Epicurus on the Self 1138633852, 9781138633858

Citation preview

Epicurus on the Self

Epicurus on the Self reconstructs a part of Epicurean ethics that only survives on the fragmentary papyrus rolls excavated from an ancient library in Herculaneum, On Nature XXV. The aim of this book is to contribute to a deeper understanding of Epicurus’ moral psychology, ethics, and of its robust epistemological framework. The book also explores how the notion of the self emerges in Epicurus’ struggle to express the individual perspective of oneself in the process of one’s holistic self-reflection as an individual psychophysical being. Attila Németh is a postdoctoral fellow in the Department of Philosophy at Eötvös Loránd University, Hungary.

Issues in Ancient Philosophy Series editor: George Boys-Stones, Durham University, UK

Routledge’s Issues in Ancient Philosophy exists to bring fresh light to the central themes of ancient philosophy through original studies which focus especially on texts and authors which lie outside the central ‘canon’. Contributions to the series are characterised by rigorous scholarship presented in an accessible manner; they are designed to be essential and invigorating reading for all advanced students in the field of ancient philosophy. Flow and Flux in Plato’s Philosophy Andrew J. Mason Forms, Souls, and Embryos James Wilberding Forthcoming titles: The Hieroglyphics of Horapollo Nilous Mark Wildish Body and Mind in Ancient Thought Peter N. Singer Taurus of Beirut and the Other Side of Middle Platonism Federico M. Petrucci

Epicurus on the Self

Attila Németh

First published 2017 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2017 Attila Németh The right of Attila Németh to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-1-138-63385-8 (hbk) ISBN: 978-1-315-20700-1 (ebk) Typeset in Bembo by Sunrise Setting Ltd, Brixham, UK

Contents

Acknowledgements List of abbreviations Introduction 1

Self-awareness

vii ix xi 1

The ethical background 1 The papyrus evidence 8 Becoming aware of oneself in the pathologikos tropos 11 Responsible agency: The aitiologikos tropos 24 The elements of thinking 27 Prole psis 27  Memory 48 Taking stock 51 2

Agency and atomism

70

Introduction 70 Reductionism vs. anti-reductionism in Epicurus’ philosophy of mind 72 The central fragments of On Nature XXV 75 A conceptual analysis:  86 The  of the soul and the hypothesis of the swerve 92 Conclusion 99 3

Self-narratives Introduction 108 Animal and human causal capacities 108 The narrative self 116 Sensing responsibility 120 Methodological considerations 122 Strengthening the connections 124 A blurry ending 125 Conclusion 128

108

vi

Contents

4

Lucretius’ cosmological perspective

133

Introduction 133 The cosmogonical crux 134 Epitome variations 134 An arche  of collisions 136 On atomic motion 137 The movements of Natura 138 DRN II 251–93 141 Reconstructing the argument 141 Example 1 144 Example 2 146 The conclusion of the argument 148 Res vs. mens 150 (B) – Causal determinism and fatalism 152 Conclusion 157 5

The pleasures of friendship

166

Friendship as a natural association 166 The pleasures of virtue 167 Rational emotions 169 Philodemus and the therapeutics of friendship 177 Conclusion 182 Epilogue Appendix Index

190 201 203

Acknowledgements

My excitement for learning about Hellenistic philosophy goes back to a seminar by Mary Margaret McCabe on the ethics of the Hellenistic philosophers at University College London in early 1999. Although I had already been thrilled by the antagonism of ancient philosophers who attempted to rationalize their cosmos despite it always somehow remaining full of gods, an idea which was central to their lives, the deterministic implications of cosmic matter concerning human beings had never seemed so pressing before as it became for the Stoics and the Epicureans. The idea to write about Epicurus’ conception of the self, however, appeared to me much later, when I was already doing my PhD at the Royal Holloway, University of London. It was ignited by Richard Sorabji’s book, Self: Ancient and Modern (Oxford, 2006). Although Sorabji addressed the question of personal identity in his examination of the Epicurean Lucretius, he quite understandably did not include those fragments from Epicurus’ On Nature, which are strongly concerned with the topic, and which I knew had been available but had never been discussed. My curiosity to comprehend them as fully as possible led me to write a dissertation on the subject, which was the seed of this publication: only a seed though because this book is not only a fully revised version of it, but it also includes a completely new and extensive discussion and interpretation of Epicurus’ technical concept, the prole psis, an interpretation of  his narrative ideas of the self, and a novel examination of his practice and theory of friendship, as well as a discussion of the Epicurean ideal of godlikeness in relation to the self and its literary implications, all of which were absent from my original conception of the subject. It is my pleasure to acknowledge here the invaluable intellectual and personal debts I have incurred along the way. Kornél Steiger was my first teacher of ancient philosophy, who brought a distant world closer to many of us in Hungary. István M. Bodnár was a great mentor, both during my MA and PhD studies: reading Greek texts with him deepened my understanding of ancient philosophy in all sorts of ways. I am greatly indebted to Anne Sheppard for her supervision during my PhD; our conversations were always very focused and stimulating, and her continuing encouragement and endorsement have been important to me ever since. Attila Ferenczi has been a great teacher and friend: our readings and discussions of Latin philosophy have always been inspiring and thought-provoking.

viii

Acknowledgements

I am most grateful to David Sedley for his enthusiasm about my dissertation and my book, both of which he examined meticulously and enriched through his precious reflections. I am most thankful to James Warren for reading multiple versions of various drafts of this book over the years and for his constantly comprehensive comments. I am filled with great gratitude to Voula Tsouna for her extensive and supportive remarks. I feel fortunate to have been able to discuss various issues while still doing my PhD with Gábor Betegh, and when already working on the book with Péter Lautner. I am grateful to George Boys-Stones for embracing the idea to publish this monograph in his series and to Elizabeth Thomasson, the editor of the classical studies, for her guidance. I appreciate the help of Péter Agócs, Benjamin Harriman and Sean McConnell; my warm thanks to all of them and to many others who have contributed to discussions at different stages. Likewise to audiences who have responded to the presentations of my ideas at the Central European University, Budapest, at Eötvös Loránd University, Budapest, at Humboldt University, Berlin, at the Institute of Philosophy, Bratislava, at Pázmány Péter Catholic University, Piliscsaba, at University College London, and at the University of Edinburgh. I am grateful for the scholarships or funds I have benefited from during my research at Central European University, at Royal Holloway, University of London and at the Institute of Philosophy in Bratislava, and also for the postdoctoral scholarships I have received for shorter or longer periods (TÁMOP 4.2.1/B-09/1/KMR-2010-0003, OTKA 100418 and OTKA 112253). I dedicate this book to my friends and family because it is from them my eudaimonia springs. Let me express my gratitude first of all to my mother for her help and support, and most of all for her patience. I appreciate the encouragement and care of my wife, Hana, who has always reassured me in my pursuits. Our daughter, Kaede, has also been a wonderful source for the practising of Democritean cheerfulness.

Abbreviations

AP

Palatine Anthology

Alexander of Aphrodisias De Mixt. De Mixtione (On Mixture) Aristotle A. Po. De An. N. E. Phys.

Analytica Posteriora De Anima (On the Soul ) Nicomachean Ethics Physics

Athenaeus Deipn.

Deipnosophistae (Deipnosophists)

Cicero (Marcus Tullius) Acad. Academica (Academics) De Fat. De Fato (On Fate) De Fin. De Finibus Bonorum et Malorum (On Moral Ends) DND De Natura Deorum (On the Nature of Gods) D. L.

Diogenes Laertius (Lives and Opinions of Eminent Philosophers)

Diog. Oin.

Diogenes of Oinoanda

Dox. Gr.

Doxographi Graeci, edited by Hermann Diels

Epicurus Ep. Hdt. Ep. Men. Ep. Pyth. K. D. S.V.

Epistula ad Herodotum (Letter to Herodotus) Epistula ad Menoeceum (Letter to Menoeceus) Epistula ad Pythoclem (Letter to Pythocles) Kyriai Doxai (Principal Doctrines) Sententiae Vaticanae (Vatican Sayings)

L&S

Long, A. A. and Sedley, D. N. (1987) The Hellenistic Philosophers, 2 vols., Cambridge

x

Abbreviations

Lucretius DRN

De Rerum Natura (On the Nature of Things)

PHerc.

Papyrus Herculaneum

Philodemus De Oec. De Mort. De Piet. De Poem. De Sign.

De Oeconomia (On Property Management) De Morte (On Death) De Pietate (On Piety) De Poemata (On Poems) De Signis (On Signs)

Plato Alc. Leg. Phaedr. Theat. Tim. ∗ Plato (?) Pliny (the Elder) N. H. Plutarch Adv. Col. De Frat. Am. De Lat.Viv. De Tranq. Anim. Non Poss.

First Alcibiades∗ De Legibus (On Laws) Phaedrus Theaetetus Timaeus

Naturalis Historia (The History of Nature) Adversus Colotem (Against Colotes) De Fraterno Amore (On Brotherly Love) De Latenter Vivendo (On Epicurus’ Advice to ‘Live Unknown’) De Tranquilitate Animi (On the Tranquility of the Mind ) Non Posse Suaviter vivi secundum Epicurum (That Epicurus Actually Makes a Pleasant Life Impossible)

Porphyry Ad Marc.

Ad Marcellam (To Marcellus)

Sextus Empiricus Adv. Math. P. H.

Adversus Mathematicos Pyrroneion Hypotuposeon (Outlines of Pyrrhonism)

Seneca (Lucius Annaeus iunior) Ep. Mor. Epistulae Morales (Moral Letters to Lucillius) SVF

Stoicorum Veterum Fragmenta, edited by Hans von Arnim

Us.

Usener (Epicurea, edited by Hermann Usener)

Introduction

Epicurus was a pragmatic thinker. He maintained that all speculation is worthwhile in as much as it helps to turn human suffering into pleasure, and, in fact, into a very special sort of pleasure: the absence of pain. Epicurus thought that the diverse constituent conditions and their consequent pleasurable affections are valuable only if they dispel pain and help the living being to attain a balanced, healthy and, therefore, tranquil condition. Epicurus’ hedonism was, indeed, unique. He pointed to the cradle and took infant behaviour as sufficient evidence to establish the cornerstone of his ethics, the thesis that every living being instinctively pursues pleasure and avoids pain. Although it seems far from evident what conclusions can be drawn from infant behaviour, Epicurus took the Cradle Argument to justify that humans are naturally motivated to pursue pleasure and to avoid pain and he built his normative theory around this idea, in accordance with the formal requirements of the ancient ethical traditions. Consequently, most modern discussions of his ethics have revolved around the relationship between how the descriptive and normative parts of his theory are compatible with each other, but there has been much less focus on the relationship between his conception of the psychological development of living beings and his ethics. The evidence is meagre and fragmentary, but it is available, and I aim to bring it back to life in the following examination. As a result, the controversial notion of the ‘self ’ will emerge from the surviving fragmentary papyrus texts of Epicurus, which are at the heart of my discussion. Hence in this book I piece together the evidence available on the Epicurean conception of the self. I shall start with the interpretation of the relevant papyrus fragments, telling the story of Epicurus’ conception of self-awareness, developed through one’s internal affections, as he argued. For Epicurus, self-awareness was a process that started at the cradle and continued through one’s whole life. Other than creating one’s emotional states, affections also help form one’s linguistic competence, a set of pre-conceptions and conceptions. And most importantly – according to Epicurus’ principles of epistemological methodology – affections are the criteria of truth along with perceptions and pre-conceptions or prole pseis. For Epicurus,  self-awareness, under the influence of the affections, was consequently essentially related to one’s developing conceptions of what is of positive or negative

xii

Introduction

value. In the Letter to Menoeceus, an epitome of his ethics, he took such a developed, desiring agent for granted, and he concentrated, in arguing for his normative theory, on the classification of human desires. However, in his major and most extensive work, On Nature, which consisted of thirty-seven books, he seems to have explained in detail the formative stages of one’s conception of oneself as a desiring agent. This is where my investigation of the conception of self in his philosophy will start. As will become clear, Epicurus maintained that the affective realization of oneself naturally goes hand in hand with its outward orientation, namely with an understanding of oneself in the midst of the person’s causal network, i.e. in one’s natural and social environment. This was a necessary implication since – as I will argue – Epicurus entertained a general sensationalist view according to which affections are not only the internal criteria for action, but are also the criteria of truth. What people conceive through their affections, therefore, is strongly connected with the external world; thus, how people conceive of their own agency depends naturally on the origin of their affections in the environment. Observing their social networks and contemplating the causal connections of the perceived phenomena through their own naturally acquired pre-conceptual and conceptual set, people can expand their understanding, and with the help of self-reflective thinking can attribute to their own agency the pre-conception of the cause, forming the idea of their own responsibility.1 The reconstruction of this complex process is difficult because of the fragmentary state of the evidence, and it is further complicated by Epicurus’ highly technical vocabulary, most of which requires careful interpretation. Consequently, in Chapter 1 I will have to divert my attention from Epicurus’ ethics to its epistemological background. We shall learn, for example, the function of memory in thinking, which I will also investigate further in the following chapters. This digression is necessary in order to show how Epicurus built a philosophically tenable explanation of self-awareness, taking his cue from the phenomena, the foundation of his empiricist philosophy. In the later fragments of On Nature XXV, Epicurus turned to account for the possibility of responsible agency in the framework of his atomist physics, to show how the concept of a responsible agent was compatible with his atomism. There is no scholarly agreement about whether he was successful, or indeed about how he set about doing this, or even what exactly he wished to achieve in this part of his work. These fragments have been at the forefront of debates concerning his philosophy of mind. I will survey only the two strongest competing interpretations in detail because they represent diametrically opposed positions and thus examining them together provides the best insight into the issue. Tim O’Keefe, on the one hand, attributes a reductionist position to Epicurus’ philosophy of mind. David Sedley, on the other, thinks Epicurus had a non-reductionist philosophy of mind, even allowing for the mental to have non-physical instantiations, which may activate patterns of atomic motion that, due to the swerve, are physically possible ones. Both interpretations are compatible with some understandings of human responsibility. However, are they really compatible with Epicurus’ texts? I re-examine the fragments in detail and find

Introduction

xiii

both competing interpretations unsatisfactory: reductionism fails to explain the irreducible causal faculty of oneself, for which Epicurus clearly argues, while emergentism will prove to be too extravagant, because Epicurus does not allow for non-physical causation as Sedley must. Hence I will argue that Epicurus took a non-reductive physicalist position in his philosophy of mind. In Chapter 3, I will connect the early and the later fragments of On Nature XXV to justify the independent philosophical work achieved by Epicurus’ concept of the self expressed by the reflexive pronoun heauton in some of the fragments. As will become clear from my analysis, the self in Epicurus denotes, in my understanding, the essential and accidental qualities of an individual living being based on the person’s particular state of mental and bodily character. These comprise, on the one hand, one’s awareness of one’s distinctive psychophysical self based on his own affections, memories and present conditions; on the other, of his conception of himself in the cosmos as an individual human being in a particular social network. These features are acquired through a person’s psychological development, which creates an individual conception of oneself as a unified psychophysical being in harmony with its continuous self-awareness, partially due to one’s physical make-up. This self-awareness also leads the person to the understanding of his or her own responsible agency.The conception of the self thus acquired does the job of accounting for personal identity over time (see Lucretius on memory in: DRN III 843–64). All these questions will be examined in detail during the course of my analysis of the difference between animal and human selves and their narrative creation as detectable in the surviving papyri. But, one might reasonably ask, what indicates in these texts that Epicurus had a conception of the ‘self ’? What conception of self does the Greek reflexive pronoun heauton, used regularly by Epicurus, designate? Does this pronoun really pick out a conception of the self? Some scholars have even questioned if the ancient philosophers had a notion of the self at all.2 Their scepticism, I think, has already been sufficiently refuted.3 The various accounts of ancient philosophers, studied inter alia by Richard Sorabji, have shown how diverse the usage of the Greek pronouns (the tokens of our English word ‘self ’) can be in the ancient parlance – the personal pronouns ‘I’, ‘you’, ‘we’ in general, or the reflexive pronoun heauton, meaning ‘oneself ’, or the pronoun autos used in an emphatic way meaning ‘someone him-/herself ’ or even ‘one’s true self ’. These accounts have also made it clear how much these pronouns express the strong interest of the ancients in the individual person, especially from the individual’s point of view. Yet it would truly not be sufficient to attribute a conception of the self to Epicurus just on the strength of some fragmentary evidence in which he recurrently uses a reflexive pronoun, even if he most likely applies it in some sense of ‘selfhood’ in On Nature XXV. But if we take into account his entire philosophy, which advocates hedonism based on one’s psychophysical ‘wellbeing’ or ‘happiness’ (eudaimonia) attained by one’s self-reflective rational considerations, the central position of the self in his account becomes unquestionable. The goal (telos) of Epicureanism was – very much in the vein of the ancient ethical

xiv

Introduction

tradition – to understand one’s own nature and transform it (of course, in Epicurus’ case, this was in harmony with his teachings, which offered the possibility of a perfect human life, no less valuable than that of the gods (see DRN III 322)). An individual conception of one’s self is essential to such an ethics. Epicurus’ ethical end was in line with the principal schools of ancient philosophy, all of which, as Tony Long puts it, raised the simple question of what to make of oneself in an ontological as well as in a practical and normative sense.4 On the one hand, you need to get to know yourself in order to see your place in the cosmos. First you evaluate your nature in order to be able to understand the means by which you can take care of yourself, as Socrates taught Alcibiades in harmony with the Delphic imperative to Know Thyself (Plato (?) Alc. 124a–b). Only by understanding his true character and his related social status could Alcibiades have arranged his political career in a way that would help others.5 Aristotle’s ethical theory can be considered along these lines as well in light of his ergon argument in Book 1 of his Nicomachean Ethics: first we need to identify the human function so we can define the best form of human life. Ancient ethics, thus, seems to take on a twofold aspect: a descriptive and a normative part. Both parts are normally supported by some elaborate metaphysics, which lend a natural foundation to ancient theories and are the framework on which a normative ethics can be built. But in so far as one defines oneself in relation to the world, starting from one’s own perspective, applied ethics naturally requires a common ground for both the descriptive and normative components of an ethical theory. That ground is the subjective and individual conception of the self. As was the case with the Epicureans – as we will see clearly from Epicurus’ Letter to Menoeceus – self-introspection took its cue from the observation of one’s own desires and from the scrutiny of the ways to achieve their satisfaction. This necessarily focuses on the individual’s current conditions. For Epicurus the only direct and objective ground for the evaluation of one’s bodily and mental states was the person’s pleasurable and painful bodily and mental affections (pathe ). Consequently, to form a conception of oneself one first needs to understand how one’s current disposition is related to the satisfaction of one’s desires – whether pleasurably or painfully – and then one needs to evaluate one’s own disposition, by comparing how much one’s desires conform to Epicurus’ threefold classification of desires, the basic building blocks of his normative theory (Ep. Men. 127–8 and K. D. 29). That such self-evaluation was not possible for the Epicureans without social relationships is clearly shown by how essential friendship and the teacher–pupil relationship were for them. It would have been impossible for the individual to apply Epicurus’ normative theory without a community of friends – the idea simply being that one cannot judge objectively the value of one’s pleasurable or painful conditions in the light of an ethical theory. Of course, as a simple egoist you could decide and act in correspondence with the desires you find worth satisfying, and keep satisfying them on a regular basis, yet you may be completely mistaken concerning their natural value. Your consequent pleasures may be pleasurable qua pleasures, but they may not be at all worthy of choice, lacking the support of Epicurus’ ethical theory as a

Introduction

xv

standard. It is essential to recognize what this implied for anyone who wished to become an Epicurean: in so far as he was willing to subscribe to the Master’s tenets, leading a life along egoistic principles was no longer any help, since by joining the Garden, Epicurus’ school in Athens, one had to accept the criticism of the other Epicureans in order to accomplish the Epicurean natural telos by the constant adjustment of one’s value judgements in harmony with Epicurus’ teachings.You may, for example, consider fulfilling your desire for some Lesbian wine an imperative when longing for some, without recognizing that your natural desire for wine includes a non-necessary aspect, i.e. that the wine should come from Lesbos. And if you had a desire for wine merely to quench your thirst then you would have needed to admit that in such a case the desire is not even natural, since water would suit your intention much better, given the dehydrating effect of wine. And you may also drift constantly between your home and country residences believing that it is your circumstances which provide you with a tranquil life, without realizing that it depends first and foremost on the internal dispositions of your body and soul (see DRN III 1053–75), recognitions that could be brought to life with the help of an Epicurean community.6 But if the ethical end depends so much on the conditions of body and soul, where does the concept of the self come in? Is it not just a convenient explanatory term, a ‘modern shortcut’ for summarizing Epicurus’ ideas? Did Epicurus really need such a concept at all to explain hedonic psychophysical ‘well-being’? Yes, I will argue, he did, even if we do not find a definition of the self in his writings, and perhaps such a definition never even existed, which is very likely given Epicurus’ lack of interest in definitions in general. Yet this does not discount the possibility of a conceptual framework for the self. The double aspect of his ethical end – the dependence of one’s well-being on the conditions of both one’s body and soul – and the psychophysical holism of his ethics strongly points to the place of selfhood in Epicurus’ theory. For Epicurus the self was not only one’s intellectual and moral commitments: he did not identify the true self singly with the soul, but included all aspects of the human condition, most importantly from an egoistic ‘I’ perspective.7 It is from his holistic conception of a person’s own disposition that his conception of the self takes its cue, through such observations, for example, as that one’s own causal power in the world is dependent on both bodily and mental conditions. According to Epicurus’ hedonism, one should make choices and avoid things with the help of a hedonic calculus, constantly taking into account bodily and mental conditions and constantly referencing one’s individual psychophysical self.8 Without an ‘I’ perspective that brings all these factors together such a calculus could not work, since it is my psychophysical condition that must be reflected on in harmony with a normative theory in order to carry out a consequent action. That is to say, the hedonic calculus is applied in relation to me or to my self.9 It is also an important aspect of the Epicurean self that this ‘I’ or rather ‘my self ’ has to be identical over time for the calculus to work. Without any memories of the interactions between my psychophysical self and the world, I have no

xvi

Introduction

basis to calculate an appropriate action since a momentary conception of the self simply would not be sufficient.10 Not even Epicurus’ normative ethical theory would suffice on its own as a canon for the calculation to work since its telos is strongly connected to one’s individual psychophysical condition, implying personal identity over time. As Sorabji has pointed out concerning Lucretius’ idea of palingenesis,11 the fact that we lack a memory of our former assemblies is not a sufficient reason to think that we need not be concerned about our future reassemblies either, especially if we take into account the asymmetrical nature between concerns for the future and the past, the former being ‘typically an object of concern in a way that the forgotten is not’.12 Consequently he finds Lucretius’ other solution to the problem more adequate, interpreting lines 861–3 of DRN III in harmony with lines 677–8 of DRN III that it is the uninterrupted retention of memory, in the sense of the possibility to exercise retained memory, which is the necessary condition for personal identity. This is very much in harmony with the kind of personal identity necessary for the hedonic calculus to work and this is the kind of conception of the self that surfaces in the fragments of On Nature XXV. There we come again into contact with the part memory plays in reference to the Epicurean natural telos, and it only makes sense to interpret memory as playing a part in one’s personal identity and consequently being an indispensable part of one’s self. Take the famous example of Epicurus counterbalancing his extreme pains on his last day with the remembrance of his past conversations, as mentioned in a letter to Idomeneus, one of the major figures of his school (D. L. X 22). In order to compensate for his pain, Epicurus needed memories of his own, which implies both that the person who lived those memories was the same as the person suffering in the present, and that memories build one’s personal identity and one’s conception of one’s psychophysical self. In this particular example it is mental pleasure, which counterbalances bodily pain – both of which are felt by the very same person. Such an operation would not be possible, nor could it be important, if one identified oneself only with the soul, since (e.g. on a Platonic view of the body as the prison of the soul in the Phaedo) one would rather look forward to death in a case such as this one. But Epicurus chose to optimize his present condition by remembering past circumstances, and even if it sounds somewhat counter-intuitive or even counter-experiential that it is possible to compensate one’s excruciating sufferings so easily, he inadvertently pointed towards some necessary elements of his concept of the self. His idea of self-reflective thinking in accordance with both the pathologikos tropos and the aitiologikos tropos (the manners dealing with being affected and with the causes of things, respectively) in the early parts of On Nature XXV will also make clear how broad his concept of the self was: he found it important to also absorb into it the opinion of others. Fortunately, my analysis, unavoidably conjectural owning to the fragmentary nature of the evidence, is corroborated by some other fragmentary papyri: writings of the first-century BCE Epicurean Philodemus, whose work On Frank Criticism describes the constantly corrective methodology of the teacher–pupil relationship in the Garden which also

Introduction

xvii

pertained among friends of the Epicurean community, will be considered in Chapter 5. Frank criticism was first and foremost a therapeutic method, applying different manners of verbal correction, obviously incorporating personal judgements about others as well. Sharing the evaluative perspectives of others meant that this constant and mutual exchange of beliefs about one another augmented the conception of one’s own identity and self-knowledge. Consequently, selfcontemplation was essentially connected with value judgements made of one by others in the Epicurean community. These judgements obviously helped to shape one’s self-conception in relation to the ethical commitments of an Epicurean social group all working towards an ethical ideal provided by Epicurus himself. One may still wish to ask the question ‘what sort of philosophical function does the concept of the “self ” fulfil in Epicurus’ philosophy?’ I think the Greek reflexive pronoun  indicates a ‘thin’ idea of the self when it is a paraphrase for ‘the subject of awareness of an individual psychophysical being’. It incorporates the findings of one’s self-reflective thinking (), which is the understanding of one’s specific qualities as a human being () based on his or her particular features. These features can consist, on the one hand, of one’s awareness of oneself through one’s affections in harmony with the pathologikos tropos; of one’s self-conception as a unity of body and soul (); of personal identity over time, conceived through one’s psychophysical being based on one’s memories and experience, all of which are important elements in psychological development and in the consequent formation of conceptual apparatus. On the other hand, such a ‘thin’ idea of the self is formed in immediate relation to one’s environment, social and cultural interaction, both on a micro and macro level, which leads to an understanding of oneself as a human individual among living beings ().13 Therefore, the philosophical work of the reflexive pronoun is to express collectively the distinctive features of a self-reflective being without spelling out each and every time all the relevant features that are implied by the discussion.This is what I mean by the ‘thin’ idea of the self and its function in Epicurus. Nonetheless, since the ‘thin’ idea of the self does not refer to an entity separate from the psychophysical human being, the reference of the self in Epicurus is a thick one. It refers to the conception one creates of oneself based on the above listed particular, essential and accidental features. Depending on the context, Epicurus places much emphasis on taking into account all the relevant factors concerning one’s causal agency, and I will argue that his infamous term ta apogegennemena –  found particularly in those frustrating fragments from On Nature XXV – picks out a temporally limited, occurrent mental state belonging uniquely to the individual living beings, both humans and wild animals, with further restrictions concerning the latter. Even in such contexts Epicurus uses the reflexive pronoun, which is the ‘thin’ idea of the self, but its reference is a thick one. For example, in Fr. 11 Epicurus says the following: ‘But many [i.e. people], though by their nature they are able to become creative of these and those [results], they fail because of themselves () and not because of the same cause of the atoms and of themselves ()’.

xviii

Introduction

The ability not exploited by those people who have the capacity to act and behave in certain ways also points to some occurrent mental states of those who are not willing to act in responsible ways. It is not simply the question of their moral agency or their abilities as human beings, but the question of their dispositions and the conception of themselves they have developed which determine their actions. Their slackness is not based on lack of human ability but it flows from their characters and from their realized conceptualization of themselves. Therefore, the ‘thin’ concept of the self could not be replaced by the concept ‘human’ in all contexts – as, for example, my discussion of Epicurus’ narrative idea of the self in Chapter 3 will also show more particularly – because the latter concept is already presupposed in the perception of one’s individual self. The notion of the self thus emerges in Epicurus’ struggle to express this individual perspective of oneself in the process of one’s holistic self-reflection of oneself as an individual psychophysical being. Chapter 1 starts out with the Epicurean methodology of ataraxia, which was a reflective conceptualization of one’s desires. It required the rational being to absorb Epicurus’ classification of desires and integrate it into the recommended hedonic calculus, which was closely related to one of the criteria of truth, the pathos (plural pathe ) . Some of our papyrological evidence expands our understanding of the function of pathe  in Epicurus’ theoretical framework, or so I shall argue, according to which his conception of self-awareness in the pathologikos tropos was the psychological development of oneself through one’s own affections (pathe ) . As will become clear, Epicurus also conceived of self-awareness in the aitiologikos tropos, which was the means for one’s self-conception through one’s social and cultural interactions. These two ways of self-intellection are augmented by the full consideration of the causal powers of moral agency in the later fragments of book XXV of On Nature. In Chapter 2 I shall scrutinize these fragments in detail and put forward my non-reductive physicalist interpretation of Epicurus’ philosophy of mind based on them in harmony with a new theoretical explanation of the atomic swerve. In Chapter 3 I will tighten the threads of the various discussions of the fragments from book XXV and will evaluate the book as a whole. I will corroborate my understanding of the reflexive pronoun  as a ‘thin’ idea of the self by reconstructing Epicurus’ narrative conception of the self. I devote Chapter 4 to a full-scale evaluation of my interpretation in light of Lucretius’ evidence, including that of the swerve. Lucretius will also emerge as a friend in the Epilogue, realizing a literary version of the ideal Epicurean friendship – discussed in Chapter 5 – by sharing the cognitive disposition of Epicurus in his poetic apotheosis.

Notes 1 For the difference between self-reflexivity and self-reflectivity see Seigel 2005, pp. 12–13. I take it that for Epicurus self-awareness was based on the rational reflection of the soul, and therefore it can be best characterized as self-reflective. 2 See Anscombe 1975, Kenny 1988 or Taylor 1989 for the differences between the ancient and the modern conception of the self.

Introduction 3 4 5 6

7 8 9

10 11 12 13

xix

See Gill 1990, 1996, 2006 and 2008; Sorabji 2006; Remes and Sihvola 2008. See Long 2001. See Annas 1985. One may contest the methodology of my interpretation by saying that it is based on a combined consideration of theory and practice. But it is based, in fact, on Epicurus’ imperative that the value of a theory can be measured only by its applicability; see Porph. Ad Marc. 31 = fr. 221 Us., quoted in full in the Epilogue, p. 196. Contra Gill 2006, p. xv. For the hedonic calculus see Chapter 1, pp. 23–4. In my view this constant self-reflectivity is exactly why we need to use the concept of self in our analysis of Epicurus’ psychophysical holism rather than another modern notion, namely that of ‘agency’, which appears to me to be limited to picking out one’s causal responsibility. See also Warren 2014, p. 182, who draws attention to the new, individual perspectives of Epicurean assessment. Thus the playwright Epicharmus’ idea of growth being a continuous destruction of one’s self would not be adequate; see Sorabji 2006, p. 57, n. 1 for references. The idea of palingenesis is, briefly, that given the infinity of time it is necessary that one’s atomic constitution will reassemble and, in fact, has reassembled infinitely many times without ever constituting our present selves. Sorabji 2006, p. 98. There are twenty-seven instances of the reflexive pronoun in the fragments of On Nature XXV, five instances of the word human and four instances of living being; see Laursen 1997, pp. 75–7.

Bibliography Annas, J. (1985) ‘Plato on Self-Knowledge’, in D. J. O’Meara, ed., Platonic Investigations, Washington: 111–38. Anscombe, E. (1975) ‘The First Person’, in S. Guttenplan, ed., Mind and Language, Wolfson College lectures, Oxford: 45–65. Gill, C. (1990) The Person and the Human Mind: Issues in Ancient and Modern Philosophy, Oxford. —. (1996) Personality in Greek Epic,Tragedy, and Philosophy:The Self in Dialogue, Oxford. —. (2006) The Structured Self in Hellenistic and Roman Thought, Oxford. —. (2008) ‘The Ancient Self – Where Now?’, Antiquorum Philosophia 2, 2008: 77–99. Kenny, A. (1988) The Self, The Aquinas Lecture, Milwaukee, WI. Laursen, S. (1997) ‘The Later Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanese 27: 5–82. Long, A. A. (2001) ‘Ancient Philosophy’s Hardest Question: What to Make of Oneself?’, Representations 74: 19–36. Remes, P. and Sihvola, J., eds., (2008) Ancient Philosophy of the Self, Dordrecht. Seigel, J. (2005) The Idea of the Self: Thought and Experience in Western Europe since the Seventeenth Century, Cambridge. Sorabji, R. (2006) Self: Ancient and Modern Insights about Individuality, Life, and Death, Oxford. Taylor, C. (1989) Sources of the Self, Cambridge, MA. Warren, J. (2014) The Pleasures of Reason in Plato, Aristotle, and the Hellenistic Hedonists, Cambridge.

1

Self-awareness

The ethical background We must rehearse the things which produce happiness (), seeing that when happiness is present we have everything, while when it is absent the one aim of our actions is to have it. (Ep. Men. 122) At the beginning of the Letter to Menoeceus, a brief summary of his ethical teachings, Epicurus states the general ethical principle that happiness or eudaimonia is the ultimate end of our actions and aligns himself with the ancient ethical tradition stemming from Socrates.1 A common characteristic of ancient philosophy was the presentation of distinctive conceptions of the good life, which include normative theories of action producing the desired end (telos) in particular.2 Crucially, the desired end, since most of the ancient thinkers worked with the underlying assumption that everyone desires to live well, to have a ‘happy’ (eudaimon ) life. This assumption originates in Socrates’ ethical intellectualism, according to which everybody aims for a good end for oneself, and nobody knowingly aims for anything bad. According to what has been termed the prudential reading of Socrates’ ethical intellectualism, good and bad things are to be understood as advantageous and disadvantageous, and, consequently, nobody acts against his best interests voluntarily. The moral reading understands Socrates’ intellectualism to define good and bad in terms of right and wrong and says that nobody aims at doing the wrong thing voluntarily. The second reading follows from the first with the additional Socratic premise that ‘it is never in one’s interest to act unjustly ()’.3 Obviously, the second formulation is open to the objection that if what is wrong is never done deliberately or voluntarily then wrong action is involuntary and no one can be held responsible or blamed for his actions. However, Socrates maintained that if we are driven by certain natural desires towards something we believe to be good, some good end, it is our first and foremost responsibility to make sure that our actions are in harmony with the health of our souls, which can only be achieved by reason. Reason defines the real conditions of a eudaimon life, and the consequent morally correct and responsible action.4 Self-knowledge via responsibility thus became a

2

Self-awareness

necessary condition within Socrates’ radically internal conception of happiness without which ethically correct behaviour was considered to be impossible. Consequently, self-knowledge, in a broad sense, was also an ethical requirement within the eudaimonistic ethical tradition. It called for the knowledge of one’s particular mental states, including one’s beliefs, desires and sensations. More often than not it also consisted of the knowledge of one’s persisting self – its ontological structure, identity conditions or character traits.5 The opening sentence of Epicurus’ letter to Menoeceus, possibly a student or follower of his, reflects these requirements concomitantly, since it is maintained that primarily one can recognize one’s current condition of happiness by reflecting on one’s realized conceptualization of oneself – the comprehension of which was also a complex, composite process according to Epicurus, who, in sympathy with Socrates, had an important place for self-knowledge in his ethics. Epicurus thought that people are responsible for their bodily and mental health, the latter of which is often disturbed by the fear of death and gods, both fears affecting one’s overall condition.6 To eradicate these basic fears Epicurus recommended philosophy as the medicine of souls, which if administered properly started with a reflection on one’s own set of beliefs about oneself and the world. The scrutiny of beliefs was essential according to Epicurus because he thought that all disturbances were structured in the following way: False Beliefs (about the world/values) → Empty Desires → Disturbances. Consequently, the procedure for Epicurus’ therapy was to remove false beliefs in order to remove all the rest that depend on them (i.e. empty desires and consequent disturbances).7 Therapy was not, of course, practised in terms of such a crude theory but a conjecture like this was the basis on which the Epicurean practitioners built their strategies when dealing with individual patients, friends and students of the Epicurean community.8 This chapter starts by setting the ethical background from which Epicurus’ ideas of self-knowledge originated. I will argue that Epicurus’ notion of self-knowledge stemmed from his conception of self-awareness through one’s affections (), which govern every animal, including humans, immediately from birth. We animals become aware of ourselves through our affections (pathe ) , and instinctively pursue pleasure and shun pain. Our close textual focus on some of the early parts of Epicurus’ On Nature XXV from the Herculaneum papyri will not only corroborate these ideas but will extend the scope and depth of our view about how Epicurus conceived of one’s conceptualization of oneself in relation to the external world, influenced and stimulated constantly by the correlative affections. We will see various, related topics emerging such as the function of memory that are concerned with Epicurus’ accounts of how humans come to think of themselves as agents that are affected and experience pleasure and pain; what the underlying physical mechanisms are for these processes and how Epicurus conceptualized them – well-known terminology indicating crucial concepts such as the epaisthe s is or prole p sis will be scrutinized

Self-awareness

3

in detail; during the investigation of the latter we will pursue how Epicurus thought agents conceive of themselves as responsible agents and how they conceive of other agents in the same ways. Securing the Epicurean goal of freedom from bodily pain () and mental disturbance () in order to live a joyful life did not begin from an immediate questioning of one’s beliefs. Epicurus taught that for all our actions to aim at the end or telos we need to observe and understand our desires first. That is, proper action requires reflection on the characteristics of our desires, which helps to reveal the content of our beliefs. As adult rational beings, we can only achieve or maintain the healthy conditions of the body and the soul if we classify our desires first and satisfy only those which are natural and necessary for the accomplishment of our happiness. We must reckon that some desires are natural and others empty, and of the natural some are necessary, others natural only; and of the necessary some are necessary for happiness, others for the body’s freedom from stress, and others for life itself. For the steady observation of these things makes it possible to refer every choice and avoidance to the health of the body and the soul’s freedom from disturbance, since this is the end belonging to the blessed life. For this is what we aim at in all our actions – to be free from pain and anxiety. Once we have got this, all the soul’s tumult is released, since the creature cannot go as if in pursuit of something it needs and search for any second thing as the means of maximizing the good of the soul and the body. (Ep. Men. 127–8)9 Having distinguished the sub-class of the natural–necessary desires from desires which are natural only and those desires which are empty, Epicurus lists the three different ends of natural–necessary desires from the top down: (1) for happiness; (2) for the body’s freedom from stress; (3) for life itself. Desires related to the basic conditions for living such as eating and drinking, but not of the excessive kind,10 belong to (3), and the fulfilment of (3) is the necessary but not the sufficient condition of (2), since the fulfilment of (3) is not sufficient to satisfy, for example, the desire to get rid of bodily pain, such as wishing to avoid hearing a painful noise. The asymmetrical relationship between (2) and (1) is similar, since (1) obviously includes not only bodily desires but also desires related to one’s soul, in harmony with the Epicurean ethical maxim. Epicurus’ claim that we aim for freedom from pain as our end or telos in all our actions is, nonetheless, not argued for in this passage. It is a reformed restatement of the eudaimonistic ethical principle according to which happiness is the ultimate end of our actions. However, it is not merely something to be pointed at, like fire, but it is to be inferred from the steady observation of our desires. How can we reach such an inference by simply observing our desires? Epicurus seems to be saying that, having grasped the real nature of desires, the living being cannot pursue just anything in order to maximize his or her good of the soul and the

4

Self-awareness

body – or at least it would not be rational or prudent for him or her to do so once he or she has correct beliefs about how to obtain pleasure.11 However, up until this point, Epicurus has not shown why we should conceive the satisfaction of our natural–necessary desires as good, why recognizing and following them would lead to proper action and to the maximizing of the good of the soul and the body. For that he needs to point out the real driving forces of our nature. For the time when we need pleasure is when we are in pain from the absence of pleasure. we no longer need pleasure. This is why we say that pleasure is the beginning and end of the blessed life. For we recognize pleasure as the good, which is primary and congenital; from it we begin every choice and avoidance, and we come back to it, using the feeling as the yardstick for judging every good thing. Since pleasure is the good which is primary and congenital, for this reason we do not choose every pleasure either, but we sometimes pass over many pleasures in cases when their outcome for us is a greater quantity of discomfort; and we regard many pains as better than pleasures in cases when our endurance of pains is followed by a greater and long-lasting pleasure. Every pleasure, then, because of its natural affinity, is something good, yet not every pleasure is choiceworthy. Correspondingly, every pain is something bad, but not every pain is by nature to be avoided. However, we have to make our judgment on all these points by a calculation and survey of advantages and disadvantages. For at certain times we treat the good as bad and conversely the bad as good. (Ep. Men. 128–30)12 Epicurus recognizes pleasure as the ignition of our actions.13 He does not give an argument for this claim here, and the Stoics already questioned whether the Cradle Argument substantiates hedonism.14 Epicurus thinks we, as humans, are endowed with a natural desire to seek pleasures.15 In this sense the desire to seek pleasures is congenital, but what makes them primary and good? If this passage is read as strongly connected with the previous one then it seems clear that pleasure as good is to be conceived as something constituted by the natural and, consequently, healthy conditions of the body and the soul. If once the health of the body () and the soul’s freedom from disturbance () are secured and, consequently, the end or telos of a joyful life is obtained, being in the state of (katastematic) pleasure, characterized as no longer in pain or in search of more pleasure, seems to provide one with all the pleasure one needs for happiness.Thus, if (kinetic) pleasure is the means or variation – or both – of the joyful life, and being in the state of (katastematic) pleasure is equivalent to having a joyful life, pleasure is surely good as well as primary: primary as in being the most valuable rather than in a chronological sense.16 Yet, not all the ancient critics were satisfied with such an explanation: Cicero complained (De Fin. II 28) that the name ‘pleasure’ cannot be univocally applied to both kinetic and katastematic pleasures. The former corresponds to how

Self-awareness

5

pleasure is understood in the common sense, involving some motion and change, some sweet sensation and being either some titillation of the body or joyful state of the mind. However, katastematic pleasure should not be called pleasure at all according to Cicero, but is simply the absence of pain. Cicero’s worry concerns the question of the Epicurean telos of the good life: he claims that Epicurus should have held that the end is either a composite of pleasure and the absence of pain, or simply only the latter, which would deprive Epicurus of being a hedonist. John Cooper contends convincingly that Cicero, in order to serve his argumentative purposes, misrepresents the views of Epicurus by distinguishing two kinds of pleasure, where there is actually only one, realized under different circumstances.17 Cooper argues that the distinction between kinetic and katastematic pleasures is not between two genera of pleasure, but rather between the causal relations of how the phenomenon of pleasurable affective states are realized. That is to say, pleasure is a single form of awareness coming about in different ways, under different conditions. In the case of kinetic pleasures, they come into life alongside some bodily or psychic movements, for example when eating or drinking or satisfying one’s sexual desire, or when listening to some pleasant music or reading Homer. In such cases one is affected pleasurably in one’s body or soul as a result of one’s activity. These are, in other words, the kinetic conditions for pleasure and are described as pleasures of the body and the joys of the soul by Epicurus. According to Cooper, the affective phenomenon of pleasure itself brought about by one’s katastematic conditions is not any different; the distinction lies in what constitutes one’s pleasure and how it is evaluated. In the case of katastematic pleasure it is the condition of one’s overall disposition, or the disposition of one’s self constituted by one’s bodily and psychic conditions, which determines one’s pleasurable affective conditions. However, not just any condition brings about such pleasure, such as those dependent merely on one’s subjective judgements, but there are formal requirements to quantify and qualify a pleasurable life, which was considered even worthy of the gods according to the Epicureans.18 I agree with Cooper to the extent that Epicurus did not distinguish two genera of pleasures. However, he seems to go too far by claiming that kinetic and katastematic pleasures form a single state of consciousness.19 It seems obvious that the affective contents of pleasures can be very different, depending on what brings them about. I think Cooper’s interpretation makes more sense if we say that Epicurus considered every pleasure uniform only to the extent of being pain-free states, realized under different conditions. Epicurus seems to have distinguished two major conditions of the constitution of pleasure, namely the kinetic and katastematic circumstances, as Cooper argues, but he did not go as far as to identify the affective content of sexual pleasure with the pleasures of learning or having a healthy disposition, which would have been quite counterintuitive indeed.20 If this understanding is correct, Cicero’s other objection, namely that the Cradle Argument can prove at most that newly born animals pursue merely

6

Self-awareness

kinetic pleasures, is flawed, since Epicurus could have answered Cicero in the following way: when we see infants pursuing (kinetic) pleasures, they are not simply after kinetic pleasures but, in fact, are instinctively motivated to regain their natural, painless conditions, which substantiate katastematic pleasures as well, articulated appropriately to their age. That is to say, in the case of a newly born animal, having no developed rational capacities yet, kinetic and katastematic pleasures coincide because they are fulfilled by the same activities. It is clear then that Epicurus held pleasure to be the end as the supreme good. But how are we to take it as the beginning? The straightforward answer – based on Ep. Men. 128–30 quoted earlier – is that the affection (pathos) of pleasure as a criterion helps us to distinguish what is good and bad,21 at least at the irrational level – in the sense that it represents unfiltered information about one’s physical states.22 All pleasure is good and all pain qua pain is bad, and therefore they can serve as criteria for action. Furthermore, if affections already govern us from birth with the consequent naturally appropriate desires, appropriate not only superficially but in the case of truly natural–necessary desires, then the evaluation of desires by classifying them rationally provides us with a set of good natural–necessary desires related to the three differentiated ends or tele  of natural–necessary desires. By pointing out that these natural–necessary desires are hardwired in us by the naturalistic grounding of pleasures as good, Epicurus might seem to be objecting to the necessary link between pleasures and our own subjective responses or attitudes towards the satisfaction of the related desires.23 As Phillip Mitsis has shown, Epicurus’ central concern when assessing pleasure is whether our objective needs are satisfied or not. Consequently, the truth of hedonism is only given immediately in experience as far as the affections or the pathe  go; that is, as far as we can tell, whether we have a pleasurable or painful pathos. Nonetheless, it is a matter of evaluation and calculus whether we take a pleasurable pathos as choice-worthy, since there are further formal requirements for pleasure to qualify as good according to the normative part of Epicurus’ ethical theory. One of the important characteristics of such an evaluation is the turn towards oneself to find objective values through the knowledge of one’s self. The steady observation of all of our desires will lead to the selection of the natural–necessary desires, which break down according to their three different ends or tele . Having distinguished these ends, we can calculate on each occasion whether the satisfaction of a particular desire belongs to one of the sub-classes of the natural– necessary desires or not. Accordingly, we can decide that their satisfaction and the consequent pleasures or pains are choice-worthy or to be avoided. But to learn the good, first we have to attend to ourselves and understand the different characteristics of our various desires, otherwise we will never be able to act in harmony with Epicurus’ ethical principles. In Epicurus’ theory, the only objective grounds for evaluation of good and bad rely on the claim that pleasure qua pleasure is good and pain qua pain is bad. It is our pathe , our affections, which inform us about our immediate pleasurable and qua pleasurable, good or painful, and qua painful, bad experiences.24 Pathe 

Self-awareness

7

are real causal events in direct causal connection with the world, and the information we obtain through them serves as signs for further inference, awaiting further witnessing or counter-witnessing by subsequent evidence. There is a strong causal connection between our internal pathe  and the external world. Unlike the Cyrenaics who held that our pathe  inform us only about our internal states, for the Epicureans, due to their theory of perception and thinking in which perception or aisthe s is is coordinated with our internal pathe , pathe  are necessarily causally connected with the outside, giving us information about the external world.25 Consequently, the Epicureans had a much more outwardly extended conception of the self than the Cyrenaics. According to Epicurus we are not merely aware of our pathe  but as rational, autonomous agents we pay attention to them in direct connection with our environment. By virtue of our immediate affections we sort out not only our desires and calculate the consequences of their satisfaction but we also take into account the long-term effects, the further consequences concerning ourselves. In this sense, pathe  do not necessarily initiate only automatic impulses in us to choose or to avoid something but they also prompt us to become aware of ourselves, and to understand ourselves in relation to our environment. ‘Sober reasoning ( ) . . . tracks down the causes of every choice and avoidance and . . . banishes the opinions that beset souls with the greatest confusion’.26 If we are prudent enough to realize how different pathe  are related to us by actualizing our rational faculties, eventually we will create a conception of our own selves. I take it that the achievement of an ideal conception of ourselves turns on the right understanding of our pathe , without which, in and of itself, Epicurus’ theory is not able to address how to classify desires appropriately. There is strong evidence that Epicurus considered the idea of self-awareness and the awareness of the external realm within the conceptual framework of pathe  in his hedonism. According to the Epicurean Cradle Argument, every animal, once born, rejoices in pleasure and rejects pain, as their yet uncorrupted nature directs them. Since their affections are entirely irrational, their pathe  provide a natural internal starting point for instinctive choice and avoidance. Having no developed rational capacities yet, Epicurus could claim that infants act according to their affections as they realize the external world through them. On the strength of the surviving fragments of book XXV of On Nature, more can be said about the functional role of affections in one’s awareness of oneself and of one’s environment: Epicurus seems to have claimed that one’s essential means to conceive of oneself – or, as he put it, of one’s self-reflective thinking – are one’s pathe . His claim seems to be strongly connected with his view of how one develops a self-conception and self-evaluation in the scrutiny of desires. Thus, it seems to be a stronger claim than what we can assert about the functional role of affections in one’s awareness, at least on the strength of the Cradle Argument. In what follows, I will investigate the primary evidence first, and then I will support my interpretation of it with further textual evidence. Some analysis of the relevant terminology will be undertaken along the

8

Self-awareness

way, which will also help in comprehending the theoretical structure of thinking in Epicurus’ philosophy. As a result the different mechanisms of the Epicurean conception of self-knowledge through others will appear, which will turn out to be essential means for achieving eudaimonia in Epicurus’ intellectual hedonism.

The papyrus evidence Chance events are less reliable forms of textual survival from antiquity than consciously institutionalized copying of personal canons, yet the former played a large role in our inheritance of some primary and invaluable evidence of Epicureanism. In the ruins of Herculaneum, a town destroyed together with Pompeii and other settlements surrounding Mount Vesuvius as a result of its eruption in 79 CE, along with hundreds of other papyrus scrolls, some remnants of Epicurus’ major work, On Nature, were found in a villa, which has been named after its contents, the Villa dei Papiri. The villa itself most likely belonged to Lucius Calpurnius Piso in the first century BCE, who was the father-in-law of Julius Caesar and a keen associate of a prominent Epicurean, Philodemus. Since the bulk of the findings contain Philodemus’ works, he is generally thought to have been a resident himself, owning a library at the site, but there is little ancient testimony to support this conjecture. Also, the largest number of books by far were found in a storage area unsuitable for sitting down and reading in, which has given the impression to some that, even if there was a library in the building, the unimportant works were intentionally left behind during the eruption of Mount Vesuvius27 – although one could also simply assume that the boxed scrolls were prepared to be saved but abandoned whilst escaping the area, even if our testimony of the events varies about how the unfolding events were perceived.28 It was not until 1709 that excavations started on the site, as instructed by the Bourbon rulers of Naples. The villa itself was found during the 1750s and the discovery of the papyri in it caused a significant stir, as people hoped for lost works of Latin and Greek literature to surface. The initial findings, nevertheless, resulted in disappointment. The works of an Epicurean philosopher, mostly known for his epigrams, did not appeal to many, and somewhat surprisingly – considering that the Epicureans generally neglected liberal arts – Philodemus turned out to have had a great interest in aesthetics as well.29 But the focus of our interest is the papyri from Epicurus’ major work On Nature (), originally consisting of thirty-seven books. Of these books only remnants of a few have survived, although, somewhat surprisingly, three copies of book XXV have been found. The dates when the scrolls were unrolled vary and the quality of their conservation – partially owing to the different techniques applied at different junctures – is uneven. Initially, the carbonized papyrus scrolls were simply cut and split along their length to separate layers, which did not result in intelligible texts. In 1754 a Vatican writer, Antonio Piaggo, was employed, who – with the help of a machine he built – made it possible for the

Self-awareness

9

papyri to unroll under their own weight. Needless to say, it was a painfully slow method, and by Piaggo’s death in 1796 only eighteen rolls had been unfurled. In 1802, John Hayter picked up the work for the following four years, extending employment to a neighbouring town, Portici. He had many Piaggo machines built and increased the rate of progress by financially motivating the draughtsmen. Hayter supervised his employees rather well but he could not control their greed, so the more fragile papyri, or the ones needing more expertise, were often quickly abandoned and left to perish. Hayter was forced to leave at the time of the French invasion of Naples and took his facsimile copies to Oxford. These facsimiles, or disegni, were made at the time of opening the rolls and, in many cases, preserve texts no longer present in the original papyri due to deterioration. If the text being copied was stuck to a lower layer, it was scraped off after its contents had been drawn in order to reveal the underlying text. Such fragments are only known from these drawings, which are now kept in the Bodleian Library in Oxford. But most of the papyri remain in Naples, where work on them has continued ever since.30 A less serious complication, yet significantly influencing textual interpretation, has been the difficulty of recognizing the matching pieces of the resulting textual puzzle and of putting the opened scrolls in order relative to each other – and even identifying from which book certain fragments derive.31 Owing to these complications, the evidence from the three copies of book XXV of Epicurus’ On Nature was originally catalogued under six different numbers in the Neapolitan library. But, as it has been established, the relevant papyri of book XXV can be grouped as PHerc. 419/1634/697, PHerc. 1420/1056 and PHerc. 1191, based on their contents, their three different scribal hands and orthography, each group forming a single copy of book XXV.32 The discovery of three presumably authoritative copies of a single book in the same library is unique, and it seems to suggest that it had a pre-eminent status in Epicurus’ curriculum as one of ‘the more important books in a sequence’ (Ep. Hdt. 35).33 The whole work of On Nature was a lecture course, which becomes clear not only from Epicurus’ explicit statement at the end of book XXVIII but also from its expansive, colloquial style, which, as David Sedley has suggested, was meant to invite the reader to join Epicurus in his quest of physical investigation.34 As the evidence of Diogenes Laertius (X 29–30) testifies, On Nature was a treatise on the study of physics, a part of Epicurus’ curriculum, the whole of which he divided into three major components: (1) epistemology; (2) physics, i.e. the study of nature, a discipline which is subordinated to (3) ethics. Thus, given the strong connection between the different parts of Epicurus’ philosophy, it is natural to find all sorts of questions coming up in the fragments of this rather expansive work even if its primary focus was physics.35 Book XXV, in particular, includes a wide range of themes such as mental development, memory, self-reflective thinking, proper conduct, the telos of life and human behaviour and its relationship to the external world.

10

Self-awareness

At the end of book XXV, Epicurus says that he has given an account of two manners of explanation, the pathologikos tropos and the aitiologikos tropos.36 The former was most likely a discussion of topics in terms of pleasure and pain, and the latter a discussion in terms of causes. Epicurus does not say, as Sedley put it, that these are ‘successive levels’37 – tackling his topic first in one tropos then in the other – but Epicurus simply claims that these two manners from those proposed at the beginning of the discussion (on the weaker and more likely reading, at least from the beginning of book XXV) have been explained. However, it does not necessarily exclude the possibility that some of the questions of book XXV may have been discussed in both manners, even if from different perspectives or with different emphases. As the fragments will in fact show they were used as explanatory means or essential methodological tools for the treatment of different topics sometimes simultaneously. Self-awareness, one of the major themes of book XXV on which I shall focus, was one such topic, which can take a variety of objects, such as one’s current activity, one’s motivation or desires, one’s character or one’s essence.38 Epicurus, at least, seems to have been concerned with one’s character in relation to one’s mental development in both manners, the pathologikos tropos, as one’s self-conception through one’s own affections or pathe , and the aitiologikos tropos, as the awareness of one’s responsible self by means of one’s prole p sis of the cause.39 The three copies of book XXV are all very lacunose. The readings of the papyri have been significantly improved by David Sedley and further refined by Simon Laursen. Laursen’s work has been published in two volumes of Cronache Ercolanesi.40 As Laursen acknowledges, the English translations he gives are not to be taken as the final version, therefore I provide new translations for all of those fragments I discuss. Needless to say, given the state of the evidence, it is hard to make good sense of it in some cases, and consequently much caution is required for anyone giving an interpretation of the sometimes incomplete sentences, which are in parts fragmentary themselves. For such work it is inevitable that one has strong preliminary ‘extra-textual assumptions’ as Laursen put it.41 However, I think there is much to learn from these fragments, and if a coherent and plausible interpretation of them can be given, fresh discussions of Epicureanism can be stimulated. In this chapter, I will present all the fragments on the subject of selfreflective thinking – which I take to be the last eight fragments of the early parts of book XXV in Laursen’s edition – even though I do not intend to give a detailed analysis of all of them. Such an approach makes the presentation more coherent because some of these eight fragments, even without a detailed examination, lend further support to my argument. It also does more justice to the fragments themselves, since the reader does not have to read them in isolation, but he or she can become familiar with a group of the extant fragments from book XXV in their complexity. I progress with the discussion of the fragments in order. I do not give my full understanding of each of them in the first instance, if further arguments to come can establish the interpretations

Self-awareness

11

more safely and if more support can be given to my interpretations on the strength of the overall understanding of the context.42 Thus for support I will sometimes bring in fragments even prior or posterior to the topic of selfreflective thinking. I have numbered the fragments I have taken to constitute the section on self-reflective thought as Fr. 1–Fr. 8 for the sake of easier reference. The texts I provide are based on Laursen’s editions;43 nonetheless, for some of them I also propose different readings. Two of them (Fr. 4 and Fr. 6) I present in footnotes for the sake of completeness. I refer to all the sources of the papyri and to Laursen’s publications, as well as to Arrighetti’s second edition as Arr. [. . .], unless he did not have certain texts in his publication.44

Becoming aware of oneself in the pathologikos tropos Fr. 1 [+/– 10/12 ][...][ +/– 6/7 ]  [  ]  []   [.].[ +/– 5/6 ][...][ (Fr. 1)45 . . . I must say that he is said to think of himself by means of himself. By means of this . . . Fr. 2   [[]]         [[]]       [] []    [+3/4].   []  [ ]  [ ]  [ ]  [.].[+/– 6/7][.][+/– 12/14]. . .[[[.]]][+/– 8/10][+/–10/12][ -][-]

(Fr. 2)46 . . . made out of corporeal entities, then it studied the common affect of itself in addition, since it cannot even think of anything else besides these, whether we of bodies, or even of place by analogy . . .

The following primary questions arise from Fr. 2: (a) What is made out of corporeal entities? (b) What is the subject of  in line 2, that is, what studies the common affect of itself? (c) What is a common affect of itself? (d) What is the reference of [] at the beginning of line 3? Considering (b) and (c) first: as opposed to those who take  to be the subject of , I agree with Laursen that  seems to be a more likely candidate.47 The distinction between the rational () and the irrational () parts of the soul seems to have been a genuine Epicurean distinction48 – in Lucretius’ terminology between the animus and the

12

Self-awareness

anima – and thinking is located in the rational part, that is, it is not the capacity of the entire . Nonetheless, with such a solution it is hard to make much sense of  . . ., since if the mind studies its own affect, what is the affect common with? In my understanding the overall sense is that the  studies the common affect of the  and the body, that is, the affect they both suffer and so have in common. Thus,  might seem to fit better as the subject of the clause if we take  . . . meaning the common affect of the (whole of the) soul and the body, but by the application of the above distinction the logical subject is . How can we establish the understanding of . . . as the affect of the soul in common with the body? In the Letter to Herodotus 63–4, where Epicurus is talking about the conjoined birth of body and soul, he uses the words  and  to express some kind of genuine coaffection between the two. He says that it is the fine texture of the soul, which makes it liable to co-affection with the rest of the aggregate, i.e. with the body, and it is also its fineness, which makes it interactive with the body. The soul seems to have the major responsibility for sensation on account of the peculiar  or  of its elements;49 nonetheless the soul would not have the capacity for sensation without the body. They both share sensation as some kind of accidental attribute and the soul can bestow sensation on the body due to their contiguity and co-affection. Consequently, they need to share the resulting  of sensation and have them as something common. This point is nicely elucidated by Lucretius: Besides, you can see that the mind is affected jointly with the body and shares our bodily sensations. If the frightful force of a spear, when it penetrates to wrench apart the bones and sinews, fails to strike at life itself, there nevertheless follows relaxation and an agreeable descent to the ground, and on the ground a turmoil which develops in the mind and at times a halfhearted will as if to rise. Hence the nature of the mind must be corporeal, since it suffers under the impact of corporeal spears. (DRN III 168–76)50 To turn now to questions (a) and (d): [] might be connected with  in line 3, but that is rather uncertain.There are two further options: we can either think that it refers back to something which is lost before the passage, or take the  and the  together as its reference. The text seems to make the best sense if we understand it as the  not being able to think anything else beside these, that is, what is made out of corporeal entities, presumably the object of thought, and its resulting affection which the body and soul suffer as a common affection. Accordingly, this passage is about how the  functions in perception or thinking after the body and soul have been affected. Such discussion, as we shall see, suits the context of self-reflective thinking very well.51

Self-awareness

13

Fr. 3 []              [[]]     []    []     []   [ ][]  [+/– 5/6].[.]..[+/– 7/8][.].[ ]  [ ] [.][ +/– 7/8 ][ ] [ Fr. 3 = 1056 corn. 3 z. 3)  [] Németh: [] Laursen, (pars sin., possis ) [..] P, [.][.][..] O, ipse supplevi, sed quae praebet O fortasse aliud suadent; [] quoque possis, quae supplerunt edd. pr. cum eis quae restant non congruunt.

(Fr. 3)52 . . . it will be said 53 that he thinks of himself by means of himself in accordance with the similar and non-different, as if the thought is some unity with this, but also towards himself from other things [i.e. what are caused by other things] by means of himself [i.e. by his pathos/pathe ] , as he is said to think of himself in some affection . . .

As Laursen notes, reading [] (‘out of ’) instead of [] (‘within’) in line 3 is also possible, since it is uncertain whether a kappa or a nu is written on the papyrus. I suggest filling in the missing bit with the genitive plural of the definite article, that is, with ‘’, which palaeographically should not cause any problem and read ‘ []’ (‘from [..]’) seperatim instead of []. Furthermore, if we put a comma after  in line 3 instead of a full stop as in Laursen54 and take the next clause beginning with  ’ as a parenthesis connected to   []  in lines 2–3, then the text becomes more coherent. I take  []  to carry the same meaning as , which is taken as an instrumental dative throughout the fragments on self-reflective thinking.55 On the strength of the new reading, from  the text concerns the same topic, that is, self-reflective thinking through one’s pathos. Accordingly,  is qualified in this part of the fragment by pathos and the most likely sense of someone thinking of oneself by means of oneself from is thinking of oneself by means of/in one’s own /. I think Fr. 5 corroborates this understanding very well:56 Fr. 5 [] [][ ]’ [ ] []         ’         [[]]   [.] [[]] [] ,  [] [] [][][][.., ] [] [][’]  [+/– 8/10][ ][+/– 8/10] .[.][ ][+/– 9/11].[.] (Fr. 5)57 . . . he [/it] will 58 be said that he thinks but rather in the way in which I expressed it [above], in such

14

Self-awareness a way it is said to think of itself by means of itself, (for otherwise every living being would be unable to sense itself and it would be difficult to get a reasoned consideration of these things) – not so as to mark with sensory recognition, but I say, in this way, by which . . .

I take it that Epicurus refers back to his earlier explanation of what it means to think of oneself by means of oneself, and the point of reference may plausibly have been the second part of Fr. 3. This point, on its own, does not strengthen my understanding of , but the qualification of  in lines 2–3 of Fr. 5 – the translation in the brackets – certainly seems to do so. If a living being could not think of itself by means of itself, it could not even sense itself. The strongest candidate for that without which living beings would not be able to sense themselves is their pathe . 59 We find further evidence in our sources concerning living beings becoming aware of themselves through themselves, so to speak, through the experience of their own affections, and by such experiences becoming aware of the external realm. These texts reveal that Epicurus conceived of pathe  as the means of psychological development. There is an incidental reference to such rational development in the Letter to Herodotus, though at this point of the text without explicit reference to the affections: Again, we must suppose that nature too has been taught and forced to learn many and various lessons by the things themselves, that reason subsequently develops what it has thus received and makes fresh discoveries, among some tribes more quickly, among others more slowly, the progress thus made being at certain times and seasons greater, at others less. (Ep. Hdt. 75)60 The simple reading of this passage is that humans gradually develop their understanding of the outside world from frequent encounters with things and the information received in this way helps them to advance their knowledge, in different tribes at different speeds. Epicurus says that things themselves must have instructed human nature.61 To understand this statement one needs to take into account the fact that, on the evidence of Epicurus’ theory of perception, every event in the world can be described as a causal interaction between bodies. That is to say, our perception of the external world is characterized as a causal interaction between us as receptive bodies and the world as the source of external atomic movement which conveys information. The objects of perception affect us in the form of constant atomic streams. Talking in the pathologikos tropos, the reception of external information, which is based on direct bodily contact, instructs human nature by causing affections in our bodies. The ‘teachings of things and the forced lessons’ thus are to be understood as direct impacts upon the living beings, causing pleasant or painful affections either by the immediate impact of the things themselves or by their atomic flux of films or eido l a in the case of vision, or by atomic streams in the case of hearing.They are

Self-awareness

15

all causally responsible for our interior stimulation, according to the various processes of the different modes of sense-perception. The resulting, specific pathe , in turn, contribute to the direct awareness of both the character of one’s immediate affections and the source of the relevant kind of sense-perception.62 Moreover, the general concern of the passage for social progress and the following paragraph on the origin of language together suggest that it is the nature of primitive human beings which is taken as the basis of development. The hypothetical edge of Epicurus’ language, when he says ‘we must suppose’ that things happened in such and such a way can be simply understood as imagining the very beginning of the history of the human race, what their initial state must have been like – of which, naturally, we have no evidence. Nonetheless, it also illuminates individual psychological development in general, which starts for everyone by becoming self-aware through one’s pathe . 63 The following discussion about the origin of language actually provides further substantiation to such an understanding: Thus names too did not originally come into being by coining, but men’s own natures underwent affections () and received impressions which varied peculiarly from tribe to tribe, and each of the individual affections () and impressions caused them to exhale breath peculiarly, according to the racial differences from place to place. (Ep. Hdt. 75)64 Epicurus envisaged a naturalistic starting point for language, where ‘men’s own nature () underwent affections and received impressions’ as a consequence of which they uttered distinct sounds instinctively.65 The causal connection between the external things and the consequent reaction to the impressions and affections they cause in us is explicit in this passage, and I think it expresses Epicurus’ conception of the awareness of our internal constitutions in relation to the external world, formulated by primitive men in impulsive utterances. Just as in the previous case, I think this instinctive reaction of the first human beings in the form of meaningless utterances can also be carried over to illuminate individual cases of behaviour and psychological development of human beings immediately after birth.66 If this is correct, it clearly demonstrates that, according to Epicurus, humans become aware of themselves through their affections or pathe , conceiving of their own selves gradually by means of their pathe  and, thus, strongly related to the external realm. It is their pleasurable or painful affections, which make them utter sounds, i.e. it is their pathe  they react to by articulating distinct voices. That is to say, to become aware of something concerning either his or her internal state or an external object or the relation between the two, a person needs to undergo some affections and eventually to recognize them as his or her own. Needless to say, I do not mean to attribute to Epicurus the invention of some kind of Cartesian cogito. Rather the point is that Epicurus saw an essential connection between pathe  and one’s progressive expansion of awareness of one’s self

16

Self-awareness

and that of the external realm. Since in his view we are not born with full-blown rationality, intelligence must be an acquired property of the human being. We have ‘seeds’ to develop our rationality as a result of external sensory stimulation which acts to provide intelligible information. Just as after the naturalistic beginnings of language, people started to express themselves more concisely, we progressively expand our rationality and the organization of our internal and external data. How we learn to speak is very different from how primitive men did given the social environment we are born in. Nonetheless, how the external world affects us, both our natural and social environment, is structurally similar for everyone, at least within the boundaries of a tribe or a society. Even if a society is constantly changing and the content of our conception of the external world can also change with it too – consider Epicurus’ flexible conception of justice based on the temporally determined notion of usefulness in mutual associations (K. D. 37–8) – the way language originates from our experiences and our consecutive agreements depends upon how our psychological development is affected by the external world, just as in the case of primitive humans. That is to say, how we learn to speak – which makes use of both natural and conventional names – depends on our common or at least similar experiences of the external world. How the external world affects us in common guarantees that our words share the same references. This point will gain further support by our consideration of the empirical formation of conceptions. The general idea of rational development is nicely brought out in Maximus of Tyre’s discussion of Epicurus’ conception of pleasure: All other things that are pursued by men are either accepted because discovered by experience or respected because tried and tested by science, or trusted because examined by reason, or welcomed because they have stood the test of time; but pleasure, which needs no assistance from reason and is more venerable than any human science, forestalls experience and does not need to wait for the passing of time; human affection for it is overwhelming and as old as body itself; it is set down, so to speak, as the foundation-stone of the survival of the species. If anyone removes it, all that has accrued must inevitably vanish. Knowledge and reason and the very thing that all harp on about so, intelligence itself, men have assembled for themselves over time, gathering them through experience, by gradual enrichment of their perceptual encounters; but pleasure comes to them naturally and without teaching, and they have it at their disposal from the very start. Pleasure they embrace, pain they strive against; pleasure ensures their survival, pain destroys them. (Oration 32.2)67 The cumulative aspect of acquiring knowledge in this passage clearly designates not only the general tendency of the human race towards increasing comprehension, but the same tendency of the individual as well. Comprehension increases because of the ‘gradual enrichment’ of perceptual encounters, which is primarily responsible for the constitution of one’s prole p seis or pre-conceptions,

Self-awareness

17

without which knowledge would be impossible. Unlike knowledge, pleasure comes effortlessly, men naturally ‘embrace’ it while shunning pain, in Maximus’ poetical reformulation of the pleasure principle. I hope to have substantiated, with the help of these passages, the central characteristics of mental development and to have corroborated my understanding of the second part of Fr. 3, that is, that self-reflective thinking in some affection means self-reflective thinking by means of one’s pathe. It is a matter of controversy, however, whether Epicurus conceived of affections as only constitutive of one’s awareness of one’s own sensory states, i.e. affections are constitutive of nothing else over and above one’s awareness of physiological states in one’s body, and, consequently, whether he denied that aisthesis  and pathos are essentially connected.68 Alternatively he may have conjoined aisthesis  with pathos and held the general sensationalist view that every perception is also an affection,69 and, thus, affections are criteria not only for action but for truth as well.70 In the next, and last, section of my reconstruction of the pathologikos tropos I will investigate this question using the example of hearing, which will also open up the interpretative issues of the Epicurean concept of epaisthes is. I will argue for the general sensationalist view, which I find is in harmony with Epicurus’ second rule of inquiry, and I will support my understanding with a papyrus fragment from an anonymous Epicurean. David K. Glidden, the adherent of the first interpretation, sharply distinguishes between the different objects of the various senses and corporeal feelings and claims that the latter is not really an aisthesis  at all according to Epicurus. The affection, for example, in one’s nose, is not a perception of what the person is smelling; it is simply the awareness of pleasure or pain in the nose. The pleasurable or painful affection gives a report about one’s condition in the nose but not about the object of sense, the odour. ‘There is no place for any intentional attitude within one’s sensory consciousness in terms of Epicurean pathe . There are simply the atomic blows constituting pleasures and pains.’71 I think Glidden is correct in saying that pathe  do not make use of intentional attitudes, but that is not a sufficient reason for not conceiving them to be the essential elements of sensations. Glidden takes Epicurus’ understanding of hearing to be decisive for his interpretation; hence I start my investigation concerning this question, which is strongly connected to the nature of Epicurean self-awareness. The key text is Ep. Hdt. 52–3: Hearing too results from a sort of wind travelling from the object which speaks, rings, bangs, or produces an auditory sensation () in whatever way it may be. This current is dispersed into similarlyconstituted particles.These at the same time preserve a certain co-affection in relation to each other, and a distinctive unity which extends right to the source, and which usually causes the sensory recognition ( ) appropriate to that source, or, failing that, just reveals what is external to us. For without a certain co-affection brought back from the source to us such sensory recognition ( ) could not occur.72

18

Self-awareness

This text distinguishes between two cases of hearing: (1) hearing something when the stream of atoms producing the hearing is continuous with the external source of the sound, and due to its co-affection and distinctive unity there is a sensory recognition (epaisthesis)  ‘appropriate to the source’; (2) hearing something, which is simply ‘outside’ the perceiver, a stream of atoms producing the hearing without preserving its peculiar continuity with its source, and thus failing to produce a sensory recognition (epaisthesis).  Lucretius’ discussion of hearing sheds some further light on this distinction: When no great distance separates the source of each utterance from the hearer, the actual words are bound to be clearly audible (plane exaudiri), with every syllable distinct; for the vocal particles maintain their arrangement and configuration. But if the intervening distance is excessively great, it is inevitable that the words become indistinct and the voice distorted in the course of their lengthy flight through the breezy air. Consequently you receive an impression of the sound without being able to distinguish the signification (sententia) of the words: so confused and hampered is the voice when it reaches you. (DRN IV 553–62)73 The term ‘plane exaudiri’ seems to be the closest expression describing the condition of epaisthesis  in the case of hearing, yet in this passage Lucretius does not seem to take pains to preserve Epicurus’ original terminology. Nevertheless, the passage clarifies that at a near distance between the source of an utterance and the person who hears it the hearer is presented with a clear understanding of a vocalization signifying the things referred to by words, while a great distance results in simply hearing a vocalization without understanding, most likely implying the lack of epaisthesis  or sensory recognition. Epaisthesis  is one of the many technical concepts of Epicurus starting with the prefix epi- and there is some question about how to understand this prefix in the case of sensation. Elizabeth Asmis has proposed that epaisthesis  is a synonym for aisthesis  and that it indicates a direct acquaintance by the perceptual organ with the external object of perception, this immediacy expressed by the ep- prefix and, consequently, the term should stand for a kind of perception that is ‘on’ the source of the perceptual data.74 There seem to be at least two concerns with regard to this interpretation: on the one hand, Lucretius’ evidence – which admittedly does not translate epaisthesis  – presents the immediacy between source and the hearer as a condition for understanding vocalization. Epaisthesis  is not the connection between a hearer and the source either in Epicurus’ text, but the connection is the condition of an epaisthe sis  appropriate to the source. Clarity depends in both texts on immediacy and, therefore, both immediacy and clarity turn out to be the conditions of epaisthesis.  The other worry seems

Self-awareness

19

to be that if we take, as Asmis does, epaisthesis  to mean such a sensation which is directly ‘on’ the source, then it sounds as if, due to this connection, such a kind of sensory recognition represents its objects more truthfully than others.75 One could try to support such an understanding by appealing to Diogenes (D. L. X 32) who attributes to Epicurus the view that ‘the fact of sensory recognitions (epaisthe mata) confirms the truth of sensations’,76 and claims that  the results of the specific class of sensation directly connected with its objects are such sensory recognitions (epaisthe mata), which construct direct bridges  between perceivers and the perceived objects, and consequently are conceived of as more truthful. To this extent they would be more accurate and therefore more precious than other sense-perceptions, which do not connect the perceivers with their source. The perception of this latter class, however, is still to be taken as true, just as even the figments (phantasmata) of the madmen are, because they cause movement and, thus, they also cause perception.77 Nonetheless, since Epicurus placed the foundation of his theory of truth on irrational sense-perception, he could not have allowed for degrees of truth in sensation itself, and Asmis herself clearly denies that it would be a proper interpretation of Epicurus’ perceptual theory.78 I think a better alternative is to understand epaisthesis  as sensory recognition, but only in a limited sense, as an exclusive capacity of the irrational part of the soul.79 Why is such a limitation necessary? First of all, I think Epicurus’ motivation for the division of the soul into rational and irrational parts was based on one of the principles of his perceptual theory, according to which ‘all sensation is irrational and does not accommodate memory’ (D. L. X 31–2). Sensation is irrational in the sense of presenting information not involving any rational activity, which is judged only subsequently by the mind, as a result of which we acquire true or false beliefs concerning our perceptions and of the external world. If the soul is to receive some information prior to judgement, then it seems to be correct to say that for the Epicureans – who did not think that the sense-organs are windows through which the rational soul sees,80 but are vitalized by the soul in a way that the sense-organs perceive non-assessed information – sense-data are registered first by the sense-organs vitalized by the irrational part of the soul.The rational part evaluates the registered information only subsequently. Thus if epaisthesis  is the sensory recognition of the irrational part of the soul, it seems to function in the sense of recognizing the represented sense-data, which are judged subsequently by the rational part of the soul. So the clarity of representation does not concern the truth-value of a given sensedatum, but rather the degree of the harmony between the sense-data and our sensory functioning. When I hear something, I can either just simply have an auditory sensation without an epaisthesis  – without recognizing what sort of thing I hear – or when I am also aware of what sort of thing I hear I have an epaisthesis  – i.e. the sensory recognition of hearing, for example, some spoken words. If we take epaisthesis  along these lines, it still remains only a basic element of recognition within perception without a rational process of identification of the heard words with the words of, for example, the Odyssey. If this is correct,

20

Self-awareness

then epaisthesis  as sensory recognition is the Epicurean term for an irrational process of our sense-data appearing to us identifiably.81 Accordingly, there is a certain physiological activity to produce hearing, which brings about the akoustikon pathos in the audience,82 and on the condition that the auditory stream reaching the perceiver preserves its distinctive unity with its source, we have a corresponding recognition of what sort of thing we are hearing and whether we are pleasurably or painfully affected. If this condition is not fulfilled, then we only hear pleasurably or painfully ‘what is external to us’, without being able to tell, for example, whether we are hearing some animals crying or humans speaking. According to Glidden, the process of hearing is something completely different:83 in his view, the awareness of one’s akoustikon pathos in the Epicurean theory did not amount to the awareness of the fact that one is hearing someone else speaking. Instead, Glidden thinks that the distinction Epicurus made is between hearing something and hearing pleasurably or painfully, the latter case merely being an awareness of the condition of one’s sense-organ in accordance with its associated pathos, but nothing over and above. Thus, Glidden’s account seems to imply that the consciousness of hearing another speaking – with or without understanding – is distinct from the state of one’s awareness of the relevant sense–organ, for example, in the case of hearing, one’s ears. It seems to me, according to his view, that if one is aware of oneself hearing, for example, someone speaking, it requires an additional kind of awareness to that of self-awareness; that is, if the awareness of hearing something and the awareness of the mode of hearing are distinct. However, if that were the case, it would be even hard to see how pathe  could be the criteria for actions at all.84 I suggest that it is correct to say that according to Epicurus one’s pathe  in life initially were nothing over and above being aware of one’s physical condition; that is, becoming aware of oneself through one’s pathe . Infants do not possess developed rational capacities; thus, if an infant, who is yet to develop his or her rationality, hears a disturbing noise and finds it painful and consequently wants to reject it, it is not that the infant wishes to reject a disturbing noise but that he or she wants to get rid of his or her uncomfortable affection in his or her body. At birth not only is pain entirely irrational for the infant but also the source of pain; he or she does not know if, for example, the sun or a nearby fire is making him or her uncomfortable, and presumably he or she could not even make a distinction. An infant can be described as having certain desires but it is only a natural instinct rather than a rational desire at this stage of the infant’s life. It is even questionable if one says that the infant wishes to get rid of the pain in his or her body. Rather, Epicurus claimed that the sensations of pleasure and pain are the first steps for an animal to become aware of what he or she would later conceive as his or her own body. That is to say, the starting points of one’s self-awareness are pleasures and pains.85 During one’s psychological development, one’s rationality expands under the continuous influence of external and internal factors. Nonetheless, one’s affections themselves still remain irrational, in the sense that they represent information about one’s non-rational physical states. However, for an adult

Self-awareness

21

being, with fully acquired rational capacities, a pleasurable akoustikon pathos in the ear realized by someone else speaking does not merely make him or her aware of being in a state of hearing, but is also constitutive of being aware of hearing something. When listening to something, what he or she hears produces some physiological process in the ears. This very physiological process in the ears is what we would call constitutive of hearing what we hear.The physiological structure causes us to have the respective sensory affections of hearing, in our case the akoustikon pathos. Affections can make us aware of their particular locations but it is not as if affections themselves make us aware of being in a particular physiological state, distinguishable from that of the preceding physiological process of hearing someone speaking. Such a theory would seem to require two different underlying atomic structures in the framework of Epicurean atomism; one for the sensation, hearing someone speaking, which in turn causes an akoustikon pathos with a different underlying atomic structure from that of hearing. But the theory does not seem to distinguish between such underlying atomic structures. The very same atomic stream which brings about hearing someone speaking constitutes a pleasant affection simultaneously. It can be rightly said that the pathos itself does not have any intentional content, but it certainly works to generate intentional content – and if there were no common physical basis for the sensation and the concomitant affection, Epicurus would have needed to account for an extra underlying atomic structure for the awareness of hearing actual sounds, distinct from one’s awareness of hearing. To be sure Glidden never claims that there are two distinct atomic structures. Nonetheless, the sharp line he draws between perceptions and affections seems to imply such a consequence. Glidden rests his argument primarily on Epicurus’ second procedural note in the Letter to Herodotus: Next, we must by all means stick to our sensations, that is, simply to the present applications, whether of the mind or of any criteria, and similarly to the affections () that obtain, so that we may have the means to infer both what is expected [to appear] and what is non-apparent. (Ep. Hdt. 38)86 He finds three aspects of this passage relevant to the contrast between aisthe s is and pathe . First, he suggests that the distinction between perceptions and pathe  seems to exclude their assimilation. Second, pathe  are nowhere described as apprehending. Third, pathe  are left out of the list of sensations to which we must attend and, thus, to be aware of one’s feelings is not an aisthe s is at all.87 I think the distinction between aisthe s is and pathe  in this context relates to the two different types of observations, relying directly on our perceptions or affections, in relation to drawing inferences about what is expected to appear or the non-apparent. Sensations and affections are both crucial for observations of

22

Self-awareness

these kinds, since they are irrational – in the sense that they represent information unfiltered by rational activity about the object of sense and one’s physical states – and, hence, they are reliable data about our external environment and internal constitutions, respectively. The distinction between them in Epicurus’ second methodological principle, however, does not exclude that having a pleasurable pathos in one’s ears because of hearing someone speaking is essentially connected to the awareness of the cause, and is not only an awareness of the physical constitution of one’s pleasure. In other words, Epicurus is not compelled to separate the awareness of sensation from the awareness of oneself through one’s pathe. We can understand him, instead, as taking perceptions and affections as the complementary contents of hearing in the sense that they are different aspects of the very same awareness of a perceiver. In my interpretation, Epicurus distinguishes between the active and passive aspects of sense-perception due to the outwardly directed, receptive sense-organs, and the concomitant affections of sense-perception, internal to the sense-organs. Epicurus’ second rule of inquiry in his procedural note supports this claim: different kinds of perceptions, whether of the sensory organs or of the mind, and affections both provide irrational information about the external world and our inner physical states, respectively, which are yet to be evaluated by reason.88 Nonetheless, there is only one, simultaneous awareness of how the external world appears to us through the internal constitution of our sense-organs; Epicurus only distinguishes the basis of inquiry into the environment, requiring different starting-points.89 It would indeed be hard to imagine in what way we could form different desires if we had not been aware of the particular connections of our pathe  to the external realm. Such an understanding has strong textual support from an anonymous Epicurean: . . . ‹›’ (), ,   .’. . . Therefore, when we say that qualities are apprehended by the affections, we do not assign apprehensions of their own to the affections, but [we assign them] to the perceptual organs through the affections; nor when [we say that] the affections are sensory recognitions of themselves . . . (PHerc. 19/698 col. 9.1–10)90 The expression ‘apprehensions’ () is certainly Stoic in vocabulary flavour, however the text itself is clearly Epicurean because of the attribution of this apprehension to the sense-organs. This evidence is perfectly compatible with my interpretation of epaisthe sis  and also supports my interpretation that the Epicureans thought there to be a genuine connection between the active and passive operations of the sense-organs – a basic requirement for choice and avoidance. And so the connection between sense-perception and the pleasurable and painful affections of a perceiver can provide the required information

Self-awareness

23

for calculating and surveying advantages and disadvantages,91 and consequently, one can lead a rational life in conformity with them. Thinking of oneself through one’s pathe  thus turns out to be an activity in which one pays attention to one’s affections in strong connection with the other aspects of the aisthe sis  of the external world. As Fr. 5 reveals from line 3 onwards,92 Fr. 5 [] [][ ]’ [ ] []         ’         [[]]   [.] [[]] [] ,  [] [] [][][][.., ] [] [][’]  [+/– 8/10][ ][+/– 8/10] .[.][ ][+/– 9/11].[.]

the activity of self-reflective thinking does not merely indicate the process of becoming aware of oneself but it also has an ethical function. If one did not conceive of oneself through one’s pathe, not only would self-perception be impossible, it would also be difficult to get an , a rational calculus of ‘these things ()’. It is impossible to tell with certainty what  refers to in Fr. 5, line 3, but given the context we can plausibly take Epicurus to be talking about the rational calculus of one’s pathe . If this is correct, then Fr. 5 seems to assert more than just the awareness of one’s self through one’s pathe: it seems that for ‘a living being to think itself by means of itself (i.e. by its pathe ) ’ the living being senses itself and can acquire a reasoned consideration of the means (i.e. its pathe ) by which it has self-awareness. That is to say it does not merely become aware of itself through its pathe , but the living being can evaluate those pathe  as its own both in relation to its own self and the external world. Such an understanding of Fr. 5 is in harmony with the ethical function of  in some other lines of the papyri, where  stands for the rational calculus based on human behaviour and affections.93 Importantly, according to Fr. 5, such a mode of investigation is not like ‘marking’ with the help of the extra element of awareness of different things, i.e. with sensory recognition ([] [] [][] []). But by being aware of the source of my pleasurable akoustikon pathos (i.e. ‘marking with sensory recognition’), hearing the sounds of somebody speaking is a completely different, more basic recognition than the inferences I may make with the help of my rational calculus concerning my pathe . , the rational calculus, is a method in which we can take into account our pathe  according to our telos in harmony with the formal requirements of the hedonistic understanding of eudaimonia. It is a reasoning concerning which pleasurable or painful pathe  are choice-worthy and to be pursued. Self-reflective thinking in accordance with the pathologikos tropos – the objective element of Epicurus’ ethical theory as far as our pleasures and pains are considered per se – paired with the essential key to one’s happiness, a calculus based on one’s prudence (phrone sis), functions as one of the most significant elements for living the life 

24

Self-awareness

of an Epicurean sage. By virtue of prudence in particular we can overcome psychological impediments. An inspection according to the pathologikos tropos paired with the corresponding classification of our desires can help us to attain the hedonistic happiness we all naturally want and strive for.94

Responsible agency: The aitiologikos tropos The evaluative element of self-knowledge seems to appear not only along the lines of the pathologikos tropos but also in accordance with the aitiologikos tropos, the ‘causal manner’ of explanation in Epicurus’ theory. The first half of Fr. 3 clearly reveals that Epicurus did not merely consider one version of self-reflective thinking. Let us return to Fr. 3 once again: []              [[]]    ’ []   [] ’  [][][] [+/– 5/6].[.]..[+/– 7/8][.].[ ]  [ ] [.][+/– 7/8][ ] [ (Fr. 3) . . . it will be said that he thinks of himself by means of himself in accordance with the similar and non-different as if the thought is some unity with this, but also towards himself from other things [i.e. what are caused by other things] by means of himself [i.e. by his pathos/ pathe ] , as he is said to think of himself in some affection . . . I agree with Laursen that in view of   in line 2 of my presentation of Fr. 3, we must assume that an   preceded,95 which makes it clear that Epicurus was concerned with a different way of self-reflective thinking in the first two lines, different that is to say from the interpretation of lines 2–4 of Fr. 3 given earlier. I think Laursen is also right to point out that  in line 2 is a reference to . That is to say, Epicurus appears to be talking about some kind of identity between the mind and its thought content. Such an identity would not be surprising if we consider the Epicurean parallelism between perceiving and thinking; to a certain extent both are the outcomes of external stimuli brought about by images of different quality streaming off objects. Perception and thinking, consequently, are both physical processes and the images of thinking, however fine they are, also possess normal bodily powers.96 On the analogy of perceiving, images of thinking make an impact on the mind, just as in perception certain images () emanate from objects, penetrate our sense faculties and imprint their nature.97 As Julia Annas has pointed out, there is no need to suppose that when in the case of perception images affect the soul atoms these images get absorbed. Instead, after they have made their impacts they pass out of the body.98 On the physical plane, perception is constituted by the rearranged atomic motions of the sense-organs upon such impacts, producing the relevant phantasia.99 In the case of thinking, as Annas notes, such external stimulation is needed much less due to the extreme fineness of the mind and of the images of thinking than to the dispositional characteristics of

Self-awareness

25

memory: ‘some images affect us in such a way that we are prone to repeat the thought of their object’.100 On the analogy of perception, thinking is also the result – at least to a certain extent – of external stimuli striking the mind atoms, which on the physical level can be described as the rearrangement of the mind atoms according to the imprints of the external images. In other words, if the rearrangements of the mind atoms brought about by the stream of atomic images or eidol a are constitutive of the thought content, there is some unity between the two – even if the mind itself cannot be identified with its content. Nonetheless, and most importantly, the mind also has an active power of focusing ( ), which Annas has aptly described as a power of the mind without which the mind would be overwhelmed by the large amounts of thought images readily available. This active characteristic of thinking, thus, helps us to overcome the problem of image overload.101 How is this unity comparable, as  in the first line of Fr. 3 seems to suggest, to self-reflective thinking by one’s own means ‘in accordance with the similar and non-different’? To understand the comparison, first we need to see what Epicurus may have meant by  in the first line of Fr. 3.102 The knowledge of self through others appears prominently in Plato’s (?) First Alcibiades, 132–3C. Socrates interprets the Delphic imperative ‘Know Thyself ’ using the analogy of perception as the order ‘to see yourself ’, indicating looking at something in which we can see ourselves. He asks Alcibiades whether he has ever noticed that when someone looks at the eye of another he sees his face in it as if in a mirror, thanks to looking at the place where the excellence of the eye is generated, that is, in this case, sight. Then Socrates applies this to our knowledge of ourselves: if a soul is to know itself, it must look at another soul, and especially at the place where its wisdom is generated. Plato’s discussion seems to have influenced Aristotle in many ways.103 Aristotle in the Nicomachean Ethics IX 1169b3–70a4 says that one reason why we need friends is that we can more easily look at them than ourselves, and at their actions than at our own, so for good and just men the actions of good and just friends are equally as pleasant as theirs, because sharing in these characteristics they can view and take pleasure in their friends’ actions as their own. Epicurean self-reflective thinking in accordance with the similar and nondifferent fits well into this historical context, especially if we introduce another fragment from an earlier part of book XXV: Fr. (a) . . . [ ]    [ ] {}  []‹›  []    []   []104    []  [ ] [ ]. (∗) .[.]  ’ [.][+/– 2/3][+/– 3/4][ (Fr. a)105 . . . that the soul and the rest of nature make the living being, and which we said to consist of the rational and remaining nature – which is thought to be a unity – thinks nothing but the other . . . not in the eye . . .

26

Self-awareness

This fragment lends support to the suggestion that we should interpret Epicurean self-reflective thinking (Fr. 3) in the historical context of self-knowledge through others. The chance (?) occurrence of the word ‘eye’ in the context of Fr. (a) is startling (cf. Plato’s (?) First Alcibiades 132–3C), however we cannot make much use of it in our interpretation. The expression ‘[that] which is thought of in unity ( [])’ seems to address directly what we would describe by the modern concept of self, denoted in this fragment simply by [], ‘the living being’. The expression ‘ []’ opens an alternative interpretation for the clause ‘as if the thought is some unity with this (    [[]] )’ in Fr. 3. Accordingly, we could rather understand it as ‘as if that which is thought is in some unity with this [i.e. with the mind]’. In light of Fr. (a), the relevant part of Fr. 3 could be translated as ‘he thinks of himself by means of himself in accordance with the similar and non-different as if that which is thought is in some unity with this’, and interpreted in the following way: one can think about oneself by means of oneself, most likely through one’s memories of a similar and non-different living being, whose bodily appearance as a living being is thought to be in a unity with this [i.e. the mind, resulting in a unity between body and soul]. Since the conception of a living being must be based on its visual perception,106 the comparison would make sense: our thinking of such living beings comes from sense-perception and the memories thereof, and we see the perceived living beings primarily as acting in certain ways. Based on our visual experience of their behaviour, we think of them as living beings having a soul, that is to say they appear to us as a unity with this, i.e. with the mind. Based on Fr. (a) it would mean that their living being, their self, is conceived of as a unity of body and soul, and that should help one, according to Epicurus, to think of oneself and conceive of his or her self in analogy with them – in the aitiologikos tropos – as responsible agents.107 At this point, this is still a very speculative understanding of Fr. 3, but there is in fact much evidence that supports this interpretation.Therefore, I devote the rest of this chapter to the reconstruction of self-reflective thinking in accordance with the aitiologikos tropos in harmony with the available evidence. For that we need to take into account Epicurus’ notion of the prolepsis  of our own responsible self, found in the later parts of book XXV (On Nature, book XXV PHerc. 1191-10 sup. 5). First of all, we need to consider Epicurus’ conception of prole psis  in general to make good sense of the prole psis  of one’s responsible self in particular in my interpretation of Fr. 3, and in my overall argument. We will pursue the discussion of prole psis  at significant length in order to reconstruct the Epicurean mechanism of thinking, so that we can not only build a possible interpretation of Fr. 3 but also understand the fundamental process involved in self-reflective thinking. The following discussion addresses a diffuse set of texts, so it may be helpful to outline its structure. First I will state briefly my concerns about previous interpretations of the Epicurean prole psis  and propose my understanding, which I will substantiate mainly by the analysis of two texts. One of them is part of a fragment from Diogenes of Oinoanda, which contains a detailed description of Epicurus’ theory of perception. I will utilize the findings of this analysis in my following interpretation of our main evidence on the Epicurean prole psis, that is, 

Self-awareness

27

Diogenes Laertius’ account at D. L. X 33. I will discuss his evidence in two waves: first I will present its general structure with some immediate questions and comments and second I will scrutinize some of its terminology. My final interpretation will take into account the elements and mechanisms of thinking, in which prole p sis performs a major role. To clarify its proper function, I will need to discuss how Epicurus’ method of verification and falsification and his conception of the prole psis  relate to each other. Once I have presented my comprehensive account of what I take the Epicurean prole psis  to have been, I shall not only corroborate it using some primary evidence but I will also apply it in the exegesis of the fragments in which we find Epicurus’ notion of the prole psis  of our own responsible self – and I shall call this group of consecutive and cohesive fragments Text P: P for prolepsis.  In order to substantiate my necessarily conjectural interpretation of Fr. 3 on the aitiologikos tropos, I will also take into account the major fragment on memory from the later parts of On Nature XXV – and I shall call this fragment Text M (for memory, naturally). After presenting the robust theoretical background for my eikos logos of Fr. 3, I will conclude this chapter on Epicurean self-awareness with a brief look at the last two fragments I take to be the parts of Epicurus’ discussion of self-reflective thinking.

The elements of thinking Prolep  sis The Epicurean notion of prole psis, literally meaning ‘taking before’, resists a  clear interpretation because of the ambiguity of our primary and secondary evidence. There are six instances of its use in Epicurus’ extant works, in quite different contexts, and although none of them gives a definition of prole psis,  some important conclusions can be drawn concerning the meaning of prole psis  with the aid of our secondary sources on the matter.108 There seems to be an irresolvable uncertainty about how to conceive of prole psis  in general: as mental acts, images or propositions, or a combination of all three?109 Since prole p sis is one of the three standards of truth (see D. L. X 31), in reference to which we express our beliefs, it seems to be very unlikely that we refer propositional beliefs to propositions. A proposition can either be true or false and if we express our beliefs in reference to propositions, which themselves need justification, we may end up with an infinite regress and could not count prole p sis as a criterion of truth. The image interpretation also sounds very weak for several reasons, one of which is the problem of how to conceive the prole p sis of abstract concepts as images.110 That prole p sis would be a sort of mental act has not been clearly expounded in the literature, since its interpretation along this line has been mixed with other difficult Epicurean concepts and also, since Epicurus’ division of the soul into rational and irrational parts,111 has not been taken into consideration exhaustively. I wish to propose that the original conception of the notion was neither of the first two above, but Epicurus himself thought of prole psis as the means of 

28

Self-awareness

recognition in the rational soul. I wish to argue that: (1) the proleptic process of recognition was an activity prior to the rational judgement of the soul (hence the word prole p sis or ‘taking before [i.e. before the mind]’)112 and that (2) prole p sis fulfilled its function in recognition as a mental process that unifies the various proper contents of the individual sense-organs in the rational part of the soul113 – or it complements some of the missing information if an object is perceived only by one of the sense-organs. For this understanding, it is of great import to understand how the sense-organs function in perception, the basis for the entire mechanism of thinking for the Epicureans. As Lucretius makes clear, every sense has its proper perception that cannot be corrected by the other senses because every sense has its own sphere of discrimination (see DRN IV 479–99).The shape perceived by sight, for example, is not the shape one differentiates by touch. The former is merely the outline of colour, an effect brought about in the perceiver; thus, sight does not represent the actual shape of the object but only that of colour.114 That is to say, sight does not merely discriminate colour but everything related to colours, for example, shapes, distances, and so on. The apparent tension of a possible contradiction between sight and touch representing the same object differently – e.g. a stick seemingly bent in water but appearing straight to the touch – is only the result of a mistake based on a false analogy, as an anonymous Epicurean pointed out, given that touch discriminates the shape of the body, while sight only that of colour.Thus, the senses do not have a common sphere of discrimination and, as Hahmann has observed, this leads to the difficulty that the individual senses perceive only phenomenal objects of their own discrimination and not the object that is the source of these appearances.115 Therefore, there needs to be something which unifies the proper perceptions of the different sense-organs and prole psis  seems to be the best candidate for such a function. It sounds, nonetheless, somewhat problematic to think that prole p sis, which is itself a result of repeated former perceptions, is what unifies one’s perception. In order to understand how that is possible, we need to pay close attention to two texts: that of Diogenes of Oinoanda for some detailed description of the mechanism of Epicurus’ theory of perception, and that of Diogenes Laertius for the most abundant evidence on Epicurus’ conception of prole psis.  Diogenes of Oinoanda (second century CE) promoted his master’s teaching in a stone inscription, carved onto a portico wall in the city of Oinoanda in Lycia. The inscriptions included topics on all three major branches of Epicurus’ philosophy – physics, epistemology and ethics – originally consisting of about 25,000 words, less than a third of which has survived. The crucial passage relevant for our investigation comes from Fr. 9, from within the context of two longer fragments (Fr. 9–10), which seem to be strongly connected to each other by their main topic, Epicurus’ conception of dreams. (1) Now the images that flow from objects, by impinging on our eyes, cause us both to see external realities and [through entering our soul to think of them. So it is through impingements] that the soul receives in turn

Self-awareness

29

the things seen by the eyes; (2) and after the impingements of the first images, our nature is rendered porous in such a manner that, even if the objects which it first saw are no longer present, images similar to the first ones are received by the mind, [creating visions both when we are awake and in sleep.] (3) [And let us not be surprised] that this happens even when we are asleep; for images flow to us in the same way at that time too. How so? (4) When we are asleep, with all the sense-organs as it were paralysed and extinguished [again in] sleep, the soul, which is [still wide] awake [and yet is unable to recognize] the predicament and condition of the sense-organs at that time, on receiving the images that approach it, conceives an untested and false opinion concerning them, as if it were actually apprehending the solid nature of true realities; (5) for the means of testing the opinion are asleep at that time.These are the sense-organs; for the rule and standard [of truth] with respect to [our dreams] remain [these].116 (Fr. 9 II 9–VI 3) This Epicurean description of how perceptual and cognitive processes are connected fills in important gaps in the theory, which are rather vaguely described in the surviving primary evidence of Epicurus and even in the relatively detailed analysis of Lucretius. First of all, the quoted passage contains explicit information concerning the soul taking over () the images of the eyes in (1). This means that however fine the images of thinking might be compared to the images of sensation, there are some from sensation itself, which become direct material in the activity of the rational soul.117 In (2) we learn that the images of sensation open up channels between our sense-organs and the soul, and that they render our nature porous ( . . . ) in such a way that the soul, maintaining the continuity of the visualization of dreaming, is receptive to similar images after the disappearance of the original ones. The capturing of similar images seems to be based on the focusing function of the mind ( ) both when we are awake or asleep; but the physical traces brought about in us by the images initiating the particular perception also seem to set a physical condition for the continuity of imagination and therefore for the selection by similarity: only those kinds of images will be admitted which fit one way or another into those physical traces.118 There is, however, a question of whether the images of thinking are finer per se than the images of sensation or whether they become finer because of their journey between our sense-organs and the soul.119 The former option is more desirable if we want to make sense of how images can enter through one’s body in dreams when the person is asleep but the mind is still active: since these dream-images do not affect the mind directly but travel through the body (see DRN IV 749–56) and flow to us in similar ways as they do when we are awake (see Diogenes (3): . . .       []  . . .), it makes the best sense to suppose that it is due to their fineness that the images can enter through the paralyzed and inert sense-organs (see Diogenes

30

Self-awareness

(4): . . .   . . .), and that they stimulate the soul after they have reached it through the physical traces in us. That it is the rational part of the soul in question becomes clear from the fact that the ‘soul . . . conceives an untested and false opinion concerning them’ ((4):  [] . . .    ). That is to say, the soul is unprotected from making false judgements when we are dreaming because it conceives dream images ‘as if it were actually apprehending the solid nature of true realities’. I will frequently refer to these findings in the following analysis of Diogenes Laertius’ characterization of Epicurus’ conception of prole p sis to which I now turn. (1) Prole p sis, they [the Epicureans] say, is something like apprehension (), or right opinion ( ), or conception (), or a stored general thought (), that is, a memory of that which has often appeared from outside, for example, that man is this sort of thing. (2) For at the same time that “man” is spoken, immediately in accordance with prole p sis the outline ( ) of man also is thought of, as a result of preceding perceptions. (3) In the case of every name, then, that which is first subordinate is evident. (4) And we would not have sought what we seek, if we had not previously been aware of it. For example, is the standing thing in the distance a horse or a cow? (5) We must have learned at some time by prole p sis the form of a horse and a cow. (6) Nor would we have named anything if we had not previously been aware of its outline ( ) by prole p sis. (7) Prole p seis, then, are evident. (8) Further, an object of belief depends on something prior that is evident, by reference to which we state [the belief]; for example, how do we know whether this is a man?120 (D. L. X 33) First, let me spell out the structure of the passage with some immediate questions and comments. doxa orthe , ennoia and noesis. (1) Prole psis    is something like () katale psis, (2) For once the word ‘man’ is uttered, immediately by means of prole psis its  typos comes to mind, as a result of preceding perceptions/‘in which/since the senses take/give the lead’– in Hick’s 1931, Morel’s 2007 and Fine’s 2014 alternative translations. In other words, the reason why prole p sis can be characterized as something like katalepsis, doxa orthe , ennoia and noe sis   is that as soon as we hear the word ‘man’ there is an automatic process bringing the outline of ‘man’ to one’s mind, and this is a process in accordance with prole psis. Prole p sis seems to be the motor of  the process or, rather, the following process itself when hearing an utterance

Self-awareness

31

and realizing the outline content of that utterance. The question is: in what ways can katalepsis, doxa orthe , ennoia and noe sis   be described similarly to prole p sis as catalysts of such processes? (If the reconstruction of prole psis  as a mental process is correct, it already seems sufficient to refute the image-like interpretation of prole p sis.) (3) Thus the thing primarily denoted by every name is enarges. Why? The nominal phrase in the following sentence from (3) – ‘ ’ – immediately brings to mind the first principle of methodology in the Letter to Herodotus121 and the related questions concerning Epicurus on meaning. It is reasonably clear because of the context and (7) later that what are subordinate to names are prole pseis. According to  Glidden, Diogenes plays on an ambiguity, however, when not clarifying whether prole pseis are subordinate to names as their meaning, that is, when uttering ‘man’  we signify some conception of ‘man’; or prole pseis are subordinate to names in  the way that our words fundamentally refer to real things we can name when talking about them, that is, when uttering ‘man’ we signify some perceived form in the world.122 Stephen Everson has shown convincingly that Plutarch’s and Sextus’ evidence makes it clear that for the Epicureans it is words themselves that have semantic properties and not something underlying them.123 It would have been unnecessary to introduce a set of meanings for Epicurus in order to show how words are meaningful because for words or sentences to be meaningful for him there do not need to be separate sets of corresponding meanings over and above the words themselves. Thus as Everson concludes: ‘. . .we should cautiously accept that the semantic values of the words a speaker utters are a 124 It seems to me that Everson’s solution has the merit function of his prole pseis.’  of bringing out the basic function of prole psis  in recognition. But how are we to understand what exactly the underlying things are? If we take it that prole p seis are subordinate to every name, Diogenes’ argument reads as follows: (1) Prole p seis are like such and such things. (2) For by prole p seis we get the outlines (typoi) of things as soon as their names are uttered. (3) Thus the primarily denoted prole pseis are evident.  Point (2) does not say that the products of the cognitive process of noe sis  are prole p seis, nor that we think of prole p seis, but that by the means of prole pseis we  think of certain typoi, some representations of the general characteristics of things (     ). That is to say, what is enarges is the process producing these representations. Furthermore, if this is correct, the typos certainly cannot be the representation of the sense-organs, but rather that of the rational part of the soul.

32

Self-awareness

I wish to argue that perception is a tripartite transitive process starting with the activation of the sense-organ: in Diogenes’ example, the uttered word ‘man’ stimulating one’s ears vitalized by one’s irrational soul, which given the sensory recognition of the word ‘man’ initiates a proleptic process of recognition in the rational part of the soul presenting a typos of the perception for the rational soul for judgement. Such a proleptic process is only initiated by a sensory recognition (epaisthesis),  which is a recognition limited to the proper object of the specific sense-organ or, alternatively, the proleptic process is initiated because of the stimulation of many simultaneous sensory recognitions of the relevant sense-organs. The consequent proleptic process either complements the information present through one of the sense-organs – in the case of hearing the word ‘man’, the proleptic process supplies all the relevant extra information for its outline representation (typos) – or it connects the separate recognitions of proper objects of each sense-organ into a holistic representation in the form of a typos. And there is also the possibility that there is no sensory recognition (epaisthesis),  e.g. hearing something without understanding it or without knowing what it is, which does not initiate a proleptic process, but instead, as I will argue, leads to one making the judgement of hearing something not on the strength of a proleptic process but with the help of one’s separate stock of conceptual apparatus. It has been often pointed out that the word prole psis sounds like an active  substantive, the act of ‘grasping in advance’, indicating a cognitive operation or is described in (1) as a ‘memory of that a movement of thought.125 Prole psis  which has often appeared from outside’ and as we shall see in a major fragment concerning memory (Text M later on pp. 48–9), Epicurus took memories to be analogous to affections or physical movements of the soul. I think these movements of prole p seis can be considered to be activated by the physical traces of sense-perception.126 These physical traces of sense-perception in us seem also to help generalize the content of sense-perception by the fact that the senseorgans can only maintain the flow of perception if they let in similarly sized images which fit their channels or physical traces.127 Thus, the physical condition for the selection between images inevitably has to involve a selection by some kind of similarity, otherwise the continuous flow of perception could not be maintained. Lucretius’ evidence confirms that this selection has a basis in physical structures (see DRN IV 973–83, especially 976–7). If we accept that prole pseis are activated through these physical traces in the sense-organs, which  are the direct outcomes of sense-perception, then it would explain the clarity of prole pseis, which were originally formed by their dependence on similar  physical traces, imprinted in us during clear acts of sense-perception, testifying directly and physically to the actual things sensed. Given that there is a natural continuity between sensation and prole psis  (see (2)), the clarity of the former is carried over to the latter. How is the immediate association between the uttered word and the consequent proleptic process in (2) guaranteed? If we connect the thesis of psychological development set out in the passages on the origin of language – the way people have been trained and taught by nature, leading them to utter words

Self-awareness

33

based on the lessons learnt from their natural and social environments – with prole p sis as a process coming about as a consequence of many previous senseperceptions leaving physical traces in us, then we can see that owing to the essential connection between experience and the natural formation of language, prole p seis are something like memories of former experiences and are directly associated with one’s acquisition of the ability to speak the language of one’s social environment. It should be pointed out that the acquisition of language now as opposed to in the time of primitive men is a cultural phenomenon, but the theory seems to take it for granted that the condition for acquiring the language we speak is that our natural and social environment affect us in very similar ways.128 Word-use and understanding are shared by individuals in the way that the prole p sis associated with them has a similar causal history in all users within a community.129 Therefore,‘’ in (2) seems to stand for the natural, empirical encounters which determine proleptic processes resulting in representations for the rational part of the soul, and these natural encounters or experiences are the guarantees for the representations being clear or enarges, ‘since the senses give the lead’.130 (4) Inquiry into something presupposes prior awareness (egnok eimen) of the inquired thing. (5) We had become aware (egnokenai) of shapes (morphe ) of things formerly by means of prole psis.  (6) Naming presupposes learning the outline (typos) of things by means of prole psis.  (7) Thus prole p seis are enargeis. In this part of the argument (6) makes it clear that prole pseis are pre-verbal131  and, thus, the wide consensus that prole pseis are true beliefs – given Diogenes’  characterization of prole pseis as doxa orthe , seemingly corroborated by various  passages (Ep. Men. 123–4; On Nature 28)132 – is seriously questionable.133 However, if we accept that the lessons of nature are responsible for the processes of sensation, which create physical traces in us, based on which prole pseis come  about by repeated sense-perceptions, by means of which we learn the outline (typos) of things (see (6)) and, in turn, we record the constantly and, to some extent, coherently uttered sounds in relation to the objects of those repeated sense-perceptions, which utterances given their utility become fixed as names,134 then we are one step closer to understanding how in the paradigmatic case of vision phantasia translates into a linguistic phenomenon. It seems that it does through the products of proleptic processes, the typoi. (8) The object of the doxaston depends on something prior and evident, by reference to which we state the belief.135

34

Self-awareness

With this general outline in mind, let us look now at the Stoic terminology Diogenes uses in his analysis of Epicurean prole psis. The above sketch seems to  make it clear that prole p sis is at most ‘something like’ () all the things he lists. I also prefer to stress that prole p seis are just ‘something like’ () them because the parallel terms stand for different things even within the Stoics’ terminology. Let us look for in what sense prole p sis is similar to each of these Stoic terms listed by Diogenes at (1) to further support my thesis of prole psis  as being a process of recognition unifying the various proper contents of the individual sense-organs in the rational part of the soul. The Stoic use of katale psis  signifies an assent to an apprehensive presentation, necessarily displaying its cause exactly as it is; that is, a necessarily true presentation of a thing. Put in terms of Epicurean prole p sis it indicates that what one thinks of is necessarily real, as Asmis put it, and that prole psis  necessarily provides an accurate representation of the general characteristics of the thing being thought of.136 Such an account is in complete harmony with my understanding of prole psis  as being essentially connected to the physical traces of sense-perception in us, on the basis of which the evidence and the accurate representation of the general characteristics of the perceived thing are guaranteed at the end of the proleptic processes of recognition, in the form of some typoi.137 Doxa orthe  is somewhat curious if we wish to preserve the function of prole p sis as a criterion,138 and it is in inherent tension with (8). The object of the doxaston depends on something prior and evident, by reference to which we state the belief, and it is hard not to notice the parallel description of prole psis  in (3) and (7) as evident, not to mention its priority – thus the answer to the question what the doxaston depends on should be obvious. However, even if we accept Diogenes’ description, the most that seems to follow is that there seem to be, according to Epicurus, some self-evidently true opinions. Ennoia is one of the main elements of Diogenes’ text which leads to a major ambiguity in the overall understanding of Epicurean prole psis. As Glidden has  pointed out,139 there seems to be an important difference between the Stoic and Epicurean understanding of the term ennoia. Glidden uses Sextus’ evidence to draw attention to this fact. When Sextus argues against the usefulness of definitions, he ridicules the Stoics for having the view that definition was just ‘an account which acts as a brief reminder, bringing us to conception of the things which underlie our utterances (         ’.140 As Glidden suggests, in the phrase ‘conception of the things which underlie our utterances’ ennoia can be understood either de dicto or de re.That is to say, according to the first reading of the phrase, ennoia indicates how we describe by propositions the things that underlie our utterances, facts or state of affairs. In speech we refer to the intended description of what we are talking about. On the strength of such an understanding of ennoia, definition is a kind of mnemonic instrument designed to bring about the full recollection of a particular ennoia under one particular viewpoint; that is, one’s subjective description of things, facts or states of affairs. According to the de re interpretation, ennoia in the phrase

Self-awareness

35

‘conception of the things which underlie our utterances’ indicates the recognition of what the things, facts or states of affairs are which our utterances designate. Here ennoia represents the actualization of a mental act of recognition. It is an ontological reading of ennoia according to which the world triggers our recognition of things, facts or states of affairs. Sextus attacks the former understanding and the example he provides reveals that he entertains the latter, ontological, understanding of ennoia in the phrase. He says that if someone were asked if he had passed a man on the road riding a horse and leading a dog, it would be beside the point to trigger one’s memory to remember his descriptions of ‘man’, ‘horse’ and ‘dog’: the important point would be the ontological one – whether he had seen a rider with a dog or not. For that, one does not need to know how the other’s conception of these natural things describes them. Sextus, then, concludes that if we understand ‘bringing us to conception of the things which underlie our utterances’ as a recognition of things which we refer to in speech, then definitions are useless. As opposed to the Stoics, who believed that ennoia determines what we should refer to in speech and thus that definition becomes the prerequisite for enquiry, Sextus emphasizes that knowledge is of what exists and our assignment is to recognize the existing things. Now, Glidden identifies Sextus’ attack on the Stoics as an Epicurean objection on the strength of a remark by the anonymous commentator on Plato’s Theaetetus, who said: ‘Epicurus says that words are clearer than definitions’. Though I disagree that this is a sufficient reason to identify Sextus’ criticism with an Epicurean one, the difference between the conceptual character of ennoia, whether it is representational or the means of recognition, seems to identify a distinction between the Stoics and the Epicureans.141 Obviously, Diogenes in the above passage on Epicurean prole p sis, using Stoic terminology, may have simply meant the Stoic understanding of the term when characterizing prole p sis as ennoia. However, Glidden’s observation on the above contrast nicely fits the interpretation of Epicurean prole p sis, as the process of awareness in recognition and thus a criterion of truth – the understanding I have been labouring to support. It also helps to answer Question (1) concerning ennoia: on the de re interpretation ennoia represents the actualization of a process of recognition. The last Stoic phrase used in the description of prole p sis by Diogenes in D. L. X 33 (1) is katholike  noe sis  enapokeimene . As Asmis describes it, ‘universal’ or in my preferred translation of  ‘general’ indicates the gathering of a number of impressions into an overall view.142 It is the result of countless previous perceptions, and Asmis thinks that ‘preceding perceptions’ at D. L. X 33 (2) is actually an abbreviation for Diogenes’ claim in his summary of Epicurus’ canon of thinking:143 Also, all thoughts arise from sense-perceptions by means of (1) confrontation, (2) analogy, (3) similarity and (4) combination, with some contribution from reasoning too. The figments of madmen and dreaming are true. For they cause movement, whereas the non-existent does not move anything. (D. L. X 32)144

36

Self-awareness

As Asmis says, ‘neither the general claim about perceptions nor the fourfold division of thoughts was unique to the Epicureans; indeed, the Epicureans seem to have borrowed their analysis from other philosophers’, and she refers us to Sextus’ presentation according to which ‘experience through sense-perception . . . has to precede every conception’.145 Sextus also provides a similar scheme of fourfold division to that of Diogenes.146 According to Sextus: (1) by way of resemblance we think of things, which we have been presented in experience, e.g. we think of the absent Socrates as a result of seeing a likeness of him. By analogy with perceptible objects either (2) by way of enlargement thereof we think of a Cyclops by mentally increasing the ordinary human being, or (3) by diminution, of a pygmy by a corresponding decrease, and (4) by combination of perceptible objects we think of a centaur. The two lists are different to the extent that Sextus’ typology speaks only about non-immediate ways of concept formation. There is no explicit information about how Epicurus or the Epicureans related the fourfold division of thoughts to prole pseis. But Asmis speculates:  since they held that presumptions147 are memories of frequent appearances from the outside that have occurred not only to the five senses but also to the mind, and since appearances of the latter kind admit of all three stated relationships to sense perception, it seems plausible that the Epicureans held that presumptions are formed in all four ways. In adapting the common scheme to their own epistemology, however, they maintained the distinctive view that all four types of presumption are formed by direct contact with the external world: although not all are memories of appearances obtained by the senses, all are memories of appearances presented directly from outside.148 However, in a more recent piece on the subject, she admits that ‘Diogenes’ identification of a memory as a concept seems to be a rather spectacular leap’,149 as opposed to Aristotle and the Stoics, who conceived of memories as a stage of cognition prior to concept formation and held a tripartite scheme of concept formation (memories – experience – universal concepts).150 If we take into consideration Diogenes’ qualification at D. L. X 32 of the four ways of thought formation, ‘with some contribution from reasoning too’, then we may note the prominent role of perception in thinking ( ). Not only in confrontation but also in the other cases – analogy, similarity and combination – thoughts arise from sense-perception and it seems all the different ways of thinking are, thus, caused externally. In confrontation ( ) thinking is caused directly by some eidol a falling into the mind; by analogy () when we see a small, spherical natural object like an orange we may think of smaller or larger spherical things like atoms or the Moon; by similarity ( ) we think of Socrates when seeing a bust of him or when seeing a likeness of him or we create a thought of a Cyclops or we use this method about theorizing the

Self-awareness

37

unseen level of the atoms as Lucretius does in countless examples (see DRN I–II); and by combination ( ) we may connect a human shape with an animal form to create an image of a centaur,151 or, again, by combination we create good poems152 or construct inferences.153 Most likely in all these thought processes thinking starts off with the focusing of the mind ( ) on some objects of sense-perception, which is the basis and the material for further thinking.154 These various thought processes, in turn, can lead to the creation of abstract concepts. In Philodemus we read: Through empirical assessment of phenomena (    []) I shall reach the conclusion that similarity must exist also in this respect. For since men in our experience possess this characteristic, I shall deem all men in general to hold it, by concluding through empirical reasoning that also in this respect similarity must exist. (De Sign. XXII 28) In our experience, for example, we find that man is a mortal animal, therefore we conclude by similarity that all men are mortal. It is only a step further to construct the abstract concept of death and to verify it on each and every occasion when we see someone die. Our abstract concept of death, therefore, relies on experience, but it is not itself a prole psis. It may come, however, from the  prole p sis or recognition of dying based on the repeated sensory experiences of dying. If we return now to Diogenes Laertius’ discussion in D. L. X 33, which does not concern the prole p seis related to abstract concepts, I think it is most important to consider how prole p sis, described as a ‘stored, general thought, that is, memory of that which has often appeared from outside’, relates to its status as the criterion of truth and to epaisthe s is. As we shall see in a fragment on memory from the later parts of On Nature XXV (see Text M on pp. 48–9), Epicurus took memories to be analogous to affections or movements of the soul, on the strength of which it is tempting to take the Stoic noe sis  in this context in the sense of a process.155 If granted, we could understand this cognitive process as something ‘general’ in the sense of producing typoi, which represent the general characteristics of the content of sense-perception. Their generality is reached by the flexibility of our sense-organs, which let similar images into themselves because of their porous nature, originally opened up by the first images of sense-perception.156 It is primarily a function, which is responsible for the continuous flow of sense-perception. Since the physical traces of sense-perception in our sense-organs seem to be the means for selecting similar images, these traces can also be the factor of generalization on the irrational level.157 This generality is to be emphasized since if typoi were portrait-like impressions in the sense of the paradigmatic case of seeing then Epicurus would have simply doubled the content of vision. Instead, if we look at other uses of the word typos, we find that it carries two pregnant meanings of ‘outline’: on the one hand, it means the contours or bounds of

38

Self-awareness

an object (see Ep. Hdt. 46) and, on the other hand, the essential or general features of something (see Ep. Hdt. 35–6).158 Given the distinct characteristics of sense-perception and of thinking, these two significant meanings are not only fully utilized in the Epicurean context but also help us to understand how the objects of perception are translated into mental and linguistic contents. Accordingly, the products of the proleptic processes are such typoi, which represent the general characteristics of the proper objects of the senses.The objects of the senses are recognized only to the extent on the irrational level that the proper objects of sense-perception trigger proleptic processes in harmony with and in parallel with perception in the irrational soul, whose processes, in turn, are causally responsible for the holistic representation of the sensed objects in the form of some general impressions or typoi. That is to say, first of all, that the phenomenal perception of the world is directly available and true (see, for example, DRN IV 469–521). Let us say I see a horse. In parallel, the particular proper object of perception, the horse, generates a proleptic process in accordance with my repeated sensory and emotional experiences of horses. It causes an epaisthe s is, still on the irrational level – a co-perception, a sensory recognition in the sense that the perceived object triggers a cognitive process, which is ‘stored’ in the sense-organ by being based on some fixed underlying physical patterns, fundamentally stemming from repeated experiences of similar sense-perceptions.159 This proleptic process brings about the categorization of the sensed object in the rational soul as belonging to a kind or type or class of some things, which represents the general characteristics of the object in the form of a typos, such as something having four legs, a flowing tail, etc. With the help of such a typos I can form an assertion or opinion about the sensed object, such as ‘I see a horse’. The sensory recognition activates my capacity to form a propositional attitude concerning what I believe to be seeing. The rational activity of perception only kicks in upon the activation of my rational capacity, which obviously includes the ability to speak, so we can form an opinion about the sense-perception. Such an account also helps the exegetical analysis of a fragment from book XXV concerning thinking: Fr. (b) ’  []                  [[]]                   []         [][ ]  [][][] . . .

(Fr. b)160 . . . less, but moulding impressions on some to a very small extent and on some not at all, and [the eidol a] similar in shape to those [eido l a] which [impact] these sense-organs fall into the rational aggregate as the way has been prepared for them from over there, for in the most cases the same constitution

Self-awareness

39

has the cause through the elements operating on the difference between the atoms and the pre-existing pores . . . but the thought content of the product [i.e. of the occurrent mental state] also . . .

Francesca Masi has drawn attention to the relevant parallel Epicurean passages that help us to understand that this fragment concerns how the mind is made receptive to external images.161 Nonetheless, those passages from Lucretius and Diogenes of Oinoanda represent the process only at the physical level of description, drawing attention to the fundamental material of both seeing and thinking, the simulacra or eido l a, respectively. They do not differentiate the mental from the physical nor between the applicable terminology to describe how the paradigmatic case of vision translates into the mental process of thinking. Although it is very difficult to tell precisely what the second half of the quoted fragment amounts to, it seems likely that Epicurus addressed this difference. Furthermore, if we apply my account of epaisthe sis  to the construal of this fragment, then we can speculate that the beginning of the fragment concerns how the eido l a mould impressions (  []) on the rational aggregate,162 and that it concerns whether the eido l a manage to represent themselves to the mind or not and, if so, to what extent. It might be the case that the constitution of the rational aggregate – the primary evidence for Epicurus’ distinction between the rational and the irrational soul – is causally responsible in some ways for the mental representation of the respective perception. If this much is correct, then the verb , although primarily describing the physical process of how the eidol a are prepared for the mind, would also imply the translation of the eidol a for the mind into some typoi through the preceding proleptic process.163 That what the mind is to interpret is different from what the sense-organs receive is also indicated by : the eido l a the mind receives are only similar to those of the sense-organs, which may imply that they had been generalized through the process into some representations of general characteristics.164 Why are the representations of the sense-organs not sufficient for the rational part of the soul to make judgements? Simply because of the immediate, irrational – in the sense of non-interpretational – representations of the sense-organs. Our sense-organs are in close contact with the external world and even if by the active focusing of our mind we can direct them and concentrate only on those things we wish to, this direction is impossible without the recognition and awareness of our environment, or without a recognition and awareness of ourselves. But how do we know that what we perceive is what we interpret it to be? In other words, where is error located? But falsehood and error are always located in the opinion which we add. For the portrait-like resemblance of the impressions which we gain either in sleep or through certain other focusing of thought or of the other discriminatory faculties, to the things we call existent and true, would not

40

Self-awareness

exist if the thing with which we come into contact were not themselves something. And error would not exist if we did not also get a certain other process within ourselves, one, which although causally connected, possesses differentiation. It is through this that, if it is unattested or contested, falsehood arises, and if attested or uncontested, truth. (Ep. Hdt. 50–1)165 Falsehood and error come about when we form opinions about the phantasmoi, which we perceive either in sleep or through the application of the mind or the other criteria, with the help of the accompanying proleptic process activating our rational capacities and our affections (pathe ) .166 Forming an opinion is, thus, a causally connected yet at the same time distinct process from perception or from proleptic processes. It is a kind of stage where, based on sensory recognition (epaisthe s is) or the lack thereof, the evaluation of our perceptions is formed. As Sextus’ evidence makes clear, Epicurus’ method of testing non-scientific beliefs was witnessing ( ) or not counter-witnessing (  ) by (further) evidence ():167 Of opinions, then, according to Epicurus, some are true and some are false. The true ones are those that are witnessed in favour of, and not witnessed against, by plain evidence, while false ones are those that are testified against and not testified in favour of plain evidence. (Adv. Math.VII 211–12)168 Since it is also clear from Sextus that Epicurus used the term ‘evidence’ as a synonym for presentation ( . . . in: Adv. Math. VII 203), it is apparent that Epicurus thought that our opinions or judgements are tested by further perceptions.169 K. D. 24 not only supports but also extends this explanation: (1) If you simply throw out some perception (2) and do not discriminate between a matter of opinion in accordance with that which awaits confirmation and that which is already present in accordance with the perception and the affections and every presentational application of the mind, (3) you will confuse your remaining sensations by your groundless belief, so that you will throw out every criterion. (4) But if you uphold also everything that is expected [to appear] in your supposed recognitions and it has not been witnessed, you will not escape falsehood, since every debate will be preserved in the case of every judgment whether it is correct or not. (K. D. 24) Epicurus tells us to make a distinction between what is present to us in accordance with the three criteria of truth, and in accordance with our opinion of it, which needs to be confirmed. He warns us that unless we preserve all our perceptions, we are bound to throw our lives into confusion. What perceptions

Self-awareness

41

might he be referring to in (1)? Since what is already present to one is apparent in accordance with the three criteria of truth (as in (2)), based on which one forms an opinion, it seems to follow that the perceptions one may wish to re-evaluate are the ones which are being tested (as is confirmed in (4)). That is to say, we can understand K. D. 24, in harmony with Sextus’ explicit testimony, as claiming that beliefs are tested by further perceptions. This is the element of Epicurus’ theory that Plutarch found inconsistent, and he attacked the Epicureans for it. He said that as opposed to the Cyrenaics, who maintained a logical consistency holding that regardless of whether we have a distant or a near view of a tower there are only ‘opinion and seeming’, the Epicureans, who claimed that all presentations are evident, still selected nearby ones as more reliable than those from afar.170 As Asmis has pointed out, Plutarch failed to take into account the Epicurean distinction between presentation and opinion: the presentations do not differ in validity but there is certainly a difference in opinion, and the distinction between presentations that verify and those that do not is entirely dependent on this difference.171 In other words, the truth or falsity of opinions depends on the consistency of one’s perceptual experience, and if one submits too readily to one’s opinions then that is akin to throwing out some perceptions, threatening our lives with confusion.172 It is obvious, however, that if I am to form an opinion about something, that opinion cannot be of something which I have not recognized as such and such a thing. If I have a phantasia of a horse, in order to utter or think a statement such as ‘there is a horse’, I have to recognize to a certain extent what I am looking at. Falsehood arises if I misinterpret the phantasia and the typos of the perceived object produced by the proleptic process and the related affections (pathe ) , and I take the phantasia, the general impression and the affection stimulated in me for a cow or the other way around. I do not make a mistake concerning what I see because of the prole p sis; it is not an inaccurate process of recognition, but given that the typos represents only the general outlines and/or that my visual experience of how I am affected may not be the clearest due to the given circumstances, there is a possibility of misinterpreting the real content of my perception. On the rational level, I may think that I am seeing a horse rather than a cow from a distance (phantasia) and take the respective typos of a four-legged shape with a flowing tail for a horse instead of a cow. Eventually, as I get closer, if I still keep focusing my attention on the animal that I am approaching, my initial assumption may turn out wrong, yet, the reason for my initial, false perception was not the perception itself but how I evaluated that perception. It is important to notice that there is no explicit evidence for the requirement of prole p sis in this method of evaluation, or, as a matter of fact, for a separate parallel method of evaluation based on prole psis.  173 In addition, K. D. 24 lists the presentational application of the mind (  ) instead of prole psis  along with the other two criteria, i.e. perception and affections.174 Most likely the presentational application of the mind was understood as a passive process based on the reception of the external world.175 Since, as

42

Self-awareness

I have argued, the presentations of the presentational faculty of the mind are handed over to the rational part of the soul in the form of some typoi (see D. L. X 33 (2) earlier), it seems to be a capacity of the mind in direct connection with proleptic processes, otherwise, it would be hard to conceive how it could stand in the place of prole p sis and fulfil its function. Most scholars who take prole p sis to be some mental images or some kinds of rational concepts, nevertheless, do not reduce the function of prole p sis to recognition, but argue that it must have been a criterion concerning the truth and falsity of opinions as well. According to the standard understanding there are two ways in which falsehood comes about: •

•

Either, (1) I mismatch the data of my sense-perception and the relevant prole p sis; that is, prole p sis is required for the formation of opinion based on one’s sense-data and affections;176 or, (2) I make a judgement on the basis of my perception and that prole psis  is required for testing the consequent opinion.177

I think a number of difficulties arise from these claims. Point (1) seems to imply that one may have a false judgement because of a mismatch between two different criteria of truth.178 For example, my sensation is paired with a prole psis  of a horse, therefore I make an assertion that ‘there is a horse’, which upon closer inspection may turn out to be a cow.179 Point (1) seems to be mistaken because it cannot account for the very function it wishes to attribute to prole psis, namely recognition: if the truth or falsity of a judgement depends on  matching sense-data with one’s proleptic conceptual apparatus, it certainly sounds like a rational activity, which needs to depend on recognition, otherwise it is not clear how sense-data could be paired with one’s conceptual prole pseis.  Furthermore, as we read in Diogenes Laertius X 33 (2), as soon as a word is spoken in accordance with prole psis  the typos of the word is thought of, where the Greek  seems to carry the force of an automatic response, which seems to exclude the possibility of a rational reflection on one’s conceptual apparatus activated by the perception – at least on the fairly straightforward assumption that self-reflectivity does not imply immediacy or automaticity. Immediacy excludes the possibility of mismatching between the apparent sense-data and the relevant conceptual content because mismatching implies rational activity, and a reflective, rational activity cannot be immediate and automatic. Therefore (1) comes short of explaining recognition. If one wishes to say, instead, that there is an automatic response between perception and prole psis, sensual expe rience activating one’s conceptual content automatically, an automatism which would also account for recognition, then that understanding excludes mismatching, since it is automatic – a line the supporters of (2) have to take to avoid the previous objection. However, granted this automatic process of recognition when I make a judgement based on my relevant perception and the parallel automatic process, if, as (2) claims, in order to check the truth or falsity of my judgement, I have to reflect on my conceptual apparatus of prole pseis, 

Self-awareness

43

which brought about the relevant typos in me, in such a process the production of typos by means of prole p sis would be completely redundant: it would be sufficient to place judgement simply in one’s perceptual data and the relevant, automatically activated prole pseis, which if they were concepts would them selves already represent the general characteristics of their objects.180 Instead, it seems to me to make much better sense to say that with the help of proleptic processes we have a sensory recognition, which the mind subsequently interprets and if our judgement is unattested or contested by further perceptual data, that is to say, if we do not have consistent perceptions then falsehood arises concerning our judgement, and if our subsequent perceptions witness our former judgements, then they turn out to be true. It sounds just like the basic backbone of an empiricist theory and it is well documented in our evidence. The requirement of prole psis for various epistemological operations,  such as comprehension, investigation, forming doubts, discussion, and so forth as a criterion of truth, thus consists exclusively in recognition: prole psis  is needed in all cases to recognize the subject of investigation without which, indeed, none of these operations could be executed.181 To summarize my findings so far: prole psis  is a criterion by means of which we can recognize our experience holistically, based on our previous perceptions, from which it originates. It is the criterion of truth immediately connected to the other two criteria: it unifies our perceptual experience of the world and as far as it is based on physical processes of the irrational soul, it is also in immediate connection with our internal affections. It automatically articulates the irrational phenomena of the external and of the internal for the mind, so that the mind can place its judgement. But it would be too extravagant to attribute to it an extra function, namely to test opinions, especially since Epicurus already had a well-documented method of evaluation. In my interpretation the external world directly affects our sense-organs, which are vitalized by the irrational part of the soul. The irrational part of the soul is responsible for processing the sensually recognized data for the rational part of the soul, which makes judgements based on the phantasia, in the paradigmatic case of vision, and on the typoi, the products of prole pseis and, further more, based on the related affections (pathe ) that we have. This is the process of translating perceptual experience into cognitive understanding. Yet, one may still wish to object to my overall interpretation that it is hard to conceive of prole p sis exclusively as an element of recognition in light of some other evidence. However, if we narrow our focus to the primary evidence about prole p sis, we will find that it lends rather more support to my interpretation than it questions it. Let me marshal first the relevant discussion of time: We should not inquire into time in the same way as other things, which we inquire into in an object by referring them to the preconceptions seen/ perceived/conceived in our own experience ( ’ ). But the self-evident thing in virtue of which we articulate the words ‘long time’ and ‘short time’ conferring a uniform cycle on

44

Self-awareness

it, must itself be grasped by analogy. And we should neither adopt alternative terminology for it as being better – we should use that which is current – nor predicate anything else of it as having the same essence as this peculiar thing – for this too is done by some – but we must merely work out empirically what we associate this peculiarity with and tend to measure against it. After all, it requires no additional proof but simply empirical reasoning to find that with days, nights and fractions thereof, and likewise with the presence or absence of feelings, with motion and rest, we associate a certain peculiar accident, and that the meaning of this accident, which we express with the word ‘time’, is itself in turn related to these phenomena . . . (Ep. Hdt. 72–3) The understanding of  is decisive because much hangs on it. Obviously, the non-committal sense ‘to perceive’ would suit my interpretation best. However, in a parallel use in Ep. Hdt. 38 it certainly seems to carry its dominant meaning of seeing in the sense of ‘to conceive’. And not much earlier, in Ep. Hdt. 35 a cognate form, , is also used in the sense of ‘conception’. Thus, Gail Fine seems to be correct when suggesting that Epicurus is speaking in a similar way to Plato who was talking about ‘looking at’ Forms in the sense of considering them.182 Such an understanding suits my interpretation equally well, since we can take the first sentence as saying that we refer our investigations to the conceived proleptic processes in our experience, which stems from sense-perception. For the meaning of ’ , there are many parallel passages (see Ep. Pyth. 88, 91, 95). However, the quoted passage is not just compatible with my interpretation: we also learn that time cannot be directly observed, however clear and common a phenomenon it is, but we have to conceive of it through analogy. To understand how the concept of time is generated we need to see that first we draw on those things which we perceive directly in our experience and consequently identify them – most likely through our proleptic processes – as day, night, motion and rest and so forth, and only successively do we abstract the concept of time as the common measure of them all. It is easy to see that this process is in harmony with my reconstruction – based on the evidence of Philodemus – of how an Epicurean may arrive at the abstract concept of death. It also corroborates my presumption that for Epicurus there are no prole p seis of abstract concepts, but abstract concepts rely on the proleptic processes of phenomena, which can be assessed empirically ( ). What Epicurus himself says about prole psis in the context of his theology  (Ep. Men. 123–4) concerns how most people are mistaken about the outline of their common conceptions of god (): they do not preserve their notion of god, but they attribute to it false suppositions instead of prole pseis. We have to see that this much in and of itself does not clar ify in the least how Epicurus conceived of prole psis. On the idealist interpreta tion of the Epicurean gods, however, this issue is reduced to the enquiry of how our notion of god is based on some proleptic processes.183 Since the rest of the

Self-awareness

45

primary evidence also concerns the relation between abstract concepts and prole p seis (see n. 109), let me now turn to Epicurus’ primary evidence on the prole p sis of one’s responsible self and demonstrate on this one example, most relevant for my discussion, the relation and distinction between abstract concepts and prole p seis. This peculiar concept appears in book XXV of On Nature, where in a digression Epicurus offers a fully fledged argument against determinism, since he found it self-refuting (cf. S.V. 40).184 Epicurus argues that by rejecting freedom the determinist must dismiss the possibility of praise and blame and, consequently, he cannot criticize others for their actions and beliefs, or more particularly, for their disbelief in determinism. However, the determinist, by holding us responsible for our non-determinist views, attributes the very notion of responsibility to us for having reasoned incorrectly. Consequently, a determinist cannot convince us that we should subscribe to his position since he cannot even argue for it consistently – one of three major reasons why Epicurus thought determinist arguments self-refuting. Text P185 (0) . . . by which we never cease to be affected, the fact that we rebuke, oppose and reform each other as if the responsibility lay also in ourselves, and not just in our original constitution and in the accidental necessity of that which surrounds and penetrates us. (1) For if someone were to attribute to the very process of rebuking and being rebuked the accidental necessity of whatever happens to be present to oneself at the time, I’m afraid he can never in this way understand (2) blame or praise. But if he were to act in this way he would be leaving intact the very same behaviour which we have in mind in the case of ourselves in accordance with our prole p sis of the cause, and he would have changed the name only. (’ []     [] []   [][]   ’  [ ]     [] ,  ’ [][] [][]) (3) such error. For this sort of account is selfrefuting and can never prove that everything is of the kind called ‘necessitated’; but he debates this very question on the assumption that his opponent is himself responsible for talking nonsense. (4) And even if he goes on to infinity saying that this action of his is in turn necessitated, always appealing to arguments, he is not reasoning it empirically so long as he goes on imputing to himself the responsibility for having reasoned correctly (     [][]) and to his opponent that for having reasoned incorrectly. (5) But unless he were to stop attributing his action to himself and to pin it on necessity instead, he would not even (6) refute . . . [by] calling what is said [to be done] by ourselves by the name of necessity, it is only the name that is changed. But it is necessary for him to demonstrate [instead] that the proleptic outlines to this

46

Self-awareness thing we call the cause by ourselves are defective. (.[]’ []       []      ’ []     [  ]     ’   [ ] [] [], [. . .) (7) but even to call necessitation empty as a result of your claim. If someone will not explain this, and has no auxiliary element or impulse in us to dissuade us from those actions which we perform, calling the cause for them ‘through us ourselves’ (’) but if for everything which we desire to do and we call the cause ‘through us ourselves’ he is giving the name of foolish necessity, he will merely be changing a name; (8) he will not be modifying any of our actions in the way in which in some cases the man who sees what sort of actions are of necessity regularly dissuades those who desire to do something in the face of compulsion. (9) And the mind will be inquisitive to learn what sort of action it should then consider that one to be which we perform in some way ‘because of us ourselves’ ([][]) by desiring to perform it.

It is somewhat curious how the concept of responsibility could be conveyed in some typoi, as the text restored by Sedley seems to suggest, if we conceive of typoi as some kind of image; however, if we understand them as impressions of general characteristics open to linguistic interpretation – the products of a proleptic process – then this tension evaporates.186 What are these general characteristics in this case? We may conjecture that certain individual actions of human beings are regarded as immediately connected in our experience. We recognize them as causally responsible efforts because the relation between certain actions and their outcomes present themselves to us as immediately connected, and consequently perceiving their repeatedly observed relations we have our prole psis  of the ‘cause’. We observe the actions of others and ourselves, and first by means of combination – the fourth type of thought formation on Diogenes’ list at D. L. X 32 – we combine the modes of those actions, and second by means of similarity – the third type of thought formation on Diogenes’ list – we apply the proleptic experience of the cause attributed to those modes of actions done by others to the modes of actions done by ourselves, which eventually results in the notion of ourselves being causal factors in the world, i.e. in the conception of our responsible selves. This would be a process similar to how we conceive of time or death or even of justice (see K. D. 37–8), in which case the proleptic experience of observing the utility of certain actions results in our conception of justice. The text seems to suggest – (0)–(2) – that it is our behaviour which makes us connect our prole p sis of the cause with ourselves. Accordingly, if the Epicurean prole pseis are built up from recurrent experiences, that is, if they are the results  of repeated encounters with things through sense-perception, the prole p sis of our own responsible self must have primarily come from frequent observations of people behaving in ways that are in harmony with explicit causal processes. The memories of such actions, in turn, give rise to the conception of responsible

Self-awareness

47

agency, since we also start reflecting on our own behaviour, as it is represented in our repeated and immediate sense-perception of ourselves. Consequently, the prole p sis of the cause concerning ourselves has an empirical basis, as opposed to the determinist’s argument, which has no empirical background (see (4)), and is even self-refuting. If the determinist attempts to convince us by describing actions which we consider ourselves responsible for as necessitated by tagging the word ‘necessity’ to them, he will not prove that the proleptic outlines () of what we call ‘the cause by ourselves’ ( ’[][] []) are false; if the determinist changes a name it does not affect the empirical background information. Epicurus’ empirical argument is founded on showing why the determinist’s argument is self-refuting – (0)–(4); why he cannot substantiate his thesis in light of our prole p sis of the causality of our self – (5)–(6); and why his attempt to convince us has no pragmatic consequences – (7)–(9). The last leg of the argument is deeply embedded in the idea discussed earlier concerning naming in strong relation to our experience of the external world. As I have argued earlier, the psychological and physical effects of nature in Epicurus’ atomism is directly connected to one’s psychological development and ability to speak the language of one’s tribe. Nonetheless, just because we think that we observe causally connected events in the world, e.g. the causal power of the wind knocking over a vase or someone killing a slave, events which we would all agree to exhibit causal connections, it does not necessarily follow that these observations could not be compatible with determinism. The reason why the determinist cannot convince us, therefore, depends on what empirical basis we have come to describe the phenomena in certain particular ways, and since for the most part naming is a natural phenomenon based on common experience, it cannot be that everybody is wrong but the determinist.Yet, Epicurus seems to understand that he needs to corroborate his refusal of determinism by our immediately available observations of our internal desires. He adds that it is the task of the mind to seek out those actions we consider ‘because of ourselves’ in harmony with our desire to perform them. This line of thought immediately leads us back to Fr. 3, for the interpretation of which we entered the arena of Epicurean thought processes. As far as I conceive of my responsibility by looking at the actions of others, that sounds very much like thinking of myself through others, or, as it is briefly put in Fr. 3, thinking of oneself in accordance with the similar and non-different. Fr. 3 []         [[]]’[]  [] ’ [] [][] [+/– 5/6 ][.]..[ +/– 7/8 ][.].[ ]  [ ] [.][ +/– 7/8 ][ ] [

But as we can see in Fr. 3 Epicurus is not simply talking about self-reflective thinking in accordance with the similar and non-different, but also about

48

Self-awareness

self-reflective thinking by means of oneself in accordance with the similar and non-different, like that which is thought to be in some unity with this [i.e. the mind]. The all-important question is how to qualify ‘by means of oneself ’, that is,  in the first line of Fr. 3. As I have conjectured,187 Fr. 3 preserves two types of self-reflective thinking, one in accordance with the aitiologikos tropos – lines 1–2 – and another in accordance with the pathologikos tropos – lines 2–4, corroborated by Fr. 5. As I have argued, in the second line of Fr. 3  is qualified by what is said in lines 3–4 and accordingly it is to be understood as ‘by means of / in one’s own pathos/ pathe ’ .188 I suggest in the first half of Fr. 3 the meaning of  also depends on a qualification, which I take to be ‘in accordance with the similar and nondifferent’. Consequently,  in line 1 and  in line 2 are to be understood differently on the strength of their different qualifications. The historical context in which I have placed the first half of Fr. 3 and Epicurus’ discussion of the prole p sis of the cause and our strongly related conception of our responsible self in the later parts of book XXV together hint at the very likely answer that ‘[]   ’ paraphrases the idea of one’s self-reflective thinking by one’s memory in accordance with the similar and non-different – thinking of oneself through the memory of similar agents to oneself – which, as a matter of fact, can lead one to the conception of oneself as a responsible agent. Such an understanding of  in line 1 of Fr. 3 is also arguable on other grounds. Epicurus’ discussion of self-reflective thinking itself is embedded in the wider context of an examination of memory: prior to Fr. 3 and even within the context of self-reflective thinking in Fr. 7 and Fr. 8, Epicurus is talking about the role of memory in one’s conception of the ethical end, and right after the context of self-reflective thinking we have some fragments which are about how memory originates. In his discussion, Epicurus repeatedly uses the phrase ‘memory or some pathos analogous with memory’ by means of which we are able to remember things and presumably by means of which we are able to perform activities such as self-reflective thinking. Therefore, as the next step in our reconstruction, we need to turn to the major fragment on memory in book XXV. Memory The key passage for understanding Epicurus’ conception of how memories come about appears at the beginning of the later parts of book XXV. Though the passage is undoubtedly difficult, I think with some minor modifications of the text we can acquire a more accurate understanding of what is going on. Text M189 []    []                                     []

Self-awareness 5

49

                     []    [] [ ] [] []      [] [ ] [ ][ ] ’  [] [] . . .

(Text M)190 . . . the memory or the affection (pathos) analogous to the memory of the more necessary things came to be/exist within? () in reference to the well-defined and that is used to test all things and not in accordance with things that cannot be defined but need judgment.This memory of that, or the movement analogous to memory, was again in one aspect co-generated immediately, and under another it had grown, being the beginning and the cause for, in the first case, the first constitution (  ) of both the atoms and what is produced ( [o]) [i.e. the occurrent mental state191], in the other case, for the on-growing [constitution] (  []), by means of which we perform all our actions, of the atoms and the product itself (    []) [i.e. the occurrent mental state itself] that in some cases is necessarily opposed to what produced . . .

This passage is about how a memory comes to be with reference to the well defined and all refuting, that is, to one of the criteria of truth. If that reference (anaphora) implies that the emergence of a memory involves the appearance of the related affection to that memory, it also indicates – in agreement with my analysis – that memory belongs to the irrational part of the soul – which Lucretius (DRN IV 765–7) also indicates. To what extent is that a well-founded assumption? A memory is either immediately present and in this case is constitutive of both the arrangements of the atoms and the first constitution of a given mental state, or recalling a memory requires some time during which the developing arrangement of the atoms as well as the emergence of a particular mental state takes place; in this latter case, the memory has a causal aspect in the structured formation of the occurrent mental state. The immediate co-generation seems to refer to the rearranged constitution of the atoms and the mental state; these are co-generated simultaneously. That fact gives us some insight into the interpretation of the phrases ‘the memory or the affect analogous to the memory’ and the ‘memory or the movement analogous to memory’. For Epicurus, memory seems to also include the affective contents of memories connected to past perceptions or thinking. In other words, remembering seems to entail that a memory of a past perception or thought is also re-constitutive of the respective affection. Thus we may assume that the rearranged atomic constitution, which must be a rearranged pattern of atomic motion, also represents the affection of the perception or thought remembered. The fact that memory or the movement analogous to memory is ‘the beginning and the cause for’ the simultaneous constitution of a new atomic pattern of the mind or a segment thereof, and also of a new mental state strengthens such a suggestion.192

50

Self-awareness

The strong causal connection between the mental and the physical in this short passage indicates a physicalist theory of mind. Nonetheless, I do not think it invites reductionism on these grounds.193 I agree with those who say that to apply the term ‘reductionism’ when interpreting Epicurean texts is misleading, since Epicurus’ distinction between the level of observable reality and the level of atomic reality does not imply the reduction of the former to the latter. As Laursen put it, ‘the atomic nature of the world is a metaphysical theory that is never contradicted by experience, but it cannot possibly be “confirmed” by a “reduction”, since we cannot verify any statement about the atoms directly’.194 Consequently, I do not think that we need to understand this passage as a reductionist claim concerning Epicurus’ idea of memory. The word  clearly shows that Epicurus is not saying that a certain memory is a certain affection () or a certain (atomic) motion () and nothing over and above it, but, I take it,  simply means, ‘at the other level of description’ memory is some kind of affection () or, as Epicurus readily substituted it, some (atomic) motion (). Nonetheless, that is just the affective content of the memory, which comes with remembering the visual or intellectual content. As I shall argue in the next chapter, the extremely detailed and technical discussion of the two stages of memory represents a psychological process between the irrational and rational soul parts in parallel with those I have analysed in my account of the Epicurean understanding of thought processes canvassed earlier. It suffices for us now to see the bipartite stages in the account of the psychological process of memory, which explains why it is a good assumption to understand ‘[]   ’ in the first line of Fr. 3 as self-reflective thinking ‘by one’s memory in accordance with the similar and non-different’. My analysis has shown that perception and thinking were both transitive processes for Epicurus, which meant that each and every kind of intellectual process has to arise from sense-perception (D. L. X 32), a condition probably meant to exclude the charge of solipsism or distinguish Epicurus from the Cyrenaics to the extent that he wanted to explain our internal states in strong connection to the external world. If memories are activated because of external stimulation, it makes perfect sense to say – especially in the context of a passage (Fr. 3) which includes a kind of self-reflective thought based on one’s affections – that whatever precisely stimulates one to think of oneself in the aitiologikos tropos, this self-reflective process is a contemplation of oneself as a causally responsible living being constituted by an interdependent unity of body and soul – the content of the contemplation formed by one’s memories of the behaviour of similar human beings. If memory is a prerequisite for the building of a rational subject’s identity in his relation to the surrounding environment, as Masi has put it,195 then just as the pathologikos tropos has an ethical function, the aitiologikos tropos is essential for Epicurus’ ethical theory: the recognition of myself as a causal agent in harmony with my memories of similar agents brings about the conception of my responsibility, on the strength of which Epicurus argued against the advocates of determinism in the later parts of book XXV. Thus, self-reflective thinking has turned out to be an element in

Self-awareness

51

the scrutinizing of what it takes to have a proper idea of oneself in the world, what is a necessary condition for one to live a eudaimon life.While the pathologikos tropos helps one classify one’s desires, the aitiologikos tropos makes one realize the responsibility for one’s actions, which are only prudent if in accord with the health of one’s body and soul.

Taking stock I conclude this chapter by looking at the last two fragments of the topic on self-intellection, Fr. 7 and Fr. 8. Although they are admittedly opaque, they seem to corroborate my interpretation of Fr. 3 and what I have said on Epicurus’ two modes of explanation. Fr. 7 ’ [][]   ’  [[]] {}   ’ (       [] [] []  [][] []  [][] []), .[+/– 4/5][.] [.].[+/– 5/6][.]  [.]  [+/– 4/5].[.]  [ +/– 7/8 ][ Fr. 7 = 1056 corn. 4 z. 2 8–9) [][] Németh: [][] Laursen.

(Fr. 7)196 . . . some products [i.e. some occurrent mental states] are brought about by being produced in accordance with external influx, others in accordance with vicinity – in a way also that part of our nature that is not held in abeyance, but remembers and determines, to a greater or smaller degree, what is our inner end and also . . . {} indeed sounds somewhat odd in line 1 and by deleting {} we get an intelligible passive present participle form of . As I will argue in Chapter 2, the different grammatical tenses of  are coordinated with the different phases of a given mental state, the passive present forms standing for the first constitution of a given mental state, which is to be further structured in accordance with one’s cognitive disposition – the result of which is expressed by the passive perfect form. If we accept Laursen’s suggestion then, in my interpretation in this text, we read about the immediate formation of some mental states, , either due to the external influx of some things (most likely of some atomic eido l a) or in accordance with vicinity. One of the only other two instances of  in the surviving works of Epicurus seems to suggest that vicinity here is to be taken as the vicinity of body and soul (see Ep. Hdt. 64 and Ep. Pyth. 106). If that is correct, the explanans of vicinity, the mental states of the rational soul, do not come about by the vicinity of an indefinite part of the body, but by some of its activities, namely by

52

Self-awareness

remembering and determining. However, these two activities cannot be attributed to the body, therefore we have to assume that they are those of the irrational soul, making the body sentient. This reconstruction is also supported by the qualification ‘to a greater or smaller degree’, a qualification which also holds up my interpretation that the activity of remembering is a transitive process between the irrational and rational parts of the soul, lending further support to my reconstruction of the Epicurean view of thought processes in general. Fr. 8  [][][] [ +/– 0/1] ...[.]. . ..... []   .. [.].[ +/– 1/2 ]        [  ]       [][]   []    [ ]        [  ][] []     [].  []    [-]. . .  . [[]] [.] [..][. . .] [.]..[

(Fr. 8)197 . . . the natural goal following . . . took by nature [sc. by its own nature ?] . . . thought . . . with attention he remembered the measuring ways and that grasping in memory that submitted itself to the natural goal, the pleasant or painful, in beliefs or otherwise, as I wrote so very much above. And to what extent, then . . .

Fr. 8 is rather difficult to make sense of, but it is in line with Fr. 7, and it seems to extend it with the next stage of remembering, that is, with the formation of a belief or judgement. As I have hoped to make clear, Epicurus’ conception of self-awareness rested on a robust theoretical groundwork.Therefore, it is no accident that the discussion of self-awareness appears at a relatively late stage, the twenty-fifth book, of On Nature. As has been suggested, the three copies of book XXV preserved in the ashes of the Villa dei Papiri indicate its illustrious status among Epicurus’ works198 and although we have only covered some of its fragments so far, we may already have a picture of the reason why it was so thrilling for the Epicureans. The explanation of self-awareness in the pathologikos tropos and aitiologikos tropos coordinates several areas and concepts of Epicurus’ philosophy199 and draws together the Epicurean framework into the focal point of self-intellection, a significant philosophical topic since Plato. In sympathy with Plato’s and Aristotle’s epistemological and ethical arguments, Epicurus’ conception of the self was strongly related to the natural sociability of human beings: self-awareness in the pathologikos tropos was not only an essential element in one’s mental development but, naturally, it was also related to one of the criteria of truth (pathe ) and thus helped to conceptualize this criterion for Epicurus’ ethical hedonism. In consequence, the pathologikos tropos seems to have rationalized the reference points of self-reflective thinking in relation to the external world, that is, in relation to socially relevant contexts. The aitiologikos tropos seems to indicate a further notion in Epicurus’ exposition of self-intellection: the idea of self-knowledge through others. It is not only an essential element for one’s

Self-awareness

53

self-conception of moral agency in an empirical way, creating a connection between one’s self and the proleptic recognition of causality, but it will also prove essential in the ethical context of friendship, as we shall see in Chapter 5, as the means of one’s shared moral perception. I hope to have unpacked in Chapter 1, at least to some extent, the terminological jungle encompassing these fragments of Epicurus’ ideas on self-intellection. It will be the task of the next two chapters to establish how Epicurus’ discussion of self-reflective thinking is related to the later parts of book XXV, which I believe will contribute to the proper understanding of Epicurus’ conception of moral agency and that of the self in the rest of book XXV.

Notes 1 Epicurus does not outline his ethics in the Letter in an inferential manner, as the Epicurean Torquatus does in Cicero’s De Finibus (see Sedley 1996), but instead he gives a longer treatment of a few points, also found in his Principal Doctrines, a forty-piece collection of his most important teachings, and in the so-called Vatican Sayings, a collection of maxims discovered in 1888 by C. Wotke in a Vatican MS (Codices Vaticani Graeci 1950) and published by him and Usener for the first time in Wiener Studien, 1888. The Letter to Menoeceus is concerned with holy beliefs in gods, the elimination of the fear of death, doing away with the common opinions in regard to destiny, and with the reflection on our desire to grasp what is by nature desirable. Although some of the claims in the Letter are not argued for, its intent is clearly more than simply to provide dogmatic advice. Just as in the case of his Letter to Herodotus, in which Epicurus explicitly states that he means to give a comprehensive view of his physics (Ep. Hdt. 35), in the Letter to Menoeceus Epicurus seems to have meant to give a comprehensive view of what constitutes the happy life for the wise. 2 Cf. Striker 1996, pp. 170–1. 3 See Kahn 1998, p. 247. 4 See Long 1999, pp. 618–23. 5 See Plato’s (?) First Alcibiades. 6 For the Epicurean conception of the thoroughly united body and soul, see Lucretius DRN III 323–36 and Diogenes of Oinoanda Fr. 37. 7 As Nussbaum has observed: ‘What he [i.e. Epicurus] needs, to make this diagnosis compelling and to commend his cure, is, first, a procedure for separating good desires from bad, healthy from sick; then, a diagnosis of the genesis of bad desires that will show that and how they are based on false belief; finally, a therapeutic treatment for false belief that will show us how, through a modification of belief, we may get rid of bad desire’ (Nussbaum 1994, p. 105). 8 See Tsouna 2009, p. 252, where she also makes the important point that Epicurean therapy was also available to the whole of mankind with the aid of the Epicurean writings. For a discussion of one of its special media see Warren 2000. 9   ’                 

54

Self-awareness   ’               Text and translation from

Long and Sedley 1987, vol. II, pp. 115–16 and vol. I, pp. 113–14, respectively. 10 See the scholion on Epicurus’ K. D. 29. 11 See Epicurus’ conception of the living being () in Fr. (a), on p. 25. 12                                 ’                                 Text and translation from Long and Sedley 1987,

vol. II, p. 116 and vol. I, pp. 113–14, respectively. 13 Cooper 1999 argues for an ethical reading of Epicurus’ position, according to which Epicurus thought that pleasure ought to be the goal of human action – ethical hedonism – an interpretation effectively refuted by Woolf 2004, who argues for the psychological reading, according to which pleasure is the goal of all human action – psychological hedonism. 14 See D. L. VII 85–6. For the so-called ‘Cradle Argument’ see Cicero De Fin. I 30: ‘Every animal as soon as it is born seeks pleasure and rejoices in it, while shunning pain as the highest evil and avoiding it as much as possible. This is behaviour that has not yet been corrupted, when nature’s judgment is pure and whole. Hence he [Epicurus] denies that there is any need for justification or debate as to why pleasure should be sought, and pain shunned. He thinks that this truth is perceived by the senses, as fire is perceived to be hot, snow white, and honey sweet.’ Translation from Annas and Woolf 2001, p. 13. Also see D. L. X 137 and Sextus Adv. Math. XI 96. For an extensive discussion of the argument see Brunschwig 1986, and also Inwood 2016. 15 See n. 13. 16 See Brunschwig 1986, p. 115. The distinction between katastematic and kinetic pleasures is not apparent in the quoted text, or even in the letter as a whole, yet we have ample evidence for it, and Epicurus’ reliance on this distinction is clear; see D. L. X 136. Painlessness () and undisturbedness () are characterized as katastematic pleasures, which qualifies the kind of pleasure Epicurus takes to be the telos of the blessed life. Kinetic pleasures, on the other hand, are pleasures in motion, involving change or the titillation of the senses; see, for example, fr. 67 Us. For my precise understanding of the distinction between katastematic and kinetic pleasures see the following paragraphs. 17 See Cooper 1999, pp. 511–14 and Cooper 2012, pp. 229–46. His main textual support is the evidence of Epicurus’ K. D. 18, where Epicurus talks about the variation of the pain or disturbance free condition being a state itself getting varied, and not merely juxtaposed, with kinetic pleasure. 18 Cf. Lucretius DRN III 320–2. On the Epicurean conception of homoiosis theoi see the Epilogue.

Self-awareness

55

19 See especially Cooper 1999, p. 514 and Cooper 2012, p. 236. 20 One may wish to object here that my interpretation collapses into the traditional two genera view. However I do not think it does because it considers pleasures uniform in a certain sense. 21 Cf. Ep. Men. 129:  (i.e. pleasure)  . 22 Cf. D. L. X 31: . I argue for the view – see pp. 17–24 – that Epicurus connected aisthe s is and pathos and held the general sensationalist view that every perception is conjoined with a respective affection. 23 This point is made by Mitsis 1988, p. 38. 24 The term pathos and its plural pathe  are best left untranslated in most cases, since in English ‘undergoing’ would be too broad, but ‘affection’ or ‘passion’ may be misleading in some contexts. I agree with Konstan that pathe  are not rational in the sense that they do not involve the logical element or the mind and thus they are very different from emotions; see Konstan 2006 and Konstan 2008, pp. 1–25. 25 See more on this connection on pp. 11–23. See also Konstan 2008, p. 6. 26 Ep. Men. 132, translation from Long and Sedley 1987, vol. I, p. 114. 27 See Blank 2013. 28 The description of the death of Pliny the Elder by his nephew, Pliny the Younger (see Letters VI 16) – though it partially sounds like a piece of fiction – testifies to how unprepared or unwilling people were to leave their settlements at the time of the eruption. 29 Cf. Philodemus’ works On Rhetoric, On Music (PHerc. 1497), On Poems and On the Good King according to Homer (PHerc. 1507). 30 The evidence to be discussed comes from all these sources; see abbreviations in n. 44. 31 For an overview of the editorial history of book XXV see Masi 2006, pp. 21–6. 32 For the discovery of the identity of 1056 and 697 see Gomperz 1876; and of these with 1191 see Gomperz 1879. Identifying book XXV as book XXV see Laursen 1987. For the grouping of PHerc. 419/1634/697 see Dorandi 1983, for PHerc. 1420/1056 see Puglia 1987. 33 See Sedley 1998, pp. 99–102. 34 Sedley 1998, p. 107. 35 For a detailed outline of the confirmed and speculated contents of all the thirtyseven books see Sedley 1998, pp. 109–34. Also see Delattre and Pigeaud 2010, pp. 79–117. 36 Fragment 1056 corn. 8z. 3 = Arr. [34.33] lines 4–7:   []  []        . ‘So, of those which we proposed at the beginning, we have explained both the pathologikos tropos and the aitiologikos tropos.’ David Sedley’s new translation, who suggested to me that this should imply that Epicurus set out more than two tropoi at the beginning (presumably of XXV), and by the end of book XXV he has explored only two of them. One strong possibility, as Sedley suggests, ought to be then that what remained was the tropos which would bring in the swerve, which if so was in book XXVI. 37 Sedley 1983, p. 17. Also see on this Masi 2006, p. 57. 38 Cf. Sorabji 2006, p. 202. 39 According to the standard interpretation prole p sis, generally translated as preconception, is a concept based on recurrent perceptions of things. I leave the term untranslated here; for my different interpretation see pp. 27–48. 40 Laursen 1995 and 1997. 41 Laursen 1995, p. 43.

56

Self-awareness

42 This methodological principle applies to the analysis of all the fragments throughout the book. 43 Laursen 1995, pp. 104–8. 44 Text sources: P: papyrus O: Oxford apographon/disegni N: Naples apographon/disegni VH:Volumnia Herculanensia (copper plate repr. of N) Ph: Photograph On the apparatus: [a]: letter lost in lacuna. [+/– 5/6 ] means that the lacuna is not supplied with supplement and either 5 broad or 6 thin letters would have fitted in the empty space. «a»: letter omitted by the ancient scribe and added by the editor. ‹a›: letter altered by the editor. [[a]]: letter deleted by the ancient scribe. {a}: letter deleted by the editor. : spatium vacuum. 45 Fr. 1 = P 1191 corn. 4 pz. 1 z. 2 col. 4, as in Laursen 1995, p. 104. Translations of the fragments are mine, unless otherwise stated. 46 Fr. 2 = P 1056 corn. 3 z. 2 = 5 II = 9 N = 889 O = Arr. [34.14], as in Laursen 1995, pp. 104–5. 47 Laursen 1995, p. 59. 48 Though it is missing from the main text of the Letter to Herodotus, cf. Fr. (b) on p. 38 and the scholion to Ep. Hdt. 66; also cf. Diogenes of Oinoanda Fr. 37, col. 1. 49 For my discussion of the Epicurean theory of  see Chapter 2, pp. 92–8. 50 Translation from Long and Sedley 1987, vol. I, p. 67. Also cf. DRN III 323–49. 51 David Sedley has pointed out to me that the end of the fragment (lines 3–5) suggests some connection with Ep. Hdt. 40, and then we could also say that with DRN I 445–8 as well. However it is very difficult to make sense of the last few words along this line in the context of self-reflective thinking. 52 Fr. 3 = 1056 corn. 3 z. 3 = 5 III = 10 N = 890 O = Arr. [34.15], as in Laursen 1995, pp. 105–6. However, I modified the text in an important respect as apparent from the critical apparatus. 53 I agree with Laursen that in view of  in line 2 an preceded. 54 Laursen 1995, p. 105. 55 Cf. Fr. 3: (3/a) [] (3/b) . . .’[][]  Fr. 5 

  

56 For the sake of completeness: Fr. 4 [][  ] [ ][ - ]. [ - ] [ - ] [ - ]. . . [ - ].[ Fr. 4 = 1191 corn. 4pz. 1 z. 2 col. 6, as in Laursen 1995, pp. 105–6. 57 Fr. 5 = 1056 corn. 4 z. 1 = 6I = 11N = 891 O = Arr. [34.16], as in Laursen 1995, p. 106. 58 In view of []’. 59 For translating the neuter pronouns as ‘living being’ cf. Fr. (a) on p. 25, which also explains how Epicurus conceived of it.

Self-awareness

57

60                                            ,      Text and translation from Hicks 1931,

pp. 604–5, with some minor modifications. 61 See Brunschwig’s 1994, pp. 23–4 ingenious solution how to resolve the reasonable interpretational tangle, for which cf. Bollack et al. 1971, pp. 22 and 236. 62 This understanding gains further support by Ep. Hdt. 49: ‘For external objects would not imprint their own nature . . . as effectively as they can through certain delineations penetrating us from objects (     ).’ Translation from Long and Sedley 1987, vol. I, p. 73. 63 Such a claim is very similar to the Stoics’ conception of oikeiosis,  according to which the fact that the dearest thing to every animal is its own constitution and its consciousness thereof is to be maintained upon the hypothesis that it is very unlikely that nature would make a living thing a stranger to itself (cf. D. L.VII 85–6). In both cases, one’s awareness of oneself or of one’s own constitution needs to be theoretically conceived; it cannot just be pointed at as in the case of the evident nature of the Epicurean pleasure principle, which needs no proving, cf. Cicero De Fin. I 30, quoted in n. 14. Both, the Stoic and the Epicurean claims can be established only hypothetically given that they are founded on introspection. But other than their similarity their different backgrounds are actually even more interesting, for the Stoics, the all-embracing, teleological Nature, and for the Epicureans the antiteleological, evolutionist characteristics of things. The latter lends support to the Epicurean hypothesis of self-awareness in the pathologikos tropos, to the hypothesis that irrational pathe  are one’s instinctive master in becoming aware of oneself, and also that they are the further means of psychological development. 64          ’      ’           ’     

65

66

67 68 69 70 71

Text and translation – with some minor modification – from Long and Sedley 1987, vol. II, p. 98 and vol. I, p. 97, respectively. For the naturalness of names and naming cf. also Proclus in Cratylus 17.13–16, where we read that language users ‘did not impose names knowledgeably, but as being moved naturally, like coughers, sneezers, bellowers, howlers and groaners’. (Reference and translation from Atherton 2009, p. 210.) It is a matter of debate over Epicurus’ semantic theory whether the initial utterances were meaningless or not: for the psychological reading see Long 1971, for the extensional see De Lacy 1939, and for its complete denial see Glidden 1983. My particular findings will support the psychological interpretation, and I find Everson’s (1994) solution the most attractive concerning the question. Everson draws attention to a passage in Lucretius (DRN V 1028–32) in which Lucretius claims ‘it was nature that compelled the utterance of the various noises of the tongue, and usefulness (utilitas) that forged them into the names of things’. His explanation introduces an extra stage in the development of language, perfectly compatible with Ep. Hdt. 75; cf. Everson 1994, pp. 97–9. Translation from Trapp 1997, pp. 254–5. Glidden 1979. Rist 1972, p. 30. Asmis 1984. Glidden 1979, p. 301.

58

Self-awareness

72                   ’              .Text and translation from Long and Sedley 1987, vol. II,

p. 76-7 and vol. I, p. 73, respectively. 73 hoc ubi non longum spatiumst unde illa profecta / perveniat vox quaeque, necessest verba quoque ipsa / plane exaudiri discernique articulatim; (555) / servat enim formaturam servatque figuram. / at si interpositum spatium sit longius aequo, / aëra per multum confundi verba necessest / et conturbari vocem, dum transvolat auras. / ergo fit, sonitum ut possis sentire neque illam (560) / internoscere, verborum sententia quae sit; / usque adeo confusa venit vox inque pedita (DRN IV 553–62).Translation from Smith 2001, p. 115, slightly modified. 74 Cf. Asmis 1984, pp. 162–4 and Asmis 2009, p. 102. 75 Sextus in Adv. Math. VIII 65 actually charges Epicurus to have fallen into an impasse without such a distinction. 76     ’       (D. L. X 32). Text and translation from Long and Sedley 1987, vol. II, p. 85 and vol. I, p. 79, respectively. 77 For the three different conceptions of truth in Epicurus cf. Hahmann 2015. 78 Cf. Asmis 1984, pp. 151–3 for her disagreement with those modern interpreters (Bailey 1928; De Witt 1943; Bourgey 1969; Glidden 1971; Rist 1972; Striker 1974) who introduce various distinctions into Epicurus’ theory of perception and verification. 79 For epaisthe s is being irrational cf. Philodemus De musica 4 col. 2. 18 Kemke 1884, p. 64. and PHerc. 19/698 col. 10. 1–4 in Scott 1885, p. 266. Also cf. Long and Sedley 1987 vol. II, p. 85. 80 Cf. Lucretius DRN III 359–69. 81 Thus I disagree with Bailey that the term  is related to the recognition of the mind (cf. Bailey 1928, pp. 420, 440) and Arrighetti (1973), who usually translates the term with perception, or Masi (2006, p. 46), who in order to explain one odd apperance of the term in the fragments (which I will analyse as Fr. 19 on pp. 125–7 in Chapter 3) understands it as an ‘apprehension of something evident’. I agree with Long (1971, p. 130, n. 11) to the extent that the term  is a kind of recognition, the recognitional element of perception, but it does not imply judgement, rather just a natural process connected to prole p sis. What that means precisely I will investigate in the section on prole p sis, pp. 27–48. 82 Cf. Ep. Hdt. 53 for the physiological description: ‘. . .we should hold that the impact which occurs inside us () when we emit our voice immediately squeezes out certain particles constitutive of a wind current in a way which produce the auditory feeling in us (   )’. Translation from Long and Sedley 1987, vol. I, p. 74. 83 Glidden 1979, pp. 299–300. 84 Glidden thinks it is completely the other way around; cf. Glidden 1979, p. 298. 85 ‘Now whoever is to any degree conscious of how he is feeling must to that extent be either in pleasure or pain.’ Cicero De Fin. I 38, translated by Woolf, in Annas and Woolf 2001. 86       ’         

Self-awareness

59

. Trans-

lation from Asmis 1984, p. 83, modified. 87 This point is simply contrary to the available evidence; cf. Cicero De Fin. I 37–9. 88 It may sound out of place to talk about the mind having irrational perceptions subsequently evaluated by reason. But this worry is vitiated if we consider Epicurus’ theory of dreams, in which the mind perceives the external world directly without rational reflection; cf. Diogenes of Oinoanda Fr. 9, quoted on pp. 28–9. 89 Consider the different approaches of the pathologikos tropos and the aitiologikos tropos. 90 Scott 1885, p. 265, however I quote the text and translation (slightly modified) from Asmis 1984, p. 161, n. 47. For the most recent edition of the papyrus see Monet 1996. Also cf. Aetius, according to whom Epicurus maintained that ‘both the affections and the perceptions are in the affected places and the mind is unaffected’, as quoted in Asmis 1984, p. 162. 91 Cf. Ep. Men. 130. 92 For translation and references see p. 14 and n. 57. 93 Cf. Sedley 1973, pp. 28–30, and Schofield 1996. For the summary on the history of the understanding of the concept in the secondary literature cf. Masi 2006, p. 41. The next fragment in the surviving sequence on self-reflective thinking has a cognate form of , which in the light of Sedley’s analysis (cf. Sedley 1973, pp. 32–3) should refer to the faculty of the rational calculus, however [][], if Laursen’s conjecture is correct, seems to refer to the activity rather than the ability in Fr. 6. Fr. 6 []   []   [][]   [ ]  . . .[.]      [] [] [ +/– 4/5][. . .][ +/– 3/4 ]. [. . .] [+/– 2 ][ +/– 3/4 ].. [. . .][-][. . .][ (Fr. 6). . . I said, and the rational calculus is produced concerning the whole, having its causes in the movements that I mentioned above.

94 95 96 97 98 99 100 101 102 103

Fr. 6 = 1191 corn. 4 pz. 1 z. 3 col. 1 = Arr. [34.3.2-6], and 697 corn. 2 pz. 2 z. 1 = 5 I = fr. 8 N (fr. 6 VH.) = + (350) O = Arr. [34.3], as in Laursen 1995, pp. 106–7. Cf. also Long 1986, repr. in: Long 2006, pp. 187–8; and see more on phrone s is in Chapter 5, pp. 166–8. Laursen 1995, p. 60. Cf. DRN IV 110–28 and Diogenes of Oinoanda Fr. 10 in Smith 1993. Aristotle had also noted in De Anima that if thinking is characteristically like perception, then thinking is brought about by the intelligible acting upon the mind; cf. De An. III 429a 13–15. For her discussion of perception and thinking cf.Annas 1992, pp. 157–75.Although one could equally think that they need to be absorbed in one way or the other as a supply for the images our body is constantly throwing off. Ep. Hdt. 49–50; cf. Lee 1978, p. 29. Annas 1992, p. 164, giving the textual reference to this point as DRN IV 962–1036. Annas 1992, p. 165. For the complex historical background of my brief discussion see Sorabji 2006, pp. 230–45, and for a detailed discussion on the history of self-intellection cf. Crystal 2002. Sorabji considers four versions, all of which are different from Plato’s. For now, I just introduce the one, which can be the most obviously compared to the Epicurean fragment.

60

Self-awareness

104 105 106 107 108

Laursen supplevit. Fr. (a) = 419 fr. 7, as in Laursen 1995, p. 72. Cf. D. L. X 32, discussed in detail on pp. 35–7. Cf. Fr. 4 quoted in n. 56, where the word analogy pops up. The primary instances of prole p sis are: Ep. Hdt. 72, the discussion of time, where we read that time is not to be investigated as other accidents, which we envisage by means of prole p seis of those objects that they are the accidents of. Sedley argues (Sedley 1973) that the introduction of the concept post-dates the writing of the Letter to Herodotus and this passage is a later introduction into the Letter; cf. n. 175, as well as my discussion on pp. 43–4. Ep. Men. 124, the prole p sis concerning gods; K. D. 37, 38, the prole p sis in relation to justice, where we do not learn anything about how they are formed; On Nature, book XXV PHerc. 1191-10 sup. 5, also published in Long and Sedley 1987, 20 C where, thanks to Sedley’s reading, the evidence of the prole p sis of the cause applied to our selves is retained; On Nature book XXVIII Fr. 12 col. III, where we read that ‘all human error is exclusively of the form that arises in relation to preconceptions and appearances because of the manifold conventions of language’ (Sedley 1973, pp. 44–5), although the preceding words in the fragment leave it entirely unclear whether Epicurus ever held this view. In later Epicurean sources, the papyri of Philodemus have also preserved us the prole pseis of human being and body at De signis XXXIV 5–11; the goodness  of a poem at De Poem. 5 XXXIII 34–6; and that of a good household manager at De Oec. XX 8–32. Furthermore, Sextus’ discussion at Adv. Math.VII 208–15 implies the prole pseis of Plato, round and square.  For the view that prole p seis should be considered as some visual images cf. Long 1971; for that they have visual and propositional content, cf. Manuwald 1972, pp. 103–5; Barnes 1996; Striker 1996, p. 41; Konstan 2007, pp. 49–54; Masi 2015, p. 210, n. 23 and most recently Tsouna 2016, pp. 162–74; for that they are outline accounts, cf. Fine 2014, pp. 226–56; for that they are both mental acts and contents Asmis 1984, p. 63 and Morel 2007; and for that they are just mental acts, cf. Glidden 1985. As it will turn out, it is only a pseudo-problem; cf. pp. 30–3 and pp. 39–42. Cf. Epicurus’ distinction for the mental aggregate () in PHerc. 1420 corn. 2 z. 2 = 6 II = 3 N = Arr. [35.10], quoted on p. 38 as Fr. (b), the scholion to Ep. Hdt. 66 and Lucretius’ distinction between the animus (and mens) and anima. I agree with Masi 2015, p. 211 that it is not an ontological but a functional bipartition. And not prior to an investigation as it is conceived generally in the literature, cf., e.g., Asmis 2009, p. 86, which interpretation does not exclude the common scholarly agreement that prole p sis also offers a solution to Plato’s Meno paradox. That prole p sis would be the unification of the respective representations of the different sense-organs is the ingenious suggestion of Andree Hahmann (cf. Hahmann 2015, pp. 171–2). However he only hints at such a possibility and understands the function of prole p sis very differently from me, along some standard lines of interpretation, the criticism of which I provide on pp. 42–4. Cf. PHerc. 19/698, the anonymous Epicurean treatise on the senses, Monet 1996. Also cf. O’Keefe 2010, p. 39 on properties. Hahmann 2015, p. 167. (1)   [ ’][] []      (2)   

109

110 111

112 113

114 115 116

Self-awareness

61

        []       [ ] [  (3) ] [][ ] []       []    (4)    [ ’]  [], [] [][]     . (5)   [[]] [] [ ]  [.] Text and translation from Smith 1993, pp. 160–2 and p. 371,

respectively, with minor modifications in the translation. 117 . . .  [] [ ] [ ] in (2) according to Smith’s reconstruction. This might be the , a faculty of the soul for visualizing or imagining, which appears to be the consequence of the focusing of the mind () in this context, based on the inflowing images of perception. (For the function of the  in dreams cf. Ep. Hdt. 51.) 118 The process is even more complicated given the problem of compacting. 119 Lucretius DRN IV 728–31 leaves both possibilities open, however the context of these lines seems to imply that the distinction concerns the quality of simulacra per se. 120 (1)            (2)         (3)   . (4)     (5) . (6) ’            (7)     : (8)         ’          ; Text and translation from Asmis 1984, pp. 21–2, with modifications in

121 122 123 124 125 126 127 128

the translation. I do not follow the punctuation of the text in (1) – putting a comma after   – as in Long and Sedley 1987 vol. II, p. 92, and I agree with their comment that omitting the comma weakens the further definitions of prole p sis, which suits my interpretation of section (1) well. Cf. my discussion in Chapter 3, pp. 122–4. Glidden 1985, p. 181. In Glidden 1983, p. 187 he says: ‘Epicurean linguistics show little concern for how words acquire meanings and thereby describe what is happening in the world’. Cf. the reservations of Atherton 2009, pp. 199–200. Everson 1994, pp. 100–1 and p. 106. See most recently Morel 2007, p. 34. For the physical traces cf. Diogenes of Oinoanda, Fr. 9. III. 6–14 Smith 1993, quoted on pp. 28–9. Cf. pp. 29–30. Cf. Lucretius speaking importantly in the present tense about the connection between feelings and denotation in DRN V 1056–8: ‘Lastly, why is it so very remarkable that human beings, with their power of voice and tongue, should

62

129 130 131 132 133

134 135 136 137 138 139 140 141

142 143

Self-awareness designate things by different sounds according to their different feelings?’ Also cf. D. L. X 31: ‘They reject dialectic as superfluous, saying that it is sufficient that natural philosophers should proceed in accordance with the sounds belonging to things ()’, where admittedly ‘sounds belonging to things’ may imply merely the relation of signification. For thinking that such uniformity is implausible cf. Long 1971, p. 121. For the psychologist theory of meaning, although with some reservations, cf. Everson 1994. Scott 1995, p. 167. Cf. Everson 1994, p. 103, n. 75; Fine 2014, pp. 231–3. Cf. the interpretation of Barnes 1996, according to which prole p seis are meanings; see p. 201. If that was correct, Diogenes would be saying in (6) something like that the meaning of a name is primary to the name, which I cannot make sense of. Cf. DRN V 1028–32 for differentiating the natural phase of language by the concept of utility; cf. also Everson 1994, p. 98. What the doxaston depends on will become clear from my analysis of doxa orthe , three paragraphs below. Asmis 1984, pp. 61–83 for the formation of prole p sis. For katale p sis in an Epicurean irrational process cf. PHerc. 19/698 col. 9.1–10, quoted on p. 22. Cf. D. L. X 31. Glidden 1985, pp. 175–80. Sextus P. H. II 212. Glidden 1985, p. 179: ‘and just as Epicureans and Stoics apparently differ about ennoiai, so it might just well be that the two schools were not of one mind with respect to prole p seis either, despite the nearly unanimous consensus of scholars that the Stoics simply appropriated Epicurean prole p sis for their own, a most unlikely turn of events, if one stops to think about it’. Cf. also Manuwald 1972, pp. 103–5, who pointed out that the content of prole p sis is always something general. Showing much similarity with what Cicero’s Torquatus says, that ‘everything that comes before our mind has its origin in sense-perception’; De Fin. I 64, translated by Woolf, in Annas and Woolf 2001.

144     . Text and translation (with minor modification) from Long and Sedley 1987,

vol. II, p. 82 and p. 85, and vol. I, p. 76, respectively. Sextus Adv. Math. VIII 60, translation from Bett 2005, p. 100. Cf. also D. L.VII 52–3, the examples of which are very similar to those of Sextus. Asmis’ translation of prole p sis. Asmis 1984, p. 66. Asmis 2009, p. 89. Cf. Aristotle Metaph. 980b28–981a12 and A. Po. 100a3–9, and for the Stoics SVF 2.83. 151 However, we read in Lucretius DRN IV 724–44 that the eidol a of centaurs are produced in the air and hit the mind already in the shape of a centaur, so this example for the Epicureans needs to be in the category of the first kind of thought formation. 152 Cf. Philodemus De Poem. V, col. XXXV 11–14. 145 146 147 148 149 150

Self-awareness

63

153 Cf. Philodemus De Sign. Frg. 1, where Philodemus is talking about the construction of inferences ([  ][ ]) in harmony with the three Epicurean criteria. 154 Lucretius describes this focusing in analogy to the paradigmatic case of vision: ‘And because they [i.e. the images] are delicate the mind can only see (cernere) sharply those of them which it strains to see. Hence the remainder all perish, beyond those for which the mind has prepared itself ’ (DRN IV 802–4). Epicurus also applies the term to the sense-organs –  (Ep. Hdt. 50) – and as Sedley has convincingly shown we should take these   as restricted to the deliberate apprehension of images vs. the  , which is broader and in addition it includes the unintentional apprehension of images – i.e. seeing without looking or hearing without listening; cf. Sedley 1973, p. 25. 155 Cf. my previous, general reconstruction of the argument of D. L. X 33, where in (4) and (5) its verbal cognates express the flow or process of awareness in my understanding. 156 This flexibility or fluidity of the sense-organs is explained in detail by Masi 2015, pp. 220–2. 157 Cf. pp. 29–30. 158 It also appears once as a synonym for eidol a in Ep. Hdt. 46, however with some restrictive qualification (     ). For the reason why I disagree with Masi 2015, p. 210 that typoi should be propositional, see the following paragraph. 159 Thus, the standard of what makes a criterion is perfectly satisfied for prole p sis, given its strong and immediate connection with sense-perception. 160 Fr. (b) = 1420 corn. 2 z. 2 = 6 II = 3 N = Arr. [35.10], as in Laursen 1995, p. 91. 161 Cf. Masi 2006, p. 50, n. 144. The texts in question are: Lucretius DRN IV 973–83 and Diogenes of Oinoanda Fr. 9. Smith 1993, partially quoted on pp. 28–9. 162 For the same understanding of  cf. Laursen 1995, p. 91 and Masi 2015, p. 210. 163 Philodemus’ phrasing is also noteworthy in his De Poem. V, XXX 30–2:         []   []  , [’]  (in Mangoni 1993, p. 157). Philodemus is definitely not discussing proleptic processes in this context, yet his language – forming an outline of the prole pseis – at least alludes to it. For a strikingly similar use of  in  the context of expressing oneself cf. Plato’s Theat. 206d. Also cf. Leg. 775d. 164 Ep. Hdt. 46, where the word occurs in Epicurus, also implies a lesser degree of representation. 165               ’ ’                . Text and translation from Long and Sedley 1987,

vol. II, p. 76 and vol. I, p. 73, respectively. 166 For the evidence of the inclusion of pathe  in these processes cf. p. 22. 167 Cf. also D. L. X 34. 168            .

64

169 170

171 172

173 174 175 176

177

178 179 180

Self-awareness Sextus Adv. Math. VII 211–16 also makes it clear that the methods of verification and of falsification belong to two different types of judgements: the ones about what is yet to appear ( ) and the ones about what is non-apparent (). Thus, the first type is verified by witnessing and falsified by not witnessing while a judgement or opinion of the second type is verified by not counterwitnessing or counter-witnessing. Also cf. Adv. Math. VII 212, where Sextus makes it explicit that a judgement that p is true only if it is confirmed by further perception that it is p. Cf. Plutarch Adv. Col. 1121C–D and 1124B. In 1121D–E he says: ‘If there is a difference in the affection of those who stand afar and those who have come close, it is false that neither a presentation nor a perception is more evident than a different one’. Asmis 1984, p. 151. See a list of modern scholars, who have sided with Plutarch, and hold that only some presentations were recognized by the Epicureans as evidence, in Asmis 1984, pp 151–2, nn. 25 and 26. I agree with Asmis’ observation, ibid.: ‘These interpretations are all . . . just like Plutarch’s criticism, critical opinions about what Epicurus should be saying rather than explanations of what he is saying’. As even Striker admits; cf. Striker 1996, p. 49. D. L. X 31 suggests the application of the mind as a criterion which was introduced by later Epicureans; cf. Tsouna 2016, pp. 186–93 and pp. 201–17. Cf. n. 156. In On Nature book XXVIII Fr. 12 col. III, Epicurus states that they used to think that ‘all human error is exclusively of the form that arises in relation to preconceptions and appearances because of the manifold conventions of language’, cf. Sedley 1973, pp. 44–5.The context of the fragment does not confirm if this was really the opinion of Epicurus, but if it had been, it would be compatible with both my interpretation and the standard explanation of error in Epicurus’ theory. Cf. Long 1971, pp. 119–22, who explicitly argues for both claims on p. 120: ‘The position about the criteria, as I understand it, is that  are necessary for the formation and testing of all assertions and objective judgements. Sensations and feelings provide us with data for making judgements. But the test of whether a judgement about such data is true requires a check, under optimum conditions, that the data match or are not inconsistent with our preconceptions of what they are data of’ (Long’s italics). Asmis 1984, p. 51 talks about (mis)matching in nonscientific inquiry in the sense of (2); in scientific inquiry her view is that scientific concepts are the extension of prole pseis (p. 52). Striker 1996 argues for (2). Why  I disagree with this formulation should be clear from how I conceive of the Epicurean formation of the abstract concept of death, on p. 56 and see more on this later on pp. 64–6. Cf. Bailey 1928, p. 248 and Long 1971, p. 122. For the contestation of prole p sis being an independent criterion see Furley 1967, p. 206. For its irreducibility to perception cf. Striker 1996, p. 30, n 13. One may wish to object that hearing a word ‘man’ or ‘horse’ has nothing to do with the visual (or, for that matter, aural) recognition of actual men or horses. However, D. L. X 32 – according to which all thoughts arise from senseperceptions – D. L. X 33 (6) – according to which naming presupposes learning the outline (typos) of things by means of prole p sis – and that prole p sis are immediately based on our sensations, all lend support to the idea that even upon hearing a word such as the word ‘man’ or ‘horse’, its interpretation initiates processes analogous to sense-perceptions.

Self-awareness

65

181 I accept Sedley’s hypothesis (cf. Sedley 1973, pp. 14–5) that the primary concept ( ) distinguished in Epicurus’ first principle of methodology (Ep. Hdt. 37–8) is an earlier formulation and conception of prole p sis, however, with the important restriction that the Epicurean theory, in its fully fledged form, does not denote exactly the same things. Even if proton  ennoe m  a and prole p sis certainly were intended to fulfil the same function, when Epicurus developed his theory he also slightly modified what that function was: the primary concept, most likely, indicated a concept, which was literally taken to be the primarily formed concept of a thing during one’s psychological development as opposed to prole p sis, which, in the mature version of the theory, stands for the process of recognition in my interpretation. If such an understanding is correct, then the change of terminology – Sedley’s hypothesis – also makes more sense since Epicurus’ reason for the change must have been that prole psis  is not only a more easily but automatically accessible criterion because of its immediacy, in comparison with one’s concepts formed earlier on in one’s life. For how an investigation comes short of an infinite regress, which also seems to threaten the method of witnessing/counter-witnessing, cf. my discussion of Epicurus’ first principle of methodology (Ep. Hdt. 37–8) in Chapter 3, pp. 122–4. 182 Fine 2014, p. 230, n. 15. 183 For the two most current formulations of the realist and idealist interpretations of Epicurus’ theology cf. Konstan 2011 and Sedley 2011, respectively; also cf. Hammerstaedt 1996. On the discussion of the prole p sis of gods cf. Tsouna 2016, pp. 174–85. 184 It is generally agreed that Epicurus is trying to avoid consequences of Democritean atomism, cf. Sedley 1983, p. 32, Long and Sedley 1987 vol. I, p. 108, Annas 1992, p. 128, Warren 2002, p. 197. Some have suggested that Epicurus reacts against the entire Democritean tradition, including Metrodorus, Anaxarchus and Nausiphanes, cf.Warren 2002, p. 197, n. 16. For identifying the opponent as a fatalist, cf. O’Keefe 2005, p. 81. 185 I call this connected and cohesive set of fragments Text P (P for prole p sis) to indicate my main interest in them. Text from Masi 2006, p. 95. Since the text has received a lot of attention ever since Sedley 1983 (cf. Laursen 1988, Annas 1992 and 1993, Sharples 1991–1993, Hankinson 1998, Bobzien 2000, O’Keefe 2002 and 2005 and Masi 2006), I provide only those parts of the Greek original that are the most important for my discussion. 186 Thus, the part. pl. perf. active form of  ([]  ), literally meaning ‘the pre-conceiving outlines’ corroborates my interpretation according to which the typoi are the products of the irrational proleptic processes, initiated by sense-perception. 187 Cf. pp. 24–6. 188 Cf. pp. 13–15. 189 I call this fragment Text M (M for memory) to indicate that this is the key fragment for Epicurus’ conception of memory. Text M = 1191 corn. 4 pz. 2 z. 1 = –24 inf./1191 corn. 7 pz. 1 z. 2-3 = –23 sup., and 697 corn. 2 pz. 2 z. 4 (= col.5), and 1056 corn. 5 z. 1, from Laursen 1997, pp. 16–17, modified along my suggestions. 190 I suggest connecting the genitives of lines 7–8 to    [] in lines 6–7. It is a pointed parallel of lines 5–6 and provides a more natural reading of the Greek with the conjunctive particle  in line 5 answered by  in line 6 in the corresponding clause. 191 I will establish my understanding of  in Chapter 2, pp. 86–92.

66

Self-awareness

192 Also cf. PHerc. 1056 corn. 4 z. 4 in Masi 2015, pp. 224–5 in her translation: ‘At some point [the mind] remembered or got an affect analogous to the remembering and stayed in a state where with calm as well as in the investigation of the things that would have produced the [most everlasting] and greatest fear according to the circumstances . . . mainly physical pain, researches for the truth, on the basis of what is defined, regarding both human beings and the things above human beings . . .’ Also cf. Fr. 8 on p. 52 later. I will discuss Text M in more detail in Chapter 2, pp. 80–1 and pp. 87–9. For a completely different understanding, partially due to our different understanding of , cf. Masi 2006, pp. 183–201 and 2015, p. 217. 193 I will argue against the reductionist understanding of Epicurus’ philosophy of mind at length in Chapter 2. 194 Laursen 1995, p. 42. 195 Masi 2015, p. 203. 196 Fr. 7 = 1056 corn. 4 z. 2 = 6 II = 12 N = 892 O = Arr. [34.17], as in Laursen 1995, pp. 107–8. 197 Fr. 8 = 697 corn. 2 pz. 2 z. 3 = col. 3= Arr. [34.18.1-3], and 1056 corn. 4 z. 3 = 6 III = 13 N = 893 O = Arr. [34.18], as in Laursen 1995, pp. 108–9, except I inserted a full stop after []. 198 Cf. n. 33. 199 Hence, my findings corroborate Remes’ observation that ‘in antiquity selfhood has to be traced in the junctures of metaphysics, philosophical psychology and ethics’; cf. Remes and Sihvola 2008, p. 3.

Bibliography Algra, K. A., Barnes, J., Mansfeld, J. and Schofield, M., eds., (1999) The Cambridge History of Hellenistic Philosophy, Cambridge. Annas, J. (1985) ‘Plato on Self-Knowledge’, in D. J. O’Meara, ed., Platonic Investigations, Washington: 111–38. —. (1992) Hellenistic Philosophy of Mind, Berkeley. —. (1993) ‘Epicurus on Agency’, in J. Brunschwig and M. Nussbaum, eds., Passions and Perceptions, Cambridge: 53–71. Annas, J. and Betegh, G., eds., (2016) Cicero’s De Finibus, Cambridge. Annas, J., ed., Woolf, R. transl. (2001) Cicero: On Moral Ends, Cambridge. Arrighetti, G. (1973) Epicuro. Opere: Introduzione, testo critico, traduzione e note, 2nd ed., Turin. Asmis, E. (1984) Epicurus’ Scientific Method, Ithaca/London. —. (2009) ‘Epicurean Empiricism’, in J. Warren, ed., The Cambridge Companion to Epicureanism, Cambridge: ch. 5, 84–105. Atherton, C. (2009) ‘Epicurean Philosophy of Language’, in J.Warren, ed., The Cambridge Companion to Epicureanism, Cambridge: ch. 11, 197–215. Bailey, C. (1928) The Greek Atomists and Epicurus, Oxford. Barnes, J. (1996) ‘Epicurus: Meaning and Thinking’, in G. Giannantoni and G. Gigante, eds., Epicureismo greco e roman: Atti del congresso internazionale Napoli 19–26 Maggio 1993, 3 vols., Naples: 197–220. Bett, R., ed., transl. (2005) Sextus Empiricus: Against the Logicians, Cambridge. Blank, D. (2013) Philodemus, at http://plato.stanford.edu/entries/philodemus/ Bobzien, S. (2000) ‘Did Epicurus Discover the Free Will Problem?’, Oxford Studies in Ancient Philosophy 19: 287–337.

Self-awareness

67

Bollack, J., Bollack, M. and Wismann, H., eds., transl. (1971) La Lettre d’Epicure, Paris. Bourgey, L. (1969) ‘La doctrine épicurienne sur le role de la sensation dans la connaissance et la tradition grecque’, in Association Guillaume Budé, Actes du VIII Congrès, Paris 5–10 avril 1968, Paris: 252–8. Brunschwig, J. (1986) ‘The Cradle Argument in Epicureanism and Stoicism’, in M. Schofield and G. Striker, eds., The Norms of Nature, Cambridge: 113–44. —. (1994) Papers in Hellenistic Philosophy, Cambridge. Cooper, J. M. (1999) Reason and Emotion: Essays on Ancient Moral Psychology and Ethical Theory, Princeton. —. (2012) Pursuits of Wisdom, Princeton. Crystal, I. M. (2002) Self-Intellection and its Epistemological Origins in Ancient Greek Thought, Aldershot. De Lacy, P. H. (1939) ‘The Epicurean Analysis of Language’, American Journal of Philology 60: 85–92. —. (1958) ‘Pigs and Epicureans’, Classical Bulletin 34: 55–6. Delattre, D. and Pigeaud, J. eds., (2010) Les Épicuriens, Paris. De Witt (1943) ‘Epicurus: All Sensations are True’, Transactions of the American Philological Association 74: 19–32. Dorandi, T. (1983) ‘Sudhaus editore di Epicuro’, Cronache Ercolanesi 13: 183–90. Everson, S. (1994) ‘Epicurus on Mind and Language’, in S. Everson, ed., Language (Companions to Ancient Thought 3), Cambridge: 74–108. Everson, S., ed., (1998) Ethics (Companions to Ancient Thought 4), Cambridge. Fine, G. (2014) The Possibility of Inquiry: Meno’s Paradox from Socrates to Sextus, Oxford. Fish, J. and Sanders, K. R., eds., (2011) Epicurus and the Epicurean Tradition, Cambridge. Flashar, H. and Gigon, O. (1986) Aspects de la Philosophie Hellenistique, Fondation Hardt 32,Vandoeuvres/Geneva. Frede, M. and Striker, G., eds., (1996) Rationality in Greek Thought, Oxford. Furley, D. (1967) Two Studies in the Greek Atomists, Princeton. Giannantoni, G. and Gigante, G., eds., (1996) Epicureismo greco e roman: Atti del congresso internazionale Napoli 19–26 Maggio 1993, 3 vols., Naples. Glidden, D. K. (1971) The Epicurean Theory of Knowledge, PhD dissertation, Princeton University. —. (1979) ‘Epicurus on Self-Perception’, American Philosophical Quarterly 16: 297–306. —. (1983) ‘Epicurean Semantics’, in : Studi sull’ Epicureismo greco e latino offerti a Marcello Gigante, 2 vols., Naples: vol. I: 185–226. —. (1985) ‘Epicurean prole p sis’, Oxford Studies of Ancient Philosophy 3: 175–217. Gomperz, T. (1876) ‘Neue Bruchstücke Epikur’s insbesondere über die Willensfrage’, Sitzungsberichte der Kaiserlichen Akademie der Wissenschaften in Wien, PhilosophischHistorischen Classe, 83: 87–98. —. (1879) ‘Die Überreste eines Buches von Epikur  ’, Wiener Studien 1: 27–31. Hahmann, A. (2015) ‘Epicurus on Truth and Phantasia’, Ancient Philosophy 35: 155–82. Hammerstaedt, J. (1996) ‘Il ruolo della  epicurea nell’ interpretazione de Epicuro, Epistula ad Herodotum 37f ’, in G. Giannantoni and G. Gigante, eds., Epicureismo greco e roman: Atti del congresso internazionale Napoli 19–26 Maggio 1993, 3 vols., Naples: 221–37. Hankinson, R. J. (1998) Cause and Explanation in Ancient Greek Thought, Oxford. Hicks, R. D. (1931) Diogenes Laertius: Lives of Eminent Philosophers, 2 vols., Loeb edition, London and Cambridge, MA.

68

Self-awareness

Inwood, B. (2016) ‘The Voice of Nature’, in J. Annas and G. Betegh, eds., Cicero’s De Finibus, Cambridge: 147–66. Kahn, C. (1998) Plato and the Socratic Dialogue: The Philosophical Use of a Literary Form, Cambridge. Kemke, I., ed. (1884) Philodemi De musica quae exstant, Leipzig. Konstan, D. (2006) ‘Epicurean “Passions” and the Good Life’, in B. Reis and S. Haffmans, eds., The Virtuous Life in Greek Ethics, Cambridge: 194–205. —. (2007) ‘Response to Morel’, Proceedings of the Boston Area Colloquium in Ancient Philosophy 23: 49–54. —. (2008) A Life Worthy of the Gods, Las Vegas, Zurich, Athens. —. (2011) ‘Epicurus on the Gods’, in J. Fish and K. R. Sanders, eds., Epicurus and the Epicurean Tradition, Cambridge: 53–72. Laursen, S. (1987) ‘Epicurus’ On Nature Book XXV’, Cronache Ercolanesi 17: 77–8. —. (1988) ‘Epicurus On Nature XXV (Long–Sedley 20, B, C and j)’, Cronache Ercolanesi 18: 7–18. —. (1995) ‘The Early Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanesi 25: 5–109. —. (1997) ‘The Later Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanesi 27: 5–82. Lee, E. N. (1978) ‘The Sense of an Object: Epicurus on Seeing and Hearing’, in P. K. Machamer and R. G. Turnbull, eds., Studies in Perception: Interrelations in the History of Philosophy and Science, Ohio: 27–59. Long, A. A. (1971) ‘Aisthesis, Prolépsis and Linguistic Theory in Epicurus’, Bulletin of the Institute of Classical Studies, University of London 18: 114–33. —. (1986) ‘Pleasure and Social Utility’, in H. Flashar and O. Gigon, eds., Aspects de la Philosophie Hellenistique, Fondation Hardt 32, Vandoeuvres/Geneva: 283–316, repr. in Long 2006: 178–201. —. (1999) ‘The Socratic Legacy’, in K. A. Algra, J. Barnes, J. Mansfeld and M. Schofield, eds., The Cambridge History of Hellenistic Philosophy, Cambridge: 617–41. —. (2006) From Epicurus to Epictetus: Studies in Hellenistic and Roman Philosophy, Oxford. Long, A. A. and Sedley, D. N. (1987) The Hellenistic Philosophers, 2 vols., Cambridge. Mangoni, C. (1993) Filodemo: Il quinto libro della Poetica (PHerc. 1425 e 1538), La scuola di Epicuro 14, Naples. Manuwald, A. (1972) Die Prole p sislehre Epikurs, Bonn. Masi, G. F. (2006) Epicuro e la Filosofia della Mente: Il XXV Libro dell’Opera “Sulla Natura”, Sankt Augustin. —. (2015) ‘Memory, Self and Self-Determination. The Mind–Body Relation in Epicurus’ Psychology’, in D. De Brasi and S. Föllinger, eds., Anthropologie in Antike und Gegenwart: Biologische und philosophische Entwürfe vom Menschen, Sankt Augustin: 203–30. Mitsis, P. (1988) Epicurus’ Ethical Theory:The Pleasures of Invulnerability, Ithaca/London. Monet, A. (1996) ‘[Philod`eme, Sur les sensations]: P.Herc. 19/698’ Cronache Ercolanesi 26: 27–126. Morel, P.-M. (2007) ‘Method and Evidence: On Epicurean Preconception’, Proceedings of the Boston Area Colloquium in Ancient Philosophy 23: 25–48. Nussbaum, M. C. (1994) The Therapy of Desire: Theory and Practice in Hellenistic Ethics, Princeton. O’Keefe, T. (2002) ‘The Reductionist and Compatibilist Argument of Epicurus’ On Nature Book 25’, Phronesis 47: 153–86.

Self-awareness

69

—. (2005) Epicurus on Freedom, Cambridge. —. (2010) Epicureanism, Berkeley. Puglia, E. (1987) ‘PHerc. 1420/1056: Un volume dell’ opera “Della natura” di Epicuro’, Cronache Ercolanesi 17: 81–3. Reis, B. and Haffmans, S., eds., (2006) The Virtuous Life in Greek Ethics, Cambridge. Remes, P. and Sihvola, J., eds., (2008) Ancient Philosophy of the Self, Dordrecht. Rist, J. M. (1972) Epicurus: An Introduction, Cambridge. Schofield, M. (1996) ‘Epilogismos: An Appraisal’, in M. Frede and G. Striker, eds., Rationality in Greek Thought, Oxford: 221–39. Schofield, M. and Striker, G., eds., (1986) The Norms of Nature, Cambridge. Scott, D. (1995) Recollection and Experience, Cambridge. Scott, W., ed., (1885) Fragmenta Herculanensia, Oxford. Sedley, D. N. (1973) ‘Epicurus, On Nature Book XXVIII’, Cronache Ercolanesi 3: 5–83. —. (1983) ‘Epicurus’ Refutation of Determinism’ in : Studi sull’ Epicureismo greco e latino offerti a Marcello Gigante, 2 vols., Naples: vol. I, pp. 11–51. —. (1996) ‘The Inferential Foundations of Epicurean Ethics’, in G. Giannantoni and G. Gigante, eds., Epicureismo greco e roman: Atti del congresso internazionale Napoli 19–26 Maggio 1993, 3 vols., Naples: 313–19; repr. in Everson, ed., 1998: 129–50. —. (1998) Lucretius and the Transformation of Greek Wisdom, Cambridge. —. (2011) ‘Epicurus’ Theological Innatism’, in J. Fish and K. R. Sanders, eds., Epicurus and the Epicurean Tradition, Cambridge: 29–52. Sharples, R. W. (1991–1993) ‘Epicurus, Carneades and the Atomic Swerve’, Bulletin of the Institute of Classical Studies, University of London 38–9: 174–90. Smith, M. F. (1966) ‘Some Lucretian Thought Processes’, Hermathena 102: 73–83. —. (1993) Diogenes of Oinoanda:The Epicurean Inscription, Naples. —. (2001) Lucretius: On the Nature of Things, Indianapolis/Cambridge. Sorabji, R. (2006) Self: Ancient and Modern Insights about Individuality, Life, and Death, Oxford. Striker, G. (1974) ‘ ’, Nachrichten der Akademie der Wissenschaften in Göttingen, Philologisch-Historische Klasse, no. 2, Göttingen, repr. in Striker 1996: 22–76. —. (1996) Essays on Hellenistic Epistemology and Ethics, Cambridge. Trapp, M. B. (1997) Maximus of Tyre:The Philosophical Orations, Oxford. Tsouna, V. (2009) ‘Epicurean Therapeutic Strategies’, in J. Warren, ed., The Cambridge Companion to Epicureanism, Cambridge: ch. 14, 249–65. —. (2016) ‘Epicurean Preconceptions’, Phronesis 61: 160–221. Warren, J. (2000) ‘Diogenes Epikourios: Keep Taking the Tablets’, Journal of Hellenic Studies 120: 144–8. —. (2002) Epicurus and Democritean Ethics: An Archeology of Ataraxia, Cambridge. Warren, J., ed., (2009) The Cambridge Companion to Epicureanism, Cambridge. Woolf, R. (2004) ‘What Kind of Hedonist was Epicurus?’, Phronesis 49: 303–22.

2

Agency and atomism

Introduction No means of predicting the future really exists, and if it did, we must regard what happens according to it as nothing to us. (D. L. X 135) The early parts of book XXV of On Nature give us a picture of how we conceive of ourselves according to the pathologikos tropos and aitiologikos tropos. Self-reflective thinking, it is maintained, derives from self-awareness through one’s pathe  and it eventually leads to the proper conception of oneself as a morally responsible agent in the cosmos, recognition of which is essential for one to properly conceive of oneself and to live and conduct a eudaimon life. The ‘thin’ idea of the self – which marks collectively the distinctive mental and bodily character of a self-reflexive being, and has emerged from the early parts of the fragments of book XXV – is substantiated to a further extent by an analysis of causal agency in the later parts of the same book. It is no surprise that Epicurus, a committed atomist, desired to explain the possibility of such a conception of oneself as compatible with his physics, since he wanted to liberate his theory from the determinist implications of the earlier atomists’ physics such as Democritus’. In the later parts of book XXV, Epicurus continues with the aitiologikos tropos and gives an account of the causal powers of the human self compatible with his atomist theory. It is of paramount importance for him to accommodate the result of the intellectual activity of self-reflective thinking, that is, the conception of responsibility. The later fragments of On Nature XXV have received a great amount of scholarly attention since David Sedley’s analysis of some of the major fragments in a seminal article.1 The major debate concerning these fragments has been whether they attribute a reductionist or an anti-reductionist theory of mind to Epicurus.2 Although many fragments have been taken into account from the later parts of book XXV to clarify this issue, very little has been said about the strength of their relationship to the earlier fragments of the same book.3 I have already examined their connection to a certain extent during the course of my analysis of the aitiologikos tropos.4 My primary intentions in this and the next

Agency and atomism

71

chapters are to do that work in more detail and to argue that based on a comprehensive analysis of the fragments of book XXV we should attribute a version of non-reductive physicalism to Epicurus. Much of what I wish to say about book XXV as a whole is in Chapter 3, while this chapter is devoted mainly to scrutinizing Epicurus’ philosophy of mind. This investigation will start with a general outline of reductionism and antireductionism and their attribution to Epicurus. It appears that the disagreement regarding Epicurus’ commitment to reductionism or anti-reductionism is fundamentally based on the assumptions with which one approaches the relevant fragments. I think it is impossible not to have some extra-textual assumptions; however, if one applies modern terminology when analysing ancient texts, one unavoidably runs the risk of placing too much weight on the philosophical reconstruction of the fragments. This leads on occasions to unavoidable distortions of the historical reconstruction – often done under the flag of the principle of charity. Obviously, the latter cannot be an empty bag of tenets having no philosophical relevance. However, if we attempt to analyse ancient texts in terms of their philosophical importance by applying our modern conceptions, we may miss out on the real historical perspective which simply may not be represented in our modern positions, because it has either proved to be false or at least is thought to be fallacious these days or it has not been solved. Thus, I think it is very useful to present the general outline of the two most influential readings of Epicurus’ theory as reductionist and anti-reductionist and the ideas of their strongest advocates, Tim O’Keefe and David Sedley, respectively before introducing any textual evidence. Presenting these interpretations at such a level of generality will clearly show the assumptions with which these modern commentators approach the later fragments of book XXV, and the introduction through their antithetical interpretations will provide the best insight into the key philosophical difficulties of these fragments. Furthermore, such a procedure allows me to reflect on them more easily during my subsequent, textual analysis on the strength of which I will evaluate the competing interpretations and put forward my own understanding of Epicurus’ theory. Since there is also disagreement concerning how Epicurus understood himself in relation to his forerunners, I will discuss briefly how Epicurus’ reductionist or anti-reductionist position is thought to have been a response to Democritus. It will only be after all these proceedings that I can turn to the fragments and hopefully advance our understanding of them at least one step further.5 The first part of the discussion will revolve around the central fragments of On Nature XXV concerning moral agency, starting with the fragments I numbered as Fr. 10–14. Since the analysis of the early fragments of book XXV in the previous chapter has already raised a number of interpretative issues concerning some of the concepts, it is the task of this chapter to finish that work, which will naturally include revisiting the major fragment on memory,Text M. In the conceptual analysis of the idiosyncratic Epicurean term, , I will also use Fr. 15 and Fr. 17. For the complete philosophical reconstruction of Epicurus’ conception of agency I will introduce my hypothesis of how the

72

Agency and atomism

atomic swerve and the conception of the soul as a mixture supports the nonreductive physicalist interpretation of Epicurus’ philosophy of mind.

Reductionism vs. anti-reductionism in Epicurus’ philosophy of mind Reductionism may be conceived as ontological or as merely theoretical. In the first case, such a theory holds that an entity is exhaustively reducible to another group of entities. The standard example is ‘Water = H2O’. Nonetheless, the example includes the conventional mark of identity, and, as Sedley has importantly noted, one needs to be cautious when explaining ontological reductionism in terms of identity, since identity is a reciprocal relation, whereas the relation of being reducible to is not.6 Sedley’s worry is more obvious when reductionism is applied in the philosophy of mind. Describing some mental state as identical with some underlying physical processes, and arguing for the reducibility of the former to the latter, does not imply a reciprocal relation – a belief that the physical process could be reduced in just the same way to the mental. Hence, as Sedley explains, the reduction of the mental to the physical does not imply an identity relation between the two, but it is an asymmetrical relation between them in the sense that, in this case, the physical is ontologically more fundamental to one’s theory and has a greater explanatory power. Accordingly, a more precise description of ontological reductionism reads as follows: the theory must hold that an entity is exhaustively reducible to another entity, and that the latter entity, to which the former is being reduced, is ontologically more fundamental. Another type of reductionism is theoretical reductionism, which holds that one theory can be reduced to another and that all scientific theories eventually may be reduced to a single theory through the process of theoretical reduction. Such reduction may have the merits of gaining more explanatory power if the basic theory, to which all the others are reduced, has a more general toolkit and can explain more events. In O’Keefe’s understanding Epicurus held that macroscopic entities can be exhaustively reduced to their component atoms. Macroscopic bodies do not have any causal powers beyond those of the atoms that they are composed of, because they are nothing more than the compounds () and aggregates () of these atoms; consequently, their properties can be exhaustively accounted for on the basis of the shape, structure and pattern of motions of their constituent atoms. Such a reading of the Epicurean theory is along the lines of ontological reductionism. O’Keefe argues that his interpretation applied in the Epicurean philosophy of mind does not make the mind into something merely ‘epiphenomenal’. Mental facts in the reductionist philosophy of mind are not ‘mere appearances’ of atomic facts, e.g. to describe pain as c-fiber stimulation is nothing else but to admit that pain is something, which makes it clear that pain is, indeed, something real. ‘If the mind is identified with a group of atoms, and the group of atoms is real, so too is the mind’.7

Agency and atomism

73

O’Keefe argues for the reductionist interpretation of Epicurus’ philosophy of mind from the obvious fact that wholes have properties that their parts do not. He states that just because the mind has many capacities – such as perception, deliberation and emotions – which its constituent atoms do not possess, it does not follow that Epicurus’ philosophy of mind is not reductionist. If we can explain phenomena consistently in terms of their constituent atoms, then reductionism is not to be dismissed for this reason. And since Lucretius’ poem (DRN) is full of such explanations, argues O’Keefe, reductionism neatly fits. O’Keefe’s third main consideration is that reductionism can equally allow for the appeal to structural elements in one’s explanation of the mind if these elements are eventually reducible to relationships among atoms, and O’Keefe thinks that Epicurus only allowed for structural elements of such a kind. Anti-reductionism may be conceived as ontological or epistemological and is often a reaction to reductionism in the philosophy of mind. Such an ontological theory holds, at a minimum, that an entity is not exhaustively or not at all reducible to another entity and, thus, requires additional principles that are not entailed by the laws of the basic constituents. Epistemological anti-reductionism holds that since our mental capacities are limited we would not be able to grasp all the physical explanations of the phenomena even if we knew all the laws of physics. In Sedley’s understanding of Epicurean anti-reductionism, though bodies and the void are indeed the fundamental elements of Epicurus’ metaphysics, their primacy in the causal realm is not to be carried over to Epicurean ontology, since his philosophy treats phenomenal entities equally with the basic building blocks of the universe, without privileging the reality of the latter over the former. Consequently, there must be some properties in the world, which are irreducible to the atomic level, namely some secondary properties. In the philosophy of mind, mental properties and events are not merely irreducible to the atomic level but are emergent, non-physical properties, not governed by the laws of physics, having their own causal powers, as a consequence of which causal interactions are not merely limited to bodies – according to Sedley’s interpretation – but are ongoing between bodies and their non-corporeal properties. That is to say, not all the causal powers can be reduced to the atomic level of description. According to Sedley, reducibility does not straightforwardly follow from the endorsement of an atomist theory, since although we all may be atomists, not all of us would reduce beauty, love and so forth to the atomic level. Second, it is extremely hard to unearth reductionist remarks from the extant writings of Epicureanism.8 Third, Epicurus must have been self-consciously avoiding reductionist lines of thought, since he objected to Democritus’ theory for its sceptical and deterministic implications, both of which he probably found to arise directly from Democritus’ reductionism. As opposed to Sedley, in O’Keefe’s understanding Democritus does not question the reality of sensible qualities as a result of his reductionism, since secondary qualities such as shape and resistance can also be identified with the primary properties of atoms; hence, if there are qualities on the macroscopic

74

Agency and atomism

level which are also to be found on the microscopic level, reductionism alone does not suffice to question the reality of sensible qualities. Rather, O’Keefe thinks Democritus concluded the unreality of sensible qualities because of the impossibility of identifying them with physical states; that is, he conceded scepticism on the strength of being an eliminative materialist. According to O’Keefe, Epicurus’ reductionist explanation is a response to Democritus’ ensuing scepticism, by means of which Epicurus believed in accounting for and preserving the reality of the phainomena. As we can see, how we explain Epicurus’ reaction to Democritus’ theory has been thought to be crucial for our understanding of Epicurus’ position. If Democritus’ alleged scepticism is to follow from his reductionism, there seems to be a good historical reason for us to think that Epicurus wanted to avoid such a reductionist theory in his philosophy of mind, since in direct opposition to scepticism he believed not only in the reality of the phainomena but he even endorsed the claim that every perception is true – and this is the line of argument Sedley takes. Nonetheless, if Epicurus did not view Democritus’ supposed scepticism as a direct result of his reductionism, then the road is open to interpreting Epicurus as entertaining an out-and-out reductionist position – the direction O’Keefe takes. I am not going to pursue the question of Epicurus’ motivation in detail because I do not believe that its correct conception – if it can be reconstructed at all – will grant victory to one of the above interpretations, and also because a comprehensive account of Democritean epistemology and its associated metaphysics is not within our purview here.9 Yet it should suffice to say that to understand Democritus as a sceptic along the lines of O’Keefe’s understanding has to face a simple but strong objection: had Democritus indeed thought it impossible to describe sensible qualities by appealing to certain atomic configurations, then his atomism would not have been able to account for the regularities of the phenomenal world, which he held to be the outcome of an atomic vortex. Even if Democritus discounted the reality of the phenomenal world only in the account of his metaphysical theory, he could not have entertained a metaphysical point of view which is ontologically prior and is unable to account for the structure of the universe, thus leading to atomic chaos. One may rightly note that Democritus’   contained many topics without any associated atomic explanation. At least in the paraphrases of these topics we do not encounter any atomic explanations but only the elements and elemental qualities. As Kornél Steiger has argued, unless we entertain the very unlikely hypothesis that the doxographers entirely reworked originally atomic explanations, Democritus seems to have had a natural philosophy not explicated on the basis of his atomist theory.10 As Steiger notes, to describe the world from two different perspectives was nothing alien in early Greek philosophy, if we think of Parmenides and Empedocles. Accordingly, Steiger concludes, the world, on the one hand, as the object of noetic vision has two principles: the full and the empty. Everything we perceive is, in reality, just the result of certain atomic motions, complex structures and of the void. On the other hand, the world as the object of our perceptions

Agency and atomism

75

and sensory experience is built up by elements and elemental qualities. In consequence, Democritus does not seem to have thought it necessary to take his metaphysics to be the explanatory principle for his natural philosophy. I find these points persuasive and, therefore, I tend to disagree with the conception of Democritus as a sceptic: Democritus questioned the reality of the phainomena merely on the strength of his metaphysical theory, but having laid down the matrix of atomism he was ready to embark on a different kind of explanation of our sensory experiences, without taking on a Lucretian type of project of attempting to give a minute atomic analysis of most of the phenomena in our world, with a few exceptions, because Democritus may have found it sufficient to indicate that he had laid down the foundations of the theoretical possibility of a complete explanation.11 If this is correct, we may also notice that, according to the reductionist interpretation of Epicurus, he would have altered Democritus’ theory merely by saying that since the atoms are real, whatever they constitute is also real; or put differently, if an atomist metaphysical theory is paired with an ontologically reductionist programme, then we have no other option left but to accept the reality of the phenomenal world. But that seems to amount to no more compared to the Democritean explanation than finding a way to accommodate some primarily linguistic conventions of the perceptible world. Was this really the Epicurean project? I believe not.12 To see the alternatives, let me now turn to the later parts of book XXV of On Nature.

The central fragments of On Nature XXV The causal manner of explanation is dominant in the language of the later parts of book XXV. It is already traceable in the early parts even prior to the pathologikos tropos,13 and, as I have said in Chapter 1, I do not think that Epicurus used the aitiologikos tropos as a successive level of explanation, which is to follow the pathologikos tropos in due course. The fragments of book XXV instead give us a picture where these modes are used without such a restricted procedure, even where both are applied concerning the same subject, as I have argued in the case of self-intellection. Having discussed our own conception of ourselves in the early parts of book XXV, in the later parts Epicurus is ready to embark on the explanation of what it takes for an atomist to entertain the possibility of responsible action. For such an explanation, the aitiologikos tropos seems to be the most appropriate mode, since what is at stake for Epicurus is to show that one’s self is not inefficacious, i.e. it is not merely epiphenomenal. This much is uncontentious. I will start the analysis with some fragments, which have been taken to be decisive in judging the reductionist and anti-reductionist interpretations, and I will show that none of the individual interpretations of O’Keefe and Sedley fit Epicurus very neatly. First I present a group of consecutive fragments as a continuous text based on Laursen’s 1997 edition as well as my translation of them.14

76

Agency and atomism Fr. 10–Fr. 14

5

10

15

20

25

30

35

(Fr. 10) . . .[]  []  []      []   []  []  [] []  [] []  [][]  [  ] . [] [] [][]. . . (Lacuna of a few letters) (Fr. 11–Fr. 12) [ ]   []                  ’    [] (         )       , .[.].        []                      ’     [ ]  [] [][]   []  ’ [] []  []   [] [] []    [ ] []  [][  ] [][][ ’   ]  []  [  +/–5/6 ][+/–1/2][. . .].[ ]  [ ]..[ - ] [] [    ] [  ] [ ]   [ - ].[ - ][] [ - ]   []  [ - ]. . .[.]. . . (Lacuna of roughly 30 letters) (Fr. 13). . . [ ]   []  []     []   ’  [] []  [].    []   [ ]                ’   []   []                          []             []   []      [ ’]    [  ’ ] []  [ ][]   [ - ] [+/–5/6 ].[+/–1/2 ].[ - ].[.][+/–7/8 ]. . .[+/– 1/2 ][ - ] .[.][+/– 2/3] (Fr. 14)      ,   []    [  - ] [.] [[. . .]] .[ +/–7/8 ]  [] [][ ]  [], ’ [ ][],  [] [  ] []     []   []    [][]   .[ - ] .      <  >    . . .[ +/–1/2 ][+/–3/4]   []  [ ] [] [+/– 8/10 ][ +/–9/11][ . . . (Fr. 10) . . . [not only] coming about in accordance with the [atoms] thrust together, but also the [atoms] thrust together, clearly, in the same way. For if thrust together in this way, the same atoms ([])15 have the nature to produce ([]) such things and to start such a process16 . . . in accordance with the manner ‘from the same distance’. . . (Lacuna of a few letters)

Agency and atomism

77

(Fr. 11–Fr. 12) [such products [i.e. occurrent mental states] come to be]17 in the way described and are able to create the same [results]. But many,18 though by their nature they are able to become creative of these and those [results], because of themselves (’) do not become so (not because of the same cause of the atoms and of themselves ( )). And with these we especially do battle, and rebuke them, . . . behaving in accordance with the original nature which is chaotic, just as in the case of every living being. For the nature of their atoms has not contributed anything to some of their actions or to the extent of actions or dispositions, but the products themselves () [i.e. the occurrent mental states themselves] possess all or most of the responsibility for certain things. It is a result of that nature [i.e. the disordered congenital nature they have] that some of their atoms move with disordered motions but it is not, however, entirely on these atoms () [because of the things necessarily] entering [from the] environment [into] the natural . . . that all the causal responsibility should be placed, [and on the atoms of the original constitution and of the compound, but] out of what is produced himself ([]   []) [i.e. the occurrent mental state itself] . . . (Lacuna of roughly 30 letters) (Fr. 13) . . . battling and admonishing at the same time with many people, which is contrary to the necessary cause of the same mode.19 Thus, whenever something is produced ( [])20 [i.e. an occurrent mental state] that takes on some otherness from the atoms according to some differentiating mode, not in the way as from another distance, it gets the cause out of itself ( []); then it immediately gives it on to the first natures and somehow makes the whole of it one. Hence, accordingly, those who cannot distinguish these very things cor rectly confuse themselves about what is said of the causes. And in these matters we battle and rebuke some more, some less. . .(Fr. 14) (Lacuna). . . .not because of the cause out of themselves ( ), but because of the cause of the best actions and thoughts . . . then we, on our part, do not battle them, but (they fight) themselves, implying in the atoms themselves the cause.21 And the cause out of themselves ( []) even though they hate or at least do not admonish all that . . . unless they still have what is in common with the original and . . . for a short while out of the same [causes (?)]. . .

Epicurus’ causal manner of explanation is clearly at work here and shows how he distinguishes between different causal factors, namely the cause of the atoms (), the cause of one’s agency or one’s self (  /   ) and the cause of the , that is, the causally efficacious, occurrent mental states of the mind in particular, as I will argue; we also find some traces of the environment being a causal factor in Fr. 12, as well as in Fr. 17.22 In these fragments, Epicurus is explicating why he finds it unacceptable to think, as some do, that all the causal work is to be placed on the atoms. Who

78

Agency and atomism

these theorists were is debated, but on the strength of this evidence, it seems rather likely that they were some earlier atomists. His criticism seems to make sense only if his opponents attributed the responsibility of their actions to their atomic constituents – or, at least, if their views had such implications – which suggests that they must have been atomists themselves. In harmony with some later fragments, the object of Epicurus’ criticism is the mistaken attitude towards one’s conception of responsibility, which is strengthened by Fr. 13: these atomists confuse themselves concerning the explanation of the causes, meaning most likely that they did not recognize the Epicurean division of causality or, as it appears from some later fragments, Epicurus thought they failed to realize what distinctions should be made to save the rationality of the atomist theory and find it compatible with the reality of the phainomena.23 What their failure implies is the subject of some later discussion. My primary interest is to see what the causal faculty of the self is according to Epicurus and how his use of terminology fits into a comprehensive theory. It is worth examining Fr. 13, in which we read about how the causal power of the mind comes about and how that causality is executed. Needless to say, the meaning of Epicurus’ description of the relationship between what is produced and the atoms is debated. Let us take the key sentence of the fragment first. Fr. 13    []   [ ]         ’   []   []       [ ] .

(Fr. 13) . . . Thus, whenever something is produced ( []) [i.e. an occurrent mental state] that takes on some otherness from the atoms according to some differentiating mode (), not in the way as from another distance, it gets the cause out of itself; then it immediately gives it on to the first natures and somehow makes the whole of it one.

The following questions are the most significant for the comprehension of this sentence: (a) In what sense is the product distinct from the atoms?24 (b) What are the two apparent ways, that is, the  and the way ‘as from another distance’? (c) If the product has some otherness from the atoms, in what sense is the causal power of oneself, which one acquires from oneself distinct from that of the atoms? (d) How are we to understand the passing of the causal powers to the first natures?

Agency and atomism

79

Let me start with question (b). Jeffrey Purinton has suggested that the way ‘as from another distance’ stands for being at a different place or location, that is, for two things to be different spatially by being in two different places.25 Accordingly, Purinton takes it that if the product and the atoms are not spatially different, then the spatial extensions of the atoms and what is produced occupy the same space. Purinton supports his understanding by appealing to Fr. 10, in which we are told that atoms of the mind thrust together have the capacity to produce such things ‘in accordance with the manner “from the same distance”’. Nonetheless, this point is a non sequitur. Epicurus does not say in Fr. 13 that the atoms and the product are spatially co-extended, only that they do not differ from each other in the way two spatially distinct objects do. That formulation would be true of a Cartesian dualist position as well: my mind is distinct from my body, but not in the way two spatially distinct bodies are different from one another.26 Epicurus’ attempt to describe this peculiar difference of the atoms and the product is already apparent in an earlier fragment of book XXV, 419 fr. 5: ‘  [] [][ ], [] [] [], [] . . .’ ‘For it is neither one thing as the property and what it is a property of must be said to be, nor is it another thing as the thing that comes in from some distance, nor . . .’27 Nonetheless, Epicurus’ earlier description of the status of what is produced28 does not help to solve the puzzle on its own either, and to substantiate my positive interpretation I first need to establish a number of points.29 The atoms and what is produced are clearly different in accordance with the  (questions ((a)/(b)). In O’Keefe’s understanding, the   has the overtone of some sort of mental grasp, since  derives from , which means ‘division’, ‘distinction’, ‘opinion’, ‘judgement’, and that comes, in turn, from the verb , ‘to grasp’. Consequently, all the above meanings of  are taken to suggest something done theoretically, that is, through some mental activity, which is rather obvious in the case of the latter two. Thus, the distinction Epicurus is making between the atoms and what is produced is first of all a theoretical one according to O’Keefe, whose translation of the relevant part of the sentence is: ‘Whenever something is produced that takes on some otherness from the atoms in a (mentally) distinguishable way, not in the way as when things are at different spatial locations’.30 Nonetheless, the suffix of  clearly indicates an active meaning of the word as opposed to the way it is understood by O’Keefe who, I think, mistakenly takes  in the passive sense. As opposed to that, the active understanding of the phrase as ‘differentiating mode’ nicely coordinates with the statement of the main clause, according to which one gets the cause out of oneself. If the product takes on some otherness from the atoms according to some differentiating mode, it seems to mean that we acquire a distinct causal power, which is able to produce a difference with regard to something else. Accordingly, Epicurus proposes a much stronger claim than just a theoretical distinction between the mind atoms and their product. Epicurus’ claim is not

80

Agency and atomism

that what is produced is merely mentally distinguishable from the atoms but rather that what is produced acquires some otherness in respect to the atoms in accordance with some differentiating mode; that is to say, it acquires an efficacious capacity.31 Consequently, the straightforward answer to the question – ‘In what sense is the product distinct from the atoms, in what does this “otherness” from the atoms which we can distinguish consist?’ – is provided by the surviving piece of the text: the product takes on some otherness from its constituent atoms in the causal power it provides to oneself whenever what is produced comes about according to some differentiating mode. It is certainly a very emphatic statement concerning causal power, in the sense that Epicurus was not willing to conceive of it as a mere epiphenomenon. But what does this distinctive power consist of (question (c)) or, in other words, in what sense is this power different from that of the atoms? What makes this power distinctive? And how should we understand its transmission to the first natures () (question (d))? O’Keefe has placed significant – though not decisive – emphasis on the apprehension of the expression ‘first natures’ for the understanding of the sentence in Fr. 13. Let us answer this last question first and come to the others afterwards. O’Keefe argues that the ‘first natures’ in question are not the atoms but that ‘first natures’ is a term interchangeable with the term ‘first constitution’, both referring to the original, congenital dispositions of our souls, and, accordingly, the sentence amounts to no more than saying that the cause one gets out of oneself – which O’Keefe identifies with our reason – can overcome our original dispositions.32 His motivation for thinking so is that there is a constant contrast made between our ‘original/first constitution’ ( / ) and their later developments throughout the passage. O’Keefe supports his claim with Laursen’s observation according to which     seems to be identical to     and given the apparent interchangeability of these terms O’Keefe thinks that ‘the “first natures” are really the same as the “first constitutions”, just as the original constitution and the original nature are interchangeable’.33 O’Keefe’s conclusion would be granted only if    were a synonym for     as he seems to assume. Nonetheless, the remaining fragments reveal a more complex picture: first of all, it is difficult to make sense of  as a synonym for on the evidence of Text M, where we read the following: This memory of that, or the movement analogous to memory, was again in one aspect co-generated immediately, and under another it had grown, being the beginning and the cause for, in the first case, the first constitution () of both the atoms and what is produced ([]) [i.e. the occurrent mental state], in the other case, for the on-growing [constitution] ( []), by means of which we perform all our actions, of the atoms and the product itself ([]) [i.e. the occurrent mental state itself].34

Agency and atomism

81

This piece of text clearly distinguishes between two different types of constitutions, which are brought about by some memories. A memory is either co-generated immediately and in this case is constitutive of both the ‘first constitution’ of the atoms and ‘what is produced’, or our memories constitute the on-growing constitution of the atoms and the product as well.35 Most importantly for our present purposes, the memory is the beginning and the cause for these respective constitutions, that is,  can hardly be understood as analogous to or as a synonym of ‘the original constitution’ (). Furthermore, both applications of  in Text M clearly represent the atoms as well as the mental states that the atoms constitute; that is, they stand for the atomic arrangements constitutive of some occurrent mental states. So if the expression ‘first natures’ is interchangeable with anything in Text M, it is with the atoms themselves. Consequently, it seems safest to maintain some distinction between   and . If the term ‘first natures’ is neither interchangeable with the ‘first constitution’ nor is it a synonym for the ‘original/congenital constitution’, then the best proposal remains that the ‘first natures’ are indeed the atoms. Furthermore,  ,  , and   are always used in the singular, while Epicurus is talking about the ‘first natures’ in the plural in Fr. 13, just as in the other two fragments where the plural is used when the expression appears, a fact which lends additional support for identifying the term with the atoms.36 O’Keefe considers that possibility: However, even if the ‘first natures’ here are atoms, the phrase does not imply that the products are radically emergent as Sedley claims. Even if, in some sense, the products do move the atoms, they need not move them by some sort of ‘downwards causation’. Epicurus could simply be stressing that the products do have causal efficacy such that, once they have arisen, the product will make a difference in the way that atoms move, as they would have to in order to affect one’s behaviour. If the products are simply certain aspects of the mind (itself a body), products which are acquired when the soul atoms are arranged in a particular way, Epicurus could consistently maintain that these features of the atomic arrangement do make a causal difference in the way the atoms of the body move, while denying that there is any sort of special ‘non-atomic’ causation going on.37 I completely agree with O’Keefe that even if we are to take the ‘first natures’ to mean the atoms, that does not provide us with evidence for reading into ‘the otherness’ in question the kind of radically emergent self, for which Sedley argues on the strength of this fragment. In addition, I think what O’Keefe says here is very close to the correct reading of the text. However, I believe that if one is prepared to privilege what is the historically plausible interpretation of the fragment at this point instead of making the text

82

Agency and atomism

compatible with a tenable standpoint in the modern philosophy of mind, then it will be hard to deny that Epicurus is talking about some sort of downward causation in this fragment.38 To summarize my findings of the analysis of Fr. 13 so far: whenever something in the mind is produced that takes on some otherness from the atoms, it is not a kind of product which is entirely reducible to the mind atoms, since the otherness it acquires is not just the same thing, a group of atoms, described on a theoretically different level, but what is produced generates an independent causal power in accordance with some differentiating mode, which clearly indicates that the product is not a mere epiphenomenon of the atoms. Once one acquires the cause out of oneself, the product (i.e. an occurrent mental state) immediately passes it on to the atoms and makes the whole of it one.39 But what does this distinctive causal power consist of – question (c) – that is to say, in what sense is the causal power generated out of oneself independently? Sedley has suggested that what Epicurus envisaged is the modern notion of emergence: ‘in Epicurus’ view matter in certain complex states can take on non-physical properties which in turn bring genuinely new behavioural laws into operation’,40 and that: the causal influence which emergent properties, like the self and its volitions, exert upon the atoms of which they are properties should be one which can transcend the purely physical laws of atomic motion and bring about motions which these atoms would probably not have followed if left to their own devices. Yet that is what Epicurus appears to hold when he drives a wedge between causation by the self and causation by the atoms, and indeed any weaker interpretation of the doctrine will reduce psychological states to supervenient consequences of atomic motions and return Epicurus to the clutches of mechanism. In these passages, Sedley clearly reveals his motivation for believing in a very novel interpretation of Epicurus, which would allow causal interactions between physical and non-physical entities. He finds only a property-dualist position tenable, since if mental states were merely supervenient on their constitutive atoms, then he thinks Epicurus’ position would inevitably collapse into reductionism and would be purely mechanistic. Nonetheless, is it really property dualism Fr. 13 is telling us about when we read about one’s independent causal power? That is to say, does the mental really need to be non-physical in order to make sense of the genuine causality which the product, an occurrent mental state, acquires out of itself as something independent? I believe if we understand the otherness which the product acquires relative to the atoms and the causal power it obtains as something happening simultaneously, then the distinctness of the product from its constituent atoms is that it acquires its causal power independently of the atoms, as I have suggested above. That the distinction is not merely theoretical is apparent from the fact that once the product acquires its independent causal power, it

Agency and atomism

83

immediately transmits that to the atoms – an activity which would not make sense if the distinction were simply theoretical. If the otherness of what is produced from its constituent atoms were a causal power for the self distinct from that of the (underlying) atoms, but at the same time if this otherness were exhaustively reducible to its constituent atoms, then we would end up with an apparent clash in the theory and Sedley’s property-dualist interpretation would helpfully offer a last solution.Yet I do not think that this is necessarily the case. If a product, that is, an occurrent mental state is constituted by mind atoms, that fact does not exclude the possibility that the occurrent mental state in question allows one to possess causal powers that the constituent atoms of one’s mental state do not have. The fact that the product makes space for one’s independent causal power amounts to no more than saying that one possesses one’s own causal power by having a certain occurrent mental state, e.g. a volition, which can have a causal influence on one’s action, and in the light of Fr. 13 that influence is exercised through the atoms. Through the atoms, because the product gives its causal power back immediately to the first natures, i.e. to the atoms, making the whole of the cause, the mental and  [ ] later41 – the physical aspects – as I will argue for this meaning of  somehow one at once. Consequently, whatever sort of influence a particular volition can have on my action, that effect is executed through some particular, atomic motions of my mind. Nonetheless, but most importantly, the particular arrangements of the atoms of my mind at T1 (at a given point in time) constituting an occurrent mental state would not determine this or that arrangement at T2 on their own without the influence of the occurrent mental state, that is, the mind and its ‘actions’ are not just epiphenomenal operations corresponding one-to-one with the underlying atomic motions.42 Our mind acquires its causal power out of itself, which is the secondary property of the mind atoms constituting particular, causally efficacious products or occurrent mental states, which in turn, give their causal power back to their underlying atoms; these products having a causal influence over the atomic motions of minds. In other words, the secondary property of some mind atoms, an occurrent mental state becomes causally efficacious and determines the further arrangements of its constitutive atoms. The mind atoms, constituting an occurrent mental state at T1, constitute a state which has a causal capacity to determine further atomic motions and, hence, the arrangements of the atoms at T2. Such an interpretation, nevertheless, still leaves the question open that if this secondary property is not a non-physical entity as I think, then, could Epicurus really have avoided the charge of reductionism, as he certainly seems to have desired? Overall, if we say that this capacity must be physical, in the end it must still be reducible to its constituent atoms, the independent character of the causal power seems to collapse, and my interpretation seems to amount to no more than O’Keefe’s compatibilist understanding of the Epicurean philosophy of mind. In other words, how could Epicurus conceive of this causal power as non-reductive and answer why we are not to say instead that in such a causal

84

Agency and atomism

matrix the particular arrangements of the mind atoms at T2 are the mere outcome of their arrangements at T1? How could Epicurus have reasonably believed that the product can generate an independent causal power on the strength of the theory he put forward? I think independent causal efficacy is to be understood as something not exclusively produced by the atoms.43 As I have suggested, the clause stating that the product generates this causal power out of itself and not directly from its constituent atoms emphasizes that this capacity is of one’s own.That is, although the product itself is brought about by its constituent atoms, the causal efficacy of one’s mental state qua some occurrent mental state differs from the causal efficacy of the atoms and, therefore, it cannot be reduced to its constituent atoms, otherwise this capacity would not need to be brought back to the atoms.44 Consequently, if some causally efficacious mental events are not exhaustively produced by atomic events, then not every effect throughout nature has a sufficient atomic cause. Does this commit Epicurus to entertain the non-physicality of, at least, some mental states? I do not think it does. If Epicurus entertained the view of what we would call today token monism in the philosophy of mind, according to which every mental event is also a physical event, a single event falling under both types, that fact also includes, or at least does not exclude, the possibility that he could have also entertained the position described today as type dualism, according to which mental types are different from physical or, more precisely in Epicurus’ case, from atomic types.45 Token monism on its own cannot guarantee what Epicurus was attempting to argue because one could simply object that if mental events were causes solely in virtue of being physical, and not at all qua mental, then it seems we can only speak of the efficacy of the mental by courtesy.46 However, if mental events are not merely causes in virtue of being the outcomes of the underlying atomic patterns of motion but in some cases causes qua mental as well, then it seems to be possible to entertain a non-reductionist point of view without taking into account the non-physicality of the mental and without ending up with Sedley’s radical emergentist point of view. According to this understanding of Epicurus’ position, all the properties of a living being are based on atomic events; nevertheless, not every behaviour is caused by atomic powers, since in certain cases some mental states are not exhaustively produced by the atomic events of the mind. One may object that, on this view, even if some behaviours are only partially produced by atomic powers, they will eventually be reducible to the atoms and then it does not make sense to talk about the independent causal efficacy of the mental or, if one does not admit that the causal efficacy of the mental is completely dependent on the atoms, then one ought to jump to the conclusion of the non-physicality of the mental in order to maintain a philosophically relevant theory. However, it seems that what Epicurus is saying is that when the product takes on some otherness from the atoms according to some differentiating mode, generating the cause out of itself, although this capacity is realized in something

Agency and atomism

85

atomic, qua something mental, it has the capacity to get the cause out of itself, as a consequence of which it influences the movement of the atoms at T2 in a way which does not necessarily follow from the atomic patterns of motions at T1. On the strength of this understanding of Epicurus, the mental cannot be exhaustively reduced to the underlying atomic movements because certain mental states seem to have their own causal efficacy acquiring this causal efficacy out of themselves qua mental states; and in this sense, although these mental states are partially brought about by the atoms, the atomic motions do not provide a complete causal explanation, nor can they explain the underlying atomic motions of the mental state of the mind at T2. The theory, thus, turns out to be a non-reductionist position because, on the strength of Fr. 10–Fr. 14, it does not merely claim that we cannot give a coherent causal story of human agency based on the causal history of the atomic level, but also that some causally efficacious (occurrent) mental states cannot be accounted for exhaustively by their underlying atoms. That is to say, the underlying atomic level is not causally comprehensive in certain cases, as a consequence of which there is an essential interaction going on between the two levels, including downward causation. The strongest evidence for some mental states not being completely determined by the causal powers of the atoms is Epicurus’ account of the . Note that Epicurus describes the product having its peculiar causal power in general terms, ‘whenever something is produced . . .’, without specifying what exactly is being produced, instead of talking about some causally efficacious desires or volitions. It obviously does not mean that Epicurus conceived the latter as inefficacious, but rather points towards an extra complication concerning how to conceive of these as causally efficacious and what their causal efficacy amounts to.47 I think this indefinite aspect of the causal power is precisely what Epicurus intended to emphasize and pick out by the verb , of which one of its frequent forms is the infamous  . Some of the mind atoms are productive of some (occurrent) mental states with some peculiar causal efficacy48 and Epicurus left some loose ends by not specifying what these products (in my interpretation [occurrent] mental states) exactly are, that is, by not specifying their exact cognitive or other content – at least in the surviving parts of the text. By leaving this question open, Epicurus seems to have set up a rather generic term for a special kind of mental state with its peculiar causal efficacy, product(s) which I interpret as occurrent mental state(s). These mental states themselves are further influenced by one’s character based on one’s available stock of memories (see Text M), beliefs or dispositions, as well as by one’s actual state of mind due to one’s given circumstances (see Fr. 17); that is, by what sort of eidola  are influencing someone at the given moment either due to his present perception or his present thoughts. That is to say, the atomic structures constituting an  are the outcome of many different factors, including such an occurrent mental state ( []), which has its own peculiar causal efficacy and which can account for the causal efficacy of the mental. For example, if I wish to drink a coffee, my desire is a kind of structured result of many additional factors, such

86

Agency and atomism

as feeling sleepy and having a belief that my condition can be changed by drinking a coffee, as well as perhaps also the result of thinking of some relevant images or bringing out some memory of a similar case. In other words, once someone wants something, the volition cannot be described as a mere outcome of the underlying pattern of atomic motions. What a particular volition is, is not only determined by some atomic motions but it is further specified by various mental factors, because for the constituted causal capacity to gain its independence needs to mean that it has to be further defined independently of the atoms.This is the reason why I chose to interpret the  as an occurrent mental state, the occurrence of which is dependent on many factors. Now, the most important things to consider in further detail are, on the one hand, how the text supports such an interpretation and, on the other hand, whether such an explanation leaves open the possibility that the product can generate an independent causal power on the strength of the theory put forward.

A conceptual analysis:   In Fr. 10 Epicurus describes the mind atoms as having such a nature that if thrust together in a certain way they produce and initiate some action and it is in Fr. 13 that Epicurus specifies how we are to understand this unique causal scheme. In Fr. 11–12, as I have already suggested, Epicurus makes an explicit distinction between the cause of the atoms ( ), the cause of one’s own agency () and the cause of the  . Epicurus says in Fr. 11 that although some people could be productive of certain results – most likely of laudable character – it is because of themselves (’) – most likely because of their not so laudable character flowing from their intellectual laziness or incontinence (akrasia) – that some people do not achieve these results, and not because of their atoms ( ).49 Apart from criticizing the implicit or explicit points of some other atomists, Epicurus seems to despise these people for their behaviour or at least for the kind of behaviour that follows from their theory, whose behaviour is similar to that of wild animals. On the strength of this rebuttal by Epicurus it may be suggested that the target of his criticism is those atomists, possibly the circle of Nausiphanes, who explicitly subscribed to the deterministic consequences of their theory and thought of their behaviour as being entirely determined by their constituent atoms.50 As opposed to them, in Fr. 12 Epicurus makes it explicit that according to his view it is not because of their atoms that some people act in certain ways or have certain dispositions, but their  ‘had gained all or most of the responsibility for certain things’. What are these certain things, or, in other words, what are the  causally responsible for? I think that it is quite straightforwardly on the strength of Fr. 11 that the  are responsible for people’s actions and for their dispositions. This aspect of Epicurus’ causal theory helps to restrict the scope of the meaning of

Agency and atomism

87

 and also sheds light on how my interpretation of the term diverges from that of the two major advocates of reductionism and antireductionism. I find Sedley’s translation of  as ‘developments’ a bit misleading, since ‘developments’ sounds like a synonym for ‘dispositions’ or are at least relatively long-term states, and it does not make sense to say that one’s developments are responsible for one’s dispositions. O’Keefe follows Laursen’s translation of the  as ‘products’, which is a literal translation, but without further specification it points misleadingly towards the sense of ‘developments’. Thus, although O’Keefe’s incorrect translation of  in Fr. 11 as ‘these particulars’ has the same referents as my understanding (one’s actions and dispositions), there is a tension in his understanding of one’s product in the sense of a development being responsible for one’s dispositions. I think Epicurus sets out the following causal scheme in Fr. 10–14: (soul) atoms if thrust together in some particular way ‘have the nature to produce () and start such an action’, namely voluntary action, which most likely motivated Epicurus’ discussion in connection with the analysis of one’s conception of one’s self throughout book XXV. The , which are produced by some of the mind atoms thrust together in a certain way, gain all or most of their power of actions or dispositions, that is, when something is produced ( []) according to some special conditions as set out in Fr. 13. The cause it gets out of itself is the power to determine and develop further one’s occurrent mental states leading to actions, and hence, indirectly, to one’s dispositions, the product’s entire or at least major influence for things of this kind. Now even from this short reconstruction many things become apparent concerning the term  . First, it is not to be understood as an acquired and constant psychological disposition because the term signifies first and foremost the causal capacity of the mind, responsible for one’s actions and dispositions, as Epicurus explicitly puts it. It is no accident that Epicurus mentions actions and dispositions together in Fr. 12, as he finds it natural to talk about them together, since, as it will become apparent in Chapter 3, according to his ethical theory one’s disposition is the outcome of one’s actions, just as once one acquires a stable disposition, Epicurus would maintain that one’s actions flow from one’s disposition. Second, the  are mental states, since if they are constituted by some atoms of the mind, they obviously must be some states of the mind in respect of it qua atomic as well as qua mental.The  are the major causal factors for actions and dispositions as we learn in these fragments but, nevertheless, they are not the only causally efficacious states of the mind. As we read, e.g. in Text M, memories are the beginning and the cause for some , but that does not mean that memories on their own can initiate actions. Rather, according to the causal scheme I have just set out, we should say that these memories have the causal power to bring about the , these occurrent mental states, and initiate actions with them. And this is what

88

Agency and atomism

I suggested earlier51 concerning  as a general term, having no determined mental content of its own. It seems that when an  is brought about, its mental content depends on what brings the  about in a given instance and on further additional factors that are also taken into account, such as one’s character, the given circumstances, etc; thus, the mental state is distinctive in this further, case sensitive, specified sense, and hence my choice to interpret the term as occurrent mental state/s. I think this conceptual background is detectable in Text M, and therefore, let me quote it again in its entirety.52 . . . the memory or the affection analogous to the memory of the more necessary things came to be/exist within? () in reference to the well-defined and that is used to test all things and not in accordance with things that cannot be defined but need judgment. This memory of that, or the movement analogous to memory, was again in one aspect co-generated immediately, and under another it had grown, being the beginning and the cause for, in the first case, the first constitution () of both the atoms and what is produced ( []) [i.e. the occurrent mental state], in the other case, for the on-growing [constitution] (  []), by means of which we perform all our actions, of the atoms and the product itself (  []) [i.e. the occurrent mental state itself] that in some cases is necessarily opposed to what produced . . . As we read, the memory in question ‘was again’ either co-generated or grown, in both cases the point being that after the causal force of the memory has been exercised, the memory in question is still present. When the memory is recalled, in the first case it causes some first constitution () of both the atoms and the occurrent mental state ( []), the memory being co-generated immediately. As I understand it, the atoms in question must be the atoms constitutive of the first instalment of an occurrent mental state and the memory co-generated with them, the particular memory which itself was the cause for the very existence of this first constitution of an occurrent mental state, which, in turn, can be further processed. From the other aspect, the memory in question is the cause of the on-growing constitution ( []).The standard understanding of this growing constitution is ‘one’s developing psychological dispositions’ which seems to gain further support by Epicurus’ statement that with these we perform all our actions, making perfect sense if we are to entertain a whole-person model of causal agency, as I certainly think Epicurus did.53 If we accept that the   is to be understood as one’s developing psychological disposition in the sense of this disposition being part of one’s character, then my interpretation seems to collapse, since if we take memories to also be part of how our psychological dispositions grow, or, as Epicurus quite literally put it, some memories by growing are the cause of the on-growing constitution of the

Agency and atomism

89

atoms and of the (occurrent) mental state itself ( [][]), then one may object that on the strength of this piece of evidence, it still seems safer to understand the  as mental developments than as some general term for mental states having some peculiar causal power, as my interpretation requires. Nonetheless, the interpretative strategy I offer is also left open: according to Laursen,  is a further processed ‘first constitution’.54 If we apply this to Text M, then we can establish the following understanding: when the memory is immediately co-generated with the first constitution of an occurrent mental state, there is further processing of this mental state, producing the growing constitution. It is this further processed mental state, the on-growing constitution – which in the sense of the literal translation of , I believe, ‘grows onto’ the first constitution of the atoms and the (occurrent) mental state – by which we perform most of our actions. For example, let us imagine a situation in which I am sitting under the sun and I remember the fact that staying under the sun for too long made me dizzy last time in a similar situation – a painful memory – and, therefore, my memory causes me to develop a desire, first of all, to change my current position. But simultaneously, I am also affected by many other things, e.g. seeing a shadow over there and, consequently, it is my memory and other factors which constitute my on-growing, causally efficacious desire to move under the shadow, which eventually result in an  to move under some shade. The course of this processing neatly fits my understanding of the   according to which this causal capacity for action is the result of further specification by one’s beliefs, memories and so forth each and every time and not a stable disposition of the mind atoms. The first constitution of an  comes to be when some mind atoms are thrust together in a certain way, and it can process further to an , by which we can perform our actions. The different grammatical tenses of  work very well with this under standing, since we can take the passive aorist participle of , and the passive present participle , in support of the fact that what is brought about can still be further processed by some extra factors in order to be more finalized or determined, that finalization being exactly picked out by the aspect of the passive perfect participle of .55 Namely: when an occurrent mental state is produced ( []) in the specific way of Fr. 13 and, simultaneously, it is responsible for one’s independent causality, this genuine causality that one has can develop further the produced mental state in harmony with one’s set of beliefs, and so forth into such products (), which are more complex rational mental states, desires or volitions, which are then, in turn, causally responsible for one’s actions or dispositions. Let me now turn to the textual evidence for taking the  as some occurrent mental states, being some peculiar causal powers involving many factors, the power to perform some actions, on the strength of the next surviving fragment of book XXV.

90

Agency and atomism Fr. 15 . . .     ’     []     []                                    [] []            []     [] []  []  [ ]  . . .[][] . . . (Fr. 15)56 . . . a certain nature predicated, and not referred to as ‘it’ or even ‘he’. And even if the first constitution of what is produced (  )57 [i.e. the occurrent mental state] exerts some compulsion in the mind, this sort of thing [is] not being produced () [i.e. the occurrent mental state] by necessity all the way to certain specific things, but, on the one hand, as far as a soul or rather a soul with a disposition and movement of this particular size comes about, a thing of such a kind [is] being produced ( ) [i.e. the occurrent mental state] from things of this kind [from the atoms?] by necessity, and on the other hand, as far as a soul of this or that kind comes about, it is not by necessity that this sort of thing [is] being produced () [i.e. this sort of occurrent mental state] or at least as one proceeds in age it is not by necessity that one has this sort of thing produced ([]) [i.e. this sort of occurrent mental state], but as a result of oneself and as a result of the cause out of oneself one [is] able to exert some power . . . [producing] something else . . .

In this piece of evidence, Epicurus discusses the different possible realizations of the first constitution of an occurrent mental state. The beginning of the sentence states that just because the first constitution of an occurrent mental state exerts some compulsive power in the mind, one’s first constitution of an occurrent mental state does not need to be determined. On the one hand, the sort of atoms which constitute a soul or a soul with some kind of disposition are determined. However, that fact does not determine the sort of disposition one is to have. As Epicurus says, if a soul has a certain disposition and movement of a particular size, from the soul’s disposition and movements a kind of occurrent mental state inevitably follows. That is to say, if I have the disposition of an angry lion, in certain similar circumstances I will have some similar, occurrent mental states as a reaction to certain things even upon encountering them for the first time, which correspond to my disposition; that is, I will initiate actions based on my disposition. However, it is not necessary that everyone reacts to these particular things even in the very same circumstances in the same way, since depending on the kind of soul one may have, there are different reactions initiated in response to the very same things. Even so, it is not impossible to

Agency and atomism

91

alter one’s character during one’s life, as a consequence of which our attitude to things can be changed as well. If we modify our dispositions over time, the occurring mental states flowing from our dispositions will change as well. Accordingly, Epicurus’ main interest is to show that the causal power one exerts to initiate actions differs in accordance with what sort of disposition one has. That the disposition understood as some stable psychological development of the mind is not to be identified with the  is shown by the fact that the  flows from one’s disposition according to this fragment.58 Furthermore, at the end of the fragment Epicurus talks about the possibility of changing one’s character by exerting some compulsion as a result of oneself or by means of some self-initiated cause, and the exertion of this compulsion to change one’s disposition/character is expressed by the same verb ‘’, as we find at the beginning of the fragment where Epicurus is talking about the first constitution of the occurrent mental state exerting some power in the mind. It is hard to resist taking the dynamis or the cause out of oneself as describing the , which, I believe, lends further support to my understanding of the  as a causal faculty of the mind, generated in the way described in Fr. 13. Accordingly, although we have certain dispositions, the causal faculty of the mind, the , can help us to overcome our set dispositions. That we are to understand the , the processed form of an  – or that of an [] (Fr. 13)[] (Text M) – as a complex combination of various internal and external factors is brought out quite explicitly very soon after the fragment above: Fr. 17 [  ]  [  ]    [],  ’    ’  [ ] []  []  []   []  []    ’  [’]         []’    []  ’  []      []   [].    [] . . .

(Fr. 17)59 From the first beginning [there are]60 seeds directing, some to these, others to those, others to both – in every case seeds, which may be many or few, of actions, thoughts and dispositions. Consequently, at some time it is precisely because of us that the product [i.e. the occurrent mental state] becomes such or such, and it is because of us or rather because of the beliefs of ours, which are from ourselves that the things which of necessity flow in through our passages from that which surrounds us at some point of time become (such or such). And if against nature . . .61

This piece of text has been translated in various ways. I offer a combination of several translations with minor interpretative additions.62 The main point

92

Agency and atomism

commentators bring out concerning this fragment is that our mental development is dependent on us and not on our constitutive atoms. Bobzien has used this fragment in her argument for what she called the one-sided causative understanding of ’ , according to which, if  is taken in the same sense of the one-sided causative in the second sentence all through – which gains credit from the syntactical coordination of the two phrases in which  is used, where in the second case of beliefs it certainly would not make sense to talk about a two-sided potestative reading of , since we cannot conceive of our beliefs as decision-making faculties – then Epicurus did not talk about whether we decide to develop one way or the other but, instead, he was concerned with us being the determining factor for developing one way and not the other.63 There is more to be said about this text incorporating my understanding of the  as a complex combination of various external and internal factors. Epicurus is most emphatic that it depends on us what sort of mental state we bring about, and he further characterizes this dependency in the case of external influence. On the strength of this fragment, it appears that while external things are flowing through our pores necessarily, the sort of external information we register depends on us and on our beliefs in the sense that the external information we receive is the result of our selection, filtered through our beliefs. The important question is how we should make sense of this relatively simple claim in context. Since it depends on us what sort of  we bring about, that is, it is the consequent outcome of the exertion of our causal efficacy, we are not influenced by external information or the environment automatically, since what we pay attention to depends on us and on our beliefs. Furthermore, our causality concerning our  and the reception of external information seems to be a transitive relation according to which how I apply my causal faculty of the  is essentially connected with what sort of external information I pay attention to, which also depends on my available stock of beliefs. In other words, this piece of evidence clearly points towards my interpretation of the  as the processed causal faculty of the mental resulting from the combination of various factors. Now, how does this support my claim that Epicurus was committed neither to the radical emergent property view of David Sedley nor to the reductionist reading of Tim O’Keefe?

The  of the soul and the hypothesis of the swerve The above analysis of the central fragments has provided evidence for the fact that Epicurus took the mind to be causally efficacious and that he conceived of this causal faculty as something not exhaustively determined by the atoms. In Fr. 11 Epicurus says: ‘But many, though by their nature they are able to become creative of these and those , because of themselves (’) do not become so (not because of the same cause of the atoms and of

Agency and atomism

93

themselves (   )’. A little later on in Fr. 12 Epicurus continues by saying that: the nature of their atoms has not contributed anything to some of their actions or to the extent of actions or dispositions, but the products themselves ( ) [i.e. their occurrent mental states] possess all or most of the responsibility for these things. These fragments clearly indicate, I suggest, that Epicurus denied that the efficacy of the mental could be entirely explained or determined by atomic motions, which means that Epicurus did not hold that all mental events have complete atomic causes. Accordingly: (I) he attributed a sort of causality to the mental not determined by its underlying atomic motions, and consequently; (II) he held a sort of physicalism in which the causal efficacy of the mental cannot be reduced exhaustively to the atoms. How could Epicurus have granted that some mental states are not merely causally efficacious in virtue of being atomic, but qua mental as well? I think the assumption of the causal efficacy of the human mind on the strength of the phenomena must have been the first step of Epicurus’ argumentative strategy, in harmony with his scientific method in general. The central fragments provide us with the evidence that the kind of causality Epicurus attributes to the self cannot be completely explained on the strength of atomic motions, which lends support to the claim that he attributed independent causality to mental states, independent in the sense that the causal efficacy of the mental, though being physical (token monism), is not exhaustively determined by the atoms (type dualism). But how could Epicurus substantiate his claim that there are mental states for which the atoms are not causally comprehensive? I think, through the introduction of the atomic swerve. Our sources testify that Epicurus conceived of the swerve as the necessary condition for the causal efficacy of the self 64 and I believe that he thought he had secured this causality by refuting causal determinism with the swerve.65 His consideration of the function of the swerve as a necessary condition for the causal efficacy of the mental may be formulated as an objection to the following argument: (1) (2) (3) (4)

Every mental state is realized in a physical state. There is no systematic over-determination. Every physical state has a sufficient atomic cause. Therefore – on the strength of (2) and (3), the physical excludes the mental from the circle of causal efficacy.

If (3) holds water, then the causal efficacy of the mental is redundant, therefore, it is over-determined, which it cannot be accepted on the strength of (2). However, if Epicurus denied (3) by his conception of the atomic swerve, he

94

Agency and atomism

avoided the problem of over-determination, a consequence of causal determinism he wished to refute. In principle atomic swerves are random, having no cause and, consequently, they happen at no fixed place (nec regione loci certa) – anywhere in the infinite universe – and at no fixed time (nec tempore certo, DRN II 293). Since atomic swerves also occur in one’s mind, i.e. in one’s animus (or mens), the theoretical possibility of a coherent causal explanation of the history of one’s animus (or mens) on the atomic level is not possible since there can be a gap in the causal story on the atomic level. Theoretically, I could give a coherent description of this level pointing out all the causal connections between the microscopic and macroscopic levels of the mental from the first moment of a person’s life, but once an atom of one’s mind swerves, it will no longer be possible to do that. Even if in retrospect I manage to integrate the atomic swerve somehow into my explanation of why certain mental phenomena occurred, there is no guarantee that in each and every case of a similar mental phenomenon an atomic swerve would occur again. That is to say, once an atomic swerve occurs in one’s mind/animus, there is a gap not only in the causal history between the atoms of one’s mind/animus (or mens) on the atomic level of description, but also between the microscopic and the macroscopic levels. The causal story of the mental can no longer be exhaustively accounted for by atomic motions or consistently reduced to them because not every mental event has a satisfactory atomic cause to the extent that such and such an atomic pattern in the mind will necessarily bring about such and such a mental disposition. Given the random atomic swerves, one’s mental life cannot be accounted for sufficiently by the corresponding events at the microscopic level. On the strength of this interpenetration, the thesis of multiple realizability (MR) also follows necessarily:66 if Epicurus wished to maintain that one’s mental state M does not change as a result of an atomic swerve S, he must have admitted that the same mental state can be realized in different patterns of atomic motions, otherwise undetermined swerves would result in the change of mental states.67 For example, if A is the given pattern of atomic motion for mental state M, then if a swerve S occurs while one is in mental state M, which is realized in the pattern of atomic motion A, then we will have on the atomic level A+S. If Epicurus wanted to hold that swerve S does not bring about some change in mental state M, then he needed to accept that mental state M can be realized in both underlying patterns of atomic motion, A and A+S. The thesis of MR must be a general claim concerning all the things in the world, since swerves, as a matter of principle, occur at no determined place or time. One may raise the worry, nonetheless, that if the atomic swerve undercuts the sufficiency of the atomic explanation so radically, then the atomic theory seems to lose its explanatory power.To phrase the worry differently: if the basic constituents of everything in our world and in the whole universe are the atoms and the void, and the atoms and the void constitutive of our cosmos are meant to explain the phainomena by their different types combined in certain ways in

Agency and atomism

95

accordance with their patterns of motion, it seems that the atomic swerve may make any atomic explanation senseless, since these patterns of atomic motion can be changed randomly by atomic swerves, the explanatory power of the theory may be in danger. According to our evidence Epicurus postulated that when an atom swerves it does so only by one minimum of space. Cicero questioned why not by two or three minima and I think the reason is exactly to do away with the above worry.68 An atom swerving by one minimum of space is not thought to be able to make such a big difference in the overall structure of a thing, whether inanimate or animate, as to cause the destruction of a certain characteristic of that thing. If swerves could cause such major changes, then we should see things suddenly changing with no apparent reason, e.g. a rock suddenly becoming soft and fluid. If swerves had such radical effects, then, indeed, the atomic theory would lose all its explanatory power. Nonetheless, we can give Epicurus a plausible reply: he may have wished to restrict the scope of the change a swerve can bring about by saying that atoms swerve by one minimum of space. If we recall Epicurus’ theory of compound bodies, the types of atoms in a compound – no perceptible body is made up of a single kind of atom (DRN II 581–97) – must have suitable shape and size in order to form a given compound. It seems clear on the strength of Lucretius’ account that when these different types of atoms form a compound, their different sizes and shapes determine the underlying pattern of atomic motion of the given compound, the rebounds of the atoms falling into a more or less stable internal equilibrium.That is to say, the different characteristics of compound bodies are the outcome of the patterns of atomic motions that are determined by the different shapes and proportions of the constitutive atoms.69 Now if this is correct, it seems that a random atomic swerve by one minimum of space cannot indeed result in the dissolution of a compound body or in the change of its qualities because Epicurus could say that the effects of swerves are overcome by the characteristic shapes and sizes of the atoms, which are the determining factors of the underlying pattern of atomic motions within a compound body. But let us now focus on Epicurus’ theory of the mind. Given that by the introduction of the swerve Epicurus granted that the causal efficacy of the mental could not be explained satisfactorily by the causal powers of the atoms, what evidence do we have for thesis (II) that his non-reductive theory is also a physicalist theory? The evidence lies, I suggest, in his conception of the soul. In the Letter to Herodotus 63, Epicurus presents a conception of the soul according to which the soul is a , being a , a blend.70 The soul, as a part of the body, is liable to co-affection with the rest of the aggregate, and although its capacity for sensation is a  of both the soul and the body, the body does not acquire a share in all the capacities of the soul, whose capacities yet again the soul could not possess without being in the container of the body. It is clear that soul and body are distinguished from each other as if they were

96

Agency and atomism

two substances, yet they are not metaphysically different entities, since they are both corporeal, which is the necessary condition for their interaction. Lucretius’ evidence in DRN III augments Epicurus’ theory, suggesting that the soul is made out of the atoms of four different elements – heat, air, breath, plus the unnamed element – consisting of the smallest, smoothest and the most mobile atoms. As Lucretius says: The primary particles of the elements [i.e. of these four elements] so interpenetrate each other in their motions that no one element can be distinguished and no capacity spatially separated, but they exist as multiple powers of a single body. (DRN III 262–65)71 It is evident that Lucretius is describing the  or  of the atoms of the four elements; that is to say, he represents an Epicurean view of the soul according to which it is a mixture made out of the constituent atoms of former elements, recombined in such a way that no elemental power is retained separately within the mixture, but instead the atoms of the elements form a new substance: the soul.72 Even if the initial nature of one’s soul can be characterized by the dominance of the atoms of certain elements, e.g. in the soul of an angry person the atoms of heat are dominant, and even if one’s nature cannot be completely eradicated, the atomic constituents of the soul have to interact in their mixture in ‘such a way that a unity be seen to be made of all, or else heat and wind apart and the power of air apart would destroy and dissipate the sensation by being separated’ (DRN III 285–7).73 It is not only impossible to distinguish any of the elements in the mixture of the atoms constitutive of the soul, but we cannot even pick them out by spatially separating their powers, but ‘they exist as multiple powers of a single body’. Now let us step back for a moment and recall the three different types of mixture recognized by the Stoics, for which our main source is Alexander of Aphrodisias (De Mixt. 3. 216): (A) juxtaposition () – mixture of beans and grains of wheat example, which do not blend – Democritus’ position; (B) fusion () – e.g. cooking, when eggs, flour, etc. form a new stuff with new properties; (C) total blending () – when different things are blended through and through in such a way that every bit of the blended mixture consists of all of the blended things.Yet all the original things retain their properties and in principle they are recoverable from the blending. What is the type of mixture Lucretius describes? It is certainly not (A), since the mere juxtaposition of elemental powers would not produce the peculiar capacities of the soul. (B) does not seem to be a viable option for an atomist, because

Agency and atomism

97

atoms cannot be cut and they are indestructible. Yet at the same time (C) is impossible since two atoms cannot occupy the same place either. It appears in comparison to the Stoics’ classification of mixture that Lucretius’ description is closest to (B), or somewhere between (A) and (B). How is this possible? Since Lucretius characterizes the soul as a mixture of the atoms of the four elements. That is to say, the fusion of those elements or their particular powers is a mixture of their constituent atoms. It is obviously not the atoms themselves, which get fused, but the powers of the elements the atoms constitute. Lucretius is talking about a kind of total fusion in which the fused elements do not retain their capacities separately but their atoms form a unity with new qualities, the causally efficacious soul.74 On the strength of this united conception of the soul, the theoretical explanation of the capacities of the soul as a result of the causal powers of its underlying atoms becomes impossible. The causal efficacy of the mental cannot be described on the atomic level because if we are to maintain this mixed conception of the soul, we cannot explain its capacities by the constituents of individual powers, which have been fused, but only by the unity of those formerly separate powers and consequently we cannot make a correspondence between the atomic and the mental levels. Although the capacities of the soul are realized in something atomic, they are not wholly determined by the individual powers of their constituent atoms, but the capacities of the soul are the of this very peculiar  within the container of the body. Thus, the  theory supports my thesis according to which (II) Epicurus held a physicalist theory in which the causal efficacy of the mental cannot be reduced exhaustively to the atoms, since given premises (1)–(3) of the next argument, (4) and (5) should follow: (1) The capacities of the soul are realized in a mixture of atoms, in which the atomic powers are fused. (2) Atomic powers are the outcome of definite patterns of atomic motions determined by the interaction of some types of atoms. (3) If the atoms of certain elemental powers are fused in a mixture in such a way that a unity is to emerge with its own, new capacities, namely the soul, then there can be no exhaustive correspondence made between the capacities of the soul and the atomic powers fused in the mixture. (4) On the strength of (2) and (3), the capacities of the soul cannot be allocated to definite atomic patterns of motions. (5) Therefore, the capacities of the soul, though they are realized in something atomic, are non-reductive. The explanatory gap provided by the  theory does not need to entail, however, a causal gap.The fact that the mixture of soul atoms results in the soul possessing emergent properties not possessed by any one constituent atom does not in itself preclude the said emergent properties from being wholly determined (causally) by the said constituent atoms (or their powers). Nor does the

98

Agency and atomism

fact that the soul’s emergent properties cannot be explained solely with reference to the causal powers of its atoms entail that the soul’s emergent properties are not wholly determined by the causal powers of its atoms. To undercut the possibility of a strict reductionist explanation of the mind, next to the  theory of the soul Epicurus needed to introduce the swerve. By deploying the swerve against the closure of the atomic, Epicurus necessarily injected the thesis of MR into his theory, as argued earlier. Consequently, he entertained a nonreductive position, since if the thesis of MR is true, then reductionism is false.75 One’s mental life is never completely determined by the atoms because the same mental states can be realized in different, even if not significantly different patterns of atomic motions, and therefore Epicurus’ token monism of the mind is open to a type dualism in the causal realm. Since the strict reductionist thesis is undermined by the swerve, granting the thesis of MR, there is a gap left for the independent causal efficacy of the mental. Thus, the only possibility for a complete causal explanation of mental events for Epicurus was to introduce causal interactions between the mental and the atomic levels without having to be committed to the non-physicality of some mental processes. During such thought processes, the atoms are thrust together in such way that their further motions are determined by the downward causation of one’s mind – I believe that the causality given back to the atoms in Fr. 13 clearly implies downward causation – considering that in such atomic states there is some non-determinedness involved, a gap filled out by the physically constituted mind, but not exclusively by its physical constituents. Once, by the postulation of the atomic swerve, the theoretical possibility of giving a complete atomic explanation of the mental is excluded, the ‘qua mental’ talk – a necessary concomitant of type dualism – gains credit. It is compelling because in the absence of a complete atomic explanation of the causal capacities of mental processes, e.g. those of volitions, the starting assumption of the independent causal efficacy of the mental becomes justified. If the atoms cannot completely define the causal capacity of the mental, it makes sense to talk about the further constitutive factor of that capacity, independently from the atoms.That is to say, the , having no complete atomic cause, will not only be further specified on each and every occasion when a volition occurs because of some external or atomic factors, but also because of some internal influences – dispositions, beliefs, desires, memories, etc. – qua mental if we are to have a complete causal explanation. These dispositions, beliefs, desires, memories, etc., have an influence on the mind and can bring about causally efficacious mental states, qua something mental, in virtue of their mental content, and not in virtue of them being atomic. Undoubtedly, they are also something atomic but one could not give an exhaustive account for their causal efficacy by describing the motions of their constituent atoms by some bridge laws,76 such as this one: (Reductionism) For each mental predicate F there is an atomic predicate G such that a statement of the form ‘x is F iff x is G’ expresses a bridge law.77

Agency and atomism

99

Conclusion If we connect the central fragments of book XXV to my arguments, it becomes clear that Epicurus did not need to invoke the non-physicality of the mental in order to secure the independent causal efficacy of one’s mental life. As I have argued, the causal efficacy of the mental needs to be exercised through the atoms because there is a correlation between the two levels; the mental never becomes non-physical, even if it is not completely determined by the atoms. How the independent causal power of the mental is exercised through the atoms is described in Fr. 13: what is produced, in my understanding of the first constitution of an occurrent mental state, immediately gives its causal power back to the atoms ‘and somehow makes the whole of it one (   [ ])’. The natural question to ask concerning this clause is the referent of . Looking at the whole fragment it seems fairly clear that the best candidate is , one’s self-initiated cause. If that is correct, Epicurus is saying that once one’s self-initiated cause is brought back to the atoms, the whole of the cause is made somehow into one. I think there are two ways we may understand this statement: (1) We either need to read the clause slightly differently and translate  as ‘each’, according to which the occurrent mental state makes each cause , that is, one – the subject of  clearly being what is produced, i.e. the occurrent mental state. This solution would mean that although  refers back to the cause the product gets out of itself, in this clause we are to take  in a general sense, meaning that the occurrent mental state makes the mental and the physical aspects of the cause somehow one, the fact of which would lend further support to the independent causal efficacy of the mental. (2) Alternatively, we can keep my original translation of  and take the ‘whole of the cause’ pointing towards my interpretation of the  as a generic term for an occurrent mental state with its peculiar causal efficacy further specified by one’s character based on one’s available stock of memories, beliefs, etc., as well as by one’s actual state of mind. This translation points towards my understanding of the  because the whole of the cause is the sum of the different factors bringing about a causally efficacious mental state. Accordingly, when one’s self-initiated cause is given back to the atoms, the mental aspects are made into one with the atomic aspects of all the further different factors, which are all parts of volition. This should clearly be the case of downward causation. Consequently, it seems to have made perfect sense for Epicurus to entertain a physicalist, yet non-reductive, causal faculty of the self, and we have Fr. 13 as strong evidence supporting such an understanding of his ideas. Epicurus’ quarrel with his predecessors in On Nature XXV becomes clearer and more credible when understood against the background assumptions of

100 Agency and atomism

his physics, and how the swerve was meant to function in that physics. He must have expected his readers to know its premises to a large extent by the time they began book XXV. And if his largest work On Nature was indeed a lecture course78 – which explains the recurrent impression that some of the texts are somewhat unpolished – then in light of my previous analysis we could say that, at the very least, Epicurus’ students were required to understand the non-reductionist characteristics of his arguments. Thus, although some of the fragments of book XXV may seem to be compatible with various positions in the philosophy of mind, the fact that Epicurus did not rehearse all the relevant principles of his physics in connection with his account of the causal faculty of the self in the surviving parts of book XXV does not mean that he did not draw on them. Instead, I believe, his arguments rested on the background assumption of the swerve to deny causal determinism and, hence, they were valid only within the particular framework of his complex atomist theory. Now we are in a position to start evaluating Epicurus’ conception of the self. The ‘thin’ idea of the self expressed by the reflexive pronoun in our fragments, incorporating the essential and accidental qualities of a human being, seems to provide a reference point for Epicurus’ non-reductive physicalism in his philosophy of mind. Such an idea of the self implies agent causation, which means that our actions are determined by our characters or dispositions. Our dispositions depend on a set of beliefs and values, which also partially determine our feelings and desires. Since it is essentially up to us in which way we develop, our characters are self-determined. By the introduction of the swerve Epicurus’ physics counters causal determinism at the atomic level, and consequently grants the possibility of such a token monism – everything in the universe is physical – which includes causal type dualism: there are two types of genuine causation in the universe: physical and mental. Type dualism in the causal realm is guaranteed by the thesis of the MR of identical mental states, which is a necessary concomitant of my understanding of how Epicurus deployed the atomic swerve. If we accept that with the swerve he did not merely do away with causal determinism but also granted the possibility of type dualism in the causal realm, then we also have to accept that for him self-determination implied the idea of a causally efficacious self. My interpretation of Epicurus’ philosophy of mind stops short of emergentism by not attributing non-physical causation to him, yet finds a way out of a reductionism he wished to avoid. Thus, the physicalism of the self depends on a combination of limited forms of indeterminism and determinism. As a result, Epicurus seems to have managed to put forward a philosophically credible account of the self, which not only comprises one’s self-conception of oneself based on one’s essential and accidental states of mental and bodily character, but also answers to Epicurus’ non-reductive physicalism in his philosophy of mind. It is the task of the next chapter to investigate, among others, how Epicurus conceived of the conceptual construction of one’s self in harmony with one’s self-intellection.

Agency and atomism

101

Notes 1 Sedley 1983. 2 Cf. Englert 1987, Mitsis 1988, Sedley 1988, Annas 1992, Sharples 1991–1993, Purinton 1999, O’Keefe 2002 and 2005, Hammerstaedt 2003 and Masi 2006. 3 Masi 2006 is an exception, however she does not deal with the fragments on selfreflective thinking. 4 Cf. Chapter 1, pp. 27–53. 5 For the relative order of the fragments see the Appendix. 6 Sedley 1988, p. 302. 7 O’Keefe 2005, p. 72. 8 The first two claims may appear to be rather too loose, the second as a result of my presentation of it. As I stated earlier, at first I only wish to present the general claims of the two main lines of interpretation, self-consciously avoiding textual examples. However, Sedley’s second claim gains some strength on the basis of some textual support; see Sedley 1988, pp. 299–301. 9 For a thorough survey of the evidence on Democritus and the complexities it has caused among different interpreters, see Lee 2005. 10 Steiger 1999, pp. 219–36. 11 However cf. Morel 1996 and Rudolph 2011. 12 Furley’s analysis points out the differences very neatly in Furley 1993. 13 Cf. Fr. 1191, 6,1 2, 4, in: Laursen 1995, p. 89. 14 I continue numbering the fragments with Arabic numbers for the sake of easier reference. I start from number 10 in this chapter – having introduced eight fragments from the early parts of book XXV, plus a text on memory from the later parts, which I labelled as Text M in Chapter 1. In order they are: Fr. 10 = 697 corn. 3 pz. 1 z. 1 = Arr. [34.4], as in Laursen 1997, p. 18; Fr. 11 = 1191 corn. 4 pz. 2 z. 3 = -22 inf./1191 corn. 7 pz. 1 z. 3 = –21 sup. = Arr. [34.21], and 1056 corn. 5 z. 2, as in Laursen 1997, pp. 19–20; Fr. 12 = 1191 corn. 4 pz. 2 z. 3 = -21 inf./1191 corn. 7 pz. 1 z. 4 = -20 sup., and 697 corn. 3 pz. 1 z. 2, as in Laursen 1997, pp. 20–1; Fr. 13 = 1191 corn. 4 pz. 2 z. 4 = -20 inf./1191 corn. 7 pz. 1 z. 5 = -19 sup., and 697 corn. 3 pz. 1 z. 3 = Arr. [34.22], and 1056 corn. 5 z. 3, as in Laursen 1997, pp. 22–3; Fr. 14 = 1191 corn. 7 pz. 2 z. 1 = –19 inf./1191 corn. 7 pz. 1 z. 6 = -18 sup., and 697 corn. 3 pz. 1 z. 4 = Arr. [34.23], as in Laursen 1997, pp. 23–4. 15 [] Vogliano 1936. 16 Following Masi 2006, p. 82; cf. her argument for this reading ibid., p. 85, n. 311. 17 [] [][] Masi 2006; [] [][+/–2/3][ +/–1/2] Laursen 1997;  / ] Hammerstaedt 2003; ] Laursen 1997. 18 Masi sides with those interpreters (cf. Masi 2006, p. 85, n. 314) who take the participle of  to be the subject and dismisses Sedley’s understanding (cf. Sedley 1983, p. 38). I also think that the neuter of  and  most likely refers to   as it seems to become clear from lines 12–13, but I would add that they seem to be used in this context interchangeably with one’s  and one’s causally efficacious self, thus, the overall sense of the passage does not make much of a difference. Also in the next sentence Epicurus says that ‘with these we especially do battle’ and it would sound rather odd if ‘many’ did not refer to living beings. 19 Possibly: the same method = aitiologikos tropos, the necessary cause = the atoms. 20 [] 697;  in 1056. 21  [] [] [] Masi;  [+/–2/3]. [+/– 0/1][] [+/–4/5]..[] Laursen. 22 See p. 86.

102 Agency and atomism 23 In some of these later fragments – 1191 -7 inf./1191 –6/1191 -5 sup., and 697,4,2,1, and 1056,7,4 = L&S 20C (13)–(15) – Epicurus criticizes the early atomists explicitly, and Democritus in particular, for being ‘blind to the fact that in his actions he was clashing with his doctrine’.This criticism of Democritus points towards the fact that Epicurus’ immediately preceding digression to argue against an explicit advocate of determinism – L&S 20C (3)–(12), quoted as Text P in Chapter 1, pp. 45–6 – must have been directed against someone other than Democritus or one of the early atomists, as Sedley has pointed out. I agree with him that the most likely target of Epicurus’ criticism in the digression passage was the later followers of atomism who adopted the view of universal necessitation implicit in Democritus’ theory but without the necessary modifications put forward by Epicurus. It may be the case that Epicurus’ rebuttal in Fr. 10–14 is directed against the same opponents if the criticism of Fr. 11–12 is read as an explicit attribution of responsibility to the atoms by those criticized. Accordingly, the function of the digression is to refute the explicit advocates of determinism in the face of their strongest possible arguments. Nonetheless, if Fr. 11–12 is read in a different way, as Epicurus criticizing an implicit view of the criticized theory, then Epicurus’ quarrel seems to be with the early atomists in general, possibly present all through book XXV; Epicurus indicating the superiority of his theory, occasionally pointing out that they should have made the same, necessary causal distinctions. If so, the function of the later digression passage is to refute even those who did not hold a deterministic view only by accident, but were explicit champions of determinism – whether or not they were themselves atomists. For criticism of Democritus and Nausiphanes cf. Cicero DND I 33, D. L. X 6–8; for the relevant textual interpretation cf. Sedley 1976, pp. 134–6 and Warren 2002, pp. 160–93. 24 As Laursen has pointed out – in Laursen 1988, p. 12 – the plural genitive of   can be taken either as a genitive of comparison as I took it following Sedley, meaning that the product differs from the atoms or as a simple subjective genitive meaning that the product differs within the atoms, as O’Keefe took it, at least in O’Keefe 2002, pp. 172–4, a solution he swaps with the genitive of comparison in O’Keefe 2005, pointing out that both are compatible with his understanding; cf. O’Keefe 2005 p. 96, n. 72. 25 Purinton 1999, p. 293. 26 Cf. Plato’s Timaeus where the rational soul is spatially co-extensive with the head, but still it is something non-physical. 27 Text and translation from Laursen 1995, p. 73. 28 Though undoubtedly what is produced is not explicitly mentioned in this fragment, the immediately preceding fragments (419 fr. 6, 419 fr. 7, cf. Laursen 1995, p. 72) discuss the soul, and 1634 fr. 1 (cf. Laursen 1995, p. 72), which Laursen aligns with 419 fr. 6 mentions the product. 29 For my solution cf. p. 99 and n. 78. 30 O’Keefe 2005, p. 98. 31 Cf. Fr. (d): ’ ’    []   , ,    ’    [][][], []  [ ][]   [ ]  [] ’  [] [– – –][+/– 3/4][+/– 2/3][– – –][]  []. . . My translation: ‘. . . and in some there is a separate one [action?] for each of the two [products?]. Because the atoms do nothing to the product [i.e. the occurrent mental state] that is shared, but although the action is one in accordance with the interlacement with some of the atoms, there is also a certain distinction that, not

Agency and atomism

32 33 34 35 36

37 38

39 40 41 42

43

44

45

103

otherwise, contributes a lot to the same action, and also, to a different action with some atoms. What was the part of memory . . .’ Text is from Masi 2006, p. 76. For the same understanding cf. Bobzien 2000, pp. 323–7. O’Keefe 2005, p. 99. For the Greek text and its reading see Chapter 1, pp. 48–9. For my interpretation of what exactly this means see pp. 88–9. Cf. Arr. [24.24.1-5]; Arr. [28. 7. 1] ff. O’Keefe tries to parallel this plural usage with Lucretius’ plural use of natura – in O’Keefe 2005, pp. 100–1. However, I doubt that he is correct to compare ‘first natures’ to natura. I suspect the term should rather be paralleled with primordia, one of Lucretius’ many names for atoms. Furthermore, in Philodemus’ evidence De Mort. 30.4–5 (cf. Henry 2009, pp. 68–9) ‘first natures’ definitely mean the atoms. O’Keefe 2005, pp. 101–2. One might mention the point made by Sharples (1991–1993, pp. 183–7) that even a compatibilist reductionist understanding is indeed compatible with the idea of downward causation, a point O’Keefe is eager to deny. What that downward causation meant for Epicurus will become clearer as my analysis continues. What that exactly means concerns question (d), which I have not yet answered entirely by arguing for the meaning of ‘first natures’. For the complete answer see pp. 82–3 and p. 99. Sedley 1983, p. 39. Cf. p. 99. By T1 I simply mean a given instant of time, in which case T2 means the following instant. I prefer to leave this part of the explanation rather vague as Epicurus seems to have done himself, since my primary intent is not to investigate how many times minima are exactly required for the operation of the mental state being discussed, but rather I wish to focus on its specific function, and hence I leave it open whether a given instant of time is to be taken here as a unit of  or one time minimum. In fact, if Epicurus wishes to avoid being an epiphenomenalist, then he had to suppose that these specific atomic motions – which leave space for the downward causation of the mental – need to be generated by the mental, otherwise his theory would render our volitions inert and completely dependent on the accidental motions of the atoms. He perhaps supposed that this peculiar atomic motion is characteristic of mental processes. This capacity, nonetheless, does not need to be some kind of non-physical entity, but it is manifest in affecting the mind atoms in the particular way it does. As we have learnt from Fr. 13 the ‘differentiating mode’, which is ‘not in the way as from another distance’ is meant to signify that Epicurus did not consider the  as a non-physical entity and he may have seen an essential connection between this tropos and the fact that the atoms are thrust together in a certain way () at the time of producing and initiating the kind of action according to the tropos ‘from the same distance’ as Fr. 10 indicates. Both claims are very neatly supported e.g. by Text M from book XXV. There is no single, common modern characterization of type dualism, hence, my attribution of this modern terminology to Epicurus may be just as misleading as my earlier criticism of the trend trying to fit Epicurus’ theory of the soul into the bag of one or another possible idea in modern philosophy of mind; cf. p. 71. However, I do not mean anything more by token monism and type dualism in the case of Epicurus than what I described in the sentence to which this footnote is attached. For an introduction on the modern terminology cf. Lowe 2000.

104 Agency and atomism 46 For this point cf. Caston 1997, p. 316. 47 Here I rely on a tacit assumption that there is an essential connection between one’s ‘product’ and one’s desires and volitions, for which I will argue later. 48 Peculiar in the sense explained in Fr. 13 on which see more later. 49 The neuter of  and  most likely refers to  as it seems to become clear from lines 12–13, and they seem to be used in this context interchangeably with one’s and one’s causally efficacious self. 50 Cf. Warren 2002, pp. 197–200. 51 Cf. pp. 85–6. 52 Strictly speaking, the whole text of what I could make sense of. For the Greek text and its reading see Chapter 1, p. 48–9. 53 Cf. Bobzien 2000 and 2006. 54 Laursen 1997, p. 10. I take this particular point of Laursen’s, however, I find his overall interpretation slightly confused. 55 This different temporality of mental states goes also very well with my interpretation of the different phases of sense-perception – the distinction between proleptic processes and their evaluation, as explained in Chapter 1. 56 Fr. 15 = 1191 corn. 7 pz. 2 z. 3 = -16 inf./1191 corn. 8 pz. 1 z. 2 = -15 sup. = Arr. [34.24], and 697 corn. 3 pz. 2 z. 3, and 1056 corn. 6 z. 1, as in Laursen 1997, pp. 28–9. 57  1056:  corr. Luc. in Laursen 1997, p. 28:  corr. Gomperz. 58 Concerning this text, I take it that if my suggested interpretation of ‘   ’ in line 6 is correct then the sentence will read: ‘. . .this sort of thing [is] not being produced () [i.e. the occurrent mental state] by necessity all the way to these specific things, but, on the one hand, as far as a soul or rather a soul with a disposition and movement of this particular size comes about, a thing of such a kind [is] being produced () [i.e. the occurrent mental state] from things of this kind [from a soul, or a soul with a disposition?] by necessity’. That is to say, according to this interpretation of the sentence the  flows from one’s disposition. However, there is a mutual interdependence between dispositions and the  cf. pp. 86–7. Furthermore, if one rejects my understanding of the  and takes it in the sense of mental developments, then one has to face the problem of explaining how it is possible that memories bring about the first constitution of an , as well as how we are to understand that dispositions can cause , objections I make above to the standard understanding. 59 For Fr. 16 see Chapter 3, p. 109. I do not proceed with numbering the fragments consecutively as I present them, but I number them in accordance with the order they appear in the papyri. Fr. 17 = 1191 corn. 8 pz. 1 z. 5 = -12 sup., and 697 corn. 4 pz. 1 z. 1 = Arr. [34.26], and 1056 corn. 6 z. 3, as in Laursen 1997, pp. 32–4. 60  Masi 2006;  Sedley 1983, Laursen 1997. 61 The translation is based on those of Annas 1992, Bobzien 2006, Laursen 1997 and Masi 2006. 62 Cf. Sedley 1983, Annas 1992, Bobzien 2000 and 2006. 63 Bobzien 2000, pp. 297–8. 64 Cf. my discussion of Lucretius in Chapter 4. 65 Cf., e.g., Cicero De Fato 23 and previous note. 66 The thesis of multiple realizability in the philosophy of mind asserts that a single mental kind (property, state, event) can be realized by many distinct physical kinds. 67 The reason why we are to think that for Epicurus a given mental state may remain unchanged after the occurrence of an undetermined swerve in the mind is that if

Agency and atomism

105

Epicurus held the view that undetermined swerves take part in the production of each and every volition then our volitions would be dependent on undetermined swerves, as a consequence of which our volitions would appear to be undetermined and random. We would keep changing our behaviour for no apparent reason, as if puppets moved by the atomic swerves of our minds (cf. Furley 1967, second study as well as Chapter 4, p. 140). 68 Cicero, De Fato 46. 69 Cf. Wigodsky 2007. 70             -    -    ’                                 . ‘The next thing to see – referring it to the sensa-

71

72 73

74

75

76

tions and feelings, since that will provide the strongest confirmation – is that the soul is a fine-textured body diffused through the whole aggregate, most similar to pneuma that has a certain blending of warmth, and in one way [the soul is] similar to the former [i.e. to pneuma], in another to the latter [i.e. warmth]. It [i.e. the soul] is the part, which has acquired a great capacity for change [i.e. variability] as a result of the fine texture of just these things [i.e. the atoms of the elements], a fact which makes it the more liable to co-affection with the rest of the aggregate.’ My translation, influenced by Kerferd’s interpretation; cf. Kerferd 1971. inter enim cursant primordia principiorum / motibus inter se, nihil ut secernier unum / possit nec spatio fieri divisa potestas, / sed quasi multae vis unius corporis extant. Translation from Long and Sedley 1987, vol. I, p. 67. For Epicurus conceiving of the soul as a mixture also cf. Aëtius IV. 3. 11 = fr. 315 Us. and Plutarch Adv. Col. 1118D = fr. 314 Us. Also cf. Kerferd 1971, pp. 84–5. The entire sentence is DRN III 282–7: consimili ratione necessest ventus et aër / et calor inter se vigeant commixta per artus / atque aliis aliud subsit magis emineatque, / ut quiddam fieri videatur ab omnibus unum, / ni calor ac ventus seorsum seorsumque potestas / aëris interemant sensum diductaque solvant. ‘In like manner it is necessary that wind and air and heat interact commingled throughout the frame, one element yielding place to another or rising pre-eminent in such a way that a unity be seen to be made of all, or else heat and wind apart and the power of air apart would destroy and dissipate the sensation by being separated.’ Translation from Rouse and Smith 1975, p. 209 and p. 211. This understanding is very similar to Galen’s and Arius Didymus’ description of certain Aristotelians who say that only the qualities of the four elements mix, but the substances or bodies do not. Cf. Galen On Hippocrates’ ‘Nature of Man’ (vol. 15, p. 32 Kühn) = SVF 2.463; Arius Didymus fr. 4, Dox. Gr. 449, 1–3, from Sobaeus Eclogae 1.17. Cf. some modern proponents of this thesis, Fodor 1974 and Kim 1993. However, the relationship between multiple realizability and reductionism is hotly contested; a number of serious objections have been made to the claim that reductionism is falsified by multiple realizability, cf., e.g., Lewis 1969 or Richardson 1979. I think this is the idea somewhat vaguely expressed in Fr. 13 when Epicurus says that the otherness the product takes from the atoms is not in the way as from another distance. The causal efficacy of oneself, that is, of one’s product is executed through the atoms, although the atoms are not sufficient to determine these causal powers on their own.

106 Agency and atomism 77 A bridge law, in Nagel’s formulation (Nagel 1961), expresses the idea of connecting the predicates of the reduced properties with the predicates of the reducing properties, or in Epicurus’ case, a bridge law would connect the predicates of the mental with the predicates of the atoms. 78 Cf. Sedley 1998, p. 104.

Bibliography Annas, J. (1992) Hellenistic Philosophy of Mind, Berkeley. Barnes, J. and Mignucci, M., eds., (1988) Matter and Metaphysics. Fourth Symposium Hellenisticum, Elenchos 14, Naples. Bobzien, S. (2000) ‘Did Epicurus Discover the Free Will Problem?’, Oxford Studies in Ancient Philosophy 19: 287–337. —. (2006) ‘Moral Responsibility and Moral Development in Epicurus’ Philosophy’, in B. Reis and S. Haffmans, eds., The Virtuous Life in Greek Ethics, Cambridge: 206–29. Brunschwig, J. and Nussbaum, M., eds., (1993) Passions and Perceptions, Cambridge. Caston, V. (1997) ‘Epiphenomenalisms, Ancient and Modern’, Philosophical Review 106: 309–63. Englert, W. (1987) Epicurus on the Swerve and Voluntary Action, Atlanta. Fodor, J. A. (1974) ‘Special Sciences: Or the Disunity of Science as a Working Hypothesis’, Synthese 28: 97–115. Furley, D. (1967) Two Studies in the Greek Atomists, Princeton. —. (1993) ‘Democritus and Epicurus on Sensible Qualities’, in J. Brunschwig and M. Nussbaum, eds., Passions and Perceptions, Cambridge: 72–94. Hammerstaedt, J. (2003) ‘Atomismo e liberta nel XXV libro  di Epicuro’, Cronache Ercolanesi 33: 151–8. Henry, B. W. (2009) Philodemus, On Death, Atlanta. Kerferd, G. B. (1971) ‘Epicurus’ Doctrine of the Soul’, Phronesis 16: 80–96. Kim, J. (1993) ‘Explanatory Realism, Causal Realism, and Explanatory Exclusion’, in D. H. Ruben, ed., Explanation, Oxford: 228–45. Laursen, S. (1988) ‘Epicurus On Nature XXV (Long–Sedley 20, B, C and j)’, Cronache Ercolanesi 18: 7–18. —. (1995) ‘The Early Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanesi 25: 5–109. —. (1997) ‘The Later Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanesi 27: 5–82. Lee, M.-K. (2005) Epistemology after Protagoras, Oxford. Lewis, D. (1969) ‘Review of Art, Mind and Religion’, Journal of Philosophy 66: 23–35. Long, A. A. and Sedley, D. N. (1987) The Hellenistic Philosophers, 2 vols., Cambridge. Lowe, E. J. (2000) An Introduction to the Philosophy of Mind, Cambridge. Masi, G. F. (2006) Epicuro e la Filosofia della Mente: Il XXV Libro dell’Opera “Sulla Natura”, Sankt Augustin. Mitsis, P. (1988) Epicurus’ Ethical Theory:The Pleasures of Invulnerability, Ithaca/London. Morel, P.-M. (1996) Démocrite et la recherche des causes, Paris. Nagel, E. (1961) The Structure of Science: Problems in the Logic of Explanation, New York. O’Keefe, T. (1997) ‘The Ontological Status of Sensible Qualities for Democritus and Epicurus’, Ancient Philosophy 17: 119–34. —. (2002) ‘The Reductionist and Compatibilist Argument of Epicurus’ On Nature Book 25’, Phronesis 47: 153–86.

Agency and atomism

107

—. (2005) Epicurus on Freedom, Cambridge. Purinton, J. (1999) ‘Epicurus on “Free Volition” and the Atomic Swerve’, Phronesis 44: 253–99. Reis, B. and Haffmans, S., eds., (2006) The Virtuous Life in Greek Ethics, Cambridge. Richardson, R. (1979) ‘Functionalism and Reductionism’, Philosophy of Science 46: 533–58. Rouse, W. H. D. and Smith, M. F. (1975) Lucretius, De Rerum Natura, Loeb edition, London and Cambridge, MA. Rudolph, K. (2011) ‘Democritus’ Perspectival Theory of Vision’, Journal of Hellenic Studies 131: 67–83. Sedley, D. N. (1976) ‘Epicurus and his Professional Rivals’, in J. Bollack and A. Laks, eds., Études sur l’Epicurisme antique, Lille: 119–59. —. (1983) ‘Epicurus’ Refutation of Determinism’ in : Studi sull’ Epicureismo greco e latino offerti a Marcello Gigante, 2 vols., Naples: vol. I: 11–51. —. (1988) ‘Epicurean Anti-reductionism’, in J. Barnes and M. Mignucci, eds., Matter and Metaphysics. Fourth Symposium Hellenisticum, Elenchos 14, Naples: 295–327. —. (1998) Lucretius and the Transformation of Greek Wisdom, Cambridge. Sharples, R. W. (1991–1993) ‘Epicurus, Carneades and the Atomic Swerve’, Bulletin of the Institute of Classical Studies, University of London 38–9: 174–90. Steiger, K. (1999) A Lappangó Örökség, Budapest. Vogliano, A. (1936) ‘Frammento di un nuovo “gnomologicum epicureum”’, Studi Italiani di Filologia Classica 13: 268–81. Warren, J. (2002) Epicurus and Democritean Ethics: An Archeology of Ataraxia, Cambridge. Wigodsky, M. (2007) ‘Homoiotetes, Stoicheia and Homoiomereiai in Epicurus’, Classical Quarterly 57.2: 521–42.

3

Self-narratives

Introduction For all living things which have not been able to make compacts not to harm one another or be harmed, nothing ever is either just or unjust; and likewise too for all tribes of men which have been unable or unwilling to make compacts not to harm or be harmed. (K. D. 32) The discussion of self-awareness and the study of its theoretical groundwork have already shown how rational reflectivity distances human beings from their social and cultural relations and opens up the possibility of the comparison of themselves to others. Epicurus conceptualized such reflectiveness as not only necessary for the explanation of human psychology – for example in one’s empirical acquisition of a certain set of concepts – but also for making distinctions between the living beings of the natural world or for refuting his philosophical opponents. In this chapter I shall begin with the examination of a fragment from book XXV (Fr. 16), which is our primary evidence for how Epicurus distinguished between human and animal selves. Epicurus’ differentiation includes a complex causal account of his assimilation of human wrongdoing to animal selves, and in his explanation Epicurus relates bad behaviour to a morally corrupted conception of the self. I will argue in light of this evidence that Epicurus refutes the advocate of determinism on similar grounds, the analysis of which will extend to the remaining surviving fragments of book XXV.

Animal and human causal capacities While explicating his positive account, Epicurus systematically contrasts admired and reviled human behaviours and makes it explicit that both types of outcomes are the product of the self and not the responsibility of the atoms. As we learnt in Fr. 12, unacceptable human behaviour flows from one’s moral character, which is similar to some agitating, congenital nature. As Epicurus notes in Fr. 12, such a nature is what a whole range of living beings have.1 In Fr. 16 Epicurus further scrutinizes this issue, clearly making a comparison

Self-narratives

109

and distinction between bad moral character and animal behaviour, in particular. The fragment is interesting because it adds to our information on Epicurus’ theory of animal behaviour, which he addresses to give reasons why animals do not develop the kind of selves humans do.Yet, at the same time, the fragment is concerned with the worry that if bad moral character is identified with an undeveloped congenital nature, then one could object that human selves are not distinguishable from animal selves; this seems to amount to the possibility of a life with an undeveloped congenital nature, as a quasi-Cartesian automaton. As it turns out from our look at Epicurus’ way of addressing these issues in Fr. 16, he not only avoids these problems but his solution lends support to my interpretation of his theory of mind. As Laursen has pointed out in his brief commentary on the text,2 the syntax is very uncertain. In my translation I have not aimed at a comprehensive account of everything, precisely because of the ambiguous syntax. Also, as I indicate in my translation, I follow some of the corrections made to the text by Jürgen Hammerstaedt.3 Fr. 16              []  [’] []   []        []      [’]        [+/–0/1][], ’ 5       [] ’        [ ] []  ’  []       [][ ]    [][].[ ] []  [ ] ’ [].[.]  [ ..]   [][] []  10  []       []             []    []   []         [],           []       [ ]  []   [] []        15  []  [][]   []    []  [+/–2/3] [+/–1/2]   . . .

(Fr. 16)4 . . . for the original constitution5 the product [i.e. of the occurrent mental state], and if the product [i.e. the occurrent mental state] is not able to create other things,6 neither at present, nor something else than what7 the first constitution would do8 and if it does not do some things somehow within a short time either by exerting some force and resisting, but it has all the same [characteristics of the first constitution], we [still] do not exempt the product [i.e. the occurrent mental state] from the cause, but we make it and the constitution one, similarly clearing that which we do not even admonish [i.e. the wild animals]. Many we do rearrange in accordance with some meaningful9 manner

110 Self-narratives of speaking. For that, which is exempted from causality necessarily must be exempted from it by its original constitution10 because it does not fulfil the same (cause) as this. And if, precisely because of the cause which comes from oneself, one goes in the direction of what is similar to the original constitution and this a bad one, then at times we censure him even more in an admonitory way, and not as we do indeed cleanse the living beings of what is wild in a similar case, by weaving their products [i.e. their occurrent mental states] themselves and the constitution into one thing. We do not use [in their case] either the admonitory and corrective manner nor the simply retaliatory one . . . purify . . .

The sense of this passage is far from clear, but we can say with certainty that the fragment hints at why and in what sense inappropriate human behaviour and animal nature are different from each other on Epicurus’ account. The point in general seems to be that unacceptable human behaviour comes from the cause ‘out of oneself ’ and for this reason one’s causally efficacious, occurrent mental state is not considered to be exempt from responsibility. But we do not judge animals as liable to praise and blame, because we believe the causality they exhibit is in accord with their original constitutions. From Fr. 16, we learn that animals are not exempt from having their own  – and on the strength of Lucretius’ evidence of animal mens and animus it would be rather strange if they did not have them.11 Nevertheless, in the case of the animals their  are not viewed in the same way as the  of the human self, which leads to the following aporia: if we do not find animals responsible for their actions, Epicurus has to clarify in what sense they develop differently from us, otherwise his account would be open to the objection that applying the terminology of the  to both humans and animals cannot be justified. The somewhat obscure answer Epicurus provides is that we weave the  of animals together with their constitutions. Viewing the two as one is simply based on, we may assume, our conception of wild animal behaviour – although this is not apparent in our sources. Just as in the case of humans we acquire the common conception of our own responsibility by seeing people acting in a responsible manner, in the case of wild animals our common conception of the causality of wild animals tells us that the causality of their  differs in some respect from that of humans and, thus, we do not consider them to be liable to praise or blame.This impression of causal difference is reason for us to conceive of animals the way we do, that is to say by weaving their occurrent mental states together with their constitutions. But then in what respect does the causality of their  differ from that of humans? These fragments only provide us with Epicurus’ rather weak explanation for why we should understand animal  to be woven into a unity with their original constitutions. This description points to the fact that if animal  are not the kind of causally efficacious mental states we take human  to be, there needs to be a reason which prevents

Self-narratives

111

animal  from having the same causal power. I think this reason is simply the force of their original constitution: animals do not develop their own individual selves the way we do because they have a different nature from the very beginning, naturally exhibiting a different kind of causal efficacy. This does not exclude an understanding of the animal  as a causal faculty,12 but only that animal causal power is always or at least largely in accord with the original constitution of animal minds. Epicurus says in lines 8–10 that ‘that which is exempted from causality necessarily must be exempted from it by its original constitution because it does not fulfil the same (cause) as this [i.e. human causality]’.13 The question is still: what does this difference consist of? One piece of evidence we might bring to bear in answering this question is a text from Hermarchus, one of the founding fathers, the so-called  of the Garden.14 According to him, animals do not have logos, which is clear from the fact that we cannot make contracts with them. Hermarchus says that civilization stems from human reasoning as to what is most useful. Since humans have epilogismos,15 they can calculate what is useful for them in their social relations, while animals are not capable of such rational calculations, because they merely act on the basis of their irrational memories.16 Polystratus’ evidence is in line with Hermarchus’ testimony:17 although he thinks we share some general features with animals,18 he denies that animals have logos and holds that we are, therefore, different in this respect. I think it follows that animals cannot make their lives consistent with self-intellection and even if they have some kind of reasoning, it differs from ours because they certainly do not have value concepts and inferential capacities.19 In light of these testimonies, we may say that animal  is not the same kind of causal faculty of the self as human  because animals do not have the capacity to interpret external influences in the same way we do and because their selves are not reflected upon, as a consequence of which, they have a different kind of agency from ours. That is to say, their  as the causal faculty of their selves is causally operative, but it exhibits an instinctive kind of response because animals do not behave in rationally active but rather in reactive ways based on their natural instincts. The idea in Fr. 16 is that humans have the mental capacity to revert to behaviour similar to that produced by their original (pre-rational) constitutions, which is like ‘going towards what is similar to some bad original constitution’: it only seems to make sense if we take the bad original constitution to be some kind of mental state which is like not having any control of reason. That is to say, humans may act in certain ways which are similar to the actions of instinctive infant behaviour, and which applies no epilogismos or hedonic calculus. Examples of such cases are indulgence in different kinds of pleasures without bringing to mind the Epicurean division of pleasures, or the unconditional avoidance of pain, without considering its possible, pleasurable outcomes in the long run. However, if an adult behaves in such ways, his or her occurrent

112 Self-narratives

mental state is not to be excused from responsibility, although he acts in a manner similar to those creatures we do not admonish, i.e. wild animals. In the case of the latter, we make their occurrent mental states one with their first constitution; that is to say, with the first constitution of the occurrent mental states. On the strength of the distinction between the first and the on-growing constitution of the occurrent mental state of humans, made in Chapter 2, this can mean that: (a) in such cases the further processed mental state is to be identified with its first occurrence.That is to say, the further processed human mental state is literally processed by the agent in the direction of what it was like in its earliest stage. Alternatively, (b) the first constitution of the occurrent mental state is not processed at all. If the first constitution of the mental state is not changed, it may be viewed as a ready response to our affections in a quasi-automatic fashion, which leads to some action, reactively following our instincts without any reasoned considerations or calculations (epilogismos). This would mean that while in the case of (a) Epicurus is saying that the occurrent mental state is realized because of one’s self-initiated cause, i.e. intentionally, in the case of (b) it is an instinct reaction, but both result in culpable behaviour. What does not become entirely clear from the text is whether (a) and (b) are really two different types of wrongdoing. (a), found in the second half of Fr. 16, describes intentional wrongdoing, while (b), in the first part of Fr. 16, describes wrongdoing out of ignorance. Perhaps, as seems more likely, Epicurus is talking about the same kind of thing, but while (a) emphasizes one’s responsibility, which is explicitly said to be applicable in the case of (b) as well, (b) describes the mechanism of this very same responsibility. What becomes clear, nonetheless, is that my interpretation of the  as the causal faculty of the self gains strength from Fr. 16. Epicurus’ comparison of human wrongdoing to animal behaviour reveals a kind of intentional, bad action which follows from the mechanism of the occurrent mental states regressing to some kind of bad original constitution of the mind. As I suggested earlier, it is the kind of mental state which is devoid of reasoning, and it brings to mind Epicurus’ alignment with the Socratic tradition, especially in connection with the discussion of self-reflective thinking in the early parts of book XXV. The point seems to be that if we act without rationally considering and evaluating our desires and, thus, without understanding their true nature, then our actions will be similar to those of wild animals, and bad moral character will inevitably follow.20 The cause is ‘out of ourselves’, i.e. it is our responsibility who we become and what sort of actions we deliver, and as we have seen in Fr. 17, Epicurus connects this cause of ours – ’  as he put it in Fr. 17 – with our beliefs. As Fr. 17 asserts, our beliefs function as filters of external influence. Depending on what sort of beliefs we have, each of us is differently affected by our environment – even if that environment is identical for many of us – and, consequently, we have different desires. For example, if one does not apply the hedonic calculus appropriately and believes that every pleasure is choice-worthy and every pain is to be avoided, then in many of one’s actions how one’s  is processed will not differ significantly from

Self-narratives

113

the consequent sort of animal . This fact clearly brings out the important role beliefs play as essential components of one’s character as well as one’s force of motivation in action. The following fragments are emphatic about this crucial role of beliefs in action or in the behaviour we exhibit, insofar as they are beliefs concerning ourselves. In Text P (0)21 Epicurus says: ‘we rebuke, oppose and reform each other as if we have the cause also in ourselves, and not only in our initial constitution and in the mechanical necessity of that which surrounds and penetrates us’.22 The point is clearly that our mutual praising and blaming of each other indicates the beliefs we have concerning each other, thinking of each other as being the causes of our own actions, that is, being responsible for them. This kind of explanation is in accordance with his aitiologikos tropos, and it clearly brings out once again the point we have seen in Fr. 10–14 that Epicurus’ conception of moral responsibility for actions in general is based on one’s causal responsibility for one’s actions.23 But most importantly, what becomes clear from Text P (0) as well as from the following fragments is that Epicurus found his arguments against the determinist strong because of our common beliefs concerning ourselves as responsible agents, which are derived from our common experience resulting in our proleptic recognition of responsible agency. This point reinforces the strong connection between these fragments and the fragment concerning self-reflective thinking in the earlier part of book XXV: we could see in Fr. 3 that thinking of oneself in accordance with the similar and non-different made the most sense on the interpretation according to which such a kind of self-reflective thinking involves one’s memories of responsible agents based on one’s prole p sis of the cause applied to agency. This unique concept usefully occurs in these later fragments of book XXV.24 Consequently, this connection suggests a kind of strategy on the strength of the surviving fragments of book XXV according to which, after the issue of self-reflective thinking is discussed as part of Epicurus’ ethical theory, book XXV continues with the discussion of the causal faculty of the self, ending in a digression, which shows how the result of our self-reflective thinking in the form of beliefs concerning our own responsible agency, coupled with the analysis of the atomic self, can be effectively used to argue against determinism. That is to say, most importantly, the outcome of self-reflective thinking is a conception of the self in Epicurus’ atomist theory which, on the one hand, preserves the moral responsibility of agency compatible with a non-reductive, physicalist explanation, and which, on the other hand, is used to argue against determinism. One of the main reasons why this upshot of self-reflective thinking is useful against the determinist is that we gain our conception of ourselves as responsible agents empirically. Both tropoi, the pathologikos and the aitiologikos, used in the explanation of self-reflective thinking are empirical ways to become aware of ourselves and to become aware of ourselves as responsible agents. The kind of epilogismos relating to these modes of self-reflective thinking brings Epicurus’ preferred means of investigation nicely out, taking the phainomena for granted

114 Self-narratives

and building up arguments in harmony with them; this proves to be a strong weapon in Epicurus’ arguments against the determinist. One of his complaints is that, when the determinist blames his opponent for having reasoned incorrectly and he imputes to himself the cause for having reasoned correctly, the determinist is not calculating on the strength of the phainomena ( ).25 If the determinist’s argument were to go through, it would make no sense to attribute praise and blame to each other, but the empirical fact that we see each other doing exactly that – that is, if we take the phainomena as a starting point – speaks against the idea of universal determinism. One may correctly say that the appearance of phainomena as non-deterministic is not a guarantee that it is non-deterministic at the atomic level.26 But most importantly as we learn from Lucretius, since nothing can come into being out of nothing, the swerve was introduced on the atomic level to account for the indeterminism at the phenomenal level.27 Consequently, even if the occurrence of indeterminist phainomena of the world is compatible in principle with atomic determinism or even if some of the later fragments of book XXV may seem to be compatible with such an atomic level – which I hope to have shown to be wrong by my positive interpretation of the fragments – Epicurus’ overall conception was certainly far away from a compatibilism of such a kind. The conception of ourselves accessed through the aitiologikos tropos provides us with a belief concerning ourselves, on the strength of which we conceive of ourselves as responsible agents and because of which the determinist is unable to succeed in convincing us that our actions are not our responsibility. As Epicurus puts it in Text P: (7) . . . If someone will not explain this, and has no auxiliary element or impulse in us to dissuade us from those actions which we perform, calling the cause for them ‘through us ourselves’ (’) but if for everything which we desire to do and we call the cause ‘through us ourselves’ he is giving the name of foolish necessity, he will merely be changing a name; (8) he will not be modifying any of our actions in the way in which in some cases the man who sees what sort of actions are of necessity regularly dissuades those who desire to do something in the face of compulsion. (9) And the mind will be inquisitive to learn what sort of action it should then consider that one to be which we perform in some way ‘because of us ourselves’ ( [][]) by desiring to perform it.28 When the determinist calls our actions necessitated, at most he is just misapplying the word. He does not succeed in rearranging any actions of ours. The reason why he fails is that he cannot change our impulse () towards the action we carry out. How are we to understand this failure of the determinist concerning our impulse? As Epicurus says, the man with broad vision – most likely he has himself in mind – is capable of pointing out to us in some cases that we have certain desires

Self-narratives

115

to do something, indeed, against something which is necessary. Such a case may be, for example, a longing for immortality which is truly an empty desire according to Epicurus, based on the false belief that infinite time can provide us with more pleasures – in fact, infinite pleasures – than a definite lifespan. A man with broad vision is able to change our conviction, our empty beliefs and by replacing them with the correct beliefs concerning the nature of things he frees us from vain desires. That is to say, by changing our beliefs he changes our impulses too for certain actions; however, the determinist by merely changing the name of what we call the cause by means of ourselves, calling it by the name of necessity, does not change our beliefs concerning our causal or moral responsibility for those actions which occur because of us. Rather our mind is eager to learn what sort of actions we are to consider done by ourselves driven by our desire () to act. I think this last claim also brings out the strong connection between these later fragments and the topic of self-reflective thinking, and they can be best understood in connection with each other.29 It is clear that the impulse () or desire () we have towards action is based on our beliefs. According to the sort of beliefs we have, we understand or filter the inflow of information differently and, consequently, our different reactions are based on our different sets of beliefs. If our mind wishes to learn the sort of actions which we ought to consider as done by us the only way we can find an answer is by turning towards ourselves and examining what we believe to be  . Once again our common recognition of the cause applied to ourselves – since we conceive of ourselves as responsible agents – is taken into the argument as the criterion of truth. Accordingly, the sets of beliefs we have determine our cognitive dispositions. The determinist has a kind of cognitive disposition incompatible with that of Epicurus and, as the self-refutation argument shows, it is even incompatible with his own behaviour. To that extent, the pragmatic arguments are also important to demonstrate the consistency of Epicurus’ tenets. This demonstration does not merely gain its credibility by pointing out that the determinist cannot change our beliefs concerning ourselves, since we Epicureans have more consistent sets of beliefs than theirs, but by the application of the prole psis  of the cause, it is also indicated that the cognitive dispositions of Epicureans depend on one of the criteria of truth. If these arguments against the opponents are successful, then they should urge the determinist or, as a matter of fact, anyone else to seek out the truth of Epicureanism. If the determinist fails to be persuaded by Epicurus’ arguments, then he fails to understand himself and the mistaken nature of his beliefs. That is, the pragmatic arguments are also meant to bring out the point that although the view of the determinist is already proved wrong by the selfrefutation argument, if he carries on arguing on account of his false beliefs and judgements, he is acting irrationally. His irrationality does not mean that he is incapable of reasoning, but rather as we saw in Fr. 16 in the case of the blameworthy person, he also processes his causal faculty, his  towards a

116 Self-narratives

similar disposition, which is like that of wild animals. That is to say, his cognitive disposition is deficient because it is based on empty beliefs, as Epicurus’ arguments against him have shown and, consequently, his beliefs cannot be justified. By not recognizing his deficient cognitive disposition, his lack of self-knowledge is revealed. Although Epicurus’ arguments against this determinist are only a digression, I think this digression has also demonstrated how Epicurus’ conception of the self as explicated in book XXV can be applied in a debate.30

The narrative self A further element of Fr. 16 seems to suggest – in addition to the philosophical analysis of one’s moral responsibility – a narrative idea of the self. In lines 10–14 of Fr. 16, Epicurus says the following: And if, precisely because of the cause out of oneself, one goes in the direction of what is similar to the original constitution and this a bad one, then at times we censure him even more in an admonitory way, and not as we do indeed clear the wild among the living beings in a similar case, by weaving their products [i.e. their occurrent mental states] themselves and the constitution into one thing ([ ] []  [] []  ). We exempt wild animals from responsibility by weaving their products (their ) together with the [first] constitution of their mental states. The plural active participle of the verb of  indicates that it is we who, after observing these wild animals, weave the different stages of their mental states together. This activity on our behalf seems to mean that people give a narrative as to why animal selves should not be considered morally responsible for their actions.This understanding becomes more likely if we compare this bit of Fr. 16 with a text from Plutarch who used some of the same vocabulary for a narrative conception of the self: That everyone has within himself the store-rooms of good and bad cheer (euthymia, dysthymia) and that the jars of goods and evils are laid down not on the threshold of Zeus, but in the soul, is clear from the differences in people’s emotions ( pathe ) . For foolish people overlook and neglect even present goods because they are always intent in their thoughts on the future. But wise people make even what no longer exists to exist vividly for themselves by the use of memory. The present, which allows contact with only the smallest portion of time and then escapes observation, no longer seems to the foolish to be anything to us or to be ours. But just as the man pictured in Hades plaiting a rope allows a grazing donkey to consume what he is plaiting, so forgetfulness (le t he ) , unaware of most things and ungrateful, snatches and overruns things, obliterating every action and right act, every

Self-narratives

117

pleasant discussion, meeting, or enjoyment, and does not allow our life to be unified, through the past being woven together with the future (         ). Whatever happens, it immediately consigns it to what has not happened by forgetfulness, and divides yesterday’s life from today’s as something different, and tomorrow’s similarly as not the same as today’s. Those in the schools who refute the fact of growth on the grounds that substance is perpetually flowing make each of us in theory ever different from himself. But those who do not preserve or retrieve the past in memory, but allow it to flow away from under them, make themselves needy every day in actual fact, and empty and dependent on tomorrow, as if last year and yesterday and the day before were nothing to them and had not actually happened to them. This is one thing that disturbs good cheer. Another does so more, when people drift away from cheerful and soothing things and get enmeshed in recollections of the disagreeable. It is as when flies slip off the smooth patches on mirrors and catch hold of the rough and scratched ones, or rather as they say that beetles in Olynthus, falling into a place called ‘Death to Beetles’, are unable to get out, but twist and turn there until they die. That is how people who slide into the memory of ills do not want to recover or revive.What we should do is make the bright and shining events prominent in the mind, like the colors in a picture, and hide and suppress the gloomy ones, since we cannot rub them out or get rid of them altogether. For the harmony of the cosmos, like that of a lyre or bow, involves bending in two directions, and nothing in human affairs is pure or unmixed. But as in music there are high and low notes, and in language consonants and vowels, and a musician or language specialist is not one who dislikes either of these, but one who knows how to use them all and mix them appropriately, so . . . we too should make the mixture of our life harmonious and appropriate for ourselves. (De Tranq. Anim. 473B–4B)31 From Plutarch’s perspective, to secure our tranquillity we need to create a unified conception of our life by weaving our memories together. Otherwise we will be like the man in the painting who, plaiting a rope in Hades, throwing it over his shoulder as he plaits it, does not notice that a donkey is eating it up behind him. If forgetfulness becomes dominant in our lives, we cannot achieve tranquillity. Plutarch also connects this idea of forgetfulness originating from the lack of a consistent set of memories with the Growing Argument and implies that we shall have no continuous selves if we act this way.32 Furthermore, it is not enough to remember selectively: to have a picture of ourselves we have to weave in the bad parts as well as the good ones, for a picture needs dark patches as well as bright, as music needs low notes as well as high. But we must not wallow in the bad parts, like beetles struggling until they die in the place

118 Self-narratives

called ‘Death to Beetles’, since tranquillity does not originate automatically from any memories, but from the skilful weaving in of good and bad. Consequently, it is our responsibility to create ourselves by the artful application of our memories. As Richard Sorabji has pointed out, there are two ways to understand Plutarch: on one of the interpretations, there is no continuous self until the weaving has been done; on the other, there is already a human being but until human beings weave the story of their lives, they will not have acquired an identity of their own. Sorabji thinks that the appeal to the Growing Argument may indicate that the more radical, first option is being discussed; however, he also points out that the fact that genuine memories are woven – i.e. the memories of the very same human being who originally had the experiences remembered – suggests otherwise. The Epicurean evidence on the function of weaving is certainly much slimmer and it looks very different. Nonetheless, I think it can be fruitfully compared to Plutarch’s conceptualization. The narrative the Epicureans give to animal selves by weaving their different mental states together naturally does not involve memories concerning one’s human self as in Plutarch. But the conceptual classification of wild animals, on the strength of Epicurus’ empirical theory of knowledge, must have depended on their observations and the consequent memories of their encounters with them – even if there is no explicit connection between the narrative of weaving and memories in Fr. 16 or anywhere else in Epicurus’ surviving works for that matter.33 Their observation of animal behaviour and its comparison to human behaviour, however, was naturally part of their experience (see DRN III 294–306) and Epicurus’ fragment on how they excused animals of moral wrongdoing suggests that it depended on their narrative for animal selves. As I argued in the previous chapters, building one’s human identity for the Epicureans was essentially based on one’s memories of oneself in relation to one’s surrounding social, cultural and natural environment. The central passage for memory being a psychological link for personal identity is Lucretius’ idea of palingenesis: And, if suppose, after the nature and power of soul and mind have been pulled away from our body, the soul does perceive, then this too is nothing to us – we who are formed and constituted as a whole by the joining and union of body and soul. Nor, even if time should gather our matter after death and should rearrange it once more as it is now placed, and once again the light of life should be given to us, should it matter to us in the slightest that even this had happened, when the recollection of our selves has once been broken? Moreover, we now feel no concern over those we have been in the past, nor does any pain now afflict us from them. For when you look back at the whole vast expanse of past time and at how varied the motions of matter are, then you may easily understand that these very same atoms – the ones of which we are now composed – have

Self-narratives

119

often before been placed in the order in which they now are. But we cannot recall that in our mind’s memory, for a break in life has been cast in between, and all the motions have wandered here and there, far and wide, away from the senses. (DRN III 843–61)34 Lucretius’ thought-experiment does not include the idea of weaving memories together. However the uninterrupted possibility of retaining one’s memories as the necessary condition for personal identity – on Sorabji’s interpretation of the passage, with which I find myself in agreement – displays how similar the Epicurean criterion for personal identity was to Plutarch’s description of self-intellection.35 Epicurus even argued against the determinist in On Nature XXV partly on the strength of his conception of a kind of self-reflectivity, which depends on one’s memories of similar agents to oneself, bringing about the conception of one’s responsibility. The reason, thus, why Epicurus cannot exempt the person in Fr. 16 who behaves similarly to the ‘wild of living beings’ is that he had a different account for a human self on the strength of the different contents of their respective memories than for an animal self. Consequently, the kind of narrative based on Epicurus’ conceptual apparatus used in his analysis for the causal faculties of living beings in Fr. 16 – given his empirical theory of knowledge and the content of On Nature XXV – must have been related to one’s classified stock of memories.36 As we have seen in the analysis related to Fr. 3 in Chapter 1, the prole psis  of the cause is built from the recurrent observation of human causal behaviour, and we create our conception of ourselves as responsible agents on the strength of memories of such actions. These memories are the building blocks of our own identity in relation to our society and the cosmos. The formation of the concept of our own responsible agency translates well into our classification of our memories, which makes all the more sense if we conceive it through a narrative, which seems to be just the very activity of self-reflective thinking. Epicurean self-intellection seems to have consisted, therefore, in giving a narrative to our experiences on the strength of which we not only create certain essential concepts, but we also classify our experiences by means of which we create a conception of ourselves or of the human self. Such self-intellection was extended ideally in a circle of friendly selves: just as solitary dining was animal-like for the Epicureans (Seneca Ep. Mor. X 10), they wished to share contemplation with their friends (Ep. Men. 135, Diog. Oin. Fr. 62, D. L. X 22). The shared pleasure of philosophizing was not only a therapeutic means for them, but also the core activity for their intellectual improvement. And their self-reflective thinking implied a dialogue between them – hence their rebuking, opposing and reforming each other in Text P (0) – implying an account of their conception of one another, as we shall see in Chapter 5 in detail. Thus it seems very likely that self-intellection was based on continuously renewing a portrayal of oneself for achieving a correct conception of Epicurus’ ethical ideal.

120 Self-narratives

Sensing responsibility In the last fragments of On Nature XXV we find ourselves again in the choppy waters of the technical vocabulary in which Epicurus explicates the complex causal factors of the self. Let us join him in his lecture course, in harmony with the possible intention of his investigation, to understand the causal network of nature along the way. Fr. 18 . . .[]

[]  [ ] []    []                              5 []             []                    []  [], ’      [],   ’  []    10 . . . [+/– 12/15 ]  ’  []  []  [] []  [] []    [][] [], []  [] []  [][] [+/–2/3] [+/-0/1][ - ]  [] []    [][]  [],  ’ []  [][], [][] ’  15 [] ’ ,   []   [][]  [+/– 1/2 ][+/– 3/4].[ - ]            [ ]  , []   [][] [] []   []       [][.] . . . [.]..[

(Fr. 18)37 . . . and often both possess the same causal responsibility, without one of the types being followed by the other ones, or without being drawn on or forced by each other, and depending on time, age and other causes many of such things happen. Hence both the rational calculus of the end itself [the hedonic calculus]38 and the beginning were responsible, and so were we. And the part that came from us was the sensory recognition of that (the dictum): ‘if we fail to grasp the canon and the criterion of all things done under our opinions and irrationally follow the inclinations of the crowd, all those things will be lost in relation to which we make investigations, and superfluity . . .’ This is the very same thing, which both was generated as a permanent attribute and was a kind of seed – as I have been repeating –, leading from the start to something other, and when this is present we think and form opinions . . . and much there is that happens with the help of our nature, much that happens without that help, much that happens when our nature is put in order by ourselves; partly because our nature itself has shown some guidance, not [just because of]39 the on-growing [process], but also

Self-narratives

121

because of the [eidol a] that penetrate us from the surrounding environment and serve as guides to our improvement, instead of just following. Because, from the beginning, they drive40 us out of 41 our inertia and they lead to a powerful wonder . . .

Many things are obscure in this passage. One of the most important questions is the referent of  in line 1. From the remaining text it is only possible to connect it with the hedonic calculus and the beginning in lines 4–5. Even though this connection is less than certain, it gains some support by  (‘hence’) in line 4, at least insofar as it seems to make sense to draw a conclusion concerning the hedonic calculus and the beginning if they were mentioned earlier as well. Accordingly, Epicurus may be rendering the hedonic calculus and the beginning by the word , both acquiring the same cause without having been drawn on or forced by each other, whatever that is supposed to mean. This enigmatic claim becomes somewhat clearer a few lines later in what Laursen named in his translation the ‘dictum’.42 Epicurus explains, in a somewhat odd way, that our causality is the result of the sensory recognition () of what ‘the canon and the criterion’ are, without which we could not lead a pleasant and rational life. I understand this claim as maintaining that the recognition of our responsibility depends on the sensory recognition of what the criteria are, which is a straightforward empirical claim. Our causality here has to be understood in a qualified sense, meaning the kind of causality of people who act rationally in accordance with Epicurus’ theory, otherwise Epicurus’ statement would lose all its force, since it is possible to act on the basis of other rational considerations as well, e.g. as a Stoic or as a Sceptic. Therefore, if one has subscribed to Epicurus’ views and is to attain the Epicurean end, one needs to recognize one’s causality as a result of the ‘sensory recognition’ () of what ‘the canon and the criterion’ are.43 Epicurus’ canon consists of three components,  and the , and as I have argued in Chapter 1, , the criterion of recognition. Perceptions and affections were both claimed by Epicurus to be beginnings in the process leading one to action and, as I have also argued in Chapter 1, Epicurus regarded perceptions as coordinate with affections – since for affections to qualify as criteria in one’s actions they need to be connected with the relevant perceptions. As I have also argued, the respective proleptic processes belonging to each criterion are connected in the form of a typos, representing the general characteristics of things. Now it seems that on the strength of Fr. 18 Epicurus thought that the condition of the understanding of our responsibility was the sensory recognition of these criteria, according to which if we do not recognize empirically what the criteria are, we cannot behave rationally. What does this claim and the sensory recognition of the criteria mean? In a sense, I think, that is what the remainder of Fr. 18 is explaining. ‘The cause we had . . . is the very same thing, which both was generated as a permanent attribute and was a kind of seed’, a causal capacity, which can be further

122 Self-narratives

developed in many directions (‘leading from the start to something other’), into many kinds of . And ‘when this is present, we think and form opinions . . . and much there is that happens with the help of our nature, much that happens without that help, much that happens when our nature is put in order by ourselves’. Our self-reflective thinking in accordance with the pathologikos tropos and aitiologikos tropos is, consequently, in a sense nothing else but the sensory recognition of the criteria on the strength of which we can practise our rationality and rearrange ourselves or ‘put ourselves in order’. Epicurus makes it even more explicit that this is the case in the following lines: the on-growing process of our causal efficacy is not the only way to conceive our responsibility but an atomist also has to recognize that the eido l a penetrating us from the surrounding environment serve as guides to our improvement, as we have seen in Fr. 3 in accordance with the pathologikos tropos, ‘because, from the beginning, they drive us out of our inertia and they lead to a powerful wonder’ (Fr. 18). Methodological considerations The language of the dictum (e.g. ) is reminiscent of Epicurus’ first principle of methodology (e.g. ), so let us compare them briefly: First Herodotus, we need to have grasped those which are subordinate to the utterances/vocal sounds44 (  ), so that we may have them in reference to which we judge what is believed or sought or perplexing, and so that it may not be the case that everything be undistinguishable for ourselves as we demonstrate to infinity, or that we have empty utterances. For it is necessary that the first concept ( ) in accordance with each utterance/vocal sound be seen and not require demonstration, if we are to have [a standard] to which we shall refer what is sought or perplexing or believed. (Ep. Hdt. 37–8)45 Epicurus’ basic expectation of his first methodological rule of investigation is that one who inquires into nature must have grasped what is subordinate to the utterances or vocal sounds. Epicurus speaks of these things as standards by reference to which we make judgements. He describes them as something primary that we think of in connection with our utterances or vocal sounds, as self-evident and as things, which do not require any demonstration. The obvious question presents itself: what are the things subordinate to utterances or vocal sounds? Epicurus’ shift from the to may suggest that the former are some kind of thoughts. As Asmis has pointed out, this view can be supported by a number of Epicurean texts: for example, in On Nature XXVIII, Epicurus claims that some false opinion is subordinate to articulate sounds.46 On the other hand, Cicero, Plutarch and Sextus report that the

Self-narratives

123

Epicureans correlated words directly with physical reality.47 Sextus compares the Epicurean theory with that of the Stoics and points out that while the Stoics correlated three things with each other – the ‘significant’ ( ) which is the voice (); the ‘signified’ (/ ) which is the object of thought; and the underlying external thing () – Epicurus did not recognize the category of the signified. Sextus like Plutarch used Stoic vocabulary to explain the Epicureans’ twofold distinction; Cicero’s Latin and Diogenes’ Greek, however, are really straining to render Epicurus’ own usage. For example, Diogenes reports that the Epicureans rejected dialectic by claiming that ‘it is sufficient for physicists to proceed in accordance with the utterances/vocal sounds of things ()’. As Asmis concludes on the strength of this evidence, Epicurus’ own terms for the distinction between the ‘significant’ and ‘the underlying things’ were  and , respectively.48 Consequently, physical things appear to be subordinate to the utterances. But how then are we to conceive of the primary concepts which Epicurus seems to be using as interchangeable with the term ? I agree with Asmis that they do not form a separate semantic category, because the first concepts represent physical reality just as it is. Consequently, if these first concepts depict things without the possibility of misrepresentation or distortion, there is no difference between looking at things directly and thinking of these things by means of first concepts. The obvious objection to such a view is that our words do not always designate something that exists in reality, e.g. one may think of the standard centaur example. But the Epicureans seem to have tried to undo this objection by distinguishing between primary concepts of existing things and opinions formed with the help of these concepts, which are not necessarily of existing things. Given the striking similarities between Epicurus’ first principle of methodology and Diogenes Laertius’ characterization of prole psis  (D. L. X 33), there has been a scholarly consensus that the ‘first concept’ is just an early formulation of the notion of prole psis. This seems to be fairly certain. Nevertheless, unlike  Asmis, who thinks that in the first rule of investigation Epicurus does not indicate that the initial concepts are empirically formed, I think Epicurus already implies the awareness of physical reality in the first rule by the temporal indexing of the concept. ‘First’, in my understanding, indicates that the concepts formed at the beginning of one’s psychological development are in question.49 The awareness of physical reality becomes more explicit in Epicurus’ second rule of inquiry, in which Epicurus specifies that the ultimate standard of an investigation is an awareness of the physical realm exactly as perceived by the organs of perception.50 These first concepts do not require any demonstration, otherwise there would be an infinite regress of demonstration with the result that nothing could be judged, since there would not be any standard to which we could refer in our investigations. Imagine that we would need to demonstrate the concept of ‘man’ and we could only do so by referencing some other concept of ‘man’,

124 Self-narratives

and this concept would need to be demonstrated in turn, and so on ad infinitum, without ever establishing the referent or the meaning of the word. The other alternative is that we would have empty utterances, that is, sounds without any content or meaning.51 These first concepts are the standards to which we refer what is believed, or sought, or perplexing, in order to make a judgement.They are prior in the sense that they are the starting points of investigations, without which ‘we would not have sought what we seek, unless we had previously learned it’ as Diogenes puts it. However, it is a matter of debate whether such investigations require primary, rational concepts as Epicurus’ first principle of methodology indicates or whether the reason for Epicurus’ terminological development, for changing ‘first concept’ to prole p sis, was that he thought a proleptic process of recognition would suit his principle of methodology better. Strengthening the connections As I suggested earlier, there may be a correspondence in Fr. 18 between  in line 5 and the canon in the sense that perceptions and affections, being the criteria, are the primary part of a process leading to action, as well as between the hedonic calculus and the criterion, the latter of which we could dub one’s prole psis. Now it is obvious that the hedonic calculus and the prole p sis cannot be  identified; however, one’s prole pseis certainly play a large role in one’s hedonic  calculus. Accordingly, the tripartite dependence of causality set out in Fr. 18 may be understood in the following way: we are the cause of our actions insofar as we recognize the criteria in harmony with which we ought to act as good Epicureans. And those criteria are our perceptions, the related affections and our prole pseis. However, these criteria are themselves causes of our actions,  because – as we have seen in the complex structural formation of the   – it does matter how we filter the various external and internal influences. That is to say, the outcomes, our actions are because of us, the processed  on which our actions are based are the result of various causal factors and Epicurus seems to be taking those factors into account independently of the account of the . Fr. 18 is not limited to the discussion of the atomic foundations of the causal faculty of the self, but it seems to introduce some epistemological terminology into his discussion of one’s causality and, hence, perhaps placing the ending of On Nature XXV into a wider context. I also suggested that  in line 1 of Fr. 18 may refer to  and the hedonic calculus, which would lead to the odd result following the above analysis that the hedonic calculus involves one’s prole pseis, and one’s perceptions  and affections sometimes attract each other, but often they are also equally responsible without being drawn on or forced by each other. However, it is not as odd or unclear as it may initially sound. Due to one’s present perceptions and affections, e.g. when experiencing a bodily pain while being at the doctor’s and undergoing some treatment, one may wish to give in to one’s natural

Self-narratives

125

tendencies to avoid pain by halting the process. However, one may also calculate the resulting pleasure from getting rid of one’s troubling pain and, thus, decide to keep on focusing on the calculus. Such a calculus certainly involves all the elements of one’s present perceptions, affections and some of the relevant prole p seis, all these mutually affecting each other. Nonetheless, there may be an opposite case when all these separate factors are still causally affective without being drawn on or forced by each other, though admittedly it is much harder to think of what these circumstances would be in Epicurus’ explanation of structured causality. What becomes, nonetheless, entirely clear from Fr. 18 is the following: if we follow others without examining available options, we will act irrationally () and end up losing everything in relation to our investigation, as Epicurus states explicitly in his dictum. That is to say, he conceived of anyone acting in accordance with considerations other than those he himself suggested, as acting irrationally. This point nicely connects to his comparison of animal and human behaviour in Fr. 16. As I suggested earlier, in Fr. 16 Epicurus compares the characteristics of one who acts out of the cause of himself inappropriately with that of acting according to his original constitution, which is exactly what wild animals do. Actions we generally admonish are similar to acting merely in accordance with one’s original constitution without any reasoned, i.e. Epicurean, consideration, and Epicurus takes such action as irrational, as explicitly stated in Fr. 18. In the last part of Fr. 18 after the dictum, Epicurus picks up once again the topic of seeds leading from a beginning to something else. Unfortunately, it is impossible to say with certainty what the referent of   is in line 10, which is described as . What can be stated with certainty is the connection Epicurus makes between Fr. 18 and Fr. 17,52 as indicated by   in line 11. In Fr. 17 Epicurus is explicitly talking about seeds of actions, thoughts and dispositions. Nonetheless, I think it is very hard to make sense of  in this context, which must have been an important term for the Epicureans in giving an account of natural tendencies and regularities in the cosmos. It is very odd to conceive of segregated atomic seeds, some being just the seeds of actions, some being just the seeds of thoughts, etc., and to imagine that a particular action was dependent on a group of seeds, while a particular thought reliant on another group. Such an explanation certainly would not make sense in view of Epicurus’ analysis of the causal faculty of the self. Instead, taking  in a figurative sense in this context, meaning ‘the origin’, which engenders or begets things causally, may lead to a better result; however, without any parallel passage where  is clearly used in this way, it is impossible to establish this sense.53 A blurry ending The last major fragment seems to be wrapping up the results of Epicurus’ arguments concerning the different causal factors.

126 Self-narratives Fr. 19 . . .[..][..]  [..]   [  ] [ ][]      []  []    []   []     []    [[]] []    5        [  ]  [] [ ] []   [  ][ ] ..[..][]  []    [][]  [].     [] []    []   []  10                      []  []    [] [+/– 2/3] . . . (Fr. 19)54 . . .[We learn by reflecting on]55 our sounds, thoughts, afterthoughts and representations and of the durable and non-durable disturbance or happiness in the soul – the reason for searching for the beginning, the canon and the criterion, gradually. For these [reflections] led to the rational calculus of the criterion and from the sensory recognition of the criterion56 to the consideration of these . . . gradually up to the phase of investigation that I mentioned before. In fact, these [reflections] mutually provided the cause and the need for each other, and each of the two alternately, appearing to us immediately led to another thought, first created in us gradually and then rapidly flowing outside, but then occupying more and more of our thoughts, some of them [i.e. our thoughts] because of the natural causal responsibility of the on-growth and the withdrawal of softness, some of them [i.e. our thoughts] because of the [liability] that comes from us . . .57

I think one can approach Fr. 19 using several interpretative strategies. Although the French translation based on the work of Sedley and Brunschwig has made the sense of the fragment clearer, what the subjects of the relative pronouns are is still uncertain, and that makes any interpretation very conjectural. I insert into my translation a parenthesis starting in the hypothetical, missing part of the text, making the ‘reason for searching for the beginning, the canon and the criterion’ the object of the missing and therefore, hypothetical main verb. Accordingly, the first sentence of the text would describe the gradual process of becoming aware of our epistemological principles by realizing the reason of the search after them on the basis of our mental conditions, while we are trying to make sense of ourselves. What reason is there to search for the beginning, the canon and the criterion? Most likely it is simply that without them we could not patch together a picture of ourselves because without these criteria we could not sort out our thoughts and beliefs. According to this interpretation,  in line 4, thus, might refer to all those thoughts and so on we reflected

Self-narratives

127

on, because it is those that led to ‘the rational calculus of the criterion’, most likely in the sense of coming up with the Epicurean calculus of what the criterion is, along which one can reach a certain phase of an investigation, as Epicurus explicitly points to an earlier discussion of the same subject. However,  in line 8 seems to go against such an understanding of  in line 4: in what sense do our thoughts and so on provide for each other the cause and the need and why is their number suddenly reduced to two? Plus it would indeed be odd to say that  refers to our thoughts and so forth and that they occupy our thoughts more and more. I think Masi’s solution of translating  in line 8 as ‘reflections’ is much better, although it is somewhat strange that she translates  in 4 as ‘these categories’. I think the text makes better sense if we translate both instances of  as ‘reflections’. Accordingly, the first sentence of the text would also describe the gradual process of becoming aware of our epistemological principles by realizing the reason for the search after them on the basis of our mental conditions, as we are trying to make sense of ourselves. And the point still stands that without criteria we could not piece together a picture of ourselves because without them we could not sort out our thoughts and the beliefs associated with them. And if we take  in line 4 to mean ‘these reflections’, i.e. the reflections we previously had in our thoughts and so on, this also leads to the interpretation according to which we come to realize the necessary criterion for leading a rational life through such reflections, which help us to proceed further within our investigation of ourselves and that of our causality, which is the major subject of On Nature XXV. These reflections, according to the fragment, are somehow also supposed to provide the appropriate causal scheme; however, at this point, the text becomes too damaged to give any further positive interpretation. I think it is fairly clear that Fr. 19 also supports my interpretation of the structured causal faculty of the self, with the further noteworthy twist that it connects the dependence of one’s self-conception to the technical vocabulary of Epicurus’ epistemology. The end of book XXV again: Fragment 1056 corn. 8z. 3 = Arr. [34.33] lines 4–7:   []  []        .

So, of those, which we proposed at the beginning, we have explained both the pathologikos tropos and the aitiologikos tropos (trans. David Sedley). According to Sedley the fragment can be understood as stating that we have explained at least the pathologikos tropos and the aitiologikos tropos by the end of the book, but there are other tropoi to be explained in the following book(s), perhaps bringing in the swerve in the discussion of another tropos in book XXVI. As I argued in Chapter 2, by book XXV Epicurus’ students had to

128 Self-narratives

understand the general, non-reductionist characteristics of his theory including the swerve. However, that does not exclude that the discussion of the swerve popped up again in the later parts of the thirty-seven-book-long treatise (cf. Philodemus’ references in De Piet. to many separate books of On Nature concerning Epicurus’ discussion of the gods).

Conclusion Epicurus’ conception of reflectivity originates from our capacity to make both the cosmos and our own existence in it the objects of our thinking. We contemplate their connections and in consequence our attention realizes a conception of ourselves. We become what our awareness of ourselves makes us be, to a great extent depending on our social and cultural relations. Epicurus not only provides the theoretical framework for such self-awareness and that of our causally efficacious selves, but he also takes into account the epistemological conditions of self-intellection towards the end of On Nature XXV, and constantly promotes the possibility of moral reform. Perhaps one of the most exciting elements of his account is the narrative conceptualization of living beings, which seems to have had a role in Epicurus’ understanding of what to make of our selves. Although the evidence for it is rather slim, yet, since Fr. 16 suggests that Epicurus thought that it is our narrative which helps to make a distinction between animal and human selves, it does not seem to be a huge leap therefore in the argument – taking into account as well that memories certainly played a large role in one’s Epicurean personal identity – that Epicurus also contemplated a narrative idea of the human self. Before contextualizing the different possible accounts of self-intellection in the Garden and outside, I turn to Lucretius in the next chapter to see how my findings concerning On Nature XXV converge with his evidence of the swerve. Lucretius hardly needs an introduction, so I will immediately start with the questions to which the idea of the atomic swerve presents itself as a possible answer.

Notes 1 Such behaviour may be compared to Aristotle’s complaint against young people or those who only follow their feelings at N. E. I 1095a 4–11. 2 Laursen 1997, p. 64. 3 In Hammerstaedt 2003. 4 Fr. 16 = 1191 corn. 7 pz. 2 z. 4 = -15 inf./1191 corn. 8 pz. 1 z. 3 = -14 sup., and 1191 corn. 7 pz. 2 z. 5 = -14 inf./ 1191 con. 8 pz. 1 z 4 = -13 sup., and 697 corn. 3 pz. 2 z. 4 = Arr. [34.24]/697 corn. 3 pz. 2 z. 5, and 1056 corn. 6 z. 2 = Arr. [34.25]/1056 corn. 6 z. 2 = Arr. [34.25], as in Laursen 1997, pp. 29–32. 5  Hammerstaedt;  Laursen/Masi. 6  Hammerstaedt;  Laursen. 7 ( Hammerstaedt; ’ (P Herc. 1191 col. –15 inf., 1. –2), [1/2] (P Herc. 697, 3,2,4) Laursen. 8 I agree with Laursen that . . .  seems to indicate a conditional.

Self-narratives

129

9 Taking  in connection with . 10 Hence, I disagree with Masi 2007, p. 325, according to whom we should read ‘[][ ] ’[]’ as ‘exempted from necessary causality’. 11 DRN II 265, 268 for animal mens, and 270 for animus. 12 Animal mens is certainly causally efficacious for their action, cf. the racehorses in Lucretius DRN II 263–71. 13 Consider the racehorses example again in the previous note. 14 Annas 1993, pp. 67–71. For Hermarchus Fr. 34 and Fr. 12.5–6; also cf. K. D. 32. The next two paragraphs are based on Annas’ reconstruction. 15 For epilogismos cf. Sedley 1973, pp. 27–34, and Schofield 1996. 16 Cf. K. D. 37–8 for how fundamental the notion of utility is for justice, which is – the Epicureans could have argued – the backbone of the kind of (human) societies animals do not form. 17 The successor of Hermarchus as head of the school at Athens. For his evidence, see On Irrational Contempt for Popular Opinions; in Indelli 1978. 18 On the strength of book XXV of On Nature, we both have the same kind of structured causal faculty, which needs to be in the case of the animals a structure of many various internal and external factors, with the significant difference that it lacks the kind of development humans have through their logos. 19 For Chrysippus’ quite different view see Sextus P. H. I 69. 20 It is not to deny that moral errors can also follow from using our reason and making mistakes, but it does not seem to be the point in this particular context. 21 Text P (0) =1191 -10 sup., and 697,4,1,2, and 1056,7,1 = L&S 20C(2). 22 . . . []    []          . Text from Laursen 1997,

p. 35, my translation; already quoted in Chapter 1, pp. 45–6, as part of Text P. 23 Cf. Bobzien 2000, pp. 301–2. 24 Quoted already in Chapter 1, pp. 45–6 as Text P. For the order of fragments see Appendix. 25 Text P (4). 26 To this extent I agree with O’Keefe that on its own it is not sufficient to falsify the determinist tenet; cf. O’Keefe 2002, p. 165. 27 DRN II 289–93. Also cf. Diogenes of Oinoanda, Fr. 54 in Smith 1993. 28     [][ ][] ’   []  ’  []’    ’[ ][][][]         ’   ’’  ’  [ ]  [][]    [ ]. (1191 -8 inf./1191 -7 sup./697,4,1,4/1056,7,3 = L&S 20C (9)–(11)); text from Laursen 1997, p. 39. 29 Sedley, who first added this fragment to the discussion of Epicurean agency, read [] [] at the end of the fragment – in Sedley 1983, p. 20 – and, consequently, he understood the sentence as the mind being ‘inquisitive to learn what sort of action it should then consider that one to be which we perform in some way through our own agency but without desiring to’. Sedley comments that Epicurus’ point against the determinist is that the determinist cannot characterize those actions which we are reluctant to do, but still we do in order to avoid some greater evil, actions described by Aristotle in N. E. III 1 as mixed actions – in Long

130 Self-narratives

30

31

32 33

34 35 36

37 38 39 40 41 42 43

and Sedley 1987 vol. I., p. 109. However, this reading has been superseded by Laursen and I myself find the text more meaningful based on Laursen’s reading. Thus, my analysis has revealed Epicurus as a cognitive theorist. It follows from Epicurus’ moral cognitivism that he regarded virtues as inner states, mainly consisting of beliefs or pieces of knowledge. As Tsouna has pointed out (see Tsouna 2007, p. 25), Epicurus, by referring to the cognitive components of virtues, conceived of the traditional virtues as forming a sort of unity. Epicurus’ claim that all virtues spring from prudence is a clear indicator of this fact. However, as Mitsis explained (see Mitsis 1988), this claim is compatible with both the Socratic view – according to which all the virtues are the instances of phronesis  – and the Aristotelian view – according to which all the virtues require phronesis,  but each of them has different non-cognitive roots. Sorabji’s translation from Sorabji 2006, pp. 174–5, also cf. his commentary ibid., pp. 172–7. Although the most relevant part of the text for us is in the first paragraph, it is worth quoting it at length not only for an easier comprehension of the subject but because this is one of the most beautiful texts from antiquity on the topic of self. For the Growing Argument see Sorabji 2006, pp. 83–5. See, nonetheless, Fr. (c) for memory being a product of one’s occurrent mental states to support the hypotheses that weaving for Epicurus must have included memories: ’[] [] (’[] ’ [][])  [][][][]   []. . . My translation: . . . at the same time arguing that everything receives its causal power from the previous movement and turning the argument upside-down . . . (which is precisely why it is in fact your skill that has led you to the point of stupidity). It must be said that the memories or the affects analogous to memories that accompanied them are of the products [i.e. the occurrent mental states] themselves . . . Text from Laursen 1995, p. 92. Translation from Warren 2001, p. 499. For a thorough discussion of the argument see his paper as well as Warren 2004. Cf. also Lucretius’ idea of palingenesis in the Introduction. There are only another three instances of the verb  in Epicurus’ surviving works: in the early parts of the same book of On Nature there is one in 697, 2 pz., 1 z, 1 col., 9 = fr. 4 N (Laursen 1995, p. 82), but it is not quoted because the fragment consists of only a few letters. There are two other occurrences in Ep. Hdt. 72–3 in the meaning of ‘to associate’, both instances indicating a kind of association in order to construct the concept of ‘time’; cf. Chapter 1, pp. 43–4. Fr. 18 = 1191 corn. 9 pz. 2 z. 4 = -5 inf./1191 corn. 9 pz. 1 z. 1 = -4 sup. = Arr. [34.31], and 697 corn. 4 pz. 2 z. 2, and 1056 corn. 8 z. 1, and 1191 corn. 9 pz. 1 z. 1 = -3 sup., as in Laursen 1997, pp. 43–6. In my understanding, based on that of Sedley’s, cf. Chapter 1, p. 23, and n. 93.  [] [] Sedley (cf. Masi 2006, p. 39, n. 98);  [+/– 1/2 ] [+/3/4][-] Laursen 1997. [] Sedley as indicated in Delattre and Pigeaud 2010, p. 106 and p. 1126 n. 75; []  Laursen 1997. [] Sedley as indicated in Delattre and Pigeaud 2010, p. 106 and p. 1126 n. 76; [..] Laursen 1997. Cf. Laursen’s translation in Laursen 1997, p. 55. To have an  of the dictum and understand it as a sensory recognition of the dictum sounds certainly problematic and seems to question if my understanding of  can be attributed to this Epicurean concept in all cases. I think,

Self-narratives

131

however, that  can be understood here in the same way, the claim being simply that our responsibility is based on a sensory recognition of our canon, without which we could not realize our conceptions of responsibility, since as it is spelled out in the dictum, otherwise all would be lost in relation to our investigations. 44 Cf. Everson 1994, p. 79, n. 16, as well as Atherton 2009, p. 202. 45                     ’               ’. Text and translation from Asmis 1984, pp. 20–1.

46 47 48 49 50 51 52 53

54 55 56

57

Asmis 1984, pp. 24–34. Cf. Cicero De Fin. II 6; Sextus Adv. Math.VIII 11–13; Plutarch Adv. Col. 1119F–20A. Cf. Chapter 1, pp. 34–5. And thus, as opposed to Sedley 1973, I believe Epicurus did not merely just change a name when he modified his theory by exchanging first concepts with prole p seis; cf. n. 181 in Chapter 1. Quoted on p. 21 in Chapter 1. Cf. Atherton 2009. Which does not immediately precede Fr. 18, only according to my numbering; cf. Appendix. If Epicurus is rendering the things that causally affect us by seeds in this context, then it would make perfect sense that those things are some kind of seeds of actions, thoughts or dispositions, leading to these very actions, thoughts or dispositions they are seeds of, i.e. causing them to happen by being causally responsible for them. Such an understanding would do justice to Fr. 17 in which we read that these seeds are operating factors from the very beginning. However, also cf. Sedley 1998, pp. 193–9 and Masi 2006, pp. 46–7 and pp. 196–8. Fr. 19 = 1191 corn. 9 pz. 1 z. 3 = -2 sup., and 697 corn. 4 pz. 2 z. 3, and 1056 corn. 8 z. 2 = Arr. [34.32], as in Laursen 1997, pp. 46–8 and Masi 2006, p. 38. As it is extended by Sedley in Delattre and Pigeaud 2010, p. 106. [] ] [] [] []  [][] ..[..][] Masi; [  ]   [+/–5/6] [][..].   [ ..]..[. . .]..[..][.] Laursen. I follow Masi’s text, but I translate it slightly differently. My translation depends on the French in Delattre and Pigeaud 2010, p. 106.

Bibliography Annas, J. (1993) ‘Epicurus on Agency’, in J. Brunschwig and M. Nussbaum, eds., Passions and Perceptions, Cambridge: 53–71. Asmis, E. (1984) Epicurus’ Scientific Method, Ithaca/London. Atherton, C. (2009) ‘Epicurean Philosophy of Language’, in J.Warren, ed., The Cambridge Companion to Epicureanism, Cambridge: 197–215. Bobzien, S. (2000) ‘Did Epicurus Discover the Free Will Problem?’, Oxford Studies in Ancient Philosophy 19: 287–337. Brunschwig, J. and Nussbaum, M., eds., (1993) Passions and Perceptions, Cambridge. Delattre, D. and Pigeaud, J., eds., (2010) Les Épicuriens, Paris. Everson, S. (1994) ‘Epicurus on Mind and Language’, in S. Everson, ed., Language (Companions to Ancient Thought 3), Cambridge: 74–108.

132 Self-narratives Everson, S., ed. (1994) Language, (Companions to Ancient Thought 3), Cambridge. Frede, M. and Striker, G., eds., (1996) Rationality in Greek Thought, Oxford. Hammerstaedt, J. (2003) ‘Atomismo e liberta nel XXV libro  di Epicuro’, Cronache Ercolanesi 33: 151–8. Indelli, G., ed., (1978) Polistrato: Sul disprezzo irrazionale delle opinioni popolari, Naples. Laursen, S. (1995) ‘The Early Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanesi 25: 5–109. —. (1997) ‘The Later Parts of Epicurus, On Nature, 25th Book’, Cronache Ercolanesi 27: 5–82. Long, A. A. and Sedley, D. N. (1987) The Hellenistic Philosophers, 2 vols., Cambridge. Masi, G. F. (2006) Epicuro e la filosofia della mente: Il XXV libro dell’opera “Sulla Natura”, Sankt Augustin. —. (2007) ‘Swerves and Voluntary Actions’, Rhizai 4.2: 311–28. Mitsis, P. (1988) Epicurus’ Ethical Theory:The Pleasures of Invulnerability, Ithaca/London. O’Keefe, T. (2002) ‘The Reductionist and Compatibilist Argument of Epicurus’ On Nature Book 25’, Phronesis 47: 153–86. Schofield, M. (1996) ‘Epilogismos: An Appraisal’, in M. Frede and G. Striker, eds., Rationality in Greek Thought, Oxford: 221–39. Sedley, D. N. (1973) ‘Epicurus, On Nature Book XXVIII’, Cronache Ercolanesi 3: 5–83. —. (1983) ‘Epicurus’ Refutation of Determinism’ in : Studi sull’ Epicureismo greco e latino offerti a Marcello Gigante, 2 vols., Naples: vol. I: 11–51. —. (1998) Lucretius and the Transformation of Greek Wisdom, Cambridge. Smith, M. F. (1993) Diogenes of Oinoanda:The Epicurean Inscription, Naples. Sorabji, R. (2006) Self: Ancient and Modern Insights about Individuality, Life, and Death, Oxford. Tsouna,V. (2007) The Ethics of Philodemus, Oxford. Warren, J. (2001) ‘Lucretian Palingenesis Recycled’, Classical Quarterly 51.2: 499–508. —. (2004) Facing Death: Epicurus and his Critics, Oxford. Warren, J., ed., (2009) The Cambridge Companion to Epicureanism, Cambridge.

4

Lucretius’ cosmological perspective

Introduction Do you [not] know, whoever you are, that there is actually a free movement in the atoms, which Democritus failed to discover but Epicurus brought to light, a swerving movement, as he demonstrates from evident facts? (Diog. Oin. Fr. 54) Epicurus’ discovery of the swerve refuted causal determinism at the atomic level of the theory and granted the possibility of limited indeterminism at the macro and micro levels, that is to say both in cosmology and in animal psychology as our central evidence for the atomic swerve, as Lucretius’ DRN II 216–93 attests. In this chapter we need to investigate his testimony in order to see whether it is compatible with my account of Epicurus’ conception of a causally efficacious self. First I will take into account the discrepancy – and its possible solutions – between Epicurus’ and Lucretius’ accounts of atomic collisions, in which the swerve is supposed to fulfil one of its basic roles according to the latter. Next I shall examine how Lucretius’ arguments for the swerve fit into the larger context of his discussion of atomic motion. Afterwards I need to follow a well-trodden path: the scrutiny of the arguments themselves, which is necessary for any new interpretation of Epicurean psychology. DRN II 251–93 has been the subject of great attention and is a passage often studied to comprehend how the swerve is supposed to preserve our voluntas. One of the aims of my analysis will be to point out the connections between Lucretius’ arguments and Epicurus’ conception of animal and human behaviour and the operations of their respective causal faculties. But the investigation will also consider an emendation vis à vis its manuscript reading, and Epicurus’ concerns with causal determinism and fatalism. The discussion of the latter follows naturally from Lucretius’ account and it is relevant to the larger framework of Epicurus’ conception of selfhood, since if the events of the world were causally determined or fated, it would be incompatible with the kind of freedom and possibility of moral reform Epicurus credited humans with based on his ideas of the causal capacities of one’s self.

134 Lucretius’ cosmological perspective

The cosmogonical crux One of the necessary conditions of motion for atomists is void.1 Nonetheless, void is not a sufficient condition for motion: there needs to be a further reason for motion to start. If we attribute weight to the atoms2 then our imaginary pool gets going since void is unable to lend the atoms support, and if we then postulate infinite past collisions of atoms, we can account for their generative functions and take a rest, just what atoms can never do. Why bother then to introduce a third kind of atomic motion, the swerve? Lucretius offers us three reasons in DRN II 216–93: (1) a cosmogonical one: the swerve allows there to be collisions of atoms and, hence, generation of things; (2) an anti-determinist one: the swerve prevents there being a complete causal nexus; and (3) an anti-fatalist one: the swerve breaks the tyranny of fate.3 The first seems to be redundant on the strength of Epicurus’ idea of infinite past collisions (see Ep. Hdt. 44), which would do perfectly well for cosmogony. In fact, if the swerve is meant to start a series of atomic collisions, as the standard understanding of Lucretius’ cosmogonical argument takes it, then Lucretius’ argument is incompatible with Epicurus’ explanation, since it contradicts the assumption of an infinite series of past collisions, this latter conception going back to the early atomists. There are two kinds of solutions offered to do away with the tension. Epitome variations David Sedley thinks that the Letter to Herodotus, an epitome of Epicurus’ physics, was a comparatively early work, and possibly only a partial epitome of the first thirteen books of Epicurus’ On Nature, which consisted of thirty-seven books. His thesis runs that, at the time of its composition, Epicurus was content to give an economical account of atomic collisions and it was his subsequent adoption of the swerve in his moral theory, which made him think of incorporating the swerve within his presentation of atomic collisions.4 One objection is that it seems indeed obtuse to replace an economical solution with a radical one, incorporating the swerve.5 Elsewhere Sedley argues more convincingly for the conformity of the Letter with the first thirteen books of On Nature.6 Still, I think the introduction of the Letter tells against its restriction to being an epitome of only the early part of On Nature, so let us first examine this objection before introducing an alternative solution to the above tension in our text of Lucretius. We learn from the first sentence of the Letter to Herodotus that Epicurus had already prepared an epitome of his physical theory for those who are unable to study his philosophy in detail and only occasionally take up the study of nature. The epitome of the whole  included the summary of the principal doctrines and served to instruct its audience on the most important points, most likely written in a manner accessible for a wider audience as well. It is tempting to identify this writing with the work known to us as the  , however, the only evidence for this connection is the recurrent reference

Lucretius’ cosmological perspective

135

of the scholiast to the  in the Letter to Herodotus, a fact which leaves this link merely speculative.7 The extant epitome addressed to Herodotus is for advanced students only. Epicurus emphasizes that it is important as well for advanced students to memorize the outlines of his treatises and in 37 he concludes: since such a method is useful to all who are accustomed to investigate into nature, I, who keep recommending the continuous activity of investigation into nature, and live such a calm life, have prepared for you just such an epitome and elementary exposition of the complete doctrines. (Ep. Hdt. 37)8 It is clear that Herodotus belongs to the group of those who take on the investigation into nature seriously and are used to such a kind of enquiry and not those who do not engage in the close examination of Epicurus’ larger books, most likely referring to the books of the  . And it seems also that the current epitome Epicurus made for Herodotus (  Ep. Hdt. 37) – and indeed for all the advanced students – is different from the one Epicurus prepared for a different kind of audience ( Ep. Hdt. 35), since otherwise it would not seem to make much sense for Epicurus to say explicitly that he made such an epitome for Herodotus, instead of saying that Herodotus should himself turn to studying the epitome written for that other audience or stating that this very epitome is indeed good for both of them.9 Thus, if we make a distinction between the two different summaries, the first – possibly the  – written for inexperienced students, and the Letter addressed to Herodotus and written for advanced students, then the difficult style of the Letter can be explained – at least to a certain extent – since one could say that he was writing for those who had already familiarized themselves with his detailed works and understood his presentation style. Epicurus’ conception of the different functional roles of his epitomes, explicit in the introduction, strengthens the distinction between the different audiences of the different summaries. While for inexperienced students it is beneficial to study the epitome prepared for them so they can ‘assist themselves on the most important points on every occasion’, for advanced students – who have made significant progress in looking upon () nature – the continuous memorizing of the outline () of the whole  reduced to its elements is required until the following fruitful outcomes are obtained: (a) ‘the most decisive attention () to things’, and (b) ‘the means of discovering all the exact details’. In fact, the virtue of the advanced student is that he or she can turn his or her attention quickly to the basic principles. This kind of excellence of the student is already pointing towards the next section of the Letter, where Epicurus discusses his principles of methodology; that is to say, an excellent student does well if he takes on the investigation into nature according to the Epicurean rules of methodology.

136 Lucretius’ cosmological perspective

Furthermore, while in principle it is not impossible to think that the very same epitome can have different functions for different audiences, at the end of the Letter Epicurus reaffirms the usefulness of memorizing this very epitome for those who are sufficiently or even perfectly acquainted with his detailed expositions, because it helps them to clear up the points discussed in the long treatises – just as the reduction of the longer works to such elementary apprehensions of the main points helps their researches () into nature. While there are those who are not being perfected by these (  (Ep. Hdt. 83)) – which I take to be Epicurus’ writings10 – the beginners are advised to consult the main ‘headings’ () in a silent fashion and as quickly as possible for their peace of mind, which seems to have been a lesser kind of intellectual occupation and without the urgency to apply the main headings as a means for advanced research; this is in contrast to the advanced students’ epitome, which supports such an in-depth investigation. Not to mention that it is hard to believe that Epicurus seriously meant that this particular epitome brings tranquillity to anyone who does not know his detailed works. Thus, I find it untenable that the letter written to Herodotus is only a partial summary, and, consequently, I think we need to find a different explanation for resolving the apparent textual and philosophical tension between Epicurus and Lucretius on atomic motion. An arche of collisions O’Keefe offers an alternative solution, which aims to settle the tension between Epicurus’ and Lucretius’ different cosmogonical explanations.11 He suggests that instead of taking the swerve as a starting point for atomic collisions in Lucretius, as according to the standard interpretation, we should understand its cosmogonic function as an explanatory  of collisions. On the strength of the commonly accepted claim that Epicurus was influenced by Aristotle’s criticism of Democritus’ theory of atomic motion, O’Keefe speculates that Epicurus initially introduced the swerve as a response to that criticism. O’Keefe’s account squares well with the evidence, it is charitable and, briefly, it unfolds as follows.12 Aristotle’s main complaint is that all atomic motions are the result of external force according to Democritus, since they are the outcomes of previous atomic blows, and consequently he gives no account of natural atomic motion.13 On Aristotle’s assumption, in order for there to be motion, bodies must have some sort of natural motion, but if all motion is forced as the atomists maintained, then motion is impossible. As O’Keefe points out, Aristotle does not seek an explanation of any particular motion, but rather he demands to get an explanation of why there should be motion at all: But it is a wrong assumption to suppose universally that we have an adequate first principle () in virtue of the fact that something always is so or always happens so. Thus Democritus reduces the causes that explain nature to the fact that things happened in the past in the same way as they

Lucretius’ cosmological perspective

137

happen now; but he does not think fit to seek for a principle () to explain this ‘always’: so, while his theory is right in so far as it is applied to certain individual cases, he is wrong in making it of universal application. (Phys. VIII 252a32-b2)14 In my view, O’Keefe convincingly argues that Epicurus, as a follower of Democritus, living in Athens shortly after Aristotle, introduced the swerve as a response to Aristotle’s criticism. Epicurus accepted that the infinite past collisions of atoms, a postulate he took over from Democritus, would not suffice for explaining why there are collisions at all, and why there are different types of motion. If the natural motion of atoms came about due to their weight they would be falling downwards, just like drops of rain. The properties of atoms – solidity, weight, shape and size – do not account for collisions; however, with the addition of the swerve Epicurus could appeal to a natural feature of atomic motion, which elucidates the reason why atoms collide. O’Keefe argues: ‘Of course, this explanation may be ad hoc and unsatisfactory, but it does give a reason why there are collisions that appeals to the properties of the atoms, instead of leaving the existence of the collisions “dangling” without an explanation’.15 Moreover, this interpretation fits the Lucretian evidence very neatly: If they were not accustomed to turn aside, all would fall downwards, like drops of rain, through the deep void, and neither would a collision occur, nor a blow be produced among the primary bodies: in this way nature would have never produced anything. (DRN II 221–4)16 As O’Keefe says, on the traditional reading the following conclusion of the argument that ‘the atoms swerve’ is too strong and should read instead ‘the atoms did swerve’. Thus, if we are to maintain Lucretius’ inference that ‘the atoms swerve’, we need to conceive the swerve as an explanatory  of collisions instead of as a beginning or temporal starting point. I think O’Keefe’s solution, though necessarily speculative, is attractive and it also lends great support to my thesis.

On atomic motion In my view, Epicurus’ primary concern in introducing the swerve was to allow for a cosmology, which is open to positive ethical outcomes. In his philosophy, where ethics clings so much to physics – since the main barriers to attaining tranquillity of mind are our unnecessary fears and desires, and the only way to eliminate them is to study physics17 – Epicurus was not committed to giving an explanation of everything in physical terms, however much we may expect an atomist to do so. Instead, I believe, he preferred to offer a physics which catered for related ethical concepts, one of which was in Lucretius’ words the libera voluntas.

138 Lucretius’ cosmological perspective

According to the standard understanding of the ethical function of the swerve, Epicurus introduced it to tackle the following problem: since his atomist physics only allowed physical interactions, on the strength of the inherited Democritean tenet, all interactions at the basic level of description were collisions of atoms. If everything in the cosmos is reducible to atomic motions and new atomic motion arises out of old in a fixed order then everything is determined at the atomic level and, inevitably, at the phenomenal level of our world. Consequently, all actions can be excused from praise and blame. To avoid this conclusion, Epicurus opted for the introduction of an undetermined atomic swerve into his physics and made a direct connection between the swerve and our voluntas, thus granting the possibility of moral responsibility in his ethics. However – as I have argued in Chapter 2 – this view can be challenged on the strength of some of the surviving papyrus fragments of On Nature XXV. I find it convincing that Epicurus did not conceive of moral responsibility in terms of causal indeterminacy of an agent’s actions, swerves being directly involved in the generation of each and every voluntas, but rather held that the human agent’s present overall dispositions are responsible for his or her deeds. Epicurus explicitly says in the fragments of On Nature XXV that the development of our own character is not physically predetermined, that is, our character does not turn on our atomic constitutions, but it ‘depends on us’ (’ ). Consequently, our initial make-up and the accidental necessity of that which surrounds and penetrates us are not alone the causes for our actions. If we take into account Epicurus’ conception of the self, agency need not be explained by reference to atoms or, strictly speaking, atoms and their properties, and although they are the sole constituents of a person, the complex concept of a moral agent cannot be explained by them. If so, then what are we to do with Lucretius’ ethical argument for the atomic swerve? To answer this question it is essential to recapitulate the original emphatic structure of Lucretius’ discussion of atomic motion in DRN II 62–332, instead of merely looking at the discussion of the swerve in isolation. In order to see how Lucretius arrives at the ethical function of the swerve, first I will look at his general strategy in the preceding lines, and I will argue subsequently for my thesis that the atomic swerve was intended to provide a cosmology accommodating, in Lucretius’ conception, the libera voluntas. The movements of Natura The Syllabus (62–6): (a) What motion is responsible for generation? (b) By what force are atoms compelled to generate things? (c) What speed do atoms have? Don Fowler suggests that (a) and (b) should be taken together, and are answered by 80–141, followed by the section on atomic speed in 142–66.Thus, the remaining

Lucretius’ cosmological perspective

139

lines of 184–332 are ‘essential corollaries to the basic doctrines.This explains the disparate subject matter of those parts and gives a plausible motivation for the position of the anti-providential argument [viz. 167–83]’.18 I think this is a credible reading and not only goes well with the evidence we find in the Letter to Herodotus, but with the further assumption that Lucretius’ different sources included both Epicurus’ On Nature – or parts thereof – as well as the Letters, possibly along with others.19 It also clarifies why there is no trace of the swerve in the Letter. If there had been an epitome of Lucretius’ account of atomic motion, possibly only the body of the argument – on Fowler’s reconstruction 67–141 – would have been given. The Letter, as an epitome, was only meant to reiterate the essence and the fact that it gives an ‘economical’ account of atomic motion fits nicely with the absence of further reasons for it. Thus, the question of whether the Letter was an early work20 or rather a later one, since it is an epitome,21 loses its relevance for the question of why the swerve does not figure in it. As Fowler points out, the partitio and enumeratio of the subject matter of the first section is not only a rhetorical technique but a promise of knowledge and power.22 What power shall we possess after understanding this section on atomic motion? To formulate the same question differently: what sort of practical or ethical benefits will the study of physics produce?23 We find Lucretius’ explicit answer in lines 167–83. Lucretius, by arguing against providence, instantiates a clear case where the knowledge of physics advances our understanding of the whole nature of the universe and, indirectly, the understanding of divinities in general. Although it is not explicit here, the thought runs that if the world is not run by divinities and there is no such thing as , then why should we be troubled about fearing the gods, since they do not interrupt our life? Fowler thinks the passage in isolation makes poor sense, but if we suppose that in the lengthy lacuna preceding lines 165–66 Lucretius had already given a substantial anti-theological point – presumably giving an account of the role of atomic motions in creating a cosmos – the passage becomes less opaque.24 The text reads as follows: But some in opposition to these [facts] are ignorant of matter and they believe that without gods’ consent/power nature is not able, with such regular conformity to human plans, to change the seasons of the year and produce crops and everything else besides, which divine pleasure, leader of life, persuades mortals to approach, herself (i.e. voluptas) leading them, pleasure who charms them through the acts of Venus to reproduce their generations, so the human race would not become extinct.When they imagine the gods to have created everything for the sake of men, they are seen to be slipping far away from true reasoning in every way. However much I might be ignorant of what the primary elements of things are, I would venture to confirm from the actual arrangements of the sky, as well as to exhibit from many other facts that the nature of the cosmos has by no means been made

140 Lucretius’ cosmological perspective

for us by a divinity; such are the faults [of things]. All this, Memmius, I shall clarify later; now I will explain what remains to be said about the motion [of atoms].25 Let us look first at what the different possible readings of ignari materiai in line 167 imply. It can be understood in several ways:26 (1) Ignorance of the causes of phenomena resulting in disturbance and belief in the gods.27 (2) Ignorance of the power of matter: Lucretius reverses the standard objection of the opponents, namely that the atomists draw too much attention to matter, and charges his opponents with being ignorant and, thus, lacking knowledge of the power of matter. (3) Ignorance of the nature of matter: knowledge of atomic motion reveals the real nature of matter and makes providence impossible. Misconception of the nature of matter is misconception of the nature of the cosmos itself and, hence, the cause of superstitions and fear. The conflicting views are further dramatized by naturam in line 168 with its semi-personification as if nature was quarrelling with the gods for mastery of the cosmos. Natura as a provider takes on an odd role: it conforms to human plans and actions. I think the phrasing of line 169 is strange given the atomists’ non-anthropocentric (cf. DRN V 156–234), non-teleological view of the cosmos.28 As it turns out, the natural order of the cosmos is not perfect, hence, admoderate is restrictive and it is rather a poetic association of divine governance. Wonderment about natural order and regularity was a standard argument for the existence of divinities, but Lucretius twists this reasoning and denies gods’ existence by pointing out the imperfect characteristics of the cosmos. The text ex ipsis caeli rationibus in line 178 is to be taken as a general reference point, including both astronomical and meteorological phenomena. Unlike the Peripatetics, who distinguished between the upper heavens and the sublunary world, and could still argue for the existence of divine providence by pointing to the orderly motions of the heavens, the Epicureans did not make this distinction. Consequently, they could claim that not only is the sublunary world imperfect but that the imperfection also applies to the whole of the cosmos,29 which as an obvious result cannot be run by divinities. Understanding the true nature of things is a powerful therapeutic means to eliminate our fears and superstitions. The comprehension of atomic motion in particular and its generative functions as it is outlined (DRN II 80–166) grants the foundation of a non-anthropocentric, anti-providential cosmos (DRN II 167–83). Given the ascending structure of the whole poem,30 I doubt that Lucretius in the preceding lacuna was prepared to carry his exposition as far as to present a fully armoured argument for the case at this stage.31 It seems plausible that what we see here is a typical Lucretian technique: in his procedure of listing the basic nature and function of entities, he exemplifies those functions on

Lucretius’ cosmological perspective

141

all stages of the ascent.That is, he is keen to illustrate the consequences of atomic motion to give further strength to his arguments, and to prepare his audience for what is to come. Strictly speaking, he delineates a physics which caters for ethical concepts or for ethically relevant issues, such as the anti-providential cosmos. In his illustrations, Lucretius often draws on observation, which was an accepted Epicurean method of demonstrating certain facts in a certain case, in these lines the imperfect character of the cosmos.32 It was often used by Lucretius as part of a modus tollens, frequently started with a negative proposition: ~p → q ~q (from direct observation) ~~p p This kind of argument is of significant importance in the ‘essential corollaries’ in lines 184–332, and in particular when establishing the atomic swerve, to which I now turn.33

DRN II 251–93 Reconstructing the argument The first section of Lucretius’ argument for the swerve in connection with libera voluntas (251–60) is crucial for our understanding of how voluntas, causal determinism, fatalism and the swerve are connected. This section is normally understood to contain the first premise of Lucretius’ argument, running from 251–93 – yet again another modus tollens – and this section is normally summed up in the following way: if the atoms did not swerve, there would be no libera voluntas.34 On the strength of this premise, it is easy to make a direct connection between the voluntas and the atomic swerve by identifying volitions with undetermined swerves, as the traditional interpretation does, at the price of facing the problem of how an undetermined swerve could account for volition at all.35 Nonetheless, on the basis of the first section of Lucretius’ ethical argument for the swerve, I think two slightly different modus tollens arguments can be sketched to illustrate how the swerve, the question of causal determinism, fatalism and voluntas are connected. Denique si semper motu conectitur omnis / et vetere exoritur novus ordine certo / nec declinando faciunt primordia motus /principium quoddam, quod fati foedera rumpat, /ex infinito ne causam causa sequatur, / libera per terras unde haec animantibus exstat, / unde est haec, inquam, fatis avolsa voluntas, / per quam progredimur quo ducit quemque voluptas, / declinamus item motus nec tempore certo / nec regione loci certa, sed ubi ipsa tulit mens? (251–60)

142 Lucretius’ cosmological perspective

(A) Moreover, if all motion is always linked, and new motion arises out of old in a fixed order, and atoms do not by swerving make some beginning of motion to break the decrees of fate, so that cause should not follow cause from infinity, from where does this free volition exist for animals throughout the world? (B) From where, I ask, comes this volition wrested away from the fates, through which we proceed wherever each of us is led by his pleasure, and likewise swerve off our motions at no fixed time or fixed region of space, but wherever the mind itself carries us?36 (A) 1

If the atoms did not swerve, there would be a complete causal nexus and there would be no libera voluntas. 2 There is libera voluntas. (from direct observation) 3 Therefore, it is not the case that the atoms do not swerve, as a consequence of which there is a complete causal nexus. 3′ Therefore, the atoms swerve, as a consequence of which there is no complete causal nexus. (B) 1 2 3

If the world was ruled by fate, we could not wrest volition away from fate. But we can proceed according to our pleasure. (from direct observation) Therefore, the world is not ruled by fate.

Although the aim of the whole argument (251–93), undoubtedly, is to establish the atomic swerve, and the conclusions of (A) and (B) are not explicitly stated in their entire scope, the first section of the argument provides the first premise of (A) and the first and second premises of (B) explicitly, and the structure of the whole argument (251–93) allows for this reconstruction. Consequently, by implying two modus tollens arguments Lucretius seems to be linking his argumentation directly to the problem of causal determinism and fatalism. While (A) points out that if we are to have voluntas, there cannot be a complete causal nexus at the atomic level of description, and to fulfil this condition there needs to be an atomic swerve, (B) draws attention to the relation between fate and voluntas, and the requirement for non-fatalism. This distinction points towards Cicero’s understanding of Epicurus’ taking causal determinism and fatalism (logical determinism) to be inter-entailing. That is to say, Lucretius seems to have hinted at Epicurus’ concern with the Truth-to-Necessity arguments, though he certainly did not make his understanding of fatum explicit .37 I think (A) lends direct support to my argument in Chapter 2, according to which Epicurus introduced the atomic swerve to substantiate his type-dualist thesis by denying that every token of the mental is completely determined by the physical. Otherwise, if the physical were sufficient for a complete causal explanation, talking about the causal efficacy of the mental would imply over-determination, and, hence, it would be redundant. As we shall see,

Lucretius’ cosmological perspective

143

Lucretius uses in his examples the accepted Epicurean method of  , and he takes the voluntas for granted, just as he thinks that the causality of the mental makes sense only on the strength of an indeterministic physics. That is to say, the condition for making sense of voluntas in an undetermined cosmos, and the condition for such an undetermined cosmos is the atomic swerve. Thus, in the first premise the swerve is accounted for as the necessary atomic motion to break a complete causal nexus. As Susanne Bobzien has observed,38 there is a great similarity between, on the one hand, Lucretius’ contrast between our mind being the cause of our motions through volitions and fate in the form of causal determinism pre-determining our motions, and, on the other hand, Epicurus’ criticism of the earlier atomists in the fragments of book XXV. Just as Epicurus distinguishes the causality of the self from the causality of the atoms, Lucretius attributes a causal capacity to the mind in the form of volitions and explores the conditions of the atomic level for this capacity to exist, making an inference from the existence of the voluntas to the existence of the clinamen, without ever identifying the two. However much this interpretation is in line with the texts and provides a strong philosophical point, the language of the passage proves to be good ground for formulating objections. Seemingly, the most serious worry is the two instances of unde in lines 256 and 257, because they seem to suggest that something at the phenomenal level of our world comes from something at the atomic level, introducing a bottom-up causal relation. Hence, it makes the reading of the traditional interpretation look natural, suggesting that libera and fatis avulsa voluntas are somehow to be identified with the swerve or, at least, that the swerve needs to play an essential part either in the formation or in the execution of the voluntas.39 The fact that Lucretius’ language has this bottom-up causal-relational tendency also suggests that libera and fatis avulsa voluntas are merely the by-products of undetermined atomic motions, opening the gate for reductionist readings. But such readings are the result of losing sight of the general direction of Epicurus’ atomist project. As Robert Hankinson puts it, ‘the challenge of the Epicurean physics and cosmology is to be able to produce a convincing account of the complexity of the cosmos, and crucially of its apparent order, on the basis of such a limited range of explanatory concepts’;40 such an explanation is just what is going on in the section on atomic motion, by taking the phenomena of voluntary motions for granted and explaining the framework precondition of the atomic level. Since, as Sedley pointed out,41 Epicurus regarded mental events and states as accidental properties and Epicurus was emphatic in taking accidental properties to be just what they appear to be,42 in the case of our mind’s voluntas being causally efficacious, if Lucretius was to remain in line with his master, he must have also conceived this potestas as irreducibly different from atomic states. Therefore, the bottom-up causal relation suggested by the poetical questions starting with unde can be simply taken to mean ‘from where else comes this potestas of libera voluntas for animals throughout the world if there is a fixed causal chain, if not from the fact that this innate capacity (nobis innata potestas, line 286) can only make sense on the strength of

144 Lucretius’ cosmological perspective

an indeterministic physics’, according to which the mental cannot even have a complete physical determination. Lucretius leaves us in no doubt in the following two lines that he is working with the notion of the causal primacy of the voluntas: nam dubio procul his rebus sua cuique voluntas principium dat et hinc motus per membra rigantur. (261–2) For without doubt it is volition that gives these things (i.e. our movements43) their beginnings for each of us, and it is from volition that motions are spread through the limbs. Lucretius embarks in the second section (261–83) of the argument (251–93) on establishing the second premise that volition exists and that it is the beginning of animal motions. As Bobzien has shown in her analysis of the passage, we can understand the beginning of motion in an absolute or in a relative sense. The former means that what produces motion is not itself entirely causally determined; it is not the outcome of a fixed causal nexus as it were but there is a gap in the causal chain, a spontaneous movement, which counts as a fresh beginning, while, on the other hand, the relative beginning of motion means that, given certain external and internal circumstances, the animal will initiate motion in correspondence with the animal’s actual cognitive disposition. The proponents of the traditional interpretation argue for the absolute-beginning understanding and insert the swerve into Lucretius’ explanation of voluntas as a silent assumption, being the cause and the beginning for voluntary animal motion. Against that view, Bobzien’s whole-person model of agency interpretation argues that Lucretius establishes the phenomena of volitions as the relative beginning of voluntary movements, which is in accord with the acting agent’s current cognitive disposition. When Lucretius contrasts volitional and forced motion in animals, we can see that we are not expected to read this as an absolute beginning of motion. Volitional motions have an internal beginning based on the agent’s desire according to Lucretius – the example of racehorses is presented – while forced motions are initiated from outside without the agent’s desire – being pushed in a crowd for example.44 In both cases, there is an explicit time delay between the rise of the desire to do something and its execution or the volition aligning the movement of the body with the mind’s desire, which are the obvious signs of the existence of volition for Lucretius and entirely compatible with the interpretation of the relative beginning of motion. Example 1 nonne vides etiam patefactis tempore puncto / carceribus non posse tamen prorumpere equorum / vim cupidam tam de subito quam mens avet ipsa? / omnis enim totum per corpus materiai /copia conciri debet, concita per

Lucretius’ cosmological perspective

145

artus / omnis ut studium mentis conixa sequatur; / ut videas initum motus a corde creari / ex animique voluntate id procedere primum, / inde dari porro per totum corpus et artus. (263–71) (Example 1) Don’t you see also how when at an instant the starting gates are opened the eager strength of horses can nevertheless not surge forward as suddenly as the mind wishes? For all the mass of matter has to be stirred up throughout the body, so that stirred up through all the limbs it may in a concerted effort follow the mind’s desire. Thus you may see that the beginning of motion is created from the heart and proceeds initially from the mind’s volition, and from there is spread further through the entire body and limbs. In agreement with the whole-person model of agency, I would like to propose an even more radical reading of what is going on. Lucretius is demonstrating the mind’s independent causal efficacy by drawing attention in his first example to the way in which the animal’s desire is capable of affecting the body, setting it in motion, which in light of my non-reductionist understanding of Epicurus certainly needs to mean downward causation, though not the kind of downward causation Sedley envisages; the mental causing the atoms to swerve and, hence, enforcing its efficacy on the physical.The swerve is not mentioned in the examples, since the examples are meant to establish the phenomenon of volition, only on the strength of which can Lucretius carry on concluding the existence of the atomic swerve. The downward causation in these examples is simply exhausted by the voluntas, an accidental property of the mind, causally influencing the physical. As we read in Fr.13 of On Nature XXV, what is produced (i.e. the causally efficacious mental state), once it gets the cause from itself, immediately gives its causality to the atoms – which I understand as the mental causally operating on the physical, hence the downward causation. The physical is not a sufficient cause for the mental and the physical constitution of the mind at T2 – a point which in general is secured by the atomic swerve, and consequently, does not need to be mentioned in Fr. 13. Similarly, in these lines of DRN, Lucretius presents a corresponding line of thought in his first example. He claims that the delayed reaction of the horses is a sign of their minds’ desire stirring up the mass of matter, which is clearly a case of the mental being causally operative on the physical mass of the body, expressed in the form of an entirely top-down relation. It is the mind’s voluntas, which brings the body into motion; the action moves first from the mind’s volition ‘and then is passed on in turn’45 through the entire body and limbs. I do not wish to claim that ‘dari porro’ is a direct translation of  in Fr. 13 of book XXV, thus, I do not say that Lucretius is looking at and translating ‘the original’ of that fragment in particular, since the two lines of thought do not exhibit a perfect match, yet the similarity of their verbal expression is striking. In Fr. 13 we read that   [], that is, that

146 Lucretius’ cosmological perspective

the product or the causally efficacious mental state in turn/then immediately gives or distributes its causality to the atoms and somehow makes the whole of the cause one. As I argued in Chapter 2, ‘the whole of the cause’ is to be understood as the sum of the different factors bringing about a causally efficacious mental state and, accordingly, when the causality of the mental is distributed to the atoms, the cause is made into one with the atomic aspects of all the further different factors which are part of a volition. In Lucretius’ phrasing, this is what the volition spreading through the entire body and limbs seems to be because the body obeys the mind’s desire. The distribution of the causality of the mind makes the body move in accordance with one’s voluntas, so in this sense the independent causality of the mind is passed onto the atoms of the mind, starting a chain reaction in the atomic body. I think the second example is also meant to demonstrate the same point in a different situation, a difference which is important for Lucretius to set up his analogy between the phenomenal level and the three atomic motions in his conclusion, thus, establishing his argument. Example 2 nec similest ut cum inpulsi procedimus ictu / viribus alterius magnis magnoque coactu; / nam tum materiem totius corporis omnem / perspicuumst nobis invitis ire rapique, / donec eam refrenavit per membra voluntas. (272–6) (Example 2) Nor is it the same when we move forward impelled by a blow, through another person’s great strength and great coercion. For then it is plain that all the matter of the whole body moves and is driven against our wish, until volition has reined it back throughout the limbs. The causal efficacy over matter is also explicit in this example, the voluntas reining back the movement of the body like a rider or charioteer; as noted by Fowler, the imagery relates both to the first example and to Plato’s charioteer of the soul in the Phaedrus (246 A ff.).46 The allusion continues in the next seven lines, but what is most important for our purposes is the clear description of the voluntas mastering the body. iamne vides igitur, quamquam vis extera multos /pellat et invitos cogat procedere saepe / praecipitesque rapi, tamen esse in pectore nostro /quiddam quod contra pugnare obstareque possit? / cuius ad arbitrium quoque copia materiai / cogitur inter dum flecti per membra per artus / et proiecta refrenatur retroque residit. (277–83) So do you now see that, although external force propels many along and often obliges them to proceed against their wishes and to be driven headlong,

Lucretius’ cosmological perspective

147

nevertheless there is something in our chest capable of fighting and resisting, at whose decision the mass of matter is also forced at times to be turned throughout the limbs and frame, and, when hurled forward, is reined back and settles down? Lucretius takes the nature of voluntas to have been established in 277–83, and he explicitly connects it with the second example. I think he does this in order to set up a pair of opposite motions, forced motions versus voluntary motions against some force – the latter being a further variation of the voluntas of the horses, which is not used against some force – and he is going to use this pair of opposites in the analogy he draws between the phenomenal and atomic levels, on the latter level the pair of opposite motions of the atoms being motions as a result of impact and motions as a result of the swerve. In my understanding, the tripartite parallel works by matching natural atomic motion due to weight with example 1, in which the voluntas is executed naturally, without any impediment. The connection seems to be that just as atoms fall downwards due to their weight when not colliding with other atoms, although it takes some time for the horses to execute their voluntas after the gates have been opened, they also burst forth in a natural direction, towards the goal of the hippodrome, or at least they do so naturally as racehorses, without being hindered any longer by some external factor.The connection works even better as a simile if we consider the fact that it must take some time for the atoms after taking part in collisions to gain their natural motions back. The bond between atomic motions as the outcome of impacts and forced motions as in example 2 is straightforward. It may seem tempting then to make an exclusive affinity between the capacity of executing one’s voluntas against some force and the atomic swerve, and to exclude the swerves from easily or naturally enacted volitions, so that we have a perfect match between the two levels.47 Nonetheless, I think the distinction between the two examples may be taken differently as well, with the further advantage of strengthening my understanding of the distinction between animal and human voluntas as explored in Chapter 3.48 The voluntary motions of the racehorses seem to be natural reactions qua racehorses, incorporating instinctive stimuli to run forth in response to their perception of a gate opening. Their reaction does not involve any reasoning about their situation and their , their complex causal faculty, is in harmony with their constitution, or as Epicurus put it, at least according to the way we view them their causality is woven into one with their constitution.49 In contrast the second example of a man being pushed by a crowd is a description of a case in which one presumably needs to take into account one’s situation, instead of just acting according to a quasi-automatic or conditioned desire.Thus, I do not think that in the second example the emphasis is so much on how we can fight against some force with the help of our volitions, but rather that we do not act simply under some external or internal force, since we have this potestas in our chest which is capable of influencing

148 Lucretius’ cosmological perspective

what sort of actions we actually perform.50 Consequently, Lucretius seems to be operating with different levels of rationality of voluntas between animals and humans; this squares well with the evidence, though slight, we have of the Epicurean conception of animals. From the testimonies of Hermarchus and Polystratus, it becomes fairly clear that animals lack the so-called higher rational faculties and, since without epilogismos it is impossible to attain the kind of eudaimonia humans can achieve, they cannot have a good life in the way humans can.51 The lack of epilogismos means animals only have one kind of voluntas, which is either an instinctive reaction or is based on some conditioning, as certainly is the case for racehorses. So Lucretius’ point is not to show that voluntas exercised against some force is to involve the swerve exclusively, but in his examples he demonstrates two different kinds of voluntas by the particular actions of the racehorses and the man pushed by a crowd; in the latter case, the man certainly uses his epilogismos to consider the appropriate course of action to take. Very importantly, there is no doubt that when animals are put in cages against their voluntas, for example, they exhibit behaviour which would convincingly show that their voluntas is applied against some force, i.e. being forced into the cage. But according to the earlier distinction, it is to be understood as an instinctive and natural reaction against violence or, in an Epicurean fashion, they naturally try to avoid the pain they are suffering; and conversely, human beings can exercise their voluntas on many occasions when they are not exercising it against some force, but nonetheless, they can obviously form their voluntary motions in such cases as well in accordance with their rational calculations. When they do not use epilogismos, it is then that Epicurus considers them to behave similarly to animals, as the analysis of Fr. 16 has shown.52 The conclusion of the argument Having established the nature of voluntas by setting up three types of motions apparent on the phenomenal level in his examples – voluntary motions without any impediment, forced motions and voluntary motions against some force – Lucretius concludes the section by drawing a parallel between the phenomenal and the atomic levels to justify the theoretical assumption of the atomic swerve. quare in seminibus quoque idem fateare necessest, / esse aliam praeter plagas et pondera causam / motibus, unde haec est nobis innata potestas, / de nihilo quoniam fieri nihil posse videmus. / pondus enim prohibet ne plagis omnia fiant / externa quasi vi; sed ne mens ipsa necessum / intestinum habeat cunctis in rebus agendis / et devicta quasi cogatur ferre patique, / id facit exiguum clinamen principiorum / nec regione loci certa nec tempore certo. (284–93) (Conclusion) Therefore in the seeds too you must admit the same thing, that there is another cause of motion besides impacts and weight, from

Lucretius’ cosmological perspective

149

which this power is born in us, since we see that nothing can come into being out of nothing. For weight prevents all things from coming about by impacts, by a sort of external force. But that the mind should not itself possess an internal necessity in all its behaviour, and be overcome and, as it were, forced to suffer and to be acted upon – that is brought about by a tiny swerve of atoms at no fixed region of space or fixed time. Quare in line 284 begins the conclusion that, just as in the case of volition – which must be a new beginning of motion as has been demonstrated – there must be a third kind of motion at the atomic level of description as well, since according to the Epicurean first principle of conservation ‘nothing can come into being out of nothing’, and if there are new beginnings at the phenomenal level, then we have to admit that there are new beginnings at the atomic level too.That is, the existence of the swerve is argued for in a usual Epicurean manner, from something evident towards something non-evident. It is easy to see why it has been always so tempting to connect volitions with swerves on the strength of this conclusion. Nonetheless, Lucretius is not securing the existence of voluntas by the swerve, rather he is setting up an analogy between the phenomenal and the atomic levels, which is intended to secure the existence of the swerve with the help of the first principle of conservation – according to which nothing comes out of nothing.This analogy provides room for the atomic swerve without the intention of making the swerve the necessary condition for each and every volition. Undoubtedly, the conclusion is to establish the swerve as a necessary condition for voluntas, but the sort of condition Lucretius took it to be is only stated explicitly at the foundation of the first premise of the whole argument. As it became clear there, the atomic swerve breaks the complete nexus of causal determinism, hence providing a cosmos built on undetermined atomic motions, providing room for his conception of genuine libera voluntas. As we witness the existence of new beginnings of motion at the phenomenal level, there must be such new beginnings of motions at the atomic level as well, argues Lucretius, and the atomic swerve is a necessary condition to break the causal chain resulting from natural and forced motions from time to time – without being the necessary condition for each and every new beginning of motion at the phenomenal level – a causal chain which would be fully determined if there were no swerving motions in the atoms. The correspondence I have drawn between the examples given for the three kinds of motions at the phenomenal level and the three kinds of atomic motions also indicates that a swerve is not to be appealed to in the case of each and every volition. If it is correct to connect the racehorses example with the natural (downwards) motions of atoms, the motion of being pushed by the crowd with the forced motions of the atoms resulting from impacts, and the voluntary motions against some force with the swerving motions of atoms, then it becomes very clear that the swerve is not intended to be a necessary condition for each and every volition. The racehorses example registers a new beginning

150 Lucretius’ cosmological perspective

of motion and the execution of animal voluntas in particular, without Lucretius explicating any direct connection between this type of voluntas against an impediment and the swerve. Admittedly, we cannot take the analogy at face value.53 The swerve has to be taken as a condition for any kind of animal voluntas if we are to stay in harmony with the evidence and the above considerations about the voluntas – hence we should understand the difference between animal and human voluntas as the outcome of the dissimilar animal and human nature. Nevertheless, we cannot disregard the analogy either because it forms a central part of the argument by drawing a parallel between the phenomenal and the atomic levels. The analogy helps to establish the hypothesis of the swerve by pointing out that, since human voluntas can be practised against some force – because it is not determined in all its movements – there has to be a condition for such a kind of movement at the atomic level as well, which is the swerve, a kind of motion against some force (i.e. against the deterministic causal chain of atomic motion). Lucretius is not merely demonstrating by his examples that volition exists but also that volition is a new beginning of motion. Only when the existence of voluntas is established does Lucretius conclude with confidence his argument for the swerve, where he does not directly link the atomic swerve with volition but merely says that if there is a fresh beginning at the phenomenal level – since ‘nothing comes out of nothing’ – there must be a fresh beginning at the atomic level as well. I think we need to understand the swerve as the necessary condition of an undetermined physics, which is also the necessary condition for justifying the phenomena of the causal efficacy of the mind – the existence of which Lucretius has demonstrated by his examples – so that the mental is not completely determined by the atomic motions. And we can also take the poetical analogy as a means in the philosophical argument for the atomic swerve. It lends support for the theoretical assumption of the atomic swerve on the strength of the first principle of conservation, but it does not grant any further function for the swerve than for it to be a necessary condition for there to be voluntas, which the swerve provides for by breaking the causal nexus of causal determinism. Res vs. mens Importantly, my interpretation works regardless of whether we read mens or res in line 289;54 nonetheless, I think there are good reasons to retain the manuscript reading res ipsa, meaning ‘the atom itself ’.55 In my view, the only major objection not answered properly in the literature so far has been the difficulty of making sense of the necessum intestinum of the atom and, therefore, I limit my discussion to this point.56 I think it makes very good sense to understand the internal necessity of an atom as the determinant of atomic motion due to its weight resulting in natural downward falling unless checked by another atom or group of atoms.57 This understanding has the clear advantage of keeping the tripartite analogy until the very end of the argument, and Lucretius is simply reinforcing in the last

Lucretius’ cosmological perspective

151

sentence of his conclusion what he had already said in the first: we must admit another cause of atomic motion besides impacts and weight which prevents all things (i.e. all atomic motions) from coming about by impacts by re-establishing natural motion from time to time: but that the atom should not itself possess an internal necessity [i.e. due to its weight] in all its behaviour [i.e. in all its movements due to its weight], and be overcome and, as it were, forced to suffer and to be acted upon [forced motion due to collisions] – that is brought about by a tiny swerve of atoms at no fixed region of space or fixed time. This understanding leads to the seemingly odd result that natural motions take part in the deterministic causal chain of atomic motion. Nonetheless, if we take a close look at the first premise (251–60) of Lucretius’ argument (251–93), we do not find Lucretius describing the causal chain of atomic motion as ‘new motion arising out of old in a fixed order’ exclusively as a result of atomic collisions, nor does causam causa sequatur in line 255 imply that we are to understand the causal nexus of atomic motion as determined entirely by atomic collisions. As the first sentence of the conclusion (284–93) explicitly takes weight into consideration as one of the causal factors for atomic motion besides impacts and the atomic swerve – which is meant to break the fixed order of the atomic causal nexus – we can insert natural atomic motion due to weight into the determined causal chain of atomic motion with certainty. It means that natural and forced motions are equally part of the ‘decrees of fate’, in light of which describing the atom as having an internal necessity, i.e. initiating downward natural motion due to its weight from time to time – weight preventing ‘all things (i.e. motions)’ from coming about by impacts – sounds less strange. Weight is a necessitating factor in the sense of making the atoms fall downwards sooner or later depending on the frequency of their collisions as well as on their weight.58 Undoubtedly, the speed of their falling is not determined by their weight (cf. Ep. Hdt. 62, DRN II 225–42), but on the strength of Ep. Hdt. 61 Epicurus certainly attributed a crucial role to weight in how long it may take them to gain their natural motion back after a collision, provided that they are not hurled in some other direction by another collision, and in this sense weight can be conceived as an internal necessitating factor. The other consideration for retaining the manuscript reading of res points in the direction of taking the weight of the atoms as their internal necessitating factor. When considering the function of the random atomic swerve, Maurice Pope recognizes that it must be a permanently operative cause because, if it were not, the influence of weight would reassert itself and eventually all the atoms would revert to their natural downward fall, resulting in the atomic rain described by Lucretius – a point already noticed by Knut Kleve, of which Pope gives the following reformulation: ‘the innumerable swerves have the function of preventing the atoms from remaining in the natural downward motion to which they will all inevitably return when the concilia they form part of disintegrate’.59

152 Lucretius’ cosmological perspective

Pope observes that without swerves, atoms would obey the principle of weight, already described by Cicero as an internal and necessitating cause,60 which would lead to the total disintegration of everything.61 Accordingly, the last sentence of Lucretius’ conclusion (284–93) simply states that for atomic motion not to be exhaustively determined by atomic weight and collisions – which would lead to causal determinism – we need to postulate a tiny swerve of the atoms, which results in the incompleteness of the causal explanation by the atomic; that is to say, it leads to the kind of indeterminism Epicurus sought in order to justify the causal efficacy of the mental witnessed at the phenomenal level.

(B) – Causal determinism and fatalism Lines 257–60 of book II indicate that the whole argument (251–93) implies another conclusion beyond that of (A): (B) 1 2 3

If the world were ruled by fate, we could not wrest volition away from fate. But we can proceed according to our pleasure. (from direct observation) Therefore, the world is not fatalistic.

As I have already observed, (B) points towards Cicero’s representation of Epicurus taking logical and causal determinism as inter-entailing in De Fato 17–20, according to which he held that to avoid the yoke of necessity we need to deny both, the universal applicability of future truth and a complete causal nexus, by the introduction of the atomic swerve. Cicero complicates the issue by portraying Epicurus’ worry concerning logical determinism at some places to be about the Principle of Bivalence (T[p] v F[p], where T stands for true, F for false and p for proposition), and at others by representing Epicurus to be concerned with the Law of the Excluded Middle (p v ~p). As Bobzien has argued, in De Fato 19 Cicero probably recast Epicurus’ argument in the form of the Principle of Bivalence in order to mirror the Stoic argument for determinism. What she thinks the Epicureans in fact denied was the Law of the Excluded Middle,62 and she offers the following tentative reconstruction of the original argument based on De Fato 21: (P1) Of every contradictory pair either one or the other is true now. (P2) If one of every contradictory pair is true now, it is also certain now. (P3) If it is certain now, it is also necessary now. Therefore, of every contradictory pair one is necessary now. Epicurus is thought to have found this argument threatening because if whatever happens – including future events – is already necessary now, then the

Lucretius’ cosmological perspective

153

argument would confirm both necessity and fate. As Bobzien observed, although the argument starts out with propositions, that is, with a semantic version of the Law of the Excluded Middle, there is a switch from propositions to events in the argument to substantiate a deterministic presupposition. However, since the logical argument itself is not about causation and there is also no mention of causation in the argument, she holds that the connection between the truth value of a future proposition and the corresponding event is merely an assumption of a general correspondence between propositions and events, and not a connection between propositions and causation, which she supports with textual evidence.63 She takes the argument to be merely an ontological statement about the fixedness of the future, Epicurus fearing the consequences of the Truth-to-Necessity argument independently of ‘any particular type of determinism’.64 Bobzien’s claim that in the original context Epicurus was arguing against the Law of the Excluded Middle has only minor philosophical implications,65 so it is her second claim, that Epicurus’ argument did not take into account physical causation or fate at all, which requires more attention. O’Keefe has argued, against Bobzien, that we have no reason to cast doubt on Cicero’s representation of Epicurus’ conception of logical and causal determinism as inter-entailing, which he finds supported by the evidence, as well as by the communis opinio, according to which Epicurus took causal and logical determinism to be two sides of the same coin, based on the assumption that what is true in advance is determined by pre-existing causes.66 Although O’Keefe agrees that Epicurus wished to eliminate causal determinism, he thinks that Epicurus’ ultimate object was to refute fatalism, in his understanding the thesis being that the fixity of the future renders us helpless, and decisions and deliberations pointless, as a consequence of which exercising reason becomes irrational.67 O’Keefe interprets the atomic swerve as having been introduced by Epicurus to counter the threat to the efficacy of reason presented by the thesis that the future is fixed regardless of what we do. Nonetheless, as Catherine Atherton has noted – I think correctly – O’Keefe’s understanding of the fatalism to which Epicurus objected appears to be an ambiguous thesis between causes and explanations. ‘O’Keefe’s “fatalism∗” corresponds, not to logical determinism, but to the sort of determinism that would render us helpless (20.3) by denying “that our decisions make any difference.”’68 To see why it is credible to think that both Epicurus and the Stoics took logical and causal determinism as inter-entailing, I find it best to follow Bob Sharples’ presentation of how Cicero portrayed the relation between the Epicurean and the Stoic argument.69 My primary aim is to show how Lucretius’ (B) connects to Cicero’s testimony and how it matches my interpretation of the swerve. In 17–18, Cicero argues that future truth and causation are separate issues. The fact that future events will happen or the way they will happen undoubtedly cannot be changed in the sense that what will happen will happen; however, it does not mean that they are already determined by physical causes. In other words, the things referred to in future truth are not already determined

154 Lucretius’ cosmological perspective

by physical causes and so, according to Cicero’s criticism, Epicurus had no reason to fear fate and was wrong to introduce the swerve to avoid the deterministic implications of future truth, thinking – as presented in 21 – that if every proposition is true or false, then everything will be fated. Importantly, what we learn from 17–20 is that Cicero presents Epicurus as refuting fate by the denial of the universal applicability of future truth secured by the swerve, an uncaused movement. As Sharples observed, Cicero appears to portray Epicurus in a paradoxical form, Epicurus refuting causal determinism, the major worry for human freedom, only by accident: Epicurus was primarily worried about the Principle of Bivalence and it was only his mistaken interentailment thesis which led him to the refutation of causal determinism.70 As opposed to Epicurus, Chrysippus was afraid that if not every proposition is true or false, not everything will be fated and there will be uncaused movement. Therefore, he accepted the Principle of Bivalence, and concluded that there is no uncaused movement as a consequence of which everything happens by pre-determined causes, thus, everything is fated.71 Now Cicero denies the views which both Epicurus and Chrysippus share that: (a) if there is no uncaused movement, everything is fated; and (b) if not everything is fated, not every proposition is true or false, and instead Cicero argues for the compossibility of: (i) every proposition being true or false; (ii) there being no uncaused motion; and (iii) not everything being fated; the denial of (a) and the compossibility thesis being credited to Carneades.72 As Sharples has argued, though it is true that in 21 Cicero makes no explicit reference to causation, his preceding representation of Epicurus indicates that fate has a causal force on the strength of Epicurus’ inter-entailment thesis, as a consequence of which Epicurus’ argument indeed mirrors Chrysippus’ view of the connection between future truth, causation and fate. As Sharples elaborates: that Cicero treats the Epicurean and the Stoic positions as symmetrical in this way is significant. For the wider context in Cicero shows that for him at least the question of future truth was itself connected not only with that of causation generally but specifically with that of the freedom of our actions from determination by internal antecedent causes.73 Sharples argues that Cicero’s presentation of Carneades in 23 ff. – rejecting (a) that if there is no uncaused movement, everything is fated and asserting instead the compossibility of (ii) there being no uncaused motion, and (iii) not everything is fated – was represented as Carneades’ improvement of Epicurus’ position. From this, Sharples concludes that Cicero took Epicurus’ worry – that if every proposition is true or false, then everything will be fated, and if there is no uncaused movement, everything will be fated – and Chrysippus’ position – that the Principle of Bivalence implies that there is no uncaused movement, and, therefore, everything is fated – as well as Carneades’ compossibility thesis, as being different positions of the same debate. Consequently, it follows that if the proposition of ‘not everything is fated’ of Carneades’ compossibility thesis

Lucretius’ cosmological perspective

155

involves freedom from being determined by internal antecedent causes, then Cicero very likely took Epicurus’ denial of fate to be concerned with the same kind of freedom. Sharples substantiates his point concerning Carneades’ position by noting that although the contrast between external causes and the nature of the thing itself in De Fato 25 may suggest that only the question of not being determined by external causes is discussed as Carneades’ idea, nonetheless, if it were so, Carneades’ thesis would collapse into that of Chrysippus. And that this is not what Carneades means to suggest is shown by his referring to the absence of determination by antecedent causes, as well as by external ones; the causes within us that primarily determine our actions for Chrysippus may not be external, but they are still antecedent, for the whole Stoic picture of fate as a series or network of causes depends on causes succeeding one another in time, and it is this that Carneades wants to challenge.74 Consequently, the freedom from determination by antecedent internal causes was already an issue at least for Cicero in the first century BCE, but also very likely for Carneades himself.75 O’Keefe goes on without much hesitation to attribute the libertarian problem to Carneades, whom he takes to be directly influenced by Epicurus, since Cicero presents Carneades’ position as a modification of Epicurus’ view.76 Let me consider Carneades’ criticism/position in relation to Epicurus’ theory in the light of two key passages of De Fato, which, I believe, will shed more light on Epicurus’ theory: Epicurus introduced this theory [i.e. the atomic swerve] because he was afraid that, if the atom was always carried along by its weight in a natural and necessary way, we would have no freedom, since our mind would be moved in the way which it was constrained by the movement of the atoms. Democritus, the inventor of the atoms, preferred to accept this, that all things come about through fate, rather than remove the natural movements of individual bodies from them. [XI] More acutely, Carneades taught that the Epicureans could have maintained their position without this fictitious swerve. For, seeing that (Epicurus) taught that there could be some voluntary movement of the mind, it would have been better to defend that than to introduce the swerve, especially as they cannot find a cause for it. And by defending this they could easily have resisted Chrysippus. For in having admitted that there was no movement without a cause, they would not be admitting that all things that came about did so through antecedent causes. For (they could have said), there are no external and antecedent causes of our will. (De Fat. 23–4)77 If all things come about through antecedent causes, all things come about in such a way that they are joined and woven together by a natural connection.

156 Lucretius’ cosmological perspective

But if that is so, all things are brought about by necessity; and if that is true, nothing is in our power. However, there is something in our power. But if all things come about through fate, all things come about antecedent causes. So it is not the case that whatever comes about does so through fate. (De Fat. 31)78 These two texts clearly show that Carneades thought of causal determinism as incompatible with the voluntary movement of the mind, which capacity of the mind is probably also designated by in nostra potestate in 31. The voluntary movement of the mind is most likely to be understood as the power either to perform an action or not to perform it, so what is at stake for Carneades is one’s ability to do otherwise than one does. He takes it that this power must come about without antecedent causes, because he assumes the causal independence of the mind itself – hence the attribution of the ‘traditional’ libertarian position to Carneades. If we are to take Carneades as establishing the ‘traditional’ libertarian view, then the freedom of the will from external and internal antecedent causes is to be understood as a specific faculty of the soul, which is responsible for deciding what to do, causally not determined by one’s cognitive disposition. That is to say, on this reading of the freedom of the will, the mind has an independent decision making faculty.79 If this interpretation is the correct one, then how are we to take Carneades’ criticism of Epicurus in 23? As I have argued, Epicurus introduced the swerve as the necessary condition for causal indeterminism, which, among other things, substantiates the claim that it makes sense to talk about the causal efficacy of our mind, which would not be acceptable if all the powers of the mind could be explained by atomic motions as a result of weight and collisions. Epicurus did not take swerves and volitions to be identical, but by the introduction of the swerve he hoped to avoid the position that everything is to be determined by atomic motion alone, i.e. that everything has a sufficient atomic cause. Carneades criticized Epicurus’ theory – in the light of my interpretation of the function of the swerve – by pointing out that Epicurus did not need to posit the swerving motion of atoms in order to justify this causality of the mind, but if the ‘traditional’ libertarian view of Carneades is correct, he thought that this efficacy could be granted by the postulation of an independent causal faculty of the mind, independent even of one’s own cognitive disposition in the sense that one has the possibility to do otherwise. The difference between the two positions lies then in the kind of causal faculties, since in Epicurus’ conception of the causal power of the self, it was a structured outcome of one’s cognitive disposition, a result of external and internal factors alike, which was obviously not the kind of Carneadean voluntas free of antecedent and external causes. This difference springs, I assume, from their way of challenging causal determinism. Carneades agreed with Epicurus that causal determinism cannot be right, since we see that there is something in our power. However, as opposed to Epicurus, Carneades did not accept that there can be uncaused motion and, therefore – for whatever dialectical purposes – he

Lucretius’ cosmological perspective

157

defended the voluntary motion of the mind by postulating a kind of causal faculty of the mind, which is independent of external and internal antecedent causes. I think it is important to see that Epicurus could not have given such an answer due to his theoretical assumptions, and it is hard to imagine that he would have found Carneades’ suggestion appealing, given the general desire of atomism to explain a wide range of phenomena by an economical number of explanatory means. According to my interpretation of the function of the atomic swerve, it solves a number of problems: it is not only an  of collisions, but Epicurus introduced it to challenge – on the strength of the interentailment thesis – causal and logical determinism, and to establish the causal efficacy of the mind by denying that everything has a sufficient cause based on atomic motion. Lucretius’ (A) and (B) can then be best understood as reflecting Epicurus’ idea of causal and logical determinism being inter-entailing. Undoubtedly, Lucretius does not introduce the discussion of the Principle of Bivalence or the Law of the Excluded Middle in (B). However, he hints at Epicurus’ view on the relation between causal and logical determinism, and if my reconstruction of Lucretius’ argument (251–93) implying these two different modus tollens arguments is correct, then Lucretius, in fact, refutes both, causal and logical determinism, (A) concluding that the atoms swerve and therefore there is no complete causal nexus and (B) providing the non-fatalist consequence.

Conclusion At the end of the prologue of book II, Lucretius’ aim is set: to dispel the terror and gloom of the mind by the view and by the law of nature (DRN II 59–61, also cf. DRN I 1090–1104). As he mixes the medicine for us,80 he moistens the side of the cup with the sweet notion of libera voluntas, so that we can swallow the notion of the swerve without any fatalistic consequences. Once healed, we understand how our voluntas as a new beginning of motion is possible as we witness it to happen, and as the foedera naturae provide. Significantly, it is not the same as saying that we understand how to account for our voluntas by giving its underlying atomic pattern. The analysis of the emphatic structure of the section on atomic motion has shown that Lucretius’ intent was certainly not to engage in all kinds of departments of Epicurus’ philosophy at once but to exemplify the consequences of his explanations, to explain the evident by the non-evident and vice versa, all in the service of the therapy of the animus.The theory of atomic motion teaches us on the strength of the foedera naturae why we should not fear divinities and why it does not imply fatalism. As we read in the Letter to Menoeceus,81 Epicurus would have preferred the former rather than the latter, but he did not accept either of them. The swerve as an  of collisions, by and large in the universe, is one of the underlying natural characteristics of atomic motions. As a new conceptual

158 Lucretius’ cosmological perspective

device in the Epicurean toolkit, it was primarily meant to mend his ancestors’ theory of atomic motion and, as such, it was introduced on cosmogonical grounds. Nonetheless, as the principle of new beginnings in motion, it also solves the problem of causal and logical determinism, and by providing that not everything has a complete physical cause it caters for libera voluntas – the new beginning of voluntary motion at the phenomenal level. It is hard not to notice though that if the first argument of section DRN II 216–93 is a response to Aristotle, then the second argument may have some traces of Aristotle’s influence as well. Aristotle in Physics II 1 declares that those things which have nature all have within themselves their principle of motion and rest. He restricts his claim in Physics VIII 4 by saying that things with souls have a principle of motion and rest in a stronger sense than any other natural bodies. Natural bodies without souls do not have a source of causing movement and action, but they can only be acted upon. The atomic swerve as an internal cause goes along smoothly with the Aristotelian requirement but immediately trembles at the point of restriction, since atoms obviously possess no souls.Thus, to force an interpretation of a direct nexus between swerves and volitions seems to be seriously missing the point. On that interpretation – here we are again – swerves would initiate all volitions, and not only would their given randomness make moral responsibility entirely inexplicable but it would also seem as if it was the swerve which would decide when I should have a new volition. When I am about to submit my manuscript of this book, I could just suddenly start playing the guitar and neglect my plan because of an incalculable atomic swerve. Taken together with my interpretation of Epicurus’ On Nature XXV Lucretius’ evidence poses a challenge to the standard view of the function of the swerve. Just as Epicurus reinforces his position against his opponent’s deterministic point of view by the evidence of the preconception of our own agency’s responsibility for our actions and our behaviour, Lucretius, in II 251–93, takes what seems evident for granted, that is, animals have libera voluntas. Having established libera voluntas as a new beginning of motion, he holds that it is a necessary and a sufficient condition for arguing against fatalism. However, given the locus of his argument, his primary concern is to show the underlying atomic framework. Lucretius seems to have taken the swerve as an internal condition for libera voluntas but, on the other hand, he did not seek to show how the latter supervenes on the former. It is the animus, not the atoms, which is responsible for carrying us where we desire nec regione loci nec tempore certo. I think Lucretius’ evidence not only provides a compatible larger framework theory to my reconstruction of the Epicurean self but it also reinforces it on many particular points in harmony with the relevant papyrological evidence, as we have seen during the course of the analysis. More particularly, we have found that Lucretius’ argument (A) lends direct support to my interpretation of Epicurus’ causal type dualism, according to which although everything in the universe is physical – token monism – there are two types of genuine causation: physical and mental. Consequently argument (A) also lends support to my thesis according to which Epicurus distinguished between the causal capacity of the self

Lucretius’ cosmological perspective

159

and that of the atoms. Lucretius’ characterization of the voluntas as a causally primary power also substantiates my reconstruction of the Epicurean self. His Example 1 confirms the idea of downward causation, according to which the mental can causally influence the physical and Lucretius’ language has striking similarities with Epicurus’ verbal expression in Fr. 13 of On Nature XXV, most likely concerning the same subject. Although no direct textual connection can be established, the correspondence between Epicurus’ and Lucretius’ concerns is evident. And last but not least the distinctions concerning animal and human voluntas in Examples 1 and 2 also confirm my analysis of their different causal capacities and the narrative idea the Epicureans had of their selves, as discussed in Chapter 3.

Notes 1 On how to understand the development of the concept of void in the history of atomism cf. Sedley 1982. 2 For the seemingly ridiculous view of atoms not having weight cf. Furley 1983, repr. in Furley 1989. 3 I will support the view, according to which Epicurus took causal and logical determinism as mutually entailing, therefore, (3) will collapse into (2), cf. pp. 152–7. 4 Sedley 1983, pp. 13–14. 5 O’Keefe 2005, pp. 115–16. Sedley claims, however, that once the swerve was established on other grounds, viz. to help account for psychological autonomy, it was no longer radical to try making it do some work elsewhere too, viz. in cosmogony, whereas to have introduced it in the first place in order to do that other work would have been radical; cf. section titled ‘An arche  of collisions’. 6 Sedley 1998, pp. 109–33. 7 We encounter the title  three times in the scholia to Ep. Hdt. in 39, 40 and 73. 8                      . My translation.

9 This point only stands on the strength of the acceptance of the textual emendations made by Usener, who emended  to  ; cf. Usener 1887, p. 4. 10 I base this understanding on the manuscript reading of W, lines 11–12 and not on its emendation. 11 O’Keefe 1996 and 2005, ch. 5. 12 Note, however, that not everyone is satisfied by this explanation; see Atherton 2007, pp. 197–8. 13 The atomists’ theory of atomic motion faced heavy criticism from Aristotle. He argued that there can be no such thing as void (Phys. IV 211b 20–9), that their explanation of atomic speed is untenable (Phys. IV 216a 11–21) and that there can be no direction of motion of atoms moving in the void (Phys. IV 215a 20–2). 14  ’   ·   ’    . Translated by Hardie and Gaye, in Barnes 1984. 15 O’Keefe 2005, p. 119.

160 Lucretius’ cosmological perspective 16 quod nisi declinare solerent, omnia deorsum / imbris uti guttae caderent per inane profundum / nec foret offensus natus nec plaga creata / principiis: ita nihil umquam natura creasset (221–4). Translation from O’Keefe 2005, p. 120: the italics in line 223 are mine. Like Rouse and Smith 1975 and Smith 2001, and as I do, he translates line 223 ‘nec foret offensus natus nec plaga creata’ as a present unreal conditional (foret = esset). Long and Sedley, by contrast, translate it as a past unreal condition, that is:‘collisions and impacts . . . would have not arisen’.This solution explicitly entertains the idea that there was a single temporal starting-point of the atomic collisions, thus shifting the evidence towards the traditional understanding of the passage; Long and Sedley 1987, vol. I, p. 49. 17 ‘A man cannot dispel his fear about the most important matters if he does not know what is the nature of the universe but suspects the truth of some mythical story. So that without natural science it is not possible to attain our pleasures unalloyed’ (K. D. 12). Translation from Bailey 1926, p. 97. 18 Fowler 2002, pp. 144–6. 19 Obviously many different combinations may have been possible (cf. Sedley 1998), and although we cannot trace back with certainty, it is likely that Lucretius’ emphasis on the basic tenets matched that of his sources. 20 Sedley 1983. 21 O’Keefe 2005, p. 118, does not explain the lack of the swerve in the Letter. The lack of the swerve from the basic principles has puzzled scholars, even the scholium to Ep. Hdt. 43–4, which refers to the twelve  of Epicurus, does not list it among them; cf. Clay 1973. 22 Fowler 2002, pp. 146–7. 23 At this stage I will only answer this question considering lines 62–183. 24 For the same speculation about the content of the lacuna, cf. Smith 2001, p. 39. I disagree with this intuition of the content, since we get a rather lengthy description of world formation in book V 91–508, and in particular in DRN V 195–234, but of course it does not necessarily exclude the possibility of Lucretius discussing world formation already here. 25 At quidam contra haec, ignari materiai, / naturam non posse deum sine numine credunt / tanto opere humanis rationibus admoderate / tempora mutare annorum frugesque creare, (170) / et iam cetera, mortalis quae suadet adire / ipsaque deducit dux vitae dia voluptas / et res per Veneris blanditur saecla propagent, / ne genus occidat humanum, quorum omnia causa / constituisse deos cum fingunt, omnibu’ rebus (175) / magno opere a vera lapsi ratione videntur. / nam quamvis rerum ignorem primordia quae sint, / hoc tamen ex ipsis caeli rationibus ausim / confirmare aliisque ex rebus reddere multis, / nequaquam nobis divinitus esse creatam (180) / naturam mundi: tanta stat praedita culpa. / quae tibi posterius, Memmi, faciemus aperta. / nunc id quod superest de motibus expediemus. (DRN II 167–183) 26 I am indebted to Fowler’s work, which has improved my understanding of the passage and some of the points I make below are based on the ones he made in Fowler 2002. 27 ‘. . . their ignorance of the causes obliges them to attribute everything to the government of the gods and to admit their sovereignty’ (DRN VI 54–5). Translation from Smith 2001, 179–80. 28 Strange but very fascinating at the same time.This really long one sentence of 167–76 unfolds nicely, I think, if we recognize that this conformity of nature to human plans is one and the same as the conformity of human nature to itself. 29 Cf. DRN VI 1090–1137 where the plague comes from the caelum; and DRN V 91–109, the caelum crashing down. 30 ‘The sequence is one of ascending scale: the first pair of books deals with the microscopic world of atoms, the second with human beings, the third with the cosmos as a

Lucretius’ cosmological perspective

31 32 33

34 35

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

161

whole. Within each pair of books, the first explains the basic nature of the entity or entities in question, the second goes on to examine a range of individual phenomena associated with them. A further symmetry lies in the theme of mortality, treated by the odd-numbered books. Book I stresses from the outset the indestructibility of the basic elements, while books III and V in pointed contrast give matching prominence to the perishability and transience of, respectively, the soul and the cosmos,’ Sedley 2004. Lucretius explicitly states his intention to clarify this matter later. Also cf. n. 24. ; cf. DRN II 67–79, where observations prove that the world is not a plenum. A linear reconstruction of the cosmogonical argument will suffice here because a detailed discussion of it need not detain us. On this understanding, see the section titled ‘The cosmogonical crux’ on pp. 134–7. (1) 216–220 Atoms when falling down decline slightly at undetermined times and places. 221–224 If they were not accustomed to this, they would be falling down in straight lines and no atomic compounds would ever have been formed. (2) 225–229 It is wrong to think that heavier atoms fall more quickly and strike the lighter ones and generate compounds. 230–237 Because what falls in the air or water, falls in proportion to its weight, because of the resistance of the air or water; however, nothing resists the downward motion of atoms in the void. 238–239 quapropter, all atoms move at the same speed. 240–242 igitur, the heavier atoms cannot catch up with the lighter ones and form compounds. 243–244 quare, we must suppose the existence of the clinamen. (3) 244–245 The clinamen is no more than minimal, to avoid oblique motion which is contrary to facts, 246–248 since we see that bodies cannot move obliquely when they are falling. 249–250 But nobody can say that they do not move aside with a minimum when falling. Cf., e.g., Purinton 1999, p. 265. The standard objection against the ‘traditional interpretation’, which was first formulated by Furley (cf. Furley 1967) among the modern commentators. For the traditional view see Bailey 1928, pp. 838–42, and 1947, pp. 318–23 for the standard formulation; Purinton 1999, for a more modern recapitulation of the idea and with some further variations of the basic idea Asmis 1990 and Fowler 1983. Translation of lines 251–93 is from Long and Sedley 1987, vol. I, pp. 105–6. I discuss Epicurus’ view of the inter-entailment thesis on pp. 152–7. Bobzien 2000, p. 310. Cf. Purinton 1999, p. 257 and p. 267. On p. 288 using the point against Sedley as a ‘philological’ objection. Hankinson 1999, p. 499. Sedley 1988, p. 324. Ep. Hdt. 71. Bobzien 2000, p. 313. As Smith has noted, in his second example Lucretius probably had in mind the audience at horse races. Smith 2001, pp. 116–7 note (a), originally in Smith 1966, pp. 76–7. Fowler 2002, p. 357. Fowler 2002, p. 359. The direction Saunders turned to, cf. Saunders 1984. See the section ‘Animal and human causal capacities’ in Chapter 3 of this book. See the section ‘The narrative self ’ in Chapter 3 of this book. As Bobzien has pointed out haec . . . potestas in line 286 can only have an implicit referent in lines 279–83; cf. Bobzien 2000, p. 315.

162 Lucretius’ cosmological perspective 51 Annas 1993a, pp. 67–9. 52 It is a question whether Epicurus would have allowed exceptions in cases when e.g. something is done by routine. According to Annas’ discussion of Aristotle’s conception of virtues (Annas 1993b, pp. 49–52) Aristotle would not have accepted the modern objection to his view that since virtue is a stable disposition it is something passive, because he kept emphasizing that virtue is a state which involves choice ( ), which cannot be passive; cf. N. E. II 1106b 36–1107a 2. 53 And that allows for one to connect the swerve as a new beginning of motion at the atomic level to each and every new beginning of voluntary motion at the phenomenal level. If we were to take these connections at face value, then clearly Saunders’ interpretation – according to which swerves are only included when we execute some action forcefully – would be the correct one. However, if we were to make such a direct connection between the third case, the correspondence between voluntary motion against some force and the atomic swerve, we should then expect Lucretius’ analogy to work in all cases, which would lead to the odd result that voluntary motions without any impediment necessarily stem from naturally moving atoms, or forced motions are the direct outcomes of some atomic impacts. 54 Since Epicurus did not restrict the occurrence of the swerve to the mind, I do not think it would have made much sense for Lucretius to switch suddenly to the topic of the philosophy of mind in his conclusion, but even if he did and he wrote mens in DRN II, line 289, my argument still stands. At the very beginning of (A) in DRN II 251, denique has the same force, as at the very beginning of the cosmogonical argument in DRN II 216, where Lucretius states his anxiety that Memmius should conceive a further point on the general principles of atomic motion, and then he presents reasons why the swerve is required in the formation of a cosmos. In (A) there is more to be said on the general principles of atomic motion, to postulate another reason why we need the swerving motion of atoms. The swerve is not restricted to breaking the causal nexus of liberta voluntas, but it functions to break the complete causal nexus of the universe, so that there can be libera voluntas at all. If we take mens in line 289 to be the correct reading, when in his conclusion Lucretius asserts that there must be a swerving motion at the atomic level, so that the mind does not suffer from internal necessity, once again, he is talking about a framework, which is pre-conditional and caters for the mind to be free from internal necessity and not restrictively about the framework of atomic motion in the philosophy of mind. It would indeed be strange if the latter was the case, since he continues his discussion about the permanent nature of atomic motion in the whole universe in the following section (DRN II 294–307), adding on a different argument still concerning the same discussion of atomic motion. 55 This understanding of res ipsa was suggested first by Furley 1967, p. 179. 56 Furley made five objections (Furley 1967, pp. 179–80) against reading res: (1) the oddity of the phrase, although he admits straight away that it is a possible one. (2) The stylistic shortcomings of lines 289–92, because in the translation we would read ‘that the atom itself has no internal necessity . . . is the work of a tiny swerve of the atoms’. Nonetheless, I accept Avotins’ point (Avotins 1979, p. 98), according to which the Latin text does not carry this problem. (3) cunctis in rebus agendis, an odd phrase to use in talking of an atom. I think it is hardly a decisive objection, nonetheless also made by Fowler 2002, p. 327. For my interpretation see pp. 149–51. (4) Talking about the freedom of atoms from internal necessity does not match up with anything at the phenomenal level, where Lucretius is only discussing animal freedom from external force. I believe my interpretation of the Epicurean theory makes clear why I disagree with this point. (5) In what sense are we to talk about the ‘internal necessity’ of an atom? This is what I discuss in the main text.

Lucretius’ cosmological perspective

163

Furthermore, I share Pancheri’s view (Pancheri 1974, p. 53) that the reading of mens would indicate shifting back and forth between the atomic and the phenomenal levels, which to me is unappealing not for the reason Pancheri gives that it is not worthy of Epicurus’ style, but rather because Lucretius’ conclusion focuses on the atomic level, drawing a conclusion which accommodates the phenomenon of the voluntas. 57 However strange it may sound first to describe natural motion due to atomic weight as an internal necessity. 58 Cf. Ep. Hdt. 61: Nor will their upward or their lateral motion, which is due to collisions, nor again their downward motion, due to the weight, affect their velocity. As long as either motion obtains, it must continue, quick as the speed of thought, provided there is no obstruction, whether due to external collision or to the atom’s own weight counteracting the force of the blow. Translation from Hicks 1931, p. 593; my italics. 59 Pope 1986, n. 57, p. 96; Kleve 1980, p. 28.The point is a clear and explicit antecedent of O’Keefe’s interpretation of the swerve as an  of collisions. Pope goes even as far as taking the swerves to be ‘miniature prime movers’, taking the Epicurean swerve to be ‘a kind of identical twin to the Aristotelian prime mover’, Pope 1986, pp. 88–9. 60 Cicero De Fat. 11. 23, 11. 24 and 11. 25; also in DND I 69. 61 Notwithstanding, he does not argue for the manuscript reading of res on these grounds, and as a matter of fact he does not argue for it at all but just ridicules Lambinus’ emendation by saying that it is like altering ‘word’ to ‘sword’ at the beginning of St John’s gospel. Furthermore, I think Pope’s point offers a satisfactory solution to Atherton’s complaint against O’Keefe (in Atherton 2007, pp. 197–8), according to which O’Keefe’s defence of the cosmogonic function of the swerve does not justify the assumption ‘that Epicurus would have felt philosophically obliged to provide a general, principled explanation of atomic movements of the sort Aristotle found lacking in Democritus’. 62 Cf. De Fat. 37, Acad. II 97 – although here Cicero describes Epicurus mentioning that there is no necessity in nature, which suggests that even though Cicero is talking about the Law of the Excluded Middle, his testimony does not evidently support Bobzien’s view (Bobzien 1998a) – and DND I 70. 63 Cf. previous endnote. 64 Bobzien 1998a, p. 81. 65 As noted by Sharples (2007, p. 54), which is also recognized by Bobzien once she admits that the abolition of the Principle of Bivalence follows in any case; Bobzien 1998a, p. 83. 66 Cf. Long and Sedley 1987, vol. I, p. 112. 67 O’Keefe 2005, p.17, p. 20, p. 150. 68 Atherton 2007, p. 194. 69 In my reconstruction I shall rely on Sharples 1991 and Sharples 2007. 70 Sharples 2007, p. 58. 71 De Fat. 20–1. 72 This reconstruction is based on that of Sharples 2007, pp. 58–62. 73 Sharples 2007, pp. 60–1. 74 Sharples 1991, p. 177. 75 Sharples 2007, p. 62. 76 O’Keefe 2005, pp. 153–62. Although undoubtedly Carneades’ ‘position’ seems to be ‘influenced’ by Epicurus, I am not so much convinced that we should take

164 Lucretius’ cosmological perspective

77

78

79 80 81

Carneades’ stance in the debate as a position instead of being just a criticism, if we consider him to be the head of the sceptical current of the Academy. I am not entirely convinced either that ‘Carneades should be credited as the first person to come up with a libertarian position vis-à-vis the “traditional” problem’ (O’Keefe 2005, p. 154), but a discussion of this question would go beyond the scope of this book. It remains, nonetheless, important to discuss Carneades’ criticism of Epicurus. Hanc Epicurus rationem induxit ob eam rem, quod veritus est, ne, si semper atomus gravitate ferretur naturali ac necessaria, nihil liberum nobis esset, cum ita moveretur animus, ut atomorum motu cogeretur. Id Democritus, auctor atomorum, accipere maluit, necessitate omnia fieri, quam a corporibus individuis naturalis motus avellere. Acutius Carneades, qui docebat posse Epicureos suam causam sine hac commenticia declinatione defendere. Nam cum docerent esse posse quendam animi motum voluntarium, id fuit defendi melius quam introducere declinationem, cuius praesertim causam reperire non possent; quo defenso facile Chrysippo possent resistere. Cum enim concessissent motum nullum esse sine causa, non concederent omnia, quae fierent, fieri causis antecedentibus; voluntatis enim nostrae non esse causas externas et antecedentis. Translation from Sharples 1991, p. 73. Si omnia antecedentibus causis fiunt, omnia naturali conligatione conserte contexteque fiunt; quod si ita est, omnia necessitas efficit; id si verum est, nihil est in nostra potestate; est autem aliquid in nostra potestate; at, si omnia fato fiunt, omnia causis antecedentibus fiunt; non igitur fato fiunt, quaecumque fiunt. Translation from Sharples 1991, p. 79. Cf. Bobzien 1998b, pp. 133–4. For Lucretius comparing himself to a physician, see DRN I 926–50; DRN IV 1–25. Ep. Men. 125.

Bibliography Algra, K. A., Barnes, J, Mansfeld, J. and Schofield, M., eds., (1999) The Cambridge History of Hellenistic Philosophy, Cambridge. Annas, J. (1993a) ‘Epicurus on Agency’, in J. Brunschwig and M. Nussbaum, eds., Passions and Perceptions, Cambridge: 53–71. —. (1993b) The Morality of Happiness, Oxford. Asmis, E. (1990) ‘Free Action and the Swerve’, Oxford Studies in Ancient Philosophy 8: 275–91. Atherton, C. (2007) ‘Reductionism, Rationality and Responsibility: A Discussion of Tim O’Keefe, Epicurus on Freedom’, Archiv für Geschichte der Philosophie 89: 192–230. Avotins, I. (1979) ‘The Question of Mens in Lucretius 2.289’, Classical Quarterly 29.1: 95–100. Bailey, C. (1926) Epicurus:The Extant Remains, Oxford. —. (1928) The Greek Atomists and Epicurus, Oxford. —. (1947) Titi Lucreti Cari. De Rerum Natura: Libri Sex, Oxford. Barnes, J., ed., (1984) The Complete Works of Aristotle, 2 vols., Princeton. Barnes, J. and Mignucci, M., eds., (1988) Matter and Metaphysics. Fourth Symposium Hellenisticum, Elenchos 14, Naples. Bobzien, S. (1998a) Determinism and Freedom in Stoic Philosophy, Oxford. —. (1998b) ‘The Inadvertent Conception and Late Birth of the Free-Will Problem’, Phronesis 43: 133–75. —. (2000) ‘Did Epicurus Discover the Free Will Problem?’, Oxford Studies in Ancient Philosophy 19: 287–337.

Lucretius’ cosmological perspective

165

Brunschwig, J. and Nussbaum, M., eds., (1993) Passions and Perceptions, Cambridge. Clay, D. (1973) ‘Epicurus’ Last Will and Testament’, Archiv für Geschichte der Philosophie: 252–80; repr. in Clay 1998: 3–31. —. (1998) Paradosis and Survival, Ann Arbor, MI. Fowler, D. (1983) ‘Lucretius on the Clinamen and “Free Will”’ (2.251–93)’, in : Studi sull’ Epicureismo greco e latino offerti a Marcello Gigante, 2 vols., Naples: vol. I: 329–52; repr. in Fowler 2002: 407–27. —. (2002) Lucretius on Atomic Motion, Oxford. Furley, D. (1967) Two Studies in the Greek Atomists, Princeton. —. (1983) ‘Weight and Motion in Democritus’ Theory’, Oxford Studies in Ancient Philosophy 1: 193–209; repr. in Furley 1989: 91–102. —. (1989) Cosmic Problems: Essays on Greek and Roman Philosophy of Nature, Cambridge. —. (1993) ‘Democritus and Epicurus on Sensible Qualities’, in J. Brunschwig and M. Nussbaum, eds., Passions and Perceptions, Cambridge: 72–94. Hankinson, R. J. (1999) ‘Explanation and Causation’, in K. A. Algra, J. Barnes, J. Mansfeld and M. Schofield, eds., The Cambridge History of Hellenistic Philosophy, Cambridge: 479– 512. Hicks, R. D. (1931) Diogenes Laertius: Lives of Eminent Philosophers, 2 vols., Loeb edition, London and Cambridge, MA. Kleve, K. (1980) ‘Id facit exiguum clinamen’, Symbolae Osloenses 55: 27–31. Long, A. A. and Sedley, D. N. (1987) The Hellenistic Philosophers, 2 vols., Cambridge. O’Keefe,T. (1996) ‘Does Epicurus Need the Swerve as an Arche  of Collisions?’, Phronesis 41: 305–17. —. (2005) Epicurus on Freedom, Cambridge. Pancheri, L. U. (1974) ‘On De Rerum Natura 2.289: A Philosophical Argument for a Textual Point’, Apeiron 8: 49–55. Pope, M. (1986) ‘Epicureanism and the Atomic Swerve’, Symbolae Osloenses 61: 77–97. Purinton, J. (1999) ‘Epicurus on “Free Volition” and the Atomic Swerve’, Phronesis 44: 253–99. Rouse,W. H. D. and Smith, M. F. (1975) Lucretius, De Rerum Natura, Loeb edition, London and Cambridge, MA. Saunders,T. J. (1984) ‘Free Will and the Atomic Swerve in Lucretius’, Symbolae Osloenses 59: 37–59. Sedley, D. N. (1982) ‘Two Conceptions of Vacuum’, Phronesis 27: 175–93. —. (1983) ‘Epicurus’ Refutation of Determinism’ in : Studi sull’ Epicureismo greco e latino offerti a Marcello Gigante, 2 vols., Naples: vol. I: 11–51. —. (1988) ‘Epicurean Anti-reductionism’, in J. Barnes and M. Mignucci, eds., Matter and Metaphysics. Fourth Symposium Hellenisticum, Elenchos 14, Naples: 295–327. —. (1998) Lucretius and the Transformation of Greek Wisdom, Cambridge. —. (2004) ‘Lucretius’, in E. N. Zalta, ed., The Stanford Encyclopedia of Philosophy, at https://plato.stanford.edu/entries/lucretius/. Sharples, R. W. (1991) Cicero: On Fate, with Boethius, Consolation V, Warminster. —. (2007) ‘“Sed haec hactenus: alia videamus”, De Fato 20’, Lexis 25: 53–68. Smith, M. F. (1966) ‘Some Lucretian Thought Processes’, Hermathena 102: 73–83. —. (2001) Lucretius: On the Nature of Things, Indianapolis/Cambridge. Usener, H. (1887) Epicurea, Leipzig.

5

The pleasures of friendship

Friendship as a natural association We value our characters as something peculiar to ourselves, whether they are good and we are esteemed by men, or not; so ought we to value the characters of others, if they are well-disposed to us. (S.V. 15) Philosophizing was not a solitary activity for the Epicureans. The social environment was an essential element of Epicurean self-awareness according to the aitiologikos tropos (the manner of causal explanation), which explained the recognition of oneself through others as a causal agent – or so I argued in Chapter 1. Even more importantly, friendship, springing naturally from the social context, was the quintessential means for one’s Epicurean self-conception and, thus, for securing eudaimonia. As Philodemus’ evidence will attest, the friends of the Garden shared their evaluative perspectives of the moral domain, or at least by learning and practising their master’s tenets they aimed at attaining a balance within their evaluative outlooks. This process, through the shared activities of their studies and the lives they lived together, helped them to have a positive rational influence on each other’s way of life by realizing more clearly the promoted Epicurean values through their own emotional experiences and their own moral judgements. Naturally there were some who quit the community,1 and there were many who needed some extra push or therapy, most prominently through frank speech. But part of being a good Epicurean was an association with friends, which also included sharing the evaluative perspectives of each other. Such continuous and shared interchange of beliefs played an essential part in one’s self-knowledge and in one’s self-narrative, as a consequence of which the Epicurean notion of personal identity and of the self was indispensably connected to the value judgements made of oneself by others. These evaluations shaped one’s self-conception in the context of the ethical commitments of the Epicurean society; this imitated the ethical ideal end exemplified by Epicurus himself not only by philosophizing but by performing various cult practices, creating a group identity.2 In this chapter, I will argue that Epicurean philia, as a cornerstone of the peculiar social context in the Garden, grew out of their account of the psychology of natural human associations. Although I shall maintain, based on Cicero’s evidence,

The pleasures of friendship

167

that Epicurus had a rational conception of emotions and he considered the emotion of love as necessary for friendship, my view, more particularly, is that Epicurus did not justify friendship in terms of the production of rationally calculated advantage. Rather, he thought that friendship was a natural relation based on human psychology and he wanted to advise how such natural associations should be arranged so as best to fit a good life. To substantiate my claim, I will first examine the implications of Epicurus’ inter-entailment thesis, according to which it is impossible to live pleasurably without living virtuously and vice versa, and then I will compare it with Epicurus’ inter-entailment thesis between friendship and pleasure as it is described in Cicero’s De Finibus. I will argue that just as Epicurus considered some of the natural and necessary desires inherently connected to virtuous activities, he regarded friendship as a natural association based on a mutual love for each other’s characters. Philodemus’ evidence will clarify the practical role friendship played in one’s self-conception, which will disclose the Epicurean friend as another self.Therefore the examination of Epicurean friendship is indispensable for the proper understanding of their notion of the self.

The pleasures of virtue For what produces the pleasant life is not continuous drinking and parties or pederasty or womanizing or the enjoyment of fish and the other dishes of an expensive table, but sober reasoning which tracks down the causes of every choice and avoidance, and which banishes the opinions that beset souls with the greatest confusion. Of all this, the beginning and the greatest good is prudence.Therefore prudence is even more precious than philosophy, and it is the natural source of all the remaining virtues: it teaches the impossibility of living pleasurably without living prudently, honourably and justly, without living pleasurably. For the virtues are naturally linked with living pleasurably, and living pleasurably is inseparable from them. (Ep. Men. 132)3 It is worth examining Epicurus’ wording of the second leg of his inter-entailment thesis: the reason why pleasure and the virtues cannot be considered without each other is, literally, ‘because the virtues have grown together with living pleasurably and living pleasurably is inseparable from them ( )’. This statement seems to indicate, on the one hand, that virtues evolve hand in hand with the development of what one takes pleasure in,4 and, on the other hand, that the virtues and the related virtuous activities have grown together with one’s pleasures not simply instrumentally but by being the essential constituents of them.5 Virtues and pleasures are essentially dependent on one another: a virtuous activity is pleasurable and a choice-worthy pleasure is essentially related to human virtue or excellence, though not exclusively as it is clear from the classification of natural

168 The pleasures of friendship

and necessary desires (see Ep. Men. 127–8 and K. D. 29).Therefore, if choice-worthy pleasures are katastematic pleasures, then the katastematic pleasures of freedom from bodily and mental pain – aponia and ataraxia – are related to something inherently virtuous. Thus, if happiness is related to the inherently virtuous disposition of the self, and phrone s is – the ‘natural source of virtues’ – is the primary means for securing such a virtuous disposition, then Epicurean happiness is attainable through one’s cognitive dispositions.6 These virtuous cognitive states are connected to one’s beliefs and to the taxonomy of desires in the following way: if we conceive of pleasures in the sense of the satisfaction of desires – for example, I take pleasure in the satisfaction of my desire to contemplate the Epicurean principles of natural philosophy – which are based on beliefs, in this case on the belief that contemplating physics will help me to understand the real nature of the universe, then the beliefs related to the virtuous cognitive disposition have to be concerned with the satisfaction of the natural and necessary desires exclusively if they are to contribute to the Epicurean goal of aponia and ataraxia. Understanding the nature of the universe is a natural desire because it is pleasurable, given that it dissolves fears of death and of the gods, and it is necessary for achieving pleasurable contemplative results. Thus virtues have to be limited to a certain kind of pleasure, namely to the one which is essentially related to the satisfaction of some of the natural and necessary desires.7 But there is a much more important lesson to be drawn here: this essential relation between virtues and the satisfaction of some of the natural and necessary desires seems to explain further why Epicurus put forward his inter-entailment thesis and how we are to take his statements that the virtues have grown together with the pleasurable life and that virtues and pleasures are inseparable from each other. If our virtues are related to our natural and necessary desires and with their consequent satisfaction, which produce natural pleasures constituting happiness, then the virtues themselves turn out to be natural: since for the desire and the pleasure taken in its satisfaction to be natural they must spring from a disposition, which is itself acquired naturally. It seems that, like Aristotle, Epicurus thought that ‘we are by nature fit to acquire’ virtuous dispositions during our mental development.8 And this is critical for Epicurus because of its further consequence: one does not act virtuously for some consequent pleasure – such a motivation could be seriously questioned (see De Fin. II 78) – but rather one’s pleasurable actions flow from one’s virtuous disposition because one’s motivation to act virtuously is natural. If we are phronimoi and understand that Epicurus’ hedonism articulates what sort of natural pleasures are appropriate to our age and mental development then we will take pleasure in becoming virtuous and acting on our virtuous disposition.9 Considered in this way, virtue is not merely a means to one’s own pleasure but a constituent of it. How one attends to or engages in an activity is determined by one’s character or disposition, and, therefore, the consequent pleasure of the related action is influenced by the manner in which someone realizes a perceived good. The virtuous disposition is constitutive of the end but if it

The pleasures of friendship

169

were not pleasurable naturally, attaining such a disposition would not be choiceworthy.10 Virtue is not something intrinsically good for the Epicureans, given their non-teleological metaphysics; rather it produces the pleasurable life by being the necessary condition for the Epicurean telos and, consequently, virtue becomes naturally worthy of choice because it is constitutive of the kind of hedonism Epicurus advocated.11 If we are all psychological hedonists we all care for our pleasure12 and Epicurus’ normative ethical theory formulated very precise conditions for the katastematic pleasures of rational beings: living virtuously as well as having friends. Epicurus’ inter-entailment theses – between living virtuously and living pleasurably (see Ep. Men. 132, De Fin. I 42–50) and between having friends and living pleasurably (De Fin. I 66) – imply that virtues and friends are intrinsically connected to katastematic pleasures because katastematic pleasures are the necessary concomitants of being disposed virtuously and having friends. Did Epicurus, therefore, also hold that pleasure and friendship are as naturally associated as virtues and pleasures are? I think the evidence, though slight, for Epicurus’ argumentative strategy in his philosophical conception of friendship answers this question affirmatively. In the next section I will show how this natural association of friendship with pleasure is argued for in Cicero’s De Finibus.

Rational emotions Torquatus speaking: I understand that friendship has been discussed by Epicureans in three ways. (1st claim) Some deny that the pleasures, which our friends experience are to be valued in their own right as highly as those, we experience ourselves. This position has been thought to threaten the whole basis of friendship. But its proponents defend it, and acquit themselves comfortably, so it seems to me. (The inter-entailment thesis) As in the case of the virtues, which I discussed above, so too with friendship they deny that it can be separated from pleasure. (Premise 1) Solitude, and a life without friends, is filled with fear and danger, so reason herself bids us to acquire friends. (Reason 1) Having friends strengthens the spirit, and inevitably brings with it the hope of obtaining pleasure. (Reason 2) And just as hatred, jealousy and contempt are the enemies of pleasure, so too is friendship not only its most faithful sponsor, but also the author of pleasures as much for our friends as for ourselves. (Reason 3) Friends not only enjoy the pleasures of the moment, but are cheered with hope for the near and distant future. (Premise 1′) Thus, since we cannot maintain a stable and lasting enjoyment of life without friendship; (Premise 2) nor can we maintain friendship itself unless we love our friends no less than we do ourselves, (Conclusion a) therefore it is within friendship that this attitude is created, (Conclusion b) while at the same time friendship is connected to pleasure.We delight in our friend’s happiness, and suffer at their sorrow, as much as we do our own.

170 The pleasures of friendship

(Expounding the conclusions) Hence the wise will feel the same way about their friends as they do about themselves. They would undertake the same effort to secure their friend’s pleasure as to secure their own. And what has been said about the inextricable link between the virtues and pleasure is equally applicable to friendship and pleasure. Epicurus famously put it in pretty much the following words: ‘The same doctrine that gave our hearts the strength to have no fear of ever-lasting or long-lasting evil, also identified friendship as our firmest protector in the short span of our life.’ (De Fin. I 67–8)13 This is the first of three discussions on the theme of Epicurean friendship in De Finibus I, the latter two (in De Fin. I 69 and 70) being explicitly attributed to Epicurus’ followers and not to Epicurus himself. Cicero recognized one of Epicurus’ dictums in the claim that friendship cannot be separated from pleasure ‘since without it no life can be secure and free from fear’ (De Fin. II 82), and for this reason the quoted passage is generally taken to reflect Epicurus’ theory.14 It delivers an argument, though one not without some tension, as a consequence of which it has been differently construed.15 The major contradiction is between the first claim of the passage: (1st claim) It is not the case that pleasures which our friends experience are to be valued (expetendas) in their own right as highly as those we experience ourselves, and what has generally taken to be its conclusion, which is the first half of the expounding part of the previous conclusions in my understanding: (Expounding the conclusions) Hence the wise will feel (erit affectus) the same way about their friends as they do about themselves. They would undertake the same effort to secure their friend’s pleasure as to secure their own. Since O’Keefe’s behavioural interpretation – according to which there is a major difference between the starting point of the Epicurean position and its conclusion, namely that it is not the same to value someone somehow and to be disposed to treat someone in certain ways, the latter concerning one’s motivation rather than one’s value judgements – interpreters have tried to resolve the tension and to harmonize Epicurus’ supposedly egoist hedonism with his conception of friendship, by creating some space for the latter on egoistic grounds. Accordingly friends would treat each other with as much care as they treat themselves, being disposed to each other in the same way as to themselves but without valuing each other as they value themselves. However, this direction seems somewhat problematic if we recall Epicurus’ psychological structure of human disturbance from Chapter 1: False Beliefs (about the world / values) → Empty Desires → Disturbances. Let us consider this scheme in connection with

The pleasures of friendship

171

my earlier analysis according to which virtues are cognitive dispositions. It seems to me that one’s evaluation and one’s disposition are strongly related elements of the same causal structure. The way I am disposed, from which my actions naturally flow, depends primarily on what beliefs I have about the world or, in other words, what values I attribute to certain things in the world. If evaluation would not be a part of one’s set of beliefs then beliefs could not influence desires, which drive at pleasurable things instinctively. It is Epicurus’ normative ethical theory, which helps to channel one’s beliefs and, consequently, one’s desires. Hence, how I am disposed to my friends also depends on what value I attribute to my friends and to friendship more generally. I think the key to resolving the apparent tension of the argument is to grasp the proper conception of friendship in the Epicurean telos: just as the virtues have turned out to be constitutive of katastematic pleasures, friendship seems to play an important function in attaining the best human condition possible according to Epicurus. Although it might be counter-intuitive at first to think of friendship as a relationship in which we love our friends as much as ourselves, since we would not think that one can value one’s friends’ pleasures as much as one’s own, it is within friendship that we recognize that without reforming ourselves we cannot maintain a lasting enjoyment of our life either. This point of recognition can be understood as the time when the normative part of Epicurus’ ethical theory comes into force to overturn one’s selfish, hedonic inclinations by rational calculation. This calculation implies the comprehension that by loving our friends as ourselves we foster the trust and confidence that are necessary for our own tranquillity, and we realize that such trust and confidence can be fostered only if we love our friends’ characters as much as ourselves. Consequently, we ought not to act selfishly because our rational calculation makes it clear that friendship is constitutive of pleasure in its highest and perhaps even in its most divine quality. Such an approach, however, indeed questions one’s motivation making it seem rather a matter of expediency, especially if we reconstruct the argument along the lines of O’Keefe’s behavioural interpretation and say that we are disposed to act in certain, other-regarding ways in order to maximize our pleasure: the only thing valuable per se in Epicurus’ universe. However, Epicurus explicitly advocated a similar inter-entailment between friendship and pleasure (see De Fin. I 66) as between virtues and pleasure (see Ep. Men. 132, De Fin. I 42–50). For virtues this inter-entailment implies that if pleasures are conceived of as the satisfaction of desires arising from one’s cognitive dispositions, then virtues constitutive of the Epicurean aponia and ataraxia must be related to some of the natural and necessary desires and their satisfaction, since the Epicurean telos is only attainable through the fulfilment of these kinds of desires. Does the inter-entailment between friendship and pleasure imply a similar argument? If it does, as Cicero’s text seems to indicate, then it not only suggests that there is a similar relation between friendship and pleasure as there is between virtues and pleasure, but that it also allows us to draw the further conclusion that just as the virtues have a natural affinity with one’s natural and necessary desires,

172 The pleasures of friendship

so does friendship. However, the desire for having friends seems not only to be natural and necessary for rational beings but is of a kind that also implies the emotion of other-regarding love. First of all, we are by nature inclined to make friends – just as we are to become virtuous – but the understanding of what real friendship implies – to love our friends as much as we love ourselves – makes us realize that such a relation must have a very strong emotive initiation, over and above rational calculation. This extra element of emotive initiation would also explain why the virtues and friendship, despite all of them being or depending on some cognitive dispositions, are treated separately by Torquatus. There is still a question about whether we should take Torquatus’ first claim as part of the argument itself, or rather take Torquatus’ assertion – that its proponents maintain it and easily acquit themselves (tuentur tamen eum locum seque facile . . . expediunt) of the charge that their position would threaten the basis of friendship – as introducing an argument, which is supposed to justify the first claim.The latter option could easily collapse into the former, but Cicero’s language (seque facile . . . expediunt) indicates that the speaker Torquatus may not be entirely convinced of the success of the supporting argument, which I reconstruct as follows: Thesis: P1: P1′: P2: C(a):

Friendship cannot be separated from pleasure. Reason herself bids us to acquire friends. R1–R3 We cannot maintain a stable and lasting enjoyment of life without friendship. Nor can we maintain friendship itself unless we love our friends no less than we do ourselves. Therefore, it is within friendship that this attitude is created;

and C(b):

Friendship is connected to pleasure.

Cicero himself was certainly not persuaded by this argument (see De Fin. II 78). He puts forward the objection that love (amor) does not commonly result from any calculation of expediency. But have we not seen that some affection indeed results from rational calculations according to the Epicureans? Is the normative part of Epicurus’ ethical theory not a set of logical rational calculations in harmony with its descriptive part, which, if we are ready to follow it, should substantiate the Epicurean end of aponia and ataraxia, including the most perfect and stable forms of pleasurable affection? If my arguments so far for the interpretation of virtues and friendship as the cognitive conditions of katastematic pleasures are correct, then Epicurus, indeed, found it possible on theoretical grounds that certain cognitive dispositions are responsible for certain affective states, and even if we do not value our friends’ pleasure per se as highly as our own, we can learn in the course of a friendship that the real substance for its bond, if a friendship is to last, is emotionally conditioning ourselves towards our friends to love/value (diligamus) them as much as ourselves.16

The pleasures of friendship

173

Therefore, if we take the verb diligo to express an emotion based on evaluation, it might resolve the internal contradiction of the passage. The emphasis of the first claim is that we are not to value the pleasures belonging to our friends per se as highly as our own – the pleasures per se we feel – which does not exclude the possibility of valuing and consequently loving our friends themselves and their character as much as ourselves, i.e. our own characters. Friends can share their pleasure to the extent of aligning the cognitive dispositions by which their pleasures are constituted, but they cannot share their internal affective states, their pleasurable sensations.17 Consequently, there can be a distinction between the value we attach to our friends’ pleasurable affections and to their characters, if we recognize the same inter-entailment between friendship and pleasure as there is between the virtues and pleasure. The way in which this argument defends the first claim seems to be simply that although our pleasures are private and, therefore, they are to be valued higher than the pleasures of our friends, the way in which those pleasures come about is essentially other-regarding. Someone might object that it makes no difference if one values one’s friends’ pleasures or one’s friends, if friendship and pleasure are inter-entailing. But in fact this inter-entailment does not imply causal reciprocity any more than it implies it in the case of the inter-entailment between virtues and pleasure. It is true that there is no pleasurable life without virtues just as there is no virtuous life without pleasure, but the causality implied in the relation is only one-way. Virtues are productive of pleasures, but pleasures are not productive of virtues. That is to say, if one only cared about the pleasures of one’s friends and not about their characters then there is no guarantee that the way one values the other’s pleasure will be choice-worthy, since one can try to please a friend with things constitutive merely of kinetic and not katastematic pleasures. As Torquatus concludes, by valuing or loving friends as much as oneself, the wise will consequently feel or be affected in the same way about their friends as about themselves (eodem modo sapiens erit affectus erga amicum, quo in se ipsum). Since this consequence is a clarification of the conclusion that friendship – based on a chain of rational considerations – is connected to pleasure, these feelings must have depended on one’s cognitive disposition. Thus, when Torquatus claims that the wise would undertake ‘the same effort to secure their friend’s pleasure as to secure their own’ (De Fin. I 67), the effort in question is not the same as saying that they value their friends’ pleasure as much as their own, but rather he seems to claim that they secure their friends’ pleasure by taking care of their cognitive dispositions constitutive of pleasure. The effort they put into securing the constitutive conditions of their friends’ pleasure, in light of the activities of the Garden, seems to have been an intellectual improvement of each other by a constant contemplation of Epicurus’ tenets in company with the like-minded within that community (see Ep. Men. 135; also Sextus Adv. Math. XI 169–70 = fr. 219 Us.). This intellectual attitude to pleasure implies that for the Epicureans to maintain the security of friendships for their aponia and ataraxia, they had to develop love towards their friends’ characters on the strength of their cognitive dispositions

174 The pleasures of friendship

and care for them as much as they do for themselves.Thus, the emotion of amor must have been based on some evaluation and, consequently, the first group of Epicureans seems to have wished to clear themselves of the charge that they present friendship as only a matter of expediency by arguing for the inseparability of friendship from pleasure. But how successfully can such an explanation be free from the standard objections that it does not imply real affections or includes a serious doublethink? If I only care for my friends’ character so that I have someone to care for my character, to make sure that I am staying on the right Epicurean track (which is in turn constitutive of my pleasure), this does not seem to imply genuine other-concern. It seems to me, nonetheless, that everything turns on what we take genuine other-concern to be. On the strength of my explanation, Epicurus is entitled to draw the conclusion that the wise man will be affected the same way in relation to himself as in relation to his friends because he took affection (in the sense of being affected pleasurably or painfully in one’s mind) as an intellectual matter dependent on one’s cognitive disposition. And somewhat like Chrysippus, who thought that feelings and emotions were rational beliefs, Epicurus also took cognitive contents to be constitutive of affective states. His intellectual conception of some of the affections may also have been the reason why Epicurus did not leave any room for intermediary states between pleasure and pain in general (see D. L. II 89), a position for which he was constantly criticized: there is no time when one is not affected either pleasurably or painfully in one’s mind because one either has the appropriate intellectual disposition or one does not.18 Similarly when someone is properly disposed, even in the face of physical torture or a fearful situation, one is able to overcome one’s seemingly painful affections not by unexplainable tricks producing more dominant pleasant affections but also by one’s proper intellectual disposition.19 If one has the correct Epicurean mind-set, one will always feel pleasurably affected no matter whether one is seriously sick (D. L. X 22), on the rack (D. L. X 118) or even dying for a friend (D. L. X 120). Genuine Epicurean affection or love, therefore, will always be an intellectual, cognitive matter and, consequently, it is not our friends’ pleasure per se that matters but making sure that their dispositions are of the right kind.20 Still we ought to be careful to maintain a distinction between affective states and emotions because the latter seem to be of a more complex nature. While by now it should be clearer how Epicurus conceived of the affective effects of virtues or of friendship, I think there is a significant distinction between being disposed virtuously and having the consequent constant affection of pleasure and being disposed naturally to make friends out of which the friendly emotion of love can flow. In the case of the latter, there has to be a separate, external cause bringing about the (naturally pleasurable) emotion of loving our friends as much as ourselves, and it is not hard to guess that such external causes must be our friends themselves. However, if the cognitive content of our natural dispositions to make friends rests on the rational consideration of how to foster trust and confidence as outlined in the argument for friendship, these beliefs are also causally responsible for the emotion of loving our friends as much as ourselves.

The pleasures of friendship

175

Yet, an angrily disposed person, for example, is not always angry: thus, it seems that our pleasurable emotion of love towards our friends requires some form of presence of the friends in question, either actually or at least in one’s thoughts, just as, for example, the emotion of the fear of death would require one to see exemplifications of death or to contemplate it in order for this particular emotion to occur. Thus, it seems that the Epicureans treated friendship separately from the virtues because the emotion of loving one’s friends is causally more complex and not merely dependent on one’s disposition.21 And, actually, the causal complexity of emotions seems to be a sufficient reason for genuine otherconcern, since such an emotion can only arise and be maintained – regardless of its initial motivation – when at least two parties are mentally or physically present to each other and value each other in the same way. The pleasurable emotion of love towards our friends does not merely come about because of our dispositions, which could indeed be egoistic and self-centred, but also because of the causally relevant presence of our friends. We cannot pull off faking the emotion of love towards them because, if it is non-genuine, then that leads to a discrepancy in evaluation – most likely even to a non-pleasurable emotion – as a consequence of which, the friendship breaks its bonds and ceases to exist. Very similarly to Aristotle’s considerations, one’s  or goodwill must be recognized and reciprocated for there to be a friendship (see N. E. VIII 1155b27–56a6). Therefore,  on its own is not a sufficient condition for friendship either for Aristotle or for Epicurus, since for the emotion of friendship to last and be the bond between people, mutual well-wishing and welldoing in a community of friends are necessary requirements.22 And just as it was the case for Aristotle that friends ‘take pleasure in having a shared perception of what is good in its own right’ (N. E. IX 1170b4–5), which implied a perception of one’s self as well as a shared perception of the friend (see N. E. IX 1170b4–14), this emotive connection can be maintained for the Epicureans if their pleasure originates from similar grounds as well:   .

Most beautiful is the sight of those close to us, if the harmony-of-mind of the primary kinship exists – the one [i.e. the harmony-of-mind] that gives great concern to this thing [i.e. to have the sight of those close to us]. (S.V. 61)23 In conclusion, Epicurean friendship turns out to be the pleasurable emotion of mutual love and affection between people, based on a harmony of mind.24 The question of whether friendship originates from some kind of affection or liking towards the other by reference to advantage, pleasure or the other’s character did not seem to be as emphatic for Epicurus as has often been conceived, perhaps under the influence of Aristotle’s theory. His main concern seems to have been, instead, the strong emotional relation between the parties involved in the

176 The pleasures of friendship

friendship. The emotion of love includes genuine other-concern because of its causal inter-dependence and, thus, even if it includes any of the motivations of the three basic types of Aristotelian friendship, it will necessarily be the kind of friendship ranked as the highest by Aristotle, since it will rest on genuine love towards the other’s character, which safeguards a friendly relation stemming from natural and necessary desires, at least in light of my understanding of Cicero’s text. As Cicero testifies (De Fin. I 66–70), the later followers of Epicurus were not of the same mind about friendship. In spite of the dogmatism of the Garden, they were ready to present their own different conceptions; one of the groups changed its mind because of the pressure of criticism of the New Academy (see De Fin. I 69). They did not find the distinction between valuing someone and someone’s pleasure adequate, and consequently they worried that if they held the view that in friendship only one’s own pleasure is valued the whole notion of friendship would look utterly untenable. Such a limited understanding of the original Epicurean claim that ‘the pleasures our friends experience are not to be valued per se as highly as those we experience ourselves’ excludes the interentailment thesis of friendship and pleasure, since it does not take into account the possibility that our highest end, katastematic pleasure, is attainable through its dependence on friendship. Instead, the modified view requires the abandonment of such an inter-entailment by claiming that friends are choice-worthy per se, ‘regardless of any expectation of pleasure’ (tum ipsum amari per se etiam omissa spe voluptatis; De Fin. II 82). But such a modification runs into the question of how to make a distinction between kinetic and katastematic pleasures because as soon as there are subsidiary allowances made for not identifying our telos with pleasure, the Epicurean theory collapses (see Ep. Men. 128, K. D. 25). The third group of Epicureans (De Fin. I 70) also seem to have been motivated to solve the seeming lack of genuine motivation by reducing friendship to a kind of pact between the wise. As I have shown, the kind of friendship entertained by the first group of Epicureans can imply genuine other-concern on the strength of the rational and natural conditions of katastematic pleasure; and perhaps it is no accident that Cicero does not devote substantial criticism to either of the two later groups, being satisfied to use these deviations as a sign of confusion among the followers, and indicating how confusing the Master himself was even to his disciples. Admittedly, Cicero’s presentation is not disinterestedly expressed, and we would probably be in a better situation to evaluate it if we had such primary evidence as Epicurus’   or Metrodorus’  , works essentially related to friendship. But we do have evidence of the firstcentury BCE Philodemus, who in a series of works addresses matters related to friendship.25 One of them in particular, On Frank Criticism ( ), lends great support to the understanding that the foundation of Epicurean friendship was affectionate love. Philodemus also sheds further light on how the development of the conception of oneself as a desiring agent responsible for certain actions was essentially related to interpersonal relationships and concerns,

The pleasures of friendship

177

as well as on the epistemological method of shared identity for one’s selfconception and eudaimonia.26 Therefore investigating the part friendship played in Epicurean moral reform will bring to light the development of the Epicurean conception of one’s self.

Philodemus and the therapeutics of friendship The central subject of On Frank Criticism was the method of moral reform, which consisted of hypothetical question and answer scenarios in the area of psychagogic theory and practice within or outside a moral community. It implied praise and blame in the service of moral development, the negative assignment of responsibility often signifying specific errors and possible ways of correction.  was a therapeutic means and it involved both measured (; see Fr. 6) and harsh (; see Fr. 7) criticism, depending on the errors and personalities in question. It included arguments to address the rational character of the students, and also applied some rhetorical features intended to stir emotions.27 Even in general, outside the Epicurean school, it was conceived of as not only akin to friendship but as the language and the most potent medicine of friendship.28 For the Epicureans, frank criticism as a method was important because, as part of Fr. 25 testifies, ‘through frankness, we shall heighten the goodwill () towards ourselves of those who are being instructed by the very fact of speaking frankly’. That is to say, frank criticism, if practised and received correctly, strengthens one of the necessary conditions of friendship, . It is essential to see that it was primarily the wise man who criticized the pupils – teachers of various ranks were also entitled to practise it or even promising beginners as well as advanced students – and in certain cases the wise man was even ready to go so far as attributing some errors to himself hypothetically, by saying ‘that it occurred in his youth’ (see Fr. 9), thereby artificially sharing the student’s mental disposition to a certain extent. Such practice mirrors the idea of shared identity not only in a negative sense, but by its implications also as a desired pleasurable telos in the Epicurean community: when the wise man ascribes an error to himself hypothetically, he imagines the sort of mental disposition required to make the specific error in question and corrects his pupil by pointing out the false beliefs involved in such a disposition. But psychological community does not extend only to this imaginary point, since given the characteristics of the relation, the wise man is aiming to make the students themselves wise, in order to have the same or at least similar mental characteristics to his own and to share with them the cognitive content of his disposition by implanting the correct ideas about nature in his students. And most importantly, he acts out of goodwill and love: (Fr. 80) . . . {although they say that they} differ from them, both in bearing a resemblance to the teachers, and further in being favorably disposed toward us, one must bear those who have

178 The pleasures of friendship

scrutinized {one}. For these {the students} have profited unhesitatingly on account of their {the teachers’} love (), and practically on account of their [goodwill] ( [ ]). . . .29 Philodemus appears to be advising students – if we follow the translators’ clarification of the passage – how to behave in view of the wise man’s attitude towards them. This attitude is essentially other-regarding since it originates from the teachers’ goodwill and love without any obvious benefits accruing to him from the students. Furthermore, the Epicurean teachers appear to take care only of those towards whom they are already disposed with both the necessary and sufficient conditions for the strongest bonds of friendship, goodwill and love, even if some of these students have not got the same attitude just yet (also see Col. IIIb). Most likely, if the students do not change their dispositions and do not start to resemble their teachers in their attitudes, their relations will be short-lived. How is such a psychological reform of the students possible according to Philodemus? The student has to give him- or herself over to the ‘saviour’ (Fr. 40) since, if one has chosen to live according to Epicurus’ teachings, one needs to obey him (Fr. 45). Such compliance is founded on trust30 and leads at a certain point even to voluntary disclosure, as part of Fr. 42 testifies: (Fr. 42). . . And many of the companions will somehow voluntarily disclose {their secrets}, even without the teacher interrogating {them}, because of their concern. . .31 Such a voluntary exposure is perhaps a further phase in one’s moral progress compared to opening up under the pressure of admonishment. It is a vital one since it implies a desire to share one’s feelings and thoughts with someone else who – in response – can polish the other’s character in an essentially epistemological relation of self-knowledge. (Fr. 28) We can show by reasons that as numerous and beautiful as are the things that come to us by friendship, none is so great as having someone to whom one shall tell what is in one’s heart and whom one shall hear speaking back. For very greatly does our nature desire to reveal to others what it is thinking.32 This fragment does not distinguish between teachers and pupils and consequently it seems to exemplify a friendly relationship when both parties are completely open to each other.33 The passage does not focus on the epistemological aspects of such a relation, however, but rather emphasizes the pleasure drawn from the characteristics of such an association. David Armstrong compares the idea of the desire to share one’s thoughts with one’s friends to a similar thought in the Epicurean community of the gods, a comparison which proves to be very fruitful for various reasons. The textual evidence of Philodemus’

The pleasures of friendship

179

On the Gods III attests that gods ‘share their affections/feelings’ and, even more interestingly, it also says that ‘for us, the weak, who require friendship for external needs in addition, one has no needs in relation to friends he has lost . . . [and yet] our feeling of wonder at their similar character to our own . . . holds [us] together [in even] the highest affection’.34 As Armstrong notes, the latter part of the text is uncertain, but on such an understanding, it corroborates to a large extent my interpretation of Epicurean friendship. It seems that friendly affections can be maintained even towards those who are already deceased by having an active memory of your friends, which seems manageable to the highest extent only if you share their character: and thus, through your own continued living you can maintain your affections towards your lost friends even in relation to yourself, i.e. to your own character, given that it implies your theoretical understanding of the possibility of such a self-reflection. Keeping one’s friends alive in one’s memory seems to have been directly connected with the Epicurean cults and festivals, something not properly understood by Pliny (see Pliny, N. H. 35.5) or later on by Plutarch (see De Lat. Viv. 1129A). A very important aspect of such remembrance is brought out by another piece of evidence from Philodemus, a fragment of his short treatise On Epicurus (PHerc. 1232): . . . as concerns those who experience turmoil and difficulty in their conceptions of natures that are best and most blessed. [But Epicurus says] that he invites these very people to join in a feast, just as he invites others – all those who are members of his household and he asks them to exclude none of the ‘outsiders’ who are well disposed ( [] [] . . . ) both to him and to his friends. In doing this [he says], they will not be engaged in gathering the masses, something which is a form of meaningless ‘demagogy’ and unworthy of the natural philosopher; rather, in practicing what is congenial to their nature, they will remember all those who are well disposed to us (   []  ), so they help to sanctify things that are fitting to their own blessedness ( []    [ ] [. . . . . . ]). . .35 Although the beginning of the fragment clearly refers to the gods, I think Diskin Clay is correct to argue that at its end those to be remembered (‘who are well disposed to us’) are the heroic dead of the first generation of the Epicurean community in Athens.36 Thus the strongest candidate for the kind of memorial service in question is the so-called , which was held on the twentieth of each month to commemorate Metrodorus and Epicurus, a practice set up by Epicurus himself (D. L. X 18).37 As Clay has pointed out, by the time of Philodemus the number of people commemorated could have been even more, indicated by . That suggests that we include at least the four founding fathers, Epicurus, Metrodorus, Polyaenus and Hermarchus, the so-called .38

180 The pleasures of friendship

The all-important question, nevertheless, is: who are remembering? People confused about the best and most blessed natures (the gods), the members of the household or the outsiders? Concerning the status of the first two groups of people it seems fairly clear that they were most likely all well disposed to Epicurus, even if confused to some extent, but the feasts were open even to those – the third group – who were not members of the community, but were also well disposed to the Epicureans ( []  ). As I have argued,  was the necessary but not sufficient condition for Epicurean friendship and it also seems to have been a catchword for possible recruitment.39 The people outside the circles, obviously, did not need to have a personal or other sort of knowledge of the people to be remembered, but it does not seem to exclude the possibility that even they can learn to call to mind the Epicurean wise men who were disposed well not only to the already converted but even to that part of the world which was ready to listen.Thus anyone reciprocating such a disposition could join the Epicurean feasts, which must have implied not only a festive celebration and remembrance of the founding fathers but also a contemplation of their teachings, through which everybody who joined could also ‘help to sanctify things that are fitting to their own blessedness’ or, in other words, attain a correct conception of nature, of themselves and so on. These feasts seem to have served as an introduction to the Epicurean way of life, which in its most complete and perfect form led to the assimilation to the Epicurean wise men.40 As we can apprehend in view of On the Gods III, even though divine beings are in no need of anything, they enjoy sharing their affections with each other once they happen to live in a society. They naturally derive pleasure from their connections. It seems that in such cases the gods are satisfying natural but non-necessary desires, as even in some cases we ourselves accept things from others we do not need, as Philodemus puts it in the remainder of the passage (not quoted). But this would only be a kinetic variation of one’s condition whether human or divine. However, Fr. 28 seems to imply a much stronger, necessary desire for sharing one’s thoughts in the case of humans. Elizabeth Asmis takes this as a degree of intimacy, although even in this context, which amplifies a restricted understanding, it is not clear if we should take the idea only in such a limited sense.41 The text seems to imply only that the desire for sharing one’s thoughts with others is a necessary characteristic of human nature in general, and in comparison with the evidence of On the Gods III it is the understanding of this as necessary, which distinguishes human from divine behaviour. It is a more complex matter what consequences such a desire may lead to. As we read in Fr. 43: (Fr. 43). . . as it will come to be . But if {on account} of [{our} good {qualities}], how not also of {our} bad ones? For, just as it is suitable on account of the good cheer of the former, so too thanks to sympathy for the latter, through which we are helped. . . .42 It is not clear if Philodemus is reflecting on his lecture notes based on Zeno of Sidon’s lessons or if the subject of this fragment was part of a lecture. It is also unclear who the parties involved are, but since Voula Tsouna has made a strong case based on the evidence for the imperfections of the Epicurean teachers practicing their therapeutic methods of frank criticism, both options of interpretation, the teachers or the students having a bad influence on others are left open.43 The latter option seems to be more plausible in the next fragment because of the asymmetrical dispositions between the different members of the school, since it is very unlikely in light of Fr. 80 (quoted earlier; see also Col. IIIb) that even if a teacher fails to heal a student, he would be unfavourably disposed to him or her: (Fr. 44) . . . they further inflame {them} whenever they are involved with those same men, who do not like {them} nor know how to correct {them} nor will persuade those who are much better, instead of {being involved} with one who is pure and loves {them} () and is better and knows how to treat {them}. And if he, with [handshakes], without obtaining the finest wages . . .44 The different emotional dispositions between the members of the school teach us an important lesson about the Epicurean community itself: not all the members were forced to love each other and to live together as a happy family, but more realistically, there existed social tensions even within the Epicurean community. It is not a discrepancy in the theory, but it exemplifies the natural work-in-progress status quo of their society, which is open to anyone well disposed towards the founding fathers and ready to reform his or her moral commitments. It would be too idealistic to expect everyone, however, to attain perfect assimilation with Epicurus and to make friends with all the members. Consequently, the idea of fellowship seems to have existed in the community, but only as a social phenomenon and not as a sufficient condition for friendship. Fellowship on its own could not have qualified as friendship because of its insufficient emotional input, but it must have been a valid form of practical relationship between students not favourably disposed to each other. These fellows were ready to learn but not from each other, or perhaps not just yet since, ideally, the students’ moral improvements must have been in harmony with their dispositional changes.45 One of the most striking feature of administering the students’ dispositional changes seems to have been the teachers’ self-disposal not only in front of their colleagues but their students: (Fr. 55. 1–6) [They present] with frankness their own errors in front of the students, so that they are put before Epicurus and for the sake of correction.

182 The pleasures of friendship

If we assume that the practice described is not limited to Epicurus’ time, it resembles to a large extent the ethical function of the commemorative cults: the teachers shared their intellectual errors and corrected them in the light of Epicurus’ moral philosophy. A community discussion, whether of a small or a large scale, strengthened the identity of its participants and served as a therapeutic means to moral improvement, most of all by heightening the goodwill () towards each other (see Fr. 25 and 36), a necessary condition of Epicurean friendship.

Conclusion Feeling for friends, living or dead, seems to have been primarily a rational disposition for the Epicureans. Their friendships were fundamentally based on the emotion of love and even if, as we have seen, the normative part of Epicurus’ ethics rationalized this emotion, he conceived its emergence as natural and even necessary as the inter-entailment thesis between friendship and pleasure implies. As psychological hedonists we naturally seek the pleasure of friendly associations. Our psychological development is necessarily embedded in our social and cultural environment, which involves one’s own self-reflection through others. We learn much about ourselves not only by observing the behaviour of others but also by listening to the opinions of those who are well disposed to us. Thus our self-knowledge is essentially articulated through others. If there is mutual love involved between friends, for Epicurus it entailed that they care about each other’s characters. That required a kind of intellectual exchange necessary not only for one’s self-knowledge but for one’s moral improvement. In its most perfect form Epicurean friendships seem to have implied a shared personal identity, which consisted of the attainment of a similar cognitive disposition implying an intellectual self-narrative in harmony with Epicurus’ philosophy, which substantiated the familiar ancient theme that a friend is ‘another self ’. Thus Epicurus’ theory seems to have blurred the differentiation Aristotle made earlier concerning how friendship originates from some kind of affection or liking, by reference to one of the following three concerns: (1) advantage, (2) pleasure and (3) the other person’s moral goodness. Aristotle and Epicurus agreed that mutual attraction is not a sufficient condition of philia because two people can be attracted to each other without any further active connection or mutual well-wishing and mutual well-doing out of concern for one another, which are characteristic of friendship (see N. E. VIII 1155b31). Were mutual well-wishing and mutual well-doing out of concern for one another implied in all of the forms of Aristotelian friendships? In the cases of advantage and pleasure friendships, which appear to be more self-centred, Aristotle thinks this mutuality is present only incidentally (kata sumbebe k os), e.g. as in erotic relationships people do not love each other but only their incidental features (N. E. IX 1164a10–12) or in advantage friendships once the advantage ceases, so does the friendship (N. E. VIII 1157a14–16 and IX 1165b3–4). In the case of

The pleasures of friendship

183

character friendship, however, we maintain a close relationship and community with our friends and we wish and do good things for them for their own sake. It seems that for Epicurus friendship, even if it starts from a need for help (see S. V. 23), the reason why it is choice-worthy is that our ataraxia could not be complete without self-knowledge, which is only attainable through others. But this condition still does not make friendship an advantage relation because without a genuine love of each other’s characters friends cannot maintain a naturally pleasurable association. Unless they maintain a genuine affection to cultivate each other’s characters their friendship dissolves. However, if they can secure their mutual affection based on a harmony of mind (see S. V. 61 earlier) then a friend becomes another self to a friend because to realize their ataraxia they imitate the same ideal self, that of Epicurus. In a similar spirit, Aristotle had not supposed that virtue friendship is only possible between moral heroes. Even if the reason for the attraction is the character of the other, it was not necessarily an attraction towards complete and perfect virtue already present in one of the parties. Thus there may be some limits or even some degree of deficiency involved in the relation and it is quite natural that friends do not possess moral goodness to the same extent, as is also clearly described in Epicurean school practice by Philodemus.46 In such cases we are not necessarily talking about inequality in natural status between friends, such as that between husband and wife for Aristotle, but about simple inequality. Ideally, friendship demanded absolute equality for Aristotle (N. E. VIII 1158b30–33), but that would have actually excluded the place for moral improvement in cases of character friendships, which was also an essential element for the Epicureans. I think this latter point is highly important in light of the mutual motivation, which Epicurus thought was available within a circle of friends, since if a friend is another self then doing good to one’s friend is like doing good to oneself, which can be taken in two ways: (a) it is impossible to eradicate selfish or egoistic motivations from friendship; or (b) it is all the same if one does good to one’s friend or to oneself since it makes no sense to consider these actions in one or the other way. If one lives in a community with one’s friend, which was a very emphatic requirement for both the Aristotelian and Epicurean concepts of friendship, one is working towards the common good anyway. And the virtuous person actually needs such a community since one could not possess the virtue of character friendship if one could not practise it. Therefore friendship was a necessary condition for eudaimonia for both the Aristotelians and the Epicureans: observing the good and hence virtuous actions of one’s friends inadvertently contributes to one’s self-knowledge and pleasure, since the actions of virtuous friends are like those of a virtuous person and the latter feels elated at observing dear and familiar actions constitutive of happiness for all the parties involved in this kind of friendly relation. As Aristotle famously put it: they take pleasure in having a shared perception of what is good in its own right. But as a good person is related to himself, so ‹must he be› also to his

184 The pleasures of friendship

friend, since a friend is another self.Therefore, in just the same way that his own existing is worth choosing to each, so too is his friend’s, or nearly so. But his existing was worth choosing on account of his perceiving himself as being good; and this sort of perception is pleasant in its own right. Hence there must also be a shared perception of his friend, that he exists, and this would occur by their living life together and sharing in discussion and thought, since that is what ‘living life together’ means for human beings, and not, as with cattle, feeding in the same place. (N. E. IX 1170b4–14)47 Aristotle’s and Epicurus’ conceptions of character friendships thus turn out to be very similar in important aspects. But Epicurus was less concerned with the question of how the affection originated; he was interested rather in the psychology underlying human associations and its rationalization. For that he set himself as an ideal to which his student could assimilate.Thus Epicurus became the constant ‘other’ self.

Notes 1 For Timocrates see Sedley 1976, pp. 127–32. 2 See Clay 1998, pp. 101–2. For the Epicurean cults, see Clay 1986, repr. in 1998, pp. 75–104. I will take some of the cults into account only when necessary while dealing with the psychology of friendship later, pp. 184–6. 3   ’ ’                          ,            . Text and translation from Long and

Sedley 1987, vol. II, pp. 116–7 and vol. I, p. 114, respectively. 4 Cf. Alexander of Aphrodisias: ‘The Epicureans maintain that what is first congenial to us, unqualifiedly, is pleasure, but they say that as we get older this pleasure becomes more fully articulated’. (Supplement to On the Soul 150.33–34.) Translation from Wolfsdorf 2013, p. 177. 5 It is important to note that Epicurus’ inter-entailment thesis does not imply identity; cf. fr. 508 Us. and Diogenes of Oinoanda Fr. 32 in Smith 1993; and Cicero’s complaint in De Fin. II 50. 6 In K. D. 17 the virtue of justice is also connected with the mental disposition of ataraxia. Freedom from mental trouble is not simply a subsequent, later effect of the just person, but one is , free from trouble if just. Consequently the Epicurean goal of bodily and mental pleasures understood as aponia and ataraxia are constituted by some virtuous disposition, which is the katastematic condition of pleasure. Katastematic pleasures flow from virtuous characters. Also see Chapter 3, n. 30. 7 It does not follow that all the pleasures stemming from the satisfaction of natural and necessary desires are the outcomes of virtuous activities because I can satisfy my

The pleasures of friendship

8 9

10

11

12 13

185

natural and necessary desire for food, for example, regardless of whether I am virtuous or not. However, Epicurus’ idea seems to have been that the way I satisfy such hunger must be a matter of rational consideration for the wise. Cf. Aristotle N. E. II 1103a18–26, who says that virtues do not come about in us by nature nor contrary to nature, but it is our nature to acquire them through continuous habituation. One may object that experience does not justify the claim that virtuous activity and pleasure are always realized simultaneously. Even though Epicurus pointed out in Ep. Men. 128–30 that one has to forgo pleasure in certain cases to act properly, as a result of which one will actually maximize one’s pleasure, and to this extent when acting virtuously even if unpleasant at the time of the action, yet still pleasant on account of the result. Some cases seem harder, nevertheless, e.g. taking care of one’s elderly or dying parents may not always be pleasant, yet certainly just, and in my understanding of Epicurus’ account it has to also be pleasurable, at least in retrospect. But it is somewhat counter-intuitive to think that even in retrospect anyone would feel pleasurably affected. Epicurus also seems to have supported this position by its implication: if virtuous dispositions are natural to the extent that they are pleasurable then it should follow that the disposition of evil or wicked men is painful because it is a cognitive error. In S.V. 7 e.g. we read that an evil-doer cannot secure his or her tranquillity because he or she can never be sure of not being caught eventually, therefore he or she will never have the pleasure of ataraxia. Also cf. S.V. 16, 21, 37, 46 and 79. I think Epicurus’ provocative comments, considered outrageous by all the ancient commentators, according to which he spits on any virtue ‘and those who emptily admire it’ if it does not produce any pleasure (Athenaeus Deipn. XII 547a = fr. 512 Us.), can be considered along these lines. Virtues have no intrinsic value in Epicurus’ metaphysics, thus he cannot regard them as valuable on their own, but his statements can be taken counterfactually: if they produced no pleasure, ‘one should leave them alone’ (Athenaeus Deipn. XII 546f. = fr. 70 Us.). But as a matter of fact they are constituent of the pleasurable life, thus they are natural and to be cultivated. Cf. Chapter 1, n. 13. (I 66) Tribus igitur modis video esse a nostris de amicitia disputatum. (1st claim) alii cum eas voluptates, quae ad amicos pertinerent, negarent esse per se ipsas tam expetendas, quam nostras expeteremus, quo loco videtur quibusdam stabilitas amicitiae vacillare, tuentur tamen eum locum seque facile, ut mihi videtur, expediunt. (The inter-entailment thesis) ut enim virtutes, de quibus ante dictum est, sic amicitiam negant posse a voluptate discedere. (Premise 1) nam cum solitudo et vita sine amicis insidiarum et metus plena sit, ratio ipsa monet amicitias comparare, (Reason 1) quibus partis confirmatur animus et a spe pariendarum voluptatum seiungi non potest. (I 67) (Reason 2) atque ut odia, invidiae, despicationes adversantur voluptatibus, sic amicitiae non modo fautrices fidelissimae, sed etiam effectrices sunt voluptatum tam amicis quam sibi, quibus non solum praesentibus fruuntur, sed etiam spe eriguntur consequentis ac posteri temporis. (Premise 1′) quod quia nullo modo sine amicitia firmam et perpetuam iucunditatem vitae tenere possumus (Premise 2) neque vero ipsam amicitiam tueri, nisi aeque amicos et nosmet ipsos diligamus, (Conclusion a) idcirco et hoc ipsum efficitur in amicitia, (Conclusion b) et amicitia cum voluptate conectitur. nam et laetamur amicorum laetitia aeque atque nostra et pariter dolemus angoribus. (I 68) (Expounding the conclusions) quocirca eodem modo sapiens erit affectus erga amicum, quo in se ipsum, quosque labores propter suam voluptatem susciperet, eosdem suscipiet propter amici voluptatem. quaeque de virtutibus dicta sunt, quem ad modum eae semper voluptatibus inhaererent,

186 The pleasures of friendship

14

15 16

17 18 19

20 21

22

23 24

eadem de amicitia dicenda sunt. praeclare enim Epicurus his paene verbis: ‘Eadem’, inquit, ‘scientia confirmavit animum, ne quod aut sempiternum aut diuturnum timeret malum, quae perspexit in hoc ipso vitae spatio amicitiae praesidium esse firmissimum.’ Translation from Annas and Woolf 2001, pp. 23–4. O’Connor 1989 (pp. 184–5) cast doubt on whether more than the first claim of the passage could be directly attributed to Epicurus since it seems to be in contradiction with the rest, thus he prefers to attribute the rest of the argument to later Epicureans. O’Keefe 2001 and Evans 2004 see some apparent tension, but the ways they wish to resolve it depend on their presumption of Epicurus’ being an egoistic hedonist, and thus I do not find their solutions satisfactory. Cf. also Frede 2016 (p. 102) who thinks that the fact that Epicurus is not named as the adherent of the first position suggests that he did not have an extended explanation and defence of friendship. Cf. O’Keefe 2001, p. 290, Brown 2002, pp. 70–1, Evans 2004, p. 411 and Frede 2016, pp. 102–5. Hence even though there is a shift from diligo to amor between Cicero’s representation and his subsequent criticism, it is not so decisive as it might seem at first. For more on the elements of distortion in Cicero’s representation, however, see O’Keefe 2001, pp. 296–7. For aligning cognitive dispositions see S. V. 61, quoted on p. 175, as well as in the section on Philodemus later. Cf. Warren 2016, pp. 47–55. Although K. D. 3 does not indicate the different kinds of pleasures, there is no question that only those people can be free from both bodily and mental pain whose mental disposition is based on a proper understanding of nature. As stated in K. D. 12, such knowledge is the condition of dispelling fear about the most important matters and without natural science it is not even possible to attain unmixed pleasures ().This phrase opens the possibility of comparison with Plato’s Philebus, nonetheless, let it suffice here to say that in Epicurus’ case, given the subject matter of his natural science, it is clear that unmixed pleasures can only be those of the mind, since for bodily pleasure to be registered it needs to be sensed by the mind as well, as a consequence of which it is necessarily mixed. Since one could be pleasurably affected without the right disposition by things constitutive of kinetic pleasures. Cf. Tsouna 2007, p. 39 on the nature of Epicurean emotion based on her analysis of Philodemus: ‘We might put it this way: one’s disposition is the necessary antecedent cause of an emotion, whereas some external event is the auxiliary cause of a particular occurrence of it’. Also cf. Annas 1989 and 1992, pp. 189–99. For the textual evidence of  for the Epicureans as a condition of friendship and how it essentially appears in the context of a community see Philodemus’ evidence quoted later, as well as in Konstan et al. 1998 Fr. 31; Fr. 74; Col.Va.; Col. Xb; Col. XIb; Col. XVIIb. Also cf. Plutarch Non Poss. 1097C = fr. 183 Us. There are various texts expressing the importance of goodwill and sympathy in general for the Epicureans; cf. PHerc. 1418 col. XII 11 and XIV 5, Spina 1977; Metrodorus, fr. 38, Körte 1890; Diog. Oin. Fr. 127 I 2–3, Smith 1993; Porphyry, On Abstinence 52.4, Bouffartigue and Patillon 1977. For a parallel passage for the usage of  by Epicurus cf. Ep. Pyth. 84. Contra O’Connor 1989 and Konstan 1996, who both consider Epicurean friendship to be based on some kind of loose relation, as a kind of fellowship. This is not to say that looser, fellowship-like relations were impossible for the Epicureans but they did not qualify for different reasons as friendships, (cf. e.g. S. V. 39), they just existed as legitimate social relations within their community.

The pleasures of friendship

187

25 For the most recent survey of this evidence see Armstrong 2011 and cf. p. 108 where Armstrong argues that Torquatus’ account of Epicurean friendship in De Fin. is not complete. 26 For the modern concept of shared identity cf. Bennett 2009. Although the evidence could belong to the group of the more timid Epicureans at the time of Cicero, Philodemus was not necessarily one of them. Philodemus – who was a former pupil of both Demetrius of Laconia and Zeno of Sidon, the scholarch of the Epicurean school in Athens until 75 BCE – may have distinguished himself from the circle of Siro in Naples. Both are referred to as Epicurean authorities at the end of Cicero’s criticism of the Epicureans in De Fin. II 119. Philodemus’ evidence fits my interpretation well, only deviating from it in minor respects. On Frank Criticism was actually a transcript of lecture notes he took at Zeno’s classes in Athens, so it appears to be a testimony of an older pedigree; cf. Konstan et al. 1998, pp. 1–2. 27 For how the wise man relates himself to the method of frank criticism cf. Fr. 47 and Fr. 48. In general cf. Tsouna 2007, pp. 91–125. 28 Cf. Konstan et al. 1998, p. 5 for textual references. 29 Konstan et al. 1998, p. 85. For Greek text see ibid. p. 84. Sigla used in their translations: < > text based on the disegni as inspected by Philippson and others and compared with Olivieri’s text; [ ] Olivieri’s supplements; < [ ] > supplements suggested by Philippson or Gigante; {[ ]}translators’ supplements; { } translators’ additions or clarifications; ( ) parentheses in Olivieri’s text; ? indicates grave doubt about a restoration; italics indicates section heading in the text. 30 See Tsouna 2007, p. 94. Also D. L. X 11. 31 Konstan et al. 1998, p. 55. For Greek text see ibid. p. 54. 32 Translation from Armstrong 2011, p. 126. Greek text in Konstan et al. 1998, p. 44. Their translations (ibid. p. 45) differ in that they take the speaker and the listener to have different subjects, while Armstrong identifies them, thus understanding the text as describing two people in dialogue. 33 Most likely   described the teacher–pupil relation primarily, however not necessarily exclusively as it will become clear from Fr. 43, quoted later. Nonetheless, if we put emphasis on the nature of the relation, i.e. that it is a friendship, the participants’ social status becomes irrelevant. 34 From fragments 87 and 83; translation from Armstrong 2011, p. 126, who quotes the passage more extensively. For the Greek and other translations cf. Diels 1917 and Essler 2011. 35 Text and translation from Clay 1986, repr. in Clay 1998, pp. 80–2, end of his translation slightly modified. 36 Since Vogliano’s emendation has been challenged by Tepedino Guerra’s reading it can be safely established that the first lines refer to the gods; cf. Clay 1998, p. 84. While ‘all those who are well disposed to us’ cannot be the gods because the Epicurean divinities do not care about us, since that is incompatible with their nature (K. D. 1). 37 For the various other cults of the school, cf. Clay 1998, ch. 5. 38 Cf. Longo Auricchio, F. 1978. 39 For the problems of recruitment cf. Frischer 1982, pp. 46–52. 40 As Clay has suggested based on Plutarch’s evidence (Non Poss. 1097E) memorial pamphlets such as Epicurus’ Metrodorus and Neocles or Carneiscus’ Philistas (cf. Capasso 1988 and Konstan 1997, pp. 109–10) were read during such commemorations in order to hold up to the worshipper an ideal for the conduct of his or her life; cf. Clay 1998, p. 87 and p. 96 – although in his ‘Individual and Community . . .’ Clay also notes that Metrodorus could not have been a mere pamphlet since it consisted

188 The pleasures of friendship

41 42 43 44 45 46 47

of five rolls (D. L. X 28), in Clay 1998, p. 64. And Clay has also drawn attention to the evidence of Philodemus (AP XI 44), which implies that a discourse also took place at such occasions; cf. Clay 1998, pp. 89–90 and p. 99. Cf. Asmis 1990, p. 2395 n. 60. Cf. also Fr. 49, and Tsouna 2007, p. 94 for Fr. 49. Konstan et al. 1998, p. 57. For Greek text see ibid. p. 56. Cf. Fr. 56 as well as Tsouna 2007, pp. 99–101. Konstan et al. 1998, p. 57. For Greek text see ibid. p. 56. For the explanation of how open-ended such a relation can be in both ways cf. Tsouna 2007, pp. 114–15. As well as the fact that one cannot have and does not even need too many friends to be happy. It is just Aristotle’s ‘teleological bias’ as Cooper has put it (Cooper 1980, p. 308) that makes him always consider the best and most fully realized case. Translation from Pakaluk 1998, p. 39, with minor corrections.

Bibliography Annas, J. (1989) ‘Epicurean Emotions’, Greek, Roman and Byzantine Studies 30: 145–64. —. (1992) Hellenistic Philosophy of Mind, Berkeley. Annas, J. and Betegh, G. eds., (2016) Cicero’s De Finibus, Cambridge. Annas, J., ed., Woolf, R. transl. (2001) Cicero: On Moral Ends, Cambridge. Armstrong, D. (2011) ‘EpicureanVirtues, Epicurean Friendship: Cicero vs. the Herculaneum Papyri’, in J. Fish and K. R. Sanders, eds., Epicurus and the Epicurean Tradition, Cambridge: 105–28. Asmis, E. (1990) ‘Philodemus’ Epicureanism’, Aufstieg und Niedergang der Römischen Welt II 36: 42369–406. Bennett, W. H. (2009) Love, Friendship and the Self, Oxford. Bollack, J. and Laks, A. eds., (1976) Études sur l’épicurisme antique, Lille. Bouffartigue, J. and Patillon, M., eds., (1977) De l’abstinence / Porphyre, Paris. Brown, E. (2002) ‘Epicurus on the Value of Friendship (Sententia Vaticana 23)’, Classical Philology 97: 68–80. Capasso, M., ed., (1988) Carneisco: Il secondo libro del Filista (PHerc. 1027), Naples. Clay, D. (1986) ‘The Cults of Epicurus’, Cronache Ercolanesi 16: 11–28; repr. in Clay 1998: 75–102. —. (1998) Paradosis and Survival, Ann Arbor, MI. Cooper, J. M. (1980) ‘Aristotle on Friendship’, Essays on Aristotle’s Ethics, in A. O. Rorty, ed., Essays on Aristotle’s Ethics, London: 301–40. Diels, H. (1917) Philodemos über die Götter: Drittes Buch, Berlin. Essler, H. (2011) Glückselig und unsterblich: Epikureische Theologie bei Cicero und Philodem (mit einer Edition von Pherc. 152/157, Kol. 8–10). Schwabe Epicurea, 2. Evans, M. (2004) ‘Can Epicureans Be Friends?’, Ancient Philosophy 24: 407–24. Fish, J. and Sanders, K. R., eds., (2011) Epicurus and the Epicurean Tradition, Cambridge. Frede, D. (2016) ‘Epicurus on the Importance of Friendship in the Good Life (De Finibus 1. 65–70; 2. 78–85)’ in J. Annas and G. Betegh, eds., Cicero’s De Finibus, Cambridge: 96–117. Frischer, B. (1982) The Sculpted Word, Berkeley, CA. Konstan, D. (1996) ‘Friendship from Epicurus to Philodemus’, in G. Giannantoni and G. Gigante, eds., Epicureismo greco e romano, Naples, 3 vols., vol. I: 387–96. —. (1997) Friendship in the Classical World, Cambridge. Konstan, D., Clay, D., Glad, C. E., Thom, J. C. and Ware, J., eds., (1998) Philodemus: On Frank Criticism. Introduction,Translation and Notes, Atlanta.

The pleasures of friendship

189

Körte, A. (1890) ‘Metrodori Epicurei fragmenta,’ Jahrbücher für Classische Philologie Suppl. 17: 531–97. Long, A. A. and Sedley, D. N. (1987) The Hellenistic Philosophers, 2 vols., Cambridge. Longo Auricchio, F. (1978) ‘La scuola di Epicuro’ Cronache Ercolansesi 8: 21–37. O’Connor, D. (1989) ‘The Invulnerable Pleasures of Epicurean Friendship’, Greek, Roman, and Byzantine Studies 30: 165–86. O’Keefe, T. (2001) ‘Is Epicurean Friendship Altruistic?’ Apeiron 34: 269–305. Pakaluk, M. (1998) Aristotle: Nicomachean Ethics, Books VIII and IX, Oxford. Rorty, A. O., ed., (1980) Essays on Aristotle’s Ethics, London. Sedley, D. N. (1976) ‘Epicurus and his Professional Rivals’, in J. Bollack and A. Laks, eds., Études sur l’Epicurisme antique, Lille: 119–59. Smith, M. F. (1993) Diogenes of Oinoanda:The Epicurean Inscription, Naples. Spina, L. (1977) ‘Il trattato di Filodemo su Epicuro e altri (PHerc. 1418)’, Cronache Ercolanesi 7: 43–83. Tsouna,V. (2007) The Ethics of Philodemus, Oxford. Warren, J. (2016) ‘Epicurean Pleasure in Cicero’s De Finibus’ in J. Annas and G. Betegh, eds., Cicero’s De Finibus, Cambridge: 41–76. Wolfsdorf, D. (2013) Pleasure in Ancient Greek Philosophy, Cambridge.

Epilogue

The noble soul occupies itself with wisdom and friendship: of these the one is a mortal good, the other immortal. (S.V. 78) Diogenes Laertius reports (D. L. X 3) that Epicurus encouraged his three brothers to join him, as a result of which they went together on a life-long journey of philosophizing, benefiting later not only a small community but a number of friends, so large that it ‘could hardly be counted by whole cities’ (D. L. X 9).1 This story also contains one of the essential elements of Epicurus’ philosophy, its communal aspect based on a strong, affective relationship. And it is perhaps no accident that Philodemus – from whose work, On Philosophers book X, the above story comes – draws attention to Epicurus’ strong personal influence on his brothers. It seems to suggest a natural intellectual supremacy capable of reanimating organic bonds through philosophy and turning them into mutual benevolence (D. L. X 10) supported by theoretical reasoning. Even if later authors were ready to mock the original characteristic of this association by describing Neocles, Epicurus’ oldest brother, who emphasized the wisdom of Epicurus from childhood, and related that their mother, when pregnant with Epicurus, had atoms in her belly of a kind which inevitably produced a sage (Plutarch, Non Poss. 1100A–B), the original intent stuck: Epicurus’ spiritual presence and its religious overtones were later developed as a separate subject by both parties, followers and opponents.2 We have evidence that Epicurus even played on his blessed nature, and even if we understand this fragment as Epicurus’ mere preference to speak as an oracle rather than a demagogue,3 it invites us to consider the aspects of the Epicurean apotheosis,  the highest possible outcome of his ethical theory. In this fragment, in which he appears to be speaking himself, we read:                    

Epilogue

191

[Since] I would rather prefer to speak frankly in investigating nature and to give oracles of those things, which are useful to all men even if no one would understand me, than to conform to popular opinions and so win the swift praise given by the crowd. (S.V. 29)4 Epicurus, by invoking a traditional element of religious practice, a priest or priestess being a medium between gods and people, giving some useful, holy advice, entertains, if only hypothetically, his own image of definite authority in natural philosophy as of an oracle.5 He compares himself to a prophet, who has the insights of the divine, yet who is not a divinity. He implies that, even from a human perspective, a divine understanding of the world is possible, not as the result of the madness of a priestess from Delphi or Dodona, or the application of the faculty of divination by the Sibyl, but as a result of sober investigation, using frank speech.6 Epicurus, thus, in maintaining the possibility of the possession of the divine perspective rejects conventional opinion and the praise of the crowd – possibly opposing some (philosophers) who enjoyed such popularity. His statement is clearly polemical and is also most likely a direct refusal of the traditional conception of divinities playing their roles in the creation and government of the cosmos. Epicurus infamously denied them these roles because such activities, in his conception, are incompatible with divine blessedness and imperishability (cf. Ep. Hdt. 76–7). As a consequence, he is in a delicate position in this fragment: he is a genuine philosopher to whom nothing is dearer than the truth, spoken frankly, and though only a human with a godlike vision he can clearly see that popular opinion is mistaken.7 In conclusion of my investigation concerning Epicurus’ conception of the self, I wish to investigate how Epicurus’ follower Lucretius developed the theme of his master’s spirituality and used it for his own metamorphosis. In the proem of book V of his De Rerum Natura, Lucretius asks if there is anyone in possession of such powerful inspiration that they could compose a poem worthy of Epicurus’ discoveries in natural science, since ‘a god he was, a god’ – deus ille fuit, deus (DRN V 8). Of course, having written more than 4,000 lines, the question is merely poetic at the beginning of the last third of his magnificent six-book poem, a chance for the poet to bathe in some divine light, despite asserting that there is no one of mortal birth in command of such a language which is capable of forming a carmen worthy of Epicurus’ deeds.Yet Lucretius’ poetry transforms Epicurus’ wisdom in such a way that it must be based on his own transformation through some kind of divine inspiration, or so he seems to claim in the same book in lines 110–12 where he writes that he is about to utter oracles ( fata) much more reliable than those of the Pythia. As a result, he inevitably and intentionally invokes his ancestry.8 His words, indeed, could be interpreted merely as referencing the Epicurean doctrine he is passing on, yet the first person singular strongly indicates the perfection of his own intellectual insight based on Epicurus’ teachings.9 It appears that Epicurus’ spiritual presence not only helped him to establish a community around himself, but that it

192 Epilogue

radiated through centuries, stimulating his followers to emulate him by appropriating his ideas to become like him. The desire to imitate a superior being was not an alien idea in antiquity, but it was largely connected with gods and not with human beings.The concept of  , of ‘becoming like god so far as it is possible’, was a common ethical ideal of the philosophical schools from Plato onwards, and it already appeared with different emphases within Plato’s dialogues: as a theme of achieving immortality in the Symposium (207e–9e), as the realization of moral ideas, taking god to be the canon in the Theaetetus (176b–c) or just simply focusing on the intellectual capacities of the soul longing for the intelligible domain in the Republic (611c–12a). In the Timaeus, however, Plato presents the study of nature as the method of homoiosis  theoi. By studying cosmology, we turn our intellect from the realm of becoming to that of being and, in consequence, we understand the parallel between the nature of the immortal part of the soul and the divine order of the cosmos. The investigation leads to happiness () because it helps to re-establish the original nature of the immortal part of the soul by way of contemplating the Platonic Forms. Since just as the gods themselves gain their divine status by seeing the world of the Forms (cf. Phaedr. 247d1–2), the immortal part of the soul has to observe them in order to achieve its divine-like status and its consequent moral perfection. Happiness () and the ideal of godlikeness are directly connected, as Plato points out, by the etymology of the word , explaining that the man who always takes care of that which is divine has his daimon () well () ordered and, thus, he will be supremely happy.10 Another two aspects of the Platonic notion of homoiosis  theoi have also proved to be influential most prominently in its Epicurean conception, as Michael Erler has pointed out in his seminal paper.11 These originate from Plato’s discussion of the subject in his Laws (715e–17a), a passage in which the focus switches back to moral practice, not least because of the context. Two ideas were prominent: god being the true measure of all things, a measure to be imposed on one’s entire soul in order to assimilate oneself to god; and proper religious behaviour, in which not one’s performance of the ritual but one’s right opinion, one’s character, is important for the possibility of assimilation. Now Epicurus’ self-referential apotheosis  (S. V. 29) implies that right opinion – as in Plato – is the result of the proper investigation and understanding of nature leading to his divine perspective. Lucretius’ praise of his master and his parallel accomplishments are based on a view and perspective of nature that comes from the appropriation of Epicurean theory. That is to say, as for Plato, the study of nature was the primary means of homoiosis  theoi, it also had divine consequences for the Epicureans. Unlike Plato, who recognized divine teleology in the structure of the cosmos and who believed that the assimilation of the immortal part of the soul to god’s intentions is the way to moral perfection, Epicurus attained his ethical ideal by grasping and applying the limits of nature and, consequently, perfecting his mortal self by imposing on himself those

Epilogue

193

limits and measures, instead of trying to transcend his mortality. Thus, he managed to achieve a divine-like existence only temporarily.12 So it seems that for Lucretius to attain ‘divine delight’ (divina voluptas DRN III 28), it would have sufficed to understand and impose the same limits of nature on himself as well. Yet his emulation of Epicurus, and not that of the gods, indicates a kind of commitment, which is essential to his own voluptas, a sort of extra and indispensable supplement. In the proem of book III, he writes: O you who first amid so great a darkness were able to raise aloft a light so clear, illumining the blessings of life, you I follow, O glory of the Grecian race, and now on the marks you have left I plant my own footsteps firm, not so much desiring to be your rival, as for love, because I yearn to imitate you: for why should a swallow vie with swans, or what could a kid with its shaking limbs do in running to match himself with the strong horse’s vigour? You are our father, the discoverer of truths, you supply us with a father’s precepts, from your pages, illustrious man, as bees in the flowery glades sip all the sweets, so we likewise feed on all your golden words, your words of gold, ever most worthy of life eternal. For as soon as your reasoning begins to proclaim the nature of things revealed by your divine mind, away flee the mind’s terrors, the walls of the world open out, I see action going on throughout the whole void. (DRN III 1–17)13 Lucretius not only follows in Epicurus’ footsteps but desires to imitate him as well. Where does his desire stem from? As he explicitly states, not from rivalry but from love. He longs to imitate his father, in fact the father of all the Epicureans – hence the plural in line 9 – out of love, a love, which by the simile of comparing a kid to a strong horse invokes the image of a familial relationship, and indirectly that of friendship. As it is commonly known, the Greek term for friendship, philia (), had a much wider scope in meaning compared to our general conception of friendship, denoting any affective bond between individuals, whether family members, friends or fellow residents.14 I want to suggest that even though the Latin word for friendship, amicitia, is not present in this part of Lucretius’ text, his depiction of a familial relationship between a father and son, and their mutual love – the saviour imagery of lines 1–2 and lines 14–17 implies mutuality – indirectly bring out an important element and aspect of Epicurean friendship, which can be described as an aspiration for shared identity, consisting in a similar cognitive disposition. Obviously, Lucretius’ case is delicate and special in countless ways but, even so, in his declamation to Epicurus it is easy to detect an echo of all the characteristics of Epicurean love and friendship discussed in Chapter 5. Although Lucretius was a poet about whose life and affiliations we know next to nothing, and thus at first it seems hard to credit him with practising his Epicureanism in a social context such as the Epicurean Garden once seems to have been, by granting a place to himself in his poem among Epicurus’ children,

194 Epilogue

Lucretius offers a literary solution to find himself in a net of relations, which are brought back to life through his poetry. I am not suggesting that these relations appear as some kind of interactive ghosts or memories of former Epicurean friendships, but as Lucretius explicitly puts it, ideas ‘worthy of life eternal’ are available from the pages of his father’s ‘precepts’, which are supplied by Epicurus’ benevolence: tenets which are to be sipped like some excellent sort of intellectual nectar. His particularly Epicurean social context emerges through his intellectual engagement with the Master by transforming his recorded teachings in a continuous interaction between Lucretius, the pupil, and Epicurus, the teacher. We do not know the details of this metamorphosis, but the poem itself is the proof of his successful conversion.That this interaction can be interpreted along the lines of friendship having a mutual, positive and rational influence on the way of life of the other is justified, on the one hand, by Lucretius’ promotion of Epicurean values, which keep Epicurean ideas alive, and on the other, through the transformation of Lucretius’ own emotional experiences and his own moral judgements as, for example, can be seen in his views on sexual pleasures at the end of the book IV, or in his love towards Epicurus. Lucretius’ studies are fuelled by a child’s love, that of someone who wishes to imitate his father (‘propter amorem quod te imitari aveo’, DRN III 5–6), and although granting superiority to Epicurus, he seems to accomplish his emulation rather well: he himself achieves the status of an oracle and, thus, eventually shares his cognitive disposition with Epicurus. As children grow and become parental figures, so does Lucretius become an authority himself, a physician and an oracle through his poem, educating and healing no less than the Roman populace. Just how well his own transformation has turned out through emulating Epicurus becomes clear from the proem of book IV: A pathless country of the Pierides I traverse, where no other foot has ever trod. I love to approach virgin springs, and there to drink; I love to pluck new flowers, and to seek an illustrious chaplet for my head from fields whence before this the Muses have crowned the brows of none: first because my teaching is of high matters, and I proceed to set free the mind from the close knots of superstition; next because the subject is so dark and the verses I write so clear, touching every part with the Muses’ grace. (DRN IV 1–9)15 Lucretius continues with his famous simile: just like a physician who sweetens the rim of a cup with honey so that children drink the bitter medicine without noticing and regain their health, so he chose to set forth the Epicurean doctrine in sweet-speaking Pierian song.16 Lucretius’ transformation is superb: he does not plant his footsteps in the traces left by Epicurus anymore, but he explores virgin territories. Obviously, the emphasis is primarily on his poetic achievements but this part of the proem to book IV is in sharp contrast with the proem of book V, in which, as I have already claimed, his modesty is only artificial. In the proem of book IV,

Epilogue

195

Lucretius is not an Epicurean pupil but the teacher, the healer and, not least, the friend, who by addressing his poem to his patron and friend Memmius (‘Memmiadae nostro’ I 25) wishes to engage Memmius’ mind in his verses, so that Memmius can ‘understand the whole nature of things and perceive its utility’.17 So Lucretius’ desire for imitation stems from a filial or friendly love as was the case between Epicurus and his brothers and among the members of those communities they gathered around themselves. As Cicero’s Epicurean speaker Torquatus put it: Epicurus, however, in a single household, and one of slender means at that, maintained a whole host of friends, united by a wonderful bond of affection (amoris conspiratione consentientis). And this is still a feature of present-day Epicureanism. (De Fin. I 65)18 It is not hard to imagine that Epicurus’ spiritual presence was appealing to his immediate followers and that his philosophy offering a holistic explanation of the universe by the most economical number of principles – the atoms and the void – attracted more and more adherents in the ancient world. But the question about why they chose to imitate Epicurus instead of the gods themselves is still open. I think the solution to this issue is connected with Epicurus’ metaphysics and the answer is simply the lack of any cosmic reason or teleology in his cosmology. It is not the Platonic benevolent and divine demiurge who arranges an anthropocentric cosmos in the best possible order or, for that matter, the Stoic Zeus you affiliate with, but the discoverer of truth, Epicurus. It is not the divine sphere of the world or the recognition of some divine immanent principle or the anamne s is of Forms, but Epicurus where your knowledge of the nonteleological explanation of the universe comes from. Natural phenomena do not exhibit divine orderliness according to him but testify to the complete opposite, the corruptness of our cosmos (DRN V 100–234). And the recognition of that fact itself is indeed something worth appropriating because it resolves the worry about divine control, intervention and retribution, or in short, it does away with the fear of the gods and leads to a kind of tranquillity, which is akin to divine blessedness. Yet one may object that Epicurus seems to have founded his ethical ideal on his conception of the divine, which is blessed and immortal and knows no trouble itself nor causes trouble to any other (K. D. 1) when he claimed that: 

The man who is tranquil causes no disturbance to himself or to another. (S.V. 79)

196 Epilogue

And consequently the question still seems to remain: why should we imitate him who managed to achieve a temporarily divine condition instead of trying to simply affiliate with him and under his direction and influence try to reach a cognitive state similar to that of the gods themselves? What does Epicurus’ divine-like condition imply for which reason he is to be imitated and not the divinities themselves? Consider the anthropomorphic Epicurean gods living outside our world:19 other than knowing that they speak Greek and that they are free of every trouble, we have no explicit information about what they are doing. We just know that they are blessed and everlasting. It sounded to some rather boring for eternity.20 Is ‘torpid idleness’ something people are meant to imitate? No, of course not. Epicurus was the discoverer, he claimed, of the perfect human condition, which implies a kind of blessedness only similar to that of the gods, whose conditions are everlasting regardless of the different phases of mankind’s history. Thus it is not so much some divine activity (e.g. contemplation) the Epicureans imitated, but that they rather used contemplation – contemplating Epicurus’ teachings – as a means to achieve human tranquillity, which is only similar to that of the gods. That is to say, the Epicureans attained their blessedness by imposing on themselves the proper limits and measures of nature – which Epicurus had already enjoyed (cf. S. V. 33 and fr. 602 Us.) – through practising Epicurus’ philosophical arguments together (cf. Ep. Men. 135). It seems philosophy only had an instrumental value for them, in harmony with Epicurus’ claim that practical wisdom (phrone s is) is even more valuable than philosophy itself (Ep. Men. 132). And their activity was essentially the imitation of Epicurus and not that of the gods because it was through the praxis of his philosophy that they could achieve their ideal of godlikeness, being aware of their own mortality and their different conditions.21 As Lucretius’ example has shown – who historically did not have a personal connection to Epicurus – imitation offered the possibility of aligning oneself with Epicurus through the promotion and internalization of his ideas: something open to anyone after Epicurus’ death. In light of Philodemus’ evidence On Epicurus (PHerc. 1232), Lucretius could have thought of himself as maintaining his affectionate loving towards Epicurus by becoming like him, fully aware of the theoretical requirements of such an identity and drawing on the pleasurable consequences of ataraxia and aponia.22 Epicurus famously proclaimed: Empty is that philosopher’s argument by which no human suffering is therapeutically treated. For just as there is no use in a medical art that does not cast out the sickness of bodies, so too there is no use in philosophy, if it does not throw out suffering from the soul. (Porphyry Ad Marc. 31 = fr. 221 Us.)23 His philosophical arguments were used for mapping out the natural human condition and served no other function than helping to reconstitute that

Epilogue

197

condition once corrupted. Even the privative alpha in the conceptions of the goal or end of Epicurus’ ethics, a-ponia (), freedom from bodily pain, and a-taraxia, (), freedom from mental disturbance, imply a sort of recovery from an unnatural state of being, thus, building his ethical theory all around a therapeutic imagery based on the natural end (telos) of humankind. This recovery required the Epicurean conception of the self. As I have argued in this book Epicurus’ notion of the self was primarily the product of self-reflective thinking. According to his theory we gradually become aware of ourselves through our experience of the external world; we see and conceptualize ourselves in relation to others and in relation to our immediate natural, social and cultural environment. As we grow up and develop our rational capacities we have a conception of being causally efficacious agents, an abstraction of our human selves as morally responsible, based on the narrative we give to our memories and to our immediate observations. The way we conceive of ourselves is consequently the product of our own self-reflective thinking. This narrative is not the creation of a single life by an already present self as in Plutarch (cf. Chapter 3). Instead, one’s rational narrative or weaving creates one’s conception of his or her single self. This is a ‘thin idea’ of the self, which unifies the various subjects, one’s essential and accidental characteristics, of one’s self-reflective thinking. The philosophical analysis Epicurus conducted of our self-conception seems to have been a complex account in harmony with his three criteria of truth: sensation, prole p seis, strongly related to his aitiologikos tropos or causal manner of investigation, and affections naturally connected to his pathologikos tropos or the investigation according to one’s affections. As the textual analysis of On Nature XXV has shown, Epicurus’ exposition of self-reflective thinking was embedded in his atomist account of one’s occurrent mental states ( ). The state of our evidence is not good enough to be decisive in itself about whether Epicurus had a reductionist or anti-reductionist philosophy of mind, the major debate revolving around these fragments. However, if we read the surviving fragments coupled with the background assumption of the atomic swerve and the idea of multiple realizability attested by Lucretius, they arguably substantiate a non-reductive physicalist theory of mind, one which is also compatible with Lucretius’ account. Thus it was of primary importance for Epicurus to introduce the atomic swerve, which not only helped him to refute causal determinism, but it lent great support to his ethics, advertising Epicurus’ ideal of the self.

Notes 1 Cf. Philodemus De Piet. Vol. Herc. 2, II 107 (in Obbink 1996, I. 33. 944) for Neocles being an advanced student of Epicurus. Also cf. Plutarch De Frat. Am. 487D = fr. 178 Us. for the brothers’ respect. 2 Cf. fr. 141 Us. as well as fr. 602 Us. For a good collection of the evidence and an introduction on the subject of the cults in the Garden cf. Clay 1986, reprinted in Clay 1998.

198 Epilogue 3 Cf. Warren 2002, pp. 183–8. 4 The fragment seems to be taken out of its context because of the explicative conjunction at the beginning of the sentence. Also cf. Usener’s Epicurea pp. xlv–vi where he claims that many of Epicurus’ Principal Doctrines are extractions from certain letters. 5 Also cf. Philodemus De Piet., col. 71, in Obbink 1996, pp. 246–7. Contra Obbink 1996, p. 568 and Sedley 1998, p. 13, n. 59, who interpret both Greek verbs expressing the act of giving an oracle by Epicurus and Philodemus as ironical – an interpretation which is also possible: I think this evidence still teaches us much about the Epicurean conception of apotheosis.  6 Cf. Plato’s Phaedr. 244a8–b5 for divine authorities and a similar use of language. On frank speech see Chapter 5. 7 For the refusal of popular opinion also cf. Ep. Men. 123–4, as well as PHerc. 1232 Fr. 8, quoted earlier. For the refusal of popularity cf. fr. 187 Us. 8 ‘Qua prius adgrediar quam de re fundere fata / sanctius et multo certa ratione magis quam / Pythia quae tripode a Phoebi lauroque profatur, / multa tibi expediam doctis solacia dictis’ (DRN V 110–14). Lines 111–12 already appear in book I 738–9 in a different context, where Lucretius compares Empedocles, his literary forebear, to Empedocles’ predecessors; cf. Sedley 1998, pp. 12–14. 9 The fact that he goes on to explain that it is not impious to believe that the world is not divine (DRN V 110–234) corroborates my understanding of Epicurus’ parallel passage (S.V. 29) and the possible interpretation of its missing context, as Philodemus’ evidence does, ibid. in n. 5. For Lucretius taking Epicurus seriously as an oracle also cf. DRN VI 6. 10 See Tim. 90c. Sextus uses the same etymology in criticizing the Epicureans for attributing divine happiness to man, cf. Adv. Math. IX 43–7. 11 Erler 2002, pp. 165–7. I will not address the entire history of the subject here, yet it is worth noting the difference in emphasis between the schools at least by the time of Epicurus: Aristotle and the Peripatetics emphasized the contemplative way of life, thus downplaying the importance of moral virtue (Sedley 1999). Stoics contemporary with Epicurus understood by assimilation to god the fulfilment of moral duties and obedience to virtue, on an extremely ideal level which no one but perhaps Cato alone, their Roman ideal sage, was capable of fulfilling. 12 The temporal limitedness of his divine status is also indicated by the past tense of Lucretius’ own words (‘deus ille fuit, deus’, DRN V 8) – already pointed out by Warren 2000, pp. 235–6 and Erler 2002, p. 162. For a thorough comparison of the Platonic and Epicurean doctrines cf. Erler 2002. 13 O tenebris tantis tam clarum extollere lumen / qui primus potuisti inlustrans commoda vitae, / te sequor, o Graiae gentis decus, inque tuis nunc / ficta pedum pono pressis vestigia signis, / non ita certandi cupidus quam propter amorem (5) / quod te imitari aveo: quid enim contendat hirundo / cycnis, aut quidnam tremulis facere artubus haedi /consimile in cursu possint et fortis equis vis? / tu pater es, rerum inventor, tu patria nobis / suppeditas praecepta, tuisque ex, inculte, chartis, (10) / floriferis ut apes in saltibus omnia libant, / omnia nos itidem depascimur aurea dicta, / aurea, perpetua semper dignissima vita. / nam simul ac ratio tua coepit vociferari / naturam rerum, divina mente coortam, (15) / diffugiunt animi terrores, moenia mundi / discedunt, totum video per inane geri res (DRN III 1–17). Translation from Rouse and Smith 1975 and Smith 2001. 14 Cf. Konstan 1996, and Konstan 1997 more generally. 15 Avia Pieridum peragro loca nullius ante / trita solo. iuvat integros accedere fontis / atque haurire, iuvatque novos decerpere flores / insignemque meo capiti petere inde coronam / unde prius nulli velarint tempora Musae: (5) / primum quod magnis

Epilogue

16

17 18 19 20 21

22

199

doceo de rebus et artis / religionum animum nodis exsolvere pergo, / deinde quod obscura de re tam lucida pango / carmina, musaeo contingens cuncta lepore (DRN IV 1–9). Translation from Rouse and Smith 1975. ‘volui tibi suaviloquenti / carmine Pierio rationem exponere nostram’ (DRN IV 20–1). Notice the plural nostram, indicating the full comprehension and appropriation of the Epicurean doctrines by Lucretius, hence himself being a proper Epicurean authority. But the focus of my interest lies rather in the poet’s metamorphosis through the emulation of Epicurus, since just as the poetic imagery of his proems, the therapeutic aspects of Epicureanism have already been discussed in many excellent studies, cf. e.g. Gale 1994 and Nussbaum 1994. ‘percipis omnem / naturam rerum ac persentis utilitatem’, DRN IV 24–5. Note how utilitatem echoes the usefulness of sincere natural studies in S.V. 29. Translation from Annas and Woolf 2001, p. 23. Cf. Eusebius Praeparatio Evangelica 15. 5. 11, 800B = fr. 362 Us. Cf. Cicero DND I 67, 102–4, 110 and 116. It would be possible to substantiate to some further extent the distinction between imitating Epicurus or the gods on the idealist interpretation of his theology (the hypothesis being that gods are only thought-constructs, the projections of ethical ideals; for its most current formulation cf. Sedley 2011; and for the opposite position, the realist understanding, according to which gods are real, existing beings living outside the cosmos cf. Konstan 2011), but for that a detailed discussion of his theology is necessary, which I do not wish to pursue here. Nonetheless, Lucretius’ solution is certainly devoid of real inter-personality, the foundation of self-knowledge. One’s moral development is dependent on knowing oneself through others, which aspect, even if it is not emphatically present in Philodemus’ evidence, is not merely compatible with but even implied by his ideas of friendship and the method of frank criticism, as we have seen in Chapter 5. However, if we grant that Lucretius’ solution was literary, we also need to accept a poetic solution of inter-personality, which seems to appear in the opening proem: Lucretius invokes Venus and asks her to be his partner in writing the verses on the nature of things, addressed to his friend, Memmius. Note that Venus as the goddess of love, thus, offers the perfect solution for friendly affections for Lucretius as well as the literary solution of inter-personality given her divine status, which helps Lucretius to overcome the lack of real inter-personality, cf. DRN I 21–8.

23                                 Translation from Nussbaum 1994, p. 102. The technical term, , is used in a narrow sense in this

passage, but as we have seen it also means ‘feeling’, ‘affection’ or ‘emotion’.

Bibliography Annas, J., ed., Woolf, R. transl. (2001) Cicero: On Moral Ends, Cambridge. Clay, D. (1986) ‘The Cults of Epicurus’, Cronache Ercolanesi 16: 11–28; repr. in Clay 1998: 75–102. —. (1998) Paradosis and Survival, Ann Arbor, MI. Erler, M. (2002) ‘Homoiosis theoi and Epicurean Self-Cultivation’, in D. Frede and A. Laks, eds., Traditions of Theology, Leiden: 159–82. Fish, J. and Sanders, K. R., eds., (2011) Epicurus and the Epicurean Tradition, Cambridge. Frede, D. and Laks, A., eds., (2002) Traditions of Theology, Leiden. Gale, M. (1994) Myth and Poetry in Lucretius, Cambridge.

200 Epilogue Giannantoni, G. and Gigante, M., eds., (1993) Epicureismo greco e romano: Atti del Congresso Internazionale Napoli, 19–26 Maggio 1993, 2 vols., Naples. Konstan, D. (1996) ‘Friendship from Epicurus to Philodemus’, in G. Giannantoni and G. Gigante, eds., Epicureismo greco e romano, 3 vols., Naples: vol. I: 387–96. —. (1997) Friendship in the Classical World, Cambridge. —. (2011) ‘Epicurus on the Gods’, in J. Fish and K. R. Sanders, eds., Epicurus and the Epicurean Tradition, Cambridge: 53–72. Nussbaum, M. C. (1994) The Therapy of Desire: Theory and Practice in Hellenistic Ethics, Princeton. Obbink, D. (1996) Philodemus: On Piety, Part I, Oxford. Rouse, W. H. D. and Smith, M. F. (1975) Lucretius, De Rerum Natura, Loeb edition, London and Cambridge, MA. Sedley, D. N. (1998) Lucretius and the Transformation of Greek Wisdom, Cambridge. —. (1999) ‘The Ideal of Godlikeness’, in G. Fine, ed., Plato 2: Ethics, Politics, Religion, and the Soul, Oxford: 309–28. —. (2011) ‘Epicurus’ Theological Innatism’, in J. Fish and K. R. Sanders, eds., Epicurus and the Epicurean Tradition, Cambridge: 29–52. Smith, M. F. (2001) Lucretius: On the Nature of Things, Indianapolis/Cambridge. Warren, J. (2000) ‘Epicurean Immortality’, Oxford Studies of Ancient Philosophy 18: 231–61. —. (2002) Epicurus and Democritean Ethics: An Archeology of Ataraxia, Cambridge.

Appendix

The order of the fragments of On Nature XXV There are three papyri of book XXV, but the catalogue numbers of the Neapolitan papyrus collection list them under six numbers. They can be grouped as PHerc. 419/1634/697, PHerc. 1420/1056 and PHerc. 1191. Laursen published all the fragments of PHerc. 419, PHerc. 1634 and PHerc. 1420 in Laursen 1995 and all the ‘legible’ fragments of the rest in Laursen 1995/1997. I think it is fair to say that Laursen was very generous since his 1995 publication contains a lot of incomprehensible texts. Therefore we can rest assured that we have access to all the comprehensible texts of book XXV. In my book I used and interpreted most of the fragments, and I wish to orientate the reader about their relative order of appearance compared to Laursen’s publications. Corn. is an abbreviation for the Italian word ‘cornici’ or ‘frames’ in English, which contain fragmentary ‘pezzi’ or ‘pieces’, the fragments themselves. Laursen 1995 provides a thorough and in-depth analysis of the papyrological evidence. By listing their order of appearance in Laursen 1995/1997 in an abbreviated way I tried to avoid giving the impression that one could find a lot more in Laursen’s publications. I did not address a few fragments including a few abrupt sentences, which, nevertheless, also seem to be compatible with my interpretation of Epicurus’ philosophy. PHerc. 1634 PHerc. 419 sc. fr. 12 – fr. 5 fr. 4 fr. 9 /no comprehensible text thus far/ fr. 3 fr. 2 fr. 1

fr. 8 fr. 7 fr. 6 fr. 5 /few fragmentary sentences altogether/ fr. 4 – fr. 1 (only a few words)

Fr. (a) (p. 25) (quoted on p. 79)

202 Appendix

PHerc. 1191 PHerc. 697 PHerc. 1420 PHerc. 1056 corn. 1–2

corn. 2

corn. 3 corn. 6

/no comprehensible text/ (only a few words) Fr. (b) (p. 38), Fr. (c) (p. 130, n. 33) Fr. (d) (p. 102, n. 31)

corn. 1 corn. 2

. . corn. 7 corn. 4 .

corn. 1 corn. 2 corn. 3

.

corn. 2

corn. 4

.

corn. 2

corn. 4

corn. 2

corn. 5

corn. 4/7

corn. 3 corn. 4/7 corn. 4/7 corn. 4/7 corn. 7/7 corn. 7/7 corn. 7/8

corn. 5 corn. 3 corn. 3 corn. 3 corn. 3 corn. 3

corn. 5 corn. 5 corn. 5

corn. 7/8 corn. 7/8 corn. 8 corn. 8

corn. 3 corn. 3 corn. 4

corn. 6 corn. 6 corn. 6

corn. 8 corn. 9/8

corn. 4 corn. 4

corn. 7 corn. 7

corn. 9/9 corn. 9/9 corn. 9

corn. 4 corn. 4

corn. 8 corn. 8 corn. 8

Fr. 1 (p. 11) Fr. 2 (p. 11), Fr. 3 (p. 13), Fr. 4 (p. 56, n. 56) Fr. 5 (p. 13), Fr. 6 (p. 59, n. 93), Fr. 7 (p. 51), Fr. 8 (p. 52) (a fragmentary sentence related to memory, p. 66, n. 192) Text M [= Fr. 9] (pp. 48–9) Fr. 10 (p. 76) Fr. 11 (p. 76) Fr. 12 (p. 76) Fr. 13 (p. 76) Fr. 14 (p. 76) – two fragmentary sentences on exhortation – Fr. 15 (p. 90) Fr. 16 (p. 109) Fr. 17 (p. 91) : Text P (pp. 45–6 and p. 113) : Text P (pp. 45–6) : Text P (pp. 45–6, 114, 129, n. 22, n. 28) Fr. 18 (p. 120) Fr. 19 (p. 126) (p. 55, n. 36)

Index

aitiologikos tropos xvi, xviii, 10, 24–7, 48, 50–52, 55n36, 70, 75, 113–14, 122, 127, 166, 197 akoustikon pathos 20–1, 23 Alexander of Aphrodisias 96, 184n4 animals 108–11, 129n12 Annas, J. 24–5, 129n14, 162n52 anonymous Epicurean 22 Aristotle xiv, 25, 36, 52, 59n97, 128n1, 129n29, 136–7, 158, 159n13, 162n52, 163n61, 168, 175–6, 182–4, 185n8, 188n46, 198n11; N. E. IX 1170b4–14: 184; Phys. VIII 252a32–b2: 136–7, 159n14 Armstrong, D. 178–9, 187n25 Asmis, E. 18–9, 34–6, 41, 64n172, 122–3, 180 Atherton, C. 153, 163n61 Avotins, I. 162n56 Bailey, C. 58n81, 161n35 Bobzien, S. 143–4, 152–3, 161n50 Brunschwig, J. 57n61 Carneades 154–7, 163n76 Cicero 4–6, 53n1, 62n143, 95, 122–3, 142, 152–6, 163n62, 166–173, 176, 186n16, 187n26, 195; De Fat. 23–4: 155, 164n77, De Fat. 31: 155–6, 164n78; De Fin. I 30: 54n14, De Fin. I 38: 58n85, De Fin. I 65: 195, De Fin. I 67–8: 169–70, 185n13 Clay, D. 179, 187n40 Cooper, J. M. 5, 54n13 cradle argument xi, 4–7, 54n14, 188n46 Cyrenaics 7, 41, 50

Democritus 73–5, 102n23 desires xiv, xviii, 1–7, 53n7, 112, 167–8, 170–1, 180, 184n7 78–80 Diogenes of Oinoanda 26, Fr. 9 II 9–VI 3: 28–30, 60n116; 39, 60n116, Fr. 54 133 D. L. X 32: 19, 35, 58n76, 62n144, D. L. X 33: 30, 61n120, D. L. X 135: 70 doxa orthe  30–1, 33–4 dreams/dreaming 28–30, 35, 59n88, 61n117 eidola  14, 25, 36, 38–9, 51, 62n151, 63n158, 85, 121–2 ennoia 30–1, 34–5, 62n141 epaisthe s is/sensory recognition 17–24, 32, 37–40, 58n79, 120–1, 126, 130n43 25, 29, 37, 39, 41, 61n117, 63n154, 63n165 Epicurus Ep. Hdt. 37: 135, 159n8; Ep. Hdt. 37–8: 122, 131n45; Ep. Hdt. 38: 21, 58n86; Ep. Hdt. 50–1: 39–40, 63n165; Ep. Hdt. 52–3: 17, 58n72; Ep. Hdt. 53: 58n82; Ep. Hdt. 61: 163n58; Ep. Hdt. 63: 105n70; Ep. Hdt. 72–3: 43–4; Ep. Hdt. 75: 14, 15, 57n60, 57n64; Ep. Men.122: 1; Ep. Men.127–8: 3, 53n9; Ep. Men.128–30: 4, 54n12; Ep. Men.132: 137, 184n3; K. D. 24: 40, K. D. 32: 108; On Nature XXVIII: 9, 60n108, 64n176, 122; S.V. 15: 166, S.V. 29: 190–1, S.V. 61: 175, S.V. 78: 190, S.V. 79: 195 epilogismos 13–14, 23–4, 111–13, 148 Erler, M. 192

204 Index eudaimonia/ xiii, 1–2, 8, 23, 126, 148, 166, 177, 183, 192 Everson, S. 31, 57n66 falsehood 39–43 Fine, G. 30, 44 Fowler, D. 138–9, 146, 162n56 friendship 166–184, 190–5 Furley, D. J. 64n179, 101n12, 104n67, 159n2, 161n35, 162n55–6 Glidden, D. K. 17–21, 31, 34–5, 61n122, 62n141 Hahmann, A. 28, 60n113 Hankinson, R. J. 143 Hayter, J. 9 hearing 17–23, 64n180 hedonic/rational calculus xv–xvi, xviii, 6, 23–4, 59n93, 111–12, 120–1, 124–7 Hermarchus 111, 148, 179 homoiosis  theoi 192–6 Idomeneus xvi katale psis  22, 30–1, 34, 62n137 Kleve, K. 151 Konstan, D. 55n24, 186n24, 199n21 language 15–16; semantic properties 31, pre-verbal 33 Laursen, S. 10–11, 13, 24, 50–1, 75, 80, 87, 89, 102n24, 109, 121, 201 Law of the Excluded Middle 152–3, 157, 163n62 Long, A. A. xiv, 61n120, 160n16 Lucretius DRN II 167–83: 139–40, 160n25; DRN II 221–4: 137, 160n16; DRN II 251–60: 141–2; DRN II 261–2: 144; DRN II 263–71: 144–5; DRN II 272–6: 146; DRN II 272–83: 146–7; DRN II 284–93: 148–9; DRN III 1–17: 193, 198n13; DRN III 168–76: 12; DRN 262–65: 96, 105n71; DRN III 285–7: 96; DRN III 282–7: 105n73; DRN III 843–61: 118–19; DRN IV 1–9: 194, 198n15; DRN IV 20–1: 199n16; DRN IV 24–5: 199n17; DRN IV 553–62: 18, 58n73; DRN IV 802–4: 63n154; DRN V 110–14: 198n8; DRN V 1028–32: 57n66;

DRN V 1056–8: 61n128; DRN VI 54–5: 160n27 Masi, G. F. 39, 50, 58n81, 60n111, 63n158, 66n192, 101n18, 127, 129n10 Maximus of Tyre 16 memory xvi, 19, 24–5, 30 32, 35–7, 48–53, 65n189, 80–1, 88–9, 102n31, 116–19, 130n33, 179 Metrodorus 65n184, 176, 179, 187n40 Mitsis, P. 6, 130n30 multiple realizability 94, 98, 100, 104n66, 105n75, 197 Nausiphanes 65n184, 86, 102n23 Nussbaum, M. 53n7 Obbink, D. 198n5 O’Connor, D. 186n24 O’Keefe,T. xii, 70–5, 79–83, 87, 102n24, 103n36, 103n38, 136–7, 153, 155, 170–1 palingenesis xvi, xixn11, 118–19 Pancheri, L. U. 163n56 pathologikos tropos xvi–xviii, 10, 14, 11–24, 48, 50–2, 57n63, 70, 75, 113, 122, 127, 197 phantasia 24, 33, 41, 43 Philodemus xvi, 8, 55n29, 58n79, 60n108, 62n152, 63n153, 63n163, 103n36, 128, 166–7, 176–83, 187n26, 190, 197n1, 198n5; De Sign. XXII 28: 37; On Epicurus (PHerc. 1232): 179; On Frank Criticism Fr. 28: 178, Fr. 42: 178, Fr. 43: 180–1, Fr. 44: 181, Fr. 55 (1–6): 181, Fr. 80: 177–8; On the Gods III Fr. 83: 179, 187n34, Fr. 87, 179, 187n34 phrone s is 23, 130n30, 166–8, 196 Piaggo, A. 8–9 Plato xiv, xvi, 25, 35, 44, 52, 102n26, 146, 186n19, 192, 195; First Alcibiades xiv, 25–6; Phaedo xvi pleasure xi, xiv–xvi, 2–7, 10, 16–7, 22–3, 25, 54n13, 55n20, 57n63, 58n85, 111–12, 115, 119, 125, 139, 142, 152, 160n17, 167–78, 180–3, 184n4, 184n6–7, 185n9–11, 186n19, 194 Plutarch 31, 41, 64n170, 64n172, 116–19, 122–3, 179, 197; De Tranq. Anim. 473B–4B 116–17

Index

205

Polyaenus 179 Polystratus 111, 148 Pope, M. 151–2, 163n59, 163n61 Porphyry Ad Marc. 31: 196, 199n23 Principle of Bivalence 152, 154, 157, 163n65 prole p sis 27–48 proton  ennoe m  a 65n181, 122–4 prudence (phrone s is) 23–4, 59n94, 130n30, 167–9, 196 Purinton, J. 79, 161n35

Sextus 31, 34–6, 40–1, 58n75, 60n108, 63n168, 64n169, 122–3, 198n10; Adv. Math. VII 211–12: 40, 63n168 Sharples, R. W. 103n38, 153–5, 163n65 Socrates xiv, 1–2, 25, 36 Sorabji, R. xiii, xvi, 59n103, 118–19 Steiger, K. 74 swerve xii, xviii, 55n36, 72, 92–8, 100, 104n67, 114, 127–8, 133–58, 159n5, 160n21, 162n53–4, 163n59, 163n61, 197

Remes, P. 66n199

 86–92

Saunders, T. J. 162n53 Sedley, D. N. xii–xiii, 9–10, 46, 55n36, 56n51, 59n93, 60n108, 61n120 63n154, 64n176, 65n181, 70–5, 81–4, 87, 92, 101n8, 101n18, 102n23, 102n24, 126–7, 129n29, 134, 143, 145, 159n5, 160n16, 160n30, 198n5, 199n21 self-knowledge xvii, 1–2, 8, 24–6, 52–3, 57n63, 116, 119, 166, 178, 182–3, 199n22 self-reflective thinking xii, xvi–xvii, 7, 9, 11–13, 17, 23–7, 47–8, 50, 52–3, 59n93, 70, 112–13, 115, 119, 122, 197

telos xiii–xv, 1, 3–5, 9, 23, 54n16, 169, 171, 176–7, 197 token monism 84, 93, 98, 100, 103n45, 158 Truth-to-Necessity argument 142, 153 Tsouna,V. 53n8, 130n30, 181, 186n21 type dualism 84, 93, 98, 100, 103n45, 158 typos/typoi/outline(s) 30–4, 37–9, 41–3, 45–7, 63n158, 63n163, 64n180, 65n186, 121, 135 Warren, J. xixn9, 65n184 weaving 110, 116–19, 130n33, 130n36, 197 Woolf, R. 54n13