is’, and clearly implying at 76b that he does not know what the Equal is. In the first passage, Simmias is referring to rudimentary L3-level knowledge, while in the second passage, he is referring to complete L4-level knowledge. So, there is no contradiction between these statements when they are correctly understood. As I have explained, my account of Plato’s theory of knowledge is a version of the traditional TW interpretation. I shall now address some objections to the TW ­interpretation. One such objection that might be raised is that Plato’s ontology is not in fact divided into just two worlds—the world of the Forms and the world of perceptible concrete things. There are other items in his ontology—most notably, souls, which as we have seen are ‘akin’ to the Forms in being imperceptible and eternal, but are clearly not themselves identical to Forms, like the Equal itself or the Beautiful itself. In fact, however, this point is quite consistent with my version of the TW interpretation. My argument for this interpretation only needs the assumption that the Forms and the concrete perceptible things constitute disjoint sets of items: the two sets do not need to be exhaustive as well as disjoint. Moreover, on my version of this interpretation, Plato’s claim that the Forms are the ‘objects’ of knowledge and concrete perceptible things the ‘objects’ of belief does not imply that beliefs can only be about concrete perceptible things. It implies that all our beliefs are ultimately explained by our sensory experiences, which consist in a cognitive connection to such concrete perceptible things. As I shall explain, this allows for both knowledge and beliefs about the soul. First, there are presumably some necessary truths about the soul as such, which can be known a priori (for example, it seems that according to Plato, we can know a priori that the soul is immortal). Such knowledge is explained by our cognitive connection with the Forms (presumably including the universal Form of the Soul as such). Second, our beliefs in any contingent truths (and in any false propositions) about souls must in Plato’s view be ultimately explained by sensory experiences—which

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

52  Ralph Wedgwood c­ onsist in a kind of cognitive connection with concrete perceptible things. So, Plato is committed to the view that all beliefs in contingent truths (and in false propositions) about souls are in this way empirical. Admittedly, this view is not obviously correct; but it certainly does not seem obviously wrong either. At one point, Fine (1990) enumerates five objections to the TW interpretation. I shall now defend my interpretation against these five objections. First, Fine (1999, 215) objects that according to the TW interpretation, ‘the objects of knowledge and belief are . . . disjoint; one cannot move from belief to knowledge about some single thing.’ On my interpretation, however, the ‘object’ of a belief must be distinguished from the belief ’s content. So, in principle, one could have an unjustified belief in a necessary truth—which would be ultimately explained by one’s sensory experiences, which ­consist in one’s cognitive connection with perceptible concrete things—and then move from this unjustified belief to knowing the truth. In addition, one can also move from L3-level knowledge to L4-level knowledge of a necessary truth, and from L1-level belief to L2-level belief in a contingent truth. As we have seen, in some texts, Plato uses the term doxa as a generic term for the genus of which all four levels are species. So, when we are using the term doxa in this way, it will be correct to describe someone who moves from L3 to L4 as moving from a kind of doxa to epistēmē. In Meno, the slave’s recollections of mathematical truths are described as ‘beliefs aroused in him in a dream-like way’ (85c). In the Republic, these dream-like beliefs—since they arise in the process of recollection, and so are explained by the slave’s cognitive connection with the Forms—would have to count as L3-level cognitive states, of an elementary and rudimentary kind. It is certainly possible in Plato’s view to move from L3-level beliefs of this kind to L4-level knowledge. Second, Fine (1999, 215f.) objects that the TW interpretation implies that Plato ‘radically rejects the Meno’s account of knowledge, according to which true beliefs become knowledge when they are adequately bound to an explanatory account. For the Meno, knowledge implies true belief; on TW, knowledge excludes true belief ’. However, it is only when the term “doxa” is used in the narrower way, to refer only to L1 and L2 taken together, that it excludes knowledge. When it is used in the broader way, it refers to the genus of which all four levels are species. When the term ‘belief ’ is used in the larger sense, knowledge—even according to the Republic—does not exclude true belief. Thus, my version of the TW interpretation is entirely compatible with Meno’s account of knowledge. The reader may suspect that the TW interpretation conflicts with Meno’s account of knowledge in a different way. The TW interpretation restricts knowledge—strictly so called—to a priori knowledge of necessary truths, explained by our cognitive connection to the Forms. But in Meno, one of the examples that Plato gives, seemingly as one of the centerpieces of his discussion of knowledge, is the case of someone’s knowing the road to Larissa (97a). But the truth about which road leads to Larissa is evidently a contingent truth (if the bridges and tunnels had been built in different locations, the

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  53 road in question might not have led to Larissa at all), and it is also a truth that can surely only be known empirically. In fact, however, if we analyze the dialectical context of this example, it is not clear that Plato does commit himself to the assumption that one can genuinely know the road to Larissa. At this point, Socrates is raising an objection to the premise that he invoked earlier that ‘only someone with knowledge (phronimos) can guide correctly’. Proponents of this premise will presumably accept that it is possible to ‘guide’ travelers to Larissa; and so, proponents of this premise are committed to the assumption that it is possible to know the road to Larissa. But Socrates rebuts this premise by pointing out that one can also guide correctly if one has mere correct belief (97b). So, Socrates need not accept the assumption that anyone can genuinely know the road to Larissa: for Socrates, it is an entirely live possibility that the only condition that ever enables anyone to guide travelers to Larissa is mere correct belief, and not knowledge. The most that Socrates need accept—purely for the sake of argument—is that if anyone (perhaps per impossibile) knew the road to Larissa, then they would be able to guide travelers successfully. So, when read carefully, the text of Meno does not commit Plato to recognizing the possibility of anyone’s knowing the road to Larissa.13 Third, Fine (1999, 216) objects that the TW interpretation implies that ‘Plato is quite skeptical about the limits of knowledge; . . . no one can know items in the sensible world’. Fundamentally, this point is correct: according to my version of the TW interpretation, no one can know any contingent truths at all—including contingent truths about the sensible world. However, the difference between L1 (imagination) and L2 (confidence)—that is, between the inferior and superior levels of belief—is of great importance to Plato. An L2-level belief flows from the deployment of a reliable ability to distinguish between misleading and veridical appearances; if such a belief is also true, it is at least fairly close, if not identical, to what contemporary reliabilists like Alvin Goldman (1986) would categorize as full-blown knowledge. It is not surprising, then, that algebraically the ratios of the segments of the Divided Line suggest that L2-level belief is in no way inferior to L3-level knowledge. So, Plato is not denying that we can have reliably true beliefs about the sensible world. This hardly makes him into a radical skeptic about our cognitive relations with the sensible world. Fourth, Fine (1999, 216) objects that ‘this sceptical result would be quite surprising in the context of the Republic, which aims to persuade us that philosophers should rule, since only they have knowledge, and knowledge is necessary for good ruling. If their knowledge is only of Forms—if like the rest of us, they only have belief about the sensible world—it is unclear why they are specially equipped to rule in this world’. 13   At one point in Theaetetus (201b6), Socrates seems to presuppose—at least for the sake of argument— that an eyewitness can know (eidenai) the truth ‘about what happened to people who have been robbed of their money or have suffered other acts of violence’. In this way, it seems to me that in Theaetetus, Plato is exploring the possibility of broader conception of knowledge than the one that he developed in the Republic and Phaedo.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

54  Ralph Wedgwood However, it seems that mathematical knowledge can underwrite an ability for ‘­calculation and measurement’, and one of the main ways in which someone can acquire an ability for reliably discriminating between misleading and veridical ­appearances is by having such an ability for calculation and measurement. In other words, having L4-level mathematical knowledge will help the philosopher-kings to have good L2-level beliefs rather than bad L1-level beliefs. So, this is why those who have k­ nowledge of necessary truths are better equipped to rule in the world of contingent sensible things. Still, it is unfortunately true, in Plato’s view, that beliefs about ­contingent concrete things can never be infallible in the same way as a priori k­ nowledge of necessities. But Plato explicitly accepts this. This is why in the end the kallipolis will decline, when the philosopher-kings (in spite of their knowledge of all the necessary truths) form some mistaken beliefs about contingent perceptible matters (546b–c). So, Plato’s restriction of genuine knowledge to a priori knowledge of necessary truths is not s­ urprising in the context of the Republic. Finally, Fine (1999, 216) objects that ‘the text of the Republic seems to contradict TW. At 506c, Plato says that he has beliefs about, but no knowledge of, the Form of the good; and at 520c he says that the philosopher who returns to the cave will know the things there, i.e. sensibles’. The first point is easy for my account to handle. When Socrates says at 506c that he has doxa without epistēmē, he is using the term doxa in its broad sense, to refer to the genus of which all four levels of cognitive states are ­species, and he is using epistēmē in its narrow sense, to refer to L4 (intellection) alone. The point of this passage, then, is that Socrates has some kind of doxa (perhaps an elementary form of L3-level awareness) of the Form of the Good, but that this falls far short of the kind of L4-level knowledge that the philosopher-king would have. The second point is less easy for my version of the TW interpretation to handle. But Fine (2008) has herself emphasized that in the Apology Plato uses some of the Greek cognitive verbs for knowledge (such as eidenai and gignōskein and their cognates) in an idiomatic way, to refer to what by the lights of Plato’s official theory is only a reliably true belief—that is, a L2-level belief (confidence or pistis). It seems plausible to me that something similar is happening with Plato’s use of these terms in the Republic as well. If L2-level beliefs are not in fact inferior in value to L3-level knowledge, it is not too misleading for Plato sometimes to use these terms in this idiomatic way. With the term epistēmē and its cognates, there are a couple of passages where Plato seems to use these terms in other idiomatic senses. First, these terms sometimes refer to expert know-how—that is, to an expert’s knowledge of how to do something. We find this usage in a couple of places, especially in Republic IV: first, in an ironic passage at 420e (‘We know how to order the farmers to be clothed in purple and gold’); then at 422c (‘the rich have more knowledge and experience of boxing than of warfare’); and, finally, in the passage where Socrates defines what it is for the city to be wise (428c–e), which moves in a way that is reminiscent of the early Socratic dialogues, from speaking of knowledge of the crafts of building, metal-working, and farming, to speaking of the knowledge that enables the guardians to deliberate about what is good and bad for the

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

plato’s theory of knowledge  55 city as a whole. Second, a more specialized idiomatic use of the verb epistasthai refers to knowing a text or a story by heart, as when Socrates says that he knows Aesop’s fables by heart (Phaedo 61b), or asks Glaucon whether he knows the beginning of the Iliad by heart (Republic 392e). Otherwise, however, in the middle-period dialogues, the term epistēmē and its cognates seem to refer principally to some level or other (according to context) of a priori knowledge of necessary truths.14 In these ways, I propose, we can answer Fine’s objections to the TW interpretation. The TW interpretation is much less outré than generally supposed. Plato’s goal in classifying cognitive states is to ‘carve at the joints’ (as he puts it in Phaedrus 265e). In his view, the distinction between a priori knowledge of necessary truths and our somewhat reliably true beliefs about contingent matters marks a crucial ‘joint’—that is, a crucial dissimilarity between two different kinds of cognitive states. He is willing to allow that in ordinary language we use terms like as eidenai and gignōskein to refer to the latter as well as the former; and in some contexts, he is willing to slip into that form of speech too. But what is of fundamental importance to him is not what ­terminology we use, but the essential nature of and differences between these different kinds of ­cognitive states. My fundamental conclusion, however, is that Fine is right on the most important point of all. Plato is not concerned with utterly different questions from contemporary epistemologists, nor is he defending a fantastic pre modern theory that nowadays no one could take seriously. He is brilliant and insightful epistemologist, with whom we can still have a productive dialogue today.15

References Ackrill, J. L. (1997). Essays on Plato and Aristotle (Oxford: Clarendon Press). Bailey, D. T. J. (2006). ‘Plato and Aristotle on the Unhypothetical’, Oxford Studies in Ancient Philosophy 30: 101–26. Bird, Alexander (2003). ‘Nozick’s fourth condition’, Facta Philosophica 4: 141–51. Burnyeat, Myles (1980). ‘Socrates and the Jury: Paradoxes in Plato’s Distinction between Knowledge and True Belief ’, Aristotelian Society Supplementary Volume 54: 173–206. Cohen, Stewart (1999). ‘Contextualism, skepticism, and the structure of reasons’. Philosophical Perspectives 13: 57–89. DeRose, Keith (2009). The Case for Contextualism: Knowledge, Skepticism, and Context (Oxford: Oxford University Press). 14   This is especially clear of the occurrences of epistēmē and its cognates in Phaedo 73c, 74c, 75e, and 76d. I would read the occurrences of these terms in Meno 85c and Republic 426d and 522c–d as referring to a priori knowledge of necessary mathematical truths, and the occurrences in Meno 93b, Republic 409b, 505a, 536a, and 579e as referring to a priori knowledge of necessary ethical truths. (Nothing about Plato’s own views of knowledge can be inferred from Phaedo 96b or Republic 488d or 598c, since in these passages these terms appear in direct or indirect speech, expressing views that Plato rejects.) 15   An earlier version of this essay was presented at a conference on Epistemology in Bled, Slovenia, in June 2015. I am grateful to the members of that audience, and also to Gail Fine and Christopher Shields, for helpful comments on that earlier draft.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

56  Ralph Wedgwood Fine, Gail (1990). ‘Knowledge and Belief in Republic 5–7’. In Stephen Everson, ed., Cambridge Companions to Ancient Thought (Cambridge: Cambridge University Press): 85–115. Fine, Gail, ed. (1999). Plato 1: Metaphysics and Epistemology, Oxford Readings in Philosophy (Oxford: Oxford University Press). Fine, Gail (2003). Plato on Knowledge and Forms: Selected Essays (Oxford: Oxford University Press). Fine, Gail (2004). ‘Knowledge and True Belief in the Meno’. Oxford Studies in Ancient Philosophy 27: 41–81. Fine, Gail (2008). ‘Does Socrates Claim to Know that He Knows Nothing?’ Oxford Studies in Ancient Philosophy 35: 49–85. Fine, Gail (2010). ‘Aristotle’s Two Worlds: Posterior Analytics 1.33’. Proceedings of the Aristotelian Society 110 (3): 323–46. Gettier, Edmund (1963). ‘Is justified true belief knowledge?’ Analysis 23: 121–3. Goldman, Alvin (1986). Epistemology and Cognition (Cambridge, MA: Harvard University Press). Harman, Gilbert (1973). Thought (Princeton, NJ: Princeton University Press). Nozick, Robert (1981). Philosophical Explanations (Oxford: Clarendon Press). Sainsbury, R. M. (1997). ‘Easy Possibilities’. Philosophy and Phenomenological Research 57: 907–19. Sosa, Ernest (1999). ‘How to Defeat Opposition to Moore’. Philosophical Perspectives 13: 141–53. Williams, Bernard (1978). Descartes: The Project of Pure Enquiry (Harmondsworth: Penguin). Williamson, Timothy (2000). Knowledge and its Limits (Oxford: Clarendon Press).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

3 Justice and Persuasion in the Republic Dominic Scott

In Plato’s Apology, Socrates is prepared to discuss moral philosophy with anyone, ‘young or old, citizen or foreigner’ (30a2–4). By contrast, Aristotle in the Nicomachean Ethics seems to confine his audience to those who have been brought up with the right habits and intuitions (I 4, 1095b3–8; cf. X 9, 1179b9 and 30); he also excludes the ­inexperienced and the young (I 3, 1095a2–6). And ‘the many’ follow passion, not reason; they therefore require punishment rather than argument (X 9, 1179b10–16). But where does Plato, as author of the Republic, stand? To judge from evidence internal to the work, whom did he think most likely to benefit from engaging in moral argument, especially about the value of justice? As broad a group as Socrates had in mind in the Apology, or a more restricted class? This is the main topic of this essay. But before turning to it directly, I need to draw attention to the fact that the Republic makes an explicit distinction between two ways of investigating its central topics: the ‘shorter’ and ‘longer’ routes (cf. IV 435c–d and VI 504a–e). I do this in Section 1. The shorter route is the one actually taken in books II–IV, Socrates’ first attempt at answering the challenge of Glaucon and Adeimantus to articulate the value of justice. In it Socrates gives an analysis of the virtues, including justice, confining himself to political theory and moral psychology. The longer route, first outlined in books VI–VII, would involve an investigation into the same topics, viz., the definition of justice and the other virtues, but it would do so by means of a long excursion into metaphysics, involving an understanding of the Forms, particularly the Form of the Good, which would act as the ultimate first principle of the inquiry. But, in addition to the longer and shorter routes, which are explicitly distinguished in the work, I shall also argue that there is the possibility of a third, the ‘middle route’, in which one argues for the value of justice by appealing to the metaphysics of the ­central books, but without pursuing the rigour demanded of the philosopher-rulers in their lengthy education. Once I have discussed the three routes in Section 1, and the extent to which Socrates actually follows them within the work, it will be clear that my original question about

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

58  Dominic Scott the suitable audience for moral arguments breaks into three: whom did Plato think able to benefit from following the longer, middle, and shorter routes, respectively? Answering these questions will be the task of Section 2, the main part of this essay.1

1.  Three Ways of Arguing for the Value of Justice In Rep. II–IV, Socrates proposes to analyse justice in the individual by first looking at it in the state. This leads to the long description of the ideal city with its three classes, leading to an account of its four cardinal virtues—wisdom, courage, temperance, and justice (427d–434c). Socrates is now ready to return to the individual, and this is the  first place where he warns his interlocutors that he has been taking a shorter route (435c9–d8): ‘You should be aware, Glaucon, that in my view we shall never come to grasp this precisely from the methods that we’re currently using in our arguments. For there is another longer and more extensive route that leads to it. But perhaps we might grasp it in a way that’s up to the standard of our earlier statements and inquiries.’ ‘Can’t we be content with that?’ he said. ‘For the time being, it would enough for me.’ ‘Well in that case’, I said, ‘I’ll be entirely satisfied with it too.’

At this stage he does not say what the longer route would consist in, just that it is ­possible and desirable. But for now, they will make do with the shorter. Accordingly, he first argues that, like the state, the soul is tripartite (436–441c) and then gives a parallel account of its cardinal virtues. Justice turns out to involve the three parts of the soul interacting harmoniously, something analogous to bodily health. As such, it is readily agreed to be valuable to the agent. This brings us to the end of book IV, where Glaucon sounds quite satisfied with the response to his initial challenge. Socrates, however, wants to confirm their conclusion by looking more closely at the various kinds of vice or injustice in state and soul. But he is interrupted by the other interlocutors, who want to find out more about the ideal city, in particular about the life and living arrangements of the guardians. In the event Socrates lays out three proposals: the inclusion of women among the rulers (and auxiliaries), the abolition of the family, and the need for philosopher-rulers. Books V–VI consist in an attempt to defend each of these proposals. His defence of philosopher-rulers is complete by 502a, at which point he turns to consider their education.

1   As a whole, this paper is a reconsideration of a thesis originally proposed in Scott (1999). Section 1 draws on material from Scott (2015), esp. chs 3–4, about the different levels of argument in the Republic. Versions of the essay have been presented to audiences at Trinity College Dublin, the University of California at Berkeley, Cornell, Durham, Oxford, and São Paulo universities. I am grateful to all those who provided feedback. In the final stages of writing, I received extremely helpful comments from Christopher Shields. In the long time I have been thinking about these issues, I have benefited enormously from conversations with Gail Fine and Terry Irwin, and of course, from reading their work on the Republic.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  59 In doing so, he stresses the extraordinary intellectual demands to be placed on the rulers. This leads him to make a back-reference to the argument followed in books II–IV, recalling the distinction between the shorter and longer routes at IV 435c9–d8, and effectively disparaging the methods they followed in comparison with those to be pursued by the philosopher-rulers. Here is the text in question: ‘You recall’, I said, ‘that once we had distinguished three kinds in the soul we drew conclusions about the nature of justice, temperance, courage and wisdom.’ ‘If I didn’t recall, I wouldn’t deserve to hear the rest’, he said. ‘And do you also recall what came before?’ ‘What was that?’ ‘We were saying that, in order to view these things in the best way possible, there would be another, longer way round, which would make them clear to anyone who took it. Nonetheless, it was possible to give arguments on a level with what we’d said before. You claimed that was enough for you. And so what we then said, it seemed to me, lacked precision, but whether you found it acceptable is for you to say.’  (504a4–b7)

However, Socrates now insists that the guardians of the ideal state will have to follow the longer route, even if he was more lenient on himself and his interlocutors. This path involves proceeding to the Form of the Good, a process that will take them from the age of 20 to 50. They start with ten years of mathematics (between the ages of 20 and 30), which will help them make the transition from the sensible world of becoming to the intelligible world of being. Only at 30 will they start philosophical dialectic, the study of Forms. After five years, they engage in administrative and military tasks until the age of 50, when they are finally ready to apprehend the Form of the Good, the ultimate first principle of everything. This completes their education, and they are now ready to rule. The point I wish to highlight here is that the understanding gained in this ­extraordinarily long and rigorous process of education will enable them to achieve far more exact answers to the questions about the virtues discussed in books II–IV (as well as to make practical decisions in the running of the state). By contrast, the arguments followed by Socrates’ interlocutors in II–IV operated at the lower level, confined to political analysis and moral psychology. There was no reference to the Forms, or more generally to the metaphysics and epistemology to be unveiled in the  central books. The two levels of argument are quite distinct. Since there are two different ways of arguing for the value of justice, the longer and the shorter routes, a further question arises. Socrates follows the shorter route up to the end of book IV, which provisionally concludes that justice is a benefit to those who possess it. But as soon as he reaches this conclusion, he signals his desire to continue the defence of justice further (IV 445b5–7). So, in what remains of the work, does he do so by staying on the shorter route, or does he start to follow the longer one (once he has announced its existence to his interlocutors in book VI)? In a sense this question is easy enough to answer. Neither Socrates nor his interlocutors are in any position to take the longer route. To do so requires that one has attained knowledge of the Form of the Good, which can then be used as a first principle to derive knowledge of the other Forms, including justice, and to explain their value.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

60  Dominic Scott But Socrates disclaims knowledge of the good at 505a5–6 (shortly after the passage previously quoted, 504a4–b7), and declines to speak to his interlocutors as if they had knowledge of it either. So, following the longer route is not an option for any of them.2 However, there is the possibility of another route, a sort of hybrid method, which stands between the longer and shorter routes. This would be to operate at a mezzanine level, as it were, and I shall call it the ‘middle route’. Initially it would involve analysing the nature of justice along the lines of the shorter route: first, following the state–soul parallel, we analyse justice in the state as the proper interaction of the three classes in the ideal city, producers, auxiliaries, and guardians; next, we confirm that the soul also has three corresponding parts, appetite, spirit, and reason; then we apply the same account of justice to the soul: the just soul is one characterized by psychic harmony or health, in which each part sticks to its proper function. In the case of reason, this requires knowledge of what is good for the soul, which it can use in its leadership over the other two parts. But so far, this account is extremely vague about the activity of reason. By following the middle route, we can go some way towards rectifying this. In the central books of the Republic, we find out much more about the knowledge that reason will possess: it is knowledge of the Forms, transcendent entities that exist separately from particulars. Supreme among them is the Form of the Good, which acts as a first principle for understanding all the other Forms, including justice. It is this level of understanding that characterizes the activity of the rational part, and which it will use as it tries to manage the life of the soul. So far, this shows how even a modest grasp of the metaphysics and epistemology of the central books can start to deepen our understanding of the nature of justice. But, crucially, it also throws further light on the value of justice. In the shorter route, this was expressed in terms of harmony or health. But in the central books, we realize that justice involves the rational part attaining knowledge of the Forms, and this is something that has a transformative effect on the soul. The Forms are divine in nature, and the activity of knowing them, intellectual contemplation, sometimes characterized as a Form of love (eros), involves the soul assimilating itself to the Forms, and hence to the divine (cf. VI 500b8–d2).3 So to follow this line of argument, which is to take the middle route, is to change our understanding of the value of justice from something that sounded relatively mundane, a form of psychic health, to something transcendent. Following the middle route can also throw light on another issue. At IV 442d11–443b2, Socrates asserts that the person who is just in the sense he has described, i.e. whose psychic parts stand in the correct relations to each other, will also be someone who performs actions conventionally thought to be just: that is, he will refrain from theft, adultery, treachery, and so on. This claim has puzzled scholars for a long time: why, they ask, should Socrates be so confident that his innovative account of justice will

  As Shorey (1937) I 378 n. b points out, Plato proposes the longer route, but does not follow it.   See Kraut (1992a) 319–21.

2

3

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  61 converge with justice as conventionally conceived?4 Perhaps the source of the problem is that in book IV, when Socrates says that reason will have knowledge of what benefits the soul (442c4–7), he gives no indication as to what this benefit actually is. Until we have a more determinate conception of the good, we do not really know what sort of life will be led by the just person, guided by his knowledge of the good. So, one way of solving the problem would be to take the middle route. In the central books we begin to hear of a good life that involves the contemplation of the Forms. Part of what I called the transformative effect of contemplation will be a radical change in what one values: goals such as appetitive pleasure, power, or honour will seem utterly insignificant in comparison with the divine order of the Forms. Hence the philosopher will simply lack the inclination to pursue such goals, which means that they will no longer have any motivation to commit acts conventionally considered unjust (cf. VI 485a1–487a5). This is one way in which the middle route allows us to narrow the gap between Platonic and conventional justice. Another is to focus on the Form of the Good. Though Socrates tells us little about its nature in the central books, scholars have speculated that it is a source of unity and order. The more the philosopher understands the good, the greater their love for it, and the greater their desire to see that order replicated elsewhere, specifically among human beings. Thus, the philosopher, i.e. the just person, will always desire that their relations with others reflect the harmony and order they see in the Form of the good, and will therefore abstain from any action at odds with this. Again, this implies that their actions will also be conventionally just.5 It should be clear how the middle and longer routes differ from each other. The longer route involves knowledge that starts from a supreme first principle, the Form of the Good, and uses this to derive knowledge of the other Forms. So, one would only understand the value of justice by understanding the Form of justice itself, which would in turn require seeing how its nature and value are explained by the Form of the Good. In the middle route, by contrast, we start out as if on the shorter route, ­characterizing justice as the proper interaction of the three parts of the soul, but then supplement our understanding by having recourse to claims made in the central books.6 But we do not ourselves derive these claims from the Form of the Good, and our grasp of them falls far short of philosophical dialectic as it is described in book VII. Indeed, many of the key metaphysical claims in the central books take the form of announcements (often supported by images or analogies) rather than conclusions derived from extended argument. So, the middle route is a hybrid because it follows the basic pattern of the shorter route while ‘borrowing’ a number of claims that only a 4   For discussion, see Sachs (1963), Kraut (1973), Cooper (1977), Irwin (1977) 233–9, Vlastos (1981) 111–139, Annas (1981) 153–69, Dahl (1999), Kraut (1992a), Irwin (1995) 256–61, Brown (2004), Singpurwalla (2006) and Kamtekar (2010). 5   See Annas (1981) 260–71. 6   Cooper (1977) 151 gives a good articulation of the middle route. According to him, Plato’s defence of justice proceeds in two phases: the first comes in book IV; books V–VII then ‘fill out’ (his words) the earlier phase.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

62  Dominic Scott philosopher who has followed the curriculum described in the central books would be able to master. So far, I have been describing the middle route as something that it would be ­possible to take. Whether Socrates actually follows it at any point in the Republic in defending the value of justice is another question. But before answering it, I should clarify one further point. This question is about whether he uses the middle route to support his ethical argument. But where the political theory of the work is concerned, there is no doubt that he follows a middle route to develop his conception of the ideal state initially set out in books II–IV. The impetus for books V–VI is to defend a group of proposals about the rulers of the ideal state, including the thesis that they should be philosophers. It is this that leads into the metaphysical and epistemological excursus that lasts from V 473c to the end of book VII, introducing the reader to the kind of knowledge that the philosophers will have and the education they will need to acquire it. But if it is obvious that Socrates follows a political middle route to extend our understanding of the ideal state, we cannot simply assume that he also uses an ethical one to deepen his defence of justice in the individual soul. So as regards the question of whether Socrates actually follows an ethical middle route in the Republic, let me briefly sketch my own answer.7 I do not think that he does so in the central books, in large part because of the way they are framed. At the end of book IV, Socrates had announced his intention to continue the defence of justice by studying the various types of injustice (445c4–e3). Then, at the beginning of book V he is interrupted by some of his interlocutors, who want to know more about the ideal state, in particular certain details about the life of its rulers. This request prompts him to launch into a long discussion about women rulers, the abolition of the family, and the need for philosopher-rulers (and the nature of their education). But I stress that this whole section is framed as a digression—a digression from the task of defending justice. This is confirmed from the way book VIII begins: But now that we’ve dealt with this topic, let’s recall the point at which we began the digression that led us here, so that we may follow the same path again.  (543c4–6)

So, just from the way the central books are presented, we should be very wary of reading them as part of the defence of justice. Although they contain the materials that one might use in following the ethical middle route, the way they are framed gives us a strong reason to resist the idea that Socrates does actually follow this route as a way of continuing defence of justice, even though he clearly follows the political middle route. As for books VIII–IX, my view is that Socrates only starts to follow the middle route towards the end of book IX. The long argument that runs throughout most of VIII–IX, the description of the degenerate characters, does not make any use of the metaphysics 7   I have discussed it at length in Scott (2015) 58–83. I should point out, however, that we do not actually need to answer the question for the purposes of this essay. It is enough to show that the middle route is a possibility in order to raise the issue of who would actually benefit from taking it.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  63 of the central books, and is a continuation of the shorter route as found in books II–IV. The two pleasure arguments of IX 580d–588a, on the other hand, both make reference to the just life as a life of philosophy, and as such might seem more promising candidates. Nonetheless I think that the first pleasure argument (580d–583b) ranks the philosophical life the highest without actually alluding to the content of what the philosopher does: it makes no reference to the Forms, or to the metaphysics and epistemology of the central books. In other words, Socrates is claiming that the philosophical life is the most pleasurable without actually committing himself to a distinctively transcendent conception of philosophy. It is only with the second pleasure argument (583b–588a) that we get a clear reference to the Forms. The philosophical life is said to involve the ‘truest’ pleasures, which are intellectual pleasures because they involve a relation to what is more real and so are ‘more real pleasures’ (585b11–c6). It would be impossible to make sense of this argument without referring to the Forms. However, this argument, which is remarkably brief and allusive, does not count as an example of the longer route: it is not presented as proceeding from an unhypothetical first principle. Rather, it is an excellent example of the middle route: it explicates the value of the just life by referring to unargued claims about the Forms. In the final argument of book IX (588b–592b), which acts as a summation of the moral argument of the Republic as a whole, Socrates likens the soul to a composite of three different animals: reason is a human being, spirit a lion, and appetite a manyheaded hydra-like beast. In what follows, he characterizes the defender of injustice as recommending that we allow the human to be dominated by either by the lion or the hydra (588e–589c). By contrast, justice is a matter of allowing reason, which is the best part of us and perhaps the most divine, to rule over the others. As such, it is clearly preferable. This is an extension of the psychological argument mounted at the end of book IV and developed in the first argument of VIII–IX: it stresses the conflict that results from injustice, and the harmony that accompanies justice. There is no explicit mention of the metaphysics of the central books, and I think the argument should be read as the culmination of the shorter route.

2.  Questions about the Audience I now turn to the main question of this essay, about the range of people who can benefit from exposure to philosophical argument. As I stated in the introduction, this is complicated by the existence of different levels of argument, leaving us with three distinct questions: who is suitable for the longer, middle, and shorter routes, respectively?

[a]  The Longer Route First, then, we need to ask, on evidence internal to the text, whether everyone, whatever their age or background, should (ideally) be expected to take the longer route, proceeding via an understanding of the Forms towards knowledge of the value of justice.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

64  Dominic Scott It would seem not, because there are certain passages in the central books stating that most people are incapable of knowing the Forms:8 Will not those be few who would be able to approach beauty itself and contemplate it in and by itself?  (V 476b9–10) ‘Can the multitude possibly tolerate or believe in the reality of the beautiful in itself as opposed to the multiplicity of beautiful things, or can they believe in anything conceived in its essence as opposed to the many particulars?’ ‘Not in the least’, he said. ‘Philosophy, then, the love of wisdom, is impossible for the multitude.’  (VI 493e2–494a3)

I take these passages to show that most people could not follow the longer route. But there is an ambiguity here. Does Socrates mean that, as things are (e.g. in Athens), most people are incapable of apprehending the Forms, but could do so if exposed to the right upbringing in an ideal state? Or does he think that even in an ideal state the number of people capable of following the longer route will be small? It seems the latter, since even in the ideal state the guardian class is small (428e7–429a3): this only makes sense on the assumption that only a very few have the right aptitude. This also ties in with a passage in VI 491a8–b3, where Socrates says that those who have a philosophical nature are rare. There is plenty more to be said about this issue—about why Plato takes such a view. It seems inconsistent with his approach in the Meno, where at the end of the slave boy demonstration Socrates states that, if he kept working, the boy would have as much knowledge about geometry and all other disciplines as anyone (85c10–d1).9 On the other hand, the Phaedrus seems to assume that knowledge of Forms from a prior existence is buried more deeply in some souls than others (cf. 248c2–e3), which is perhaps more in tune with the Republic as I have interpreted it. Why Plato shifted his stance in this way is an interesting question, but I do not have space to pursue it here.

[b]  The Shorter Route If the longer route is not open to the majority of people, what about the shorter? Let me start with a prima facie case for holding that they can follow it. First, one might think the whole point of insisting upon the possibility of a shorter route alongside the longer is to allow for an open-house approach to moral argument, despite the unavailability of the longer. (Without the possibility of a shorter route, Socrates would have left us with a very bleak state of affairs.) Second, the argument about justice in Republic I is dominated by the argument with Thrasymachus, where Socrates seems to be doing exactly what he promises in the Apology: debating with an open critic of the just life, and attempting to win him over. If he were prepared to debate with so extreme an opponent as Thrasymachus, he might be willing to argue with anyone, ‘young or old, citizen or foreigner’. Note that the position espoused by Thrasymachus is also that   Both translations are by Grube, rev. Reeve in Cooper (1997).   See Popper (1966) 129.

8

9

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  65 of ‘the many’ (358a4 and c7). So, on these grounds we might think that Plato maintained the approach of the Apology right through the Republic, although he applied it only to the shorter route. But this interpretation is open to question. I agree that Socrates’ willingness to argue with Thrasymachus in book I suggests prima facie that he would be prepared to discuss the nature and value of justice with anyone. But consider what happens after that ­argument concludes. At the beginning of book II, the work undergoes a hugely important dialectical shift. Thrasymachus goes silent, and remains so for the rest of the dialogue (apart from one very brief comment at the beginning of book V: 450a5–6). Glaucon and Adeimantus revive his challenge at the beginning of book II (cf. 358b8), because they do not think Socrates has adequately answered it. But in doing so, they stress that they disagree with Thrasymachus; what they want are stronger reasons for something they already believe—viz., that justice is good in itself. They want reinforcement, not conversion. So, they play devil’s advocates, setting out the case against justice as forcefully as they can. Indeed, so powerful is their case, Socrates is amazed that they do not actually believe it; that they do not is something he attributes to their character or upbringing (367e5–368a7). If one looks at the defence of justice as it unfolds in books II–IV, there is no doubt that Socrates remains focused on Glaucon and Adeimantus, and hence on reinforcement rather than conversion. This is particularly clear, for instance, when the argument comes to a temporary close at the end of book IV: Socrates talks directly to Glaucon to see if he is now convinced about the value of justice. He does not suddenly address Thrasymachus again, or wonder if ‘the many’ might have been converted to his point of view. The same applies when he resumes the argument at the beginning of book VIII. He presents the argument as continuous with books II–IV and talks as if the ­conversation will be carried on as before (543c4–6). There is no indication that the dialectical context has changed and that a different audience is being addressed. So too when the main argument of books VIII–IX draws to a close, i.e. where Socrates gives his ranking of the five different lives, IX 580b: having now gone through the list of vicious characters in detail, Socrates asks Glaucon to rank them in terms of happiness or misery, and it is Glaucon who answers—and not on behalf of anyone else. Again, the same can be said of the two pleasure arguments that immediately follow. To appreciate what is happening dialectically in the Republic here, we need to make a sharp distinction between Plato as author and Socrates as principal interlocutor. Plato might be using the interactions between Socrates and the other interlocutors to provide the reader with a commentary on some of the positions that Socrates adopts, particularly those concerning education and dialectic. In book I, Plato presents Socrates as willing to engage with a hostile critic, and so (we may assume) with anyone, whatever their background or assumptions about justice. But Plato presents the conversation with Thrasymachus as a failure: Thrasymachus becomes increasingly annoyed, and even at the end of the argument seems unconvinced about Socrates’ view of the relation between justice and happiness (cf. 350c12–354a11). But it is Plato who decides to

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

66  Dominic Scott jettison Thrasymachus and replace him with a quite different species of interlocutor, much closer to what Aristotle had in mind in the Nicomachean Ethics: someone who is well brought up, already believes what is being argued for, but wants deeper reasons for doing this. So, what I am suggesting is that Plato, as the author of the work, even as he presents Socrates in book I as willing to engage with anyone (as in the Apology), had already started to move in the direction of Aristotle’s approach to the audience and purpose of moral argument. Why would Plato have made this shift? The answer, I suggest, is that by the time of the Republic he had come to think that our susceptibility to rational persuasion is quite fragile, dependent on our cultural upbringing and moral character. As far as upbringing is concerned, there is an important passage at the very beginning of the discussion of poetry in book II, where Socrates claims that they need to keep a careful watch over the content of poetry. The stories children hear when very young are able to instil beliefs that become indelible. At II 377a12–b3 he states that the mind is very malleable when young, and able to absorb any pattern that one might want to impress on it. He then says that beliefs taken in at that age tend to become unalterable (378d7–e3). As his programme of censorship goes along, he makes clear what sort of tales he has in mind: those that portray the gods or heroes in an unseemly way, so that they appear to commit acts of deceit, violence, and intemperance. Such stories will instil the belief that this sort of behaviour is acceptable. He also thinks poetry has the same power to instil false beliefs about the value of justice (392a12–b6). In the ideal state, the trainee guardians (and presumably all the citizens) will not be exposed to such content. But in Plato’s own time, many people would have been brought up listening to Homer. This point about the effects of poetry on the very young is central to our interests. Such content might leave a residue of indelible false beliefs about the virtues, including justice, making people unresponsive to rational argument even of the kind Socrates mounts in the shorter route of the Republic. So, the lack of the right early education is not just relevant to following the longer route, but the shorter one as well. The other factor that affects our susceptibility to rational argument is the strength of non-rational appetites and emotions. By the time of the Republic, Plato had decided that, alongside reason, the soul has appetitive and spirited parts, which are able to disrupt reason in various ways. At certain points in the work, he points to the phenomenon in which a strong desire or emotion might prevent reason from forming a belief, or force it to abandon a belief it used to hold. He first mentions this phenomenon (even before the formal introduction of the tripartite psychology of book IV) at 429c7–430b5, where he talks of (physical) pleasure having the capacity to displace true beliefs, like a solvent washing away a colour dyed into a fabric. Then, in his account of the democratic character in book VIII, he describes how the growing power of his appetites prevents him from even listening to the admonitions of his household, who are warning him to live a more frugal lifestyle (560c6–d1). All this is of course highly relevant to our question. If the defence of justice (and the other virtues) would require one to moderate and even subjugate non-rational desires and emotions to reason, those who are already

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  67 in the grip of such passions might well refuse even to listen to the argument, just like the democratic character. As a result, their behaviour might resemble the intransigence of Thrasymachus in book I. So here are two distinct reasons why Plato could have become pessimistic about engaging just anyone in moral argument, and why he might have already adopted the position we associate with Aristotle. This in turn would explain why he depicts the conversation of the Republic as he does—starting with an unsuccessful attempt to ­convert Thrasymachus, and then shifting to the quite different project of reinforcing the values of the well brought up. But the many, just like Thrasymachus, are no longer the intended beneficiaries of the argument. Note that the failure to respond to argument applies as much to the shorter route as to the longer. It is not the intellectual difficulty of the argument that is at issue—whether it appeals to a level of reality too abstract to grasp—but simply that the conclusions of the argument are in conflict with a person’s upbringing and moral character. I have argued for this rather pessimistic reading of the Republic elsewhere.10 However, my view now is that things are (still) more complicated. I think that Plato was indeed wary of engaging with just anyone in moral argument, for the reasons stated, and I still think that he uses the behaviour of Thrasymachus to draw our attention to the problem. But I now think that he may not have completely closed the door on the sort of open-house or protreptic argument endorsed in the Apology. My reasons for this change of view come from looking at the very closing passage of book IX (588b–592b). This is how it begins: Now that we’ve reached this point in our argument, let’s take up what was said at the outset, which is what led us here: acting unjustly, it was claimed, benefits the person who is completely unjust, but who is thought to be just . . . . Since we’re agreed on the power that just and unjust action each has, let’s now take up the argument with him.  (588b1–8)

As we saw when discussing this argument (p. 63), Socrates goes on to use an image in which reason is compared to a human, spirit to a lion, and appetite to a hydra-like beast. He then redescribes injustice as the domination of the human by either the lion or the hydra (588e–589c), with all the consequent turmoil and conflict, and justice as the human dominating the others, which results in friendship and harmony. What is so interesting for our purposes is that Socrates here wishes to address the out-and-out sceptic, i.e. someone who genuinely believes that injustice benefits the agent. This is already suggested by the passage just quoted, but it comes out even more clearly at 589c. Talking of the defender of injustice, he says: S.: Then let’s persuade him gently—for he isn’t wrong of his own will—by asking him these questions.

  See Scott (1999).

10

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

68  Dominic Scott He then interprets conventional terms of censure in terms of the image—of one or other of the animals dominating the human—and vice versa for terms of praise. Initially, the argument is addressed directly to the imaginary interlocutor, though after a while Glaucon says that he will reply on his behalf (590a). In the rest of the ­passage, Socrates goes through a whole list of traits conventionally praised or blamed, translating them into the terms of the image, and thereby explaining why they either benefit or harm the soul. That Socrates is addressing the sceptic, albeit indirectly through Glaucon, is crucial for our interests: contrary to my previous suggestion, he does not quite confine all his attention to those like Glaucon and Adeimantus, who have been well brought up and merely need reinforcement. So, I think it safe to assume that Plato himself was still interested in addressing an unrestricted audience in the defence of justice. On the other hand, because he has Socrates engage in a form of indirect advocacy here, I would suggest that he is wary about the difficulties of such a project, for all the reasons I stated when setting out the pessimistic view of Plato’s dialectical ambitions.11 In support of the idea that Plato is only cautiously optimistic about the prospects of converting an unrestricted audience, I would suggest that in this final argument he does something to destabilize our confidence in the success of his attempt to win over the sceptic. An obvious question that arises in this passage is why Socrates does not turn back to Thrasymachus. After all, they are talking about someone who defends the value of injustice and Thrasymachus is sitting there among them. Yet he makes no comment. It is very interesting that in this argument, at 590d2–3, we have one of the few references to him after book I: his belief that rulers should exploit their subjects. This reminds us of his presence and at the same time highlights his refusal to step in and state that he is now beginning to be won around. He doesn’t, because he isn’t. That fact is disconcerting, placed there by Plato to remind us of the difficulty of ever breaking out beyond the circle of friends like Glaucon and Adeimantus. In sum: the final argument of book IX does show that Plato is prepared to consider arguing with a sceptic who holds the view favoured by the many, that justice is another’s good and not one’s own. To this extent, it would be wrong to say that, at the time of writing the Republic, Plato had gone all the way over to the position of the Nicomachean Ethics, abandoning any attempt to reach out to the sceptic. On the other hand, he shows great caution: he does not address the sceptic directly, but consults with Glaucon about the prospects of indirect advocacy, and he is extremely careful about the rhetorical means needed for any successful conversion. So, the pessimistic interpretation I have been reconsidering remains important: it can be seen as a point to which Plato might have been tended, even if he left it to his greatest student to make the final move.

11   The reference to the need for gentleness at 589c6 also shows a concern about exactly how we engage with the sceptic.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  69

[c]  The Middle Route So, with some caution, we can say that most people could be expected to follow a version of the shorter route. The longer route, though, is out of the question. But given the points made in Section 1, there is a third possibility: just because the many cannot follow the longer route in its entirety, it does not follow that they could not follow the middle route as described. But why might one think even this a possibility? Here is one reason. When discussing this route, I distinguished two contexts in which it might be followed: ethical and political. The ethical involves articulating the value of justice in the individual, as ­actually happens in the second pleasure argument of book IX. But in the second context, the political, the middle route is followed in order to justify the rule of philosophers in the ideal state. Now there is no doubt that Socrates follows the middle route to do this at the end of book V and for much of book VI. As part of this argument, he needs to explain what qualifies the philosophers for their role as guardians, and this in turn requires him to introduce the Forms as the objects of the knowledge that enables the philosophers to rule. This much is obvious. But what is interesting from our ­perspective is that he sees himself as needing to convince not just Glaucon and Adeimantus of the need for philosopher-rulers but those who are deeply hostile to the idea (VI 498c6–8), an audience that includes the likes of Thrasymachus. Hence it seems that, in order to convince such an unrestricted audience, he has to persuade them of the existence of Forms. If so, they need to follow—and hence be able to follow—the middle route in its political Form. Even though they can never come to have knowledge of the Forms, they can presumably have some sort of grasp, sufficient to follow the case for philosopher-rulers. So, if they can follow the political version of the middle route, the possibility is there for them to follow the ethical as well. In what follows, I shall start by asking whether Socrates really does envisage exposing an unrestricted audience to the Forms in V–VI: although I said this seems to be the case, we need to look at the relevant texts quite closely to be sure; perhaps instead he thinks there is a way of defending the proposal for philosopher-rulers without mentioning the metaphysics at all, and the political middle route is only for Glaucon and Adeimantus. However, I shall maintain that he does envisage doing so, for an unrestricted audience as well the two brothers. I then turn to the question of whether this means that he would also expect an unrestricted audience to follow the ethical middle route. [I]  The Defence of Philosopher-Rulers in Books V–VI The discussion of philosopher-rulers in Rep. V–VI breaks into roughly two sections: the first a discussion with Glaucon, the second (beginning at 487b1) with Adeimantus. The proposal for philosopher-rulers emerges as part of a discussion about the feasibility of the ideal state, when Socrates claims that it will never come into existence until either the current rulers turn to philosophy or philosophers acquire power themselves

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

70  Dominic Scott (473c9–e4). The announcement of the proposal is dramatic: immediately beforehand, Socrates warns Glaucon that what he has to say may provoke ridicule and contempt (473c8). Glaucon also focuses upon the paradoxical nature of the proposal: Socrates, after hurling a speech and statement like that at us, you must expect that a great many people (and not undistinguished ones either) will cast off their cloaks and, stripped for action, snatch any available weapon, and make a determined rush at you, ready to do terrible things. So, unless you can hold them off by argument and escape, you really will pay the penalty of general derision.12

Curiously, he does not say what he himself thinks of the proposal, but offers to support Socrates in its defence. This turns out to mean offering to answer Socrates’ questions in the argument that follows. What Socrates appears to be promising from now on is as much to address the sceptics just mentioned as to persuade Glaucon. (That said, the outrage and ridicule of the unbelievers is not actually referred to a great deal in the discussion with Glaucon.) The first part of their discussion centres on the distinction between the philosopher and the non-philosopher. At 474c8–480a13 philosophers are distinguished as having knowledge of Forms; as such, only they have the requisite expertise to rule in the state (cf. 484a1–d9). At 485a1–487a8 Socrates makes the additional point that philosophers, because they are lovers of wisdom, will have all the character traits and talents that are expected of a leader, e.g. courage, high-mindedness, temperance, and justice. During most of this section (474c8–487a8), Socrates and Glaucon talk with each other, without making reference to the external audience so vividly portrayed at 473e5–474a4. There is one exception to this: having defined the philosopher as a lover of Forms for Glaucon’s benefit at 475e5–476d5, Socrates then mounts the famous argument against the ‘lovers of sights and sounds’ at 476e4–480a13. This is a group of people who pride themselves on their knowledge of beauty, attending as many cultural festivals as they can. Socrates’ claim is that, immersed in the empirical world and ­unable to recognize the Form of beauty itself (476b6–7), they count as having only belief, not knowledge. And yet, they themselves think they have knowledge; so, the purpose of the argument of 476e4–480a13 is to convince them otherwise. Socrates introduces the passage as follows: What if the person who has belief but not knowledge is angry with us and disputes what we are saying? Is there some way to console him and persuade him gently, while hiding from him that he isn’t in his right mind?13

This is a different group from the sceptics mentioned at the beginning of the passage, though they may perhaps be a subset; and their anger is not directed at the proposal for philosopher-rulers, but at the claim that they lack knowledge.   473e5–474a4, trans. Grube, rev. Reeve in Cooper (1997).   476d7–e2, trans. Grube, rev. Reeve in Cooper (1997) modified.

12 13

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  71 But the passage is of interest to us because in it Socrates appears to introduce the Forms to a highly sceptical, not to say hostile audience. There is room for doubt as to how deeply the lovers of sights and sounds are to engage in metaphysics. Do they, by the end of the argument, have as clear a conception of the Forms as Glaucon at the very beginning of this whole section? Or do they just have some vague sense that certain things ‘are’ without qualification, as distinct from the many things that ‘roll around between being and not-being’ (479d3)? How far is Socrates really explaining his ­ontology to them? Or is his main purpose just to induce a sense of intellectual humility in them, even at the cost of leaving them a little baffled? Even if we did claim that they end up having as developed an understanding of the ontology of Forms as Glaucon, there is the further question of how narrow the ­audience is. Most scholars take it as confined to a strange group of people, those who frequent cultural festivals and pride themselves on their knowledge of beauty (475d4–8). Socrates broadens the audience to include the ‘lovers of crafts’ (476a11; cf. 475e1); but these could be a very similar group: those who admire beautiful objects created by craftsmen. He also includes certain ‘practical people’ (praktikoi, 476a11), whom I take to be a distinct class of people engaged in public affairs. Since he implies the same argument could have been used with the example of justice (476a5 and 479e2), I imagine the praktikoi would be people who think they have knowledge of justice based upon their extensive experience of practical affairs (in law courts, for instance). So far, then, it is difficult to determine to what extent and how widely Socrates intends to share his views about metaphysics and the Forms, beyond talking to the likes of Glaucon.14 But we can settle some of these questions if we look through the rest of the discussion of philosopher-rulers, when Adeimantus takes over the conversation from Glaucon. After the argument with the lovers of sights and sounds, Socrates returns to discussing directly with Glaucon and no longer talks explicitly of the external critics of the proposal. But once he has tried to draw the argument to a close, concluding that  philosophers alone have both the expertise and character to rule in the state, Adeimantus breaks in with an objection. No one, he admits, will be able to dispute Socrates’ argument step by step. But when people look at its ultimate conclusion, they will object that those who study philosophy are either cranks, corrupt or, if actually decent, useless to their cities (487d1–5). To Adeimantus’ surprise, Socrates agrees with this description—but only as a description of the way philosophers actually are, not as they should be. He begins by offering the ship of state analogy, which aims to show how philosophy, while having a reputation for irrelevance, is nonetheless crucial for running the state. In the image, the sailors who fight with each other to take control of the ship stand for rival democratic politicians; meanwhile the person who stands apart from the fray and instead studies the stars, the seasons, and the winds, is dismissed as useless (489a1). This figure, of course, stands for the philosopher, whose highly abstruse   For further discussion, see Vasiliou (2008) 240–6, opposing Bobonich (2002) esp. 58–66.

14

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

72  Dominic Scott knowledge is vital to the well-being of the state (paradoxical as that may seem), just as knowledge of the stars is vital for navigating the seas (paradoxical as that may seem). In  what follows, Socrates attempts to explain the views of his opponents, and to acknowledge what truth there is in them. Each of the categories listed by Adeimantus at 487d1–5—the cranks, the vicious, and the useless—receive an extended treatment in the course of 490e4–496e3. In actual states, anyone with a nature genuinely suited to philosophy, possessing the full list of characteristics and talents listed at the beginning of the book, is likely to be prevailed upon by his relatives to relinquish philosophy and pursue political power instead. This accounts for the reputation of philosophy to ­corrupt its students. Because they desert philosophy, another group, the sophists and eristics, rush in to fill the void. These pseudo-philosophers are responsible for giving the subject a reputation for crankiness. The few decent people who stay with the subject throughout their lives (such as Socrates himself) stay out of politics, for fear of having to resist the wishes of the multitude; hence their reputation for uselessness. In all this, Socrates shows how one can address the objection posed by Adeimantus, explaining how the opponents of philosophy came to form their opinions and acknowledging there is a certain truth in them. At the same time, he tries to show that, present conditions notwithstanding, philosophers need not be as they currently are, and that it is possible for decent people to be both philosophers and useful to their cities. So far, i.e. up to 498c4, he has been talking directly to Adeimantus, just as most of the earlier discussion of the proposal was addressed directly to Glaucon. But now, in a short ­section, 498d1–502a3, he brings the opponents of the proposal much more clearly into view, and presses the question of whether and how they might be persuaded to change their minds. This section is central to our interests, as it is specifically concerned with the sorts of arguments with which Socrates intends to win a hostile audience round to his way of thinking. Before going any further, we should pause to note how Plato describes the audience at issue. At the very beginning of the passage, Adeimantus points out that ‘the majority of the auditors’, starting with Thrasymachus, will be unconvinced by Socrates (498c6–8). As the passage goes on, we hear simply about ‘the majority’ or ‘the many’ (499d8, 10 and 500d11). Is this just a shorthand way of referring back to the same group at 498c6–8, ‘the majority of the auditors’, or is he now talking about hoi polloi in our sense, which is presumably a particular socio-economic group? Pursuing this latter question would take us too far afield. But I think it safe to assume that in what follows Socrates is addressing an unrestricted audience, i.e. he is not excluding anyone from the argument because of their prior intuitions, upbringing, philosophical abilities, or character. In this passage he lists a number of strategies for dealing with an unrestricted audience. My specific question here is whether any of these involve reference to the Forms. One reason for assuming this to be the case stems from the structure of the discussion with Adeimantus as a whole. From the outset, the concern was to persuade the hostile critics mentioned in his initial objection (487b1–d5). Even though Socrates spends

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  73 most of the section as though he were only trying to win over Adeimantus, we should not forget that, ultimately, he wishes to use these arguments on a wider audience. Thus, at 500a1, he tells Adeimantus to persuade the audience15 of the need for philosopherrulers by defining them ‘just as we did even now, both their nature and practice’. This will address the original concern that philosophers are corrupt, cranky, or useless by arguing that this is based on a wrong definition of true philosophers. Now, if we take the phrase, ‘just as we did even now’, to refer to the entire discussion of philosophy, i.e. all the way back to argument with the lovers of sights and sounds in book V, we are home and dry.16 But we need to be cautious. This quotation certainly refers to what has preceded in book VI, which focuses precisely upon the nature and practice of the ­philosophers. But it is not clear whether the reference stretches all the way back to book V and includes the specific points about ontology and epistemology made against the lovers of sights and sounds. On the other hand, it does seem plausible to say that, if we are to persuade an unrestricted audience, we need to give them some idea of what philosophers actually do, which means some reference to the Forms. The best and most direct evidence in favour of this comes at 500b8–501c9, where Socrates tries to illustrate the nature and activity of philosopher-ruler by means of an analogy with an artist. I would suggest that this image is especially intended for an unrestricted audience that includes people unable to understand the Forms for themselves. It involves the artist looking to a model, which is explicitly said to be the world of Forms. Here is his description of the process: No one whose thoughts are truly directed towards the things that are, Adeimantus, has the leisure to look down at human affairs or to be filled with envy and hatred by competing with people. Instead, as he looks and studies things that are organized and always the same, that neither do injustice to one another nor suffer it, being all in a rational order, he imitates them and tries to become as like them as he can . . . . Then the philosopher, by consorting with what is ordered and divine and despite all the slanders around that say otherwise, himself becomes as divine and ordered as a human being can . . . . And if he should come to be compelled to put what he sees there into people’s characters, whether into a single person or into a populace, instead of shaping only his own, do you think that he will be a poor craftsman of temperance, justice, and the whole of popular virtue?17

A few lines later on (501b–4), he stresses that the philosophers will look to such Forms as justice, beauty, and temperance as they try to shape the characters of their citizens. This image will be used as a device to persuade an unrestricted audience to accept the idea of philosopher-rulers. So, the fact that Socrates refers here to the Forms as the model that the philosopher-ruler uses strongly suggests that non-philosophers18 need to be given some outline of what it is that the philosophers use as a model, though at 16   Here described as ‘the many’: cf. 499d10.   See Vasiliou (2008) 234ff.   500b8–d9, trans. Grube, rev. Reeve in Cooper (1997), modified. 18   Whether people in actual states like Athens, or perhaps also the producers and auxiliaries in the ideal state: see again Vasiliou (2008) 234ff., with references to Kamtekar (2004). 15 17

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

74  Dominic Scott what level of detail remains unclear. In characterizing the ruler as an artist, we say he has a model that he seeks to realize in his political activity. This model entails some kind of order and harmony that exists eternally. This much might well be communicated; but whether one starts talking about the metaphysics of the Forms to any extent is a further question. [II]  The Ethical Middle Route Now we can return to the possibility of the ethical middle route. Assume for the sake of  argument that Socrates does envisage exposing an unrestricted audience to the Forms in the political argument. Why then would he not do this where the defence of individual justice is concerned? If they are to be exposed to some account of the Forms when being persuaded of the need for philosopher-rulers, they might be expected to follow the middle route in defending the value of justice. In fact, however, I claimed that Socrates did not actually follow the middle route in the Republic even for Glaucon, except almost at the end of IX (the second pleasure argument). So, we should be wary of the idea that he would be prepared to do so for a much wider audience. But why not? The question remains: if he expects people to follow the political version of the middle route, why not the ethical as well? Let us pose the question in more detail: if a person were capable of grasping, though not fully knowing, the Forms, why would they not benefit from following a defence of justice that made appeal to them? One answer is that a feeble grasp of the Forms would only yield a weak commitment to the conclusions of any argument based on them. For rhetorical, psychological, and pedagogical reasons, therefore, Socrates embraces the safety of the shorter route. (This would help explain why Socrates is reticent in his use of the ethical middle route even for Glaucon.) But there is another consideration. Anyone incapable of knowing the Forms would, unfortunately, be unable to live the sort of just life built around contemplation. They would not be able to experience the transformation of their desires, which, according to the beginning of book VI (485a1–487a8), makes the philosopher indifferent to appetitive and spirited concerns, and hence uninterested in performing unjust actions. So, such an account of justice would not apply to someone incapable of acquiring philosophical knowledge. It would give them no reason to be just, hence there is no point trying to take them on the ethical middle route (as a prelude, perhaps, to eventually taking the longer route). However, it is still useful to take the political middle route and to use whatever grasp of the Forms they do have to explain why philosophers should rule. By contrast, the shorter route does not suffer from this problem. Whether justice is described in terms of psychic health (as at the end of book IV) or harmony (as at the end of book IX), these descriptions can easily be used to convey the value of justice to an audience incapable of following the longer route and grasping the intricacies of  Platonic metaphysics. But there is an objection here. In my view, the character described at the end of book IV as psychologically healthy is still the perfectly just person—one and the same character who will be described in metaphysical terms in

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Justice and Persuasion in the Republic  75 the central books. So how can even the description given at the end of IV provide a motivation for someone who could not aspire to become such a character? The solution to this problem lies in the fact that the description of the just person at the end of IV—call it the ‘psychological’ description—is capable of imperfect Forms that still provide non-philosophers with reasons to aspire to the just life. The same is not true of the ‘metaphysical’ description given of the just person in the central books. Let me explain. If perfect justice is conceived in terms of harmony, it is not difficult to imagine less perfect Forms: someone who may not have wisdom, as the just person of book IV does, but borrows true belief from others, and harmoniously abides by that true belief in their actions and feelings. Such a person has a quality recognizably similar to the person described at the end of book IV, though it is an imperfect version. Nonetheless, if one is attracted by perfect justice as an ideal (because it involves health or harmony), one can be attracted to the imperfect Form. And if that is all one is capable of, it is none the less attractive for that.19 But now try the same with justice as conceived in the central books. Imagine someone who will never be able to achieve full philosophical understanding, and is at best only able to pursue the middle route. They could understand the ideal Socrates is ­proposing. But why would it be an ideal for them? It wouldn’t, for the simple reason that they could never become philosophers in that way. But is there an imperfect version of the just person so conceived? This would be someone who does a bit of philosophy. Yet this level of achievement would not be such as to generate the sort of eros that transforms the philosopher, and makes them the kind of person described at the beginning of VI, with weak appetites, and hence an indifference to the sorts of goals that typically drive the unjust person. Imperfect philosophers are not imperfect lovers; they are not lovers at all. So, the whole basis of the ethical middle route is undermined: such a person will not, qua imperfect philosopher, be just in the conventional sense. So, there is no point in using the middle route if your aim is to motivate the large majority of people to turn to justice. But there is reason to use it to convince sceptics (e.g. in Athens) that the proposal for philosopher-rulers is plausible; and to  persuade the producers in the ideal state to be ruled by philosophers. Hence there is a justified asymmetry in Plato’s attitude to the ethical and political versions of the middle route.

19   The concept of imperfect virtue, and lesser grades of happiness, has been extensively discussed. Recent treatments include Kamtekar (1998 and 2004), who embraces the idea that non-philosophers can attain a form of virtue and value it for its own sake. By contrast, Bobonich (2002) 46–88 takes a far more pessimistic view, and thinks that non-philosophers (in the Republic), whether the two lower classes in the ideal state, or non-philosophers in an actual state, do not value virtue for its own sake, but only instrumentally (like the oligarch in book VIII, perhaps). I think this too pessimistic, and would argue that non-philosophers can value a form of psychic health intrinsically. For earlier discussions of the issue see Vlastos (1981) 111–139, Cooper (1977), and Kraut (1973).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

76  Dominic Scott

References Annas, J. (1981) An Introduction to Plato’s Republic. Oxford: Oxford University Press. Bobonich, C. (2002) Plato’s Utopia Recast: His Later Ethics and Politics. Oxford: Oxford University Press. Brown, E. (2004) ‘Minding the gap in Plato’s Republic’. Philosophical Studies 117 (1-2): 275–302. Cooper, J. M. (1977) ‘The psychology of justice in Plato’. American Philosophical Quarterly 14 (2): 151–7. Cooper, J. M. ed. (1997) Plato: Complete Works. Indianapolis: Hackett Publishing. Dahl, N. O. (1999) ‘Plato’s defense of justice’. In Fine ed. (1999), 207–34. Fine, G. ed. (1999) Plato II: Ethics, Politics, Religion and the Soul. Oxford: Oxford University Press. Irwin, T. (1977) Plato’s Moral Theory. Oxford: Oxford University Press. Irwin, T. (1995) Plato’s Ethics. Oxford: Oxford University Press. Kamtekar, R. (1998) ‘Imperfect Virtue’. Ancient Philosophy 18: 315–39. Kamtekar, R. (2004) ‘What’s the Good of Agreeing? Homonoia in Platonic Politics’. Oxford Studies in Ancient Philosophy 26: 131–70. Kamtekar, R. (2010) ‘Ethics and politics in Socrates’ defense of justice’. In McPherran ed. (2010), 65–82. Kraut, R. (1973) ‘Reason and justice in Plato’s Republic’. In Lee, Mourelatos & Rorty, eds. (1973), 207–24. Kraut, R. (1992a) ‘The defense of justice in Plato’s Republic’. In Kraut ed. (1992b), 311–37. Kraut, R. ed. (1992b) The Cambridge Companion to Plato. Cambridge: Cambridge University Press. Lee, E. N., Mourelatos, A. D. P., and Rorty, R. M. eds. (1973) Exegesis and Argument. Phronesis Suppl. I. McPherran, M. (2010) Plato’s Republic: A Critical Guide. Cambridge: Cambridge University Press. Popper, K. R. (1966) The Open Society and its Enemies. 5th edn. London: Routledge and Kegan Paul Ltd. Sachs, D. (1963) ‘A fallacy in Plato’s Republic’. Philosophical Review 72 (2): 141–58. Santas, G. ed. (2006) The Blackwell Guide to Plato’s Republic. Oxford: Blackwell Publishing. Scott, D. (1999) ‘Platonic pessimism and moral education’. Oxford Studies in Ancient Philosophy 17: 15–36. Scott, D. (2015) Levels of Argument: A Comparative Study of Plato’s Republic and Aristotle’s Nicomachean Ethics. Oxford: Oxford University Press. Shorey, P. (1937) Plato: The Republic. 2 vols. Loeb Classical Library. Cambridge, Mass.: Harvard University Press. Singpurwalla, R. (2006) ‘Plato’s defense of justice in the Republic’. In Santas ed. (2006), 263–82. Vasiliou, I. (2008) Aiming at Virtue in Plato. Cambridge: Cambridge University Press. Vlastos, G. (1981) Platonic Studies. 2nd ed. Princeton: Princeton University Press.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

4 Plato against Democracy A Defense Richard Kraut

1. Preliminaries ‘I have tried both to write about moral philosophy and to engage in moral philosophy through discussion of its history.’ Thus, Terence Irwin, on the first page of his magisterial three-volume work, The Development of Ethics.1 Irwin’s subtitle—A Historical and Critical Study—emphasizes that his history of ethics from Socrates to Rawls is no mere summary of what was said by this figure or that, but a defense of a philosophy— more specifically, of the Socratic method of inquiry and the Aristotelian naturalism to which that method leads, when properly pursued. His immense achievement serves as an inspiring model for the far more limited task I undertake here. Although my topic belongs not to moral but to political philosophy, I hope to investigate it ‘through discussion of its history.’ As my title indicates, I am going to argue that Plato’s attack on democracy in the Republic is warranted. There is today a vast literature in the field known as ‘democratic theory,’ and much of it treats democracy as the best kind of political system—many authors, in fact, hold that only such a regime is legitimate. Although I will address a small portion of this literature, I realize that within the small compass of this essay I cannot come close to doing it justice. My aim is not to refute any defender of democracy, but to  explain why I believe that Plato’s critique deserves serious consideration. If his objection can be answered at all, it cannot be answered easily. In the Laws, Plato’s longest and most important contribution to political philosophy, he seems to drop his criticism of democracy.2 In fact, the model city-state he portrays   Oxford University Press, 2007, p. 1.   For the view that Plato softened his opposition to democracy after writing the Republic, see Christopher Bobonich, Plato’s Utopia Recast: His Later Ethics and Politics, Oxford: Clarendon Press, 2002; Thanassis Samaras, Plato on Democracy, New York: Peter Lang, 2002. It is beyond dispute, however, that in the Laws the common good is the basis of the constitution (see e.g., 875a). This principle, I will claim, plays a central role in Plato’s critique of democracy. 1 2

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

78  Richard Kraut adapts several institutions of democratic Athens. This work is the beginning of the mixed regime tradition of political philosophy—a tradition that finds a place for the rule of the many as one component of a larger whole comprised of monarchical and aristocratic elements. Aristotle, Polybius, and Cicero develop some of the themes of the Laws, and their ideas had a profound effect on modern political thought and the founding of the American republic. But I will leave aside this development in Plato’s thought, and concentrate on the uncompromisingly anti-democratic element of his thought. It might be asked: Why suppose that Plato was a critic of the political system that we now call ‘democracy’? What he attacked was dēmokratia (more specifically: Athenian dēmokratia), not democracy—he wasn’t acquainted with the regimes that are so-called. In Athens, major decisions were made by an assembly in which every male citizen was entitled to speak and vote. By contrast, in modern democracies a small number of elected officials make decisions on behalf of the rest of the citizenry. In Athens, many offices were filled by lot; although we may use lotteries to raise state revenue, we do not use the lot to choose holders of public office. Furthermore, the Greeks of Plato’s time think of dēmokratia as rule by the poor—because dēmos could refer not only to the whole body of citizens, but also to the largest portion of it, namely, those non-elite citizens whose resources were small enough to require them to work for a living.3 For us, by contrast, democracy is rule by ‘the people,’ that is, the whole body of citizens, rich and poor, elite and non-elite.4 The Greeks used dēmokratia in that way, but their term also conveyed the idea that power was in the hands of one portion of the whole, namely ordinary people, plebeians, commoners, those with fewer resources. It is important to keep these differences in mind when we read Plato, but they do not by themselves show that his critique of Athenian democracy is inapplicable to democracy in our own time. Obviously, to settle that question we must look carefully at what that critique is. Only then can we determine whether the deficiencies he thinks he sees in the dēmokratia of Athens are features that continue to exist in modern democracies. Of course, no one believes that the democracies of today’s world are perfect regimes. The question that we should ask about them is this: even if the flawed democracies of our world were transformed into perfect democracies, would they nonetheless be distressingly unsatisfactory political systems precisely because they would still exemplify features that Plato identified as the defects of Athenian democracy? And if so, is it possible, as he thought, for a superior form of government to come into existence?   See Aristotle, Politics III.8 1279b18–19. For fuller treatments of democracy see IV.4–5 and VI.4.   I will assume, as a matter of definition, that a democratic state is one in which citizens are treated as  equals and their equality plays a significant role in collective decisions. (Corporations, then, are not democratically organized; one has as many votes as one owns shares.) Here I follow Tom Christiano, Stanford Encyclopedia of Philosophy (Spring 2015 edition), s.v. ‘democracy,’ section 1. Democracy is ‘a method of group decision making characterized by a kind of equality among the participants at an essential stage of the collective decision making.’ 3 4

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  79

2.  What is Plato’s Critique? What is the Republic’s critique of dēmokratia? Where in the text does it begin? Where does it end? What are its premises? Is there one argument or several? These questions are not as easy to answer as one might suppose. But let’s begin with a few obvious points. In Books VIII and IX, Plato describes a series of increasingly deficient regimes. Democracy comes next to last. Over the course of several pages, he ridicules the way in which freedom and equality are generally understood in a democracy: freedom is treated as equivalent to the right not to be governed by anyone; and equality is understood as the equal value of every whim. Obviously, these pages contain at least part of Plato’s anti-democratic argument. But it would be a mistake to suppose that there is nothing anti-democratic in the Republic prior to books VIII and IX. For in earlier books he claims that the best-governed city is one in which all political power is placed in the hands of a small philosophical elite, an unelected group that is not subject to the control of the larger population. That depiction of the best political system implies that rule by the many is defective rule. Democracy is an inferior system because philosophers should rule, not the many. But a moment’s thought is all it takes to make us realize that the foundation of Plato’s critique of democracy is laid at an even earlier point in the Republic. The thesis of Book VI that philosophers are best able to rule is derived in part from the principle of specialization that Plato proposes when his construction of the ideal city begins in book II. Each citizen is to have just one job: the one best suited to his abilities. Farmers, artisans, doctors, cooks, soldiers—all of them should contribute to the good of the community solely in one way, by doing the job for which they are suited and for which they have received training. The conclusion implicit in this foundational principle is that ordinary workers who compose the electorate of any democracy should stay out of the business of political decision-making. Political rule is a task that should be undertaken only by those who have the ability and education that makes them excel at it. As a rough approximation, then, the Republic’s argument against democracy goes as follows: 1. In every political community, the common good requires that certain roles be filled.5 2. Those who play those roles should be qualified to do so (otherwise the common good will not be served). 3. Ruling is one of those roles. 4. The dēmos is not qualified to rule. 5. Conclusion: the dēmos should not rule.6   I will return to the notion of the common good in Section 10.  There is an echo of this argument in Jason Brennan, The Ethics of Voting, Princeton: Princeton University Press, 2011. He argues that citizens have a duty to reflect on the common good and ought to refrain from voting purely on self-interested grounds or out of group loyalty. However, he does not suggest that this duty should be enforced by law. Note too this statement of Jeremy Waldron: ‘in order for law to 5 6

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

80  Richard Kraut I don’t claim that these statements exactly reproduce in full detail Plato’s actual argument. One omission is this: Plato holds that in an ideal community each citizen should fulfill one and only one social role. If you are best qualified as a farmer, for example, you should stick to that one job; you will not do it well enough if you are allowed to try your hand at ruling. The argument just sketched omits this aspect of Plato’s thinking. It leaves it open that any citizen might play several different roles; it requires of him only that he be well qualified for each of them. Similarly, the previous argument does not include the statement that to be qualified to rule one must have knowledge of the Form of the Good. I omit these aspects of Plato’s argument because I want to extract from what he says something that still has considerable plausibility for us. For our current purposes, it does not matter that many more premises than these are advanced in the Republic as part of its critique of democracy. What matters is that Plato affirms each of these four premises, that all of them have at least some plausibility, and that if we accept them we come to an anti-democratic conclusion. Obviously, the crucial step in the argument—because it is most controversial—is the fourth statement: the dēmos is not qualified to rule. What does this mean? What evidence can be given, by Plato or anyone else, to support it? Let’s begin with this empirical observation: in all of the democracies that were known to the Greeks in the fifth and fourth centuries and that have arisen in the modern world, most citizens have not sought or received any special training that would have given them political understanding, political knowledge, or political skills. None has required that citizens receive that kind of education before they are allowed to vote or run for office. Furthermore, it is no accident that democracies have not required this of their citizens. It would be contrary to the way most people understand the word ‘democracy’ and its equivalents in other languages (including the Greek word, dēmokratia) for a state that calls itself democratic to prevent citizens from voting or holding office unless they had first become qualified—for example, by taking an exam in which they demonstrate their knowledge of matters pertaining to the common good. The root idea of democracy is that the people shall rule—not that only qualified people shall rule. Their legal right to vote is not something that they have to earn or deserve or qualify for. In light of these observations about the history of democratic states and the common use of ‘democracy’ and other words that have the same meaning, we can restate Plato’s argument more perspicuously as follows: 1. In every political community, the common good requires that certain roles be filled. 2. Those who play those roles should be qualified to do so (otherwise the common good will not be served). claim authority . . . those who participate in making it ought to do their best to address in good faith the issues on which they know they disagree.’ As the context shows, he means that they must not (even sometimes) ‘vote their interests.’ See Law and Disagreement, Oxford: Clarendon Press, 1999, p. 14.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  81 3. Ruling is one of those roles. 4. There are qualifications one ought to have, if one is to be good at political decision-making. 5. Most citizens lack those qualifications. 6. Support for (5): this is precisely what we should expect in democracies. It is their nature not to require such qualifications. As a result, citizens have little or no incentive to acquire them. 7. Conclusion: the people (that is, most people) should not rule. [Notice that ‘the people (that is, most people)’ here replaces the term dēmos. I will soon comment on this substitution.] Ideally, premise six would be replaced by a long inductive argument that looked at each democratic citizen who has ever existed, one by one: this one was not qualified to play a role in governance; nor that one; nor he; nor she; and so on. But since life is too short for anyone to carry out this induction, premise five has to be supported in some other way. Evidence drawn from carefully conducted surveys of representative populations could play a useful role here, provided they make reasonable assumptions about which qualifications citizens ought to have—which is a normative rather than an empirical question. What premise five claims is that were we to make a list of the qualifications needed for ruling well, and then count how many citizens actually have those qualifications, few would pass our tests. When people are permitted to do a job but not required to do the hard work that would make them qualified for it, many of them will be deficient. Even apart from this point about permissions and r­ equirements, it is reasonable to expect that what is difficult to do well will be done well by few. Therefore, there is some basis for accepting Plato’s conclusion that most citizens are not especially well qualified to rule. Plato’s argument against democracy rests on an argument from analogy; that is what is going on when premises one and two are put together. When we think of the various social roles that serve the common good—doctors, lawyers, teachers, soldiers, cooks, accountants, police officers, airline pilots—we expect or require those who occupy these positions to have taken special steps so that they excel at their tasks. But voting is treated differently in democracies. Friends of democracy say that everyone has an equal right to vote, even though they would never say that everyone has an equal right to be a doctor, or an accountant, or a police officer. The challenge Plato poses to these democrats is this: is there any reasonable basis for this radical difference? I pointed out in section one that the word dēmos could refer not only to the whole body of citizens, but to the largest portion of it—those who did not belong to the elite class of leisured and wealthy citizens who often used their time and resources to play a prominent role in public life. That linguistic fact might mislead us into supposing that it is poor people—and not the rich—who are the target of his critique of democracy. But nothing in my reconstruction of his argument takes him to be assuming that a typical elite citizen is qualified to rule, or more qualified than a typical commoner. I take him

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

82  Richard Kraut to be opposed to rule by the dēmos in the larger sense—the sense in which that term designates the whole body of citizens, not just its non-elite component. Rule should not be given to the whole body of citizens because most (perhaps even all) of them are not qualified to rule. Why should this task be given to a collective body that is entirely or almost entirely unqualified to perform it? Why not instead give it to the few (if there are a few) who are willing and able to serve the common good?

3.  The Ubiquity of Hierarchies As I noted earlier, part of Plato’s argument uses analogical reasoning. Special training and preparation is needed for many jobs—farming, cooking, house-building, and so on—that serve the common good. Political tasks like choosing officers, speaking in the assembly, hearing courtroom cases, and so on—also serve the common good. Therefore, those who acquire the necessary political qualifications are the only ones who are likely to perform those tasks well; or at any rate they will do their jobs better than others. Therefore, people should not participate in politics unless they have done whatever it takes to acquire the relevant political skills. Here is another argument from analogy that reaches the same conclusion. Nearly all of the social structures with which we are familiar are hierarchically ordered: power is not equally distributed; instead, some members have greater authority and responsibilities than others. Businesses, churches, universities, hospitals, orchestras, armies are some common examples. These are social institutions in which democracy is at best ill-advised and at worst entirely unworkable. But it would be implausible to suppose that the non-democratic governance of these social groups constitutes a defect in them. They are not inherently unjust institutions despite the fact that some people have a say in how they are run and others have none. That suggests that it is not necessarily an injustice if the nation-state is organized in such a way that some citizens have great powers and responsibilities, other citizens fewer, and some citizens none. At any rate, the democratic theorist needs to explain why that kind of hierarchy is impermissible in the nation-state, but permissible in every other important social organization. Non-democratic ways of operating are often an excellent way to get things done: so why not adopt these non-democratic ways in the political realm as well? If the general public should not be empowered to choose university professors, or doctors, or the heads of corporations, why should the ultimate political authority rest solely in its hands? I doubt that there are any easy answers to this question. It might occur to you that what makes the nation-state different from the other organizations mentioned is that we do not become affiliated with it through a voluntary choice. We are born as citizens of this or that country; but we voluntarily decide to attend a university or join a church. Accordingly, if a university or church does not give you a say in the way it makes decisions, then you have an easy remedy: don’t join it. The same cannot be said if the nation-state

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  83 in which you have citizen rights does not give you a say in decision-making. Either you are not legally permitted to renounce your citizenship; or, if you are, the consequences of your doing so would be so devastating that such renunciation is not really a viable option that you can freely chose or reject. The weaknesses of this attempt to explain why the nation-state is special is obvious. Parents have legitimate authority over their young children, but children did not choose to be born or choose to be born to those particular parents. Furthermore, if you are someone who does not want to belong to any hierarchically organized social structure, it is going to be practically impossible for you to live a decent life, because non-democratic organizations are ubiquitous in the modern world. You would have to withdraw from the academic world, the business world, the military world, and so on. What if you need to belong to one of these organizations to earn your daily bread? You would not be freely opting for a position in a non-democratically run community. And yet no one could plausibly suppose that all of these organizations must alter their constitutions and become democracies.

4.  A Thought Experiment: Restricting the Franchise My next point is that all of the democracies with which we are familiar, both in the ancient and the modern world, give the vote to citizens very soon after they have emerged from their childhood. That is a time of life in which we are not yet as experienced, mature, or knowledgeable as we hope that someday we will be. Suppose a political community decided to require citizens to wait much longer before they can vote—let’s say, until they are 50 years old.7 And since age often brings with it the deterioration of our cognitive functioning, let’s suppose that when you turn 60, your political responsibilities and rights come to an end. Furthermore, let’s imagine that in this society you can exercise political power during your 50s only if you have enrolled in a special program that is designed to increase your cognitive functioning, your moral maturity, and your political understanding. The special education is something you will receive from the ages of 20 to 50, and it will be paid for by the state. All that is required of you is that you do the work involved (studying hard after the work-day has ended) and demonstrate a certain level of competence. Let me say more about what the kind of training I am imagining will be given to citizens who are studying to qualify as members of the political elite, who hold office in their 50s. They are taught how to be on guard against the cognitive biases that bedevil 7   This is the age at which those selected as rulers of Plato’s ideal city are allowed to come to the last stage of their education—the one in which they finally study the Form of the Good (Republic 540a). But that is not why I have picked this number (which is somewhat arbitrary). My conjecture is that people in their 50s can easily identify both with those much younger (it was not so long ago that they were young) and those much older (those days are not so far away). Also, it would be hard to believe that those in their 50s are not yet experienced enough to fill important political offices.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

84  Richard Kraut the thought processes of the rest of us. They are trained to think about issues from an impartial perspective. They know something about history, economics, the life sciences, and the physical sciences. They have read many of the works of the poets, dramatists, and novelists. Those who exhibit anti-social or psychopathic tendencies have been disqualified. So too those who have an ineliminable antipathy to certain groups simply because they are different or outsiders. These leaders-in-training are capable of compassion, sympathy, and charitable feelings. They know how to listen to differences of opinion with an open mind. They are humane, morally sensitive, and politically sophisticated; they know how to solicit and evaluate technical advice, but are not themselves narrow specialists. This is not a technocratic elite. Let us assume that the political structure of this society resembles the ones with which we are familiar. There are legislative, executive, and judicial offices. Those who play these roles are drawn from and elected by citizens who are in their 50s and who have successfully completed the training program that makes them qualified to vote and to hold office. Familiar ‘checks and balances’ are in place so that all members of this elite group are held accountable to other members of the group. In this respect, my imaginary society is unlike the one Plato portrays in the Republic: in Kallipolis, those who have proved themselves qualified to rule are given absolute power, and there is no mechanism for removing them from office. That is one respect in which Magnesia, the city portrayed in the Laws, differs from Kallipolis.

5.  Democracy and Aristocracy It might be said that this political system should be called a democracy, because no legal or economic barrier stops any citizen from sharing in political office during a certain period of his or her life. If everyone demonstrates, at the age of 50, that he or she is qualified to exercise political power, then every citizen is, for a period of ten years, a member of the political class in which all power resides. This is rule by all the people (who belong to that age group)—because all the people (in that group) can become qualified, if they so choose. (I leave unspecified what percentage of the citizenry opts to receive the education required of the ruling class. Those who choose not to prepare themselves to take on political responsibilities do so because they are not interested in politics, and they trust others to do this job.) But there is good reason not to call this political system a democracy: it is a community ruled by few rather than many. Those between the ages of 50 and 60 are inevitably going to be a small proportion of the whole. They constitute an elite, although it is an elite to which everyone may belong for a short while. This is a state that bars citizens from voting or holding office unless they have first become qualified—we can imagine them taking an examination in which they demonstrate their knowledge of matters pertaining to the common good. As I said in Section 2, the root idea of democracy is

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  85 that the people rule—not that only qualified people rule.8 All citizens have an equal right to share in governance. It is part of our core understanding of democracy that one’s legal right to vote is not something one must earn, or deserve, or qualify for. It belongs to one by right when one reaches the age that marks the end of childhood and the beginning of adulthood. And it is not a right that is taken away merely on the grounds that one is likely to be less qualified than one used to be. There is a better word than ‘democracy’ for this regime: since it is ruled by the best— that is, those who are best qualified to engage in political activity—it would be appropriate to call it an ‘aristocracy.’ At any rate, if we want to extend the term ‘democracy’ so that this hypothetical state receives that label, it would be appropriate to call it a Platonic democracy, because it differs significantly from those regimes in which political power is guaranteed to all adult citizens whether they are qualified to exercise it or not. And if we believe that Platonic democracies are likely to be better governed than non-Platonic democracies, we would in effect have accepted the core of Plato’s critique of democracy. We would be in favor of reforming democracies of the present age so that voting is no longer a right but a privilege (open to all) that must be earned by showing that one is qualified for it. Whether we continue to call ourselves democrats or not, Plato would have been a powerful influence on our politics. The society I have imagined resembles Plato’s ideal city in its requirement that only the qualified may rule. But there is also an important respect in which it is unlike the one imagined in the Republic. In that dialogue, most citizens are not given the opportunity to receive the kind of training that will equip them some day to become rulers. Rather, that kind of education is reserved for a small number who are observed to have the requisite natural talent to learn the subjects that will be taught. But in the society I am imagining, every citizen has this opportunity, because I am supposing that all of them have by nature the capability to acquire the skills and to master the subjects that a qualified ruler will need. The significance of this dissimilarity, however, is slight. For Plato, it is essential not that those who govern be few but that they be well qualified. That the well qualified will in fact be few is a further assumption he makes, but we should be careful not to attribute to him the absurd view that smallness of number is by itself a desirable property for those who govern to have, or largeness by itself undesirable. What is fundamental to his critique of democracy is not that it is the rule of the many but that it is the rule of the unqualified. If we could convince him that there is a way in which the many can become qualified, he would drop his objection to that kind of regime.

8   Many people who think of themselves as democrats strongly support universal college education because they believe that only then will democracy work well. I do not deny that these people are democrats. My claim is that they are not properly classified as democrats if they think the uneducated should not have the right to share in democratic deliberation and voting. It is one thing to say that those who have this right should be educated; quite another to say that the uneducated should not have this right.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

86  Richard Kraut

6.  A Realistic Utopia Another point should now come into view, in the light of the thought experiment I have proposed. It is sometimes conceded by defenders of democracy that it is indeed a defective political system. But, these defenders of democracy go on to claim, this admission makes no practical difference, because other forms of government are worse. We are all familiar with Winston Churchill’s remark, ‘It has been said that democracy is the worst form of government, except all the others that have been tried.’ For practical purposes, the important question is comparative: if democracy is the least defective form of government, then we just have to live with whatever weaknesses are inherent to it. Plato may be right that there will always be unqualified rulers in democracies, but (according to this way of thinking) non-democratic regimes have even worse problems. But this way of defending democracy is no longer viable, if it is agreed that the nondemocratic society I have imagined is both possible to achieve and better governed than any democracy could be. We can do better than democracy, if we offer everyone the kind of education that would qualify them to exercise power wisely and well, and if we restrict political power to those who have chosen to become qualified to rule. I must now face the objection that the thought experiment I have carried out—the construction of an imaginary aristocracy ruled by qualified middle-aged citizens— proves nothing because it is a utopian fantasy. Political philosophy, according to this line of criticism, must restrict itself to ideas that stand some chance of being realized. Admittedly, there is nothing impossible in the supposition that this kind of aristocracy (this ‘New Kallipolis’ as it might be called) could exist. There is no contradiction or violation of any law of nature in my description of it. People do sometimes bring to politics the kind of wisdom that makes them eminently qualified to act in the political arena. If some have already done this, it is not a feat beyond human nature. We can therefore try to bring it about that more people achieve this high level of moral and political competence. There is no good reason to assume that a whole citizenry cannot accomplish this. Perhaps it will be said at this point: my description of the New Kallipolis assumes that virtue—political virtue, at any rate—can be taught, and this is where it goes wrong. Political wisdom, according to this objection, arises now and then through a fortunate combination of circumstances. There is no way to mass produce it—no way to educate children and young adults so that we know that all (or nearly all) of them will acquire the high level of moral and political competence demanded of rulers in the New Kallipolis. It should be conceded that the kind of virtue they will need cannot be ‘mass produced,’ because that phrase suggests a mechanical process in which one passively absorbs material without reflection and understanding, like a garment taking on a dye in a factory. That is not how political wisdom grows when one reads works of history, literature, philosophy, and so on. One acquires the habit of looking at moral and political

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  87 issues from an impartial perspective by actively reflecting on philosophical works which articulate and employ that perspective, and by training oneself to guard against biases that block or hinder that way of thinking. What my depiction of a New Kallipolis presupposes is that when these and other subjects are taught in the right way, that will make students more qualified to solve moral and political problems than they would otherwise be. It is not perfection that their 50 years of education aims at, but significant improvements in their ability to think intelligently about ethical and political matters. That is not a utopian fantasy—it is a reasonable hope that many teachers already have for their students. A New Kallipolis might arise from present circumstances gradually and peacefully. Here is one such scenario: an influential group of citizens announces that although they are not withdrawing from political activity and do not counsel others to do so, they are distressed by the shallowness and meanness of political discourse, and seek to take it to a higher level. They organize study groups in which they examine the severe problems facing their society, and read widely in philosophy, politics, economics, and other subjects. They try to persuade their fellow citizens not to vote or make other political decisions on the basis of self- or group-interest, but to reflect as citizens devoted above all to the common good (a perspective that is not restricted to national concerns). When they reach middle age, they become more politically active than they were before, because their confidence about political matters has increased. They are admired for their seriousness, intelligence, and maturity, and respected for their nonpartisanship. Their movement grows in size, and eventually their approach to politics is enshrined in law. The voting age is increasingly narrowed, and eventually only those in their 50s are allowed to vote. Voting is no longer a right but a position restricted to those who are qualified— although everyone is legally entitled to the wherewithal needed to qualify. Many citizens decide to prepare themselves for political life, but many others decide to leave that job to others, because they are confident that the problems facing the nation will be handled by those who are best equipped to address them. I concede that this scenario is a fantasy. It is not going to happen. But my defense of a Platonic alternative to democracy does not require me to show that a New Kallipolis will someday be established, or even that there is some chance, however small, that it lies in our future. My thesis is that it would be good for this story to become a reality. The fact that it will not do so does not impugn it.

7.  Democracy and Fairness How might a defender of democracy respond to the argument of the Republic that I have been proposing? One clever response is advanced in the third book of Aristotle’s Politics, and has been revived in recent times. It holds that a large group can be politically wiser than a small one, just as a group composed entirely of the poor can be collectively

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

88  Richard Kraut richer than any single wealthy individual. Many people can pool the limited knowledge each has, thus producing a large body of knowledge out of smaller bits. This argument accepts the key premise of Plato’s anti-democratic critique: people ought to rule only if they are qualified to do so. Aristotle and his contemporary followers reply: well, collectively they can be qualified, even if individually they are not. I will soon examine the merits of this approach, but before doing so, I would like to consider a more radical strategy that democrats might pursue: they might argue that there are good reasons to give political power to people even when they are not qualified to use it well. Democrats of this stripe concede that the quality of decisionmaking is not as high as could be achieved by other kinds of political regimes. What they insist upon is that there are more important things in politics than making good decisions. The way decisions are made, they will say, is of greater moral significance. If they are made democratically, that by itself is a decisive point in their favor. Democratic decision-making is justified because it is inherently fair; it need not be the best means to the solution of the problems facing decision-makers.9 I agree that democracy is a way of achieving a fair distribution of political power, and that this is one of its great merits. Here is what I have in mind: one of the great evils of political life is the domination of the masses by elite individuals or groups—kings, nobles, plutocrats, churches, businesses, armies, and so on. When a democracy is not in place, ordinary individuals—the commoners or plebes—can only hope that they will not be victimized or regarded as mere means to the ends of the powerful. Democracy is a way of diminishing this evil—it makes the competition for power and other resources less unfair, because it allows members of the dēmos to aggregate the small bits of power each has so that collectively they become a force that can compete in a fair fight against the power of elite individuals. According to Plato and Aristotle, in democracies ‘the people’ had so much power that they unfairly dominated elites, whereas in oligarchies that relationship was reversed. Whether or not that was an astute analysis of Greek politics, we can adapt their way of thinking about the competition between masses and elites for our own purposes. We can agree that the domination of some groups by other groups or by individuals is a problem that political institutions ought to solve, and that when democratic institutions work properly they at least make the problem less severe. Looking at the many nations of our world that are called democracies, and asking whether there would be less fairness in them if their democratic institutions were eliminated and replaced by 9   For defenses of democracy that rest on its intrinsic moral appropriateness rather than its efficacy as means to an end, see Amy Gutmann and Dennis Thompson, Why Deliberative Democracy? Princeton: Princeton University Press, 2004; Jeremy Waldron, Law and Disagreement; Thomas Christiano, The Constitution of Equality, Oxford: Oxford University Press, 2008; Ronald Dworkin, Justice for Hedgehogs, Cambridge, Mass.: Belknap, 2011, chapter 18. Other democratic theorists hold that a justification of democracy must show that it has at least some value as a way of finding good solutions to collective problems. Thus, David Estlund, Democratic Authority: A Philosophical Framework, Princeton: Princeton University Press, 2008; Hélène Landemore, Democratic Reason: Politics, Collective Intelligence, and the Rule of the Many, Princeton: Princeton University Press, 2013.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  89 the rule of military forces, or corporations, or religious leaders, the answer is clearly yes. Even if Plato is right to hold that democracy is an inherently defective regime because it is ruled by the unqualified, it remains the case that a political system in which collective decisions are made exclusively by those who rule simply because it so happens that they have the power to rule is unfair. Any system that gives power to the weak so that they are not dominated by the strong has much in its favor—it is not unfair, even if the weak make poor collective decisions. In New Kallipolis, many people choose to become qualified to rule, and many others choose not to pursue that path. Although all have a right to receive political training— a right they may freely decline to exercise—there is no political equality between the rulers and the ruled, for the rulers have all the power, and the ruled have none. But there is no unfairness in this system. So, a central thesis of many democratic thinkers must be rejected: political equality is not something that fairness requires. What is objectionable in a regime in which all power is held by such elites as corporations and armies is that neither wealth nor military might qualify one to sole possession of the nation’s collective decision-making apparatus. There is no legitimate basis in these regimes for excluding other citizens from sharing power. By contrast, in New Kallipolis, there is: some have chosen to become qualified to rule, and others have chosen not to. One further point about fairness: it is common in nations that are called democracies to place significant limits on the ability of the public to influence collective decisions. Ordinary citizens do not themselves make most of the laws—they merely elect others to do so on their behalf. Elections are not held every month or every year, but much less frequently. A constitutional convention charged with creating a new political order cannot be legally convened whenever a majority of citizens favor doing so—it is much more difficult to change the constitution than that. The reason why the public is so limited in its power is obvious: we are fearful that it will make bad decisions. In a sense, then, our democracies are mixtures of democracy and anti-democracy—institutions that give some power to the people are combined with restrictions that make sure they do not have too much power. But it would be implausible to claim that these restrictions are unfair. For there is nothing arbitrary about them: they are accepted for a good reason, namely, the quality of collective decisionmaking is higher when the power of the people is limited. We already believe that it is not unfair to give people only as much power as is compatible with good decisionmaking. In the New Kallipolis, that idea is taken one step farther: people can either trust others to make decisions for them, or they can become qualified to make decisions; but they have no right to participate in collective deliberations if they are not qualified to do so. The notion of fairness is used by John Rawls to secure the conclusion that a just society is one in which all citizens have a right to the political liberties that are characteristic of democracies. In the original position, the parties who choose principles of justice (and, at a later stage, a constitution) behind a ‘veil of ignorance’ have an equal say about which principles should be adopted, and this equality represents the fact that

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

90  Richard Kraut in life beyond the veil citizens must be treated as political equals. Political power is the collective possession of the body of all citizens. But I am doubtful that his framework for thinking about politics can be used to undermine Plato’s critique of democracy. Suppose the contracting parties are offered a choice between New Kallipolis and the characteristic legal structure of democratic societies, which guarantees regular and fair elections and makes voting a right rather than an office for which one must be qualified. Each party in the original position is to ask: which of these two systems will best advance my good (or the good of my family or group)? A plausible answer, I suggest, is that it is in everyone’s interest to be governed by a group all of whose members are qualified to deliberate rather than by a mixed group few of whom are qualified and most of whom are not. In that case, Rawls’s framework supports a Platonic aristocracy rather than a typical democracy! On the other hand, it could be argued that the usual kind of democracy is the better choice because it allows one to vote even though one has done no work to earn that power and one lacks the ability to vote well. But that should strike us as a way of thinking that should be disallowed—whether in the original position or outside it. Those who are not qualified to vote have a duty to refrain from voting. It is not a legitimate reason to establish a standard democracy that gives one the liberty to exercise power irresponsibly. But suppose there is no determinate answer to this question. Suppose, in other words, there is no way to know whether one will be better served by one of these political systems or the other. If that is so, it reveals a defect in Rawls’s way of approaching political philosophy. Whether and why democracy is valuable is precisely the sort of question that we look to political philosophy to answer.

8.  Aristotle and the Wisdom of Crowds Let’s turn now to Aristotle’s strategy for responding to Plato: he accepts the premise that the many ought to rule only if they are qualified, but argues that when they are considered as a group they can achieve the political wisdom that each individually fails to have. Unfortunately, when we look closely at the Politics, we see that what Aristotle has in mind is far from a full-throated defense of rule by the many. To begin with, he thinks that democracies should enfranchise only those who have sufficient material resources. He does not think the urban poor who live hand-to-mouth are capable of making worthwhile contributions to politics. So, when he claims that people who are on their own not fully virtuous can collectively be highly qualified as citizens, he is careful to say that this point will not apply to every multitude whatsoever. The citizens must meet some standard of good citizenship before it is the case that pooling their resources will produce good results. If they meet a modest property qualification, that is a rough indicator that they can collectively achieve a kind of wisdom: pooling the information they have, they can do an adequate job of electing their leaders, and of serving as jurors should any of those leaders be accused of corruption.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  91 There is an even greater limitation to Aristotle’s argument: it shows merely that it is possible for a multitude to be superior to a small group. Whether any given multitude actually is superior depends entirely on the facts of the case. Aristotle’s argument leaves it open that certain small groups are so superior that no multitude can ever match them in wisdom. It even leaves it open that kingship is the best of all political systems! That will be the case when the king is sufficiently impressive and his subjects sufficiently unimpressive.10

9.  Condorcet, Galton, and Surowiecki Let’s look, then, for another way to show that the many should rule because collectively they are better qualified than the few. In 1785, the Marquis de Condorcet published an ‘Essay on the Application of Analysis to the Probability of Majority Decision,’ in which he argued that if certain conditions are met, the opinion of a group of voters is more likely to be correct than the opinion of a single individual; further, the larger the group is, the greater is the likelihood that it is correct. Those conditions are that each individual has a better than even chance of being correct, each makes an independent decision, and the matter about which the decision is being made has an objectively correct answer. The theorem is intuitively obvious. One single voter has a merely better than even chance of voting correctly; but if each of one hundred voters has a better than even chance of being correct, and they converge on the same choice independently of each other, then it would be extremely unlikely that they would all be wrong. But defenders of democracy should not be comforted. For if voters are such that each is more likely than not to be wrong, then an option favored by a large number is much more likely to be the wrong choice than an option favored by just one individual. In effect, in these circumstances the fact that an option is favored by a voter is a counterindicator of its correctness. An option favored by one voter may be only somewhat unlikely to be the correct one, but an option to which there are many independent counter-indicators is for that reason much more likely to be incorrect.11 An idea resembling Aristotle’s was revived by Francis Galton near the beginning of the twentieth century and ‘the wisdom of crowds’ has recently been touted by The New Yorker columnist James Surowiecki.12 Galton, according to Surowiecki, observed a 10   Here is another reason why Aristotle cannot be interpreted as an opponent of Plato’s thesis that those who rule should be educated and qualified to do so. In his ideal city, all citizens are educated, and this is regarded as a public responsibility (Politics VIII.1). I take him to mean they are morally educated, so that they become good human beings—individuals who know how to rule well. That is part of what makes his city better than all others. He would not agree that a much more populous city than this would be better ruled, even if its citizens were defective, because it would gain so much from containing a larger multitude that the sacrifice of individual quality would be outweighed. 11   As Landemore points out, Condorcet initially meant his theorem to apply to small assemblies of superior men. ‘[I]t can be dangerous to give a democratic constitution to an unenlightened people. A pure democracy, indeed, would only be appropriate to a people . . . much freer from prejudices than any of those known to history’ (Democratic Reason, p. 151). 12   The Wisdom of Crowds, New York: Anchor Books, 2005.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

92  Richard Kraut competition at a Plymouth local fair in which a crowd bet on the accuracy with which they could guess the weight of an ox selected for display. The crowd was diverse, containing not only butchers and farmers, who had considerable experience estimating the weight of farm animals, but also many others who lacked such expertise. Galton was aware of the similarity between the diversity of such a crowd and the diversity of an electorate in a democracy. As reported by Surowiecki, Galton wrote: ‘the average competitor was probably as well fitted for making a just estimate of the dressed weight of the ox, as an average voter is of judging the merits of most political issues on which he votes’ (xii).13 In a sense, then, the local fair was conceived as a test of the instrumental value of democracy. After the contest was over and the winners received their prizes, Galton obtained all 787 tickets that had been purchased, and on which the estimates of the weight of the ox were recorded. The mean estimate, according to his calculation, was 1,197 pounds— just one pound short of the true weight of the ox (after it had been slaughtered and dressed). Galton wrote (as reported by Surowiecki): ‘The result seems more creditable to the trustworthiness of a democratic judgment than might have been expected’.14 American sociologists and psychologists took a keen interest in this phenomenon and achieved results similar to Galton’s. As Surowiecki notes: ‘A classic demonstration of group intelligence is the jelly-beans-in-the-jar experiment, in which invariably the group’s estimate is superior to the vast majority of the individual guesses. When finance professor Jack Treynor ran the experiment with a jar that held 850 beans, the group estimate was 871. Only one of the 56 people in the class made a better guess’.15 Surowiecki’s book is a defense of the general thesis that it is often better to rely on a crowd (that is, a large group that includes a few experts but many non-experts) than any small group of experts. It is a highly paradoxical idea, because it seem obvious that if one knows who the experts are in a given area, it will be better to follow their advice, if they achieve a consensus, than it would be to abide by some alternative piece of advice that is a mechanical result of combining the views of those same experts with the views of a large number of non-experts. In the examples just cited—Galton’s country fair and Treynor’s jar of jelly beans—getting the correct answer involves much guess work. No one knows the correct answer in advance of measurement or counting the weight of the ox or the number of beans in the jar. The distinction between expert and non-expert either fails entirely (as in the case of the beans) or is not hard and fast (in the case of the ox). Where the notion of expertise is applicable, surely following the consensus of experts (when there are some) is more likely to lead to the truth than is following a recommendation that results from the aggregation of those experts’ views and the views of many people who are far less knowledgeable. In his final chapter, Surowiecki returns to democratic theory and more specifically to the idea, endorsed by some members of the academic world, that the opinions of voters need to be better informed and more thoughtful, and that they should participate   Ibid., p. xii.

13

  Ibid., p. xiii.

14

  Ibid., p. 5.

15

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  93 in face-to-face discussions both with experts and with other ordinary people who have conflicting views.16 Polling and voting should be the outcome of a sound, collective, deliberative procedure rather than the expression of a mere preference or a purely self-interested desire to shape the public will. These ideas express a Platonic point: unless special measures are taken to produce a more enlightened citizenry, democratic politics will simply reflect the strongest force, not the best argument. Surowiecki, however, rejects these ideas because he believes that representative democracies already do a good enough job of finding solutions to political problems. ‘The point of a representative democracy,’ he writes, ‘is that it allows the same kind of cognitive division of labor that operates in the rest of society. Politicians can specialize and acquire the knowledge they need to make informed decisions, and citizens can monitor them to see how those decisions turn out . . . . Competition makes it more likely that politicians will make good decisions by making it more likely that they will be punished when they don’t.’17 Remarkably, Surowiecki here abandons the idea that in politics we should rely on the crowd and ignore the experts. He is saying that in the civic realm the politicians are the experts, and the role of the non-expert voter is not to become better qualified but merely to select among the experts. It is hard to share his confidence that in modern democracies successful candidates for office are likely to be more qualified than those who receive fewer votes, because the citizenry is, more often than not, qualified enough to know who the best candidates are.18

10.  Impersonal Good and Ethical Insight I will make two final remarks before coming to a conclusion. First, something must be said about the notion of the common good, which plays an important role in my reconstruction of Plato’s argument. (Recall that his first premise is: ‘In every political community, the common good requires that certain roles be filled.’) But it might be asked: What is this? Why assume that there is such a thing? We know what we are talking about when we say: ‘this is good for me, or for this group of individuals.’ But is there such a thing as a ‘common,’ and does this mysterious entity   He refers to James Fishkin, Democracy and Deliberation, New Haven: Yale University Press, 1992.   Ibid., p. 266. 18  Landemore, Democratic Reason, provides a more thorough case for the epistemic merits of democratic deliberation and majority rule. She argues that a large and cognitively diverse group is more likely to solve collective problems than a smaller group, even if each member of the smaller group is smarter. See too Josiah Ober, Democracy and Knowledge: Innovation and Learning in Classical Athens, Princeton: Princeton University Press, 2008. He argues that Athens competed successfully against its rivals because it was a participatory democracy. For a number of papers devoted to these issues, see Hélène Landemore and Jon Elster (eds.), Collective Wisdom: Principles and Mechanisms, Cambridge: Cambridge University Press, 2012. I regret that I cannot comment here on the arguments of this literature, but I suspect that defenders of the wisdom of the multitude ignore the problem of widespread ethical deficiencies that distort democratic decisions. 16 17

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

94  Richard Kraut have a good of its own? Why not speak in a less mysterious way, by talking about what is good for some, or for many, or for all? I will set aside the question of what Plato means when he refers to the common good, since my goal is to defend a revised form of his argument rather than to convey his meaning as faithfully as possible. For this purpose, the important point is that there is a way of thinking about ethical and political controversies that can be called impersonal, non-partisan, impartial, or neutral. When one reasons in this way, one does not confine oneself to such questions as: what is best for me, or for some individual to whom I am attached, or some group to which I belong? One asks, in addition: what other reasons apply in these circumstances? And one acts in light of the balance of all of those reasons. In some cases, the weightiest reasons favor acting in a way that serves the good of one’s own group; but in other cases, they do not. There is nothing mysterious about this way of thinking—no weird entity called the ‘common’ whose good must be served. It is not a way of thinking that every major philosopher or philosophical school endorses. Epicurus and his followers hold that we should think solely of our well-being and that of our friends. Nietzsche is hostile to utilitarianism and Kantianism precisely because of their acceptance of impersonal reasoning. But Plato’s notion of the common good—or the use I make of it—is endorsed by an impressive group of political philosophers—including Aristotle, the Stoics, Hobbes, Spinoza, Locke, and Rawls. When citizens think and act in the public arena in purely personal terms, aggregating their separate inputs is very unlikely to result in a wise collective decision. In a conflict between group A and group B that is resolved by majority rule, the larger or better organized or more passionate group is the one that will win, if everyone thinks in partisan terms. Perhaps that resolution of the conflict will also be the one that is best supported by reasons, impartially considered; but if so, that will be a mere accident. It would be far better if all citizens were committed to thinking about such disputes impersonally. But that cannot occur unless it is part of the ethos of the political community that citizens should try to resolve their differences by reasoning in this way. Ethical insight cannot be achieved by a group by aggregating individual preferences. And collective political decisions can go disastrously wrong if they are divorced from such insight. My second point is that if substantive ethical, political, or legal truths are not widely recognized in a political community, because the public is not educated enough to recognize them, that too will lead to poor decisions. Here is an example: most American citizens (including most students beginning law school) believe that the proper role of a judge is simply to find the law that is applicable to the issue brought before them.19 They do not realize that it is inherent in the general scope of the law and the open texture of language that there will be hard cases brought before a judge, and that he or she must therefore engage in a process of thought that goes beyond the mere application of   Scott J. Shapiro, Legality, Cambridge, Mass.: Belknap, 2011, pp. 234–8.

19

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato Against Democracy: A Defense  95 a rule to a case that falls under it. The familiar example used to make this point in jurisprudence is the rule: no vehicles are permitted in the park. Does that mean that ambulances are not permitted? Lawn mowers that must be operated by a rider? To answer those questions well, one must have a plausible judicial philosophy. In democracies, judges are chosen, directly or indirectly, by citizens, but the general public cannot choose them intelligently unless it recognizes this fact about the character of law. The public itself needs a judicial philosophy if it is to make intelligent choices about who should fill judicial offices, and it is unlikely to have one if it has never been trained to think philosophically.

11. Conclusion My conclusion is that Plato presents a formidable challenge to democratic theory. There may be a defense of democracy powerful enough to overcome that challenge, but it is surprisingly difficult to find or construct one. Since we now live in a democratic age, we are likely to take it for granted that it is the best or only legitimate form of government, and that we have nothing to learn from such critics as Plato. We need to shed this complacency. One way to put the Republic to good use is to take its critique of democracy seriously.20

Bibliography Christopher Bobonich, Plato’s Utopia Recast: His Later Ethics and Politics, Oxford: Clarendon Press, 2002. Jason Brennan, The Ethics of Voting, Princeton: Princeton University Press, 2011. Thomas Christiano, The Constitution of Equality, Oxford: Oxford University Press, 2008. Tom Christiano, ‘Democracy’, Stanford Encyclopedia of Philosophy (Spring 2015 Edition), Edward N. Zalta (ed.), section 1. Ronald Dworkin, Justice for Hedgehogs, Cambridge, Mass.: Belknap Press of Harvard University, 2011, chapter 18. David Estlund, Democratic Authority: A Philosophical Framework, Princeton: Princeton University Press, 2008. James Fishkin, Democracy and Deliberation, New Haven: Yale University Press, 1992. Amy Gutmann and Dennis Thompson, Why Deliberative Democracy? Princeton: Princeton University Press, 2004. Terence Irwin, The Development of Ethics, Oxford: Oxford University Press, 2007. Hélène Landemore, Democratic Reason: Politics, Collective Intelligence, and the Rule of the Many, Princeton: Princeton University Press, 2013. Hélène Landemore and Jon Elster (eds.), Collective Wisdom: Principles and Mechanisms, Cambridge: Cambridge University Press, 2012. 20   I am grateful to Seth Mayer, Robert Wallace, and David Brink for their helpful comments on an earlier draft.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

96  Richard Kraut Josiah Ober, Democracy and Knowledge: Innovation and Learning in Classical Athens, Princeton: Princeton University Press, 2008. Thanassis Samaras, Plato on Democracy, New York: Peter Lang, 2002. Scott J. Shapiro, Legality, Cambridge, Mass.: Belknap Press of Harvard University, 2011. James Surowiecki, The Wisdom of Crowds, New York: Anchor Books, 2005. Jeremy Waldron, Law and Disagreement, Clarendon: Oxford University Press, 1999.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

5 Self-Mastery and Self-Rule in Plato’s Laws Susan Sauvé Meyer

I begin with a text that occurs at a pivotal moment in book I of Plato’s Laws.1 The Athenian has just concluded his account of paideia, the process of training citizens to be good (644a7–8)—not good carpenters, or good sailors, or good businesspeople—but good tout court (643d6–644a5). He has in mind the perfect (teleon) citizen, ‘who knows how to rule and be ruled with justice’ (643e5–6). Turning now to the question of what such goodness consists in, he announces that he and his interlocutors have already agreed on a criterion of goodness: ATH: Now, we previously agreed that those who are able to rule themselves (archein hautōn) are good, while those who are unable to do so are bad. CL: Exactly. ATH: Let us return to this phenomenon and clarify further what we take it to be. I hope you will allow me, if I can, to illustrate by means of an image.2 (644b6–c2)

The image he offers is the famous comparison of human beings to divine puppets, pulled by various ‘iron cords’ (pleasure, pain, and their anticipations) and by the ‘golden cord’ of calculation or reasoning (logismos): Consider each of us, living beings that we are, to be a divine puppet—whether constituted as the gods’ plaything, or for a serious purpose, we have no idea. What we do know is that these various experiences in us are like cords or strings that tug at us and oppose each other. They pull against each other towards opposing actions across the field where virtue is marked off from vice. Our account singles out one of these pulls (helxeōn) and says that each of us must 1   It gives me great pleasure to dedicate this essay to my teachers Gail Fine and Terry Irwin, in thanks for their generous support and guidance over many years and in appreciation of the fine example they have set in their own work, both as teachers and as scholars. I would also like to thank David Brink for his careful written comments on an earlier draft of this essay. 2   Translations of the Laws are from Meyer 2015, with occasional modifications. The text translated is that of des Places (1951).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

98  Susan Sauvé Meyer follow it and pull against (anthelkein) the other cords, never loosening our grip on it. This is the sacred and golden guidance (agōgē) of calculation, also called the city’s common law.3 The other pulls are hard as iron, but this one is soft because it is golden . . . . One must always pitch in with the noblest guidance, that of law, since calculation—although it is noble—is gentle rather than violent, so its guidance requires helpers if our golden element is to be victorious (nika(i)) over the others.  (644d7–645b1)

He concludes that the point of this ‘tale of virtue’ is to clarify what is meant by the expressions ‘self-mastery’ (to kreittō heautou) and ‘self-defeat’ (to hēttō heautou): Here is how we may vindicate the tale of virtue (muthos aretēs) that likens us to puppets. It would become clear, in a way, what is meant by ‘self-mastery’ and ‘self-defeat’ (to kreittō heautou kai hēttō einai), as well as the manner in which a city and an individual ought to live.  (645b1–4)

On the most natural reading of this stretch of text (644b6–645b4), the Athenian ­proposes to illuminate the goodness or virtue of the perfect citizen as follows: 1. Virtue is to be understood as self-rule (archein heautou) (previously agreed) 2. Self-rule is to be understood as self-mastery (to kreittō heautou einai) (unstated assumption) 3. Self-mastery is to be understood as the victory of our golden cord over our iron cords. While I think that this is the correct way to read the Athenian’s remarks, it raises several puzzles for the reader of the dialogue. One puzzle concerns the putative agreement about (1). The Athenian states that he and his interlocutors have previously (palai) agreed that ‘those who are able to rule themselves (archein hautōn) are good, while those who are unable to do so are bad’ (644b6–7). But where has this putative agreement taken place? At no point in the prior conversation have the Athenian and his interlocutors agreed to any claim using the expression ‘self-rule’ (archein hautōn).4 Even if we take (2) to indicate that ‘self-rule’ (archein heautou) and ‘self-mastery’ (to kreittō heautou) may be used interchangeably, and we look for places in the prior conversation where the interlocutors have agreed that virtue is to be understood as self-mastery (to kreittō heautou), we will still come up short. While Clinias has emphatically endorsed such a conception of human excellence at 626e2–627a2, the Athenian has subjected it to devastating criticism (627c3–628e1).5

3   According to the alternative translation defended by Schofield 2016: 140–1, by Schofield and Griffith 2016: 59-60, and by Nightingale 1999: 104n12, logismos ‘calls in aid the law common to the city’. 4   At 627e1 the Athenian uses the expression ‘rule themselves’ (archein autous hautōn), but in that ­context, it means political self-determination, not a psychological condition that the puppet image can illuminate. I discuss the psychological use of the expression (which we find at Meno 86d6 and Gorgias 491d4–8) in Meyer 2015: 169–72. 5  626e–627a is generally taken to be the locus of the prior agreement, although Schofield and Griffin 2016: 58 and Schöpsdau 1994: 228 rightly mention 633d–e as well, a passage that I will discuss.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Self-Mastery and Self-Rule in Plato’s Laws  99 A second, and related, puzzle is that self-mastery, as illustrated in the puppet image, involves internal conflict. It consists of reason’s victory in a struggle against the  conflicting pulls of pleasure, pain, and their anticipations. Thus (1), (2), and (3) entail a conception of virtue that involves psychological conflict.6 But the Athenian, in opposing Clinias’ contention that self-mastery is the highest achievement, has rejected the thesis that excellence consists in victory in battle. On the contrary, he maintains, peaceful relations among potentially conflicting elements are far better than victory over internal opposition (628c9–e1). My project in this essay is to argue that these puzzles cannot be resolved in any ­satisfactory way. There is no interpretation of (1) and (3) on which (3) both is a plausible interpretation of the puppet passage and also clarifies a conception of goodness on which the Athenian and his interlocutors have previously agreed. I will argue that there are two competing models of virtue in books 1 and 2 of the Laws. One is what I will call the CONFLICT model, which is introduced by Clinias, criticized and rejected by the Athenian, and captured by the victory of the golden cord in the puppet image. The other is what I will call the HARMONY model. It receives its fullest articulation at the beginning of book 2 where the Athenian describes virtue as the ‘agreement’ (sumphōnia) between one’s reason and one’s pleasures and pains (653a5–c4), but has its seeds in the Athenian’s criticisms of the conflict model early in Book 1, and in the account of paideia that precedes the puppet passage (643b1–644b5). Readers of the dialogue thus must face the question of why the Athenian continues to appeal to the conflict model even though he does not endorse it and has discredited it. I will argue that he does so as part of a dialectical strategy that is defensible in the light of the legislative project of the dialogue, even if it involves arguing from premises that the Athenian does not accept.

Victory, Conflict, and Self-Mastery Self-mastery and self-defeat loom large in the discussion from very early in the Laws. Right after the Athenian and his two interlocutors have agreed to undertake their ­leisurely discussion of ‘constitutions and legislation’ (625a6) Clinias is asked to identify the goal of the institutions characteristic of cities on Crete. The latter’s response presents a very dark assessment of the human condition: [W]hat most people call peace exists only in name, while in fact every city is by nature always in an undeclared war against every other. If you look at it this way, you will discover, I submit, that it is with war in mind that the Cretan lawgiver established all the institutions that govern our public and private life.  (626a2–7) 6   I use the label ‘psychological’ for convenience, as a general characterization of the pleasures, pains, anticipations, and calculations (644c4–d3) that are modelled by the puppet’s iron and gold strings. It is, however, worth noting that neither the term ‘psuchē’ nor its cognates is used in the puppet passage, or in the other passages where self-mastery is invoked.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

100  Susan Sauvé Meyer When questioned by the Athenian on the scope of the war he has in mind, Clinias readily agrees that it is not just cities that are at war with each other, but also  villages, households, and individual people (626c3–13). Indeed, he agrees, even within an ­individual person conflict is pervasive: ‘each person is pitted against himself ’ (626d8–9). Victory in this latter conflict, he claims, is the greatest of all victories: CL: In the latter contest, Stranger, victory over oneself (to nikan auton hauton) is the highest and most excellent of all victories and being defeated by oneself (to hēttasthai auton huph’ heautou) is the most shameful and wretched of defeats. These ways of speaking indicate that each of us is engaged in an internal war against himself.  (626e2–6)

This ‘internal war’ waged by each person ‘against himself ’ is the struggle that will later be illuminated in the figure of the puppets. While Clinias does not here use the locution for self-mastery that we find in the puppet passage (to kreittō heautou)— speaking instead of ‘victory over oneself ’—the Athenian immediately uses the ­language of self-mastery to restate Clinias’ thesis: ‘each of us as a single individual is either master of or defeated by himself ’ (ho men kreittōn hautou, ho de hēttōn, 626e8–627a1). Clinias has put on the table the thesis that self-mastery is the ultimate human achievement.7 Of course, it is the Athenian’s questioning that has led Clinias to think about ­self-mastery. His immediate purpose in doing so, however, is not to illuminate the psychology of self-mastery; this he turns to only much later, when he introduces the figure of the puppets. His strategy at 626e–628d is to broaden the notion of selfmastery beyond its primary application—the case of an individual ‘pitted against himself ’—and extend it to cases where there is conflict within groups, specifically, cases where one faction within a city or a family fights for control against another. While the Athenian is prepared to concede that it may be odd to use the expressions ‘self-mastery’ and ‘self-defeat’ in the case of familial or political faction, his main point is not about ­linguistic usage (627d1–2), but the structure of the underlying phenomenon. A crucial feature of this phenomenon, he and Clinias agree, is that the struggle is between a better party and a worse. Self-mastery is the victory of the better party, while self-defeat is the victory of the inferior party. ATH: . . . I take you to be saying . . . the following. It sometimes happens that a large number of unjust citizens collectively undertake to force their will on the just minority and enslave them—even though all are members of the same tribe and born into the same city. When the 7   One might object that Clinias’ precise claim is only about the greatest victory, leaving open that other achievements may be greater. But the reference back to this passage at 633d5–e2 indicates that the Athenian interprets it as a general criterion of goodness and badness and at his paraphrase of Clinias’ remarks at 627b7–8 glosses ‘master of itself ’ as ‘good’.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Self-Mastery and Self-Rule in Plato’s Laws  101 former prevail you say it is right to call the city ‘defeated by itself ’ (hēttōn hē polis hautēs) and bad, and when they are defeated, ‘master of itself ’ (kreittōn) and good. CL: While this way of speaking is decidedly odd, Stranger, it is undeniable that these things do happen. ATH: Now, let’s stay with that point. Presumably there could be many brothers born to a single man and wife and it would not be unheard of if the majority of them turned out to be unjust and the minority just? CL: Not at all. ATH: We are not seeking to mandate that the household or tribe as a whole be called ‘defeated by itself ’ (hēttōn hautēs) when the wicked parties win, and that it be called ‘master of itself ’ (kreittōn) when the latter are defeated (hēttōmenōn). That wouldn’t be fitting, since our concern is not with the felicity of this popular expression, but with what is correct and mistaken in the nature of laws.  (627b2–d4)

The Athenian’s concluding sentence reminds his interlocutors that their investigation  of self-mastery is part of a broader inquiry into laws, with the background assumption that a law code is correct if it aims at the right objectives. In response to Clinias’ opening proposal that victory in war against other cities is the goal of Cretan legislation (626b5–c2), the Athenian is now embarked upon an extended argument whose conclusion will be that a good legislator must aim at cultivating internal peace and good relations within the city. As the first step in this argument, he invokes the case of faction within a family and asks Clinias which is the best outcome for an adjudicator to aim at (627e–628a): • The good brothers defeat and kill the bad brothers. • The good brothers win, and the bad brothers are made to submit to their rule. • The brothers are reconciled with each other in peace and friendship. While the precise nature of the difference between the second and third options is disputed, it is undeniable that the Athenian takes the third to be better than either the first or the second.8 The best outcome for the family is not to have the conflict decided in favour of the better party (a scenario that would count as ‘self-mastery’ in the extended sense), but to avoid the conflict by cultivating peace and goodwill among the different parties. The Athenian immediately applies this result to the case of a faction-ridden city and asks whether it would be better for the city to face its external enemies with ‘one side destroyed and the other victorious, or with friendship and peace achieved through

8   Against Leo Strauss’ contention that the third option is ranked the worst by the Athenian (Strauss 1975: 5), see Meyer 2015: 91.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

102  Susan Sauvé Meyer reconciliation’ (628b6–c1). The latter is clearly preferable, Clinias agrees (c2–3). On this basis, the Athenian draw the crucial conclusion: The best is neither war nor faction (one should pray to be spared the necessity of either) but rather peace and mutual good will. Victory of a city over itself, it would seem, is not an excellent but a necessary achievement. To think otherwise is like supposing that a disease-ridden body is performing at its best after being flushed out by a purgative—with no thought to the case of a body that needs no such treatment. For the same reason, no proper statesman will assess the happiness of either a city or an individual solely and primarily with a view to war against external enemies, and no lawgiver is any good unless he regulates military matters for the sake of peace, rather than regulating peacetime for the sake of war.  (628c9–e1)

He drives home his point by comparing the strife-ridden city to a disease-ridden body, and the city where one faction is victorious over the others to the sick body that has been ‘flushed out by a purgative’. Far better than the latter is a body that has no need of the purgative, and the same goes for a city. Thus, a proper legislator will aim to cultivate peace and mutual goodwill in the city. At this point, readers of the Republic might expect the Athenian to make the analogous point about victory over oneself in the personal case. In the Republic, justice is presented as a condition of harmony between the various elements in the soul, analogous both to a faction-free city (441e–444b) and to health in the body (444c–445b). That is, we might expect the Athenian to say to Clinias something like, ‘A person who has defeated a rebellion by the inferior elements within him is as depleted as a sick body flushed out by a purgative, or a city that has just emerged from a civil war. Far better is the person who experiences no internal insurrection in the first place’. The Athenian, however, fails to draw such an inference, at least not explicitly. He does indicate that his conclusion applies to an individual as well as a city (kai pros poleōs . . . ē kai idiōtou 628d4–5), but he fails to state explicitly that this means the individual must have ­internally harmonious relations. Instead, Plato keeps the conversation focused on the case of the city, by having Clinias respond with the observation that his city and Sparta do seem to be organized with a view to war against external enemies (628e2–5). The Athenian will answer this concern with a discussion of the Spartan poet Tyrtaeus and a lengthy disquisition on the virtues that citizens must have in order for the city to avoid faction (629a1–630d1). While none of the interlocutors denies that virtue requires internal harmony, they do veer away from the subject. Although the Athenian has the opportunity here at 628e1 to affirm the HARMONY model for virtue, he conspicuously fails to avail himself of the opportunity. The Athenian’s failure here to draw any conclusions about psychological self-mastery is in keeping with his apparent lack of interest in the phenomenon so far in the dialogue. Immediately after he draws Clinias’ attention to the phenomenon of an individual at war with himself, the Athenian immediately turns to apply the notion ‘back to the previous cases’ (626e7), that is, to the familial and political versions of self-mastery. One might suspect that he has led Clinias to the topic of personal self-mastery simply

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Self-Mastery and Self-Rule in Plato’s Laws  103 as a device to introduce the topic of political strife. The latter topic is important for the Athenian’s political argument, since his major contention at this stage in the discussion is that the legislator’s priority must be to avoid political faction. Having devoted no consideration to the sort of internal complexity involved in cases where an individual person is pitted against himself, one might suppose that the interlocutors are not in a position, at 628e1, to articulate the conception of internal psychological harmony that is analogous to the peaceful relations in the factionfree city described in 628c9–e1. Indeed, it is only much later, in the puppet passage (644d7–645b1), that the Athenian considers the internal complexity involved in cases of individual self-mastery and self-defeat. The iron cords of pleasure, pain, and their anticipations and the golden cord of logismos are the parties to the conflict in the ­person previously described as ‘pitted against himself ’ (626d8), and self-mastery is analysed as the victory of the golden over the iron cords. With this understanding in hand of the internal complexity involved in psychological self-mastery, the Athenian is in a position to make the point he failed to make at 628e, that a condition even better than self-mastery is agreement between the golden and iron cords. But the Athenian makes no such claim about the puppets. As I previously argued, the most straightforward reading of the puppet passage is that the Athenian is  construing ­virtue as self-mastery (the victory of the golden cord over the iron cords), with no mention that a better condition would involve agreement between these psychological forces. Has he forgotten his prior remarks that victory over d ­ issenting opponents is a poor second (a necessity rather than an excellence, as he put it at 628d1), while the best condition will involve peaceful internal relations? In what follows, I will refer to these alternative models of excellence as the CONFLICT model and the HARMONY model.

Does Virtue Allow for Conflict? Might we avoid these difficulties by supposing that in the earlier passage, where the Athenian rejects the political version of the CONFLICT model (628c9–e1), his failure to draw the analogous conclusion about an individual person indicates that he does not accept the conclusion in the individual case? Perhaps he agrees with Clinias that victory over oneself (in the psychological sense) really is what goodness (human excellence) consists in. In that case, virtue as the Athenian conceives it does allow for internal conflict. If this were his position, then it would not be misleading for him to claim at 644b6–7 that the figure of the puppets illuminates a conception of goodness to  which he and his interlocutors have already agreed. While he didn’t explicitly endorse the CONFLICT model for personal virtue at 628e, he left it unchallenged. Some interpreters have indeed proposed that, for the Athenian, a virtuous person may still experience psychological conflict. For example, Elizabeth Belfiore has proposed that, in contrast to the Republic, the virtue of temperance (sōphrosunē) in the Laws involves a balance of opposing powers, in which there is a ‘constantly renewed

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

104  Susan Sauvé Meyer combat against the anti-rational’ (Belfiore 1986: 428–33, 429). Christopher Bobonich finds that, on the Athenian’s view, ‘a courageous person may not be entirely free of misdirected fears and desires . . . but is able to resist them’ (Bobonich  2002: 289; cf. 350, 546n122). However, it is implausible to suppose that this is the Athenian’s view tout court, since at the beginning of book II (653a–c) he outlines in great detail a ­conception of virtue on which there is agreement (sumphōnia) between reason on the one hand and one’s pleasures and pains, loves, and hates, etc., on the other hand. This is exactly the model of individual virtue (the HARMONY model) that we would expect him to endorse, given that he has criticized the CONFLICT model and endorsed the HARMONY model in the political case—criticisms that he articulates before the puppet passage. Moreover, the account of paideia at 643b–644b, which immediately precedes the puppet passage, speaks of shaping the ‘pleasures and desires’ of citizens (643c7–8; cf. 643e5), so it too fits the HARMONY model rather than the CONFLICT model. To be sure, the Athenian can grant, in cases of conflict between the ‘golden’ and ‘iron’ elements in a person’s psyche, that it is better for the rational element to win the struggle than to be defeated. Along these lines one might argue that even if the best psychic condition is one of harmony, victory of the sort modelled in the puppet passage will still be good and admirable.9 Since the prior agreement that the Athenian proposes to elucidate in terms of the puppet image does not explicitly invoke virtue (aretē), but instead makes claims about what a good person must be able to do (644b6–7), one might propose to resolve the puzzle by taking the self-mastery illuminated by the ­puppet passage to model a condition that is good, even if it is not virtue. However, this is not a plausible solution to our interpretive puzzle. First of all, the Athenian describes the puppet image as a ‘tale of virtue’ (muthos aretēs, 645b2)—so we are entitled to construe it as proposing a conception of excellence, not a lesser degree of goodness. Second, the criticisms that the Athenian has explicitly levelled against the CONFLICT model and in favour of the HARMONY model in the political case do not treat victory over internal opposition as something good and desirable, even if second best. Rather, he concludes that victory in such cases is ‘a necessity’ (628d1), that is, a drastic measure called for in a dire situation. Victory in such situations is choice worthy only in the way a purgative is. If we take our cue from the Athenian’s assessment of political self-mastery, then we should see the person who is victorious over conflicting internal impulses as debilitated and weakened—hardly in a position to ­perform well in the challenges he faces. This is a much less positive assessment of the CONFLICT model than we find in the puppet passage. So, we cannot dissolve our puzzle along these lines. 9   David Riesbeck has argued (in private correspondence elaborating upon his remarks in Riesbeck 2016) that even if virtue is a condition of HARMONY, it must be cultivated in a process whose earlier stages will involve battling against conflicting inclinations. So even if the Athenian does not take self-mastery to be virtue, he still takes it to be an important achievement.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Self-Mastery and Self-Rule in Plato’s Laws  105

Does Reason’s ‘Victory’ Eliminate Internal Conflict? An alternative strategy for dissolving the puzzle is to deny that the victory of the golden over the iron strings in the puppet is an instance of the CONFLICT model. On this reading, we should accept that the Athenian and his interlocutors agree that the superiority of the HARMONY model to the CONFLICT model applies both to the political and the personal case, even though they do not explicitly draw the latter ­conclusion at 628c9–e1. We may accordingly suppose that this tacit endorsement is the ‘previous agreement’ about goodness referred to at 644b6–7. With these presuppositions, we should expect that the puppet image expresses (or is at least consistent with) the HARMONY model. It has become popular in recent years to interpret the puppet image along these lines (even if not explicitly for these reasons). For example, Emmanuelle Jouët-Pastré construes the self-mastery recommended by the Athenian as a ‘harmonization’ of the puppet’s movements.10 Dorothea Frede proposes that only the iron strings pull against each other, while reasoning exerts its control over them by giving them their content and direction (Frede 2010: 217–20). Julia Annas proposes that the golden string does not oppose the iron strings with brute force, but rather manipulates them, shaping them and giving them direction (Annas 1999: 142–4).11 Joshua Wilburn has denied that logismos pulls the agent towards specific actions (Wilburn 2012: 32–3). Most recently, Malcolm Schofield has proposed, against my own construal of the golden cord as pulling against the iron strings (Meyer 2015: 181–2), that the moral of the puppet passage is that logismos is ‘not a force pulling at us like pleasure and pain’; victory over oneself is a matter of a person ‘translat[ing] into action the deliberations he has himself engaged in’ (Schofield 2016: 146, 148).12 However, these interpretations are hardly the most straightforward reading of our passage, which refers to all of pleasure, pain, anticipations, and logismos as ‘pulls’ (helxeōn, e5). It says they ‘draw against each other (anthelkousin) towards opposing actions’ (644e3; the verb is repeated at 644e6). While reason’s pull is also labelled ‘guidance’ (agogē) at 645a1, 7, which might suggest it is something other than a pull,13 Plato elsewhere uses agogē to refer to the non-rational pull of appetites and emotions.14 Granted, he does here characterize reason’s pull as different from that of the iron strings. It is ‘gentle’ rather than ‘violent’, and consequently needs ‘helpers’ if it is to be 10   ‘L’homme doit donc sans cesse essayer de devenir maître de lui-même en harmonisant les mouvements de la marionnette qu’il est. La victoire sur soi est l’établissement d’une harmonie et d’un accord entre ses affects et le logismos. (Jouët-Pastré 2006: 42). 11   Terence Irwin’s description of the golden cord as ‘controlling [a person’s] feelings of pleasure and pain’ (Irwin 1995: 351) also seems to adopt this line of interpretation. 12   Still, Schofield appears to allow that this involves resisting the pull of pleasures and pains, and hence does not after all do away with the conflict model (Schofield 2016: 148). His main contention, in any case, is that the puppet passage invites the careful reader to identify the verdict of logismos as the activity of the self. 13   Indeed, England  1921: 256 construes it as education, invoking 659d2 and 819a5; against this see Meyer 2015: 181, and on the iron strings more generally see Meyer 2012. 14   Republic 604b1; Phaedrus 238c3.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

106  Susan Sauvé Meyer ‘victorious’ over the others (645a6–b1). But in identifying victory as the objective he is invoking the CONFLICT model. Indeed, the very necessity of ‘helpers’ underscores the fact that logismos is the weaker party in a struggle. I think that the ‘harmonizing’ interpretations of the puppet image that I have just mentioned are most plausible as proposals about what the Athenian really thinks about virtue, the latter being the HARMONY model that he articulates explicitly at 653a–c. I am also prepared to concede that the further comments the Athenian offers in elucidation of the morals to be drawn from the puppet passage are compatible with the HARMONY model. Here is the fuller statement of those lessons (including the opening lines, 645b1–2, previously quoted): Here is how we may vindicate the tale of virtue (muthos aretēs) that likens us to puppets. It would become clear, in a way, what is meant by ‘self-mastery’ and ‘self-defeat’ (to kreittō heautou kai hēttō einai), as well as the manner in which a city and an individual ought to live. A person must grasp within himself the true account concerning these pulls and live in accordance with it, while a city, having received this account from a god or from the person who understands these matters, must establish it as law and conduct its internal and external affairs accordingly. In addition, the tale gives us a more lucid articulation of virtue and vice, and greater clarity here will perhaps shed light on education and various other practices, particularly the drinking party. (645b1–c4)

The ‘true account concerning these pulls’ (b4–5) seems to capture the description of logismos as something that evaluates the iron strings (644d1–2). The life in ‘accordance with’ such an account invoked in the present passage need not be a life in which the golden cord wins a struggle against the iron cords. It could equally characterize a life in which there is no struggle between the cords, or even one in which (on the interpretations of Frede and Annas) the golden cord informs and gives content to the iron cords.  As such, it would be consistent with the HARMONY model of virtue that the Athenian will articulate explicitly at the beginning of book II (653a–c), and for which he has prepared the theoretical grounding by rejecting the CONFLICT model advocated by Clinias earlier in book I. However, I still deny that the HARMONY model captures or is consistent with the image of the puppets (644d7–645b1),15 for the Athenian portrays the puppets as subject to internal conflict and advocates victory as the solution to that conflict. One might object that interpretive charity requires us to seek an interpretation of the puppet image that accords with the HARMONY model, since the Athenian has already laid the groundwork for that model and will soon explicitly affirm it. However, even if we interpret the puppet image as illustrating HARMONY rather than CONFLICT, we will still be left with the first of the two puzzles identified at the beginning of this essay. That puzzle concerns the Athenian’s assertion that he and his interlocutors have 15   It is important to distinguish between the image itself (644d7–645b1, quoted in full at the opening of my essay) from the psychological preamble (644c3–d6) and from the subsequent commentary that draws morals from the image (645b1–c6).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Self-Mastery and Self-Rule in Plato’s Laws  107 previously agreed that a good person is able to ‘rule himself ’ (644b6–7). This criterion of goodness is what the puppets image is supposed to clarify (644b9–c2), and if we are to interpret that image as illuminating the HARMONY model for virtue, then the Athenian must mean at 644b6–7 that he and his interlocutors have previously endorsed the HARMONY model. If the implication is correct, then Clinias, who has grudgingly16 conceded that the HARMONY model is superior to the CONFLICT model in the ­political case, has tacitly accepted the same point about the individual case. But once again, we must ask where this tacit agreement has taken place. Presumably it would have occurred by 628e5, which marks the end of the critique of the CONFLICT model. There is, however, ample textual evidence that well beyond this point in the conversation, Clinias and Megillus continue to endorse the CONFLICT model for persons, and the Athenian continues to offer it for their approval.

The Persistence of the CONFLICT Model When Megillus is asked to enumerate the institutions that train Spartan citizens in courage, he invokes ‘a large set of activities that involve endurance,’ including ‘boxing matches and the raids that regularly involve a severe beating, . . . going barefoot and sleeping without bedding in the winter’ and competing in the ‘naked games’ that take place ‘in the worst heat of summer’ (633b5–c7). The Athenian responds by making explicit Megillus’ assumption that courage involves struggle, and asks him whether the struggle is only against pains, but also against pleasures: ATH: . . . Let’s consider what we take courage to be. Does it amount quite simply to a battle against fears and pains alone? Or does it also oppose yearning and pleasures . . . ? MEG: That’s what I think. It opposes all these things. ATH: Now, if I recall our previous discussion correctly, our friend here spoke of a city being ‘defeated by itself ’ and likewise a man. Isn’t that so, Stranger from Knossos? CL: Quite true. ATH: So now, is it the person defeated by pains whom we call bad, or also the one defeated by pleasures? CL: Even more so, I think, in the case of pleasures . . . .  (633c8–e3)

The Athenian is clearly invoking a CONFLICT model for virtue. The ‘previous discussion’ to which he refers, in which Clinias ‘spoke of a city being “defeated by itself ” and likewise a man’ must be 626d–e, where Clinias first invokes self-mastery. Most interesting for our present purposes, the Athenian indicates here that he and his interlocutors agree with Clinias’ proposal. (Note the first-person plural: ‘the person whom   ‘While (men) your argument appears sound in a way (pōs), Stranger . . . ’ (628e2–3).

16

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

108  Susan Sauvé Meyer we call bad’—633e1.) The present passage, then, must underwrite the Athenian’s later claim, at 644b6–7, that he and his interlocutors have already agreed that ‘those who are able to rule themselves are good, while those who are unable to do so are bad.’17 But in that case, the prior agreement was to the CONFLICT model. Since the puppet image is offered to elucidate the phenomena invoked in that prior agreement, we should expect the image to model an individual person’s victory over himself (self-mastery in the personal sense). And this, I have argued, is the most natural way to read the image.

Conclusion I conclude that there is no satisfactory resolution of the interpretive puzzles with which we began. There is no interpretation of the puppet passage on which it both captures the HARMONY model advocated by the Athenian and illuminates a conception of virtue that both the Athenian and his interlocutors endorse. Rather, we have seen that the Athenian persistently encourages his interlocutors to think of virtue in terms of the CONFLICT model long after he has established the HARMONY model’s superiority.18 Why does Plato have the Athenian proceed in this puzzling manner? No doubt the CONFLICT model for virtue is attractive to Clinias and Megillus, since it resonates with their militaristic ethos, and it is hard to deny that the Athenian finds that model deeply misguided. Nonetheless, the difference between the CONFLICT model and the HARMONY model turns out to be unimportant for the purposes of establishing the main results of the legislative inquiry in book I. We may summarize those results as follows. A legislator must aim not simply at making a city victorious in war against its enemies, for which it must cultivate courage in its citizens. His more important aim is to keep the city free from faction. To that end, the city must cultivate the virtues of ­justice and temperance in its citizens, not just courage. To cultivate this broader palette of virtues, a society must have a system of education that addresses people’s pleasures and desires, not just their pains and fears. Whether one conceives of the virtues to be cultivated by such an education in terms of the CONFLICT model or in terms of the HARMONY model, one has reason to endorse a system of education that pays attention to all the iron strings, not just to pain and its anticipation. Indeed, one can make the case that the self-mastery captured by the CONFLICT model is a developmental stage on the way to the full virtue captured by the HARMONY model—that repeated success at resisting the pull of opposing desires and fears will result in retraining those desires and fears so that they agree with rather than oppose the golden cord.19 If this is the case, then one can justify, to proponents of the CONFLICT model, educational institutions required by the HARMONY model.   Schofield also notes this connection (Schofield and Griffith 2016: 58n13).   Indeed, he continues to invoke the struggle against pleasure, pains, desires, and fears in book I even after the puppet passage. See, for example, 647a4–6, 647c7–d8, 648c8–e5. 19   Aristotle’s thesis that virtue of character is formed by habituation would be an instance of such a view (Nicomachean Ethics 2.1–4). 17 18

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Self-Mastery and Self-Rule in Plato’s Laws  109 This, I propose, is what the Athenian is doing in book I of the Laws when he appeals to  normative beliefs that are deeply held by his interlocutors but that he h ­ imself does not accept.

References Annas, J. 1999. Platonic Ethics Old and New. Ithaca: Cornell University Press. Belfiore, E. 1986. ‘Wine and the Catharsis of Emotions in Plato’s Laws’. Classical Quarterly 36 (2): 421–37. Bobonich, C. 2002. Plato’s Utopia Recast. New York: Oxford University Press. des Places, E. (ed.) 1951. Platon: Oeuvres Complètes: Les Lois, Livres I et II. Paris: Les Belles Lettres. England, E. B. 1921. The Laws of Plato. Text edited with introduction and notes. 2 vols. London: Longman, Greene, & Company. Frede, D. 2010. ‘Puppets on Strings’, in C. Bobonich, ed. Plato’s Laws: a critical guide. New York: Cambridge University Press, pp. 108–26. Irwin, T. H. 1995. Plato’s Ethics. Oxford: Oxford University Press. Jouët-Pastré, E. 2006. Le jeu et le sérieux dans les Lois de Platon. Sankt Augustin: Academia Verlag. Meyer, S. S. 2012. ‘Pleasure, Pain, and “Anticipation” in Plato’s Laws, Book I’, in R. Patterson, V. Karasmanis and A. Hermann, eds. Presocratics and Plato Las Vegas: Parmenides Publishing, pp. 349–66. Meyer, S. S. 2015. Plato: Laws 1 & 2, translated with introduction and commentary. Oxford: Clarendon Press. Nightingale, A. 1999. ‘Plato’s Law Code in Context: Rule by Written Law in Athens and Magnesia’. Classical Quarterly 49 (1): 100–22. Riesbeck, D. 2016. Review of Meyer 2015. Notre Dame Philosophical Reviews. 24 May. Schofield, M. 2016. ‘Plato’s Marionette’. Rhizomata 4 (2): 128–53. Schofield, M., and Griffith, T. 2016. Plato: The Laws, ed. M. Schofield and translated by T. Griffin. Cambridge: Cambridge University Press. Schöpsdau, K. 1994. Platon: Nomoi (Gesetze): Buch I–III. Übersetzung und Kommentar. Göttingen: Vandenhoek und Ruprecht. Strauss, L. 1975. The Argument and the Action of Plato’s Laws. Chicago: University of Chicago Press. Wilburn, J. 2012. ‘Akrasia and Self-Rule in Plato’s Laws’. Oxford Studies in Ancient Philosophy (43): 25–53.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

6 Plato’s Philebus and the Value of Idle Pleasure Verity Harte

I.  Idle Pleasure Early in his Methods of Ethics Sidgwick quotes with approval William Lecky’s claim that ‘the pleasure of virtue can be obtained only on the express condition of its not being sought’.*,1 This makes the pleasure of virtue idle in the sense in which I will understand it in this essay: a pleasure that is not sought, that falls outside of the motivational framework of the activity in which it arises at least in the sense that pleasure is not adopted as the intended goal of the activity.2 Such pleasures bear on Sidgwick’s well-known paradox of hedonism: 1. Here comes into view what we may call the fundamental paradox of Hedonism, that the impulse towards pleasure, if too predominant, defeats its own aim. This effect is not visible, or at any rate is scarcely visible, in the case of passive sensual pleasures. But of our active enjoyments generally, whether the activities on which they attend are classed as ‘bodily’ or as ‘intellectual’ (as well as of many emotional pleasures), it may certainly be said that we cannot attain them, at least in their highest degree, so long as we keep our main conscious aim concentrated upon them.  (Methods of Ethics I.4, pp. 48–9)

The existence of this paradox is not in itself an objection to hedonism. (That it is not is suggested by its being a paradox of hedonism.) The problem it generates is, as *  It is a true pleasure to contribute this essay to a volume in honour of Gail Fine and Terry Irwin. No one who came of scholarly age, as I did, in the early 1990s could fail to have learnt much from both their work. But I also value the friendship born during a semester-long visit at Cornell in Spring 2003, where they were warm and generous hosts, and cemented through subsequent years’ discussion of ancient philosophical matters as well as a shared enthusiasm for travel, theatre, and art. 1  Sidgwick The Methods of Ethics I.4, 50, citing W.E.H. Lecky The History of European Morals from Augustus to Charlemagne Part I, New York: D. Appleton & co. 1869, 36. 2   As Frances Kamm points out to me, pleasure might motivate an activity by being amongst the conditions under which the activity is undertaken. For relevant distinctions, see Frances Kamm ‘The Doctrine of Triple Effect and Why a Rational Agent Need not Intend the Means to his End’, Proceedings of the Aristotelian Society, Supplementary volume 74, 2000, 21–39. Much more needs to be said about all this. I will try to make things at least somewhat clearer in what follows.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  111 Peter Railton has said, ‘pragmatic’: ‘the hedonist . . . ought not to be a hedonist’.3 This problem is most acute for the position Railton calls ‘subjective hedonism’: ‘the view that one should adopt the hedonistic point of view in action, that is, that one should whenever possible attempt to determine which act seems most likely to contribute optimally to one’s happiness, and behave accordingly’.4 My interest in the paradox is in connection with pleasure’s role in the practical guidance of action, in how we should think about the pursuit of pleasure, both what it is and how it should be evaluated. In Plato’s Philebus, I shall argue, Socrates seeks to defend a thesis in some ways stronger and stranger than Sidgwick’s paradox. He identifies certain true, pure ­pleasures as the only pleasures that make a contribution to the goodness of the best human life. These pleasures turn out to be idle in the relevant sense. Further, their idleness, while not constitutive of their goodness, is a prerequisite of it. I shall reflect on why Socrates should characterize the pleasures he regards as good as idle in this way in the context of the dialogue’s broader encounter with hedonism, and what this characterization suggests about his view of the relation between pleasure, motivation, and goodness in practical reasoning. Socrates develops this view in conversation with a specific sort of hedonist ­opponent  and in a particular framework. That framework is eudaimonist, where eudaimonism may be understood as ‘the view that each person should seek his own good or happiness’.5 And it involves a broadly internalist assumption that a person’s own good, when recognized, will be practically motivating. The Philebus states such a broadly internalist assumption explicitly: 2. SOC:  This, as I suppose, [we shall agree] to be most necessary regarding it [the (human) good], that everything that recognizes it hunts for and aims at it, wanting to get hold of and take possession of it, and cares nothing for other things except what come to an end simultaneous with goods.  (Phlb. 20d7–10)6

Here is where a difficulty arises. It is a difficulty connected to understanding the good as something at which we aim.7 Does this mean merely that the good is an evaluative criterion in light of which I assess my actions? Or does the good also shape my actions practically, so that each act is self-consciously chosen with it in view?8 In 2, in his talk 3   P. Railton ‘Alienation, Consequentialism and the Demands of Morality’, Philosophy and Public Affairs 13.2 1984, 134–71, at 141. 4   ‘Alienation’ 142–3. Railton appears to want to argue that the paradox is nonetheless at least destabilizing for other, more sophisticated brands of hedonism. 5  So, for example, S. Darwall Philosophical Ethics, Dimensions of Philosophy Series, Boulder, CO: Westview, Glossary, 235. 6   Translations of the Philebus are my own following Burnet’s OCT (I. Burnet Platonis Opera II, Oxford: Clarendon 1901) unless otherwise noted. 7   A parallel difficulty arises for Lecky’s talk of the ‘object sought’. 8   Compare Sidgwick’s talk of ‘our main conscious aim’ in passage 1. The distinction I draw relates to and is indebted to Velleman’s distinction between two different senses of ‘a person’s good’ (‘Is Motivation Internal to Value?’, in J. David Velleman The Possibility of Practical Reason Oxford: Oxford University Press 2000, 85–98): (1) the person’s well-being; (2) the person’s practical good. Velleman uses this distinction to

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

112  Verity Harte of hunting and aiming, Socrates seems to me to be urging the latter position: the good is to be, when recognized, the practical guide of our actions. But the paradox of ­hedonism might be thought to give a hedonist opponent reason to reject this kind of internalism.9 Nevertheless, in context, Socrates’ hedonist opponent does not deny the internalist thesis. This is in part because Socrates’ hedonist opponent seems in thrall to a pattern of argument that appears to have been common in Antiquity.10 This is the pattern of argument according to which the purported fact that we (people in general) or human infants or animals seek pleasure is offered as evidence that pleasure is in fact the human good. This argument evidently makes claims about the motivational structure of ­practical behaviour: pleasure is the practical object of pursuit. It is against such a hedonist that Socrates sets his sights in the Philebus. But I hope that his position may at the same time raise general questions of interest about the way in which pleasure is and is not an explanatorily final practical goal of action.11 By ‘explanatorily final’, I mean the view that the pursuit of pleasure may in some sense put an end to all further questions about one’s reason for acting.12

II. The Philebus Since my focus is on a relatively obscure part of a relatively obscure Platonic work, some introduction is needed. Plato’s Philebus is a late-written dialogue addressed to the question of what is the best human life. It is less well known than other Platonic works addressed to similar topics. Given its topic, it might seem surprising that it is not more read outside of specialists. Actually reading the dialogue goes some way towards explaining this relative neglect. Its discussion of its reader-friendly topic is burdened by digressions into the arcane world of late Platonic metaphysics and the dialogue’s overall structure is very obscure. This is a pity, for the dialogue is philosophically very rewarding. In its historical context, it was enormously influential. The Philebus is at put pressure on any especially strong form of internalism about the good, which is where, if anywhere, he thinks, internalism has some foundation. 9   Crisp, in his recent defence of hedonism, adopts externalism about reasons: R. Crisp Reasons and the Good Oxford University Press 2006, Ch. 2. 10   For thoughtful discussion of the version of this style of argument that appeals to the pleasure-seeking behaviour of human infants and young animals, see J. Brunschwig ‘The Cradle Argument in Epicureanism and Stoicism’, The Norms of Nature, ed. M. Schofield and G. Striker, Cambridge University Press 1986, 113–45. 11   For general discussion that I have found profitable to consider in this connection, see E. Millgram ‘Pleasure in Practical Reasoning’, The Monist 76.3, 1993, 394–415, which I return to briefly below. The article is cited in connection with the Philebus by M. Evans ‘Plato’s Anti-Hedonism’ in Proceedings of the Boston Area Colloquium in Ancient Philosophy XXIII, 2007, ed. J.J. Cleary and G.M. Gurtler S.J. (Brill, 2008), 121–45, at 141–2 with n28. 12   Compare Anscombe’s famous remark that ‘the chain of “Why’s” comes to an end’ with the answer ‘It’s  pleasant’ (G.E.M. Anscombe Intention §40 Second Edition, Cambridge: Harvard University Press 1963, 78). I return to Anscombe in §VI.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  113 least as important a precursor to Aristotle’s Nicomachean Ethics as the Republic, for example.13 Indeed, the Philebus comes closer than any other Platonic work to the ­precise framework of the Nicomachean Ethics. Like the Ethics, its central question is identified both as the question of what is the human good and as the more specific question of what state or disposition of soul is capable of providing the eudaimōn (happy) life (Phlb. 11d4–6). The Philebus’s investigation of the human good is structured around a contest between two candidates: pleasure and knowledge. Each in fact represents a family of candidates. This contest, of course, is familiar from elsewhere in Plato. Spoiler alert! Pleasure will lose. But knowledge loses too. At least the dialogue also denies ­knowledge the status of being the good, but it is, so the dialogue will argue, more akin to the good than pleasure. Why, however, is it that the question of the human good is framed as a contest between pleasure and knowledge? One reason, and the reason most germane to my present interests, is that each can be regarded as something that is chosen for its own sake and for whose sake other things are chosen. Each, that is, can be regarded as a goal or end. That pleasure was regarded as an end in Plato’s environment, indeed as a candidate for being the ultimate or final such end for the sake of which all other things are ­chosen,  is supported by Aristotle’s report of Eudoxus in Nicomachean Ethics X.2. Eudoxus—a highly regarded mathematician—was an associate of Plato.14 According to Aristotle, Eudoxus was tempted by a version of the argument for hedonism I mentioned: 3. Eudoxus thought pleasure was the good because he saw all things, whether rational or not, aim at it. And in everything, he says, what is worthy of choice is good, and what is most worthy of choice is best; thus the fact that everything is borne towards the same thing shows that this is what is best for all, since each thing finds its own good, as it finds its own food; and that which is good for all things and at which all aim is the good.  (EN X.2, 1172b9–15, tr. Crisp)

Knowledge, by contrast, may seem a less natural candidate for something chosen for its own sake. One paradigmatic model for valuing knowledge is instrumental: ­knowledge is good because of the other good things it brings about; absent this, knowledge would not be valuable. (Academics, particularly in the humanities, are surely all too familiar with this way of valuing knowledge.) Nevertheless, Aristotle, at least, 13   The influence of the Philebus on Aristotle’s ethical thought is rightly emphasized in a brief, but illuminating discussion of the Philebus in Terence Irwin Plato’s Ethics, Oxford University Press 1995, 318–38. 14   Eudoxan hedonism has sometimes been taken as the target of or at least the motivation for the Philebus’s argument against hedonism. See, for example, J.C.B. Gosling Plato Philebus Oxford: Clarendon 1975, 139–42. This historical claim is not something I intend to evaluate. For general doubts about the Eudoxan background to the Philebus, see G.E.R. Lloyd, ‘Review of J.C.B. Gosling Philebus’, The Classical Review 27, 1977, 173–5 and R.A.H. Waterfield ‘The Place of the Philebus in Plato’s Dialogues’, Phronesis 25.3 1980, 270–305, at 273.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

114  Verity Harte thought of knowledge as something valued and chosen for its own sake, famously beginning his Metaphysics with the claim that all human beings desire knowledge, and do so for its own sake, aside from any utility it may have.15 4. All men by nature desire to know. An indication of this is the delight we take in our senses; for even apart from their usefulness they are loved for themselves; and above all others the sense of sight. For not only with a view to action, but even when we are not going to do ­anything, we prefer sight to almost everything else. The reason is that this, most of all of the senses, makes us know and brings to light many differences between things. (Metaph. I.1, 980a21–7, tr. Ross)

One way to understand the overall structure of the Philebus—as I see it—is as one long argument that, though neither pleasure nor knowledge is the good, and despite the fact that certain pleasures—like knowledge—are good, it is not, as the opponent thinks, pursuit of pleasure, but pursuit of knowledge that offers a practical guide to the human good. My focus here will be the Philebus’s argument that it is not through pursuit of pleasure that we secure the good. More particularly, I shall be focused on the contrasting pictures it develops. On one side, there is Socrates’ hedonist opponent, who takes practical pursuit of pleasure to show that pleasure is the good. On the other side, there is Socrates who denies that pleasure is the good and that pleasure is a practical guide to the good and who accordingly offers a different account of the practical motivation of the narrow class of pleasures that he allows do nevertheless make a valuable contribution to the best human life.

III.  Appetite and Pleasure The hedonist position of the Philebus is ascribed to the eponymous Philebus. But after Philebus begs off, its defence is left to the interlocutor Protarchus. The hedonist ­position he is to defend is Eudoxan in character, whether or not as the result of actual historical influence. It holds that the fact that we and other animals do as a matter of fact seek pleasure is evidence both that pleasure is the good and that to get the good, to be happy, we should pursue (maximal) pleasure. The best evidence for this being the hedonist’s position comes close to the end of the dialogue, when Socrates concludes: 5. SOC: But [pleasure does] not [gain] first [prize, i.e. is not the good], not even if all the cattle and horses and all the other beasts should declare it so by pursuing enjoyment, trusting to whom—as the seers trust birds—the majority of people judge pleasures to be mightiest as regards our living well and regard the desires (erōtas) of beasts as authoritative witnesses rather

15   In Cicero De Finibus III.17, as Susan Sauvé Meyer points out to me (pers. comm.), Cato, the Stoic speaker, makes a comparable point, asserting the value of cognitions (cognitiones) and of the sciences (artes), and giving as evidence the delight young people take in working something out for themselves, even when no other interest is furthered.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  115 than the desires in16 arguments (tous [erōtas]17 tōn . . . logōn) repeatedly testified to in the p ­ hilosophical muse.  (Phlb. 67b1–7)

The conclusion that pleasure is not the good, reiterated here, was in fact arrived at early in the dialogue. Socrates argues that neither pleasure nor knowledge can be the good since no one would choose a life that consisted solely in pleasure without any of Socrates’ rival family of candidates (knowledge, judgment, or memory) and no one would choose a life that consisted solely in knowledge without any pleasure (Phlb. 20c–22c). This earlier argument may not seem decisive and the dialogue does not, I think, regard it as decisive, since a second independent argument against the claim of pleasure is offered later on (Phlb. 53c–55c). Nevertheless, this early agreement that the good is to be found in a life that contains some mixture of pleasure and knowledge provides the framework for the rest of the dialogue and it inaugurates a second contest between pleasure and knowledge. This second contest is framed in two different ways: first as the question which of pleasure or knowledge is more akin to the good (which shares the qualities the good has);18 and second as the question which of pleasure or knowledge is causally responsible for the good of the mixed life.19 Here I am interested in a version of the second question: Which of pleasure and knowledge is practically effective in achieving the good? That is to say, if I want to achieve the good (and recall that Socrates and his opponent assume that all of us do (2 above)), should my life and my activity be organized around the acquisition of pleasure or the acquisition of knowledge?20 Socrates’ hedonist opponent has some reason to be confident in his candidate’s ability to win this second contest. After all, it is part of his view that pursuit of pleasure is a key source of evidence of pleasure’s value. Further, Socrates’ initial discussion of pleasure gives him no reason to doubt this. This, I think, is one reason why Protarchus does not object to Socrates’ proposal of what seems a controversial model of pleasure. This

16   I translate thus with the intention of leaving an open question about the status of this genitive as subjective (as at b5), so the arguments would have the desires, or objective, so the desires are for the arguments (as one might more naturally assume). See R.G. Bury The Philebus of Plato Cambridge 1897 ad loc. In fact, this passage, occurring at the end of the dialogue, seems to me interestingly emblematic of the central line of argument I am interested in, in which the characteristics of pleasure that are indeed evidenced by the erōtes of beasts are precisely the reason why pleasure fails to be an adequate, practical guide to the good. In turn, the question of whether to read a subjective or objective genitive at b6–7 (on which see also Dorothea Frede, Philebos Platon Werke: Übersetzung und Kommentar III.2, Göttingen: Vandenhoeck & Ruprecht, ad loc.) captures an important point about the way in which knowledge can successfully function as a practical guide to the good: the desire for knowledge is a desire to be responsive to the ways that knowledge is responsive to (and might, metaphorically, be said to have a desire for) the way things really are. 17   The text has sometimes been thought corrupt (see discussion in Bury ad loc.), but the construction can be made sense of. 18 19  See Phlb. 11d11–12a5, 22d4–8, 64c7–9, 65a7–b2, 67a10–12.  See Phlb. 22d1–4. 20   This tracks, but is distinct from an aspect of the second question that might be formulated as the question which of pleasure and knowledge (if either) has the right, teleologically structured responsibility for goodness, picked up regarding pleasure in the genesis argument of 53c–54d. The argument is illuminatingly discussed by Evans ‘Plato’s Anti-Hedonism’.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

116  Verity Harte model—call it ‘the Replenishment Model’—illustrated by the pleasures of eating and drinking, is stated in outline here: 6. SOC: I say that when we find the harmony being destroyed in animals, there comes to be, at that time, both a destruction of the nature and an occurrence of pains. PRO: What you say is very plausible. SOC: And, in turn, when the harmony is being restored and is returning to its own nature, pleasure must be said to occur, if we must speak as economically and swiftly as possible regarding the most important things.  (Phlb. 31d4–10)

Building on an earlier metaphysical passage, the Replenishment Model assumes the  existence of naturally harmonious conditions for animals. Think of these as the  optimal conditions for animals. It then ties the occurrence of pleasure to the ­restoration of these optimal conditions, though not to the occurrence of those ­conditions themselves. Pain arises as a consequence of an animal’s deterioration from some optimum condition. Pleasure arises when the animal is in the process of being restored to the optimal condition. When in optimal condition, there is neither ­pleasure nor pain. There is a danger lurking here for the hedonist. The Replenishment Model locates value (harmony, optimality) in the neutral conditions that pleasant processes reinstate. This is consistent with those pleasant processes being good, if not, as they have agreed, the good. But the value of the pleasures might be thought dependent on the value of the neutral conditions. This is a point that will be exploited in the dialogue’s second direct argument that pleasure is not the good. But the Replenishment Model is nevertheless one in which pleasure, when it occurs, coincides with the production of good, and this seems in keeping with the opponent’s view that pleasure is a practical guide to the good and helps to explain why the opponent does not object to the Model. But what is the structure of pleasure’s practical guidance? On the opponent’s view, pleasure is the aim of eating-when-hungry—we are motivated to eat when hungry both by the pain of hunger and by the pleasure of eating. Thus, if eating-when-hungry is a means to the good, pursuit of pleasure is a practical guide to that good. But the view that pleasure is the aim of eating-when-hungry is in tension with the view that the pleasure of eating-when-hungry is tied to the restoration of health, since health, the optimal condition restored, involves neither pleasure nor pain. Tied to the production of health, the pleasure of eating-when-hungry aims at its own destruction. If the aim of  eating-when-hungry is pleasure, the conditions of pleasure’s occurrence must be sought. This is a step in the direction of the Gorgias argument that the life of a hedonist is a life with the proverbial hole in one’s bucket: one must be forever creating those deteriorations from optimal condition that set the stage for experience of ­pleasure (Gorgias 492d–494b). For present purposes, the point to be noticed is that aiming at pleasure threatens to come apart from the pleasant production of health. But if ­aiming at pleasure comes apart from the pleasant production of health and health is the locus of value, pleasure will lose its promised internal practical connection to the production of good.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  117 Now, it is a striking feature of the introduction of the Replenishment Model that it offers no psychology, no account of pleasure experience. In the lengthy examination of pleasure that follows the introduction of the Replenishment Model, Socrates offers a rich and detailed psychology. The main object of this detailed psychology turns out to be to establish various ways in which pleasures can be, as he says, false. I take his object to be to support the plausible view that our felt experience of pleasure can and often does come apart from processes that as a matter of fact restore optimality. Socrates continues to focus on pleasures that instantiate the Model or at least are closely related to such pleasures, being anticipations, recollections, or imaginings of replenishment pleasures. And he proceeds, I would argue, to establish various ways in which ­maintaining the Model’s link between pleasure and goodness—between the pleasure of e­ ating-when-hungry and the good of producing and maintaining our health— depends on possession of Socrates’ contestants in the family of knowledge. We need accurate memories to give us correctly oriented desires. We need true judgments and self-knowledge. Take, for example, Socrates’ proffered psychology of appetitive desire (epithumia), the motivational force involved in, for example, eating-when-hungry and drinkingwhen-thirsty. It is surely quite obvious that one can experience pleasure in eating or drinking quite independently of whether that eating and drinking conduces to the production and maintenance of health. Socrates offers a portrait of appetitive desire when things work in the way that they should. Through appetite, an animal whose condition is currently not optimal and whose awareness of this is registered as discomfort or pain can not only register this fact, but also through memory identify the appropriate replenishment pleasure and set it as target to aim at. Memory was, somewhat surprisingly, one of Socrates’ original examples of things in the family of knowledge to  be set against pleasure as rivals to being the good. Socrates’ emphasis on the ­importance of memory to appetite is part of a broader argument that a pleasant ­process’s production of goodness is reliant on knowledge.21 One might see in Socrates’ insistence on pleasure’s dependence on possession of knowledge an implicit response to the type of worry provoked by reflection on the ‘experience machine’22—I do not want merely mistakenly to suppose I am getting the things that would please me and have the corresponding pleasure experiences; I want really to get what is pleasing. Pleasure experience that is not reality founded is not enough. Socrates’ position is, I think, even stronger than this. His view is not just that I want—or at least should want—what would as a matter of fact be pleasing to me. I want—or should want—what is actually pleasant—that is, what as a matter of fact tracks the restoration of my optimal condition. In this way, satisfying a desire for pleasure will rely on knowledge, of myself and of the relevant external factors. But if Socrates’ 21   I discuss the Philebus’s treatment of epithumia at 34d–36c in greater detail in ‘Desire, Memory and the Authority of Soul: Plato Philebus 35c–d’, Oxford Studies in Ancient Philosophy XLVI, Summer 2014, 33–72. 22   Robert Nozick, Anarchy, State and Utopia Basic Books 1974, pp. 42–5. Cf. Shelly Kagan Normative Ethics, Dimensions of Philosophy Series, Boulder, CO: Westview, 34–6.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

118  Verity Harte position—the weak or the strong one—gives his hedonist opponent some reason to  value possession of knowledge, the opponent can still regard this value for ­knowledge as being in the service of pleasure. An appetite for the pleasure of eating that is knowledge-based is still an appetite for pleasure. Hence, the opponent may ­reason, it is still pleasure, albeit knowledge-based pleasure that is the practical guide to the production of good.

IV.  True Pleasure The true pleasures—the ones that I propose be understood as idle pleasures—stand in contrast to the various other kinds of pleasure considered in the dialogue insofar as they are given a different motivational structure. The true pleasures, by definition, are not the object of an appetitive desire (epithumia). For Socrates stipulates that true pleasures are not preceded by and not mixed up with any pain. But appetitive desire (epithumia) was introduced by him as a species of pain.23 Further, in the context of the Philebus (as elsewhere in Plato), appetitive desire (epithumia) just is motivation in ­pursuit of pleasure, the kind of motivation that, in Socrates’ view, we share with nonhuman animals.24 Socrates’ discussion of the appetitive desires (epithumiai) of hunger and thirst makes clear that the account that he offers is a general one applying to human and non-human animals.25 And, as we have seen, in appetitive desire, the role accorded to memory is to remember as pleasant the replenishment process (35e9–10); the ­pleasure of eating, not health, is the explicit motivational object of hunger. Thus, in identifying a class of pleasures that are not the object of an appetitive desire, Socrates identifies the true pleasures in such a way that it follows that their motiv­ ational structure is not, as I shall say, epithumetic; it is not the kind of motivation in which pleasure is the object of pursuit. It does not immediately follow that they are idle in the way I suggest. For, while I have argued that appetitive desire just is motivation in pursuit of pleasure, I have not shown that other motivated activities cannot be. In what follows, I shall point to evidence that Socrates does indeed have in mind a different 23  At Phlb. 31e6–32a5 hunger and thirst are introduced as examples of pains. At 34d10–e1, explicitly referring back to this passage (‘we said’), they are called appetitive desires (epithumiai). I take this to indicate a refinement in our understanding of them made available by Socrates’ proffered psychology: appetitive desires are a species of pain. That appetitive desire (epithumia) in and of itself is constructed as a pain is supported by the options considered for the situation of being hungry and anticipating the satisfaction of hunger through eating. Protarchus initially thinks it a double pain; Socrates corrects him: a double pain would be to be hungry and to anticipate not being able to satisfy one’s hunger. Anticipating the satisfaction of hunger involves a mixture of pleasure and pain: hunger is pain; anticipated satisfaction a pleasure. See 35e7–36c2. 24   For more general Platonic association of epithumia and pleasure see, for example, Republic 439d6–8. While there are pleasures associated with all three parts of the soul, the particular relation between pleasure and epithumia is captured in the identification of the epithumetikon as such; this is the part of the soul that is motivated by pleasure as such, and it is the only such part of soul. 25   He talks in context both of ‘us’ (referring to the assembled group of people, Phlb. 35a3) and of animals in general (Phlb. 35c9–10, 35d3).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  119 structure of motivation for the true pleasures, though he is not at this stage of the ­dialogue all that explicit about this. How we should characterize and what we should make of this difference in their structure of motivation is the question I am interested in. Before introducing Socrates’ class of true pleasures, let me mention, but set aside a puzzle that his characterization of them as true gives rise to. As I have said, in arguing for the central importance of knowledge to maintaining the Replenishment Model’s connection between pleasure and goodness, Socrates has been pointing to the various ways in which pleasures can be false. But the class of pleasures identified as the true pleasures are true not only in such a way as to involve avoidance of one of these v­ arieties of falsehood. They are true in the sense of being entirely unmixed with pain. I set aside the question of how Socrates’ uses of ‘true’ fit together and why the unmixed pleasures should co-opt the label. Socrates offers two broad classes of true, unmixed pleasures. The first is associated with perceptual beauty, but of a particular sort; it includes pleasures involved in seeing and hearing and what are said to be a ‘less divine’ class involving the sense of smell (Phlb. 51e1–2). The second broad class is associated with intellectual activity: ‘the pleasures associated with learning’ (peri ta mathēmata, Phlb. 51e7–52a1). Neither ­ leasure is altogether transparent and the diet of pleasure they offer is class of p a­ pparently sparse. Consider, for example, how Socrates explains the class of perceptual true pleasures and the kind of perceptual beauty that they involve: 7. SOC: [The pleasures that can rightly be regarded as true are] those that are connected to the colours called beautiful and to shapes and the majority [of pleasures] in smells and the [pleasures] of sounds and all those that involve unperceived and pain-free lacks and deliver up perceived and pleasant fillings. PRO: But how indeed, Socrates, are we in turn to understand these in this way? SOC: Certainly, what I am saying may not be immediately clear, so an attempt must be made to make it clear. By beauty of shapes, I don’t mean to indicate what the majority of people would suppose, viz. [the beauty] of animals or of certain paintings; rather I mean— the argument declares—something straight or curved and the planes and solids that come to be out of these by means of compasses and those by means of straight edges and angles, if you understand me. For these I say not to be beautiful relative to something (pros ti), as others are, but to be naturally such as always to be beautiful in themselves (kath’ hauta) and to involve certain proper pleasures in no way comparable to those of scratching.  (Phlb. 51b3–d1)

Opaque though it may seem, Socrates gives his most expansive illustration of the class of perceptual true pleasures involving sight and hearing here in passage 7, detailing the  visual pleasures of certain beautiful shapes. He has relatively little to say more ­specifically about the relevant pleasures of colour.26 And what he says about the 26   The sum total of his expansion on the pleasures connected to visual perception of beautiful colours is found at Phlb. 51d2–3. It is often supposed that the discussion of purity of whiteness at Phlb. 52e6–53b7 is a discussion of the pleasure of perceiving pure white (for example, Dorothea Frede Plato Philebus Indianapolis: Hackett 1993, 63 n1 appears to assume this; and her Philebos, 303). But, whether or not

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

120  Verity Harte ­ leasures of hearing is comparatively brief (the single sentence of passage 8). p Nevertheless, discussions of the true pleasures of the Philebus have tended to focus on visual perception of colour and on hearing and have represented the dialogue as approving a narrow diet of pleasure indeed: visual pleasures associated with seeing single, unadulterated colour patches27 and aural pleasures associated with hearing ­single tones. This results from what I think is a misreading of various passages, including passage 8, though I do think it plausible that Socrates means to exclude the beauty of sounds in the context of ­performed music,28 as opposed to musical phrases. 8. SOC: I say that those amongst sounds that are smooth and bright that issue some single pure strain (melos) are not beautiful in relation to something else (pros heteron) but themselves by themselves (autas kath’ hautas) and pleasures akin to them follow.  (Phlb. 51d6–9)29

However, it is not my intention to get deep into these exegetical details here. For my purposes, what matters is not what pleasures Socrates identifies as true in this way— nor indeed whether such an identification is independently plausible; what matters is the structure of motivation that he ascribes to these pleasures. I do not suggest that this structure of motivation is an explicit part of their characterization; in context, Socrates’ focus is on their status as pure and as unmixed with pain, a characterization no doubt in part designed to appeal to the hedonist. But the structure emerges, I argue, once the passage is seen in the light of some additional context. Some of this context comes from later on in the dialogue, some from a passage of the Gorgias, a dialogue that, as so often in the Philebus, contributes important background to this later re-engagement with hedonism.30 Begin with passage 7. Socrates points to the pleasures of certain perceptually beautiful shapes characterized in geometrical terms: straight, curved, plane, and solid. Though these are characterized geometrically, the reference is not to abstract ­geometrical objects. This is clear, first, from the fact that they are perceptible, and, ­ leasure in perceiving pure white (if this ever occurs) would count as an example of pure pleasure, it is a p mistake to think that that is what is intended here. Rather, pure white serves as a paradigm for a general point about purity then applied to pure pleasure (at Phlb. 53b8–c2). Given this argumentative structure, the discussion of white should not itself be a discussion of some pure pleasure. 27   The assumption that this is (all and only) what Socrates has in mind may be driven by the mistaken understanding of Phlb. 52e6–53b7 mentioned in the previous note. 28   Performed music would be the aural analogue of paintings, pleasure in whose visual beauty was explicitly excluded in passage 7. 29   The conventional understanding is premised on understanding the term I have translated ‘strain’— melos—as referring to a single tone. I agree with Constance Meinwald (‘The Philebus’, in G. Fine ed. The Oxford Handbook of Plato, Oxford: Oxford University Press 2008, 484–503 at 498) that this is a forced and unnatural rendering of the term. Though the term is not one of the technical terms given detailed discussion, Meinwald’s view is supported by the evidence of E. Rocconi, Le parole delle muse: La formazione del lessico tecnico musicale nella Grecia antica, Rome 2003, whose glossary entry for melos is ‘canto, melodia’ (138). Thanks to Pauline LeVen for advice on this point and for direction to Rocconi. 30   The importance of the Gorgias as background to the Philebus is well emphasized by Jennifer Whiting, ‘Fools’ Pleasures in Plato’s Philebus’, in Mi-Kyoung Lee ed. Strategies of Argument: Essays in Ancient Ethics, Epistemology and Logic Oxford: Oxford University Press 2014, 21–59.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  121 s­ econd, from the fact that they are associated with production by means of certain craftsmen’s tools: the compass, straight edge, and angle. We, of course, have a use of the term ‘compass’ for an object that is used in producing visual representations of geometrical objects as part of doing geometry. But the Greek terms that Socrates is using refer to tools that are in the purview of crafts such as carpentry and masonry. This is confirmed by the fact that crafts of this sort are exactly the context in which the same terms will be mentioned again, later on in the dialogue, in the course of the dialogue’s discussion of varieties of knowledge. There, they are given as examples of the instruments of measurement used in various forms of carpentry (house building, shipbuilding, and so on) (Phlb. 56b4–c3). In this later passage, carpentry will be contrasted for the accuracy it has in virtue of use of these measuring tools with music performance,31 the example is flute-playing, said to involve a kind of experiential, trained knack (Phlb. 55e5–56a7). This is a version of a contrast familiar from elsewhere in Plato as a contrast between genuine skills or crafts (such as philosophy, medicine, gymnastics) and certain debased counterparts (such as rhetoric, pastry-making, cosmetics), which hardly merit the name of ‘skill’.32 The contrast is set out in its most stark and elaborate form in a passage of the Gorgias (462b3–465e1) and it is this contrast that offers the key to the non-epithumetic motivational structure of the true perceptual pleasures.33 Central to the contrast, as drawn in the passage, is that genuine skills aim at the good, where their debased counterparts instead aim at pleasure. 9. These [gymnastics, medicine, legislation and justice34], then, are the four parts, and they always provide care, in the one case for the body, in the other for the soul, with a view to what’s best. Now flattery takes notice of them, and—I won’t say by knowing, but only by guessing— divides itself into four, masks itself with each of the parts, and then pretends to be the characters of the masks. It takes no thought at all of whatever is best; with the lure of what’s most pleasant at the moment, it sniffs out folly and hoodwinks it, so that it gives the impression of being most deserving. Pastry baking has put on the mask of medicine, and pretends to know the foods that are best for the body, so that if a pastry baker and a doctor had to compete in front of children, or in front of men just as foolish as children, to determine which of the two, the doctor or the pastry baker, had expert knowledge of good food and bad, the doctor would die of starvation. I call this flattery, and I say that such a thing is shameful, Polus—it’s you I’m

31   He has in mind music performance rather than music theory of the kind used in illustration of the divine method at Phlb. 17b–d. 32   It has, however, some important differences in the Philebus version, not least in admitting flute playing, said to involve a ‘knack’, to the hierarchy of crafts, albeit at its lowest level. I do not address these differences here. 33   This need not mean that craft activities are the only context in which such pure pleasures can be experienced. 34   In context, Socrates sets out an elaborate fourfold comparison. There are four skills or crafts whose subject matter is the fitness of body or soul: (body) gymnastics and medicine; (soul) legislation and justice. These four as a whole have a counterpart; Socrates calls it ‘flattery’, which in turn divides into four pretender counterparts to gymnastics, medicine, legislation, and justice.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

122  Verity Harte saying this to—because it guesses at what’s pleasant with no consideration for what is best. (Grg. 464c3–465a2, tr. Zeyl)35

This Gorgias passage both illustrates and supports the idea of an alternative motiv­ ational structure for activities that, though they may result in pleasure (the Gorgias need not deny this to be true of genuine skills) are not aimed at pleasure, and that ­contrast with activities for which pleasure is indeed their aim. In the Philebus, in passage 7, Socrates contrasts pleasures of a kind associated with the production or appreciation of beauty in the objects of a geometrically informed craft such as carpentry with pleasures associated with the majority’s vulgar notion of beauty, that of animals or of paintings.36 But the beauty of animals (in particular, the human animal) and the beauty of paintings are central to two canonical Platonic examples of debased counterparts to crafts: cosmetics37 and mimetic artistry quite generally (music, drama, painting). Framed in terms of the passage’s contrast between relative and non-relative beauty, I propose to understand this as a contrast between beauty in the service of pleasure (and relative to context, to taste, to the person) and beauty in the service of something other than pleasure (what is best), where the latter is fixed, for example, by the objective standards of some craft. In the context of house building, one might think of this as Socrates displaying some sympathy for the Modernist architect’s credo: form follows function. The language of ‘following’ is exactly how Socrates later captures the relation between the true pleasures and the perceptual and intellectual activities to which they are related. At the end of the dialogue, Socrates gives prizes for the contribution made to the good of the best human life by its various features and ingredients. The true pleasures are the only pleasures to win a prize.38 They come in fifth and last; but they do get a prize. Here is how he identifies them. 10. SOC: Fifth, then, are the pleasures that we set down and marked off as painless, naming them pure, belonging to soul itself, following (hepomenas) kinds of knowledge,39 but also ­ erceptions.  (Phlb. 66c4–6) p 35  Note that ancient medicine was as interested in prescribing regimens (dietary advice) as drugs, incorporating tasks we might ascribe to a nutritionist, for example. 36   Strictly, passage 7 merely contrasts two kinds of beauty, but, since it does so in a context where we are seeking to identify more precisely certain pleasures that accompany perceptual experiences of beauty, it seems reasonable to suppose that, along with allowing that there is a (vulgar) beauty belonging to animals and paintings, so there are (vulgar) pleasures associated with its perceptual experience. 37   Though Socrates talks, in passage 7, quite generally of ‘animals’ (zōa), I take him to have the human animal in mind first and foremost, but to use the term ‘animal’ both to resonate with the term for painting (zōgraphēma) and to keep the focus on the kind of appetite-driven pleasures we share with other animals, as the reference to animals (zōa) in his statement of the Replenishment Model, in passage 6, makes clear. 38   They are not, however, the only pleasures to be included in the dialogue’s picture of the best human life, which includes also certain pleasures regarded as necessary (Phlb. 62e8–10). I take the difference between the prize-winning pleasures and those that are included without winning a prize to mark the difference between pleasures that make a contribution to the goodness of the best human life and those that do not. 39   It is noteworthy that Socrates talks here of epistemai, branches of knowledge or expertise. This bears on a question as to whether his earlier talk of mathēmata, in connection with the ‘pleasures of learning’,

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  123 This talk of following indicates some kind of dependence of the true pleasures on the relevant sorts of knowledge or perception. But this is a dependence of a different sort from that of the epithumetic, but knowledge-based pleasures considered earlier. There, pleasure depended upon accurate memory or true judgment for achievement of the restoration of optimal condition associated with the production of pleasure. But such intellectual accomplishments had the job of properly directing epithumetic pursuit of pleasure and hence might still be regarded as in the service of pleasure. In contrast, when pure pleasures follow and are guided by the associated perceptual and intellectual activities, these perceptual and intellectual activities are directed at pleasureindependent norms which they pass on to pleasure. Both these relationships can be framed in terms of guidance of pleasure by knowledge. The difference between them is the difference between my hiring a guide to take me where I want to go and my choosing as guide a leader who will be in charge of what we do. It is to pleasures guided in the latter way, led by standards not themselves determined by pleasure, that Socrates attributes genuine value. This is the pay-off of the alternative motivational structure accorded the true pleasures. The motivational structure accorded the true pleasures is not explicit when they are first introduced. Socrates’ focus there is on the fact of these pleasures being unmixed with pain. This is confirmed by his assertion at the end of passage 7 that the pleasures that he is interested in are ‘in no way comparable to those of scratching’ (Phlb. 51d1). The pleasure of scratching an itch was held up earlier (Phlb. 46d10, referring back to 46a8–13) as a core example of a bodily mixed pleasure (there are also examples of mixed pleasures of body and soul and of soul alone). It is something of a puzzle, if I am right, as to why the class of pleasures that have this alternative motivational structure should also be identified as unmixed with pain. It is certainly not obvious that pleasant activities that are not preceded by any pain cannot be pursued as pleasures. Perhaps Socrates need not deny this. Perhaps he will simply stipulate that the pleasures he regards as worthy of a prize are those that are not so pursued. And he might justify this by pointing to the way in which pursuit of pleasure can come apart from tracking actual goodness, the argumentative point of his earlier defence of the possibility of false pleasure. Putative true pleasures pursued as pleasures would be merely accidentally good. Socrates may also hold it important to the goodness of true pleasures that pleasure be taken in them in the right way, as a consequence of the value of the concomitant activities and not for their pleasure. What is clear, given the dialogue’s earlier discussion of appetite (epithumia), is that Socrates’ insistence that true pleasures are not preceded by pain makes them non-­ epithumetic. This gives the careful reader reason to question the structure of their motiv­ation. It is this question, I argue, that gets answered against the background of the Gorgias contrast and by the later Philebus context: the true pleasures are idle Phlb. 51e7–52a3, intended to incorporate achievement as well as acquisition of knowledge, despite their framing as ‘fillings’.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

124  Verity Harte ­ leasures; the pleasure they give rise to is not the intended object of their subject’s p ­pursuit. But they are not for that reason unmotivated. Consider Socrates’ example of the true ­pleasure of learning. Later in the dialogue he is prepared to talk in terms of desire in an intellectual context. Perhaps significantly, he uses a different Greek term, not epithumia but erōs (Phlb. 58d5). Certainly significantly, if I am right, what he talks of in ­connection with knowledge is a desire for truth and not for the pleasure arising from pursuit of truth.

V.  True Pleasures and Value Though Socrates’ true pleasures are not the only pleasures to be included in the best human life described by the dialogue, they are the only pleasures to be awarded a prize. I take this to show that they are the only pleasures that contribute to the goodness of that life. (The other pleasures included are explicitly characterized as being necessary, Phlb. 62e8–10). The alternative motivational structure given to the true pleasures does not constitute their value, but it is, I argue, non-accidentally related to it. For the hedonist, the value of the true, pure pleasures is most naturally sourced in their lack of admixture with pain. For Socrates, this lack of admixture of pain is a cue to their non-epithumetic character as pleasures. This in turn is connected to their value, which consists in their having a character that is measured and not in excess, where measure, throughout the dialogue, is a constituent of value. Even before they feature amongst the dialogue’s prizewinners, the unique situation of the true, pure pleasures as regards value is signalled by what is in context a surprising classification of them. 11. SOC: Since, then, we have distinguished separately the pure pleasures and those that might correctly be said to be impure, let us add to our story lack of due measure (ametria) for the excessive pleasures [i.e. the impure pleasures], but, in contrast, due measure (emmetria) to those that are not. That is to say, the big and the excessive, whether such [pleasures] occur frequently or infrequently, let us assign them to that class of what is without limit, admitting less and more through both body and soul, but the [pleasures] that do not, [let us assign] to the class of things that have due measure.  (Phlb. 52c1–d1)

Socrates here appeals to two of the four classes he identified in an earlier metaphysical passage in which he divided beings into four classes: (1) things without limit, whose hallmark is to be subject to variations of ‘more and less’; (2) limits; (3) the mixtures that result from the imposition of limit on things that are without limit; and (4) the cause of said mixtures (Phlb. 23c–27c). In his talk, in passage 11, of ‘things that have due measure’, he picks out the third class, mixtures. His description of the class is a function of his earlier identification of limits with measures (specifically ratios and proportions), where measure should be understood as a normative notion: the right ratio or proportion for some specified context. For example, health, in this passage, consists in the right combination of elements such as hot and cold, wet and dry, their combination in

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  125 the due proportion for bodily good condition. In general, in the Philebus, goodness is constituted by measure: appropriate proportions and ratios. Passage 11 assigns true pleasures to a different place in this metaphysical classification from other pleasures.40 True pleasures are in the class of mixtures—the class associated with goods. (To prevent confusion, note that this is a different and quasi-technical use of ‘mixture’; the true pleasures are nevertheless unmixed with pain.) Other pleasures are in the class of things without limit—a class of things that, in and of themselves, do not possess the relevant good-making, measure properties. What is significant about this set of assignments is that it is at least apparently inconsistent with something that Socrates previously said in the dialogue and specifically instructed us to remember. This was that the genus of pleasure in general was the class of things without limit (Phlb. 31a7–10). What are we to make of this apparent inconsistency? One possible response is to suppose that Socrates’ dispersed assignment of pleasures indicates that, despite his earlier insistence, pleasure is not in fact a unitary class with a single genus.41 My own preferred explanation is that the conjunction of Socrates’ claims illustrates the point that true pleasures (and only true pleasures) are, without qualification, good (because in possession of appropriate measure), while also showing that their goodness is not derived from their being pleasures, but is inherited from the value of the normproviding activities on which the true pleasures depend. For a later example of the wedge I see Socrates driving between the standard of pleasure and value standards of correctness or benefit in what are nonetheless ­pleasure-providing activities, one might compare the following passage from Plato’s Laws.42 Here Plato has the Athenian (the Laws’ main speaker) offer a terminological 40   I take this to be a different place from all other pleasures. Against this, one might object that the talk of intensity picks up the particular pathological mixed pleasures discussed in Phlb. 45aff. and not, for ­example, the (nevertheless, mixed) pleasure of eating the appropriate amount when hungry. To this, there are, I think, two strategies of response: 1. Bite the bullet: He does mean all mixed pleasures. This is what we would expect given his understanding of purity and given that the contrast is between pure and impure. The discussion of intense pleasures simply reveals the inherent weakness/problem of all mixed pleasure. 2. Take there to be two ways of thinking about the pleasure of eating the right amount when hungry: In one, such pleasure is accidentally moderate, but is guided by pleasure—it is simply that one’s pleasure happened to be satisfied at the moderate point. This being the case, though the pleasure is not obviously bad on this occasion, it has the relevant bad-making feature. In the other, the pleasure occurs as an expression of moderation and is in some sense a pleasure in moderation. (Note, in this connection, the interest in moderation at Phlb. 45de and earlier at 12cd, where the datives give the option to understand this as claiming that the pleasure of moderation is a pleasure, in some sense, in moderation, that is, in moderate acts.) So, pleasures that look like they should be mixed, on one way of thinking about things, could on this second strategy be incorporated into the good side of the equation. The dialogue clearly says that there is such a thing as the pleasure of moderation, but leaves us to construct the account of it for ourselves. This may be in favour of strategy 2. 41   This view is defended—at length and based on a rich reading of the whole dialogue—by Emily Fletcher in her Toronto PhD: Plato on Pleasure, Intelligence and the Human Good: An Interpretation of the Philebus (2012). 42   My thanks to Susan Sauvé Meyer for suggesting this passage as illuminating comparison and for sharing her then forthcoming translation and commentary, Plato Laws Books I and II Translated with a Commentary, Oxford: Clarendon Plato Series.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

126  Verity Harte distinction to mark out pleasures arising from activities governed by standards ­independent of pleasure and those that are governed by the standard of pleasure itself. 12. ATH: Now, in the case of anything that brings us enjoyment, mustn’t we first of all say the following? Either this very feature on its own is the most important thing about it, or else what is most important is a certain correctness or, third, a benefit? Consider eating and drinking, for example, and taking in nourishment quite generally. These bring us enjoyment, which we might call ‘pleasure,’ but if we are talking about correctness and benefit, we talk of what is healthy in a given meal, and this is what is most correct in it. CL: Absolutely. ATH: Learning too has its pleasure, an element of enjoyment that comes along with it. But its correctness and benefit, what is good and beautiful in it, derive from its truth. CL: That is so. ATH: Now consider the representational arts.  Their business is to produce likenesses, and when they succeed, any pleasure that accompanies their product, should any pleasure arise, is appropriately called ‘enjoyment’. CL: Yes. ATH: But presumably the correctness in such cases would be produced, roughly speaking, by accuracy in quantity and quality, not by pleasure. CL: You are right. ATH: So the only thing that would be correctly judged by pleasure is what provides neither benefit nor truth nor likeness when it is achieved (nor harm either, for that matter). Rather, it comes to be solely for the sake of this very thing, enjoyment, which those other things bring in their wake, but when it occurs without them, is best called ‘pleasure’.  (Laws 667b5–e4, tr. Meyer)

VI.  The Value of Idle Pleasure The intuition that pleasure is a final goal of practical reasoning is a persistent one. Anscombe famously wrote: ‘ “It’s pleasant” is an adequate answer to “What’s the good of it?” or “What do you want that for?” I.e., the chain of “Why’s” comes to an end with this answer.’43 One might think this an argument for hedonism. Against this view, Elijah Millgram has offered what he calls an ‘observationalist account’ of pleasure: the answer ‘it’s pleasant’ reflects an experiential statement of the grounds of my practical reasoning insofar as the pleasure I take in something is a defeasible indicator of its desirability.44 Socrates allows, with Millgram, that pleasure can be an indicator of value. But, for Socrates, practical pursuit of pleasure precisely destabilizes the relation between pleasure and goodness, and this is the reason he will not follow Millgram in allowing the pleasure taken in something to be the grounds of any practical reasoning. The Socrates of the Philebus does not deny a connection between pleasure and goodness; the true pleasures make a valuable contribution to the best human life. Plato, we may suspect, is no more persuaded by his nephew Speusippus (a notorious grouch about pleasure)45 than by his colleague, Eudoxus. But the only way to maintain a reliable connection between pleasure and goodness is not to take pleasure as one’s aim. 44  Anscombe Intention2 §40, at 78.   ‘Pleasure in Practical Reasoning’, especially 403–5.   And often seen lurking behind the dialogue’s reference to ‘the real enemies of Philebus’, also known as ‘the grumpies’ (the duschereis), Phlb. 44b6–d6, introduced at a critical moment en route to the discussion of true pleasures. 43 45

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Plato’s Philebus and the Value of Idle Pleasure  127 Thus, in Socrates’ view, not only is his hedonist opponent wrong that pursuit of ­pleasure is evidence for hedonism; pursuit of pleasure is a sure fire way to ensure that the good escapes one. This is the value of idle pleasure.46

Bibliography J. Brunschwig, ‘The Cradle Argument in Epicureanism and Stoicism’, The Norms of Nature: Studies in Hellenistic Ethics ed. M. Schofield and G. Striker, Cambridge: Cambridge University Press 1986, 113–45. R.G. Bury, The Philebus of Plato. Cambridge: Cambridge University Press 1897. R. Crisp, Reasons and the Good. Oxford: Oxford University Press 2006. S. Darwall, Philosophical Ethics. Dimensions of Philosophy Series, Boulder, CO: Westview, Glossary. M. Evans, ‘Plato’s Anti-Hedonism’ in Proceedings of the Boston Area Colloquium in Ancient Philosophy XXIII, 2007, ed. J.J. Cleary, and G.M. Gurtler, S.J., Leiden: Brill 2008, 121–45. Emily Fletcher, Plato on Pleasure, Intelligence and the Human Good: An Interpretation of the Philebus. Doctoral dissertation, University of Toronto 2012. Dorothea Frede, Plato Philebus. Indianapolis: Hackett 1993. Dorothea Frede, Philebos. Platon Werke: Übersetzung und Kommentar III.2, Göttingen: Vandenhoeck & Ruprecht 1997. G.E.M. Anscombe, Intention. Second Edition, Cambridge: Harvard University Press 1963. J.C.B. Gosling, Plato Philebus. Oxford: Clarendon Press 1975. Verity Harte, ‘Desire, Memory and the Authority of Soul: Plato Philebus 35c–d’, Oxford Studies in Ancient Philosophy XLVI, Summer 2014, 33–72. Terence Irwin, Plato’s Ethics. Oxford: Oxford University Press 1995. Shelly Kagan, Normative Ethics. Dimensions of Philosophy Series, Boulder, CO: Westview 1998. Frances Kamm, ‘The Doctrine of Triple Effect and Why a Rational Agent Need not Intend the Means to his End’, Proceedings of the Aristotelian Society, Supplementary volume 74, 2000, 21–39. W.E.H. Lecky, The History of European Morals from Augustus to Charlemagne Part I. New York: D. Appleton & Co. 1869. G.E.R. Lloyd, ‘Review of J.C.B. Gosling Philebus’, The Classical Review 27, 1977, 173–5.

46   This essay was given as the first of two Whitehead Lectures at Harvard University in April 2013 and, in a revised form, as the Rosamond Kent Sprague Lecture at University of South Carolina in March 2015. I am grateful to the faculties in Philosophy at Harvard and at South Carolina for the honour of the invitations, for the departments’ hospitality, and for their lively and stimulating discussion of my lectures from which I greatly profited. The essay has benefited from comments on earlier drafts by David Charles, Ursula Coope, Shelly Kagan, Rachana Kamtekar, M.M. McCabe, Susan Sauvé Meyer, Barbara Sattler, Jennifer Whiting, and Raphael Woolf. I owe debts of longer-standing to conversations—in some cases, multiple conversations—about the Philebus with Joachim Aufderheide, Charles Brittain, Amber Carpenter, Matt Evans, Carl Ginet, Terry Irwin, Gabriel Lear, Hendrik Lorenz, M.M. McCabe, and Jennifer Whiting as well as to the participants in seminars on the Philebus at Cornell, King’s College London, and Yale. For research assistance at various points during my work on this essay, thanks to Maya Gupta.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

128  Verity Harte Constance Meinwald, ‘The Philebus’, in G. Fine ed. The Oxford Handbook of Plato. Oxford: Oxford University Press 2008. Susan Sauvé Meyer, Plato Laws Books I and II Translated with a Commentary. Oxford: Clarendon Plato Series 2015. E. Millgram, ‘Pleasure in Practical Reasoning’, The Monist 76(3), 1993, 394–415. Robert Nozick, Anarchy, State and Utopia. New York: Basic Books 1974. P. Railton, ‘Alienation, Consequentialism and the Demands of Morality’, Philosophy and Public Affairs 13(2), 1984, 134–71. E. Rocconi, Le parole delle muse: La formazione del lessico tecnico musicale nella Grecia antica. Rome: Edizione Quasar 2003. Henry Sidgwick, The Methods of Ethics, Seventh Edition, Indianapolis: Hackett 1981. J. David Velleman, ‘Is Motivation Internal to Value?’, in J. David Velleman The Possibility of Practical Reason. Oxford: Oxford University Press 2000. R.A.H. Waterfield, ‘The Place of the Philebus in Plato’s Dialogues’, Phronesis 25(3), 1980, 270–305. Jennifer Whiting, ‘Fools’ Pleasures in Plato’s Philebus’, in Mi-Kyoung Lee ed. Strategies of Argument: Essays in Ancient Ethics, Epistemology and Logic. Oxford: Oxford University Press 2014, 21–59.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

7 A Series of Goods Christopher Shields

I.  Good, Better, Best Near the outset of his exposition of the human good in the Nicomachean Ethics, Aristotle pauses to inform his readers that he finds himself saddled with an unwelcome duty: he must undertake a critical investigation of the Academic approach to goodness, where we may presume that he means to focus his attention on the Academic approach as it is realized most markedly in Plato’s postulation of the Form of the Good.1 Plato’s Form of the Good, as Aristotle presents it, is meant to be ‘something common, universal, and one’ (koinon ti katholou kai hen; EN 1096a28), universal and common, that is, to all good things. This common good, postulated by Plato in Republic v, is also, again as Aristotle characterizes it, ‘something existing in its own right’ (ti kath’ auto einai), serving as the cause (aition) of the goodness of all the things we call good (EN 1095a26–8).

1   This is in one way a contentious presumption, since Aristotle noticeably does not name Plato in the central chapter in which he assails the Academic conception of goodness, Nicomachean Ethics i 6, speaking instead only of those who introduced, or perhaps, those who embraced (hoi komisantes, EN 1096a17) the theory of Forms and mentioning by name only Speusippus (EN 1096b7), Plato’s nephew and successor as head of the Academy. Some (e.g. Stewart (1892)) think that the presumption is more than contentious, but is in fact positively false, contending instead that Aristotle fails to mention Plato for the very good reason that he is talking not about him or his views, but rather about the views of Speusippus or Xenocrates, or perhaps of the other, lesser members of the Academy. This judgment is motivated in part by the thought that Aristotle’s criticisms cannot be directed against someone whose views about the final good Aristotle in fact shares (e.g. Stewart (1892 vol. I, 74)). Three obstacles militate against this view. First, Stewart is wrong to suppose that Aristotle finds Plato’s conception of goodness congenial in the relevant respects. Second, Speusippus, contrary to the view of Diogenes (Vitae iv 1), seems to have been one of the Academicians who in fact rejected Plato’s theory of Forms; still less, then, was he one of those who introduced or promulgated it. Third, the evidence is tenuous, but Speusippus seems to have rejected in particular the principle of causal synonymy (that only φ things can cause non-φ things to become φ) relied upon by the author in Aristotle’s sights (Met. 1028b21, 1072b31, 1091a34). Although the details of this doxography merit fuller discussion, the current chapter maintains the presumption, however contentious, that Aristotle targets Plato and not Speusippus or some other, unnamed Academic in his polemic. Be that as it may, those rejecting this presumption may safely treat ‘Plato’ as a place-holder for ‘the proponent of the view of goodness here characterized and rejected.’ As we shall see, the precise content of the view assailed by Aristotle is already itself a matter of controversy. See also Graesar (1997) and (1999).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

130  Christopher Shields According to this Academic view, at least as mooted,2 all good things—whether they are good symphonies, good citizens, good gestures, or good tries—are all univocally good, good in the same sense, made good by dint of their participation in the same single Form, namely the Form of the Good. Such a theory, whatever its ultimate demerits, seems to have at least the merit of economy: the Form of the Good, on the Academic theory, at once grounds and explains all the goodness we may ever encounter. Unfortunately, Aristotle avers, this theory has at best a false economy to it: it is not true. Saying so grieves him, he submits, and perhaps understandably so. After all, as he explains, those who first promulgated the theory of Forms were friends of his. Yet the requisites of piety outstrip even those of friendship: We had perhaps better consider the universal good and run through the puzzles concerning what is meant by it—even though this sort of investigation is unwelcome to us, because those who introduced the Forms are friends of ours. Yet presumably it would be the better course to destroy even what is close to us, as something necessary for preserving the truth—and all the more so, given that we are philosophers. For though we love them both, piety bids us to honour the truth before our friends.  (EN 1096a11–16)

The allegiances of philosophers are, then, first and last to the truth. This is not to say, of course, that friendship should be discarded when we find our friends saying what is false or even daft; rather, though friendship is valuable, we should not shield our friends from the truth—especially when our friends are, as we ourselves are, philosophers. This same maxim, however, commends a sober look at Aristotle’s presentation of the truth about the Good as he sees it: what precisely is it that he finds objectionable in Plato’s view? More to the point, should we ourselves, based upon the evidence we have available to us, accept Aristotle’s view of the situation, or should we reject his antiPlatonic rhetoric as revealing little more than a pretension to understanding a Platonic insight about the Good which has eluded him altogether? Or does our state of evidence preclude our arriving at reasonably secure answers to these sorts of questions? In fact, these questions are large and multivalent, since Aristotle retails a series of some six arguments against Plato’s conception of the Good in both Nicomachean Ethics i 6 and Eudemian Ethics i 8.3 We may make some modest progress by considering the 2   There is a further question, on the assumption that the situation is as described in n. 1, as to whether Aristotle’s characterization of Plato’s conception the Good is fair and accurate. Although I will not argue the point here, I assume that it is in all essentials correct, especially as regards the thesis of univocity. More fully, I assume in this chapter that Plato holds—as Aristotle contends that he holds—that the Form of the Good is a simple, univocal, non-natural, indefinable, irreducible, non-indexed property (non-indexed in the sense that it is good simpliciter, and not good-for-someone or -something, or merely good-as-a-φ, for any predicate φ). For a fuller explication of the notion of an indexed good, see Shields (2015, 93, with n. 27). For detailed arguments in support, see Shields (2008) and (2015). 3   For an overview of the arguments, whose number is disputed, as they appear in Nicomachean Ethics i 6, see Shields (2015 §IV). Briefly, they are: (i) an argument based upon the lack of Forms for items arranged in series (EN 1096a17–23); (ii) an argument based on the theory of categories (EN 1096a23–9; cf. Top. 107a3–17); (iii) an argument from the diversity of sciences (EN 1096a29–34); (iv) two primarily ad hominem arguments concerning the abstract and eternal character of Forms (EN 1096a34–1096b5); (v) a dilemmic argument based on the Academic commitment to the existence of a plurality of intrinsic goods

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  131 arguments at first individually and thereafter in their interlocking totality. Here we consider only the first argument in this larger set, since doing so already suffices to indicate the extreme complexity of the questions we wish to address. This first argument is peculiar in so far as it represents itself as standing on ground common to both critic and target, namely that—as Aristotle represents them—the Platonists themselves agree with him in holding that there are no Forms set over items arranged in a series. If, then, as Aristotle goes on to urge, good things are in fact arranged in some manner of series, Plato will be constrained to agree that there is no Form of Goodness after all. The argument is, then, at least to this degree ad hominem: according to Aristotle, Plato’s own commitments serve to show up the falsity of his adherence to the Form of the Good. We proceed in four stages. We begin by stating and analysing the argument as it is presented in Nicomachean Ethics i 6, while clearing the ground and settling some preliminary issues. In assessing the argument thus presented, we find it almost wholly inadequate—to such a degree, in fact, that this already gives us reason to suppose that this cannot be the whole of the story. Fortunately, a parallel passage in Eudemian Ethics i 8 provides some additional data which serves to rescue the argument from the shortcomings of the version extracted from Nicomachean Ethics. Based upon the treatment of the Eudemian Ethics, the argument may then be expanded, with a view towards reappraisal. By contrast with the earlier version, the more sophisticated formulation reveals a good deal more structure behind Aristotle’s objection to Plato. Unfortunately, it too, for subtler reasons, falls wide of its mark. Even so, one might, by releasing on both texts in conjunction, hazard a conjecture about the source of Aristotle’s confidence regarding a series of goods. For once we have shown it in its best, textually anchored version, the argument against there being a Form set over a series of goods proves entirely reasonable, though not entirely devastating. It is not that Plato, or a Platonist, will have a ready riposte; rather, Aristotle’s argument turns out to raise a familiar problem for a Platonist regarding the explanatory efficacy of the theory of Forms. In this sense, Aristotle’s complaint about the Form of the Good turns out to be not so much a complaint about the Form of Goodness, but about there being a Form of Goodness, one which plays the explanatory role Plato would like to see it play. Indeed, Aristotle’s argument against a series of goods proves to be a special case of a broader argument against all Forms, and not merely an argument against the Form of Goodness per se. The criticisms launched against the Form of the Good are thus peculiar neither to Goodness nor even to Forms (putatively) arrayed in series. To be clear, however, we will discover not that the (putative) fact that good things form a series plays no role in Aristotle’s criticism of the Form of the Good, but only that this fact plays a role which requires supplementation by other common anti-Platonic theses. This in a way should (EN 1096b8–26); and (vi) two arguments on the practical irrelevance of the Form of the Good (EN 1096b35–1097a13).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

132  Christopher Shields turn out to be a welcome result for those sympathetic to Aristotle’s anti-Platonic ­animadversions. For his argument against a series of goods proves to but one of several launching pads trained on the same destination, namely a dismissal of the Form of the Good.

II.  A Complicated Simple Argument in Nicomachean Ethics i 6 Aristotle’s first argument against Plato’s Form of the Good in Nicomachean Ethics i 6 is deceptively simple in presentation; in fact, it proves concise to the point of obscurity. Immediately after expressing his rueful duty to piety, Aristotle says: Those who introduced this view did not posit Ideas in cases in which they recognized the prior and the posterior—this is the very reason they did not establish an Idea of numbers; but something is called good both in [the category of] what it is [scil. substance, or ousia] and in that of quality and in that of relation, and that which is per se, i.e. substance, is prior in nature to the relative (for the latter seems like an offshoot co-incident with what is), so that there could not be some common Idea set over all these goods.  (EN 1096a17–23)

In its simplest formulation, the argument closest to the text here is easy to state, but hard to fathom: 1. Where things φ1, φ2 . . . φn are related as prior and posterior (proteron kai husteron), there is no Form φ-ness set over them. 2. Good things are related as prior and posterior. 3. So, there is no Form Goodness set over good things. What might be said on behalf of either of these premises? Beginning with the second first, we are initially left perplexed. If we consider a random variety of good things, or, to put matters in a linguistic idiom, if we consider a random set of things of which we predicate ‘. . . is good’,4 we discern no immediate ordering. Consider, for instance, the sentences: 1. Gelato on a hot evening is always good. 2. Well, no-one can deny that Damrau’s rendition of Strauss’s Morgen is good. 3. Some shortcuts are good—some are just shortcuts. This one is good. 4. Spite is soul-draining and self-destructive; genuine forgiveness is, well, good. 4   We may sometimes speak in a linguistic idiom for clarity or ease of expression, but in each case we should bear in mind that Plato is ultimately thinking in metaphysical terms, as we can surmise from Republic 596a and, especially, Parmenides 130e–131a: ‘So, tell me this: is it your view that, as you say, there are certain Forms whose names these other things have through getting a share of them as, for instance, they came to be like by getting a share of Likeness, large by getting a share of Largeness, and just and beautiful by getting a share of Justice and Beauty?’ ‘It certainly is,’ Socrates replied.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  133 The four general kinds of things here called good—evening treats, renditions, directions of travel, and psychological processes—are all, let us agree, good things. But they stand in no discernible ordering relation. Certainly, they are not readily likened to the series of numbers, to which Aristotle implicitly does liken good things when he notes that ‘this is the very reason they [scil. the Academics] did not establish an Idea of numbers.’ If we take the comparison to the number series at face value, then we should expect each predication of goodness to stand in some determinate priority–posteriority relation to every other, such that each has a fixed and unalterable position in a determinate, changeless series. It is difficult to fathom how one could even begin to construct an argument for any such view. It seems patently outlandish. Fortunately, as Aristotle’s own brief illustration indicates, in thinking of the orderings of good things, he is not taking the comparison to the number series at face value. Rather, he is contending that predications of goodness, like the numbers, stand to one another in priority and posterity relations, a thesis which is compatible with the thought that predications of goodness collect themselves into kinds, such that some kinds are, along some dimension to be specified, prior to others, while others are posterior to them. Consistent with this picture, some kinds of goods could be prior to all other kinds, others posterior to all other kinds, and some prior to some and posterior to others. On the assumption that a suitable priority relation could be specified, this would suffice to justify a general comparison with the number series, for which, we are told, not even the Platonists want Forms. On the assumption that this general orientation is correct, there remains, then, the task of specifying the priority and posteriority relations obtaining between kinds of good things. Here too Aristotle provides some orientation within this brief passage. He appeals to the doctrine of categories, contending that substances are ‘prior in nature’ (proteron kata phusin) to other categories of beings, like qualities or relations. This ­specifies the priority relation clearly as a familiar sort of priority, namely priority in nature, as it obtains in the categories, where his form of priority receives a full (if disputed)5 characterization in Metaphysics Δ. There it is explicitly aligned with priority in substance (kata phusin kai ousian; Met. 1094a1–2): Some things are called prior and posterior in this way, while others are called so in nature and substance, those for which it is possible to be without other things, but not the latter without them. (Met. 1019a1–4)6 5   See Peramatzis (2011), with the review of Malink (2013) for various understandings of Aristotle’s notion of priority in nature. I here accept, without argument, the traditional understanding, which has been discussed and criticized by Peramatzis. As far as the argument of this essay is concerned, the same results could be attained, in a more round-about way, by accepting Peramatzis’s interpretation and making the necessary adjustments. 6   In the Physics, we find priority in nature not so closely connected to priority in substance: ‘A thing is said to be prior to other things when, if it is not, the others will not be, whereas it can be without the others; and there is also priority in time and priority in substance (kat’ ousian)’ (Phys. 260b17–19).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

134  Christopher Shields At least initially, this passage seems to gloss priority in nature or substance in terms of an asymmetry of existence implication: a is prior to b just in case (i) a can exist without b, while (ii) b cannot exist without a. Thus, to take an illustration initially favourable to this way of understanding priority in nature and substance, Socrates can exist without his pallor, while his pallor cannot exist without him. Generalizing, then, substances can exist without their intrinsic accidents whereas intrinsic accidents cannot exist without the substances of which they are predicated. Applying this understanding to the case at hand, we might be inclined to think that predications of goodness in the category of substance are prior in nature to predications of goodness in other categories of being. So, for instance, Aristotle mentions the nonsubstantial categories of quality and relative (pros ti), where a relative, we know from Categories 7, is one of a pair of correlatively existing beings, each of whose existence implies the existence of the other, and indeed, each of whose definitions or accounts must mention the existence of the others (Cat. 8a28–34).7 So, for instance, slaves and masters are relatives, since there are no slaves without masters and no masters without slaves, and, moreover, to provide an account (a logos) of the one requires mentioning, at least implicitly, the other. Tracing this same thought a bit further, one might say that the goodness of substance is prior to the goodness of non-substance in the way that relatives are posterior to substances. That is, just as nothing is a slave without first being some particular human being, a substance, perhaps no quality or relative is good, thinks Aristotle, without there being a good substance realizing that quality. To begin again with a favourable illustration, one might say that the virtues are good, but that a given virtue, like courage, is good only when it is realized by a good human being, like Socrates. Thus, on this approach, one might attempt, in effect, to allow the priority relations among predications of goodness to supervene on the priority relations of the categories. This yields the following argument, also close to the text, but now a bit more expansive: 1. Beings in the category of substance (ousiai) are called good. 2. Beings in the categories of (a) quality (poios) and (b) relative (pros ti) are called good. 3. Substances are prior in nature (proteron kata phusin) to qualities and relatives. 4. If substances are prior to the other categories of being, and if φ is predicated of substances and beings in the other categories of being, then the φ substances are prior to the φ non-substances. 5. So, good things are related as prior and posterior. The conclusion of this mildly expanded argument then provides all the reason we might need to accept our second premise, namely that good things are arranged in a series. 7   Closer to Aristotle’s actual formulation: x is a relative (pros ti) =df what it is for x to be F is the same as what it is for x to stand in R to y. The locution ‘what it is for x to be F’ implies the notion of an essencespecifying account.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  135 It might be a short series, beginning with substances and ending with non-substances, but a short series is a series all the same. In any event, this is just what Aristotle seems to imply when he explicates the priority of substances with regard to relatives by noting that ‘the latter seems like an offshoot co-incident with what is [scil. with the former, namely substance, or ousia]’ (EN 1096a22–3). Unfortunately, this suggestion advances the discussion but little. To begin, it is unclear why the case of virtue, if it is itself a suitable illustration of the contention Aristotle is meaning to advance here, should be thought to generalize. If we consider a slave named Marcus, we might predicate pallor of him in at least two ways, by saying (i) ‘Marcus is pale’ or (ii) ‘The slave is pale.’ One feels no inclination, however, to think that the second predication is somehow unlike the first, and still less to think that it is posterior to it. Pallor is pallor, and both instances of its predication, both equally in the category of quality, seem plainly on par. Indeed, Aristotle himself seems to concede precisely this point when averring that whiteness is predicated univocally irrespective of its subject. This he does when setting a dilemma as a rejoinder to a Platonic response to his basic argument against the univocity of goodness. He imagines a Platonist complaining that the Academic view is restricted to intrinsic goods, such that only things good in their own right are univocally good (goods kath’ hauta). On this approach, the Form of the Good corresponds univocally only to intrinsic predications of goodness (EN 1096b7–13). Aristotle’s response: Which goods should one regard as goods in their own right (kath’ hauta)? Those pursued even when considered individually, like intelligence, seeing, certain pleasures, and honours? For even if we pursue these because of something else, one would none the less regard them as goods in their own right. Or is nothing good in its own right beyond the Idea (idea) ? If the latter, the Form (eidos) will be otiose. If the former, and these are counted as among things good in their own right, then the account of goodness (ton tagathou logon) in all of them will need to be shown to be the same, just as the account of whiteness is the same in snow and in white lead.  (EN 1096b16–24)

So, whiteness remains univocal irrespective of its subject of predication, and so, presumably, remains univocal when predicated of Marcus (the slave), and the slave (Marcus). This remains true even if we are perfectly content to agree that since a slave is a relative (a pros ti), nothing is a slave without first being a human being (an ousia), and indeed even if we regard substances and relatives as standing in some discernible priority–posteriority relation. In sum, φ-entities can be ordered as prior and posterior, without being prior and posterior qua-φ. Consequently, if this approach correctly explicates Aristotle’s appeal to the categories in this connection, then he appears guilty of either conflating subjects of predications with the predications of which they are subjects or of wrongly presuming priority–posteriority relations obtaining between subjects of predictions to determine supervening priority–posteriority relations among their predications.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

136  Christopher Shields One might grant this, and still hold out hope for Aristotle’s argument. For the case of goodness may seem more complicated. If we say (i) Marcus is good, we seem to say something unlike what we say when we say (ii) the slave is good. We may, that is, implicitly be presuming that the predicate ‘. . . is good’ in these applications functions in the manner denominated by Geach as implicitly attributive. According to Geach, every sentence of the structure: ‘x is good’ is implicitly an abridged sentence of the form ‘x is a good φ’ where φ is some sortal or other. Thus, ‘a is good’ when said of Abe  the baker, means, assuming various contextual facts, ‘Abe is a good baker,’ from which we could not infer that Abe was a good father or a good driver or, most generally, a good man. In Geach’s view this is a completely general phenomenon. Indeed, in Geach’s rather extreme appeal to the attributive–predicative distinction, the predicate ‘. . . is good’ is always spurious: ‘There is no such thing as being just good or bad, there is only being a good or bad so and so.’8 There is no such thing, that is, as a purely predicative ‘. . . good.’ If something like this contention is underlying Aristotle’s reservations about the univocity of goodness, then (on the assumption that there is something to Geach’s contention9), turn-about would now be fair-play. For Aristotle might now reasonably protest that it is the Platonist who only haphazardly generalizes; perhaps it is true that whiteness predicates univocally of the categories, but this would by itself give us no reason to suppose that goodness follows suit. On the contrary, if all goodness is attributive goodness, then all goodness would be sortally bound, such that the full range of intrinsic goods, however many they may be, will be good in distinct ways. On this assumption, there simply would be no general predicate goodness whose essence-specifying account might prove upon analysis, as the Platonist presumes, to be univocal. In this goodness would contrast with whiteness, such that whiteness might prove univocal when predicated across the categories while goodness simply could not be so predicated—there being, in point of fact, no such predicate. Although some of Aristotle’s remarks may tend in this direction (so, e.g., EN 1097b), we cannot say that he has drawn Geach’s distinction in any clear or direct way. Be that as it may, we seem in the face of such a contention to have reached an unproductive stalemate. If Aristotle, or an Aristotelian (or Geachean), wishes to insist that there is no such predicate as ‘. . . being good’ and so, trivially, no such univocal predicate as ‘. . .  being good’, then the Platonist seems well within her rights to demur. At the very least   Geach (1956, 65).   This seems to me a most dubious assumption. Geach offers in effect just four arguments: (i) The words ‘good’ and ‘bad’ function like an alienans adjective, but an adjective can be alienans only if it is attributive. (ii) In predicative cases, one can detach the predicate and deploy it in inference. For example: If she has an old red car, then she has a red car; but iff he is a good thief, it does not follow that he is good. (iii) Conversely, one can ‘pool’ information when the predicative use is in view. For example, if her car is red and old, then she has an old red car. In contrast, if he is basically good but occasionally a thief, we cannot infer that he is basically a good thief. (iv) Finally, if ‘good’ were predicative then it would be hopelessly homonymous. None of these arguments succeeds, and the last one seems simply hard to fathom. Pigden (1990) offers a thorough critical discussion, which goes a good distance to undermining Geach’s arguments. 8 9

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  137 nothing in Aristotle’s anti-Platonic argument of Nicomachean Ethics i 6 offers any ­compelling or even clear support for such a contention. In view of that, although the Platonist could not continue blithely to assume univocity neither should she capitulate. What is needed, then, is some way to adjudicate disputed contentions regarding univocity for the range of intrinsic goods. Neither side, at this juncture, has an easy way forward. Taking that all together, we can see that the second premise of Aristotle’s simple anti-Platonic argument, that good things are related as prior and posterior, is at best inconclusive. Without some ancillary support, it hardly merits consideration. With the ancillary support it may reasonably expect from the doctrine of categories, this premise does raise some provocative issues. It does not, however, settle these issues decisively in Aristotle’s favour. On the contrary, the best thing that can be said so far on its behalf is that it raises some issues which the Platonist might fairly be asked to address. This finding may seem so far underwhelming, since one would like to move more rapidly into a position where the argument from a series of goods might be more finally appraised. If apt, however, this finding shows that any such final appraisal is premature. Indeed, one might note, any such appraisal would be not only premature but in fact woefully so: despite the argument’s extreme simplicity, we have so far managed to consider just one of its two premises. Moreover, the premise so far considered, in stark contrast to the first premise of the argument, is itself relatively simple. That premise, recall, is far more general, and not meant to involve any idiosyncratic features of goodness as such: where things φ1, φ2 . . . φn are related as prior and posterior (proteron kai husteron), there is no Form φ-ness set over them. Recall further that this is a premise to which the Platonists, as Aristotle represents them, have already acceded. Why might they have done so? Here Nicomachean Ethics i 6 provides not even a hint of an answer. Fortunately, a partially parallel passage in Eudemian Ethics i 8 goes a good deal further.

III.  Series, Before, and After In the fuller discussion of the Eudemian Ethics, Aristotle makes several points omitted from the more compact statement of the argument in Nicomachean Ethics i 6. The text is unfortunately in a somewhat distressed state,10 but it is none the less clear in its essential contentions: Further, in things having an earlier and a later, there is no common element beyond, where this is separable from them, for then there would be something prior to the first; for the common and separable element would be prior, because with its destruction the first would be destroyed as well; e.g. if the double is the first of the multiples, then the universal multiple cannot be separable, for it would be prior to the double . . . if the common element turns out to be the Idea, as it would be if one made the common element separable: for if justice is good, and so also is   On the state of the text see Susemihl (1884) and Kenny (1978).

10

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

138  Christopher Shields bravery, there is then, they say, a good per se—for which reason they add per se to the general definition; but what could this mean except that it is eternal and separable? But what is white for many days is no whiter than that which is white for a single day; so the good will not be any more good by being eternal. Hence, the common good is not identical with the Idea, for the common good belongs to all.  (EE 1218a1–15)

The statement of the argument here is noteworthy for its contrasts no less than for its similarities to the parallel passage in the Nicomachaen Ethics. Most importantly, here Aristotle makes some effort to explain what is meant to be problematic about there being a Form postulated over a series. Although unlike the previous passage there is no  representation of the first premise as ad hominem, we now, crucially, are given some grounds for supposing there are no Forms over series: a common Form, if separate (chōriston) from the entities over which it is set, would, precisely because separate, ‘be something prior to the first’ element of the series. Indeed, the first premise of the argument is not stated expressly at all; but it is clear that Aristotle has the same premise in view.11 The argument given on its behalf is tantalizing, but once again a bit elusive. Its exact structure is, unfortunately, more suggested than stated. One may begin by noting three discrete contentions not mentioned in the previous passage but introduced overtly here: the Idea is described (i) as something common (koinon), and (ii) as something beyond or beside (para tauta) its instances, where (iii) this is something separable (chōriston). As Fine has noted carefully and to good effect in several contexts, the contention that the Form is something beyond or beside (para) is not the same as its being separate (chōriston).12 The latter, but not the former, implies a capacity for independent existence. In Aristotle’s idiom, being beside or beyond (para) something suggests only that the item so described is not the same as any item in the set or indeed with the entire set taken in its entirety. This is consistent with its being existentially dependent on those items, and so with its not being, in the terms of the version of the argument given in the Nicomachean Ethics, prior in nature (proteron kata phusin) to them. By contrast, the non-pleonastic addition of ‘where this is separable from them’ (or, more simply, ‘and this is separable’; kai touto chōriston; EE 1218a1), adds the thought that Forms are more than merely distinct: they are also prior in nature.

11   Lloyd’s (1962, 70) observation here is apposite: ‘The Eudemian Ethics too argues against an Idea of the good from the premiss that a P-series [scil. a series of entities arranged by some priority–posteriority relation] has no universal separable from its terms. Here the premiss is asserted categorically. The argument is not expressly ad hominem. It is radically elliptical since it omits the whole step of asserting, let alone arguing, that the good forms a P-series.’ 12   See especially Fine (1984) and (1993). Aristotle is quite clear that a core source (perhaps the core source) of difficulty for Plato results from his commitment to separation. So, for instance, in comparison with Socrates, who equally sought out universal definitions: ‘Nevertheless, Socrates surely never separated them from particulars; and in not separating them, he thought rightly’ (Met. 1086b3–5). That he thought rightly, Aristotle insists, can be appreciated by observing how those who do separate universals from particulars, the Platonists, go awry (Met. 1086b3–5).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  139 Aristotle may reasonably ascribe to the Platonists just such a view. Indeed, for our purposes, we may grant that the Platonists hold, just as Aristotle represents them as holding, a full panoply of realist views about the Form of the Good. They hold, on this representation, that the Form of the Good exists; that it is common (koinon) to all good things; that it is set over (epi) the good things there are; that it is univocally predicated of the good things to which it is common; that it is paradigmatic for them; that it is a formal cause of the goodness of the good things it is set over; and, most pertinently in the present context, that it is separate (chōriston) from them, and so capable of existing without them, though they are dependent upon it for their existence as good things (cf. Met. 1078b16, 1086a25). They hold, that is, that the Form of the Good is a non-natural, non-disjunctive, non-reductive, mind- and language-independent property which is essentially and context-invariantly what it is, namely the good itself—or, as Aristotle chides the Platonists for holding,13 it is goodness itself, goodness in its own right (t’agathon kath’ hauto). Aristotle’s complaint in the present context is simple: even the Platonists deny that any such Form φ exists where a series of φ-things is concerned. What precisely, though, is the problem for series? Although this is an occasional trope of Aristotle’s,14 and although it is clear that at least in some cases he represents the Platonists as ­concurring in this judgment, it is not clear why they should—or, indeed, in terms of the surviving Platonic corpus, that they in fact do. Let us, though, waive this last consideration, and assume, as seems suitable, that Aristotle is here representing them fairly and accurately. On this assumption, we may ask: what is the precise problem about a series of goods? What, in particular, does the Eudemian Ethics add to the picture given by the Nicomachean Ethics? Here A.C. Lloyd offers a reasonable first conjecture: . . . we are provided with what the ad hominem argument of the Nicomachean Ethics did not and did not formally need to provide, a demonstration of the general premiss about P-series [scil. a series of entities arranged by some priority–posteriority relation]. It is this. In such a series one of the terms is first in the order of priority (for example, the double in a series of multiples); but if there were a separate Idea (say, multiplicity) which embraced this series it would be prior to all the terms, so that what had ex hypothesi been first would no longer be first. The simplicity of this demonstration must not make us suspect it: it is perfectly valid, since according to the theory of Ideas the Idea would have to rank as one of the terms of the series.  Lloyd (1962, 70)

 See EN 1096a33–1096b2.   He advances similar considerations regarding: the better and the worse (EN 1096a19–35; EE 1218a1–8); the capacities of the soul (DA 414b19–415a11); citizens and constitutions (Pol. 1275a34–8); and, in a passage to which he draws our attention in EN i 6, to numbers (Met. 999a6–10). We do not come away with a clear notion of the sort of series here envisaged, nor even of the sort of priority–posteriority relation he (or he, together with the relevant Platonists) have in view. On the contrary, we have a sort of motley. 13 14

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

140  Christopher Shields Here, as Lloyd notes, Aristotle suggests that if a Form φ were both beyond (para) and separate from (chōriston) the members of a φ series, then ‘there would be something prior to the first’ member of the series. Lloyd is content that this much provides ‘perfectly valid’ demonstration. He does not reproduce this demonstration, but seems to have something of the following expansion in view: 1. If a, b, c . . . n are arranged as prior and posterior, then there is a φ set over them only if it is prior to the entire series. 2. If φ is prior to the entire series, then it is also prior to each of its members. 3. If (2), then φ will be prior to a, the first member of the series. 4. If φ is prior to a, then a, the first member of the series a, b, c . . . n will be posterior to φ. 5. If a, the first member of the series a, b, c . . . n, is posterior to φ, then a will not be the first member of the series. 6. So, if there is a Form φ over the series a, b, c . . . n, a will and will not be the first item in that series. 7. So, there is no Form φ set over the series a, b, c . . . n. To this we may add the categorial appeal of Nicomachean Ethics i 6, thereby applying this result to goodness. 8. The categories of being form a series arranged as prior and posterior. 9. If (8), then if goodness is predicated across the categories, predications of goodness will form a series. 10. So, there is no Form φ set over the categories. (7, 9) 11. So, there is no single Form, Goodness, set over the series of goods predicted across the categories. It may be that Lloyd has a different argument in mind, but this seems to be what is suggested by his observation ‘that what had ex hypothesi been first would no longer be first.’ It behoves us to lay out this expanded form of the argument, since—if this is the argument Lloyd has in view—we can now see that Lloyd is wrong to contend that ‘[t]he simplicity of this demonstration must not make us suspect it.’ While a simple, valid argument would be most welcome in the context, it takes little effort to see that the argument here expanded is neither. The question of simplicity is partly a function of its clarity and directness, but we can see that we have already come a good distance from the simple two-premise argument of the Nicomachean Ethics. As for validity, the matter is more complex. It would, perhaps, be useful if some such argument could be shown valid, because then we would have a good reason to see why the Platonists might have agreed, as Aristotle represents them as having agreed, to there being a problem about Forms for series. After all, the strong implication seems to be that this would be a special problem, one attending to members of a series, or more broadly for members standing in

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  141 priority–posteriority relations, which is not also a problem for level-specific kinds, that is, kinds like various species under the genus animal. Thus, for instance, if nothing which is a human is any more or less a human than any other human, then we would expect the Platonist to posit a Form of Humanity, arrayed over those humans, even if, due to the reasons currently under exploration, she demurred for level-ordered kinds, like numbers. One virtue of the expansion suggested by Lloyd is precisely that it has this result: it reveals, if sound, a problem for Forms for series which is not also a problem for level-specific kinds. It locates a contradiction by making the (putatively) first member of a series both first and not first in that series. Unfortunately, the argument as expanded has no such consequence: it is invalid. The problem owes to an equivocation on the notion of priority. One notion of priority we may term sequential-priority. On this notion, priority is given by a sort of successor function, such that, for instance, any given natural number is prior to the one which follows it, and then, by transitivity, to every subsequent number in the series. This is a far cry from the notion of priority to which Aristotle has appealed in the Nicomachean Ethics, namely priority in nature (kata phusin), which we have glossed in terms of an asymmetrical implication of existence. More to the point, the notion of priority in nature is at play in the first three premises of our expanded argument. This, then, must the sense of (4), in which the first member of a given series will be (along with every other member of the series) posterior to the postulated Form set over them. The notion of priority required for (5), however, shifts abruptly to sequential priority: nothing can be first in a series if it is sequentially prior to another member of the series. Nothing, however, precludes a first member (like any other member) of a series from being posterior in nature to a postulated universal set over it. Given, then, that we have distinct notions at play in the argument—neither of which entails the other—the expanded version equivocates, and so turns out invalid. A Platonist may then say that the putative contradiction recorded in (6) is simply false, so that there will be no contradiction involved in the postulation of a Form for series, as (6) concludes: so, if there is a Form φ over the series a, b, c . . . n, a will and will not be the first item in that series. The Platonist will now rightly reject (6), because a will after all be the first member in the series. This first member of the series a, b, c . . . n will indeed be posterior to the postulated Form set over the series; but this posteriority will be posteriority in nature (kata phusin). Hence, to avoid a manifest equivocation, (4) will now have to read: (4*) If φ is prior in nature to a, then a, the first member of the series a, b, c . . . n will be posterior in nature to φ. and (5) will turn out false. (5) holds that if a, the first member of the series a, b, c . . . n, is posterior to φ, then a will not be the first member of the series. For a can be posterior in nature to φ, even while a is the first in a sequence ordered in terms of sequential priority. It follows then that (6) cannot be derived: it is not the case that if there is a Form φ over the series a, b, c . . . n, a will and will not be the first item in that series.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

142  Christopher Shields In any event, this is what a Platonist can and indeed should say in response to the expanded argument of the Eudemian Ethics. There is simply no contradiction derivable from the commitments ascribed to the Platonist thus far; and there is thus no reason for her to concur with the judgment that there is a special problem about goodness. Indeed, the arguments offered thus far invite rejoinders on all fronts. First, good things do not seem to be ordered in any discernible sequential-priority order. In this, good things do not seem relevantly like the numbers series to which Aristotle likens them. Indeed, they seem, on the contrary, unlike the number series in this respect. Second, any ­priority–posterior relation which may seem to obtain between members of a sequence and a Form φ set over them fails to generate anything approaching a contradiction forcing the abandonment of Platonic realism as characterized. So far, then, Aristotle’s polemics seem unjustified. Or, more mildly, nothing justifying Aristotle’s confidence regarding the security of his anti-Platonic conception of goodness is readily ascertainable from the texts in which he articulates them.

IV.  A Conjectural Reconstruction This result may seem in various ways unhappy. If Aristotle’s polemics are hollow, then perhaps his anti-Academic postures about goodness should simply be set aside. Perhaps, then, we should revert to the earlier finding of Cook Wilson, who concluded that the discussion of Forms in Eudemian Ethics i 8 ‘is an entire misunderstanding’.15 Others have been more strident still. In the nineteenth century, speaking of the parallel discussion of Nicomachean Ethics i 6, the Aristotelian commentator Alexander Grant protested: ‘Everyone has felt the unsatisfactoriness of these arguments; they seem captious, verbal, unreal, and not to touch the point at issue.’16 The point at issue in our discussion is this: why do series present a special problem for good things? As we have seen, this issue breaks down into a variety of sub-issues, each involving its own constraints. On the one hand, we would like to respect the report that the Academics themselves acknowledged a special problem about Forms set over series. It seems prima facie unlikely that Aristotle would make up such a report out of whole cloth. Indeed, the tenor of the argument of Nicomachean Ethics i 6 seems to be just this: given that the Academics have already agreed to there being a problem about Forms set over series, they need now only to be shown that predictions of goodness stand in a sufficiently clear priority–posteriority relation to qualify as relevantly series-like. So, it would be good to accept Aristotle’s report as basically apt. Still, on the other hand, we would not wish to determine in advance that the Platonists faced an insurmountable objection in this regard, especially as regards the univocity of 15   Wilson (1904, 256 n. 1): ‘Eudemus’ attempt to explain this case in Eud. Eth. I. viii, where he reproduces Nic. Eth. I. vi.1096a 17 sqq., is an entire misunderstanding.’ (Wilson is here assuming, as was common in his day, that the Eudemian Ethics was not by Aristotle, but by Aristotle’s pupil, Eudemus of Rhodes.) 16   Grant (1885 vol. I, 208). Compare Ackrill (1972).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  143 goodness. Surely, in any case, we are not entitled to any such conclusion before coming to terms with the argument to which Aristotle is alluding but not reproducing. We are in effect invited to peer into the Academy to witness a discussion we cannot hear. Without the benefit of this information, we are unlikely to come to determinate answers about the force of Aristotle’s criticisms. That acknowledged, we might yet follow the example of Gail Fine, whose methodology in On Ideas: Aristotle’s Criticisms of Plato’s Theory of Forms provides a way forward. It may yet be that we have the resources in Aristotle’s texts to offer a conjectural reconstruction of his objection, without being forced to embrace the wholly disparaging conclusions of Grant or Wilson. Rather, following in Fine’s footsteps, we may yet discover a plausible but non-devastating argument against Plato, one which registers a legitimate philosophical objection, which is yet also one to which Plato may have the resources to reply. The reconstruction to be offered is in one way no more conjectural than anything we have so far encountered; yet it too finds some mooring in the text of Eudemian Ethics i 8; and it too takes seriously the suggestion of Nicomachean Ethics i 6 that any postulated Form would need to be prior in nature to the entities over which it is set, in precisely this sense: the Form φ could exist without there being any φ-things, but there could be no φ-things were there no Form φ. So much conforms precisely to the version of Platonic Realism we have characterized and which seems, on the balance, the understanding of Platonic Forms shared by both Plato and Aristotle.17 More to the point, it reflects the fact that in the Eudemian Ethics, no less than in the Nicomachean Ethics, Aristotle has a notion of priority in nature (kata phusin) in view: ‘for the common and separable element would be prior, because with its destruction the first would be ­destroyed as well’ (EE 1218a3–4). Taking these commitments together with three pregnant remarks, one from each text, we find a plausible but non-devastating argument emerging from our texts. The pregnant remarks are these. First in Nicomachean Ethics i 6 we find Aristotle concluding that there is no Form of Goodness such that it is ‘something common, universal, and one’ (koinon ti katholou kai hen; EN 1096a28). In Eudemian Ethics i 8, we have Aristotle characterizing the Form of the Good as ‘common’ (koinon), ‘beside’ (or ‘beyond’; para), and separate (chōriston) (EE 1218a2–3). Finally, third, the Form of the Good is itself good (‘they add itself (auto) to the common account’; to oun auto prokeitai pros ton logon ton koinon; EE 1218a11). We have, then, the following constellation of characterizations, emerging from the two texts taken in tandem: • Commonality: a Form Goodness itself is set over all good things. • Univocity: this common form is single, such that it provides the non-disjunctive, essence-specifying account of what it is to be good. 17   Fine (1983) provides a very clear, textually informed account of the points of agreement between Plato and Aristotle as regards form and substance.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

144  Christopher Shields • Together, Commonality and Univocity entail a one-over-many thesis: that there is exactly one Form of Goodness set over all good things. • Priority in nature: The Form of Goodness could exist without there being any good things, but there could be no good things were there no Form of Goodness. • Essential self-predication: The Form of the Good is itself good in its own right, or good per se (kath’ hauto). With these resources, one can develop an argument which fits the basic requisites of an adequate understanding of the dialectic between Plato and Aristotle on this point. The conjectural reconstruction is, then, this: 1. If a, b, c . . . n φ-things are arranged as prior and posterior, then there is a φ set over them only if (a) it is prior to the entire series (by priority in nature) and (b) it is itself φ (by essential self-predication). 2. If φ is itself φ, and all φ-things are arranged in a series, then φ is a member of that series of φ-things. 3. So, if a, b, c . . . n φ-things are arranged as prior and posterior and there is a φ set over them, then φ is a member of the series prior to the first member of the series. 4. There is exactly one Form of Goodness set over all good things. 5. So, if φ is a member of the series prior to the first member of the series, then there is a φ* over the series φ, a, b, c . . . n. 6. If φ* is itself φ, then it is a member of the series φ, a, b, c . . . n prior to the first member of the series. 7. By repeated application of (6), there is forever another member the series, which is itself a member of the series. 8. So, there is a Form φ set over the series of φ-things only if there is forever another member of that series. 9. The categories of being form a series of good things arranged as prior and posterior. 10. So, there is no single Form of Goodness itself set over the categories of being. In sum, as Aristotle concludes, there is no Form of Goodness such that it is ‘something common, universal, and one’ (koinon ti katholou kai hen; EN 1096a28). Each projected common universal generates another in its wake. So, there will not be one, but, just as a critic of the theory of Forms says in the Parmenides, ‘indefinitely many’ (apeira to plēthos; Parm. 132a10–b2). So much, of course, would also be a familiar Aristotelian complaint against the theory of Forms. Note too that it no longer implicates Aristotle in the obvious equivocation that we have identified in Lloyd’s reconstruction. That equivocation involved two senses of  priority, one the sort Forms bear to the instances arrayed under them (priority in nature), and the other the sort the number three bears to the number four and all subsequent numbers (sequential priority). The point is rather that self-predication

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  145 implicates the Form into the series itself, thereby generating an infinite regress. This it does in effect by using the Form’s self-predication to force it into a sequential priority which it would not otherwise be thought to have. That is, if we are Platonists and are otherwise not inclined to treat the Form of Number as itself a number alongside the other numbers, then we will none the less be constrained to do so, if we also treat it as self-predicative and then also assume, sensibly enough before Cantor, that anything which is a number must be located somewhere on the number line, before all the numbers which follow it and after all which proceed it, if any. If we think that no Form can be so located on the number line, then that is only a reason for doubting that the Form of Number, if there is such a Form, is self-predicative. So, as far as this argument goes, Plato is not jockeyed into any immediate contradiction. Instead, an unpleasant if familiar regress looms. If this is correct, then we are pushed in other directions when we come to evaluate the argument. We may set aside discussion of (9), which concerns the issue already assayed of how or whether good things are arrayed as by some priority–posteriority relation. This reconstruction, like those which have already been attempted, has various merits and demerits. On the one hand, Aristotle does not allude to a regress in either of our passages. Still, each of the core premises and assumptions in this argument is stated in our passages, and in this reconstruction, we do end up with an anti-Academic polemic which is both familiar and formidable. It is familiar, in that—as has already been noted— it bears a family resemblance to the third-man argument of the Parmenides, which argument Aristotle himself discusses several times over (Met. 990b17, 1039a2, 1079a13; Soph. Ref. 178b36). So, here at least, we have some trace of a live intra-Academic debate, one which we know exercised both Plato and Aristotle, if in different ways. What we do not know, if we may indulge this conjecture as a hypothesis, is this: why should the existence of a series make a discernible difference? One thought would be that the third-man argument has always, in different ways, traded upon a precarious interaction between three theses, namely a one-over-many thesis, to the effect that over any group of φ-things a, b, c . . . n there is a φ set over them; a uniqueness thesis, to the effect that there is exactly one Form φ set over the φ-things; and some version of self-predication, such that the Form φ is itself a φ-thing. The interaction is precarious, because taken some ways, these theses are actually contradictory, whereas taken other ways they are consistent, but difficult to apply such that they generate a vicious regress—especially a regress which is both vicious and non-epistemic.18 Moreover, it is difficult to see, on some formulations of these theses, why there is a regress at all. This is where the presence of a series alters the terrain somewhat. Once a series is in view, with each generation of a new φ-thing there is something new, given essential self-predication, which is itself perforce a member of the series. That is, since everything 18   The literature on this topic is extensive. For an orientation, see Vlastos (1954) and (1955); for a very clear presentation of the alternatives, see Fine (1993); and for an excellent survey of the literature focussing helpfully on the logic of the various formulation of the third-man argument, see Rickless (2007).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

146  Christopher Shields which is a member of a series is essentially so ordered, then in order to qualify as a φ-thing in the first instance, the newly generated φ-thing must itself also stand in the ordering relation. To take an example favourable to Aristotle’s point so understood, if there is a Form of number set over the naturals, then, given essential self-predication, it is itself a natural number. Every natural number stands in some essential ordering relation to every other. Hence, the generated natural must be itself somewhere in the series. There is, then, of course, given the one-over-many thesis, something set over this set of naturals. With each new application of the theses, we see that there cannot be exactly one entity set over the series, but indefinitely many. The consequence of the repeated application is, so to speak, to shove each instance of the φ-thing-itself back down into the series. This φ-thing will be both prior to the other members of the series in the sense of being prior in nature (kata phusin), and it will be sequentially prior to all those which follow it. To be clear, the point here is not that the conjectured reconstruction for series represents an improvement to the third-man argument; rather, it is an argument in the family of the third-man argument, and so one whose provenance we can readily recognize. More importantly, it is a reconstruction which does not trade on any equivocation regarding priority or posteriority. Earlier, the Platonist had an easy rebuttal—one would say a rather too easy rebuttal—to Aristotle’s argument, namely that the Form of the Good might be prior in nature to the other good things without thereby being sequentially prior or posterior to them. There was no contradiction, and this was the core shortcoming of the proposed reconstruction. Here we find no contradiction either, but that is a point in its favour: there is no contradiction. There is none the less a genuine problem for the Platonists, and it is one with which we know they have wrestled.

V. Conclusions By the time we have reached this conjectural reconstruction, we have travelled a good distance from the simple, two-premise argument of Nicomachean Ethics i 6. That simple argument held (i) that there are no Forms set over items arranged in a series; and (ii) that good things are themselves so arranged; with the result that (iii) there is no Form of the Good. When reflecting on the relative complexity of our final conjecture, one may have the sense that we have travelled rather too far, that the meagre texts before us cannot bear the hermeneutical weight we have heaped upon them. This may be so. On the other hand, we do find all of the elements of our final reconstruction in our texts; we find, moreover, in their interaction, an argument worthy of both Plato and Aristotle. This is an argument which should give the Platonists a pause, but it is not an argument which Aristotle can promulgate as decisive against them. He might, though, as he does, fairly report the Platonists as having been conciliatory in response to it, even though they might well, as indeed they should, push back in particular against Aristotle’s contention that good things form a discernible series given by some suitably precise priority–posteriority relation.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

a series of goods  147 As for the complexity of the final formulation: we can hardly expect the Academic dialectic pertaining to the metaphysics of goodness to have been anything but complex, however simple and economical those friends of ours who introduced the Forms meant the Form of the Good to be.*

References Ackrill, J. (1972), ‘Aristotle on “Good” and the Categories’, in Islamic Philosophy and the Classical Tradition, ed. S. Stern, V. Brown and A. Hourani (Oxford: Oxford University Press), 7–25. Fine, Gail (1983), ‘Plato and Aristotle on Form and Substance’, Proceedings of the Cambridge Philological Society 209, 23–47. Fine, Gail (1984), ‘Separation’, Oxford Studies in Ancient Philosophy 2, 31–87. Fine, Gail (1993), On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms (Oxford: Oxford University Press). Geach, P. T. (1956), ‘Good and Evil’, Analysis 17 (2), 32–42. Graeser, Andreas (1997), ‘Platon gegen Speusipp: Bemerkungen zur ersten Hypothese des Platonischen Parmenides’, Museum Helveticum 54, 45–7. Graeser, Andreas (1999), ‘Anhang: Probleme der Speusipp-Interpretation’, Prolegomena zu einer Interpretation des zweiten Teils des Platonischen Parmenides [Berner Reihe philosophischer Studien 25] (Bern: Verlag Paul Haupt), 41–53. Grant, A, (1885), The Ethics of Aristotle, 4th and rev. edn, vols I and II (London: John W. Parker and Son). Halper, Edward (2006), ‘Metaphysics I and the Difference it Makes’, Proceedings of the Boston Area Colloquium in Ancient Philosophy 22, 69–103. Kenny, A. (1978), The Aristotelian Ethics (Oxford: Oxford University Press). Lloyd, A. C. (1962), ‘Genus, Species, and Ordered Series in Aristotle’, Phronesis 7, 68–9. Malink, Marko (2013), ‘Essence and Being: A Discussion of Michail Peramatzis, Priority in Aristotle’s Metaphysics’, Oxford Studies in Ancient Philosophy 45, 341–62. Peramatzis Michail (2011), Priority in Aristotle’s Metaphysics (Oxford: Oxford University Press). Pigden, C. R. (1990), ‘Geach on “Good” ’, Philosophical Quarterly, 40 (159), 129–54. Rickless, S. (2007), Plato’s Forms in Transition: A Reading of the Parmenides (Cambridge: Cambridge University Press). Shields, Christopher (2008), ‘Surpassing in Dignity and Power: the Metaphysics of Goodness in Plato’s Republic’, Philosophical Inquiry XXX, 1–17. Shields, Christopher (2015), ‘Fractured Goodness: the Summum Bonum in Aristotle,’ in The Highest Good in Aristotle and Kant, ed. R. Bader and J. Aufderheide (Oxford: Oxford University Press), 83–111. *  Versions of this essay were presented at Harvard University, the Chicago Colloquium for Ancient Philosophy, and Cornell University. I am grateful to the members of those audiences for their probing questions. I thank also David Brink for his astute and instructive comments on the penultimate draft. The occasion for the presentation at Cornell University was a conference in honour of Gail Fine and Terry Irwin, where I was fortunate enough to receive still more schooling from them both; their humane intellectual guidance has been a mainstay of my personal and professional life from the time of my graduate days down to the present.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

148  Christopher Shields Stewart, J. A. (1892), Notes on the Nicomachean Ethics of Aristotle (Oxford: Oxford University Press). Susemihl, F. (1884), [Aristotelis Eudemia Ethica] Eudemii Rhodii Ethica (Leipzig: B. G. Teubner). Vlastos, G. (1954), ‘The Third Man Argument in the Parmenides’, Philosophical Review 63 (3), 319–49. Vlastos, G. (1955), ‘Addenda to the Third Man Argument: A Reply to Professor Sellars’, Philosophical Review 64 (3), 438–48. Wilson, J. Cook (1904), ‘On the Platonic Doctrine of the ἀσύμβλατοι ἀριθμοί ’, Classical Review 18, 247–60.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

8 Practical Truth An Interpretation of Parts of NE VI David Charles

In this essay I shall consider Aristotle’s discussion of practical truth in NE VI.2, 1139a 17–b5. The essay falls into five sections. In the first I shall outline two styles of interpretation of Aristotle’s remarks, in the second show how one of these (which I shall call ‘the third way’) applies to these lines and in the third compare it with the major alternative (which I shall call ‘a two-component’ view). In the fourth I shall consider some texts in the remainder of NE VI which provide additional support for the third way reading. In a brief concluding section, I shall place Aristotle’s view of practical truth, so understood, in a broader philosophical context.

1.  Two Styles of Interpretation Aristotle concludes his discussion of practical truth by commenting (1139b4–5): This is why preferential choice (prohairesis) is either desiderative nous or intellectual desire. . . . 

There are two distinct ways to interpret this remark, which rest on differing views of the relation, in Aristotle’s thought, between desire and perception or intellectual thought. According to the third way (3rd way) interpretation, he holds that: [A]  To see something as pleasant is inseparable in definition from sensuously ­desiring it. [B]  To think of something as good is inseparable in definition from pursuing it. (De Anima 431a8–17) This interpretation makes two claims. First, that one cannot, in Aristotle’s view, define the type of seeing in [A] without reference to desire or the type of thinking in [B] without reference to pursuit (itself a form of desire). The relevant kind of seeing in [A] is, on this view, desiderative-perception, the kind of thinking in [B] desiderativethinking. Second, that one cannot define the type of desire in [A] without reference to

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

150  David Charles seeing the object as pleasant or the type of pursuit in [B] without reference to thinking of the object as good. They may be characterized respectively as perceptual-desire and intellectual-desire.1 What is being denied? On the 3rd way reading, Aristotle does not believe that there is, in the cases just mentioned, a kind of seeing or thinking, defined without ­reference to desire, which co-occurs (perhaps invariably) with desire (itself defined without reference to seeing or thinking). He thinks of the relevant type of seeing or thinking as desiderative-seeing or desiderative-thinking and the relevant types of desiring as perceptual- or intellectual-desire. In this account, seeing (or thinking) could be the genus (or determinable) and desiderative the differentia (or determinant) to yield a distinctive species of seeing: desiderative-perceiving. But equally, for all that has been said so far, desire (or pursuit) could be the genus (the determinable) and perceptual or intellectual as the differentiae to yield distinctive types of desiring: perceptual-desire and intellectual-desire. 3rd way interpretations differ from (what I shall call) two-component accounts in that the latter define one (or both) of desire and perception in [A] and desire and thought in [B] without reference to the other.2 In some two-component accounts, the type of desire in [A] and [B] is defined independently of perception or thought and taken as attending to (or following) one or both of them either by necessity or contingently. However, in all such accounts, the type of perceiving in [A] or the type of thinking in [B] is defined independently of the subsequent desire. They are not defined as desiderative-thinking or desiderative-perceiving.3 There are two different ways to develop a 3rd way account of preferential choice. In one, desiderative nous and intellectual desire share exactly the same definition. There is just one type of state, albeit one which can be thought of in different ways (as, in some accounts, one geometrical shape can be thought of either as convex or as concave).4 In the second, even though desiderative nous and intellectual desire necessarily co-occur and are both defined with reference to each other, they do not have exactly the same definition. While in the definition of the desiderative nous, nous (or intellectual thought) is the genus (or determinable) and desiderative the differentia (or determinant), in the definition of intellectual desire these roles are reversed. On this view desiderative nous and intellectual desire will be distinct types of state which necessarily co-occur 1   A full defence of this viewpoint rests on an analysis of texts in De Anima and De Motu. See my ‘Aristotle’s Desire’ Mind and Modality: Studies in the History of Philosophy in Honour of Simo Knuuttila, ed. V. Hironen, T. Holpaianen, M. Tuominen, Leiden, 2006). 2   Nearly all twentieth-century writers on this topic (with the possible exceptions of Ryle (discussed below) and Heidegger) are two-component interpreters: Greenwood, Ross, Hardie, Irwin, McDowell, Wiggins, Broadie, Bostock, Reeve, Segvic, Lorenz, Moss, Olfert, my earlier self, and many others. The two-component approach sets up the debate between those for whom Aristotle’s ethical theory is intellectualist (whether intuitionist or rationalist) and those for whom it is Humean or (more generally) sentimentalist. 3   Desiderative-perceiving, so understood, is not defined as a type of perceiving (defined independently of desire) which (efficiently) causes (or is caused by) desiring (defined independently of perception). 4   In Aristotle’s terms, things can be separable in thought even if error results (Phys. 191b33–4): examples of this approach are given by Meta Z.3 and 11.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  151 but can only be defined with reference to the other. Their relation would be akin to that between being concave and being convex, if these were understood as different states. On either view, Aristotle could describe desiderative nous and intellectual desire as being (in some way) the same, a unified state, even though (in his terminology) the type of sameness involved will differ.5 In neither will desiderative nous be the efficient cause or effect of intellectual desire. In this essay, although I argue for a 3rd way interpretation of preferential choice, I shall not attempt to determine which of these two versions best captures Aristotle’s remarks in Nicomachean Ethics VI. Indeed, I am inclined to suspect that he did not commit himself to one or other account in this book, even if he did so elsewhere in his scientific writings.6

2.  A 3rd Way Reading of 1139a17–b5 In sketching a 3rd way interpretation of these lines, I shall consider each of the sections of this passage one by one, giving a 3rd way reading of each in turn. [1] There are three things in the soul responsible both for action [praxis] and truthfulness [alētheia]: perception, intellect, and desire. Of these, perception is not in control [an archē] of any action [praxis]. It is clear that wild animals have perception but do not share in action [praxis]. (1139a17–20). Aristotle considers perception, intellect, and desire as responsible both for action (praxis) and truth.7 They all play some role in both. Alētheia (often translated as ‘truth’) should be taken to mean hitting or grasping the truth: truthfulness, in Aristotle’s own terminology (see NE 1108a20). It is ‘correctness of belief ’ (1142b11). It consists in meeting or conforming to the norm of truth and is not the norm itself: truth. These three psychological states are responsible (when they function well) for grasping the truth, not for the truth they grasp. While perception may be responsible for my seeing the storm approaching, it is not responsible for its being true that the storm is approaching. Nor is what makes it true that the storm is approaching.8 5   For discussions of different types of sameness, see Meta 1003b22ff. In the first case, they will be one in account, in the second not. On this topic, see my ‘Aristotle’s Processes’, Aristotle’s Physics: a Critical Guide, ed. Mariska Leunissen, Cambridge 2015, pp. 196ff. 6   Elsewhere in the Ethics, Aristotle suggests that he will not seek exactitude on such ontological issues in his ethical writings (1102a30ff.). 7   I shall take ‘praxis’ to refer to ethically significant actions throughout. While wild animals clearly act in some way, they do not engage in praxis (so understood). In the present passage Aristotle is concerned with the virtues of the rational faculty (logistikon: 1139a14–16) and resulting actions, not with all cases of action. 8   Nor is perception responsible for there being a true proposition that I grasp to the effect that the storm is approaching. While perception or thought may be responsible for there being a proposition that I grasp (Meta 1027b33ff.), they are not responsible for the proposition I grasp being true. The proposition in question is not true because I think [or perceive] it to be true. Indeed, its being true is what makes it the case that the proposition I grasp is true.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

152  David Charles Perception, however, is excluded from further discussion because it is not a controller (archē) of action (praxis) although presumably it is a controller of truthfulness.9 Aristotle now seeks for one phenomenon which controls (archē) both action and truthfulness. Desire and intellect are the only candidates to be considered for this demanding (‘two jobs’) role. If desire controls our grasping the truth in the same way as perception and thought do, it—like them—will be truth assessable. However, it is not required (by this sentence alone) that this be so. Perhaps it does not control truthfulness in precisely the same way. Maybe, unlike thought and perception, it controls only whether or not we are in some further state in which we grasp the truth. So, [1], although consistent with a 3rd way reading, does not require it. The next lines, however, offer more direct support for this style of reading. [2] The very thing assertion and denial are in the intellect, this pursuit and avoidance are in desire. As a consequence, since ethical virtue is a state in which one is disposed to choose preferentially, and preferential choice is deliberative desire, it is necessary on these very grounds10—if the choice is to be excellent—for the reasoning [or argument] to be true and the desire correct, and the former to say what the latter pursues. The first sentence of [2] states (somewhat elliptically): That which (in the practical intellect) is just what assertion is,11 this (in desire) is (just what) pursuit is.

One way to understand this sentence is as follows: There is something which is just what assertion is (a type of practical intellectual state) and is also pursuit (a type of desire).

and so as equivalent to: Pursuit (a type of desire) is the same as assertion (a type of practical intellectual state).

From this it will follow that: There is a type of desire which is the same as a type of practical intellectual state.

and Pursuit is the same as assertion (when understood as a state of practical intellect). 9   Aristotle’s actual argument is weak. He suggests that perception is not the controller of action (praxis) because wild animals perceive but do not engage in actions (praxeis). However, since wild animals have desire, he could equally well have concluded that desire is not the controller of action! He is claiming, it seems, that perception by itself (even human perception) is never the controller of action (praxis). Something more needs to be added. (The same could, no doubt, also be said of desire.) 10   The text is in dispute here: ‘dia tauta men’ is Bywater’s conjecture. LO read ‘dia men tauta’ and M ‘dia tauta’ while K reads ‘de tauta men’, which makes no sense. I am inclined to read either ‘dia tauta’ or ‘dia men tauta’ (with ‘men’ as a case of a solitary ‘men’ used for emphasis: Denniston, p. 359). 11   Intellect should (in this context) be confined to the practical intellect: the logistikon (1139a12), the type of intellectual activity (nous: 1139a18) which focuses on action.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  153 This sentence, so understood, expresses Aristotle’s commitment to one version of the 3rd way account. To think of something as good/to be done is the same in definition as to pursue it.  (431a6–9)

There is one type (or species) of phenomenon which is common to thought and desire, which can be thought of either as a type of thought or a type of desire. (This is the more demanding form of 3rd way interpretation.) However, there is, as noted above, a less demanding alternative: Aristotle may be claiming only that assertion and pursuit cannot be defined without reference to each other without insisting that they have exactly the same definition. If desiderative nous and intellectual desire belong to different genera, they are inseparable, but not the same, in definition. The first account requires the stronger claim, introduced without support in this context. Aristotle, some may suggest, is calling on statements he has made elsewhere. Perhaps he is relying on the view taken in De Anima 431a6–9 (already cited). Some may point to the complex discussion in De Anima 433a22–26 or to De Motu 700b22ff., where preferential choice is said to be the common property of thought and desire. However, since none of these passages is explicitly invoked in the present context, it seems best to retain both interpretations as possibilities at this point. Not all desires, it should be noted, will be of the type specified: there can be desires (e.g. sensuous desires) which are not identical with any assertion of the practical intellect. Nor are all assertions involving value desires. There can, it seems clear, be assertions about value without desire being involved, such as assertions concerning the remote past. (See also De Anima 432b27ff.: perhaps this is true also of ‘understanding’: 1143a9–10.) At issue in the present passage are the assertions of the practical intellect, ones which lead to action. The first sentence in [2] has been interpreted differently. Some take it as expressing merely an analogy: As assertion and denial are in thought, so pursuit and avoidance are in desire.12 However, it is a disadvantage of this reading that the sentence lacks the trademark signs of an analogy: phrases such as ‘as in . . . so in . . . ’. Nor do analogies standardly have a ‘this’ referring back to a previously identified object.13 Others suggest that the sentence means only: There is a generic type of state instantiated by assertion in the intellect and by pursuit in the case of desire.

On this view, pursuit and assertion will be different species which share a genus. For some assertion and pursuing are both members of the genus: being the last stages   For this reading, see Irwin’ s translation.   Analogies are standardly introduced with ‘as’: ‘as A in B, C in D, and something else in something else’ (see NE 1096b28–30). 12 13

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

154  David Charles in  their respective processes: thinking and desiring. However, it is a disadvantage of this reading that no such genus is mentioned in the present context. Nor do we find the terms used elsewhere to indicate a reference to genus (hoper . . . ti): we have only the term (hoper: 1139a21), standardly used to indicate the essence of the thing.14 Had Aristotle intended to rely on the claim that pursuit and assertion are members of the same genus, there should have been some grammatical indication of this intention or some suggestion as to what the (unobvious) genus is. Given the absence of both, ‘that which (hoper)’ is most naturally taken to indicate one of the two interpretations given above. The next sentence clarifies the issue, affording a better grip on what has been claimed in the lines just considered. Aristotle writes: ‘as a consequence, since ethical virtue is a state which issues in preferential choice and preferential choice is deliberative desire, it is necessary—if the preferential choice is to be excellent—that (1) the reasoning (or argument) be true and the desire correct and (2) the reasoning (or argument) say and the desire pursue the same thing.’ How do these two conclusions follow? On a 3rd way reading, Aristotle’s argument runs as follows: [A]  Pursuit is inseparable in definition from the assertion of the practical intellect. [B]  Pursuit (in the case at issue) is inseparable in definition from deliberative desire. So: (1)  The argument/reasoning involved (in the deliberation) must be true and the desire (present in preferential choice) correct (if the choice is to be excellent). (2)  The true argument/reasoning and the correct desire (present in preferential choice) say and pursue the same things. The first conclusion (1) rests on the idea that the assertions of practical thinking (or  possibly deliberation) involve deliberative reasoning. This together with [B] yields (1). However, (1) does not specify how the relevant reasoning and desire are ­connected.15 What Aristotle needs to support (2) is that true reasoning and correct desire are both engaged with the very same object. Aristotle’s use of the phrase ‘as a 14   For this use, see Meta 1030a3–5, where the essence (to ti ēn einai) is said to be exactly what (hoper) an individual (tode ti) is. Post An 83a24ff. says that predicates indicating essence express just what something (hoper) is or just what it is (hoper) a species of. In both passages, accidental predicates do not say what something is or what species it belongs to. 15   While ‘deliberative desire’ indicates a type of deliberation-involving desire, Aristotle does not mention here any particular way to understand the connection between deliberation and desire. Further, had he been simply relying on the view that reasoning terminates with a judgment that a is to be done which generates a desire to do a, [A] would not have been needed as a premise in the present context (as it seems to be) to arrive at (2). (I am also inclined to interpret ‘deliberative desire’, if this is meant to capture the thesis in NE III. 3, in a 3rd way fashion. But that is another story!)

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  155 consequence’ provides an important clue, indicating that [A] is required to arrive at the combination of (1) and (2). Nor is this surprising: (2) does not follow from [B] alone: for all [B] says is that desire could end with a correct general desire which causes the agent to reason in such a way as to generate some specific act in line with their general desire. Or, alternatively, thought could have stopped with a general conclusion and caused the agent to choose one specific action.16 In the present context [A] is required, as is suggested by the phrase ‘as a consequence’, to derive (2). If assertion and desire are inseparable in definition (in the ways required by either version of the 3rd way account) they must both be concerned with the same action. There is no possibility of their focusing on different types of action (at different levels of generality) or playing differing roles at different stages of practical reasoning. Without a 3rd way reading of the previous sentence (1139a21–2), it is difficult to see how (2) can be derived. Were assertion and desire simply the last stages in their relevant processes, there would have been nothing to require that they finish by focusing on the very same object. They might, as has just been suggested, have ended in states with different objects, causally connected to each other.17 Nor will it do to suggest that pursuit and assertion are of the same genus (in 1139a21–2) because they concern the same objects (or have the same content). There is, of course, no hint of this proposal in  the text. Nor is this surprising: if it were presupposed as a premiss, this would be equivalent to the conclusion (in 1139a25–6) that the same thing is asserted and ­pursued. If so, (2) would not be a consequence of [A] as is suggested by the text but a restatement of the original claim. That the 3rd way interpretation of 1139a21–2 makes good sense of Aristotle’s argument in these lines speaks strongly in its favour and against the alternatives proposed. On the 3rd way view, asserting the conclusion of the reasoning (e.g. that a is to be done or is best to do) is inseparable in definition from pursuing a (desiring to do a). The r­ elevant type of assertion is of a desiderative type, to be contrasted with other types of assertion (theoretical assertion etc.). When desire and the practical intellect are working well, preferential choice will be excellent in that it asserts in a correct desiderative way what is (truly) to be done. One achieves practical truthfulness when one regularly and reliably attains excellent preferential choices. Aristotle goes on to spell this out in the next lines. Preferential choice will be excellent if it asserts in a correct desiderative way what is—in truth—to be done. It will be excellent not simply because it asserts what is—in truth—to be done but because it accepts it in the correct desiderative way. It is important to consider two objections to the 3rd way interpretation of these lines. 16   As in Buridan’s cases, one might finish one’s reasoning with the thought that one should do a or b, and desire take over and lead one to do a. 17  ‘Phasis’ may refer to the particular action chosen (see De Anima 431a7) or to it as the action to be pursued/as good etc.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

156  David Charles [1]  Some will say that had Aristotle been a 3rd way thinker, he would not have talked about correct desire and true reasoning (or argument). A 3rd way thinker, it will be said, would speak (unequivocally) of both desires and reasoning as true.18 There are, however, two reasons for Aristotle’s use of this terminology, both fully consistent with the 3rd way interpretation. (a)  There is more to desire’s being correct than simply its accepting a true claim about what is to be done. In the present context—if the preferential choice is to be excellent—the desire must not simply be for what is, in truth, to be desired. The self-controlled and the akratic also may have preferential choices of this type, although they lack excellent ones. To have an excellent preferential choice requires that one’s desire be of the type possessed by the virtuous rather than the self-controlled or the akratic. There is more to desire being correct in the present case than its being for what is, in truth, to be desired. (b)  Preferential choice is introduced in this passage as a desire (of a deliberative type: 1139a23) and desires (and choices) are characterized in Greek (as in English) as desires (or choices) to do a given type of action. So described, they are said to be correct or incorrect depending on whether their objects are (or are not) the ones to be pursued. However, this grammatical point is fully consistent with Aristotle’s holding that the relevant desire to do a cannot be defined without reference to thinking-in-the-desiderative-way that a is to be done. In the most demanding version of the 3rd way interpretation, the self-same state is both a case of desiderative thinking and thoughtful desire.19 Described in the latter way it is assessed as correct (as a type of desire), described in the former (as a type of thought) it is assessed as true or false. The state itself, described in these different ways, is both truth and correctness assessable. Its fullest specification might be truly thinking-in-the-correct-desiderative-way that a is to be done. [2]  Some may object that, notwithstanding the arguments above, Aristotle must be committed to the presence of two definitionally separate components by his talk of correct desire and true logos. If ‘logos’ means ‘intellect’ or ‘reason’, Aristotle must be envisaging two separate components at work: desire and an additional rational component. 18  In NE 1111b32ff. Aristotle may allow that preferential choice is characterized by the true and the false, although more by being good and bad (taking ‘mallon’ as more often/preferably). However, in EE 1226a4 he does say that ‘preferential choice is not true or false’. In EE II. 10 (unlike VI.2) Aristotle seems to be thinking of preferential choice solely as a choice (1226b7ff.) to do something (walk, sit down, etc. 1226a10ff.) and not as form of nous/intellect. So described, it is clearly not true or false. There is an important question still to be resolved: is preferential choice in the EE to be understood in a two-component way? Or is 1226b7ff., which emphasizes that preferential choice—described as a choice to act—is not true or false, consistent with the state itself being a type of desiderative thinking (and so truth-assessable)? More work is needed on passages such as EE 1226b19ff. and 1227a3–5 to answer this question. 19   There will be one state with differing descriptions: compare De Anima 431a9ff. where one state is both a type of perception and of desire.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  157 This objection can be met: the crucial term ‘logos’ in the present context is most naturally taken as referring to reasoning or argument, not to reason as a faculty of the soul. It is reasoning or argument, not faculties, that are true (or false). The rational faculty may be what *has* correct logos (see 1138b20, 25, 29–30) but is not itself true or false (or correct or incorrect). By contrast, arguments (and reasoning) are true when the premises and conclusion accepted are true. Further, if desire itself can say and accept premises and conclusions of the reasoning (De Motu 700b32), so too can the practical intellect (i.e. the desire-involving intellect).20 There is no need to take what accepts the relevant premises and conclusions to be purely intellectual states, awaiting the injection of desire to generate action (as in the Humean twocomponent account). The 3rd way reading is further supported by the next three sentences: [3]–[5]. [3] Thus (men oun) this is what practical intellect in the sense of (kai) practical truthfulness is. In the case of the theoretical intellect (which is neither practical nor productive), doing well and doing badly is grasping the true and grasping the false (for this is the task of any intellectual capacity). In the case of the practical intellect, its doing well is truthfulness agreeing with correct desire. The first sentence, taking ‘kai’ as clarificatory, refers to truthfulness to clarify the good state of the practical intellect.21 Practical truthfulness is not just the practical intellect doing well (as some have suggested). It is rather the type of truthfulness in virtue of whose achievement the practical intellect does well. Practical truthfulness rests, it seems, on an ordinary notion of truth and a distinct way of grasping that truth: intellectual assertion in a correct desiderative way (the way which leads the agent to act virtuously). What is practical truthfulness? It is, we are now being told, the kind of truthfulness exemplified by an excellent preferential choice: one defined by the presence of true argument and correct desire (the desire previously introduced as the one present in the choice). The presence of this correct desire is part of what constitutes the relevant type of truthfulness. Practical truthfulness is asserting-in-the-right-desiring-way the results of true reasoning. This suggestion spells out the way in which desire (of a given type) is responsible for both truthfulness and action: it simultaneously controls both (as their archē) in virtue of its being identical in definition with (or the same in role as) asserting-in-the-correct-desire-way what is to be done. As such, it is both truth assessable and the starting point of action. It is important to note that the phrase ‘truthfulness in agreement with correct desire’ does not itself require the presence of two definitionally independent components: 20   One can, it should be noted, consistent with the 3rd way view, regard desire-involving intellect and intellectual desire as separate in thought even if not separate in definition. 21   ‘This’ may be taken grammatically to refer forward to the account to be given or backwards to excellent preferential choice: 1139a25. Whichever way the grammar is construed, the inferential phrase ‘thus’ (men oun) must refer back to what is required for an excellent preferential choice: true argument and correct desire saying the same thing.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

158  David Charles asserting the true and rightly desiring it. If asserting that a is to be done is inseparable in definition from desiring to do a, the type of assertion will (on the 3rd way interpretation) be a desire-involving type of assertion. True assertion and correct desire will both be constitutive of practical truthfulness. It is not that practical truthfulness is defined independently of correct desire and—in reality—always attended by it. Instead, one cannot define what truthfulness of this type is (in contrast with theoretical or poietic truthfulness) without reference to correct desire. While the self-controlled could reason truly, they would not attain practical truthfulness because they lack the relevant type of assertion: asserting-truly-in-a-correct-desiring-way. (They may assert truly but would not attain practical truthfulness.) It may be helpful, at this point, to summarize the 3rd way interpretation of these lines: (i) Practical reasoning issues in assertions/thoughts of the form ‘a is to be done’ or perhaps ‘a is to be done as the pleasant/fine/useful thing to do’. (Let us call these practical thoughts.) (ii) Such thoughts are true if and only if a is to be done (ordinary notion of truth). (iii) The practical thought that a is to be done is inseparable in definition from the intellectual desire to do a (The Third Way thesis). (iv) Desires are correct when and because they are directed in the right way at what is to be done (eupraxia). (v) Truthfulness of practical thought is grasping what is—in truth—to be done in a correct desiring way. (vi) This type of truthfulness cannot be defined independently of correct desire and good logos: it is present when correct desire pursues and true logos [reasoning] says the same things. The agent is practically truthful if and only if he/she asserts in the correct desire involving way what is (in truth) to be done. On this account Aristotle relies on the ordinary idea of truth (as introduced in 1139a24 and presupposed in 1139a28) but a distinct type of practical truthfulness: it is a kind of truthfulness that cannot be defined independently of good desire (because the relevant type of assertion/thought is practical thought, inseparable in definition from desire). Why, one might reasonably ask, is—in Aristotle’s account—a to be done? The most straightforward answer is: a is to be done because it is a case of acting well (eupraxia). Its being a case of eupraxia is what makes it the case that this action is to be done. One might call its being a case of eupraxia the fundamental explanatory ground which explains why it is true that it is to be done. If we desire to do a, we will also think in the desiderative mode that a is to be done. So, our desire will be correct only if we think truly in the desiderative mode that a is to be done. The very same thing (a’s being the thing to do) is what makes our assertions true and is part of what makes our desires correct. A’s being a case of eupraxia is what grounds the truth of our assertion that a is to be done and what makes our desire to do a correct.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  159 [4] Thus preferential choice is what controls action (praxis)—the efficient cause not the final—and desire and goal-directed reasoning controls preferential choice. This is why preferential choice is not without intellect in the sense of reasoning and a state of ­character. For doing well (eupraxia) and its opposite in action do not occur without intellect and character. Intellect itself moves nothing, but goal-directed practical ­intellect does. This directs the productive intellect; for each maker makes something for a goal and what is made is not a goal without further qualification (except relative to something else and for something else) but what is done (is a goal without qualification). For doing well is the goal and the desire in question is for this. Preferential choice is the type of state which is both truthful and the controller (archē) of action (praxis). It controls action (praxis) because it is what is to be done. The inferential ‘thus’ (1139a31: men oun) rests on the following line of argument: (1) Preferential choice is what controls truthfulness (suggested in 1139a27–31). (2) There is one type of state which controls truthfulness and action (praxis): 1139a17f. (3) So: preferential choice controls action (praxis). Point (2) is in line with the 3rd way understanding of preferential choice as a distinctively desiderative type of intellectual assertion or a distinctive type of intellectually assertive desire. (Aristotle makes this assumption explicit in 1139b4–5.) In the present context, it is important to note—as Aristotle points out—that the type of intellect involved (practical intellect) cannot be defined without reference to desire for the goal (1139b2: see also De Anima 433a20ff.) and so must be a distinctive type of desire-involving intellect. While Aristotle does not specify here all the ways in which desire is involved in the practical intellect, he has done enough to mark it out as a desiderative type of intellect which (as such) leads to action (although intellect of the non-desiderative, theoretical, type does not). This reference is enough to mark out a distinctive type of intellect, different from theoretical intellect and not decomposable into two definitionally separable components: theoretical intellect plus desire. In sum: the type of (practical) truthfulness possessed by the agent with excellent preferential choice cannot be characterized without reference to desire (for the goal of doing well) or the true reasoning which leads to the true conclusion (asserted in the correct-desiderative way by the excellent agent). [5] This is why preferential choice is either desiderative intellect or intellectual desire, and this is the type of controller a human being is. This unusual expression is exactly what the 3rd way viewpoint would predict: preferential choice can be described either as being a desire-of-an-intellectual kind or an intellectual state of a desiderative type.22 22   For another example of this unusual type of expression, designed to point to two ways of capturing one phenomenon, see Meta 1036b23–4.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

160  David Charles This sentence, so understood, follows from what has preceded it: the type of intellect (as we have seen) is of a desiderative type and the type of desire is of an intellectual (assertive of the true) type. It is not the type of formulation a two-component theorist would use: they would say preferential choice is a combination of desire and intellect, the one accompanying (or perhaps being with) the other. Further, on their view, the relevant type of intellect would be defined independently of desire for the goal (contrary to Aristotle’s procedure in the preceding lines: 1139b2). The type of starting point which characterizes human beings is a distinctive type of desiderative thinking, which we share neither with other animals (who fail to be thinkers) nor with the gods (who do not possess the relevant type of desire that involves ethical virtues: 1178a17–21). Further, to be a starting point a human must be one unified whole. In Aristotle’s preferred way of thinking, we are not such wholes by simply having a combination of intellect and desire. Rather, our intellect affects the nature of the relevant types of desire, making them intellectual desires. These are not the types of desire non-rational animals could have. Conversely, our type of intellect is affected by its essential involvement with desire. Its goals involve desire (and so—in the best case—ethical virtue) as does its distinctive form of practical reasoning in which desire takes the place of thought in asserting the relevant claims (see De Motu 700b31f.). This type of intellect (desiderative intellect or practical intellect) is not something we share with the gods.

3.  Two-Component Accounts and Other Alternatives On the 3rd way reading desiderative-assertion and assertive-desire are both truthassessable and responsible for action. The notion of truth is Aristotle’s standard one: what is asserted-in-the-desiderative-way as what is to be done is, in truth, what is to be done. However, truthfulness (the best condition of such a state) consists in asserting what is, in truth, to be done in the correct desiderative way. This way not only (we may assume) reliably leads to action but to action done in the way the virtuous person acts. There is one unified type of assertion (desiderative-assertion), not an amalgam of two distinct definitionally independent components. According to two-component theories there is a type of thought or assertion defined independently of any connection to desire which is accompanied by desire. There is not one type of state which is both truth assessable and a type of desire. (This will be true even if the two types always co-occur and are necessarily co-instantiated by the same ‘token’ states.) I shall consider some versions of the two-component story. [A]  One version of two-component theory runs as follows: in cases of excellent ­preferential choice the assertion is true, and the desire correct (but not true). Truthfulness consists only in the assertion’s being true. The desire has to be correct, but its correctness is a matter of its leading to good action. Desires are correct (in this version of the two-component account) because the actions they generate are ones which are, in truth, good.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  161 This version of two-component theory, in effect, retains the ordinary notion of truth (as I have done) but adds that the same type of truthfulness is to be found in the ­theoretical and practical case (albeit in different domains). There are, however, several problems for this account: (i) The goal of the practical intellect is, it appears, the same as that of good desire: doing well (eupraxia). Aristotle spells out the connection clearly: the goal of practical reason is one which is desired. Indeed, it is a goal of such reason because it is desired. The thought and assertion involved in practical reasoning are desire-involving. They cannot be defined without reference to desire. (ii) Practical truthfulness is to be contrasted with the truthfulness characteristic of theoretical and productive intellect. But, on the present account, they are fundamentally the same because the truthfulness of the practical intellect is defined independently of desire. (Both productive and practical intellects, it should be noted, will be concerned with what can be otherwise.) [B]  A subtler version of the two-component account runs as follows: practical ­truthfulness is distinguished in the following way: what makes the assertion true is the same thing as what makes the desire correct. What makes the assertions true is that they state truly which actions it is good for humans to do. What makes desires correct is that, guided by true assertions of this kind, they lead us to act well. Indeed, on this view, what is distinctive about practical truth is that once grasped it generates desires which lead us to act well.23 But this account is also problematic. First: on this account the assertions of the self-controlled and the virtuous are both possessed of practical truthfulness. Both assert truly that the actions are good/to be done which are the object of correct desire. But the self-controlled do not achieve the good state of the practical intellect: practical wisdom.24 Since the good state of the practical intellect is truthfulness, they do not attain truthfulness. Hence, there must be more to truthfulness than assertions made true because the actions they commend are the objects of correct desire. Some might seek to address this concern as follows: it is not enough for practical truthfulness that the assertions are made true because the actions they say are good are the objects of correct desire. Practical truthfulness requires also that the assertion made true in this way is actually accompanied by a correct desire to do it (which the self-controlled, we shall assume, lack).

23   Christiana Olfert has suggested an account along these lines: her ‘Aristotle’s Conception of Practical Truth’, Journal of the History of Philosophy, 2014. 24   I assume that in VI.2 Aristotle is focusing on those with excellent preferential choices (1139a25: the practically wise/fully virtuous) and excluding the akratic (whose preferential choice is less than excellent and merely ‘OK’: 1152 a17) and the self-controlled (who differ in excellence from the virtuous in having some base desires: 1152a1ff.). This is why he focuses on the good state of the practical intellect (1139a28ff.), the one in which it achieves its characteristic excellence (or function).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

162  David Charles This reply, however, leaves the crucial connection unexplained: why does being attended by correct desire have anything to do with truthfulness? On the present account, truthfulness consists in making or grasping true assertions (where making or grasping is defined independently of desire). If so, why should it be affected in any way by the nature of the desires that accompany the relevant assertion? Surely, this form of truthfulness should be independent of the nature of the accompanying desire. After all, both the virtuous and the self-controlled track truths in their judgments. A two-component interpreter may be tempted, at this point, simply to stipulate that a truthful assertion is (amongst other things) one actually attended by correct desire. Given this stipulation, the self-controlled will not make truthful assertions because they lack correct desires. But this stipulation seems ad hoc. It fails to indicate what it is about the nature of a truthful assertion which explains why it is followed by correct desire (when the intellect functions well). Why should true grasp of assertions which are true and as such the objects of correct desire bring with them in the case of the virtuous but not the self-controlled (or the akratic) correct desires? They both may make true assertions about what is good to do. There seems to be nothing about truthful assertion itself that—on the present account—explains why correct desire is present in one case but not the other. There is a gap precisely where an answer is most needed. Further, even if the virtuous wish to make their desires fall in line with their true judgments about what is to be done and succeed in doing so, neither this wish nor its achievement appears to affect the truthfulness of their judgments. It affects only the extent to which they succeed in bringing their desires in line with their truthful assertions. There is a second problem for this account: it takes as its starting point true claims about what actions it is good for a human to do. From this starting point, it denies that what makes such actions ones which it is good for us to do is that they are actions which we should desire. In its view, our desires are correct only because they are aimed at actions which are—on grounds independent of desire—good for us to do. However, in the present context, Aristotle takes actions as the ones which are, in truth, to be done because they are cases of doing well (eupraxia: ones which we should desire to do: 1139b3–4). It is not that true judgment about the good is more basic than correct desire. Both rest equally and non-derivatively on what doing well (eupraxia) is. Desires are correct and judgments true in virtue of tracking (in their ­differing ways) the very same features of these actions.25 In sum: two-component interpretations fail to capture the idea of one unified state which meets the three following desiderata: it is (i) truthful (ii) action guiding and (iii) specific to the practically wise (not shared with the self-controlled and the akratic).26 25   Nor is it the case that judgments are true because they track those features which make our desires about what actions to do correct. That is, it is not that our having correct desire is primary and our judgments are true when they are based on these desires. What makes our desires correct is the very same thing as makes our judgments about what is to be done true. 26   This is not to say that the self-controlled’s or the akratic’s better judgment as to what to do is false. It is rather that they grasp it in a less than truthful way: their desiderative intellects are not governed by the truth in the right way. This is why they have less than ideal desires.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  163 Miss Anscombe, in her innovative contribution, differs from the two-component interpreters just considered in taking there to be one state which is truthful and action guiding: desire or preferential choice.27 On this, she and the 3rd way interpreter agree. However, her account achieves this result by presenting Aristotle as introducing in these lines a wholly new (and unparalleled) type of truth: truth-in-action. It is not because—as in the third way account—preferential choice is both truth assessable (in terms of the ordinary notion of truth) and at the same time determinative of action. Preferential choice is true, for Anscombe, in that it leads to truth-in-action: practical truth. The agent in acting in accordance with correct desire achieves truth-in-action. So understood, truth and falsity apply directly to actions ‘strictly and properly’. In Anscombe’s account, the practical intellect aims at practical truth and the practically wise succeed in achieving this goal by making certain claims (those that agree with correct desire) true by acting on them. Theoretical intellect, by contrast, aims at grasping what is true. There is—in her view—a distinctive type of truthfulness for ­preferential choice, based on a different type of truth: truth-in-action which the agent brings about by acting. This is how, in her view, preferential choice controls both action and truth (-in-action). It makes the relevant claims true by initiating action on their basis. If preferential choice is to be excellent, the relevant desires must be correct ones: the ones made true by acting well. There are several problematic features of Anscombe’s account, at least when considered as an interpretation of Aristotle’s viewpoint. First, it breaks Aristotle’s tight connection between truthfulness and true accounts or arguments (1139a24f.), even though the latter were used initially to introduce the idea of practical truthfulness. Instead, according to Anscombe, Aristotle invokes a wholly new notion of truth grasped by doing what accords with correct desire (practical truth). More specifically, in her view, what makes a preferential choice true is that it leads to achieving truthin-action. It is not true because it asserts truly that a certain action is to be done. On her reading, Aristotle introduces this radically new idea of truth in this passage, divorced from his earlier connections with true accounts, ‘out of the blue’, with no support in the intervening lines. This degree of isolation from context suggests a view read into the text for further philosophical reasons than one based on a careful study of its actual argument. There is a further, more theoretical, problem: it seems, as already noted, that what makes certain actions the objects of correct desire is that, for Aristotle, they are—in truth—desirable. However, it is unclear what, in Anscombe’s account, makes actions the objects of correct desire. It cannot simply be the fact that we act (and can act) on them. It must be that so acting is doing well. But what makes so acting a case of doing (or acting) well? It is hard to see how Aristotle could have answered that question except by saying: what makes so acting a case of doing well is that it is actually a case of eupraxia. That is, there is some truth about what eupraxia is that makes so acting a case   ‘Thought and Action in Aristotle’, Essays on Plato and Aristotle, ed. R. Bamborough, London 1965.

27

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

164  David Charles of acting well. However, so understood, pace Anscombe, practical truth rests (is grounded) on truths about what is to be done (or what acting well is) and does not itself constitute them.28

4.  Some Relevant Texts in NE VI A full defence of the 3rd way reading of NE VI lies beyond the range of this essay. It would consist in arguing that this account offers the best explanation of a variety of passages and themes in the remainder of the book. I shall conclude by mentioning three of these, which bear directly on the issues so far discussed.29 (1) Practical wisdom (phronesis), although a virtue of the doxastic part of the soul, ­differs from belief (doxa: 1140b26ff.) and is not simply a state involving reason (meta logou). It differs from belief in that it does not admit of forgetfulness. Practical wisdom, on a 3rd way reading, does not admit of forgetfulness because (even though it is—in a way—cognitive) it is not itself simply a cognitive state nor a compound of a cognitive state and a separate desiderative one. If it were simply a ­cognitive state, it would admit of forgetfulness in the way ordinary cognitive states do. If it were a combination of a cognitive state and a separate desiderative one, it would seem possible for the cognitive component to suffer memory loss. A more complex story would be needed to explain why the cognitive component cannot suffer memory loss because of the presence of a good conative component. Perhaps, it will be said, the presence of the latter in some way prevents the former being adversely affected by the processes that normally induce forgetfulness. But why would this be? And why would the presence of a conative state only have in this effect in the case of practical wisdom and not in that of other pieces of knowledge one would dearly love to retain (such as peoples’ names etc.). There is, certainly, no indication of this line of thought in the present context. Ryle pursued this line of thought, suggesting that grasping the difference between right and wrong ‘goes hand in hand with approving and disapproving, relishing and disrelishing, admiring and despising, pursuing and avoiding.’ Indeed, Ryle went on to suggest that the ‘connection seems closer than hand-in-hand concomitance. . . . In this 28   I assume, in this argument, that it is not something’s being a case of truth-in-action that itself makes it a case of acting well or of something to be done by the virtuous. However, perhaps Anscombe was ­making the opposite assumption, taking truth-in-action as her (and Aristotle’s) a basic notion. But, if so, she has to address the question: what makes it the case that the actions of the virtuous are the ones to be done (cases of truth-in-action)? The obvious answer is: they are to be done (cases of truth-in-action) because they are, in fact, cases of good action, acting well (eupraxia). It will not do to say: doing such actions just is doing well. This is because what is at stake is what makes it true that such actions are cases of doing well. Desires if correct may aim at doing well. But doing well is not the right kind of thing to make assertions true or desires correct. What makes such assertions true (and such desires correct) is the relevant actions’ being—in truth—cases of doing well. 29  Some others are discussed in my ‘Aristotle on practical and theoretical knowledge’, Bridging the Gap between Aristotle’s Science and Ethics, ed. D. Henry and K.M. Nielsen, Cambridge 2015, pp. 71–93.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  165 area the partitions are down between the Faculties of Cognition, Conation and Feeling.’30 While Ryle did not develop, or defend, a 3rd way reading of Aristotle’s account of practical wisdom in any detail, he emphasized (in a way wholly congenial to this reading) that practical wisdom is not to be regarded as an intellectual state which is simply with reason (meta logou). Practical wisdom may, it should be noted, differ from understanding (sunesis) in a  similar way since the latter involves belief (doxa: 1143a14f.). Understanding, in ­contrast with practical wisdom, seems to belong to the doxastic part of the soul which has beliefs, albeit ones about what is beneficial or good to do (see 1112a4). Practical wisdom is said to be prescriptive (1143a8), saying what is to be done, and (so understood) can be compared with ordering. What emerges from these passages is a limited claim: practical wisdom is a virtue of the doxastic part of the soul which differs from belief and is not simply a reasoninvolving state. What is needed is an explanation of why it (a) is prescriptive (b) not subject to forgetfulness and (c) leads to action (1140b21f.) but none is provided in NE VI.5 and VI.10. Further, there is no stated connection between this specific aspect of the soul and desire in these chapters. Indeed, ‘desire’ is not explicitly mentioned in VI.3–11. However, in the light of VI.2, it seems appropriate to see (a)–(c) as arising because practical wisdom (understood as the best state of the practical intellect) is, in its nature, desiderative. Further, the 3rd way reading offers a good explanation of these claims: to think of something as good is inseparable in definition from desiring it. When thought of as a desire, the state can be described as saying that ‘a is to be done’ or ‘let me do a’. When thought of as a type of thought, the state can be described as thinking-in-the-desiderative-mode that a is to be done. In De Motu, the relevant type of thinking is described as thinking that a is to be done ‘through the good’. The best description of the relevant state may be: thinking that a—a good action—is to be done or thinking that a is to be done seen from the perspective of what is good to do (under the guise of what is good to do). (2) While desire is not mentioned in these chapters, pleasure and pain are. In N.E. VI. 5, shortly before his discussion of forgetfulness, Aristotle notes that: while pleasure and pain do not destroy or warp every judgment, such as ones about whether triangles have two sides, they do destroy or warp judgments about what is done. For the starting points for actions are the agent’s goals in action. To the person corrupted by pleasure or pain the goal is—as a direct consequence—not visible, nor that he should choose and act in all cases for its sake and because of it. For vice is destructive of the starting point.  (1140b12–20)

30   Gilbert Ryle, ‘On Forgetting the Difference between Right and Wrong’, Aristotle’s Ethics, ed. Walsh and Shapiro, 1967, pp. 73–4. Ryle raises serious objections against two other possible explanations of the phenomenon: (a) such knowledge is in constant use (b) it is simply a form of ‘know how’. Ryle argues that there is forgetfulness of a type in both (a) and (b) (e.g. ‘getting rusty’) but not in the case of practical wisdom.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

166  David Charles He returns to this point in VI. 12 when he claims that: wickedness warps [one’s grasp of the goal] and makes one be mistaken about the practical starting points.  (1144a34–5)

Why do pleasure and pain affect one’s grasp of the practical goal but not one’s judgments in geometry? The 3rd way account offers a simple and satisfying explanation. To see something as a good goal is inseparable in definition from being attracted to it and desiring to act on its basis. If other pleasures attract you to different actions or if  acting virtuously becomes painful, acting in this way will cease to be attractive. However, to the extent that they cease to be attractive, you will cease to see so acting as a good goal to be achieved by action. Indeed, such goals ceasing to be attractive will just be their ceasing to appear to you as your practical goal, the good to be pursued. By contrast, in the case of mathematics, pleasure and pain do not, as a direct causal consequence, make one forget the relevant truths. This is because grasping a mathematical truth does not essentially involve being attracted to it. Nor does its grasp essentially involve thinking of something as good. Standard two-component accounts need to tell a far more complicated story. If the grasp of goals is a type of belief, it is not clear why it is affected by pleasure and pain in this way. Aristotle would have to be relying on a complex but untold causal story of why (and how) pleasure and pain affect our intellectual grasp of valuational goals but not of mathematical truths. Given that both the valuational and the mathematical are beliefs, why does pleasure affect the former and not the latter? What is it about valuational beliefs that makes them uniquely vulnerable to pleasure and pain in this way, if it is not the distinctive manner in which (according to the 3rd way interpretation) we grasp them? Aristotle does not develop or even suggest an answer to this question. From the perspective of the two-component interpretation, there is a gap in his account at the precise point at which an explanation is most needed. Scholars have (generously) attempted to fill this gap by developing their own causal stories of how change in desires causes change in some areas but not in others. However, since Aristotle himself did not do so, the simplest account is that offered in the 3rd interpretation: there is no such gap to be filled. On this view, to cease to find acting virtuously pleasant and to find so acting painful just is to cease to see it as one’s goal. (3) ‘Virtue is not just a state which is in line with correct argument (reasoning) but rather a state with correct argument (reasoning)’ (1144b27ff.). On a 3rd way reading: to be fully virtuous requires one to have correct grasp of the goals. To have a correct grasp of the goal is to have a truthful grasp of them and therein to have correct desires. This is why one cannot have virtue without a truthful grasp of the goals. Possessing a grasp of this type on the relevant goals is, on the 3rd way reading, (partially) constitutive of what it is to be virtuous. Virtue does not consist in having ends which are in line with correct argument. Part of what it is to be virtuous is grasp such goals for oneself in the truthful way characteristic of the practically wise.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Practical Truth: An Interpretation of Parts of NE VI  167 Full virtue requires practical wisdom both in requiring a truthful grasp of the goals and of what is to be done to achieve them. But the type of practical wisdom involved is itself inextricably connected with correct desire. One cannot have either a truthful grasp of the goal or the means without desiderative modes of grasping each of them. But equally one could not have the correct desires for the goal or the means without practical wisdom. In sum: (1)–(3) develop, in a 3rd way style, the claims about practical truthfulness and correct desire in VI.2 as analysed in the opening sections of this essay. Indeed, these three claims reveal the coherence and appealing economy of Aristotle’s account of practical rationality when (and, as I have argued, only when) interpreted in this way.

5. Conclusions If truthfulness is a secure way of hitting the truth, practical truthfulness will be a secure way of hitting the truth about what is to be done. The practically wise possess, and manifest, practical truthfulness. Two-component philosophers (whether they be interpreters or independent thinkers) are inclined to understand practical truthfulness in one of the following ways. For one group, it consists in (i) having true thoughts (whether beliefs or knowledge) about what is to be done (and why), where the thoughts are defined independently of desire and (ii) the thoughts’ being true in a straightforward way (‘a is to be done’ is true in L iff a is to be done). On this view, the role of desire is taken as a background condition: it enables one to grasp, or act on, the truth one grasps. Perhaps, if all goes well, one can ensure that one’s desires match one’s intellectual grasp of the truth. (This is the ‘belief-based’ version.) For a second group, practical truthfulness consists in (i) having correct desires about what to do (and correct desires for goals), where the relevant desires are defined independently of one’s thoughts and (ii) the desires are correct if and only if they issue in good actions. On this view, the role of an intellectual grasp of the truth (straightforwardly understood) is to facilitate the emergence of good general desires and direct them appropriately through means-ends (or rule case) reasoning. (This is the ‘desire-based’ version.) For a third group, practical truthfulness consists in two separate achievements: ­having true thoughts (as for the first group) and correct desires (as for the second). Practical wisdom decomposes into two definitionally independent and equally basic components, which are thought to interact with each other in a variety of ways (e.g. by efficient causality). The two components must be equally basic if this view is not to be a notational variant of one of the first two accounts. The 3rd way account differs, fundamentally, from all of these: while it accepts that practical thoughts (such as that a is to be done) are true (or false) in the straightforward manner suggested by belief-based theories, it does not understand them to be separable in definition from the desire to do a. Nor does it regard the relevant desires

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

168  David Charles as separable in definition from practical thoughts. Success in practical thought and ­correctness in desire are, so understood, inseparable. Practical wisdom is a strongly unified phenomenon, not decomposable into two separate achievements of two definitionally separate, equally basic, causally connected components. Similarly, ­failure in practical thought and incorrectness in desire are, for the 3rd way account, inseparable in definition. The 3rd way, so understood, offers a way of thinking about a wide variety of issues in moral psychology (such as the acquisition and grasp of valued goals, a­ krasia, and self-control) distinct from those available within the widely accepted ­two-component mindset. Aristotle’s alternative (if I have understood it correctly) presents a challenge to contemporary orthodoxies. Indeed, in this area at least, there may be no better way!31

Bibliography Anscombe, G.E.M. ‘Thought and Action in Aristotle’, Essays on Plato and Aristotle, ed. R. Bamborough, London: Routledge and Kegan Paul 1965, pp. 143–58. Charles, David. ‘Aristotle’s Desire’, Mind and Modality: Studies in the History of Philosophy in Honour of Simo Knuuttila, ed. V. Hironen, T. Holpaianen, M. Tuominen, Leiden: Brill 2006. Charles, David. ‘Aristotle’s Processes’, Aristotle’s Physics: A Critical Guide, ed. Mariska Leunissen, Cambridge: Cambridge University Press 2015, pp. 186–205. Charles, David. ‘Aristotle on practical and theoretical knowledge’, Bridging the Gap between Aristotle’s Science and Ethics, ed. D. Henry and K.M. Nielsen, Cambridge: Cambridge University Press 2015, pp. 71–93. Greenwood, L.H.G. Aristotle: Nicomachean Ethics Book Six, Cambridge: Cambridge University Press 1909. Irwin, T.H. ‘Aristotle on Reason, Desire and Virtue’, Journal of Philosophy 72: pp. 567–78. Olfert, Christinana. ‘Aristotle’s Conception of Practical Truth’, Journal of the History of Philosophy 52 (2), 2014, pp. 205–31. Ryle, Gilbert. ‘On Forgetting the Difference between Right and Wrong’, Aristotle’s Ethics, Issues and Interpretations, ed. J.J Walsh and H.L Shapiro, California: Wadsworth 1967, pp. 70–9.

31   I am indebted to discussions with Christiana Olfert, Verity Harte, and the participants in the joint seminar she and I gave on Nicomachean Ethics VI in Yale in autumn 2014. I also gained from Christopher Shields’ helpful comments on the penultimate draft. Since the 1970s, Gail Fine, Terry Irwin, and I have had many lively and enjoyable discussions of issues in ancient philosophy. It is a pleasure to contribute this essay to a volume in their honour.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

9 Aristotelian Feelings in the Rhetoric Paula Gottlieb

The 1980s were exciting years at Cornell, with fellow students from among the ­contributors to this volume, and the opportunity to study both with Terry Irwin, the only philosopher since Aristotle to write the same brilliant work twice (Plato’s Moral Theory and Plato’s Ethics), but in a completely different style and with extra argumentation (Future scholars will no doubt argue about the chronology) and with Gail Fine, who was exacting and encouraging at the same time, and whose crystalclear explanations made my job as teaching assistant redundant. For Terry and Gail, who are just as awe-inspiring today as they were thirty years ago, with thanks. Feelings play an important role in Aristotle’s Nicomachean Ethics, but the discussion of them is scanty. A much longer discussion is to be found in Aristotle’s Rhetoric, but as Irwin, among others, points out (Irwin 1996), there are important differences between the context and content of the two works. Despite these differences, I should like to consider how far the discussion of the feelings in Aristotle’s Rhetoric can be used to elucidate the feelings mentioned in the Nicomachean Ethics. The answer is that we get a richer picture of the emotional life as Aristotle understands it if we draw on material from the Rhetoric. The Rhetoric shows how most feelings are prompted by concrete circumstances, providing indexical insight. The Rhetoric also reveals the interplay between different feelings, especially fear and sympathy, not explicitly described in Aristotle’s Nicomachean Ethics. Evidence taken from the Rhetoric can also be used to make clear the neutral nature of the feelings. Feelings are not inherently bad and therefore do not need to be quashed by an Aristotelian good person. The Rhetoric also raises a puzzle about the feelings and the desire to act. While anger is defined as a desire for retaliation, for example, many definitions of feelings in the Rhetoric do not mention any specific desire to act.1 This silence reinforces the similar absence in the Ethics, supporting the hypothesis that the content of the desire is 1   Here I use the term ‘define’ loosely. It is debated whether Aristotle means to define the feelings in his Rhetoric. The ‘definitions’ do not fit Aristotle’s strictures in the Topics, for example, but nor are they simply

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

170  Paula Gottlieb dependent both on context, as explained in the discussion of indexical insight, and on the character of the agent. There is, however, an important difficulty with using the Rhetoric to understand Aristotle’s Nicomachean Ethics, and that difficulty concerns the development of good character. However, understanding why the Rhetoric falls short in that regard shows what philosophical work would need to be done to gain a full picture. One upshot of my general discussion is that Aristotle’s views about feelings in the Rhetoric, while superficially similar to Hobbes’s discussion of what Hobbes calls ‘the passions’ in the Leviathan, contrast with those of Hobbes, especially on the issues of the interplay of sympathy and fear, the neutrality of the feelings, and the relationship between feeling and the desire for action. In sum, the important differences between Aristotelian naturalism and Hobbesian voluntarism, highlighted in Irwin’s trilogy (Irwin 2007–9), first appear in the contrast between Aristotle’s account of feelings and Hobbes’s account of the passions.2

1. Aristotle’s Nicomachean Ethics and Rhetoric In his Nicomachean Ethics Aristotle says that virtues of character relate to the feelings. Virtues of character are neither capacities to feel, nor mere feelings, but it is according to them that we are disposed well or badly in relation to the feelings (pros ta pathē) (EN II 5 1105b26). Virtuous people act and feel correctly, we are told,3 but with no further analysis of the feelings, what it is to do so is not specified by the text. True, the Nicomachean Ethics provides a list of the feelings as follows: ‘By feelings pathē) I mean appetite (epithumia), anger (orgē), fear (phobos), confidence (tharsos), envy (phthonos), joy (chara), love (philia), hate (misos), longing (pothos), emulation (zēlos), sympathy (eleos), in general whatever implies pleasure and pain’ (EN II 5 1105b21–3). However, it would be helpful to have a more detailed account. Such an account is available in Aristotle’s Rhetoric II 1–11.4 However, the use of passages in the Rhetoric in order to elucidate Aristotle’s ethical works is problematic. Here I shall concentrate on the Nicomachean Ethics, but similar points may apply to  Aristotle’s other ethical works. First, the aims and context of the Rhetoric and Nicomachean Ethics are very different. The aim of the Rhetoric is to show the orator how to persuade his audience to support his cause, whereas the aim of the Nicomachean Ethics relates to moral improvement. Second, as Irwin argues, the Rhetoric often accepts without argument commonplace beliefs that are not accepted in the Nicomachean

what people happen to think. The latter would not be sufficiently helpful to the rhetorician. For more on definition, see, for example Rapp 2009, 586–7. 2   For the purposes of this essay, I am restricting my discussion of the passions to Hobbes’s Leviathan, since it forms part of the introduction to his social contract theory. 3   See especially Kosman 1980. 4   The accounts in Aristotle’s De Anima and De Motu Animalium are beyond the scope of this essay.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  171 Ethics and elsewhere,5 and shows no awareness of the metaphysical distinctions between capacity (dunamis) and disposition (hexis), essential to Aristotle’s account of virtues of character in his Nicomachean Ethics (Irwin 1996). Aristotle’s famous doctrine of the mean is also missing. In the first book of the Rhetoric each virtue of character is opposed to one and not two vices (Rh. I 9). A possible response to these problems is to consider Rhetoric II 1–11 as part of a separate treatise, perhaps incorporated into the Rhetoric at a later time than the rest,6 and philosophical enough to be taken more seriously than the rest of the work, indeed a ‘credible philosophy’ in itself.7 Its account can therefore be considered self-contained, and not contaminated by other discussions in the same work. However, there are difficulties with this view. First, the discussion of the feelings is not in fact confined to Rhetoric II 1–11.8 Second, the feelings are presented as affecting people’s judgments, a salient fact for the orator. Most importantly, the feelings invoked in Rhetoric II 1–11 themselves are a somewhat motley crew, including phenomena that are or would be classified differently in the Nicomachean Ethics. For example, mildness (praotēs), a feeling of calming down in the Rhetoric, refers to a full-blown virtue of character in the Nicomachean Ethics. Kindliness or gratitude (charis) is characterized as a feeling in the Rhetoric, but is then described as an action, a service to someone in need.9 According to the Nicomachean Ethics, feelings and actions are not the same. In the Nicomachean Ethics appetite (epithumia) includes physical appetites for food, drink, and sex which especially involve the sense of touch (EN III 10). Appetites for health and learning are mentioned as well (EN III 1 1111a30–1).10 As Leighton has emphasized, appetite (epithumia) is not explicitly included in the Rhetoric’s list,11 although Aristotle does argue that kindliness (charis) involves noting the needs and especially the epithumiai (translated ‘longings’ by Kennedy 1991) of others. These epithumiai involve pain for something absent, for example the longing of sexual

5   Merely commonplace beliefs do not of course cover beliefs about feelings. As Striker 1996, 287 argues, the orator needs true beliefs about the feelings in order to be successful. 6   Kennedy 1996, especially 417–21, argues that the Rhetoric was written and revised over quite a long period of time, with the discussion on feelings and ethos incorporated quite late. 7   See Dow 2011 and Dow 2013, 145–81, especially 155–61. 8   Konstan  2006, 245, makes much of the omission of grief from the discussion of Rhetoric II 1–11, which he takes to show that grief is not considered a proper feeling. However, grief appears in Rhetoric I 11. In Rhetoric III 16 Aristotle even describes the body language typical of those with particular feelings so that the orator can describe it successfully in his narration. For example, an angry person may spit with fury and a grieving person may bury her head in her hands. 9   Kindliness, gratitude (or grace) (charis) is not on Aristotle’s Nicomachean list of feelings, but it plays an important role in Aristotle’s account of reciprocal justice in Nicomachean Ethics (EN V 5, especially 1133a1–5). See Young 2013. 10   In Annas 2011’s modern account of virtue, aspiring to learn is very important. 11   Leighton 1996, 224–5 argues that Aristotle does not list epithumiai among the feelings in his Rhetoric because they do not alter judgments. But Aristotle is concerned that the pleasures related to temperance can destroy the starting points of action (EN VI 5 1140b12–18).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

172  Paula Gottlieb desire (erōs), or cases in which one experiences physical suffering or is in danger (Rh. II 7 1385a23–4).12 Therefore, the aims and context of the Rhetoric and the Nicomachean Ethics differ, and even leaving aside the rest of the Rhetoric, Rhetoric II 1–11 is at odds with the Nicomachean Ethics in terms of its list of feelings and how certain phenomena are classified. With this large caveat, I should like to consider several aspects of the Rhetoric’s discussion of the feelings that are helpful in understanding the ethics, and one which would require major supplementation to be useful. I shall conclude with some brief reflections contrasting Aristotle and Hobbes.

2.  Indexical Insight The issue of indexical insight described below arises because the feelings described in the Rhetoric are occurrent episodes; they are not general moods,13 and in most cases, they are directed at particular objects or particular times. In the clearest example, Aristotle defines anger as an impulse involving pain for an apparent (phainomenēs)14 revenge for an apparent slight (phainomenēn oligōrian) towards oneself or what concerns one’s friends, when the slight is not fitting. ‘If this [is a proper definition of] anger’, Aristotle says, ‘it must always be felt towards some individual, e.g., Cleon, and not “human being” in general’ (Rh. II 2 1378a30–4). To take some other examples, fear is defined as ‘a pain or disturbance (lupē tis ē tarachē) due to an impression (ek phantasias15) of some imminent destructive or painful bad [thing]’ (Rh. II 5 1382a21–2). Confidence (tharsos) is the hope involving an impression that safety is near or that fearful things are far away or nonexistent (Rh. II 5 1383a16–19). Hatred is an exception. In the Rhetoric, hatred is directed at types of people such as the thief or the informer, and the hater wishes those he hates to cease to exist (Rh. II 4 1382a1–15) (perhaps as the opposite to the friend who wishes her friend to exist, see 12   Appetites are also discussed in book 1 as desires (orexeis) for the pleasant, and are divided into the nonrational (alogoi) and those that involve reason (meta logou), arising from persuasion (Rh. I 11 1370a17–27). Pearson 2012 discusses Rhetoric I 10–11. 13   Aristotle explains elsewhere how they can colour the situation:

. . . we are easily deceived respecting the operations of sense-perception when we are excited by emotions, and different persons according to their different emotions, for example, the coward when excited by fear, the amorous person by amorous desire, so that, with little resemblance to go upon, the former thinks he sees his foes approaching, the latter, that he sees his beloved, and the more deeply one is under the influence of the emotion, the less similarity is required for these to appear.  (On Dreams 3 460b3–9, tr. Beare)   This word is not in all the manuscripts.   In Section 5, I discuss Cooper’s view that the phainomenon is what is apparent to the agent, not what is particularly conspicuous. This view does not rule out veridical appearances (see Cooper 1996, 47; 1999, 417). 14

15

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  173 (EE VII 6 1240a25–8). Hatred is exceptional in another way also. It does not fit the ‘definition’ of a feeling as a pleasure or pain. Not surprisingly, particulars are emphasized in the Rhetoric since the orator may wish to direct the audience’s attention to particular people now. But particulars are just as important for making ethical decisions that are made in particular circumstances and relate to particular people or things. In short, feelings are mainly directed to things in the here and now, providing what may be called indexical insight, an important requirement for good Aristotelian practical reasoning.16 Of course, the good practical reasoner will have to have been well brought up in order to have veridical feelings. This is an issue for Section 5. Despite the absence from the Rhetoric of Aristotle’s signature ethical doctrine, the doctrine of the mean, which includes the idea that the agent should have his character in equilibrium so that he has the correct feelings in the correct way at the correct time and so forth, Aristotle’s discussion of unkindliness (acharistia) in the Rhetoric foreshadows such a view. In that discussion, Aristotle invokes all his categories to show the different facets of the situation. ‘For unkindliness is either determined by substance or quantity or quality or time or place’ (Rh. II 7 1385b5–6). The comment is obscure, but presumably means that one may be unkindly because at least one factor, for example, time or place is incorrect, just as virtue and vice are determined by many factors in Aristotle’s ethical works.

3.  Linking the Feelings: Fear and Sympathy Apart from showing the importance of indexical insight, Rhetoric II 1–11 also shows the interplay of different feelings, which is missing from the ethical works. In the Rhetoric, Aristotle often presents one feeling as being opposed to another, but I should like to draw attention to fear and sympathy, which are described as a complementary duo. Aristotle says, ‘Sympathy (eleos) may be defined as a certain feeling of pain (lupē tis) at an apparent bad (epi phainomenō(i) kakō(i)), destructive and painful, which is unworthy and which we might expect to befall ourselves or some friend of ours, and moreover to befall us soon’ (Rh. II 8 1385b13–16). ‘Sympathy’ here is a better translation than ‘pity’ because ‘pity’ often implies that the agent is in a superior position to those pitied and will not be similarly harmed, a point which contradicts Aristotle’s own definition. However, ‘sympathy’ is also inexact, even though I shall continue to use the term. Aristotle is only discussing sympathetic pain and not also sympathetic pleasure, even though he recognizes both elsewhere when discussing friends (Rh. II 2 1381a3–7; EN IX 10, 11). Moreover, in his account of sympathy, unlike Hume,17 Aristotle is not 16   See, for example, Gottlieb 2009, 151–72. I am grateful to Matthew Darlingum for the term ‘indexical insight’ to describe this interpretation of Aristotle. 17   For example, Hume’s Treatise on Human Nature, Book II, Part I, Section XI 1–4 (Sayre-McCord 2006), although Hume thinks that one can ‘catch’ opinions from other people too. Hume’s account of how one becomes sympathetic to the public good may or may not be consistent with this view.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

174  Paula Gottlieb thinking of one person’s feelings resonating with another’s, so that each catches the other’s mood.18 Rather sympathy is directed at the bad situation of another that one might expect to be experienced by oneself or by one’s friends soon. Aristotle’s account of sympathy has an advantage over those modern accounts that falter in explaining how one can imagine being someone else. Bernard Williams memorably commented that to imagine being someone else is to imagine the world just as it is, but with yourself no longer in it.19 There is also a certain kind of hubris in imagining that you are someone else. The Aristotelian sympathizer does not need to imagine that she is the other person. Instead she is distressed by the situation that the other person is in, being able to imagine being in it herself. According to Aristotle, while fear is not essential to sympathy—it is not included in its definition—we sympathize with others for what we fear for ourselves and vice versa. As Aristotle says, ‘speaking generally, fearful things (phobera) are things that when they happen or are about to happen to others are sympathy-provoking (eleeina)’ (Rh. II 5 1382b24–6). He says, ‘In general, we should grasp here too that people feel sympathy with things happening to others in so far as they fear them for themselves’ (Rh. II 8 1386a27–9). As we shall see, Hobbes misses the connection between fear and sympathy and its significance. Even so, one might wonder why fearing for oneself is relevant to having sympathy for another. It may seem too narrow because it hardly explains Aristotle’s view that the aim of ancient Greek tragedies is to provoke fear and sympathy in the audience, since members of the audience and their friends are unlikely to expect to be in the position of those tragedies’ protagonists. However, perhaps it is not so much the status of those tragedies’ protagonists that awakens the audience’s feelings, but other factors. For example, as Aristotle explains, the tragedian puts the undeserved suffering of the characters ‘in front of our eyes’ (Rh. II 1386b5–7), and it may be that seeing their sufferings here and now is what causes fear and sympathy, in accord with the indexical insight described earlier.20 Indeed, Aristotle makes a similar point about shame in the Rhetoric. Shame, like sympathy, is more easily aroused in the presence of the relevant parties. Aristotle defines shame as phantasia about a loss of reputation. He describes a particularly effective speech when the Athenians were asked to imagine all the Greeks standing in a circle and seeing what is going on (Rh. II 6). Aristotle shows that having aggrieved parties in front of one’s eyes, or imagining them as being in front of oneself, does much to arouse certain feelings.

18   This is sometimes described as ‘emotional contagion’. Sober and Wilson 1998, 234 describe a case of infants a few days old who cry when they hear other infants crying. 19   Williams attributes the thought to Leibniz, referring to someone who wished to be the king of China (Williams 1973, 42–3). 20   Seeing the sufferings presumably does not mean seeing the events that caused the sufferings, as these occur offstage. It must mean seeing the characters suffer on stage.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  175 Aristotle reasonably assumes that we feel sympathy for those nearby even if we are not related to them. He also assumes that cynics will not feel sympathy for anyone, since they see no suffering as undeserved. In making these two assumptions in the Rhetoric, Aristotle is not explaining why people should feel sympathy, only how they usually do. This is especially important for the orator to know, since he needs friendly feeling and sympathy for himself (Rh. III 14 1415b27–8). Nor does Aristotle say what desires for action those with sympathy should have. Some may help the person who provokes sympathy, while others may turn away, or even make sure that that person or those people are far away. This shows that the pain involved in the very same feelings may provoke desires either for pursuit or for aversion. This omission is significant. I will return to it forthwith. According to Aristotle, the orator can also arouse other feelings to counteract sympathy. For example, a feeling of indignation can counteract sympathy. Good people will feel indignation at those bad people in trouble who deserved what they got, while malicious people, who are envious of others, will be pleased at the downfall of those others (Rh. II 9 1386b26–1387a5). As Cooper mentions, the idea that there are connections between different feelings is an important one.21 Aristotle describes how different appetites can give rise to a particular feeling. For example, if people are ill, poor, at war, in love, thirsty, and so on and someone belittles their condition, they will be angry accordingly ‘for each has prepared a path22 for his own anger because of some underlying feeling (huparchontos pathous)’ (Rh. II 2 1379a23–4, my emphasis). Aristotle says that an angry person can be brought to have sympathy for the object of anger, whereas someone who hates cannot (Rh. II 4 1382a1–15). Aristotle provides no explanation at this point, but presumably this is because there is no personal relationship between hater and hated (as described in the earlier section on indexical insight) as there is between the Aristotelian angry person and the object of the anger, and so it will be much harder to rectify the situation between the two.

4.  Neutrality of the Feelings In Aristotle’s Rhetoric, most feelings are described in a neutral fashion; that is, they are not presented as good nor bad independently of the concrete situation. Envy and spite are exceptions, but they may have a base that is neutral.23 Envy is a feeling directed at peers who have goods that one lacks. So is emulation, but envy, as Aristotle explains, is a bad feeling had by bad people whereas emulation is a good feeling had by good people. Aristotle describes envy as ‘pain at the sight of good fortune . . . and is felt towards those we consider our equals, not with the idea of getting something ourselves, but 22   Cooper 1996, 250; 1999, 422.  Reading proōdopoeitai.   Leighton denies the possibility of this, especially with reference to the Eudemian Ethics, but there Aristotle is discussing emotional propensities, not mere feelings (Leighton 2011). 21 23

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

176  Paula Gottlieb because the other people have it’ (Rh. II 10). ‘Emulation is pain caused at the apparent presence (lupē tis epi phainomenē parousiai agathōn), in persons who are our equals by nature, of good things that are highly valued and are possible for ourselves to acquire; but it is felt not because others have these goods, but because we have not got them ourselves. Emulation is a good thing characteristic of decent people, whereas envy is bad and felt by bad people’ (Rh. II 11 1388a32–5, my emphasis). According to Aristotle, emulation makes one aim to get the good things, including virtue, oneself, whereas envy makes one want to prevent one’s neighbor from having good things (Rh. II 11 1388a36–8). The same impression, of someone doing better than oneself, is part of a different overall affect. Understanding envy and spite in this way meshes with Aristotle’s listing envy with murder as something whose name always denotes something wrong (EN II 1107a9–12). In the case of murder, there is a neutral base, killing, that, according to Aristotle, may be good or bad depending on the circumstances. Similarly, there is a neutral base for envy, a feeling towards peers who have goods that one lacks. Spite too can be explained as having a misplaced basis for joy. Such views contribute to Aristotle’s important view in his ethics that people are not born with bad feelings that need to be quashed in order that they may become good adults. In addition, what are described as painful feelings can have positive repercussions. For example, fear makes us ‘able to deliberate’ (Rh. II 5 1383a6–7). Once again, feelings are not inherently bad. As Aristotle says, virtues of character in humans are not by nature nor contrary to nature as it would be if there were many bad emotions that we need to overcome (EN II 1 1103a24).

5.  Feelings and the Desire for Action In the Rhetoric Aristotle says ‘The feelings (ta pathē) are all those that so change [human beings] as to affect their judgments (tas kriseis), and that involve (meta) pain or pleasure, for example, anger (orgē), sympathy, fear, all such things and their opposites’ (Rh. II 1 1378a19–21).24 The reference to pleasure and pain chimes with the passage in the Nicomachean Ethics (EN II 5 1105b21–3), although it is controversial whether pleasure and pain receive the same treatment in or even within each work, and how the analysis of pleasure in either work fits the feelings.25 Are the pleasures and pains involved in feelings physical or mental disturbances or both? In the same book of the Rhetoric, Aristotle says that fear is a kind of chilling, which suggests that a physical reaction is at issue (Rh. II 13 1389b32), but not every 24   On hatred as the exception that does not involve pleasure and pain, see Cooper  1996, 248;  1999, 417–18. As noted, earlier, hatred is also the exception to the point about indexical insight. 25   Pace Frede 1996, it is questionable whether the remedial view of pleasure and pain quoted in Rhetoric I is consistent with all the pleasures and pains discussed in that and the following book, just as it is questionable whether the remedial view of pleasure and pain that appears in Plato’s Philebus applies to all the pleasures described there. Thanks to Emily Fletcher for this point.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  177 kind of chilling is fear, so it is reasonable to suppose that there is a psychic side to the distress as well.26 Whatever the precise psycho-physical analysis of pleasure and pain in the Rhetoric, the issue of interest is whether pleasure automatically leads to a desire in favor of some action and pain against. In his Rhetoric, Aristotle does not mention such desires in the definition of any of the feelings except anger and friendliness,27 although desire (orexis) is essential in his analysis of anger in De Anima. In the Rhetoric, anger is defined as a desire (orexis), accompanied by pain for retaliation (Rh. II 2 1378b30). In the De Anima, the desire is said to be for retaliation or some such thing [my emphasis] (De An. I 1 403a26–7). Significantly Aristotle does not say that fear is a desire to flee or that sympathy is a desire to help. At first sight, the omission is puzzling. Perhaps Aristotle thinks it obvious that some desire is involved, or perhaps he thinks that a particular feeling does not give the agent a desire to act in any particular direction independent of the particular context or the previous experiences and character of the agent. For example, even in the case of anger where a desire is described, Achilles’s anger leads him not to desire to attack Agamemnon directly, but to desire to withdraw from the battle.28 Perhaps if he had been a poet, he might have wanted to write a satirical account instead, or perhaps he might even have channeled his anger to beneficial results.29 Of course, one might argue that the omission merely reflects the context: the orator is not interested in the audience wanting to take particular action; rather he wants them to have the feelings that will make them favorable to his cause. As in the case of watching a play, the audience would not have desires to take particular actions, for example, a desire to provide help to Agamemnon, because that would be impossible, and, as Aristotle points out ‘no one aims at (ephietai) things that appear impossible for himself to get’ (Rh. II 2 1378b3–4).30 What is important is that the specific desire involved in a particular feeling may be  underdetermined by the feeling itself. As Lawrence Becker notes, ‘conation does  not always track emotion. One may be “paralyzed” by fear as well as set in motion by it’ (Becker 2005, 5). In fact, it is left open that the feelings themselves 26   Taylor translates ‘lupē’ as ‘distress’, and he emphasizes frustration (Taylor 2006, 7, 98). On Aristotle’s psychophysical account of the feelings, see especially Charles 2011. 27   Friendliness raises particular difficulties, because the desire involved, boulēsis, is elsewhere treated as a rational desire by Aristotle. 28   Achilles provides a counterexample to Aristotle’s earlier claim that people are not angry or are less angry with those who are more powerful (Rh I 2 1370b14–15), but this comment refers to those who are unable to retaliate successfully. Aristotle does not point out that those who are unable to retaliate against the powerful may retaliate against others less powerful than themselves. See, for example, the account of the prisoners transported from England to Australia (Hughes 1986). The maltreated male prisoners mistreated the female prisoners and they all mistreated the Aborigines. 29   In other words, there may even be room for what Martha Nussbaum calls ‘transitional anger’ where the agent turns to more profitable action than retaliation, although this may not count as anger on her own account (Nussbaum 2015). 30   This is no doubt related to Aristotle’s claim in the Nicomachean Ethics that wish (boulēsis) can be of what is impossible, and is therefore distinct from appetite and other desires and from deliberative desire (prohairesis).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

178  Paula Gottlieb can be directed in different ways depending on the character of the agent and the particular circumstances. That would be a welcome conclusion for the position of the Nicomachean Ethics.31

6.  Becoming Good While Aristotle’s account of feelings in the Rhetoric can give the reader a descriptive account of some of the feelings of the average adult, given its context it cannot show how feelings themselves may differ from a young infant to an adult and even from adult to adult. For example, while a crying baby may display anger, and perhaps is even angry at being neglected and thus, from a third-person point of view, belittled, it is implausible to say that the baby wants to retaliate as opposed merely to express frustration or to get attention, and this is true of many adults too. Hence, it is hard to glean an account of moral development from the Rhetoric. One might think that the account of phantasia in the Rhetoric would provide help here, but there are reasons to be dubious, even if the account of phantasia can be clarified as follows. As has often been noted, in the Rhetoric, it is significant that the feelings are not themselves described as judgments—decisions, or verdicts—or beliefs. Instead, Aristotle describes many of the feelings as involving phantasia. This elusive phenomenon in the Aristotelian corpus, sometimes applying to a psychological faculty and sometimes to something occurrent, has been variously translated ‘imagination’, ‘imagining’, and ‘impression’.32 Aristotle mentions that the style of oratory used to address public assemblies is impressionistic, just like scene-painting (skiagraphia) (Rh. III 12). Presumably he means that the orator paints a scene in broad brush strokes so that it is sketchy enough to be plausible if one does not look too closely, i.e., if one feels without thinking too hard. Political and forensic oratory therefore often involves telling stories about particular people, thus exciting the imagination of the hearers and gaining their support. When something is described as ‘phainomenon’, this relates to the way something appears to the agent. Kennedy and the Oxford translator use the word ‘conspicuous’, implying that there is something out there in the world that has to be made clear to the agent in order to have the appropriate feeling. For example, he takes Aristotle to be arguing that someone will be angry at a slight that is conspicuous. By contrast, Cooper more plausibly argues that the slight may be merely apparent, and the agent may know 31   The discussion in EN X 9 of those living kata pathos (merely in accord with feeling) reflects Aristotle in a more pessimistic mood than in the rest of the Ethics, but the fact that there are those who are habituated in such a way that no arguments can move them does not show that feelings cannot be habituated so that they involve desires with the correct directions. Feelings are not inherently bad. 32   According to Dow 2011 and 2013 pleasure and pain are themselves representational states. Whether phantasia in the Rhetoric and in De Anima are the same is a matter of contention. Moss  2012, 82 and Scheiter 2012 argue that it is. On phantasia in general, see Schofield 1978 and 2011, Lorenz 2006, 119–85.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  179 that fact but continue to be angry, for example.33 There may be no slight, and even if there is, it need not be conspicuous. Cooper emphasizes the fact that how things appear is not necessarily how they are, but just as things may strike someone incorrectly, they might strike a person correctly too. Obviously, the orator will want things that are salient from his or his client’s point of view to strike the audience, but in ordinary life, the agent will be struck by particular things based on prior experience and her actual character.34 Perhaps a more felicitous way to put this is to say that the agent notices various things based on her character and experience, a point that avoids a sense of mere passivity. Even given this account, it is still unclear how an adult reaches the correct stage of moral development. For an infant, the ‘phantasia’ embodied in a feeling is not something expressed or expressible in linguistic terms. It may just be a vague sense of something that an adult would describe as being safe or awry. In an adult are impressions also ­pre-linguistic or are there no impressions that are, to use Ryle’s expression, ‘neat’? If desires and impressions are malleable, then much more needs to be added to the account of the feelings in the Rhetoric to explain how a human being develops good character ‘in relation to the feelings.’ However, this should not be a surprising conclusion, since in Aristotle’s account of moral development in his ethical works, practical thinking needs to be combined with feeling, and both need to be related to an objectively happy life for the agent.35

7.  Aristotelian Feelings, Hobbesian Passions Feelings, which Hobbes calls ‘passions’, play an important role in his Leviathan, and Hobbes provides a long list, superficially reminiscent of Aristotle’s Rhetoric. For example, to return to Section 3, pity has affinities with Aristotelian sympathy: ‘Grief for the calamity of another is PITY, and ariseth from the imagination that the like calamity may befall himself; and therefore, is also called COMPASSION’ [46] (see Curley 1994.) In the introduction to the Leviathan, Hobbes asks his reader to ‘read thy self ’. The aim is to discover the thoughts and passions of mankind in general by looking at one’s own passions (Introduction [3] and [4]). Hobbes notes that he is not talking about the objects of those passions which may vary depending on the individual’s constitution and particular education. He later notes that some passions are innate while others are acquired (Part I Chapter VI [4]). Having noted these points, though, Hobbes goes on to ignore them. His list of passions contain some that Aristotle would classify as feelings and others that he would classify as virtues of character. For example, the Aristotelian virtue of courage is a mere   Cooper 1996, 247; 1999, 417.   For example, Aristotle says that ‘we feel angry with our friends. . . if they do not perceive our needs . . . for this want of perception shows that they are slighting us—we do not fail to perceive the needs of those for whom we care’ (Rh. II 1379b13–17). 35   See Sherman 2013, 749–50. 33

34

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

180  Paula Gottlieb passion on Hobbes’s account, although Hobbes does distinguish it from sudden courage which he calls anger [13–18]. According to Hobbes, the passions are voluntary motions as opposed to the involuntary motions of the blood, nutrition, and so forth (Part I Chapter VI). Several of the passions are connected with physical manifestations, sudden glory with laughter, sudden dejection with weeping, and shame with blushing [42–4]. Hobbes associates appetite and aversion with the Greek hormē and aphormē. The simple passions are appetite, desire, love (a possessive phenomenon unlike Aristotelian friendship which is for the sake of the friend), aversion, hate, joy, and grief. Opinions added to some of these passions create new passions—hope (appetite with the opinion of attaining), despair (the same without such opinion), fear (aversion with opinion of hurt from the object), courage (the same with hope of avoiding that hurt by resistance), and so on. When it comes to describing the state of nature, in which there is a war of every man against every man, and life is ‘solitary, poor, nasty, brutish and short’, Hobbes says that there are three salient motivations: competition, distrust (which he calls ‘diffidence’), and glory (Part I [6]–[9]). Hobbes refers to this situation as an ‘inference made from the passions.’ This is a puzzling conclusion on Hobbes’s own account, for several reasons, not least the existence of other more cooperative passions in his list, for example, benevolence. The passions that incline men to peace are ‘the fear of death, desire of such things as are necessary for commodious living and a hope by their industry to attain them’ [14], none of them particularly social desires. Significantly, while Hobbes does explain pity in a way reminiscent of Aristotle, as we saw above, he does not take into account that fear and sympathy might be two sides of a single coin, and so, in Hobbes’s scheme, sympathy plays no role in any of the motivations that men have in the state of nature, and it also plays no role in the motivation to leave the state of nature. Hobbes fails to understand the interplay of feelings in Aristotle’s Rhetoric. While Hobbes also mentions in his introduction that education and experience can affect the passions [4], his ultimate view is that only the strong power of a state can keep them under control. Therefore, against Aristotle, his official view is that the demands of society are contrary to human nature, rather than in accord with it. His salient feelings are antisocial, and can only be effectively restrained from outside. In the middle of Hobbes’s discussion of the feelings is his famous claim that whatever a man desires counts as good and whatever he is averse to counts as evil, since there are no natural or objective goods and evils, until a representative of a commonwealth is there to pronounce them so [7]. Later in the Leviathan, Hobbes comments, ‘Aristotle and other heathen philosophers defined good and evil by the appetite of men; and well enough, as long as we consider them governed everyone by his own law. For in the condition of men that have no other law than their own appetites, there can be no general rule of good and evil actions. But in a commonwealth this measure is false’ (Part IV Ch. xlvi [32]).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  181 This remark is doubly confused. There is nothing in Aristotle’s account of the feelings that suggests that something is good or bad merely because that is how anyone feels about it. Aristotle’s excellent person, the spoudaios, is the measure, but that is because he has the appropriate practical thinking, virtues of character, and feelings to discern what is naturally good and bad, such things being so independently of what he or any particular society happens to think. The fact that the phronimos can come up with new solutions to new problems does not vitiate this view. Nor does the fact that different things may be appropriate to do on different occasions. Aristotle is no relativist. What is good or bad is not relative to just what any individual thinks, according to Aristotle. The omission of the Aristotelian good person as measure leaves Hobbes’s individuals with a mishmash of passions that leads to an unstructured and arbitrary way of making decisions [49]. Unlike the Aristotelian feelings, the passions are simply tugs and pulls in opposite directions. If my interpretation of the Rhetoric is right, for Aristotle, the feelings are neutral but can gain direction and structure by being put into a larger context of experience, character, and ultimately the happy life. Although Hobbes says that the passions can gain new objects through education, he nowhere explains how this would work. The war of all against all would seem to apply as much to a man’s inner passions as to the outward state of nature.

Conclusion Even though the contexts and metaphysical assumptions of Rhetoric II 1–11 and Aristotle’s Nicomachean Ethics may be different, the Rhetoric provides various insights into the feelings that are relevant to his ethics, and raises questions that a full account of Aristotelian moral development would have to answer. While superficially similar to Aristotle’s account of the feelings, Hobbes’s account of the feelings in his Leviathan already contains the seeds of a philosophical viewpoint very different from Aristotle’s. As Irwin shows, the differences between Aristotle and Hobbes permeate many centuries of moral philosophy (Irwin 2007–9).36

References Annas, Julia. 2011. Intelligent Virtue. Oxford: Oxford University Press. Becker, Lawrence C. 2005. ‘Stoic Emotion.’ In Steven K. Strange and Jack Zupko (eds.), Stoicism: Traditions and Transformations, Cambridge: Cambridge University Press, 250–75. Charles, David. 2011. ‘Desire in Action: Aristotle’s Move.’ In Pakuluk and Pearson (eds.), Moral Psychology and Human Action in Aristotle. Oxford: Oxford University Press, 75–93.

36   Thanks to David Ebrey, Emily Fletcher, Terry Penner, Ruth Saunders, and ‘the Greeks’ for help with the Rhetoric, and to the participants of the IHP conference on Aristotle and the Emotions at Emory, June 2015, for questions relating to the material of Sections 3 and 5. Thanks especially to Susan Sauvé Meyer for detailed comments and suggestions that have led to numerous improvements.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

182  Paula Gottlieb Cooper, John. 1996. ‘An Aristotelian Theory of the Emotions.’ In A. O. Rorty (ed.), Essays on Aristotle’s Rhetoric. Berkeley; Los Angeles; London: University of California Press, 238–57. Cooper, John. 1999. Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory. Princeton: Princeton University Press. Curley, Edwin. 1994. Hobbes Leviathan with selected variants from the Latin edition of 1668. Edition and introduction. Indianapolis, Indiana: Hackett. Dow, Jamie. 2011. ‘Aristotle’s Theory of the Emotions: Emotions as Pleasures and Pains.’ In Pakaluk and Pearson (eds.), Moral Psychology and Human Action in Aristotle. Oxford: Oxford University Press, 47–74. Dow, Jamie. 2013. Passions and Persuasion in Aristotle’s Rhetoric. Oxford: Oxford University Press. Frede, Dorothea. 1996. ‘Mixed Feelings in Aristotle’s Rhetoric.’ In A. O. Rorty (ed.), Essays on Aristotle’s Rhetoric. Berkeley; Los Angeles; London: University of California Press, 258–85. Gottlieb, P. 2009. The Virtue of Aristotle’s Ethics. Cambridge: Cambridge University Press. Hughes, Robert. 1986. The Fatal Shore: The Epic of Australia’s Founding. U.S.A: Alfred A. Kopf. Irwin, T. H. 1996. ‘Ethics in the Rhetoric and in the Ethics.’ In A. O. Rorty (ed.), Essays on Aristotle’s Rhetoric. Berkeley; Los Angeles; London: University of California Press, 142–74. Irwin, T. H. 2007–09. The Development of Ethics: A Historical and Critical Study. Oxford: Oxford University Press, vols. 1–3. Kennedy, George. 1991. Aristotle: A Theory of Civic Discourse, On Rhetoric. Translation, with introduction, notes and appendices. Oxford: Oxford University Press. Kennedy, George. 1996. ‘The Composition and Influence of Aristotle’s Rhetoric.’ In A. O. Rorty (ed.), Essays on Aristotle’s Rhetoric. Berkeley; Los Angeles; London: University of California Press, 416–24. Konstan, David. 2006. The Emotions of the Ancient Greeks: Studies in Aristotle and Classical Literature. Toronto: University of Toronto Press. Kosman, Aryeh. 1980. ‘Being Properly Affected: Virtues and Feelings in Aristotle’s Ethics.’ In:  A. O. Rorty (ed.), Essays on Aristotle’s Ethics. Berkeley and Los Angeles: University of California Press, 103–16. Leighton, Stephen R. 1996. ‘Aristotle and the Emotions.’ In A. O. Rorty (ed.), Essays on Aristotle’s Rhetoric. Berkeley and Los Angeles: University of California Press, 206–37. Leighton, Stephen R. 2011. ‘Inappropriate Passion.’ In J. Miller (ed.), Aristotle’s Nicomachean Ethics: A Critical Guide. Cambridge: Cambridge University Press, 211–36. Lorenz, Hendrik. 2006. The Brute Within: Appetitive Desire in Plato and Aristotle. Oxford: Oxford University Press. Moss, Jessica. 2012. Aristotle on the Apparent Good. Oxford: Oxford University Press. Nussbaum, Martha. 2015. ‘Transitional Anger.’ Journal of the American Philosophical Association, vol 1.1: 41–56. Pearson, Giles. 2012. Aristotle on Desire. Cambridge: Cambridge University. Rapp, Christof. 2009. ‘The Nature and Goals of Rhetoric.’ In Georgios Anagnostopoulos (ed.), A Companion to Aristotle, Oxford: Wiley-Blackwell. Sayre-McCord, Geoffrey. 2006. David Hume Moral Philosophy. Edition and introduction. Indianapolis, Indiana: Hackett.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Aristotelian Feelings in the Rhetoric  183 Scheiter, Krisanna. 2012. Emotion, Imagination, and Feeling in Aristotle. Ph.D. Dissertation, University of Pennsylvania. Schofield, Malcolm. 1978. ‘Aristotle on the Imagination.’ Reprinted in M. Nussbaum and A. O. Rorty (eds.), Essays on Aristotle’s De Anima 1992, 249–78. Schofield, Malcolm. 2011. ‘Phantasia in De Motu Animalium.’ In Pakaluk and Pearson (eds.), Moral Psychology and Human Action in Aristotle. Oxford: Oxford University Press, 119–34. Sherman, Nancy. 2013. ‘Moral Psychology and Virtue.’ In Crisp (ed.), The Oxford Handbook of the History of Ethics. Oxford: Oxford University Press, 744–67. Sober, Elliott and Wilson, David Sloan. 1988. Unto Others: The Evolution and Psychology of Unselfish Behavior. Cambridge, Mass: London: Harvard University Press. Striker, Gisela. 1996. ‘Emotions in Context: Aristotle’s Treatment of the Passions in the Rhetoric and His Moral Psychology.’ In A. O. Rorty (ed.), Essays on Aristotle’s Rhetoric. Berkeley; Los Angeles; London: University of California Press, 286–302. Taylor, C. C. W. 2006. Aristotle Nicomachean Ethics Books II–IV. Translation and commentary. Oxford: Clarendon Press. Williams, Bernard. 1973. ‘Imagination and the Self.’ In Problems of the Self: Philosophical Papers 1956–72. Cambridge: Cambridge University Press, 26–45. Young, Charles. 2013. ‘Aristotelian Grace.’ In G. Anagnostopoulos and Fred Miller Jr. (eds.), Reason and Analysis in Ancient Greek Philosophy: Essays in honor of David Keyt. New York: Springer.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

10 ‘Ought’ in Aristotle’s Nicomachean Ethics Julia Annas

Many issues remain live in the interpretation (and the translation) of Aristotle’s Nicomachean Ethics; in this essay I offer a contribution to one of these, the nature and role of deontic concepts in the work. One way of helpfully approaching this issue is to look at Aristotle’s use of the Greek terms dei and chrē and various gerundive verb forms in the work, and see what this contributes to our understanding. I offer this essay with pleasure to a celebration of Terry and Gail’s work, which has so enriched and enlarged Aristotelian studies among others. We are all grateful for what we owe to them in the understanding of Aristotle’s huge and difficult corpus (quite apart from their extensive other work!). It is hardly news to point out that Aristotle’s ethics is developed in terms of ­virtue, but we should not overstate the role of virtue and think of his as an ­ethics entirely of the good, using, in Sidgwick’s terms, only ‘attractive’ rather than ‘imperative’ concepts.1 Claims that Aristotle lacks deontic language are clearly wrong; he uses the Greek words dei and chrē, as well as gerundive forms of the verb, and these are standardly taken, when they occur in Greek in various genres, to have deontic force; they are translated by ‘ought’, ‘should’, and ‘must’. Sometimes they are translated as ‘it is necessary’, an unfortunate view that has spread to some textbooks; but this is a bad idea, since Greek has a readily usable group of words to express necessity, the anagke-group, and dire confusion is caused if ‘necessary’ is used for this group and also for words like dei and chrē, especially in ethical contexts. Aristotle uses the vocabulary of virtue, and also words for ‘ought’, ‘should’, and ‘must’. What is the relation between them? This is a complex and difficult question, because 1   See Irwin’s valuable article, ‘Eminent Victorians and Greek ethics: Sidgwick, Green and Aristotle', in B. Schultz (ed.), Essays on Henry Sidgwick, Cambridge University Press 1992, 279–310. Sidgwick’s view on attractive and imperative concepts is also criticized effectively by Nicholas White, ‘The attractive and the imperative: Sidgwick’s view of Greek ethics', in the same volume, 311–30. I cannot take up here the issue of whether Sidgwick’s position can be defended against these criticisms.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

‘Ought’ in Aristotle’s Nicomachean Ethics  185 it inevitably involves our own views about the nature and function of deontic terms. I will set up the issue in Aristotelian fashion by laying out opposing positions on the issue, leading to an aporia, and then marshalling further resources with which to tackle the issue in ways that I hope will do justice to the different contributions. One view, defended at length by Richard Kraut,2 argues that Aristotle uses dei in the Ethics3 in ways that should be translated as ‘should’, rather than, as sometimes done, ‘duty’.4 Kraut argues that Aristotle is giving ‘a general theory of what one should do’, in terms of the good that our deliberations are aimed at achieving. Dei expresses what one should do in any area to achieve one’s good, and this is no different in the area of virtue, since Aristotle does not introduce a special kind of moral goodness to which duty would be appropriate.5 Rather, ‘Rightness and wrongness in every sphere of ­practical activity are explained in terms of success and failure to achieve . . . some good.’ (p. 175). Aristotle does not make a study of dein, as he does of virtue, pleasure, and so on, because dein does not introduce a distinct kind of reason for acting. ‘[N]o one aims at must-ness or ought-ness or should-ness’ (p. 181). Several scholars have argued on the other side. Here there is a strong and a moderate version. The strong view can conveniently be represented by R. Gauthier in his article ‘On the Nature of Aristotle’s Ethics’.6 Gauthier claims that, ‘The “mean” of virtue is thus in the final analysis for Aristotle only the conformity of action to the moral rule.’7 On this view, Aristotle sees deontic demands as prior to and limiting the concern of ­virtue to achieve the good. Gauthier sees these deontic demands in terms of duty. The ­doctrine of the mean does not merely tell us to eat what we need— ‘[t]he moral precept is to eat what it is our duty to eat.’8 On this view strong deontic concepts are prior to and delimit the development of virtue and achievement of the good. The moderate view can be found most clearly in Nicholas White, who argues that Aristotle employs distinctive deontic concepts in his use of terms like dei, ‘should’, but 2  Richard Kraut, ‘Doing Without Morality: Reflections on the Meaning of Dein in Aristotle’s Nicomachean Ethics’, Oxford Studies in Ancient Philosophy 2006, 159–200. 3   In this essay, Ethics refers to the Nicomachean Ethics. 4   Kraut foregrounds Roger Crisp’s translation (Aristotle, Nicomachean Ethics, Cambridge Texts in the History of Philosophy 2000) as an example of translating dein as ‘one’s duty’ (p. 159 and elsewhere); this is, however, as far as I can tell the only passage in which Crisp does use ‘duty’ for dein; elsewhere he uses ‘should’, ‘must’, ‘ought’, ‘required’, ‘right’, and ‘needs’ (in keeping with the relation of dein to deisthai). 5   Kraut insists that Aristotle does not introduce the notion of ‘moral duty’ but unfortunately does not articulate what he understands by ‘moral’. On p. 162 he claims, mistakenly, that Anscombe has had no effect on my book The Morality of Happiness because I use the term ‘moral’ there. This is to mistake what the book was doing; it aimed to show that ancient ethical theories can in fact (against the assumptions of the time) do what contemporary theories do which use the notion of the moral. I would now specify which aspects of ‘the moral’ this is true of. 6  R. Gauthier, ‘On the Nature of Aristotle’s Ethics', in James J. Walsh and Henry L. Shapiro (eds), Aristotle’s Ethics: Issues and Interpretations, Wadsworth 1967, pp. 10–29. 7   ‘On the Nature’, p. 14, italics in original. 8   ‘On the Nature', p. 13, italics in original. Cf. p. 23, where Gauthier says that it is beyond doubt ‘that the imperative which the rule is, does not merely arrange or regulate, but it expresses an obligation properly so-called’ (italics in original). Also p. 21: ‘I believe, then, that one cannot fail to recognize that Aristotle placed a distinct idea of moral “duty” (obligation) at the very center of his ethics.’

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

186  Julia Annas that Aristotle employs both deontic terms and terms for virtue and the good without giving either priority in his ethical thought.9 White insists, rightly as we shall see, that Aristotle often uses dei with the force of ‘ought’ and thus thinks of the achievement of virtue in imperatival terms as well as the idea of attraction to the good. Our understanding of virtue ‘is based on an understanding about what one ought to do’; this is an ‘essential component of his definition of virtue’.10 White concludes that Aristotle is indifferent to what might seem to us a crucial issue. ‘Aristotle is simply not concerned to pronounce decisively on whether the specification of the mean, or of virtuous action, is to be thought of as couched in imperatival terms or attractive ones.’11 Aristotle is not confused, White argues, since he sees no need to find a priority here; his is an ethics of the right and of the good. The real difference between his and contemporary ethics is that Aristotle analyses virtue, happiness, pleasure and friendship, but never turns a similar philosophical scrutiny on ‘imperatival’ terms such as ‘ought’, taking them for granted.12 We have here a conflict of a kind familiar from Aristotle’s own works. We have ­conflicting views, each of which is well grounded and interprets the text in ways that cannot be simply dismissed. I have made use of Kraut and of Gauthier and White to set the issue out clearly, but it can also be seen in the variety of ways in which dein is translated in the many translations that we now have to hand.13 We have a classic Aristotelian aporia—we are stuck and can’t proceed until we provide fresh resources with which to tackle the problem. I propose taking a closer look at Aristotle’s deontic language in the Ethics, starting with dei. A work sometimes cited for Aristotle’s use of dei in the Ethics is Gauthier’s article which has already been cited. Gauthier claims that ‘Setting aside its occurrences in a non-moral sense (approximately 100), Aristotle, in the Nicomachean Ethics alone, employs it [dei] about 170 times in a sense which is uncontestably moral, as is ­confirmed by the many times we are told of the closeness and equivalence of duty and the moral good’.14 Gauthier’s figures for the occurrence of ‘moral’ and ‘non-moral’ dei in the Ethics have been taken over by some philosophers who accept a similarly ‘duty’ centred interpretation of the Ethics.15 The weakness of Gauthier’s claim lies of course in 9   See the article cited in note 1, as well as Individual and Conflict in Greek Ethics, Oxford University Press 2002, chapters 3 and 6. 10 11   Individual and Conflict pp. 113–15.   Individual and Conflict pp. 118–19. 12   Individual and Conflict pp. 122–3. Gauthier denies this, though unsuccessfully (‘On the Nature’ p. 21), claiming that Aristotle gives us his ‘clear and technically elaborated idea of “duty” (obligation) by connecting it with the mean of virtue and what “the right rule” commands’. This is, however, using terms like dei, not giving an Aristotelian account of them. 13   Perhaps surprisingly, the deontologist Ross, though he sometimes inserts the word moral into his translation, does not use ‘duty’ for dein, except in the I 6 passage where Crisp also does (presumably because it is a solemn claim about truth and friendship). Ross generally uses ‘ought’, particularly in passages about the mean. (Ross’s original translation, free from subsequent revisions by others, can be found in volume IX of the Works of Aristotle, Oxford University Press 1966.) 14   ‘On the Nature’ p. 21. 15   Robert B. Louden, Morality and Moral Theory, Oxford University Press 1992, pp. 34–9, especially p. 36 and p. 174 n 35, where he notices that Gauthier does not specify his criteria for a ‘moral’ use; Andrés

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

‘Ought’ in Aristotle’s Nicomachean Ethics  187 the fact that he does not give us any indication of the criteria he employs to distinguish ‘moral’ from ‘non-moral’ uses of dei in the Ethics. I will return to his claim about the use of dei in connection with virtue (Gauthier’s ‘moral good’), which others have made. The figures of 100 and 170 uses are puzzling in themselves, and I have followed up the issue by reading through the Ethics noting occurrences of dei (and other forms of the verb), adding a TLG search which revealed the ones I had missed. Reading the Ethics in this way reveals that besides dei Aristotle occasionally uses chrē (and other forms of the verb) and also often uses gerundive forms of the verb. If we are looking for places where a modern translation of Aristotle’s Ethics can ­reasonably use ‘should’ or ‘ought’, we find the following: There are 316 occurrences of dei (and other forms of the verb16) in the NE: Bk I: 9 Bk II: 28 Bk III: 43 Bk IV: 91 Bk V: 23 Bk VI: 26 Bk VII akrasia: 18 Bk VII pleasure: 3 Bk VIII: 23 Bk IX: 35 Bk X pleasure: 4 Bk X happiness: 15 There are 12 instances of chrē (and other forms of the verb17): Bk I: 5 Bk VI: 1 Bk IX: 3 Bk X happiness: 3 And there are 152 instances of gerundives (poiēteon, etc.) in the NE as a whole: Bk I: 27 Bk II: 12 Bk III: 12 Bk IV: 7 Bk V: 9 Bk VI: 8 Rosler, Political Authority and Obligation in Aristotle (Oxford Aristotle Series), Oxford University Press 2005 pp. 116–44, who seems to be reliant on Louden.   Edei, dein, deoi, deei, deēsei, deon (also in other cases and in plural).   Echrē, chrēon, chrēonta.

16 17

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

188  Julia Annas Bk VII akrasia: 12 Bk VII pleasure: 3 Bk VIII: 13 Bk IX: 33 Bk X pleasure: 7 Bk X happiness: 9 These figures come from my own reading and use of the TLG, and further diligent study might reveal mistakes and disagreements, but some patterns do seem clear. Dei is used far more often than chrē, and I have not found any significant difference in their use, or significance in the parts of the Ethics in which they turn up. A surprise is ­perhaps the number of times we find a gerundive form of the verb, rather than dei or chrē, where the translation requires ‘should’ or ‘ought’. Sometimes a form of a verb is more natural in English also, as for example, ‘We have to . . . ’, and sometimes when a chapter begins with a gerundive we would say, ‘Let us turn to . . . ’ rather than ‘We should turn to . . . ’. Nonetheless if we look for constructions that correspond to ‘should’ or ‘ought’ we find gerundives roughly half as often as dei.18 These results are only suggestive, but they indicate a few interesting points. First, it becomes quite clear that Aristotle’s deontic vocabulary is entirely different from ours. Whereas we have ‘ought’, ‘should’, ‘must’ and ‘required’, as well as ‘right’ in ‘that is right’ or ‘that is the right thing to do’, Aristotle has only the two impersonal forms dei and chrē plus the fact that Greek verbs have a gerundive form which has the deontic force we express by ‘should’ or ‘ought’. It has been remarked, as already noted, that Aristotle analyses virtue, pleasure, friendship, practical wisdom, and akrasia as well as flourishing, but never turns his attention in this way to his deontic notions. It seems persuasive that he takes them for granted because he does not see them as expressing an independently important ethical idea. But even before we get to such a general position we can see that the nature of his deontic terms would not make it obvious to him that there is a single consideration here which can readily be articulated and studied in a unified way. Let us look at a passage at the end of Book 2, where Aristotle is giving advice as to how to achieve the mean. ‘The person aiming at the mean should (dei) first steer away from the extreme that is more opposed to it . . . for of the extremes the one is more and the other less a missing of the mark. And since hitting the mean is extremely difficult, we should take (lēpteon) the lesser evil, doing the second best, as they say . . . We should (dei) consider what the things are that we are easily led astray by; we have different natural tendencies from one another, and this is recognizable from the pleasure and pain that we get. We should (dei) drag ourselves in the opposite direction, since it is by pulling far away from going wrong, as people do who straighten warped planks, that we ­ leasant will arrive at the mean. In everything we should beware (phulakteon) of the p and pleasure; for our judgements of it are not unbiased’ (1109a30–1109b9). 18   Irwin is among the few who have noticed the importance of the gerundive here: see his entry on RIGHT, pages 346 of his translation of the Nicomachean Ethics (Hackett, 1999, second edition).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

‘Ought’ in Aristotle’s Nicomachean Ethics  189 One thing that is clear right away from this and many other passages is that it seems to be indifferent whether Aristotle uses dei or a gerundive to express what we should do. They appear to have for him exactly the same deontic force. Further, if we were to ask Aristotle what exactly this force is—does it, for example, correspond to our ‘ought’ or to our ‘duty’? — he would have to extract the required deontic force from dei and from the gerundive form of the verb. What is it that both dei (and chrē) and the gerundives have in common that makes them both suitable to express deontic force, telling people that they should or ought to do whatever it is? Aristotle never extracts and ­comments on such a single unified notion of deontic force, though it is what he uses throughout the Ethics. It does not occur to him that he needs to do such a thing. We can see from this that Gauthier’s method of going through uses of dei and checking them off as ‘moral’ or ‘non-moral’ is flawed. Apart from the obvious question of what his criteria are for counting an occurrence of dei as ‘moral’ or ‘non-moral', ­singling out dei alone (or even adding the 12 occurrences of chrē) picks out only one way in which Aristotle expresses deontic force; it leaves unexamined what is happening with the gerundives, and what the relation of this is to the force of dei. One place in the Ethics where dei has often been translated in terms of ‘right’ rather than ‘should’ or ‘ought’ is in the accounts of the particular virtues in the third and fourth books. Aristotle often says, of a virtue, that the virtuous person will act cour­ ageously, or feel angry, ‘at the right time, in the right way’ and with several other qualifications. For example, ‘The person who is angry at the right things (eph’hois dei) and with the right people (hois dei) and also as is right (hōs dei) and when and for how long is right (hote kai hoson chronon) gets praised’ (1125b31–3). Some translations have ‘as he ought, when he ought', or ‘as he should, when he should', and so on. We can also find ‘at the right time, in the right way’, and so on. In passages like this, and doubtless in some others, it seems reasonable to use ‘right’ to translate this use of dei. We should, however, be cautious about translations which use ‘in the right way’ and the like, because Aristotle’s terms cannot be translated into the common contemporary idiom of ‘doing the right thing’ or ‘doing the right action’. Aristotle’s deontic vocabulary simply does not let him express this, as we can see by trying to match Aristotle’s terms up with philosophically prominent use of the idea of ‘doing the right thing.’ Ross, in chapter 2 of his most famous book,19 begins by asking, ‘What makes right acts right?’ He goes on to answer this immediately in terms of duty, as though the two were obviously equivalent. Whatever we think of Ross’s question (or his answer to it) it is significant that it cannot be expressed in the Greek of Aristotle’s Ethics. Aristotle can certainly ask, ‘What unifies actions such that either dei do them or they are such-as-tobe-done?’ But the answer to this is, ‘They are virtuous actions’, which is obviously not the answer to Ross’s question. Aristotle has no deontic word which can function, as ‘right’ can in English, as an adjective qualifying a noun for an action. He has no ­‘adjectival’ way of grouping the actions which Ross is grouping as ‘right actions’ in   W.D. Ross, The Right and the Good, Oxford: Clarendon Press 1930, chapter 2.

19

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

190  Julia Annas order to look for a ‘right-making characteristic’ that makes them all right. One good reason for thinking that Aristotle does not have anything like Ross’s question in mind (still less an answer in terms of duty) is that he could not talk about a ‘right action’ without coining new Greek terms. Given Aristotle’s achievement in thinking up new terms in a variety of philosophical fields we can suppose that he could have done it in the Ethics had he felt any need. The fact that he doesn’t, but stays content with his familiar deontic terms, is a good indication that the use of ‘right’ in ‘right action’ is not in his repertoire. Still less is the idea of looking for something common to all right actions making them right. (This closes off one (though only one) reason for holding that Aristotle thinks in terms of duty, since it is right action, or the right thing to do, which is most often taken to be duty.) Some interesting points arise from what has been said so far. First, we might think that dei and chrē are the central deontic words, and that gerundives are in some way weaker, making a less strident deontic claim. Settling this issue would require a long and complex study of all the gerundives in the Ethics, and I cannot produce that here. However, one section of the Ethics is of interest here, and indicates that the above ­suggestion is mistaken. In discussing types of friendly or love relation and their implications, Aristotle discusses cases at Book 9 chapter 2, where one relationship appears to make a demand that conflicts with the demand from another. Should you, for ­example, ransom from slavery the person who ransomed you earlier, or your father (assuming you can’t do both)? What weight should you give to obeying your doctor as against your father? Here we seem to have cases which for us are clearly conflicts of duty and obligation, and if dei has a stronger force than gerundives we would expect it to predominate. However, in this chapter we find only three occurrences of dei, whereas there are 20 gerundives. We should conclude from this, I think, that our intuitions about the use of ‘duty’ and ‘obligation’ are useless here; our use of these terms, which for us are stronger than ‘should’ and ‘ought’, can turn out to correspond to Aristotle’s use of gerundives, rather than dei or chrē. Gerundives can, then, make claims that are deontically strong, and it also turns out that dei can be used for claims that are deontically weak. There are, for example, contexts where dei corresponds to the idea of need rather than the deontic force of ‘should’ or ‘ought’. One example is 1112b24–7, where Aristotle is discussing the nature of deliberation, and comments that if people, in working back from a goal to ways of achieving it, encounter something impossible, they give up—‘for example, if dei money, but it can’t be produced; but if [the way] appears possible, they will try to do it.’ Here the thought is clearly that you give up if you need money but can’t get it; ‘should’, ‘ought’, and so on just do not fit here, and ‘duty’ would be utterly baffling. Similarly, it is completely clear from the continuation of the passage that the query whether the good person dei friends at 1169b3–4 is not whether he ought [to have] friends, or has a duty to have friends, but whether he needs them. Examples like this20 20   Cf. 1154a2–3: we would dei [have need of] phronēsis, even if were not practical. Why would the happy person need (deoi) pleasure, if it were not a good (1154a1–3)?

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

‘Ought’ in Aristotle’s Nicomachean Ethics  191 remind us that dei is linguistically related to the verb deisthai and cognates, whose ­central meaning is that of needing. The notion of need does not always exclude the idea of deontic force, and sometimes, as in contemporary America, can be used as a polite substitute for ‘ought’ or ‘should’, as in, ‘You need to stop doing that’ where there is no suggestion that you actually have a need to stop doing the action, but rather a directive that you ought to stop doing it. But the examples in the Ethics do not seem to be like this. Further, since dei indicates ‘ought’, ‘should’, or ‘must’ with a force which extends to that of need, it does not look like a promising term for duty, or a ‘moral ought’. Can dei ever be used to indicate that something ought to be the case? This is a common contemporary use of deontic concepts. At 1170a4–7 it looks as though we have a case of this. ‘People think that the happy person dein live pleasantly, but life is difficult for someone on their own, for it’s not easy to be continuously active on one’s own, and it’s easier together with others and in relation to them. His activity will then be more continuous, and is pleasant in itself, which dei be the case concerning the happy ­person’. Here the thought is reported from others, and it is simplest to take the final clause as being a restatement of the initial one. This is usually taken as saying that the happy person should or ought to live pleasantly to be happy, but could as well be taken as saying that he needs to live pleasantly to be happy. The claim that something ought (or needs) to be the case concerning the happy person seems thus to be simply a ­syntactical alternative to saying that the happy person ought (or needs) to have it. Aristotle in the Ethics does not appear to say that something ought to be the case in a way that is not merely a restatement of the point that somebody ought to have it, or do it. This is one reason for doubting that Aristotle has our technical concept of a state of affairs, as that is often used in ethical discussions. Of great importance, of course, are the passages where the use of dei is linked with ­virtue and/or the fine, kalon. This is clear from the huge number of dei’s in book 4, where ­particular virtues are discussed, and the comparatively large number in Book 3, where two virtues are discussed at length. As we have seen, vastly diverse interpretations have been made of these passages, from Gauthier’s claim that the deontic terms ­introduce moral duty which constrains the idea of Aristotelian virtue to Kraut’s claim that the deontic terms indicate how to succeed in achieving one’s relevant aim, in v­ irtue and in other endeavours. I will give two examples; the reader has to take it on trust that these are typical. In Books 3 and 4 together I find nine examples of relevant passages, including the following two,21 and in the rest of the work eight.22 Some of these are brief (a deontic term together with a reference to virtue or the fine (kalon), or both), some more substantial. In Book 3 chapter 7 there is a discussion of the brave person, who is undaunted as far as is possible for a human being. ‘He will, then, fear the relevant kinds of thing, but he will endure as he should (dei) and as reason tells him, for the sake of the fine (kalon), 21   These nine (including the ones given here) are: 1115a 10–13, 1115b 11–21, 1116b 2–3, 1117a 29–32, 1119a 11–20, 1119b 15–18, 1120a 23–33, 1121a 1–4, 1121b 1–7. 22   1096a 14, 1106b 21–3, 1107a 12–17, 1108a 26–8, 1146a 27–31, 1165b 15–17, 1169a 11–15, 1171a 24–7.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

192  Julia Annas for this is the goal of virtue (aretē). It’s possible to fear these [frightening] things both too much and too little, and also to fear things that aren’t frightening as though they were. Mistakes come about in fearing something one shouldn’t (ho ou dei), or in the way one shouldn’t (ouch’hōs dei) or when one shouldn’t (ouch’ hote dei) or something similar; the same is true of things to be confident about. The person who endures, and is afraid of (and similarly is confident about) what, and for the sake of what, he should (ha dei kai hou heneka), and also as and when he should (hōs dei kai hote)—that is the brave person (andreios). For the brave person feels and acts appropriately, and as reason tells him. The goal of every activity is acting in accordance with the (relevant) disposition. To the brave person, courage is fine (kalon), and so its goal is likewise, since everything is determined by its goal.23 It is the fine (kalon), then, for whose sake the brave person endures and acts in accordance with courage’ (1115b 7–24). In Book 4 chapter 1 we find the following discussion of people who are generous (eleutherios). ‘Actions in accordance with virtue (aretē) are fine (kalai), and done for the sake of the fine (kalon). The generous person will then give for the sake of the fine (kalon) and correctly (orthōs), to those to whom he should (dei), and as much as he should and when, and whatever else follows on correct giving. This he will do with pleasure, or without pain, for what accords with virtue (aretē) is pleasant, or without pain, but least of all painful. The person who gives to people to whom he should not (dei), or for the sake not of the fine (kalon) but something else, will be called not generous, but some other kind of person. Nor will the person for whom it is painful to give, for he would choose wealth over fine (kalēs) action, and this is not characteristic of a generous person. Nor will he take from a source that he should not (dei); such taking is not characteristic of a person who does not honour wealth’ (1120a 23–34). In these passages, where Aristotle is giving an account of virtues, he sets out for us the nature of the virtue in question by appealing to the deontic term dei, what the person should do or ought to do, and also by appealing to the fine, kalon, the virtuous person’s aim. When we read through these and other passages we find that Aristotle is using both these approaches freely, assuming that they complement each other. It is ­ lausible to read these passages as Aristotle using one approach, the deontic one, to not p set constraints on the other, the appeal to the fine. Nor is it plausible that the appeal to the fine is setting constraints on the force of the deontic terms. Rather, they work together in a coordinated way to make clear what virtue is and what it is for someone to exercise a particular virtue, rather than failing to achieve it. (For Aristotle, this will be in one of two opposed ways, that of excess and that of defect.) Neither the deontic approach nor the approach in terms of the virtuous person’s aim has priority over the other. They both illuminate virtue, and work together.24   The text at 1115b21 is corrupt.   See Irwin’s article, ‘The Sense and Reference of Kalon in Aristotle', Classical Philology 105 (10|0), 381–402. Irwin points out that kalon in Aristotle has a wide range over different kinds of context, but that there is no reason not to take the uses in the Ethics to be distinctively ethical (rather than, as is sometimes suggested, regarding them as assimilated to aesthetic uses). 23 24

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

‘Ought’ in Aristotle’s Nicomachean Ethics  193 One thing is clear from these and other passages, namely that when he uses dei Aristotle is not giving us rules (still less ‘moral rules’) that we can extricate and use if we wish to become virtuous. This has indeed been a complaint of many interpreters, who object that in saying that the brave or generous person does what he should, or ought, Aristotle is merely gesturing vaguely and giving no real guidance as to how to act. However, this is not an objection if Aristotle never intended us to understand virtue and becoming virtuous via the deontic terms on their own, or via the deontic terms functioning independently of, and constraining, the virtuous person’s aim. It is fairly easy to see how both kinds of approach—the virtuous person’s aim and what she should do—work together as a child learns to become virtuous. At first, she has to learn what she should do and not do, without further explanation, but then she wonders why, and is given answers in terms of what the brave or generous person ­characteristically aims at. The idea of becoming a person of a certain character is internalized by coming to see that what you should and should not do is to be understood not on its own but in the context of having certain aims and not others. The ideas of the fine, and of virtue, give substance to, and specify, the idea of doing what you should or ought to. In isolation from this context, directives about what you should do are indeed vague and useless; but Aristotle never intends us to isolate them from this kind of context. Moreover, as the person further develops in becoming virtuous, she will reflect further on what she should and should not do, in the light of virtue and the aims of the virtuous person. The two approaches keep on ­illuminating each other, and expanding the person’s developing understanding of being virtuous. Kraut is right, then, that for Aristotle the deontic terms like dei do not introduce a new kind of reason which is quite distinct from reasons of virtue and the fine. What they do is to introduce a kind of reason which works not on its own but together with reasons of virtue and the fine; we need not follow Kraut, then, in holding that Aristotle has merely a general theory of the good, without regard for the distinctive force of virtue. White is right to point out that Aristotle does not analyse the idea of deontic force to be found in dei, chrē, and the gerundives in the way he analyses virtue, happiness, pleasure, and so on, and also to point out that the reason Aristotle does not do this is because there is no distinct kind of motivational force to be found by looking at them. We can also see why ‘duty’ and ‘obligation’ are unsuited as translations of dei, chrē, and the gerundives. If we try to substitute ‘duty’ or ‘obligation’ for ‘should’ in the above passages we can see immediately that they are quite inappropriate. At this point I would like to step back and ask what larger issue emerges from examining Aristotle’s use of certain terms. It seems to me that Aristotle’s use of terms with deontic force, and their relation to terms for virtue and the fine, has often been approached from a perspective too influenced by contemporary issues which have inclined us to look for answers to questions which Aristotle does not ask, and, as we have seen, sometimes cannot ask. Examining his use of these terms closely may enable us to see why, even when he can, he does not ask them.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

194  Julia Annas One contemporary issue has recently been restated under the heading of ‘virtue and right action’, but it is older than this restatement. It is the issue of whether virtue is to be understood in terms of right action or the other way around. A compact way of expressing the point is to set up an alleged dilemma: Either right action is defined in terms of what the virtuous person does, so that the virtuous person is prior to right action; or the virtuous person is the person who (systematically etc.) does actions which are right. Both options, however, have systematic problems, a result which looks bad for virtue ethics. This dilemma, however, openly relies on the assumption that ­virtue must be prior to right action or vice versa, and we have seen that that is not the way Aristotle thinks. He is not a foundationalist about the structure of ethics (though foundationalist about many other parts of philosophy). For him neither virtue nor action as you should (‘right action’) has priority; to understand ethics we have to understand both, and also how they work together. Another assumption is often made, and it also complicates the issue in a way foreign to Aristotle. This is the assumption that what we should do, or ought to do, has to be our duty or obligation. Gauthier, for example, openly makes this assumption. As we saw above, for him, Aristotle’s position that the temperate person will eat what he should is a position about what it is the temperate person’s duty to eat. This is what I call ‘the fusion assumption’,25 the thought that what I should do, or the right thing to do, has to be a duty. This assumption often includes the thought that what I should do has to be a duty if it is to be moral; otherwise it can merely be prudential advice-giving. In this essay I have avoided using the term ‘moral’, since the issues arising from its use are so large and unwieldy.26 Since Aristotle has no corresponding term or set of terms, introducing the notion of morality into translation or interpretation of him risks failing to get him right. Here I shall merely point out that interpreting Aristotle as having a moral theory, or a theory of moral duty, is a theory-laden project; if we want to understand Aristotle in a way that begins from the terms he uses to give structure to his ethics, it is best avoided. It is a bad idea, likely to lead to mistakes, to translate or interpret Aristotle in terms of morality without first making clear exactly what is meant by morality. There are many theories of morality, which differ greatly, and there is no clear unified view of morality that is pre-theoretical.27 If we refrain from interpreting Aristotle in terms of morality, duty, or obligation, are we doing justice to his ethics? Scholars and philosophers who hold, for a variety of reasons, that Aristotle does lack these ideas sometimes make this a reproach, claiming that virtue alone, and deontic terms which do not express a kind of reason independent of virtue, cannot account for the special particular force that we find in morality,

25   I discuss this in ‘Why virtue ethics does not have a problem with right action', Oxford Studies in Normative Ethics vol.4, 2014, 13–33. 26   I have begun to investigate these issues in ‘Ancient Eudaimonism and Modern Morality', in The Blackwell Handbook of Eudaimonism, ed. Christopher Bobonich 2018, 265–80. 27   I raise this issue also in ‘Ancient Eudaimonism and Modern Morality’.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

‘Ought’ in Aristotle’s Nicomachean Ethics  195 duty, and obligation.28 But if we look at the way Aristotle thinks that virtue and deontic notions function together, we can see that for him virtue is adequately demanding, and his deontic notions strong enough, for him not to need a special kind of reason independent of virtue to make them. He does not think that all examples where the ­virtuous person should do something are cases of what we call duty; this is the characteristically modern fusion assumption. But virtue can sometimes be demanding in a way that we recognize as having the force for us of duty, as when the virtuous person readily risks his life for his city. This is an under-explored area to which I hope to return.29 This essay opens up, I hope, further exploration of how Aristotle’s ethical language works. I hope at least to have raised the issue that we cannot interpret Aristotle ­adequately as a theorist of virtue as long as we bring to bear notions like morality and duty without doing two things first: examining his own language of virtue and of ‘ought’, and exploring our own conceptions of duty, virtue, and what it is for an action to be demanded of us.30

Bibliography Julia Annas. ‘Ancient Eudaimonism and Modern Morality’, in Christopher Bobonich (ed.), The Blackwell Handbook of Eudaimonism, Oxford: Blackwell 2018, 265–80. Julia Annas. ‘Why virtue ethics does not have a problem with right action’, in Mark Timmons (ed.), Oxford Studies in Normative Ethics vol. 4, 2014, 13–33. Roger Crisp (ed. and trans.). Aristotle: Nicomachean Ethics (Cambridge Text in the History of Philosophy), Cambridge: Cambridge University Press 2000. R. Gauthier. ‘On the Nature of Aristotle’s Ethics’, in James J. Walsh and Henry L. Shapiro (eds.), Aristotle’s Ethics: Issues and Interpretations, California: Wadsworth 1967, 10–29. Terence Irwin. ‘Eminent Victorians and Greek ethics: Sidgwick, Green and Aristotle’, in B. Schultz (ed.), Essays on Henry Sidgwick, Cambridge: Cambridge University Press 1992, 279–310. Terence Irwin. Nicomachean Ethics, 2nd ed. Indianapolis: Hackett 1999. Terence Irwin. ‘The Sense and Reference of Kalon in Aristotle’, Classical Philology 105 (2010), 381–96. Richard Kraut. ‘Doing Without Morality: Reflections on the Meaning of Dein in Aristotle’s Nicomachean Ethics’, in David Sedley (ed.), Oxford Studies in Ancient Philosophy vol. 30, Summer 2006, 159–200. Robert B. Louden. Morality and Moral Theory, Oxford: Oxford University Press 1992.

28   Not all accounts of morality do make the special force of duty and obligation central to it (one of the many complications when trying to comprehend the contemporary understanding of morality), but it is central to many of them. 29   It is wrong to hold, with Gauthier, that Aristotle thinks that the virtuous person is, just by virtue of eating what he should eat, eating what it is his duty to eat; but we can think up circumstances when eating something might be demanded by virtue (politeness to a foreign host in a fraught situation, for example). 30   I am grateful for comments to an editor of this volume and also to many friends.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

196  Julia Annas Andrés Rosler. Political Authority and Obligation in Aristotle (Oxford Aristotle Series), Oxford: Oxford University Press 2005. Nicholas White. ‘The attractive and the imperative: Sidgwick’s view of Greek ethics’, in B. Schultz (ed.), Essays on Henry Sidgwick, Cambridge: Cambridge University Press 1992, 311–30. Nicholas White. Individual and Conflict in Greek Ethics, Oxford: Clarendon Press 2002, 83–123 and 216–89. W.D. Ross. The Right and the Good, Oxford: Clarendon Press 1930, 17–64. W.D. Ross. Works of Aristotle, Oxford: Clarendon Press 1966.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

11 Deliberation and Decision in the Magna Moralia and Eudemian Ethics Karen Margrethe Nielsen

The Magna Moralia has long been the ugly duckling in the Aristotelian pond, shunned by critics on account of its ungainly composition, flatfooted argument, and peculiar linguistic habits.1 While the Eudemian Ethics reclaimed its rightful place in the Aristotelian corpus in the early twentieth century after centuries of scholarly neglect,2 and while the Nicomachean Ethics’ authenticity was never in serious doubt, despite Schleiermacher’s attempt to revive Cicero’s apparent view that it is due to Nicomachus Jr., rather than his more famous father,3 the Magna Moralia has yet to be fully integrated into Aristotle scholarship. In the twentieth century, von Arnim, Dirlmeier, Düring, and Cooper all tried to reclaim a place for the MM in the history of the development of Aristotle’s thought4 in response to attacks against its authenticity launched 1   I would like to take this opportunity to thank Gail and Terry for their life-transforming friendship and counsel over the years. Aristotle remarks that ‘it is proper to good people to avoid error and not to permit it in their friends’. Being the very best of people, Terry and Gail have taken a long view on my imperfections, an indulgence for which I will always be grateful. As advisors, they have certainly made a better shoe of this hide than the material portended. 2   Anthony Kenny defends the EE over the EN as the canonical statement of Aristotle’s thought in The Aristotelian Ethics (Oxford: Clarendon Press, 1978). He maintains that the common books (EE IV–VI=EN V–VII) originated in the EE, and were transferred to the EN as late as the first century BC. Kenny does not, however, challenge the Aristotelian provenance of any of the ten Nicomachean books. 3   On Moral Ends V.12, reveals that Cicero is familiar with at least two treatises on ethics by Aristotle. He attributes one to Aristotle’s son Nicomachus on account of its name. Whether this is a version of our EN is ambiguous, however, since some of the MSS of our MM refer to its books as the Megala Nikomacheia, and the books comprising the EN as the Mikra Nikomacheia, presumably on account of their relative size (see also Elias’ commentary, CAG 18,32.34–33.2). Schleiermacher idiosyncratically holds that the MM is the only authentic treatise we have by Aristotle, a view that has not caught on. See F. Schleiermacher, Samtliche Werke, Abteilung III: ‘Über die ethischen Werke des Aristoteles’ (1835), pp. 306–33. 4   See e.g. Hans von Arnim, Die drei aristotelischen Ethiken. Sitzungsberichte der Wiener Akademie der Wissenschaften, Phil.-hist Klasse 202 (1924) 2. Abhandlung; also, his ‘Nochmals die aristotelichen Ethiken (Gegen W Jaeger. Zur Abwehr.)’, Ak. der Wissenschaften in Wien (phil.-hist. Klasse) SB 211.2 (1929); and  ‘Der neuste Versuch, die Magna Moralia als unecht zu erweisen’, Ak. der Wissenschaften in Wien

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

198  Karen Margrethe Nielsen by Jaeger, Walzer, Brink, and Allan.5 Because of its contested status, few scholars ­nowadays pay more than cursory attention to its argument.6 What little attention it receives is typically in the service of showing that it can’t be by Aristotle. ‘Look!’ skeptics say, ‘It’s trying to quack like a proper duck (the EE or the EN), but it doesn’t quite succeed! It must be an impostor!’ Debunking arguments of this type tend to start from the observation that the MM diverges from the EE and EN on points of substance, and conclude that it cannot be written by Aristotle, for he would then have misunderstood his own arguments. Stylistic and lexical arguments are subsequently invoked to support alternative accounts of its provenance, for instance that it is a later epitome of Aristotelian ethics written by a sympathetic but independent-minded thinker, a ­student’s inept summary of the Eudemian or Nicomachean Ethics, or both, or a lightly edited set of notes reflecting a third lecture course on ethics by Aristotle which has since been lost. Views about the authenticity of the MM fall on a continuum, running from outright dismissal to carefully qualified defenses of its authenticity. In a recent survey, Terry Irwin outlines the positions that have been taken on the authenticity of the MM since the 1950s. Allan7 and Donini8 argue that it is spurious: the author is a Peripatetic in the generation after Aristotle, perhaps even Aristotle’s successor Theophrastus. The author used Aristotelian material—either the EE or EN, or a third course, or all three courses—and combined it with his own ideas to produce a new course on ethics. On this view, the MM is a compendium of Aristotelian ethics written in the school of Aristotle, but it is not by Aristotle. Dirlmeier (1964) argues that it was written by Aristotle as a third lecture course, but edited by a later hand (a ‘redactor’). This explains the stylistic and linguistic peculiarities that Dirlmeier thinks must stem from a later date. In place of Dirlmeier’s redactor, Cooper proposes that the MM is a student’s notes from a third lecture course of Aristotle’s that has since been lost. On both hypotheses, the stylistic and linguistic (phil.-hist. Klasse) SB 211.2 (1929); Franz Dirlmeier, Aristoteles Magna Moralia. Übersetzt und kommentiert (Berlin: Akademie Verlag, 1958); Ingmar Düring, Aristoteles. Darstellung und Interpretation seiner Denkens (Heidelberg: C. Winters Universitätsverlag, 1966), esp. pp. 138–44; John M. Cooper, ‘The Magna Moralia and Aristotle’s Moral Theory’, American Journal of Philology (1973), reprinted in his Reason and Emotion (Princeton: Princeton University Press, 1999), pp. 195–211. 5  W. Jaeger, Aristoteles. Grundlegung einer Geschichte seiner Entwicklung (Berlin: Weidemannsche Buchhandlung, 1923) pp. 237–70; also Über Ursprung und Kreislauf des philosophischen Lebensideals, Sitzungsberichte der. Pruss. A. d. W. Philos.-hist. Kl. (1928); R. Walzer, Magna Moralia und Aristotelische Ethik (Berlin, 1929); K. O. Brink, Stil und Form der pseudoaristotelischen Magna Moralia (Ohlau: Eschenhagen, 1933) D. J. Allan, ‘Magna Moralia and Nicomachean Ethics’, Journal of Hellenic Studies 77 (1957), pp. 7–11; Kenny, The Aristotelian Ethics (Oxford: Clarendon Press, 1978). For a full overview of the articles and reviews in the Jaeger—von Arnim controversy, see A.-H. Chroust, ‘The First Thirty Years of Modern Aristotelian Scholarship (1912–1942)’ (1963), reprinted in W. R. Wians (ed.), Aristotle’s Philosophical Development: Problems and Prospects (Rowman and Littlefield, 1995) pp. 41–66, especially the references in note 44 (p. 54). 6   For a survey of ancient and Medieval sources, see F. Dirlmeier, Einleitung, Aristoteles Magna Moralia (1964), pp. 93–115. 7   See D. J. Allan, ‘Magna Moralia and Nicomachean Ethics’, Journal of Hellenic Studies 77 (1957), pp. 7–11. 8  Donini, L’etica dei Magna moralia (Torino: Fondazione Parini Chirio, Giappichelli 1965).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  199 peculiarities of the MM are explained by the presence of a second hand revising Aristotle’s manuscript. By contrast, Kenny has argued that the author is a student of Aristotle who wrote a summary with the EE and EN in front of him. His only independent contribution is the presentation, as well as the mistakes he makes while attempting to faithfully capture Aristotle’s argument.9 A fifth view, outlined by Irwin, is that the MM just is the third course: ‘Aristotle is the author in the sense in which he is the author of EEN’. Allan’s conjecture—the author of the MM stands at a ‘considerable distance’ from Aristotle—rests on what he takes to be a series of literal citations in the MM from the EN (specifically at MM 1185b14–87a4, which closely parallels EN 1104a11–19 and b8–12, 1103a17, 1105b19, and 1109a1), its supposedly post-Aristotelian terminology, as well as stylistic and doctrinal discrepancies.10 Textual parallels could equally be explained by the reversing the causal arrow, however: if the MM is an earlier treatise, the EN could incorporate material from it, rather than the other way around. And the idiosyncratic lexis that authors like Dirlmeier claims must date from Hellenistic times is in fact attested in contemporary sources. In some cases, purportedly ‘Stoic’ terminology can also be found in the EE and EN. In others, we can’t preclude that Stoic use is influenced by the MM, or that the popularity of some of the terms may be explained by a common Platonic source.11 What about the MM’s distinctive style? Dirlmeier explains its peculiar style with recourse to a hypothesized skilled redactor working at some remove from a conjectured third lecture course that has since been lost. Dirlmeier’s hypothesis is not very compelling, however. To introduce the stylistic peculiarities of the MM, the hypothetical redactor would have had to go through the manuscript systematically, replacing peri with huper, adding rhetorical questions, direct speech, and all the other distinctive features of the MM. But this is hardly the 9   Because he thinks the MM is based on the EE, and because he thinks the EE is Aristotle’s final work on ethics, written in the last years of his life, Kenny rejects the MM as spurious—it falls off the temporal cliff, so to speak. 10   Allan quotes the list compiled by Dirlmeier in his earlier article ‘Die Zeit der Grossen Ethik’, Rheinisches Museum, 88 (1939), pp. 214–43, including (1) prothetikos (two instances in MM I. 18, of virtue as setting an end before itself); (2) epiteuktikos; (3) katorthōma; (4) apokatastasis. Re (1): The EN uses ‘prosthesis’ at 1115a1 and kata prosthesin at 1148a10 and 49a17 in slightly different senses; the EE prostithenai at 1215b29 and 1224b6. More relevant is the EE’s reference to the person capable of living in accordance with decision (prohairesis) ‘setting up’ (thesthai) a target to aim at (1241b7) and the EN’s reference to deliberation occurring when the end has been laid down (themenoi to telos, 1112b15). Re (2): epiteuktikē is used once in the MM, at II 3 1, and not as a technical term. It is a hapax in the corpus, but epiteuxis can be found in the Platonic Definitiones at 413c in a similar sense. Re (3), we find the verb katorthoun in the sense associated with katorthōma in the EE 1247a4, 13, 15, 24, 30, 47b25, 27, 31, 35, 48a14, 30, 48b4, and in the EN at 1098b29, 1106b31, 1107a14, 1142b30; katorthousthai at 1106b26, and katorthōtikos at 1104b33. Re (4): In Aristotle’s Categories, we find apokathistanai (9b15), apokathistamenōn (9b28), and apokathistēsin (13a30). The latter describes changes into contraries, specifically how a person is changed over from vice into a virtuous state. We thus have an attested Aristotelian use of the verb in a very early work, and also in the context of ethics. The noun is attested in Epicurus’ first letter, and we also find it in Platonic spuria. 11   I here paraphrase one of Irwin’s unpublished MS for the seminar on Aristotle’s Three Ethics (Hilary 2014), titled ‘doc.1b’, p. 5. I am grateful to Terry for sharing his work in progress on the three Ethics.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

200  Karen Margrethe Nielsen way to revise a manuscript for publication. As Cooper notes: ‘Dirlmeier’s hypothesis of an “editor” who published Aristotle’s own lecture notes with a few light alterations of language is untenable. The very mention of an editor who not only touches and patches here and there, but regularly supplies some of the prepositions and adverbs is really quite suspect. No other works of Aristotle, all of which, one supposes, were at least lightly edited, shows any such features’ (p. 333). But by parity of argument, it is hard to see how a student taking notes on a third course could have ended up adding all the stylistic peculiarities, as Cooper himself proposes. Why, in revising his notes, should this student change prepositions at the start of sentences? Why add direct speech? In the end, Cooper’s hypothesis faces the same problems as Dirlmeier’s. In light of these problems, we should consider a fifth possibility: that the MM is by Aristotle ‘in the same way that the EE and the EN are by Aristotle’ (Irwin). Instead of placing the MM in the Hellenistic period, explaining its style and lexis by a late date of composition, it may be an early treatise, perhaps stemming from Aristotle’s period in the Academy, but, in any case, predating the EE and the EN. Its style would then reflect Aristotle’s youth and (perhaps) his Macedonian origins. A recent immigrant to Athens, he may not yet have developed his trademark style.12 Whether we accept this explanation or not, we should agree with Cooper when he remarks: [I]f Dirlmeier is right, and the Magna Moralia, at any rate in content, is the work of Aristotle himself, composed earlier than either of the other two extant expositions of his moral ­philosophy, we have in this work an extremely valuable tool for tracing out Aristotle’s philosophical development, as also for the interpretation and evaluation of his moral philosophy itself (p. 195).

In this essay, I will leave the Nicomachean Ethics to one side, and examine the relationship between the Magna Moralia and the Eudemian Ethics.13 The MM and EE seem 12   Though some scholars think proper names entail a date after 335 BC, the evidence is inconclusive. Kenny (1978, pp. 216–20) takes the past-tense reference to ‘Mentor’s cleverness’ at 1197b21 to suggest a date after the death of Mentor the Rhodian in 337/6. He thinks the ‘cleverness’ alludes to Mentor’s entrapment of Aristotle’s patron Hermeias in 341/0 rather than to some generic trait of some Mentor or other. ‘Darius’ at 1212a4 Kenny presumes to concern Darius III (reigned 336–30) rather than I or II or a namesake. ‘Lampros and Neleus’ at 1205a19–23 picks out Coriscus’ son Neleus of Scepsis (born c. 350), Kenny argues, and since he is said to be knowledgeable of grammar, he can’t be much younger than ten. Some MSS have ‘Ileus’ instead of ‘Neleus’, however. For a critique of Kenny’s inferences, see Irwin, doc. 1b (unpublished). In fact, it is harder to pin down the referents of proper names than is often assumed. We can’t take for granted that the (to us) most famous bearer or the (to us) most famous incident is meant. Witness Aristotle’s disambiguation in the EN between ‘Socrates the Elder’ and ‘Socrates the Younger’, an obscure member of Plato’s Academy. The Greek tradition of naming sons after their paternal grandfathers (exemplified in Aristotle’s own family) complicates matters further. And even if we were to grant some of Kenny’s arguments, this does not preclude a date of composition in the 350s. Aristotle may have revised his manuscript late in life, adding new examples to keep his lecture course current. 13   Translations of the three Ethics mostly follow Irwin; for the Nicomachean Ethics, I rely on T. H. Irwin (transl.), Aristotle. Nicomachean Ethics, 2nd ed. (Indianapolis: Hackett, 1999);  for the Eudemian Ethics, I also consult A. Kenny (transl.), Aristotle. The Eudemian Ethics, (Oxford: Oxford University Press, 2011). For the MM, I draw on Irwin (unpublished).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  201 closer in structure and argument than the MM and EN, and so one might be led to infer, with Kenny, that the MM is a summary of the EE, with a handful of passages interspersed from the EN. Against Kenny’s conjecture, I wish to draw attention to a significant shift in Aristotle’s conception of deliberation (bouleusis) and decision (prohairesis), from the MM to the EE. The overlap between the real definitions of decision as a deliberate desire (orexis bouleutikē)14 hides substantial disagreement over the meaning of prohairesis, captured in the divergent nominal definitions at MM 1189a12– 17 and EE 1226b5–9. This disagreement can in turn be traced back to different accounts of the nature of deliberation, with the EE defending an account of decision that is closely aligned with Aristotle’s teleological account of action (praxis) and his account of happiness as the final end. The model of deliberation in the MM is not so closely aligned.15 Most strikingly, perhaps, only the EE defines virtue as a state that decides (hexis prohairetikē), a definition that marks the culmination of Aristotle’s analysis of virtue of character in EE book II (it plays a similar role in the EN). The MM omits this phrase altogether from its definition of virtue of character, leaving the significance of decision for virtue obscure.16 Instead, it is prudence that is defined as a hexis tis prohairetikē in the brief and disjointed chapter (MM I 33) corresponding to the analysis of prudence in EE V. Kenny holds that ‘with a few exceptions’ differences in content between the MM and EE ‘can be explained by a combination of a pardonable degree of incompetence in understanding, and a modest talent for editorial revision, on the part of a disciple of Aristotle present at the lectures’ (Kenny, p. 220). The MM, says Kenny, follows the EE ‘slavishly in the order of topics and the manner of their treatments, for almost the whole of its length’ (p. 220). I believe these claims are unfounded. It is simply not possible to construe the analysis in the MM as an attempt—however inept—to capture the gist of Aristotle’s discussion of prohairesis in the EE. Kenny’s conjectured note-taker would have produced a treatise that bears little more than a passing resemblance to the original.17 Kenny accuses the author of the MM of ‘constant botching of the details of a 14   Both treatises define decision as a deliberate desire (orexis bouleutikē, EE II.6, 1226b17; orexis tis bouleutikē, MM I.17, 1139a33–4), with the EE adding that it is of acts that are up to the agent (tōn eph’hautōi), and the MM that is arises through thought (meta dianoias). 15   Indeed, Allan argues that virtue, rather than happiness, is the fundamental concept of the MM (p. 7). This means that the MM is closer than the EE or EN to Plato’s Republic. 16   My own view is that the EN is the latest of Aristotle’s three lecture courses on ethics, but I will not defend that view here. 17   I have elsewhere attempted to show that Aristotle’s analysis of virtue of character in general in the MM is less detailed and less circumspectly worked out than the parallel passages in the EE and EN (unpublished MS). Compare, for instance, the stone and adultery examples. The MM does not distinguish between the ‘arithmetic’ mean and the mean ‘relative to us’ (contrast EE and EN) and there is no mention of the parameters ‘at the right time, about the right things, etc’. The striking example of Milo the wrestler is absent. The author of the MM furthermore does not give a name to the virtues of the rational part of the soul (what the EE and EN call dianoetikē arēte) but instead leaves it an open question in book I whether there are such things as virtues of the rational part of the soul. The weight of these omissions makes it hard to believe that the author of the MM had an MS of either the EE or EN available to him or heard a lecture course based on either.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

202  Karen Margrethe Nielsen philosophical argument’ (p. 219). But botching is as botching does—it is only if we agree that the author was trying to reproduce the argument of the EE that he can be said to ‘botch’ the job. If the purpose of the author was not to reproduce Aristotle’s argument in the EE or EN, quite simply because the EE and EN had not yet been written, then he can hardly be accused of ‘botching’ the details. Kenny reads the MM as a treatise that tries to be like, but falls short of, the EE’s perfection. But that may be a delusion of perspective—the MM may resemble the EE because it is the teenage version of the grown treatise—less nuanced, more bombastic, and with unevenly developed limbs. If its voice breaks here and there, that’s simply because it is immature.

Prohairesis in the History of Greek Philosophy In their landmark commentary on the Nicomachean Ethics, the French scholars Gauthier and Jolif remark that Aristotle’s role in the history of prohairesis was to transform a term that was ‘still rare and with an imprecise sense’ into ‘a technical term, expressing a new concept’ (Gauthier and Jolif, II, 1, p. 190).18 Aristotle is the first Greek philosopher to put the noun prohairesis to systematic use. Plato uses it only once, at Parmenides 143c. He uses the verb prohaireisthai 33 times, but not systematically, and outside the context of ethics.19 There are scattered uses of the verb in the fragments of Democritus, and the occasional use in Xenophon’s Memorabilia.20 But overall, there is no evidence to suggest that any philosopher prior to Aristotle made the noun—or the verb—the linchpin of their ethics or psychology of action.21 By contrast, in the EE prohairesis and its cognates occur 133 times in 101 Bekker pages (incl. CB). In the MM we find 50 instances in 63 pages, all but one in book I. Thus, the EE uses prohairesis and its cognates with almost twice the frequency of the MM. This is not simply a matter of style. If the MM is a student’s attempt to summarize the EE, the student would have had 18   It is striking that De Anima does not make use of the notion at all in its explication of the origin of action. The term only occurs once, in book I, and not in the narrow technical sense specified in any of the ethical treatises. Instead, De Anima makes use of the Platonic tripartite distinction between epithumia, thumos, and boulēsis. 19   Examples of use of the verb in the context of the method of division (diairesis): Sophist 251e1; 255e1; Statesman 257c1; 279b5; Parmenides 136c3 (with the noun at 143c3). 20   Of the 16 instances of the verb in the Memorabilia, the most striking are Mem. 2.7.10; 3.9.4; 4.5.7; and 4.8.11. These correspond to the ethical usage in the MM. The Memorabilia was likely composed in the years after 371 BC, and in any case before Xenophon’s death in 354. 21   A qualification: Aristotle’s treatment of prohairesis in the EE II.10 starts from the views of ‘some people’ (tines) that prohairesis is desire (orexis) or belief (doxa), a dialectical approach that is mirrored in the EN (III.2. 1111b12: ‘ . . . those who say . . . ’; 1112a1: ‘ . . . now presumably no one even claims that decision is the same as belief in general . . . ’). This suggests that Aristotle intervenes in an ongoing debate about the nature of prohairesis, though no extant texts or testimonia chronicle the positions taken before Aristotle. The surviving speeches of the Attic orators contain a wealth of instances from the 350s on, as Danielle Allen observes in ‘Talking about Revolution: A Political Change in Fourth-Century Athens and Historiographic Method’, in S. Goldhill and R. Osborne (eds.), Rethinking Revolution through Ancient Greece (CUP, 2006), p. 184. Pace Allan, I do not think Aristotle’s public lectures on rhetoric inspired the orators. A more plausible hypothesis is that Aristotle gives a precise, technical sense to a term already current in forensic speeches and the post-Platonic (Speusippean) Academy.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  203 to systematically excise references to prohairesis from the text, removing the concept altogether from the discussion of akrasia and friendship, and dramatically downplay the significance of prohairesis in the discussion of virtue of character at the core of book I. But what would his motive be? If the author thought the theory of prohairesis in the EE added nothing of substance to Aristotle’s argument, then he would surely have removed more of them, and it is in any case clear that Aristotle would beg to differ. It is more likely that the theory of prohairesis in the EE is Aristotle’s own attempt to systematize and revise scattered remarks in book I of the MM.22 Each of the three ethics contains a section devoted specifically to the clarification of the nature of prohairesis: EN III. 2–3, EE II.10, and MM I.17. But whereas the EE paves the way for the analysis by first making substantive claims about the relationship between virtue and decision in the books leading up to the topical discussion, the remarks that precede the topical analysis of prohairesis in MM I.17 are unsystematic and scattered. And whereas the EE follows the topical discussion of prohairesis by defining virtue as a hexis prohairetikē, the MM omits the differentia ‘prohairetikē’ from its definition of virtue of character. All we learn from the analysis of virtue of character in MM I.5–9 is that virtue is a mean state of our feelings (mesotēs tōn pathōn), positioned between ‘excess’ (huperbolē) and ‘deficiency’ (elleipsis) (MM I.8 1186a33–4), where excess is described in purely quantitative terms as ‘too much’ (pollōn) and deficiency as ‘too little’ (oligōn). There is no mention of prohairesis. In addition to these discrepancies, the analysis of prohairesis in the MM seems poorly integrated into the discussion of individual virtues of character. The MM makes only two mentions of prohairesis in connection with specific virtues in I.18–33—both in the discussion of justice—whereas the other treatises integrate the general theory of virtue as a state that decides into the treatment of individual virtues of character (e.g. EE 1230a27 on courage; EE 1233a32 and 37 on magnificence; EN 1117a5 on courage; EN 1125b11 on the virtue concerned with small honours; EN 1127b14 on truthfulness; CB 1134a2; 20; b12; 1135b25; 1136a1; 1137b35; 1138a21 on justice). The EE and EN further distinguish mere feelings from virtues by observing that, whereas states rely on decision, feelings do not (EE 1234a24–32; EN 1106a3–7). In the EE, this leads Aristotle to exclude righteous indignation, friendliness, truthfulness, and wit from the list of virtues (EE 1234a24–5), since they do not involve prohairesis, and are mean qualities rather than mean states.23 And whereas the analysis of akrasia in MM ends without a single mention of prohairesis or its cognates—the term is entirely absent—prohairesis is central to the definition and discussion of akrasia in CB VI/VII. The EE and EN 22   Aristotle uses the verb in a methodological sense in the first line of the MM, referring to the object of inquiry (epeidē prohairoumetha legein huper ēthikōn, proton an eiē skepteon tinos esti meros to ēthos). This terminology reflects Plato’s use of the verb in describing a step in the method of division in his later dialogues (e.g. Parmenides, Sophist, and Statesman). Dirlmeier mentions the opening of Politics II (1260b27) as a parallel, but there are others, e.g. EN I. 13, 1102a13, with the back reference at EN X. 9 1179a35. 23   In the EN, friendliness, truthfulness, and wit are treated as if they were virtues (note the mention of prohairesis at 1127b14). Righteous indignation (nemesis) is included in the catalogue of virtues in EN II.7, but not discussed later.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

204  Karen Margrethe Nielsen f­ urthermore make repeated use of the term in the discussions of friendship, while the MM only uses it once in this connection (1212a21).24 In short, the theory of decision that we encounter in the MM seems underdeveloped and poorly integrated into the treatise as a whole compared to its EE and EN counterparts. If the author has important insights about the relationship between decision and virtue, he does not show his hand, though one could see how highlighting the connection will help clarify several claims the author makes in the MM. The topical analysis of prohairesis in EE II.10 (and EN III.2–3) is expanded in the common book treatment of virtue of thought. The analysis in MM I.17 is likewise developed in the discussion of prudence and right reason (orthos logos) in MM I.34.25 The MM here connects prohairesis closely to prudence, though without offering any explanation for this choice. There is no parallel in the EE or EN to the reference in MM I.34 to a ‘prohairetic’ part of the soul, a label the MM uses unselfconsciously. The EE and EN call this the ‘deliberative (bouleutikon)’ or ‘calculating (logistikon)’ part, whereas the MM alternates between ‘to prohairetikon’ and ‘to bouleutikon’. More strikingly yet, the MM defines prudence, rather than virtue of character, as a ‘state that decides’. It is ‘a certain state that decides on and does those acts that are up to us to both do or not to do that promote advantage’ (hē phronēsis an eiē hexis tis prohairetikē kai praktikē tōn eph’hēmin ontōn kai praxai kai mē praxai, hosa eis to sumpheron ēdē sunteinei, MM I.34, 1097a14–16). But why decision should be included in the definition is never explained. The CB definition of prudence, by contrast, omits any explicit reference to decision—the connection is only made obliquely, by way of the claim that prudence presupposes virtue of character (VI. 12–13). But that is as it should be, as the EE and EN have already defined virtue of character as a state that decides. It would be remarkable, however, if a compendium writer aiming to capture the gist of the EE were to remove the definition of virtue as a state that decides and instead insert this as the definiens of prudence in the analysis of what the EE and EN (but not the MM) call ‘virtues of thought’. This is not editorial revision—whether talented or not. Instead, it appears to be the result of deliberate reconsideration and revision of an argument. And given the propensity of ancient commentators and compendium writers for paraphrase and direct quotation, it would be surprising if the revisions were carried out by someone other than the author. A line-by-line comparison of the paragraphs concerning prohairesis would take us too far afield, and so I shall limit my analysis of the differences between the MM and EE 24  The EE makes extensive use of the notion of ‘antiprohairesis’ (1236b3; 37a31; 32) in describing the friendship of good men. The lone remark about prohairesis at MM II.12, 1212a21 concerns concord; we are in concord not simply by sharing the same beliefs, but by having the same practical intention (prohairesis) (e.g. that one of us should hold office). There is no indication that this prohairesis is ethical in nature. 25   MM I.34 aims to justify treating wisdom and other perfections of the rational part of the soul as virtues, whereas CB V/VI simply assumes that virtues of thought exist. Note the apparent contradiction between MM I.34, which treats prudence as praiseworthy, and hence a virtue, and the claim seemingly endorsed in MM I.5, namely that no one is praised with respect to the rational part of his soul—for being wise or prudent. MM I.5 appears to locate all virtues in the non-rational part of the soul insofar as it is subservient to the rational.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  205 to the accounts of the meaning of the name prohairesis in the two treatises (the ‘nominal’ definitions that shape the search for real definitions in each treatise). Insofar as the nominal definitions offer an insight into Aristotle’s conception of deliberation, they are not merely nominal, but affect Aristotle’s substantive conception of prohairesis. I will show that the MM understands the prefix pro in the preferential sense, analysing a decision as a choice of one course of action in preference to another, whereas the MSS of the EE instead treat the ‘pro’ in prohairesis as equivalent to pros: what we decide to do is chosen for the sake of (pros) the end we have laid down. This shift places deliberation and decision in an explicitly teleological framework in the EE: we make a decision, not when we form a preference for ‘this course of action instead of that’, but rather when we undertake to pursue an action for the sake of an end.26 Modern translators and editors elide the difference, amending the MSS of the EE to fit the preferential sense evident in the MM. But if we instead trust the manuscripts of the EE, it becomes clear that the author of the MM cannot be attempting to reproduce or ­summarize the EE analysis of deliberation and decision, since he would then be systematically misrepresenting the text. There are limits to how inept the note-taker can be before we start doubting his existence.

Prohairesis in the Magna Moralia In its topical analysis of prohairesis, the MM attempts to establish the nature of decision by elimination (an approach mirrored by the EE and EN). The author first notes that decision is not the same as orexis unqualifiedly. For non-human animals have orexis, but prohairesis is ‘meta logou’, or with reason. The MM here only considers only two candidates—desire ‘in general’ and wish. In contrast to the EE and EN, it does not mention the threefold Platonic classification of orexis into epithumia, thumos, and boulēsis, but implicitly dismisses the identification of decision with the first two on the grounds that they are shared with non-rational animals. Thus, decision cannot be the same as orexis unqualifiedly. But neither is it the same as wish (boulēsis) (a rational kind of orexis), for wishes are also for impossible things (like being immortal), whereas decisions are not. Furthermore, we decide, not about the end, but about the things that are toward the end (pros to telos). We wish, by contrast, for ends. Though the MM here rules out identifying decision with wish unqualifiedly speaking, nothing in the argument precludes the possibility that decision may be—or rely on—a particular kind of wish, namely a wish for something that is possible. The EE is more circumspect, for there, wish is said to be malista or ‘most of all’ for ends (1226a13). This is repeated in the EN, which states that it is mallon or ‘primarily’ for ends (1111b26). This leaves open the possibility that decision may be a kind of wish, though its primary object is the end. 26   I examine Aristotle’s conception of deliberation in ‘Deliberation as Inquiry: Aristotle’s Alternative to the Presumption of Open Alternatives’, The Philosophical Review, vol. 120, no. 3 (2011), pp. 383–421.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

206  Karen Margrethe Nielsen The MM now introduces a consideration of the name prohairesis in an attempt to clarify its nature (I quote the Greek for ease of reference). Though it is not an attempt to give an etymology, it serves as a nominal definition. NOMINAL DEFINITION MM: But decision would seem to be as its name is also—namely, we decide on this instead of that— namely, the better instead of the worse. Whenever, then, we take the better instead of the worse that is in our choice, in that case deciding would seem to be appropriate (MM1189a12–17). (ἀλλὰ ἡ προαίρεσις ἔοικεν οὕτως ἔχειν ὥσπερ καὶ τοὔνομα αὐτῆς ἔχει, οἷον προαιρούμεθα τόδε ἀντὶ τοῦδε, οἷον τὸ βέλτιον ἀντὶ τοῦ χείρονος. ὅταν οὖν ἀντικαταλλαττώμεθα τὸ βέλτιον ἀντὶ τοῦ χείρονος ἐν αἱρέσει ὄντος, ἐνταῦθα τὸ προαιρεῖσθαι δόξειεν ἂν οἰκεῖον εἶναι) (alla hē prohairesis eoiken houtōs echein hōsper kai tounoma autēs echei, hoion prohairoumetha tode anti toude, hoion to beltion anti tou cheironos, hotan oun antikatallattōmetha to beltion anti tou cheironos en hairesei ontos, entautha to prohaireisthai doxeien an oikeion einai)

The word prohairesis indicates that what we decide to do is τόδε ἀντὶ τοῦδε (tode anti toude)—more specifically, the better in preference to the worse. If we take the first hoion to mean ‘i.e.’ rather than ‘e.g.’, Aristotle’s point is that the pro in prohairesis essentially denotes having a preference for one thing over (anti) another. The anti thus gives the sense of the preposition pro. This is a sense of the verb prohaireisthai that we find e.g. in Xenophon. In his Memorabilia, prohaireisthai is used to denote the act of preferring one thing to another, and consequently the verbal adjective prohaireton/prohaireteon is used to state that one course of action or outcome is preferable to another.27 Here, the verb prohaireisthai denotes the (mental) act of choosing between alternative courses of action after comparing their merits. At Mem. 4.5.7 and 4.8.11, the act of making a decision is likewise said to consist in forming a preference for one thing over another—ti anti tinos. This interpretation finds its counterpart in the MM’s analysis of deliberation. The MM next asks whether decision is the same as what is in accordance with thought (kata dianoian). It rejects that proposal as well. For we think about many things, including the affairs of the Indians, but we do not decide on them. In the EE, being kata dianoian is only considered as a possible definition of the voluntary, rather than, as in the MM, as a possible definition first of the voluntary, and subsequently of decision. In lieu of kata dianoian, the EE offers doxa as a possible definition of ­prohairesis (the EN follows suit). It is hard to see why a note-taker would switch the terms around—it cannot be a mere scribal error. Another difference: in the EE, Aristotle uses the example of the Indians later in the analysis, when he elaborates on his conception of deliberation. We do not deliberate about the affairs of the Indians. 27   Thus, Socrates is said to have held that ‘if someone were to do something shameful, death would be preferable (prohaireteon ēn)’ (Mem. 2.7.10). Xenophon later employs the verb to express the Socratic paradox that temperance is wisdom: ‘I believe everyone does whatever actions he prefers (prohairoumenous) among those available, insofar as he thinks they are most beneficial to himself ’ (Mem. 3.9.4).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  207 The point is not made in the parallel context of the EE’s arguments against identifying prohairesis with doxa (1226a29).28 Having excluded desire and thought individually, the MM now considers whether decision could be the product of both of them combined, a claim it affirms. This yields the MM ‘real’ definition of decision as ‘a certain kind of deliberate desire with thought’: orexis tis bouleutikē meta dianoias (1289a33–4). Thus, very quickly, the MM has established the nature of decision. In the next paragraph, the MM proceeds to explain how decision arises by a process of thought and deliberation. It argues that decision concerns actions—things that promote the end rather than the end—but it does not analyse deliberation as an inquiry that seeks means to ends, instead describing deliberation as the act of weighing alternative courses of action. This leads us to form a preference for one over another: Since, therefore, as was said before, decision is about the goods towards the end and not about the end, and about things that are possible for us, and about things that present arguments on each side about whether this or that is choiceworthy (kai tōn antilogian paradidontōn poteron touto ē touto haireton), it is clear that one must previously think on them and deliberate, and then, whenever it appears better to us after having thought, in this way there is some impulse to do it, and in doing this we seem to act in accord with decision  (MM 1189a25–32).

The MM assumes that deliberation amounts to comparing at least two options to find the best. There must be arguments on either side about whether this or that is choiceworthy. This lends support to treating the first ‘hoion’ above in the passage presenting the nominal definition as equivalent to ‘i.e.’ When the MM defines decision as a certain kind of deliberative desire with thought, it invites us to understand the deliberation in question to concern whether this act or that act ought to be done, when settling this question happens as a result of thought. The MM concludes that: ‘Decision, therefore, is in things that can be done, and in those in which it is up to us to do and not to do, and this way or not this way, and in which it is possible to grasp the why (to dia ti)’. Specifying ‘the why’ (to dia ti) is a task that the EE takes far more seriously than the MM, however, for the MM does not explain how the preferential aspect of prohairesis relates to this ‘teleological’ aspect. Instead, the MM treats deciding between alternative as the characteristic task of deliberation, and defines prohairesis in line with this comprehension. But at the same time, the author also recognizes that there’s a difference between wish and decision, not just because the former extends to things that are impossible, and ‘not in our decision’, as the MM puts it, but more importantly because we wish for ends but decide on things that promote ends. Deliberation, then, aims to ‘grasp a cause’. Reaching back to a geometrical example that prefaced the discussion of the voluntary, the MM observes: Now the why is not of just one type. For in geometry, whenever someone says that the quadrilateral has equal to four right angles, and one asks ‘Why?’, ‘Because’, he says, ‘also the 28  The EN omits mention of the affairs of the Indians altogether. Instead, it says that Scythians do not deliberate about the affairs of the Spartans. Alexander’s Indian campaign began in 326 BC—Aristotle died in 322 BC.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

208  Karen Margrethe Nielsen triangle has equal to two right angles’. In such cases, then, they grasped the why from the defined origin. But in matters of action, in which decision is, it is not so (for no defined origin is laid down), but if one asks ‘Why did you do that?’, ‘Because it was not possible any other way’, or ‘Because it was best this way’. From the results themselves, whichever appear to be better, he decides on these things, and because of these things.  (MM 1189b9–18)

The act of deciding here follows from the consideration of alternative courses of action. We decide, either by noting that only one act leads to the desired end, or by forming a preferential judgment for one course of action over another. Despite the overture regarding principles, which one might expect to lead to considerations about the desire for happiness as the ultimate principle, the MM never explicitly connects the dots. Instead, finding the best course of action among alternatives is treated as the essential task of deliberation. It may be that the ‘best’ is determined with reference to the agent’s conception of eudaimonia, but the MM does not say.

Prohairesis in the Eudemian Ethics Turning to the EE, one is immediately struck by its constant emphasis on the teleological aspect of prohairesis, which is here said to be what defines it. The EE introduces the notion of prohairesis at the start, in book I.2, where Aristotle alludes to his archer ­analogy to justify his inquiry into the nature of happiness: About these things, then, having attended to the fact about these things that every one who has the power to live according to his own decision (hapanta ton dunamenon zēn kata tēn hautou prohairesin) sets up some target of living finely (thesthai tina skopon tou kalōs zēn) (whether honour or reputation or wealth or culture), with reference to which he will then do all his actions, since not to have one’s life organized towards some end is a sign of much folly, we ought, then, above all first to determine in oneself without rashness or dilatoriness in which of the things of ours living well is, and what are the things without which it cannot belong to human beings.  (EE I.2 1214b7–14)29

What does it mean to ‘live in accordance with one’s own prohairesis’? The answer is not evident from the passage, but it will emerge later in the treatise that it requires acting on the basis of one’s own choice of means towards an end (telos) that one takes to be fine. Thus, the agent’s prohairesis reveals both what kind of end the agent takes to be fine, and his reasons for pursuing specific courses of action in the pursuit of that end. A person who lacks this ability may be either someone who is deficiently rational, for instance due to youth; because he is a natural slave; or because his choices are ­constrained through coercion, as in the case of non-natural slaves compelled to obey. By Aristotle’s argument in the EE, to be able to decide upon a life is not simply to endorse a certain overarching end as good—to adopt a conception of living well—but also to be able to act on it. Aristotle later states that a true politician is someone who 29   There is a dispute about the exact reading (dei thesthai versus thesthai) that I won’t go into here. I give the version supported by the MSS.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  209 decides on fine acts for their own sake (ho men gar politikos tōn kalōn esti praxeōn prohairetikos autōn charin, 1216a24–5). And he maintains that although an act can be coerced, a prohairesis cannot, for it is a necessary condition for coercion that the act runs contrary to the agent’s conception of the fine. This places Aristotle’s theory of prohairesis in the EE squarely within a eudaemonist framework. What the MM called ‘to dia ti’ and the ‘best’ is here identified with the end that we seek as fine when we act on the on the basis of reason. Before he turns to the topical analysis of prohairesis in EE II.10, Aristotle presents a preliminary definition of virtue of character as a state that accords with correct reason (orthos logos), which lies in the mean relative to us between excess and deficiency with regard to pleasures and pains. This definition is refined in concluding paragraphs of II.10 to include the specification that virtue is a prohairetic state. Thus, the topical analysis of prohairesis in the EE furthers the main line of inquiry in the treatise in a way that the MM analysis does not. In the topical analysis in the EE, we again encounter an eliminative argument. Aristotle first considers proposals made by ‘some people’ (tines): that prohairesis is either desire (orexis) or belief (doxa). Either of these views seems correct to the extent that decision is accompanied by both, but Aristotle rejects their identification. It isn’t desire, since it is neither appetite (epithumia), or spirit (thumos), or wish (boulēsis) in an unqualified sense. But neither is it belief or belief in some qualified sense—viz., belief about things that can be done that are up to us. This is common to belief and wish. In contrast to the MM, which explains the nature of prohairesis without first examining the nature of deliberation, the EE presents an elaborate account of the objects of deliberation, and only offers its nominal definition at the end of this analysis. This indicates that Aristotle thinks we can only understand the meaning of prohairesis once we understand what deliberation is, and that the answer to this question is not immediately evident—it requires extensive philosophical analysis. We should thus expect the theory of deliberation in the EE to be philosophically more sophisticated than the largely unanalysed notion of choosing between alternatives prevalent in the MM. Unlike the MM, the EE construes deliberation as an inquiry (zētēsis) that seeks to bring the cause of an end that we want back to us by identifying an act that we can do. Because deliberation is a kind of causal inquiry, Aristotle says that ‘the deliberative part of the soul (to bouleutikon) is the part that observes some cause (to theōretikon aitias tinos)’ (EE 1226b25–6). Just as theoretical inquiry seeks to articulate the because, practical inquiry seeks to uncover not simply that we should do something, but why we should perform an act, and that justification will always make reference to our end—in the final analysis to our conception of living finely. The aim of deliberation is to trace the principle of an end that is the object of wish back to us—to an act we can do for its sake: in general, in every case when one decides one shows what one decides on and for the sake of what one decides on it; and that for the sake is that for the sake of which one decides on something else, and what one decides on is what one decides on for the sake of something else (kai

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

210  Karen Margrethe Nielsen holōs dēloi aei prohairoumenos ti te kai tinos heneka prohaireitai, esti de to men tinos, hou heneka prohaireitai allo, to de ti, ho prohaireitai heneka allou). (EE II.10. 1226a11–13)

This explanation uses explicitly teleological terminology like ‘to hou heneka’ about the end, and explains the object of decision as what we do for the sake of this end. This prepares the way for the nominal definition of prohairesis in the EE, which departs from that of the MM. The departure is easily missed, since the OCT and modern translations all trust an emendation proposed by Sylburg in the sixteenth century at a critical point, thus making a hash both of Aristotle’s explanation and the surrounding argument.30 With Sylburg’s misguided emendation, the nominal definition at first seems to conform to what we find in the MM, yielding the analysis below: NOMINAL DEFINITION EE (Sylburg’s emendation in bold): Since, therefore, decision is neither belief nor wish individually nor both together (for no one decides on the spur of the moment, but it seems to do it and they wish . Then it must be out of both; for both of these belong to the one who is deciding. But we must inquire how it is out of these. Now the name itself makes it clear in a way. For decision is choice, but not without qualification, but the choice of one thing before another (dēloi de pōs kai to onoma auto. hē gar prohairesis hairesis men estin, ouch haplōs de, all’heterou pro heterou). And this is not possible without inquiry and deliberation. That is why decision is out of deliberative belief. Now about the end no one deliberates, but this is assumed for all, but about the things that tend towards it—whether this or that tends towards it, or, once it is settled, how this will come about. And we all deliberate about this, until we bring the origin of its coming into being back to ourselves.

If we trust the manuscripts, however, the nominal definition in the EE does not contain the preposition pro at all. The MSS all print pros at 1226b8 (in bold above). If this is right, Aristotle’s nominal definition in the EE instead reads: Now the name itself makes it clear in a way. For decision is choice, but not without ­qualification, but the choice of one thing for the sake of another (heterou pros heteron) And this is not possible without inquiry and deliberation. That is why decision is out of ­deliberative belief. (EE II.10 1226b6–9)

30   The Greek text on which Sylburg comments, included in his 1587 edition of the EE, in fact gives the L-family reading, ‘heterou pros heteron’. But in a section titled ‘Variae lectiones’, Sylburg expresses unhappiness with this: ‘P. 108, v.I, rectius forsan heterou pro heterou: ut significetur alterius prae altero electio: sicut scilicet in Nicomacheis, 40, 13, prohaireton dictum tradit hōs on pro heterōn haireton. Aretius tamen vulgatam scripturam secutus vertit, alterius ad alterum’ (p. 271). (‘More correct, possibly, heterou pro heterou: in order that it means alterius prae altero electio (the choice of one thing before another): just as in the Nicomachean Ethics 40, 13, where the expression hōs on pro heterōn haireton renders prohaireton. Aretius, notwithstanding, having followed the received manuscript, translates alterius ad alterum (of one thing for the sake of another)’.) The entry is, as far as I can tell, not cited by any of the scholars who accept the emendation, though their rationale is likely similar to Sylburg’s: to make the nominal definition in the EE correspond to that in the EN. Now, Sylburg was no upstart, having studied with Henri Estienne (Stephanus) in Paris. Nevertheless, he rejects the L manuscript reading, which was respected by Aretius (the renaissance scholar Leonardo Bruni, 1369–1444). I discuss the textual issues in more detail in a separate paper.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  211 In other words, depending on whether we trust the MSS or Sylburg, the nominal ­definition in the EE either contains the preposition pros (MSS), thus explaining the sense of prohairesis as ‘pros-hairesis’, or pro (Sylburg). Walzer and Mingay’s OCT follows Sylburg’s emendation of pros to pro, as do Inwood and Woolf in their recent translation of the EE (2013) and Kenny in his translation (2011). Thinking that Aristotle means ‘choosing this before that’ (pro in the temporal or preferential sense, requiring the genitive), they are then required to supply the PC family MSS reading heterou rather than L’s heteron after pro. The existence of L makes this choice questionable, especially since the manuscripts are unanimous in reading pros rather than pro, making the reading of PC (‘heterou’) impossible. I take Sylburg’s emendation to be unnecessary. ‘Hairesis (. . .) heterou pros heteron’ makes sense both grammatically and philosophically—­meaning, simply, a choice of something for the sake of something else.31 However, even if we remain undecided, I don’t think the choice of pro or pros forces us to choose between either the teleological or preferential interpretation of prohairesis in the EE, since pro + genitive can convey the teleological sense in and of itself. On the EE conception, unlike the MM conception, a prohairesis is a hairesis of something for the sake of (pros) something else (viz., the end), just as Aristotle makes clear when he says that we deliberate about what is for the sake of the end (to pros to telos) rather than the end itself. Few readers show any awareness of the teleological use of ‘pro’, and so in the scholarship on prohairesis in the ethics, scholars typically insist that the sense must be either preferential or temporal. But that does not exhaust the alternatives.32 The following points merit attention. First, pro + genitive can express the ground or motive of an action—that because of which it is performed. This is what I have called the ‘teleological’ sense of pro. Second, in this sense, pros + accusative and pro + genitive can be used interchangeably. While the MM highlights the preferential sense of prohaireis, Aristotle in the EE exploits the teleological sense, the sense in which the object of a prohairesis is an act that we do for the sake of an end. Consequently, the EE does not treat the connection between making a choice and forming a preference for an alternative as fundamental. Instead, it highlights the teleological structure of prohairesis—it is always of something 31   Kenny (2011) translates: ‘Its very name already tells us something. Choice is a matter of picking – not random picking, but picking one thing in preference to another, and this is not possible without examination and deliberation’. Inwood and Woolf follow suit in their 2013 translation for the Cambridge Texts in the History of Philosophy: ‘Decision is choice, not unqualifiedly so, but of one thing in preference to another’. 32   The teleological usage is covered by e.g. G. Harrison’s A Treatise on the Greek Prepositions and on the Cases of Nouns with which These are Used (Philadelphia: J. B. Lippincott & Co, 1858). Consider, for example, this excerpt from the entry on pro: ‘Pro corresponds to the Sanskrit pra, the Latin pro and prae, the German vor, the Slavonic pro and pri, the Gothic fra and fri, the Irish fur, for, foir, and the English for, and fro, as seen in the compound fro-ward (. . .) In some examples pro with the genitive case expresses the ground or the motive of an action. The object introduced by pro is that “in view of ” which or “looking to” which, the action is performed, and so is regarded as its ground or motive. And, again, it is obvious that the notion of “looking to”, or “having in view”, is to be referred immediately to the primary sense of “before”, “in front of ”, belonging to pro’.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

212  Karen Margrethe Nielsen (an act) for the sake of something (an end). This usage is reflected in Aristotle’s use of the simple noun hairesis. When Aristotle speaks about happiness as an end, he speaks about the end as an object of hairesis, not of a prohairesis. This stands to reason: we never choose happiness for the sake of anything else—it isn’t for the sake of—pros— anything. A decision, then, which results from deliberation, is ‘of this for the sake of that’—not of a ‘this’—a particular end or action—unqualifiedly. That would be a bare hairesis.33 Our decision thus reflects both what we decide to do—which course of action we are prepared to undertake—and why we think it is appropriate given the circumstances, and our conception of the fine. Notice, too, that Aristotle in the nominal definition of the EE claims that a prohairesis is not a hairesis haplōs—an unqualified choice. On the preferential reading, we should expect his point to be that we don’t decide on x simpliciter, but rather on x in preference to y, some alternative course of action. But this isn’t how Aristotle explains the relationship between choosing x unqualifiedly and choosing it qualifiedly in common book VII.9. For in common book VII.9 he gives a teleological explanation of the contrast between choosing something unqualifiedly and choosing it qualifiedly, or kata sumbebēkos: If someone chooses (haireitai) or pursues one thing because of a second (todi dia todi), he pursues and chooses the second in itself (kath’auto) and the first coincidentally (kata sumbebēkos). Now when we speak of something without qualification (haplōs), we speak of it because of itself (kath’hauto). (EN VII =EE VI.9 1151a35–b3)

The end, then, is an object of choice in its own right, while the means are chosen because they are for the sake of this end. So, ‘haplōs’ in the etymological explication in EE means kath’hauto, while todi dia todi means x for the sake of y. It does not mean, as we should expect on the preferential reading, x rather than—anti—y. Preference is not what is at issue.34 It is thus unlikely that the author of the MM had an MS of the EE in front of him and attempted to capture its argument, for the theory of prohairesis in the MM does not draw out the same connections as the theory developed and defended in the EE.35 33   To see how systematically Aristotle uses hairesis of the end in the EE, see e.g. the discussion of the choice of lives and ultimate ends in EE 1214a32; 1215b17; 21; 30; 35; 1216a13; 15; 21. The EN likewise opens with a consideration of what end is most choiceworthy (1094a15; also 1097a26 and following). In discussing the completeness of different ends, he systematically uses hairesis without pro. 34   Aristotle uses the prepositions pro and pros interchangeably elsewhere in the EE. Examples include the expression pro ergou, roughly, ‘furthering the inquiry’ (EE 1215a8; EE 1220a22; also, EN 1113b28 (ouden pro ergou means ‘pointless’)), which has the same sense as the expression pros to ergon (with sigma) at EE 1242a16 and EN 1098a31. This reveals that Aristotle is prepared to use pro and pros interchangeably, to refer to what promotes some end. Just like the Latin pro in expressions like pro bono and pro forma, Aristotle’s pro in prohairesis in the EE has a sense that is equivalent to pros. 35   A critic might object that EE II.10 1226b9–18 defends a preferential notion of decision and a weighing conception of deliberation. In the OCT, the line reads ‘no one decides without having prepared or without having deliberated whether it is better or worse’. In fact, ‘whether it is better or worse’ is not supported by the MSS, which have ê instead of ei, the latter an emendation proposed by Bonitz. In this case, Inwood and Woolf (2012) preserve the sense correctly when they translate ‘if then, no one makes a decision without

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  213

Decision and Virtue of Character Aristotle concludes the topical analysis in EE II.10 by offering a refined definition of virtue of character as a hexis prohairetikē. He thereby incorporates the concept into the provisional definition of virtue of character he presented at the outset. This step is never taken in the MM, and so the EE shows signs of far greater compositional control: It is necessary, therefore, since virtue of character is itself a mean and all of it is about both pleasures and pains, and since vice is in excess or deficiency and about the same things as virtue, for virtue of character to be a state that decides (hexin prohairetikēn) on the mean relative to us in pleasant and painful things, in accord with however many one is said be a certain sort of person in his character by his enjoyment or pain—for the lover of sweet or bitter things is not said to be a certain sort of person in his character. (EE II.10 1227b5–12)

In II.11, which can be read as the coda to II.10, Aristotle draws out some important implications of the definition he has just presented. He argues that virtue makes the target (skopos) correct, rather than the reasoning about what promotes the target. This, he says, belongs to another capacity. He does not say what that capacity is, though in CB EN VI.12 =EE V.12. 1144a22–b1, it appears to be identified with the calculating aspect of prudence. This means that virtue of character concerns only one aspect of prohairesis, the aspect having to do with the end. When we decide, we decide on some something-for-the-sake-of-something, and virtue makes the end—that for the sake of which we act—correct.36 Hitting on what we should do for the sake of it belongs to a different faculty. This is why, says Aristotle, we can tell a man’s character from his decision (ek tēs prohaireseōs krinomen poios tis, II.11 1228a2–3), for we can identify both what end he pursues and the actions that he chooses for its sake on the basis of his decision. It is only because ‘it is not easy to see what sort of decision it is’ that we are ‘compelled to judge what sort of person someone is from the products’—in other words, from acts (EE II.11 1228a4–16). The virtuous man’s end is the intermediate, of which virtue is the cause by deciding for the sake of it, whereas the vicious man’s (actual rather than intended) end is excess or deficiency. Whether this means that there can be no rational justification for the adoption of an end is doubtful. For, if Aristotle’s point had been that rational thought about ends was impossible, or without practical consepreparation and deliberation, doing so either in a better or worse way’. If this is right, Aristotle does not here suggest that deliberation essentially consists in weighing alternatives. 36   The wording in CB EN VI.12 =EE V.12. 1144a22–b1 echoes that of EE II.11: ‘Now virtue makes the decision correct, but the actions that are naturally to be done to fulfill the decision are the concern not of virtue, but of a different capacity’ (1144a22–4). The claim that ‘virtue makes the decision correct’ is a less precise way of stating the point of II.11, viz., that virtue makes the end of the decision correct—Aristotle has just stated that ‘virtue makes the end correct’ at 1144a9 (see also the claim in EN VI/EE V.13 1145a3–6). These remarks thus have a natural antecedent in the analysis of EE II.11, but they have no precise antecedents in the preceding analysis of the EN. However, the paragraph between the two claims in II.12 appears to refer back to EN II.4, which has no parallel in the EE. This indicates that even if CB VI/V originated in the EE, Aristotle revised it to fit with the analysis of the EN.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

214  Karen Margrethe Nielsen quence, he would presumably not have written EE or EN book I. This interpretation also ignores Aristotle’s several remarks in EE V/EN VI about the role of understanding (nous) in grasping the right end, and furnishing the agent with the right supposition about the good he should aim at. Aristotle’s point, rather, seems to be that virtue makes us inclined to pursue the right ends; it gives us a desire to promote the mean in our action rather than a purely intellection comprehension of what the mean is. However, his claims should be interpreted, there is little doubt that the points Aristotle makes in  this passage presuppose a teleological rather than preferential interpretation of ­prohairesis. That Aristotle can exploit his topical analysis in the EE to make these observations is testament to the distance travelled from the MM to the EE. I have argued that the different treatments of prohairesis in the MM and EE should make us question whether EE II.10 could possibly be the template from which MM I.17 emerged. If the author of the MM had the EE in front of him, he was a very sloppy editor. If he was a student taking notes from a lecture course, he either failed to fasten on what Aristotle was saying, or he heard a course differing radically from the EE and EN. Maybe the student heard a public lecture? If so, our MM may be based on a lecture course that diverges from the EE and EN, and would be of Aristotelian origin, albeit filtered through the conjectured note-taker, as Cooper suggests. The odd disconnects in the progression of the argument may then be the result of the ineptitude of the note taker or bathroom breaks. The more likely explanation is, I think, that the MM does not post-date the EE and the EN, but is an early version of the views that we find in mature form in the EE. It’s not that a moderately clever undergraduate distilled the theory from a far more complex template; rather, the ‘undergraduate’ in question is Aristotle himself, working his way to the articulation of a theory that only came into its own in his later treatises.37, 38

Bibliography Allan, D. J., ‘Magna Moralia and Nicomachean Ethics’, Journal of Hellenic Studies 77 (1957), pp. 7–11. Allen, D., ‘Talking about Revolution: A Political Change in Fourth-Century Athens and Historiographic Method’, in S. Goldhill and R. Osborne (eds.), Rethinking Revolution through Ancient Greece (Cambridge: Cambridge University Press, 2006). 37   In painting, early sketches are sometimes mistaken for later copies. Consider the four versions of ‘The Scream’ painted by Edvard Munch from 1893 to 1910. If one didn’t know better, one could think of the1893 crayon on cardboard version as a bad imitation made by a student. The colours are more washed out, and the clouds less menacing. It looks like something a talented art student could produce after a visit to the galleries. It is in fact a very early version—some claim the very first version Munch drew of the motive. 38   This essay first took shape in a graduate seminar I co-taught with Terry Irwin in Oxford in Hilary term of 2014. I am indebted to all the participants for their valuable contributions, and especially to Terry for getting me interested in the Magna Moralia and for sharing drafts of work in progress on the three ­ethics. A first draft of the paper was first presented at the Ancient Philosophy Work in Progress Workshop in Oxford on 20 Jan. 2014. I am grateful to all the participants for their excellent questions and comments. Susan Sauvé Meyer provided a characteristically insightful set of comments on a later draft.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Deliberation and Decision  215 Brink, K. O., Stil und Form der pseudoaristotelischen Magna Moralia (Ohlau: Eschenhagen, 1933). Bywater, I., Aristotelis Ethica Nicomachea (OCT) (Oxford: Clarendon Press, 1894). Chroust, A.-H. ‘The First Thirty Years of Modern Aristotelian Scholarship (1912–1942)’ (1963), reprinted in W. R. Wians (ed.), Aristotle’s Philosophical Development: Problems and Prospects (Lanham, MD.: Rowman and Littlefield, 1995), pp. 41–66. Cooper, J. M.,‘The Magna Moralia and Aristotle’s Moral Theory’, American Journal of Philology (1973), reprinted in his Reason and Emotion (Princeton: Princeton University Press, 1999), pp. 195–211. Dirlmeier, F., ‘Zeit der Grossen Ethik’, Rheinisches Museum, 88 (1939), pp. 214–43. Dirlmeier, F., Aristoteles Magna Moralia. Übersetzt und kommentiert (Berlin: Akademie Verlag, 1964). Donini, P. L. L’etica dei Magna moralia (Torino: Fondazione Parini Chirio, Giappichelli 1965). Düring, I., Aristoteles. Darstellung und Interpretation seiner Denkens (Heidelberg: C. Winters Univeristätsverlag, 1966). Gauthier, R.-A. and Jolif, J. Y., L’Ethique à Nicomaque: Introduction, Traduction et Commentaire, vol. II, 1, 2nd ed. (Louvain: Publications Universitaires 1970). Harrison, G., A Treatise on the Greek Prepositions and on the Cases of Nouns with which These are Used (Philadelphia: J. B. Lippincott & Co, 1858). Inwood, B., & Woolf, R., Aristotle Eudemian Ethics (Cambridge: Cambridge University Press, 2012). Irwin, T. H. Aristotle Nicomachean Ethics, 2nd ed. (Indianapolis: Hackett, 1999). Irwin, T. H., unpublished MS for the seminar on Aristotle’s Three Ethics (Oxford, Hilary Term, 2014), titled ‘doc.1b’. Jaeger, W., Aristoteles. Grundlegung einer Geschichte seiner Entwicklung (Berlin: Weidemannsche Buchhandlung, 1923), pp. 237–70. Jaeger, W., ‘Über Ursprung und Kreislauf des philosophischen Lebensideals’, Sitzungsberichte der Preussischen Akademie der Wissenschaften, Philosophisch-Historische Klasse, 25 (1928). Kenny, A., The Aristotelian Ethics (Oxford: Clarendon Press, 1978). Kenny, A., Aristotle. The Eudemian Ethics (Oxford: Oxford University Press, 2011). Nielsen, K. M., ‘Deliberation as Inquiry: Aristotle’s Alternative to the Presumption of Open Alternatives’, The Philosophical Review, 120, no 3 (2011), pp. 383–421. Schleiermacher, F., Samtliche Werke, Abteilung III: ‘Über die ethischen Werke des Aristoteles’ (Berlin: G. Reimer, 1835). von Arnim, H., ‘Die drei aristotelischen Ethiken’, Sitzungsberichte der Preussischen Akademie der Wissenschaften, philosophisch-historische Klasse 202.2 (Wien/Leipzig: Hölder-PichlerTempsky, 1924). von Arnim, H., ‘Der neuste Versuch, die Magna Moralia als unecht zu erweisen’, Akademie der Wissenschaften in Wien, philosophisch-historische Klasse, Sitzungsberichte 211.2 (Wien/ Leipzig: Hölder-Pichler-Tempsky, 1929). von Arnim, H., ‘Nochmals die aristotelichen Ethiken (Gegen W Jaeger. Zur Abwehr.)’, Akademie der Wissenschaften in Wien, philosophisch-historische Klasse, Sitzungsberichte 211.2 (Wien/Leipzig: Hölder-Pichler-Tempsky, 1929). Walzer, R., Magna Moralia und Aristotelische Ethik (Berlin: Weidemann, 1929). Walzer, R. R. & Mingay, J. M., Aristotelis Ethica Eudemia (OCT) (Oxford: Clarendon Press, 1991).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

12 The Freedom Required for Moral Responsibility John Martin Fischer

T.H. Irwin was extraordinarily influential in my development as a philosopher. He was a mentor par excellence at Cornell, where I was fortunate enough to work with him extensively in graduate school in philosophy. Here I shall select one area in which his work has had a big impact on me: moral responsibility theory. I learned a tremendous amount from Irwin’s classic essay (1980). Irwin interprets Aristotle as offering a (broadly speaking) reasons-responsiveness approach to moral responsibility. I shall begin by laying out (in a very brief form) some of the main ideas in Irwin’s ‘reconstruction’ of Aristotle’s views on moral responsibility. I then point out ways in which my own approach shares central features of this Aristotelian account, while also departing in some ways. Finally, I seek to defend my theory of moral responsibility in light of some salient and important recent challenges.

Irwin’s Aristotle on Moral Responsibility Irwin’s ‘Reason and Responsibility’ is a brilliant and important contribution, not just to Aristotle scholarship, but to moral responsibility theory more broadly. Irwin is careful to point out that the account of moral responsibility he is attributing to Aristotle is not explicitly in any of Aristotle’s writings. Rather, he is taking key passages from Aristotle’s texts and putting them together in ways that capture what he takes to be central ideas in Aristotle, as well as ideas to which Aristotle, upon reflection, would see that he is committed. Irwin writes: Despite the apparent simplicity and negative procedure of Aristotle’s account, I believe he needs and assumes a positive account that raises more complex questions about the general conditions for responsibility. When we consider difficulties in the account as it stands and try to solve these difficulties by appeal to other Aristotelian doctrines, the resulting theory will be less simple, but more interesting, and more capable of answering some hard questions about responsible action. I will set out this ‘Aristotelian’ theory—I will call it Aristotle’s ‘complex theory’—partly for an exegetical purpose, to show what Aristotle is committed to, whether he

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   217 realizes it or not. But I have a philosophical purpose too, to show that the account is worth consideration in its own right.  (Irwin 1980: 117–18)

In evaluating various Aristotelian ideas about moral responsibility (and constructing an Aristotelian account of responsibility), Irwin is guided by the following extremely important insight: We normally suppose that a responsible person can do something about the desire he acts on—that it is not compulsively strong but is responsive to his reasoning and deliberation. We assume that rational agents are capable of doing more than simply acting on the desire they find to be strongest; they can also affect the strength of their desires by further deliberation. (Irwin 1980: 131)

I fully agree with Irwin’s point here, and I have made the notion of ‘reasons-responsiveness’ central to my own ideas about moral responsibility. It has always seemed plausible to me that a compulsive desire is indeed one that is not responsive to reason; given the strength of the desire, one would still act on it, no matter what the reasons to do otherwise were. Reasons-responsiveness represents a significant way in which we are not just passive, but active; we are not just ‘passive subjects or spectators of [our] desires’, but, rather, we are agents. We don’t just act on our strongest desires, where the strength is fixed independently of deliberation and reflection. As Irwin puts it, ‘Here Aristotle is right to suggest that decision and rational desire, as he understands them, are important for responsibility’ (Irwin 1980: 131). After considering various simpler accounts of moral responsibility suggested by Aristotle, Irwin ‘restates’ Aristotle’s view as follows: (18) A is responsible for doing x if and only if (a) there is some deliberative argument which, if it were presented to A, would be effective about A’s doing x; (b) A does x voluntarily. (Irwin 1980: 138; the number ‘18’ represents the numbered proposition in Irwin: 1980; I shall employ his numbered propositions throughout.)

He then elaborates upon (b) to produce this ‘summary’ of Aristotle’s conception of moral responsibility: (19) A is responsible for doing x at a time t if and only if (a) there is some deliberative argument which, if it were presented to A before t, would be effective at t about A’s doing x; (b) A believes at t that x is F; (c) x is F; (d) A’s beliefs and desires, as reasons, cause A to do x.  (Irwin 1980: 139)

This conception captures a reasons-responsiveness approach to moral responsibility. Clause (d) expresses the idea that the relevant action must be caused by reasons; and clause (a) conveys (perhaps among other things) the idea that the agent was responsive to considerations involving reasons in such a way that he could have done otherwise. This interpretation of (a) seems to be what Irwin has in mind. Consider, for instance, the remarks he offers in support of this conception: Suppose A did x yesterday, and today we are considering his action. We might believe from what we know about A’s past that a deliberative argument applied yesterday before he did

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

218  John Martin Fischer x would have prevented him from doing x and that therefore he was responsible for doing x when he did it.  (Irwin: 138)

It is important that, on Irwin’s Aristotelian approach, moral responsibility requires more than causation (of the right sort) by reasons; it involves responsiveness to reasons and freedom to do otherwise.1

Guidance Control and the Appeal to Mechanisms I shall begin with a challenge to Irwin’s Aristotelian conception of moral responsibility, as captured in (19). I think that there could be a person who is, intuitively, morally responsible for performing an action at a time, but whose deliberative capacities would be impaired or bypassed, in any scenario in which he were presented (just prior to the time of the action) with a deliberative argument against doing it. That is, as things actually are, the individual’s deliberative capacities are intact, and he performs the action as a result of his reasons (in the right sort of way). But in all counterfactual scenarios in which he is presented a compelling deliberative argument against performing the action, he would still perform it, because of a disruption in his deliberative capacities— perhaps a nefarious neurosurgeon would be directly manipulating his brain only in the counterfactual scenarios. An adequate account of moral responsibility should imply that the individual is morally responsible for the action; but Irwin’s Aristotelian proposal does not have this result. I believe that this sort of problem helps to motivate a relativization to the ‘actualsequence mechanism’ in a reasons-responsiveness account of moral responsibility. This is indeed how I seek to accommodate the famous ‘Frankfurt-style cases’, which are putative counterexamples to the Principle of Alternative Possibilities, according to which moral responsibility for an action requires freedom to do otherwise (Frankfurt 1969). Here is a version of the cases, all of which contain a signature structure involving pre-emptive overdetermination: Black is a stalwart defender of the Democratic party. He has secretly inserted a chip in Jones’s brain, which enables Black to monitor and control Jones’s activities. Black can exercise this control through a sophisticated computer that he has programmed so that, among other things, it monitors Jones’s voting behavior. If Jones were to show any inclination to vote for anyone other than the Democrat, then the computer, through the chip in Jones’s brain, would intervene to assure that he actually decides to vote for the Democrat and does so vote. But if Jones decides on his own to vote for the Democrat (as Black, the old progressive would prefer), the computer does nothing but continue to monitor—without affecting—the goings-on in Jones’s head.  (Fischer 1982)

1   Irwin’s pathbreaking interpretation of Aristotle’s account of moral responsibility continues to be influential (and controversial) today. It is the subject of sustained critical discussion in (Echenique 2012), and, more recently, in (Cooper 2013).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   219 Now suppose that Jones decides to vote for the Democrat on his own, just as he would have if Black had not inserted the chip in his head. It seems, upon first thinking about this case, that Jones can be held morally responsible for his choice and act of voting for the Democrat, although he could not have chosen otherwise, and he could not have done otherwise. In this sort of case, the ordinary deliberative mechanism actually operates, and, intuitively, the individual is morally responsible for his action. But in the alternative scenario, there would be direct intervention in the brain producing the same action as in the actual sequence. We can of course generalize the case so that there would be such an intervention in all counterfactual scenarios. In any case, my preferred account of moral responsibility relativizes the question of reasonsresponsiveness to the actual-sequence mechanism. In the Frankfurt-style case above, the actual-sequence mechanism (ordinary practical reasoning, which does not include Black and his device) is responsive to reasons, even though the alternative scenarios all involve non-responsive mechanisms. Given that the actual-sequence mechanism is suitably responsive to reasons, the agent is morally responsible (on my approach), even though I claim that the agent himself is not responsive to reasons and he cannot do otherwise. On my view (following Aristotle), one is morally responsible for one’s behavior insofar as one meets both an epistemic and a ‘freedom-relevant’ condition. On my view, we can learn from the Frankfurt-style cases that the freedom-relevant condition involves ‘acting freely’, rather than freedom to do otherwise; and Jones acts freely in the actual sequence. That is, the freedom-relevant condition involves the exhibiting of a distinctive kind of control: ‘guidance control’, which does not require control over one’s behavior, or ‘regulative control’ (Fischer and Ravizza 1998). What follows is a very brief account of guidance control (Fischer and Ravizza 1998). An individual exhibits guidance control to the extent that he acts from his own, suitably reasons-responsive mechanism. There are thus two important components of the analysis: ownership of the actual-sequence mechanism, and reasons-responsiveness of that mechanism. Clearly, the analysis presupposes that we can distinguish (in a reasonable way) the actual-sequence mechanism from various alternative-scenario mechanisms. As I have previously explained, this is the key to accommodating the Frankfurt cases. That is, in these cases, the actual-sequence mechanism may be the agent’s own, suitably reasons-responsive mechanism, even though the alternative-scenarios involve mechanisms that are not the agent’s own or not appropriately reasons-responsive (or both). So, for example, in the Frankfurt case sketched, Jones’s actual-sequence mechanism is ‘ordinary practical reasoning’ or ‘the human deliberative mechanism’, whereas the alternative scenario features a manipulation mechanism. On my view, an agent’s mechanism becomes his own when he takes responsibility for it; he thereby makes the mechanism his own. Note that a mechanism could be inculcated in an agent in such a way as to be reasons-responsive; in this case it might not be the agent’s own mechanism. Thus, we need to supplement reasons-responsiveness with ownership to get an adequate account of the kind of control implicated in moral

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

220  John Martin Fischer responsibility. Here are the three conditions for taking responsibility and thereby making one’s mechanism one’s own: a) The individual must see himself as an agent; he must see that his choices and actions are efficacious in the world. This condition includes the claim that he sees that if he were to choose and act differently, different upshots would occur in the world. b) The individual must accept that he is a fair target of the reactive attitudes as a result of how he exercises this agency in certain contexts. c) The individual’s view of himself specified in the first two conditions must be based, in an appropriate way, on the evidence. (Fischer and Ravizza 1998: 210–14) In order for an agent to meet the freedom-relevant condition on moral responsibility for his behavior, it must be the result of his own, suitably reasons-responsive mechanism. The crucial idea here is that we must hold fixed the actual-sequence mechanism (say, unimpaired human practical reasoning), and we then ask whether they agent would (given the operation of that same mechanism) recognize reasons to do otherwise and respond to those reasons. This is a very brief and oversimplified account, but I do not believe that the details matter for the purposes of this essay. (For a full exposition, see Fischer and Ravizza: 1998: 28–91.) The key departures of my approach from Irwin’s Aristotelian account are the mechanism-relativization of reasons-responsiveness and the explicit rejection of the requirement of alternative possibilities (freedom to do otherwise). In the following sections, I will defend these features of my reasons-responsiveness theory against important criticisms. But before I turn to these replies, I shall seek to clarify my contention that I have offered an ‘actual-sequence’ account of moral responsibility, that is, an account that does not require alternative possibilities (i.e., freedom to do otherwise).

An Actual-Sequence Approach to Moral Responsibility My account of moral responsibility is an ‘actual-sequence’ model of moral responsibility. That is, it fixes on (possibly modal or dispositional) properties of the actual sequence, and features of other possible worlds are relevant only insofar as they bear on these properties of the actual sequence, not insofar as they indicate that the agent has freedom to do otherwise. Recently, Christopher Franklin has contended that everyone (and that includes me) holds that an ability to do otherwise is necessary for moral responsibility (Franklin 2015). A brief examination of this point will help to clarify my actual-sequence approach. Franklin writes: The issue is not whether the ability to do otherwise is necessary for freedom and responsibility: the issue is which ability to do otherwise is necessary. Everyone thinks that some ability to do otherwise is necessary for freedom and responsibility. . . .  (Franklin 2015: 2092)

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   221 Later in Franklin’s interesting and thoughtful piece, he writes: Fischer’s repeated denial that alternative possibilities are relevant [to moral responsibility] is apt to mislead (and has misled). What he really intends (or should intend) by these statements is that certain species of abilities are irrelevant. . . . He employs FSCs [Frankfurt-style cases] to show that the ability to do otherwise . . . is not necessary for moral responsibility. Fischer’s phrasing of the point in terms of ‘alternative possibilities’ has the unfortunate suggestion that he does not think that an ability to do otherwise is required in any sense. . . . That such confusions have arisen in the literature becomes clear when we consider recent compatibilists who argue that Fischer is mistaken in concluding that the ability to do otherwise is irrelevant to moral responsibility (or at least blameworthiness). This is most clearly seen in the case of the new dispositionalists [such as Michael Smith (2003), Kadri Vihvelin (2004) and Michael Fara (2008)].  (Franklin 2015: 2097)

I completely agree with Franklin that I do indeed believe that various kinds of ­alternative possibilities are required for moral responsibility (although not for the ‘grounding’ or explanation of moral responsibility), and thus that my repeated contention that alternative possibilities are not required for moral responsibility might well have caused confusion. Of course, in the Frankfurt-style cases Jones does not lose his general ability to vote for a Republican (i.e., to do otherwise in the context), just as we do not lose various general abilities when we are asleep. One might say that we keep our general abilities in these contexts, although we do not have the opportunity to exercise them (immediately). Again: I do not lose my general ability to play the piano when there is no piano in my vicinity; I simply do not have the opportunity to exercise it. So Franklin is certainly right: we all believe that some kinds of alternative possibilities are present in any case in which an individual is morally responsible for his behavior (even if the presence of such alternative possibilities does not ground or explain the agent’s moral responsibility). But, as Franklin also notes, these were not the sorts of alternative possibilities I had in mind in contending that moral responsibility does not require alternative possibilities. I have absolutely no interest in showing that moral responsibility does not require general capacities or abilities to do otherwise, or various other kinds of abilities to do otherwise that abstract away from the particulars of the agent’s history and/or present situation. After all, there is no reason to suppose that these sorts of abilities to do otherwise are incompatible with causal determinism! I have always been interested in the sort of alternative possibility that would be (or could plausibly be thought to be) ruled out by causal determinism. And, clearly, general abilities and indeed any sort of ability to do otherwise that abstracts away from features of the agent’s past and/or current situation need not be inconsistent with causal determinism. So perhaps I should have been clearer all along. As Franklin points out, I take the Consequence Argument very seriously; this argument purports to show that causal determinism would rule out alternative possibilities (Van Inwagen 1983; Fischer 1994). But here it is obvious that the alternative possibilities in question are conceptualized as

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

222  John Martin Fischer requiring the past and current situation of the agent to be held fixed. It is this sort of alternative possibility that the incompatibilist believes is both ruled out by causal determinism and required for moral responsibility. I have sought to engage with the incompatibilist by stipulating, for the sake of the argument, that this is the sort of alternative possibility that is envisaged by Principle of Alternative Possibilities: that is, an alternative possibility that is not abstract in the way a general capacity is abstract or any similar way, i.e., by subtracting features of the past or current situation of the agent. Given this stipulation, I have argued that Frankfurt-style cases show that precisely this sort of alternative possibility is not required for moral responsibility. Given my dialectical project, it is completely unobjectionable that abstract alternative possibilities are present whenever there is moral responsibility; after all, there is no reason to suppose that such alternative possibilities are ruled out by causal determinism. I shall return to this point below (when I discuss the ‘new dispositionalists’). But first I shall consider objections to my mechanism-based version of a reasons-responsiveness approach to moral responsibility.

A Defense of the Mechanism-Based Approach to Moral Responsibility In a way, my reasons-responsiveness approach to moral responsibility is the ‘offspring’ of Irwin’s Aristotle and Robert Nozick, because I was also deeply influenced (in developing my account) by Robert Nozick’s accounts of knowledge and ‘tracking goodness’ or ‘aligning with value’ (Nozick 1981; Fischer 1994 and Fischer and Ravizza 1998). Nozick’s theory of knowledge is an ‘actual-sequence’ theory of knowledge, just as mine is an actual-sequence theory of moral responsibility; in both cases we focus on the (dispositional) properties of the actual-sequence mechanisms (or, ‘methods’, in Nozick’s terminology). As Nozick points out, it is possible for an individual to know that p, where he believes that p via a certain method. And yet it might also be the case that if p were false, the individual would use another method instead—a method that would lead him to believe that p, even though it is false. Nozick writes, ‘A grandmother sees her grandson is well when he comes to visit; but if he were sick or dead, others would tell her he was well to spare her upset’ (Nozick 1981: 179). In this sort of case, Nozick holds that the individual does indeed know that p in virtue of the relevant properties of the actual method of belief-production: this method ‘tracks truth’, even though the method employed in the alternative scenario does not (Nozick 1981: 179). Nozick is not the only philosopher who employs a ‘method-based’ (or ‘mechanismbased’) theory of knowledge. Because of the possibility of cases of the sort just mentioned, and for various other reasons, many epistemologists have adopted ‘mechanism-based’ theories of knowledge (and/or justified belief). Just two prominent examples are Alvin Goldman and Fred Dretske, both of whom defend ‘reliabilist’ accounts of knowledge.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   223 (The classic early presentations of theirs views are in: Goldman 1967 and 1976; and Dretske 1970 and 1971.) On these approaches (which differ in their details), an individual’s knowledge that p requires that the mechanism that produces the individual’s belief that p be reliable. Many epistemologists have concluded that the best way to analyse knowledge is via mechanism-based accounts. Just as Nozick argued that one could—and should—give an account of a certain kind of exemplary agency (‘tracking value or bestness’) via a method-based approach, which is structurally similar to his method-based account of knowledge, I have argued that the best way to give an account of moral responsibility is via a mechanism-based model (Nozick  1981: 317–26; Fischer  1994). I agree with Nozick that the analogy between the epistemological context and the context of agency is deeply illuminating, and I have sought further to develop the analogy. Various philosophers have criticized my employment of a mechanism-based approach to moral responsibility (McKenna 2001 and 2013; Ginet 2006; Watson 2001; and Brink and Nelkin 2013). Perhaps the fullest development of this objection is in insightful work by Michael McKenna, and it will be helpful to lay out salient features of McKenna’s nuanced critique here. McKenna is skeptical about the Fischer/Ravizza mechanism-based approach, because (in his view) we offer too little explicit guidance on mechanism-individuation. (And, of course, relativization of reasons-responsiveness to the actual-sequence mechanism is a crucial ingredient in my account, so it is important to be clear on what counts as the same mechanism.) He concedes that it is relatively clear that there is a different kind of mechanism operating in the actual sequence and alternative scenario in a Frankfurt-style case (McKenna 2013: 161). But he writes: The problem, however, is that there are some contexts crucial to the development of their theory where, absent some principled bases for mechanism individuation, it is hard to assess their broader claims. . . . it is an indispensable part of their theory that we are to test how the same mechanism behaves in response to different reasons. How are we to settle whether, when different reasons are put to an agent, the same mechanism is operative? We need some purchase on what it is that we are holding fixed when we hold fixed ‘the same mechanism’ while testing its degree of responsiveness.  (McKenna 2013: 161)

McKenna goes on to express the worry that because I (and my co-author, Mark Ravizza) have offered no ‘principled’ basis for mechanism-individuation, ‘we are left to settle the matter exclusively by appeal to our intuitive reactions to varying cases’ (McKenna 2013: 162). And, of course, such intuitions will vary widely, and may be influenced by one’s theoretical commitments about the relationship between causal determinism and moral responsibility (McKenna 2013: 162–3). I have replied to these sorts of worries (especially as articulated in McKenna 2001) by emphasizing that my reliance on intuitive ideas about mechanism-individuation is  no different from similar reliance on non-reductive views about mechanismindividuation (or similar ideas) in other areas of philosophy (Fischer 2004). Will one

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

224  John Martin Fischer have to give up a Nozickean method-based epistemology, because one offers no fully reductive account of method-individuation? Similarly, will one have to give up a Dretske/Goldman sort of reliabilism, because one offers no fully reductive account of mechanism-individuation? Perhaps in the end this lacuna will be deemed a significant cost of such theories; but should they be ruled out from the start? I doubt it; these are  very important and widely influential epistemological theories—approaches that deserve careful attention, even though they are not fully reductive (as regards mechanism-individuation). And a similar point applies to generalization strategies in ethics—consequentialist and Kantian. Will one give up rule-consequentialism or the universalizability version of Kant’s Categorical Imperative because one has no fully reductive account of rule-individuation or maxim-individuation? To this reply to his objection, McKenna has offered this thoughtful and challenging response: Fischer has responded to my objections by likening the problem I raise here to generality problems in other areas of philosophy. A certain level of generality, and thus lack of specification, is acceptable in other domains of inquiry, such as ethical theory. So, too, Fischer argues, is a degree of generality acceptable when theorizing in terms of mechanisms of agency. (Fischer 2004: 169) Fischer also points out that it is unreasonable to demand of a successful theory that all of its elements are fully analysed (168). Thus, he resists the burden of offering any ‘purely “principled” account of mechanism individuation—an account that did not at some level appeal to intuition’ (167–8). He thus remains committed to relying exclusively on appeal to intuitions in response to difference [sic] cases. While I sympathize with Fischer’s thoughtful reply, it is not enough to put the objection to rest. Requesting some principled basis of mechanism individuation is not the same as demanding a full analysis of mechanisms—one that is ‘purely’ principled. And a degree of generality is of course acceptable in theorizing in areas such as this one. But when the degree of generality and an exclusive reliance on intuition will not help adjudicate differences between cases crucial to assessing the theory, then demanding some principled basis for settling the dispute is only dialectically fair and reasonable.  (McKenna 2013: 180)

McKenna’s critique here, and his worries about the role of mechanism-individuation in my theory, are eminently fair and reasonable. Still, I think more can be said in defense of the sort of mechanism-based theory I have presented. McKenna is willing to concede that there can be a certain residual ‘generality’ in a theory of responsibility, and he is also willing to accept a role for intuition. But he seems to think of these components as entirely distinct. That is, he seems to think that I am somehow forced to an ‘exclusive reliance’ on intuitions in applying my theory (which involves a ‘general’ component) to cases. This would certainly be problematic; here the theory would not be doing any significant conceptual work, but all the important work would be done by intuitions about cases. The theory would just be a façade lending some (illicit) credibility to the intuitive judgments. But I am not committed (as far as I can see) to thinking of the generality and intuition components as separate. Rather, I think of the level of generality of the mechanism as

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   225 in part determined by intuitions, although not ‘first-order’ intuitions about certain cases (that is, intuitions about whether some individual is—or is not—morally responsible in a particular case); thus, the reliance on intuition is not separate and entirely distinct from the issue of the level of generality or abstraction of the mechanism. On my view, we start with some clear intuitions about cases in which individuals are morally responsible, and other clear intuitions about cases in which individuals are not morally responsible. Based on these cases, we ‘try out’ a certain intuitively plausible answer to the question of mechanism-individuation. We then apply this provisional notion of mechanism-individuation (a provisional answer to the question of how specific the specification of the mechanism must be) to other cases, adjusting it if it yields intuitively implausible results in these cases. This process of adjustment results in a notion of mechanism-individuation that is not reductively defined (insofar as there is a residuum of guidance by intuition). The key is that it has to be applied consistently across the various cases, in some intuitive sense. That is, intuition also plays a role here in testing the consistency of the level of generality or abstraction of the relevant mechanism, as the theory gets applied across a range of cases. Similarly, how does one individuate ‘maxims’ in relation to Kant’s universalizability formulation of the categorical imperative? (A parallel question: how does one individuate moral rules, in relation to rule-consequentialism?) This is a notoriously vexing problem—but it has not diminished the interest or importance of Kantian approaches to ethics! Is it permissible to tell a lie in a certain situation? Well, what is the maxim of the action? Telling a lie? Telling a lie on Monday? Telling a lie on Monday while wearing a blue shirt? Telling a lie when no one else will? Telling a lie when no one will know about it? The level of generality or abstraction of the maxim is not somehow ‘given’ or written into the fabric of the moral universe. Rather, the proponent of Kantian universalizability might try out a level of description and see whether it yields the result it is ‘supposed to’; since Kant holds that it is never permissible to tell a lie, this should be the result of the universalizability test. Now the key is to ‘hold fixed’ roughly this view of maxim-individuation when applying the test to other kinds of actions. Although the level of specificity of maxims, for the purposes of the universalizability test, is not somehow fixed in advance, it is nevertheless important to be consistent in the application of the test; otherwise, it would not be the test, but one’s antecedent ‘first-order’ intuitions and views about what actions should pass the test, that are doing all of the work. Consider the similar point made by Nozick: Since we have not specified a precise theory of subjunctives or specified precisely how to identify a method and tell when it is held fixed, there is some leeway in our account. It may be this leeway that enables the account to cope with these examples and other cases, by using the constituent notions loosely and intuitively. This is not an objection but a reason to think the notions can be specified more precisely to handle the cases—a condition on their specification is that they handle the cases adequately—provided the discussion of the cases did not exploit the leeway or wobble inconsistently, first leaning in one direction, then in another.  (Nozick 1981: 193)

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

226  John Martin Fischer The last part of the quotation gets at the worry that an ‘exclusive reliance’ on intuition is what is really doing all of the work in my theory. If what drives the approach are in the first instance intuitions about cases, and then one leans toward a narrow specification of mechanisms to accommodate some intuitions, and a wide specification of mechanisms to accommodate others, this would be objectionable. But this is not how I conceptualize my methodology. A crucial part of my approach requires that mechanisms be individuated consistently. At this point in the development of the overall theory I must rely to some extent on intuition in evaluating whether this is indeed the case, i.e., that mechanisms are being individuated consistently in the application to various cases. Perhaps Nozick is correct that if it ‘feels’ like the application of the theory across cases involves a consistent view of mechanism-individuation, then this provides reason to hope that in principle there could be a more precise (and reductive) account of mechanism-individuation. I am, however, a bit less sanguine than Nozick about the possibility of a reductive analysis of mechanism-individuation. In sum, I do not think it is fair to saddle me with the view that I must rely exclusively on intuitions about the moral responsibility (or lack of it) of individuals in various actual and hypothetical cases, then gerrymandering the level of generality of the mechanisms so as to fit the intuitive judgments. Intuition plays a role in evaluating the consistency of the inchoate view of mechanism-individuation involved in the theory. Of course, I am still not out of the woods. It still must turn out that an overall evaluation of the application of my theory to cases must pass the test; that is, it must seem intuitively that the level of specificity of the mechanism is consistent across cases. It would be completely legitimate for a critic to inquire more specifically into this issue; but this is a different problem than the alleged sole reliance on intuitions about cases. As I have previously written, I am less optimistic than Nozick about the possibility of a fully reductive account of mechanism-individuation. In the end, I would not be surprised if we will have to settle for a theory of moral responsibility that leaves an important role (of the sort identified above) for intuition; it will have at least a part (or parts) that are not specified in an algorithm that employs only unproblematic (or uncontested) ingredients. McKenna prefers an agent-based to a mechanism-based approach to moral responsibility. But I shall now contend that his own preferred account does not succeed in avoiding the invocation of a component that is ‘unspecified reductively’ in just the sort of way my notion of ‘same mechanism’ is. So the complaint cannot be that mechanism-individuation is not susceptible to reductive, algorithmic explanation. To take stock, McKenna argues that in a Frankfurt-style case, the agent (Jones) is indeed responsive to reasons, although he cannot do otherwise. Jones cannot do otherwise because of the presence and dispositions of the counterfactual intervener, Black. Nevertheless, Jones is responsive to reasons because in ascertaining an individual’s reasons-responsiveness, one attends only to intrinsic features of the agent. That is, one abstracts away from extrinsic features, such as Black. And if we subtract Black from the equation, then Jones is evidently reasons-responsive. Although McKenna agrees with

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   227 me that the agent in Frankfurt-style cases cannot do otherwise (in the relevant sense of ‘cannot’), he nevertheless maintains that the agent is reasons-responsive. We are both semicompatibilists; that is, we both hold that causal determinism is compatible with moral responsibility, quite apart from whether causal determinism rules out freedom to do otherwise. But whereas I explain moral responsibility by reference to the reasonsresponsiveness of the actual-sequence mechanism, McKenna does so by reference to the reasons-responsiveness of the agent. David Brink and Dana Nelkin agree with McKenna in preferring an agent-based, rather than a mechanism-based, reasonsresponsiveness approach to moral responsibility (Brink and Nelkin 2013). McKenna works out his specific suggestion for an analysis of Jones’s reasonsresponsiveness in light of similar approaches to the analysis of notions, such as water-solubility or fragility, offered by the so-called ‘new dispositionalists’ (McKenna 2013, Vihvelin 2004 and 2013, Fara 2008, and Smith 2003; the name, ‘new dispositionalists’, comes from a penetrating article by Randolph Clarke 2009). The new dispositionalists hold that we can analyse a passive power (such as solubility or fragility) in terms of a complex set of conditionals. According to these theorists, we can thus avoid the problems posed by ‘masks’ and ‘finks’. (A mask would block the manifestation of the power in the ordinarily triggering circumstances, and a fink would eliminate the power in such circumstances.) McKenna starts with this sort of (rough and approximate) conditional analysis of the ‘disposition’ or ‘passive power’ fragility: If this vase were toppled, and if it retained during the relevant duration of time its intrinsic properties P1-Pn, and if it were not interfered with in a way that would impede the causal efficacy of those intrinsic properties, then it would break.  (McKenna 2013: 167)

Modeling his account of reasons-responsiveness on this sort of approach, he offers the following conditional pertinent to Jones’s ‘active’ power of reasons-responsiveness in the Frankfurt-style cases (here McKenna employs the original Frankfurt-style case in which Jones shoots Smith, but nothing hangs on the difference between this version and the one I presented above in which Jones votes for a Democrat): RR1: If Jones were to become aware of R1, and if Jones retained during the relevant duration of time intrinsic agential properties P1-Pn, and if Jones were not interfered with in a way that would impede the causal efficacy of those properties, then Jones would not shoot Smith. (McKenna 2013: 168; ‘RR1’ indicates that the conditional in question is one of a collection of counterfactuals relevant to an agents reasons-responsiveness)

Employing RR1 as part of a general analysis of reasons-responsiveness, McKenna contends that Jones is reasons-responsive in the Frankfurt-style case. The crucial move is to abstract away from Black, who is ‘external’ to Jones; Black’s presence is not an ‘intrinsic’ agential property of Jones. But now the problem for McKenna is that he relies on an ingredient, the distinction between intrinsic and extrinsic (agential) properties, which he does not explicitly define or explain. This is a notoriously difficult

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

228  John Martin Fischer distinction to characterize in an algorithmic, reductive fashion. There is a huge literature discussing this sort of distinction, with no clear upshot about how reductively to explain it. Two important papers in this literature are Langton and Lewis 1998 and Lewis 2000; it is noteworthy that the distinction between intrinsic and extrinsic spatial properties is parallel in important ways to the equally vexed distinction between hard (temporally nonrelational or intrinsic) and soft (temporally relational or extrinsic) facts about times. (See Fischer 1989 and Fischer and Todd 2015.) Of course, I do not object to an account of reasons-responsiveness or moral responsibility that employs an ingredient that is not given an explicit reduction to unproblematic notions. I do not object to invoking a component that is left to some degree up to guidance by intuitions (of the appropriate kind). Here I wish simply to indicate that McKenna’s agent-based approach to reasons-responsiveness is, arguably, in the same boat with my mechanism-based approach insofar as they both employ ingredients that are not explicitly reduced to unproblematic notions. It might be that we have a better purchase on the intrinsic–extrinsic property distinction than we do on mechanism-individuation, but this is not clear, and we would need an argument for this conclusion. When one seeks to be ‘constructive’ in moral responsibility theory—or in pretty much any other area of philosophy—it is hard to avoid employing some elements that are not reducible in algorithmic fashion to unproblematic or uncontentious elements. It is perhaps a cost of doing business. In any case, the presence in one’s theory of such an element cannot in itself be a fatal flaw, even if it represents a kind of incompleteness or failure to achieve a philosophical ideal. We might again seek wisdom from Aristotle, who pointed out that we should not seek more precision in our theorizing than the data allow.

Moral Responsibility and Freedom to do Otherwise Whereas McKenna seeks to employ resources of the ‘new dispositionalists’ (such as Fara, Vihvelin, and Smith) to give an account of reasons-responsiveness, according to which the agent in the Frankfurt-style cases is reasons-responsive, although he could not have done otherwise, the new dispositionalists employ their analyses to give an account of freedom, such that the agent in the Frankfurt cases is indeed free to do otherwise. Dana Nelkin joins the new dispositionalists in giving an account of freedom that has the result that Jones is free to do otherwise in the Frankfurt-style cases (Nelkin 2011:115). For simplicity’s sake, I shall include Nelkin in the group I refer to as the ‘new dispositionalists’; perhaps I should employ a different name (such as ‘New Alternative Possibilities Compatibilists’) for this inclusive group, but it will be simpler to use ‘new dispositionalists’ with the caveat the Nelkin departs from Smith, Fara, and Vihvelin in important ways. Although these philosophers offer different particular accounts of freedom, it will suffice here to note that they all abstract away from Black’s presence; they analyse Jones’s freedom while subtracting away Black. In Nelkin’s terminology, it is an ‘interference-free’ notion of freedom.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   229 So the new dispositionalists offer an account of freedom that is ‘abstract’—it does not require that all features of the past and current situation of the agent be held fixed. As I have previously written, this kind of account of freedom is perfectly consistent with causal determinism, and it has thus not been my focus of attention. I agree that agents have various kinds of abilities to do otherwise in the Frankfurt-style cases, including general abilities. I further agree that they possess the kind of ability to do otherwise identified by the new dispositionalists; this might be thought to be an ability or freedom that is ‘in-between’ a general capacity and the incompatibilists’ preferred notion of freedom, which holds everything about the past and current situation fixed. Just as there are no worries about the compatibility of general abilities with causal determinism, there are no worries about the compatibility of the new dispositionalists’ notion of freedom with causal determinism. If one is primarily concerned to defend the compatibility of causal determinism and moral responsibility, one can embrace the contention that, in order to be morally responsible, an agent must possess the ability to do otherwise, as conceptualized by the new dispositionalists. Not a problem! But now we have two routes to compatibilism about causal determinism and moral responsibility. On one approach, moral responsibility does not require freedom to do  otherwise—this is the actual-sequence model. On the other approach, moral responsibility does require freedom to do otherwise. Of course, the relevant freedom is interpreted differently by proponents of each route to compatibilism. On the first route, freedom is interpreted as requiring that all facts about the past and current situation be held fixed; on the second route, the fixity requirements are not so stringent, and one does not hold fixed the presence and inclinations of Black. Is there some reason to prefer one approach to another? Why should we adopt an actual-sequence model, rather than the new dispositionalists’ strategy? Indeed, the two approaches can seem to be mere ‘notational variants’ of each other. To explain. On my actual-sequence model, I contend that Jones is not free to do otherwise (McKenna agrees). In evaluating Jones’s freedom to do otherwise (in the relevant sense), I hold fixed the presence and inclinations of Black. However, I also contend that Jones’s actual-sequence mechanism is appropriately reasons-responsive, and thus he can be morally responsible. In evaluating the responsiveness characteristics of the actual-sequence mechanism, I abstract away from Black—he is not part of the actual-sequence mechanism. (McKenna abstracts away from Black in evaluating Jones’s [agential] reasons-responsiveness, and thus gets to the conclusion that Jones is morally responsible in the Frankfurt-case, despite lacking freedom to do otherwise.) The new dispositionalists do not hold fixed Black’s presence and dispositions in evaluating Jones’s freedom to do otherwise. So they get to the very same conclusion (that Jones is morally responsible in the Frankfurt cases), and they do so by abstracting away from Black. The only difference between the two approaches is that the actualsequence theorist abstracts away from Black when ascertaining reasons-responsiveness (either of the agent or mechanism), and the new dispositionalists abstract away from Black when evaluating whether Jones has freedom to do otherwise. But this might

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

230  John Martin Fischer seem to be a superficial difference. At a deep level, both approaches deem Jones morally responsible in the Frankfurt cases because Jones would act differently in certain scenarios in which Black is out of the equation. Why prefer (say) an actual-sequence approach to moral responsibility? Aren’t the two approaches only superficially different? This is a difficult question, but I think at least some progress can be made toward answering it. The first step here is to note that it is desirable to have a theory of moral responsibility that fixes on features that ground or explain moral responsibility, not merely features that are necessary conditions for moral responsibility. It might be that freedom, as interpreted by the new dispositionalists, is a necessary condition of moral responsibility. But in my view the presence of such freedom is not part of the best explanation of an agent’s moral responsibility. Rather, the best explanation of moral responsibility—what grounds moral responsibility—will fix on features of the actual sequence. What explains an agent’s moral responsibility is that she makes a certain kind of statement (I explain this point in Fischer 1990), not that she makes a difference (of a certain kind). After all, interpreted naturally and straightforwardly, Jones does not make a difference in the Frankfurt-style case, but he does make a statement (in an important sense) (Fischer 1990). The new dispositionalists accept a kind of alternative-possibilities requirement on moral responsibility, and thus, I believe, it is natural to suppose that they believe that moral responsibility rests on the capacity to make a difference to the world. But if one wants one’s theory to track closely what explains or grounds moral responsibility, the new dispositionalist will have to say that the freedom to do otherwise (interpreted as in new dispositionalism) does in fact explain or ground moral responsibility. But this just seems implausible to me. That is, it is implausible to explain Jones’s moral responsibility for voting for the Democrat by contending that he selected this path rather than a path in which he voted for a Republican, where both of these paths were genuinely available to him. On the new dispositionalists’ approach, this is precisely what would explain Jones’s moral responsibility; Jones is selecting from among alternative ways the world could go, where the ‘could’ is interpreted via abstraction away from Black. On this view, Jones is indeed making a difference—he is selecting from among paths that are available to him. It is in virtue of this selection that he makes a difference and is thus morally responsible. But this picture just seems incorrect to me, even though it is not straightforward to say why. It does not seem to me that Jones is best described as selecting from among paths the world could genuinely take—either the path in which he votes for the Democrat or the path in which he votes for the Republican. This kind of selection cannot explain why Jones is morally responsible—he doesn’t really select from among such alternatives. He does not know about Black, so he believes that he can vote for either the Republican or the Democrat. So, he makes a selection from among ‘epistemic possibilities’—possibilities which are not ruled out, for all Jones knows. But it is infelicitous to conceive of Jones as selecting from metaphysically open options, thus making a difference, and thereby being morally responsible—or so it seems to me.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   231 Perhaps the point could be put as follows. The proponent of the alternative-possibilities model (presented by the new dispositonalists) thinks that Jones makes a difference to how the world goes in the sense that he selects a path in which he votes for the Democrat, rather than a path in which he votes for the Republican, where both paths are open to him. But, unbeknownst to Jones, if he were about to choose to vote for the Republican, he would be blocked from so choosing (and acting on such a choice). This fact—the fact that Jones would be blocked from choosing to vote for a Republican or acting on that choice—would seem to etiolate the claim that Jones really has a power to make a difference to the world, to select a path along which he votes for the Democrat, rather than a path along which he votes for the Republican. And this is so, even though the new dispositionalist holds that Jones has the freedom in a Frankfurt-style case to vote for the Republican. This suggests that the ‘freedom’ identified by the new dispositionalist does not correspond to the notion of ‘power’ in the idea that Jones has a power to make a difference to the world in the sense that matters for the grounding of moral responsibility. Think of it this way. Suppose Jones reflected on his behavior and proclaimed, ‘In this context, I was a difference-maker. Through my choice and action, I made a difference to the way the world unfolded—I made the difference between the world’s taking the Jones-votes-for-the-Democrat path and the world’s taking the Jones-votes-for-theRepublican path.’ It would be natural to be skeptical, and to point out (gently, of course) that Jones is being rather grandiose here. After all, he really didn’t make the indicated difference to the way the world unfolds, since if he were about to choose to vote for the Republican, he would be occluded from doing so (and from acting on such a choice). Perhaps Jones makes a difference—but not this difference, and it is this difference that is putatively relevant to the explanation of his moral responsibility. It is better to think of Jones as making a statement, rather than making a difference. (Although I do not have space to elaborate on this and the following claims here, please see Fischer 1990.) Insofar as he acts from his own, suitably reasons-responsive mechanism, he makes a statement: he writes a sentence in the narrative of his life. In acting freely, agents transform the chronicles of their lives into genuine narratives or stories (strictly speaking). That is, agents who act freely have life-stories that can evoke prototypical emotional reactions in appropriate audiences, and that are evaluated in part via a consideration of narrative relationships (and not just a function that adds up the welfare of the moments of the life). As the author of this narrative, the agent makes a statement, even if he does not make a difference. I have sought to develop these rudimentary ideas a bit more fully elsewhere (Fischer 2009). Here, unfortunately, given space constraints, this will have to suffice. I certainly do not think that I have a knockdown argument for preferring an actual-sequence approach to the new dispositionalists’ approach. Rather, my argument here has been tentative and merely suggestive. I hold that it is desirable to have a theory of moral responsibility that gives not just necessary (and sufficient) conditions, but closely tracks what grounds (or explains) moral responsibility. And I have further suggested

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

232  John Martin Fischer that selection from among metaphysically available options—making a difference—is not a felicitous way of explaining moral responsibility in general (as illustrated by the Frankfurt-style cases). In contrast to the alternative-possibilities model (and the new dispositionalism), the actual-sequence approach to moral responsibility has it that what explains an agent’s moral responsibility is not selection from among available paths, but freely adding to the narrative of one’s life. In acting freely, one writes a sentence in the narrative of one’s life—one makes a statement, even if one doesn’t make a difference.

Conclusion I have sought to defend a reasons-responsiveness model of moral responsibility. I am in this respect deeply indebted to Irwin’s interpretation of Aristotle. I depart from Irwin’s Aristotle in at least two important ways. First, I adopt a mechanism-based account of moral responsibility, that is, an account that relativizes the issue of reasonsresponsiveness to mechanisms (rather than agents). This is why I previously explained that my theory is the offspring of Irwin’s Aristotle and Nozick. In adopting a mechanismbased model, one can embrace the deep insights of certain epistemologists—and the parallel morals about agency. Second, I adopt an ‘actual-sequence’ model of moral responsibility, according to which moral responsibility does not require freedom to do otherwise. I thus should also mention another important part of my ancestry: Chrysippus, Locke, and Frankfurt. Embracing an actual-sequence model of moral responsibility gives one an elegant answer to the question of what grounds moral responsibility: it is not that an individual makes a difference, but that she makes a statement.2

References Brink, D. and D. Nelkin. 2013. ‘Fairness and the Architecture of Responsibility,’ in D. Shoemaker, ed., Oxford Studies in Agency and Responsibility Volume 1: 284–313. Clarke, R. 2009. ‘Dispositions, Abilities to Act, and Free Will: The New Dispositionalism,’ Mind 118 (470): 323–51. Cooper, J. ‘Aristotle on Responsibility,’ Oxford Studies in Ancient Philosophy 45 (2013): 265–312. Dretske, F. 1970. ‘Epistemic Operators,’ Journal of Philosophy 67 (24): 1007–23. Dretske, F. 1971. ‘Conclusive Reason,’ Australasian Journal of Philosophy 49 (1): 1–22. Echenique, J. 2012. Aristotle’s Ethics and Moral Responsibility. Cambridge: Cambridge University Press. Fara, M. 2008. ‘Masked Abilities and Compatibilism,’ Mind 117 (468): 843–65. Fischer, J.M. 1982. ‘Responsibility and Control,’ Journal of Philosophy 79 (1): 24–40. Fischer, J.M. 1990. ‘Responsibility and Self-Expression,’ Journal of Ethics 3 (4): 277–97. 2   I am very grateful to helpful comments on a previous version of this essay by David Beglin, Benjamin Mitchell-Yellin, Dana Nelkin, and Susan Sauvé Meyer.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

the freedom required for moral responsibility   233 Fischer, J.M. 1994. The Metaphysics of Free Will: An Essay on Control. Oxford: Blackwell Publishers. Fischer, J.M. 2004. ‘Responsibility and Manipulation,’ Journal of Ethics 8 (2): 145–77. Fischer, J.M. 2009. Our Stories: Essays on Life, Death, and Free Will. New York: Oxford University Press. Fischer, J.M., ed. 1989. God, Foreknowledge, and Freedom. Stanford: Stanford University Press. Fischer, J.M. and P. Todd, eds. 2015. Freedom, Fatalism, and Foreknowledge. New York: Oxford University Press. Fischer, J.M. and M. Ravizza. 1998. Responsibility and Control: A Theory of Moral Responsibility. New York: Cambridge University Press. Franklin, C. 2015. ‘Everyone Thinks that an Ability to Do Otherwise is Necessary for Free Will and Moral Responsibility,’ Philosophical Studies 172 (8): 2091–107. Frankfurt, H. 1969. ‘Alternate Possibilities and Moral Responsibility,’ Journal of Philosophy 66 (23): 829–39. Frankfurt, H. 1971. ‘Freedom of the Will and the Concept of a Person,’ Journal of Philosophy 68 (1): 5–20. Frankfurt, H. 2002. ‘Reply to John Martin Fischer’, in S. Buss and L. Overton, eds., Contours of Agency: Essays on Themes from Harry Frankfurt. Cambridge, Mass.: MIT Press: 27–31. Ginet, C. 2006. ‘Working with Fischer and Ravizza’s Account of Moral Responsiblity,’ Journal of Ethics 10 (3): 229–53. Goldman, A. 1967. ‘A Causal Theory of Knowledge,’ Journal of Philosophy 64 (12): 357–72. Goldman, A. 1976. ‘Discrimination and Perceptual Knowledge,’ Journal of Philosophy 73 (20): 771–91. Irwin, T. H. 1980. ‘Reason and Responsibility in Aristotle,’ in A. Rorty, ed., Essays on Aristotle’s Ethics (Berkeley: University of California Press): 117–55. Langton, R. and D. Lewis. 1998. ‘Defining “Intrinsic”,’ Philosophy and Phenomenological Research 58 (2): 333–45. Lewis, D. 1983. ‘Extrinsic Properties,’ Philosophical Studies 44 (2): 197–200. McKenna, M. 2001. ‘Review of John Martin Fischer and Mark Ravizza’s Responsibility and Control: A Theory of Moral Responsibility,’ Journal of Philosophy 98 (2): 93–100. McKenna, M. 2013. ‘Reasons-Responsiveness, Agents, and Mechanisms,’ in D. Shoemaker, ed., Oxford Studies in Agency and Responsibility Volume 1: 151–83. Nelkin, D. 2011. Making Sense of Freedom and Responsibility. New York: Oxford University Press. Nozick, R. 1981. Philosophical Explanations. Cambridge, Mass.: Harvard University Press. Smith, M. 2003. ‘Rational Capacities, or How to Distinguish Recklessness, Weakness, and Compulsion,’ in S. Stroud and C. Tappolet, eds., Weakness of Will and Practical Irrationality. Oxford: Clarendon Press, pp. 17–38. Van Inwagen, P. 1983. An Essay on Free Will. Oxford: Clarendon Press. Vihvelin, K. 2004. ‘Free Will Demystified: A Dispositional Account,’ Philosophical Topics 32(1/2): 427–50. Vihevelin, K. 2013. Causes, Laws, and Free Will: Why Determinism Doesn’t Matter. New York: Oxford University Press. Watson, G. 2001. ‘Reason and Responsibility [Critical Notice of John Martin Fischer and Mark Ravizza, Responsibility and Control: A Theory of Moral Responsibility],’ Ethics 111 (2): 374–94.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

13 Virtue Aristotle and Kant Allen W. Wood

The standard way of classifying ethical theories is to divide them into consequentialist, deontological, and virtue theories. These classifications are not an innocent taxonomy, but a set of invidious stereotypes. When they are applied to great historical philo­ sophers, such Procrustean treatments tend to lop off what constitutes their greatness. Aristotle is read as a virtue theorist, Kant as a deontologist. But Kant’s own final and most comprehensive work on ethics was entitled ‘The Doctrine of Virtue’ (MS 6: 373–493). Aristotle’s doctrine of the mean holds that the mean determines which action follows ‘correct reason’, or is the right action. So why isn’t Kant a virtue theorist, or Aristotle a deontologist? In comparing Aristotle and Kant on virtue, I will be aided by some of the excellent work that has been done on both philosophers by Terence Irwin, in the relevant ­chapters of The Development of Ethics.1 The comparison of Aristotle and Kant on virtue is made more difficult, yet also more interesting, because Kant exhibits almost no knowledge of Aristotle’s ethics. Kant did, in his lectures, formulate interesting views on the  history of ethics (a fact too seldom appreciated).2 Kant discusses Cynicism, Epicureanism, and Stoicism as ‘naturalist’ and Platonism and Christianity as ‘super­ naturalist’ theories of the highest good or ideal kind of life. But Aristotle does not ­figure in his account at all. What little Kant says about Aristotelian ethics was apparently based entirely on J. J. Brucker’s cursory account.3 It displays an almost astonishing degree of incomprehension.4 1   Terence Irwin, The Development of Ethics: A Historical and Critical Study. 3 volumes. (Oxford: Oxford University Press, 2009). Cited subsequently as: ‘Irwin’, by Volume: page number, or else by chapter, or by section (§) number. 2   See Allen W. Wood, ‘Kant’s History of Ethics’, in O. Sensen and L. Denis (eds.) Kant’s Lectures on Ethics: A Critical Guide (Cambridge: Cambridge University Press), 2015. 3   Johann Jakob Brucker, Historia Critica Philosophiae a mundi incunabulis ad nostrum aetatum deducta, 5 vols. Leipzig: Weidemann and Reich, 1742–1744. Pars II, Cap. II, Lib. VII. The brief discussion of Aristotle’s ethics is in §XIX, pp. 835–9. 4   Kant’s explicit references to Aristotle’s ethics are concerned exclusively with the doctrine of the mean. He attacks the doctrine both on the ground that it is an analytic principle, and thus uninformative (VE Collins

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  235

Aristotle vs. Kant I begin by presenting the understandings of Aristotle and Kant according to which they might fit the accepted taxonomy. It could be based on some observations about both philosophers that would be hard for anyone to challenge. Aristotle is a naturalist, in the sense that he bases his ethics on a theory of human nature—the proper human function: rational activity in accordance with virtue or excellence. The fulfillment of this function constitutes happiness (eudaimonia), which is our final and complete end. The virtue or excellence which directs the others is practical wisdom or prudence (phronēsis). Prudence consists in excellence in deliberating about the selection of the correct means to the end. Virtue consists in the agreement of our desires and feelings with the results of prudence. Accordingly, virtue requires being pleased and pained by the right things, and letting these feelings influence our actions. We develop virtues through habituation; this involves shaping the things in which we take pleasure or ­displeasure so that our feelings as well as our desires harmonize with prudence and with the virtuous actions that constitute happiness. Kant’s ethics are not based on any end, but on a moral principle or law, a categorical imperative, which commands us to choose certain actions and refrain from others purely for their own sakes, irrespective of our happiness. This principle is not based on empirical human nature, but on pure reason. The moral imperative commands us to act only on maxims that can be willed to be universal laws, and to treat all persons as ends in themselves. Kant holds that actions in obedience to the imperative sometimes require us to renounce our happiness. Actions have moral worth only when done from duty, not from inclination. We can have no duty to feel pleasure or pain. Virtue consists solely in the strength of our character in obeying the moral law. Further, Kant’s criti­ cisms of eudaimonism are well known (KpV 5:20–8, 35–6, 111–12; cf. G 4:432–3, 442–4). I will argue later that these criticisms are not such direct assaults on Aristotle’s position as they might seem to be. Still, there is clearly disagreement between Kant and Aristotle about the place of happiness in life and its role in practical reason. So presented, the contrast between Aristotle and Kant looks stark indeed. For Aristotle, happiness, an end associated with human nature, is the foundation of ethics, 27:277), but also on the ground that it entails that the right action would have to be one combining two equally corrupt principles (e.g. good management combines prodigality with avarice; courage combines cowardice with recklessness) (MS 6:404–5, 433, VE Vigilantius 27:611, 659). Kant then also complains that neither Aristotle nor his followers were ever able to determine where between the two vicious principles the correct principle is to be found (VE Vigilantius 27:654). Kant does not mention Aristotle by name when rejecting the view that virtue is habit (assuetudo), ‘a uniformity of action that has become a necessity through repetition’ (MS 6:407), but he may have had Aristotle in mind. Kant and Aristotle are probably the two moral philosophers who most made friendship into an important moral topic. Kant seems to have been influenced, at least indirectly, by Aristotle’s threefold classification of friendships. See Allen W. Wood, ‘Kant on Friendship’, Oliver Sensen and Stefano Bacin (eds.) Kant’s Ethics in Context (Cambridge: Cambridge University Press, forthcoming). But on the subject of friendship, Kant quotes from Aristotle only the wry saying ascribed to him by Diogenes Laertius, ‘O my friends, there is no friend’ (MS 6:470, VE 27:680). In fact, this was probably not a quotation from Aristotle at all, but only a textual corruption of a passage which really means: ‘He who has [many] friends has no friend’ (cf. Eudemian Ethics 1245b20).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

236  Allen W. Wood and virtue involves the harmony of desires and feelings with the actions leading to happiness. For Kant, the foundation of morality is a moral law independent of our empirical nature and sometimes opposed to our happiness. Virtue is the power to curb our natural desires when they resist moral reason. Proponents of virtue ethics often find Aristotle’s conception of virtue friendly to our emotional nature, and Kant’s ­hostile to it. Those who think morality sometimes conflicts with self-interest think Kant’s moral philosophy gets something right that Aristotle gets wrong. Kant and Aristotle seem irreconcilable.

Aristotle: ‘Rationalist’ or ‘Anti-rationalist’? We can begin to see that there is more to it than this if we ask whether Aristotle’s view of virtue is what Irwin calls ‘rationalist’ or ‘anti-rationalist.’ Irwin patiently develops the ‘anti-rationalist’ interpretation, which has been highly influential, and for which there is much apparent support in Aristotle’s texts (Irwin §§87–93). There are two main components to this interpretation. First, Aristotle conceives of virtue as the state of an agent in which the non-rational or appetitive part of the soul is in harmony with the rational part. The anti-rationalist takes virtue to be mainly a condition of the nonrational part of the soul, namely its harmony with the deliverances of rational deliber­ ation and election (prohairesis). Second, the anti-rationalist understands Aristotle’s doctrine that we deliberate not about ends, but only about means, to entail that the ends of action are not determined by reason, but instead depend on the way the human character has been shaped by education and habituation. The role of reason in action, on the anti-rationalist view, is limited to selecting the means. This way of reading Aristotle leads directly to a position like Hume’s, in which reason plays the role of ser­ vant to the passions and, to put it in Kantian terms, gives only hypothetical imperatives. Virtue for the anti-rationalist Aristotle would consist in having been conditioned and habituated to have the right ends, and in having the non-rational part of one’s soul habituated to be in harmony with the means to them elected by the rational part of the soul. To translate it into Kantian terms, on this interpretation, virtue for Aristotle con­ sists in having one’s inclinations harmonize spontaneously with the choices of reason, and the ends by which those choices are determined are set not by reason itself, but through education and habituation. Irwin argues, however, that there are significant elements in Aristotle’s position that cannot be reconciled with the anti-rationalist interpretation. To begin with, the antirationalist interpretation threatens to conflict with Aristotle’s naturalism—his view that happiness is an activity of the rational part of the soul in accordance with virtue (Irwin, 1:158). Virtue must then consist essentially (though perhaps not exclusively) in the activity of practical reason. ‘Someone who simply has well-trained non-rational desires does not satisfy the conditions of [EN] I 13, which implies that a person’s func­ tion is fulfilled by agreement between the rational and non-rational parts, under the control of the rational part (1102a5–7; b23–8)’ (Irwin, 1:168).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  237 A related problem for the anti-rationalist interpretation is that ‘The virtuous person must elect virtuous action for its own sake (1105a32; 1144a19)’ (Irwin, 1:168). Virtuous activity must be elected by correct reason, and valued for its own sake. This is related to Aristotle’s doctrines about the relation of virtue to pleasure and pain. It will not do to identify Aristotelian virtue simply with the habituation to be pleasured and pained by the right things. For Aristotle, pleasure is ‘supervenient’ on activities (1174b31–3). Pleasure must depend on the value of the activity on which the pleasure supervenes (1176a3–29). This value is appreciated only by r­ eason. The virtuous agent is pleased by virtuous action because reason values this action for its own sake. We do not have virtue of character in the Aristotelian sense, Irwin argues, if ‘(like the non-ruling classes in the Republic) our emotions are guided by someone else’s ­reason and prudence’ (Irwin 1:168)—still less if they are guided by what Kant calls ‘the whisperings of who knows what tutelary nature’ (G 4:425). For Aristotle, we have v­ irtue only if our actions are chosen for their own sakes by our own reason. Habituation leading to virtue cannot be a merely fortunate or benevolent shaping of non-rational desires and feelings of pleasure and displeasure so that they happen to be directed to objects of moral approval, or to objects conducive to human happiness. ‘Intellectual and deliberative processes are needed for the acquisition of . . . virtue, [and] habituation includes the acquisition and exercise of the relevant deliberative capacities’ (Irwin 1:181). The same considerations imply that the virtuous agent cannot be seen as choos­ ing only the means to ends set independently of reason. Happiness for Aristotle consists essentially of actions. In Aristotle’s claim that we deliberate about means and not ends, the notion of ‘means’ must include not merely instruments for pro­ curing the final end of happiness, but even more importantly, those actions that are essential components of happiness. ‘Deliberation is not about ends, insofar as it must begin with some end in view; the means that deliberation finds to promote that first end is an end in itself, if deliberation finds that it is a component of the first end’ (Irwin, 1:172). If happiness is essentially constituted by virtuous activity (praxis), and if this consists of actions done for their own sakes, then what happiness consists in are actions chosen by correct reason, not independently of it, and for their own sakes, not as mere ways of achieving ends set by non-rational desire. In fact, non-rational desire as such could never be directed to the essential ­components of happiness. It is the function of reason, not of non-rational desire, to determine the actions which turn out to be the essential ­components of happi­ ness (Irwin, § 96–8). These points require a rationalist interpretation of Aristotle on virtue. A rationalist Aristotle, however, begins to look a lot more like Kant. A rationalist Aristotle takes the rule of correct reason over our conduct to be what is most essential to virtue. This is not so different from Kantian rationalism, which also takes rational self-government (or autonomy) to be the foundation of moral virtue.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

238  Allen W. Wood

Options in the Relation between Reason and Non-rational Desire Before we turn to Kant, it may help if we introduce the comparison by distinguishing, in Aristotle, as Irwin helpfully does, four possible conditions of the soul regarding its rational and non-rational parts: (1) Vice: the rational and non-rational parts agree in pursuing the wrong ends. (2) Incontinence: the rational and non-rational parts disagree; the rational part pursues the right ends, but the non-rational part pursues the wrong ends, and the non-rational part overcomes the rational part. (3) Continence: [same disagreement], but the rational part overcomes the non-rational part. (4) Virtue: the rational and non-rational parts agree, under the control of the rational part, in pursuing the right ends.  (Irwin 1:154)

In looking at (2)–(4), we may usefully consider the parallel fourfold set of options that Irwin offers us regarding the relation of the rational part of the soul to the non-rational or appetitive part: ‘(1) Indulgence: leaving it completely unchecked. (2) Suppression, as far as possible. (3) Control, as far as possible. (Control seems different from suppres­ sion, in that it approves some satisfaction of inclinations, but puts limits on permissible satisfaction: for Aristotle, the limits seem to be set by the virtue of temperance, while for Kant they are set by our duties.) (4) Harmony and agreement with the rational part’ (Irwin 1:155). Aristotle and Kant of course equally reject (1): Indulgence. In the Greek context, Irwin says (2) ‘Suppression was associated with Antisthenes and the Cynics; perhaps it influences Plato in the Philebus.’ (3) Control is apparently the relation constituting continence, while only (4) Harmony is Aristotelian virtue. We might think that Kant rules out (4) Harmony in the first set of alternatives, and therefore also (4) in the second set. For harmony seems not to be even a possibility if (as Kant asserts) inclinations cannot be commanded (G 4:399), and therefore cannot be obedient to reason. Even if Aristotelian virtue is admitted to be possible, Kant might be thought to regard it as morally undesirable (or at best morally indifferent), since he holds that the good will must act from duty alone, never from inclinations (or empirical feelings). Even when duty and inclination contingently agree in providing incentives to the same action, Kant says that action from inclination has no moral worth. Some even think Kant’s real position is (2) Suppression. In the Groundwork he declares that ‘Inclinations themselves, however, as sources of needs, are so little of absolute worth, to be wished for in themselves, that rather to be entirely free of them must be the universal wish of every rational being’ (G 4:428, cf. 4:454). This remark, however, is often misunderstood and needs to be read with care. It speaks only of a wish; it is not a recommendation that inclinations actually be suppressed. Further, it considers undesirable only one aspect of inclinations: ‘as sources of needs.’ There might well be other aspects that would lead us to welcome our inclinations, or even give us reasons to forbid their suppression. The thought stated here is one we know that Kant

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  239 himself associates with the ancient Cynics (the least plausible, in his view, of the ancient schools of ethics) (see VE Collins 27:100–6, 247–50, Vigilantius 27:483–5). In the Groundwork, the Cynic claim is not fully accepted but is being used in a limited way, only to show that inclinations themselves cannot be candidates for the end in itself that could motivate rational obedience to a categorical imperative. We know from what Kant says elsewhere that he definitely does not accept it: ‘Considered in themselves’, Kant says, ‘natural inclinations are good, i.e., not reprehensible, and to want to extirpate them would not only be futile, but harmful and blameworthy as well; we must rather only curb them, so that they will not wear each other out but will instead be harmonized into a whole called “happiness” ’ (R 6:58).5 A more plausible position might be to say that Kant subscribes to (3) Control, as far as possible. The conclusion would then be that although he rejects Aristotelian virtue, Kant embraces instead, under the name of ‘virtue’, only Aristotelian continence. This might be supported by the claims already cited from the Groundwork, as well as the remark just quoted that inclinations need to be ‘curbed.’ It might also be suggested by Kant’s characterization of virtue as ‘the moral disposition in the struggle (im Kampfe)’ (KpV 5:84), as well as his account in the Doctrine of Virtue, which describes virtue as ‘moral fortitude’: ‘the capacity and considered resolve to withstand a strong but unjust opponent’ (MS 6:380). But we will see later both that Aristotelian continence falls short of being Kantian virtue, and that it is wrong to think of Kant as rejecting (4): Harmony.

Duty and Moral Worth The image that most philosophers seem to have of Kant’s ethics comes mainly from one brief discussion early in the Groundwork. What is most famous there are Kant’s claims that the good will is the only thing good without limitation and that actions have moral worth only if done from duty. These two claims, and the relation between them, are commonly subjected to disastrous misinterpretation.6 They are misunderstood right away if Kant’s position is taken to be that only the will that acts from duty is a good will. For after introducing the notion of the good will, Kant explicitly narrows his focus to a special case of the good will, [action from] duty, which ‘contains that of a good will, although under certain subjective limitations and hindrances which, however, far from concealing it and making it unrecognizable, rather elevate it by contrast and let it shine forth all the more brightly’ (G 4:397).

5   At G 4:454, Kant also attributes to a scoundrel both the wish to be a person of good will (without ­ aving to do anything to make himself into such a person) and also the possible wish to be free of burden­ h some inclinations. To say that a scoundrel might think this way can hardly be regarded as Kant’s endorse­ ment of the latter wish. 6   For a fuller account of what I am about to say here, see Wood, The Free Development of Each: Studies on Freedom, Right and Ethics in Classical German Philosophy (Oxford: Oxford University Press, 2014), chapter 1.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

240  Allen W. Wood For Kant there are clearly more favorably situated cases of the good will—not beset with limitations and hindrances—which, for just this reason, are not cases of acting from duty. One of these is obviously that of the ‘holy will’, e.g., the divine will: the will that necessarily acts according to reason (G 4:414). The acts of a holy will have no moral worth, but it is the best conceivable will. The human will in which the moral law harmonizes with inclination would be another case of the good will that does not, even cannot, act from duty, because it is also not encumbered with the limitations and ­hindrances that make possible action from duty. It is extremely common for readers to read into this discussion something that is simply not there: a Kantian account of virtue.7 This error is conspicuously exemplified by Alasdair MacIntyre: ‘To act virtuously is not, as Kant [held], to act against inclin­ ation; it is to act from inclination formed by cultivation of the virtues.’8 Virtue for Kant is the strength of character to resist inclinations when they conflict with duty. But that resistance is easiest when, through rational habit, we have cultivated natural inclin­ ations to be in harmony with moral reason. So for Kant, as for Aristotle, the best rela­ tion between reason and desire is harmony. We can properly address the topic of virtue, however, only after clearing up some common errors about moral worth, acting from duty, and also about Kant’s conceptions of desire, inclination, and feeling. Kant’s purpose in the discussion early in the Groundwork is to identify the special kind of value—moral worth, which lies right at the heart of morality—with the aim of using it to identify the supreme principle of morality. The governance of reason, its capacity for self-constraint, shines forth most brightly, under unfavorable circum­ stances, in the face of ‘limitations and hindrances’, that is, when the action has to be performed in the absence of any co-operating inclination. Both holiness and virtue would make it impossible, or at least less likely, to display this kind of rational control. Virtue might be found in actions with moral worth, but they are certainly not the same. Kant thinks there are many other things that are valuable from the moral standpoint, some instances of them doubtless more valuable than some instances of actions with true or genuine moral worth. His aim in this discussion is to argue that authentic moral worth, the value that lies right at the heart of morality, consists in the control of reason over our actions. That is not contrary to a rationalist Aristotle. Nothing Kant says in the Groundwork is correctly understood if it is taken to express (as people often put it) ‘hostility to the emotions.’9 We will soon see that Kant regards feelings and desires—both rational and empirical—as essential to moral ­motivation, 7   Kant does refer indirectly to virtue when he discusses a good character (G 4:393, 398–9), since virtue for Kant consists in a good character. But the main topic is not character but action—actions done from duty and actions with moral worth. Even a person weak in virtue and bad in character could perform such actions. 8  MacIntyre, After Virtue (Notre Dame: Notre Dame Press, 1984), p. 149. 9   There is no German word that means precisely what we mean by ‘emotion’. Kant’s theory of ‘the emo­ tions’ requires us to consider his account of feelings, inclinations, and dispositions. See Wiebke Deimling, ‘Kant’s Pragmatic Concept of Emotions’, A. Cohen (ed.) Kant on Emotion and Value (London: Palgrave Macmillan, 2015).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  241 and also to both virtue and vice. A perceptive reader of the Groundwork should have known this already. For Kant observed quite early on that the ‘cold-bloodedness of a villain’—his rational self-control over the natural pity he might feel toward his ­victims—makes him not only more dangerous but also more abominable (G 4:394). Kant thinks he would have been less abominable if his healthy human emotions, the emotions felt by the virtuous person, had not been curbed by his instrumental or p ­ rudential reason. Kant’s act of abstraction in considering only the moral worth of actions was done for the purpose of developing his first formula of the moral law in the First Section of the Groundwork. But whatever the argumentative advantages of this strategy, it has often led to the disastrous misunderstanding of the entire spirit of his ethical theory. Sometimes people seem to think that Kant’s term ‘inclination’ is simply a quaint ­eighteenth-century word for ‘desire.’ So when Kant says that actions from inclination lack moral worth, it is sometimes concluded that Kant counts as morally virtuous only someone who displays the ghoulishly icy disposition to act with no desire at all. In such a person, the mechanical obedience to a categorical imperative thwarts all their ­loathsome natural desires—and this is what they suppose Kant thinks virtue consists in. One instance of this misreading so extreme that it rises to the level of witty selfparody, is the following absurd remark from the posthumously published lectures of G. A. Cohen: ‘Kant believes that he is most morally admirable who has a rotten soul, but acts against its natural dictates.’10 This colourful interpretation no doubt makes the Groundwork more exciting to read, but there is nothing else to be said in its favor.

Feeling and Desire in Rational Action Both ‘desire’ and ‘inclination’ are technical terms in Kant’s vocabulary. He defines ‘desire’ (Begierde, appetitio) as ‘the self-determination of a subject’s power through the representation of something future as an effect of this representation’ (Anth 7:251). A theoretical cognition is a representation of an object that corresponds to the object (KrV A57–58/B82–3), and receptivity to the object (KrV A19–20/B33–4). Desire, by contrast, has the opposite direction of fit. It is the causality of a subject’s representation, having a future object as its effect. Here the object is to be made to correspond to the representation of it present in desire. For Kant, every desire or impulse (Antrieb) is the representation of an object accompanied by a feeling of pleasure (Lust) (G 4:444, MS 6:211, 213, KpV 5:72, 116). There can be no desire without feeling. All action, moreover, must proceed from desire of some kind. An action is always chosen as a means to an end set by the agent. In some desires (empirical desires) the feeling of pleasure in the representation of an object precedes and grounds the desire. In others, rational desires, the desire itself—as a self-determination of the subject’s 10   G. A. Cohen Lectures on the History of Moral and Political Philosophy, ed. Jonathan Wolff (Princeton, NJ: Princeton University Press, 2014), p. 150.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

242  Allen W. Wood rational faculty of desire—comes first, and the pleasure in its object is the effect of this determination (KpV 5:8–9n, MS 6:211–13). We decide rationally to do something—to bring about some object—and this decision brings about a feeling of pleasure in our representation of the object. Rational desires, therefore, every bit as much as empirical desires, involve feelings. Of course, rational desires, and the feelings necessary to them, may proceed directly from moral reason (G 4:399, 418, MS 6:215–16, 480). Even rational desires based on non-moral uses of practical reason, such as instrumental rea­ son, are not inclinations because they are grounded neither on empirical impulses nor empirical desires. Rational desire to take the means to an end one has set must be ­distinguished from any inclination that might have led the agent either to set the end or take the means to it. In inclination, the desire follows from the feeling of pleasure accompanying a repre­ sentation of the object. Inclinations rest on instinct, which is the predisposition to be pleased by something even before one possesses it (Anth 7:265; VA-Menschenkunde 25: 1111–12, VA-Mrongovius 25:1339, VA-Busolt 25:1518). But inclination requires acquaintance with the object. ‘Inclination is a determinate principle to desire an object, in so far as it is already known to me’ (VA-Friedländer 25:584). Inclination is ‘habitual sensible desire’ (Anth 7:251; cf. VA-Mrongovius 25:1339, VA-Menschenkunde 25:1109, VA-Busolt 25:1519, MS 6:212). ‘Habitual’ in this formulation is meant as the claim that the desire has been taken up into a rule or maxim of the subject’s will. ‘A sensible desire that serves the subject as a rule (habit) is called “inclination” (inclinatio)’ (Anth 7:265). Inclinations always involve an element of habituation, for good or ill, involving free, rational agency. The fact that inclination involves the rational will, and not merely ­sensibility or instinct, follows merely from the fact that inclination is a species of desire. Desire, as we have seen, is the self-determination of the subject to bring about an object corresponding to its representation. Inclinations are never mere occurrences (in us or to us), to which our free agency is simply passive. Even in empirical desires we are already self-determining; we are always voluntarily complicit in desiring. We are not responsible for our instincts, but we are always responsible for our inclinations. Kant’s usual focus in his discussion of the faculty of desire is on both the feelings and the desires that pose obstacles to the rational agent’s self-mastery, from both the stand­ point of prudence and the standpoint of morality. But certain inclinations are also seen as virtue’s allies. Some rational desires, together with the feelings associated with them, are even indispensable to virtue. In the Groundwork, Kant points to the essential role played by the feeling of respect for the moral law (G 4:401). In the Doctrine of Virtue, he distinguishes four such feelings: ‘moral feeling’, ‘conscience’, ‘love of human beings’, and ‘[self-]respect’ (MS 6:399–403). These are not empirical feelings, but effects of the immediate influence of our reason on our sensibility (MS 6:399). There is no obliga­ tion to acquire these feelings, he says, because without the predisposition to them one could not be a rational agent at all and could not be put under moral obligation. But there is an obligation—and it also belongs to virtue—to respond to these rational feel­ ings and to cultivate them.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  243 Kant does not think of the human soul, as Aristotle does, in vivid Platonic terms, as having parts which, like homunculi, may command and obey one another, or take advice from one another like a son from a wise father (1102b30–2). That is why he does not describe the responsiveness of feelings and desires to reason in the way Aristotle does, and why he does not think feelings or desires, such as empirical (or ‘pathological’) love, can be the object of moral commands. Despite this difference in their moral psychologies, however, we should now begin to see the convergence of Kant’s and Aristotle’s conceptions of virtue. Kant does think that some feelings and desires, those most directly involved in moral motivation and moral virtue, arise dir­ ectly from reason, as a response to the moral law. Like Aristotle, Kant locates virtue not only in dutiful actions done for their own sake, but also in the feelings and desires that go with the performance of these actions, under the guidance of practical reason.

Kant on Virtue Now we are at last ready to take up Kant’s conception of moral virtue. Virtue is a state of our character or ‘way of thinking’ (Denkungsart). This is our nature itself, insofar as we do not get it from nature but make it ourselves as free and rational beings (Anth 7:119, 291–2). Virtue is the agent’s character regarded as the strength of will to con­ form conduct to moral reason (MS 6:405). This strength consists in ‘an aptitude [Fertigkeit] (habitus)’, which is a ‘facility [Leichtigkeit] in acting and a subjective per­ fection of choice [Willkür]. Not every such facility is a free aptitude (habitus libertatis); for if it is a habit [Gewohnheit] (assuetudo) that has become a necessity through fre­ quent repetition, then it is not . . . a moral aptitude’ (MS 6:407). These aptitudes cannot be acquired merely by contemplating the dignity of the pure rational law (though Kant thinks this might help), but also require ‘practicing virtue’ (exercitio) (MS 6:397). Does Aristotle disagree? I think not. Even the appetitive part of the soul does not merely obey a command given by the rational part; it must be habituated to obey correct reason, through actions guided by the rational part, involving rational deliber­ ation and election. Dutiful actions do not have authentic, true, or inner moral worth unless they are done from duty. This means: being done from one of the rational feelings—respect for the law, moral feeling, conscience, love of humanity, self-respect . . . (I see no reason why there could not be other such feelings in addition to these, appropriate to different circumstances, which Kant does not take the trouble to spell out in detail). The rational feelings of love and respect are necessary accompaniments of carrying out duties of love and respect (MS 6:448). Our duties also include the cultivation of empirical feel­ ings, especially those of (empirical or ‘pathological’) love and sympathy. Kant’s discus­ sion in the Groundwork of the ‘friend of humanity’ who is beneficent from sympathetic inclination (G 4:398) is often read as a moral criticism of such people and of their empirical feelings and inclinations. But this misses the entire point of that discussion. Authentic or true moral worth is central to morality, but it is obviously not the only

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

244  Allen W. Wood thing morality cares about.11 We cultivate the right empirical feelings through the appropriate actions: if we do good to others, we come to love them (MS 6:402). Active participation in the situation of another (Teilnehmung)—a mode of conduct to which we have a duty—produces sympathetic feelings (MS 6:456–7). This, and not through obeying a command, is the way Kant thinks the virtuous person acquires the right empirical feelings of love and sympathy. Moral virtue is therefore most perfect when, through rational habit, natural inclinations have been brought into harmony with the commands of reason: ‘Perfect art again becomes nature, which is the ultimate goal of the moral vocation of the human species’ (MA 8:117–18). We can appreciate the role of feelings or emotions even better if we attend to some of the negative ones—to Kant’s account of vices, or the evil disposition. Evil and vice, as Kant describes them, consist not only in a disposition to do the wrong things, but also in having bad desires, bad attitudes and the wrong feelings. For Kant, the chief enemies of both virtue and prudence are the passions—desires that resist the control of reason, and in general consist in wanting superiority over others (in power, wealth, and ­honour, as well as sexual subordination and freedom from all constraints imposed on us by the existence of others—including their rights) (Anth 7:265–75). The radical propensity to evil in human nature is a freely willed disposition to place incentives of self-love or inclination ahead of those of morality. This shows itself not only in actions but also in a variety of feelings and attitudes (R 6:32–44). The vices of hatred toward others include the gloating enjoyment of the sufferings of others (Schadenfreude) (R 6:27, MS 6:458). Other vices of hatred (envy and ingratitude) involve being dis­ pleased at the success of another, and resenting the obligation under which the other’s beneficence has placed you (MS 6:458–60, VE Vigilantius 27:686–8, 693–5). Vices of disrespect to others (arrogance, defamation, ridicule) involve not only actions, but also feelings and attitudes: the attitude of condescension or contempt toward ­others, enjoyment in being able to lord it over another, taking satisfaction or mirth in the fact that something has been made public that turns another into an object of ­disapproval or derision (MS 6:462–8, VE Collins 27:408–12, Vigilantius 27:624–5). Virtue and vice for Kant, every bit as much as for Aristotle, involve being pleased and pained at the right or the wrong things.

Kant and Aristotle Agree about Virtue Virtue for Kant is therefore located in the human character, and is a free or rational habit (not a mere mechanical one), just as it is for Aristotle (at least when he is inter­ preted rationalistically). Virtue is a facility (something that makes easy what would be harder without it) in performing the actions that reason commands. It does this 11   Notice too that the moral worth that belongs only to action from duty is exclusively an appraiser’s value, not an agent’s value—it is not something virtuous agents are ever interested in giving to their actions unless acting from duty is the only way duty will get done at all. And then it is only the doing of duty (not the moral worth of the action) that the agent cares about.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  245 through feelings under the control of reason—some of them rational feelings that are direct responses to reason, others empirical feelings cultivated through the practice of virtuous actions. The agreement between Aristotle and Kant can be seen as still closer when we note Kant’s insistence that the typical temperament of virtue is joyous, not fearful, dejected, and ascetical (‘Carthusian’) (R 6:23–4). He regularly praises Epicurus for holding that a cheerful heart is a mark of virtue (R 6:60, KpV 5:111–13, 116, MS 6:485, VE 27:249–50, 483, 29:603). If the spirit of virtue is courageous and joyous, then it is not that of the continent person, who (Aristotle says) does the right thing, but is grieved by doing it (1104b7–9). We might think that Kantian virtue can never rise to more than Aristotelian contin­ ence because Kant says that we do our duty ‘reluctantly’ and that we must be ‘­constrained’ or ‘necessitated’ by our own reason to obey the moral law (KpV 5:83, G 4:400, MS 6:379). But when Kant says that we submit to the moral law reluctantly, what he means is merely that our will must be rationally self-constrained to obey. What ­ istinguishes virtuous people is that this act of self-overcoming is something willed d and enjoyed—it is not something that pains them, or that they begrudge or resent. If they do resent it, feeling regret or chagrin at not having done the wrong thing, that wouldn’t be virtue after all, or at best it would be an inferior kind of virtue. Virtue gives us an ‘inner freedom’ (MS 6: 406–7)—the freedom of self-mastery—which pleases us by lifting us up and is experienced aesthetically in the feeling of the sublime (KU 5:257–62). This is very close to what Aristotle says about virtues involving self-­control, such as temperance (sophrosyne) (1102b25, 1118b30–1119a20). Irwin agrees that once they are understood, the two conceptions of virtue are not at odds: We would be entitled to mark a sharp contrast between Aristotle and Kant on the place of the emotions in moral virtue if each accepted an extreme claim, so that (i) Kant believed that emo­ tions have no part in virtue, and (ii) Aristotle believed that virtue requires the right emotions, even when they are not up to us. Neither Kant nor Aristotle, however, accepts the extreme claim; acceptance of it would cause serious difficulties for the rest of their positions. A reason­ able assumption about the adaptability of emotions reconciles the Kantian and the Aristotelian position  (Irwin 3:58).

Eudaimonism But what about happiness? There certainly seems to be some disagreement between Aristotle and Kant about that. Aristotle regards happiness as the all-encompassing and final human end: it unifies the system of the ends of crafts, lines of inquiry, actions, and decisions (1094a1). It is complete: we wish for it only for itself, and we wish for all other things because of it (1094a18, 1097a28). It is self-sufficient: it makes life choiceworthy and lacking in nothing (1097b7). It is the end of the highest ruling science (1094a30).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

246  Allen W. Wood Kant believes none of these things about happiness. He thinks happiness is our ­ atural good, distinct from our moral good. It is the end of one species of practical n reason: pragmatic or prudential reason (G 4:417–19). This is a species of practical reason that stands above mere technical reason. But it is not moral reason, and there is no categorical imperative to pursue your own happiness (though there is to pursue the happiness of others (MS 6:385–8). Happiness for Kant is not the whole good, or even the supreme good. Happiness forms a part of the summum bonum, but only con­ ditionally (KpV 5: 110–13). The principle of one’s own happiness, in fact, is the direct opposite of the principle of morality (KpV 5:25). Pragmatic reason directs us, and moral reason permits us, to pursue happiness, as far as we can consistent with duty. But complete happiness is also impossible for us humans (KU 5:430). We humans are so constitutionally fickle and discontented that we are incapable even of forming a con­ cept of happiness that is self-consistent or enduring over time (G 4:417). If, however, we look past the word ‘happiness’ (or its Greek and German equiva­ lents—eudaimonia, Glückseligkeit), it is easy to see important areas of agreement between Aristotle and Kant about the human good. Both distinguish acting well, virtu­ ously, according to reason, from other, lesser goods. For Kant, the good will, which manifests itself in virtuous activity, is the good of one’s person—of one’s will or power of choice (Willkür)—rather than of one’s state or condition (Zustand): they include pleas­ ure and freedom from pain, and what Kant also calls ‘gifts of fortune’—health, wealth, power, and honour (G 4:393). Goods of one’s state seem to correspond most closely in Aristotle to the goods of the body and external goods, which may depend on good for­ tune (1099a30–b5). These are at most necessary conditions for happiness, or auxiliary to a life of virtuous activity. Kant and Aristotle both hold that acting well is decisively preferable to these other goods with which it is contrasted. Both think that virtuous activity constitutes our supreme good and is something always good, whereas these other goods have at best conditional value—as conditions for virtuous action or condi­ tional on virtuous action. We might be tempted to reconcile Aristotle and Kant on the human good by ­suggesting that ‘happiness’ is not the right translation for both eudaimonia and Glückseligkeit. We might at least propose that these words, as used by Aristotle and Kant, should not be treated as equivalent in meaning. Some have suggested, for instance, that ‘happiness’ refers to a subjective state of mind or the goods that produce it (which would be fairly close to Kant’s Glückseligkeit), whereas eudiamonia ought to be translated as something like ‘flourishing.’ I think this strategy is wrongheaded. The words eudaimonia, Glückseligkeit, and ‘happiness’ (and related words in other European languages) all belong to the same philosophical tradition. Further, they do all have the same meaning in ordinary language. All refer, namely, to what we may call your individual good, to what is best for you, or in your interest, or what makes you well off. This entails that there is an important philosophical disagreement between Aristotle and Kant—about the individual human good. Aristotle holds that what is supremely good for you—acting well or virtuously—is also the dominant component of your

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  247 individual good. It is the chief thing that makes you well off, is in your interest, or is best for you. For Kant, however, there is something better than your individual good, better than your self-interest or your being well off. Your individual good (expressed in any of these ways) is merely the good of your state whereas the supreme human good is the good of your person. This is different from your individual good; it is what makes you worthy of possessing your individual good (G 4:393). Kant, therefore, regards the human good as heterogeneous, consisting of two distinct objects of desire and choice: the moral good (Gut), and the natural or physical good or well-being (Wohl). He criticizes the ancients, especially the Stoics and Epicureans, for identifying the two kinds of goods, and scolds the moderns for confusing them (KpV 5:64–5). The Epicurean identifies virtue with what is pleasant (what affords the best and purest pleasures, those of virtuous activity and friendship) while the Stoic (perhaps closer to Aristotle) insists that virtue itself constitutes happiness (KpV 5:111–12, cf. Anth 7:235). This criticism is important, because it is essential to Kant’s own doctrine of the highest good (­summum bonum). The highest good requires two distinct elements, virtue or morality and happiness which are related conditionally. Aristotle and Kant therefore disagree. But we haven’t yet located precisely where. There is also one further significant point of agreement that must be appreciated. It is not Aristotelian to hold that virtuous action is sufficient for happiness. That would be the extreme Stoic position, perhaps also that of Socrates. Aristotle’s view is that being virtu­ ous and acting virtuously, as the dominant component of happiness, is, on the whole, ­ nfortunate circum­ the best way of achieving happiness, even if (sometimes owing to u stances) virtue does not lead invariably to happiness. Here Kant may actually come close to agreeing with Aristotle and with much present day virtue ethics. For he thinks that under the actual conditions of human life, the maxim most likely on the whole to lead to happiness or well-being is that of ‘reasonable self-love’ (vernünftige Selbstliebe) (KpV 5:73, R 6:45n)—a pursuit of happiness moderated to accord with duty.12 It would also be a mistake to think that the point of disagreement is to be found in the Kantian thesis that even if it is true that virtue leads to happiness, to advocate a vir­ tuous life on this ground is to give the wrong reason for leading a virtuous life. Kant holds that you should live virtuously not because it will make you happy but rather because that is simply the way you ought to live, whereas Aristotle thinks you should live virtuously because it is good for you, or will make you happy. But it is a mistake to see any disagreement here. For Aristotle too would deny that the reason you should live virtuously is that living virtuously is apt to provide you with what Kant means by ‘happiness’—a desirable state or condition (favourable circumstances, health, wealth, 12   On this, see Stephen Engstrom, ‘The Concept of the Highest Good in Kant’s Moral Theory’, Philosophy and Phenomenological Research 52 (4) (1992). The agreement is only a limited one, however. For Kant it would be a true empirical claim that your best chance at happiness consists in pursuing ‘reasonable happi­ ness.’ For Aristotle, however, it is not merely an empirical fact that your individual good is best achieved by pursuing eudaimonia; on the contrary, eudaimonia simply is the human good in its entirety, and this is a necessary truth.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

248  Allen W. Wood power, honour, etc.). Like Kant, Aristotle would say that that the reason one should live virtuously is that living virtuously is living the way you ought to live. Aristotle thinks virtuous activity is done for its own sake, or (what is equivalent for Aristotle) for the sake of the fine (the kalon). Kant too thinks dutiful or virtuous activity is done for its own sake and because it is simply good in itself. But Kant infers from this that virtuous activity must be different from happiness. For Aristotle, however, this is why living this way makes you happy. That raises the question: Why doesn’t Kant agree with this last inference? The real disagreement between Kant and Aristotle is best framed, I think, as a ­disagreement over the structure of practical reason. Kant holds that prudence or ­pragmatic reason, or skill in conceiving of and pursuing our individual good or happi­ ness, is a function of practical reason distinct from and independent of moral reason, and inferior to it (G 4:416–18). It is significant that Aristotle’s term phronesis (the practical wisdom that lies at the center of Aristotelian virtue) was translated into Latin as prudentia; and then also that this term later came to mean what Kant means by prudence (prudentia, Klugheit)—namely shrewdness, circumspection, and caution involved in pursuing one’s own self-interest.13 My diagnosis of this change in the meaning of the term ‘prudentia’ (and its cognates or equivalents) is that it went along with a profound change in the intended referent (but not in the meaning) of ‘happiness’ and its equivalents in Greek, German, and other languages. This is a deep change in the way people came to view human life. Kant and Aristotle agree that our highest aim in life is (or should be) virtuous action (or goodness of will). But Kant denies that this should, or even can, be a finite rational being’s entire or sole aim in life. For him our self-love and self-interest cannot be directed toward the higher moral goodness of our person; it must be directed toward the goodness of our state or condition. Although Kant recognizes that awareness of the goodness of our person produces a feeling of pleasure with ourselves, or at least of negative contentment with our person, he nevertheless denies that this feeling could constitute our happiness, or even the smallest part of it (MS 6:387; KpV 5:38, 88, 117, 156). This is how Kant recognizes a dualism of practical reason regarding human ends. It is different from the dualism later claimed by Sidgwick. It avoids Sidgwick’s troubling dilemma by maintaining a decisive superiority of the one kind (moral reason) over the other (prudential reason). In comparison to Aristotle, however, Kant represents a step downward, in the direction of Sidgwick. Further steps down that path would take us to a merely instrumental conception of reason, or a pluralistic and relativistic conception of it. The bottom of the slippery slope would be the ultimate depravity of saying that reason is and ought to be only a slave of the passions. Aristotle himself, however, has 13   Kant even recognizes a related meaning of these terms that refers to skill in manipulating people, using them as means for your own ends—even when this might involve the violation of the moral principle that forbids treating other persons as mere means (Anth 7:201, 272, VP 9:486, KpV 5:37, R 6:24, MS 6:455).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  249 already taken a step down the slope compared with Socrates before him, or the Stoics after him. For he acknowledges at least a conditional and auxiliary role for goods other than virtue in the constitution of human happiness. The Socratic and Stoic position, for Aristotle, is too high-minded to get at the whole truth about the human good. This is precisely what Kant thinks about the Aristotelian position, along with other ancient views. We might want to ask: what is the right position to stop on this slope? Or where is the correct mean on the continuum? I won’t try to settle this difficult question. Instead, I want to understand better the disagreement between Kant and Aristotle by consider­ ing a way someone might defend the Aristotelian position against the Kantian. The Aristotelian might claim that there is a kind of irrationality or inconsistency built into the Kantian dualism of practical reason. Why shouldn’t you identify your individual good—what is best for you, what is most in your interest—with what you yourself say is the most rational course, the course of the most excellent deliberation and action? If there are other goods besides virtuous action, then those who place their happiness in these lesser goods seem foolish and irrational. They are base souls who identify happi­ ness with the brutish pleasures of Sardanapallus, or the fickle and ignorant rabble who capriciously identify happiness with health when they are sick, with wealth when they are poor, and so on (1095a20–5, 1095b20)? As we have seen, Kant himself does hold that people in general—even the best of us—are like this rabble. We are incapable of forming any constant and coherent conception of the goodness of our state or condition. Kant even denies, on just these grounds, that achieving such a state could be the natural function of our practical reason (G 4:395–6). Why, then, does Kant think that the ‘wavering idea’ of this desirable or pleasant state (G 4:399) is the same as what is best for us, even when virtuous action is not the way to reach it? Isn’t it irrational, or even inconsistent to think this? However compelling this argument may seem in the abstract, to most of us moderns it seems strangely blind—high-minded, perhaps, but naïve and hopelessly cut off from the reality of human life. We moderns know that in this nasty world many people are in a less desirable state or condition (less well off, less wealthy, honoured, or powerful) than they should be just because they have done the right thing. Others obtain more power, wealth, and honour, and the other things we all want, and achieve greater ­satisfaction with their state, precisely by doing things that are morally wrong. This is true even if we agree with both Kant and Aristotle that the virtuous life is, on the whole, the one most likely to lead to happiness. In cases where the world has taken better care of the happiness, the individual good or self-interest, of bad people than of good ­people, we think things ought to be different. Moreover, we don’t think we are being depraved or foolish to think this. Rather, along with Kant we think the principles of virtuous action are different from those that often to our individual good, self-interest, or happiness. Does Kant have a philosophical defence of this modern view? I think he does. There is too little space here for me to do more than sketch it. It lies not in his moral theory,

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

250  Allen W. Wood the a priori part of his philosophy, but rather in the empirical part—his philosophical anthropology and philosophy of history. Kant’s conception of human nature is that it is self-made by humans out of materials given by nature. More particularly, our nature consists above all in a certain task or vocation nature has set our species: ‘In compari­ son with the idea of possible rational beings on earth in general, the characteristic of the human species is this: that nature has planted in it the seeds of discord, and has willed that its own reason bring out of this concord, or at least the constant approxima­ tion to it’ (Anth 7:322). This means that we must look at human nature from two standpoints. I don’t mean the notorious metaphysical distinction between sensible and intelligible standpoints. I mean the historical and anthropological distinction between our starting point as natural beings and our historical goal as rational beings. As natural and social beings, we are discordant, both with ourselves—always dissatisfied—and with one another— antagonistic, competitive, beset with unsociable sociability. But as rational beings, our moral vocation is to bring about concord. We should bring our feelings and desires into accord with reason, our state into harmony with reasonable self-love, and our ends into harmony with the ends of others in an ideal community or ‘realm of ends.’ What we are is a troubled mixture. We are the selfish, discordant, irrational beings ­conditioned by our society and also the rational, concordant, moral beings of our aspirations. Our good is heterogeneous because it includes both parts of our selfalienated nature. Kant is not known for making concessions to human weakness and irrationality. He thinks the moral law makes none, but Kant’s philosophy as a whole certainly does. We necessarily find our individual good tangled up with our comparative–competitive relations with others: our rational predisposition to humanity ‘can be brought under the general title of a self-love which is physical and yet involves comparison (for which reason is required); that is, only in comparison with others does one consider oneself happy or unhappy’ (R 6:27). Our self-love necessarily involves ‘seeking superior ­status . . . it involves jealousy and rivalry; we consider ourselves poor only when we think that others will consider us so, and despise us for it’ (R 6:27, 93). It is our voca­ tion, our endless task, to lead a conflictual life, pursuing our individual good, but also subordinating it to our rational moral vocation. It is from this standpoint that all ancient eudaimonism appears naïve and out of touch with our human condition. Ancient ethics behaves as if our nature, and our good, were determined exclusively by the rational end we ought to seek. It imagines that we can afford simply to ignore the inevitable but contrasting aspect of what we are, the social and psychological whence of our striving. It supposes too optimistically that our present moral condition is close enough to satisfactory that we can think about our good as if this good were already natural to us. But in our present state, moral progress is not a process of natural growth; it is a hard struggle, against ourselves and even partly against our nature. Virtue partakes not only of happiness and the fine, but also of the terrible and the sublime.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

virtue: Aristotle and Kant  251 This is not a dispute between two great philosophers as much as it is a yawning gulf between ancient high culture and modern Enlightenment culture. Aristotle holds that happiness consists in virtuous activity (praxis). We learn from the Nicomachean Ethics that true human happiness, however, is found only in theoria. This is a life as close as humans can come to that of God. Perhaps in ancient high culture, the privileged lives of the upper class enabled them to aspire to live like gods, consigning the menial tasks of the human condition to the rabble. We moderns, however, have come to see that we are all rabble. Many of us even believe in what Kant called human dignity—that is, we believe that the highest status anyone attains is just that of equality with the most wretched and humble human being. We don’t look kindly on those who want to live like gods at the expense of the rest of us. Kant too, of course, devoted himself to the life of theory, science, and philosophy. Here is what he wrote about that, quite early in his career: I am myself by inclination an investigator. I feel a complete thirst for knowledge and an eager unrest to go further in it as well as satisfaction at every acquisition. There was a time when I believed that this alone could constitute the honour of mankind, and I had contempt for the ignorant rabble who know nothing. Rousseau brought me around. This blinding superiority disappeared. I learned to honour human beings, and I would find myself far more useless than the common laborer if I did not believe that this consideration could impart to all others a value establishing the rights of humanity.  (Ca Notes and Fragments, p. 7)

The life of theory—learning, science, philosophy—has no value for Kant unless it helps establish the universal and equal rights of all human beings. It is only the philosopher’s possible contribution to the social and historical cause of universal human equality that enables him to look himself in the mirror and tell himself that his life is worth even as much as that of the most common laborer. That’s why the standpoint of the rabble must not be excluded from the structure of practical reason. It is also why the wisdom of the ancients seems to us out of touch with the human condition.

Citations Aristotle: Nicomachean Ethics, translated by Terence Irwin, second edition. Indianapolis: Hackett, 1999. Cited by Becker number. Kant: Ak Immanuel Kants Schriften. Ausgabe der königlich preussischen Akademie der Wissenschaften (Berlin: W. de Gruyter, 1902). Unless otherwise footnoted, writings of Immanuel Kant are cited by volume:page number in this edition. Ca Cambridge Edition of the Writings of Immanuel Kant (New York: Cambridge University Press, 1992). This edition provides marginal Ak volume:page citations. Anth Anthropologie in pragmatischer Hinsicht (1798), Ak 7.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

252  Allen W. Wood Anthropology from a pragmatic point of view, Ca Anthropology, History and Education. G Grundlegung zur Metaphysik der Sitten (1785), Ak 4. Groundwork of the metaphysics of morals, Ca Practical Philosophy. KpV Kritik der praktischen Vernunft (1788), Ak 5. Critique of practical reason, Ca Practical Philosophy. KrV Critique of Pure Reason. Ca Critique of Pure Reason. KU Kritik der Urteilskraft (1790), Ak 5. Critique of the power of judgment, Ca Critique of the Power of Judgment. MA Mutmasslicher Anfang der Menschengeschichte (1786), Ak 8. Conjectural Beginning of Human History, Ca Anthropology, History and Education. MS Metaphysik der Sitten (1797–8), Ak 6. Metaphysics of morals, Ca Practical Philosophy. R Religion innerhalb der Grenzen der bloßen Vernunft (1793–4), Ak 6. Religion within the boundaries of mere reason, Ca Religion and Rational Theology. VA Vorlesungen über Anthropologie, VA 25; transcribed texts are also supplied. Lectures on Anthropology, Ca Lectures on Anthropology, VE Vorlesungen über Ethik, Ak 27, 29; transcribed texts are also supplied. Lectures on Ethics, Ca Lectures on Ethics. VP [Vorlesungen über] Pädagogik, Ak 9. Lectures on Pedagogy, Ca Anthropology, History and Education.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

14 Richard Price on Virtue Roger Crisp

Richard Price (1723–91), a Welsh dissenting minister, theologian, and philosopher, set out clearly and forcefully an intuitionist epistemology and deontological ethics which provided the foundations for later work by W.D. Ross and others. In 2014, I had the great privilege of convening with Terry Irwin a seminar in Oxford on Price’s moral philosophy. As I have found on many other occasions, the combination of a lessstudied text with Terry’s insight and remarkable breadth of knowledge revealed a huge amount about the origin of several contemporary positions, their strengths and weaknesses, and their potential for further development.1 In this essay, I shall focus in particular on Price’s view of virtue, bearing in mind the subtitle of his Review of the Principal Questions of Morals, where Price says that he will concern himself with questions ‘respecting the origin of our ideas of virtue, relation to the deity, obligation, subject-matter, and sanctions’.2 Price speaks broadly of the ‘obligations of virtue’ (11), seeing virtue as in effect equivalent to morality, obligation, or what is morally right (15; 105–6; 110).3 So an examination of his views on virtue will engage with his firstorder or normative ethics as a whole.

1.  Virtue and Happiness Price is convinced that, in general, choosing a life of virtue over one of vice will advance the agent’s own happiness. This is in large part because ‘[s]elf-approbation and selfreproach are the chief sources of private happiness and misery’ (59). The virtuous 1   This essay is dedicated to both Gail and Terry, in admiration of their work, and with gratitude for their friendship and encouragement. I wish also to thank those who attended our seminars, especially David Brink, who also provided extensive and extremely helpful comments on an earlier draft of this essay, and Laura Callahan, who offered an insightful presentation on Price’s views on virtue. 2   I shall use the 1974 reproduction of the 1787 3rd edn of the Review. All unattributed references in the text are to this edition. 3   Note that Price distinguishes ‘the virtue of the action’ from ‘the virtue of the agent’, the former being equivalent to what is ‘objectively right’, independently of the intention of the agent, and the latter to the agent’s willing what she believes to be right (184; see Sects. 3 and 5 below). As he points out, this distinction allows him to claim that doing what is objectively wrong (as far as the virtue of the action is concerned) can be virtuous (since the agent’s motivation is correct).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

254  Roger Crisp person loves virtue more than anything else, and pursuing and engaging with the objects one values most highly will produce the greatest pleasures (222). Nor is it any objection to this claim that making virtue one’s primary goal will be self-defeating, since virtue requires disinterestedness (222–3). The virtuous person aims not at his own happiness, but to act rightly; pleasure is not a cause of his desires, but an effect (see Butler 2006: Sermon 11.6–7; Irwin 2008: 716). Vice, however, if its possessor is not entirely evil, will produce in the agent a state of ‘constant intestine war’ (259), in which the agent is torn between rational and irrational desire. But what if the agent is entirely evil? Price admits that such a person may be happier than the partially virtuous agent (258–9). Further, as far as happiness or pleasure are concerned, in the present state of the world these vicious individuals—and presumably on occasion the conflicted individuals—are the happiest. Price does not say explicitly that these vicious individuals could not have been even happier had they chosen virtue, but this is clearly what he means: ‘A person who sacrifices his life, rather than violate his conscience, or betray his country, gives up all possibility of any present reward, and loses the more in proportion as his virtue is more glorious’ (258). Price’s view here contrasts strongly with Aristotle’s: It is true also of the good person that he does a great deal for his friends and his country, and will die for them if he must; he will sacrifice money, honours, and in general the goods for which people compete, procuring for himself what is noble . . . he gets what is noble, and therefore assigns himself the greater good.  (2014: 1169a)

On Aristotle’s view, virtue is always the best choice to make as far as the eudaimonia— the happiness, or good—of the virtuous agent is concerned. It cannot guarantee happiness—as in case of Priam (2014: 1101a)—but even when it does not result in happiness it still produces the best outcome for the agent, since the alternative would be vicious and hence unhappy rather than ‘non-happy’. Price sometimes uses language which suggests the Aristotelian position. Having mentioned the ‘tranquility and bliss’ consequent on virtue (229), he goes on—with explicit reference to Plato’s Republic—to praise the comparison of the mind of the virtuous person to a well-governed and harmonious state, asking: ‘Is there any thing that deserves ambition, besides acquiring such a mind? In what else can the true blessedness and perfection of a man consist?’ A virtuous mind is the ‘first and highest good’. Later, however, having allowed that virtue should always be chosen in preference to vice, whatever one’s circumstances, Price makes it clear that this is not because virtue is ‘more profitable than vice (or attended with more pleasure)’, but because it is of ‘intrinsick excellence and obligation’ and ‘is to be chosen for itself, independently of its utility’ (257–8). The contrast here between, on the one hand, what is ‘profitable’ (presumably either in itself or for its consequences) and, on the other, what is intrinsically good independently of any such profit or utility suggests that we must interpret Price in the earlier passage to be implying that a person’s highest good is not always identical with

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  255 what is best for them. This is another place, then, where he parts company with Aristotle, denying the central and unstated premise of the ‘function’ argument of Nicomachean Ethics 1.7 (2014: 1097b–1098a).4 Price might appear, then, to be attracted towards a hedonistic account of the good, such that consistency with the view that virtue always benefits the agent could be achieved only through the adoption of some extremely implausible claims about the pleasantness of acting virtuously. But in fact he has an independent reason for denying the Aristotelian position. On Price’s view, we have a great deal of evidence for the existence of God. In the Dissertation on the Being and the Attributes of the Deity, published with the 3rd edition of the Review, Price claims that ‘this visible universe, of which we are a part, is, self-evidently, an exhibition of the power and wisdom of a powerful and wise cause’ (285). This allows him to use the fact that the wicked prosper in this world as itself an argument for an afterlife, in which all will receive their just deserts: Can it be conceived, that the wisdom and equity of providence should fail only in the instance of virtue? That here, where we should expect the exactest order, there should be the least? . . . ’Tis of little consequence, how much at any time virtue suffers and vice triumphs here, if hereafter there is to be a just distinction between them, and every inequality is to be set right.  (260–1)

But what is to prevent God’s always immediately rewarding the virtuous and punishing the vicious in this world? This, Price suggests (261), would make it impossible for him properly to test our characters, since we would always choose virtue for ‘mercenary’ reasons. Here, however, we can see how close to the wind Price is sailing. On the one hand, as we have seen, he wishes to claim that moral obligations are overriding, and should be acted on whatever the consequences for one’s happiness. On the other, his theology involves the idea that virtue will always be rewarded and vice punished, perhaps to an infinite degree: virtue, that is to say, is not only ultimately cost-free, but also ultimately maximally beneficial to the agent. But if I perform the virtuous action now with an eye on heavenly reward perhaps many years hence, why is that any less mercenary than my doing so for an immediate reward? Price’s response would doubtless be to return to the Butlerian distinction between pleasure as cause and pleasure as effect. But again it is not clear why a clear grasp of that distinction would not allow God to create a system in which immediate reward follows only on virtuous actions motivated independently of any desire for that reward itself. I have just mentioned the possibility that rewards and punishments in the afterlife may be infinite. Price was a brilliant probability theorist, and in the Review’s conclusion attempts a further happiness-based, Pascalian argument for virtue grounded on the implications of infinite goods and bads for expected utility theory: ‘when the good is infinite the value of any chance for it must be likewise infinite’ (271). This argument is problematic in at least two ways. First, it is dangerous pragmatically for the sorts of reasons discussed above. If Price persuades me that I will be punished   Irwin (2008: 717) suggests that Price seems not to have understood what Aristotle had in mind.

4

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

256  Roger Crisp for eternity for vice, how can I avoid being motivated, to some degree, by that belief, which may itself put me at risk of that very punishment for failing to act in the morally correct way? Second, there is a possibility (it does not matter how small) that God admires those self-reliant individuals who pursue their own good at any cost to that of others. That possibility cancels out that of Price’s God, and Price will have to rely on non-probability-based arguments to make his case (see Mackie 1982: 203).

2.  Virtue and Self-Love The sixth chapter of the Review begins with an explanation of the close relationship between the concepts of fitness, rightness, moral goodness, reasonableness, obligation, ought, should, duty, and virtue. Price then considers the view that we are not obligated to refrain from anything, even if it is wrong, unless refraining from it advances our own self-interest or is commanded by a superior power (105–6). He charges the view with self-contradiction, on the ground that wrongness just is what we are obligated to avoid independently of self-interest or the will of any higher power. The charge of self-contradiction is an odd one, however, since Price’s opponent will of course deny that she is using wrongness in Price’s sense. Nor does her usage seem to involve any kind of conceptual error. Price goes on to claim that: those who maintain that all obligation is to be deduced from . . . self-love, assert what . . . implies, that the words right and just stand for no real and distinct characters of actions; but signify merely what is . . . conducive to private advantage, whatever that may be . . . ; and any the most pernicious effects will become just . . . if but the smallest degree of clear advantage or pleasure may result to him from them.

Price’s argument here is not straightforward. On one interpretation, he is once more assuming a conceptual truth: that rightness cannot be identical with what maximally promotes the agent’s self-interest. This may be because he accepts the Kantian view that duty must provide reasons in at least potential competition with self-interest (Kant 2002: 4: 414). But this again is certainly not a conceptual truth, and further argument is required. On the face of it, there is nothing incoherent about a form of ethical egoism, according to which each agent is morally required only to do what most advances her self-interest. Nor does there seem to be anything to prevent such an egoist from adopting Price’s realist metaethics, and arguing that the real property of rightness is in fact identical with the real property of maximizing the agent’s good. Another interpretation puts more weight on Price’s reference to ‘deduction’. It may be that he has in mind a form of normative egoism stated in non-moral terms, from which position are then ‘deduced’ various claims about moral obligation. One way to make such a deduction valid, of course, will be to include a further premise, such as ethical egoism. But it may be that Price has in mind a position according to which the word ‘right’ is used just to mean what is in the agent’s self-interest. If so, one might have

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  257 expected him to point out that this position relies on a clear misuse of language, rather than that it leads to morally unacceptable conclusions. The ‘unacceptable conclusions’ argument also seems to rely on something close to a hedonist conception of the good. Price fails to consider the view—which may plausibly be ascribed to certain ancient philosophers, including Aristotle—that the foundational normative principle is egoistic and can be stated in non-moral terms: any agent has reason only to do what best promotes her own good. The good itself may then be said to be identical with doing what is required by morality. (These requirements may themselves be reason-giving, but need not come into conflict with the egoistic principle. Indeed it may be a necessary truth that the agent’s good will be maximally promoted by acting in conformity with morality. Equally, however, the requirements may not be reason-giving, and then some other account of them will be required, such perhaps as a divine command view.) Nevertheless, Price’s point about the implausibility of the implications of ethical egoism is a strong one, and he later uses an analogous argument against consequentialism (131–8; see Sect. 4 below). Both egoism and utilitarianism are strongly counter-intuitive, and Price rightly puts the onus of justification at this point on proponents of either. Price’s conceptual claims appear to commit him to what some have called ‘moral rationalism’, the view that moral reasons are always overriding. That position emerges also in a further revaluation argument against egoism (107–8). Consider a case in which what morality requires is against my self-interest. According to the egoist, I have here no reason to obey morality. This, Price claims, implies not only ‘that virtue itself would cease’, but it ‘would be changed into vice’. This is because, on the egoist position, my obligation must be to do what is best for myself; so, going against that obligation would itself be vicious. Again, to respond to this argument the egoist may insist on a distinction between rationality and morality, that is, between what one has a reason to do and what is required by morality. According to the most plausible form of normative egoism, my only reason is to advance my own self-interest, and this reason is not itself an ‘ethical’ or ‘moral’ reason. This allows for a conception of morality as a set of demands or requirements that are not in themselves normative, that is, do not constitute practical reasons. On this view, the requirements of morality may be seen as analogous in certain ways to the requirements of logic or epistemic rationality. So they may be genuine requirements, but the question of whether they ground practical reasons remains open, a possibility ruled out within Price’s somewhat narrow conceptual scheme.

3.  Virtue as Law, and Supererogation Later in the same chapter, Price says of rectitude that it is ‘a law as well as a rule to us’ (108). A rule merely directs, and a moral rule certainly does that; but it also binds us, in the sense of having ultimate and primary authority (108–10). Further, morality is allencompassing and overriding: any other alleged rule of action can only assist in our

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

258  Roger Crisp discovering what it is right to do. And when one judges that some action is ‘fit to be done’, one is: tied in the most strict and absolute manner, in bonds that no power can dissolve, and from which he can at no time, or in any single instance, break loose, without offering the most unnatural violence to himself; without making an inroad into his own soul, and immediately pronouncing his own sentence.

This seems to me an especially clear expression of what I have elsewhere (Crisp 2006: 35) called ‘the myth of bindingness’, the idea that something normatively significant is achieved through adding to the claim that one has overall reason to ø the further claim that one is ‘bound’ to ø, where the notion of bindingness is not being used merely to point to the existence of sanctions for non-ø-ing. The passage also enables us to anticipate the problem Price is going to have with supererogation. Standard cases of supererogation seem to involve the agent in going beyond the bounds of reason and morality, and it is not clear why this will not involve her doing the most unnatural violence to herself. As one might expect from a writer with the insight and integrity of Price, the general problem is recognized a few pages later, with a reference to Lord Kames’s Essays (119–24). The first form of the objection is put in terms of rightness. How can the right be what we are bound to do, if there are actions that are right, and even more praiseworthy than if they had been our duty, and yet we are not bound to perform them? Price here draws a distinction between two senses of ‘right’. One is the sense he has been using, which is opposed to ‘wrong’ and means ‘such that one is obliged’. The right action is that which one is rationally required to perform. But there is another, broader sense in which what is right is what is merely permitted. The right action will, of course, always be a right action; but there will be actions that are right but not required. This allows Price to draw a distinction between perfect and imperfect duties, though he does not use this Kantian terminology (120–1). His example is beneficence. We may say of someone who does something kind that they were not obliged to do this particular act. Beneficence in general is a duty, and so the person who never helps others will be blamed, but we have a certain amount of discretion to decide whom we benefit and how. Further, echoing Aristotle (2014: 1098a–b), Price suggests that ethics is somewhat indeterminate. No one can tell, for example, exactly how much of her wealth she should donate to charity, though there is clarity at the extremes: giving a few pence will be insufficient, while giving everything (if she has duties to her children, for example) is excessive. But, as Price has already insisted, everyone is obligated ‘to do all the good he can for his fellow-creatures’ (120), and in general ‘whenever any behaviour . . . appear, all things considered, BEST’, it becomes obligatory (122). So any particular action, Price claims, will be either right (in the sense of ‘obligatory’), wrong, or indifferent. And a particular action, such as helping someone in need, may be indifferent, although there may be a duty to perform such actions in general.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  259 But allowing for imperfect duties is of course not yet to allow for supererogation, as Price again sees, asking how it can be that certain indifferent actions are more ‘amiable’. The answer, he says, is ‘very easy’. As we shall see, Price’s view is that the virtuous agent is motivated by concern for the good and the right. He first notes that this motivation can arise even in cases of indifference. So, when helping someone, I may be motivated by an argument along the following lines: ‘There is a general duty to help people. This action is a case of helping someone. So I will help this person’. Price then asks how it is that we often see ‘greater merit’ in actions of this kind. I take him not to be implying that we generally find beneficent actions more meritorious than, say, just actions. Rather, he is referring to the apparent phenomenon of supererogation. I say this because of his own response to the problem: As every action of an agent is in him so far virtuous, as he was determined to it by a regard to virtue; so the more of this regard it discovers, the more we must admire it. And it is plain, it is more discovered, and a stronger virtuous principle proved, by fixing (in cases where the limits of duty are not exactly defined) upon the greater rather than the less.  (123)

This discovery argument shows that Price’s position is in fact inimical to the idea of supererogation. He allows for indifference only in cases of discretion within the limits of imperfect duty. In the case of a person who appears to be going beyond duty, Price will insist—if the person is acting rationally—that this is in fact a case of doing what is best, and so required rather than optional. As he puts it, ‘acting beyond obligation’ is equivalent to ‘acting contrary to obligation’ (124). But there is a further problem, internal to his own strictly deontological position. Consider a case in which I know that I am required to help someone financially. I know that I am required to give them £90–100 or so, but cannot be more precise. According to Price, if I give £100 then this proves a ‘stronger virtuous principle’ in me, and so I am to that extent more praiseworthy. But then it is not clear why any agent in a similar position is not required to give £100. Why would we not blame someone for—meanly— giving only £90 when they could have given £100? Now Price does note here the distinction between the virtue of the action and the virtue of the agent. So his thought could be the following. It is permissible for any agent, when it is required of them to give £90–100, to decide that any amount within that range is required of them. So, the person who gives £90 is doing what she believes is right, and so cannot be blamed. But because her action is less demanding it is evidence of a ‘weaker virtuous principle’ and her action is therefore less ‘amiable’ than that of the person who gives £100. This seems problematic in two ways. First, Price seems to be allowing that a virtuous agent—even the most virtuous agent in such circumstances—is not only ignorant of the true normative circumstances, but ignorant in a way that allows her to do something morally more admirable. Second, it remains unclear why we are not required to act on the strongest virtuous principle in every case—that is, how it can be morally permissible to do less than the morally best. As Henry Sidgwick later put it (1907: 220), this notion will seem to common sense an

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

260  Roger Crisp ‘immoral paradox’. On Price’s view, it seems that it will be significantly easier for the naturally stingy person, for example, to avoid wrongdoing than a person with abundant natural generosity.

4. Consequentialism Price was well aware of the utilitarian view that morality is to be understood as consisting solely in a principle of impartial beneficence. His immediate response, with liberal quotation from Joseph Butler, tends to consist in counter-examples.5 We think that the virtuous deserve to be happy, and the vicious unhappy, and would choose to bring that about even if the opposite distribution produced no less happiness overall (79–80). Gratitude and promise-keeping are good, and stealing the fruits of someone’s labour bad, independently of consequences (131–2). We think that lying is wrong even if it has no bad effects, and must conclude that there is ‘intrinsick rectitude’ in veracity (133). And, in a thought experiment reminiscent of later work on future generations by Sidgwick (1907: 414–16) and Derek Parfit (1984: part 4), Price notes that on the utilitarian view: any number of innocent beings might be placed in a state of absolute and eternal misery, provided amends is made for their misery by producing at the same time a greater number of beings in a greater degree happy.  (159–60)6

Price does not wish to play down the significance of beneficence. He admits that it is ‘the most general and leading consideration in all our enquiries concerning right’, and even that in extreme cases it can override all other obligations, including justice (153). The promotion of happiness, he suggests, is God’s only reason for creating and sustaining the world, though he pursues that end under the (genuine) constraints of justice, veracity, and other principles of rectitude (249–52). In his conception of beneficence, then, Price anticipates Ross, whose conception of beneficence is essentially equivalent to the utilitarian principle, but whose ethics also contains other principles in possible tension with beneficence. One can imagine Philippa Foot, for example, asking Price whether he really believed that bringing about the scenario in which innocent suffering is outweighed by the creation of happy beings is an exercise of beneficence (Foot 1985: 205). (And imagining this brings out how Foot’s question is really just raising in a secular context the problem of natural evil.) There is also here in Price an implicit response to those who claim that any ethical ­theory can be ‘consequentialized’ (e.g. Dreier 1993). Justice, for example, is not itself an end or a good that can ground teleological reasons. Of course, one could describe a consequentialist view according to which justice was such a good and which had the   See also his criticism of egoism at 105–6 (discussed in Sect. 2 above).  Here Price is also anticipating John Findlay’s claim (1961: 235–6) that ‘the separateness of ­ ersons . . . is . . . the basic fact for morals’. p 5 6

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  261 same practical conclusions as Price’s view. But the reasons for those conclusions would be quite different in each case, and so therefore would the underlying moral theories. Price is fully aware of the ‘two-level’ consequentialist strategy, according to which the apparently non-consequentialist judgements we make in the cases he describes are: owing to the idea of a plan or system of common utility established by custom in the mind with which these vices are apprehended to be inconsistent; or to a habit acquired of considering them as of general pernicious tendency, by which we are insensibly influenced, when, in any particular circumstances or instances, we contemplate them.  (134–5)

Price’s subtle response to this strategy is essentially twofold. First, he objects to the pressure for normative monism within consequentialism (137–8; see Irwin 2008: 750). Why should, for example, truth or gratitude not matter independently of happiness, and ground their own independent principles? This is a powerful point. Of course, some principle of parsimony must operate within ethical theory as elsewhere. But there is no reason a priori to assume that ethics must rest on a single principle. Nevertheless, it must be admitted that by making the promotion of happiness so central within his own account, Price provides something of a hostage to consequentialism. If happiness is really the only reason for our being on this earth, should we not take very seriously the idea that other things we think important are so only because of their relation to happiness? The idea that our judgements themselves rest on the recognition of such a relation Price dismisses by describing cases in which consequences are no longer relevant, and in which the inclination to object to ingratitude or injustice remains (135). He also suggests (136) that children make such deontological judgements at an early age, when they cannot plausibly be assumed to have a conception of the long-term consequences of such actions on human happiness.7 Further, when we introspect, we shall find that we—even though we ourselves may be aware of these consequences and see them as significant—disapprove of ingratitude, injustice, and so on in their own right. But what about the second two-level position, according to which we are ‘insensibly influenced’ by our habit of disapproval of certain actions on the ground of their consequences? Here it emerges that Price has offered another hostage to the consequentialist through his recognition of the utilitarian case for deontological principles. Given that God’s only reason for creating us is one of beneficence, and that ‘there is the greatest occasion for [deontological principles] to secure the general welfare, . . . therefore it might antecedently be expected that a good Being would give them to us’ (137). This position seems to ascribe to God ‘pragmatic’ reasons for bringing it about that we believe deontological principles (see Butler 2006: Sermon 12.31n13). These reasons, the consequentialist might claim, could conflict with our epistemic reasons, and this would 7   Here I cannot resist quoting Hudson (1970: 99): ‘This is a very debatable theory. Children frequently lie, or at least romance, without showing any signs of aversion to their own action. On the contrary, they seem to derive positive satisfaction from it, and their parents have to discourage such conduct by impressing upon them the unfortunate consequences of telling tall stories’. Hudson may speak from experience: he dedicates his book on Price to his own children.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

262  Roger Crisp raise the question of how a God committed to veracity could allow us to believe what is false for the sake of the overall good. But perhaps even in the case of God the principle of beneficence can override all other principles. And if that is so, since our believing deontological principles appears to produce so much good and prevent so much harm, the consequentialist might argue that God would ensure that we believed them anyway, thus undercutting any epistemic case for them based on our own judgements.8

5.  Moral Motivation and Moral Worth Consider a case in which some one individual has committed a crime, and two ­people A and B have each been charged. A judge is making a decision on whom if either to punish, and finds a great deal of persuasive evidence that A is guilty, and little evidence against B; so A is punished. In this case, according to Price, the judge does what is practically virtuous, this notion being equivalent to what many would later call subjectively right (see Russell 1970: 10–15). If B is in fact guilty, however, the judge’s action is abstractly (or ‘objectively’) vicious or wrong. Abstract virtue denotes what an agent, ‘if he judged truly, . . . would judge that he ought to do’ (177).9 One might claim that the practically virtuous action might be described as, to some extent, abstractly virtuous, since any agent is required—independently of their beliefs on such a requirement—to do what they sincerely believe to be right. But what Price has in mind by abstract virtue is presumably the action a virtuous and fully informed agent would do in the circumstances. Which notion of virtue is the basis of moral assessment? Price is clear that it is practical virtue (184). Our judge is certainly not to be blamed for the decision she came to; indeed she would have been blameworthy for punishing B rather than A. This seems plausible, though it might perhaps have been wiser for Price to avoid talk of ‘virtue’ at all in connection with what he has in mind by ‘abstract virtue’. The actions he has mind could be described as the ‘correct’ rather than the ‘morally right’ actions, the concept of morality being so closely connected to those of blame and praise. Price also makes it clear that, strictly speaking, morality does not concern itself with actions as external events, as opposed to the intentions behind those events (184–5; see 200–1; 207).10 If someone intends something, [w]hat arises beyond or contrary to his intention, however it may eventually happen, or be derived, by the connexion of natural causes, from his determination, should not be imputed to him. 8   There are analogies here with Sharon Street’s (2006) argument that the case for moral realism is undercut by evolutionary explanations of our moral beliefs. 9   Hudson (1970: 132) suggests that Price was developing Hutcheson’s use of the material–formal distinction. For helpful discussion of the abstract–practical distinction, see Thomas 1977: ch. 4. 10   Price occasionally appears to deny what one might call volitionalism. See e.g. the weight he puts on what I do rather than what I intend in his discussion of veracity (155–6); here I suggest we should understand him, strictly, to be distinguishing my intention now to fulfil my promise from an intention (that is, an intentional attempt) to fulfil my promise when appropriate.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  263 Among other things, Price probably has in mind what was later called by Bernard Williams the problem of ‘moral luck’, and which was stated especially clearly by Adam Smith: Whatever praise or blame can be due to any action, must belong either, first, to the intention or affection of the heart, from which it proceeds; or, secondly, the external action or movement of the body, which this affection gives occasion to; or, lastly, to the good or bad consequences, which actually, and in fact, proceed from it. . . . That the last two of these three circumstances cannot be the foundation of any praise or blame is abundantly evident. . . . The external action or movement of the body is often the same in the most innocent and in the most blameable actions. . . . The consequences which actually, and in fact, happen to proceed from any action are, if possible, still more indifferent either to praise or blame, than even the external movement of the body. As they depend, not upon the agent, but upon fortune, they cannot be the proper foundation for any sentiment, of which his character and conduct are the objects. . . . To the intention or affection of the heart, therefore . . . all praise and blame . . ., which can justly be bestowed upon any action, must ultimately belong. When this maxim is thus proposed, in abstract and general terms, there is nobody who does not agree to it. Its self-evident justice is acknowledged by all the world.  (Smith 1982: 2.3.intro.1–4)

Price, then, in defending volitionalism is accepting Smith’s ‘equitable maxim’. Further, reflection on cases of moral luck show that he is right to do so. Consider the following well-known case, adapted from Thomas Nagel (1979: 25): The reckless drivers. Two drivers, through culpable lack of attention, fail to notice a red traffic light. The unlucky driver kills a pedestrian who has started to cross the road. The lucky driver does not, but only because there is no pedestrian for her to hit. All else is equal.

As Smith implies, in this case the unlucky driver would in fact be blamed a good deal more than the lucky one, and this blame would represent itself as appropriate and deserved. Since in general degree of blame tracks degree of wrongness, it seems that our sentiments incline us to the view that the unlucky driver has committed a significantly greater wrong than the lucky one. And of course we expect that it is only reasonable that the unlucky driver be wracked with guilt, while the lucky one—if she becomes aware of what she has done—will feel if anything only a twinge of guilt, mixed with relief. But our sentiments are at odds with the equitable maxim, and Smith and Price are correct that we should reject the beliefs they imply. We have to revise our everyday moral judgements not only about negligence, but also, for example, about attempts which fail for reasons independent of the agent, or about cases in which decisions are made under uncertainty and turn out one way or the other. Price has seen that morality represents itself to us as set of norms governing the wills of rational agents. And he sees further than Smith, who allows moral status to depend not only on intention, but also on the ‘affection of the heart’.11 Price’s argument here is reminiscent of Kant’s view of moral worth(2002: 4: 398):   Price occasionally slips, allowing at 140 for example that affections can be ‘owed’ (see also 143).

11

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

264  Roger Crisp Wherever the influence of mere natural temper or inclination appears, and a particular ­conduct is known to proceed from hence, we may, it is true, love the person, as we commonly do the inferior creatures when they discover mildness and tractableness of disposition; but no regard to him as a virtuous agent will arise within us.  (191)

Our passions, Price believes, depend on our rational deficiencies (76). If, for example, parents were fully aware of their reasons for acting in certain ways towards their children, there would be no need for parental affection. Price also correctly sees that rational volitionalism leads directly to a strictly rationalist conception of moral worth. An action’s being virtuous and praiseworthy consists in its being done from ‘a regard to goodness and right’ (123). And this regard is a matter of degree, and it is the degree of regard for rectitude which determines the degree of moral worth (200). It may seem as if Price also agrees with Kant that moral worth is exemplified in the overcoming of the passions by reason: [T]he more remote a good is, and the more temptations we have to forego our own interest, the greater is our virtue in maintaining a proper regard to it. In these cases, reason is necessarily more called forth to interpose and decide; our passions concur less with its dictates; and our determinations are more derived from its authority.  (194)

But it becomes clear from other passages that Price sees moral worth in the exercise of reason itself, not in the conquering of the passions by reason: How unreasonable then is it to affirm, that human virtue exceeds that of angels, because of the opposition it encounters; or to regard it as a question of difficulty, whether the excellence of the moral character of the Deity would not be encreased, if he had within him some dispositions contrary to goodness?—Can the very circumstances which argue imperfection in virtue, add to the merit of it?  (205)

God is morally perfect, in that he always wills what is right and is incapable of temptation otherwise, and the closer we come to God the more virtuous we are—though of course we will never achieve perfection itself (221–2). Price seems to offer two quite different pictures of an agent’s virtue. On one, which we might call the actualist position, an agent’s degree of virtue over her life depends on the degree to which she in fact wills what is right during that life. This allows for ‘good’ moral luck, since the more opportunities an agent has for willing rightly, the more ­virtuous she will be. The actualist view is primary in Price’s account. But he also, in discussing the significance of overcoming passions, suggests what we might call the potentialist account: As long as the degree of virtuous attachment is the same, it matters not whether or no any opposition is subdued: The character remains equally worthy. The man, who in a course of goodness, meets with less hindrance than another from his passions and temper, may be equally virtuous, if he has in him that affection to goodness, which would engage him, if he had the same opportunities and trials with another, equally to master the same hindrances. Difficulties and

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  265 inconveniencies attending virtue are the means of shewing to others, who cannot see immediately into our hearts, what is in us, and what our moral temper is.  (203)

The reference here to ‘opportunities’, the use of the subjunctive in ‘would engage’, and the reference to ‘moral temper’ might be taken as the claim that what matters in virtue is not actual willing but the power to will, which may remain unexercised.12 Price should perhaps have made it clearer that there are in fact two ways of morally assessing an agent, and that these are not mutually exclusive. As far as actualist virtue is concerned, the highest degree of this will be achieved by always willing what is right. But if one agent is given more opportunity to perform morally admirable actions than another, she will, from the actualist perspective, have lived a more virtuous life. But this is not, the potentialist will point out, to say that she is a less virtuous person. A maximally virtuous person has the power always to will what is right in every circumstance. Someone who is morally lucky in actualist terms is not thereby a more virtuous person. So the focus of volitionalist morality is twofold: on the willing of actions, and the will itself.

6.  Heads of Virtue: Piety and Unity The seventh chapter of the Review includes Price’s discussion of what he calls the ‘heads’ of virtue—that is, of particular virtues. He lists six: piety, self-love, beneficence, gratitude, veracity, and justice. Interestingly, he excludes courage and other executive virtues. This is perhaps because his moral ideal is ‘saintly’ rather than ‘heroic’.13 On the standard conception of courage, it is required for overcoming recalcitrant fears. Price’s ideal moral agent will presumably feel fear when appropriate; but there will no temptation to give in to such fear. Nor does he speak of humility, but perhaps this he would see as falling under the head of piety. Price would doubtless have held that the pious person, properly understanding the status of God, would see herself as nearly worthless in comparison. Price’s discussion is rich and suggestive, and much could be said about it. I shall restrict myself here to what he says about the priority of piety, and about the unity of virtue, which I shall suggest may be related. Price has very little time for atheists: ‘as long as men continue void of religion and piety, there is great reason to apprehend they are destitute of the genuine principle of virtue’ (144). He goes on to suggest that the virtuous action of such people is probably the result of ‘instinct and natural temper’, and is of little moral value. On the face of it, this seems somewhat unfair. Let us assume that Price is an ‘intellectualist’ about virtue, 12   See 206n. Here Price suggests substituting ‘power’ for ‘virtue’ and continues: ‘The power of a being is the same, whether it meets with any opposition or not’. 13   David Brink suggested to me that Price’s neglect of the executive virtues might be thought to reflect his strong internalism about moral motivation (59, 186, 213). But, as Brink himself noted, Price’s internalism seems to amount to the view that moral belief is sufficient only for pro tanto, not overall, motivation.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

266  Roger Crisp who believes that the virtuous agent must be aware of her reasons for acting (see e.g. Hills 2015: 9). If he were a divine command theorist, who saw all moral reasons resting ultimately on the will of God, his harsh judgement of the atheist would have some justification. But Price explicitly rejects the divine command theory: will ‘implies nothing of a rule, direction, or motive, but is entirely ministerial to these, and supposes them’ (148). Morality, then, is prior to God’s will. So why should we obey God? Because he is God—our ‘universal and almighty Parent, benefactor, and ruler’, and because his will is ‘in necessary union with perfect rectitude’ (ibid.). The second, epistemic reason given here does not justify Price’s harsh judgement of atheists, since all they are doing is ignoring one, admittedly significant, source of moral knowledge or understanding. The first reason is substantive, and explains Price’s claim that those who merely pretend to piety for their own advantage are ‘the most inexcusable and wicked’ (145). But two serious problems with Price’s position remain. First, his claim that atheists are motivated by ‘instinct’ seems unfounded. It is not clear why an atheist should be in any worse a position than a theist to recognize fundamental moral principles, such as that gratitude is appropriate to a benefactor. Second, his weighting of impiety against other vices is disproportional. How can it plausibly be claimed that the deceit of a Tartuffe is more vicious than, say, the actions of a sadistic serial murderer? Price appears to believe that a pious belief in the infinite goodness of God implies such a moral ranking, but in the Review itself he offers little justification for this position. In an appendix, however, Price argues that truth is not itself independent of God; rather, he constitutes it (290). This may be thought to explain Price’s ranking of virtue and vice. If I violate some non-religious moral rule, then I am violating God, and this is a form of impiety. The most wicked are those who violate moral rules to the greatest degree, so it follows analytically that the most wicked will be the most impious, and vice versa. But this does not seem to be Price’s view in the Review itself, where piety and impiety are clearly restricted to the sphere of religious belief and observance. Having discussed the various heads of virtue, Price notes (164–5) that he has covered only ‘the main and leading branches of virtue’ and claims that it may not be possible to ‘comprehend’ all the particular instances of virtue within any number of heads. What Price seems to mean is not that there is some endless list of virtues, but that any list will be not be sufficient to provide guidance in individual cases, because of the huge variety of circumstances in which agents find themselves. This leads Price to suggest that the various heads of virtue all ‘run up to one general idea’ and are all modifications of a single universal rational law (see 252–3). Price here refers to the passage in the Meno in which Socrates tells Meno that he is seeking the single essential property (eidos) through possession of which all the different virtues are virtue. But Price’s position is independent of any such metaphysics. Even a Wittgensteinian might allow that there is a single universal moral principle (‘act rightly’), which can be given further substance when applied in particular spheres (‘act rightly in your relation to God’, ‘act rightly in relation to your own well-being’, and so on).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  267 But Price then claims that ‘[n]o part of [virtue] can be separated from another’.14 In the sense just explained, this is unobjectionable: virtue in relation to God and virtue in relation to one’s own well-being are both virtue. But Price’s claim is significantly more substantive than merely this. He suggests that any degree of ‘partial virtue’ is morally equivalent to any other, and that any agent who neglects any single virtue is ‘an apostate from righteousness and order, as if he neglected them all’. Again, there are weaker and stronger interpretations available. Perhaps Price is merely claiming that the neglect of a single virtue cannot be morally cancelled out by action in accordance with the others. Neglecting a single virtue is blameworthy just as clearly as neglecting all virtues is blameworthy. But what Price goes on to say suggests a much stronger interpretation: To transgress . . . in one point . . . is to throw off effectually our allegiance, and to trample on the whole authority of the law. . . . As long as any evil habit is retained, . . . we continue under the curse of guilt . . . and unqualified for bliss.  (165–6)

On the face of it, Price’s position seems implausibly harsh: a person who lives in accordance with all the virtues except, say, gratitude is morally as bad as a sadistic serial murderer. As Price himself later says: ‘He that is just, kind, meek, and humble, but at the same time an habitual drunkard, can have no pretence to genuine virtue’ (220). But he then goes on to offer a further argument (220–1). The habitual breach of one moral law shows that, if the agent had equally strong temptation to do wrong in other areas, he would do so. As far as virtue is concerned: ‘He then loves her not at all, who loves her not first’. Price’s reliance here on counterfactuals is problematic for at least two related ­reasons. First, it ignores the distinction between action and character, central to volitionalism properly construed. According to Price’s own volitionalism, morality consists in a certain set of rational requirements on an agent’s will. If any agent, properly motivated by concern for the moral law, acts correctly in any case, then her action is maximally morally worthy—that is, she herself is praiseworthy to a certain degree for acting in that way on that occasion. No agent in similar circumstances who acted similarly, from similar motivation, could deserve more praise. Of course, she may deserve blame for not acting rightly on other occasions, which means that her life will have been lived less morally well than that of some other agent who always acted rightly. But equally if her lapses are not universal her life will be more morally worthy than that of some other agent who always acts wrongly, even if the circumstances she finds herself in are similar to those of our partially virtuous agent. This is all a matter of the moral worth of one’s actual actions. One’s moral character, however, can be understood dispositionally, as a matter of what one would do in certain 14   For critical discussion of Price on inseparability, see Irwin 2008: 753. I take it that Price cannot be understood as committing himself to a strong identity claim, according to which, say, benevolence and justice are identical in all significant respects. This is ruled out by, among other things, his clear distinctions between the heads of virtue.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

268  Roger Crisp circumstances. Here, then, we may claim that the character of an agent who would act wrongly if the temptation to do so increased is morally less valuable than the character of an agent who would resist. And Price is right to say that acting wrongly in one sphere might be taken as some evidence that, under greater temptation, one would be more prepared to act wrongly in other spheres. But this raises a second problem for Price: he appears to have raised the bar so high that not only those with partial virtue, but also many, and perhaps all of, those with apparently full virtue—that is, those who always act rightly—are unqualified for bliss. For it is not implausible that there is some degree of temptation—which may of course be to avoid some bad outcome, such as severe suffering—sufficient to motivate any agent in any circumstances to act wrongly. So the only individual with true moral worth would be God himself. Price himself may have recognized this problem, for he goes on to reassure us that he does not mean that we must be perfect: A work of any kind may have all its essentials, and be complete in all its parts, when yet it may be unfinished . . . . Some infirmities will cleave to the best, and it is impossible at present always to hold our passions under such strict discipline, as that they shall never surprize or hurry us into any thing which our hearts shall disapprove. But whenever this happens, it is essential to the character of a good man, that it is his greatest trouble, and that he is put by it upon more future vigilance. His settled prevailing regard in heart and life is to truth, piety, and goodness.  (221–2)

In effect, Price is retreating in the face of the second objection raised above, and allowing that partial virtue—as long as the agent greatly regrets her lapses and tries harder to avoid them in future—is sufficient for virtue and a place in heaven.15 But the reference to the agent’s regret may also provide a clue as to why he was initially tempted to such a harsh judgement of partial virtue. An agent who violates a duty is violating the moral law, going against God’s will, and in a sense violating God himself. But her regret shows that her ‘prevailing’ attitude to God is one of respect. The sin of disrespect to God is not one that Price is ready to see as a matter of degree (presumably above some fairly low threshold). Any significant disrespect at any time damages one’s moral character to the point that only sincere repentance provides hope of saving it, and hence oneself.

References Aristotle 2014. Nicomachean Ethics, ed. and trans. R. Crisp, rev. edn. Cambridge: Cambridge University Press. Butler, J. 2006. The Works of Bishop Butler, ed. D. White. Rochester, NY: University of Rochester Press. Crisp, R. 2006. Reasons and the Good. Oxford: Clarendon Press. Dreier, J. 1993. ‘Structure of Normative Theories’, The Monist 76 (1): 22–40. Findlay, J.N. 1961. Values and Intentions. London: George Allen and Unwin.   On Price’s view of repentance, see Thomas 34; 37–40.

15

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

richard price on virtue  269 Foot, P. 1985. ‘Utilitarianism and the Virtues’, Mind N.S. 94 (374): 196–209. Hills, A. 2015. ‘The Intellectuals and the Virtues’. Ethics 126: 7–36. Hudson, W.D. 1970. Reason and Right: A Critical Examination of Richard Price’s Moral Philosophy. London: Macmillan. Irwin, T. 2008. The Development of Ethics: A Historical and Critical Study. Vol. 2: From Suarez to Roussseau. Oxford: Oxford University Press. Kant, I. 2002. Groundwork for the Metaphysics of Morals, ed. T. Hill and A. Zweig, trans. A. Zweig. Oxford: Oxford University Press. Mackie, J.L. 1982. The Miracle of Theism. Oxford: Oxford University Press. Nagel, T. 1979. ‘Moral Luck’. Repr. in his Mortal Questions, 24–38. Cambridge: Cambridge University Press. Parfit, D. 1984. Reasons and Persons. Oxford: Clarendon Press. Price, R. 1974. A Review of the Principal Questions in Morals, ed. D.D. Raphael. Oxford: Clarendon Press. Russell, B. 1970. ‘The Elements of Ethics’. Repr. in W. Sellars and J. Hospers (ed.), Readings in Ethical Theory, 2nd edn, 3–28. New York: Appleton-Century-Crofts. Sidgwick, H. 1907. The Methods of Ethics. 7th edn. London: Macmillan. Smith, A. 1982. The Theory of Moral Sentiments, ed. D.D. Raphael and A.L. Macfie. Indianapolis: Liberty Fund. Street, S. 2006. ‘A Darwinian Dilemma for Realist Theories of Value’. Philosophical Studies 127 (1): 109–66. Thomas, D.O. 1977. The Honest Mind: The Thought and Work of Richard Price. Oxford: Clarendon Press.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

15 Eudaimonism and Cosmopolitan Concern David O. Brink

In The Methods of Ethics Henry Sidgwick draws more than one contrast between ancient and modern ethical conceptions.1 He contrasts attractive and imperatival ­ethical concepts, concluding that Greek ethics is fundamentally attractive whereas modern ethics is fundamentally imperatival (ME 105–6).2 He glosses this contrast in terms of a contrast between the good and the right; whereas Greek ethics treats the good as the fundamental ethical concept, modern ethical conceptions take deontic or juridical concepts about duty and obligation to be fundamental. But Sidgwick also identifies Greek ethics as egocentric in a way that modern ethics is not (ME 91–2). Ancient ethical conceptions tend to be oriented around the question ‘What sort of life should I live?’ and assume that the answer to that question would be a life that promotes the agent’s own eudaimonia or happiness.3 Sidgwick does not explicitly identify the modern ethical assumption that contrasts with Greek egocentrism. But we might suppose that it is a kind of impartiality and cosmopolitanism that does not filter other-regarding concern through the lens of the agent’s own eudaimonia but recognizes the claims that common humanity imposes on agents. It is this second contrast between ancient eudaimonism and modern cosmopolitanism on which I will focus. Ancient eudaimonism recognizes some familiar moral virtues and other-regarding duties, but these tuistic concerns must contribute to the agent’s own eudaimonia in some way. For instance, in the Republic Plato defends the claim that justice is a genuine 1   It is a pleasure and honour for me to help celebrate Gail Fine’s and Terry Irwin’s careers, especially since they played such a crucial role in forming my philosophical character and nurturing my love for the history of ethics, both ancient and modern. This essay engages issues that have concerned me since graduate school at Cornell and on which I have learned so much from Terry’s own work on related topics. 2   Henry Sidgwick, The Methods of Ethics 7th ed. (London: Macmillan, 1907). 3   In equating eudaimonia and happiness, I assume that the subjective connotations that happiness may have for some modern ears is not an insurmountable obstacle to us understanding the ancient concept of  eudaimonia as happiness. See Richard Kraut, ‘Two Conceptions of Happiness’ Philosophical Review 88  (2), 1979: 167–97. Those who are not persuaded should feel free to understand eudaimonia as the ­personal good, which may or may not be equivalent to happiness.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  271 virtue, contributing to the agent’s own eudaimonia, by arguing that justice and its other-regarding demands contribute to the psychic order of the agent’s soul. As this example illustrates, eudaimonist justifications of other-regarding or tuistic concern have an inside-out structure, grounding other-regarding concern in the agent’s own eudaimonia. An inside-out constraint on tuistic justification seems likely to limit the scope of ethical concern. As we will see, Aristotle equates ethical concern for others with general justice and conceives of justice as predicated on a kind of friendship. While friendship does explain eudaimonistic concern for others, it limits this concern to those with whom one is appropriately related, and this explains ways in which Aristotle’s conception of ethical concern is parochial. Sidgwick is surely right in claiming that ancient ethics is egocentric. However, his ­contrast between ancient and modern ethics is misguided if he believes that modern ethics does not display any of the egocentrism of the ancients. In different ways, Hobbes, Spinoza, Leibniz, Bradley, and Green all carry egocentrism into the modern period. Nonetheless, with the exception of the Stoics, it seems true that only in the modern period do we see the emergence of ethical conceptions that are explicitly universalistic in a fundamental way, insisting that everyone has moral standing and can make valid claims on others. This kind of impartial or cosmopolitan ethical concern receives clear philosophical expression in Kantian and utilitarian conceptions of morality. Even if modern ethics is not consistently cosmopolitan, Sidgwick may well believe that this is the predominant and distinctive tendency of modern ethics, in contrast with ancient ethics. But before we accept this contrast between ancient and modern conceptions, we should ask if the apparent tension between eudaimonism and cosmopolitan concern is genuine or, at least, inevitable. One way to test this perceived tension between eudaimonism and cosmopolitan concern is by contrasting Aristotelian parochialism with Stoic cosmopolitanism. The Stoics share Aristotle’s eudaimonist commitments. But whereas Aristotle’s eudaimonism leads him to pursue an inside-out justification of tuistic concern that is limited in scope, the Stoics believe that we have eudaimonist reason to be concerned with any rational being. Indeed, it is precisely Stoic cosmopolitanism that has led many to view them as anticipating modern moral conceptions and ­representing a bridge between ancient and modern ethics (cf. ME 105). However, I am skeptical that the Stoics succeed in reconciling their cosmopolitan and eudaimonistic commitments. This conclusion might seem to reinforce the difficulty of reconciling eudaimonism and cosmopolitan concern and so may seem to support Sidgwick’s contrast between ancient and modern ethics. But the Aristotelian justification of tuistic concern can take a less parochial form than Aristotle himself recognized. Building on resources contained in Aristotle’s justification of tuistic concern, we can defend a form of ethical concern that is universal in scope. If so, we can reconcile eudaimonism with one form of cosmopolitan concern, perhaps surprisingly, using Aristotle, rather than the Stoics, as our guide. This is obviously an ambitious project with many moving parts, both interpretive and systematic. In the interest of comparing and assessing Aristotelian and Stoic

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

272  David O. Brink commitments about eudaimonism and cosmopolitan concern, I must be selective in my treatment of each, ignoring interesting complexities that would be appropriate to discuss in a longer or narrower treatment. I hope that the interest of the issues that will be my focus compensates for these simplifications.

1. Eudaimonism Virtually all of Greek ethics is eudaimonist in character, treating the agent’s own eudaimonia or happiness as the central or foundational element.4 In particular, eudaimonism implies that the agent’s happiness is the final good—other things are pursued for its sake, and it is not pursued for the sake of anything else. This is a claim, at least in the first instance, about the structure of justification among various kinds of goods or ends.5 As such, eudaimonism implies that the virtues are to be chosen for the sake of happiness. For instance, the eudaimonist can claim that the virtues contribute either instrumentally or constitutively to happiness. Socrates, Plato, and Aristotle are all eudaimonists. In Plato’s early dialogues, Socrates assumes that the virtues would improve and benefit young men if they were to acquire them. This seems reasonably plausible about self-regarding virtues, such as temperance. But it might seem less obvious as applied to other-regarding virtues, such as justice. The Gorgias addresses eudaimonist concerns about justice. There, Callicles accepts the eudaimonist constraint on the virtues, but he contrasts real and conventional justice, arguing that real justice, unlike conventional justice, benefits the agent, rather than others (483b–488b). Socrates responds by arguing that familiar, other-regarding justice is a virtue, because it is necessary for psychic restraint and a well-ordered soul. In the Republic Thrasymachus denies that justice is a virtue because he thinks it benefits others, rather than the agent. Glaucon and Adeimantus develop Thrasymachus’s doubts about justice, conceding that justice is often instrumentally valuable but challenging Socrates to show that justice is valuable for its own sake, as well as its consequences (357b–367e). In Books IV and VIII–IX Socrates defends the claim that justice is a ­virtue to be admired and practiced, by arguing that justice contributes constitutively to the happiness of the agent who is just. Aristotle is also a eudaimonist. In the Nicomachean Ethics he claims that although people have different conceptions of eudaimonia, we all treat eudaimonia as the final good (1095a17–21).6 Eudaimonia is the only 4   Gregory Vlastos, Socrates: Ironist and Moral Philosopher (Ithaca: Cornell University Press, 1991), 203. Also see Terence Irwin, Plato’s Moral Theory (Oxford: Clarendon Press, 1977), esp. 51–4, 249–80 and Plato’s Ethics (Oxford: Clarendon Press, 1995), §§36–7, 142. The main exception to the generalization that Greek ethics is eudaimonist is the Cyrenaics. See Terence Irwin, The Development of Ethics, 3 vols. (Oxford: Clarendon Press, 2007–09), §§28–34. The Cyrenaics resist the appeal to the agent’s overall good, appealing instead to her momentary good (pleasure). If anything, this sort of solipsism of the moment makes an even greater contrast with modern cosmopolitan demands than eudaimonism does. 5   Hence, I understand eudaimonism primarily as a claim about justification, rather than motivation. 6   Three works on ethics are often attributed to Aristotle—the Nicomachean Ethics (NE), Eudemian Ethics (EE), and the Magna Moralia (MM). Here, I focus primarily on NE, but supplement it sometimes with the EE. Unless context specifies otherwise, references to and quotes from Aristotle’s Ethics are to the

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  273 unconditionally complete good. It alone is choiceworthy for its own sake and not for the sake of something else; all other things are choiceworthy for the sake of eudaimonia (1097a27–b6). In particular, the virtues are chosen for the sake of eudaimonia. Eudaimonism is also accepted in most of the Hellenistic schools. In De Finibus Bonorum et Malorum, Cicero’s Epicurean spokesman Torquatus describes their hedonistic conception of eudaimonia and argues that we have reason to cultivate otherregarding virtues such as justice because of their instrumental advantages.7 The Stoics also take the human good to be central. Unlike the Epicureans, they agree with Aristotle in claiming that the human good consists in a form of virtue that involves living in accord with nature. Because the human function consists in rational activity, they conclude that the good involves a life of virtue regulated by the rational part of the agent’s soul (DL vii 84–9; Fin iii 20–1). So, like Aristotle and unlike the Epicureans, the Stoics assign intrinsic, and not merely instrumental, value to the virtues. But unlike Aristotle, who sees virtue as the controlling ingredient in a life that also contains goods of fortune, the Stoics identify happiness with virtue, claiming that virtue is sufficient for happiness and famously treating goods of fortune as preferred indifferents (DL vii 89, 102, 107, 127; Fin iii 11, 20–39).

2.  Aristotle on Justice and Friendship Aristotle does not expressly engage eudaimonist worries about the other-regarding virtues such as justice in the way that Plato does. But, as a eudaimonist, Aristotle owes us an explanation of how familiar tuistic traits, such as courage and justice, are genuine virtues, contributing to the agent’s happiness. The resources for such an explanation lie in his accounts of justice and friendship.8 In the Rhetoric (I 9) Aristotle links virtue with what is fine (kalon). To be fine, something must be both choiceworthy in itself and praiseworthy. He suggests that actions that benefit others are most likely to elicit praise and seem praiseworthy. Virtue is, according to the usual view, a faculty of providing and preserving good things; or a faculty of conferring many great benefits, and benefits of all kinds on all occasions. The parts of virtue are justice, courage, temperance, magnificence, magnanimity, liberality, gentleness, prudence, wisdom.  If virtue is a faculty of beneficence, the highest kinds of it must be those which are most useful to others, and for this reason men honor most the just and the courageous . . . [1366a36–b6] NE. For translation of the NE, I rely on Nicomachean Ethics, trs. T. Irwin, 2d ed. (Indianapolis: Hackett, 1999). References to and quotes from other works of Aristotle are to The Revised Oxford Translation of the Complete Works of Aristotle, 2 vols., ed. J. Barnes (Princeton: Princeton University Press, 1984). 7  Cicero, De Finibus Bonorum et Malorum, trs. H. Rackham (Cambridge: Loeb, 1914) [Fin] i 47–53, ii 78–85 and Kuriai Doxai [KD] in Diogenes Laertius, The Lives of Eminent Philosophers, 2 vols., R.D. Hicks (Cambridge: Loeb, 1925) [DL] 33–6. 8   My understanding of Aristotle’s account of friendship and its role in justifying justice has been influenced by Terence Irwin, Aristotle’s First Principles (Oxford: Clarendon Press, 1988), esp. ch. 18. Here, I draw on ideas in David O. Brink, ‘Self-love and Altruism’ Social Philosophy & Policy 14 (1), 1997: 122–57 and ‘Eudaimonism, Love and Friendship, and Political Community’ Social Philosophy & Policy 16 (1), 1999: 252–89.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

274  David O. Brink Indeed, Aristotle suggests that the virtues are not just other-regarding traits but involve concern for others, rather than oneself. [Fine] also are . . . all actions done for the sake of others, since these less than other actions are done for one’s own sake; and all successes which benefit others and not oneself; and services done to one’s benefactors, for this is just; and good deeds generally, since they are not directed to one’s own profit. [1367a1–5]

But if we identify the virtues with selfless altruism, as Aristotle seems to here, this ­reinforces, rather than resolves, the worry that they won’t satisfy the eudaimonist ­constraint. It is significant, therefore, that in the Ethics Aristotle claims that the fine action benefits not only others but also the agent. Virtues must aim at and secure a common good, common to the agent and others. For instance, general justice, which is complete virtue in relation to another (1129b20–30), aims at the benefit of the ­community and the common good (1129b15–18). These connections among the fine, the common good, and the agent’s own good are clearest in Aristotle’s discussion of true self-love in Ethics IX 8. And when everyone competes to achieve what is fine and strains to do the finest actions, ­everything that is right will be done for the common good, and each person individually will receive the greatest of goods, since that is the character of virtue. [1169a8–12]

Whereas the Rhetoric treats virtues as involving selfless altruism, the Ethics sees them as being good for others and the agent. For Aristotle, as for Socrates, and Plato, the real test case for this claim is justice, because justice is perhaps the most clearly other-regarding virtue (Rhetoric 1366a34–1367b6). Aristotle’s insistence on the connection among justice, the good of a community, and the common good suggests that we look to his justification of friendship for help in justifying justice, because friendship is the virtue appropriate to communities or associations and includes the perfection of justice (NE 1155a22–8, 1159b25–1160a8; EE 1242a19–b1). Moreover, the appeal to friendship seems promising, because Aristotle makes two important claims about some forms of friendship. 1. The best sort of friendship involves concern for the other’s own sake. 2. One’s friend is another or second self (heteros autos). Together, these two claims may permit a eudaimonist justification of tuistic concern. (1) promises to secure other-regarding concern, while (2) promises to show that such concern is in the agent’s own interest. This gives an importance to Aristotle’s discussion of friendship that could explain why he devotes what might otherwise seem to be disproportionate attention (two whole books) to friendship.9 9   Even justice gets only one book (NE V), and it is not uncommon to regard friendship, unlike justice, as a comparatively minor virtue.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  275 Initially, Aristotle distinguishes three main kinds of friendship: (1) friendship for advantage, (2) friendship for pleasure, and (3) complete friendship found between ­virtuous people (VIII 3–8). Both advantage-friendship and pleasure-friendship, Aristotle claims, can involve something less than concern for the other’s own sake (1156a11–13). He insists that virtue-friendship supplies the ‘focal meaning’ of friendship. In calling virtue-friendship the best or most complete kind of friendship, he signals that it is friendship to the fullest extent and that other associations are friendships insofar as they approximate it (1157a25–33; EE 1236a16–b27). Virtue-friendship cannot be widespread inasmuch as virtuous people are rare (1156b25), and this sort of friendship requires a degree of intensity that cannot be maintained on a large scale (1158a11–17, 1171a1–20). Complete friends share similar psychological states, such as aims and goals (1170b16–17) and live together, sharing thought and discussion (1157b8–19, 1159b25–33, 1166a1–12, 1171b30–1172a6). Virtue-friendship ‘reflects the comparative worth of the friends’ (1158b28). The true friend aims at what is good (1162a5, b12, 1165b14–16) and fine (1168b28–1169a12). Because virtue is fine, the friend is concerned with his friend’s virtue. This explains why Aristotle thinks that one cannot remain friends with someone who becomes irredeemably vicious (1165b14–21), that the vicious cannot even love themselves (1166b2–27), and that the person who values and aims to promote his own virtue is the true self-lover (1168a28–1169a12). Aristotle anticipates some of his claims about the justification of virtue-friendship (which begins at IX 4) in VIII 12, where he suggests that we should take parental friendship as our model of friendship. The parent is concerned with the child’s welfare for the child’s own sake. This concern is appropriate on eudaimonist grounds, because the parent can regard the child as ‘another self ’ (1161b19, 28). This is apparently because the child owes its existence and physical and psychological nature in significant part to the parent. This both echoes and helps explain the common view that a parent’s interests are extended by the life of the child. Aristotle suggests similar claims can be made about friendship between siblings. In virtue of living together, siblings causally interact in important ways and share many things in common and so can regard each other as second selves (1161b30–5). The account of familial-friendship brings out what is crucial to justifying the otherregarding concern of virtue-friendship. Aristotle explains the justification of virtuefriendship in terms of proper self-love (1166a1–2, 10, 1166a30–2, 1168b1–1169a12). The excellent person is related to his friend in the same way as he is related to himself, since a friend is another self; and therefore, just as his own being is choiceworthy for him, the friend’s being is choiceworthy for him in the same or a similar way. [1170b6–9; cf. 1168b2–6]

This passage expresses the eudaimonist’s inside-out perspective in which interpersonal relations are modeled on intrapersonal ones. Proper self-love requires a proper conception of the self and of what is beneficial for the self, as we can see when we distinguish it from vulgar self-love.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

276  David O. Brink However, it is this [the virtuous person] more than any other sort of person who seems to be a self-lover. At any rate, he awards himself what is finest and best of all, and gratifies the most controlling part of himself, obeying it in everything. And just as a city and every other composite system seems to be above all its most controlling part, the same is true of a human being; hence someone loves himself most if he likes and gratifies this part. [1168b28–34; cf. 1166a15–23]

Here, Aristotle identifies a person with the controlling part of his soul. Because a human is essentially a psycho-physical compound in which reason can regulate thought and action (1097b24–1098a16, 1102b13–1103a3), the persistence of an individual consists in the continuous employment of his rational faculties to regulate his thought and action. I preserve or extend myself by exercising my practical reason—forming beliefs and desires, deliberating about them, and acting as the result of deliberate choice. If this is what underlies Aristotle’s account of intrapersonal love, we can see how interpersonal love or friendship might be modeled on it. Psychological interaction and influence explain interpersonal, as well as intrapersonal, unity. This sort of unity can be found, presumably to a lesser extent, between two different persons who are friends, because friends share similar psychological states, such as aims and goals (1170b16–17) and live together (1159b25–33, 1166a1–12, 1171b30–1172a6). This suggests the following understanding of Aristotle’s inside-out strategy for justifying tuistic concern. 1. Concern for myself should take the form of concern for the rational part of my soul. 2. Concern for myself involves concern for my future self, which is a self that is psychologically dependent on my present self in a reasons-responsive way. 3. I am related to my friend by psychological influence and interdependence in much the same way my future self is related to my present self. 4. Just as my future self extends my interests, so too do the interests of my friend. 5. Hence, I should regard my friend as a second self and care about her for her own sake just as I care about my own future self for its own sake. This explains why Aristotle thinks we can view a friend as a second self and how he can view the justification of friendship in terms of self-love.10

3.  The Scope of the Common Good If tuistic concern justified on eudaimonist grounds is limited to friendship, the scope of other-regarding concern will be quite limited. However, Aristotle can extend the scope of his eudaimonist justification of interpersonal concern from friends to other members 10   This interpretation of how Aristotle thinks that the relation between lover and beloved makes the beloved a second self to the lover is similar in important ways to the way in which Plato thinks that philosophical eros is the next best thing to immortality for the lover (Symposium 206c–208b). For discussion of the Platonic account, see Richard Kraut, ‘Egoism, Love, and Political Office in Plato’ Philosophical Review 82 (3), 1973: 330–44; Irwin, Plato’s Moral Theory, 241–2, 267–73, and Plato’s Ethics, ch. 18.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  277 of a just political community. It is true that he recognizes that virtue-friendship can’t hold on the scale of a political community that is just (1158a11–12, 1170b29–1171a20; Politics 1262b3–20) and that political communities are associations for mutual advantage and do not involve the best sort of friendship (1160a11–15). These are parts of Aristotle’s critique of Plato’s account of political association in the Republic as reflecting a dilute form of friendship (Politics 1262b14–18). Nonetheless, political communities that are just have to a significant degree the two features that are crucial to the justification of virtue-friendship and familial-friendship: there is commonality of aims and aspirations among members of the political association, and this commonality is produced and sustained by members of the association living together in the right way, in particular by defining their aims and goals consensually (1155a24–8, 1167a25–b8). Insofar as this is true, members of such a political association can see the interests of other members implicated in their own interests and can aim at justice for its own sake, because it promotes the common good, which is presumably the good common to them insofar as they are members of an interdependent political community (1129b15–18). This begins to explain Aristotle’s reasons for his well-known belief that we are essentially political animals (1097b9–12; Politics 1253a2) and that, as a result, the complete good for an individual can only be realized in a political community. Though the scope of the common good extends beyond intimate personal friendship to civic friendship, it seems to be limited to those with whom the agent is psychologically connected. If so, the scope of Aristotelian concern may still seem too narrow. We can see this by considering the reservations expressed by the nineteenth-century British idealist T.H. Green about Aristotle’s conception of the common good. In his Prolegomena to Ethics, Green develops a perfectionist ethical theory that aims to provide a synthesis of the best elements in the ancient and modern traditions.11 Whereas he thought that Aristotle was right to ground an agent’s duties in an account of eudaimonia the principal ingredient of which is a conception of virtue regulated by the common good (§§253, 256, 263, 271, 279), he thought the Greeks had too narrow a conception of various virtues and the common good (§§257, 261–2, 265–6, 270, 279–80). The idea of a society of free and law-abiding persons, each his own master yet each his brother’s keeper, was first definitely formed among the Greeks, and its formation was the condition of all subsequent progress in the direction described; but with them . . . it was limited in its application to select groups of men surrounded by populations of aliens and slaves. In its universality, as capable of application to the whole human race, an attempt has first been made to act upon it in modern Christendom. [§271]

Green’s own conception of the common good is universal or cosmopolitan, which he regards as a distinctively modern idea. Full self-realization occurs only when each respects the claims made by other members of a maximally inclusive community of ends (§§214, 216, 244, 332). Indeed, Green endorses the Humanity Formula of Kant’s   T.H. Green, Prolegomena to Ethics [1883], ed. D. Brink (Oxford: Clarendon Press, 2003).

11

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

278  David O. Brink Categorical Imperative, requiring that one treat all rational agents as ends in themselves and never merely as means, though he believes that this form of cosmopolitanism is a demand of self-realization. One might try to meet Green’s objection to Aristotle by appealing to resources in Aristotelian eudaimonism. There already are significant forms of personal, social, and economic interaction and interdependence between Aristotle’s citizens, on the one hand, and women, slaves, manual laborers, and resident aliens, on the other hand. This interaction provides the basis for including them in a common good. If they are part of the common good, they ought be given a share in citizenship, inasmuch as Aristotle believes that political activity is part of the good of rational animals (Politics V 8–9, esp. 1329a35–8).12 This possible expansion of Aristotle’s conception of the scope of the common good is arguably motivated by Aristotelian ideas and so could be viewed as a friendly amendment to Aristotle. But it falls short of delivering a genuinely universal conception of the common good of the sort Green wants. Predicated as it is on a shared ­history of interaction, the Aristotelian justification of tuistic concern seems destined to be parochial. For there must be someone—the proverbial remotest Mysian—with whom one has no prior history, however indirect, who comes within one’s causal orbit.13 It seems unclear how the Aristotelian could justify concern for the remotest Mysian, for instance, to save her from a runaway olive cart, at least if one could do so at little cost or risk to oneself. Presumably, this is the point of the parable about the Good Samaritan (Luke 10: 29–37). Aristotle may believe that a virtuous person should have concern for the remotest Mysian. In Book IV of the Ethics he mentions the nameless virtue of the person who is beneficent and well disposed to others, as friends are to each other, even though he has no shared history and no prior special concern with those to whom he is beneficent (1126b20–1127a8). We might call this virtue ‘friendliness’, to note its similarities and differences with friendship. Then at the beginning of Book VIII Aristotle suggests that there can be a form of friendship based on common humanity (philanthrōpia) (1155a20–3). Because philanthrōpia is based on common humanity, it would presumably extend to those, such as the remotest Mysian, with whom one has no prior history or concern. Indeed, perhaps philanthrōpia just is the sort of friendliness introduced in Book IV. It is reasonably clear that philanthrōpia would entail cosmopolitan concern with wide scope. What is not clear is how philanthrōpia would satisfy the eudaimonist constraint. Like friendship, philanthrōpia involves good will toward others; but, unlike friendship, philanthrōpia does not require any prior relationship between benefactor and beneficiary. For this reason, it is hard to see how philanthrōpia could be a form of friendship. Insofar as friendship is the key to understanding Aristotle’s eudaimonist rationale for tuistic concern, it is hard to see how philanthrōpia could be justified on   For discussion, see my ‘Eudaimonism, Love and Friendship, and Political Community’, 282–8.   Plato mentions the remotest Mysian in the Theaetetus (209b8). The relevance of the remotest Mysian to the scope of ethical concern is nicely explored in Julia Annas, The Morality of Happiness (Oxford: Clarendon Press, 1993), ch. 12. 12 13

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  279 eudaimonist grounds. As long as the inside-out strategy for justifying tuistic concern appeals to shared history, it looks like it cannot justify cosmopolitan concern. For Aristotle, ethical concern for others is limited in scope. But it appears limited in another way as well. Because Aristotle’s justification of tuistic concern justifies A’s concern for B insofar as B is psychologically connected with A, it appears to support a form of partiality. Presumably, an agent is party to various forms of association, but the bonds in some associations are stronger than others. For instance, I am more closely connected to my spouse and children than I am to neighbors who are comparative strangers. Aristotle’s inside-out strategy of justifying tuistic concern appears to support a form of partiality toward the near and dear. He believes that, all else being equal, it is better to help and worse to harm those to whom one stands in special relations than it is to do these things to others (NE 1160a1–6; Politics 1262a27–30).

4.  Stoic Cosmopolitanism As part of the Greek eudaimonist tradition, the Stoics also face the question about how they can recognize other-regarding traits, such as justice, as genuine virtues. Like Aristotle, they think that justice aims at a common good, but, unlike Aristotle, they think that the scope of the common good should be universal, extending to all members of humanity. The Stoics recognize that most people tend to have tuistic concern that is limited in two ways. First, normal concern is limited in scope to those with whom we are familiar in some way. Second, normal concern is of variable intensity, favoring those who are near and dear to the agent. On their view, we tend to think of our relations to others in terms of a set of concentric circles, with ourselves in the innermost circle, those near and dear to us in intermediate circles, and weaker relations in outer ­circles. The circles represent different levels of ethical concern, with the result that we accept partiality in the form of an interpersonal discount rate of concern. Stobaeus reports the views of Hierocles. Each one of us is as it were entirely encompassed by many circles, some smaller, others larger, the latter enclosing the former on the basis of their different and unequal dispositions relative to each other. The first and foremost circle is the one which a person has drawn as though around a centre, his own mind. . . . Next, the second one further removed from the centre but enclosing the first circle; this contains parents, siblings, wife, and children. The third one has in it uncles and aunts, grandparents, nephews, nieces, and cousins. The next circle includes the other relatives, and this is followed by the circle of local residents, and then the circle of fellow tribesmen, next that of fellow citizens, and then in the same way the circle of people from neigbouring towns, and the circle of fellow countrymen. The outermost and largest circle, which encompasses all the rest, is that of the whole human race.14

We could represent this pattern of conventional ethical concern as in the diagram.   The Hellenistic Philosophers, vol. 1, trs. A.A. Long and D.N. Sedley (Cambridge: Cambridge University Press, 1997), 349. 14

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

280  David O. Brink First Self Second Selves Third Selves Fourth Selves Remotest Mysian

Hierocles implies that conventional ethical concern is universal in scope but variable in weight. In fact, if Aristotle is an exemplar, there is reason to think that normal ethical concern is limited in scope and variable in weight. But Hierocles rejects these limitations in ordinary patterns of concern. He insists on concern for other rational beings with universal scope and appears to reject an interpersonal discount rate. The previous passage continues as follows. Once these [concentric circles] have all been surveyed, it is the task of a well tempered man, in his proper treatment of each group, to draw the circles somehow toward the center, and to keep  zealously transferring those from the enclosing circles into the enclosed ones. . . . It is incumbent on us to respect people from the third circle, as if they were those from the second, and again to respect our other relatives as if they were from the third circle. . . . The right point will be reached if, through our own initiative, we reduce the distance of the relationship with each person.

Hierocles concludes this passage by claiming that the right perspective contracts, rather than collapses, ordinary distinctions. On one reading, the passage revises ordinary concern by reducing the number of circles and decreasing the distance between them. On this reading, the passage advocates a revisionary pattern of concern but nonetheless maintains some form of interpersonal partiality.15 However, the first part of the passage seems to require moving those in outer circles into inner circles. But if we follow this counsel consistently, we must treat those in outer circles the same as those in the innermost circle. If we must bring those in the nth orbit into the nth-1 orbit, then we must eventually bring everyone into the first circle. On this second 15   Irwin suggests that Stoic cosmopolitanism is compatible with interpersonal partiality, The Development of Ethics, §195. For instance, Cicero implies that duties of beneficence should reflect the relationship or ties between benefactor and beneficiary, as well as the importance of the benefit; see Cicero, De Officiis, trs. M. Atkins (Cambridge: Cambridge University Press, 1991) i 45–59. But, as we will see, fundamentally impartial theories, such as utilitarianism, can offer derivative justifications of partiality, which means that partiality, as such, does not require variable weighting at a fundamental level. Moreover, the biggest worry about Stoic cosmopolitanism concerns its claims about the scope, rather than the weight, of interpersonal concern.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  281 r­ eading, the passage rejects both limitations in Aristotelian ethical concern. Though there is some question whether Hierocles is representative of Stoic ethical views here, he articulates a conception of Stoic cosmopolitanism that is closer to some modern conceptions and is worth our attention.16 We can explain what is distinctive of this form of cosmopolitanism by distinguishing the two claims that Aristotle denies and that this strand in Stoicism affirms—that ­ethical concern extends to any rational being and that all else being equal every rational being makes an equal claim on the agent’s moral attention. The first claim is one about scope, whereas the second is one about weight. Aristotle’s position is doubly parochial, because he denies both claims. He thinks that ethical concern should be limited to those with whom one has a history of interaction, and he thinks that the degree or weight of ethical concern that one should have for others should be proportional to the strength of the bonds one has with them. By contrast, the Stoic position that we are examining is doubly cosmopolitan, because it affirms that ethical concern should have both universal scope and equal weight. In one sense, this makes Aristotle’s position purely parochial and the Stoic position purely cosmopolitan. Sidgwick and Green may be right that cosmopolitan concern is a characteristically modern ethical commitment, but it is not a commitment that emerges only in the modern period. The Stoics recognize the appeal of cosmopolitanism and implicitly criticize Aristotle’s more parochial form of tuistic concern. They believe that this ­cosmopolitan commitment can be reconciled with eudaimonism, because they think that eudaimonism directs us to live in accordance with rational nature, wherever that is found, whether in ourselves or others.

5.  The Stoic Reconciliation Unlike Aristotle, who combines eudaimonism with parochial ethical concern, the Stoics combine eudaimonism with cosmopolitan ethical concern. But how can the Stoics reconcile these two commitments? Why should the virtuous person have an equal regard for all humanity? Because they identify human nature with reason, they think that we each have a natural affinity for reason as such, wherever we find it in humanity. As Cato, Cicero’s Stoic spokesman, explains: Hence it follows that mutual attraction between men is also something natural. Consequently, the mere fact someone is a man makes it incumbent on another man not to regard him as alien. . . . We are therefore by nature suited to form unions, societies, and states. [Fin iii 62–8] 16   Indeed, at one point, Cicero represents the Stoics as concluding that the just person will sometimes have to favor the common good in relation to his own good (Fin iii 64). This could suggest an inversion of the normal interpersonal discount rate, in which an agent should count the good of others for more than his own. But in such cases the good of others may count for more collectively, as opposed to individually, than the agent’s own good. If so, Cicero’s claim is compatible with the Stoics rejecting any interpersonal discount rate.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

282  David O. Brink But it is hard to square Stoic cosmopolitanism with their eudaimonism. Shared rationality itself does not seem to provide a eudaimonist reason for caring about other rational agents. The eudaimonist claims that I have reason to promote my own eudaimonia, which depends upon my nature as a rational being. But then what I have reason to care about, in the first instance, is my own rational agency. The fact that other people are also rational agents does not, without further argument, provide me with eudaimonist reason to care about them. The Stoics would need to show that promoting the agency of other rational beings would promote my own. The dialectic between eudaimonism and cosmopolitanism here seems relevantly like the dialectic Sidgwick explores between egoism and utilitarianism. In discussing the proof of utilitarianism, Sidgwick considers a related claim that egoism must give way to utilitarianism (ME 382, 403, 420–1, 497–8). The reasoning begins with something the egoist concedes but argues that consistency requires the utilitarian conclusion. 1. I have reason to promote happiness in my life. 2. But I can promote happiness in other lives. 3. The happiness of others is no less important than my happiness. 4. Hence, I have reason to promote happiness generally. Sidgwick’s considered view is that this argument is problematic. Its cogency turns on the interpretation of (1). That premise is ambiguous between two claims: (a) I have reason to promote my happiness qua happiness, and (b) I have reason to promote my happiness qua my happiness. The argument is valid just in case (1) is read as (1a), not if it is read as (1b). But the egoist will insist on (1b). Appeals to common happiness should not move the egoist, and appeals to common rationality should not move the eudaimonist. Perhaps the Stoics think that Aristotelian friendship should extend beyond one’s own rational agency to any rational agent. Here, the similarity between the virtuous agent and his virtuous friend may seem to provide the ground of friendship. If what makes anyone virtuous is proper control by the rational part of his soul, then perhaps the virtuous person should be friends with any other rational being. It is unclear how much similarity Aristotle does or should require among friends. However that issue is resolved, it is clear that similarity is not sufficient for Aristotelian friendship; friendship must be produced and sustained by living together and sharing thought and discussion (1157b5–12, 18–21). He must, then, perceive his friend's being together [with his own], and he will do this when they live together and share conversation and thought.  For in the case of human beings what seems to count as living together is this sharing of conversation and thought, not sharing the same pasture, as in the case of grazing animals. [1170b10–14]

Even maximal similarity, by itself, would not make for friendship. Even if the remotest Mysian is my doppelgänger, the fact that we have no shared history means that he is not thereby my friend. Our common rationality cannot itself constitute a friendship.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  283

6.  The Common Good and Divine Design However, the Stoics might try to defend their claims about the common good by appeal to their natural teleology, in particular, their assumptions about divine design (DL vii 138–47).17 Cicero gives expression to the Stoic argument for design in De Natura Deorum (ii 17). If you see a large and beautiful house, you could not be induced to think that it was built by mice and polecats, even if you do not see the master of the house. If, then, you were to think that the great ornament of the cosmos, the great variety and beauty of the heavenly bodies, the great power and vastness of the sea and land, were your own house and not that of the immortal gods, would you not seem to be downright crazy?18

How might divine design help with the eudaimonist defense of justice? Of course, if god is benevolent, then he may care equally that all his creatures fare well, and he may want individual agents to act on cosmopolitan concern and may assign rewards and penalties in an afterlife based on the agent’s record of cosmopolitan concern. But, as Plato observes in the Republic, this would be an instrumental defense of ­virtue, albeit a perfect instrumental defense.19 This would not explain how philanthrōpia was a virtue to be chosen for its own sake. But divine design makes possible the idea of theodicy—the idea that the best whole may contain parts that, considered apart from the whole, are imperfect. One might then claim that because individuals are parts of a larger beneficial cosmos what they do for the sake of others makes them better off.20 This requires interpreting the good of each in terms of the whole of which they are parts. This is an interesting argument. Perhaps I could see my sacrifices for others compensated if I knew that our world is the product of divine design, that god has commanded justice, and that justice requires me to sacrifice my interests for the sake of another. Then I might be able to see this sacrifice as compensated, provided I interpret my own good in terms of my role in the best community. But I would have to know each of these things independently. Theodicy may give me reason to believe that I live in the best possible world, but it does not itself tell me whether sacrifices are necessary or, if they are, who should bear them. But then I cannot conclude simply from the 17   See John Cooper, ‘Eudaimonism, the Appeal to Nature, and Moral Duty’ reprinted in his Reason and Emotion (Princeton: Princeton University Press, 1999). 18   Hellenistic Philosophy, 2d ed., trs. B. Inwood and L. Gerson (Indianapolis: Hackett, 1997), 143. 19   Glaucon and Adeimantus ask Socrates to show that justice is the best sort of good, good both for its consequences and in itself. In Book II they concede that justice is instrumentally good and ask Socrates to show that it is good in itself. But their account of the instrumental value of justice shows it to be an imperfect instrumental good, not instrumentally valuable in those circumstances in which one can commit injustice with impunity. There is a sense in which the defense of justice is not complete when Socrates later argues that justice is good for its intrinsic benefits. It is not complete until Book X when Socrates invokes the myth of Er to provide a perfect instrumental defense of justice; the other-worldly rewards of justice and penalties of injustice ensure that justice is always instrumentally best. 20  See Fin iii 64 and Seneca, De Providentia in Seneca, Moral Essays, 3 vols., trs. J. Basore (Cambridge: Loeb, 1928) iii 1–2.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

284  David O. Brink fact that the world is ordered for the best that I have reason to care for you or to make sacrifices for your sake. Moreover, it is not clear that the good of each should be understood in terms of the good of the whole. Perhaps the best whole is one in which I have a very short and miserable life. If so, should we conclude that my life is much better than it appears to be because it has made some contribution to the best whole? Why not conclude instead that the world is best despite or perhaps because of my suffering? But then we couldn’t infer that I fare better for participating in a better whole. Indeed, this argument seems to rest on a confusion of parts and wholes. Consider an intrapersonal analogy to this interpersonal issue. Persons have histories that have parts or stages. It may well be that the best life for Socrates requires some very significant sacrifice early in his life. If so, we can see how Socrates, the person, has reason to make this sacrifice now for the sake of greater later benefits; he is both benefactor and beneficiary, and this explains why there is compensation for the sacrifice. Though Socrates is compensated, his parts are not. Younger Socrates sacrifices and older Socrates benefits. Similarly, the cosmos may be better if some of its parts are worse off than they might otherwise be, but this does not provide compensation to the parts—in this case, particular individuals—who sacrifice for the good of the whole. Of course, even if the argument were valid, it would still rest on the premise of divine design. As the Epicureans note, the Stoics are arguably too selective in their analysis of the evidence. If one is going to count efficient and beneficial processes as evidence of design, then one should count waste, hardship, and evil as flaws in the design.21 As Sidgwick hoped, it would be best to find a reconciliation that did not depend on ­controverted theistic commitments.22

7.  An Aristotelian Reconciliation The Aristotelian justification of tuistic concern seemed promising, as far as it went. The problem was that it didn’t go far enough. Its scope was too parochial. Given our reservations about the Stoic reconciliation of eudaimonism and cosmopolitan concern, it might be worth revisiting the Aristotelian reconciliation to see if there is any way to expand its scope. Recall that the Aristotelian reconciliation pursues the inside-out strategy, justifying concern for another insofar as she stands in the same sort of relations of psychological interaction and interdependence to the agent that the agent’s own future self stands to her present self. This explains why Aristotle regards friends and associates as second and third selves, and we said that it allows him to claim that agents have reason to  Lucretius, On the Nature of the Universe, trs. R. Latham (Harmondsworth: Penguin, 1951) v 156–234.   I allude here, of course, to Sidgwick’s doubts about the possibility of reconciling his own dualism of practical reason between egoism and utilitarianism without resort to a belief in the existence of a God who would reward beneficence and punish indifference to others in an afterlife and his (Sidgwick’s) reluctance to accept any reconciliation that transcended an ‘independent ethical science’ (ME 507–8). 21

22

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  285 be concerned about their fellow citizens. But because the Aristotelian rationale for concern seems to appeal to shared history, it looks like it cannot recognize concern for other rational beings with whom one has no prior history. It must leave concern for the remotest Mysian unjustified. Is that right? With existing friends and associates, there is shared history and previous interaction, providing a backward-looking rationale for concern. By hypothesis, this is absent in my relation to the remotest Mysian. But consider the case where I can save her life by ushering her out of the way of the runaway olive cart at little cost or risk to myself. Though I have no previous connection with her, my assistance would constitute such a connection. For my assistance will enable or facilitate her pursuit of her own projects and plans, and this will make her subsequent actions and mental states dependent in certain ways on my assistance. This provides a forward-looking rationale for concern for the remotest Mysian. It is true that by itself this one connection may establish a comparatively weak connection between her and me, but it might well be sufficient to justify my making small sacrifices on her behalf. Moreover, other things being equal, the greater the assistance I provide to the remotest Mysian, the greater the dependence of her subsequent actions and states on my assistance, and so the greater the share I earn in her happiness. So, the constitutive connections that beneficence establishes explain how I can benefit from helping the remotest Mysian, and the greater the help I provide, the greater the benefit it provides me and the stronger my reason to help. This rationale appeals to the idea of forging a connection with a bigger or more permanent good by contributing to it, and thereby earning a share of that good for oneself.23 Here, we have eudaimonist reason for tuistic concern that does not appeal to shared history or prior relationship. Instead, it appeals to the very same kind of psychological interdependence that was the ground of concern in cases where there is shared history and points out how beneficence itself establishes that sort of connection. If successful, this rationale would allow Aristotle to justify ethical concern with ­universal scope. Indeed, it would explain otherwise puzzling claims that he makes. We saw that Aristotle recognizes a (nameless) virtue of friendliness that involves beneficence in the absence of shared history and affective attachment (1126b2–23) and a form of friendship (philanthrōpia) toward con-specifics (1155a20–3). But it was hard to understand how philanthrōpia could be a form of friendship, inasmuch as it did not require the shared history that is part of friendship in all its other forms, and it was hard to see how friendliness satisfies the inside-out constraint of eudaimonism. But now that we understand how the psychological interaction and interdependence that is part of shared history in friendship makes the interests of friends interdependent, we can see how beneficence itself, even in the absence of shared history, can earn the 23   This is the extension of Aristotle’s conception of the common good that Green endorses, though I think that he (Green) is clearer about the conclusion of the argument than about the premises. For further discussion, see David O. Brink, Perfectionism and the Common Good: Themes in the Philosophy of T.H. Green (Oxford: Clarendon Press, 2003), esp. §§XV–XXIII.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

286  David O. Brink benefactor a share of the beneficiary’s happiness. This allows us to see how Aristotle’s eudaimonistic rationale for tuistic concern can indeed be extended to friendliness and philanthrōpia. This gives us a principled Aristotelian rationale for mixed cosmopolitanism. It also provides a possible rationale for Stoic cosmopolitanism, though only if the Stoics accept the Aristotelian claim that the right sort of psychological interaction and interdependence is the ground of tuistic concern.

8.  Eudaimonism and Derivative Concern for Others Even if Aristotelian eudaimonism can recognize ethical concern with universal scope, it may seem to be an imperfect sort of cosmopolitanism. One concern is that eudaimonism represents other-regarding concern as derivative and instrumental. It seems to justify tuistic concern as a special case of self-love: I have reason to be concerned about others insofar as they do or have the potential to extend my interests. This may seem to imply an inappropriately instrumental and mercenary attitude toward others and toward virtues such as courage and justice. But whereas the Epicureans embrace an instrumental defense of the value of justice (KD 33–6; Fin i 47–53, ii 78–85), Plato, Aristotle, and the Stoics all insist that virtues are choiceworthy for their own sakes. Famously, Glaucon and Adeimantus demand that Socrates show that justice is good in itself and not just for its causal consequences (Rep 357b–358d). And Aristotle agrees, insisting that virtues be choiceworthy for their own sake (NE 1097b2–5). Similarly, the Stoics insist that virtue is intrinsically good (Fin iii 11, 21). Moreover, if eudaimonism requires instrumental concern for one’s friends, this will violate Aristotle’s own requirement that virtuous friends care about one another for the other’s own sake. Even when applied to other forms of friendship or association whose bonds are weaker, eudaimonism may seem to impose an inappropriately colonial or imperialistic perspective on ethical concern.24 It is true that the inside-out strategy that Aristotle (on this interpretation) employs represents a concern for others as derivative. But a derivative concern need not be instrumental or otherwise insufficiently robust. Aristotle’s discussion of complete goods makes room for goods that are derivatively justified that are nonetheless good for their own sakes and not simply instrumental goods (1097a26–b7). Incomplete goods are not chosen for their own sakes; they are chosen only for the sake of something else and are mere instrumental goods. By contrast, complete goods are chosen for their own sakes; they are intrinsic goods. Unconditionally complete goods are chosen for their own sakes and not chosen for the sake of anything else. Eudaimonia is the only unconditionally complete good. This means that merely complete goods are goods in themselves but are also chosen for the sake of eudaimonia, perhaps as parts are chosen for the sakes of the wholes of which they are parts. Here, x is valuable as constituent   See, for example, Jennifer Whiting, ‘Impersonal Friends’ The Monist 74 (1), 1991: 3–29.

24

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  287 of y; it has contributory value and is good in itself.25 This idea that something might be good both in itself and for the sake of the larger whole of which it is a part is not unfamiliar. For instance, my philosophy articles have structures; they defend larger aims by a series of arguments and so have constituent sub-aims. When I work on a particular sub-argument, I want to get that argument right both for its own sake and for or because of its constituent role in my larger argument. Aristotle makes a similar claim about the value of the virtues. The virtues are complete, but not unconditionally complete, goods (1097a35–b7, 1100b8–11, 1176b1–8). They are choiceworthy in themselves as parts of happiness. In making this claim, he makes explicit the sort of assumptions Plato must make about the relationship between justice and eudaimonia in Republic II, where he (Plato) values justice for its own sake and for its constitutive contribution to the agent’s own eudaimonia. If the agent treats the good of another as a complete good, then the fact that the ­other’s own good is choiceworthy for the sake of the agent’s own eudaimonia won’t prevent the agent from caring about the other for her own sake. We can see this more clearly if we consider again the analogy between intrapersonal and interpersonal ­concern. When I undertake a present sacrifice for a future benefit, I do so because the future interests are mine. The on-balance justification of the sacrifice depends on its promoting my overall good. But because my future good is a part of this overall good, concern for my overall good requires, as a constituent part, a concern for my future good. In this way, concern for my future self for its own sake seems compatible with and, indeed, essential to self-love. Here, the justification of concern for my own future self is derivative, but not instrumental. Similarly, if interpersonal psychological relations can extend the agent’s interests, then the good of others can be a part of my overall good, just as my own future good can be. Though the inside-out strategy of justification derives tuistic concern from self-love, it can recognize concern for the other’s own sake.

9.  Mixed Cosmopolitanism A different concern about the Aristotelian reconciliation of eudaimonism and cosmopolitanism is that it is an imperfect kind of cosmopolitanism. For the Aristotelian ­justification may be able to recognize tuistic concern with universal scope, but it seems committed to recognizing concern with variable, rather than equal, weight. Because the agent has reason to be concerned for others insofar as they are (or can be) ­psychologically connected to her, her level of concern should be proportional to the degree of connection. But even if there is some actual or potential connection with anyone who is in one’s power to affect, the degree of connection must be variable. Presumably, it is this variable degree of connection that characterizes different forms   Cf. Clarence Irving Lewis, An Analysis of Knowledge and Valuation (La Salle: Open Court, 1946), ch. 16.

25

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

288  David O. Brink of association. My psychological interaction and interdependence are greater with those in my inner circle than with those in my outer circles. Indeed, this is what distinguishes who is in my inner circle and who is in my outer circles. So, the Aristotelian rationale for tuistic concern seems committed to embracing a discount rate of ­concern of the sort reflected in the model of concentric circles that the Stoics associate with commonsense morality. So even if Aristotle can deliver the sort of universal concern that the Stoics require, he cannot deliver the equal weighting of everyone’s interests that some of them also require. To assess this concern about Aristotelian cosmopolitanism, we should revisit our earlier contrast between the scope and weight of ethical concern (§4 above). There, we saw that Aristotle’s position is doubly parochial because it recognizes ethical concern with limited scope and variable weight. By contrast, the Stoic view articulated by Hierocles is doubly cosmopolitan because it recognizes ethical concern that combines universal scope and equal weight. But the extended or amended Aristotelian view recognizes ethical concern that combines universal scope and variable weight. In this way, it represents a third possibility, intermediate between parochial and cosmopolitan purebreds. We might call it mixed cosmopolitanism. The question is whether mixed cosmopolitanism is cosmopolitan enough. In understanding and assessing different cosmopolitan commitments, we might consider C.D. Broad’s defense of a mixed cosmopolitan view that he calls self-referential altruism. In ‘Self and Others’ Broad discusses the contrast between Sidgwick’s two main methods of ethics, viz. egoism and utilitarianism.26 He contrasts egoism’s limited scope and ­partiality with utilitarianism’s universal scope and egalitarian concern. But he thinks that neither adequately captures the demands of commonsense morality. Though Broad thinks that utilitarianism is right to insist that moral concern should extend to any rational (or sentient) creature that it is within one’s power to help or harm, he thinks that its insistence that everyone matters equally fails to recognize special obligations we have toward those to whom we stand in special relationships. Broad thinks commonsense morality recognizes special concern at a fundamental level, reflected in self-referential altruism. On the other hand, the altruism which common sense approves is always limited in scope. It holds that each of us has specially urgent obligations to benefit certain individuals and groups which stand in certain special relations to himself, e.g., his parents, his children, his fellowcountry-men, etc. And it holds that these special relationships are the ultimate and sufficient ground for these specially urgent claims on one’s beneficence. [280]

This passage is ambiguous. On the one hand, it begins by insisting that special relations limit the scope of one’s duties toward others. On the other hand, it also suggests that special relations affect the special urgency or stringency of those obligations. I think 26   C.D. Broad, ‘Self and Others’ reprinted in Broad’s Critical Essays in Moral Philosophy, ed. D. Cheney (London: George Allen and Unwin, 1971).

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  289 Broad’s considered view is to understand special relations as qualifying the weight, rather than the scope, of moral concern. Elsewhere in this essay, he suggests that any plausible ethical theory must recognize wide scope duties (264). We may debate just how stringent our duties to comparative strangers are, but it seems less controversial that we have non-derivative reason, however weak, to be concerned about the remotest Mysian, such that we would have reason to provide significant aid to her if we could do so at little or no cost to ourselves. On this reading of self-referential altruism, there is a non-derivative impartial demand to be concerned about anyone whom it is within one’s power to benefit but it is also non-derivatively true that the strength or stringency of one’s obligations to benefit others is a function of the nature of the relationship between benefactor and beneficiary.27 This is an ethical conception that has universal scope but variable weight. As such, it is a mixed conception that is cosmopolitan about scope, not weight. It occupies a position intermediate between pure parochialism and pure cosmopolitanism. Broad defends self-referential altruism’s mixed cosmopolitanism against the pure cosmopolitanism of the utilitarians. He is aware that utilitarians can try to provide a derivative justification of partiality. In The Methods of Ethics Sidgwick argues that one can accommodate special concern within a utilitarian framework (ME 432–9). He claims that we have natural affections toward the near and dear, such that focusing our attention on them tends to yield greater utility. Moreover, our knowledge of what others want and need and our causal powers to benefit others are greater in cases involving those near and dear to us and other associates with whom we have regular contact, with the result that as individuals we do better overall by focusing our energies and actions on associates of one kind or another, rather than the world at large. But Broad thinks that this derivative justification of special obligations is inadequate. It is not hard to see why. First, even if I get a utility boost from helping my friends, rather than strangers, the strangers whom I could help have their own friends who get a utility boost from seeing their lot improved. So, it seems that I can often produce more utility by providing a greater benefit to strangers (and their friends) than by providing a lesser benefit to my own friends. Second, intimate knowledge is not necessary to benefit ­others. I may be likely to get a better birthday present for my friend than the remotest Mysian, but it doesn’t take special knowledge about personal tastes to know that strangers benefit from basic nutrition, healthcare, education, and life’s essentials. Third, our causal reach is not limited in the relevant ways. Even if it is costly to benefit people in far-away places, it is often true that there are needy strangers in our midst who 27   This version of self-referential altruism treats special relationships as a thumb in the scales of what would otherwise be a utilitarian reckoning. If we think of special relationships as forms of association, we might claim that self-referential altruism adds an associate-multiplier to an otherwise utilitarian analysis. On this view, an agent is required to perform that action whose value is greatest after the consequences for everyone have been recorded and multiplied by the relevant factor (equal to or greater than one) corresponding to the significance of the association between the agent and potential beneficiaries. This would be a form of consequentialism that combined agent-neutral and agent-relative aspects.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

290  David O. Brink would benefit more from our assistance than our comparatively prosperous friends and associates. For these reasons, the utilitarian justification of special concern seems insufficiently robust. Insofar as we agree with Broad that commonsense morality contains a fundamental commitment to partiality, we have reason to take seriously the sort of mixed cosmopolitan commitment to ethical concern with universal scope and variable weight that the Aristotelian view can embrace. The interpersonal discount rate recognized by Aristotelian cosmopolitanism may be a virtue, rather than a vice.

10.  Eudaimonist Architecture The Aristotelian rationale and self-referential altruism agree in preferring mixed to pure cosmopolitanism. But they disagree about the exact architecture of mixed cosmopolitanism. For Broad, universal concern for others and special concern for associates are equally fundamental moral demands, and there is no other demand that is more fundamental. However, on the Aristotelian rationale that we have been exploring, ­universal concern and special concern are both derived from proper self-love. The Aristotelian can conclude that Broad has underestimated the resources of eudaimonism and egoism. The differences between eudaimonism and theories that build universal concern in as an axiom are bound to look unbridgeable as long as we assume a hedonist conception of happiness, as Sidgwick does and Broad does not dispute. But on Aristotle’s perfectionist conception of happiness, eudaimonism and self-referential altruism can agree on matters of ethical substance. The central claims of self-referential altruism do not need to be represented as ethical axioms, as Broad seems to believe, but can be represented as theorems derived from a eudaimonist axiom. The derivative character of these theorems does not prevent them from being fundamental commitments. According to one of Sidgwick’s contrasts between ancient and modern ethics, they are fundamentally opposed conceptions, because whereas ancient ethics is eudaimonist, modern ethics is impartial or cosmopolitan. Their opposition stems from the tension between the inside-out character of the eudaimonist concern for others and the commitments to universal and equal moral standing of all rational beings characteristic of impartiality and (pure) cosmopolitanism. However, Aristotelian eudaimonism can provide a rationale for tuistic concern with universal scope, including the proverbial remotest Mysian. Though this delivers a cosmopolitan conclusion about the proper scope of ethical concern, it must justify tuistic concern with variable weight, depending on the nature of the relationship between the agent and potential beneficiaries. This sort of mixed cosmopolitanism, combining ethical concern with universal scope and variable weight, falls short of the sort of pure cosmopolitanism that insists on ethical concern with both wide scope and equal weight. For this reason, Aristotelian eudaimonism cannot deliver the sort of cosmopolitanism endorsed by some Stoics and some moderns, including Kant and the utilitarians. But, as Broad’s disagreement with Sidgwick shows, it is unclear if pure cosmopolitanism does justice to our views

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

Eudaimonism and Cosmopolitan Concern  291 about special concern and special obligations. Perhaps mixed cosmopolitanism is cosmopolitanism enough. Whether we prefer egocentric or impartialist architecture may depend on how demanding a conception of the common good we accept. Pure cosmopolitanism may require impartialist architecture, but mixed cosmopolitanism can be agnostic between egocentric and impartialist architecture.28

Bibliography Julia Annas, The Morality of Happiness (Oxford: Clarendon Press, 1993). David O. Brink, ‘Eudaimonism, Love and Friendship, and Political Community’ Social Philosophy & Policy 16 (1), 1999: 252–89. David O. Brink, ‘Self-love and Altruism’ Social Philosophy & Policy 14 (1), 1997: 122–57. David O. Brink, Perfectionism and the Common Good: Themes in the Philosophy of T.H. Green (Oxford: Clarendon Press, 2003). C.D. Broad, ‘Self and Others’ reprinted in Broad’s Critical Essays in Moral Philosophy, ed. D. Cheney (London: George Allen and Unwin, 1971). Cicero, De Finibus Bonorum et Malorum, trs. H. Rackham (Cambridge: Loeb, 1914). John Cooper, ‘Eudaimonism, the Appeal to Nature, and Moral Duty’ reprinted in his Reason and Emotion (Princeton: Princeton University Press, 1999). T.H. Green, Prolegomena to Ethics [1883], ed. D. Brink (Oxford: Clarendon Press, 2003). Hellenistic Philosophy, 2nd ed., trs. B. Inwood and L. Gerson (Indianapolis: Hackett, 1997). Terence Irwin, Plato’s Moral Theory (Oxford: Clarendon Press, 1977). Terence Irwin, Aristotle’s First Principles (Oxford: Clarendon Press, 1988). Terence Irwin, Plato’s Ethics (Oxford: Clarendon Press, 1995). Terence Irwin, The Development of Ethics, 3 vols. (Oxford: Clarendon Press, 2007–09). Richard Kraut, ‘Egoism, Love, and Political Office in Plato’ Philosophical Review 82 (3), 1973: 330–44. Richard Kraut, ‘Two Conceptions of Happiness’ Philosophical Review 88 (2), 1979: 167–97. The Hellenistic Philosophers, vol. 1, trs. A.A. Long and D.N. Sedley (Cambridge: Cambridge University Press, 1997). Kuriai Doxai [KD] in Diogenes Laertius, The Lives of Eminent Philosophers, 2 vols., R.D. Hicks (Cambridge: Loeb, 1925) [DL] 33–6. Clarence Irving Lewis, An Analysis of Knowledge and Valuation (La Salle: Open Court, 1946). 28   Presented at the Ethics and Epistemology in the History of Philosophy conference in honor of Gail Fine and Terence Irwin at Cornell University in September 2013, at the University of St. Andrews, the University of California, San Diego Summer Program for Women in Philosophy, Washington University in St. Louis, California State University, Sacramento, and the University of California, Riverside. I would like to thank members of those audiences for useful discussion and especially Marcia Baron, Anne Margaret Baxley, Richard Boyd, Sarah Broadie, Eric Brown, Leslie Brown, Julia Driver, Gail Fine, John Fischer, Carl Ginet, Paula Gottlieb, Verity Harte, Terence Irwin, Richard Kraut, Charlie Kurth, Susan Sauvé Meyer, Richard Miller, Jozef Müller, Andy Reath, Christopher Shields, and Eric Schwitzgebel. Special thanks to Susan and Christopher for their detailed comments on the penultimate draft.

OUP CORRECTED PROOF – FINAL, 06/08/2018, SPi

292  David O. Brink Lucretius, On the Nature of the Universe, trs. R. Latham (Harmondsworth: Penguin, 1951). Seneca, De Providentia in Seneca, Moral Essays, 3 vols., trs. J. Basore (Cambridge: Loeb, 1928) iii 1–2. Henry Sidgwick, The Methods of Ethics, 7th ed. (London: Macmillan, 1907). Gregory Vlastos, Socrates: Ironist and Moral Philosopher (Ithaca: Cornell university Press, 1991). Jennifer Whiting, ‘Impersonal Friends’ The Monist 74 (1), 1991: 3–29.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Bibliographies of Gail Fine and Terence Irwin Gail Fine Books (1993a), On Ideas: Aristotle’s Criticism of Plato’s Theory of Forms. Oxford: Oxford University Press. (1995), Aristotle: Selections. With T. H. Irwin. Indianapolis: Hackett. (1996a), Aristotle: Introductory Readings. Indianapolis: Hackett (an abridgment of Aristotle: Selections). (2003a), Plato on Knowledge and Forms: Selected Essays. Oxford: Oxford University Press. (2014), The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus. Oxford: Oxford University Press. Edited Books (1999a), Plato, Volume 1: Metaphysics and Epistemology. Oxford: Oxford University Press. (1999b), Plato, Volume 2: Ethics, Politics, Religious and the Soul. Oxford: Oxford University Press. (2008a), The Oxford Handbook of Plato. Oxford: Oxford University Press. Articles (including critical notices) (1977a), ‘Plato on Naming.’ Philosophical Quarterly 27, 289–301. Reprinted in Plato on Knowledge and Forms: Selected Essays, 117–31. (1978), ‘Knowledge and Belief in Republic V.’ Archiv Für Geschichte der Philosophie 60, no. 2, 121–39. Reprinted in Plato on Knowledge and Forms: Selected Essays, 66–84. (1979a), ‘False Belief in the Theaetetus.’ Phronesis 24, 70–80. Reprinted in Plato on Knowledge and Forms: Selected Essays, 213–24. (1979b), ‘Knowledge and Logos in the Theaetetus.’ Philosophical Review 88, 366–97. Reprinted in Plato on Knowledge and Forms: Selected Essays, 225–51. (1980), ‘The One Over Many.’ Philosophical Review 89, no. 2, 197–240. (1981a), ‘Aristotle on Determinism: A Review of Richard Sorabji’s Necessity, Cause and Blame.’ Philosophical Review 90, 561–80. (1981b), ‘Armstrong on Relational and Nonrelational Realism.’ Pacific Philosophical Quarterly 62, 262–71. (1982a), ‘Aristotle and the More Accurate Arguments.’ In Language and Logos, edited by Martha Nussbaum and Malcolm Schofield. Cambridge: Cambridge University Press. (1982b), ‘Owen, Aristotle, and the Third Man.’ Phronesis 27, 13–33. (1983a), ‘Relational Entities.’ Archiv Für Geschichte der Philosophie 65, no. 3, 225–49. Reprinted in Plato on Knowledge and Forms: Selected Essays, 326–49.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

294  Bibliographies of Gail Fine and Terence Irwin (1983b), ‘Plato and Aristotle on Form and Substance.’ Proceedings of the Cambridge Philological Society 209, 23–47. Reprinted in Plato on Knowledge and Forms: Selected Essays, 397–425. (1984a), ‘Separation.’ Oxford Studies in Ancient Philosophy 2, 31–87. Reprinted in Plato on Knowledge and Forms: Selected Essays, 252–300. (1984b), ‘Truth and Necessity in De Interpretatione 9.’ History of Philosophy Quarterly 1, 23–47. (1985), ‘Separation: A Reply to Morrison.’ Oxford Studies in Ancient Philosophy 3, 159–65. (1986), ‘Immanence.’ Oxford Studies in Ancient Philosophy 4, 71–97. Reprinted in Plato on Knowledge and Forms: Selected Essays, 301–25. (1987), ‘Forms as Causes: Plato and Aristotle.’ In Mathematik und Metaphysik bei Aristoteles, edited by A. Graeser. Bern: Haupt Verlag, 69–112. (This is the Proceedings of the Xth Symposium Aristotelicum.) Reprinted in Plato on Knowledge and Forms: Selected Essays, 350–96. (1988a), ‘Owen’s Progress: Logic, Science, and Dialectic.’ Philosophical Review 97, 373–99. (1988b), ‘Plato on Perception: A Reply to Professor Turnbull, “Becoming and Intelligibility.” ’ Oxford Studies in Ancient Philosophy 6, 15–28. (1988c), ‘The Object of Thought Argument: Forms and Thoughts.’ Apeiron 21, 105–45. (1989a), ‘Commentary on Sedley.’ Boston Area Colloquium in Ancient Philosophy 5, 384–98. (1989b), ‘Knowledge and Belief in Republic V–VII.’ In Companions to Ancient Thought 1: Epistemology, edited by S. Everson. Cambridge: Cambridge University Press, 85–115. Reprinted in Plato on Knowledge and Forms: Selected Essays, 85–116. (1992a), ‘Aristotle’s Criticisms of Plato.’ Oxford Studies in Ancient Philosophy 10, 13–41. (1992b), ‘Inquiry in the Meno.’ In The Cambridge Companion to Plato, edited by Richard Kraut. Cambridge: Cambridge University Press, 200–26. Reprinted in Plato on Knowledge and Forms: Selected Essays, 44–65. (1992c), ‘Critical Notice of R.M. Dancy, Two Studies in the Early Academy.’ Canadian Journal of Philosophy 22, no. 3, 393–409. (1993b), ‘Vlastos on Socratic and Platonic Forms.’ Apeiron 26, no. 3, 67–83. (1994), ‘Protagorean Relativisms.’ Boston Area Colloquium in Ancient Philosophy 10, 211–43. Reprinted in Plato on Knowledge and Forms: Selected Essays, 132–59. (1996b), ‘Conflicting Appearances.’ In Form and Argument in Late Plato, edited by C. Gill and M. McCabe, 105–33. Oxford: Oxford University Press. Reprinted in Plato on Knowledge and Forms: Selected Essays, 160–83. (1996c), ‘Nozick’s Socrates.’ Phronesis 41, no. 3, 233–44. (1996d), ‘Scepticism, Existence, and Belief: A Discussion of R.J. Hankinson, The Sceptics.’ Oxford Studies in Ancient Philosophy 14, 273–90. (1998a), ‘Plato’s Refutation of Protagoras in the Theaetetus.’ Apeiron 31, no. 3, 201–34. Reprinted in Plato on Knowledge and Forms: Selected Essays, 184–212. (1998b), ‘Relativism and Self-Refutation: Plato, Protagoras, and Burnyeat.’ In Method in Ancient Greek Philosophy, edited by J. Gentzler. Oxford: Oxford University Press, 137–63. (1999c), ‘Introduction.’ In Plato, Volume 1: Metaphysics and Epistemology, edited by Gail Fine. Oxford: Oxford University Press, 1–35. (1999d), ‘Introduction.’ In Plato, Volume 2: Ethics, Politics, Religious and the Soul, edited by Gail Fine. Oxford: Oxford University Press, 1–33. (2000a), ‘Descartes and Ancient Skepticism: Reheated Cabbage?’ Philosophical Review 109, no. 2, 195–234.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Bibliographies of Gail Fine and Terence Irwin   295 (2000b), ‘Sceptical Dogmata: Outlines of Pyrrhonism I 13.’ Méthexis 13, 81–105. (2003b), ‘Sextus and External World Scepticism.’ Oxford Studies in Ancient Philosophy 24, 341–85. (2003c), ‘Subjectivity, Ancient and Modern: The Cyrenaics, Sextus, and Descartes.’ In Hellenistic and Early Modern Philosophy, edited by J. Miller and B. Inwood. Cambridge: Cambridge University Press, 192–231. (2004a), ‘Knowledge and True Belief in the Meno.’ Oxford Studies in Ancient Philosophy 27, no. 1, 41–81. (2004b), ‘The Subjective Appearance of Cyrenaic Pathê.’ In Socrates: 2400 Hundred Years Since His Death, edited by V. Karasmanis. Delphi: European Cultural Center of Delphi, 383–94. (2007), ‘Enquiry and Discovery: A Discussion of Dominic Scott’s Plato’s Meno.’ Oxford Studies in Ancient Philosophy 32, 331–67. (2008b), ‘Does Socrates Claim to Know That He Knows Nothing?’ Oxford Studies in Ancient Philosophy 35, 49–85. (2008c), ‘Introduction.’ In The Oxford Handbook of Plato, edited by Gail Fine. Oxford: Oxford University Press, 3–35. (2010a), ‘Aristotle on Knowledge.’ Elenchos 14, 121–56. (2010b), ‘Aristotle’s Reply to the Aprêma in the Meno.’ In Aristotle and the Stoics Reading Plato, edited by V. Harte and M. M. McCabe. New York: Bulletin of the Institute of Classical Studies, 45–72. (2010c), ‘Aristotle’s Two Worlds: Knowledge and Belief in Posterior Analytics I.33.’ Proceedings of the Aristotelian Society 110, 323–46. (2010d), ‘Sceptical Inquiry.’ In Definition in Ancient Philosophy, edited by D. Charles. Oxford: Oxford University Press, 493–525. (2010e), ‘Signification, Essence, and Meno’s Paradox: A Reply to David Charles’s “Types of Definition in the Meno.’ ” Phronesis 55, no. 2, 125–52. (2011), ‘Concepts and Inquiry: Sextus and the Epicureans.’ In Episteme, Etc.: Essays in Honour of Jonathan Barnes, edited by B. Morison and K. Ierodiakonou. Oxford: Oxford University Press, 90–114. (2013a), ‘Colloquium 4: Meno’s Paradox and The Sisyphus.’ Boston Area Colloquium of Ancient Philosophy 28, no. 1, 113–46. (2013b), ‘Critical Notice of Katja Vogt, Belief and Truth: A Skeptic Reading of Plato.’ International Journal for the Study of Skepticism 3, no. 2, 131–44. (2016), ‘The “Two Worlds” Theory in the Phaedo.’ British Journal for the History of Philosophy 24, 557–72. Reviews (1977b), Review of Plato: Phaedo, by David Gallop. Philosophical Review 86, no. 1, 101–5. (1983c), Review of A History of Greek Philosophy: Volume Six: Aristotle: An Encounter, by W. K. C. Guthrie and Aristotle the Philosopher, by J. L. Ackrill. With T.H. Irwin. Philosophical Review 92, no. 3, 426–31. (1984c), Review of The Development of Plato’s Metaphysics, by H. Teloh. Philosophical Review 93, no. 1, 143–7. (1991a), Review of The Ascent from Nominalism, by Terry Penner. Noûs 25, no. 1, 126–32.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

296  Bibliographies of Gail Fine and Terence Irwin (1991b), Review of Plato’s Theaetetus, by David Bostock. Philosophical Review 100, no. 4, 687–92. (1992d), Review of The Theaetetus of Plato, by Myles Burnyeat. Philosophical Review 101, no. 4, 830–34. (1996e), Review of Substance and Separation in Aristotle, by Lynne Spellman. Philosophical Review 105, no. 4, 527–30. (2005), Review of The Midwife of Platonism by David Sedley. Philosophical Quarterly 55, no. 221, 662–5. Terence Irwin Books (1977a), Plato’s Moral Theory. Oxford: Oxford University Press. (1979a), Plato’s Gorgias (translation and notes). Oxford: Oxford University Press. (1985a), Aristotle: Nicomachean Ethics. Indianapolis: Hackett. (2nd. edition, 1999). (1988a), Aristotle’s First Principles. Oxford: Oxford University Press. Translated into Italian as I principi primi di Aristotele (tr. A. Giordani), Vita e Pensiero (Milan), 1996. Greek translation (University of Crete Press), in preparation. (1988b), Classical Thought. Oxford: Oxford University Press. Translated into Chinese (1999); Japanese (2000); Greek (2005); Portugese (Brazil, in preparation). (1995a), Plato’s Ethics. Oxford: Oxford University Press. Translated into Spanish as La Etica de Platon (tr. A.I. Stellino), Instituto de Investigaciones Filosoficas (Mexico City), 2000. (1995b), Aristotle: Selections. With Gail Fine. Indianapolis: Hackett. (1996a), Aristotle: Introductory Readings. Indianapolis: Hackett. (An abridgment of Aristotle: Selections) (2000a), Oxford Reader in Classical Philosophy. Oxford: Oxford University Press. (2007a), The Development of Ethics, Volume 1: From Socrates to the Reformation. Oxford: Oxford University Press. (2008a), The Development of Ethics, Volume 2: From Suárez to Rousseau. Oxford: Oxford University Press. (2009a), The Development of Ethics, Volume 3: From Kant to Rawls. Oxford: Oxford University Press. Edited Books (1995c), Classical Philosophy. Oxford Readers. Oxford: Oxford University Press. Articles (including critical notices) (1974a), ‘Recollection and Plato’s Moral Theory.’ Review of Metaphysics 27, 752–72. (1975a), ‘Aristotle on Reason, Desire, and Virtue.’ Journal of Philosophy 72, 567–78. (1977b), ‘Aristotle’s Discovery of Metaphysics.’ Review of Metaphysics 31, 210–29. (1977c), ‘Plato’s Heracleiteanism.’ Philosophical Quarterly 27 (1977): 1–13. (1978), ‘First Principles in Aristotle’s Ethics.’ Midwest Studies in Philosophy 3, 252–72.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Bibliographies of Gail Fine and Terence Irwin   297 (1980a), ‘The Metaphysical and Psychological Basis of Aristotle’s Ethics.’ In Essays on Aristotle’s Ethics, edited by A.O. Rorty, Berkeley: University of California Press, 35–54. (1980b), ‘Reason and Responsibility in Aristotle.’ In Essays on Aristotle’s Ethics, edited by A.O. Rorty. Berkeley: University of California Press, 117–56. (1981a), ‘Aristotle’s Methods of Ethics.’ In Studies in Aristotle, edited by D.J. O’Meara. Washington DC: Catholic University of America Press, 193–223. (1981b), ‘Homonymy in Aristotle.’ Review of Metaphysics 34, 523–44. (1982), ‘Aristotle’s Concept of Signification.’ In Language and Logos (Festschrift for G.E.L. Owen,), edited by Malcolm Schofield and Martha Nussbaum. New York: Cambridge University Press, 241–66. (1983a), ‘Euripides and Socrates.’ Classical Philology 78, no. 3, 183–97. (1984), ‘Morality and Personality: Kant and Green.’ In Self and Nature in Kant’s Philosophy, edited by Allen W. Wood, Ithaca: Cornell University Press, 31–56. (1985b), ‘Aristotle’s Concept of Morality.’ Boston Area Colloquium in Ancient Philosophy 1, no. 1, 115–43. (1985c), ‘Moral Science and Political Theory in Aristotle’ in Crux (festschrift for G.E.M. de Ste Croix) edited by F.D. Havey and P.A. Cartledge (London: Duckworth). Reprinted in History of Political Thought 6, 150–68. (1985d), ‘Permanent Happiness: Aristotle and Solon.’ Oxford Studies in Ancient Philosophy 3, 89–124. (1986a), ‘Aristotelian Actions.’ Phronesis 31, no. 1, 68–89. (1986b), ‘Coercion and Objectivity in Plato’s Dialectic.’ Revue Internationale de Philosophie 40, 49–74. (1986c), ‘Socrates the Epicurean?’ Illinois Classical Studies 11, no. 1, 85–112. Reprinted in Essays on the Philosophy of Socrates, edited by H. Benson, Oxford: Oxford University Press (1992), 198–219. (1986d), ‘Stoic and Aristotelian Conceptions of Happiness.’ In The Norms of Nature: Studies in Hellenistic Ethics, edited by Malcolm Schofield and Gisela Striker. Cambridge: Cambridge University Press, 205–44. (1986e), ‘Socratic Inquiry and Politics’ Ethics 96, no. 2, 400–15. (1987a), ‘Generosity and Property in Aristotle’s “Politics.” ’ Social Philosophy and Policy 4, 37–54. (1987b), ‘Ways to First Principles: Aristotle’s Methods of Discovery.’ Philosophical Topics 15, no. 2, 109–34. (1988c), ‘Disunity in the Aristotelian Virtues.’ Oxford Studies in Ancient Philosophy suppl. vol., 61–78. (1988d), ‘Reply to David L Roochnik’s “Irwin’s Reading of Plato”.’ In Platonic Writings, Platonic Readings, edited by Charles L. Griswold Jr. New York: Routledge, 194–9. (1988e), ‘Socrates and the Tragic Hero.’ In Language and the Tragic Hero. Essays in Honor of Gorden M. Kirkwood, edited by P. Pucci and G.M. Kirkwood. Atlanta: Scholar’s Press, 55–83. (1988f), ‘Some Rational Aspects of Incontinence.’ Southern Journal of Philosophy Suppl. 27, 49–88. (1989a), ‘La Conception stoïcienne et la conception Aristotélicienne du bonheur.’ Revue de Metaphysique et de Morale 94, 535–76. Translation of (1986d). (1989b), ‘Socrates and Athenian Democracy.’ Philosophy and Public Affairs 18, 184–205.

OUP CORRECTED PROOF – FINAL, 06/07/2018, SPi

298  Bibliographies of Gail Fine and Terence Irwin (1989c), ‘Tradition and Reason in the History of Ethics.’ Social Philosophy and Policy 7, no. 1, 45–68. (1990a), ‘Le Caractère Aporétique de la Métaphysique d’Aristote.’ Revue de Me ́taphysique et de Morale 95, no. 2, 221–48. (1990b), ‘The Good of Political Activity.’ In Aristoteles’ Politik, edited by G. Patzig, Göttingen: Vandenhoeck & Ruprecht, 73–98. (1990c), ‘Comments on Christoph Eucken, “Die Aristotelische Demokratiebegriff Und Sein Historisches Umfeld.” ’ In Aristoteles’ Politik, edited by Otfried Höffe. Göttingen: Vandenhoeck & Ruprecht, 292–5. (1990d), ‘The Scope of Deliberation: A Conflict in Aquinas.’ Review of Metaphysics 44, no.1, 21–42. (1990e), ‘Virtue, Praise and Success: Stoic Responses to Aristotle.’ Monist 73, no. 1, 59–79. (1991a), ‘Aristippus against Happiness.’ Monist 74, no.1, 55–82. (1991b), ‘Aristotle’s Defence of Private Property.’ In A Companion to Aristotle’s Politics, edited by David Keyt and Fred D. Miller. Hoboken: Blackwell Publishing, 200–25. Includes a revision of ‘Generosity and Property in Aristotle’s Politics.’ (1991c), ‘Aristotle’s Philosophy of Mind.’ In Psychology (Companions to Ancient Thought: 2), edited by Stephen Everson. Cambridge: Cambridge University Press, 56–83. (1991d), ‘The Structure of Aristotelian Happiness.’ Ethics 101, no. 2, 382–91. (1992a), ‘Eminent Victorians and Greek Ethics.’ In Essays on Henry Sidgwick, edited by Bart Schultz. Cambridge: Cambridge University Press, 279–310. (1992b), ‘Introduction’ in Plato’s Republic, translated by A.D. Lindsay. London: Dent: Everyman’s Library. (1992c), ‘Plato: The Intellectual Background.’ In The Cambridge Companion to Plato, edited by Richard Kraut. Cambridge: Cambridge University Press, 51–89. (1992d), ‘Quelques apories de la science de l’être.’ In Nos Grecs et leurs modernes, edited by B. Cassin. Paris: Editions de Seuil, 417–31. (1992e), ‘Who Discovered the Will?’ Philosophical Perspectives 6, 453–73. (1992f), ‘Socratic Puzzles’ (A Review of G. Vlastos, Socrates: Ironist and Moral Philosopher.) Oxford Studies in Ancient Philosophy X, 241–66. (1992), Articles ‘Aristotle’, ‘Plato,’ and ‘Socrates’ In Encyclopedia of Ethics, edited by Lawrence C. Becker and Charlotte B. Becker. New York: Routledge, 91–8; 1318–26; 1622–8 (pages refer to 2nd edition, 2001). (1993a), ‘Aristotle’ in A Companion to Epistemology, edited by Jonathan Dancy, Ernest Sosa, and Matthias Steup. Chichester: Wiley-Blackwell, 240–3. (1993b), ‘ “Say What You Believe.” ’ Apeiron 26, no. 3, 1–16. (1994a), ‘Critical Notice of Bernard Williams, Shame and Necessity’ Apeiron 27, no. 1, 45–76. (1994b), ‘Happiness, Virtue, and Morality.’ Ethics 105, no. 1, 153–77. (1995d), ‘Plato’s Objections to the Sophists.’ In The Greek World, edited by C. Anton Powell, 568–90. London: Routledge. (1995e), ‘Prudence and Morality in Greek Ethics.’ Ethics 105, no. 2, 284–95. (1996b), ‘Art and Philosophy in Plato’s Dialogues.’ Phronesis 41, no. 3, 335–50. (1996c), ‘Ethics in the Rhetoric and in the Ethics.’ In Essays on Aristotle’s Rhetoric, edited by A.O. Rorty. Berkeley: University of California Press, 142–74. (1996d), ‘Kant’s Criticism of Eudaemonism.’ In Aristotle, Kant, and the Stoics edited by Stephen Engstrom and Jennifer Whiting. Cambridge: Cambridge University Press, 63–101.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Bibliographies of Gail Fine and Terence Irwin   299 (1996e), ‘Stoic Individuals.’ Philosophical Perspectives 10, 459–80. (1996f), ‘The Virtues: Theory and Common Sense in Greek Philosophy.’ In How Should One Live?: Essays on the Virtues, 37–56. Oxford: Oxford University Press. (1996g), ‘A etica como uma ciencia inexata.’ Analytica 1 no. 3, 13–73. (1997a), ‘Aristotelian Substances and Stoic Subjects.’ Revue Internationale de Philosophie 51, no. 201, 397–415. (1997b), ‘Practical Reason Divided: Aquinas and His Successors.’ In Ethics and Practical Reason, edited by Garrett Cullity and Berys Gaut. Oxford: Oxford University Press, 189–214. (1997c), ‘The parts of the soul and the cardinal virtues,’ in Klassiker Auslegen: Platon, Politeia, edited by O. Hoeffe (Akademie-Verlag, Berlin), 89–104. (1998a), ‘Aristotle.’ In Routledge Encyclopedia of Philosophy, edited by E.J. Craig, vol. 1. London: Routledge, 414–35. (1998b), ‘Common Sense and Socratic Method.’ In Method in Ancient Philosophy, edited by Jyl Gentzler. Oxford: Oxford University Press, 29–66. (1998c), ‘Mill and the Classical World.’ In The Cambridge Companion to Mill, edited by John Skorupski. Cambridge: Cambridge University Press, 423–63. (1998d), ‘Socratic Paradox and Stoic Theory.’ In Companions to Ancient Thought Volume 4: Ethics, edited by Stephen Everson. Cambridge: Cambridge University Press, 151–92. (1998e), ‘Stoic Inhumanity.’ In The Emotions in Hellenistic Philosophy, edited by T. EngbergPederson and J. Sihvola. Dordrecht: Kluwer Academic Publishers, 219–42. (1999a), ‘Republic II: Questions about Justice.’ In Plato 2 edited by Gail Fine. Oxford: Oxford University Press. Revision of part of Plato’s Ethics, 164–83. (1999b), ‘Splendid Vices? Augustine For and Against Pagan Virtues.’ Medieval Philosophy and Theology 8, no. 2, 105–27. (1999c), ‘The Theory of Forms.’ In Plato 1: Metaphysics and Epistemology, edited by Gail Fine. Oxford: Oxford University Press, 143–70. (1999d), ‘Some Developments in Aristotle’s Conception of Magnanimity.’ Areté. Revista de Filosofía 11, no. 1, 173–94. (2000b), ‘Ethics as an Inexact Science: Aristotle’s Ambitions for Moral Theory.’ In Moral Particularism, edited by B. Hooker and M.O. Little, 100–29. Oxford: Oxford University Press. (2001d), ‘Vice and Reason.’ Journal of Ethics. 5, no. 1, 73–97. (2003a), ‘Aristotle’s Nicomachean Ethics: A Sort of Political Science.’ In The Classics of Western Philosophy edited by Jorge J. E. Gracia, Gregory M. Reichberg, and Bernard N. Schumacher. Oxford: Wiley-Blackwell, 56–69. (2003b), ‘Augustine’s Criticisms of the Stoic Theory of Passions.’ Faith and Philosophy: 20, no. 4, 430–47. (2003c), ‘Glaucon’s Challenge: Does Aristotle Change His Mind?’ In Plato and Aristotle’s Ethics, edited by Robert Heinaman. Aldershot: Ashgate Publishing, 87–108. (2003d), ‘Morality and Immutability: A Platonic Contribution to Meta-Ethics.’ In Perspectives on Greek Philosophy: S. V. Keeling Memorial Lectures in Ancient Philosophy 1992–2002, edited by R.W. Sharples. Aldershot: Ashgate Publishing, 19–37. (2003e), ‘Stoic and Scholastic Naturalism in Butler.’ In Hellenistic and Early Modern Philosophy, edited by Jon Miller and Brad Inwood. Cambridge: Cambridge University Press, 274–300.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

300  Bibliographies of Gail Fine and Terence Irwin (2003f), ‘Stoic Naturalism and Its Critics.’ In The Cambridge Companion to the Stoics, edited by Brad Inwood. Cambridge: Cambridge University Press, 345–64. (2004a), ‘Cnoty W Filozofi Greckiej.’ In Etyka I Charakter, edited by Jacek Jastal. Krakow: Aureus, 65–84. Translation of (1996f). (2004b), ‘Kantian Autonomy and “Kantian Constructivism”: A Conflict.’ In Agency and Action, edited by J. Hyman and H. Steward. Cambridge: Cambridge University Press, 137–64. (2004c), ‘O Bom E O Certo: Aristoteles E Seus Interpretes Sobre Kalon E Honestum.’ Analytica 8, no. 2, 31–46. (2004d), ‘Was Socrates Against Democracy?’ In Plato’s Euthyphro, Apology, and Crito: Critical Essays, edited by Rachana Kamtekar. Lanham, Md: Rowman & Littlefield Publishers, 127–49. Includes a revision of some of “Socrates and Athenian Democracy.” (2005a), ‘Do Virtues Conflict? Aquinas’s Answer.’ In Virtue Ethics, Old and New, edited by Stephen Mark Gardiner. Ithaca: Cornell University Press, 60–79. (2005b), ‘O Caráter Aporético Da Metafísica de Aristóteles.’ In Sobre a Metafísica de Aristóteles, edited by M. Zingano, 341–70. Sao Paulo: Odysseus. Translation of (1990b). (2005c), ‘Platon et le monisme de la raison pratique.’ In Études Sur La République: Vol. 1: de La Justice, Education, Psychologie Et Politique, edited by Monique Dixsaut. Paris: Librarie Philosophique J. Vrin, 307–26. (2006a), ‘Anachronism and the Concept of Morality.’ In Antike Philosophie Verstehen, edited by M. van Ackeren and J. Müller. Darmstadt: Wissenschaftliche Buchgesellschaft, 149–66. (2006b), ‘Aristotle’s Use of Prudential Concepts.’ In McDowell and His Critics, edited by Cynthia Macdonald and Graham Macdonald. Oxford: Blackwell, 180–97. (2006c), ‘Aquinas, Natural Law, and Aristotelian Eudaimonism.’ In The Blackwell Guide to Aristotle’s Nicomachean Ethics, edited by Richard Kraut. Oxford: Blackwell, 323–41. (2006d), ‘Green’s Criticism of the British Moralists.’ In T. H. Green: Ethics, Metaphysics, and Political Philosophy, edited by M. Dimova-Cookson and W. Mander. Oxford: Oxford University Press, 106–36. (2006e), ‘Socrates and Euthyphro: The Argument and Its Revival.’ In Remembering Socrates: Philosophical Essays, edited by Lindsay Judson and Vassilis Karasmanis. Oxford: Oxford University Press, 58–71. (2006f), ‘Will, Responsibility, and Ignorance: Aristotelian Accounts of Incontinence.’ In Das Problem der Willensschwäche in der Mittelalterlichen Philosophie, edited by T. Hoffmann, J. Müller, and M. Perkams. Leuven: Peeters, 37–56. (2007b), ‘A “Fundamental Misunderstanding”?’ Utilitas 19, no. 1, 78–90. (2007c), ‘Scotus and the Possibility of Moral Motivation.’ In Morality and Self-Interest, edited by Paul Bloomfield. Oxford: Oxford University Press, 159–76. (2008b), ‘Aristotle Reads the Protagoras.’ In Weakness of Will from Plato to the Present, edited by  T. Hoffmann. Studies in Philosophy and the History of Philosophy. Washington, DC: Catholic University of America Press, 22–41. (2008c), ‘The Platonic Corpus.’ In The Oxford Handbook of Plato, edited by Gail Fine. Oxford: Oxford University Press, 63–87. (2008d), ‘The Threefold Cord: Reconciling Strategies in Moral Theory.’ Proceedings of the Aristotelian Society 108, 121–33. (2008e), ‘Introduction.’ In Platão: Carta VII, translated by José Trindade Santos and Juvino Maia Jr. São Paulo: Biblioteca Antiqua PUC. Rio, Edições Loyola, 7–43.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Bibliographies of Gail Fine and Terence Irwin   301 (2009b), ‘The Inside Story of the Seventh Platonic Letter: A Sceptical Introduction.’ Rhizai 6, no. 2, 7–40. A revised version of “Introduction to Plato, Epistle VII” (Platão: Carta VII). (2010a), ‘Morality as Law and Morality in the Laws.’ In Plato’s ‘Laws’: A Critical Guide, edited by Christopher Bobonich, Cambridge: Cambridge University Press, 92–107. (2010b), ‘Sidgwick, Green, and Bradley.’ In The Routledge Companion to Ethics, edited by John Skorupski, London and New York: Routledge, 192–203. (2010c), ‘Stoics, Epicureans, and Aristotelians.’ In A Companion to the Philosophy of Action, edited by Timothy O’Connor and Constantine Sandis. Chichester: Wiley-Blackwell, 447–58. (2010d), ‘The Role of Consent in Aquinas’ Theory of Action.’ In Mind, Method, and Morality: Essays in Honour of Anthony Kenny, edited by John Cottingham and Peter Hacker, 1st ed. Oxford: Oxford University Press, 97–118. (2010e), ‘The Sense and Reference of ‘Kalon’ in Aristotle.’ Classical Philology 105, no. 4, 381–96. (2011a), ‘Beauty and Morality in Aristotle.’ In Aristotle’s Nicomachean Ethics: A Critical Guide, edited by Jon Miller, Cambridge Critical Guides. Cambridge: Cambridge University Press, 239–53. (2011b), ‘Continuity in the History of Autonomy.’ Inquiry 54, no. 5, 442–59. (2011c), ‘From Essence to Form: Metaphysics 1029b1-14 (in that order).’ In Aristotle: Metaphysics and Practical Philosophy: Essays in Honour of Enrico Berti, edited by arlo Natali. Louvain-la-Neuve: Peeters, 95–110. (2011d), ‘Mistakes about Good: Prichard, Carritt, and Aristotle.’ In Underivative Duty: British Moral Philosophers from Sidgwick to Ewing, edited by Thomas Hurka. Oxford: Oxford University Press, 106–25. (2011e), ‘Obligation, Rightness, and Natural Law: Suárez and Some Critics.’ In Interpreting Suárez: Critical Essays, edited by Daniel Schwartz. Cambridge: Cambridge University Press, 142–62. (2012a), ‘Aristotle’s Ethics.’ In Philosophy Bites Back, edited by D. Edmonds and N. Warburton. Oxford: Oxford University Press, 20–8. (2012b), ‘Luther’s Attack on Self-Love: The Failure of Pagan Virtue.’ Journal of Medieval and Early Modern Studies 42, no. 1, 131–55. (2012c), ‘Conceptions of Happiness in the Nicomachean Ethics.’ In The Oxford Handbook of Aristotle, edited by Christopher Shields. Oxford: Oxford University Press, 495–528. (2012d), ‘Virtue and Law in Aquinas.’ In The Oxford Handbook of Medieval Philosophy, edited by John Marenbon. Oxford: Oxford University Press, 605–21. (2012e), ‘Antiochus, Aristotle, and the Stoics on Degrees of Happiness.’ In The Philosophy of Antiochus, edited by David Sedley. Cambridge: Cambridge University Press, 151–72. (2013a), ‘Historical Accuracy in Aquinas’s Commentary on the Ethics.’ In Aquinas and the Nicomachean Ethics, edited by T. Hoffmann, J. Müller, and M. Perkams, Cambridge: Cambridge University Press, 13–32. (2013b), ‘Later Christian Ethics.’ In The Oxford Handbook of the History of Ethics, edited by Roger Crisp, Oxford: Oxford University Press, 184–205. (2013c), ‘Nature, Law, and Natural Law.’ In The Oxford Handbook of the History of Ethics, edited by Roger Crisp. Oxford: Oxford University Press, 206–28. (2013d), ‘Sympathy and the Basis of Morality.’ In A Companion to George Eliot, edited by Amanda Anderson and Harry E. Shaw. Chichester: Wiley-Blackwell, 279–93.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

302  Bibliographies of Gail Fine and Terence Irwin (2013e), ‘Mental Health as Moral Virtue: Some Ancient Arguments.’ In The Oxford Handbook of Philosophy and Psychiatry, edited by K.W.M. Fulford, Martin Davies, Richard Gipps, George Graham, John Sadler, Giovanni Stanghellini, and Tim Thornton. Oxford: Oxford University Press, 37–46. (2014a), ‘Officia and casuistry: some episodes.’ Philosophie Antique 14, 111–28. (2015a), ‘Friendship and Altruism in the Rhetoric.’ In La retorica di Aristotele e la dottrina delle passioni, edited by Bruno Centrone Pisa: Pisa University Press, 87–120. (2015b), ‘Shaftesbury’s Place in the History of Moral Realism.’ Philosophical Studies 172, no. 4, 865–82. (2015c), ‘Nil admirari? Uses and abuses of admiration.’ Proceedings of the Aristotelian Society Supplementary Volume 89 (2015), 223–48. (2016a), ‘Conceptions of Love, Greek and Christian.’ In Love and Christian Ethics: Tradition, Theory, and Society, edited by Frederick V. Simmons and Brian C. Sorrells. Georgetown: Georgetown University Press, 36–50. (2016b), ‘Matter as subject.’ in Materia e causa material in Aristotele e oltre, ed. C. Viano, Rome: Edizioni di Storia e Letteratura, Ch. 1, 3–20. Book Reviews (1974b), Review of Xenophon’s Socrates, by Leo Strauss. Philosophical Review 83, no. 3, 409–13. (1975b), Review of The Pre-Socratics; A Collection of Critical Essays, by Alexander P. D. Mourelatos. American Journal of Philology 96, no. 4, 447–8. (1977d), Review of A History of Greek Philosophy: Volume IV by W.K.C. Guthrie. Philosophical Review 86, no. 2, 254–60. (1977e), Review of Aristotle’s Man, by Stephen R.L. Clark and Aristotle’s Deduction and Induction: Introductory Analysis and Synthesis, by Wayne N. Thompson. Classical World 70, no. 6, 401–2. (1979b), Review of Reason and Human Good in Aristotle, by John M. Cooper. Archiv für Geschichte der Philosophie 61, no. 1, 82–6. (1979c), Review of Substance, Body and Soul by Edwin Hartman. Philosophical Review 88, no. 1, 124–8. (1980c), Review of A Companion to Plato’s Republic, by Nicholas P. White. Philosophical Review 89, no. 4, 640–7. (1980d), Review of The Aristotelian Ethics and Aristotle’s Theory of the Will, by Anthony Kenny. Journal of Philosophy 77, no. 6, 338–54. (1981c), Review of Socrates: Philosophy in Plato’s Early Dialogues, by Gerasimos Xenophon Santas. Journal of Philosophy 78, no. 5, 272–9. (1983b), Review of An Introduction to Plato’s Republic, by J. Annas. Canadian Philosophical Reviews 24, 49–54. (1983c), Review of Book Zeta of Aristotle’s Metaphysics, M. Burnyeat et al. Classical Review 33, no. 2, 234–6. (1983d), Review of Doubt and Dogmatism: Studies in Hellenistic Epistemology, Malcolm Schofield, Myles Burnyeat, and Jonathan Barnes (eds.). Noûs 17, no. 1, 126–34. (1983e), Review of The Greeks on Pleasure, by J.C.B. Gosling and C.C.W. Taylor. Times Literary Supplement, no. 4198, 1003.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

Bibliographies of Gail Fine and Terence Irwin   303 (1988f), Review of The Fragility of Goodness, by Martha Nussbaum. Journal of Philosophy 85, no. 7, 376–83. (1991e), Review of Kant and the Claims of Knowledge, by Paul Guyer. Philosophical Review 100, no. 2, 332–41. (1991f), Review of Socrates: Ironist and Moral Philosopher, by Gregory Vlastos, London Review of Books, 14–15. (1993c), Review of Ethics with Aristotle, by Sarah Broadie. Journal of Philosophy 90, no. 6, 323–29. (1996h), Review of Nature, Justice and Rights in Aristotle’s ‘Politics’, by Fred D. Miller Jr. Times Literary Supplement, no. 4873, 26. (2004e), Review of Aristotle on Meaning and Essence, by D. Charles. International Philosophical Quarterly 44, no. 1, 95–105. (2004f), Review of Individual and Conflict in Greek Ethics, by Nicholas White. Ethics 114, no. 4, 849–58. (2004g), Review of Plato’s Utopia Recast, by Christopher Bobonich. Philosophical Quarterly 54, no. 217, 619–22. (2010f), Review of Essays on the History of Moral Philosophy, by J.B. Schneewind. Notre Dame Philosophical Reviews (http://ndpr.nd.edu/news/essays-on-the-history-of-moral-philosophy/). (2013f), Review of Intelligent Virtue, by J. Annas. Ethics 123, no. 3, 549–56. (2014b), Review of Pursuits of Wisdom by John M. Cooper. Archiv für Geschichte der Philosophie 96, no. 3, 389–95.

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

Index Locorum Aristotle Categories 8a28–34 134 9b15 199n10 9b28 199n10 13a30 199n10 Eudemian Ethics 1144a22–4 213n36 1144a22–b1  213, 213n36 1145a3–6 213n36 1151a35–b3 212 1214a32 212n33 1214b7–14 208 1215a8 212n34 1215b17 212n33 1215b21 212n33 1215b29 199n10 1215b30 212n33 1215b35 212n33 1216a13 212n33 1216a15 212n33 1216a21 212n33 1216a24–5 209 1218a1 138 1218a1–8 139n14 1218a1–15 138 1218a2–3 143 1218a3–4 143 1218a11 143 1220a22 212n34 1224b6 199n10 1226a4 156n18 1226a10ff. 156n18 1226a11–13 210 1226a13 205 1226a29 207 1226b5–9 201 1226b6–9 210 1226b9–18 212n35 1226b7ff. 156n18 1226b8 210 1226b19ff. 156n18 1226b25–6 209 1227a3–5 156n18 1227b5–12 213 1228a2–3 213 1228a4–16 213 1230a27 203 1233a32 203 1233a37 203

1234a24–5 203 1236a16–b27 275 1236b3 204n24 1237a31 204n24 1237a32 204n24 1240a25–8 173 1241b7 199n10 1242a16 212n34 1242a19–b1 274 1245b20 235n4 1247a4 199n10 1247a13 199n10 1247a15 199n10 1247a24 199n10 1247b25 199n10 1247b27 199n10 1247b31 199n10 1247b35 199n10 1248a14 199n10 1248a30 199n10 1248b4 199n10 Magna Moralia 1097a14–16 204 1185b14–87a4 199 1186a33–4 203 1189a12–17  201, 206 1189a25–32 207 1189b9–18 208 1197b21 200n12 1205a19–23 200n12 1212a21  204, 204n24 1289a33–4 207 Metaphysics 980a21–7 114 990b17 145 999a6–10 139n14 1003b22ff. 151n5 1019a1–4 133 1027b33ff. 151n8 1028b21 129n1 1030a3–5 154n14 1036b23–4 159n22 1039a2 145 1072b31 129n1 1078b16 139 1079a13 145 1086a25 139 1086b3–5 138n12 1091a34 129n1 1094a1–2 133

OUP CORRECTED PROOF – FINAL, 06/07/2018, SPi

306  Index Locorum Aristotle (cont.) Movement of Animals 700b22ff. 153 700b31ff. 160 700b32 157 Nicomachean Ethics 1094a1 245 1094a15 212n33 1094a18 245 1094a30 245 1095a2–6 57 1095a17–21 272 1095a20–5 249 1095a26–8 129 1095b3–8 57 1095b20 249 1096aff. 140 1096a11–16 130 1096a14 191n22 1096a17  129n1, 142n15 1096a17–23  130n3, 132 1096a19–35 139n14 1096a22–3 135 1096a23–9 130n3 1096a28  129, 143, 144 1096a29–34 130n3 1096a33–1096b2 139n13 1096a34–1096b5 130n3 1096b7 129n1 1096b7–13 135 1096b8–26 131n3 1096b16–24 135 1096b28–30 153n13 1096b35–1097a13 131n3 1097a26 212n33 1097a26–b7 286 1097a27–b6 273 1097a28 245 1097a35–b7 287 1097b 136 1097b–1098a 255 1097b2–5 286 1097b7 245 1097b9–12 277 1097b24–1098a16 276 1098a–b 258 1098a31 212n34 1098b29 199n10 1099a30–1099b5 246 1100b8–11 287 1102a5–7 236 1102a13 203n22 1102a30ff. 151n6 1102b13–1103a3 276 1102b23–8 236 1102b25 245

1102b30–2 243 1103a14–1105b16 108n19 1103a17 199 1103a24 176 1104a11–19 199 1104b7–9 245 1104b8–12 199 1104b33 199n10 1105a32 237 1105b19 199 1105b21–3  170, 176 1105b26 170 1106a3–7 203 1106b21–3 191n22 1106b26 199n10 1106b31 199n10 1107a9–12 176 1107a12–17 191n22 1107a14 199n10 1108a20 151 1109a1 199 1108a26–8 191n22 1109a30–1109b9 188 1111a30–1 171 1111b12 202n21 1111b26 205 1111b32ff. 156n18 1112a1 202n21 1112a4 165 1112b15 199n10 1112b24–7 190 1113b28 212n34 1115a1 199n10 1115a10–13 191n21 1115b7–24 192 1115b11–21 191n21 1115b21 192n23 1116b2–3 191n21 1117a5 203 1117a29–32 191n21 1170b16–17 276 1118b30–119a20 245 1119a11–20 191n21 1119b15–18 191n21 1120a23–33  191n21, 192 1121a1–4 191n21 1121b1–7 191n21 1125b11 203 1125b31–3 189 1126b2–23 285 1126b20–1127a8 278 1127b14  203, 203n23 1129b15–18  274, 277 1129b20–30 274 1133a1–5 171n9 1134a2 203

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

Index Locorum  307 1134a20 203 1134b12 203 1135b25 203 1136a1 203 1137b35 203 1138a21 203 1138b20 157 1138b25 157 1138b29–30 157 1139a12 152n11 1139a14–16 151n7 1139a17ff. 159 1139a17–b5  149, 151 1139a17–20 151 1139a18 152n11 1139a21  154, 156 1139a21–2 155 1139a24 158 1139a24ff. 163 1139a25   157n21, 161n24 1139a25–6 155 1139a27–31 159 1139a28 158 1139a28ff. 161n24 1139a31 159 1139b2  159, 160 1139b3–4 162 1139b4–5  149, 159 1140b12–18 171n11 1140b12–20 165 1140b21ff. 165 1140b26ff. 164 1142b11 151 1142b30 199n10 1143a8 165 1143a9–10 153 1143a14ff. 165 1144a19 237 1144a34–5 166 1144b27ff. 166 11454a2–3 190n20 1146a27–31 191n22 1148a10 199n10 1149a17 199n10 1152a1ff. 161n24 1152a17 161n24 1154a1–3 190n20 1155a20–3  278, 285 1155a22–8 274 1155a24–8 277 1156a11–13 275 1156b25 275 1157a25–33 275 1157b5–12 282 1157b8–19 275 1157b18–21 282

1158a11–12 277 1158a11–17 275 1158b28 275 1159b25–33  275, 276 1159b25–1160a8 274 1160a1–6 279 1160a11–15 277 1161b19 275 1161b28 275 1161b30–5 275 1162a5 275 1162b12 275 1165b14–16 275 1165b14–21 275 1165b15–17 191n22 1166a1–2 275 1166a1–12  275, 276 1166a15–23 276 1166a30–2 275 1166b2–27 275 1167a25–b8 277 1168a28–1169a12 275 1168b1–1169a12 275 1168b2–6 275 1168b28–34 276 1168b28–1169a12 275 1169a 254 1169a8–12 274 1169a11–15 191n22 1169b3–4 190 1170a4–7 191 1170b6–9 275 1170b10–14 282 1170b29–1171a20 277 1171a1–20 275 1171a24–7 191n22 1171b30–1172a6  275, 276 1172b9–15 113 1174b31–3 237 1176a3–29 237 1176b1–8 287 1178a17–21 160 1179a35 203n22 1179b9 57 1179b10–16 57 1179b30 57 On Dreams 460b3–9 172n13 On the Soul 403a26–7 177 414b19–415a11 139n14 418a6–25 44n6 431a6–9  149, 153 431a7 155n17 431a9ff. 156n19 432b27ff. 153

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

308  Index Locorum Aristotle (cont.) 433a20ff. 159 433a22–6 153 Physics 191b33–4 150n4 260b17–19 133n6 Politics 1253a2 277 1260b27 203n22 1262a27–30 279 1262b3–20 277 1262b14–18 277 1275a34–8 139n14 1279b18–19 78n3 1329a35–8 278 Posterior Analytics 83a24ff. 154n14 88b30–89b9 42 89a3 42 Rhetoric 1366a34–1367b6 274 1366a36–b6 273 1367a1–5 274 1370a17–27 172n12 1370b14–15 177n28 1378a19–21 176 1378a30–4 172 1378b3–4 177 1378b30 177 1379a23–4 175 1379b13–17 179n34 1381a3–7 173 1382a1–15  172, 175 1382a21–2 172 1386a27–9 174 1382b24–6 174 1383a6–7 176 1383a16–19 172 1385a23–4 172 1385b5–6 173 1385b13–16 173 1386b5–7 174 1386b26–1387a5 175 1388a32–5 176 1388a36–8 176 1389b32 176 1415b27–8 175 Sophistical Refutations 178b36 145 183b6–8 21 183b7–8 21n15 Topics 107a3–17 130n3 110b3 20n10 Joseph Butler Sermons 11.6–7 254 12.31n13 261

Cicero On Ends 1.47–53 286 2.78–85 286 3.11  273, 286 3.17 114n15 3.20–1 273 3.21 286 3.20–39 273 3.62–8 281 3.64  281n16, 283n20 5.12 197n3 On the Nature of the Gods 2.17 283 Demosthenes Orations 27.30 26n27 Diogenes Laertius Lives and Opinions of Eminent Philosophers 4.1 129n1 7.84–9 273 7.89 273 7.102 273 7.107 273 7.127 273 7.138–47 283 Epicurus Principal Doctrines 33–6 286 497–8 282 507–8 284n22  T.H. Green Prolegomena to Ethics § 214  277 § 216  277 § 244  277 § 253  277 § 256  277 § 257  277 §§261–2 277 § 263  277 §§ 265–6  277 § 270  277 § 271  277 § 279  277 §§279–80 277 § 332  277  Thomas Hobbes Leviathan Part 1 Ch. VI [4]  179 Part 1 Ch. VI [6]–[9]  180 Part 1 Ch. VI [7]  180 Part 1 Ch. VI [13–18]  180 Part 1 Ch. VI [14]  180 Part 1 Ch. VI [42–4]  180 Part 4 Ch. XLVI [32]  180

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

Index Locorum  309 Immanuel Kant (keyed as ‘volume: page number’ in the Academy Edition; the Critique of Pure Reason is keyed to the original first (A) and second (B) editions) Anthropology from a Pragmatic Point of View 7:119 243 7:201 248n13 7:235 247 7:251  241, 242 7:265 242 7:265–75 244 7:272 248n13 7:291–2 243 7:322 250 7:625 242 Conjectural Beginning of Human History 8:117–18 244 Critique of Practical Reason 5:8–9n 242 5:20–8 235 5:25 246 5:35–6 235 5:37 248n13 5:38 248 5:64–5 247 5:72 241 5:73 247 5:83 245 5:84 239 5:88 248 5:110–13 246 5:111–12  235, 247 5:111–13 245 5:116 241 5:117 248 5:156 248 Critique of Pure Reason A19–20/B33–4 241 A57–8/B82–3 241 Critique of the Power of Judgment 5:257–62 245 5:430 246 Groundwork of the Metaphysics of Morals 4:393  246, 247 4:394 241 4:395–6 249 4:397 239 4:398  243, 263 4:398–9 240n7 4:399  238, 242, 249 4:400 245 4:401 242 4:414  240, 256 4:416–18 248 4:417 246 4:417–19 246 4:418 242 4:425 237

4:428 238 4:432–3 235 4:442–4 235 4:444 241 4:493 240n7 4:454  238, 239n5 Lectures on Anthropology, VA-Busolt 25:1518 242 25:1519 242 Lectures on Anthropology, VA-Friedländer 25:584 242 Lectures on Anthropology, VA-Menschenkunde 25:1109 242 25:1111–12 242 Lectures on Anthropology, VA-Mrongovius 25:1339 242 Lectures on Ethics, VE Collins 27:100–6 239 27:247–50 239 27:249–50 245 27:277 235n4 27:408–12 244 Lectures on Ethics, VE Mrongovius 29:603 245 Lectures on Ethics, VE Vigilantius 27:483 245 27:483–5 239 27:611 235n4 27:624–5 244 27:654 235n4 27:659 235n4 27:680 235n4 27:686–8 244 27:693–5 244 Lectures on Pedagogy 9:486 248n13 Metaphysics of Morals 6:211 241 6:211–13 242 6:212 242 6:213 241 6:215–16 242 6: 373–93  234 6:379 245 6:380 239 6:385–8 246 6:387 248 6:397 243 6:399 242 6:399–403 242 6:402 244 6:405 243 6: 406–7  245 6:407  235n4, 243 6:433 235n4 6:433 235n4

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

310  Index Locorum Immanuel Kant (cont.) 6:448 243 6:455 248n13 6:456–7 244 6:458 244 6:458–60 244 6:462–8 244 6:470 235n4 6:480 242 6:485 245 Notes and Fragments (Cambridge Edition) p. 17   251 Religion within the Boundaries of Mere Reason 6:23–4 245 6:24 248n13 6:27  244, 250 6:32–44 244 6:45n 247 6:58 239 6:60 245 6:93 250  Lucretius On the Nature of the Universe V.156–234 284n21 Menander Samia 524 22 651 22 Plato Apology 30a2–4 57 Charmides 178c 26n28 Crito 49a 28 49d 24 49d1 27 49e3–4 29 49e6–7 28 49e7 29 50a2–3 28 50a2 29 51e 28 Euthydemus 276a–277c 26n29 280a5 28n37 281a1 28n37 281c8 28n37 294d7 26 295a  26, 27 301c–d 27 301d2–4 27n31 302e 26 Gorgias 462b3–465e1 121 464c3–465a2 122

483b–488b 272 491d4–8 98n4 492d–494b 116 499b4 27n33 501c 28n35 Laches 198b 24n19 198c 24 Laws 625a6 99 626a2–7 99 626b5–c2 101 626c3–14 100 626d–e 107 626d8 103 626d8–9 100 626e–627a 98n5 626e–628e 100 626e2–6 100 626e2–627a2 98 626e7 102 626e8–627a1 100 627b2–d4 101 627b7–8 100n7 627c2–628e1 98 627d1–2 100 627e1 98n4 627e–628a 101 628b6–c1 102 628c2–3 102 628c9–e1  99, 102, 103, 105 628d1  103, 104 628d4–5 102 628e 103 628e1  102, 103 628e2–3 107n16 628e2–5 102 628e5 107 629a1–630d1 102 633b5–c7 107 633c8–e3 107 633d–e 98n5 633d5–e2 100n7 633e1 108 643b–644b 104 643b1–644b5 99 643c7–8 104 643d6–644a5 97 643e5 104 643e5–6 97 644a7–8 97 644b6–7  98, 103, 104, 105, 107, 108 644b6–c2 97 644b6–645b3 98 644b9–c2 107 644c4–d3 99n6 644c3–d6 106n15

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

Index Locorum  311 644d1–2 106 644d7–645b1  98, 103, 106, 106n15 644e3 105 644e6 105 645a1 105 645a6–b1 106 645a7 105 645b1–2 106 645b1–3 98 645b1–c4 106 645b1–c6 106n15 645b2 104 645b4–5 106 647a4–6 108n18 647c7–d8 108n18 648c8–e5 108n18 653a–c  104, 106 653a5–c4 99 659d2 105n13 667b5–e4 126 819a5 105n13 966b 6 967e 6 Meno 75d 21n13 75d5–7 18 80e 7 81d 40 82b–85b 41 82e 44 85c  52, 55n14 85c10–d1 64 86b 40 86d6 98n4 93b 55n14 97e–98a 37 97a  40, 52 97b 52 98a  35, 46 Parmenides 130e–131a 132n4 132a10–b2 144 136c3 202n19 143c3 202n19 Phaedo 61b 55 65a–66e 40 72e–77a 50 73c 55n14 74b  50, 51 74c 55n14 75b 50 75e  55n14 76b  6, 46, 50, 51 76d 55n14 78b–80c 41 96b 55n14

Phaedrus 238c3 105n14 248c2–e3 64 265e 55 Philebus 11d4–6 113 11d11–12a5 115n18 12c–d 125n40 17b–d 121n31 20c–22c 115 20d7–10 111 22d1–4 115n19 22d4–8 115n18 23c–27c 124 31a7–10 125 31d4–10 116 31e6–32a5 118n23 34d–36c  117n21, 118 34d10–e1 118n23 35c9–10 118n25 35d3 118n25 35e7–36c2 118n23 44b6–d6 126n45 45de 125n40 46a8–13 123 46d10 123 51b3–d1 119 51d1 123 51d2–3 119n26 51d6–9 120 51e1–2 119 51e7–52a1  119, 123n39 52c1–d1 124 52e6–53b7  119n26, 120n27 53b8–c2 120n26 53c–54d 115n20 53c–55c 115 55e5–56a7 121 56b4–c3 121 58d5 124 62e8–10  122n38, 124 64c7–9 115n18 65a7–b2 115n18 66c4–6 122 67a10–12 115n18 67b1–7 115 67b5 115n16 67b6–7 115n16 Protagoras 317b4 18n3 317b6 26n28 349c7–d2 29 Republic 345b7–9 29 345b 30 350c12–354a11 65 357b–358d 286 357b–367e 272

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

312  Index Locorum Plato (cont.) 358a4 65 358b8 65 358c7 65 367e5–368a7 65 377a12–b3 66 378d7–e3 66 392a12–b6 66 392e 55 409b 55n14 420e 54 422c 54 426d 55n14 427d–434c 58 428c–e 54 428e7–429a3 64 429c7–430b5 66 434e 3 435c–d 57 435c9–d8  58, 59 436–441c 58 439d6–8 118n24 441e–444b 102 442c4–7 61 442d11–443b2 60 444c–445b 102 445b5–7 59 445c4–e3 62 450a5–6 65 473cff. 62 473c9–e4 70 473c8 70 473e5–474a4 70 474c8–487a8 70 474c8–480a13 70 475d4–8 71 4.75e1 71 475e5–476d5 70 476a5 71 476a11 71 476b6–7 70 476b9–10 64 476d–480a 43 476d7–e2 70 476e–480a  33, 35 476e4–480a13 70 477a 45 477a–b  10, 45 477e 42 478b–c 44 478d–e 45 478e 45 479a–c 45 479d3 71 479e2 71 480a 45 484a1–d9 70 484c 51

485a1–487a5 61 485a1–487a8  70, 74 487b1 69 487b1–d5 72 487d1–5  71, 72 488d 55n14 490e4–496e3 72 491a8–b3 64 493e2–494a3 64 498c6–8  69, 72 498d1–502a3 72 499d8 72 499d10  72, 73 500a1 73 500b8–501c9 73 500b8–d2 60 500b8–d9 73 500d11 72 501b–4 73 502a 58 504a–e 57 504a4–b7  59, 60 505a 55n14 505a5–6 60 506c 54 508d 41 509d 46 509d4 47n8 509d5 47n8 509d–511e  34, 35, 36, 46 510b 48 510b7 6 510b–c 48 510c 6 510c–d 48 510d 48 511a 48 511a–b 48 511b 49 511b6 6 520c 54 522c–d 55n14 529b–c 43 531e 6 533b–c 6 533d–534a  35, 36 534b 46 534b–c 6 536a 55n14 540a 83n7 543c4–6  62, 65 546b–c 54 560c6–d1 66 579e 55n14 580b 65 580d–583b 63 580d–588a 63 585b11–c6 63

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

Index Locorum  313 588b–592b  56, 63, 67 588e–589c  63, 67 589c 67 590a 68 590d2–3 68 596a 132n4 598c 55n14 602c–603b 47 602c 47 603a 48 604b1 105n14 Sophist 251e1 202n19 255e1 202n19 Statesman 257c1 202n19 279b5 202n19 Symposium 201b9 23 202a 6 206c–208b 276n10 223d2–6 31 Theaetetus 170e9–171a2 23 171a6–b8 22 171a9 22 187b 37 201c–d  37, 46 201b6 53n13 202c 6 209b8 278n13 Pseudo-Plato Definitions 413c 199n10 Richard Price Review of the Principal Questions of Morals (keyed to the 1974 reproduction of the 1787 3rd ed.) 11 253 15 253 59  253, 265 76 264 79–80 260 105–6  253, 256, 260n5 107–8 257 108 257 108–10 257 110 253 119–24 258 120 258 120–1 258 122 258 123  259, 264 124 259 131–2 260 131–8 257 133 260 134–5 261

135 261 136 261 137 261 137–8 261 140 263n11 143 263n11 144 265 145 266 148 266 153 260 155–6 262n10 159–60 260 164–5 266 165–6 267 177 262 184 262 184–5 262 186 265 191 264 194 264 200 264 200–1 262 203 265 205 264 206n 265 207 262 213 265 220 267 220–1 267 221–2  264, 268 222 254 222–3 254 229 254 249–52 260 252–3 266 257–8 254 258–9 254 259 254 260–1 255 261 255 271 255 285 255 290 266 Seneca On Providence 3.1–2 283n20 Henry Sidgwick The Methods of Ethics (keyed to the page numbers of the 7th ed., 1907) 91–2 270 105 271 105–6 270 220 259 382 282 403 282 414–16 260 420–1 282 432–39 289

OUP CORRECTED PROOF – FINAL, 05/30/2018, SPi

314  Index Locorum Adam Smith The Theory of Moral Sentiments 2.3.Intro.1–4 263 Stobaeus Anthology 4.671, 7ff.  279, 280 

Xenophon Memorabilia 2.7.10  202n20, 206n27 3.9.4  202n20, 206n27 4.5.7  202n20, 206 4.8.11  202n20, 206

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

General Index Academy (Plato)  129–30, 135, 147 actual-sequence mechanism  219–22, 229–30, 232 adherence  34, 37, 39–40, 42 akrasia  12–13, 188, 203 (see also self-control) Alexander of Aphrodisias  4 Allan, D. J.  198–9, 199n.10 altruism  288–90, 289n.27 anger  177, 177n.29 Annas, Julia  105 Anscombe, G. E. M.  126, 163–4, 164n.28 Antisthenes 238 Aquinas, Thomas  16 Aristotle (see Index Locorum) aristocracy 84–6 Arnim, H. von  197 atheism 265–6 Becker, Lawrence  177 Belfiore, Elizabeth  103–4 belief  5, 8, 11, 14–21, 30, 34, 37–9, 42, 48, 52–5, 62–4 (see also doxa, judgments, knowledge, opinion) beneficence  240, 258, 260–2, 265, 273, 285 Benson, Hugh  25, 25nn.23–5 Bobonich, Christopher  104 Brennan, Jason  79–80n.6 Brink, David  227 Brink, K. O.  198 Broad, C. D.  288–91 Brucker, J. J.  234, 234–5n.4 Butler, Joseph  260 categorical imperative  224–5, 239, 241, 246, 277–8 Chomsky, Noam  3 Chrysippus 232 Churchill, Winston  86 Cicero 78 Cohen, G. A.  241 Cohen, Stewart  46 common good  77–84, 87, 93–4, 129, 138, 274, 276–9, 283–4 commonality  143–4, 277 commoners (plebes)  88 compassion  84, 179 competition  88, 92–3, 180, 256 Condorcet, Marquis de  91, 91n.11 confidence  35–6, 46–8, 53–4, 68, 93, 132, 142, 170, 172

consequentialism  224–5, 234, 257, 260–2 constructivism 24 contextualism 35–7 continence 238–9 Cook Wilson, J.  142–3 Cooper, John  175, 178–9, 198, 200, 214 cosmopolitanism  270–2, 277, 279–82, 286–91 mixed cosmopolitanism  288 courage  58–9, 70, 107–8, 179–80 Cynics 238–9 deliberation  154, 190, 197–9, 206–13, 236–7, 243, 249 democracy 77–95 Democritus 202 deontic terms  185–95 deontological theories  234, 259, 261 derivative  286–7, 289–90 DeRose, Keith  36 desire  151–60, 154n.15, 157n.20, 165, 168, 177, 237, 241–2 desiderative nous  150–1, 153 non-rational desire  97–8, 103–6 rational desire  66, 97–9, 103–6, 108, 217, 236–8, 241–2 (see also prohairesis, motivation) determinism  221–3, 227, 229 dialectic  3, 12–14, 20–1, 35, 49, 53, 59, 61, 99, 144, 147, 224 differentia 150 Dirlmeier, Franz  197–200 dispositionalists (new)  228–32 disrespect  244, 268 Divided Line  36, 46–7, 53 divine design  283–4 Donini, Pier Luigi  215 Dover, K. J.  23 doxa  6, 8, 9–10, 33–7, 52, 54, 164, 206, 207, 209 (see also belief) doxastic constraint  25, 25n.24 Dretske, Fred  222, 224 Düring, Ingmar  197 duty  16, 79, 90, 129, 185–6, 189–91, 193–5, 235, 238–40, 244–7, 256, 259 egoism  256–7, 270–1, 288, 290–1 emulation  170, 175–6 Engstrom, Stephen  247n.12 envy  73, 170, 175–6 Epicureans  247, 284, 286

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

316  General index Epicurus  9, 94, 245 eudaimonia (see happiness, eupraxia) eudaimonism  15–16, 235, 245–51, 254, 270–91, 270n.3, 272nn.4–5 eupraxia  158, 163–4 (see also happiness) evil  180, 192, 244, 254, 284 Fara, M.  228 fear  169–70, 172–7, 192 Fine, Gail  1–18, 21, 21n.13, 36, 43, 45, 49, 51–3, 54–5, 143, 169 Fischer, J. M.  218, 221, 223–4 flute-playing 121 Foot, Philippa  260 Forms (Plato)  4, 5, 10, 41–5, 48, 51, 53, 57–61, 63, 69–75, 133, 138, 141–2, 143, 145, 147 of Beauty  51 of the Good  6, 10, 46, 49, 54, 57–60, 64, 129–32, 129n.1, 135, 139, 143–7 of Equal  50–1 of Number  145 of Humanity  141 Frankfurt-style cases  218–19, 221–3, 226, 228–32 Franklin, Christopher  220–1 Frede, Dorothea  105, 176, 176n.25 Frede, Michael  25 freedom  79, 219–20, 228–9 friendliness  177, 177n.27, 278 friendship  274–5, 282 Galton, Francis  91–2 Gauthier, R. A.  185–7, 189–91, 194, 195n.29, 202 Geach, P. T.  136, 136n.9 gerundives 187–90 Gettier, Edmund  38 god  255–6, 260–2, 264–8 Goldman, Alvin  53, 222, 224 good, goods  129–46, 247–8, 270 common good  77–84, 87, 93–4, 129, 138, 274, 276–9, 283–4 (in)complete goods  286 Grant, Alexander  142–3 grief 179–80 guidance control  219 habituation 237 happiness  235–7, 239, 245–7, 251, 253, 260–1, 261n.7, 270–8, 281, 282, 285–90 (see also eudaimonia) Harman, Gilbert  39 harmony  75, 99, 102–8, 104n.9, 238–40, 250 hatred  172–3, 244 hedonism  110–15, 112n.9, 113n.14, 120, 124 hierarchies 82–3

Hobbes, Thomas  170, 172–4, 179–81, 271 Homer 66 human nature  16, 86, 180, 235, 244–5, 250, 281 Hume, David  173–4, 173n.17, 236 imagination  17, 35–6, 46–7, 53, 178–9, 178n.32 indulgence 238 insight  172–4, 174n.20 intellect  35, 36, 47, 150–4, 160–3 (see also practical intellect) intuition  126, 224–6, 228 Inwood, Brad  211 Irwin, Terence  1–2, 12–18, 20, 22, 24–6, 29, 77, 105n.11, 169–70, 181, 192n.24, 198, 216–18, 220, 222, 232, 234–8, 253, 280n.15 Jaeger, Werner  3–4, 198 jelly-beans-in-the-jar experiment  92 Jolif, J. Y.  202 Jouët-Pastré, Emmanuelle  105 judgments  117, 162, 162n.25, 165–6, 176, 178, 224, 261 justice  58–64, 75, 270–1, 287 justified true belief (JTB)  38–9 Kallipolis  54, 84, 86–7, 89–90 kalon 191–2 Kames, Lord (Henry Home)  258 Kamm, Frances  110n.2 Kant, Immanuel (see Index Locorum) Kenny, Anthony  197n.2, 199, 199n.9, 201, 211, 211n.31 kindliness  171, 171n.9 knowledge  33–5, 41, 42, 43n.5, 44n.6, 47n.8, 52–4, 222 Kraut, Richard  29, 185–6, 185n.2, 193 Landemore, Hélène  93n.18 Langton, Rae  228 learning  7–8, 35, 40, 43, 119, 124, 171, 251 Lecky, William  110 Leighton, Stephen R.  171, 171n.11 Lewis, David  228 Lloyd, A. C.  138–41, 138n.11, 144 logismos  103, 105–6 logos  5–6, 28, 134, 156–8 love  22–3, 45, 60–1, 64, 190, 242–4, 276 MacIntyre, Alasdair  240 McKenna, Michael  223–4, 226–8 mechanism-based model  222–6, 228 Menander 22 Meno’s Paradox  8–9 Millgram, Elijah  126 Mingay, Jean  211 monism 261 moral luck  263

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

General Index  317 moral reason  242 moral responsibility  216–32, 232 moral worth  262–5 morality  194–5, 195n.28, 240, 242–6, 244n.11, 257, 262, 271 motivation  61, 75, 118–24, 180, 240–3, 259, 267 moral motivation  262–5 nation-states 82–3 naturalism  15–16, 170, 235 Nelkin, Dana  227–8 Nietzsche, Friedrich  94 Nozick, Robert  39, 222–6, 232 opinion  22–3, 27, 33–7, 46–7 (see also belief) Owen, G. E. L.  4 pain  103, 165–6, 237 paradox of inquiry  7 Parfit, Derek  260 parochialism  271, 289 Pascal, Blaise  255 Peramatzis, Michail  133, 133n.5 perception  5, 13, 50, 119–20, 122–3, 149–52, 151n.8 phantasia (Aristotle)  178 philanthrōpia  278, 283, 285–6 pity  173, 179 Plato (see Index Locorum) pleasure  103, 110–27, 122n.38, 125n.40, 165–6, 176, 176n.25, 237 Polybius 78 practical intellect  152–7, 157n.21, 159, 161 practical reason  11, 126, 155, 158, 160–1, 173, 219–20, 235–6, 242–3, 246–9, 251 practical truth  149–68 practical wisdom  161, 164–8, 235 (see also prudence) preferential choices  150, 154, 156, 159–60 Price, Richard (see Index Locorum) Principle of Alternative Possibilities  218, 222 priority  133–5, 139, 141–6 priority in nature  133–4, 141, 143–4 prohairesis  201–12, 202nn.18 and 21, 211n.32 prudence  201, 204, 213, 235, 237, 244, 248 puppets (Plato)  97–103 Railton, Peter  111 rationalism/anti-rationalism  236–8, 257, 264 Ravizza, Mark  223 Rawls, John  14, 89–90 reason (see intellect, logismos) reasons-responsiveness  222, 227–8, 231–2 recollection (Plato)  34, 40, 49–50 Riesbeck, David  104n.9

Ross, W. D.  189–90, 253, 260 Rousseau, Jean-Jacques  251 Ryle, Gilbert  164–6, 165n.30 scepticism  2, 5, 11 Schleiermacher, F.  197 Schofield, Malcolm  105, 105n.12 Scott, Dominic  20–1 self-control  97–109 (see also temperance) self-determination 241–2 sequential-priority 141 Sextus Empiricus  12 Sidgwick, Henry (see Index Lorocum) Smith, Adam  263 Smith, Michael  233 soul  7–8, 14, 37, 40–3, 47–8, 51–2, 58–60, 151, 243 Stoics  247, 249, 271–3, 279–80n.15, 281–4, 281n.16, 286, 288 supererogation 257–9 Surowiecki, James  91–3 Sylburg, Friedrich  210–11, 210n.30 sympathy  173–4, 243 Taylor, C. C. W.  177n.26 temperance  58–9, 73, 103–4, 108 (see also self-control) theodicy 283 Theophrastus 198 third man argument  4, 145–6 third way thinking  155–7, 159, 163, 165–8 Treynor, Jack  92 truth  41–6, 49, 52, 55, 151–2, 151n.7, 154, 157–64, 162n.26, 167, 261 (see also practical truth) truth-in-action 163 tuistic concern  271, 276, 279, 287–8, 290 two-component theory  150, 150n.2, 160–3 Two-Worlds Theory (TW)  9–11, 45, 51–2, 54–5 universalizability test  225 univocity  135–7, 142–4 utilitarianism  257, 261, 271, 282, 288–90 Utopia 86–7 value  261 (see also good) Vihevelin, Kadri  228 virtue abstract virtue  262 moral virtue  243–4 virtue theories  234–40, 243–9, 251, 253–68, 286–7 virtuous agents  259 (see also practical wisdom) Vlastos, Gregory  13, 24–5, 29–30 volitionalism  263–5, 267 voluntarism 170

OUP CORRECTED PROOF – FINAL, 06/04/2018, SPi

318  General index Waldron, Jeremy  79–80n.6 Walzer, R. R.  198, 211 weakness of will (see akrasia) White, Nicholas  185–6, 193 Wilburn, Joshua  105

Williams, Bernard  37–8, 174, 174n.19, 263 Wittgenstein, Ludwig  266 Woolf, Raphael  211 Xenophon 206