David N. Sedley Lucretius and the Transformation of Greek Wisdom 1998

http://www.cambridge.org/0521570328

This page intentionally left blank

LUCRETIUS AND THE TRANSFORMATION OF

GREEK WISDOM

This book is designed to appeal both to those interested in Romanpoetry and to specialists in ancient philosophy. In it David Sedleyexplores Lucretius’ complex relationship with Greek culture, inparticular with Empedocles, whose poetry was the model for hisown, with Epicurus, the source of his philosophical inspiration, andwith the Greek language itself. He includes a detailed reconstruc-tion of Epicurus’ great treatise On nature, and seeks to show howLucretius worked with this as his sole philosophical source, butgradually emancipated himself from its structure, transforming itsraw contents into something radically new. By pursuing thesethemes, the book uncovers many unrecognised aspects ofLucretius’ methods and achievements as a poetic craftsman.

David Sedley is Professor of Ancient Philosophy, University ofCambridge, and Fellow of Christ’s College. He is the author, withA. A. Long, of The Hellenistic Philosophers (!"#$).

LUCRETIUS AND THETRANSFORMATION OF

GREEK WISDOM

DAVID SEDLEY

!"#$%&'() #* +'( !,(&& &*-)%./+( 01 +'( "-%2(,&%+* 01 ./3#,%)4(The Pitt Building, Trumpington Street, Cambridge, United Kingdom

./3#,%)4( "-%2(,&%+* !,(&&The Edinburgh Building, Cambridge CB2 2RU, UK40 West 20th Street, New York, NY 10011-4211, USA477 Williamstown Road, Port Melbourne, VIC 3207, AustraliaRuiz de Alarcón 13, 28014 Madrid, SpainDock House, The Waterfront, Cape Town 8001, South Africa

http://www.cambridge.org

First published in printed format

ISBN 0-521-57032-8 hardbackISBN 0-521-54214-6 paperback

ISBN 0-511-02083-X eBook

David N. Sedley 2004

1998

(netLibrary)

©

For Tony Long

Contents

Preface page xiIntroduction xv

! The Empedoclean opening !!. Cicero’s letter !%. Empedocles’ two poems %&. The provenance of Empedocles '!!( #). Lucretius and Empedocles !*(. The enigma of Lucretius’ proem !(+. Furley’s thesis !+$. Empedocles as literary forebear %!#. Empedocles’ proem &%". Conclusion &)

% Two languages, two worlds &(!. Linguistic poverty &(%. The technicalities of physics &(&. Atomic vocabulary &#). Simulacra &"(. Prose and verse contrasted )&+. Distorted reasoning )+$. The price of failure )##. Bailey’s complaint )"". Evoking Greece (*

!*. Magnets in Samothrace (%!!. The familiar and the exotic ()!%. The swallow and the swan ($!&. The resolution (#Appendix ("

& Lucretius the fundamentalist +%!. Philosophy in Italy +%%. The school of Philodemus +(&. The location of the mind +#

vii

). Philosophical opponents $&(. Creationism $(+. Geocentrism $#$. Contemporary Stoicism #%#. Scepticism #(". Berosus "*

!*. Fundamentalism "!

) Epicurus, On nature ")!. The discovery ")%. The papyri "#&. Favourite books? ""). Length !*%(. Style !*)+. Reconstructing the contents !*"$. Books ,–,- !!*#. Books -–. !!+". Books .,–.,,, !!"

!*. Books .,-–.- !%&!!. Books .-,–...-,, !%+!%. Chronology !%#!&. Conclusions !&%

( Lucretius’ plan and its execution !&)!. The thesis !&)%. Lucretius’ source !&(&. The structure of De rerum natura !))). Books ,–,,, !)((. Book ,- !)#+. Book - !(%$. Lucretius’ method !((#. Book -, !($

+ The imprint of Theophrastus !++!. Theophrastus and the world’s destructibility !++%. The fourth argument !+#&. The third argument !$)). The first and second arguments !$+(. The provenance of Theophrastus fr. !#) !$$+. Meteorology !$"$. Theophrastus, Epicurus and Lucretius !#%

$ The transformation of book , !#+!. The contents !#+%. DRN , and On nature ,–,, !#+&. Seeds !"&

viii Contents

). Divinity !"#(. Other changes !""+. Back to Empedocles %*!

Epilogue %*&

Bibliography %*(Index locorum %!$General index %%#Index of modern scholars %&%

Contents ix

Preface

This book is the partial repayment of a debt. It was my desire to under-stand Lucretius better that led me into postgraduate research onEpicureanism. And, even more than the philosophy component of myGreats course at Oxford, it was that postgraduate research onEpicureanism that emboldened me to pursue the study of ancient phi-losophy as a career. It would therefore be only a small exaggeration tosay that I learnt ancient philosophy in order to understand Lucretius.Until recently I have ventured little about Lucretius in print, but I havebeen thinking about him throughout my teaching career at Cambridge.This book is the outcome, and my way of thanking its eponymous hero.

My fascination with Lucretius was fuelled when as an Oxford under-graduate I had the good fortune, in !"++–$, to attend the wonderful lec-tures on Lucretius by the then Corpus Professor of Latin, Sir RogerMynors. Mynors told us that he had himself in his early days beenenthralled by Cyril Bailey’s Lucretius lectures, none of whose brilliance,he remarked, showed through into Bailey’s later monumental edition ofthe poet (‘He had gone o/ the boil’). I like to think that some excitementfrom the real Bailey filtered through to me in those lectures.

Another debt is to David Furley, whose book Two Studies in the GreekAtomists I came across in Blackwell’s while studying Aristotle for Greats.It was that book – which I bought for the then shocking sum of threepounds and nine shillings – that taught me not only how much interestAristotle gained when he was read alongside other philosophers from avery di/erent tradition, but also how much philosophical depth and sub-tlety there were to be found in Epicureanism, including that of Lucretiushimself.

There are two other friends I should also like especially to thank. Mycopy of Martin Ferguson Smith’s Loeb edition of Lucretius has finallyfallen to bits during the writing of this book, a tribute to the fact that Irely on it at all times. His pioneering work on Diogenes of Oenoanda

xi

has also been a constant inspiration to me in my own studies ofEpicureanism. And Diskin Clay, with his book Lucretius and Epicurus, hasset a dauntingly high standard for anyone hoping to shed new light onLucretius’ poetry through the study of ancient philosophy. My book maybe very di/erent from his, but I have been constantly conscious ofLucretius and Epicurus as a model.

Many of my Cambridge colleagues, past and present – especially theancient philosophers Myles Burnyeat, Geo/rey Lloyd, MalcolmSchofield, Nick Denyer, Robert Wardy, Mary Margaret McCabe andDominic Scott, but also Ted Kenney and others – have engaged with mein debate about Lucretius at various times; and we have had three splen-did Lucretius seminars. I have also learnt much from my students, espe-cially from James Warren, with whom I have discussed Lucretian issueson many occasions.

Much of the background to this book lies in the Herculaneum papyri.In the nine months of !"$! I spent in Naples working on these uniquelydi0cult but rewarding texts, and during numerous return visits there-after, I benefited from the help and hospitality of many, notablyMarcello Gigante, Francesca Longo Auricchio, Albert Henrichs, AdeleTepedino Guerra, Giovanni Indelli, Gioia Rispoli, Salvatore Cerasuoloand Tiziano Dorandi. All of them, and others too numerous to mention,I thank warmly.

Those who have commented on earlier drafts of the material thatfound its way into this book include Jim Adams, Han Baltussen, CharlesBrittain, Myles Burnyeat, Diskin Clay, Tiziano Dorandi, Don Fowler,Bill Furley, David Furley, Monica Gale, Philip Hardie, Ted Kenney,Mieke Koenen, Geo/rey Lloyd, Jaap Mansfeld, Roland Mayer,Catherine Osborne, Michael Reeve, David Runia, Samuel Scolnicov,Bob Sharples, Martin Smith, Voula Tsouna, Paul Vander Waerdt,Richard Wallace, Robert Wardy and David West. The penultimate draftof the entire book was read and commented on by Myles Burnyeat,Tony Long, Tom Rosenmeyer, Malcolm Schofield, Gisela Striker, VoulaTsouna and Robert Wardy. My warm thanks to all of these, and to manyothers who have contributed to discussion at various stages. Likewise toaudiences who have responded to presentations of various parts of thebook’s thesis: at Berkeley, UCLA, Stanford University, CornellUniversity, the University of Wales, Duke University, the BibliothecaClassica at St Petersburg, the Institute of Classical Studies in London,the Oxford Philological Society, the British Academy, the University ofLeiden, the University of Durham, the Royal Netherlands Academy of

xii Preface

Arts and Sciences, and the University of Nottingham. Miriam Gri0nwas kind enough to lend me some valuable unpublished work of her ownon Cicero’s philosophical vocabulary. I have been most grateful for theadvice of two anonymous referees for Cambridge University Press: evenif, as I suspect, they have both been named at least once above, let metake the opportunity to thank them once again. Susan Moore’s scrupu-lous copy-editing has saved me from numerous errors, unclarities andinconsistencies. Finally, warm thanks to Pauline Hire of the CambridgeUniversity Press for all her advice and encouragement.

To the University of Cambridge and to Christ’s College I am grate-ful for granting me sabbatical leave during Michaelmas Term !""+,when the bulk of the book was written.

Over the last twenty-seven years I have enjoyed innumerable conver-sations about Lucretius and Epicureanism with Tony Long – first myresearch supervisor, then my collaborator, and at all times a wonderfulfriend and supporter. It is to him that I have chosen to dedicate this book,with gratitude and a/ection.

David SedleyCambridge

Preface xiii

Introduction

The old quarrel between poetry and philosophy may have simmereddown, but in Lucretian studies the two do not always manage to be aswilling allies as they ought to be. Lucretius used poetry to illuminate phi-losophy. My aim in this book is to use philosophy to illuminate poetry.

Lucretius’ achievements as a poet to a large extent lie in his genius fortransforming Epicurean philosophy to fit a language, a culture and a lit-erary medium for which it was never intended. In order to understandhow he has brought about this transformation, we need to know all wecan about what he was transforming and how he set about his task.

In Chapter !, ‘The Empedoclean opening’, I try to show how hedefines the pedigree of his literary medium. It is the poetic genre of thehexameter poem on physics, pioneered by Empedocles. Lucretius’ wayof proclaiming this, I argue, is to write a proem which emphasises thenature and extent of his debt to Empedocles.

In Chapter %, ‘Two languages, two worlds’, I turn to a neglected lin-guistic aspect of Lucretius’ enterprise, his ambiguous relationship withthe Greek language. The transition from Epicurus’ technical Greekprose to Lucretius’ largely untechnical Latin verse is not merely a for-midable task of conversion, it is also an opportunity for Lucretius to mapout an interrelation between two cultures. The result, I argue, is a pow-erful message, encoded in his linguistic imagery, about the true univer-sality of Epicureanism, a universality demonstrated by its unique abilityto transcend linguistic and cultural boundaries.

In Chapter &, ‘Lucretius the fundamentalist’, I defend a quite tradi-tional view, albeit one which is increasingly out of favour. It is thatLucretius had no significant contact with, or knowledge of, contempo-rary philosophy or science. I argue for a strong version of this claim:Lucretius was a true fundamentalist, nourished on the unmediated scrip-tures of his school’s revered founder. To refute systematically every claimever made about recent or contemporary influences on Lucretius would

xv

have resulted in a massive and tedious chapter. But equally, it is impor-tant not to content ourselves with impressionistic assertions or appealsto mere likelihood. I have therefore sought in this chapter to present acomprehensive argument for my case – at any rate, one fuller and moresystematic than has previously been attempted so far as I am aware. Theupshot is that Lucretius really does rely directly on Epicurus’ own writ-ings, just as he tells us he does in the proem to book ,,,. His reverence forthe master’s scriptures is so all-consuming as to obviate any interest inlater philosophical or scientific developments.

Chapters )–( form a single block. Between them they try to answerthe questions (a) what was the hallowed Epicurean material whichLucretius was transforming, and (b) how did he proceed with the task oftransforming it? This leads me into another rather traditional activity,one which many will think recent Lucretian scholarship to be well rid of.I mean the activity of Quellenforschung. But I hope what I have come upwith will not seem like a return to the endless and inconclusive joustingsof Lucretian scholarship in the first half of this century. The text whicheveryone agrees was in some sense Lucretius’ ultimate source for physics,and which I among others believe to be his single direct source – I meanEpicurus’ great treatise On nature – is one about which we possess a hugeamount of information. Yet, extraordinarily, this information has neverbeen assembled into a coherent overview, let alone adequately exploitedby Lucretian scholars.

Therefore my Chapter ), ‘Epicurus, On nature’, is devoted to a full-scale evaluation of this work. I try to show its probable structure, all theway down to the sequence of contents in individual books. I also o/er acharacterisation of its style, and try to explain why it held a unique placein the a/ections of Epicureans. Finally, I o/er a partial chronology of itscomposition.

In Chapter (, ‘Lucretius’ plan and its execution’, I give reasons forregarding the first fifteen books of Epicurus’ On nature as Lucretius’ directsource on physics. This leads me on to what I consider the single mostsignificant proposal in my book. I argue that we can discern in Lucretius’text his actual procedure when composing the poem. (Others have madethe same claim, with very di/erent results from mine, but they havenever based it on an adequate look at On nature.) Initially he worked hisway through On nature fairly systematically, following the order of topicswhich Epicurus himself had said was the correct one. He omitted anumber of topics and individual arguments, but rarely deviated fromEpicurus’ sequence. However, as he wrote he began to see how the mate-

xvi Introduction

rial should be eventually reordered, into something very much like thesix-book structure in which we now know it. This crucially included thedecision, taken quite early on, to reverse the material of books ,,, and ,-from Epicurus’ order into that which we now find in the poem.

Much of the fine detail of this restructuring, however, was undertakenin a second phase, in which he only got half-way through reworking thepoem. Books ,–,,, are, to all intents and purposes, fully integrated intoLucretius’ master plan. But books ,-–-, as we have them are, I am con-vinced, not fully reworked. In Lucretius’ proems, which represent thelatest stages of his work, I claim to be able to detect what his further planswere for books ,- and - – plans which remain unfulfilled in the text as ithas come down to us. This in turn seems to me to o/er strong supportto those who have found themselves unable to believe that book -,,including its closing description of the Athenian plague, is in the finalstate that Lucretius envisaged for it. I thus end Chapter ( with a proposalabout how far he had got with his plans for the plague passage, basedpartly on what has proved to be his method of composition in the pre-ceding books of the poem, and partly on a moral motif which I believeto play an important part in Lucretius’ grand design.

I thus strongly resist the view, which is threatening to become anorthodoxy of Lucretian scholarship, that the De rerum natura is in factfinished. But although I am by implication endorsing the ancient tradi-tion that Lucretius died before putting the final touches to the poem, Ihave nothing new to say about that tradition, including Jerome’s storythat Cicero was the posthumous editor. My sole contribution toLucretian biography is to be found in Chapter %: Lucretius had been toGreece.

One finding of Chapters )–( is that when the voice of Epicurus showsthrough in Lucretius’ text, a primary source used by Epicurus sometimesshows through too. This is Theophrastus’ great pioneering doxograph-ical treatise, Physical opinions. In Chapter +, ‘The imprint ofTheophrastus’, I take the same theme forward, charting particularLucretian passages where Theophrastus is being either borrowed fromor implicitly criticised.

Chapter $ rounds o/ the story by looking close-up at the structure andargument of a single book, the first. Doing so makes it possible to see insome detail how Lucretius’ reworking of his Epicurean material hastransformed Epicurus’ primarily deductive chain of reasoning into aradically new style of discourse, governed even more by the require-ments of rhetoric than by those of philosophical dialectic.

Introduction xvii

My single most earnest goal in writing this book is to be able to addressreaders who have themselves come to Lucretius through the study ofLatin poetry. I hope to persuade some of them that there is much tolearn about Lucretius, even as a poetic craftsman, by scrutinising thephilosophical background to his poem in ways in which it is not usuallyscrutinised. I recognise that a certain proportion of the material in thelater part of the book may be tough going for some readers. But I dovery strongly urge even them at the very least to read the first two chap-ters, to skim the third and fourth, and to read the fifth and seventh. Ifthey so prefer, they have my permission to ignore Chapter + altogether.

None of the chapters, with the exception of & and +, assumes muchprior philosophical knowledge on the part of the reader. All chaptersinvolve some use of both Greek and Latin, but I have tried to translateall words and excerpts quoted in the main text.

Some of the material for this book can also be found in articles whichI have already published or which are currently in press. They are theones listed in the bibliography under my name for !"#) (Chapter )),!"#"a (Chapter !), forthcoming (Chapter $), !""$b (Chapter (), !""#a(Chapter +), and !""#b (Chapter %). In all cases the material has beenreworked and expanded for the book, and a good deal of it is entirelynew.

xviii Introduction

1234567 !

The Empedoclean opening

! . 1,1678 ’9 :65567

Lucreti poemata ut scribis ita sunt, multis luminibus ingeni, multae tamen artis.sed cum veneris, virum te putabo si Sallusti Empedoclea legeris, hominem nonputabo.

Writing to his brother in () '1, Cicero supplies two unique testimonies(Ad Q. fr. ,, ".)). In the first sentence he echoes Quintus’ admiration forLucretius’ poem, thus providing the sole allusion to the De rerum naturalikely to be more or less contemporary with its publication. In thesecond, he attests the publication of an Empedoclea by a certain Sallustius,presumably a Latin translation or imitation of Empedocles (compareCicero’s own near-contemporary use of the title Aratea for his translationof Aratus).

But even more striking than the two individual testimonies is theirjuxtaposition. Modern editors have taken to printing a full stop after sedcum veneris, understanding ‘But when you come . . . (sc. we will discuss it).’This suppresses any overt link between the two literary judgements: thefirst breaks o/ abruptly with an aposiopesis, and the second, juxtaposed,is to all appearances a quite independent observation. On the equallynatural and more fluent reading that can be obtained simply by revert-ing to the older punctuation,1 as printed above, with a comma instead ofthe full stop, the letter is an explicit comparison between the DRN andthe Empedoclea:

Lucretius’ poetry shows, as you say in your letter, many flashes of genius, yet alsomuch craftsmanship. On the other hand, when you come, I shall consider you aman if you have read Sallustius’ Empedoclea, though I won’t consider you human.

!

1 This was the standard punctuation until the late nineteenth century. The repunctuation, with itsaposiopesis sed cum veneris . . . (unique, but cf. partial parallels at Ad Att. .,, (a and .,- %*.&), appearsto have been introduced by R. Y. Tyrrell in !##+, in his revised text of Cicero’s Letters (Tyrrell(!##(–!"*!)), but without o/ering any evidence or argument – since when it has been repeated,without comment, by all editors.

If this is right, the two works were being directly compared at the timeof their publication, and Cicero, at least, judged the Lucretian poemvastly superior.

Why did this particular comparison suggest itself ? It is well recognisedthat Empedocles is, along with Homer, Ennius, and others,2 an impor-tant literary influence on Lucretius, and it has even been claimed that hewas a philosophical influence.3 But I do not believe that the depth andsignificance of the poem’s Empedoclean character have yet been prop-erly understood. If what I shall argue in this chapter is right, Cicero’scomparison of the DRN with the Empedoclea will turn out to be an entirelynatural one, which Lucretius would have welcomed and indeed invited.My case will be centred on the relation of Lucretius’ proem to the proemof Empedocles’ On nature.

% 6;46<81:69 ’ 5=8 486;9

There is plentiful evidence that it was principally if not exclusively in thehexameter poem usually known in antiquity as the On nature (!"#$!%&" '"()) or the Physics (*+! %&'$,+" ) – I shall discuss its actual title in §$ –that Empedocles expounded his world system. The central features ofthe cosmic cycle it described are well known: four enduring elements –earth, air (called ‘aether’),4 fire, and water – are periodically united intoa homogeneous sphere by a constructive force called Love, then againseparated out into the familiar stratified world by the polar force, Strife.5But there is a longstanding scholarly tradition, deriving primarily fromDiels’ editions published in !"*! and !"*&, of attributing all the frag-ments concerning Empedocles’ theories on the pollution and trans-migration of the individual spirit, or ‘daimon’, to a second hexameterpoem, the Katharmoi, or Purifications.

The original ground for this segregation was the belief that the phys-ical doctrine of the cosmic cycle and the ‘religious’ doctrine of trans-migration belonged to radically distinct and probably incompatibleareas of Empedocles’ thought. But Empedoclean studies have nowreached a curious stage. On the one hand, the old dogma has been sub-jected to searching criticism, and is regarded by many as an anachron-

% !. The Empedoclean opening

2 The range of literary influences on Lucretius was considerably enlarged by the findings ofKenney (!"$*). 3 Furley (!"$*), discussed below; also Bollack (!"(").

4 For ‘aether’, rather than ‘air’, as Empedocles’ chosen designation of this element, see Kingsley(!""(), ch. %.

5 The traditional belief that zoogony took place in both halves of this cycle, for which see espe-cially O’Brien (!"+"), has been powerfully challenged by Bollack (!"+(–"), Hölscher (!"+(),Solmsen (!"+(), and Long (!"$)), and ably defended by Graham (!"##).

istic imposition on fifth-century thought.6 On the other hand, theconventional apportionment of fragments between the two poems,which was founded on that dogma, remains largely unchallenged, as ifit had some independent authority. I believe that it has none.

One radical challenge to this picture, however, has been developedrecently. Catherine Osborne7 proposes that there were never two poems:rather, both titles name one and the same work. Although this proposalhas found some favour,8 and has certainly inspired some importantreassessment of the doctrinal relation between the two sides ofEmpedocles’ thought, I do not think that it can be right. Diogenes Laertiusis unambiguously speaking of two separate poems when he tells us that‘On nature and the Katharmoi (-,,, $$, -+! µ"!. /&# . !"#$! %&" '"() ,+$! /$$0+1+#µ/$" . . .) run to (,*** lines.’9 Moreover, a number of the survivingfragments of Empedocles are reported with explicit assignations to one orthe other poem, yet not a single one with attributions to both the physicalpoem and the Katharmoi. Finally, as Jaap Mansfeld has brought to light,Giovanni Aurispa is known to have had a manuscript entitled (in Greek)‘Empedocles’ Katharmoi ’ (now tragically lost) in his library at Venice in!)%).10 Even if this evidence were thought insu0cient, I hope that thematter will be put beyond doubt by my next section, where it will turn outthat one major fragment cannot be placed in the Katharmoi without glaringinconsistency: Empedocles must have written at least two poems.

If we simply stick to the hard and the relatively hard evidence for whatwas in the Katharmoi, a di/erent picture will emerge. We do at least haveits opening lines.11

". Empedocles’ two poems &

16 E.g. Kahn (!"+*), Barnes (!"$") ,, "&, Wright (!"#!), Osborne (!"#$), Inwood (!""%), Kingsley(!""(); reservations in Long (!"++). 17 Osborne (!"#$).

18 Cf. its further development in Inwood (!""%), pp. #–!". The reply to Osborne and Inwood inO’Brien (!""() is unfortunately timed: it contains news of the recent papyrus find (see pp. !* and%# below), but not the specific information that this now virtually proves at least one ‘Katharmic’fragment to belong to On nature.

19 See Osborne (!"#$), pp. %#–" on the unreliability of the figure (,***. But as for the separationof the two titles, there is no compelling reason to doubt Diogenes’ reliability, especially when noancient source contradicts him on the point.

10 Mansfeld (!"")b), which should also be consulted for its further arguments for the existence oftwo separate poems. Of course his evidence is not strictly incompatible with the thesis that therewas one poem, whose proponents may reply that this was that one poem. But it is uncomfortablefor them, since it means that, if they are right, Katharmoi was the o0cial title, contrary to the greatbulk of the ancient citations.

11 Empedocles '!!%. The square-bracketed words represent Greek words apparently corrupt ormissing in the quotation as preserved. Here and elsewhere, I use the Diels/Kranz (!"(!–%) num-bering of Empedocles’ fragments, although a significantly better text is now available in the valu-able edition of Wright (!"#!). Since the many available numerations are, as I shall argue, allequally misleading as regards the apportionment of fragments between the two poems, it isbetter for now simply to stick to the standard one.

Friends, who in the great town of the yellow Acragas dwell on the city’s heights,caring about good deeds, I greet you. You see me going about as a divine god,no longer a mortal, honoured amongst all, it seems, and wreathed in ribbonsand verdant garlands. [Whenever] I arrive in prosperous towns I am revered bymen and women. They follow me in their thousands, asking me where lies theirroad to advantage, some requesting oracles, while others have asked to hear ahealing utterance for ailments of all kinds, long pierced by troublesome [pains].

Thus Empedocles addresses the citizens of his native Acragas, tellinghow they revere him as a living god, ‘no longer a mortal’. Men andwomen flock to follow him, pressing him with enquiries, requestingoracles and cures.

Why should we not suppose that the poem was nothing more nor lessthan a response to these requests, a set of purificatory oracles and‘healing utterances’?12

There is immediate support for this conjecture in the pseudo-Pythagorean Carmen aureum: ‘But abstain from the foods that I spoke ofin my Katharmoi and Absolution of the soul.’13 This citation, or pseudo-cita-tion, of the author’s own Katharmoi invokes it for just the kind of self-purificatory advice that the title itself suggests. And that the allusion isinspired by Empedocles’ work of the same name is confirmed just threelines later, where the poem closes with the words ‘You will be an immor-tal, divine god, no longer a mortal’ ("%''"+$ +& 1+" .+-/) 1"/! ) +% µ2#/-/),/&& ,""-$ 1.3-/" )), pointedly recalling the famous opening of Empedocles’Katharmoi, ‘You see me going about as a divine god, no longer a mortal’('!!%.)–(, "&4(! 5 & &$ µ$' . 1"/! ) +% µ2#/-/), /&& ,""-$ 1.3-/" ),|6(7"&' µ+$).Whatever the date of this forgery may be, its author clearly knowsEmpedocles’ Katharmoi, and associates it with advice to abstain fromcertain kinds of food.

That a work with this title should be one dedicated to purificatoryadvice is unsurprising, since the very word katharmoi means ritual acts ofpurification. To adherents of the traditional interpretation, it is easy toassume that the poem was one about the wandering spirit’s processes ofpurification, but I know no evidence that the word can mean that:14 suchprocesses would normally be called katharseis.

) !. The Empedoclean opening

12 For the scope and content of the relevant notions of pollution and purification, see Parker (!"#&).I have no particular suggestion to make about the function of the ‘oracles’. The evidence of apurificatory role for oracles is meagre (Parker (!"#&), p. #+), and I would guess that it isEmpedocles’ assumed divinity that makes this an appropriate designation for his pronounce-ments.

13 Carmen aureum +$–#, in Young (!"$!), !*&–): +& 77 & "$%#4/& 2#(-(' . (( . "$%6/µ". "% . -" 0+1+#µ/$ )8 "% . -" 9&" '"$ :&;3).

14 The use of ,+1+#µ/$" is usefully surveyed by Guthrie (!"+"), pp. %))–(.

Better still, the hypothesis also fits the other two items of evidenceknown to me for Katharmoi as a literary genre. These two references alsoresemble the Carmen aureum in fathering the works in question on archaicfigures of semi-legendary status. First, Epimenides the Cretan is said tohave written Katharmoi, in verse and perhaps also prose,15 and, althoughtheir content is not reported, it can hardly be a coincidence thatEpimenides was celebrated above all for his ritual purifications, anexpertise that led the Athenians to send for him to purify their city ofplague.16 Second, the remark at Aristophanes, Frogs !*&& that Musaeustaught ‘healing and oracles’ is glossed by a scholiast with the commentthat Musaeus ‘composed absolutions [?], initiations, and katharmoi’.17

Healing and oracles are precisely the two services mentioned byEmpedocles at the opening of his Katharmoi. Then why look further forthe content of the poem?

Certainly no fragment explicitly attributed to the Katharmoi forces usto look further. Apart from the proem, there are just two such cases. Oneis '!(&a: according to Theon of Smyrna (!*).!>&), Empedocles ‘hints’(+$&.$"--"-+$) in the Katharmoi that the foetus achieves full human form inseven times seven days. Aetius18 confirms the report – though not theattribution to the Katharmoi – with the further information that thedi/erentiation of limbs starts at thirty-six days. That Empedocles shouldonly have ‘hinted’ this in the Katharmoi suggests that we are not dealingwith an expository account of embryology. We learn from Censorinus19

(third century 3<) that in Greece the pregnant woman does not go outto a shrine before the fortieth day of her pregnancy. This is thought tobe linked to the widespread belief that miscarriages are likeliest to occurin the first forty days.20 There is a strong possibility that Empedocles’original remark occurred in the context of ritual advice to pregnantwomen, perhaps to avoid shrines for the first ‘seven times seven’ days.Here it is important to remember the opening of the Katharmoi, where itis made explicit that the demands for healing and oracles to whichEmpedocles is responding come from women as well as men.

The other explicit attribution to the Katharmoi – in fact to book ,, ofthe poem – occurs in a fragment first published in !"+$, fr. !(% Wright:21

". Empedocles’ two poems (

15 &3%–& DK. 16 &3!, %, ), # DK.17 %3+ DK. There is a close parallel at Plato, Rep. ,, &+)e–&+(a: Adimantus, as evidence of the belief

that the gods can be bought o/, cites the books of Musaeus and Orpheus, on the basis of whichrituals are performed to bring about the 7&" '"$) -" ,+$! ,+1+#µ/$" of wrongs done by both theliving and the dead. 18 Aetius - %!.!!Empedocles 3#&.

19 Censorinus, De die natali !!.$. 20 See Parker (!"#&), p. )#.21 Wright (!"#!), pp. !(! and %"#; not, of course, to be found in Diels/Kranz (!"(!–%).

‘For those of them which grow with their roots denser below but theirbranches more thinly spread . . .’ Trees, or more generally plants, of thiskind were singled out for a reason which cannot now be recovered.22 Thecontext may well have been one concerning the avoidance of certainleaves. According to Plutarch, in a probable but unprovable citation ofthe Katharmoi, Empedocles urged that all trees should be ‘spared’, butespecially the laurel:23 ‘Keep completely away from the laurel’s leaves’('!)*). This has every chance of tying in with Empedocles’ views ontransmigration – he holds, for example, that the laurel is the best tree totransmigrate into ('!%$)! But it is significant that here once again, if thelink with the injunction about laurel leaves is accepted, the actual frag-ment may well contain moral or purificatory advice rather than the doc-trinal exposition characteristic of the physical poem. To repeat, ritualadvice is just what we should expect in a work entitled Katharmoi.

The expectation finds further strong support in the story surroundingfragment '!!!. We learn that the biographer Satyrus quoted this frag-ment as confirming the suspicion that Empedocles dabbled in magic.24

Since, according to Apuleius,25 it was Empedocles’ Katharmoi thatbrought upon him just such a suspicion, there is a strong likelihood that'!!! is from this poem.26 Significantly, the fragment is once again not adoctrinal exposition but ritual advice: how to influence the weather andto summon up the dead.

'!!! uses the second person singular: ‘You [singular] will learn . . .’Because the On nature was addressed to an individual, Pausanias, whereasthe opening lines of the Katharmoi address the citizens of Acragas in theplural, it has often been thought that any fragments containing thesecond person singular must be assigned to the former poem. This is avery dubious criterion, since changes of address within a single didacticpoem are quite normal. Hesiod’s Works and days switches in its first threehundred lines between addresses to the Muses, to Perses, and to the‘bribe-swallowing princes’.27 That the Katharmoi should, after itsopening, move into the second person singular may merely reflect thefact that Empedocles is by now answering the individual requests fromhis audience of which the proem spoke.

+ !. The Empedoclean opening

22 According to Theophrastus, HP , +.), all plants have their roots more densely packed than theirparts above ground, but some, e.g. the olive tree, have a particularly dense mass of slender roots.

23 Plut. Quaest. conv. +)+<, see preamble to '!)* DK. 24 DL -,,, (". 25 Apuleius, Apol. %$.26 This attribution is supported, as Inwood (!""%), p. !+ has shown, by the fact that Clement (Strom.

-, &*.!–&) directly associates '!!! with the opening lines of the Katharmoi.27 See further, Osborne (!"#$), pp. &!–%, who appositely compares Lucretius’ own switches of

address.

There are no further unambiguously attested fragments of theKatharmoi. But we may, with caution,28 consider as potential fragments ofit any citations of Empedocles whose sources explicitly call them kathar-moi. The clearest case of this is in Hippolytus,29 who describes prohibi-tions on marriage and on certain foods as tantamount to teaching thekatharmoi of Empedocles. Given this remark, along with the associationof the Katharmoi with food prohibitions in the Carmen aureum, it seems safeto assume that the poem carried Empedocles’ advice to abstain fromslaughter, meat-eating, and perhaps even beans.30 And it seems thatabstention from marriage was a further injunction to be found in thesame work.31

Another plausible such candidate is a fragment preserved by Theonof Smyrna.32 Comparing philosophy as a whole to a religious ritual,Theon calls Plato’s five propaedeutic mathematical studies in Republic -,,a katharmos, which he immediately proceeds to link with Empedocles’injunction to cleanse oneself by ‘cutting from five springs (in a bowl of)indestructible bronze’ ('!)&).33 We are here firmly in the territory ofritual self-purification. Theophrastus’ godfearing character, forexample, refuses to set out on his daily rounds until he has washed hishands at three springs.34

Deciding just which other verbatim fragments should be assigned tothe Katharmoi is a problem to pursue on another occasion. The argumentto which I shall now turn relies on a primarily negative conclusion: there

". Empedocles’ two poems $

28 '!&", which in Sedley (!"#"a) I incautiously left in the Katharmoi, can now be shown to belong tothe physical poem: see p. &* below.

29 Hippolytus, Ref. -,, &*.&–); see preamble to '!!* in Diels/Kranz.30 Empedocles '!)!, carrying the Pythagorean advice to abstain from beans, is condemned as inau-

thentic by Wright (!"#!), p. %#", perhaps rightly.31 Hippolytus loc. cit. presents the advice not to marry as itself Empedoclean: ‘You are dissolving

marriages made by God, following the doctrines of Empedocles, in order to preserve the workof Love as one and undivided. For according to Empedocles, marriage divides the one andmakes many.’ This is a curious view to take of marriage, although it could well apply to the family.

32 Theon of Smyrna !)–!(.33 I here translate the Diels/Kranz text, based on Theon, ,#3.+" (. +% 6/ 6"" .-" -+µ/" .- & <"& .>

+& -"$#"" $ ;+7,(. . Aristotle, Poet.!)($b!& quotes (without attribution) the words -"µ(! . +& -"$#"" $ (A,-+.+,"" $ B) ;+7,(. , explaining that ‘cutting’ here is used to mean ‘drawing’. This leads van derBen (!"$(), %*&–#, and Wright (!"#!), %#"–"*, to follow the lead of Maas and conflate the twoquotations in the form ,#3.+" (. +% 6/ 6"" .-" -"µ(! . (or -+µ(! .) -+.+3" ,"$ ;+7,(. , with thefurther inevitable conclusion that the reference is to drawing blood with a knife – which ofcourse Empedocles would be condemning. This seems to me too high a price to pay, since ittotally contradicts Theon’s report that Empedocles with these words is advising us to cleanseourselves.

34 Theophrastus, Char. !+.%. See Parker (!"#&), pp. %%+–$. Cf. Apollonius Rhodius ,,, #+*, whereMedea, before preparing an ointment which confers invulnerability, bathes herself in sevenstreams.

is no reason to attribute to this poem any fragments of Empedoclesbeyond those o/ering ritual advice.35

& . 526 478-6?3?16 8@ 6;46<81:69 '!!(

There is a decree of necessity, an ancient resolution of the gods, sworn by broadoaths, that when one of the daimons which have a share of long life defiles . . .its own limbs, or does wrong and swears a false oath, for thirty thousand yearsit must wander, away from the blessed ones, being born during that time asevery form of mortal creature, exchanging for each other the arduous paths oflife. The might of the aether drives it to the sea, the sea spits it out onto thethreshold of land, the earth sends it into the rays of the gleaming sun, and thesun hurls it into the whirling aether. One receives it from another, and all hateit. I too am now one of these, a fugitive from the gods and a wanderer, who trustin raving Strife.

These lines ('!!(),36 which are crucial for explaining the daimon’s migra-tions, have been assigned to the Katharmoi by every editor of Empedoclessince Diels.37 The attribution has been questioned by N. van der Ben,and subsequently defended by D. O’Brien.38 But this renewed debatehas so far focused excessively on the contexts in which the lines arequoted by our sources, as if one could settle the question of their prove-nance by counting the allusions in those contexts to katharsis and cognateterms and likewise those to the cosmic cycle. Given the improbabilitythat any ancient reader of Empedocles might have expected the phys-ical poem and the Katharmoi to conflict doctrinally, the provenance of thelines will have mattered less to those who cited them than their value asevidence for Empedocles’ views on the katharsis of the soul – a topic onwhich Platonism had conferred an absolutely pivotal importance.

Plutarch reports that Empedocles used these lines ‘as a preface at thebeginning of his philosophy’.39 Is this too vague to be helpful?‘Philosophy’ certainly might describe the content of the physicalpoem.40 It might also be appropriate to the Katharmoi, on the tradi-

# !. The Empedoclean opening

35 I agree with Kingsley (!""+), p. !*" that the Katharmoi must have contained some indication ofhow it is the facts of transmigration that make meat-eating a sin. But Empedocles’ declaredcelebrity at the time of writing this poem hardly suggests that he would need to do very muchexplaining of his doctrine. I certainly see no necessity on this ground to attribute any specificknown fragment (e.g. '!&$, as Kingsley suggests) to it, beyond those I have listed.

36 I have avoided engaging with the textual di0culties of this passage, which are well discussed byWright (!"#!). They do not a/ect any of the issues I am addressing here.

37 This of course applies to Inwood (!""%) only in so far as he identifies the Katharmoi with the wholeof Empedocles’ poetic œuvre. 38 Van der Ben (!"$(), pp. !+/.; O’Brien (!"#!).

39 Plut., De exilio +*$1: "&. +& #;<' -3' ) %$7/'/%$"+) 6#/+6/%(.3" '+).40 Kingsley (!""+) argues, in reply to Sedley (!"#"a), that ‘philosophy’ to Plutarch would normally

tional view of that poem’s content as expository and doctrinal. But itis very much less appropriate if, as I have argued, the Katharmoi was nota doctrinal work but a set of purificatory pronouncements. Indeed, ifthat suggestion is correct, Plutarch’s expression ‘at the beginning of hisphilosophy’ would immediately gain a much clearer sense. IfEmpedocles wrote two doctrinal poems, the words ‘his philosophy’ area desperately vague way of referring to either one of them. But if hewrote just one, they become an entirely natural way of referring to thatone.41

Plutarch’s description in no way indicates that these were the veryopening lines of the poem to which they belonged, just that they pre-ceded the philosophy proper. Hence there is little value in the argu-ment42 that since we have the opening of the Katharmoi and it di/ers fromthese lines, they must have opened the physical poem instead. Muchmore mileage can be got out of the content of the disputed lines. First,it is hardly insignificant that they name five of the six cosmic entities onwhich Empedocles’ physical system is based: the daimon’s wanderingsare graphically described in terms of its being tossed into and out ofeach of the four elements in turn; and Strife is named as the cause of itsdownfall. This at least supports the coherence of the passage with the phys-ical poem.

But far more important, and strangely absent from the debate aboutits provenance, is the following consideration. In these disputed lines,Empedocles is himself a fallen daimon: ‘I too am now one of these, afugitive from the gods and a wanderer, who trust in raving Strife.’ Is itcredible that these words came in the introductory passage of a poem inwhose opening lines Empedocles had moments earlier described himself

#. The provenance of Empedocles B!!$ "

mean the kind of moral precepts, tinged with myth and religion, that are associated with theKatharmoi. This may not seem much of a challenge to my position, since I argue that there wasa good deal of this kind of material in On nature. But Kingsley’s claim is that ‘philosophy’ is pre-cisely the word Plutarch would use to distinguish the ‘philosophical’ Katharmoi from the other,merely ‘physical’ poem. However, his evidence crumbles on examination. At De gen. Socr. (#*1Plutarch’s speaker Galaxidorus does (on a plausible restoration of the text) say that Pythagoras’philosophy, already full of ‘visions and myths and religious dread’, became positively ‘Bacchic’in the hands of Empedocles. But in no way does this, as Kingsley seems to think, delimit whatPlutarch would mean by the expression ‘Empedocles’ philosophy’, and thus exclude physicsfrom it. Plutarch’s other speakers often make it abundantly clear that, like anybody else, theyregard ‘philosophy’ as including physics (De def. or. )%*', De facie ")%') and logic (De Is. et Os.&#$3), as well as contemplation of first principles (ib. &#%<–6). And although, as Kingsley notes,at De poet. aud. !)6 and !(@, Plutarch recommends the couching of philosophy in versified mythas a didactic device, that tells us nothing about what he means by the word ‘philosophy’, espe-cially when at least one of his speakers, Theon (De Pyth. or. )*+6), takes an almost diametricallyopposed view of philosophy. 41 Cf. Osborne (!"#$), pp. %"/.

42 Van der Ben (!"$(), p. !+.

as ‘a divine god, no longer a mortal’?43 Without the straitjacket of theold prejudice that science and religion do not mix, it is hard to believethat anyone would ever have thought of assigning the former text to theKatharmoi. The most natural interpretation is that '!!( comes from apoem in which Empedocles classed himself as a fallen daimon stillworking through its long cycle of transmigrations, whereas in theKatharmoi, opening as it does with his confident self-proclamation as agod, ‘no longer a mortal’, he presented himself as having now completedthe cycle and recovered his divinity. I therefore feel a reasonable degreeof confidence in placing Empedocles’ major fragment on the wander-ings of the daimon somewhere in the proem to the On Nature.

Since I first developed this argument several years ago,44 it hasreceived welcome confirmation in the discovery of papyrus fragmentsfrom book , of Empedocles’ On nature.45 They include lines denouncinganimal slaughter46 – lines which editors have always hitherto assigned tothe Katharmoi. The taboo on slaughter is, famously, one whichEmpedocles based on his doctrine of transmigration. Hence the trans-fer of these lines to the opening book of the On nature should do much toobviate any remaining resistance to the conclusion that '!!(, on themigrations of the daimon, belongs to the proem of that same book.

This conclusion will prove important at a later stage in my argument.Earmarking it for future use, we can now at last turn to Lucretius.

) . :A1765,A9 3?< 6;46<81:69

Numerous echoes of Empedoclean passages have been recognised inLucretius’ poem, with varying degrees of certainty.47 It is no part of mypurpose to catalogue these. But two observations seem in order. First, the(** or so extant lines of Empedocles48 represent around one-tenth of his

!* !. The Empedoclean opening

43 '!!%.), reinforced by '!!&.% (‘if I am superior to frequently-perishing mortal human beings’), if,as Sextus’ juxtaposition of '!!& with '!!% suggests, it is also from the Katharmoi. In Empedocles’world, even the generated gods perish eventually, i.e. at the end of each cosmic cycle: hence theyare not immortal but ‘long-lived’ ('%!.!%, '%&.#; cf. '!!(.( on the daimons). By contrast, mortalsare ‘frequently-perishing’, 6/7&%1"#"(., see Wright (!"#!), p. %+". 44 In Sedley (!"#"a).

45 The exciting new Strasbourg papyrus of Empedocles has its editio princeps in Martin/Primavesi(!""#). Although, at the time of completing the present book, I had not seen this edition, OliverPrimavesi was kind enough to send me a copy of his habilitationsschrift (the basis of Primavesi(forthcoming)), and both he and Alain Martin have been extremely generous in keeping meinformed about their work. 46 '!&", see n. !*$ below.

47 Esp. Furley (!"$*); also Kranz (!"))), Castner (!"#$), Gale (!"")a), pp. ("–$(. I have not seenJobst (!"*$), but I understand from Don Fowler that he anticipated Kranz’s most important find-ings. For other studies, see Tatum (!"#)), p. !$# n. (.

48 This figure tries to take some account of the new papyrus find. I understand from the editors,

poetic output, if we are to trust Diogenes Laertius’ figure of (,*** linesin total,49 and even on the most conservative estimates of Empedocles’total output,50 not more than one-fifth. Or supposing (as I am inclinedto suppose) that Lucretius’ interest was exclusively in the On nature, whatis extant of that is still likely to be less than a quarter – roughly )(* linesout of %,***.51 This raises the probability that if we had Empedocles’poems intact a great deal more Empedoclean influence would come tolight, and our understanding of the DRN be immensely enriched.

Second, I would suggest that Lucretius is likely to owe rather more toEmpedocles in terms of poetic technique than is generally recognised.For example, at , %$!–"$ Lucretius argues for the corporeality of air bymeans of an intricate analogy between the destructive power of windand that of water. David West has observed that the number of distinctpoints of correspondence between the description of the wind and thedescription of the water greatly exceeds that normally found in thesimiles of Homer and Apollonius.52 Lucretius is thus, in West’s termi-nology, a practitioner of the ‘multiple-correspondence simile’, a legacythat he was to pass on to Virgil. What I would myself add is that,although Homer and Apollonius may o/er no adequate model for thetechnique, Empedocles does. In his description of the eye’s structureand function as analogous to those of a lantern,53 Empedocles rein-forces the idea with a set of carefully engineered correspondencesbetween the two halves of the simile.54 As in Lucretius, so already inEmpedocles, the multiplicity of correspondences has an argumentativemotive, and not merely a descriptive one: the more correspondencesthere are, the more persuasive the analogy becomes. Here then is a tech-nique, singularly at home in philosophical poetry, which has almost cer-tainly passed from Empedocles, through Lucretius, into the Latin poetictradition.

Lucretius’ reverence for Empedocles is evident in the paean of praisewith which he prefaces his criticism of Empedocles’ four-element theoryat , $!+–)!:

%. Lucretius and Empedocles !!

Alain Martin and Oliver Primavesi, that they have detected in them some new examples of locu-tions imitated by Lucretius. 49 DL -,,, $$; for discussion see Osborne (!"#$), pp. %#–".

50 Wright (!"#!), p. %!.51 %,*** lines seems to be the figure for the length of the physical poem given by the Suda, s.v.

‘Empedocles’ (!Empedocles 3% DK), despite the slightly odd grammar.52 West (!"$*).53 Empedocles '#). For discussion see Wright (!"#!), pp. %)*–&, Sedley (!""%b).54 These are contained principally in the close linguistic parallelism of lines )–( with the final two

lines. For comparable prose uses of complex analogy in Hippocratic authors, cf. Lloyd (!"++),pp. &)(–#.

quorum Acragantinus cum primis Empedocles estinsula quem triquetris terrarum gessit in oris,quam fluitans circum magnis anfractibus aequorIonium glaucis aspargit virus ab undis,angustoque fretu rapidum mare dividit undis $%*Aeoliae terrarum oras a finibus eius.hic est vasta Charybdis et hic Aetnaea minanturmurmura flammarum rursum se colligere iras,faucibus eruptos iterum vis ut vomat ignisad caelumque ferat flammai fulgura rursum. $%(quae cum magna modis multis miranda videturgentibus humanis regio visendaque fertur,rebus opima bonis, multa munita virum vi,nil tamen hoc habuisse viro praeclarius in senec sanctum magis et mirum carumque videtur. $&*carmina quin etiam divini pectoris eiusvociferantur et exponunt praeclara reperta,ut vix humana videatur stirpe creatus.hic tamen et supra quos diximus inferiorespartibus egregie multis multoque minores, $&(quamquam multa bene ac divinitus invenientesex adyto tamquam cordis responsa dederesanctius et multo certa ratione magis quamPythia quae tripodi a Phoebi lauroque profatur,principiis tamen in rerum fecere ruinas $)*et graviter magni magno cecidere ibi casu.

Of these [sc. the four-element theorists] the foremost isEmpedocles of Acragas, born within the three-cornered terres-trial coasts of the island [Sicily] around which the Ionian Sea,flowing with its great windings, sprays the brine from its greenwaves, and from whose boundaries the rushing sea with itsnarrow strait divides the coasts of the Aeolian land with itswaves. Here is destructive Charybdis, and here the rumblings ofEtna give warning that they are once more gathering the wrathof their flames so that her violence may again spew out the fireflung from her jaws and hurl once more to the sky the lightningflashes of flame. Although this great region seems in many waysworthy of admiration by the human races, and is said to deservevisiting for its wealth of good things and the great stock of menthat fortify it, yet it appears to have had in it nothing moreillustrious than this man, nor more holy, admirable, and pre-cious. What is more, the poems sprung from his godlike mindcall out and expound his illustrious discoveries, so that hescarcely seems to be born of mortal stock.

But this man and the greatly inferior and far lesser ones whom

!% !. The Empedoclean opening

I mentioned above, although in making their many excellentand godlike discoveries they gave responses, as from the shrineof the mind, in a holier and much more certain way than thePythia who makes her pronouncements from Apollo’s tripodand laurel, nevertheless came crashing down when they dealtwith the elementary principles of things. Great as they were,their fall here was a great and heavy one.

This is remarkable praise55 to lavish on a philosopher who did, after all,radically misconceive the underlying nature of the world. Where doesthe emphasis lie? Lucretius speaks highly both of Empedocles’ ‘illustri-ous discoveries’ (praeclara reperta, $&%), and of his poetry, which is sosublime as almost to prove his divinity – an honour that in the endLucretius will reserve for Epicurus alone.56 With regard to Empedocles’‘discoveries’, I am inclined to agree with those who hold that Lucretiusis implicitly commending, among other things, the clarity of theirexposition, especially by contrast with the obscurities of Heraclitusdenounced in the preceding passage.57 This, I would further suggest, issupported by the closing remarks in the passage quoted above, whereLucretius expresses his approval both of Empedocles and of his ‘lesser’colleagues in the pluralist tradition58 for revealing their findings ‘in aholier and much more certain way than the Pythia who makes her pro-nouncements from Apollo’s tripod and laurel’ ($&#–"). This has stan-dardly been understood as crediting those philosophers with anauthority comparable to that of an oracle. It would be safer, however, tosay that it relies on a contrast – between religious oracles, whichLucretius like any good Epicurean deplores, and the philosopher’s ratio-nal alternative, delivered ‘as from the shrine of the mind’ ($&$).59 That

%. Lucretius and Empedocles !&

55 Contrast Edwards (!"#"), who takes this passage and others in Lucretius as treating Empedocleswith a certain disdain.

56 First at ,,, !(. It is unwise to be too confident that Lucretius is alluding to Empedocles’ own pro-fession of divinity at the beginning of the Katharmoi, if, as I would maintain, his interest is other-wise focused entirely on Empedocles’ On nature. But the legend of Empedocles’ plunge into Etnain a bid to establish his own divinity was probably well enough known by this date to give theremark extra point (cf. Wright (!"#!), pp. !(–!+ and Hor. Ars poet. )+&–+).

57 , +&(>)), cf. Kollmann (!"$!), and especially Tatum (!"#)).58 The reference is vague, but perhaps picks up the proponents of two elements in , $!%–!& as well

as the four-element theorists of $!)–!(. On the Epicurean background to their belittling descrip-tion, see pp. !)%>& below.

59 On this reading, Lucretius’ words distance him from approval of (literal) oracles as e/ectively asthe way in which, for example, those who praise the ‘university of life’ distance themselves fromapproval of (literal) universities. Thus Lucretius’ application of oracular language to his ownpronouncements, here and at - !!!–!% ( fundere fata), is ironic: cf. Obbink (!""+), pp. (+#–", com-menting on the irony in Philodemus, Piet. %*))–( ("& ;#3'µ(. [$]53'+µ".) and in Epicurus SV %",with a comprehensive set of Epicurean parallel uses of oracular language. The evidence listed

would amount to a contrast between, on the one hand, the clear, ratio-nal and unambiguous assertions of the pluralists, and, on the other, theDelphic ambiguities so characteristic of Heraclitus.60 If so, we must bewary of exaggerating the extent to which this eulogy of Empedoclesexpresses special admiration for his teaching as such. It is largely as aneloquent and straight-talking expositor of his teaching that he is canon-ised. Empedocles’ language may be densely metaphorical (as isLucretius’ own), but at least, as Lucretius sees it, it lacks the multi-layeredevasiveness and trickery of Heraclitean prose. About Lucretius’ veryreserved evaluation of Empedocles’ actual teachings I shall say morebelow.

What purpose is served in this passage by the fulsome praise of Sicily?One object, no doubt, is to compare Empedocles favourably with thatother wonder of Sicily, Etna.61 But it also has the job of illustrating whySicily was the birthplace of the four-element theory.62 The four elementsare intricately worked into the travelogue. Empedocles was born withinSicily’s ‘terrestrial coasts’ (terrarum . . . in oris, $!$: literally ‘coasts of lands’)– and here terrarum is no ‘otiose addition’ (Bailey), but Lucretius’ way ofidentifying the land of Sicily with the element earth. The elements waterand fire are abundantly in evidence in the descriptions of the surround-ing sea, of the whirlpool Charybdis, and of the flames of Etna ($!#–%().Finally ($%(), those flames are borne ‘to the sky’ (caelum). Now the sky, asthe abode both of air and of the heavenly bodies, might in principlesymbolise either of the elements air and fire. What surely clinches itsidentification with air, and thus completes the catalogue of four ele-

!) !. The Empedoclean opening

by Smith (!"$(), pp. +*–!, note b, does not militate against this picture: in Epicurus SV %",;#3'µ(

$5"$' . is associated with unintelligibility; Cic. Fin. ,, %*, !*% and ND , ++ do use oracula of

philosophical pronouncements (some of them Epicurean), but only in the mouths of Epicurus’critics; the epigram of Athenaeus (ap. DL . !%) speaks of Epicurus not as himself oracular butas inspired either by the Muses or by the Delphic oracle. Cf. Smith (!""+), p. !&* n. $( for furthercomment.

60 For certus!‘unambiguous’ see OLD s.v., ". The same sense fits perfectly into - !!!–!%, wherethese lines recur: Lucretius is saying that his quasi-oracular prediction that the world will oneday perish (see Chapter +) is a firm and unambiguous one, unlike those associated with theDelphic oracle. For Heraclitus’ ‘Delphic’ ambiguity, cf. his '"& DK. As for sanctius, in a compari-son with an oracle this must primarily imply ‘holier’, but the basic meaning of sanctus (fromsancire) is ‘ratified’ or ‘confirmed’, and it also has connotations of ‘above board’ or ‘honourable’(OLD s.v., )).

61 If the thesis developed below about Lucretius’ literary debt to Empedocles is right, it may notbe too fanciful to see in the imminent new eruption of Etna ($%%/.) a hint at the scheduled rebirthof Empedoclean poetry. And is it really just a coincidence that at $&* Lucretius praisesEmpedocles as ‘carus’, his own cognomen (for the point, see Fowler (!""+), p. ###)? The adjec-tive is not part of his regular vocabulary, this being one of only two occurrences in his poem.

62 This was well spotted by MacKay (!"(() and Snyder (!"$%).

ments, is the fact that Empedocles himself uses ‘sky’ (/&& #+./" )) as a namefor his element air ('%%.%).63

And the Empedoclean influence goes deeper still. The very idea ofusing individual phenomena like sea, rain, wind and sun to symbolisethe four elemental stu/s is thoroughly Empedoclean. So too is the poeticdevice of interweaving the four elements into the language of a descrip-tive passage: we have already seen Empedocles do the same at '!!(,when he described the tossing of the fallen daimon from aether (!air)to sea, to land, to the sun’s rays, and then back once more into the eddiesof the aether.

At the very least, then, Lucretius’ description of Sicily reveals his inti-mate knowledge and exploitation of Empedoclean poetry. And it wouldbe unwise to rule out a further possibility: that it is itself a direct imita-tion of a lost passage of Empedocles.

( . 526 6?,B;3 8@ :A1765,A9 4786;

We are now ready to turn to the most hotly and inconclusively debatedpassage in Lucretius, the proem to book ,.64 It is structured as follows:!–%*: praise of Venus as Aeneadum genetrix and the life force of all

nature;%!–#: prayer to Venus to inspire Lucretius’ poem, because she alone

is responsible for making things pleasing, and becauseMemmius has always been her favourite;

%"–)&: prayer to Venus to intercede with her lover Mars and bringpeace to the Roman republic;

))–": it is not in the divine nature to concern itself with our a/airs;(*–+!: programmatic address to Memmius about the content of the

poem;+%–$": praise of Epicurus’ intellectual achievement;#*–!*!: attack on the evils of religion, as illustrated by the sacrifice of

Iphigeneia;!*%–&(: warning to Memmius not to be enticed by false religious tales

about the survival and transmigration of the soul;!&+–)(: the di0culty of Lucretius’ poetic task.

$. The enigma of Lucretius’ proem !(

63 As Kingsley (!""(), ch. %, shows, Empedocles’ own designation of air is ‘aether’, and aether inearly Greek epic is intimately associated with /&& #+./" ).

64 The huge bibliography on this passage prominently includes Giancotti (!"("), Kleve (!"++),Kenney (!"$$), pp. !&–!$; Clay (!"#&), pp. #%–!!*, Gale (!"")a) ch. +, and all the major com-mentaries.

The most enigmatic feature of the proem lies in the first three sub-divisions, !–)&. How can Lucretius, as an Epicurean, praise Venus as acontrolling force in nature, and even beg her to intervene in humana/airs? In Epicureanism, the gods emphatically do not intervene in anyway in human a/airs – as Lucretius himself paradoxically goes onimmediately to point out ())–"!,, +)+–(!).

To respond that the proem’s treatment of Venus is allegorical is not initself a solution to the puzzle. As Lucretius himself warns at ,, +((–+*,allegorical use of divinities’ names, e.g. ‘Neptune’ for the sea and ‘Ceres’for corn, is permissible only if one avoids any false religious implications.Although Venus might, on this principle, get away with symbolisingnature, or even perhaps Epicurean pleasure,65 the opening address to heras ancestress of the Romans can hardly be judged equally innocent, norcan the prayers to her to intervene in Roman a/airs and to inspireLucretius’ poetry.

It is not that these allegorical explanations do not carry any weight atall. I think there is much truth in them. But the most they can do, forreaders who have read on and been surprised to learn that this is anEpicurean poem, is mitigate their baCement. The question remains,what can have impelled Lucretius to start out so misleadingly, under-mining exactly that attitude to the gods that the rest of the poem will soenergetically promote? It would scarcely be an exaggeration to say thathe spends the remainder of the poem undoing the damage done by thefirst forty-three lines.

+ . @A7:6D ’9 5269,9

In short, the opening of the proem simply is not like Lucretius. But it isvery like Empedocles. In his outstandingly important study of theproem, David Furley has observed the high level of Empedocleancontent to be found in it.66 My object here will be to augment hisobservations with further evidence of Empedoclean echoes, but then, inthe remainder of the chapter, to propose a very di/erent explanationfrom his for their presence here.

!+ !. The Empedoclean opening

65 The suggestion of Bignone (!")(), pp. )&$–)), but one which faces the di0culty that Lucretius’Venus controls all natural coming-to-be (esp. %!/.), not just animal reproduction. Asmis (!"#%)proposes that Venus is here an Epicurean deity invented to take over the role assigned to Zeusby the Stoics; but against the supposition that Lucretius is concerned to resist the Stoics, see Ch.& below.

66 Furley (!"$*). The range and depth of Empedoclean nuances in the proem are further enrichedby Clay (!"#&), pp. %%–&, )"/., #%–!!*, %(&–$.

First, notice the by now familiar technique of working the four ele-ments into a descriptive passage. The poem begins as follows (!–():

Aeneadum genetrix, hominum divomque voluptas,alma Venus, caeli subter labentia signaquae mare navigerum, quae terras frugiferentisconcelebras, per te quoniam genus omne animantumconcipitur visitque exortum lumina solis. (

Ancestress of the race of Aeneas, delight of humans and gods,nurturing Venus, who beneath the gliding beacons of the skypervade the ship-bearing sea and the crop-carrying lands,because it is due to you that every race of living beings is con-ceived, and born to look upon the sunlight.

Planted in the text already are references to the sky (which we have seento represent the element air in Empedoclean imagery),67 to the heavenlybodies and the sunlight (i.e. fire), to the sea, and to the land. We thenlaunch into a second catalogue of the same four (+–"):

te dea, te fugiunt venti, te nubila caeliadventumque tuum, tibi suavis daedala tellussummittit flores, tibi rident aequora pontiplacatumque nitet di/uso lumine caelum.

From you, goddess, and your approach the winds and the cloudsof the sky flee away. For you the creative earth pushes up sweetflowers. For you the sea’s surface laughs, and the sky, made calm,shines with di/used light.

Again, the four elements feature: the winds and clouds of the sky, theearth, the sea, the sunlight. And if all this is still not enough, we needonly move on to %"–)&, Lucretius’ prayer to Venus to intercede with herlover Mars. It has long been recognised that here we have a striking allu-sion to the joint-protagonists of Empedocles’ physical poem, Love andStrife – whom Empedocles himself sometimes calls Aphrodite and Ares.

Furley has noted two other Empedoclean echoes in the proem, towhich we will come shortly. But first the question must be asked: whyshould an Epicurean poem start with an Empedoclean prologue?

It is here that I part company with Furley. He argues that Lucretius’act of piety to Empedocles is the acknowledgement of a philosophicaldebt. Although Lucretius was himself a committed follower of Epicurus,Furley suggests, he recognised Empedocles as the inaugurator or cham-pion of two traditions to which, as an Epicurean, he too adhered. The

&. Furley’s thesis !$

67 I o/er this as a ground for going beyond Furley and detecting all four elements even in lines !–(.

first of these is the insistence on absolutely unchanging physical ele-ments. The second is the rejection of a teleological world-view, with allits implications of divine intervention.

But this could hardly explain Lucretius’ decision to open with a tributeto Empedocles. No reader of the proems to books ,,,, -, and -, can doubtthat Lucretius’ other philosophical debts pale into insignificance whencompared with his acknowledged dependence upon Epicurus. Why thenwould he give his putative philosophical obligation to Empedocles theundeserved and thoroughly misleading prominence that it gains from aposition at the poem’s opening?

Moreover, the unwritten rules of philosophical allegiance in theancient world do not normally permit the imputation of authority toanyone other than the founder of your own school, or, at most, to hisown acknowledged forerunners.68 The Epicurean school was second tonone in observing this principle. It seems certain that Empedocles wasnot regarded by Epicurus or his successors as any sort of philosophicalforerunner; and even an acknowledged forerunner like Democritus wastreated with limited respect in the school.69 Now Lucretius is admittedlyin certain ways a non-standard Epicurean, and I shall be arguing inChapter & that he was not a participating member of any Epicureangroup. Even so, his declarations of absolute loyalty to Epicurus as thevery first philosopher to liberate the human race from fear of thedivine70 hardly suggest that he was an exception to this usual style ofschool loyalty. In any case, he certainly knew his Epicurean source textswell enough to be aware of Epicurus’ own reserve with regard to hisforerunners.

Even on the two philosophical issues picked out by Furley, elementtheory and anti-teleology, it is doubtful whether Lucretius or any otherEpicurean would have been as generous in acknowledging Empedocles’contribution as Furley proposes. Indeed, so far as concerns elementtheory, Lucretius is emphatic at , $&)–)! (translated above pp. !%>!&) thatthis is not a topic on which Empedocles acquitted himself with distinction.

!# !. The Empedoclean opening

68 As argued in Sedley (!"#"b).69 For Democritus as an acknowledged precursor of Epicurus, see Plut. Col. !!*#6–@; for Epicurus’

reserved praise of him in On nature, see pp. !)%>& below. Epicurean attacks on Empedoclesinclude those of Hermarchus (see Longo Auricchio (!"##), pp. ++–$&, "%–", !%(–(*, and VanderWaerdt (!"##), pp. #"–"*, n. !&) and Colotes (Plut. Col. !!!!@ /.); see also Cic. ND , %", Diogenesof Oenoanda + ,,–,,, Smith (!""%), with the further passages assembled by Vander Waerdt. Inmy view (Sedley (!"$+a)) Epicurus’ attitude to his predecessors was more respectful and lenientthan that adopted by his followers, but it undoubtedly showed enough coolness to authorise andencourage their attacks. 70 , +%–$", ,,, !–%%, - "–!&.

That there is something, singular or plural, that somehow persiststhrough all cosmogonical and other changes is common ground for allphysical philosophers from Anaximander on. No doubt Empedocles’elements were more emphatically unchanging than those of his prede-cessors. At least, he says that as the elements intermingle they bothbecome di/erent things at di/erent times and remain always alike('!$.&)–(). He probably means that they form di/erent compound sub-stances but nevertheless retain their own distinctive properties in themixture. But other interpretations were possible – for example, that inmixtures the elements do retain their original properties, but that theseremain dormant until the compounds separate out again. And, at anyrate, I see little sign that Lucretius was prepared to give him the benefitof the doubt on this point. In criticising the four-element theory, hemakes no gesture of respect even for the well-advertised indestructibil-ity of Empedocles’ elements ('#, '", '!%): on the contrary, his principalground for rejecting the theory is that stu/s like earth, air, fire, and waterare inevitably perishable (, $(&–+%). As for their unchangeability, he men-tions this as no more than a possible interpretation of the theory, andone that would rob it of what little explanatory power it has (, $$*–#!).

Does Empedocles fare any better in Lucretius’ eyes as a champion ofanti-teleology? It cannot be denied that Aristotle casts him in that role:in defending the teleological structure of organisms, Aristotle contrastshis view with the zoogonical thesis of Empedocles that originally a set ofrandomly composed monsters sprang up – graphically described byEmpedocles as ‘ox-children man-faced’71 – of which only the fittest sur-vived. This anticipation of one of the principles of Darwinism hasearned Empedocles widespread respect, including, it is sometimes sug-gested, the respect of the Epicureans. For Lucretius testifies (- #&$–$$)that they adopted a similar-sounding theory of the survival of the fittestas their basis for the origin of species.

I would not want to deny the probability of a historical link betweenthe Empedoclean and Epicurean theories. But it is a large leap from thatto the supposition that the Epicureans acknowledged a debt toEmpedocles. Indeed, it can be precisely in those cases where a school isdrawing on the ideas of another that it is most at pains to minimise theresemblance and to stress its own originality. This appears to have beenthe Epicurean attitude to the Empedoclean theory of evolution.Plutarch72 tells us explicitly that the Epicureans derided Empedocles’

&. Furley’s thesis !"

71 Empedocles '+!.%. Cf. Aristotle, Phys. !"#b&%, !""b!*–!%, PA +)*a!"/. 72 Plut. Col. !!%&'.

‘ox-children man-faced’. And well they might, for Empedocles’ mon-sters were themselves the bizarre product of random combinations oflimbs and organs that in an even earlier stage had sprung up and wan-dered about on their own!73 There is nothing like this in the Epicureantheory, as we hear about it from Lucretius; and I can see no attempt inLucretius book - to restore to Empedocles the credit which theEpicurean school traditionally denied him.74

Indeed, since Lucretius certainly knew Empedocles’ physical poem atfirst hand and did not have to rely exclusively on Aristotelian-influenceddoxography,75 it certainly should not be assumed that he read Empedoclesas a pioneering opponent of teleology. If Aristotle chooses Empedoclesrather than the far more suitable Democritus for that role, it is surelybecause Empedocles, perhaps alone among the Presocratics, has actuallysupplied him with an illustration of what a non-teleological explanationof an organism would look like. It does not follow that Empedocles’ ownintention, taken in context, came over as anti-teleological.76 As is wellknown, he is supposed to have postulated four stages of animal evolution,of which the compounding of the ox-children man-faced was only thesecond. Either in the first stage, that of solitary animal parts, or perhapsin the third stage, that of the so-called ‘whole-natured forms’, hedescribed the creation of individual animal parts in terms that couldhardly have won him the friendship of an anti-teleologist like Lucretius.In '#), already mentioned above, Empedocles describes how Aphrodite77

cunningly created the eye, just like someone fitting together a lantern forthe preconceived purpose of lighting their way at night. Even if one stripsfrom this the figurative personification of Love as a divine artisan, one isleft with the impression of an intelligent and purposive creative force. Thearchitectonic role of Love in Empedocles’ cosmic cycle makes it a veryhard task indeed to portray him as a pure mechanist.

%* !. The Empedoclean opening

73 Empedocles 3$%, '($.74 Furley (!"$*), p. +! with n. !(, supports his thesis with the claim that Lucretius - #&$–)! is a trans-

lation of Empedocles '($. Although it may pointedly recall the Empedoclean lines, it is hardlya translation. Where Empedocies describes isolated limbs, Lucretius describes whole organismswith congenital defects – and that represents a crucial di/erence between the two zoogonicaltheories.

75 Cf. Clay (!"#&), pp. %%–&, %#"–"* nn. )&–). Rösler (!"$&) correctly stresses Lucretius’ use of dox-ography in his critique of Empedocles at , $!)–#%"; but this is, I believe, a special case, in so faras the passage is almost certainly based on Epicurus’ own criticism of earlier physical theories inOn nature .,- and .-, which in turn will have relied heavily on Theophrastus’ Physical opinions (seeCh. ), §!*; Ch. (, §); Ch. +, §$).

76 Teleology was not in Empedocles’ day an issue on which sides had to be taken. In what follows,I am describing the impression he was likely to make on later readers attuned to such a debate.

77 '#+ confirms that Aphrodite was the artisan in question; see Sedley (!""%b).

Why, then, does Lucretius nevertheless speak approvingly ofEmpedocles’ ‘discoveries’ (, $&%–&)? To see this in perspective, it is impor-tant to note that only four lines later he speaks with equal approval ofthe ‘discoveries’ of other, unnamed natural philosophers whom hebrackets with Empedocles. Lucretius is not, in e/ect, singling outEmpedocles as a uniquely important authority but is expressing anEpicurean’s qualified respect for the work of the Presocratic naturalphilosophers in general. Following Epicurus, he applauds the Presocratictendency to seek physical, as opposed to theological, explanations forsuch cosmic phenomena as celestial motions, eclipses, and earthquakes.The Epicurean school’s method of handling these phenomena was tocatalogue with approval all the available physical explanations of each,adding that any or all might be correct, so that to choose between themwould be arbitrary and unscientific. Both Epicurus, in his Letter toPythocles, and Lucretius, in books - ((*"–$$*) and -,, thus come to list aspossibilities a range of explanatory theses deriving in large measure fromthe Presocratic philosophers, including Empedocles. For example, bothEpicurus (Letter to Pythocles !*!) and Lucretius (-, %*)>!%) accept as one ofthe possible explanations of lightning the thesis of Empedocles (3+&)that it is fire from the sun trapped in the clouds. It is, I am convinced,only at this level of detail that the Epicureans, Lucretius included, areprepared to applaud the ‘discoveries’ of Empedocles.

$. 6;46<81:69 39 :,56737D @876'637

If, then, Lucretius is not thanking Empedocles for the content of theDRN, perhaps he is thanking him for its form. There are, after all, well-recognised formal correspondences between the two hexameter poems.

Take first their titles. De rerum natura is usually thought to translate !"#$!%&" '"(), a title conventionally assigned to many Greek cosmologicaltexts, including Empedocles’ physical poem, as well of course as beingthe title of Epicurus’ great prose treatise on which, I shall argue inChapters )–(, Lucretius was relying. As a matter of fact, though, one latesource78 reports Empedocles’ title as On the nature of the things there are (!"#$!%&" '"() -(' . /% .-(.), which would be closer still to De rerum natura. Thereis no independent evidence to confirm this title, but it seems highly plau-sible. The simple ‘On nature’ is so widespread that it has been suspectedof being, at least for fifth-century '1 authors, no more than a standard

'. Empedocles as literary forebear %!

78 Suda, s.v. ‘Empedocles’!&!3% DK.

title assigned to their works by later scholars.79 But someone likeEmpedocles who wrote at least two poems, not to mention prose works,is less likely to have left them untitled, and we have seen no reason notto accept the title Katharmoi as entirely authentic. As for On the nature ofthe things there are, his near-contemporary Melissus published a work enti-tled On nature, or on what there is (!"#$! %&" '"() 3) 6"#$! -/&' /% .-/)), the sin-gular ‘what there is’ proclaiming his Eleatic monism. That Melissus’reported title is authentic is confirmed by the parody published byEmpedocles’ follower Gorgias, On what there is not, or on nature (!"#$! -/&' µ3!/% .-/) 3) 6"#$! %&" '"()). Against this background, Empedocles’ choice ofOn the nature of the things there are, with its plural form -(' . /% .-(., as a titlefor what was above all else a pluralist manifesto, makes ready sense.

Apart from the titles, there are other striking formal resemblances. Inparticular, Lucretius’ poem is addressed to a friend, Memmius, asEmpedocles’ physical poem is to his friend Pausanias. And both atcertain points turn to address an invocation to the muse Calliope.80

I am now ready to unveil my own hypothesis: the proem of the DRN is,and is meant to be recognised as, an imitation of the proem to Empedocles’ physicalpoem.

The letter of Cicero with which I opened the present chapter consti-tutes strong evidence that contemporary readers could be expected torecognise this imitation, if such it was. For it attests a literary climate inwhich Empedocles was on the list of acknowledged Greek authors,81

familiar to the well-educated either through direct acquaintance orthrough Latin translations and imitations. (Even if other Roman literatishared Cicero’s inability to struggle through to the end of Sallustius’Empedoclea, many could be assumed, like him, at least to have read theopening.) Above all, it shows us Lucretius being thought about by hiscontemporaries in an Empedoclean context.

On my hypothesis, Lucretius’ purpose is to establish from the outsetthe precise Greek literary mantle he is assuming (rather as Virgil’s Aeneidannounces with the opening words arma virumque cano that it will be acombined Iliad–Odyssey). Lucretius, in his poetic manifesto at , "%!–(*and his appreciation of Ennius’ pedigree at , !!$–%+ (see pp. &!>% below),shows himself to be no less concerned with literary pedigree than otherRoman poets of his era.

%% !. The Empedoclean opening

79 Schmalzriedt (!"$*). 80 Empedocles '&, '!&!; Lucretius -, "%–(. See Clay (!"#&), pp. %(&–$.81 For a judicious discussion of Ennius’ possible use of Empedocles, see Skutsch (!"#(), pp. !+*, !+)

n. !#, %+*, &")/., $(#. Ovid’s extensive use of Empedocles in the speech of Pythagoras in Met..- is impressively discussed in Hardie (!""().

To amplify the hypothesis: Lucretius is imitating Empedocles’ proem,but adapting it, as he goes along, (a) to a Roman patriotic theme, and (b)to Epicurean philosophy, at the same time steering us gently away fromEmpedocles’ actual doctrines. His object? To announce himself as theRoman Empedocles – the great Roman poet of nature. In short, he is layingclaim to a literary, not a philosophical, heritage. For there can be littledoubt that it was to Empedocles, rather than to the only other availablecandidate, Parmenides, that Lucretius looked as his great Greek fore-bear in the tradition of cosmological poetry. This was certainly thecomparison that regularly occurred to Roman readers,82 and rightly so.(I shall be exploring Empedoclean features of Lucretius’ writing not onlyin the present chapter but also in Chapter %, §(, and in Chapter $, §+.)

A glaring weakness of this hypothesis will already be obvious. We donot have the proem to Empedocles’ On nature.83 How then can we sayanything at all about its resemblance or otherwise to Lucretius’ proem?

My answer is twofold. First, we are not altogether without evidenceabout its content, as I hope to show. Indeed, there is little doubt thatsome of our familiar fragments of Empedocles are in fact from it.Moreover, thanks to the exciting papyrus find that has been made sinceI first formulated the argument of this chapter,84 we now have consider-ably more of Empedocles’ proem than was available even a few yearsago. The new fragments are believed all to come from a single scroll,which contained book , of Empedocles’ physical poem.85

Second, if the proposed hypothesis proves capable of explaining fea-tures of Lucretius’ proem that otherwise remain inexplicable, that initself would provide some degree of confirmation.

I !–"#. I shall begin my defence of the hypothesis with a re-examinationof the opening lines (translated above p. !$):

Aeneadum genetrix, hominum divomque voluptas,alma Venus, caeli subter labentia signaquae mare navigerum, quae terras frugiferentisconcelebras . . .

The linguistic case for a direct Empedoclean model seems to me a ratherstrong one. The first two words are, of course, a distinctively Roman

'. Empedocles as literary forebear %&

82 E.g. Quintilian Inst. , ).); Lactantius, Div. inst. ,, !%–!).83 Van der Ben (!"$() o/ers his own wholesale reconstruction of Empedocles’ proem. Most of it

rests on guesswork. My grounds for rejecting it will simply be the arguments I o/er below foraccepting a di/erent reconstruction, based largely on Lucretius. 84 In Sedley (!"#"a)

85 See above p. !*.

invocation. But hominum divomque voluptas already bears an Empedocleanfingerprint. The identical phrase recurs, with a small change of syntax,at -, "%–(, in an address to Calliope that has long been recognised as anEmpedoclean touch on Lucretius’ part.86 Hominum divomque could trans-late some variant on the regular hexameter ending +& .5#(' . -" 1"(' . -",used in Homer’s formulaic designation of Zeus as ‘father of men andgods’. Such reworkings of Homeric locutions are an integral feature ofEmpedocles’ poetry.87 And voluptas picks up =31/'&" .3, ‘Delight’, usedby Empedocles, like ‘Aphrodite’, as a title for his goddess Love ('!$.%)).Next, alma, ‘nurturing’, might represent >"$"5(#/), ‘life-giving’, anattested Empedoclean epithet for Aphrodite ('!(!); but as a matter offact there is a much better candidate among the new fragments ofEmpedocles. These include (in a fragmentary context) the adjective%&-+" 7µ$/), ‘nurturing’, commonly used in Greek poetry as a stockepithet for divinities.88 In addition to being virtually synonymous withthe Latin almus, it also shares its leading syllable. It would be easy toimagine ‘0&" 6#$ %&-+" 7µ$" . . .’ as an Empedoclean line-beginning,matching Lucretius’ alma Venus . . .

We then proceed, in %–" (quoted p. !$ above), to the elaborate doubleinterweaving of the four elements into the hymn. For Lucretius to expectany reader to identify these as the Empedoclean four in the very openinglines of the poem, without any prior clue, would be wildly optimistic. Itis far more credible that he found them already present in hisEmpedoclean original. We have already noted that interweaving the fourelements into a descriptive passage is an authentic Empedoclean device.

Line & is remarkable for its pair of compound adjectives: quae mare nav-igerum, quae terras frugiferentis . . . Lucretius has a well-known penchant forthese quasi-Greek formations,89 and we will see in the next chapter howthey sometimes combine with Greek loan words to build up an evocativecontext that transports his reader to the Greek world. But there are twounusual features of this particular pair. First, both accurately translateactual Greek compound adjectives – respectively, navigerum!.+&'$"6/#/.and frugiferentis!,+#6/%/" #/&) (or a participial equivalent from,+#6/%/#"$'.).90 Second, although the bold deployment of compound

%) !. The Empedoclean opening

86 -, "), Calliope, requies hominum divomque voluptas.87 See the seminal study by Bollack (!"+(–"), , %$$/. Aristotle, in his lost On poets, called Empedocles

$?µ3#$,/" ) (DL -,,, ($).88 There seems little possibility that in the actual fragment the adjective is serving this role.89 See Bailey (!")$), , !&%/. for a convenient catalogue of Lucretius’ compound adjectives.90 This is not unique in Lucretius – for instance florifer (,,, !!) corresponds to +& .1/%/" #/) and ignifer

(- )(" etc.) to 6&#%/" #/) – but I have spotted very few such cases.

adjectives in pairs, or even in trios, is among the most prominent featuresof Empedocles’ verse,91 it is one Empedoclean practice which elsewhereLucretius studiously avoids. In his whole poem, in fact, such a groupingoccurs uniquely here.92 The double idiosyncrasy suggests that in line &some exceptional motivation is at work. The supposition that Lucretiuswas consciously seeking to capture and reproduce in Latin an actualEmpedoclean line would provide such a motivation. In fact, his line prac-tically tumbles unaided into a characteristically Empedoclean Greekhexameter: 6/" .-/. .+&'$"6/#/. ,+$! 4+$"+) ,+#6/%/#/&" '+).93

Lines !*–%* present Lucretius’ entrancing portrait of the repro-ductive frenzy which Venus inspires throughout the animal kingdom inspring. Then in line %! Venus emerges as the controller of all natura, inthe passage (%!–#) which also, in line %(, e/ectively delivers to us the titleof the poem:

quae quoniam rerum naturam sola gubernasnec sine te quicquam dias in luminis orasexoritur neque fit laetum neque amabile quicquam,te sociam studeo scribendis versibus essequos ego de rerum natura pangere conor %(Memmiadae nostro, quem tu, dea, tempore in omniomnibus ornatum voluisti excellere rebus.quo magis aeternum da dictis, diva, leporem.

Since you alone control the natura of things, and without younothing springs forth into the realm of light or becomes joyfuland delectable, I am eager for you to be my partner in writingthe verses which I am trying to set out about the nature of things(de rerum natura) for our friend Memmius, whom you, goddess,have wanted to be at all times outstanding in all things. All themore then, goddess, bestow on my words an everlasting charm.

As Diskin Clay has pointed out, in this context natura (%!) tends towardsthe sense ‘birth’ (through its association with nasci, ‘be born’) rather than

'. Empedocles as literary forebear %(

91 Empedocles '%*.+–$, '%!.!!–!%, ')*, '+*–!, '$+.!–%, etc. The forms ""µ6"5/" %&77+ and ""µ6"5/" ,+#6+ ('$$.!) are unique to Empedocles, and suggest a heightened consciousness of theetymology of his own name as a further compound adjective, ‘eternally renowned’.

92 The closest groupings of compound adjectives, outside , &, are at ,, !*#!–& and - #+)–+. In bothcases they occur two lines apart, qualifying items which are respectively first and third in a list.

93 In Sedley (!"#"a) I placed the first two words in the reverse order (perhaps metrically preferable– but cf. e.g. Empedocles '#).+) .+&'$"6/#/. 6/" .-/. ,+$! 4+$"+) ,+#6/%/#/&" '+). Gisela Strikerhas persuaded me to avoid the jingle which this creates in the second to third foot. Her ownsuggestion is to retain my original order but change the cases throughout to genitives. This couldwell be preferable. I retain the accusative merely in order to maximise the isomorphism with theLucretian line.

simply ‘nature’, thus echoing Empedocles’ characteristic use of %&" '$)with precisely the same shift from the more familiar ‘nature’ to ‘birth’.94

Leaving aside these linguistic and conceptual echoes, it is in any caseeminently plausible that Empedocles’ poem should have opened with ahymn to Aphrodite. Hesiod’s Works and days, with its opening hymn toZeus, would constitute ample precedent within the tradition of didacticpoetry; and it goes without saying that Aphrodite would be Empedocles’preferred divinity. In '!%# he makes it a mark of the Golden Age, inwhich among other things there was no animal slaughter, that Aphroditewas the only divinity worshipped:

Nor did they have Ares or Strife as a god, nor was Zeus or Cronos or Poseidontheir king, but Cypris was queen . . . Her they propitiated with pious images . . .

I am not suggesting that this fragment itself comes from Empedocles’proem. But it does reveal a feature of his religious thought that Lucretiuscould himself use to advantage – namely the idea that the identity of aperson’s divinities is a function of that person’s own moral state.95 If youare a peaceful person, Ares is not your god, but Aphrodite is. Lucretius,as an Epicurean, must hold the somewhat similar view that the gods’true nature is peaceful, and that people’s tendency to endow them withangry and warlike temperaments is a projection of their own moralmaladjustment.96 The essence of god is blessed detachment; anger, jeal-ousy and the like are accretions misleadingly superimposed by us on thatessence.

This may o/er us a lead on the much debated lines ))–", in whichLucretius presents the correct Epicurean view of the gods as tranquiland detached:

omnis enim per se divom natura necessestimmortali aevo summa cum pace fruatur )(semota ab nostris rebus seiunctaque longe;nam privata dolore omni, privata periclis,ipsa suis pollens opibus, nil indiga nostri,nec bene promeritis capitur neque tangitur ira.

For the entire nature of the gods, in its essence, must of neces-sity enjoy everlasting life along with perfect peace, removed andfar separated from our a/airs. Without any pain, without

%+ !. The Empedoclean opening

94 Empedocles '#. See Clay (!"#&), pp. #&–"(, with the parallels he cites at p. &*# n. %".95 Cf. '!$.%&, where Love is ‘she by whom mortals think friendly thoughts and perform peaceful

deeds’.96 See Long/Sedley (!"#$), , !&"–)". The point stands whether or not, as argued there, Lucretius

was wrong to understand Epicurus’ gods as objectively real life-forms.

dangers, strong through its own resources, with no need of us, itis neither won over by favours nor touched by anger.

These lines occur also at ,, +)+–(!, where they are superficially muchmore at home, and many editors believe that they are an intrusive glossin the proem:97 it seems anomalous for Lucretius to stress the totaldetachment of the gods from human a/airs directly after his prayer toVenus to intervene and save the Roman republic from war. And yet thesudden reversal is too characteristic of Lucretius to be lightly dismissed.Even when these same lines recur in book ,,, they are used similarly toreverse the religious implications of what precedes, this time a seductiveallegorical reading of the worship of Cybele as symbolising (at least inits last lines) her direct interest in human a/airs.

Imagine now in addition an original Empedoclean proem in whichAphrodite, as Love, is asked to propitiate Ares, as Strife. WhatEmpedocles would have intended by this is not so much an attempt tointerfere with the inevitable progression of the cosmic cycle, as a pleato human beings to let their peaceful tendencies calm and suppress thebloodthirsty side of their nature.98 If so, Lucretius would welcome thisessentially moral use of myth and prayer, and could readily apply it tothe current war-torn state of his own country. But since Empedoclesregards Ares/Strife as a real, if less palatable, god, Lucretius might verynaturally want to add an Epicurean corrective: that Venus’ hoped-forpropitiation of Mars represents no more than people’s return to the onetrue conception of the divine nature as tranquil and detached, insteadof angry and warlike. Hence the connexion of thought found in thetext: Venus, make Mars peaceful, because that alone is the essentialnature of divinity (omnis enim per se divom natura . . .). Or, translated intoEpicurean moral terms: Romans, let your belief in a peaceful god over-come your belief in a warlike god, because peacefulness is the trueessence of godlike happiness. The connexion of thought could nodoubt have been made clearer; but I would be reluctant to robLucretius of this important Epicurean modification to Empedocleantheology.

'. Empedocles as literary forebear %$

97 See the arguments marshalled by Courtney (!"#$), pp. !!/.98 Eustathius (Od. &!*.&&/., ad Hom. Od. -,,, &+$) may imply that Empedocles used the myth of the

union of Aphrodite and Ares as an allegory for friendship; and since there is no stage within thecosmic cycle itself at which Love and Strife unite, the likeliest location for that piece of symbol-ism would indeed be his proem. However, Eustathius’ words may mean no more than that someallegorists proposed an Empedoclean interpretation of the myth; cf. Heraclit. Alleg. Hom. +".#,and Bu0ère (!"(+), pp. !+#–$%.

By this stage, it should be noted, I am no longer suggesting direct trans-lation or line-by-line imitation of Empedocles’ proem on Lucretius’part, but the deployment of the same sequence of themes as occurredin it, for increasingly Epicurean purposes.

I $%–&!. The next section of Lucretius’ proem is a programmaticaddress to Memmius. He asks Memmius to give him his full attention –perhaps an echo of the passage that contained Empedocles’ survivingline ‘Listen to me, Pausanias, son of wise Anchiteus’ ('!). He then pro-ceeds to outline the content of the poem. He will explain to Memmiusthe character of the heaven and the gods, and the elements ((+–$)

unde omnis natura creet res auctet alatquequove eadem rursum natura perempta resolvat.

. . . from which nature creates, increases, and nurtures all things,and into which that same nature once more resolves them whenthey are destroyed.

After this he spends four lines naming his cosmic first principles (geni-talia corpora, semina rerum, etc.).99

This dual process, whereby things combine and are once more dis-solved into their constituents, bears a strong formal resemblance toEmpedocles’ own programmatic description in '!$, a passage that isexplicitly attested by Simplicius as coming from the opening of the phys-ical poem.100 Not only do we have Simplicius’ attestation to that e/ect,but the new papyrus find includes a fragment which at its beginningcoincides with lines &!–( of '!$ and then continues it for another &) lines.Thanks to a line number (=!&**) preserved in the margin at the end ofthe new fragment, we can now say that the opening couplet of fragment!$ was probably lines %&&–) of the poem:

5$"67 & "&#""(· -/-"! µ"! . 4+! # "*. 3&& @3" 13 µ/" ./. "$# .+$"&, 67"/" .(., -/-"! 5 & +&# 5$""%& 67""/. & "&@ "$./! ) "$# .+$.

I will tell a double tale. For at one time there grew to be just onething from many, and at another it grew apart once more to bemany out of one.

The symmetrical two-way nature of the process is emphasised repeat-edly in similarly balanced antitheses for a further fifteen lines,101 after

%# !. The Empedoclean opening

199 See further, Ch. %, §&; Ch. $, §§& and +.100 Simplicius, In Phys. !+!.!)–!(; see preamble to '!$ in Diels/Kranz.101 In the new fragment which continues the passage there appear to be further returns to much

which Empedocles, like Lucretius, proceeds to name the cosmic princi-ples underlying the process – the four elements, plus Love and Strife.Empedocles’ repetitiveness, a device for emphasising the eternality ofhis cosmic cycle, is understandably not reproduced by Lucretius. But inother respects the formal parallelism of the two programmatic passagesis striking.

It begins to look highly plausible that Empedocles’ proem to On nature,having opened with a hymn to Aphrodite, then continued with a pro-grammatic address to Pausanias, of which '!$ formed a part. Such astructure would, naturally enough, mimic the opening of Hesiod’s Worksand days, where a short hymn to Zeus is immediately followed by a per-sonal address to Perses.

Given our new knowledge of Empedocles’ line numbering, thishypothesis would mean that the hymn to Aphrodite and the personaladdress to Pausanias occupied around %&* lines.102 Is this implausiblylong? I do not think so. Empedocles is a wordy and repetitive writer, asthe new fragments amply confirm. And we have no way of guessing howmuch personal detail was included in the address to Pausanias (theHesiodic model would have permitted plenty).103

!.&'–(#. Lucretius’ next section is his praise of Epicurus’ intellectualachievement.104 At a time when mankind was wretchedly oppressed byreligion, a certain Greek became the first (primum Graius homo, ++) to standup against its tyranny. Such were his mental powers that he was able tobreak through the ‘flaming walls of the world’ and traverse with his intel-lect the measureless universe. By reporting back to us the laws that bindand limit natural processes, he has broken the power of religion.

Once more there is a clear Empedoclean model, '!%", almost cer-tainly referring to Pythagoras:

'. Empedocles as literary forebear %"

the same two-way description of change, as well as a brief preview of the beginning of the cycle,to be resumed in earnest in '&(.

102 The numbering was not a system of textual citation but the scribe’s way of keeping count ofthe number of lines he was due to be paid for. It is therefore (I understand from Drs. Martinand Primavesi) possible that his numeration included the title – in which case the number oflines preceding '!$ would be slightly reduced. Hence I have rounded my figure down to ‘around%&* lines’.

103 Full discussion of the opening part of the poem and its possible contents must await publica-tion of the new fragments in Martin/Primavesi (!""#).

104 I am unpersuaded by the proposal of Edelstein (!")*) that the reference is a general one to thePresocratic physical tradition. The proems to books ,,,, - and -, supply ample evidence ofLucretius’ belief that Epicurus was the first to make the crucial breakthrough, scientific as wellas moral.

There was among them a man of extraordinary knowledge, possessing a vasttreasury of understanding, and master of every kind of wise deed. For when hereached out with all his understanding he easily saw everything there is, overten and twenty human generations.

As Furley has pointed out,105 the Lucretian passage unmistakably recallsthe Empedoclean. Both men are great historical figures, too august to benamed. And both are praised for their intellectual achievement in break-ing through the boundaries of ordinary human experience – Pythagorasfor his recollection of his former incarnations,106 Epicurus for his graspof the nature of the infinite universe beyond our own world.

Doctrinally, it should be noticed, Lucretius and Empedocles areveering ever further apart. Epicurus’ discoveries, which secured hisvictory over religion, are taking the place of an Empedoclean religiousdoctrine that is anathema to Lucretius, the doctrine of transmigration.

I )%–!%!. There follows Lucretius’ direct attack on the evils of religion,illustrated with the example of Agamemnon’s sacrifice of his owndaughter Iphigeneia.

Furley is right to point out the clear reminiscence of Empedocles '!&$,in which Empedocles attacks the sin of animal slaughter with theexample of a father unwittingly sacrificing his own son, who has trans-migrated into the body of an ox. There is no detailed linguistic imita-tion, but the close functional parallelism of the two pathetic scenes ofsacrifice should leave little doubt that the one passage is written with theother in mind. (Lucretius’ description does not, incidentally, appear tobe directly modelled on any of the accounts of Iphigeneia’s sacrificeextant in Greek tragedy.)

That '!&$ came from Empedocles’ physical poem, and not from theKatharmoi, was until recently a highly unorthodox proposal. Now,however, we have the new papyrus fragments from the opening book ofthe physical poem, and they include '!&", where the speaker expresseshis regret that he did not perish before his wicked complicity in theslaughter and consumption of animals. So close is the thematic linkbetween '!&$ and '!&" that they have regularly been assumed to derivefrom a single original context. That context can now be identified asbook , of the physical poem, and very probably its proem.107

&* !. The Empedoclean opening

105 Furley (!"$*). See also Burkert (!"$%), p. !&$; Hardie (!"#+), p. &" n.!$.106 For the tradition of Pythagoras’ multiple incarnations, see Burkert (!"$%), pp. !&$/.107 It would be unwise, in the present transitional state of Empedoclean scholarship, to insist that

'!&" itself came from the proem. The work of Martin and Primavesi, based on its admittedly

I !%'–*$. Lucretius continues with a warning to Memmius not to beconfused by superstitious tales, such as those about the survival andtransmigration of the soul.

Why did he choose to include the topic of transmigration in hisproem? In view of all the Empedoclean echoes we have already wit-nessed, it can hardly be a coincidence that Empedocles likewise outlinedhis beliefs about transmigration in his proem. It is here that I can at lastcall upon the findings of §§%–& of this chapter, in which I defended theattribution of '!!(, Empedocles’ explanation of his doctrine of trans-migration, to the proem of his On nature. If I am right, and Lucretius’attack on transmigration is an intended counterpart to Empedocles’exposition of the doctrine at the corresponding point in his own proem,he has now moved yet further in distancing himself philosophically fromhis principal literary model. Where previously we saw him adaptingthemes from Empedocles’ proem to his Epicurean philosophy, he is nowpresenting his own matching passage not as an adaptation ofEmpedocles but as a direct antidote to his teachings.

In the course of making this point, Lucretius names Ennius as theauthor of just the kind of confusion that he is condemning. SomehowEnnius managed to believe both in transmigration and in the sojourn ofdeparted souls in Hades. The latter is an explicit reference to the dreamin which Ennius, in his own proem to the Annales, had described meetingthe shade of Homer. However, Lucretius allows, Ennius must be givenhis due as the great innovator who brought Greek poetry to the mediumof the Latin language: ‘. . . our own Ennius, the first to bring down fromlovely Helicon the enduringly-leaved crown which was to achieve glorythroughout the Italian peoples’.

Here we should note Lucretius’ concern with literary pedigree, andspecifically with Ennius’ pioneering role in the task which he is himselfnow engaged in, that of re-creating for Latin readers the poetic genresof the Greeks. There is in fact little doubt that the dream passage inEnnius’ own proem is being directly echoed in Lucretius’ lines.108

Lucretius is here distancing himself from Ennius’ beliefs, while reveringhis poetry, in a way that pointedly parallels his treatment of

'. Empedocles as literary forebear &!

fragmentary context in the papyrus, is currently favouring a location later in the poem, but withthe further inference that the daimon’s original sin and subsequent fate must already have beendescribed in the proem. Given that '!!(, which I have argued comes from the proem of thephysical poem, ends up with Empedocles’ declaration that he is himself one of the fallendaimons – ‘I too am now one of these, a fugitive from the gods and a wanderer, who trust inraving Strife’ (above p. #) – the story of his own downfall could very naturally accompany it.

108 See Skutsch (!"#(), pp. !%, !((, Clay (!"#&), p. &!* n. )#, Waszink (!"(*), pp. %%)–(.

Empedocles.109 Anyone who may doubt the appropriateness of my dis-tinction between a ‘literary’ and a ‘philosophical’ debt to Empedoclesshould note that just such a distinction is operating here with regard toEnnius.

I !*&–"$. Finally we come to the closing section of Lucretius’ proem, inwhich he stresses the magnitude of his poetic task – a task made harder,he says, by the deficiencies of the Latin language and the novelty of thesubject matter. It is overwhelmingly tempting to correlate this with thegroup of fragments ('#–!!, '!() in which Empedocles deplores theimprecision of ordinary language in speaking of things’ being born anddying, where there is in reality only combination and separation, butadds that he will nevertheless follow the convention. The shared themeof how to cope with the deficiencies of one’s own language110 constitutesa strong link between the two passages. We have no explicit attributionof these fragments to Empedocles’ proem, but '# is at least cited bySimplicius as coming from the opening book of On nature; and evenwithout the Lucretian parallel the proem has always seemed the likeliestlocation.

# . 6;46<81:69 ’ 4786;

A little earlier we arrived at the informed guess that Empedocles’ proemto On nature opened with a hymn to Aphrodite, followed by a program-matic address to Pausanias. We can now, in the light of our subsequentfindings, ask how it went on.

Lucretius’ proem o/ers the following sequence of topics in its latterpart (+%–!)():(a) Epicurus’ intellectual achievement and defeat of religion;(b) the evils of religion;(c) the folly of uncritically believing religious tales, such as those about

transmigration;(d) the magnitude of Lucretius’ poetic task.The thematic link between the first three is a perfectly satisfactory one,and the last is, if not directly connected, still an appropriate enough

&% !. The Empedoclean opening

109 The point is redoubled if, as seems likely, Ennius’ beliefs were themselves influenced byEmpedocles: cf. Hardie (!"#+), pp. !$–%%, $"–#&.

110 Empedocles does not in the surviving fragments specify that the deficiency is one of his ownlanguage, Greek, rather than of language as such. But his contemporary Anaxagoras ('!$)makes the same point with explicit reference to Greek usage, and that was a natural enough wayto understand Empedocles too.

topic to address in a proem. And yet there is something disquietinglyspecific, not to say arbitrary, about the third topic. Why go to suchlengths to criticise the transmigration thesis in particular, when there arecountless other o/ending doctrines? Is it merely in order to introduce aheavily qualified tribute to Ennius?111 My preferred explanation hasbeen that the choice and sequence of topics was in some measure dic-tated by a further consideration, Lucretius’ desire to reproduce the the-matic structure of Empedocles’ proem. One incidental by-product hasbeen the materials for a scissors-and-paste reconstruction of the latterpart of Empedocles’ own proem. Now stand back and look at the result.We have supplied Empedocles with the following fluent sequence oftopics:(a) Pythagoras’ achievement in recalling past incarnations: an appeal to

authority for the doctrine of transmigration;(b) the evils of animal slaughter, illustrated by the unwitting sacrifice of

a deceased and transmigrated son: the moral importance of the doc-trine of transmigration;

(c) the origin and nature of transmigration itself;(d) the folly of being misled by ordinary linguistic usage into supposing

that anything literally dies.This time the thematic coherence of the sequence (a)–(d) is extraordi-nary. It is much more tight-knit than the corresponding passage inLucretius, and tells a complete story of its own, one thematically paral-lel to the Lucretian passage, yet utterly unlike it in detailed content.What is more, the denial of literal birth and death with which it ends notonly gives a philosophical basis to the transmigration doctrine that pre-cedes it, but also prepares the ground for the physical exposition tofollow, which will likewise be founded on the Parmenidean tenet thatnothing literally comes to be or perishes.112

This emergence of a reconstructed Empedoclean proem with acoherence and vitality of its own is an additional windfall, which lends

(. Empedocles’ proem &&

111 I do not mean to deny that direct reaction to Ennius plays a significant part in this passage. Myquestion concerns the overall thematic structure of the passage. I would tentatively add that,even if Lucretius were thought to be reacting to current philosophical trends (which I doubt –see Ch. &), he would still be unlikely to feel impelled to pick transmigration as a target. To judgefrom the evidence of Cicero’s Tusculan disputations, the current revival of interest in Plato’simmortality doctrine played down reincarnation in favour of discarnate survival. Nor doestransmigration appear to be an attested feature of first-century '1 neo-Pythagoreanism (forwhich see Dillon (!"$$), pp. !!$–%!, and cf. Vander Waerdt (!"#(), esp. pp. &##–").

112 See especially '!%. The Parmenidean tenet seems to be applied by Empedocles indiscriminatelyto the soul’s survival and to the permanence of the elements: both equally are separated, notdestroyed. How coherent this conflation is is another question. Cf. especially Kahn (!"+*).

welcome support to my hypothesis about Lucretius’ proem, quite apartfrom what it promises to teach us about Empedocles himself.

" . 18?1:A9,8?

The nature of my case has been essentially cumulative. Every main stageof Lucretius’ proem has proved to correlate with an Empedoclean orig-inal. The first part reads as if it were closely imitating an Empedocleanhymn, while the remainder sustains a virtually unbroken series of the-matic links with known or attested passages of Empedocles. Moreover,every one of those Empedoclean originals can plausibly be located in theproem of his On nature, either on independent evidence, or through itsthematic coherence with passages that have already been located there.

Lucretius is the servant of two masters. Epicurus is the founder of hisphilosophy; Empedocles is the father of his genre. It is the unique taskof Epicureanism’s first poet to combine these two loyalties. And that taskis what gives his proem its very distinctive character.

&) !. The Empedoclean opening

1234567 %

Two languages, two worlds

! . :,?BA,9 5,1 48-675D

In the preceding chapter, we met at the end of Lucretius’ proem hisfamous apology on behalf of the Latin language (, !&+–)(), in which helaments the linguistic struggle that he faces (, !&+–"):

nec me animi fallit Graiorum obscura repertadi0cile inlustrare Latinis versibus esse,multa novis verbis praesertim cum sit agendumpropter egestatem linguae et rerum novitatem.

Nor do I fail to appreciate that it is di0cult to illuminate in Latinverse the dark discoveries of the Greeks, especially becausemuch use must be made of new words, given the poverty of ourlanguage and the newness of the subject matter.

In §§%–$ of this chapter I shall be considering how he handles this taskof Latinising the technical terms of Epicurean philosophy. In §§#–!& Ishall turn to his own poetic use of Greek loan-words and idioms. Thetwo practices will come out looking antithetical to each other. At the endI shall suggest how we are meant to interpret this antithesis. What maystart out looking like an issue of linguistic mechanics will turn out, if Iam right, to reveal a fundamental tension in Lucretius’ evaluation of hisown poetic and philosophical task.

% . 526 5612?,13:,5,69 8@ 42D9,19

By a ‘technical term’ I intend a single word or phrase, either speciallycoined or adapted from existing usage, and at least implicitly earmarkedby the author as his standard and more or less invariable way of desig-nating a specific item or concept within a discipline. Its sense must berecognisably di/erent from, or at least recognisably more precise than,any distinct sense that the same term may bear in ordinary usage. While

&(

medicine and mathematics were disciplines which had long possessedtechnical vocabularies, philosophy had been slow to catch up, acquiringlittle technical terminology before Aristotle. Nevertheless, Hellenisticphilosophies had become thoroughly technical in their terminology, andEpicureanism, despite its (misplaced) reputation as an ordinary-lan-guage philosophy, was very nearly as jargon-ridden as Stoicism. It mightin fact plausibly be maintained that the atomistic tradition from whichEpicureanism emerged had, in the hands of its fifth-century exponents,been the first philosophical movement to forge itself a technical vocab-ulary.

The Latinisation of technical Greek, at least in rhetorical treatises,was a familiar practice by the mid first century '1, when Lucretiuswrote. But from Cicero’s letters one may get the impression that wheneducated Romans were locked in philosophical discussion they preferredsimply to pepper their Latin prose with the authentic Greek terms. It wasnot until more than a decade after Lucretius’ death that Cicero com-posed his principal philosophical works, in which the Latin philosoph-ical vocabulary was largely forged.

A full-scale study of Cicero’s handling of this task is, as far as I know,yet to be written.1 Among many things it might help teach us is just whatis distinctive about Lucretius’ own near-contemporary e/orts to accom-modate Epicureanism within the Latin language. For the present letCicero speak for himself, as his cast of philosophical Romans muse ontheir task of Latinisation in the first book of the Academica (, %)–+):

‘. . . But the combination of the two they called “body” and, as one might putit, “quality” [qualitas]. You will permit us occasionally to use unknown wordswhen dealing with unfamiliar subject matter, just as is done by the Greeks, whohave been dealing with these subjects for a long time.’

‘We will,’ replied Atticus. ‘In fact it will even be permissible for you to useGreek words when you want, if you happen to find no Latin ones available.’

‘Thanks, but I’ll do my best to speak in Latin, except that I’ll use words like“philosophy”, “rhetoric”, “physics” or “dialectic” [philosophiam aut rhetoricam autphysicam aut dialecticam] – words which along with many others are now habitu-ally used as Latin ones. I have therefore named “qualities” the things which theGreeks call 6/$/" -3-"), a word which among the Greeks is itself not an every-day one but belongs to the philosophers. The same applies in many cases. Noneof the dialecticians’ words are from public language: they use their own. Andthat is a common feature of virtually all disciplines: for new things either newnames must be created, or metaphors must be drawn from other fields. If that

&+ ". Two languages, two worlds

1 However, I have not yet been able to consult Hartung (!"$*). There is some useful preliminarygroundwork in Reiley (!"*").

is the practice of the Greeks, who have already been engaged in these things forso many centuries, how much more should it be allowed to us, who are nowtrying to deal with these things for the first time.’

‘Actually, Varro,’ I said, ‘it looks as if you will deserve well of your fellowcountrymen, if you are going to enrich them not only with facts, as you havedone, but also with words.’

‘On your instigation then,’ he said, ‘we will venture to use new words if itbecomes necessary.’

Two features deserve particular attention. First, the simple translitera-tion of Greek words was, as the speaker Varro acknowledges, a familiarand accepted practice, albeit confined largely to the names of the disci-plines themselves, such as ‘dialectic’ and ‘rhetoric’.2 Second, Cicero pre-sents his colleagues as considering it highly commendable whendiscussing philosophy in Latin to coin the necessary technical jargon, ifpossible on the analogy of the Greek original, as in the pro/eredexample of qualitas for Greek 6/$/" -3) (pointedly reminding us of thewell-known passage in Plato, Theaetetus !#%a, where the technical term6/$/" -3) is itself invented).

In both respects Lucretius o/ers a stark contrast. Take the names ofdisciplines once more. The DRN is a poem about physics, whatLucretius’ own contemporaries were calling physica, yet nowhere in itcan that term or its cognates be found. Does Lucretius then have noname for the physical science he is practising? One clear case in whichhe does is at , !)#, where the proper Epicurean justification for the studyof physics is given: ignorant and superstitious fears are to be dispelledby naturae species ratioque. The phrase captures quite closely Epicurus’preferred term for physics, %&'$/7/4$"+, with naturae and ratio picking upits constituents %&" '$) and 7/" 4/) respectively.3 But in Lucretius’ rendi-tion it has lost all terminological technicality, and become a subtlydescriptive formula for the poem’s theme. Read actively, naturae speciesratioque no doubt denotes the rational philosophical procedure of‘looking at nature and reasoning about it’. But at the same time theLatin permits and even encourages the additional reading, ‘the appear-ance and rationale of nature’: such a rendition emphasises the power ofnature herself to confront us with the truth – a motif which Lucretiuswill be turning to good use in the poem. No strand in this web of

". The technicalities of physics &$

2 Cf. Cic. Fin. ,,, (. Weise (!##%) remains a very valuable survey of Greek loan-words in Latin, butdisappointingly brief on philosophical vocabulary (including, p. %)!, the de natura deorum (sic) ofLucretius!).

3 The link is confirmed by Cic. ND , %*, where the Epicurean Velleius speaks of physiologiam, id estnaturae rationem, an equivalence also repeated by his Stoic speaker at Div. , "*.

connotations goes beyond the potential significance of the one Greekword %&'$/7/4$"+.

& . 358;,1 -813'A:37D

Similarly with individual technical terms within his chosen discipline,Lucretius’ constant practice is to render Greek technicality neither withLatin technicality nor with mere transliteration, but with a range of hisown live metaphors. Take the case of ‘atoms’.4 Of the earlier Latin prosewriters on Epicureanism, we know only that Amafinius had rendered theterm corpuscula,5 although Lucilius’ reference to atomus . . . Epicuri ($(&Marx) shows that simple transliteration had long been another availableexpedient. Cicero, for his part, actually shows a strong preference for thistransliterated form, with occasional resort to corpuscula6 or to his ownprobable coinage individua, ‘indivisibles’. None of these is ideal.Transliteration of a term from within a discipline – as distinct from thename of the discipline itself – is a rare resort for Cicero, and savours ofdefeat. Corpuscula captures the minuteness of the atoms but not their all-important indivisibility. And individua su/ers in Cicero’s philosophicalprose from having to stand in for too many di/erent Greek originals: hehad already, in his paraphrase of Plato’s Timaeus (%!, %(, %$), used it torepresent +& µ""#$'-/), +& µ"#3" ) and +% ';$'-/), all terms with importantlydi/erent technical connotations both from each other and from ‘atom’.

Lucretius introduces his own set of terms for atoms in the proem tobook ,, ()–+!, more than )** lines before his first proof of their exis-tence:7 rerum primordia, materies, genitalia corpora, semina rerum, corpora prima.Unlike corpuscula, all these concentrate not on the smallness of atoms buton their role as the primitive starting-points from which other entities arebuilt up. In introducing them, he places the chief emphasis on theirdynamic generative powers, already indicated in the procreative implica-tions of materies (a derivative of mater), genitalia and semina. These implica-tions he then exploits in his first set of arguments, those againstgeneration ex nihilo, in the course of which he seeks to persuade us thatthe biological regularities which are evident at the macroscopic leveldepend on fixed materies or semina at the microscopic level. (I shall beexamining the strategy in detail in Chapter $, §&.) The metrically conve-

&# ". Two languages, two worlds

4 For the main evidence on Latin atomic vocabulary, cf. Reiley (!"*"), pp. &(–++.5 Cicero, Ac. , +. 6 ND , ++–$, ,, "), Ac. , +, Tusc. , %%.7 For the role of this passage, cf. pp. %#>" above, and pp. !"), %*! below, where the lines are quoted

in full.

nient transliteration atomi never so much as puts in an appearance. Butcorpuscula does crop up as an occasional variant in later books, especiallywhere their generative powers are not at issue.8 So does elementa, ‘letters’,a convenient equivalent for '-/$;"$'+ (‘elements’ but also morespecifically ‘letters’), which helps to reinforce Lucretius’ favoured analogybetween atomic rearrangement and alphabetic realignment.9 Hence ittends to occur in contexts where the ordering of atoms is in focus.10

) . S I M U L AC R A

A similar but more cautious metaphorical diversification of a single orig-inal Greek term is illustrated in book ,- by Lucretius’ range of renditionsfor "$%5(7+, ‘images’, the thin films of atoms which stream o/ bodies andcause vision. Lucilius, once again, had simply transliterated the word asidola ($(& Marx). Cicero and his Epicurean correspondent Cassius, dis-cussing the topic in )( '1,11 agreed to be appalled at the RomanEpicurean Catius for his translation of "$%5(7+ as spectra. Spectrum is oth-erwise unattested in Latin before the seventeenth century (when it seemsto mean ‘appearance’ or ‘aspect’). It probably represents Catius’ attemptto invent an o/-the-peg jargon for Latin Epicureanism. I have no ideawhat connotations it conveyed to a Roman ear, but Cicero and Cassiusseem to have found them comic.

Lucretius, at any rate, is considerably more subtle. He conveys"$%5(7/. with a range of words which collectively capture the idea,already present in the Greek, of a painted or sculpted image preservingthe surface features of its subject. His most regular term for this is simu-lacrum, but he also commonly uses imago, with the occasional further vari-ants e)gies and figura. (All four renditions were to enjoy at least somesuccess with later Latin writers on Epicureanism.)12

By an extraordinary stroke of luck, the text of book ,- preserves sideby side Lucretius’ earlier and later versions of the introductory lines inwhich his range of terms is sketched.13 (The significance of this forLucretius’ change of plan during the composition of the poem will be

%. Simulacra &"

18 ,, !(&, (%", ,- !"", #"", -, !*+&. At ,- #"" it is specifically their smallness that he wishes to empha-sise with the diminutive. 9 , !"+–#, "*$–!), ,, +##–"), !*!&–%%. See pp. !"*>! below.

10 E.g. ,, &"&, )+&, ,- ")!, -, !**". 11 Fam. .- !+.!, !".!.12 Simulacrum, Vitruvius -, %.&, Gellius - !+.&; imago, Cicero, e.g. ND , !!), and often; e)gies, Cicero,

ND , !!*; figura, Seneca, NQ , (.!, Quintilian, Inst. . %.!(.13 The matter is lucidly explained by Smith (!"$(), p. %#*, who sets out the text properly, as does

Bailey (!")$) , &+). The OCT of Bailey (!"%%) must be avoided at all costs: there he disastrouslyfollows Marullus in a wholesale re-ordering of the lines. For further discussion, see pp. !&$># below.

fully discussed in Chapter (.) In the earlier version ()(–(&), the existenceof "$%5(7+ is first broached with the words ()"–(&)

nunc agere incipiam tibi, quod vementer ad has resattinet, esse ea quae rerum simulacra vocamus, (*quae quasi membranae vel cortex nominitandast,quod speciem ac formam similem gerit eius imagocuiuscumque cluet de corpore fusa vagari.

I shall now begin to deal with what is closely relevant to this: thatthere are what we call images of things, which are to be termed‘like membranes or bark’, because the image bears a shape andform similar to those of whatever thing’s body we say it has beenshed from and travelled.

He thereby recruits, in addition to the family of artistic metaphors, thebiological vocabulary of ‘membranes’ and ‘bark’ as helping to conveythe ready detachability of these ultra-fine surface-layers of bodies. In theevent, neither of the biological terms is brought into play in this role any-where in book ,-.14 And that must be why, when for other reasons hecame to rewrite the proem, Lucretius edited them out, limiting hisvocabulary for "$%5(7+ exclusively to the iconic imagery.15

In the rewritten passage (%+–))) the existence of "$%5(7+ is broachedin language which starts out identical to the first version, but thendeparts significantly from it (%"–&%):

nunc agere incipiam tibi, quod vementer ad has resattinet, esse ea quae rerum simulacra vocamus,quae quasi membranae summo de corpore rerumdereptae volitant ultroque citroque per auras . . .

I shall now begin to deal with what is closely relevant to this: thatthere are what we call images of things, which, like membranessnatched from the outermost part of things’ bodies, fly hitherand thither through the air . . .

‘Membranes’ here is no longer part of the designated vocabulary for"$%5(7+, but forms instead the basis of an extended simile, designed toconvey one specific aspect, the detachability and volatility of theseatomic films. As for the other biological term ‘bark’, a clumsily inappo-site name for a light and volatile surface layer of atoms, this has now been

)* ". Two languages, two worlds

14 On cortex, see below. Membranae occurs once, at ,- "(, but only in the descriptive phrase tenuissummi membrana coloris, where it is not left to fend for itself.

15 E)giae and figurae are in fact used only twice and three times respectively in the remainder of thebook. Imago (most commonly singular, for metrical reasons) is used some seventeen times.Curiously, it does not occur in the revised version of the proem.

deleted. It does however put in an appearance at the end of the rewrit-ten passage, in the company of the preferred sculptural imagery ()%–&):

dico igitur rerum e0gias tenuisque figurasmittier ab rebus summo de cortice eorum.

I say, therefore, that things’ e0gies and tenuous figures aredespatched from them o/ their outermost bark.

I see no justification for the standard emendation of cortice to corpore.16

Lucretius has in his revised version rightly seen that ‘bark’ mostappropriately conveys the idea of the stable outer part of an object, fromwhich the "$%5(7+ flow.17

It might seem pointless to wonder what motivated Lucretius’ originalabortive attempt to introduce the pair of biological terms. But as ithappens the question can be answered with a surprising degree ofconfidence. Alexander of Aphrodisias, in attacking the Epicureantheory of vision by simulacra, asks why, if simulacra are as volatile as theproponents assert they are, windy conditions are not su0cient to preventour seeing things.18 In describing the images’ volatility, he quotes theactual words of the theory’s proponents: "&, %7/$(5(' . ,+$! &$ µ".(5(' . (+ )%+'$., ‘[consisting]19 of “bark-like and membrane-like” stu/s, as theyput it’. Once we place this Greek phrase alongside Lucretius ,- (!, itbecomes scarcely deniable that he has quite simply translated it.20 His quaequasi membranae vel cortex nominitandast announces that ‘membrane-like’and ‘bark-like’ are appropriate descriptions to use of the simulacra.Although there is a little evidence that the Epicureans did sometimes alsocall the visual images ‘barks’ or ‘membranes’,21 it seems clear that on this

%. Simulacra )!

16 Here I am in agreement with Godwin (!"#+), pp. ")–(. Bailey (!")$), who comments ad loc. thatcortice cannot be right because cortex designates for Lucretius the "$%5(7/., is overlooking the factthat that was only in the now discarded version of the proem to book ,-.

17 Epicurus himself in the context of images uses /$ "& @(" -+-/) ;$-(" . (%).!".!–& Arr.2), possibly asa name for the fixed outer layer – although the text is too fragmentary to rule out the alterna-tive that he means the image itself.

18 De anima mantissa !&(.%)–+, "$& 5"" "&'-$. "&% ,/7/) +&& -(' . 3$ ,$".3'$), "& , %7/$(5(' . ,+$" &$ µ".(5(' .(+ ) %+'$., ,+$" 6+' '+ #$ /63! $$ ,+.3! 6+#+'&' #+$ +&& -+" , "%5"$ µ3! /$ #+' . -/&! ) ,+-+! -/! . +% ."µ/. 27""6/.-+). For discussion of the passage, see Avotins (!"#*).

19 See Avotins (!"#*), p. )&# n. )*, for discussion as to whether a participle such as 6"6/$3µ"" .(.has fallen out here. Given the Lucretian parallel, I am at least confident that the phrase describesthe composition of the simulacra themselves, not (a possibility considered sympathetically byAvotins) some external agent which moves them.

20 The Alexander passage is cited for comparison with the Lucretius line by Lackenbacher (!"!*),p. %%", and Ferrari (!"&$), p. !"&, but I have not yet found anyone who notices the actual relationin which it stands to Lucretius’ own wording.

21 For &$ µ"" .") see Diogenes of Oenoanda !* - & Smith; %7/$"/&) is one available MS reading atPlutarch, Non posse !!*+3.

occasion Lucretius’ quasi is added in order to capture the adjectival forceof the -(" 53) termination: not membranes and barks, but membrane-like and bark-like.

The conclusion from this must be that Lucretius, at least whileworking on his first draft (see Chapter (), was ready in principle simplyto draw his imagery from the technical terminology of Greek Epicureanprose, but that such borrowings survived into subsequent drafts only ifthey could prove their independent worth in the context of Latin poeticimagery. In this particular case, the technical force of the original Greekadjectives was to specify and illuminate one particular property of theimages, their detachability, by exploiting the analogy of tree bark, reptilesloughs and the like.22 Lucretius evidently found that this aspect did not,in the event, repay emphasis in the main body of book ,-, whereexploitation of it would have tended to fall between two stools, not ase/ective as the iconic imagery for capturing the extraordinarily accuraterepresentational powers of the images, and positively detracting from hisattempts to convey their all-important high velocity. In short, while thesculptural imagery survived, the biological imagery failed the test andwas edited out.

This tantalising vestige of Lucretius’ first draft, with its failed attemptto translate %7/$(" 53) and &$ µ".(" 53), suggests that at that early stage hewas more hopeful of finding simple Latin translations of at least someGreek technical terms than he appears to be in anything else that nowsurvives of the poem. Thus the policy which I am describing in thissection should be seen as one evolved only during the writing process.Lucretius’ initial conception of his linguistic task may have beensignificantly closer to the one I have described in Cicero. The policy atwhich he eventually arrived is well illustrated by the revised proem tobook ,-. Rather than follow Catius in supplying a Latin technical ter-minology for "$%5(7+,23 he seeks to embody the notion in a set of livemetaphors which will complement each other in focusing on the cardi-nal feature of "$%5(7+, their power to preserve a portrait-like resem-blance to the object emitting them, even over a considerable distancetravelled. Their detachability and volatility will be conveyed in otherways, by both simile and argument,24 without being allowed to dilute orobfuscate the dominant metaphor of portraiture.

)% ". Two languages, two worlds

22 See ,- ($–+% for further examples.23 In this same context of simulacra, Clay (!"#&), p. %*, aptly points out how Epicurus’ technical term

/$ "&@('-$,/! ) -#/" 6/) (Nat. ,,, %).)&.&, )).! Arr.2) has become in Lucretius (,- !"&–)) the descrip-tive parvola causa . . . procul a tergo quae provehat atque propellat. 24 ,- ()–%!+.

( . 47896 3?< -6796 18?5739 56<

Those familiar with Cicero’s philosophical works may at this stage retortthat there is nothing unique about Lucretius’ search for a mutuallycomplementary set of terms corresponding to a single Greek term. Asimilar-looking process can be glimpsed in Cicero’s own forging of aphilosophical vocabulary, where he often introduces a Greek term witha whole bevy of Latin equivalents. The Stoic term for infallible cogni-tion, ,+-+" 73:$), literally ‘grasping’, provides a good illustration. Its usein rhetorical theory may have earned it Latinisation at an earlier date,since already in his youthful De inventione Cicero uses perceptio in a wayprobably intended to correspond to ,+-+" 73:$).25 Yet still in the secondbook of the Academica his spokesman Lucullus can be found tinkeringwith the rendition of it, and listing a range of alternatives: . . . ‘cognitio’aut ‘perceptio’ aut (si verbum e verbo volumus) ‘comprehensio’, quam ,+-+" 73:$.illi vocant . . .’ (Ac. ,, !$, cf. !#, &!).

Normally in Cicero this little fanfare would herald the first introduc-tion of a term. But we are already here in the second book of theAcademica, and it is certain that ,+-+" 73:$) had already featured in book,.26 What Cicero in fact turns out on closer inspection to be doing hereis not creating but enlarging his stock of Latin terms for it, adding compre-hensio to the terms perceptio and cognitio which he had been using up to now(in the Academica, that is, and also in the De finibus, composedcontemporaneously with it). And one can see why. Both perceptio and cog-nitio were too widely and loosely used within the ordinary Latin cogni-tive vocabulary to capture the very special flavour of Stoic ,+-+" 73:$),whereas comprehensio and its cognates were barely yet familiar in a cogni-tive sense, so that the usage could still retain a suitably technical ring.27

Curiously enough Cicero too, just like Lucretius in book ,-, can herebe watched in the act of refining his vocabulary. Our version of book ,,comes from the Academica priora, Cicero’s first edition. In his revisededition, the Academica posteriora, from which part of book , survives, com-prehensio is heralded as the single correct translation right from the start(Ac. , )!):

‘ . . . When that impression was discerned in its own right, Zeno called it com-prehendibile. Will you accept this?’

‘Yes,’ he replied. ‘How else could you express ,+-+736-/" .?’

$. Prose and verse contrasted )&

25 Inv. , ", &+. 26 Ac. ,, %# indicates that Hortensius had already used it in book ,.27 For the various cognate forms of comprehensio in Cicero, see Lévy (!""%).

‘But when it had already been received and endorsed, he called it comprehen-sio, like things grasped with the hand . . .’

This exclusive use of comprehensio for ,+-+" 73:$) seems thereafter tobecome canonical in what survives of the revised book ,, and wasundoubtedly continued in the lost books ,,–,- of the revised version.28 Itenables Cicero to let it stand in contrast, as a term of art, with the lesstechnical ‘knowledge’ vocabulary – scire, cognoscere and percipere – which inthe ensuing chapters he puts into the mouths of pre-Stoic philosophers.29

Consequently, it would be quite misleading to assimilate the practicesof Lucretius and Cicero when each sets about establishing a group ofalternative or complementary Latin terms for a single Greek original.Cicero does it only as a step towards what will, if all goes well,30 proveto be their eventual whittling down to a single technical term. ForLucretius, on the other hand, the range of alternative terms is nostopgap or compromise, but is intrinsically desirable. By means of it, heseeks to capture the Greek original, not by substituting a Latin technicalterm for a Greek one, but by keeping in play a whole set of mutuallycomplementary live metaphors.31 The policy is one not of finding a tech-nical terminology, but of avoiding one.

In pursuing such a policy, Lucretius is doing no more than observingthe traditional practice of his genre, the hexameter poem on physics.32

The proper comparison to make is not with Cicero, but withEmpedocles, whom, as I sought to show in Chapter !, Lucretius reveresas the founder of his genre. Empedocles has no technical vocabulary forthe six primary entities in his physics – the four elements plus the twopowers Love and Strife – but deploys for each a varied set of evocativemetaphors and allegorical names: thus the element water is representednot only by the word ‘water’ (&+ 5(#), but also by ‘rain’ (/% µ2#/)), ‘sea’

)) ". Two languages, two worlds

28 It clearly suits his purposes well in a technical epistemological context like the Academica.Elsewhere he may have continued to feel uncomfortable with it, as at Fin. ,,, !$.

29 See e.g. , )).30 Hence Fin. ,,, !(, where Cato remarks equidem soleo etiam, quod uno Graeci, si aliter non possum, idem

pluribus verbis exponere. On this passage, cf. Powell (!""(a), pp. %"%–).31 Here and elsewhere I stress ‘live’ metaphors because it is not my aim to contrast the meta-

phorical, as such, with the literal. (I agree with Lloyd (!"#$), chapter ), that this is usually a mis-leading or at least unenlightening antithesis, especially as most philosophical technical terms arethemselves metaphors.) Rather, I am singling out those metaphors which actively work on thereader by exploiting the power of imagery.

32 It is well recognised that the variety of vocabulary in philosophical verse is partly dictated bymetrical constraints: a word which works well in one place may be inadmissible in anothergrammatical case or at another point in the hexameter. However, metre alone would not haveprevented Lucretius from managing exclusively with, e.g., e)gies and atomi throughout, had hepreferred to do so.

(1+" 7+''+, 6/" .-/)), and ‘Nestis’, probably a Sicilian cult name forPersephone, who was especially associated with springs.33

At ,, +))–+*, commenting on the cult of the ‘earth mother’, Lucretiuslicenses the potentially misleading vocabulary, so long as the associatedfalse beliefs are absent. Likewise Empedocles himself, as we noted inChapter !,34 had declared his readiness to continue using a philosoph-ically misleading mode of speech, once his readers had been alerted tothe implicit error. Ordinary talk of ‘birth’, ‘death’ and the like, he hadobserved, misdescribes what in reality is simply mixture and separation;but, he had added, he would himself go along with the convention.35 Hiscontemporary Anaxagoras had likewise noted that the conventional lan-guage of birth and death is misleading,36 but he, being a prose writer,had faced no di0culty in simply acting on this insight and editing suchlanguage out of his treatise, replacing it with that of mixture and separa-tion. This contrast neatly illustrates the di/ering strategies which typ-ically separate scientific exposition in verse from that in prose. Lucretius’avoidance of terminological technicality, in the tradition of Empedocles,is another symptom of that di/erence.

However, a little new light can now be thrown on the findings ofChapter !. There we witnessed the emergence of Lucretius’ poem fromhis synthesis of two very di/erent intellectual backgrounds, withEpicurus supplying the message, Empedocles the medium. But we arenow in a position to see that the exact division of labour between the twohad to be carefully negotiated. What we have recognised (above pp.)!>%) as Lucretius’ attempt, in the early stages of his work, to LatiniseEpicurus’ technical terms %7/$(" 53) and &$ µ".(" 53) suggests that themore Empedoclean approach to philosophical vocabulary was notstraightforwardly Lucretius’ first choice. His move away fromEpicurean-style technical terminology towards Empedoclean informal-ity and variety was gradual, hard won in the light of experience.

Lucretius himself likes to remind us that his poetic task is an excitinglynew one. The point is emphasised, not just in the lines from the proemto book , with which this chapter opened, but also in his celebratedpoetic manifesto at , "%!–(*, where he is a bold explorer in the untrod-den terrain of the Muses. It may be significant that he chose to repeatthis manifesto as the opening of the book ,- proem,37 the very place

$. Prose and verse contrasted )(

33 See Kingsley (!""(), ch. %%. 34 See p. &% above. 35 Empedocles '#–".36 Anaxagoras '!$.37 ,- !>%(. See Gale (!"")b), who argues e/ectively that these lines were intended by Lucretius to

open book ,-, and are not the product of editorial meddling by others.

where we have also witnessed his shift towards a more robustlyEmpedoclean treatment of technical concepts. The findings of thischapter may help us to appreciate Lucretius’ reasons for his passionatesense of being a pioneer. It is not simply that no one had previouslywritten philosophical verse in Latin38 (although that is undoubtedly thepoint made in the proem to book ,). It is also that the versification ofEpicureanism, a philosophy with a highly developed technical vocabu-lary, makes new demands on the poet quite unlike any faced even by thefounder of the genre, Empedocles.

+ . <,9 58756< 76398?,?B

I do not mean to deny that any word in Lucretius ever has a technicalsense assigned to it, although interestingly enough the most prominentcases are ones where the Greek original lacked such a term. (I am think-ing here of coniunctum for ‘permanent property’ at , ))"/.,39 and theanimus/anima distinction set out in book ,,,.) But what we have alreadyseen, the conversion of Greek technicality into flexible Latin metaphors,is a far more pervasive feature of his poetry.

One very satisfying case, which was first detected by MylesBurnyeat,40 is Lucretius’ rendition at ,- )$% of the exclusivelyEpicurean technical term for a thesis which ‘refutes itself ’, 6"#$,+" -(-#""6"-+$. Scepticism, the claim to know nothing, is dismissed as self-refuting, but Lucretius conveys the dry technicality of 6"#$,+" -(-#""6"-+$ with a picture of the sceptic as an acrobat or contortionist (,-)+"–$%):

denique nil sciri siquis putat, id quoque nescitan sciri possit, quoniam nil scire fatetur.hunc igitur contra mittam contendere causam,qui capite ipse suo in statuit vestigia sese.

If someone thinks that nothing is known, he doesn’t even knowwhether that can be known, since he admits that he knowsnothing. I therefore decline to argue my case against this personwho has stood with his own head pressed into his footprints.

)+ ". Two languages, two worlds

38 I discount Sallustius (see Ch. !, §!), whose poem, whatever its content may have been, was pre-sumably not yet published when Lucretius was writing.

39 That coniunctum does not, as commonly supposed, translate the single Greek word '&µ2"23,/" )is argued in Long/Sedley (!"#$) §$, and more fully in Sedley (!"##). As for eventum in the samepassage, it is introduced as already a familiar Latin usage ()(#).

40 Burnyeat (!"$#). On the philosophical significance of this passage, see Ch. &, §#.

The sceptic’s confusion is reinforced in the last line with the Lucretiandevice which David West41 has christened ‘syntactical onomatopoeia’:intellectual contortion is symbolised by contorted grammar, with theproper order statuit in reversed in defiance of basic syntax. I am inclinedto resist the normalising emendation of suo to sua, adopted by mosteditors since Lachmann. That substitutes a regular Lucretian type ofgrammatical inversion – sua in for in sua42 – for a highly irregular one.But irregularity may be exactly what he intends. Anyone who objectsthat the irregularity is too harsh for Lucretius to have perpetrated shouldconsider my next example.

At ,- #%%–($, Lucretius rejects another topsy-turvy piece of thinking –the teleologist’s mistake of supposing that, because a human bodily partserves a function, that function must have been conceived prior to thepart’s coming to exist. In Lucretius’ view, a thing must already existbefore any thoughts about its function can even be entertained.Teleology is back-to-front reasoning; or, as he puts it (,- #&%–&):

cetera de genere hoc inter quaecumque pretanturomnia perversa praepostera sunt ratione.

All such explanations which they o/er are back to front, due todistorted reasoning.

What was in his Greek original? My guess is that what he found therewas a description of teleological reasoning as 5$+" '-#/A/), ‘distorted’.This term, which translates literally into Lucretius’ word perversa, is onewhich, according to Sextus Empiricus,43 Epicurus used for opinionwhich imposes a distorted construal on primary empirical data. But onceagain Lucretius has backed up the accusation with syntactical ono-matopoeia. The distortion is attributed to ‘back-to-front’ (praepostera)thinking, which in turn is conveyed by the reversal of linguistic elementscontained in the tmesis inter quaecumque pretantur. Tmesis is a commonLucretian device, and one of his favourite forms of syntactical ono-matopoeia, usually given prominence by its positioning at a line end.Most commonly the line will contain an unseparated compound ofprefix plus verb stem, followed by a second compound verb with that

&. Distorted reasoning )$

41 West (!"$(), p. "+.42 For a list of postponed prepositions in Lucretius, see Bailey (!")$), , !*$.43 SE M -,, %*"; for the authenticity of the Epicurean terminology in M -,, %*+–!*, see Sedley

(!""%), pp. ))–((, and cf. Gigante (!"#!), pp. !!#–)#. For the philosophical significance of5$+'-#/%3" and its cognates, see Grilli (!"+&).

same prefix, but this time separated in tmesis. According to the context,this movement from the conjoined to the separated can equallye/ectively convey the overlapping ideas of mixture, fusion, penetration,containment, alternation and entanglement (e.g. ,, !(), sed complexa meantinter se conque globata),44 or, on the contrary, that of separation (e.g. , +(!,disiectis disque supatis).45 But our present line, ,- #&%, o/ers one of only twoLucretian tmeses in which the bare verb stem, left exposed by separationfrom its prefix, is not a Latin word at all.46 The teleological reversalcannot be contemplated, Lucretius’ message runs: it produces nonsense.

$. 526 47,16 8@ @3,:A76

In all these Lucretian strategies for the conversion of Greek technicalityinto Latin imagery, one invariable rule is observed: never transliteratethe Greek term. There are, in fact, only two significant breaches of thatrule,47 and they both speak eloquently in its favour. A leading contenderfor the title of Lucretius’ worst line (at least when taken at face value) is, #&*. Lucretius’ apology for it runs as follows (, #&*–%):

nunc et Anaxagorae scrutemur homoeomerianquam Grai memorant nec nostra dicere linguaconcedit nobis patrii sermonis egestas.

Now let us also take a look at Anaxagoras’ homoiomereia – as theGreeks call it, but which the poverty of our native language pre-vents us from saying in our own tongue.

The ungainliness conveys a point about the unacceptable consequencesof resorting to mere transliteration of the Greek. Anaxagoras’ horribleword is glaringly not at home in the Latin language; and that in turnforeshadows the fact, which Lucretius satirically develops in the sequel,that the concept underlying it is equally unwelcome.48

)# ". Two languages, two worlds

44 ,, !(), &"), ,,, %+%, &)&, )#), ,- %*&, &&%, (+%, ##$ (inque gredi helps make the mind sound like apassenger in the body), !!)", !%(*, - !%+#, -, &"), )(+, ($*, !%+) (higgledy-piggledy).

45 , &!#, )(%, +(!, ,, !!*) (separation from culpability?), ,,, %&&, #+*, ,- %)$, &##, ")#, - %#$, %"",!&$), -, &&%. The only tmeses which are (to my eye) not obviously syntactically onomatopoeic are,- $!& and - ##&. One special case is ,,, #$# (tmesis with reversal, to convey the idea of some-thing left over).

46 For the other, see the brilliant article of Hinds (!"#$), who observes that seiungi seque gregari at ,)(% is thus used to convey the idea of a fixed property (coniunctum), like the heat of fire, whoseloss its possessor cannot survive.

47 I do not count prester (-, )%)), which although in a way technical is not a philosophically contro-versial term.

48 Cf. Wardy (!"##), an article which illuminatingly explores the ideological basis for Lucretius’ tar-geting of Anaxagoras; esp. p. !%+, ‘Accordingly, the suggestion of !.#&* is that Anaxagorean phi-losophy is in a number of senses untranslatable’, an idea which Wardy develops in terms of theanalogy between elements and alphabetic letters.

This link between the alienness of a word and the alienness of theconcept it expresses is virtually explicit in the other passage where baretransliteration is resorted to. Early in book ,,, the old Greek theory thatsoul is a harmony or attunement of the bodily elements is dismissed (,,,"#–!&(). In Lucretius’ discussion of it the Greek word +$ #µ/.$"+ is simplytransliterated, not translated. This is not in itself surprising, since+$ #µ/.$"+ is as resistant to rendition into Latin as it is into English. EvenCicero, in his paraphrase of Plato’s Timaeus (%$), while attempting thetranslation concentio for +$ #µ/.$"+, is su0ciently uneasy about it to take thestep, uncharacteristic in this work, of supplying the Greek word too.Elsewhere Cicero’s own preference with regard to +$ #µ/.$"+ is for simplytransliterating it.49 But more is at stake for Lucretius: the word’s undis-guised alienness to the Latin language is symptomatic of the concept’sirrelevance (,,, !&*–():

quapropter quoniam est animi natura reperta !&*atque animae quasi pars hominis, redde harmoniainomen, ad organicos alto delatum Heliconi,sive aliunde ipsi porro traxere et in illamtranstulerunt, proprio quae tum res nomine egebat.quidquid id est, habeant. !&(

So, since the nature of mind and spirit has been found to be likea part of the human being, give back the name of harmonia,whether it was brought down to the musicians from highHelicon, or whether they themselves drew it from some othersource and transferred it to what previously lacked a name of itsown. Whatever it is, let them keep it.

An alien concept deserves an alien name. By the same token, Lucretius’habitual practice has made clear, philosophically welcome conceptsmust make themselves at home in the language too.

# . '3,:6D ’9 18;4:3,?5

Now I come to the great Lucretian anomaly. Although Lucretius studi-ously avoids using transliterated Greek terminology, his whole poem isnevertheless knee-deep in Greek loan words.

These Greek words have been usefully catalogued by Bailey,50 whoconcludes that

(. Bailey’s complaint )"

49 E.g. Rep. ,, +", Tusc. , )!. Even Lucretius himself once outside book ,,, uses the transliterated har-moniae: at ,- !%)#, where he may feel that his need for the musical metaphor leaves him no option.

50 Bailey (!")$) , !&#–".

(a) in some cases Lucretius’ hand has been forced by the unavailabilityof a suitable Latin word;

(b) in many other cases, where a perfectly good Latin word is at his dis-posal, he is using Greek out of virtual ‘caprice’;

(c) in one extreme case, ,- !!+*–", where sixteen Greek words occur inthe space of ten lines, it is impossible to resist the conclusion thatLucretius is translating a Greek original.

It is hard to think of a more implausible set of explanations. Withregard to (a), what we have already seen of Lucretius’ handling ofphilosophical terminology should put us on our guard against everassuming too readily that he has been forced to resort to Greek by thelack of a Latin word.51 As for (c), Bailey’s explanation implies a very pooropinion of Lucretius’ skills as a translator, and one totally negated by apassage like ,,, !#–%%, where we know that Lucretius is following a Greekoriginal, the Homeric description of Olympus.52 But in the remainderof this chapter I want to concentrate on (b), the kind of cases whereBailey thought the intrusion of Greek merely gratuitous. It seems to methat there are remarkably few genuine cases that fit this description (seethe appendix to this chapter). I shall start by quoting Bailey’s cataloguein full,53 since I aim to deal with every word on his list:

Many Greek words Lucretius seems to have used almost from caprice, some-times when there was a good Latin word which he could have employed instead:anademata iv. !!%", baratre iii. "((, barathrum vi. +*+, chorea ii. +&(, durateus (!ligneus)i. )$+, gigas iv. !&+, v. !!$, lampas ii. %( etc. (( times), lychni v. %"(, mitrae iv. !!%",scaphiis vi. !*)+, sceptra iii. !*&+, v. !!&$, scymni (!catuli already used in the sameline) v. !*&+, thalassina iv. !!%$, thyrso i. "%&.

" . 6-8E,?B B76616

Most of the Greek words attributed by Bailey to Lucretian ‘caprice’ donot occur in isolation. They tend to turn up in droves. And again andagain this concentration of Greek words in a passage is exploited for aspecific e/ect – to conjure up for Lucretius’ readers a Greek or an oth-erwise exotic context. When Greece joined the European CommonMarket, The Times celebrated with a competition for the reader’s poemwith the largest number of Greek-derived words. This is pretty much

(* ". Two languages, two worlds

51 However, see n. )" (p. )") on harmoniae at ,- !%)#, and n. )$ (p. )#) on prester at -, )%).52 Od. -, )%–+. See the notes on Lucretius ad loc. in Kenney (!"$!), and notice especially innubilus (%!)

for the Homeric +& ."" %"7/), an apparent Lucretian coinage designed to capture the special flavourof the original, but decidedly not a transliteration. (Cf. Seele (!""(), p. %$, who contrasts Cicero’ssine nubibus when translating the same word from Aratus.) 53 Bailey (!")$) , !&".

what Lucretius is up to too: when he uses a whole convoy of Greekwords, he is usually quite simply trying to make us think of Greece.

‘Greek words’ here should be interpreted broadly. It naturallyincludes Greek proper names as well as common nouns and adjectives.Moreover, it can be extended to include Greek linguistic idioms, such asthe formation of compound adjectives, not native to the Latin lan-guage.54 These points are well illustrated by , )+)–#%, the wonderfuldescription of the Trojan war. In the space of five lines, )$&–$, we havenot only six Greek names, but also the quasi-Greek compound adjectiveGraiugenarum:

numquam Tyndaridis forma conflatus amoreignis, Alexandri Phrygio sub pectore gliscens,clara accendisset saevi certamina belli, )$(nec clam durateus Troianis Pergama partuinflammasset equos nocturno Graiugenarum.

never would that flame kindled deep in Alexander’s Phrygianheart and fanned with love by the beauty of Tyndareus’ daugh-ter have ignited the shining battles of savage war, nor would thatwooden horse by giving birth to its Grecian o/spring at dead ofnight have stealthily set fire to the Trojans’ Pergama.

Especially telling is the authentic Homeric adjective durateus used of the‘Wooden’ Horse (where, as Bailey ruefully points out, there was the per-fectly good Latin word ligneus available). And in this already Greekcontext it is legitimate to regard the archaic Latin nominative equos, withits Greek-like termination, as yet another linguistic detail contributing tothe same cumulative e/ect. (It should therefore not, with the majority ofeditors, be normalised to equus.)55

The argumentative context of this description is Lucretius’ discussionof the metaphysical problem how facts about the past maintain theirpresent existence: what is there in existence now for them to be proper-ties of ? It therefore serves his purpose to present his example, the Trojanwar, as a remote one. The epic ring of the Greek helps locate it in acontext far removed from present-day Rome. And this brings me to ageneral observation: that the creation of a Greek context tends, inLucretius’ hands, to emphasise the remote and the exotic.

*. Evoking Greece (!

54 On compound adjectives, cf. pp. %)>( above.55 As J. N. Adams points out to me, the retention of the old -os termination is not particularly

unusual in a noun whose stem ends in -u (to avoid the collocation uu). Nevertheless, it may bejudged to acquire a Hellenising significance when contained, as here, within a broaderHellenising context.

!* . ;3B?65 9 ,? 93;8527316

Bailey’s list of gratuitous Greek imports includes scaphiis, ‘basins’, at -,!*)+. The word was a common enough one in Latin by Lucretius’ dayto pass unnoticed. Nevertheless, since it occurs here in a Greek context,flanked by Greek proper names, it does deserve consideration. It comesin the course of a long and involved discussion of the magnet, and at thispoint Lucretius is describing the phenomenon of magnetic repulsion (-,!*))–+):

exultare etiam Samothracia ferrea vidiet ramenta simul ferri furere intus ahenisin scaphiis, lapis hic Magnes cum subditus esset.I have even seen Samothracian iron objects dance, and ironfilings all simultaneously go crazy in bronze basins (scaphiis),when this Magnesian stone was placed underneath.

What are these Samothracian ferrea? Iron rings, the editors usually say. ButI doubt it. There were rings called ‘Samothracian’, but they seem to havebeen a combination of iron and gold: on one report, gold rings with aniron ‘head’; on another account, iron rings plated or decorated with gold.56

It seems unlikely that either of these is meant. The neuter ferrea cannoteasily imply the masculine complement anuli or anelli.57 Besides, someonedisplaying the powers of a magnet would not be likely to use objects con-taining gold as well as iron, since the weight of the gold would reduce theirresponsiveness to the magnetism. Finally, both types of ring clearly had apredominantly gold exterior, and would not very naturally be known as‘iron’ rings. (It would only be if you wanted to cause o/ence that you wouldbe likely to refer to someone’s gold-plated ring as their ‘iron ring’.)

Ferrea must mean just what it appears to mean, namely ‘iron objects’.But why, then, are they called ‘Samothracian’? There is only one plau-sible answer: Lucretius is describing a display he once witnessed inSamothrace. Line !*)) means ‘I have even seen the ironware ofSamothrace dance’. The natural magnet or lodestone, variously calledthe Magnesian stone and the Heracleian stone,58 was as the namessuggest predominantly associated with Magnesia or Heracleia (therewas some disagreement whether this was the Heracleia in Lydia or theone in Pontus). But according to one variant tradition the magnet was

(% ". Two languages, two worlds

56 For the evidence, see Lewis (!"("), 5&*, 5%!&.57 For similar doubts, see Godwin (!""!), ad loc.58 For an engrossing discussion of ancient views on the lodestone, see Wallace (!""+). The ancient

testimonia are collected in Radl (!"##).

first found in Samothrace, and was named after the city of Heracleia onthat island.59 If there had actually been a Heracleia in Samothrace, thisassociation of the Heracleian stone with the island might have been dis-missed as a simple error of geography, the confusion of one Heracleiawith another. But Samothracian Heracleia seems to be a fiction. Hencea better explanation for the origin of this variant tradition may be thatlodestones were indeed found on Samothrace, and that this led to a mis-conception regarding the location of the Heracleia in question.

The Lucretian passage, if I have interpreted it correctly, now stands instrong confirmation of that hypothesis. This use of a first-person eyewitness account is a rarity in Lucretius,60 and confirms that, exceptionally,he is recounting to us an exhibition of the powers of the magnet whichhe had seen when himself on the island of Samothrace – whether from avendor, or in a religious ritual, or in other circumstances is impossible toguess. The seething iron filings cannot in fact have been the product ofmagnetic repulsion, as Lucretius thinks they were.61 Their movementmust have been produced by magnetic attraction – moving a lodestonearound beneath the bronze bowl to make them alternately stand on endand lie flat again. But Lucretius’ mistake makes slightly better sense if hehad also witnessed genuine magnetic repulsion, of which the ‘dancing’iron objects may therefore represent an authentic case.62

But how likely is it that Lucretius had been to Samothrace? A pictureof Lucretius the seasoned Aegean tourist does not carry conviction, andshould become still less plausible when we proceed to explore his waryattitude to things Greek. Nor is he, of all people, very likely to have goneon a religious pilgrimage to the celebrated Kabiric mysteries heldthere.63 However, there is no obligation to see this visit as motivated by

!+. Magnets in Samothrace (&

59 Etymologicum magnum, s.v. B+4.3' -$)!Lewis (!"("), 5%*.60 The only other cases, I believe, are ,- ($$, recalling his own experience of multiple echoes, and

(if this counts) ,- $+", recalling his dreams. Given how sparing he is with them, I would take theseautopsy claims seriously. When he has not witnessed something in person, Lucretius is ready toadmit it: cf. his indication at , $%$ (quoted p. !% above) that he has never been to Sicily, and sim-ilarly - ++&, -, $(+, #)#–". 61 See Wallace (!""+), p. !#).

62 The trouble is that artificial magnets have made knowledge of natural lodestones a rarity today.The fullest and most authoritative account of the powers of the lodestone is Gilbert (!+**). Heseems to acknowledge repulsion only between two lodestones, not between a lodestone and pre-viously magnetised iron. My tentative guess is that what Lucretius actually saw ‘dance’ werepieces of lodestone shaped into rods with their poles at the tips. Lodestone being iron ore, thesecould easily have been mistaken by Lucretius for iron rods. It remains possible, alternatively, thatthis trick too was done with magnetic attraction (thus Wallace (!""+), p. !#)).

63 There is reason to think that some Romans did go to Samothrace for the mysteries, possiblyincluding one with Epicurean links. See Bloch (!")*), esp. n. !#. For Samothrace in relation tothe Roman tourist circuit, see Casson (!"$)), pp. %)&, %+!.

either tourism or religious zeal. Samothrace, lying just o/ the coast ofThrace, was a natural point of anchorage for anyone on a sea voyagebetween Europe and Asia. Acts of the Apostles !+.!! describes how StPaul put in there for the night when sailing from the Troad toMacedonia.64 Ovid changed ships there on his way to exile in Tomi.65

Any Roman sent on a tour of duty to an Asian province might well stopo/ there on the outward or homeward journey. One plausible suchjourney would be – but here I am entering the realms of fantasy – a tourof duty to Bithynia, where Lucretius’ patron Memmius was propraetorin ($ '1.

At all events, the use of the Greek word scaphiis at -, !*)+ can nowhardly be called gratuitous. It is part of the window-dressing forLucretius’ brief excursion into an exotic world – his report of tricks withmagnets in Greek bronze vessels, witnessed in person on this distantAegean island.

!! . 526 @3;,:,37 3?< 526 6.85,1

Nor are the remote and the exotic by any means always viewed withsympathy or approval. In book ,-, for example (!!%&–&*, !!+*–"), Greekvocabulary piles up to describe the absurd luxuries and euphemistic epi-thets which deluded lovers, blinded to the realities of life, bestow on theobjects of their a/ections. (These lines, incidentally, feature prominentlyin Bailey’s list of gratuitously introduced Greek words, and include theones which he thought must be translated from a Greek original.) Andbook ,, has another build-up of Greek words and names in the freneticdescription of the worship of Cybele (+**–)&),66 a cult whose theolog-ical implications we are immediately urged to shun. Just as they are cul-turally remote,67 so too they are, as Lucretius puts it (,, +)(), ‘far removedfrom true reasoning’.68

This shunning of the exotic can be felt in the important ethical proem

() ". Two languages, two worlds

64 Although Samothrace was said to be ill-provided with harbours (Pliny, NH ,-. !%.$&), it certainlyhad at least one (Livy .:- +.&).

65 Ovid, Tristia , !*.!"–%%. My thanks to Ted Kenney for this piece of information.66 As well as the Greek proper names in the passage, note tympana and cymbala (+!#), chorea (+&(),

and the compound adjectives at +*!, +!", +%$ and +&%.67 This is emphasised by Lucretius’ specific indications (+**, +%") that he is giving us a Greek por-

trayal of Cybele.68 Elsewhere, much the same sense of cultural remoteness is conveyed by the Greek triplet harmo-

niai, organicos and Heliconi at ,,, !&!–%, quoted and discussed above, p. )".

to book ,,, at lines %*–+!. The simple idyllic Epicurean lifestyle is eulo-gised in pure pastoral Latin. Greek words and formations creep in onlywhen Lucretius is describing the pointless luxuries with which it standsin contrast (lampadas igniferas and citharae in %)–#, where %)–( themselvesrecall the well-known Homeric description of Phaeacian opulence atOdyssey -,, !**–%).

One less hostile use of Greek is book ,,,’s quasi-heroic parade of thegreat men who, for all their greatness, proved mortal (,,, !*%)–))) –Scipiadas (note the Greek termination),69 the companions of theHeliconiades, i.e. the poets, including Homer, who out of all of them wasthe one who won the sceptra (!*&#), Democritus, and even Epicurus –whose actual Greek name appears nowhere but here in the entire poem.What is evoked this time is not alienness or remoteness, but the larger-than-life heroism of Homeric (as well as Ennian) epic, in a parade of thedead also reminiscent of the Homeric Nekuia.70

Homer’s own canonisation in this list does reflect a recognition onLucretius’ part of Greek superiority in both music and poetry. Thisemerges from the key Greek terms and forms which highlight his owncelebrated poetic manifesto at , "%!–(*: his poetic ambitions have struckhis heart with a thyrsus (, "%&), inspiring him to expound his philosophy‘with sweet-talking Pierian song’ (suaviloquenti | carmine Pierio, , ")(–+).We should perhaps also detect an implicit contrast of Roman and Greeknoises at ,, )!*–!&:

ne tu forte putes serrae stridentis acerbumhorrorem constare elementis levibus aequeac musaea mele, per chordas organici quaemobilibus digitis expergefacta figurant.

So you must not think that the harsh grating of a shrieking sawconsists of elements as smooth as those constituting the musicalmelodies which the instrumentalists with nimble fingers arouseand form on their strings.

!!. The familiar and the exotic ((

69 It is true that, for metrical reasons, Scipiadas becomes the standard verse name for Scipio(although Africanus was available as an alternative, and is used as such by Martial). Likewise, forthat matter, some of the Greek loan words which I am considering were fairly common in Latin.This, however, does not alter the power of such words to help evoke Greekness when they occur inthe company of other Graecisms.

70 !*%( is Ennian (see fr. !&$ Skutsch), followed immediately by the Iliadic line !*%+ (cf. Il. .., !*$).The thematic link with the Nekuia is already set up by the preceding lines, "#*–!*!%, on myths oftorture in Hades. For the dense series of further echoes of Greek literature in this passage, seeSegal (!""*), esp. pp. !$$–#.

The almost pure Greek third line contrasts with the pure Latin whichprecedes. Where Greece has given us sublime music, Rome’s morecharacteristic noise is the shrieking sawblades of a workshop.71

Sudden switches of vocabulary have this power to transport usinstantly to and fro between the Greek and the Roman worlds. They canbe used not only to praise Greek superiority, and to marginalise whatLucretius shuns as alien, but also, on the contrary, to universalise aconcept. In book - (!*%#–"*), Lucretius argues for the natural origin oflanguage partly by appeal to the way that all animals alike from infancyinstinctively know their innate powers (- !*&)–#):

cornua nata prius vitulo quam frontibus extent,illis iratus petit atque infestus inurget; !*&(at catuli pantherarum scymnique leonumunguibus ac pedibus iam tum morsuque repugnantvix etiam cum sunt dentes unguesque creati.

The calf angrily butts and charges with his incipient hornsbefore they have even protruded from his forehead. Pantherwhelps and lion cubs already fight with claws, paws and bitingat an age when their teeth and claws have barely appeared.

Scymni (!*&+), the Greek vox propria for lion cubs, occurs in Latin litera-ture only here. Bailey objected to it on the ground that there was a per-fectly good Latin word for cubs available, catuli, and one which Lucretiuscould hardly have overlooked since he used it in the very same line! Butthis once again misses the point. The butting calf, a familiar sight in theItalian countryside, is described in pure Latin. The young panthers andlions, on the other hand, those exotic inhabitants of the easternMediterranean and beyond, belong to another world. The switch to thatother world is made instantaneously with the consecutive Greek-derivedwords pantherarum scymnique in line !*&+. Lucretius neatly gets across thepoint that this instinctive use of innate powers is the same the wholeworld over, even though the nature of the powers themselves may varyfrom region to region. Likewise, he is arguing, human beings the worldover naturally express themselves in language, even though the actualsounds produced di/er according to region.

It is worth looking out for a comparable universality in the account ofdisease with which the whole poem closes (-, !*"*–!%#+). Initially,Lucretius emphasises how widely diseases di/er from one region of the

(+ ". Two languages, two worlds

71 Cf. ,, (**–+, Lucretius’ catalogue of qualitative extremes, where Graecisms indicate the exoticcharacter of the finest dyes ((**–!) and of the most sublime music ((*().

world to another (-, !!*&–!#). The diversity is brought home mainly bythe deployment of geographical names, although the exotic character ofthe Egyptian elephantiasis disease is further emphasised by its Greekname, elephas (!!!)). When Lucretius turns to his long closing descriptionof the Athenian plague, however, there is no attempt to bring out itsexotic character by the use of Greek, despite the ready availability ofsuitable vocabulary in the Thucydides text which he is following.72 Thethorny problem of Lucretius’ purpose in closing with the plague passagemust be deferred until Chapter (. Here I shall simply observe that thelinguistic pattern I have described confirms that its lessons, whateverthey are, are meant to be universal ones.

!% . 526 9=3::8= 3?< 526 9=3?

I hope that these examples have succeeded in demonstrating the wide-ranging evocative powers of strategically placed Greek names, idiomsand loan words in Lucretius’ poem. If I am right, something unexpectedhas emerged. Despite the proclaimed Greek origins of both his poeticmedium and its message, Lucretius is very far from being a philhelleneor Helleniser. Although the Greeks are acknowledged to outshine theRomans both artistically and philosophically, Greekness for him fre-quently symbolises the culturally remote, the morally dangerous, and thephilosophically obscure. Seen in this light, the wholesale Latinisation ofGreek philosophical terminology which I discussed in §§!–$ of thischapter will need careful interpretation. We can now see that Lucretius’concern is not the philosophical spoon-feeding of disadvantagedRoman readers linguistically incapable of savouring the Epicureangospels in their original Greek. On the contrary, his readers’ familiaritywith the Greek language, as with Greek literature, is assumed from theoutset, and is systematically exploited. Nor on the other hand is he trans-porting his Roman readers to Athens. He is importing to Rome fromAthens its single most precious product, which, as the proem to book -,eloquently declares, is Epicurus’ philosophy.

It is certainly no part of his strategy to play down Epicurus’Greekness. Right from the proem to book ,, Epicurus has been labelledthe great Greek discoverer (primum Graius homo . . ., , ++).73 And in the

!". The swallow and the swan ($

72 For Lucretius’ use of medical vocabulary, see Langslow (!""#).73 As Farrell (!""!), pp. &)–(, n. !$, points out, Graius homo echoes Ennius’ application of the same

phrase to Pyrrhus, thus implicitly bracketing Epicurus and Pyrrhus as formidable Greek invad-ers of Italy.

proem to book ,,, not only is Epicurus hailed as the ‘glory of the Greekrace’ (o Graiae gentis decus, ,,, &), but his Greekness is brought out with thesame linguistic device that I have been documenting. Lucretius professeshimself Epicurus’ imitator, not his rival (,,, +–#):

quid enim contendat hirundocycnis, aut quidnam tremulis facere artubus haediconsimile in cursu possint et fortis equi vis.

For how should a swallow compete with swans, or what wouldkids, with their trembling limbs, be able to do in a race tocompare with the powerful strength of a horse?

The familiar pattern emerges once again. Lucretius is the swallow, or thekid, described in his own langage, Latin. Epicurus is the swan, or thehorse. The swan is so named in Greek: the Greek cycnus became commonenough in Latin, but this may well be its earliest occurrence;74 and at allevents, the native Latin word olor was available to Lucretius as an alter-native. Even the dative form of cycnis imports a further Graecism, theindigenous Latin construction after verbs of contending being cum plusablative.75 What is more, fortis equi vis, although Latin, honours the horsewith the Greek idiom, familiar from epic, whereby a hero is periphrasti-cally called not ‘x’ but ‘the (mighty) strength of x’, e.g. Iliad ..,,, $%*,#+-"#3! . . . $) ) &?5&'3' /) (where $% ) is cognate with Lucretius’ vis).76

So at this crucial juncture Lucretius is not only emphasising Epicurus’Greekness, but even acknowledging that the Romans are, philosoph-ically, the poor relations.77 The question ‘How can a Lucretius competewith an Epicurus?’ turns out to carry the subtext ‘How can a Romanphilosopher compete with a Greek philosopher?’

!& . 526 7698:A5,8?

What are we to make of these contradictions? Lucretius considers Greekculture artistically and philosophically superior, and yet at the same timedeeply alien.78 He floods his poem with Greek words, but scrupulously

(# ". Two languages, two worlds

74 See André (!"+$), p. +(.75 I am grateful to Roland Mayer for pointing this out to me. He illuminatingly compares the device

of using a Greek nominative-plus-infinitive construction at Catullus ).!–%, where the purportedspeaker is designated by a Greek noun, phaselus.

76 This Graecism is noted by Kenney (!"$!), ad loc., and I owe to David West the further point that$% ), rather than 2$"+, is the Greek form directly echoed by Lucretius.

77 For a further twist on the ‘swan’ theme, however, see pp. !)*>! below.78 To investigate how typical a Roman attitude this is would take me far beyond my present task.

But it is safe to say that it was widespread: see e.g. Woolf (!"")).

avoids them in the course of doctrinal exposition. Let me close with asuggested explanation of these anomalies.

Epicurus is a Greek, a voice from an alien culture to which Lucretiushas no interest in acclimatising himself or his reader. Lucretius’ mapping-out of the Greek and the Roman, e/ected by his strategic interweavingof Greek and Latin vocabulary, is a constant reminder of the gulf thatdivides the two worlds. But although Epicurus’ world is alien, his philos-ophy is not. It directly addresses the universal moral needs of mankind,and to that extent it transcends all cultural barriers. Lucretius, we haveseen, is constantly emphasising the barriers. It is precisely by drawingattention to the cultural divide between the Greek and the Roman, whilemaking Epicurean philosophy nevertheless thoroughly at home in hisown native language, that he proves to us its true universality.

3446?<,.

On p. (* I quoted Bailey’s list of ‘capricious’ Greek imports. In the ensuing dis-cussion I have dealt with nearly all of them, arguing that they serve a strategicfunction within deliberately Hellenising contexts. In this appendix I shall dealwith the handful of items left over from the same list: baratre ,,, "((, barathrum-, +*+, gigas ,- !&+, - !!$, lampas ,, $", - )*%, +!*, -, !!"+, lychni - %"(, sceptra -!!&$.

I have no need to fight very hard for these. As I have emphasised, a Greekword in company with others is significant, but an isolated one may easily passunnoticed. Some of these remaining Graecisms fall into the latter category:lychni at - %"(, sceptra at - !!&$, and barathrum at -, +*+ are likely to be such cases.(However, another occurrence of sceptra, at - !*&#, did prove to be a significantGraecism, above p. ((; and so, below, will another occurrence of barathrum.)

Gigas, used as by Lucretius to designate the Greek mythical beings of thatname, is extremely common in Latin authors. What indigenous Latin word doesBailey think Lucretius might have used in its place?

Lampas is so widely used in Latin as to fall far short of the ‘caprice’ Baileyalleges. I happily concede that at - +!* and -, !!"+ its Greek origins are simplyirrelevant. On the other hand, we have already met one significantly Hellen-ising use of it at ,, %( (above p. ((), and another can now be added: at - )*%,aeternam . . . lampada mundi is the sun in the explicitly Greek context of thePhaethon myth, where it is flanked by two compound adjectives (&""–)**), twooccurrences of the name Phaethon (&"$, )**), and an attribution of the wholestory to the ‘ancient Greek poets’ ()*(). Finally, at ,, $", animal species’ swifttransition from one generation to the next is compared to the handing on of atorch: et quasi cursores vitai lampada tradunt. This time for once there is noaccumulation of Graecisms, but it has long been recognised by Lucretianeditors that the reference is to a Greek institution, the torch race, to which Plato

Appendix ("

too (Laws $$+b) compares the passage from generation to generation. TheGreek word is, to say the very least, apposite.

This leaves just baratre at ,,, "((, where an old man lamenting his imminentdeath is rebuked by a personified Nature:

aufer abhinc lacrimas, baratre, et compesce querellas!Away with your tears, baratre, and curb your complaining.

Unfortunately this word is unattested. Just one item of evidence is cited,however, for the existence of a very similar Greek noun 2+" #+1#/),79 meaningsomeone who deserves to be hurled into the barathrum, the pit. Not surprisingly,many editors have preferred to emend, usually to balatro.80

It is, at least, surely no coincidence that the neuter noun barathrum occurs justeleven lines later in Lucretius’ poem (an occurrence missed by Bailey), where,speaking now in his own voice, he endorses nature’s rebuke by observing (,,, "++–$)

nec quisquam in barathrum nec Tartara deditur atra:materies opus est ut crescant postera saecla.And no one is sent down to the pit of hell: their matter is needed,so that future generations may grow.

In the current context, then, barathrum is the pit of hell. So if baratre wereretained and held to mean someone who deserves to be hurled into thebarathrum, there would be a most surprising irony in Nature’s addressing thewhinging old man with this particular term of abuse. Hardly a productive wayof conveying Lucretius’ principal message, that there is no pit of hell to fear!

In seeking a better solution, it is worth first observing that barathrum in line"++ (quoted immediately above) may well be added to our list of significantGraecisms. After all, it is there in company with Tartara, and they jointly invokethat terrifying realm of punishment familiar from Greek myth and literature.

In seeking to make sense of the preceding baratre ("((), we could do worsethan take our lead from this line. The chances are surely that here too the orig-inal text referred directly to the pit of hell, the barathrum. What has bedevilledattempts to cure the corruption is the assumption that the word must be a voca-tive. Why not emend in the following way?

aufer abhinc lacrimas, barathri et compesce querellas.Away with your weeping, and curb your complaining about thepit of hell.81

+* ". Two languages, two worlds

79 Cited by the editors from Ammonius, De di,. adf. voc. '&.%", 2+" #+1#/) µ"! . 4+! # /$ 2+#+" 1#/& +% @$/)+% .1#(6/). LSJ note hesitantly that it may occur with this meaning at Lucian, Pseudologista !$,adding, however, that this could be the neuter noun barathrum. I take it that the latter assumptioninfluenced Bockemüller’s conjecture barathrum.

80 Smith (!"$(), p. %+% reports that balatro had, before Heinsius, already been proposed by ‘anony-mous critics in Turnebus, Adversaria %*.%+’. Other emendations proposed are blatero (Merrill) andbarathrum (Bockemüller).

81 Querella + objective genitive is cited in OLD (s.v.) from Cicero and Apuleius. If alternatively onechose to construe barathri with lacrimas, one might think of Virg. Aen. , )+%, sunt lacrimae rerum.

(Since baratre, if retained, would be taken to stand for barathre, the emendationamounts to a further change of one letter only.) The proposal has one immedi-ate advantage. It supplies a piece of information which is otherwise leftunstated, that the old man – whose words were not actually quoted – has beencomplaining partly about the prospect of hell. And without an indication to thate/ect, one might be left wondering why Lucretius, at "++–$, should o/er hisrationalistic denial of hell as directly confirming Nature’s rebuke.

Appendix +!

1234567 &

Lucretius the fundamentalist

! . 42,:89842D ,? ,53:D

Virtually no facts about Lucretius’ life have been determined by modernscholarship, beyond a consensus that it was spent mainly if not entirelyin Italy, and that it terminated in the (*s '1.1 But for a Roman withphilosophical leanings those two facts in themselves ought to speakvolumes. He could hardly have chosen a better time to be alive. The lastfifty years of the Roman Republic were a period of unsurpassedphilosophical upheaval in the Graeco-Roman world. And for the firsttime ever the philosophical centre of gravity was shifting away fromAthens, with Italy capturing more than its share of the action. Theevents of the Mithridatic War ("!–#+ '1) – in particular, by a curious his-torical irony, the regime of the Epicurean tyrant Aristion (##–#+) – haddriven many philosophers out of the city.2 The Athenian schools wereno longer guaranteed the status of international headquarters for theirrespective movements.3 And in the resultant diaspora, many philoso-phers found their way to Italy. Here a ready-made audience awaitedthem, including plenty of Romans who had already trained at Athens inone or more of the philosophical schools.4

The leading figures of the Academy, Philo of Larissa and Antiochusof Ascalon, conducted their well advertised rift over the true nature oftheir Platonic legacy not in Athens but from bases in Rome and

+%

1 For a judicious survey of Lucretian biography, see Smith (!"$(), pp. ix–xxviii . I shall have nothingto add to it in this book, beyond my argument in Ch. %, §!* that Lucretius must have been toGreece.

2 See Ferrary (!"##), pp. )&(–#+, Habicht (!""(), esp. pp. %""–&!&, Dorandi (!""$).3 This is well argued, with regard to the Academy, by Glucker (!"$#), esp. pp. &+)/.4 The philosophical engagement of influential Romans during the late Republic is well brought

out by Rawson (!"#() and Gri0n (!"#") and (!""(). Jocelyn (!"$$) supplies further invaluablematerial, which does not on the whole seem to me to support his rather negative conclusions. Ihave myself tried, in Sedley (!""$a), to demonstrate the crucial role that philosophy played inCaesar’s assassination.

Alexandria respectively.5 And both became important figures at Rome,where their influence on leading public figures was considerable. AmongLucretius’ Roman contemporaries, Cicero became a follower of Philo’sNew Academy, while Brutus, Varro, Lucullus, M. Pupinius Piso andothers attached themselves to Antiochus’ Old Academy, which professedto have rescued the true philosophy of Plato from obfuscation and dis-tortion at the hands of the Stoics. Although it is debated how muchinfluence Antiochus and his pupils exerted on later philosophicaldevelopments,6 there can be no doubt of their importance to their owncontemporaries. Partly as a result, renewed study of Plato’s own text wasat this very time becoming the basis of a revitalised Platonism:7 theMiddle Platonist era was just getting under way.

Pythagoreanism was already in some sense a native Italian philoso-phy. Around this time it was enjoying an Italian renaissance, and was toexert considerable influence on the re-emerging Platonism. It had as itsleading figure a Roman, Nigidius Figulus.

Posidonius, the most influential Stoic of the day, was especially activein Lucretius’ own chosen domain, cosmology. Although he had chosenRhodes as the main base for his own teaching, he visited Rome at leasttwice, and was an intimate acquaintance of Cicero and other eminentRomans. Stoicism certainly had a considerable impact on Romanthought in the late Republic, although card-carrying Roman Stoics atthis date seem relatively few in number. The most celebrated of them allwas of course Lucretius’ contemporary the younger Cato.

Peripatetic philosophy too was gaining in influence. Manuscripts ofAristotle’s school treatises – works largely unknown or ignored duringthe preceding two centuries8 – had arrived in Rome along with Sulla’sother booty from Athens, acquired during the Mithridatic War. Owingto recent work by Jonathan Barnes,9 it can no longer be assumed thatfrom these copies the Peripatetic philosopher Andronicus was toproduce the first full edition of Aristotle’s treatises. Nevertheless, it wasat just this time, the mid first century '1, that Aristotle’s Categoriessuddenly and spectacularly burst onto the scene, provoking a stampedeof commentators, some favourable, others highly critical. At least oneof the first generation of Aristotelian commentators, the Stoic

!. Philosophy in Italy +&

5 On this dispute see e.g. Glucker (!"$#), Sedley (!"#!), Barnes (!"#"a).6 For doubts on this, see Glucker (!"$#), Barnes (!"#"a).7 On the reasons for the return to the study of ancient texts at this date, see Hadot (!"#$). That

Brutus, as an Antiochean, relied on a close knowledge of Plato’s text in )) '1 is argued in Sedley(!""$a). 8 See Sandbach (!"#().

9 The standard story is brillantly demolished by Barnes (!""$).

Athenodorus, a friend of Cicero and teacher of Octavian, was almostcertainly working in Rome by the mid first century.10 I find it hard todoubt that the rediscovery of the Categories is one which really did emergefrom the war booty at Rome.11 And it marked the beginning of theAristotelian commentary tradition, which was to remain the dominantform of philosophical exposition for the rest of antiquity. Thus, althoughthe exact timing is uncertain, it may well be that in the last decade ofLucretius’ life, when he was working on the De rerum natura, thephilosophical intelligentsia of Rome was already turning its attention toAristotelian school texts.

The history of Aristotelian scholarship is admittedly a speculativematter. But there can be no doubt that Antiochus and others had in anycase done much to rekindle interest in Peripatetic texts in Lucretius’ life-time.12 There were a number of prominent Peripatetics in circulation, atleast some of them based in Italy.13 The Antiochean M. Puponius Pisohad, even before studying with Antiochus, for many years kept thePeripatetic Staseas of Naples as his philosopher-in-residence atTusculum.14 Crassus, too, enjoyed the services of a resident Peripatetic.15

And the leading Peripatetic of the day, Cratippus, although based inAthens, was well known to, and very highly regarded by, a number ofRomans, Cicero included.16

Even Epicureanism had at least one major base in Italy. In the nextsection I shall have more to say about Philodemus’ Epicurean school atHerculaneum, which was at its zenith in the mid first century '1 whenLucretius was writing. It would not be at all surprising if Lucretius hadhad links with it, not only because he was himself an Epicurean, but alsobecause in the next generation leading literary Romans like Virgil andhis eventual executors Varius and Plotius Tucca were to maintain closecontacts with it.17 Influential Romans in Lucretius’ day who wereEpicureans included Cassius, later the assassin of Caesar, Cicero’s friendAtticus, and L. Calpurnius Piso Caesoninus, the target of Cicero’s In

+) #. Lucretius the fundamentalist

10 See Goulet (!"#"– ), , +()–$, for the probable identity of Athenodorus the commentator withAthenodorus the teacher of Octavian.

11 The Categories and De interpretatione appear in two versions of a list of Aristotelian works held toderive from a late third-century '1 source (one in Diogenes Laertius - %+, the other inHesychius), but it is generally agreed that they are a later interpolation there (Moraux (!"(!), pp.!&!, !#+–"*, &!&, Düring (!"($), pp. +", "*).

12 Cf. Cicero’s close study of Theophrastus and Dicaearchus in +*–(" '1 (Ad Att. ,, &.), !+.&). Theissue of Cicero’s knowledge of the Peripatos is thoroughly explored in Fortenbaugh/Steinmetz(!"#"). 13 See Gri0n (!"#"), p. $. 14 Cic. Fin. - #, De or. , !*). 15 Plut. Crass. &.

16 Plut. Cic. %). 17 See Gigante/Capasso (!"#").

Pisonem and patron of Philodemus. Even Caesar himself is widely sus-pected to have been an Epicurean.18

Cicero’s philosophical dialogues, even if they are dramatic fictions,cannot be wholly fictitious in representing many others from Cicero’scircle as committed exponents of current Greek philosophical systems.19

Indeed, Cicero’s correspondence with numerous contemporaries showsthem, in general, to be remarkably well attuned to live philosophicalissues and inter-school debates.20

Lucretius was extraordinarily lucky to find such a ready-made philosoph-ical environment for the maturation of his own ideas. No previous era atRome could compare with it. And it is doubtful whether any subsequentera could either: barely a generation after his death, Alexandria wouldlargely eclipse Rome as a philosophical centre. But did Lucretius seizethis unique opportunity? Does the De rerum natura, the most brilliantphilosophical composition to survive from its period, reflect the highlycharged philosophical atmosphere of mid first-century '1 Italy?

Amazingly, it does not.

% . 526 91288: 8@ 42,:8<6;A9

The Epicurean library discovered at Herculaneum, near Naples, in themid eighteenth century (see Ch. ), §! below), is almost certainly thatfrom the school of Philodemus,21 which seems to have flourished fromthe #*s '1 until late in the century.22 The school, although located in aGreek-speaking region, was close enough to Rome for its influence to befelt there. Cicero names its leading figures, Siro and Philodemus, asEpicurean familiares to whom he can turn when in need of advice onsome point of doctrine.23 And Philodemus’ library itself attests the deepinvolvement of contemporary Epicureans in current philosophical con-troversies, debating, for instance, the proper philosophical attitudes torhetoric, music and poetry, and concerned to meet Stoic and Peripateticchallenges on these and other issues.24

". The school of Philodemus +(

18 For details of these and many other known and putative Roman Epicureans in the late Republic,see Castner (!"##), who on pp. #&–+ judiciously sifts the evidence for counting Caesar amongthem. 19 However, for a cautionary note on this point, see Gri0n (!""(), p. &%(.

20 This is admirably demonstrated by Gri0n (!""().21 For a valuable synoptic survey of the library and modern work on it, see Gigante (!""(). For some

residual doubts about its ownership, see De Lacy (!""&).22 For Philodemus’ biography, and in particular the date and circumstances of his move to Italy

from Athens, see Dorandi (!""$). 23 Cic. Fin. ,, !!". 24 For a survey, see Asmis (!""*).

There has been much debate about Lucretius’ orthodoxy or hetero-doxy as an Epicurean choosing to write poetry. Philodemus was himselfa poet, and his aesthetic writings o/er important evidence for anyonesetting out to resolve the issue.25 But the study of these texts has so faryielded no good evidence of a direct influence of Philodemus onLucretius, or vice versa.26

Knut Kleve27 has opened new avenues for speculation by arguing thatthe fragments of a Latin text found in the Herculaneum library come froma copy of the DRN. This is not the place to examine the evidence adducedby Kleve. The fragments are very tiny, but he nevertheless builds quite animpressive case. I limit myself here to Kleve’s conclusion from his findings:‘Theories building on the assumption that Lucretius had no contact withcontemporary Epicureanism, su/er a serious setback.’ I do not think thatthis in any way follows. The books people have on their shelves, unlessthese contain autograph dedications, can tell us nothing about their per-sonal acquaintance with the authors of those books. It would only be if wehad reason to suppose that Lucretius’ poem was unavailable for publicpurchase in the first centuries '1 and 3< that anything could be inferredabout personal contact between him and the school at Herculaneum. Andsince (as my arguments in Ch. ( will confirm) there is excellent reason toaccept the ancient tradition that the poem was published posthumously,the chances are that any copy in the library was obtained only at a timewhen personal contact was no longer even possible.

Evidence of contact between Lucretius and Philodemus, then, if it isto be found, would have to be sought in the content of their writings.These unfortunately have for the most part too little common thematicground to be of much help.28 But one (partially) surviving work of

++ #. Lucretius the fundamentalist

25 There are a number of important papers relating to this question in Obbink (!""().26 Milanese (!"#"), pp. !&)–", detects echoes of Philodemus’ aesthetics (and a reaction to Stoic

aesthetics) at Lucr. , +)!–), but seems to me to be exaggerating some rather unremarkable resem-blances. 27 Kleve (!"#").

28 I consign to a footnote the problematic question of their views on the gods. Lucretius, in commonwith Cicero, Philodemus and most modern interpreters, treats the Epicurean gods as actuallyliving beings. I myself believe (Long/Sedley (!"#$), §%&), albeit controversially, that Epicurusintended them primarily as thought-constructs, saying little to resolve the question whether therewere any such beings actually alive in the universe. However, he clearly spoke at length abouthow we should think of the gods as being and living (including, perhaps, that we should think ofthem as living outside our world, cf. Lucr. - !)+–((), and the inference that these were objec-tively real gods may have been a natural one for readers to make. It is striking (cf. Long/Sedley(!"#$), , !)") that Lucretius’ indications of this assumption are evident only in his proems, uni-versally agreed to be his own original compositions, while in the main body of book - (!!+!–%%(),where I shall argue that he is following Epicurus’ text closely, he provides some of the strongestevidence for the thought-construct interpretation.

Philodemus which significantly overlaps with Lucretius’ interests is hisOn signs.29 This fascinating text is testimony to a highly sophisticateddebate between Epicureans and Stoics on scientific inference – the verybranch of philosophical methodology which underlies Lucretius’ ownpoem. Philodemus is no innovator here. He is reporting, in more thanone version, the arguments of Zeno of Sidon, the Epicurean school-head under whom he studied at Athens. In other words, this is above alla teaching text from Philodemus’ own school,30 illustrating exactly thekind of issues to which we would expect Lucretius to have been exposedif he had studied there.31

The debate achieved enough publicity to leave an unmistakableimprint on the Academic critique of Epicurean theology created byCicero in De natura deorum ,.32 Yet when we turn to Lucretius, the contrastis a stark one. His methodological pronouncements on the rules of infer-ence are confined to areas which we know to have been covered byEpicurus – above all, his very clear assertion of the principle of multipleexplanations in books - and -,.33 Although he naturally enough usesinductive and analogical arguments throughout the poem, there is noindication in these that he is concerned with resisting the challenges totheir validity which Stoic critics had posed and Zeno set out to answer.34

One very clear such case is the debate about the size of the sun. I shalldeal with this in §$ of the present chapter.

Reverence for Epicurus was to be expected of any Epicurean, butLucretius’ reverence is idiosyncratic in two ways. Not only did it, as I haveargued, lead him to ignore the contemporary philosophical environment.It also focused more narrowly on Epicurus himself than school practicein his day expected. As Philodemus’ writings make clear, it was normalfor contemporary Epicureans to assign virtually biblical status not to thewritings of Epicurus alone, but jointly to those of the foursome knownsimply as /$$ +% .5#"), ‘the Great Men’. These infallible four were the

". The school of Philodemus +$

29 See the edition of De Lacy (!"$#); also Sedley (!"#%), Barnes (!"##), Long (!"##a), Asmis (!""+).30 For this characterisation, and related aspects of Philodemus’ work, see Sedley (!"#"b), !*&–!$.31 Sign. itself is too late for Lucretius to have known it as a written text. On its date, see Long/Sedley

(!"#$), ,, %+&. 32 As shown by Auvray-Assayas (!""%) and Asmis (!""+).33 - (%+–&&, -, $*&–!!.34 Cf. the judicious remarks of Clay (!"#&), pp. %)–(. The best that Schmidt (!""*), p. +", can do

to find a Lucretian echo of these innovations is to cite , %+(–&%#, where ‘Induktionsschlüsse vor-liegen’. Actually there are no inductive arguments here, and only one analogical argument, butwhat really matters is that neither here nor anywhere else do we find Lucretius using Zenonian-style formulations of sign-inferences or taking account of the contentious issues about validitythat featured in the recent debate. There is nothing in this passage that could not be directlyechoing Epicurus.

founding figures of the school, Epicurus, Metrodorus, Hermarchus andPolyaenus,35 and all four were treated as absolutely and equally authori-tative. In this context, Lucretius stands out for his singular devotion toEpicurus himself, eloquently declared in the proem to book ,,, ("–!&):

tu pater es, rerum inventor, tu patria nobissuppeditas praecepta, tuisque ex, inclute, chartis, !*floriferis ut apes in saltibus omnia libant,omnia nos itidem depascimur aurea dicta,aurea, perpetua semper dignissima vita.

You are our father, the discoverer. You it is who supply us with afather’s precepts. It is from your scrolls, illustrious one, that likebees sipping everything in flowery glades we feed on all yourgolden sayings – golden, and ever deserving of unending life.

Just as he here declares his exclusive dependence on Epicurus’ own writ-ings, so too throughout the poem he o/ers no hint of bowing to otherEpicurean authorities.

& . 526 :8135,8? 8@ 526 ;,?<

To illustrate my thesis in a less negative way, let me now focus on a debatewhich had been simmering for two and a half centuries. Both the maindoctrinal schools of the Hellenistic era, the Stoic and the Epicurean, hadbeen founded around &** '1, at a time when medical opinion tended tofavour the heart as the location of the rational mind. This was certainlythe view of the great contemporary physician Praxagoras of Cos, laterinvoked as an authority on the issue by the Stoic Chrysippus.36 The alter-native of locating it in the head, although supported by the Pythagoreantradition and accepted by Plato in the Timaeus, was not endorsed byAristotle or, in the next generation, by the first Stoics and Epicureans.Zeno of Citium, the founder of Stoicism, was so confident on the pointthat he even produced his own syllogism in confirmation:

Voice comes through the windpipe. If it were coming from the brain, it wouldnot come through the windpipe. Where speech comes from is where voice alsocomes from. But speech comes from the mind. Therefore the mind is not in thebrain.37

Now, such was the commitment in philosophical schools to the truth ofthe founder’s word that subsequent Stoics could not simply disown this

+# #. Lucretius the fundamentalist

35 See Longo Auricchio (!"$#), Sedley (!"#"b).36 Galen, On the doctrines of Hippocrates and Plato (PHP) , $.!!SVF ,, #"$.37 Galen, PHP ,, (.#!SVF , !)#.

argument and its implications.38 Philosophical debate within schools waspresented as recovery and interpretation of the founder’s true views, nottheir replacement or revision. And Zeno’s argument was too unambigu-ous in its formulation to be explained away or rendered harmless byreinterpretation. That was most unfortunate. For within a generation ofZeno’s confident syllogism, Alexandrian physicians, operating with thegrisly insights vouchsafed by the human vivisections which they werepermitted to perform on prisoners from the local jail, had discovered thenervous system and put it beyond doubt that the brain, not the heart, isthe true command centre.39 Following these discoveries by Herophilusand Erasistratus, there is no record of any further doubt within medicalopinion that, if there is any seat of reason,40 it is located in the head.41

This left Stoic philosophers of subsequent generations looking asstranded as religious zealots defending the Genesis account of the crea-tion in the face of the Big Bang and Darwin. Not only did Chrysippusdefend the outdated Zenonian anatomy – earning himself and his argu-ments a merciless tirade of ridicule from Galen42 – but even Posidonius,

#. The location of the mind +"

38 Sedley (!"#"b). The only exception was Zeno’s early Republic, which some Stoics seem to havediscounted as pre-Stoic (Philodemus, De Stoicis ,.–.-, in Dorandi (!"#%)).

39 See Herophilus frr. +&–+ in von Staden (!"#").40 I add this qualification because it was open to Lucretius’ near-contemporary Asclepiades (SE,

M -,, &#*, Aetius ,- %.#), for one, to deny that there is any localised hegemonikon. It was longassumed that Lucretius knew Asclepiades, who in the nineteenth century was frequently, on noevidence at all, referred to as his ‘master’ (see Schrijvers (!"$+), p. %&#). However, the new revi-sion of his chronology (he was dead by "! '1, see Rawson (!"#%)) no longer makes that a veryappetising guess. There is in any case not the slightest sign in Lucretius of reaction toAsclepiades’ idiosyncratic theories (on which see Vallance (!""*)) – the anarmoi onkoi, the non-localisation of the hegemonikon, or any other in which he might have shown an interest. Pigeaud(!"#!) explicates Lucretius’ ‘eclectic’ physiology largely in terms of his agreements and disagree-ments with Asclepiades, but o/ers no evidence that Lucretius was actually aware of Asclepiades’work. See also n. (! (p. $%).

41 See the wide range of doxographical material on this topic assembled and studied by Mansfeld(!""*), pp. &*"%–!*#.

42 Galen, PHP, esp. ,, (. Diogenes of Babylon is another Stoic reported as backing Zeno’s thesis:SVF ,,, Diog. %", &&. In the latter (of which a more up-to-date text can be found in Henrichs(!"$)), pp. !"–%*, or in Obbink (!""+), pp. !"–%*), Diogenes attributes this same view correctlyto Chrysippus, but also speaks of ‘some Stoics’, to whom Chrysippus appears to have been reply-ing, who had located the hegemonikon in the head. Since these are probably pre-Chrysippean(contra Mansfeld (!""*), p. &*"( n. !)(; the text does not make this explicit, but Mansfeld is defi-nitely wrong to make them later than Diogenes of Babylon, since it is Diogenes who is quotingthem), I assume them to be first-generation Stoics, writing before Zeno’s death, when schoolorthodoxy first became established. The one genuinely anomalous text is Aetius ,- %!.), where‘the Stoics’ are given the thesis that ‘the hegemonikon itself, like the world’s, resides in our spher-ical head’. This, a transparent import from Plato’s Timaeus, is out of step not only with all theother evidence for standard Stoicism, but even with the preceding sentence in Aetius’ report,which apparently assumes the Zenonian anatomy. (I am not clear why Mansfeld, loc. cit., thinksCornutus an exception to the generalisation at Aetius ,- (.+ that ‘all’ Stoics located thehegemonikon in the region of the heart.)

despite incorporating within his Stoicism the tripartite division of thesoul which had enabled Plato to separate the rational mind, in the head,from the ‘spirited’ part in the chest and the ‘appetitive’ part in theabdomen, could not follow Plato on this point, and had no choice but tolocate all three soul-parts in the heart.43

Now it appears that the Epicureans faced the same problem. Amongthe books in Philodemus’ library were several by Demetrius of Laconia,the eminent Epicurean of the late second and early first centuries '1.One of these (P.Herc. !*!%) is an exegetical treatise on a series of crucesin foundational Epicurean texts,44 and at one point in its fragments(cols. .:,,–.:-,,) Demetrius is clearly discussing a passage in whichEpicurus had argued that the rational mind is in the chest. The state ofthe text leaves it quite unclear what concession, if any, Demetrius andhis fellow-Epicureans are making to modern science. But Demetriusshows himself well aware of what he calls an ‘argument used by manydoctors to prove that reason is located in the head’ (.:-,, $–!!). And hementions the expedient, adopted by some leading Epicureans, of posit-ing a scribal error and emending Epicurus’ text on the point (.:,-).Whether the object was to justify Epicurus’ statement on the locationof the mind, or to absolve him of having made it in the first place, isunclear. But the recognition by contemporary Epicureans of thecurrent state of medical opinion and of the challenge it posed for themis not in doubt.

And where does Lucretius stand on this fascinating issue? The extra-ordinary truth is that he is so out of touch with his own school, and withmodern science, as never to have even heard of the debate. Such is hisreliance on Epicurus’ own word that he not only repeats his founder’sargument that the mind is in the chest,45 but shows himself altogetherunaware of the challenge to it that expert medical opinion now presents.How else can we explain the following argument, which he deploys ontwo separate occasions?

At ,,, $#)–"$ he argues that the mind could not exist outside the body,and at - !%#–)! that the cosmic masses like sea and earth could not them-selves be animate. On both occasions the ground is the same. A mindcannot exist in just any physical location. Not only can it exist only in anorganic body, but it cannot even exist in just any part of that body (,,,$##–"&!- !&%–$):

$* #. Lucretius the fundamentalist

43 Galen, PHP -, %.(!Posidonius @!)+ Edelstein/Kidd (!"#").44 See the edition by Puglia (!"##).45 ,,, !)*–%. For Epicurus’ own very similar argument, see scholion on Ep. Hdt. ++.

sic animi natura nequit sine corpore oririsola neque a nervis et sanguine longius esse.quod si posset enim, multo prius ipsa animi vis $"*in capite aut umeris aut imis calcibus esseposset et innasci quavis in parte soleret,tandem in eodem homine atque in eodem vase manere.

Thus the mind’s nature cannot exist in isolation, without a body,nor be separated from sinews and blood. For if it could, muchsooner would the mind’s power be able itself to exist in the heador the shoulders or the soles of the feet – and to be born in any part youlike, while still remaining in the same person and the same con-tainer.

The words which I have italicised are hardly those of someone in tunewith the contemporary debate. The rhetorical force of Lucretius’ argu-ment depends on the expectation that it will sound as silly to imagineyour mind in your head as in your foot.46 He must be astonishingly outof touch with current opinion on the matter.47 It is as if an evangelicalmissionary in twentieth-century Britain were to tell the public ‘If you aregoing to believe things like that, you might as well deny that God madethe world in six days.’

This last is the analogy which suggests the title of my chapter,‘Lucretius the fundamentalist’.48 Lucretius cannot, in my view, be aninteracting member of any current philosophical circle. His inspirationcomes from the unmediated scriptures of the philosopher he reveres,Epicurus himself. Like any fundamentalist, he does not expect thenumerous contributions made since the composition of his sacred

#. The location of the mind $!

46 Hence I am puzzled by the claim of Mansfeld (!""*), p. &!)" that Lucretius’ argument shows heis aware of the alternative view that the mind is in the head. Note too that Lucretius does,throughout the poem, consistently locate thought in the chest – although the brain is often ille-gitimately smuggled in by his translators, and interpreters often falsely assume that ‘mind’implies ‘head’ for him (e.g. the influential suggestion of Kenney (!"$!), $+–$ that at ,,, !( divinamente coortam is an implicit reference to the birth of Athena from the head of Zeus). The onlyplace where he might appear aware of the head’s claims is at ,,, !&#, where the animus is said caputesse quasi et dominari in corpore toto. However, I take it that here the metaphor of ‘head’ derives fromthe word’s common usage to designate the source of a river, as also at - +*!.

47 It is often thought (see esp. Gottschalk (!""+)) that on another aspect of the soul, namely itscomposition, Lucretius reflects developments postdating Epicurus, since to the three ingredients(wind, heat and the unnamed substance) listed at Ep. Hdt. +& he joins the doxographical tradi-tion (cf. Aetius ,- &.!!) in adding a fourth, air. My preferred guess would be that Epicurus’ fullaccount in Nat. already listed the four, but that his summary in Ep. Hdt. subsumes wind and airunder the deliberately vague description 6#/'"µ%"#""'-+-/. 5"! 6."&" µ+-$, which avoids actu-ally naming a component, no doubt because ‘wind’ and ‘air’ consist of the very same atoms,albeit in di/erent patterns of motion.

48 My thanks to David West, who first suggested to me that ‘fundamentalist’ was the word I neededto capture my view of Lucretius.

scriptures, either by his school or by its critics, to have added anythingworth taking into account.

I recognise the paradox of speaking in these quasi-religious termsabout a self-declared enemy of religio. But it is Lucretius himself who,in his proems to books ,,, and -, declares the religious nature of hisdevotion: Epicurus, whose divinely conceived philosophy is Lucretius’own inspiration and guide, was himelf a god, indeed a god whoeclipsed the traditional Olympian deities in the importance of hisbenefactions to mankind. Without exploring Lucretius’ fundamental-ism, we cannot hope to appreciate the full force of these religiousdeclarations.

I do not mean to suggest that Lucretius was a recluse, either sociallyor intellectually. In every other respect he shows himself an acuteobserver of his own society, sensitive and subtle in argument and thor-oughly versed in the literary traditions of both Italy and Greece, includ-ing Hellenistic as well as archaic Greek poetry.49 Even Hellenisticintellectual trends, such as euhemeristic and other allegorical rational-isations of religion, have left their mark on him.50 It is only in his hard-core scientific and philosophical beliefs that his fundamentalism showsitself.51

$% #. Lucretius the fundamentalist

49 See esp. Kenney (!"$*). 50 See Gale (!"")a), esp. pp. %+–&*, $(–#*.51 I here dissent, with some misgivings, from the important contributions of P. H. Schrijvers, who

has been the most eloquent modern defender of Lucretius’ familiarity with Hellenistic thought,especially scientific thought. But I shall take up – and with apologies for the extreme condensa-tion – only the cases where he o/ers explicit evidence for his position, and also ignore those wherethe alleged debt (e.g. to Dicaearchus, Schrijvers (!"")), (!""+); to Palaephatus, Schrijvers (!"#&)),even if accepted, might in principle be mediated by Epicurus, as well as those where no directdebt is even claimed. (I have not, at the time of writing, seen a text of Schrijvers (!""$).) (!)Schrijvers (!"$+), pp. %&#–", links Lucretius’ accounts of hunger and nutrition (,- #(#–+") tothose of Asclepiades, who had similarly explained them in terms of the dilation and refilling ofbodily ducts (6/" #/$). Against this (a) 6/" #/$ are already a widely used explanatory device inEpicurus (see Usener (!"$$), s.v.); (b) there is no record of Epicurus having a di,erent account ofhunger and nutrition from that in Lucretius (contra Solmsen (!"(&), p. &" n. !", Epicurus fr. %"&Usener, where absorption of food is a case of attraction of like to like analogous to magnetism,is fully consistent both with Lucr. ,- #(#–+" and – read in context – with -, ")+–$; for the roleof 6/" #/$ in Epicurus’ theory of magnetism, see fr. %"& Usener, p. %*", l. !& and p. %!*, l. %%); (c)in the third century '1, Damoxenus’ comic cook, a self-declared follower of Democritus andEpicurus, says -/$4+#/&' . "$& ) -/&! ) 6/" #/&) 8 /$ ;&µ/! ) /$ µ+7(' ) 6+.-+;/&' '&.$"'-+-+$ (ap.Athenaeus !*%'!Damoxenus fr. %, %"–&* Kassel/Austin); (d) see n. )* (p. +") on Asclepiades’date; and (e) as Pigeaud (!"#!), p. !#" points out, the account of thirst with which this passagecontinues (,- #$*–+) is very di/erent from Asclepiades’. (%) In the ‘Discussion’ appended to Furley(!"$#), p. &!, Schrijvers cites Gellius .-,, !!.!!–!& as evidence that the expression -#+;"$'++& #-3#$"+, apparently reflected in Lucretius’ asperiora . . . arteria (,- (%"), was introduced byErasistratus, i.e. too late to be taken up by Epicurus; but Gellius does not say this as far as I cansee, and Anon. Lond. -,,, %"–&* is powerful evidence that the term is much older. (&) OnAenesidemus, see §# below.

) . 42,:89842,13: 8448?6?5 9

This impression is amply confirmed when one turns to Lucretius’polemics.52

Epicurus’ own targets in his philosophical critiques had been aboveall the Presocratics – especially his chief forerunner Democritus, alongwith his own Democritean teacher Nausiphanes – and Plato.53 The con-temporary Stoic school had apparently achieved prominence too late tofeature as a further target. Later Epicureans adopted Epicurus’ enemiesas their own, but added contemporary ones, above all the Stoics.Lucretius’ contemporaries illustrate this point e/ectively. Cicero’sEpicurean spokesman Velleius in book , of the De natura deorum criticisesPlatonic and Stoic theology in tandem,54 and Philodemus’ targetsinclude not only the by now obscure Nausiphanes but also near-con-temporary Peripatetics and Stoics.

Lucretius’ most sustained polemic is at , +&(–"%*, concerning the ulti-mate constituents of things. Here he criticises the monists, exemplifiedby Heraclitus, the four-element theorists, represented by Empedocles,and finally Anaxagoras. There is little doubt that this derives fromEpicurus’ own critique in books .,- and .- of his On nature.55 None ofthe named philosophers presented, by any stretch of the imagination, alive philosophical challenge to the Epicureans in Lucretius’ own day, but,as I have said, it was quite unexceptional for contemporary Epicureansto perpetuate Epicurus’ own invectives in this way. Where Lucretiusdi/ers from his contemporaries is in the entire absence of a matchingcritique of Stoic physics.

Many editors have assumed that the Stoics would be included amongthe unnamed followers of Heraclitus’ doctrine that everything consistsof fire (, +&(–$, $*(–+).56 In fact, though, they were committed expo-nents of the four-element theory, and would be much more likely, if theywere implicitly present in Lucretius’ list, to be included amongEmpedocles’ unnamed followers in his four-element theory (, $!)–!(,

%. Philosophical opponents $&

52 For a general survey of these, see Kleve (!"$#). 53 For Epicurus’ polemics, see Sedley (!"$+a).54 Cic. ND , !#–%&. 55 See Ch. ), §!*.56 This would gratuitously presuppose a distortion of Stoic physics on Lucretius’ part – one admit-

tedly found in much of the modern literature, but in none of the ancient. In some contexts theStoics did call fire ‘the element par excellence’, but meant by this that it was privileged above theother three elements in that between world-phases everything became fire, and hence that ourworld starts from fire (SVF ,, )*&). No plausible reading of Stoic physics could make fire the soleconstituent of the world. Rather, at least for Zeno and Cleanthes, the role of fire in the presentcosmos was as an active force imbuing and controlling the world’s material constituents. Cf. Long(!"$(–+), p. !)*.

$&)–().57 But although the four elements did indeed play an importantrole in Stoic cosmology, not even they were viewed by the Epicureans asthe Stoics’ candidate for the primary constituents of things, which wererather, quite correctly, taken to be the two principles ‘matter and god’.Diogenes of Oenoanda goes through just the same list of targets asLucretius (di/ering only in that he assigns names to the ‘water’ and ‘air’monists, left anonymous at DRN , $*$–"). But, significantly, he adds theStoic principles at the end of it (+ , !*–,, " Smith):

Heraclitus of Ephesus said that fire is the element, Thales of Miletus water,Diogenes of Apollonia and Anaximenes air, Empedocles of Acragas fire, air,water and earth, Anaxagoras of Clazomenae the ‘homoeomeries’ of eachthing, and the Stoics matter and god.

It is the absence of any such updating in Lucretius’ list of targets thatmakes him an anomalous figure within the Epicurean tradition. No onegenuinely concerned to combat Stoic physics could have failed to alludeat some point in his book to these two principles, or for that matter tothe ordering role of pneuma in the Stoic cosmos.

But do the Stoics never feature as targets for Lucretian polemic?Editors long assumed that they frequently do so feature. But more thanthirty years ago David Furley, in a seminal if controversial article,worked systematically through these purported cases and brought highlypersuasive arguments to bear against the necessary identification of anyof the opponents as Stoics.58

The one systematic attempt to refute Furley’s arguments has beenmade by Jürgen Schmidt.59 In reply to Schmidt, I will content myself withthe following observations. Much of his case rests on two passages ofLucretius which, as I will be trying to show in the next two sections, havein view not Stoic cosmology but a Platonist forerunner of it.60 And a good

$) #. Lucretius the fundamentalist

57 Thus SE PH ,,, &! classes the Stoics as four-element theorists, while implicitly acknowledgingthat it might be more correct to call them matter–god dualists. 58 Furley (!"++).

59 Schmidt (!""*).60 I therefore take the same to be true of the critique of anthropocentric teleology at ,, !+$–#% (dis-

cussed by Schmidt at pp. !(%–+*), which is presented (!#%) as a foretaste of the book - passage.I am also unpersuaded by the argument of De Lacy (!")#), taken up by Schmidt, that Lucretius’replies to anthropocentric teleology at - !"(–%&) are dependent on Carneades’ critique of Stoiccosmology. Even supposing that Carneades was the source of the objection at Cic. Ac. ,, !%* andPlutarch fr. !"& Sandbach (!Porphyry, Abst. ,,, %*) that the world is too full of pests to have beenmade for human beings, it is such an easy objection to anthropocentrism that we need hardlydoubt that Epicurus or Lucretius could have hit upon it independently. Besides, Lucretius’expansion of the objection tanta stat praedita culpa (,, !#!) at - !"(–%&) has a twist not found in theAcademic tradition – that other species have a better claim than we do to be the true beneficia-ries.

deal of what remains turns on Lucretius’ interest in allegory: even if thereis a Stoic influence here,61 it is likely to be one which has come to himthrough Hellenistic literature and literary theory, not through an interestin Stoicism as such.62 That leaves just two passages where it can be madeto look plausible that Lucretius has a specifically Stoic target in mind.

( . 17635,8?,9;

One of these passages is - !(+–%&). Here Lucretius argues against thosewho maintain that the world was created by the gods for the sake ofmankind, and that it will therefore last for ever. These are standardlytaken to be the Stoics, on the ground that no one before them held thatthe world was created for the sake of man.

The thesis attacked is that the world was divinely created, and there-fore will never be destroyed. But most Stoics held that the world will oneday be destroyed, in the conflagration. Those few who dissented, prin-cipally Boethus and Panaetius, not unnaturally adopted the matchingthesis that the world had no beginning either.63 The asymmetrical thesisattacked by Lucretius – creation, but no destruction – was regularly seenas the idiosyncratic view of Plato’s Timaeus.64 This attribution was

$. Creationism $(

61 For some cautionary remarks about the extent of Stoic concern with ‘allegory’, see Long (!""+b).62 Schmidt focuses on (a) Lucretius’ critique of allegorical interpretations applied to the worship of

Cybele (,, +**–+*), (b) his image of the Golden Chain at ,, !!(&–+, and (c) his favourablecomparison of Epicurus to Heracles in - %%–(). Of these (a) seems to me rather persuasive(Schmidt (!""*), pp. !!&–))). But (b) (pp. !))–(!, following Ernout/Robin (!"+%), , &+(, and othercommentators) is very weak evidence for a specifically Stoic allusion. Allegorisation of theGolden Chain was extraordinarily widespread (see Lévêque (!"(")), and went back at least toPlato (Tht. !(&c–d). Lucretius’ remark haud, ut opinor, enim mortalia saecla superne | aurea de caelo demisitfunis in arva (,, !!(&–)) imagines the golden chain as a vehicle by which life might have descendedto earth from heaven. It has nothing whatever to do with the Stoic allegorisation of the GoldenChain as the universal causal nexus, and it is doubtful whether it deserves the name ‘allegory’ atall. I am no more persuaded by (c) (pp. !+!–%). It is frequently claimed that Heracles was a Stoichero, but there is only limited truth in this. Down to the first century '1, Heracles the legendarybenefactor of mankind (which is what Lucretius takes him to be) is no more prominent inStoicism than outside it. Heracles the allegorised moral hero is a separate tradition, associatedmainly with Prodicus and the Cynics, and apparently not absorbed into Stoicism until the firstcentury 3< (it is found in Seneca, Epictetus and Dio Chrysostom).

63 See Philo, Aet. $+–##. I take it that /$$ 6"#$! -/! . C/31/" . ($#), whose argument specifically rejectscosmic generation as well as destruction, are the entire Stoic group singled out at $+ for theirrejection of ekpyrosis.

64 E.g. Philo, Aet. !&–!+; cf. ib. !$, where the one possible exception, and a feeble one, is that Hesiodmight have held the Platonic view, because he describes the world’s beginning but says nothingabout its end. The one other exception known to me is at ps.-Plutarch, Epit. ,, ).!>%, &&*.!(–&&!.&Diels (!#$"), where in most versions the asymmetric Tim. thesis is attributed without discriminationto Pythagoras, Plato and the Stoics. But that an unwarranted conflation has occurred (cf. Diels, ib.introd., !!–!%) is clear from the parallel Stobaeus text, where the standard distinctions are observed.

accepted by, among others, the Epicureans, who expressly distinguishedthe Platonic from the Stoic position in this regard.65 What is more, theTimaeus also contains the principal thesis attacked by Lucretius – that itis thanks to the benevolence of its creator that the world will never bedestroyed (Tim. &%c).

The one element which the Timaeus does not supply is the specificthesis, so characteristic of Stoicism, that the world was created for thesake of mankind. While it would no doubt be a mistake to credit anysuch view to Plato,66 it remains entirely possible that the Timaeus wasbeing read in this way by Epicurus’ day. At &*b–c Plato says that theworld ‘truly came to be (4".""'1+$) an ensouled and intelligent animal,owing to the god’s providence (6#/" ./$+.)’.67 Divine ‘providence’ alreadyhas strongly anthropocentric connotations (e.g. prominently atXenophon, Mem. ,- !), and any such impression could well have beenstrengthened by &"b–e, where there is a clear indication that the heav-enly bodies were created for the sake of the human race. Could it be thatin the Academy of the late fourth century, under Zeno’s teacher Polemo,this anthropocentric reading of the Timaeus was already being taught, tobecome in due course a mainstay of Stoic cosmology?68

There is at least one clear parallel for such a development.Theophrastus (fr. %&* FHS&G) gave a paraphrase of the Timaeus whichsounds uncannily like the Stoic twin physical principles, ‘matter’ and‘god’: the world is analysable into a passive principle, ‘matter’, and anactive causal principle ‘which he attaches to the power of god and of thegood’. It is hard to believe that this highly revisionist interpretation is onethat Theophrastus had arrived at simply by his own reading of theTimaeus. He surely must be echoing a way in which that text was cur-rently being expounded in Plato’s own school69 – a hypothesis whichwould also comfortably explain how the matter–god doctrine passed indue course into the Stoicism of Polemo’s pupil Zeno.

In the current state of our evidence, the picture which I have justsketched of Platonist cosmology at the end of the fourth century cannot

$+ #. Lucretius the fundamentalist

65 Cic. ND , %*. 66 See Owen (!"$&), pp. &((–+.67 The problems of construing this sentence, discussed by Runia (!"#"), should not a/ect the

present issue.68 It cannot, however, have been the position of Crantor, whose commentary on the Timaeus

belongs to this period, since he rejected the genetic interpretation (Procl. In Tim. , %$$.#), whichthis version of the anthropocentric reading assumes.

69 There may conceivably be a parallel explanation for Theophrastus’ curiously inaccuratesummary of the Timaeus at Sens. (–+, #&–+, on which see Long (!""+b). He seems less faithful, inthis doxographical context, than at CP -, !.

be directly confirmed. But there is powerful indirect confirmation for itin the philosophy of Antiochus of Ascalon. Antiochus (see pp. +%>&above) presented his views as a return to the ‘Old Academy’, meaningby this not so much the text of Plato as the philosophy of the Platonicschool under Polemo70 – its final phase of doctrinal Platonism before itfell into the hands of the sceptical New Academics. Now Cicero’sAcademica includes a good deal of information on Antiochus’ ‘OldAcademic’ physics.71 It is a system which is manifestly based on theTimaeus,72 yet so closely prefigures Stoicism that some have questionablysupposed it to be nothing more than his retrojection of Stoic physicsonto Platonism. In fact, however, it di/ers from Stoic cosmology in justthe sort of ways one would have predicted for Polemo. For example, onthe one hand it already contains the analysis of the world into an active‘power’ and passive ‘matter’. On the other hand, the passive principle isdescribed in terms unmistakably echoing the ‘receptacle’ in the Timaeus;and the active ‘power’ di/ers from the Stoic one in that, although it con-stitutes the essence of an immanent god, it is not fully identified with thatgod, which is itself a further ‘power’, namely a ‘quality’ compounded outof the active power and passive matter.73 This relation between god andthe active principle seems entirely compatible with what we have justseen to be Theophrastus’ description of the active principle as one whichPlato ‘attaches to the power of god and of the good’, without any formalidentification between the two.74

Significantly, this same account of early Platonist physics contains themain ingredients of the anthropocentric thesis attacked by Lucretius.First, ‘Old Academic’ physics holds that the world is governed for thegood of mankind:75 the world soul’s most immediate concern is said tobe the motions of the heavens (a central message of the Timaeus), butafter that ‘those things on earth which relate to mankind’. Second, there

$. Creationism $$

70 Cic. Ac. ,, !&!, . . . Polemonis, quem Antiochus probat maxime . . . Cf. Barnes (!"#"a), p. $#. However,Barnes is wrong in adding ‘Antiochus is also said to have converted to Stoicism.’

71 Ac. , %)–". 72 See Dillon (!"$$), pp. #%–#, esp. #%–).73 Cic. Ac. , %), qualitas is compounded out of the active vis and passive matter; ib. %": quam vim [sc.

qualitatem] animum esse dicunt mundi, eandemque esse mentem sapientiamque perfectam, quem deum appellant,omniumque rerum quae sint ei subiectae quasi prudentiam quandam, procurantem caelestia maxime, deinde interris ea quae pertineant ad homines. (Platonist ‘quality’ thus looks like an antecedent of Stoic pneuma.)

74 Theophrastus fr. %&* FHS&G. The suspicion that the reading of the Timaeus in terms of onlytwo principles, rather than three, is itself a sign of Antiochus’ retrojection of Stoic physics ontothe early Academy (e.g. Görler (!"")), pp. ")"–(!) can be adequately countered by citing thissame fragment of Theophrastus, which shows that such a reading of the Timaeus was already incirculation by Theophrastus’ day. (I must reserve for another occasion my fuller reply to Görlerand other proponents of the Stoicising interpretation.)

75 See the end of the passage quoted in n. $& above.

are clear signs of the asymmetric thesis – creation, but no destruction.On the one hand the cosmological thesis is so worded as to imply that theworld had a beginning.76 On the other hand it also, according to Cicero,declared the world soul to be everlasting77 – on the ground that there isnothing strong enough to destroy it – and added that the world-orderwhich it governs is likewise everlasting.78 Hence the Academy of the latefourth century, at least as interpreted by Antiochus, was committed to theasymmetric view that the world was created but, contingently, will last forever. This is not the retrojection of a Stoic thesis, and it is hard to imaginewhat might have led Antiochus to invent the attribution.

Antiochus thus provides indirect but extremely strong evidence thatthe thesis attacked by Lucretius is to be identified, not with the tenets ofStoicism, but with their Platonist antecedent in the school of Polemo –in other words, with the cosmology of Plato as it was being expoundedin Epicurus’ day by Plato’s own school. This finding underlines the keyrole of Polemo’s Academy in the emergence of Hellenistic philosophy79

– a story still waiting to be told in full.

+ . B6816?57,9;

The second passage where Lucretius is widely held to have a Stoic targetin his sights is , !*(%–!!!&. Here he criticises a geocentric theory, accord-ing to which: (a) things on all sides tend inwards towards the centre ofthe universe; (b) people walk upside down in the antipodes, where it isday when we have night, winter when we have summer, etc.; (c) whileearth and water are centripetal, air and fire are actually centrifugal; (d)the upward motion of fire from the earth feeds the heavenly bodies,which are themselves fiery, and also (apparently: the text is deficient here)enables nutriment to travel up into the branches of trees.

Against the then universal assumption that these geocentrists are theStoics, Furley pointed out that the recorded Stoic theory of cosmicarrangement appears to answer one or more of the criticisms levelled by

$# #. Lucretius the fundamentalist

76 Ib. , %#, . . . unum e,ectum esse mundum. The attribution of this cosmology to Polemo is supportedby the unavailability of any other leading figure in the early Academy to be its author. Polemo’spredecessors Speusippus and Xenocrates, and his principal colleague Crantor, all rejected theliteral interpretation of creation in the Timaeus.

77 The main Platonic antecedent is Tim. &%b#–&&b!. Although there the context strictly concernsthe world body, it entails the same kind of indestructibility for the world soul too.

78 Cic. Ac. , %#–".79 Unfortunately the direct evidence for Polemo, beyond his ethics, is very thin: see Gigante (!"$+),

where fr. !%! (Aetius , $.%") is all that survives of his physics.

Lucretius – strong evidence that Lucretius’ criticisms are derived fromEpicurus, who will himself have been attacking some precursor of theStoic doctrine. I do not think that Schmidt has seriously underminedFurley’s case, or provided adequate grounds for his alternative proposalthat Lucretius is drawing on more recent Academic attacks on Stoic cos-mology.80

There is a very strong additional reason for supporting Furley’s denialthat the target is Stoic. It seems to have gone unremarked that onefeature of the theory, item (d), is definitely not Stoic. There is abundantevidence that the Stoics held that the heavens are nourished by moistureevaporating from terrestrial waters, and not by fire.81 Indeed, whereasthe moisture theory is an old one, going back at least to Heraclitus82 andcriticised at length by Aristotle in the Meteorologica,83 the idea that theheavenly bodies are nourished by terrestrial fire is very unusual. (That isa point to which I shall return in a moment.)

The fact is that we have no record of precisely this version of thetheory being held by anybody at all (including the early Aristotle, towhom Furley suggests attributing it). But that is hardly surprising. In par-ticular, we have virtually no evidence about the cosmology taught in theAcademy between the death of Plato and the advent of Stoicism, but Ihave nevertheless argued in the previous section that at least oneLucretian argument in book - is in fact targeted at a Platonist develop-ment of ideas contained in Plato’s Timaeus, one dating from the end ofthe fourth century. It is more than likely that the version of the geo-centric theory attacked by Lucretius comes from that same stable.

The Timaeus contains most of the materials needed for such a theoryto be worked out, including the paradoxical image of the antipodes.84

And, much more important, it is to the Timaeus that we can trace theextremely unusual thesis about the nourishment of the stars: terrestrialfire travels upwards through attraction, according to Plato, heading o/to join the main cosmic mass of fire in the heavens.85 This surely lies

&. Geocentrism $"

80 Schmidt (!""*), pp. %!%–%%. His assimilation of Lucretius’ arguments to Academic ones which canbe recovered from Plutarch relies on the assumption that Lucretius is trying to expose an inter-nal inconsistency in the position he is attacking, one between universal centripetalism and thecentrifugality of air and fire. But none of Lucretius’ arguments turns on any such contradiction.

81 Aetius ,, %*.), %&.(, Arius Didymus fr. && Diels, Cic. ND ,, )*, )&, #&, !!#, ,,, &$, DL -,, !)(.82 Heraclitus: Aetius ,, %*.!+ (see %%3!% DK). Thales: ib. , &.!. Innumerable further references in

Pease (!"((–#), ,, +&(–+. Cf. esp. Macrobius, Sat. , %&.%, who goes so far as to attribute it to allphilosophers. 83 Ar. Meteor. ,, %, &()b&%–&((a&%. 84 Tim. +&a%–).

85 Ib. +&b%–), e&–$. Compare Lucr. , !*"*–!, et solis flammam per caeli caerula pasci, | quod calor a mediofugiens se ibi conligat omnis with Tim. +&b%–), "&. -D' -/&' 6+.-/! ) -/" 6D ,+1 & /* . 3$ -/&' 6&#/! ) "$%73;"µ+""7$'-+ %&" '$), /&( ,+$! 67"$''-/. +) . 3& 1#/$'µ"" ./. "$%3 6#/! ) /* %""#"-+$.

behind item (d) in the theory attacked by Lucretius. The Stoics in thismatter followed an old and widely favoured tradition about the nourish-ment of the heavens by moisture, while the geocentric theorists attackedby Lucretius were following a much more unusual tradition, apparentlystemming from Plato.

There is a further Platonising ingredient in the thesis under attack,again one that distinguishes it from Stoic cosmology. According toLucretius, the opponents make the matter constituting our world gather,not around its own centre, but around the centre of space. That is whyhe is able to object, (i) that in infinite space there is no centre (, !*+#–$%),and (ii) that space has no causal properties which would allow its centreto exercise such a role (, !*$%–#%). Now, as Furley has emphasised, thiswas certainly not the Stoic position. From Zeno onwards, the Stoicsmaintained that the finite mass of body within infinite space gatheredaround its own centre. The alternative idea that space itself supplies thecentre around which matter gathers can be plausibly traced, once again,to the Timaeus (which itself in this case had earlier antecedents).86

Here (+%e!%–+&a%) we learn that – implicitly even if our world did notexist – any symmetrical body placed at the centre would remain there,because of its equidistance from the extremities. This passage certainlyinvites the interpretation that space as such has a centre. Now it is truethat Lucretius’ argument (i) is that infinite space can have no centre,whereas Plato’s ‘centre’ is said to be equidistant from the extremes, andtherefore can only be the centre of the finite space which our worldoccupies. But this should not be an obstacle. It may well be, as Furleymaintains, that the infinity of space is a premise supplied by Lucretius,not by the opponents. However, for anyone who feels that this excessivelyweakens the ad hominem force of Lucretius’ objection, it is worth point-ing out that the Timaeus easily lends itself to the interpretation that spaceis infinite.87

How else is a reader who possesses a clear notion of geometrical space(which Plato himself may have lacked)88 to make sense of Timaeus’concession that the creator could if he had wanted have made more thanone world, even infinitely many (&!a–b), so that it was by his choice that

#* #. Lucretius the fundamentalist

86 Plato, Phaedo !*#e-!*"a (and see Kingsley (!""(), esp. p. #", for the proposal that this itself is ofPythagorean origin), and Anaximander 3%+ DK!Aristotle, DC %"(b!*–!+ For doubts about theaccuracy of this attribution to Anaximander, see Furley (!"#"b); Anaximander can in any casehardly be Lucretius’ primary target, being most unlikely to have posited the antipodes.

87 I am grateful to Hendrik Lorenz for impressing this point on me.88 However, it is worth noting that the Pythagorean tradition on which the Timaeus draws already

contained the concept of ‘void’ outside the cosmos: (#'&* DK.

our world is neither endangered by, nor in need of, anything external toitself (ib. &&b–&)a)? It would have been natural to read Plato not only asattaching some kind of causal power to the central place in our world,but also as making that world itself an occupant of infinite space. Andthere is in fact independent evidence, once more thanks to Cicero, thatthe existence of extra-cosmic space came to be formally acknowledgedby early Platonist physics. In summarising the cosmology of the earlyAcademy – which I argued in the last section to mean the Academy ofEpicurus’ contemporary Polemo – Cicero attributes to these Platoniststhe tenet that outside our world there is ‘no part of matter and no body’(Academica , %#): it is hard to avoid reading this as indicating that there isextra-cosmic space, empty of all matter.

Perhaps a less Platonic-sounding component of the thesis under attackis (c) the centrifugal motion of air and fire. According to Plato, these twoelements are centrifugal only in the special sense that portions of themare attracted outwards to rejoin their main masses. Although this in itselfmay sound compatible with the position attributed to the opponents, itwould hardly seem to justify Lucretius’ objection, according to whichthese same elements would carry on into inter-cosmic space and lead tothe world’s dissolution. But Lucretius’ objection here is one which has inany case caused puzzlement,89 since for some reason he protests that onthe opponents’ view even the heavy elements earth and water would belikely to disperse, leading to the total disintegration of the cosmos. Canwe solve both problems together? The beginning of his objection is lostin a lacuna (!*")–!!*!), resuming at !!*%–) with the words

ne volucri ritu flammarum moenia mundidi/ugiant subito magnum per inane soluta,et ne cetera consimili ratione sequantur . . .

Perhaps, then, the sense of the lacuna should be filled out as follows:

[. . . They say that air and fire travel away from the centrebecause each element is attracted to its own kind. But if this wereso, there would be nothing to hold the four cosmic masses to eachother, in which case there would be a danger] that the walls of theworld will fly apart like winged flames through the great void,and that in a similar way the other parts will follow . . .

Who the geocentrist opponents are is not answerable with certainty,but there is a very reasonable chance that they are Platonists, working in

&. Geocentrism #!

89 See discussion in Furley (!"++), p. !".

the Academy in the late fourth century, and developing the cosmolog-ical picture sketched by Plato in the Timaeus. There may, for all I know,be further possibilities.90 What seems relatively safe to say is that they arenot the Stoics.

$. 18?56;48737D 9 58,1,9;

Quite apart from the di0culties of demonstrating that Lucretius everhas the Stoics in mind as a target, there is a much greater problem facinganyone who wishes to see him as responding to the contemporary Stoicchallenge. Posidonius was not only by far the most influential Stoicamong Romans in Lucretius’ own day, he was also, arguably, the mostprominent cosmologist that the Stoic movement ever produced. LikeLucretius, he made it his leading ambition to understand the causalstructure of the world. His work, some of it based on his own empiricalresearch, is undoubtedly reflected in the eloquent speech made byCicero’s Stoic spokesman Balbus (De natura deorum ,,) in defence of theprovidentially structured Stoic world and, at the same time, in ridiculeof the Epicureans’ appeal to mere accident as the cause of cosmogony.Even if we discount Lucretius’ total failure to allow for the vital contribu-tions made to cosmology since the time of Epicurus by mathematicalastronomers, how are we to explain his equal lack of concern to counterthe contemporary challenge posed by Posidonius?91 Let me select justtwo examples.

First, Posidonius followed a tradition, inaugurated in the fourthcentury '1 by Eudoxus, of demonstrating the perfect mathematicalregularity of the celestial orbits, aided by the construction of mechan-ical devices reproducing those orbits. Balbus regards such devices aspowerful support for the Stoic rationally governed cosmos (Cicero, Denatura deorum ,, ##):

Suppose someone were to bring to Scythia or Britain the armillary sphererecently built by our friend Posidonius, which revolution by revolution bringsabout in the sun, the moon and the five planets e/ects identical to those broughtabout day by day and night by night in the heavens. Who in those foreign landswould doubt that that sphere was a product of reason? And yet these people

#% #. Lucretius the fundamentalist

90 If Epicurus was here following Theophrastus, Phys. op., the findings of Ch. + below will open thepossibility that Theophrastus has himself here synthesised a theory representing the ideas of twoor more philosophers (e.g. the Pythagoreans, Anaximander, Plato).

91 For an appropriately sceptical response to the suggestion that Lucretius borrowed material inbooks - and -, from Posidonius, see Clay (!"#&), p. %+.

[the Epicureans] hesitate as to whether the world, from which all things comeinto being, is itself the product of some kind of accident or necessity or of adivine mind’s reason. And they rate Archimedes’ achievement in imitating therevolutions of the heavenly sphere higher than nature’s in creating them – andthat when the original is a vastly more brilliant creation than the copy.

Indeed, we possess one such device, possibly contemporary withPosidonius and Lucretius, in the Antikythera mechanism, fished up fromthe Aegean sea-bed in !"**.92 Now Epicurus had himself already beenfamiliar with similar mechanisms, constructed by the mathematicalschool which Eudoxus had founded at Cyzicus, and had made a pointof attacking the observational principles on which they were based.93 Wewill see in Chapter ( (item (xxix) on Chart %, p. !&+ below) that Lucretius,when drawing his own material for book - from Epicurus, On nature .,,chose to omit the critique of these machines. A perfectly reasonabledecision, no doubt, on aesthetic grounds (although there is plenty ofheavy-duty technicality which he does retain in this part of his book), buthardly a sign of concern to combat the contemporary Stoic challenge.

Second, and closely related to this, is the debate about the sizes of thesun and the other heavenly bodies. Epicurus, notoriously, had main-tained that these objects are as small as they appear.94 The main purposeserved by this widely ridiculed thesis seems to have been to exempt theheavenly bodies from the normal laws of optics, and thus to invalidatethe astronomers’ attempts to track and measure their orbits.95 The ideo-logical importance of this motive probably accounts for Epicurus’ deci-sion to state and defend the doctrine right at the beginning of his ownaccount of celestial motion. As we will learn in Chapter (, Lucretius, ina rare departure from the order established by Epicurus, moved the topicto a less emphatic position, late on in his own digest of Epicurean astron-omy.96 This tends to confirm the impression that he retained none ofEpicurus’ concern for combating mathematical astronomy, even in theface of the Posidonian challenge.

And the point can be strengthened. Lucretius’ defence of the thesisthat the heavenly bodies are as small as they look appears to be drawnfrom a set of optical arguments already sketched by Epicurus, based ontwo analogies with terrestrial objects: (a) terrestrial objects whose colourand outline are still sharp (as those of the moon are) are always near

'. Contemporary Stoicism #&

92 See the brilliant reconstruction by Price (!"$().93 See Ch. ), p. !!">%*, !&* below, and Sedley (!"$+b).94 Sedley (!"$+b), pp. )#–(&; Barnes (!"#"b). 95 Sedley ib.96 On the possible reasons for this transposition, see n. &(, p. !(& below.

enough for us to see their true size;97 (b) when terrestrial fires are seen ata distance (presumably he means especially at night) their size barelyseems to change as the distance increases.98

But the debate had moved on since Epicurus’ time. The Stoics of thelate second century '1 reported in Philodemus, On signs99 had arguedenergetically against the Epicurean thesis, alleging that it violated theEpicureans’ own reliance on analogy with familiar objects. For example,they constructed an argument mimicking the sign-inferences theEpicureans themselves now favoured:100

Familiar objects which appear slowly from behind things which screen them doso either because they are moving slowly or because they are very large.Therefore the sun too, since it appears slowly, must have one of these tworeasons applying to it. It presumably does not move slowly, since it completesthe journey from east to west in twelve hours, a huge distance. [Therefore it isvery large.]

The response of Philodemus’ teacher Zeno had been to de-emphasisethe analogy with terrestrial fires, and to focus instead on defending thepossibility that in this respect the heavenly bodies are unique. He wasable to point out that such uniqueness could itself be defended analog-ically, by a comparison with other unique items such as the magnet.101

In Lucretius’ own day Posidonius had continued the onslaught againstthe Epicureans in a monograph devoted to the topic,102 bequeathingmuch material for the later Stoic cosmologist Cleomedes to use in hisown attack on the same Epicurean position.103

Lucretius appears blissfully unaware of the entire debate ragingaround him. At no point is his defence of the Epicurean tenet adjustedto resist any known element in the Stoic critique, and he makes no useof the new generation of Epicurean counter-arguments preserved by

#) #. Lucretius the fundamentalist

197 - ($(–#); Epicurus, On nature .,, quoted in scholion on Ep. Pyth. "!.198 Lucr. - (#(–"!; Epicurus, Ep. Pyth. "!.199 For the identification of the opponents as Stoics, see Sedley (!"#%); the alternative thesis that

they are Academics is defended by Asmis (!""+). 100 Philodemus, Sign. ,. &#–. !+.101 Ib. ,. !#–&#. 102 Posidonius @#, !#, !", !!)–!+ Edelstein/Kidd.103 Cleomedes, De motu circulari ,, !. For a helpful survey of the debate, see Romeo (!"$"). Her article

is an edition of P.Herc. !*!&, a work conjecturally attributed to Demetrius of Laconia, the frag-ments of which argue the Epicurean side in this very same debate, and which she takes to be areply to Posidonius. This last point may be right (see, however, Barnes (!"#"b), p. &) n. &*), butshe is clearly wrong to take the whole book to have been on the size of the sun. Its closing columnmakes it explicit that it was a defence of the reliability of the senses against their critics, andthere is therefore no reason why the debate about the sun should have been anything more thanone in a series of examples discussed. (This weakens her argument on p. !#, derived fromPhilippson, for dating the book after that of Posidonius.)

Philodemus.104 That would be most surprising either if he had had aserious interest in combating contemporary Stoic cosmology, or if, asoften assumed, he had enjoyed significant philosophical interaction withthe Epicurean school of Philodemus.

# . 91645,1,9;

We have already met in Chapter % (p. )+ above) Lucretius’ celebratedrebuttal of scepticism as self-refuting (,- )+"–$%):

If someone thinks that nothing is known, he doesn’t even know whether that canbe known, since he admits that he knows nothing. I therefore decline to arguemy case against this person who has stood with his own head pressed into hisfootprints.

It used to be widely assumed that this argument is directly derived fromEpicurus. I shall argue in favour of the assumption, and hence againstthe recently more favoured view that it represents a later phase ofEpicureanism, postdating Epicurus himself.105

Much debate has been expended on the question whether we are heredealing with ‘Socratic’ scepticism, in which one claims to know that oneknows nothing,106 or with the more refined scepticism commonly asso-ciated with Metrodorus of Chios (mid fourth century '1) and Arcesilaus(Epicurus’ younger contemporary), according to which we know literallynothing, not even the very fact that we know nothing.107 For convenienceI will call these, respectively, ‘non-reflexive’ and ‘reflexive’ scepticism.

Actually neither kind seems at all likely to be the target of Lucretius’criticism. His argument assumes, on the contrary, that the sceptic has notstated any position on whether he does or does not know the truth of hissceptical claim. The inference that he must admit to not knowing

(. Scepticism #(

104 Kleve (!"$#), p. +#, suggests that ,, (%%–+# is Lucretius’ ‘counterpart’ to the Epicurean debatewith the Stoics about uniqueness recorded in Philodemus, On signs. But he does not succeed inshowing any significant resemblance. 105 Notably Vander Waerdt (!"#"), Schrijvers (!""%).

106 ‘Socratic’ scepticism is favoured as Lucretius’ target by Schrijvers (!""%), p. !%#, who identifiesit with the scepticism of the New Academy. This is a slightly misleading conflation. The imageof the New Academy from the first century '1 onwards was as holding +& ,+-+73:$"+ (‘unknowa-bility’) as a dogma, but not as claiming to know it. Hence the best that Antiochus could do to findan inconsistency in their position was to complain that they ought to claim to know it (Cic. Ac. ,,%").

107 $*'! DK. On Metrodorus’ scepticism, see Brunschwig (!""+). Metrodorus has been named asthe target by Bailey (!")$) and Ernout/Robin (!"+%), ad loc., and by Burnyeat (!"$#). The pro-posal is rejected by Boyancé (!"+&), p. !") n. !, Vander Waerdt (!"#"), and Schrijvers (!""%), pp.!%$–#. (In Sedley (!"#&), p. &&, I argue the target to be the entire fourth-century tradition ofDemocritean scepticism, including Metrodorus, but without explicit reference to his ‘reflexive’version of scepticism – which I shall argue below to be largely a later retrojection.)

whether it is true is the Epicurean’s own triumphant move, which we arenot meant to understand as having been anticipated by the sceptic.Hence this self-refutation argument is most likely to be one which wasconstructed before the reflexive and non-reflexive versions of scepticismhad been explicitly formulated and di/erentiated.

Consequently, there is not the slightest reason why Lucretius’ self-refutation argument should not have been first devised by Epicurus,108

and at least some reason why it should not postdate him. It is quitedoubtful whether what we think of as either the ‘Socratic’ or the‘Metrodoran’ version of scepticism had actually been formulated bythe last decade of the fourth century, when Epicurus was writing thebooks of his On nature which I shall argue to be Lucretius’ source text(see Ch. (). ‘I know that I know nothing’ is not attributed to Socratesby Plato, Xenophon or Aristotle. This non-reflexive sceptical dictum isfirst known to have been fathered on Socrates by Arcesilaus, i.e. wellinto the third century '1, in the same breath in which Arcesilaushimself opted for the reflexive alternative.109 As for Metrodorus, heprobably did not write ‘We know nothing, not even the actual fact thatwe know nothing’, but rather, as Cicero reports him, ‘I say that we donot know whether we know anything or not . . .’110 – a version whichdoes not even appear to be fully sceptical, let alone vulnerable to theself-refutation argument.

On the other hand, the simple unrefined denial that anything isknown was endemic in the atomist tradition to which Epicurus hadaligned himself, from Democritus onwards.111 Democritus himself,whatever his final position may have been on the question, had forexample written: ‘This argument too shows that in reality we know

#+ #. Lucretius the fundamentalist

108 The main objection of Vander Waerdt (!"#"), founded on the lack of explicit evidence thatEpicurus did formulate this argument, carries very little weight, given that considerably less thanone per cent of his writings has survived. Vander Waerdt himself is not proposing any explicitalternative to Epicurus’ authorship, which he rejects in indirect support of his main goal – totransfer from Epicurus to Colotes ownership of the pragmatic argument against scepticismwhich follows in Lucretius’ text. Let me therefore add that he comes up with no positive reasonfor denying Epicurus the latter either.

109 Cic. Ac. , )(: itaque Arcesilas negabat esse quidquam quod sciri posset, ne illud quidem ipsum quod Socratessibi reliquisset. See further, Long (!"##b), pp. !($–", for Arcesilaus’ role in inventing the scepticalSocrates.

110 Cic. Ac. ,, $&!$*'! DK. In favour of the Ciceronian version, see Brunschwig (!""+), and Sedley(!""%a), p. %$, where I also argue that the explicitly reflexive version of his dictum was proba-bly transmitted by Aenesidemus (I do not mean to imply that he was necessarily its inventor, andmy argument below will imply that it is likely to have Hellenistic antecedents).

111 There are even signs of scepticism in Epicurus’ own reviled teacher Nausiphanes: $(') DK!Sen. Ep. ##.)&. I leave Pyrrho out of account, since there is no reason to think the Epicureansever considered him a sceptic: on this see Vander Waerdt (!"#"), pp. &%(–+.

nothing about anything . . .’112 It seems to have been only after Epicurus’intervention that sceptics started to take a firm position on whether theirscepticism was reflexive or not. And it may well have been Epicurus’ self-refutation argument that forced the issue into the open. It was at thatstage, it seems, that Socrates and Metrodorus were, with hindsight,reinterpreted as belonging to one or the other camp.113

Once the issue of reflexivity was in the open, the self-refutation argu-ment lost its bite. If a sceptic has already told you whether or not heclaims to know that he knows nothing, it becomes pointless to draw yourown inferences on the matter. Instead, the main area of probing inHellenistic debate about scepticism was whether the sceptic’s position isinternally coherent or not. Reflexive and non-reflexive scepticism wereeach defended as coherent by some, denounced as incoherent byothers.114 Self-refutation arguments as such by no means go out offashion in this period or after,115 yet the self-refutation argument againstscepticism never puts in a reappearance – except, that is, in Lucretius.There is therefore excellent reason to see Lucretius’ anti-sceptical argu-ment as one dating back to Epicurus, rather than as reflecting thedebates of his own time.

Once we see that the self-refutaton argument is a throwback, echoingEpicurus’ own original contribution to the debate, it becomes hard notto view in the same light the remainder of the anti-sceptical argumentat ,- )+"–(%!. The overall structure is:

S! A self-refutation argument: the sceptic’s thesis undermines his owncommitment to it, in that he cannot claim to know whether it is true()+"–$%).

(. Scepticism #$

112 Democritus '$ DK. I can see no force in the objection of Vander Waerdt (!"#"), pp. %&#–", thatthe challenge of ‘atomist scepticism’ did not require the self-refutation argument as a reply butsimply a revision of the theory of perception. (a) The revision of perceptual theory could onlybe enforced if its alternative, scepticism, had first been shown to be untenable; (b) Democritusdeployed more than one argument for scepticism, as the opening of '$ proves, and the sequelto that opening (+& 77 & "&6$#&'µ$"3 "$ ,+" '-/$'$. 3$ 5/" @$)) is most plausibly interpreted as itselfrelying not on a perceptual premise but on a broader metaphysical one, the identity of all beliefwith physical rearrangement; (c) by Epicurus’ day there were non-atomist versions of scepticismtoo, e.g. that of Gorgias.

113 Once the question of reflexivity was addressed to Socrates, Plato, Apol. %!b)–( must haveseemed quite su0cient to Arcesilaus and others to reveal him as a non-reflexive sceptic.Needless to say, though, that goes well beyond what, in context, he is actually saying there.

114 Reflexive scepticism is coherent according to Arcesilaus and Carneades (Cic. Ac. , )(, ,, #) butincoherent according to Antiochus (ib. ,, %"). Non-reflexive scepticism is coherent according toAntipater but incoherent according to Carneades (ib. ,, %#). For a satisfying reconstruction ofAntipater’s defence of it, see Burnyeat (!""$).

115 This is fully documented and explored in Burnyeat (!"$+).

S'(a) An argument questioning the sceptic’s access to the conceptionsneeded to make his case coherently. By denying himself cognitive accessto conceptual distinctions like ‘true/false’ (notitiam veri . . . falsique, )$+)and ‘certain/uncertain’, the sceptic cannot claim to grasp the dependentnotion of knowledge either, or, therefore, to understand the terms inwhich he formulates his own scepticism ()$&–$).

S'(b) The conception of ‘true’ in fact comes from the senses, whichmust themselves be irrefutable ()$#–"").

S* A pragmatic argument: scepticism is unlivable in practice((**–%!).

This sequence employs a type of argumentative strategy used byEpicurus himself in On nature ..-, in a digression where he criticisesdeterminism:116

D! A self-refutation argument: the determinist’s thesis undermines hisown defence of it, since he must admit that he is compelled to defend it,regardless of its merits (Long/Sedley (!"#$), %*1(–$).

D' An argument based on the conception (prolepsis,!notitia?) involved:the determinist may supply a new name, ‘necessity’, for what we have pre-viously called personal agency, but he fails to change the accompanyingconception (%*1#–!%).

D* A pragmatic argument: if the determinist took his own doctrineseriously, he would find life unlivable (%*1!&–!)).

Details inevitably di/er as between the two arguments, but it is hardto doubt their shared authorship. Furthermore, both play the sameoverall role in the strategies of their respective books. Lucretius firstsets out in book ,- his positive doctrine of vision and its reliability, andonly then rejects the opposing sceptical viewpoint as self-undermining.Likewise Epicurus, after first setting out his own positive account ofhuman agency as autonomous, goes on, in what he explicitly labels a‘digression’, to expose the self-defeating character of its alternative,determinism. These considerations, added to the preceding evidencethat at least S! must be a throwback to Epicurus, seem to me morethan enough to establish that Lucretius’ entire anti-sceptical digression

## #. Lucretius the fundamentalist

116 I argue this fully in Sedley (!"#&), pp. !#–&!. The text and translation can be found there, or,more accessibly, in Long/Sedley (!"#$), %*1, to whose section-numbering I refer for conve-nience.

at ,- )+"–(%! originated with Epicurus as a single integrated argu-ment.117

I have saved until last the strongest case known to me for allowingLucretius a philosophical source more recent than Epicurus. In an out-standing article,118 P. H. Schrijvers has demonstrated the large amountof Lucretian material on illusions, especially in the section of book ,-leading up to his critique of scepticism, that is also to be found in the tenPyrrhonist Tropes, those massive compilations of evidence for per-ceptual relativity on which Aenesidemus’ revival of Pyrrhonism wasfounded. Given that both Aenesidemus and Lucretius were working inthe first century '1, it may seem very reasonable of Schrijvers to posit adirect borrowing from the former by the latter.

However, the picture is not quite so simple. For one thing,Aenesidemus’ dates are uncertain, and it has recently been argued thathis revival of Pyrrhonism occurred in the )*s '1, well after Lucretius’death.119 That dating is controversial,120 but the fact remains that evenCicero, writing in )(, has not heard of this Pyrrhonist revival.121 Thatshifts the onus of proof onto anyone who wants Lucretius, ten or moreyears earlier, to be in direct touch with it. But no such hypothesis is nec-essary. As Schrijvers acknowledges, there is a good deal of material inthe Tropes which we know to have been traditional, some of it evengoing back as far as Heraclitus. While it is perfectly true that many ofthe examples common to Lucretius and the Tropes cannot be traced

(. Scepticism #"

117 See Vander Waerdt (!"#") pp. %&"–)% (also nn. !*#, !!% above (pp. #+, #$)) for opposition to mythesis. I have dealt with most of his individual points in what precedes, and I would add that hisconsideration (a) in note )+, ‘the targets of these passages are very di/erent, the former beingdirected against the ethical determinist and the latter against the skeptic’, is an ignoratio elenchi,since I only claim structural analogy between the two arguments, not identity. Lévy (!""$)implicity accepts that s! and s* may well go back to Epicurus, but argues that s' reflects first-century debate between Antiochus and the New Academy, comparing Cic. Ac. ,, %". I agree thatthere are interesting analogies between Lucretius’ argument, concerning the sceptic’s non-access to the conceptions of true and false, and that of Antiochus, about the sceptic’s lack of acriterion of true and false; but they are not nearly enough to show that Lucretius’ argumentcannot predate Antiochus’. Its absence from Colotes’ third-century polemical work reported inPlut. Adv. Col., pointed out by Lévy, proves nothing. Colotes’ work, as indicated by its title Theimpossibility of even living according to the doctrines of other philosophers, as a matter of policy invokedexclusively the pragmatic arguments against scepticism: its silence about non-pragmatic argu-ments can therefore tell us nothing about their unavailability to him. 118 Schrijvers (!""%).

119 Caizzi (!""%).120 The reply by Mansfeld (!""() does not directly address the question of date, but its conclusions

implicitly threaten Caizzi’s late dating of Aenesidemus. For further replies which do questionher dating, see Görler (!"")), pp. "#&–), Lévy (!""$). 121 Cic. Tusc. - #(.

back to the fourth century or earlier, there is no reason in principle whythey should not already have been in the tradition early enough forEpicurus to have borrowed them,122 or even, in some cases, to have intro-duced them.123

If the preceding run of my argument had lent any credence to theidea that Lucretius had his finger on the philosophical pulse of his ownday, it would be easy to adopt Schrijvers’ explanation as the mosteconomical. But if my general argument is right, for Lucretius to haveborrowed directly from Aenesidemus would be so contrary to his usualmodus operandi that to assume that he did so, when our hand is not forced,would be in reality a false economy.

" . '6789A9

If the question is posed, who (apart from Epicurus) is the latest philoso-pher or scientist to whom Lucretius unambiguously refers, the answermust be Berosus, the Babylonian historian, astrologer and astronomer.The idiosyncratic theory about the moon which Lucretius considerssympathetically at - $%*–&* – that it has one bright hemisphere, androtates – is there called ‘the Babylonian doctrine of the Chaldaeans’; andour sources inform us that the Babylonian in question was Berosus.124

Now Berosus’ biography is problematic (it is not even universally agreedthat the astronomer and the historian were the same person). His workthe Babyloniaca was dedicated to Antiochus I Soter, who reigned from %#!to %+%, although he had been co-regent from %"). But this must havebeen late in Berosus’ life, since he is reported to have called himself acontemporary of Alexander the Great. There is no good reason to think

"* #. Lucretius the fundamentalist

122 I do not see why this should not apply equally to the example of the lion fearing the cock (,-$!*–%!). Even if Schrijvers is right (art. cit. p. !%", following Wellmann) that such ‘sympathies andantipathies’ were first anthologised as such by Bolos in the late third century '1, it does notfollow that the example itself ‘est clairement à dater dans une période plus récente du scepti-cisme aussi bien que de l’épicurisme’. Schrijvers also argues (pp. !%"–&*) that the anomalouspositioning of this topic within book ,- is a sign that Lucretius is drawing on his own knowl-edge; against this, in Ch. ( I shall be arguing that anomalous grouping of topics is a pervasivefeature of book ,-, reflecting its unrevised state. For important further negative argumentsregarding Schrijvers’ case, see Lévy (!""$). Lévy’s positive arguments for his own preferred alter-native, that Lucretius drew the material from recent Academic debate, seem to me no bettersupported. His suggestion that the dialectical character of the anti-sceptical arguments is morecharacteristic of late than of early Epicureanism is contradicted by the evidence, discussedabove, of Epicurus’ own anti-determinist argument, and even Cic. Fin. , &!, which he cites inhis support (n. %#), goes back to Epicurus’ own arguments in my view (Sedley (!""+), pp. &%&/.).

123 For evidence that the illusion described at Lucr. ,- )*)–!&, that the sun is setting on a nearbyhill, is Epicurus’ own contribution, reported from personal experience, see p. !&* below.

124 Aetius ,, %(.!%, %#.!, %".%; Cleomedes ,, ).

that his astronomical theories were to be found in his historical treatiseBabyloniaca,125 or, even supposing that they were, that he had never airedthem before writing this very late work.126 There is, in short, no obstacleto assuming that his theory about the moon could have been familiar toEpicurus at the end of the fourth century, or even for that matter toreaders a quarter of a century earlier.

But that is where the door closes. If one looks in Lucretius’ poem forcomparably clear references to the astronomical or other scientific the-ories of the third century and after – such as those of Aristarchus (whoseheliocentric theory was a talking-point among the third-century '1Stoics), Eratosthenes, Archimedes and Hipparchus – one looks invain.127 And the same applies to the philosophical innovations of thesame period.

!* . @A?<3;6?53:,9;

The position that I am defending can at this stage a/ord a little relaxa-tion. Even if it were in the future to turn out, in one or two isolated cases,that Lucretius was aware of some philosophical or scientific develop-ment which demonstrably postdated Epicurus, that would not alter thegeneral picture. Some such snippet of information could have reachedhim through his non-philosophical reading, through a chanceencounter, or even through marginal scholia in the copy of Epicuruswhich he consulted.128 It would still be his lack of such contact, and hispreference for the unmediated arguments of Epicurus, that pre-dominantly characterised the content of his poem.

Thirty years ago, Furley’s own attempted explanation of the phe-nomenon which he had discovered – Lucretius’ lack of interest inStoicism – was that Stoic physics was little more than a down-marketversion of Aristotelian physics, too unimportant to occupy Lucretius’time. Such an explanation would be unlikely to be put forward today,when the power, originality and contemporary influence of Stoic physicsare widely acknowledged, while it is Aristotle’s importance to Hellenistic

!+. Fundamentalism "!

125 On this see Kuhrt (!"#$), esp. pp. &+–)); for Berosus’ life and work, see also the article by J.Campos Daroca, s.v. in Goulet (!"#"– ), ,, "(–!*).

126 The argument of Runia (!""$), pp. !*!>%, that Berosus’ theory is too late to have been reflectedin Nat. ., rests on the supposition that it first appeared in the Babyloniaca.

127 On this see Furley (!"$#), pp. )–(.128 There are marginal scholia in the papyrus of book .,- of Epicurus, On nature (on which see

Leone (!"#))). There are also scholia incorporated into the versions of Epicurus’ shorter writ-ings that have been preserved in DL ..

philosophers that has become increasingly controversial.129 Neverthe-less, I believe that the conclusion which Furley’s explanation was tryingto make sense of is itself correct, and is of the utmost importance for theunderstanding of Lucretius’ intellectual position. Once again it revealsLucretius as an Epicurean fundamentalist, uninterested in pursuing thehistory of philosophy beyond the zenith it had reached when theEpicurean gospels were written.

It is important to emphasise that my claims in this chapter in no wayexclude the possibility, indeed the likelihood, that those of Lucretius’Roman readers who were better acquainted than he was with Stoicismconstrued his poem as having anti-Stoic dimensions – much as, in latercenturies, many were to read it as anti-Christian. My thesis is one aboutthe motivating forces of Lucretius’ project as he himself conceived,researched and realised it. Contemporary philosophical and scientificdebate played no part in this, and we seriously misconceive the charac-ter of his enterprise if we assume otherwise.130

Nor should this kind of fundamentalism, with its accompanying deaf-ness towards contemporary opponents, be thought historically incred-ible, as it very often is.131 One partial parallel can be found in the Scepticwriter Sextus Empiricus. Although himself almost certainly working inthe second century 3<,132 Sextus in his philosophical critiques shows vir-tually no interest in, or even knowledge of, contemporary developmentsand concerns, which were in fact dominated by Peripatetics andPlatonists. None of his philosophical targets, either named or unnamed,can plausibly be dated later than the first century '1. Significantly, thefirst century '1 was when his school’s foundational literature had beenlargely written, by its re-founder Aenesidemus. Evidently for Sextus, asfor Lucretius, the list of his school’s philosophical targets had been

"% #. Lucretius the fundamentalist

129 The rehabilitation of Stoic physics was largely the work of Sambursky (!"("), while the newchallenge to Aristotle’s Hellenistic influence has come from Sandbach (!"#().

130 It has been impossible to deal here with all the innumerable studies of Lucretius that havedetected Stoic allusions. In many of those with which I am acquainted attention is drawn tofeatures which I would happily agree might have an at least partly Stoic reference if it werealready established that Lucretius is conscious of the Stoic challenge, but which do not them-selves constitute evidence for such a consciousness. Long (!""$) is an outstanding example ofsuch a study.

131 E.g. Kleve (!"$#), although his even-handed assessment of the evidence does not point firmlyeither way, surprisingly concludes (p. $!) that it is ‘improbable that Lucretius should have felthimself bound exclusively to present the polemics from the time of Epicurus’, and that ‘wecan confidently use Lucretius as a source for the general philosophical situation of his owntime’.

132 See House (!"#*), who allows, however, that Sextus might even be dated to the early thirdcentury.

definitively drawn up, and did not need extending. Plenty of furtherexamples of this same phenomenon could be added.133

If this should still sound historically implausible, some modern analo-gies may help. Think of Christian preachers who see no need to informthemselves about Islam or any other post-Christian faith. Think of theMarxist who is content to remain altogether unversed in post-Marxianpolitical theory. Think even of traditional literary critics who arecontent, and may actually prefer, to know nothing of contemporary crit-ical theories. Ignorance of the current alternatives may signal no morethan total confidence in one’s own chosen ideology.

That Lucretius, unlike Philodemus and his associates, never consultedEpicurean texts other than those by Epicurus himself – either thosewritten by the other three authoritative founders, or those of the manylater Epicurean writers of whom we know – is of course incapable ofdemonstration. But I believe that such a hypothesis has every chance ofbeing correct. It is strongly supported by his formal declaration ofloyalty to Epicurus alone, and by his imperviousness to later develop-ments inside and outside the school. Even more than that, it makes thebest sense of his poem’s contents. As I shall argue in the next two chap-ters, his sole philosophical source and inspiration from early in book ,until late in book -, is Epicurus’ great physical treatise, On nature.

!+. Fundamentalism "&

133 Although Diogenes of Oenoanda (second cent. 3<) does, as we have seen (p. $) above), knowabout the Stoics, he shows no acquaintance with the Platonism and Peripateticism which by hisday were even more dominant: cf. his notorious mistake about the Peripatetics in fr. ( Smith.Cleomedes (datable between the second and fourth centuries 3<) is still locked into Posidonius’debate with the Epicureans, and shows little if any awareness of more contemporary issues, oreven of Hellenistic astronomy (cf. Goulet (!"#*), p. (: ‘Le système planétaire de Cléomèdeignore apparemment les développements de l’astronomie grecque depuis le IIIe siècle avant J.C.et répète pour l’essentiel la cosmologie du stoïcisme ancien’). Or again, take Calcidius, of whomDillon (!"$$), pp. )*!–#, shows that although he probably wrote in the fourth century 3< he isphilosophically living in the second.

1234567 )

Epicurus, On nature

! . 526 <,918-67D

The cataclysmic eruption of Vesuvius in 3< $" buried the town ofHerculaneum in a torrent of boiling mud. Streams of lava from sub-sequent eruptions increased its depth to some hundred feet below thesurface. Not surprisingly, the teams of excavators assigned to the sitefrom !$*" onwards by the Bourbon rulers of Naples did not consideruncovering the buildings from above, but chose instead to plunder theircontents by means of underground tunnels. The most spectacular dis-covery was made during the !$(*s – a vast suburban villa containing anart collection of unrivalled magnificence. The excavation of the villaproceeded, room by room, over a period of many years. Towards theend of the second year, workmen excavating the tablinum began tohappen upon black lumps which they mistook for charcoal. Many theythrew away or took home to kindle their fires, and it was only when thefragments of one which had been dropped were seen to contain writingthat they were recognised as rolls of papyrus.

The tablinum was a pleasant room looking out onto the garden on oneside and onto a portico on the other, with a mosaic floor and, down thecentre, a row of eight bronze busts. The papyrus scrolls were foundstrewn around the room, together with a few wax tablets. But over thenext few years papyri were found in four more rooms of the villa, in oneof which the excavators found some %(* scrolls, and three busts ofEpicurean philosophers. Another room was plainly set out as a library,with &&$ Greek scrolls arranged on shelves along the walls and in aninlaid wooden cupboard placed in the middle, as well as a bundle ofeighteen Latin scrolls in a box.

This was the first discovery of papyri in modern times, and still oneof only two major Greek papyrus finds ever made outside Egypt (theDerveni papyrus being the other).

")

The papyri were brittle, and the action of damp had tended to weldthe layers together. They were, moreover, in a carbonised condition – afact which enabled them to survive centuries of burial, but which alsofurther impeded their legibility. Naturally the outer parts of the rolls,containing the beginnings of the books, had su/ered most, while thecentres were sometimes comparatively undamaged. The first reactionwas to separate the layers by splitting the rolls along their length; but thisturned out to reveal only isolated groups of letters, and certainly nointelligible text. Hundreds were destroyed in this way.

The next stages of the story can best be told by quoting a letter froma ‘learned Gentleman of Naples’ to Monsignor Cerati, of Pisa, dated %(February !$((:1

At length signor Assemani . . . proposed to the king to send for one fatherAntonio, a writer at the Vatican, as the only man in the world, who could under-take this di0cult a/air.

This Father Antonio was Antonio Piaggio, who was to become the mostimportant figure in the early history of the Herculaneum papyri. He wassummoned, and in early !$() he set to work on one of the worst pre-served rolls, using a technique which was not to be superseded for thenext two centuries:

It is incredible to imagine what this man contrived and executed. He made amachine, with which (by means of certain threads, which being gummed, stuckto the back part of the papyrus, where there was no writing), he begins, bydegrees, to pull, while with a sort of engraver’s instrument he loosens one leaffrom the other (which is the most di0cult part of all), and then makes a sort oflining to the back of the papyrus, with exceeding thin leaves of onion (if I mistakenot [he did – it was the strips of beaten oxgut known as ‘goldbeaters’ skin’]), andwith some spirituous liquor, with which he wets the papyrus, by little and little heunfolds it. All this labour cannot be well comprehended without seeing. Withpatience superior to what a man can imagine, this good father has unrolled apretty large piece of papyrus, the worst preserved, by way of trial . . . The worstis, the work takes up so much time, that a small quantity of writing requires fiveor six days to unroll, so that a whole year is already consumed about half this roll.

What Piaggio had done was to construct a case in which the roll couldbe hung up by silk threads attached to its outer edge, its main bulk restingon an adjustable platform. Through gradual tightening of the threadsover a period of months, even years, the papyrus was allowed to unroll

!. The discovery "(

1 From Comparetti/De Petra (!##&), pp. %)(–+, quoting an English translation by John Locke ofthe letter, published in Philosophical Transactions )" (London !$(+).

under its own weight, while the layers were gently eased apart with avariety of instruments and the verso of the papyrus was strengthened bygluing on a backing of goldbeaters’ skin. Whenever a sizeable section –say two feet in length – had been unrolled, it could be cut o/ and gluedto a sheet of card.

As Piaggio’s work proceeded with agonising slowness, excitementmounted in the learned circles of Europe. There were expectations of asecond Renaissance. Surely the lost poetry of Sappho, or some exquis-ite play by Menander, was about to come to light. These expectationsturned to baCement and disappointment when rumours started toemerge from Piaggio’s workshop at Portici that the first columns deci-phered contained a Greek prose text in which harmful e/ects of musicwere criticised. A flattering mention of ‘Zeno’ led to reports that theauthor was a Stoic. But when after nearly two years the unrolling wascompleted, the title given at the end turned out to be ‘Philodemus, Onmusic’, and this Philodemus was soon identified as the Epicurean philoso-pher of that name, already known as the author of some attractive epi-grams. The Zeno whom he praised was not the Stoic Zeno of Citium,but the Epicurean Zeno of Sidon, whose bust was among those found inthe villa’s library. It has since come to be generally accepted that thelibrary was that of Philodemus’ school (see Ch. &, §%).

The next papyri opened proved to contain other works of Philodemus– on rhetoric, virtues and vices, and similar topics. Scholarly impatiencemounted. It was even proposed, by no lesser a luminary thanWinckelmann, that if any roll being opened appeared to have similarcontents it should be immediately abandoned.

When Piaggio died in !$"+, forty-four years of work had resulted inthe opening of just eighteen rolls. In !#*% the Prince of Wales (laterGeorge IV) o/ered to finance the unrolling and transcribing of thepapyri, and sent his chaplain John Hayter to organise the work at Portici.Hayter employed thirteen local draughtsmen who had already for sometime been engaged in this work. Many Piaggio machines were built, andHayter further speeded up the work by paying his men according toresults, o/ering the sum of one carlino for every line of text unrolled andcopied in facsimile. The e/ects of the system were only partiallybeneficial. Haste in copying, by draughtsmen ignorant of Greek, mighthave been expected to introduce errors, but thanks to Hayter’s attentivesupervision the standard of copying was in general very high. On theother hand, there seems little doubt that papyri were often unrolled tooquickly. Piaggio’s first attempt had been on a badly preserved roll, but by

"+ %. Epicurus, On nature

spending nearly two years on its unrolling he had succeeded in saving aremarkable proportion of the writing. Under Hayter it became commonto spend only months, or even weeks, on each roll, because his methodof payment undoubtedly made it more lucrative to abandon those por-tions whose fragility or adherence to other layers of papyrus demandedspecial skill and patience on the part of the unroller, and to save onlythose which could be quickly separated.

Hayter stayed for four years at Portici, until February !#*+, when theimpending French invasion of Naples drove him to take refuge inPalermo along with the Bourbons. He brought his facsimile copies backto Oxford, but most of the papyri remained in Naples, where work onthem continued.

Since the beginning of the scroll was invariably lost or illegible, it wasonly when the unrollers reached the end that they could hope to iden-tify the text they had been unravelling. (Fortunately it was standardancient scribal practice to repeat the name of the book at the end.)During Hayter’s four years, a substantial number of the papyri unrolledturned out to contain at the end the title

E!80?FG?F!EG8 HFIEJI

sometimes followed by a book number (,,, .,, .,-, .-, ..-,,,), by anindication of the book’s length, and in two cases even by a date ofcomposition. Later discoveries have identified a few extra pieces of thesesame rolls, a further copy of book .,, the number of one further book(..-), and one further date. But the golden age for the recovery ofEpicurus, On nature was undoubtedly Hayter’s four years.

So began the modern partial rediscovery of Epicurus’ great physicstreatise, On nature. So formidable are the problems of deciphering theseravaged papyri that even today, after almost two centuries, we do nothave adequate published editions of all of them.2 Nevertheless, thanks

!. The discovery "$

2 There are modern editions, based on a proper autopsy of the papyri with binocular microscope,of books .,- (Leone (!"#)), .- (Millot (!"$$)), and ..-,,, (Sedley (!"$&)). We are near to havingthe same for book ..-: so far there are Laursen (!"#$), (!"##), (!""%) and (!""(), and further partsof the book in Sedley (!"#&) and Long/Sedley (!"#$), §%*. Still awaited are adequate editions ofbooks ,, and ., (for which we have Vogliano (!"(&) and (!")*) respectively), each of which con-fronts editors with the irritating obstacle that some of the papyrus material is in the BritishMuseum, the remainder in Naples. In the mean time, Arrighetti (!"$&) o/ers a comprehensivecollection of the Nat. papyri, based on what were at the time the best available editions plus, insome cases, Arrighetti’s own further readings. Wherever possible I shall include an Arrighettireference (abbreviated as Arr.2) to each passage, placing it in square brackets when the currenttext di/ers significantly from his.

especially to the studies of Achille Vogliano between the wars, and to aresurgence of work on the Herculaneum papyri since the early !"$*s,3we know a vast amount more about them than we did. Virtually none ofthese gains has filtered through into Lucretian studies.

% . 526 434D7,

The library produced remnants of the following books from On nature:4Book ,,: % copies (P.Herc. !!)"/""&; P.Herc. !*!*)Book .,: % copies (P.Herc. !*)%; P.Herc. !())Book .,-: ! copy (P.Herc. !!)#)Book .-: ! copy (P.Herc. !!(!)Book ..-: & copies (P.Herc. )!"/!+&)/+"$; P.Herc. !)%*/!*(+; P.Herc.

!!"!)Book ..-,,,: ! copy (P.Herc. !)$"/!)!$)) further unidentified books (P.Herc. !)&!; P.Herc. "#"; P.Herc. !)!&;

P.Herc. &+%)One remarkable feature of these is that virtually all of them appear,

on palaeographical grounds, already to have been of great antiquitywhen they were buried by Vesuvius in 3< $".5 Compared to the papyriof Philodemus, they form a group which is distinguished by the earlydate of its scribal hands and orthography, nearly all of which can beplausibly located between the beginning and the end of the secondcentury '1. Moreover, three of the scribal hands recur.6 A single copyistis responsible for one of the two copies of book ,, (P.Herc. ""&/!!)"), oneof the three copies of book ..- (P.Herc. !!"!), and the sole copy of book..-,,,.7 Another scribe wrote a further copy of book ..- (P.Herc.!)%*/!*(+) and P.Herc. !*&", a text which is too badly damaged to have

"# %. Epicurus, On nature

3 This has followed from the creation of the Centro Internazionale per lo Studio dei PapiriErcolanesi on the initiative of Marcello Gigante, to whom we also owe the inauguration of thejournal Cronache Ercolanesi and the series of editions La scuola di Epicuro.

4 The main catalogue, with a physical description of the papyri and comprehensive bibliography,is Gigante (!"$"), supplemented by Capasso (!"#"). A multiple number (e.g. !)"/""&) indicates abroken roll whose pieces were first catalogued separately, then reunited.

5 Although the chronology of the Herculaneum papyri is a subject still in its infancy, the pioneer-ing work of Guglielmo Cavallo has, I think, put this conclusion beyond serious doubt. See Cavallo(!"#&), esp. pp. %#–($. As regards age, the one clear exception among the On nature papyri isP.Herc. &+%, which Cavallo implies may even have been written after Philodemus’ lifetime (ib. pp.((–+). 6 Cavallo (!"#&), pp. ))–+.

7 Cavallo adds to this list the unidentified book of Nat. contained in P.Herc. !)&!, but I prefer tosuspend judgement on that, since my notes on the handwriting of this book (taken in !"$!,however, and not checked since) record a number of di/erences from the scribe of P.Herc.!)$"/!)!$.

been identified, but which in the circumstances is likely also to containsome part of On nature. Finally, one and the same scribe wrote the sur-viving copies of both book .,- and book .-.

All this encourages the conclusion that these papyri of On natureformed part of a separate collection, acquired in Greece and brought toItaly by Philodemus. It has been plausibly (though unprovably) suggestedthat he inherited the book collection of his teacher Zeno.8

& . @3-8A7,56 '88E9F

Since only a small fraction of the papyri were rescued and successfullyunrolled, and yet this rich haul emerged, it is natural to assume that thelibrary contained at the very least one complete set of On nature, a workwhich we know to have run to thirty-seven volumes (see §)). Indeed,given the recurrence of the same scribal hands, it is not unlikely thatthere were one or more complete sets each written by a single scribe.9

Yet there is something decidedly odd about the distribution. Whyshould the random process of recovery have given us so high a propor-tion of duplicate or even triplicate sets? Unfortunately the excavatorskept no record of where individual papyri were found: if they had, itmight have turned out that as a matter of library policy all the copies ofeach individual book – rather than all the volumes in a single completeset – were stored together, so that where one survived because fortu-itously protected from damage the others had a better chance of sur-viving along with it.

However, I do not believe that this could be a su0cient explanation.For one thing, it can be no accident that books .,- and .-, both writtenby the same scribe, survived. And, although the assignment of numbersto individual scrolls belongs to Hayter’s time,10 and seems to have beenlargely random, it may be no accident that these two papyri receivedalmost adjacent numbers (!!)# and !!(!). The combined proximity oftheir catalogue numbers and of their book numbers, plus their scribaluniformity, suggests that they survived together and were brought out andstored in the same batch. Even if the copies of book .,- and those of book.- are assumed to have been stored close to each other, so that they sharedthe same accidental protection and one of each survived, is it just coin-cidence that these two surviving copies should both be by the same scribe?

#. Favourite books? ""

28 On this proposal of Vogliano’s, see Dorandi (!""$). 9 Cf. Cavallo (!"#&), p. (#.10 See Obbink (!""+), p. %" n.!.

A likelier explanation of the imbalance is that some books of On naturewere especially treasured and collected. This is supported by the follow-ing fact. The scholia accidentally incorporated into our surviving textsof Epicurus’ Letter to Herodotus and Letter to Pythocles (in book . of DiogenesLaertius) contain several citations of individual books of On nature.11

These are: book , (two citations), book ,, (one citation), book ., (one cita-tion),12 book .,, (two citations), and books .,- and .- (one joint-citation).This list largely coincides with that of books surviving among theHerculaneum papyri, and confirms the impression that these books wereespecially favoured by readers of Epicurus. Since in addition theHerculaneum papyri contain two citations of book .,,,, one of themjointly with book .,,,13 we seem to be dealing with two favoured groupsof books: ,–,, and .,–.-. Clearly also highly valued was book ..-, as isattested by the three copies preserved at Herculaneum: the fact that noscholia refer to it may simply reflect the fact that (as I shall argue in §!%)its subject matter postdates the Letter to Herodotus and the Letter to Pythocles,leaving the scholiasts on those letters no opportunity to cite it.

There is even some reason to suppose that Epicurus had given his ownblessing to this preferential treatment of certain books. His Letter toHerodotus, an epitome of his doctines on physics, opens with the follow-ing admittedly rather di0cult sentence (Ep. Hdt. &():

For those, Herodotus, who are unable to study closely the individual books14 ofour work On nature ("+ ,+'-+ -(' . !"#$! %&" '"() +& .+4"4#+µµ"" .(. 3$ µ$' .), or evento peruse the more important books in the sequence (-+! ) µ"$">/&) -(' . '&.--"-+4µ"" .(. 2$"27/&)), I have prepared an epitome of the entire system("&6$-/µ3! . -3' ) /+ 73) 6#+4µ+-"$"+)) . . .

Although it is not demonstrable that this is an explicit reference to Onnature, rather than simply to ‘what we have written on nature’, there isevery probability that it is. The form of citation closely resembles that

!** %. Epicurus, On nature

11 These are conveniently collected in Epicurus frr. $&–"! Usener.12 Ep. Pyth. "!, where the scholion is followed immediately by Epicurus’ own citation of what he

says ‘in the books On nature’, certainly referring to the same passage in Nat. .,.13 Philodemus in De pietate twice cites book .,,, along with a single citation of book .,,, (frr. #), #$–#

Usener!%$–# Arr.2!Philodemus, Piet. %%(–$, (%&–), !*(*–! Obbink (!""+)): both .,, and .,,,dealt inter alia with the origins of religion. .,, and .,,, are also cited jointly in P.Herc. !!!!, fr.)).!–+ (!!".( Arr.2, fr. )! Usener, but now edited in an improved text with apparatus in Obbink(!""+), pp. &**–!). Unfortunately the papyrus is lost and the putative title !"#$! %&" '"(), if present,is corrupted in the facsimile to !EG8IFK [. . . But the context leaves little doubt that the citationis of Nat.

14 It is quite unproblematic that ‘books’ (i.e. 2$27$"+) should be neuter when, as here, the noun isunspecified, despite becoming feminine (2$"27/&)) in the following clause. Citations of Nat.(including Epicurus’ own at Ep. Pyth. "!) vary between the two forms: cf. frr. $&–#" Usener.

used in book ..-,,, of On nature, where Epicurus clearly has a specifictext in mind: ‘. . . which we discuss in our work On ambiguity’ (. . . +* )7""4[/]µ". "&·. -/[$' ]) !"#$! +& µ[%]$2/7$"+) 3$ µ$'·. +& .+4"4#+µµ"" ./$)).15 Ifthat supposition is granted, then Epicurus is distinguishing theawesome task of studying the entire On nature from the lesser one ofreading its more important books. Implicitly, his readers already knowwhich are the more important books. It may, then, have been Epicurus’explicit blessing that conferred preferential status on certain books inthe sequence.

Even taken as a whole, On nature was by far Epicurus’ most widely dis-cussed work. It heads Diogenes Laertius’ list of his ‘best’ works.16 It waspicked out for special attack by Epicurus’ arch-enemy Timocrates.17 Itwas cited repeatedly, as we shall see.18 There is even reason to thinkthat, such was this entire work’s pre-eminence, Epicureans regularlyreferred to individual books of it simply by their number, withoutneeding to name the work itself: thus ‘book -,’ was understood to mean‘book -, of On nature’. There are several such citations in theHerculaneum papyri, and at least one of them can be shown to referto On nature. This is because we know from Philodemus, On piety19 (%%(/.)that in ‘On nature book .,,’ Epicurus spoke of the origins of religion, yetelsewhere ((%&–)), in a context where On nature is not still under discus-sion, and despite the fact that several works other than On nature havebeen cited in between, Philodemus refers to Epicurus’ attack on theatheists simply ‘in book .,,’, without naming the work. Since the athe-ists in question prominently included those who had attributed theorigins of religion to human invention, there is every reason to supposethat this too refers to the account of the origin of religion in book .,,

#. Favourite books? !*!

15 Nat. ..-,,,, !& - % inf.–-, ! sup. Sedley (!"$&)!&!.!).%+–!(.! Arr.2. It hardly seems likely that thisis a generic reference to a range of writings by Epicurus on ambiguity, and no editor (Voglianoand Arrighetti, as well as myself) has had any hesitation in supplying the capital letter. We haveno independent citation of this work, but our surviving list of Epicurus’ writings (DL . %$–#) isavowedly selective. 16 DL . %$. 17 Ib. $.

18 Obbink (!""+), pp. &*(, )$%, is puzzled at how few citations of Nat. there are in the Herculaneumpapyri outside De pietate, which he remarks contains $ of the !* citations. Actually this should be$ out of !#, if we add the apparent references to it from the Life of Philonides (see below). But inany case, no surprise is called for. We should not expect much reference to it in works on ethicsand aesthetics, which predominate among the Herculaneum papyri. One might have expectedit to be mentioned in Philodemus’ De signis, but there are no citations of any early Epicureanworks there, so its absence is again unremarkable. The !& references to Nat. in the Herculaneumpapyri comfortably beat those to any other work by Epicurus or anyone else. The nearest rivalsare Epicurus’ On rhetoric (+ citations), On lives (( citations), and Symposium () citations) – see Delattre(!""+). (I am discounting citations where the title is purely conjectural.)

19 Text and notes in Obbink (!""+).

of On nature,20 and confirms the impression that citations by booknumber alone were readily understood to refer to this treatise.21

Once this is accepted, a number of further citations in theHerculaneum papyri can be recognised as probably referring to thesame work. The fragments of the biography of the second-century '1Epicurean Philonides include a list of the notes he possessed from hisown studies under a series of teachers. These include notes he had takenfrom one of his teachers on ‘book -,’, and from another ‘on book , andcontinuing to those on book [?]ty-three, with some omissions’.22 This isimportant evidence that On nature was a major focus of exegetical studyin the Epicurean school, yet another sign of its philosophical pre-emi-nence.

Why was On nature so formidable a text to read, yet so highly valued?The two main factors are likely to have been its length and its style. Wewill take these in turn in the next two sections.

) . :6?B52

That there were thirty-seven books of On nature is attested in three separ-ate passages of Diogenes Laertius.23 But how long was each book? Twoof the papyri, those of books .,- and .-, still preserve at the end a sti-chometric indicator. Book .,- contained &,#** lines,24 book .- probably&,%**.25 By the ancient scribal convention these figures will indicate, notthe number of lines of writing in this actual copy, but a standardised

!*% %. Epicurus, On nature

20 See Obbink (!""+), pp. &*+, &(!–%.21 On this see the very full discussion in Obbink (!""+), pp. &*)–+, cf. )$$. One small disagreement:

he assumes too readily that the pattern was always to cite the work by title first, then by booknumber alone thereafter. I do not think that the Life of Philonides passage (below; not taken intoaccount by Obbink) encourages this assumption.

22 Anon.Vita Philonidis col. $; text in Gallo (!"#*), p. +*. The context is uncertain (for a partly di/er-ent interpretation see Gallo ad loc.), but it does at least seem probable to me that either the biog-rapher or, more likely, a contemporary of Philonides whom he is quoting, is telling us aboutPhilonides’ collection of notes, as evidence of his education. Thus: ‘. . . he makes it clear’(5$+'+]%"$' ?) ‘that he has heard [the lectures of . . .?]; while in book form he brings’ (%"" #"$: precisesense unclear, but surely not ‘cites’, as Gallo takes it) ‘the following notes: two early sets, on thelectures of Eudemus and on comments on book -,; about scientific modes of thought; on thelectures of Artemon, starting with comments on book , and continuing to those on book [?]ty-three, with some omissions; and on the classes of Dionysodorus’. The last book number seemsto be corrupt. Gallo reports the papyrus at lines "–!* as 6#/! ) -/! -#$"-[/.] ,+$! 8 /$'[.],/'-/..He emends this last word to <-#$>[+],/'-/" ., to restore the reading which Crönert had origi-nally misreported from the papyrus itself. I find this emendation rather harsh, and suppose it tobe at least as likely that the number underlying the apparent corruption was -#$"-/. ,+$! "$&,/'-/" .,‘%&rd’. 23 DL . $, %$, &*. 24 See Leone (!"#)), pp. %%–&.

25 See Millot (!"$$), p. %+, who alerts us to the possibility of a lost extra digit at the beginning orend of the number.

measure of length whereby one ‘line’ means the equivalent of one hexa-meter line.26 The average length of a hexameter line is thirty-six charac-ters. Therefore, if one takes this measure strictly, the length of book .,-was approximately &,#**"&+!!&+,#** letters, and the length of book.- was approximately &,%**"&+!!!(,%** letters. Hence, although twobooks out of thirty-seven are not as good a sample as one would like, andalthough the length of a one-line unit is in principle capable of a fairdegree of variation,27 the best working hypothesis available to us is thatthe typical length of a book was around !%(,*** letters. At an average ofsix letters per word, and allowing for the occasional blank space (Greektexts contain no spaces between words, except occasionally as punctua-tion marks), this would make a little over %*,*** words.

Then how large a written text were Epicurus’ readers confronted by?As a rough unit of measure, we can take a page from the OxfordClassical Text of Thucydides. An average page of Thucydides containsaround !,)** letters. So a typical book of On nature would have filledsome "* pages of OCT . This is just a little longer than a normal bookof Thucydides, whose book ,, for example, is ## pages long.28

This should give some idea of the monumental scale of Epicurus’magnum opus. We cannot assume that all its books were of equal length,but the book divisions of surviving prose works from antiquity do permitus to hypothesise some degree of uniformity in length. Assuming books.,- and .- to be not untypical, the treatise, if it survived intact, wouldfill nine or ten volumes of Oxford Classical Text. Even books ,–.-,which I shall later be arguing to be Lucretius’ source, would havecomfortably filled four volumes.

%. Length !*&

26 Cavallo (!"#&), pp. %*–%.27 In Philodemus, De pietate (Obbink (!""+), pp. +%–& n. !), and in the second-century '1 logical

papyrus P.Par. % (Cavini et al. (!"#(), p. !&*), the scribe uses marginal marks to divide the text intosmaller sections of %* of his own lines, and bigger sections of %**. Apparently then in both casesthe scribe equated two of his lines with the standard unit of stichometric measure – a veryapproximative device, since in P.Par. % it produces a unit of roughly &+ letters, but in Piet. one ofonly around &* letters. If the scribe of Nat. .,- and .- were following the same practice, his sti-chometric unit would be around )* letters, since his lines average roughly %* letters in length.However, there is no sign that he is keeping his own running total (a practice probably used forcalculating the payment due), and the stichometric totals which he records are likely to be takenover from his archetype. Thus &+ letters per ‘line’ remains the best available hypothesis, but weshould allow for an error of !* per cent or even more.

28 Since first writing this sentence I have learnt (from Obbink (!""+), p. $! n. !) that according toDionysius of Halicarnassus (Thuc. p. &&".!)–!+ U.R.) Thucydides , !–## (a passage running to(! pages in the Oxford Classical Text) was measured at %,*** lines long. This permits the calcula-tion that the whole book was measured at &,)(* lines long – just (* lines less than the average ofNat. .,- and .-. That this fits my calculation quite so closely is no doubt accidental, but it nev-ertheless lends it very welcome confirmation.

( . 9 5D:6

On nature was a lecture course. This is stated explicitly at the end of book..-,,,, where Epicurus announces:29

[$$,]+.(' [)] /&# . 3$ µ$' . 3& 5/7"';3" '1( "&6$! -/&' 6+#/" .-/). ,+$! &$ µ"$' ) [µ]&[#$]+" ,[$)µ.3µ/].".&" "[$]µ 6. ["$#+' ]'1" -+! "&µ/$" -" ,+$! B3-#[/]5(" #($ -(' $5" .["('-$!"$& ]#3µ"".+. /$#µ+$ 5 & &$ µ$' . /% [45/]/. ,+$! ".$&,/'-/! . "$#5/) +& ,[#/]+" '"() -3' [)] "$@3' )6"#+$.[/]µ"".3) -/&-$! .[&' ]. 3& 5/7"['];3' '1+$.

Well, let that be enough chat from us for the time being. And you others, try tenthousand times over to remember what Metrodorus here and I have just said.

I think I have now finished chatting to you this twenty-eighth instalment ofour sequential lecture-course.

If any confirmation is needed that we are dealing with a lecture, thedemonstrative reference to ‘Metrodorus here’ puts the matter beyonddoubt.

How long was each lecture? We have already learnt (§)) that a typicallength may have been around %*,*** words. If we imagine that a lecturein ancient Greek was delivered at even approximately the same rate, inwords per minute, as a lecture in modern English, the delivery of eachmust have taken Epicurus some two and a half hours.

Aristotle’s Physics was a lecture course too, according to its ancient titleH&'$,3! +& ,#/" +'$), but there is little if anything in its content to confirmthat this was so. By contrast, even what little survives of Epicurus’ Onnature confirms its status as the written text of a lecture course. Thesecond person plural, used for addressing his audience, is found not onlyin the above closure, but also in that of book ..-30 (where unfortunatelythe remainder of the sentence is too fragmentary to be of use).

The informality implicit in the verb ‘chat’ (+& 5/7"';"$' .) at the end ofbook ..-,,, may to some extent be explained by the specific content ofthe book. In what survives from its last part, Epicurus is at several pointsconducting a one-way conversation with his colleague Metrodorus, andwhat we have seen to be its closing words suggest that the whole bookhas been cast in this form. Here is a sample from an earlier passage init:31

!*) %. Epicurus, On nature

29 Nat. ..-,,,, !& .,,, +–!* sup. Sedley (!"$&) [!&!.%%.+–!* Arr.2]. The strange termination ofµ.3µ/].".&" "[$]µ reflects the scribe's regular way of modifying final . before a labial.

30 Laursen (!""%), p. !)+, lines )+, )#. For another second person plural earlier in the same book,see P.Herc. !)%*, + ,,, $, in Laursen (!""(), p. "%.

31 Nat. ..-,,, !& ,- ! inf.–- !% sup. Sedley (!"$&)![&!.!&.%&–!).!% Arr.2]. "%;"$µ!"%;"$. (see n. %"above), "&45"@-!"&,5"@-, $&1$'µ"" .(.!"$&1$'µ"" .(..

+& 77+! 4+! # $%'(. ) /&& , "&% ,+$#/" . "&'-[$ -+&' ]-[+] 6#/%""#/.-+ µ3,&""" ."$[.· ,]+$! µ+" 7’/& #. 1.(' ) [4", (# ] B3-#/" 5(#"· 6+" .& 4+! # /$#µ+$" '" 6/77+! +) . "%;"$µ6#/"[.]""4,+'1+$ +* "&1" .(" #"$) 4"7/$"() [6](" [)] -.$[.+]). "&45"@+µ""./&) ,+$!6. [+" .]-[+] µ+' 77/. 3) -/! .//&" µ"./. ,+-+! -+! ) 7""@"$), /&& , "% @( -(' . $&1$'µ"".(.7""@"(. 3$ µ(' . ;#(. µ"".(. /&& 5"! µ"-+-$1"".-(. /& ./" µ+-+ "&6$! -(' µ %+."[#](' ..

But perhaps you’d say this isn’t the right time to prolong the discussion by bring-ing these up. Quite so, Metrodorus, because I’m sure you could bring up lots ofcases of words you’ve seen certain people taking in some ridiculous sense orother, in fact in any sense rather than their actual linguistic meaning, whereasour own usage does not go beyond ordinary language, nor do we change nameswith regard to things which are evident.

Similarly, we will see later (p. !!#) that at least one book of On nature islikely to have taken the form of narrated dialogue.

But ‘chat’ is also a piece of self-deflation on Epicurus’ part, perhapseven a little reminiscent of that ultimate philosophical ‘chatterer’(+& 5/7"';3" )), Socrates. Thus elsewhere in On nature Epicurus describeshimself as ‘babbling’ (1#&7(' ) and even ‘ranting’ (73#(' ).32 Thisinformality brings with it yet another potentially endearing feature, adisarming willingness to admit to mistakes. Book ..-,,, has a good dealto say about past wrong views on language – some held by Metrodorus,some by Epicurus himself, which he now openly recants, and somejointly by the two of them.33 A brief snippet (fragmentary at both ends)which conveys this tone of self-criticism is the following:34

. . . that it is unclear whether we completely forgot this opinion and took thestandpoint opposed to ourselves, as many others have done in many cases, orremembered it but . . .

It is hard to be sure whether the greater visibility of these human touchesin books ..- and ..-,,, represents a change of tone of the later books(written years after the first, see §!%), or simply reflects the fact thatsignificantly longer stretches from these two books survive than from theearlier ones, enabling us the better to glimpse the range of Epicurus’style. What little we have from the earlier books does sound less personal.But even in them it is possible to witness a broader feature of which these

$. Style !*(

32 Nat. ..-, &).&*.+–$ Arr.2, ,+1+" 6"# 6+" 7+$ 1#&7(' ; also the passage, from the same book, reportedin Laursen (!""%), p. !)(, l. !+, (+ '[6"]# 7. [3]#(' .

33 Since Epicurus himself branded some of his early writings as mistaken, school orthodoxy hadno choice but to concur, as can be seen in Philodemus, De Stoicis ., )–%% (in Dorandi (!"#%)), whorefers to these juvenilia as -+! +& #;+$'+ or -+! +& #;+$,+" . Hence the subscriptio of book ..-,,,, ]-.(' .+& #;+$"(.. [, may perhaps be filled out as ‘[Concerning] the early [works]’ (Sedley (!"#"b), p. !*$n. &*), rather than ‘[From] the old [exemplars]’ (Vogliano (!"%#), pp. !" and !*$, Sedley (!"$&),pp. (+ and $"; some doubts in Cavallo (!"#&), p. ("). 34 Nat. ..-,,,, !% - Sedley (!"$&), p. )(.

touches form a part. That is, Epicurus does not use the crabbed prose ofhis physical epitomes, or the contrived Gorgianic symmetries of hisethical epitome the Letter to Menoeceus. Rather, he a/ects what we may callthe lecturing style. He tends towards being expansive, repetitive, and, attimes, colloquial.35

A good example of this is his repeated use of the demonstrative toconvey the notion of multiple choice. A crucial tenet defended in book..- is our initial potentiality to develop in many alternative ways:36

+& 6[/" ] -" [-3' ) 6#](" -3) +& #;3' ) '6""#µ.[+-+ 3$ µ$' . +& ]4. (4+! -+! µ"! . "$& ) -+" 5. ["],-+! 5 & "$& ) -+" 5. ", -+" 5 & "$& ) +% µ%( [-+&' ]-+" ["& ]'-$. +& "$! [,+]$! 6#+" [@]"(. [,+$! ]5$+./3" '".(. ,+$! 5$+1"" ['"](. ,.+$! 67"$" [(] ,+$! "&7+" --(.

From the very outset we have seeds directing us some toward these, sometowards those, and some towards both these and those, actions and thoughtsand characters, in greater and smaller numbers.

Already in book ., the same mannerism can be witnessed. At one pointhe asks how mathematical astronomers can choose a privileged vantagepoint for taking measurements of risings and settings:37

-[$" ] 4+! # -3! . "&.[1"".5]" ,[+]-+'-+" 1µ3'. $. 3) -3! .. "&.1"" .5" ,+-+'-+" 1µ3'$. 3) -3! ."&.1"" .5" 3) -3" .[5"] '. " [6]/$.". $'.. 5"$' 6$'. [-/-""#]+. . ,+-+'-+" 1. [µ3'$.] -(' .+& .+[-]/.7(' . [3) 5]&" '"(.. ;

Why, after all, should you make the measurement from here, or the measure-ment from here, or the one from here, or this one, a more reliable measurementof risings or settings?

The heavy repetition is a quite unliterary feature, but one fully appropri-ate to the lecturing style. It is only too easy to imagine each repetitionaccompanied by a gesture.

Other human touches include disarmingly frank talk about otherphilosophers. Sometimes Epicurus shows exasperation, as when he saysof Plato’s reasoning about the interrelation of the elements in theTimaeus38

!*+ %. Epicurus, On nature

35 The ‘chat’ in book ..-,,, even contains a totally non-literary usage, albeit one well attested inPtolemaic papyri: "&+" . with the indicative (Sedley (!"$&), pp. +"–$*).

36 Text from Sedley (!"#&), p. !" (!Long/Sedley (!"#$), %*1 !), but with the first word altered toaccommodate the new reading of Laursen (!""%), p. !(! n. !#. The importance of these ‘seeds’is confirmed by a new reading of the following passage given in Laursen (!""%), p. !)(, l. !(. Thesame concept can be paralleled at Apuleius, De Platone !*(, hominem . . . habere semina quidem quaedamutrarumque rerum cum nascendi origine copulata.

37 Nat. .,, ! ,,, !![ %+.&$.!!–!" Arr.2]; text from Sedley (!"$+b), p. &!.38 Nat. .,-, ...-,, !%–!$, Leone (!"#)), p. +!!%".%.!%–!$ Arr.2. Leone, following Schmid (!"&+),

takes this column to be a critique of Tim. (+e on elemental intertransformation. I think it is morelikely to be focused on Tim. &!b–&%c. But that issue is irrelevant to the point I am illustrating.

+& 77+! 4+! # ,+$! -/&' -./ 4"7/$"() "&, -3) %+.-+'$"+) +& .+7"7/" 4$'-+$, ,+$! /&& ,"&6$'-+µ"" .() -+& %+."! ) 5$.+! -/&' %+$./µ"" ./& '&77/4$" [>]"'1+$.

But this too is a ridiculous piece of analogical reasoning from appearances,showing ignorance of how to work out the non-evident by means of the evident.

At other times, though, his criticisms are couched in generous terms, aswhere he regrets a lapse on the part of the early atomists when theyunwittingly implied universal determinism, but nevertheless describesthem as ‘greatly excelling not only their predecessors but also, manytimes over, their successors’.39

I have illustrated these human touches in the hope of capturing,however fleetingly, the tone which must have given On nature such aspecial place in the a/ections of Epicurus’ followers. To read it, it seems,conveyed a sense of joining Epicurus in his quest.

On the other side, it must be admitted that, to modern readers spoiledby Lucretius, the main style of argument throughout the surviving frag-ments is likely to seem remarkably sparse, abstract, theoretical andimpersonal in tone. Physical theses are defended by regular appeal to theusual Epicurean requirement that they be ‘consistent with’ or ‘uncon-tested by’ phenomena, but there is no sign of Epicurus focusing his audi-ence’s minds, as Lucretius would have done, on familiar examples of theanalogous phenomena. To read these books was surely a demandingtask. When at the beginning of the Letter to Herodotus we saw Epicurusspeaking of those who were ‘unable’ to read them, he may have beenreferring to their lack of the required intellectual stamina, not just lackof time.

However, this impression must not be taken to extremes. It is easy tobe misled by the fragmentary state of these texts into overestimatingtheir original unreadability. Often our editions of them contain sen-tences which appear to make little sense, but we must always rememberthat (a) these sadly damaged texts are only as good as the fallible editorswho have restored them (myself included), and (b) we have fragmentsonly from the later parts of the books, often undoubtedly deploying allu-sions, concepts, terms and assumptions which were set up in the lostearlier parts.

Graziano Arrighetti, although his work on the papyri of On nature hasdone so much to further our synoptic understanding of them, andalthough he has himself fully appreciated their importance as a show-case for Epicurus’ human side, has also promoted the idea that the work

$. Style !*$

39 Nat. ..-, in Sedley (!"#&), p. %*!Long/Sedley (!"#$), %*1!&. For full text, see below p. !)%.

as a whole lacked a formal structure.40 He points out (as we shall alsonote in §!%) that it was written in stages over many years, with a certainheterogeneity of styles, and argues that in it Epicurus was happy toreturn to a topic for a second time if his interest in it was rekindled bysecond thoughts or polemical needs. From this he concludes that the col-lected books of On nature did not constitute a single linear investigation,but served the Epicurean community as a record of the various phasesin the evolution of Epicurus’ philosophy.

Let us pause to try a small thought experiment. Imagine that in a par-allel universe numerous small papyrus fragments of a large philosoph-ical work have been discovered. Scholars trying to reconstruct the workare perplexed by the wide variety of topics and styles. Some fragmentsappear to discuss music, some metaphysics, some politics. Some containhigh-flown mythological narrative, others are colloquial in style and takethe form of rapid dialogue. Dating criteria show the first and last booksprobably to have been written years apart. And one topic occurs twicein widely separated books, the later discussion appearing to overturn theconclusions of the earlier one. The scholars naturally incline to the sus-picion that this work lacks formal unity, that it is a compilation of a some-what ad hoc character, ranging over several phases of the author’sthought. But in this instance they are wrong. For the work which I havedescribed is a literary and philosophical masterpiece of outstandingformal unity: Plato’s Republic.

Our papyrus fragments and other testimonia do not compel us toexpect any less of On nature. Returning to a subject already treated, forexample, may be perfectly proper even within a single consecutivelystructured argument (just as it was proper for Plato to return to thesubject of poetry once he had his psychology and metaphysics in place).Besides, the philosophical logbook described by Arrighetti would be amost unusual kind of work, and that fact is enough to place the onus ofproof firmly on anyone who wishes to propose such a hypothesis.

A more direct obstacle to Arrighetti’s view is the following. In theclosing sentence of On nature ..-,,, (quoted p. !*) above), Epicurusdescribes the treatise as +& ,[#/]+" '"() -3' [)] "$ @3' ) 6"#+$.[/]µ"" .3), ‘oursequential lecture-course’. What does this description mean? The paral-

!*# %. Epicurus, On nature

40 Arrighetti (!"$&), appendixes ! and %, esp. pp. $!(–!$. Arrighetti’s own main aim here is to explainthe apparent structural incoherences of the Letter to Herodotus and the Letter to Pythocles by deriv-ing them from the structural incoherence of their source text, On nature. In my view, what struc-tural awkwardness the Letter to Herodotus displays is adequately explained by the process ofabridgement; while that of the Letter to Pythocles is likely in addition to reflect its derivation fromtwo or more di/erent but possibly overlapping source texts (cf. §" below).

lels for the expression "$ @3' ) 6"#+$"."$. show that it does not simply signify‘accomplish in a series’, but ‘accomplish in a proper order’. In Plato’sGorgias Socrates explains to Gorgias the interrogative methodology he isadopting: ‘not for your sake, but for the sake of the argument, so that itmay proceed in such a way as to make it totally clear to us what we aretalking about’ ()(&c%–)); and referring back to this explanation shortlyafterwards, he remarks (in Dodds’ translation): ‘For as I said ()(&c%), I amquestioning you in order to get the argument carried through in a coher-ent way . . .’ (-/&' "$ @3' ) "+ .",+ 6"#+$"."'1+$ -/! . 7/" 4/., )()c!–%). This isstrong evidence that what Epicurus is claiming for On nature is exactly thefeature which Arrighetti denies it, structural coherence.41 That finding isfurther confirmed by Epicurus’ concern, in the closing sentence of book,, (below p. !!!), to stress the proper thematic continuity (again expressedby "$ @3' )) between this book and what follows it.

I do not mean to insist that before he set pen to papyrus Epicurusalready had the entire work mapped out. How many writers could claimto do that? All I want to insist on is that he saw its sequence of topics,however and whenever decided upon, as forming a natural philosoph-ical progression.

+ . 7618?9 57A15,?B 526 18?56?5 9

So what was the work’s structure? What is the ‘proper order’ in whichEpicurus delivered his massive lecture series? It is important to remarkthat arguing in the right order – in particular, arguing in such a sequenceas never to presuppose what is yet to be proved – is a cardinal feature ofEpicurus’ physical methodology, one to which we will turn in Chapter$.42 What I want to attempt for the present, however, is not so much amethodological defence of Epicurus’ chosen sequence of topics, but theactual task of establishing what that sequence was. For this purpose Ishall draw on nearly all the available evidence.43 This includes of coursethe scholia and other secondary citations, as well as the papyrus frag-ments. And one crucial further guide will be the sequence of topics inthe Letter to Herodotus, which, as we have seen, is almost certainly pre-senting itself as an epitome of On nature. Even the Letter to Pythocles willcome to our rescue at one stage.

&. Reconstructing the contents !*"

41 Cf. also Plato, Phlb. !%b)–+. 42 See also Ch. (, §§)–$ below.43 There is a helpful succinct guide to the data on Nat. in Erler (!"")), pp. ")–!*&. Cf. also Steckel

(!"+#), cols. +*!–!!, and of course the texts in Arrighetti (!"$&), pp. !#"–)!#, with commentarypp. ($$–+*".

Lucretius himself, however, had better stay in the background. Theaim of this reconstruction is to be able, in the next chapter, to determinethe relation of Lucretius’ organisation of topics to that in On nature. Toavoid circularity, therefore, On nature must be reconstructed so far as pos-sible without Lucretius’ help.44 However, the occasional glance atLucretius will prove unavoidable.

One restriction which can be fairly safely asserted at the outset is thatOn nature was a treatise on physics, and not, as has been suggested,45 onecovering the whole of Epicurus’ philosophy. None of our evidence aboutit requires us to dismiss the obvious meaning of its title, confirmed bythe following passage of Diogenes Laertius (. %"–&*):

It [Epicurus’ philosophy] is divided into three parts: canonic, physics and ethics.Canonic covers the methodological approach to the system, and is found in thework entitled ‘Canon’. Physics covers the entire study of nature, and is found inthe thirty-seven books On nature, and summarised in the letters. Ethics covers theissues of choice and avoidance, and is found in the books On lives, in the letters,and in On the end. However, their practice is to class canonic with physics.

This very clearly announces that On nature was a work devoted to thephysical part of the system, but that we should expect to find some treat-ment of epistemology (‘canonic’) there too – as indeed we do. Althoughthere were undoubtedly ethical remarks in it,46 and although a great dealof Epicurus’ physics has ethical implications, there is no evidence thatany book or substantial section of a book was devoted to ethical inquiryas such.

Readers who have limited stamina and are willing to take a certainamount on trust may wish to jump to §!&. But the wary would be betteradvised to check my arguments in the intervening sections.

$. '88E9 ,> ,-

To reconstruct this first group, it is best to start at the very end of book,,, which survives virtually intact:47

!!* %. Epicurus, On nature

44 An earlier version of the reconstruction which follows appeared in Sedley (!"#)), and is repro-duced in Erler (!"")), p. "(. But there I used Lucretius as a source for the reconstruction itself.On the present occasion I shall place Lucretius alongside the reconstruction of Nat. (Ch. (, §%)only when it is virtually complete. 45 Steckel (!"+#), col. +*!, cf. +*".

46 See §!! below for some ethically-related themes in the later books, and cf. DL . !!", where thedictum that the wise man will marry and have children is attributed to On nature. For the begin-ning of an ethical digression in a non-ethical book, cf. Nat. ..-,,,, !& -, # sup. /. in Sedley (!"$&),p. )". Cf. also n."*, p. !%$ below.

47 Vogliano (!"(&), pp. "%–&!%).(*.!$–(!.# Arr.2. My text combines the letters preserved in bothof the parallel papyri.

+& [6/5"" ]5"$,-+$ /&# . 3$ µ[$' ]. 53! /+ -$ "% [']-$[. "$% ]5(7+, ,+$! /+ -$ -3! . 4"" ."'$.+&& -(' . [++ ]µ+ ./3" µ+-$ '&µ2""23,". +& 6/-"7"$''1+$, ,+$! "%-$ -+! ) %/[#]+! )+& .&6"#273" -/&) -/$' ) -+" ;"'$. ,",-3' '1+$. -+! 5 & +$ #µ/" --/.-+ "$ @3' ) -/&" -/$)#$ 313' .+$ "&. -/$' ) µ"-+! -+&' -+ 5$"" @$µ"..

Thus we have proved (a) that images ("$%5(7+) exist, (b) that they have the prop-erty of being generated as quick as thought, and (c) that they have motionsunsurpassed in speed. In what follows [or ‘in the following books’]48 we will gothrough the topics which belong in sequence after these ones.

Book ,,, then, or at least its final part, introduced the ‘images’ whichEpicurean physics postulates to explain vision and thought. In fact, wehave enough of the two papyri extant to be able to say that topic (c) occu-pies the last fourteen columns of P. Herc. !!)"/""&. That is, given thecolumn size in this particular copy, topic (c) ran to some +,&** letters.Hence, assuming that the book was comparable in length to book .,-(see §) above), (c) occupied less than one-twentieth of the book. Howlong, then, was the whole account of images?

Actually, the account appears to have come in four parts. In additionto the three which Epicurus lists in his summary, there was undoubtedlyalso a proof of the fineness of the images, which must have been placedbetween (a) and (b).49 We may call this (a)*. To help us guess the totallength of the sequence (a), (a)*, (b), (c), we must fall back on our paralleltexts – not only the Letter to Herodotus, but also Lucretius (whom we mayinvoke as attesting the relative quantities of Epicurean material on theseseveral topics, without any presupposition at this stage that On nature washis direct source).

In Lucretius, (c) occupies )% or more lines (,- !$+–%!+, followed by alacuna of one or more lines); (b) occupies &) or more lines (,- !)&–$(,with a lacuna of at least one line); (a)* occupies %* or more lines (,-!!*–%#, again including a lacuna); and finally (a) occupies (+ lines (,-()–!*"). That is, in Lucretius (c) is between a third and a quarter of the wholeargument. In the Letter to Herodotus ()+–#) (c) takes up !( lines of von derMuehll’s text, (b) !* lines, and (a) $ lines. ((a)* is not treated separately

'. Books I–IV !!!

48 Where Vogliano reports -/$' ) from both papyri, Arrighetti prints -+$$' ), but without signalling acorrection to Vogliano in his ap. crit. I therefore assume -+$' ) to be merely a typographical error.If the feminine had been correct, the reference would undoubtedly have been to ‘books’(2$"27/$)). The neuter is less explicit. It could refer to books (2$"27$"/$)), or, more vaguely, to thenext discussion (cf. also n.+#, p. !%* below).

49 %).&(–+ Arr.2 clearly marks the transition from (b) to (c). Epicurus’ very first argument in (c) forthe images’ high velocity (%).&+–$) uses their fineness as a premise (see below p. !&"). Thereforethe proofs of fineness must precede (b). But they can hardly precede (a). So they came between(a) and (b). The reason why Epicurus does not list (a)* in his closing summary is simple: the fine-ness of images was needed, not as a conclusion in its own right, but as a premise for establish-ing their speed of generation and travel.

there, simply appearing as a premise in (c).) Thus in Epicurus’ ownsummary of the same argument (c) is nearly half the argument.

These parallel texts suggest that in Epicurus’ original exposition too,the arguments for the images’ speed which make up (c) were a substantialpart of the whole, and certainly not less than a quarter of it. So since, inOn nature ,,, (c) has proved likely to be less than one-twentieth of thewhole book, it is scarcely imaginable that the entire account of imageswith which the book ends occupied more than the last fifth of it.

This conclusion has required some tedious calculations, but it is of theutmost importance. Scholars have usually assumed from its closingwords that book ,, as a whole was about images. We have now seen thisto be almost inconceivable. And, by the same move, we have freed upthe bulk of book ,, for other contents.

There is no problem about filling the space vacated. According to a scho-lion on Letter to Herodotus $&, Epicurus’ description of time as a specialkind of accidental property ('&" µ6-(µ+) was also to be found ‘in Onnature book ,,, and in the Great epitome’. This has caused some consterna-tion. The fact that the equivalent remark comes only late on in theLetter, combined with the assumption, now proved incorrect, that thewhole of book ,, was about images, has led some to suspect that the booknumber in the scholion is corrupt.50 This is unnecessary. The remarkthat time is a special kind of '&" µ6-(µ+ almost certainly first occurredat a point in On nature corresponding to Letter to Herodotus )*. There,having established the ontological dualism of body and void, Epicurusadds:

Over and above these nothing can be conceived, either by imagination or byanalogy with what can be imagined, as things grasped in terms of completenatures, and not as what we call the ‘accidents’ ('&µ6-(" µ+-+) and properties('&µ2"23,/" -+) of these.

Now the manifestly parallel passage in Lucretius (, )&*–#%) arguesagainst the independent existence, not only of ‘accidents’ and ‘perma-nent properties’, but also of time and of historical events. There need belittle doubt, then, that the full text which Letter to Herodotus )* is abridg-ing had already included a preliminary analysis of time as a special kindof ‘accident’. This early analysis of properties and time had a primarilynegative motive – to eliminate them as contenders for independent exis-tence. The more sophisticated account of them which is reflected at

!!% %. Epicurus, On nature

50 E.g. Arrighetti (!"$&), pp. ($"–#*, +(*, Erler (!"")), p. "+.

Letter to Herodotus +#–$& was presumably one which Epicurus was able toundertake only in a later book of On nature, once his metaphysics ofatomic and phenomenal properties was in place. At Letter to Herodotus )*,atoms have not even yet entered the story. (This exemplifies the pointmade above, p. !*#, that Epicurus may have had entirely legitimatereasons, within the constraints of his methodology, for returning to atopic for a second time.)

We are now in a position to list the contents of books ,–,,. Book , nodoubt started with the methodological preliminaries echoed at Letter toHerodotus &$–#. The remainder of it was devoted to establishing thefoundational principles summarised at Letter to Herodotus &#–)*, giving usthe full sequence:B O O K I

(i) methodological preliminaries;(ii) nothing comes into being out of nothing;(iii) nothing perishes into nothing;(iv) the all never changes;(v) the all consists of bodies and void;(vi) of bodies, some are compounds, others the constituents

of those compounds.Some confirmation is found in the fact that (v) is explicitly attributed toOn nature , in a scholion on Letter to Herodotus &". I have added (vi), fromthe end of Letter to Herodotus )*, because a scholion there attributes it toOn nature book ,. I shall return to this point in a moment. Finally, anunnumbered book of On nature refers back to something that has beensaid about ‘collisions (,#/&" '"$)) with each other’, apparently in book ,.51

In the absence of further information, it is hard to know where to placethis.

'. Books I–IV !!&

51 Fr. $# Usener!&+.%) Arr.2, . . . ] +&. ..+4,+$' /" . -$ +&& -+$' ) &$ 6+" #;"$. ,+-+! -+! [)] 6#/! [)] +& 773" 7+),#/&" '"$), ($ ) "&. -3' [$] 6#(" -3[$] 4#+%3' [$] "$%#3-.+[$], /&& 1"! . 3( --/[. 6]+#+! -+! ) [. . . That thisexpression might refer to ‘book’ , is suggested by Philodemus’ usage at P.Herc. !**(, .-,,, +–!%(Angeli (!"##), p. !#%); see also Capasso (!"#!), p. &"$ for Epicurean usage of 4#+%3" . I do not seehow Epicurus could possibly have reached the topic of specifically atomic collisions in book ,: weshould expect them first to occur in item (xi) below, which, as we shall see, must be well into book,,. Possibly the feminine pronoun +&& -+$' ) refers not to atoms but to compounds ('&4,#$"'"$)),whose collisions might have cropped up in the context of item (iii), (v) or (vi). Alternatively,Epicurus may be referring back to his first ‘draft’ of the current book (cf. Nat. ..-,,,, # ,- "–!*in Sedley (!"$&), p. )! for a possible reference to two drafts of the same book), to the first ‘book’of the current section (i.e. group of books) of the treatise, or to the first ‘section’ (i.e. group ofbooks) of the entire treatise. In the majority of cases, 4#+%3" in Epicurean sources means‘writing’ or ‘piece of writing’, rather than ‘book’ (see Capasso loc. cit.).

And that is all for book ,, because the next item (Letter to Herodotus )*),(vii) nothing exists independently apart from bodies and voidwas (as we saw above when considering the topic of time) argued in book,,, presumably at the beginning of it. The sequence at Letter to Herodotus)*–! implies that book ,, continued with the proof that(viii) the constituents of bodies (distinguished already in (vi)) are atomic.There is a small inconcinnity here, in that the Letter places (vi) after (vii).However, there is no di0culty in assuming that (vi), the very basic divi-sion of bodies into compounds and constituents, did indeed first occurin book , as part of the defence of (v), but that it was then re-invoked inbook ,, as a premise for (viii). It was in fact one of Epicurus’ best-knownand most repeated dicta.52

If we take our lead from Letter to Herodotus )!–(, the full sequence forbook ,, can be filled out as follows:B O O K I I

(vii) nothing exists independently apart from bodies andvoid;

(viii) the constituents of bodies (distinguished already in (vi))are atomic;

(ix) the all is infinite;(x) there are unimaginably, not infinitely, many atomic

shapes;(xi) atoms are in perpetual motion;(xii) there are infinitely many worlds;(xiii) there are images, which are swiftly generated and fast.Here (xiii) of course links up with the ending of book ,, as we havealready met it. Given our conclusion that (xiii) occupied one-fifth of thebook or less, there seems no obstacle to assuming that (vii)–(xii) fittedsnugly enough into the remaining four-fifths (or more).

Finally, we can look ahead into books ,,,–,-. As we saw, at the end ofbook ,, Epicurus promises to continue with themes consequent uponthose of the existence and speed of images. This makes it overwhelm-ingly likely that he proceeded with much the same sequence of topics asin Letter to Herodotus )"–(&, namely the functioning of vision and theother individual senses, and of thought. The wording of his promiseleaves it uncertain whether he is referring to the following ‘books’, or

!!) %. Epicurus, On nature

52 The scholion on Ep. Hdt. )* reports that it was repeated also in Nat. .,- and .- and in the Greatepitome; and in Philodemus, Piet. &$–)! (Obbink (!""+), p. !*#) it is quoted verbatim by critics ofEpicurus’ theology.

more vaguely to ‘what follows’, the latter leaving open the possibility thatthe whole of this topic was contained within book ,,,.53 However, the tinyfragment of evidence that we have for book ,- may support the expecta-tion that the same topic continued there. It is a minute papyrus frag-ment, attributed to Philodemus, and containing an explicit citation of‘On nature book ,-’, with the preceding words apparently including someform or cognate of µ.3" µ3, ‘memory’.54 Memory would certainly beincluded under the rubric of ‘thought’, so in a small way this snippetencourages the assumption that psychological phenomena constitutedthe contents of books ,,, and ,-, and that thinking was dealt with afterperception (although there is no separate treatment of thinking includedin the Letter to Herodotus). This seems to justify the following tentativeaddition to our accumulating chart:B O O K S I I I–I V

(xiv) vision, truth and falsity;(xv) the other senses;(xvi) thought.Before we leave this first group of books, a word of caution is needed.We have found an encouraging degree of fit between the overallsequence of topics in On nature and that in the Letter to Herodotus. But itcannot necessarily be traced down to the details of the individual sec-tions. Although the treatment of images occurs in each work at the samepoint in the sequence, in the Letter Epicurus has undoubtedly adjustedthe sequence of arguments within that section.Thus where On nature has(a) existence of images,(a)* their fineness,(b) their speed of generation,(c) their velocity,the Letter ()+–#) has the sequence(a) existence of images(c) their velocity – including (a)*, their fineness, merely as a premise,(b) their speed of generation.The suppression of (a)* as a separate item is insignificant, merelyreflecting the obvious fact that epitomisation involves some abridge-ment. But the reversal of (b) and (c) is not insignificant. This smalldi/erence will, as it happens, be extremely useful to us in the nextchapter.55 But how is it to be accounted for? It may be that to produce

'. Books I–IV !!(

53 See n. )#, p. !!! above.54 %( Arr.2!fr. #* Usener. The remaining restorations in the first two lines are sheer guesswork, and

should be discounted. 55 P. !)! below.

the epitome Epicurus worked his way through the books of On nature,selecting the principal topics for inclusion, but that when he settled oneach topic he then wrote his digest of it largely from memory. Thus incases where the precise order of arguments did not matter methodolog-ically, some variations could have crept in. Neither of (b) and (c) seemsto presuppose the other, so that they might easily and harmlessly bereversed. On the other hand, they do undoubtedly both presuppose (a)and (a)*, and both On nature and the Letter to Herodotus in their own wayrespect that fact.

# . '88E9 ->.

There is not much helpful evidence about these books, beyond the con-tents of Letter to Herodotus ()–$&.56 The best way forward, therefore, is tolist the main topics in the sequence found there (chapter references tothe Letter are added):(xvii) atoms lack secondary qualities (()–();(xviii) atomic dimensions (((–");(xix) up and down (+*);(xx) equal speed of atoms (+!–%);(xxi) nature of the soul (+&–();(xxii) mortality of the soul (+(–+);(xxiii) soul is not incorporeal (++–$);(xxiv) metaphysics of properties and time (+#–$&).I have already speculated (pp. !!%>!& above), on Epicurus’ motives inreturning to the topic of properties and time, as he does in (xxiv). Butwhy precisely here? The conjunction of (xxiii) and (xxiv) shows that itwas the issue of incorporeality that provided him with a natural bridgefrom the topic of soul to that of properties. To call either the soul orproperties ‘incorporeals’ is implicitly to endow them with independentincorporeal existence, and hence is an unwelcome step towardsPlatonism. The main object of (xxiv) is to say, in positive terms, whatontological status we can safely attribute to properties and time.

How far can we go in correlating topics (xvii)–(xxiv) to individualbooks of On nature? There are some obscure references to books -, (pos-sibly) and -,,, which might help us here. In On piety !*)&–#",Philodemus, engaged in cataloguing references by the authoritative

!!+ %. Epicurus, On nature

56 Starting the group -–. at Ep. Hdt. () assumes the correctness of my findings in §$, where thecognitive psychology of )"–(& is correlated with Nat. ,,–,-. Ending at $& is based on the findingof §" below that the closing part of Ep. Hdt. $& corresponds to Nat. .,.

early Epicureans to the benefit and harm that can befall humansbecause of the gods, lists a number of relevant citations from Epicurushimself. These include a sadly fragmentary passage which Obbinkrestores as follows (!*$$–#"):57

/$ µ/$" ]() ,+$! ["&. -D' ] "+ ,-([$ 6]"#$! [-/&' 5$],.+" >"'[1+$ . . . . . ]/.6[ . ] . [ . ],+$[. . . . . . . ]'-3'5+[ . . . . . ,]+$! "&& /" #,/&) [,+$! 5$],+$"/&) -+$' ) +& [#]$"'-+$)5$+5/" '"['$] ,$."$''1+$ ,+$! 6+# & +&$ -/&! ) ,+$! 6+# & "& ,"$"./&). [,]+$!6+#+[6]73'$"() "& . -(' $ /& 45/" ($.

similarly in book -,58 concerning adjudication (&–( words missing) . . . and thatthose who are oath-keeping and just are moved by the best transmissions, whichdepend both on their own selves and on them [i.e. the gods]. And likewise inbook -,,,.

It is risky to infer too much from these citations about the primarythemes of books -, and -,,,,59 since Philodemus is plundering them forhis own current purpose alone, and may be drawing on parenthetical oreven digressive remarks. Certainly we should not infer that theology wasthe main theme of these books. Philodemus implies merely that the gods– presumably meaning our mental attitude to them60 – were there addedto the ‘self ’ as a second explanatory factor for certain kinds of behav-iour. Nor can we be confident that ‘adjudication’ is the word associatedwith book -, (one alternative possibility would be "$& ],.+" >"'[1+$, ‘beingimaged’ or ‘being conjectured’). Nevertheless, the recurrence of thereported remarks in two almost juxtaposed books does suggest the pres-ence of a thematic link, which would have to be in some sense a psycho-logical one.

The remark about the best ‘transmissions’ (5$+5/" '"$)) has no preciseparallel in our surviving Greek Epicurean literature, but sounds verymuch like a technical reference to the transmission of motions within thesoul.61 Lucretius tells us that the fourth ‘nameless’ element in the soul

(. Books V–X !!$

57 Obbink (!""+), pp. !#*–!. My translation largely follows his.58 Obbink (!""+), p. )$$, hesitates before identifying this as a probable reference to Nat. -,. The

identification is, however, much strengthened by the reference to Nat. -, merely as -/! "$ ,-/" . inthe biography of Philonides (above p. !*%).

59 It is hard to decide whether it is anything more than coincidence that books -, and -,,, are like-wise coupled in the Life of Philonides: he wrote something in connexion with book -,,, (col. !&inf., Gallo (!"#*)), and possessed a set of notes on book -, (see p. !*% above).

60 For an excellent discussion of this question, see Obbink (!""+), pp. )(#–+).61 For a somewhat di/erent interpretation, see Obbink (!""+), pp. )$$–#. It is certainly possible that,

as Obbink proposes, the 5$+5/" '"$) represent the gods’ causal influence on us (in which case,however, this is a non-technical use of the word, as at Epictetus, Diss. , !%.+ and Marcus Aurelius, !$.+, not helpfully explicated by reference to Ep. Pyth. !*(–+). If so, the topic would still bepsychological.

‘transmits’ motions through the body (,,, %)(, ‘sensiferos motus quaedidit prima per artus’), and his verb ‘didit’ here sounds like a renditionof 5$+5$"5(µ$, the verb cognate with 5$+" 5/'$). The ‘best’ transmissionsperhaps contrast with the excessive or insu0cient sensitivity of thosesouls which have an unbalanced atomic composition, as described byLucretius at ,,, %##–&%%.

This makes it a tempting guess that all of items (xvii)–(xx), on atomicmatters, were contained in book -, and that books -,–-,,, were in theirentirety about the soul.

What then about books ,. and .? If we take at least the last topic,(xxiv) on the metaphysics of properties and time, to correspond to book., it is worth asking whether P.Herc. !)!&62 may not be the remains ofthat book. Its last part – perhaps as many as #* columns, which couldeasily be one-third of the whole – is about the metaphysical and episte-mological status of time. Its scribal hand places it in the samepalaeographical group as the bulk of the On nature papyri,63 and since allthe Herculaneum papyri from that group whose content has beenidentified have so far proved to be from On nature, it has long been recog-nised as probable that this is yet another book of the same work.

One remarkable feature of this text is that, at least in the survivingportion, it appears to take the form of a narrated dialogue. In view ofwhat we have noted above (§() about the stylistic heterogeneity of Onnature, including the personal address to Metrodorus in book ..-,,,, thisfact should in no way discourage us from assigning the book to it.64

Suppose that this assignment is right. If we also bear in mind (a) thatas much as the last third of the book in question is devoted to time, and(b) that the same topic, time, takes up slightly under one-third of thepassage on the metaphysics of properties and time in the Letter toHerodotus (+#–$&), it becomes a plausible inference that the whole of topic(xxiv) occupied the whole of book ..

We now have the following distribution: book - on atoms, books-,–-,,, on the soul, book . on metaphysics. What about book ,.? I cansee no easy way to decide whether it continued the discussion of soul,began that of metaphysics, or dealt with some further topic which hasfallen out of our sources. However, given the close thematic link between(xxiii), ‘Soul is not incorporeal’, and (xxiv), the metaphysics of proper-ties and time, it seems natural in the absence of counter-evidence to

!!# %. Epicurus, On nature

62 This text is &$ Arr.2. See Arrighetti (!"$&), pp. &#!–)!( and +)$–+$, which is in fact the fullest andmost up-to-date edition of it; q.v. for details of earlier editions (pp. +)#–"), and for a defence ofits provenance from On nature (p. +(*). 63 Cavallo (!"#&), p. (*.

64 Thus too Arrighetti (!"$&), p. +&*.

locate topic (xxiii) in book ,.. I therefore propose the following assign-ment of topics to books:B O O K V

(xvii) atoms lack secondary qualities;(xviii) atomic dimensions;(xix) up and down;(xx) equal speed of atoms;B O O K S V I–V I I I

(xxi) nature of the soul;(xxii) mortality of the soul;B O O K I X

(xxiii) soul is not incorporeal;B O O K X

(xxiv) metaphysics of properties and time.

" . '88E9 .,>., , ,

We now move to cosmology. The best vantage point from which to starton this group is the last part of book .,. In the papyrus fragments, book., concludes with the following trio of topics: the motions of the heav-enly bodies; an attack on astronomical devices; and, finally, the stabilityof the earth. But a scholion (on Letter to Pythocles "!) also refers us to book., for the doctrine that the sun and other heavenly bodies are as smallas they appear (see Ch. &, §$), and since this is also astronomical it is likelyto belong not much before the surviving fragments on celestial motions.Thus we get:

size of heavenly bodies;motions of heavenly bodies;attack on astronomical devices;stability of the earth.

The first three of these coincide exactly with the sequence of topics atchapters "!–& of Epicurus’ surviving epitome on cosmological, astronom-ical and meteorological matters, the Letter to Pythocles.65 There, moreover,

*. Books XI–XIII !!"

65 I have little to contribute to the issue of its authenticity (on which see esp. Bollack/Laks (!"$#),introduction; cf. also the illuminating remarks of Mansfeld (!"")a), p. %" n. %), beyond theobservation that the ancient doubts about its authenticity recorded by Philodemus (see Angeli(!"##), pp. !$+–$, with valuable commentary pp. %#"–"") probably just reflect the laterEpicureans’ problems in establishing a fully consistent canon of texts by the school’s founders(see Sedley (!"#"b), p. !*+), and are unlikely to be motivated by impartial scholarly suspicions.Without this potentially misleading information from Philodemus, I doubt if any modern scholarwould ever have questioned the Letter’s authenticity. However, even for those who persist withsuch doubts, there should be no problem in regarding it as somebody’s digest of Epicurus’ writingson the same subject matter, which is all that matters for my own purposes here.

the denunciation of astronomical devices as the ‘slavish contrivances ofthe astronomers’ (-+! ) +& .5#+6/5(" 5"$) +& '-#/7/" 4(. -";.$-"$"+), Ep. Pyth."&) echoes the wording of the equivalent passage in book .,, which like-wise speaks of the astronomers’ ‘enslavements’ ([+& .5#+]6/5"$"+)) to doc-trine in the use of their instruments.66 It starts to look extremely likely –as Arrighetti has already maintained67 – that at least this early part of theLetter to Pythocles is epitomising On nature book .,.

The Letter to Pythocles itself merely claims to be summarising whatEpicurus has said ‘elsewhere’ ("& . +% 77/$), #)).68 Hence he may well bedrawing on more than one source text, and it would be unwise to rushinto correlating the whole of the Letter with On nature. But in its earlypart, at least, we seem to be on relatively safe ground in doing so. Sincethis letter (#() refers to the Letter to Herodotus and therefore postdates it, itis reasonable to take it to be filling a gap which the first letter had left inits exposition. This will be because the Letter to Herodotus, althoughdealing at some length with the same range of topics ($+–#%), does so inentirely general terms, stressing methodological, religious and ethicalaspects, without actually supplying any specific explanations of phe-nomena. It is this gap that Pythocles, the recipient of the second letter,has apparently asked Epicurus to fill (Ep. Pyth. #)).

Now the preceding topics in the Letter to Pythocles appear to join thoselisted above in making a seamless whole, and it is at least a strongpossibility that the entire opening sequence of the Letter reflects thesame source. This assumption would allow us to add the two openingtopics at Ep. Pyth. ##–"* before the above sequence, giving us the fullsequence:B O O K X I

(xxv) origin of our world;(xxvi) origin of heavenly bodies;(xxvii) size of heavenly bodies;(xxviii) motions of heavenly bodies;(xxix) attack on astronomical devices;(xxx) stability of the earth.While it cannot be proved that this entire sequence, even if correctlyreconstructed, fell entirely into book .,, it does not seem too much

!%* %. Epicurus, On nature

66 Text in Sedley (!"$+b), p. &% [!%+.&#.#–" Arr.2]. 67 Arrighetti (!"$&), esp. chart on p. $**.68 Jaap Mansfeld has pointed out to me that this plural is sometimes used even for a precise cita-

tion of a single text. In Aristotle (see numerous instances in Bonitz (!#$*), s.v. &L#$'-/-"" 73)) cita-tions of specific passages in his own works standardly use the plural: "&. +% 77/$), "&. "$ -"" #/$), "&. -/$' )6"#$" . . . etc. For Epicurus, cf. n. !%, p. !** above. Hence ‘elsewhere’ is safer than ‘in other works’.

material to fill one book, especially when one makes the comparison withwhat we learnt above (§$) about the contents of book ,,. Tentatively, then,we may adopt the above list as the probable main contents of book .,.

For book .,, we have no papyrus remains, but several secondary cita-tions. We may initially take these in the order in which Usener lists themas fragments, although omitting those where his attribution is purelyconjectural:fr. #%, scholion on Letter to Herodotus $): in On nature .,,, Epicurus

describes and delimits the range of shapes other worlds may have.fr. #&, scholion on Letter to Pythocles "+: in On nature .,,, Epicurus gives

various explanations of eclipses.fr. #), Philodemus, De pietate %%(–&! Obbink: in On nature .,,, Epicurus

speaks of early mankind’s formation of the concept of divinity.69

fr. )! (cross-reference under fr. #(), anon. Epicurean treatise (P.Herc.!!!!): some topic which Epicurus treated in his On piety he also dis-cussed in On nature .,, and .,,,.70

fr. #$, Philodemus, De pietate (%&–&& Obbink: in On nature .,,, Epicuruscriticises Prodicus, Diagoras and Critias (the standard list of atheists)as insane.71

To these we can add the closing sentence of book .,, as we have it in thepapyrus:

"& . 5"! -/$' ) "&;/[µ"" ]./$) "% [-]$ 6"#$! -(' . [µ"]-"(" #(. -/&-(."[$" -$]6#/'",,.+. [1]+. #/&' µ"..

In what follows [or ‘in the following books’?] we will continue with some furthersorting out72 of these celestial phenomena.

Now the topic of eclipses is one example of these promised furtherexplanations, and, regardless of how we translate the beginning of theabove sentence, Epicurus’ promise strongly suggests that this continua-tion was to be found (at least) in the opening part of book .,,. This hasthe further advantage of maintaining the parallel with the order oftopics in the Letter to Pythocles which, skipping item (xxx), the stability ofthe earth, passes directly from item (xxix), the attack on astronomicaldevices, to a whole range of(xxxi) further astronomical phenomena (phases etc. of moon; eclipses;

day length).

*. Books XI–XIII !%!

69 !%$.! Arr.2 70 See n. !&, p. !** above. 71 !%$.% Arr.2.72 On the problem of ,+1+$"#"$. and its derivatives in such Epicurean contexts, cf. Mansfeld (!"")a),

pp. &)–(.

At this point we seem to rejoin the Letter to Herodotus, at $&, where we leftit. Its sequence at $&–+ is: other worlds, origin of civilisation (exemplifiedby the acquisition of language). Both of these topics are, as we have seen,attested for book .,,, although the only aspect of the history of civilisa-tion explicitly recorded for it is the beginning of religion. After fullydescribing the origin and general arrangement of our world in book .,and at the beginning of book .,,, Epicurus evidently went on to drawsome inferences about how far other worlds can be expected to resem-ble and di/er from ours. He then resumed the history of our own worldby turning to the origins and development of mankind.73 This enablesus to reconstruct book .,, as follows:B O O K X I I

(xxxi) further astronomical phenomena (phases etc. ofmoon; eclipses; day length);

(xxxii) other worlds;(xxxiii) origins of civilisation.

Book .,,, is harder to pin down. All that seems to be recorded for it is adiscussion of human relationships to god – a0nity in some cases, non-a0nity in others – and possibly of other themes relating to piety.74 Itwould certainly be an appropriate point in his treatise for Epicurus to setout some moral lessons about our relationship to the divine. Now thathe has completed his account of the rise of civilisation, including theorigins of religious belief, he is able to comment on what is true and whatfalse in popular religion in the light of the entire physical system, butespecially by invoking the mechanistic, non-theological explanations ofcelestial phenomena in books .,–.,,. And the extended final expositorysection of the Letter to Herodotus ($+–#%), immediately following (xxxiii) theorigins of civilisation, dwells on just such a set of themes. We may there-fore tentatively add to our chart(xxxiv) the correct attitude to divinity.

This admittedly leaves one major question open. Where did Epicuruscover the further range of meteorological phenomena – weather signs,thunder, earthquakes etc., which I shall loosely call ‘atmospheric and ter-restrial phenomena’ – dealt with in the later part of the Letter to Pythocles(##–!!*) and in book -, of Lucretius? At least some of the final part of

!%% %. Epicurus, On nature

73 For the probable contents of the anthropological section, see Obbink (!""+), pp. &*+, &(*–!.74 Philodemus, Piet. !*(*–) Obbink!fr. ## Usener!%# Arr.2: "%. -" -(' $. -#"[$',+$]5",+" -($ 6"#[$!

-3' )] /$&,"$/" -3-/) 3*.[. 6#/" )] -$.+) /$ 1"/! ) "%;["$ ,+$! ] -3' ) +& 77/-#$[/" -3-/)]. In addition, the admit-tedly corrupt fragment of P.Herc. !!!! in !".( Arr.2 (see n. !&, p. !** above) appears to attributeto books .,, and .,,, of On nature some topic which also occurred in Epicurus’ On piety.

the Letter may well be drawn from other source texts than On nature.75

But we do not yet seem to have enough material to fill book .,,,. Sinceweather signs, thunder, earthquakes and the like have an obvious impor-tance to a section of the treatise in which Epicurus is debunking divineexplanations of cosmic phenomena, we might well expect some treat-ment of them in this book. As we shall see next, in book .,- Epicuruswill be making a completely new beginning, and unless the atmosphericphenomena are fitted into book .,,, they are likely to fall out altogether.76

Tentatively therefore (hence the square brackets) I suggest the followingcompletion for book .,,,:B O O K X I I I

(xxxiv) the correct attitude to divinity;[(xxxv) atmospheric and terrestrial phenomena.]

!* . '88E9 .,->.-

We have already seen in §& that books .,- and .- constituted a pair.The closing stretch of .,-, of which substantial parts are decipherablein the papyrus, is undoubtedly polemical, and there is every reason tobelieve that the criticism of rival theories of the elements was the uni-fying theme of the two books. The scholion on Letter to Herodotus )* tellsus that item (vi), ‘of bodies, some are compounds, others the constitu-ents of those compounds’, recurred in books .,- and .-, as indeed itwould have done if Epicurus was here returning to the theme of theprimary bodily elements in order to demonstrate the superiority of hisown theory to the rival ones. This critique is not covered in either ofour surviving epitomes – the Letter to Herodotus and the Letter to Pythocles– but we know from Diogenes Laertius (. %$) that there was a furtherepitome, now lost, the Epitome of criticisms of the physicists ( &E6$-/µ3! -(' .6#/! ) -/&! ) %&'$,/&" )). It was presumably a summary of books .,- and.-.

Working backwards once again, the first thing to observe is that thewretchedly thin fragments of book .- strongly suggest that it contained

!+. Books XIV–XV !%&

75 The use of separate but occasionally overlapping texts would be an easy way to explain theoddity (cf. Arrighetti (!"$&), pp. +"!/.) that weather signs are treated twice in widely separatedpasages (Ep. Pyth. "#–", !!(–!+), as well as the surprising return to astronomy at !!!–!( after asection (""–!!!) on atmospheric and terrestrial phenomena.

76 We could, of course, try to squeeze the atmospheric phenomena in with (xxxi), the astronomicalphenomena in book .,,. But this would leave a surprising imbalance between the contents ofbooks .,, and .,,,.

an extended critique of Anaxagoras.77 Along with two critical referencesto ‘those who have developed physical systems’ (/$$ 6"#$! %&" '"() 6#+-4µ+-"&/" µ"./$ or 6"6#+4µ+-"&µ"" ./$),78 forms or cognates of the wordhomoiomereia occur two or three times.79 This latter was the term whichthanks to the influence of Aristotle and Theophrastus came to be stan-dardly attached to his theory of elements, although it was almost cer-tainly not used by Anaxagoras himself. And we know that theEpicureans took up this usage.80 Two further signs that Anaxagoras maybe the target throughout are as follows.81

(a) The possible reference to atomic rearrangement (-.3! .6+[#+7]7+. 43! [.] -.(' . +& [-]/" µ(.) in fr. !) of the book is, in such acontext, reminiscent of Lucretius’ argument against Anaxagoras at ,#"$–"!), that phenomena like combustion through friction are betterexplained by atomic rearrangement than by the Anaxagorean principlethat there is a portion of everything in everything (see Ch. $, §%).

(b) Anaxagoras’ Principle of Predominance ('!% DK) ran as follows:

+& 77& /+ -(. 67"$''-+ "% .$, -+&' -+ "& .537/" -+-+ "* . "+ ,+'-/" . "&'-$ ,+$! 3# ..

But what each single thing most evidently was and is is whatever it has a major-ity of in it.

Each stu/, that is, despite containing portions of everything, has thephenomenal character of whatever ingredient in it exceeds the others.Or, as Lucretius puts it in his own corresponding critique, sed illud |apparere unum cuius sint plurima mixta . . .: ‘But only that single thing isapparent which is present as a majority in the mixture . . .’ (, #$$–#). Thissame dictum of Anaxagoras is surely what underlies Epicurus’ words infr. !! of book .-:82

!%) %. Epicurus, On nature

77 Millot (!"$$) properly notes some of the Anaxagorean allusions (pp. %#–"), although she does notdraw any conclusions from them about the overall content of the book (p. !)).

78 Forms of this expression clearly underlie frr. %) and Q in Millot (!"$$).79 Frr. $ and %(, and perhaps fr. %& in Millot (!"$$).80 Cf. Diogenes of Oenoanda + ,, )–$ Smith, quoted above p. $); Lucretius , #&*/. (cf. above p. )#),

with Rösler (!"$&).81 In addition, fr. &) is about misconceiving the divine nature. It is not impossible (I can put it no

stronger) that there is some allusion here to the charge of impiety that was brought againstAnaxagoras at Athens for implicitly denying the divinity of the sun.

82 The basis of what I am suggesting is already to be found in Millot (!"$$), pp. !$ and %". What Ihave added, apart from trivial changes of punctuation, is (a) a suggested change of the final wordfrom /$.µ/.$./µ"[#(' . to /$.µ/.$./µ"[#"$(' ., in conformity to what seems to be the normal Epicureanrepresentation of the Anaxagorean theory (cf. the references in n. $" above); (b) a translation. Iassume the subject and main verb carried over from the preceding sentence to be e.g. "$ ,+" '-3/&& '$"+ . . . ./µ$">"-+$.

,+1 & /* 4+! # [6#]/.'.+4. /#. "&" "-+$, /+.-. $. 53.' [7/., "$# ].+$ ,+-’ "& ,"[$' ].. / "& , -(' .[6]7. "$"'-(. -(' .[5]."" -$..(. [6"]6/$.3.[µ]"" .3, "& @ /$.µ/.$./µ"[#"$(' . . . .

For [it is claimed that] the respect in which it is spoken of – because it is evident– is the respect according to which it is made out of a majority of such and suchthings, namely the homoiomeries . . .

I conclude that book .- probably contained an extended critique ofAnaxagoras.

The final columns (...,.–.:,,,) of book .,- seem to have beendevoted to an obscure critique of certain forms of eclecticism in physics.The critique may have been sparked by Plato’s combination ofEmpedoclean and Pythagorean principles in his theory of elements,because prior to this last section Epicurus has been criticising in somedetail the theory of elements in Plato’s Timaeus (columns ...,-–...,.).It is much less clear what preceded this in book .,-. Giuliana Leone, themodern editor of the book, notes an allusion to an argument for monismin col. ...,,,, and proposes that the critique switches from monism topluralism immediately after that.83 This is of course possible, but itwould threaten to limit the critical content to a small fraction of theentire book. This is the book, remember, which we actually know to havebeen a massive %%,*** words long (see §)). Our assumption of thematicunity for books .,-–.- suggests that the surviving sections formed partof a more extended critique, filling most of the book (albeit, no doubt,after some methodological preliminaries).

Where does this leave us regarding the overall structure of .,-–.-? Therelatively clear fact that the series of critiques dealt with Anaxagoras lastis important, because it suggests that Epicurus is already organisingphysical theories along the lines set down by Aristotle (cf. Phys. , %,!#)b!(/.) and Theophrastus:84 first monism (Thales and co.), then finitepluralism (Empedocles and co.), and finally infinite pluralism (notablyAnaxagoras) – the same sequence later adopted by Lucretius (, +&(–"%*)and Diogenes of Oenoanda.85 I shall be arguing later (Chs. ( and +) thatTheophrastus’ Physical opinions was a major source for Epicurus in Onnature. It is also worth repeating that the Epicurean reading ofAnaxagoras as the proponent of homoiomereia, which can now be tracedback to On nature .-, is itself likely to reflect a considerable degree of

!+. Books XIV–XV !%(

83 Leone (!"#)), p. &). 84 See Mansfeld (!"#"), pp. !&#–)#.85 + , !*–,, " Smith (!""%), quoted p. $) above.

dependence on Theophrastean doxography, rather than on Anaxag-oras’ unmediated text.86

Accepting provisionally that such an ordering controlled the sequenceof topics in .,-–.-, the obvious guess to make is that .,- covered themonists and finite pluralists, ending with Plato (who, significantly, seemsto have been the latest philosopher covered by Theophrastus’ doxogra-phy), while .- was devoted to Anaxagoras. If this seems to allot toomuch space to Anaxagoras, we can remind ourselves that at the end of.,- Epicurus is locked into a methodological critique concerned withphysical eclecticism: it may be that the early part of .- continued toinvestigate related methodological issues, before turning to Anaxagoras.Or, if Plato is there being accused of mixing Empedoclean elementswith Pythagorean first principles, another possibility is that the first partof .- dealt with Pythagorean physics, which seems itself to have beenclassed, like Anaxagoras’ theory, as a species of infinite pluralism.87 Imention these possibilities merely in order to show that there is no obsta-cle in principle to our filling at least the bulk of books .,- and .- withthe following sequence:B O O K X I V

(xxxvi) critique of monism;(xxxvii) critique of finite pluralism.B O O K X V

(xxxviii) critique of Anaxagoras.

!! . '88E9 .-,>...-, ,

From the later part of the treatise, we have substantial papyrus fragmentsof books ..- and ..-,,,, and three skimpy citations – one of book ...,,,one of book ...,-, and a third of a book somewhere in the thirties.88

Book ..-,89 which survives in three parallel copies, is about psycho-

!%+ %. Epicurus, On nature

86 See Schofield (!"$().87 SE PH ,,, &%. The alternative possibility that Democritus was the infinite pluralist criticised in

the early part of Nat. .- is scarcely worth considering. Epicurus saw his own physics as a cor-rected Democriteanism (cf. Plut. Col. !!*#6–@), and his view is undoubtedly reflected in Diogenesof Oenoanda’s critique of the physicists, where Democritus is separated from the regularTheophrastean list of predecessors and reserved for criticism in the course of the exposition ofEpicurean atomism (+ ,, "–,,, ! Smith).

88 A possible reference to book .-, in P.Herc. ((#, !!.)–(, noted by Obbink (!""+), p. &*(, is betterignored for present purposes. The reconstruction of Crönert (!"$(), p. !)", "& . 5"! -(' $"$ ],,+[$]5"[,+" -($ is highly conjectural, and since the context seems to be a history of Socratesand/or Plato the reference would in any case probably not be to On nature.

89 The number of the book was first recovered by Laursen (!"#$); it was previously liber incertus.There has been no complete edition of it since the one in Diano (!")+), whose textual informa-

logical development, with a good deal of emphasis on the causal e0cacyof the self and our capacity to choose what we become. It includes –miraculously preserved almost intact – a brilliant argument againstdeterminism.90 At the end of the book, Epicurus’ résumé informs us thathe has now given his exposition in two modes, the first in terms of feel-ings (the 6+1/7/4$,/! ) -#/" 6/)), the second in terms of causes (the+$&-$/7/4$,/! ) -#/" 6/)).

Now there is an old puzzle about when, where and how Epicurusarrived at his doctrine of the atomic swerve – the third principle ofmotion alongside impacts and weight, introduced to rescue us from thetyranny of physical determinism. Since this doctrine does not show upin the Letter to Herodotus, which we have now seen to cover only books,–.,,, of On nature, there is a strong possibility that it was first developedin a later book of the treatise, where Epicurus had turned his attentionto the problem of determinism. If so, we may guess that this was in book..- or a neighbouring book. There is no reference to the swerve in thefragments of book ..-, but because these only represent the later partof it, perhaps the last third, and are anyway far from complete, we mustbe extremely cautious of any argument from silence. Since the swervedoes occur in Lucretius’ exposition, we will need to bear this book, orgroup of books, in mind in the next chapter, when considering hissources.

Book ..-,,, is about language and other areas of epistemology – re-examining, as we have seen (§(), the former views of both Epicurushimself and Metrodorus. Book ...,, is cited in an unassignedHerculaneum fragment91 as containing a brief definition of ‘soul’,summing up Epicurus’ previous theorising about it. Book ...,- is cited

!!. Books XVI–XXXVII !%$

tion was seriously inadequate, but who did at least print all three papyri alongside each other.Arrighetti (!"$&) prints as fr. &) a combined text which proceeds by alternating between the threepapyri as reported by Diano; it should be used with caution, since frequently letters which heprints in square brackets are in fact attested in one or both of the other papyri. For the earliestpreserved part of the book we now have the excellent edition of Laursen (!""(). Of the remain-der, while we await Laursen’s full edition, some parts must be read in Sedley (!"#&) andLong/Sedley (!"#$), §%* (with corrections in Laursen (!"##)), others in Laursen (!""%). It is toLaursen, and also to Puglia (!"#$), that we owe the identification of additional sections ofpapyrus as belonging to the same three rolls (P.Herc. +"$, !*(+, !!"!) that provide the three par-allel texts: the numbers of these additional papyri are included in the list I give on p. "# above.

90 For this passage, see above Ch. &, §#. That the book is about psychological development isbroadly agreed by Sedley (!"#&), p. !$, and Laursen (!""(), p. )%, who there seems no longer tobe pressing his suggestion in Laursen (!""%), p. !(&, that its theme was, more specifically, whethervirtue can be taught.

91 [&% Arr.2]; but see Sedley (!"$)), p. #" n. " for a correction of the text: the word ,+$" should beadded before the book citation, "&. -(' $ 2# ,+$ 7#.

for something said there about giving a criterion for things non-evident.92 Finally, there is an indication in a barely decipherable passageof Philodemus, De pietate that in book thirty-something – probably...-93 – Epicurus clarified some unspecified kind of ‘benefiting’((& %""7"$+) – apparently meaning some way in which the right attitude tothe gods is beneficial to mankind.

This is far too little evidence to permit any conjecture regarding theoverall structure of books .-,–...-,,. The paucity of citations suggeststhat this later part of the treatise dealt largely with peripheral questions,and was not so widely studied. The one clear exception is book ..-,whose survival in triplicate implies, as we saw in §&, that it was highlyvalued in the Epicurean community. This impression is furtherconfirmed by the fact that one of its arguments, the self-refutation argu-ment against determinism, was picked out for summary in the collectionof Epicurean sayings which have come down to us as the GnomologiumVaticanum.94

!% . 1278?8:8BD

A work as long as On nature cannot be written overnight. But when, andover how long a period, was it written? We are very fortunate that thepapyri of three books from it carry at the end, below their titles, theAthenian archon-year in which they were composed. Thus we have thefollowing firm dates:95

!%# %. Epicurus, On nature

92 Text in Polyaenus fr. %$ Tepedino Guerra (!""!), p. "!.93 [&& Arr.2]!Philodemus, Piet. !*((–" Obbink; see Obbink (!""+), p. )$) in defence of the reading

"& . 5"! -(' $ 6. "" [µ6-($ ,+$! ] -.#$+,[/']-(' [$.94 SV )*, which seems to be a non-technical summary of the self-refutation argument in Nat. ..-

(!Long/Sedley (!"#$), %*1 (): see further, p. ## above.95 I say this despite the important proposal of Obbink (!""+), pp. &(!–%, that these subscriptiones give

the dates of Epicurus’ later redaction of the books in question, not that of original composition.He argues that the subscriptio to book ..-,,,,

&E6. [$,]/.&" #/&!"#$! %&" '"()

,3#"& ,] -.(' . +& #;+$"(. [

"&4[#]+" %3 "&6$! K$,$"/& -/&' µ["-]+! &L.[-$]%+" -3..means, not that book ..-,,, was composed in that archonship – for which, he suggests, we shouldhave expected '&."4#+" %3 – but that this was when it was copied into its present edited form,‘from the old exemplars’, implying that a text had already existed for some time. I have severalgrounds for disagreement. (a) The dates of books .,-, .- and ..-,,, (see immediately below) areordered and spaced out in a way which fits the relatively laborious composition of a work muchbetter than that of its editing. (b) Epicurus’ own usual word for ‘composing’ a work is 4#+" %"$.(Ep. Pyth. #), #() – or +& .+4#+" %"$. (above pp. !**>!), but not '&44#+" %"$.. (c) Even if we were to

book .,-: archonship of Clearchus!&*!/&** '1;96

book .-: archonship of Hegemachus!&**/%"" '1;97

book ..-,,,: archonship of Nicias ‘the successor of Antiphates’!%"+/%"( '1.98

It is tempting to calculate as follows. Epicurus’ production rate in theperiod &*!–%"( was two-and-a-half books a year. By projecting this ontothe remainder of the work, we find that he started it in &*+/&*(, soonafter his move to Athens, and reached book ...-,, in %"& or %"%.

But the temptation deserves to be resisted. During those fourteen orfifteen years Epicurus must have been composing many other works,towards his lifetime total of &** volumes.99 There is no possible reasonto imagine that during this time the individual books of On nature werespaced out at regular intervals, like issues of a quarterly journal.

As a matter of fact, there is quite good reason to see the treatise ashaving been composed in two or more widely separated phases. I believethat the first phase, books ,–.,,,, belongs to his period spent teaching atMytilene and Lampsacus, from &!!/&!* until &*$/&*+, the year in whichhe moved to Athens and founded the Garden.100 I o/er three considera-tions in support of this.

The first piece of evidence for such a dating is negative. While threeof the books datable after Epicurus’ move to Athens carry, as we haveseen, an Athenian archon-year, no archon-year has yet been found on

!". Chronology !%"

agree with Obbink (who does not quote evidence) that the parallel of other ancient subscriptionesconfirms that those in Nat. should be expected to ‘record information regarding copying, redac-tion, and correction, and thus date the work of a scribe or 5$/#1(-3" ) (possibly the publicationof a particular version)’, that would still give us no reason to infer that our texts are a hypothet-ical second version, rather than the original lecture version (which itself had quite possibly beenpreceded by a working draft). (d) This subscriptio probably also featured as a superscriptio at thebeginning of the book, and the assumption that the date given is that of the original composi-tion and delivery of the lecture thus finds some confirmation in the formula "&5$5+" ;13 "&6$" (plusarchon name) in the prefaces to Aristophanes’ plays. (e) Even if the fourth line is to be restoredas ‘[from] the old exemplars’ (and I have proposed an alternative, which Obbink acknowledges,see n. &&, p. !*( above), it is certainly not to be construed grammatically with "&4#+" %3, therebeing a substantial space between them (see Cavallo (!"#&), p. ("). (f) The Epicureans werescrupulous about recording the archon-dates of individual letters by Epicurus and his colleagues(see below), and these unambiguously refer to the original date of composition, not to that ofsome later edition of the letters. (Obbink, pp. )&*–!, )(*–!, is unlikely to be right that these dateswere imposed on the letters only by later Epicureans: where would they have got the informa-tion from, unless the dates were already recorded on or with the letters? The fact that privateletters from Egypt are usually undated is irrelevant to what was clearly an institutionalisedEpicurean school practice.)

196 This date was first deciphered by Leone (!"#)), see esp. pp. &$–#.197 Millot (!"$$), p. %+. 98 Sedley (!"$&), p. (+. 99 DL . %+.100 That some books of Nat. predate the move to Athens was already proposed by Bignone (!"&+),

, ++) n. !&*, ,, )!" n. !+.

any of the four papyri from books ,, and .,. We may reasonably assumethat the practice of dating his books by their Athenian archon-yearbegan only when Epicurus moved to Athens. Although many suchEpicurean datings survive in the writings of Philodemus and elsewhere,with individual letters written by the school’s first generation regularlyassigned an Athenian archon-year, no such dating has yet proved to beearlier than &*$/&*+, when the Athens school was founded. Indeed, thefirst recorded archon-year in Epicurean history is that of Epicurus’foundation of the Athens school.101

The second piece of evidence for the dating is positive in character.102

In §( we met a passage of book ., in which Epicurus challengesmathematical astronomy: ‘Why, after all, should you make the measure-ment from here, or the measurement from here, or the one from here,or this one, a more reliable measurement of risings or settings?’ The evi-dence which he has adduced in the immediately preceding lines is asfollows:

. . . ] +& .+-""7. 7(., +& .+-"$"./.-") "$& ) -/! µ""#/) -3' ) 6+" '3) 43' ), /&( µ"-""23µ". "& ,-/&" -/ .& 3$ µ$' . 5&/" µ"./) %+$"."-+$, /&& 5"! 6/773! . "& .$"/-" 6+" .& 43' .µ"[-+2]"23,/" '$..

[The sun, if we walk towards the place from which it appeared to us] to rise,directing ourselves up into the mainland zone, appears to us to set where wepreviously passed by, sometimes even when we have not passed all that muchground.

Evidently he is describing the experience of walking eastwards at thetime of sunset. You look back westwards at the sun, and it appears to besetting at a place where you were standing shortly earlier. (Virtually theidentical phenomenon is described at Lucretius ,- )*)–!&.) It follows, inEpicurus’ view, that our terrestrial vantage points are useless for theobjective tracking and measuring of a celestial orbit.

But there is something very odd about his description of walking east-wards: "$& ) -/! µ""#/) -3' ) 6+" '3) 43' ), which I have translated as ‘into themainland zone’, means literally ‘into the zone of total land’. If Epicuruswere writing at Athens, that would be a highly misleading description ofthe narrow Attic peninsula to the east. It would be equally inappropri-ate to the island of Mytilene, where Epicurus taught briefly in &!!/&!*.

!&* %. Epicurus, On nature

101 DL . %. For the full set of Epicurean archon dates, see Dorandi (!""*); also Dorandi (!""!), pp.+%–&.

102 I summarise here the argument which I presented briefly in Sedley (!"$+a), pp. !&"–)&, and fullyin Sedley (!"$+b), pp. &!–+

But it does perfectly fit Lampsacus, on the Asian mainland, whereEpicurus spent four years of his teaching career before the move toAthens. In an Asian context the prefix +& .+- (as in Xenophon’s Anabasis)standardly means ‘in the direction of central Asia’, and the unusualcompound +& .+-"$"./.-") may well carry that sense. As for ‘into the zoneof total land’, I know no exact parallels, but it does seem a highlyappropriate expression for the Asian mainland, because in Thucydides-,,, )+.& the mainland of Asia is contrasted with ‘the zone of the sea’ (-/!µ""#/) -3' ) 1+7+" ''3)), meaning the whole Aegean zone to the west.103

It looks, then, as if when book ., was being written Epicurus and hisaudience were living on the mainland of Asia,104 in fact in the Epicureanschool at Lampsacus. Given that books .,–.,,, seem to form a closelylinked group, we can be fairly confident that the entire series ,–.,,, wascomplete by the time Epicurus moved to Athens in &*$/&*+. By then hehad completed the exposition of his own cosmology, and it is not par-ticularly surprising that he should have called a halt there, deciding onlyafter an interval of six or more years to extend the work with a system-atic refutation of rival theories of the elements in books .,- and .-.Before that resumption, which started in &*!/&** '1, he may well havebeen fully occupied with other matters, such as the task of setting up theAthenian school.

A third, subsidiary argument for this chronology can now be added.The Letter to Herodotus, although purporting to epitomise On nature, in factcovers only books ,–.,,,, suggesting that at the time of its compositionthese books may have been the sum total of it. That already suggests adate before &*!/&**, when book .,- was written. But we can go further.The Letter to Herodotus can be argued on independent grounds to have &*)'1 as its terminus ante quem: this is because it must predate the Letter toPythocles, which refers to it (#(), and Pythocles’ biography makes it hardto date the letter to him later than &*).105 If that is correct, books ,–.,,,were completed at least three years before the composition of book .,-.

!". Chronology !&!

103 As I report in Sedley (!"$+b), p. &+, a detailed contour map of Lapseki (the modern site ofLampsacus) confirms that there is a path leading due east from the town, along which thewestern horizon would often be very close, sometimes less than (** metres from the observer.So we can be confident that the phenomenon which Epicurus describes would be observablethere.

104 The present tense does not sit comfortably with the alternative possibility, suggested by Erler(!"")), p. "$, that Epicurus might be recalling an observation taken at Lampsacus earlier in hiscareer.

105 For a defence of this chronology, along with a biography of Pythocles, see Sedley (!"$+b), esp.pp. )(–+.

This does not yet quite force them back into the Lampsacus period, butit is a strong further pointer in that same direction, since it confirms theexistence of a substantial chronological break between books .,,, and.,-.

!& . 18?1:A9,8?9

On nature was a published lecture course in thirty-seven instalments.Books ,–.,,, were written in the period &!!/&!*–&*$/&*+ '1, whileEpicurus was teaching at Lampsacus, and the remainder was written atAthens. Books .,- and .- were written in &*!/&** and &**/%"" '1respectively. Book ..-,,, was written in %"+/%"( '1. The dates of theremaining books are a matter of pure conjecture.

As a series of lectures delivered within the school, On nature wascharacterised by an expansiveness and an informality which gave it avery special place in the a/ections of Epicurus’ followers. Certain bookswere especially highly prized, studied and quoted. These included ,–,,,.,–.-, and ..-.

The probable sequence of contents of books ,–.- can be given inoutline as shown in Chart ! (the references in brackets are to the corre-sponding passages from Epicurus’ epitomes (Letter to Herodotus, Letter toPythocles, and the lost Epitome of criticisms of the physicists; page and linenumbers refer to the Teubner text by P. von der Muehll).

!&% %. Epicurus, On nature

!#. Conclusions !&&

Chart !Nat. Epitomes, (i) methodological preliminaries Hdt. &$–# ()."–(.%)

(ii) nothing comes into being out of nothing Hdt. &# ((.%–&)(iii) nothing perishes into nothing Hdt. &" ((.)–()(iv) the all never changes Hdt. &" ((.(–")(v) the all consists of bodies and void Hdt. &"–)* ((.!*–!$)(vi) some bodies are compounds, others

constituents,, (vii) nothing exists independently except bodies

and void Hdt. )* ((.!$–+.%)(viii) bodies’ constituents (distinguished in (vi))

are atomic Hdt. )*–! (+.&–!*)(ix) the all is infinite Hdt. )!–% (+.!!–%*)(x) unimaginably, not infinitely, many atomic

shapes Hdt. )%–& (+.%!–$.#)(xi) atoms are in perpetual motion Hdt. )&–) ($."–#.%)(xii) there are infinitely many worlds Hdt. )( (#.!*–!+)(xiii) existence and mobility of images Hdt. )+–# (#.!$–!*.#)

,,,–,- (xiv) vision, visualisation, truth and falsity Hdt. )"–(% (!*."–!!.%*)(xv) the other senses Hdt. (%–& (!%.!–%*)(xvi) thought

- (xvii) atoms lack secondary qualities Hdt. ()–( (!%.%!–!&.!))(xviii) atomic dimensions Hdt. ((–" (!&.!(–!(.!+)(xix) up and down Hdt. +* (!(.!$–!+.()(xx) equal speed of atoms Hdt. +!–% (!+.+–!$.))

-,– (xxi) nature of the soul Hdt. +&–( (!$.(–!#.!*)-,,, (xxii) mortality of the soul Hdt. +(–+ (!#.!*–!+),. (xxiii) soul is not incorporeal Hdt. ++–$ (!#.!+–!".!%). (xxiv) metaphysics of properties and time Hdt. +#–$& (!".!+–%!.%!)., (xxv) origin of our world Pyth. ##–"* (%".$–&*.$)

(xxvi) origin of heavenly bodies Pyth. "* (&*.#–!))(xxvii) size of heavenly bodies Pyth. "! (&*.!)–&!.+)(xxviii) motions of heavenly bodies Pyth. "%–& (&!.+–&%.%)(xxix) attack on astronomical devices Pyth. "& (&%.%–$)(xxx) stability of the earth

.,, (xxxi) further astronomical phenomena Pyth. ")–# (&%.#–&).!%)(xxxii) other worlds Hdt. $&–) (%!.%)–%%.!#)

(xxxiii) origins of civilisation Hdt. $(–+ (%&.!–!$).,,, (xxxiv) the correct attitude to divinity Hdt. $+–#% (%&.!#–%+.!#)

[(xxxv) atmospheric and terrestrial phenomena] Pyth. "#–!!* (&).!&–)*.!().,- (xxxvi) critique of monism Epitome of

(xxxvii) critique of finite pluralism criticisms of the.- (xxxviii) critique of Anaxagoras }physicists

1234567 (

Lucretius’ plan and its execution

! . 526 5269,9

In Chapter &, I built my case for regarding Lucretius as an Epicureanfundamentalist, feeding directly on the writings of his school’s founder,and uninterested in pursuing the history of philosophy beyond thezenith which Epicurus himself represented. In Chapter ), I have puttogether a dossier on Epicurus’ major physical treatise, On nature, seekingto understand why it was at once the most demanding and the mostvalued of all Epicurus’ texts. Drawing on these findings, the presentchapter will defend the following account of Lucretius’ procedure whencomposing the De rerum natura.

Lucretius’ sole Epicurean source, I shall argue, was Epicurus’ Onnature, and, of that, mainly the first fifteen of its thirty-seven books.Initially he followed its sequence of topics very closely, indeed almostmechanically. But to some extent as he proceeded, and to a greaterextent during a phase of rewriting, he developed a radically revisedstructure for the whole. At his death, the reorganisation of DRN ,–,,, was(so far as I can judge) complete. For books ,-–-,, however, he had planswhich can still to some extent be discerned from his proems, but whichhe did not live to put into operation.

The Lucretian material of which I am speaking is the physical exposi-tion in the main body of all six books. I am assuming the remainder tobe his own original compositions – I mean in particular the proems, thepoetic manifesto at , "%!–(* (and ,- !–%(), at least the bulk of the magnamater passage (,, +**–+*), the ethical diatribes which close books ,,, and,-, and the concluding account of the Athenian plague, of which I shallhave something to say at the end of this chapter.

Even in the central physical exposition, on which I shall be concen-trating, it is no part of my intention to minimise Lucretius’ originalcontribution. Quite apart from his numerous distinctively poetic

!&)

achievements, no one need doubt for a moment that he has done muchto develop, illustrate, expand and sharpen the arguments as he foundthem. My contention concerns the bare bones of the exposition. I shallmaintain that these were lifted, more or less in their entirety, from Onnature books ,–.-.

Recovering this process of composition requires a good deal of e/ortand patience. But Lucretius’ creative achievement in structuring hispoem as we know it can only be appreciated when we see what materialhe started with, how he set about the task of reshaping it, and whatfurther plans may have remained unfulfilled at his death.

% . :A1765,A9 ’ 98A716

It used to be widely debated whether Lucretius’ source text wasEpicurus’ On nature, his Letter to Herodotus, his Great epitome, or somecombination of these, or, alternatively, whether he relied on more recentEpicurean writings.1 Lately these questions about sources have falleninto the background, no doubt to the relief of many. There are eventhose, like Diskin Clay, who maintain that Lucretius worked directlyfrom no source at all:2

There is suggestive, if not probative, evidence that Lucretius finally dependedon no written text or texts for the philosophy he expounded in De rerum natura.He made Epicurus’ philosophy his own and his preservitude to this thoughtseems to have made him free of any slavish attachment to a handbook surveyof Epicurus’ physiology. Epicureanism involved both a period of service to thetrue philosophy and an ultimate freedom.

As a picture of Epicurean education this portrayal has its attractions. ButI do not see how it can be correct in Lucretius’ own case. One half ofmy reply to it lies in Chapter &, where I argued that Lucretius is demon-strably dependent on Epicurus’ own writings, unmediated either bymore recent philosophical developments or by training in an Epicureanschool. The other half lies in the comparison, on which this chapter willbe focused, between the structure of Lucretius’ poem and that of Onnature. The data which I shall present seem to me intelligible only ifLucretius is taken to be working directly from a text of Epicurus.

". Lucretius’ source !&(

1 For surveys of these views, see Schmidt (!""*), pp. !%–%&, and, more selective, Erler (!"")), pp.)!)–!+. Schmidt in particular defends the view that more recent Epicurean writings play a largepart – a thesis originating largely from Lück (!"&%).

2 Clay (!"#&), p. &!. For Clay’s arguments against On nature as source, see ib. pp. !#–!".

!&+ $. Lucretius’ plan and its execution

Chart 'Nat. DRN, (i) methodological preliminaries

(ii) nothing comes into being out of nothing , !)"–%!)(iii) nothing perishes into nothing , %!(–%+)(iv) the all never changes —

(L!) the existence of the invisible , %+(–&%#(L%) the existence of void , &%"–)!$

(v) the all consists of bodies and void , )!#–%"(vi) some bodies are compounds, others constituents —

,, (vii) nothing exists independently except bodies and void , )&*–#%(viii) bodies’ constituents (distinguished in (vi)) are atomic , )#&–+&)

I &#$–*"+(ix) the all is infinite , "(!–!*(!

(L&) critique of geocentric cosmology , !*(%–!!!&(x) unimaginably, not infinitely, many atomic shapes (,, &&&–(#*)

(xi) atoms are in perpetual motion ,, +%–%!((L)) the swerve of atoms ,, %!+–"&(L() more on motion ,, %")–&&%

II ###–$(+(L+) compounds ,, (#!–$%"

II '#+–**+(xii) there are infinitely many worlds. ,, !*%&–!!$)

(xiii) existence and mobility of images (,- %+–%&#),,,–,- (xiv) vision, truth and falsity (,- %&"–)+#

(L$) refutation of scepticism (,- )+"–(%!)(xv) the other senses (,- (%%–$%!)

(xvi) thought (,- $%%–#%%)(L#) critique of biological teleology (,- #%&–($)(L") nutrition, motion, sleep, dreams, sex (,- #(#–!%#$)

- (xvii) atoms lack secondary qualities (,, $&*–""*)(xviii) atomic dimensions —(xix) up and down —(xx) equal speed of atoms —

-,– (xxi) nature of the soul ,,, ")–)!+-,,, (xxii) mortality of the soul ,,, )!$–!*"),. (xxiii) soul is not incorporeal —

IV "&–!"('. (xxiv) metaphysics of properties and time —., (L!*) mortality of our world - ((–)!(

(xxv) origin of our world - )!+–(*#(xxvi) origin of heavenly bodies —(xxvii) size of heavenly bodies (- (+)–+!&)(xxviii) motions of heavenly bodies - (*"–&&(xxix) attack on astronomical devices —(xxx) stability of the earth - (&)–+&

V $&%–&!#.,, (xxxi) further astronomical phenomena - +!)–$$*

(xxxii) other worlds —(xxxiii) origins of civilisation - $$!–!)($

.,,, (xxxiv) the correct attitude to divinity -, (*–"![(xxxv) atmospheric and terrestrial phenomena] -, "+–!%#+

.,- (xxxvi) critique of monism (, +&(–$!!)(xxxvii) critique of finite pluralism (, $!%–#%")

.- (xxxviii) critique of Anaxagoras (, #&*–"%*)

!""#""$

!#$

But which text? While it is quite understandable that Lucretian schol-ars should have got bored with Quellenforschung, it is surprising that noLucretian scholar has set out to reopen the question of Lucretius’ sourcematerials in the light of our constantly improving range of evidence onEpicurus’ On nature. Even while I was completing this book, valuable newdata on the contents of On nature were continuing to appear in print.3 Asthe fruits of Chapter ), we now have before us a reconstructed table ofcontents for On nature books ,–.- (Chart !, p. !&& above). Our next movemust be to place this table of contents alongside that of the De rerumnatura. I shall be referring to Chart % (p. !&+) repeatedly in the remain-der of this chapter: please insert a bookmark.

The analysis of On nature ,–.- combines the data gathered in Chapter). On the right, I show how the sequence of topics in the DRN matchesup with it. Wherever Lucretius includes a topic which we have not so farlocated in On nature, I have added this in the second column, marked‘(L!)’, ‘(L%)’ etc. There are ten such additions.4 Where Lucretius includesa topic which we have located in On nature, but in a di/erent position inthe sequence, I have signalled this with an arrow.

Despite these disparities, it leaps to the eye that there is an extensiveand non-accidental correspondence between the sequence of topics inOn nature and that in the DRN. Lucretius includes some topics not foundin the Letter to Herodotus, and the Letter includes some topics not found inLucretius, but that need reflect no more than two partly di/erent poli-cies used in excerpting material derived (directly or indirectly) from Onnature.

Moreover – and this is of the utmost importance – one of the dispar-ities of sequence can be shown to depend on a transposition whichLucretius himself made during the writing of the poem, after having ini-tially followed the exact order found in On nature. As we saw in Chapter%, §), the proem to book ,- preserves, side by side, two alternative pro-grammatic passages. The later of the two (%+–))) was written when thebook was expected to follow what we now call book ,,,, on the soul andits mortality. The earlier version ()(–(&), accidentally left in during theprocess of editing, preserves an earlier plan in which the book followeddirectly after book ,,.5 Thus originally Lucretius took items (xii) and (xiii)

". Lucretius’ source !&$

3 In Dirk Obbink’s magisterial new edition of Philodemus, On piety (Obbink (!""+)).4 I have excluded non-expository passages, such as the proem and the poetic manifesto at , "%!–(*.5 This was long ago shown by Mewaldt (!"*#), and has been widely accepted. Of later discussions

critical of Mewaldt, see esp. Pizzani (!"("), pp. !($–+$, and Gale (!"")b), pp. )–(. But their strong-est objection, that book ,- assumes some knowledge of book ,,,, is not decisive. It may simply

in their original sequence from On nature ,,, so that DRN ,-’s topic –images, perception and other soul and body functions – followed directlyafter that of the multiplicity (and impermanence) of worlds at the endof book ,,. It was only after he had begun drafting that sequence thatLucretius decided to postpone perception and related topics until afterwhat is now called book ,,,, on the soul and its mortality. Given, then,that one of the arrowed transpositions is demonstrably Lucretius’ own,there is a serious possibility that the others were also made by him duringthe process of writing.

In the case we have just examined, Lucretius’ original sequence – (xii),(xiii) – was identical to that in On nature ,,, and also, consequently, to thatin Letter to Herodotus )(–+. In principle, then, he could be following eitherof these texts,6 or indeed some third, unknown text which took over thesame sequence. The favourite candidate for this unknown text has longbeen the Great epitome (B"4+" 73 "&6$-/µ3" ),7 a work of which we know onlybecause it is cited three times in the scholia on the Letter to Herodotus.Before proceeding further, it will be best for me explain why I believe thatOn nature is itself the unmediated source.

The hypothesis that the Letter to Herodotus may have been Lucretius’primary source o/ers few attractions. It is certainly not enough to pickout the occasional phrase from it which Lucretius can be said to havedirectly translated, since there need be little doubt that any such phraseoccurred in On nature too.8 The hypothesis must of course assume thatLucretius supplemented the Letter with a great deal of other Epicureanmaterial. But it is not easy for its proponents to explain why Lucretiusshould have so closely matched the sequence of On nature even at pointswhere the Letter o/ered no guidance. For example, a look at Chart % will

!&# $. Lucretius’ plan and its execution

show that Lucretius became aware during the course of writing book ,- that he would eventuallyneed to transpose it – see p. !(* below. And the findings of Ch. %, §), where we saw how the firstversion contained a translation of technical terminology alien to Lucretius’ eventual method,lend new confirmation to Mewaldt’s conclusion that this represents an early draft of the poem.In defence of Mewaldt against the alternative proposal of Drexler (!"&() that Lucretius’ changeof mind was the other way round – to move book ,- to a position after book ,, – see Ferrari (!"&$).

6 Those who have supported On nature as primary source include Mewaldt (!"*#), von der Muehll(!"%%), pp. ,,,–,-, Boyancé (!"+&), pp. (&–+, and Arrighetti (!"$&). The case for the Letter to Herodotusas primary source, whose groundwork was laid by Woltjer (!#$$) and Brieger (!##%), has mostrecently been urged by Fowler (!""+). Cf. also Asmis (!"#&), on which see n. (, p. !"& below.

7 That the Great epitome is Lucretius’ main source was the proposal of Giussani (!#"+–#), , !–!!.8 E.g. Bailey (!")$), , %(: ‘as will be seen from time to time in the commentary, his relation to it [the

Letter to Herodotus] is so close that it is almost impossible to resist the conclusion that he was trans-lating it’ (although Bailey does not make it Lucretius’ primary source). For an e/ective reply to thisinference, see Boyancé (!"+&), p. (( n. !.

show that he largely preserves the order of topics in On nature .,–.,,,despite the fact that neither the Letter to Herodotus nor, for that matter, theLetter to Pythocles could have o/ered su0cient reason to do so. Neitherletter includes item (xxx), yet Lucretius has preserved its original placein the sequence, as the final columns of On nature ., confirm.

Stronger evidence against the Letter to Herodotus as source is forth-coming from the end of On nature ,,. This includes perhaps the one pointin the surviving portions of the treatise which it is possible to correlateprecisely with the corresponding lines of Lucretius. Discussing images,Epicurus marks his transition from the preceding topic of their speed ofgeneration to that of their speed of travel as follows:9

6"#[$! ] 5"! -3' ) ,+-+! -3! . %/#+! . &$ 6+#;/&" '3) -+;&-3' -[/)] .&' . 7""4"$."&6$;["$#]3" '/µ".. 6#(' -/. µ["! ].. 4[+! #] 3$ 7"6-/" -3) µ+,#+! . -3' ) +& 6/! -(' .+$&'. 13" '"(. 7"6-/" -3-/) +& [6"";/&]'+, -+;&-3' -+ -(' . "$&5(" 7(. ,+-+! -[3! .%/#+! ]. +& .&6""#2[73-/. "& ].5"$",.&-+$ [ . . . (lacuna c. #$ words) . . .&$ 6"#2+77/" .]-() ,/&' %+. "$& 5 & [&$ ]6"#2+77/" .-([) ,/&' ]%+, 53' 7/. ($ ) ,+$!&$ 6"#2+77/" .-() -+;"$'+ ,+-+! -3! . [%]/#+" ..

Now we will try to speak about the speed which belongs to them in respect oftheir motion.

First, their fineness, which far exceeds the fineness revealed by the senses,indicates the images’ unsurpassed speed of motion [. . . (lacuna c. #$ words) Thatfineness10 makes them] exceedingly light, and if they are exceedingly light it isclear that they are exceedingly fast in respect of their motion.

The identical transition from the images’ speed of generation to theirspeed of motion is made by Lucretius at ,- !$+/., in lines which perfectlyillustrate his talent for putting flesh on the bare bones supplied byEpicurus:

nunc age, quam celeri motu simulacra feranturet quae mobilitas ollis tranantibus aurasreddita sit, longo spatio ut brevis hora teratur,in quem quaeque locum diverso numine tendunt,suavidicis potius quam multis versibus edam; !#*

". Lucretius’ source !&"

29 %).&+–$ Arr.2!Vogliano (!"(&), p. #!, fr. E cols. ,-–-. Arrighetti indicates, p. ($", that his read-ings of this part of P.Herc. !!)" are drawn from previous editions, not from an autopsy. Yet,strangely, he reports a number of letters as missing or uncertain which in Vogliano’s text appearcertain – even some which show up clearly in the early nineteenth-century facsimiles. I havetherefore, despite some puzzlement, reverted to Vogliano’s report of the papyrus, which, fortu-nately, does not di/er in its actual reconstruction of the text. This is one of the Herculaneumpapyri of which we still await a truly modern edition, and it remains quite likely that the even-tual text will prove to di/er in some respect.

10 Since the full proof of the images’ fineness can hardly have been contained in the short lacuna,Epicurus must be relying on an earlier such proof, for whose location in the sequence see Ch. ), §$.

parvus ut est cycni melior canor, ille gruum quamclamor in aetheriis dispersus nubibus austri.principio persaepe levis res atque minutiscorporibus factas celeris licet esse videre.in quo iam genere est solis lux et vapor eius . . . !#(

Come now, how fast is the motion with which the images travel,and what mobility they possess in swimming through the air, sothat it takes them but a short time to go a long way, whateverdirection they may be tending in with their varying inclination,I shall explain in sweet-speaking rather than numerous verses,just as the short song of the swan is better than that honking ofcranes which spreads through the airy southern clouds.

First, you can often see that things made out of light and tinyparticles are swift-moving. Among such things are the light andheat of the sun . . .

When these two passages are placed side by side, it looks entirely cred-ible that we have before us the actual text from which Lucretius wasworking. They both announce their demonstrandum, and proceed to theirfirst argument (6#(' -/. µ"" .!principio)11 in an almost identical manner.Lucretius adds the swan comparison, declaring his determination tobrighten up the argument. And his diction – both when sketching thedemonstrandum itself and when elaborating the arguments to which henow proceeds – may be judged to fulfil that promise. In the proem to ,,,Epicurus was, philosophically speaking, the swan, Lucretius a mereswallow (see Ch. %, §!%). This time however, when it is the literary prizesthat are being handed out, Lucretius casts himself as the swan. Who, bycontrast, are the cranes, whose interminably raucous honking is so subtlycaptured by the cacophonous enjambment at !#!–%? Lucretius may havein mind the earlier Roman prose writers on Epicureanism – such asCatius, whose own writing on visual images provoked such derision (seeCh. %, §)) – and perhaps even Epicurus himself.12 Lucretius thus displayslittle modesty about his own literary superiority within the Epicurean

!)* $. Lucretius’ plan and its execution

11 One prominent feature of Lucretius’ method is the clear labelling of successive arguments, prin-cipio, deinde, praeterea etc. Epicurus can be seen at %).)% Arr.2 to make his own transition betweenseparate arguments for the speed of images with the words ‘Again, then, it becomes evident thatthe images have unsurpassed speed in their motion. But also in the following kind of way it willbe possible to produce a proof about the speed of images. For since . . .’ This clear division isevidently what underlies Lucretius’ organisation of his own arguments. Epicurus’ greaterexpansiveness is a feature of the lecturing style.

12 There is no need to insist that the isolated Greek-derived word cycnus is this time a consciousGraecism. If, on the other hand, we suppose that the Greek–Roman contrast of the proem tobook ,,, (described in Ch. %, §!%) is being revived here, then Lucretius is claiming for himself aGreek literary mantle (see Ch. !) which has bypassed Epicurus himself.

tradition. The philosophical hard core of the passage is itself neverthe-less visibly being supplied by Epicurus.13

More important still, in both texts the argument in question belongsat the end of exactly the same sequence. As we noted in Chapter ), §$,Epicurus’ sequence in the final part of On nature ,, is:(a) existence of images,(a)* their fineness,(b) their speed of generation,(c) their velocity.This is identical to the sequence which we find in Lucretius ,- ()–%!+.14

But we also saw there that in this particular case the Letter toHerodotus ()+–#) has a slightly changed order(a) the existence of images,(c) their velocity – including (a)*, their fineness, merely as a premise,(b) their speed of generation.Here then there can surely be no possible doubt that Lucretius’ debt toOn nature has not been mediated by use of the Letter to Herodotus.15

Thus in a passage where, for once, we can administer a direct test, theLetter to Herodotus is ruled out as Lucretius’ source. This strengthens theimpression of its irrelevance to our task, and drives us back to a choicebetween On nature itself and the Great epitome. If the mysterious Greatepitome is assumed to have been su0ciently great, in principle almost any-thing found in On nature could have recurred in it. But it seems an awfullyimplausible candidate for Lucretius’ source text. This title (a) does notfeature among those which Diogenes Laertius catalogues as Epicurus’leading works (. %$–#) or as the sources for his physics (ib. &*, where helists not only On nature but also, implicitly, the Letter to Herodotus and Letterto Pythocles);16 nor (b) is it cited by any ancient source apart from the

". Lucretius’ source !)!

13 Contrast Clay (!"#&), pp. !#–!". His objection there to taking Nat. as Lucretius’ source – includ-ing for this passage – is ‘There is no hint of translation or paraphrase.’ But why should we haveexpected there to be?

14 Cf. Barigazzi (!"(#), p. %(). The denial of this fact by Lackenbacher (!"!*) p. %&% is based on asimple mistake. He thinks that Lucretius’ first argument for (c) is founded on the images’ ‘light-ness’, while Epicurus’ first argument for (c) is founded on their ‘fineness’ (7"6-/" -3)). But theseare both the same property! Lucretius’ first argument refers (,- !#&–)) to levis res atque minutis cor-poribus factas, and that this latter expression is equivalent to ‘fineness’ for him is made plain at!!*–!(.

15 Also, as Mieke Koenen has kindly pointed out to me, at Nat. ,,, %).!%.! Arr.2, [+& .]+. ,7+" '"$) is evi-dence that in Nat. Epicurus has cited the phenomenon of reflection earlier on in his argumentsfor "$%5(7+. There are corresponding discussions of reflection in Lucretius at ,- "#–!*", !(*–+$.He could have got these from Nat., then, but not from Ep. Hdt. )+–#. 16 Quoted p. !!* above.

scholiast on the Letter to Herodotus; nor (c) has any text that could plausi-bly be identified with it turned up in the library of Philodemus.

By contrast, as we learnt in Chapter ), On nature (a) was the mostprominent of all Epicurus’ works, (b) is by far his most widely cited workon physics, and (c) can be seen from the Herculaneum library to havebeen collected and valued by Epicureans in Lucretius’ own day. In viewof all this, the common impression that it was virtually unreadable,generated by the woeful state of the papyri, cannot be correct (see Ch.), §(). Moreover, even allowing – as I prefer to insist (see Ch. &) – thatLucretius is unlikely to have been a member or active associate ofPhilodemus’ school, we at least have incontrovertible evidence thatcopies of On nature were obtainable in Italy. Lucretius had to get his owntexts of Epicurus from somewhere, and the papyri of On nature whichsurvived at Herculaneum may well include some of the exemplars fromwhich his own copy was made. There was surely no better archetype ofEpicurus available in Italy than Philodemus’ ancient copies of On nature,imported from the Garden in Athens.

These arguments are circumstantial. But I think it is possible to gofurther. One thing that would not easily have survived into an epitome,however large, is the quirks of personality which Epicurus displayed inthe lectures On nature. In Chapter ), §(, where I said a little about per-sonal touches in On nature, one example was Epicurus’ measured way ofdealing with his illustrious and not so illustrious predecessors. In Onnature ..-, he says of the early atomists’ failure to see the consequencesof determinism17

/. .$$ 5& +$&-$/7/43" '+.-") "& @ +& #;3' ) $$,+.(' ), ,+$! /&& µ[/" .]/... [-](' . 6#/! +&& -(' .6/7&! 5$".""4,+.-") +& 77+! ,+$! -(' . &+ '-"#/. 6/77+67[+]'$" [()], "%7+1[/]."$+&-/&" ), ,+$"6"# "& . 6/77/$' ) µ"4+" 7+ ,/&%$"'+.-"), "[$& ]). -/! -[3! ]. +& .+" 4,3. ,+$!-+&& -/" µ+.-.[/]. 6+" .-+ +[$&-]$+' '1+$.

The first men to give an adequate account of causes – men greatly excelling notonly their predecessors but also, many times over, their successors – although inmany matters they alleviated great problems, turned a blind eye to themselvesin order to blame everything on necessity and accident.

It would be hard not to notice how closely this stylised mixture of praiseand criticism is echoed at Lucretius , $&)/., where Empedocles and his‘lesser’ followers are likewise gently rebuked:

!)% $. Lucretius’ plan and its execution

17 The passage is most conveniently cited as Long/Sedley (!"#$), %*1 !&. The interpretation ofµ"4+" 7+ ,/&%$"'+.-") is due to Arrighetti (!"$"). The reading "[$& ]). is uncertain, but discussion ofit is unnecessary here.

hic tamen et supra quos diximus inferiorespartibus egregie multis multoque minores, $&(quamquam multa bene ac divinitus invenientesex adyto tamquam cordis responsa dederesanctius et multo certa ratione magis quamPythia quae tripodi a Phoebi lauroque profatur,principiis tamen in rerum fecere ruinas $)*et graviter magni magno cecidere ibi casu.

But this man and the greatly inferior and far lesser ones whomI mentioned above, although in making their many excellentand godlike discoveries they gave responses, as from the shrineof the mind, in a holier and much more certain way than thePythia who makes her pronouncements from Apollo’s tripodand laurel, nevertheless came crashing down when they dealtwith the elementary principles of things. Great as they were,their fall here was a great and heavy one.

Leucippus and Democritus were ‘many times over’ (6/77+67[+]'$" [()])superior to their successors; Empedocles’ successors were inferior tohim, literally, ‘by outstandingly many degrees’ (partibus egregie multis, $&().Leucippus and Democritus solved great problems; Empedocles and hissuccessors made many excellent discoveries. Yet both parties made cat-astrophic judgements concerning the issue in hand.

There is at least one clear trace of a similar comparative critique in anearlier book of On nature.18 It seems to be a style of judgement character-istic of this lecture course. Lucretius’ comment on Empedocles and hisfollowers thus has Epicurus’ unmistakable fingerprint: in all probabilityhe is directly echoing Epicurus’ criticisms of Empedocles from book .,-,characteristically qualified by praise in the form of a historical ranking.

Another such echo is suggested by the same passage from On nature..-. There Epicurus goes on to single out Democritus, not by name, butjust as -/! . +% .5#+, ‘the great man’. This term of respect is reflected inthe later Epicureans’ use of /$$ +% .5#"), ‘the Great Men’, to designate thefour authoritative founders of their school (pp. +$># above). Once again,Epicurus’ own usage is mimicked by Lucretius, who on both occasionson which he cites a doctrine of Democritus refers to it as Democriti quodsancta viri sententia ponit.19

". Lucretius’ source !)&

18 Nat. .,, K ,- + [M] ,, Vogliano (!")*) p. )"!%+.)) Arr.2: those who have explained the earth’sstability by reference to air ‘we should consider no better than these men’ (i.e. those who appealto symmetry alone, e.g. the Platonists) ‘in many matters’ ("& . 6/77/$' ) again) – ‘but in manymatters utterly, comprehensively, vastly better’ ("& . 6/77/$' ) [5"! ,]+$! µ+" 7$'-+ -(' $ /+ 7($ -#/" 6($6/77(' $ 2"7-$"/&)). 19 ,,, &$!, - +%%.

It is hard to believe that these personal touches in the comparativeevaluation of predecessors, highly visible in Epicurus’ expansive lecturecourse, had so clearly shown through in a condensed handbook such asthe Great epitome must have been.

Of course Lucretius is selective. He has omitted many topics andarguments altogether, and has cut down considerably on the largeamount of polemical content that shows through in the fragments of Onnature. His doing so is fully accounted for by (a) the need to fit the mainargument of fifteen books of On nature into six books of Latin hexame-ters, and (b) the lower levels of philosophical expertise, and perhaps eventolerance, to be expected of his Roman readers. Nevertheless, even inwhat remains the imprint of On nature shows through clearly not only inthe philosophical content of Lucretius’ poem but also in its style.

I conclude that to posit On nature itself, rather than some derivative ofit, as Lucretius’ source is not only the most economical hypothesis, butalso the one which can best explain our data. I do not, of course, denythat for his Epicurean ethics – which he knew and understood extremelywell – he had read more widely. Nor is it capable of demonstration thathe never supplemented On nature with other works of Epicurean physics.But I know no feature of his poem which is better explained by thathypothesis, and I shall argue below (§$) that at least one feature of it,namely Lucretius’ actual method of composition, is on the contrarybetter explained by the supposition that he relied exclusively on the full-length treatise.

& . 526 9 57A15A76 8@ D E R E RU M NAT U R A

As we have already noted, it was only during the process of writing thatLucretius adopted the transposition which put books ,,, and ,- into theirpresent order. This decision to depart from Epicurus’ own order was acrucial factor in producing what now stands out as the carefully balancedsix-book structure (Chart &).

The structural features of this arrangement are very well recognised,and do not need elaborate rehearsal here. Notable aspects include thesequence of three pairs of books, expanding from (A) the microscopiclevel, through (B) the human level, to (C) the cosmic level. Within eachpair, the first book sets out the nature and lifespan of the items in ques-tion – (,) atoms, (,,,) human soul, (-) world – and the second goes on toaccount for a range of phenomena related to those same items. Thepoem also falls into two pairs of three books, each ending on the theme

!)) $. Lucretius’ plan and its execution

of death (,,,c, -,c), to counterbalance the opening celebration of birthand life in ,a. All this, and much more, hung on Lucretius’ decisionduring the course of writing to move book ,- to its present position.

The remaining transpositions cannot so decisively be shown to beLucretius’ own, rather than already imposed by the author of somehypothetical intermediary source on which Lucretius might be imaginedto rely. But now that we have acknowledged the principle that he didtranspose material during the process of composition, it becomes moreplausible (as well as economical) to suppose that these are his too.

) . '88E9 ,> , , ,

A good example is the shifting of (xxxvi)–(xxxviii), the refutation of rivaltheories of matter, from its late position in books .,-–.- of On nature tothe place it now holds in book , of Lucretius (+&(–"%*). The apparentmatch in the structure of the two passages (monism, finite pluralism,Anaxagorean homoiomereia) is encouraging evidence that On nature .,-–.-

%. Books I–III !)(

Chart *, (a) Proem: praise of Venus as life force, and of Epicurus for

discovering the nature of the infinite universe(b) The basic elements(c) The infinity of the universe(A) Atoms { ,, (a) Proem: the Epicurean good life(b) Microscopic and macroscopic properties(c) The multiplicity of worlds

,,, (a) Proem: praise of Epicurus for freeing us from fear of god and ofdeath

(b) The soul and its mortality(c) Diatribe against the fear of death(B) Man { ,- (a) Proem: Lucretius' poetic mission(b) Perception and other soul/body functions(c) Diatribe against sexual passion

- (a) Proem: Epicurus as the greatest god(b) The world and its mortality (including astronomy)(c) The origin of life, civilisation and religion

(C) World { -, (a) Proem: praise of Athens for its greatest gift to civilisation,Epicurus

(b) Cosmic phenomena(c) The Athenian plague

was Lucretius’ source. As we have seen (pp. !%(>+ above), this wholestructure, including the homoiomereia interpretation of Anaxagoras, orig-inally reflected Epicurus’ use of Theophrastean doxography. That theLucretius passage is shaped by the same doxographical approach is wellrecognised.20 For example, at , +)$/. Lucretius hypothetically attributesto Heraclitus a theory of rarefaction and condensation of fire whichappears to have entered the tradition only with Theophrastus.21

Epicurus had held over his critique of rival theories of the elementsuntil he had completed his own physical and cosmological exposition inbooks ,–.,,, of On nature. The decision to bring the critique forward tooccupy a much earlier place, immediately after the initial demonstrationof atomism in the first part of book ,, is almost certainly Lucretius’ own.From a methodological point of view the critique now comes surpris-ingly early. I shall be examining this aspect fully in Chapter $. For presentpurposes I shall concentrate on just one of the main advantagesachieved by the transposition.

The transfer of the critique to this early position helps Lucretius toengineer a major structural feature of the first pair of books (see Chart& above). It enables him to postpone until the end of book , the theme(,c) of the universe’s infinity and the absurdity of the alternative,inward-looking cosmology which constructs our world around anabsolute centre.22 This horizon-expanding motif is then mirrored at theend of the second book (,,c), where Lucretius argues for the existenceof other worlds and (an aspect of the argument not brought out at (xii)in the Letter to Herodotus condensation) the transient nature of our own.The emphasis achieved by this pair of matching closures delivers whathe has e/ectively promised since the proem (, +%–$"): a liberating intel-lectual journey with Epicurus in which we will push on through the con-stricting barriers of our own world and into the infinite universebeyond.

!)+ $. Lucretius’ plan and its execution

20 Rösler (!"$&), Mansfeld (!""*), pp. &!(&–).21 Theophrastus, Phys. op. fr. ! Diels (!#$"), p. )$(.!)–!#,!fr. %%( FHS&G (!Fortenbaugh et al.

(!""%)), p. )*+.!(–!".22 Barigazzi (!"(*), in a very valuable study of (xxx), the argument about the earth’s stability at the

end of Nat. .,, points to links between it and the critique of geocentrism in (L&). However, hedoes not explicitly suggest that Lucretius has brought this material forward from (xxx) to (L&),and thinks it likely, as I do too, that (L&) was already in place in Nat. ,, adjacent to (ix), the infin-ity of the all. The main thrust of (L&) is that space cannot have a centre or any other structuralor causal properties (see Ch. &, §+); and that is already justified by the arguments which precedeit – (v), which includes the total non-resistance of void at , )&*–)), and (ix). The target in (xxx)may conceivably include the same people (see however Panchenko (!"")), pp. )$–((), but therethe issue is the stability of the earth, not the structure of space as such.

There are other signs of Lucretius’ intervention in the sequence oftopics in book ,,. One relates to items (xii), the multiplicity of worlds, and(xvii), that atoms lack secondary qualities. These became consecutivetopics as soon as the intervening material on perception etc. had beentransferred to book ,- (see the second descending arrow in Chart %).23

Lucretius then decided to reverse their order. This was his way of post-poning the topic of the plurality and transience of worlds until the endof book ,,, producing what we have just seen to be the appropriate matchbetween its closure and that of book ,.

Another case is the atomic ‘swerve’. This theory is not yet present inthe Letter to Herodotus account of the causes of atomic motion ()&–)), andtherefore was very probably introduced only in a later book of On nature.I suggested in Chapter ), §!! that this was either in book ..-, where theissue of determinism was visibly in focus, or in a book flanking it. Book..-, we have also seen (Ch. ), §§& and !!), was apparently one of thoseespecially prized by Epicureans. Hence it should be no great surprise insuch a case that Lucretius for once drew on material from the later partof Epicurus’ treatise. For that is what he seems to have done. He hasbrought the swerve thesis forward to an early position in book ,, (L)!,,%!+–"&), where it quite properly takes its place as a third cause of atomicmotion alongside impacts and weight. Furthermore, this has in turnrequired Lucretius to anticipate at the same early point item (xx), thetopic of the equal speed of atoms, needed as a premise in the argumentfor the swerve (,, %%(–&").24

I do not in this context want to discuss the theoretical merits or demer-its of these transpositions, from the point of view of Epicurean physicalmethodology. What they illustrate for my present purposes is how books,–,, are the product of delicate restructuring by Lucretius of Epicurus’original material: I shall be discussing the nature of this transformationin Chapter $. I see no sign anywhere in books ,–,,, that the restructuring

%. Books I–III !)$

23 The original proem to ,- ()(–(&), which represents a very early stage in the composition of thepoem and precedes Lucretius’ decision to move the book’s material to its present position (see n.(, pp. !&$># above, and §( below), does not make it explicit that item (xvii) was already at thetime of writing located in book ,,. But I assume that at any rate once the transposition had beendecided on – quite early on in the composition of the book, see p. !(* below – it became clearto Lucretius that any further material on atoms in (xvii)-(xx) must either be incorporated intobook ,, or simply omitted. In fact he chose to omit most of it.

24 This is not arrowed in Chart % as a transposition, because the topic is not covered in its own rightat ,, %%(–&". He only refers there to the equal speed of atoms moving vertically downwards ininter-cosmic space. The main topic of (xx), that of the equal speed of atoms at all times, evenwithin compounds, is simply omitted by Lucretius.

was not fully carried out (although in the case of book ,,, there is no clearevidence that Lucretius actually had to make any structural change,apart from his omission of (xxiii) on the metaphysical issue ofincorporeality).25 So far as I can tell, apart from corruptions acquiredduring later transmission, the first three books may well be exactly asLucretius would have intended them to be.26

( . '88E ,-

When we turn to the last three books, the picture is very di/erent. Thestate of book ,- is the prime item of evidence. I have already mentionedthe transitional state of its proem, where the original programmatic lineshave accidentally remained in the text. The later version of these linesdi/ers from the earlier one in some very revealing ways. By comparingthe two versions we saw, in Chapter %, §), how Lucretius overhauled hisannounced range of vocabulary for eidola. Here, however, I want to con-centrate on just one aspect: his preview of the book’s contents. Unlikethe original version, the revised version tells us emphatically that thebook’s main purpose is to dispel the fear of ghosts (,- %+–)!):

atque animi quoniam docui natura quid essetet quibus e rebus cum corpore compta vigeretquove modo distracta rediret in ordia prima,nunc agere incipiam tibi, quod vementer ad has resattinet, esse ea quae rerum simulacra vocamus; &*quae quasi membranae summo de corpore rerumdereptae volitant ultroque citroque per auras,atque eadem nobis vigilantibus obvia mentesterrificant atque in somnis, cum saepe figurascontuimur miras simulacraque luce carentum, &(quae nos horrifice languentis saepe soporeexcierunt; ne forte animas Acherunte reamure/ugere aut umbras inter vivos volitareneve aliquid nostri post mortem posse relinqui,cum corpus simul atque animi natura perempta )*in sua discessum dederint primordia quaeque.

!)# $. Lucretius’ plan and its execution

25 If Lucretius’ argument in ,,, is thought to acquire its structure largely from doxography – asargued by Mansfeld (!""*), pp. &!)&–(%, although I do not myself find the grounds he o/ers par-ticularly compelling – the doxographical influence would be likely to have come throughEpicurus from Theophrastus’ Physical opinions (cf. §# and Ch. + below). That would suggest thatLucretius has in the main retained Epicurus’ own sequence.

26 If there are any residual signs of lack of a final revision in these books, they concern no morethan local fine tuning (e.g. perhaps ,,, +%*–!, #*+–!#: see Kenney (!"$!), ad loc.), not the overallarrangement of material.

And now that I have taught you what is the nature of the mind,from what things it gets its power when combined with the body,and how when torn apart it returns to its elements, I shall nowbegin to deal with what is closely relevant to this: how there arewhat we call images of things, which, like membranes snatchedfrom the outermost part of things’ bodies, fly hither and thitherthrough the air; and how these same things strike us both awakeand asleep and terrify our minds, when, as often, we see strangeshapes, and images of those who have passed on – which haveoften woken us in terror as we lay slumbering; lest we should think,perhaps, that souls are escaping from Acheron, or that shades flitaround among the living, or that something of us can survive afterdeath, when both the body and the nature of the mind have beendestroyed and dissolved, each into its own elements.

The main function of book ,- within the middle pair of books is thuslaid bare. Book ,,, has shown that the soul is mortal and death not to befeared. Book ,-’s account of psychic functions will complement this byshowing that encounters with ‘ghosts’ are not evidence that somethingof us does after all survive death. That this was to be the central messageof book ,- is confirmed by the proems to books , and - (, !&%–(, - ("–+&),both of which emphasise precisely the same role for book ,-. Yet whatwe actually find on the topic of ghosts in book ,- is a mere !! lines($($–+$). None of the important questions is addressed. Are the imagesof the dead which invade our dreams ones which emanated from thosesame people before they died, even centuries ago? Or are they imageswhich our minds pick out merely because they bear some resemblanceto those people? And how are waking visions of ghosts – referred toexplicitly in the proem – to be explained? These are important questionsfor an Epicurean to be able to answer. As far as I know, the only explicitevidence – albeit from the virulently hostile Plutarch27 – attributes to theEpicureans the belief that images can stay in circulation even long afterthe death of the people from whom they emanated. But this at the veryleast needed saying, explaining and justifying. Above all, Lucretius oweshis readers a well-reasoned assurance (which was certainly forthcomingfrom other Epicureans, and almost certainly from Epicurus himself)28

$. Book IV !)"

27 Epicurus fr. &") Usener!Plut. De def. or. )%*'–1.28 Diogenes of Oenoanda !* - %–+ Smith. Since the point is argued there as a disagreement with

Democritus, it probably stems ultimately from Epicurus himself. Moreover, there is excellentreason to suppose that this disagreement was already to be found in Lucretius’ source text, sinceit is likely to be what motivates the criticism of Democritus reproduced by Lucretius at ,,, &$*–"(,for making soul-atoms and body-atoms alternate throughout the living organism: only on suchan account could atom-thin films from the body be thought themselves to have soul.

that such images could not actually be alive. Extraordinarily, Lucretiusdevotes more lines in his proems to announcing that book ,- will explainghosts than he devotes in book ,- to actually explaining them. It seemsself-evident that book ,- is not in the final state which Lucretius envis-aged for it at the time when he wrote its revised programme of contents.

There is good reason to suspect that before he had advanced very farwith the writing of what is now book ,- Lucretius came to realise that hewould eventually be moving it to come after the account of soul. As earlyas ,- !%! there is an explicit reference to anima and animus, presupposingreaders’ familiarity with the technical distinction between these devel-oped in book ,,,.29 But it seems equally clear to me that he had not, atthis stage, worked out the pivotal role that ghosts would eventually haveto play in the book: that intention is acknowledged only in the proems,which can therefore be assumed to represent a very late stage inLucretius’ planning.

As for the actual contents of book ,- as we have it, there is reason to guessthat they still closely reflect Epicurus’ original sequence of material,without the benefits of Lucretius’ planned reworking. The list, excerptedfrom Chart %, is:

Chart 'a(xiii) existence and mobility of images ,- %+–%&#(xiv) vision, truth and falsity ,- %&"–)+#

(L$) refutation of scepticism ,- )+"–(%!(xv) the other senses ,- (%%–$%!

(xvi) thought ,- $%%–#%%(L#) critique of teleology ,- #%&–($(L") nutrition, motion, sleep, dreams, sex ,- #(#–!%#$

Although the position of book ,-, following book ,,, on the soul, and like-wise its primary content, encourage the impression that it is concentrat-ing on mental phenomena, the addition of (L#) and (L") appears tointroduce an amorphous mixture of soul and body functions. Neitherthe critique of biological teleology – where we are urged not to believethose who say that legs were made for walking with, etc. – nor theensuing account of nutrition, has anything directly to do with theEpicurean soul. Puzzlingly, as F. Solmsen noticed,30 what links the itemslisted is, if anything, that they are all functions of the Aristotelian soul.Even here, however, the critique of teleology fits such an account loosely

!(* $. Lucretius’ plan and its execution

29 Much later, at ,- #$$–"*+, the distinction is even more clearly presupposed and exploited.30 Solmsen (!"+!a).

at best. Attempts have been made to find an equally good rationale fortheir grouping based on principles purely internal to Epicureanism, butwith only limited success.31 I hope that the following conjecture is animprovement.32

The cardinal rule in the second half of the poem is this: if you wantto make sense of a puzzling sequence of topics adopted by Lucretius,since he demonstrably has not completed his own reorganisation of thematerial, ask why Epicurus should have ordered it in this way. To answersuch a question with regard to book ,-, we must consider the above listin its full context on Chart % (p. !&+ above). It is actually quite easy to seewhy, having explained perception and thought in On nature ,,,–,-,Epicurus should have continued with the remaining vital functions in(L"). Significantly, the very next topic was to be(xvii) Atoms lack secondary qualities,and this, in Lucretius’ version (much fuller than Letter to Herodotus ()–(),includes some proofs that atoms lack vital properties (,, #+(–""*).Clearly the full discussion in On nature set out to show systematically thatatoms lack not only all secondary sensible properties (colour, taste etc.)but also all vital properties. Hence Epicurus felt the necessity to analysethe full range of both sensible and vital properties before turning to(xvii), showing in particular how vital properties always depend oncomplex structures and processes which single atoms cannot possess. Asfor (L#), the critique of biological teleology, while its tenuous relevanceto psychological properties and functions makes it look curiously out ofplace in Lucretius ,-, it would have been entirely at home in its originalEpicurean context. Any full explanation of vital functions was, forEpicurus, likely to include a vehement rejection of Plato’s elaborateaccount of these in the Timaeus, where the organs in which they arelocated had been described as providentially produced for those veryfunctions by divine creators.33

Epicurus’ plan, it seems, was not to turn to the analysis of soul itself– item (xxi) on his agenda – until the nature of atoms had been fully

$. Book IV !(!

31 Cf. Furley (!"+$), p. %!& for criticisms of Giussani and Bailey’s explanations, although Furley’sown proposal – ‘The whole passage from $%% to "+! might be entitled: “No need for any explana-tion other than simulacra”’ – does not fit ,- #(#–$+ or "*$–+! very comfortably (cf. Schrijvers(!"$+), p. %&%).

32 One simple explanation for this unexpectedly Aristotelian thematic link would be the hypothe-sis that Epicurus was himself working from Theophrastus’ Phys. op., which as an Aristoteliandoxographical work naturally grouped its topics on Aristotelian principles. I have made thissuggestion in Sedley (!""$a). It is compatible with the proposal which I o/er here, but is notrequired by it.

33 For teleology in the Timaeus as Epicurus’ (and derivatively Lucretius’) target, cf. Ch. &, §(.

sorted out, in order that it should be incontrovertible that the vital prop-erties of soul depend on its complex structures and processes, not on itscomponents. It is significant here that in the great majority of thePresocratic theories to which Epicurus was reacting vital properties werealready irreducibly present in the basic stu/ or stu/s of the world. Theconcern to combat this supposition, in favour of his atomistic bottom-up model, was clearly an overriding factor in Epicurus’ organisation ofhis physical exposition.

Lucretius’ eventual decision, in the interests of his poem’s overallarchitectonic, was to analyse soul in book ,,, before turning to the indi-vidual vital properties in ,-. It is this reversal of Epicurus’ expositoryorder that accounts for the otherwise puzzling heterogeneity of theissues covered in the later part of book ,-. We may recall that, on theclear evidence of the proems, Lucretius had plans for re-focusing thecontent of book ,-. We need not doubt that these would have includedan enhancement or clarification of its thematic unity.

+ . '88E -

Given what we have learnt about Lucretius’ intentions in book ,-, it isworth asking whether the proems reveal any other unfulfilled plans onhis part. Remarkably, it will turn out that they do. But let me approachthe point indirectly.

The sequence of topics in book - is curious. Lines (*"–$$*, on astron-omy, constitute a surprising interruption between two phases of thehistory of the world, coming as they do after the development of thecosmos itself but before the emergence of life and civilisation. Onceagain there is good reason to attribute the sequence not to any concernof Lucretius’ own, but to one which may well have motivated Epicurus,namely the need to respond to the account of creation in the teleologists’bible, Plato’s Timaeus. For there the motions of the heavens and earthoccupy a closely analogous position, between the origin of the world(%#b-&)a) and the origin of mankind ()!a / .), being integral to the inter-vening construction of the world soul. One can imagine that Epicurus,in developing his own account of the world’s origin in On nature .,–.,,,felt the need to respond point by point to the Timaeus.

This anti-Timaean motive is in fact borne out by a closer look at item(L!*) on Chart % (p. !&+ above), the mortality of the world. (There is noreason to doubt that this topic occurred at just the same position in Onnature, even if we happen to have found no independent attestation of

!(% $. Lucretius’ plan and its execution

the fact.) Lucretius’ series of arguments against the divinity of the world– that the world cannot be alive (!%+–((), that the gods cannot havecreated it, for lack of either a motive or a model to give them the idea(!(+–")), and that it is not good enough to be divine handiwork(!"(–%&)), look like a co-ordinated response to Timaeus %"e–&*c, wherethe world is the good product of a benevolent creator, and is animate,being modelled on the ideal Living Being. In Chapter &, §( we found evi-dence that the literal creationist reading of the Timaeus which is pre-supposed by Lucretius’ response was the one favoured in the Academyof Epicurus’ own day, under the headship of Polemo. This too helps toconfirm that Lucretius is following Epicurus closely in book -.34

The fact that Lucretius’ order of exposition in book - derives fromEpicurus should in any case be plain simply from a glance at items(xxv)–(xxxiii) on Chart %. In the entire sequence there is just one demon-strable transposition.35

If, as by now seems overwhelmingly likely, the astronomy in Lucretius- owes its present position to Epicurus’ polemical concerns, one mighthave expected Lucretius eventually to transpose it. Outside that polem-ical context it belonged much more naturally with the discussion ofother cosmic phenomena which I have conjecturally assigned to Onnature .,,,, and which are picked up by Lucretius book -,.36 DidLucretius intend to reorder his topics in this way? When we turn to theproems, we find that indeed he did. The programmatic proem to book-, +)–"*, places the astronomy at the end, after the history of civilisa-tion: he announces that he is going to expound (a) the world’s mortalityand its origin; (b) then (tum, +") the origins of life and civilisation(exemplified here, as in the Letter to Herodotus, by language, but also bythe origin of religious terror); and (c) in addition (praeterea, $+) the celes-tial motions. I see no reason to doubt that this programme was, as

&. Book V !(&

34 See also Ch. + for evidence that the next part of (L!*), Lucretius’ arguments in favour of theworld’s destructibility at - %&(/., represents Epicurus’ response to Theophrastus.

35 Assuming that it was Lucretius himself who moved (xxvii), the smallness of the heavenly bodies,to a less prominent position, his motive can only be a matter of conjecture. Epicurus had wishedto stress its damaging implications for the accuracy of astronomical observations (cf. Sedley(!"$+b)), and Lucretius (cf. his omission of the attack on astronomical devices, Ch. &, §$) was notinterested in pursuing that kind of critique. My guess therefore is that Lucretius initially omittedit, like many other topics, but that when he got to the beginning of book .,, he found an argu-ment that he really wanted to include – resulting in - ("%–+!&, his finely crafted passage on howthe sun, though small, can illuminate the entire world – but which required that he should firstbelatedly insert the argument to prove that the sun is small.

36 Epicurus’ selection of topics for the Letter to Pythocles is itself an acknowledgement of this fact: seeChart ! p. !&& above.

elsewhere in his proems, meant to correspond to the actual order ofexposition.37

The transposition of the astronomical passage to the end of book -,had he lived to e/ect it, would admittedly have sacrificed one advantage,namely the thematic link between the history of civilisation at the endof - and the proem to -,, where Epicurus is ranked prominently amongthe gifts of Athenian civilisation. But, compensatingly, the transpositionwould have eliminated the unwelcome interruption in book -’s historyof the world, and led to a smooth continuity between the end of -, onastronomy, and the primary content of -,, the remaining cosmic phe-nomena. In fact, the proem to -, also appears to allude to that sameplanned continuity between the two books (-, )&–+):

Et quoniam docui mundi mortalia templaesse !et" nativo consistere corpore caelum,et quaecumque in eo fiunt fierique necessest,pleraque dissolui, quae restant percipe porro.

And since I have taught that the world’s regions are mortal, andthat the heaven is made of a body which had birth, and I haveaccounted for the majority of the things which go on and mustgo on in it, now hear the remainder of them . . .

It is also worth noting that the programmatic lines in the proem tobook - promise ($+–"*) a heavily theological message for the astro-nomical section, namely that failure to understand the true working ofthe heavens leads to a morally ruinous misconception of the divinenature. This is a message which the actual astronomical passage as wehave it (- (*"–$$*) fails to deliver. Just as the middle pair of books wasto have the joint function of dispelling the fear of death, so the finalpair was destined to have (even more prominently than in our version)the function of dispelling the fear of god. I imagine that in the plannedrewriting there was to be a strongly theological motif in the astronom-ical close of book - – perhaps even including the famously missingaccount of the gods promised at - !(( (quae tibi posterius largo sermoneprobabo).

!() $. Lucretius’ plan and its execution

37 Townend (!"$") o/ers a similar argument for a di/erent account of Lucretius’ original plan,arguing that , !%$–&( reflects a stage at which he intended the sequence -, -,, ,,,, ,-. I disagree.These lines are not in any normal sense programmatic. Lucretius’ reasoning is: there are dangersattached both to false views about the gods (#*–!*!), and to false beliefs about the soul (!*%–%+);therefore it is necessary to learn both (cum, !%$) about the real explanation of celestial phenom-ena (!%$–&*), and (tum) about the real nature of the soul (!&*–(). The sequence is dictated, not bythe subsequent contents of the poem, but by the order of the two warnings in #*–!%+.

$. :A1765,A9 ’ ;6528<

I have o/ered reasons for supposing book ,- to be still very much in theorganisational state in which Epicurus himself had bequeathed thematerial. As for book -, no such surmise is needed, because a glance atChart % is enough to confirm that, with possibly only one exception (item(xxvii)), Lucretius has indeed conserved Epicurus’ order. In §§(–+ I haveargued that for both books he was visibly planning a radical overhaul,but did not live to carry it out. We can speculate what the final productmight have looked like, but there is little doubt that he would havedecided on further changes as he proceeded with the rewriting. Themost that it is safe to say is that the rewriting might well in the end havebeen as subtle and complex as the reorganisation that can actually bediscerned in books , and ,, (see §) above, and Ch. $).

Before proceeding to consider book -,, it is worth putting togethersome interim conclusions on Lucretius’ method of composition and itsmotivation. According to the reconstruction which I have proposed, heinitially downloaded (if I may be forgiven the computing metaphor)large quantities of material into Latin hexameters, following thesequence of his sole Epicurean source closely, indeed almost mechan-ically. He had already at this stage radically recast at least the diction ofthe arguments into his own evolving poetic mode – we saw evidence ofthis reshaping process in book ,- (Ch. %, §)) – but had not significantlydeparted from the overall sequence as he found it. It was largely in asecond phase that he set about reorganising it into the familiar structureby which we know it today, although even then he did not live to fulfil allof his plans for books ,-–-,. Three points may now be made explicitabout the earlier draft. (a) It was probably longer than the final draft:since each of books ,-–-, is longer than any of books ,–,,,,38 we can inferthat the uncompleted second phase involved some trimming. (b) It wasalready in verse. (c) It was already divided into books. The original pro-gramme of book ,-, which we examined earlier, unambiguously testifiesto both (b) and (c).

These three facts in turn virtually rule out the hypothesis of a furtherstage, an initial draft in Latin prose. If Lucretius had started out with aprose draft, as Virgil is said to have done when writing the Aeneid,39 we

'. Lucretius’ method !((

38 Book ,, !!!$ lines; ,,, !!$) lines; ,,,, !*") lines; ,-, !%#$ lines; -, !)($ lines; -,, !%#+ lines.39 Donatus, Vita Verg. %&: Virgil made a detailed prose draft of the Aeneid, already divided into !%

books, so that when versifying it he could move from passage to passage ad lib. The questionwhether Lucretius started the same way is properly raised by Smith (!"++), p. #% n. &.

might surely expect him to have proceeded to plan the reorganisation ofhis material before he turned it into verse. In fact, though, the opening ofbook ,- is evidence of major reorganisation taking place at a time when(a) the material was already in Latin hexameters and divided into books,yet (b) it was still in an early enough draft to retain the order of materialbequeathed by Epicurus. The hypothesis of a prose draft prior to thisstage looks explanatorily redundant.

Even without a prose draft, Lucretius was adopting a circuitous routeto his final goal, since much of his first verse draft undoubtedly requiredradical rewriting in the final version. Why should he have gone about thetask in this apparently time-wasting way? One part of the answer surelylies in the size and character of the On nature. In Chapter ) we saw thatthe entire treatise would have filled nine or ten volumes of OxfordClassical Text, and that books ,–.- alone would have filled four suchvolumes. That was a very considerable quantity of material for Lucretiusto work through when making his selection of arguments. It would havebeen extremely hard for him to move back and forth between widelyseparated passages in di/erent rolls of Epicurus’ work, while trying atthe same time to keep control over the internal structure of the emerg-ing poem. Without the aid of tables of contents, indexes, chapter andpage references, etc., this would have been a daunting task, and perhapsultimately an unmanageable one.

Once the hypothesis of a prose draft is excluded, then, Lucretius’initial decision to versify the aurea dicta largely if not entirely in their owntransmitted sequence turns out to have been, from a practical point ofview, the natural one to take. Nor was it a foolish decision. Epicurus hadhimself been scrupulous about maintaining what he considered aphilosophically correct sequence for his own argument in On nature, andindeed the magnum opus had advertised itself (as we saw in the closing sen-tence of book ..-,,, – see Ch. ), §( – but no doubt the same claim wasmade in earlier books too) as being delivered in a philosophically propersequence. If Lucretius initially hoped to preserve most of that sameorder in his poem, he had Epicurus’ active encouragement to do so.

It was only during the actual process of versification, as the philosoph-ical epic took shape, that Lucretius came to see the merits of a radicalrestructuring. If the reorganisation meant sacrificing some of Epicurus’methodological rigour, any loss would be vastly outweighed by gains inarchitectonic unity and rhetorical power. The first half of the poem iseloquent testimony to the wisdom of his decision. So far as concerns theplanned restructuring of the second half, however, only a few clues

!(+ $. Lucretius’ plan and its execution

survive in the proems to a/ord us a glimpse of the new organic wholeinto which, at the time of Lucretius’ death, the De rerum natura was stillundergoing its final transformation.

# . '88E -,

So far I have deliberately left book -, almost entirely out of account. Myhope is that what we have learnt about books ,–- will be able to throwlight on a long-standing puzzle about the final book.

After an opening disquisition on the proper attitude to divinity, thegreat bulk of book -, is devoted to a series of atmospheric and terrestrialphenomena. I have suggested tentatively that the corresponding part ofOn nature, book .,,,, covered this same pair of topics ((xxxiv) and (xxxv)).Moreover, since we have learnt that at the time of Lucretius’ death books,- and - had not yet had the same overhaul that he had given to the firsthalf of the poem, it is only to be expected that book -, in its present stateshould also follow Epicurus’ original order of exposition closely.

Our only clue to the internal order in which Epicurus’ section on thesephenomena proceeded is the corresponding part of the Letter to Pythocles("#–!!+). This does not match Lucretius -, at all closely. But it is a partwhere the Letter seems to be combining two or more di/erent sources(see Ch. ), §"), and we cannot be confident how far the internal struc-ture of On nature .,,, can be recovered from it. In proposing a matchbetween On nature .,,, and DRN -,, I start with an appeal to probability,based on what we have learnt about the composition of the two pre-ceding books, where very strong reasons emerged for deriving Lucretius’order from the corresponding books of On nature.

A closer look at Lucretius’ sequence of topics in book -, proves, at thevery least, consistent with this supposition. It shows a strong likelihoodthat the sequence derives ultimately from Theophrastus. And we havealready found some reason to regard Theophrastus’ Physical opinions asone major influence on Epicurus’ own exposition, and, through him, onLucretius’. (Chapter + will amply confirm that picture.)

We can see this by placing Lucretius’ order alongside that found inAetius (Chart ), p. !(# below).40 That Aetius’ own order stems fromTheophrastus’ foundational doxographical work – as one would in anycase have predicted41 – receives some confirmation if we place on the

(. Book VI !($

40 Cf the table in Runia (!""$).41 For continuing general acceptance of the thesis of Diels (!#$"), who traced the origins of the

‘Aetius’ doxography ultimately to Theophrastus, see e.g. Mansfeld (!""*).

right-hand side the sequence of headings found, not in the Physical opin-ions itself (which is lost), but in what is now thought to be Theophrastus’Metarsiologica, which survives, abridged, in Syriac and Arabic42 (I useitalics to highlight an isolated discrepancy).

Lucretius’ sequence, although selective, conforms very closely to thatin Aetius. The one minor departure is at (%)–&), but it is clear what hashappened there. Following his explanations of cloud and rain, he dealswith the next group – the remaining aspects of weather – en bloc, byo/ering a few words in explanation of one of them, rainbows, andinstructing his readers thereafter to work out their own explanations forthe remainder. Allowing for this deliberate abridgement, the parallelismholds even there.

The material in Theophrastus is incomplete, but its degree of coin-cidence with the Aetius order is su0cient to guarantee that, as expected,

!(# $. Lucretius’ plan and its execution

42 On this text, see Ch. +, §+ below.

Chart "Lucr. VI Aetius III !–IV ( Thphr. Metars.

milky waycomets

"+–!(" thunder thunder thunder!+*–%!# lightning lightning lightning%!"–)%% thunderbolts thunderbolts thunderbolts)%&–(* prester prester

typhoons)(!–") cloud cloud, mist cloud)"(–(%& rain rain rain(%)–&) [rainbows, snow, dew, snow, frost, snow, hail, dew, frost

wind, hail, frost] hail, rainbowswind windwinter and summerthe earth (number,

size, shape, halo of moonposition, stability)

(&(–+*$ earthquakes earthquakes earthquakes+*#–&# sea sea+&"–$!! Etna

tideshalo of moon

$!%–&$ Nile Nile$&#–#&" Avernian Lake#)*–"*( springs and wells"*+–!*#" magnets!*"*–!!&$ disease

that order drew its origins from Theophrastus’ work. Moreover,Theophrastus’ exposition anticipates so much of the detail of Lucretius’that it is indisputably, even if at one or more removes, a major influenceon it: I shall say more in illustration of this in Chapter +, §+.

In combination, these facts strongly encourage the further inferencethat the Lucretian material, which matches the same order almost per-fectly, reflects Epicurus’ use of Theophrastus.43

The way in which the Theophrastean expository sequence functionsis best seen by following the complete set of topics, assembled from allthree sources. It works its way steadily downwards from the upper levelsof the atmosphere to the earth and its contents. (The one exception, forwhich I have no explanation to o/er, is the halo of the moon, di/erently– but equally anomalously – placed by Aetius and Theophrastus.)44

Thus the waterspout, prester, no doubt preceded cloud and rain in the listbecause it was usually taken to descend from above, being fiery in nature,whereas cloud was regarded as produced from the air itself or by exhala-tions from below. It is particularly significant here that Lucretius does notfully endorse this standard di/erence. Like Epicurus (Letter to Pythocles!*)–(), he discounts the fiery nature usually attributed to the prester, andderives it purely from cloud.45 One might therefore have expected him,had he been creating the order himself, to have located it after cloud inthe sequence, especially as he allows as one source of cloud formationthe arrival of matter from above, even from outside the cosmos ()#!–")).The fact that he does not merely confirms that his material came to himin an order already conferred on it by Theophrastus’ doxographicalmethod.

This does not in itself prove that Lucretius’ material came to him froma book of Epicurus shaped by its reliance on Theophrastus, rather thanfrom his own independent use of doxographical sources.46 But if I have

(. Book VI !("

43 Cf. Mansfeld (!""%a), p. &%+: ‘But what we find in Theophrastus can be paralleled point by pointfrom Lucretius. We therefore must assume that Lucretius, at least for his sections dealing withthunderbolts, did not use the Letter [sc. to Pythocles], and may surmise that a much longer epitomeof Epicurus’ views on cosmology and meteorology was available to him (assuming he did notconsult the di0cult Physics [meaning On nature]). It follows that in the books of his Physics on whichthis epitome was based Epicurus followed Theophrastus very closely: at any rate this can be provedfor the argument about thunderbolts.’ (For this comparative material on thunderbolts, see Ch. +§+.)

44 As evidence that some kind of displacement has taken place, it is worth noting that inTheophrastus this chapter includes the theological excursus which Daiber (!""%), p. %#*, andMansfeld (!"")a), p. &!+, agree to be itself misplaced.

45 This in itself may well be influenced by Theophrastus’ own view on prester; see p. !#% below.46 The latter option is favoured by Runia (!""$).

been right to explain the programmes of books ,- and - by the hypoth-esis that their sequence was derived from Epicurus’ in On nature, the sameexplanation inevitably commends itself for book -, as well.

There is, on the other hand, one item in book -, which can hardly reflectOn nature. As is well known, the book’s final part (!!&#–!%#+) is a long andgruelling description of the great plague at Athens during thePeloponnesian War, borrowed directly from Thucydides. Just as theentire poem opened with a celebration of life drawn from a fifth-centurySicilian, so it closes with a grim tableau of death drawn from a fifth-century Athenian.47

There is much unresolved dispute about this passage. While it iswidely agreed that the horror story of mass death somehow serves tocounterbalance the poem’s opening focus on birth and life and thedenunciation of the fear of death which provides the climax for the firsthalf of the poem, it is not so widely accepted that the close as we haveit can be as Lucretius meant it finally to be. In confirmation of thisdoubt, it is worth observing that the passage has what we have now seento be the hallmark of a Lucretian first draft: that is, despite occasionalomissions or additions, it retains the exact order of material found in itsGreek source, Thucydides ,, )$–().48 This time there is no way ofblaming his procedure on the sheer size of the source text, but we maystill surmise that, even when working from a relatively short Greekpassage, what had by now become his habitual method suggested itselfonce again, as the best way to ensure as full coverage of the Thucydideanmaterial as he evidently wanted, prior to reworking it to fit the poem’sstill developing architectonic.

Those who believe, as I do, that Lucretius must have intended torework the plague passage and to make its moral explicit,49 can drawcomfort from the general picture I have painted. If my story is right,Lucretius’ plans for organising his individual books around an over-arching moral framework had not, at the time of his death, been fullyput into e/ect for the second half of the poem. If book -, does not yet

!+* $. Lucretius’ plan and its execution

47 Among the many excellent studies of the plague passage, I have found those of Commager(!"($), Bright (!"$!), and Clay (!"#&) especially helpful.

48 For demonstration of this, see Bright (!"$!), p. +*#. The one apparent exception is !%)$–#, butI am persuaded by Bright’s argument in defence of Bockemüller’s transposition of !%)$–(! to aconcluding position after !%#+, which restores the correspondence.

49 Although Bockemüller’s transposition (see previous note) may, as urged by Fowler (!""+), p. ##",give the final lines some appropriate closural features, such minor repairs do not in my view comenear to supplying a morally credible closure to the poem as a whole.

have a fully finalised closure, in the light of the current state of books ,-and - that is exactly what we should expect. We are therefore at libertyto ask how he might have meant to close book -,, without shackling our-selves to a doctrinaire insistence on the integrity of the existing text as afinished product. Given that the book as we have it is already exception-ally long,50 there is also a strong possibility that the description of theplague itself was destined to be significantly reduced in the final draft,before the concluding moral was grafted on to it.

Much the most promising guess to have emerged from discussions ofthis problem is that Lucretius meant here to show how the other achieve-ments of civilisation are dwarfed by Epicurus’ contribution to it. Athens,the proem to book -, points out, was the cradle of civilisation for ‘ailingmortals’ (mortalibus aegris, -, !), giving them both corn and laws: Athens,that is, helped foster both our bodily and our moral needs. But, theproem continues, Athens also gave us Epicurus, whose godlike discover-ies have outlived him to spread ‘life’s joyful solaces’. It was he who trulysatisfied our physical and moral needs, by teaching us the limits of plea-sure and by dispelling our fears.

The return to Athens at the end of the book must have been meantto take this message forward. Lucretius surely wanted us to learn thatwhen the Athenians faced the worst that fortune could hurl againstthem,51 the other benefits of their civilisation were powerless: onlyEpicurus’ wisdom, had it yet come to birth, could have dealt with thehorrors of the plague, both physical and moral. He is the one who hastaught us to tolerate bereavement and bodily pain with genuine opti-mism, and not to cling desperately to life as if death were an evil. In otherwords, the plague must carry a message for us, and this is supported bothby my earlier finding (Ch. %, §!!) that here Lucretius is not, as often else-where, seeking to emphasise any gulf between Greece and his readers’own world, and by the way in which he omits the circumstantial detailsof the Athenian plague in order, it seems, to maximise the generality ofits lessons.52

Now it seems idle to pretend that the intended message is in fact con-veyed by the closing section of the poem.53 Some would interpret the

(. Book VI !+!

50 See n. &#, p. !(( above.51 Cf. -, %"–&% for anticipation of this theme in the book’s proem.52 This aspect is well brought out by Bright (!"$!). Cf. also Segal (!""*), p. %&!: ‘Athens is remote

enough from Lucretius’ Roman audience to be exemplary, but real enough to be terrifying.’53 Cf. Long (!""%a), p. )"#: ‘If Lucretius were concerned to disclose and cure anxieties about dying

painfully, the plague would be about as e/ective as a horror movie for inducing pleasant dreams.’

plague passage as a final test: have Lucretius’ readers really learnt theirEpicureanism? But even that interpretation must presuppose that wehave at least been taught the relevant principles, so that we are now readyto apply them for ourselves.54 I wonder how many readers of this passagehave ever been left feeling that, thanks to the poem’s lessons, if they hadbeen there they would have been less helpless than the wretchedAthenians were in the face of such grisly su/ering. The expectedEpicurean teachings about the right responses to painful death are notyet fully in place. We have learnt much from book ,,, about why we shouldnot fear being dead, and those lessons will certainly prove to bear on theconduct of the plague victims. But where have we been taught how toremain happy through severe and even terminal physical su/ering?

On this matter Epicurus’ teachings were well known, and his owndeath the great model. He had maintained that even excruciating painneed not mar our happiness, both because we can be confident that itwill be short-lived,55 usually to be followed by the totally painless state ofdeath, and because even while it is going on it can be outweighed by thejoyful recollection of past pleasures. Cicero, for one, showed how well heunderstood the pivotal importance of these tenets in his eloquent expan-sion of the Epicurean tetrapharmakos (De finibus , )*–!):

That pleasure is the ultimate good can be most easily seen from the followingpicture. Let us imagine someone who enjoys great, numerous and continuouspleasures of both mind and body, unobstructed by any pain or by the prospectof it. What state could we call more excellent or choiceworthy than this one?For someone in such a condition must possess the strength of a mind which fearsneither death nor pain, on the ground that death is painless, and that long-termpain is usually bearable, serious pain short-lived, so that intense pain is com-pensated by brevity, long-term pain by lightness. Once we have added to thisthe provision that he is not in awe of divine power, and that he does not allowpast pleasures to evaporate but enjoys constantly recalling them, what furtherimprovement could be possible?

These teachings on neither fearing, nor being made wretched by, eventhe most intense physical su/ering are absolutely central to Epicurus’ethics, and their relevance to the plague victims is obvious. Yet nothinghas yet been said about them in the poem. As it stands, Lucretius’ examsets us at least one large question to which he has nowhere hinted at theanswer.

!+% $. Lucretius’ plan and its execution

54 See Clay (!"#&), pp. %%(, %($–++, who cites , )*%–&, - !%#!–%, -, +#–$", (%$–&) in support of thedo-it-yourself interpretation. In all these cases, it seems to me, we have been supplied with thenecessary materials or explanatory model. 55 KD ), SV &–).

The tetrapharmakos or ‘fourfold remedy’, which summarised the cardi-nal first four tenets of the Epicurean Kyriai doxai, ran as follows:56

God presents no fears, death no worries. And while what is good is readilyattainable, what is terrible is readily endurable.

Up to this point in the poem the first three have all been magnificentlypreached. The first tenet, that god is not to be feared, is a central themeof the entire poem, emphasised in the proems to books , and ,,,, and con-solidated at -, )#–$"; indeed, I have argued (§+ above) that Lucretiusmeant to give it even greater emphasis in the final version of books -–-,.The second, that death is nothing to us, is of course the prime lesson ofbook ,,,, and, as we have seen, was intended to be that of book ,- as well.The third tenet, that good can be readily attained (by imposing a naturallimit on our desires), is the theme of the proem to book ,,, and the proemto book -, glorifies it as the chief benefaction bequeathed to us byEpicurus, as if Lucretius, in embarking on the final lap, were consciouslyreminding us that this lesson too is in place. But the last of the four car-dinal tenets, that pain is readily endured, is totally missing. The plaguepassage is the best possible evidence that Lucretius meant to add it.57

The genetic account of the poem which I have defended in thischapter o/ers a satisfying explanation of the omission. Lucretius, wehave seen, first took over the raw material he needed from his Greeksources en bloc, and only in a second phase reworked it to blend with thepoem’s emerging master-plan. In the case of the plague, as of much elsein the second half of the poem, the reworking simply had not yet takenplace when he himself died.

What Lucretius still owed his readers was Epicurus’ explanation ofhow tolerance of physical pain depends on our mental attitude to it. Toconfirm such an account of the poet’s intentions, the most that we canhope for in his paraphrase of Thucydides is the occasional clue to hiseventual aims. Fortunately, some excellent work has been done onLucretius’ use of Thucydides, and it does indeed confirm that he washeading in some such direction. A series of valuable analyses have

(. Book VI !+&

56 For the text, see Long/Sedley (!"#$), %(G, or Angeli (!"##), p. !$&. Although both the tetraphar-makos and the present arrangement of KD may postdate Epicurus, Ep. Men. !&& fully confirmsthat Lucretius would have known the four canonical tenets from Epicurus' own writings.

57 There is, admittedly, a disparity between the first and the second pair of ethical tenets. Theformer are themselves founded on the lessons of physics, to which the poem is formally dedi-cated, while the latter are not. Nevertheless, the third tenet receives enough emphasis in thepoem (albeit without formal argument) to encourage the expectation that Lucretius intended noless for the fourth as well.

brought out a number of points where Lucretius departs from the letterof Thucydides’ text. His strategic omissions of Thucydidean materialand inclusion of additional medical details serve, inter alia, to magnify thehorror and hopelessness of the situation. Along with this comes amarked tendency to psychologise.58

Sometimes the psychologising adjustments emphasise people’s horrorof death as such. For example, where Thucydides’ plague victims onoccasion survive thanks to the loss of diseased bodily parts (,, )".$),Lucretius’ interpretation is that the fear of death actually drove them tosever their own limbs and organs (!%*#–!%):

et graviter partim metuentes limina letivivebant ferro privati parte virili,et manibus sine nonnulli pedibusque manebant !%!*in vita tamen, et perdebant lumina partim:usque adeo mortis metus his incesserat acer.

And some of them, through the burden of fear at the onset ofdeath, stayed alive by cutting o/ their genitals with a knife;others stayed on without hands and feet, but alive; others losttheir eyes. So far had the grim fear of death entered them.

Elsewhere Lucretius elaborates on the mental distress brought about bytheir current plight itself. For instance, where Thucydides (,, )".&)describes the physical symptoms as being µ"-+! -+7+$6(#$"+) µ"4+" 73),‘with much (physical) distress’, Lucretius not only takes this as describ-ing their mental state (which in itself would be an understandable error),but expands it as follows (!!(#–"):

intolerabilibusque malis erat anxius angoradsidue comes et gemitu commixta querella.

Their unbearable su/erings were at all times accompanied by atorment of troubledness,59 and by groaning mixed with lamen-tation.

Symptoms, not in Thucydides, that the disease is in its terminal stagesinclude (!!#&–))

perturbata animi mens in maerore metuquetriste supercilium, furiosus voltus et acer . . .

!+) $. Lucretius’ plan and its execution

58 Especially Commager (!"($), Bright (!"$!). It is possible to accept Commager’s findings withoutendorsing his conclusion, that Lucretius is presenting the plague itself as moral illness.

59 That anxius angor does not specifically signify ‘anxiety’ is shown by ,,, ""&, where the same phrasedescribes the torment of one in love (cf. Commager (!"($), p. !*+). Hence there is no reason toread this passage as focusing especially on the fear of death.

a mind distraught in its grief and fear, a gloomy brow, a frenziedand grim expression . . .

One further discrepancy is that where Thucydides describes thecrowding of country-dwellers into the city, Lucretius (-, !%(%–() insteadgives the impression that the plague spread death into the countryside. Iimagine that this deliberate shift prepares the ground for a warningabout universality: such su/erings are not a hazard exclusive to civic life,but one which every human being must be prepared to confront anddeal with. Epicurus, we should recall, had established his Epicureancommunity outside the city walls of Athens, thereby reminding us of hisalready memorable dictum that ‘when it comes to death, we humanbeings all live in an unwalled city’.60

These and similar clues as to how Lucretius was beginning to shapehis source material lend strong support to the hypothesis that the even-tual moral message was to be a quintessentially Epicurean one aboutfacing terminal su/ering with the right frame of mind – a frame of mindwhich will enable us to eliminate fear and to tolerate pain cheerfully if itshould come our way. Whoever we are and wherever we live, if we havenot learnt this lesson we cannot face the future with truly Epicureanequanimity.

Like Lucretius’, so too Epicurus’ last written words had been adescription of terrible physical su/ering – his own. Yet his happiness, hewrote, was unmarred:61

I wrote this to you on that blessed day of my life which was also the last.Strangury and dysentery had set in, with all the extreme intensity of which theyare capable. But the joy in my soul at the memory of our past discussions wasenough to counterbalance all this.

This triumph of philosophical serenity over the most intense physicalpain was surely what Lucretius was preparing to bring into focus at theclose of his poem. The panic, terror and misery of the pre-EpicureanAthenians, in the face of bodily su/ering hardly worse than Epicurus’own terminal illness, are a brilliantly graphic backdrop to this final lessonin the Epicurean ethical canon.

(. Book VI !+(

60 SV &!. 61 DL . %%.

1234567 +

The imprint of Theophrastus

! . 526842739 5A9 3?< 526 =87:< ’9 <69 57A15,',:,5D

The aim of this chapter is to consolidate a picture, which has beengrowing in the preceding two chapters, of the vital role of Theophrastusin Lucretius’ poem.

I shall start with a close look at one particular text, Theophrastus, fr.!#) FHS&G,1 which I shall argue lies directly behind a series of argu-ments in Lucretius book -.

Philo (ll. !–), De aeternitate mundi !!$) reports as follows:

Theophrastus, however, says that those who assert that the world2 is subject tocoming-to-be and passing away were led astray by four principal considera-tions: (!) the unevenness of the land, (%) the withdrawal of the sea, (&) the dis-solution of each of the parts of the whole, ()) the perishing of whole kinds ofland animals.

The text then goes on to amplify these four arguments (ll. )–#", Aet.!!#–&!). Finally comes a refutation of each of the four in turn.

The four arguments attacked by Theophrastus were identified byZeller3 as belonging to Zeno of Citium. As a result, the first part of thesame text became Zeno fr. (+ in Pearson’s collection,4 and Zeno fr. !*+in vol. , of von Arnim’s Stoicorum Veterum Fragmenta.5 They are still oftenregarded as Zenonian,6 despite the reservations that have been voiced

!++

1 !Fortenbaugh/Huby/Sharples/Gutas (!""%), to which all my Theophrastus fragment numbersrefer. The fragments I shall be discussing are in fact edited by R. W. Sharples. In the main I repro-duce his translation, with a few modifications, noting any changes which could be thought to beof substance. Fr. !#) will be cited by line number in FHS&G, followed by chapter reference to Deaeternitate mundi. Since I drafted a large part of the content of this chapter (now published as Sedley(!""#a)), the discussion of this fragment in Kidd (!""+) has appeared. I am glad to find myself incomplete agreement with his conclusions.

2 Here and elsewhere I shall translate ,/" 'µ/) ‘world’ (‘universe’ Sharples). 3 Zeller (!#$+).4 Pearson (!#"!). 5 Arnim (!"*&–().6 See e.g. the long appendix in Graeser (!"$(), pp. !#$–%*+, ‘Zenons Argumente gegen Aristoteles’

These von der Ewigkeit der Welt’; Long/Sedley (!"#$), ,, %$(.

from time to time about the passage’s credentials.7 Zeller’s proposal ischronologically hard to sustain,8 and I shall be maintaining that a betterexplanation of the arguments’ origin is available. But let me now brieflydiscuss one of the items cited by Zeller in his support, since it may indi-cate something about the passage’s character.

In developing the third argument, from the perishability of theworld’s parts to the perishability of the whole, the text includes the fol-lowing (ll. )(–#, Aet. !%():

Do not the strongest (,#+-+$/" -+-/$) stones moulder and decay, and because ofthe weakness of their constitution ("+ @"()) – that is the tension of their pneuma(6."&µ+-$,/! ) -/" ./)), a bond (5"'µ/" )) which is not unbreakable but only di0cultto undo – do they not crumble and dissolve . . .?

The language is unmistakably Stoic, as Zeller was quick to point out.True, but does that show that Theophrastus (died %##–%#+ '1) in his oldage must have been defending Aristotelian cosmology against the newupstarts in the Stoa? Far from it. This formal extension of pneuma, invarying states of tension, beyond living beings to become a universalcausal agent, is hard to date earlier than Chrysippus, a full generationafter Zeno.9 And we can go further than this. The language used, speak-ing as it does of the 6."&µ+-$,/! ) -/" ./) which binds even stones together,is unmistakably the Stoic-influenced language characteristic of Philohimself.10 Indeed, it seems clear that the language of the entire passageis primarily Philo’s own.11 This must put us on our guard against anyassumptions about how much of the content goes back to Theophrastus.But we will nevertheless see reason to regard its principal arguments asauthentic, and in what follows I shall permit myself to go on calling thepassage’s author Theophrastus.

!. Theophrastus and the world’s destructibility !+$

27 Especially Diels (!#$"), pp. !*+–#; Arnim (!#"&), who, despite inveighing against the attributionof the arguments to Zeno, proceeded to print them as a fragment of Zeno in Arnim (!"*&–();Wiersma (!")*); McDiarmid (!")*).

28 In §$ below I argue that the Theophrastus text was early enough to draw a response from Epicurusin On nature .,>.,,,, datable (see Ch. ) §!%) before &*$/+. Yet Zeno was still a student till c. &**.

29 See e.g. Solmsen (!"+!b); Lapidge (!"$#).10 Cf. Philo, Deus &(–+, 7$"1(. ,+$! @&" 7(. 5"'µ/! . ,#+-+$/" -+-/. "+ @$. "$&#4+" >"-/ [sc. God], 3$ 5 & "&'-$!

6."&' µ+ +& .+'-#""%/. "&% & "$+&-/" . +% #;"-+$ µ"! . 4+! # +& 6/! -(' . µ""'(. "&6$! -/! 6""#+-+ -"$"."'1+$.Cf. Bel. #*; Heres. %)%, -(' . 6."&µ+-$,(' . -/" .(., /$* '&µ%&""'-+-/) 5"'µ/! ) 3# '+..

11 See further Wiersma (!")*). In addition to his evidence, note e.g. l. (+ (Aet. !%$) -$" ;#3! µ+,#3-4/#"$' . 6"#$! . . ., l. "& (Aet. !&%) (# 4"..+$' /$ , and l. !&% (Aet. !&#) -#$6/" 13-/., all typicallyPhilonian flourishes. Kidd (!""+), pp. !&$–#, notes the very Philonian 7"(%/" #/) l. !& (Aet. !!"),and Diskin Clay has pointed out to me the citation of Euripides fr. #&" Nauck/Kannicht at ll.!+"–$! (Aet. !))), a fragment also quoted by Philo at Aet. ( and &*.

% . 526 @8A752 37BA;6?5

I want to take the four arguments in their reverse order, since it is the lastwhich most clearly demonstrates Theophrastus’ influence onLucretius.12 This fourth argument we have already seen labelled as ‘theperishing of whole kinds of land animals’. The details, when they come,match this description crudely at best (ll. #*–&, Aet. !&*):

The fourth and remaining argument is to be stated precisely in the followingway, they say. If the world were eternal, living creatures too would be eternal,and especially the human race, in so far as it is superior to the others. But thatthe human race’s origin is recent is clear to those who wish to inquire intonatural matters.

After a brief argument for the recentness of mankind’s arrival, inter-rupted by a lacuna, the passage concludes (ll. #$–", Aet. !&!):

But if mankind is not eternal, neither is any other living creature; so neither arethe places in which these live, earth and water and air. And from this it is clearthat the world is perishable.

Clearly a more complex argument has been condensed.13 Roughly, itseems to run:(a) Mankind is of recent origin.(b) What has a temporal origin has a temporal end [a standard assump-

tion].(c) Therefore mankind will one day perish.(d) Other species are less favoured than mankind, so will perish too.(e) Since the species inhabiting earth, water and air will perish, those

elements will themselves perish [no reason given].(f) Since three of the four cosmic masses will perish, the world itself will

perish.

!+# &. The imprint of Theophrastus

12 After writing the first draft of this chapter, I unearthed many of the same points of comparisonbetween the Theophrastus fragment and Lucretius - in the long chapter " of Bignone (!"&+).His thesis is that Zeno, Theophrastus and Epicurus were all engaged in a debate sparked byAristotle’s De philosophia: Zeno produced the four arguments against Aristotle, Theophrastusreplied to them, and then Epicurus rehabilitated them. Although I shall argue that this cannotbe entirely the right story, Bignone’s discussion does contain many acute observations on the linksbetween the Theophrastus and Lucretius texts (following the lead of Norden (!#"&)). Typicallyof Bignone, they are hidden in a jungle of confused speculation which makes them hard to pickout (especially for those already repelled by the author’s incessant self-congratulation). Forinstance, although his conclusion is that the Epicureans were responding to the Theophrastustext, he frequently talks as if it were the other way round (,, )(*, )(+–#, )$%–&). All this has hadthe e/ect of making his real discoveries pass virtually unnoticed. Even Boyancé (!"+&), pp.%!)–%!, while endorsing Bignone’s general thesis about the polemical motivation of Lucretius -,totally fails to pick up the crucial role of Theophrastus fr. !#). On this topic, cf. also Solmsen(!"(!). 13 This point is well made by Runia (!"#+), p. #&.

It is not yet clear why in Theophrastus’ introductory summary of thisargument – ‘the perishing of whole kinds of land animals’ – the speciescovered by (d) and (e) should have been narrowed down to land animals,but we can safely postpone the answer to this puzzle for a while. Moreurgently, we might ask whether Philo’s paraphrase distorts the originalbalance of the argument, by concentrating almost exclusively on step (a),the recent origin of mankind. But it will become clear that this was, atleast, the step on which Theophrastus chose to concentrate his criticism,and on which he therefore probably concentrated his summary too.Significantly, it is also the part of the argument singled out for defenceby the Epicureans.

The details of step (a) are given as follows (ll. #%–$, Aet. !&*–!):

But that the human race’s origin is recent is clear to those who wish to inquireinto natural matters; for it is reasonable, indeed necessary, that the crafts shouldexist alongside mankind and be of the same age, not only because what issystematic is proper to what is rational by nature, but also because it is not pos-sible to live without these. So let us consider the date of each craft, disregardingthe stories told about the gods by the tragic poets [lacuna].

Unfortunately the promised survey of the crafts has vanished into thelacuna at the end. But we can infer something of its contents from theargument at Lucretius - &%)–&):14

praeterea si nulla fuit genitalis origoterrarum et caeli semperque aeterna fuere, &%(cur supera bellum Thebanum et funera Troiaenon alias alii quoque res cecinere poetae?quo tot facta virum totiens cecidere neque usquamaeternis famae monumentis insita florent?verum, ut opinor, habet novitatem summa recensque &&*naturast mundi neque pridem exordia cepit.quare etiam quaedam nunc artes expoliuntur,nunc etiam augescunt; nunc addita navigiis suntmulta, modo organici melicos peperere sonores.

Besides, if earth and heaven had no origin of birth, and alwaysexisted, being eternal, why is it that other poets have not alsosung of other things before the Theban War and the sack ofTroy? Into what place have all those deeds of men again andagain fallen, and nowhere been planted to flourish in the eternalmonuments of fame? But actually, I believe, the world is new

". The fourth argument !+"

14 Lucretius concludes this passage, at &&(–$, with a comment on the newness of his ownphilosophical task; this is presumably his own personal touch, and can be ignored for presentpurposes.

and of recent birth,15 and did not begin long ago. That is whyeven now some crafts are being perfected, even now developing.Only now have certain improvements been made to ships, justrecently did musicians create their melodious sounds.

Now the mere correspondence of Lucretius’ argument with the trun-cated one recorded by Theophrastus does not yet demonstrate that theEpicureans learnt it from Theophrastus. The hypothesis of a commonsource would do as well. But the supposition of direct Theophrasteaninfluence will become unavoidable as soon as we see that the Epicureanshave adjusted their version of the argument to take account ofTheophrastus’ criticisms.

Theophrastus’ first criticism is as follows (ll. !$%–+, Aet. !)():

Certainly, to continue, it is complete foolishness for the human race to be datedby the crafts. For if anyone follows this absurd argument they will show that theworld is quite new, having been put together hardly a thousand years ago, sincethose who we are told were the discoverers of the crafts do not go back morethan that number of years.16

It is not clear why Theophrastus infers that the world will be no olderthan mankind. He may be assuming, either on his own account or byinference from step (e), that the principal cosmic masses have only thesame lifespan as the species which inhabit them. But whatever hisreasons, it seems clear that his inference was accepted and accommo-dated by the Epicureans.17 Lucretius, as we have just seen, builds it intohis version of the argument that both the crafts, and the world itself, areindeed of very recent origin – barely older than the Theban and Trojanwars. It is evident that this must be an Epicurean concession renderedexpedient in order to meet the Theophrastean objection, rather than astandard tenet. For in an argument at the end of book ,,, undoubtedlyechoing an original passage of Epicurus,18 Lucretius has already arguedfor the world’s perishability on the premise that, far from being young,it has already passed its peak and entered its terminal decline.

!$* &. The imprint of Theophrastus

15 For this translation of recensque naturast mundi, see Clay (!"#&), pp. #+–$.16 The figure !,*** years confirms that we have here Theophrastus’ own formulation, not Philo’s.

Plato, in Theophrastus’ own lifetime, could speak of the arts as !,*** or %,*** years old (Laws+$$d), but no one in Philo’s day, four centuries later, could plausibly still have put the figure aslow as !,*** (especially someone living in Egypt, whose civilisation was by common consentmuch older than that of Greece).

17 This fact seems to preclude the possibility that the inference used a teleological premise, namelythat the cosmic masses exist purely for the sake of their respective inhabitants.

18 ,, !!!#–$). It relies on the analogy of the world’s growth and decline to those of a living being,!!!#–)&, an analogy already invoked for the same purpose by Epicurus, see fr. &*( Usener.

Theophrastus’ objection to the argument continues as follows (ll.!$+–%*), Aet. !)+–"):

If then we are to say that the crafts are as old as mankind, we must do so notcarelessly and lazily, but with the help of research concerning nature. Whatresearch? Destructions of things on land,19 not of all of them together but ofmost of them, are attributed to two principal causes, indescribable onslaughtsof fire and water; they say that each of these descends in turn, after very longcycles of years. So, when a conflagration occurs, a stream of fire from heavenis poured out from above and scattered far and wide, spreading over greatregions of the inhabited earth; when there is an inundation, all the substance ofwater rushes down in the form of rain;20 rivers fed by their own springs, andwinter torrents, not only flow in spate but exceed the usual level to which theyrise and either break down their banks or leap over them, rising to the greatestheight. Then they overflow and pour out over the adjacent plain. This is first ofall divided into great lakes, as the water always settles into the hollow parts, butas the water continues to flow in and submerges the intervening strips of dryland by which the lakes are separated, in the end it becomes a great expanse ofsea as the many lakes are joined together.

And by these conflicting forces those who dwell in opposite places aredestroyed in turn. The fire destroys those who dwell on the mountains and hillsand in places where water is scarce, since they do not have abundant water,which is the natural defence against fire. And conversely the water destroysthose who dwell by rivers or lakes or the sea; for evils are accustomed to fastenon those close at hand, at first or even solely.

When the greater part of mankind perishes in the ways stated, apart fromcountless other minor ways, of necessity the crafts fail too; for it is not possibleto see knowledge on its own, apart from its practitioner. When the common dis-eases21 abate, and the race begins to grow and flourish from those who were notpreviously overcome by the troubles that pressed upon them, then the crafts toobegin to arise again; they have not come into being for the first time then, buthad become rare22 because of the reduction in the number of their possessors.

Theophrastus’ point is of course that the data to which the fourth argu-ment appeals – the newness of the crafts – could equally well beaccounted for by periodic cataclysms in an eternal world. Two aspectsare worth noting in passing.

First, we can now see that it must be Theophrastus himself who chose,at the beginning of the fragment, to characterise the fourth argument as

". The fourth argument !$!

19 -(' . ,+-+! 43' . (‘things on the earth’, Sharples). Or perhaps ‘land species’.20 ‘Every sort of water rushes down’, Sharples. But I think ++ 6+'+. -3" . is more likely to mean ‘all’

than ‘every’. I also prefer not to delete ,+-/µ2#$"+., with Diels, but to emend it to ,+- & /% µ2#/.,‘in the form of rain’.

21 I translate ./" '/$ ‘diseases’, not ‘ills’ with Sharples, in order to bring out a link with Lucretius –see below. 22 &$ 6/'6+.$'1"$"'+): ‘(previously) neglected’, Sharples.

based on ‘the perishing of whole kinds of land animals’. For near thebeginning of his reply to it, just quoted, he notes that it is primarily thevirtual destruction of land species that could equally well be accountedfor by periodic conflagrations and floods. Water dwellers would clearlybe largely immune to both, nor would one expect birds to be badlya/ected by either, given the availability of water as a refuge from fire,and high ground from flood (ll. !"%–$, Aet. !)#). Revealingly, then, heturns out to be labelling the argument from the outset, not in the termsin which its proponents would present it, but with an eye to the natureof the refutation he is already planning. His project is not to catalogueothers’ arguments in the manner of a neutral observer, but to engineera dialectical confrontation. We will be able to make use of this observa-tion later.

Second, as David Runia has expertly brought out,23 the description ofconflagration and flood draws heavily on Timaeus %% – not actually withany verbatim quotation, but with an extended interpretative paraphrase.Runia argues, reasonably enough, that this ‘slavish dependence on anauthoritative text’ is more likely to betray Philo’s hand thanTheophrastus’. He points out that, while Theophrastus as anAristotelian can be expected to subscribe in general terms to the theoryof periodic cataclysms,24 Aristotle seems to have concentrated on floods(especially Meteor. , !)), to the exclusion of conflagrations.

Runia’s conclusion is certainly appealing. On the other hand, thepassage does not really read like anybody’s appeal to authority. On thecontrary, that the debt is specifically to Plato is, if anything, disguised bythe paraphrastic style and the inexplicit attribution, ‘they say’ (%+'$"., l.!#!, Aet. !)+ fin.). I strongly suspect that the Timaean material wasalready included by Theophrastus, motivated not by reverence for Platobut by a dialectical strategy. But to this too we will return later.

What can be asserted with some confidence is that, with or withoutthe Timaean details, the reference to conflagration as well as flood wasalready in the Theophrastean reply as Philo found it. That is becauseLucretius, immediately following his own appeal to the newness of thecrafts, can be seen making a direct countermove to the Theophrastean

!$% &. The imprint of Theophrastus

23 Runia (!"#+), pp. #&–).24 Aristotle (De philosophia fr. # Ross; Met. 9 #, !*$)b!*–!&) taught that the crafts and other forms of

learning were all but wiped out by periodic cataclysms, but regenerated from the remnants –much the same theory as is applied to the crafts in the Theophrastus text. For Theophrastushimself there appears to be no independent evidence, but fr. (#)3 shows that he subscribed tosome sort of genetic theory of human civilisation, and it is hard to think what that can be if notthe cataclysm theory.

reply, and doing so with explicit reference to fire as well as flood (-&&#–(*):

quod si forte fuisse antehac eadem omnia credis,sed periisse hominum torrenti saecla vapore,aut cecidisse urbis magno vexamine mundi, &)*aut ex imbribus assiduis exisse rapacesper terras amnis atque oppida coperuisse,tanto quique magis victus fateare necessestexitium quoque terrarum caelique futurum.nam cum res tantis morbis tantisque periclis &)(temptarentur, ibi si tristior incubuissetcausa, darent late cladem magnasque ruinas.nec ratione alia mortales esse videmur,inter nos nisi quod morbis aegrescimus isdematque illi quos a vita natura removit. &(*

If however you happen to believe that all the same things existedpreviously, but that the generations of humans perished throughscorching heat, or that cities fell because of a great cosmicupheaval, or that as a result of incessant rain the riversoverflowed to sweep over the land and engulfed the towns, youare all the more defeated, and must admit that earth and heavenwill also be destroyed. For when the world was beset by such ter-rible diseases and dangers, if at that time a more ruinous causehad settled on it, it would have faced destruction across itsbreadth and massive catastrophe. The reason why we are seenby each other to be mortal is precisely that we fall ill with thesame diseases as those people whom nature has removed fromlife.

Unmistakably, the Epicureans have taken account of the Theophrasteanreply and tried to turn it to their own advantage, arguing that to cite cat-aclysmic events like conflagrations and floods as explaining the newnessof the crafts is to allow that the world might perish in those events.25

Lucretius’ description even echoes Theophrastus in certain details.26 Inparticular, where Theophrastus, echoing a simile at Tim. %&a$, had spokenof the cataclysms as ‘diseases’ (./" '/$, l. %**, Aet. !)") which eventually

". The fourth argument !$&

25 But for the interruption at &(!–$", it might seem natural to read &#*–)!( as continuing this sameargument. But note that while &&#–(* has a highly hypothetical character, provoked by the needto answer Theophrastus, &#*–)!( changes tone and argues from mythical evidence that thereactually have been a conflagration and a flood. While this may indirectly echo Tim. %%c, it doesnot seem to be part of the reply to Theophrastus.

26 Cf. Bignone (!"&+) ,, )$(. The attribution of the flood to heavy rain making rivers overflow is afurther shared feature of the two passages, but no doubt too commonplace to carry any weighthere.

abate, Lucretius too calls them diseases (morbis, &)(), but adds pointedly –drawing on the lessons of book ,,, (especially )$%–&) – that in our ownexperience there is an intimate link between disease and mortality.

It emerges, then, that the palpable dialectical interplay betweenLucretius - &%)–(* and Theophrastus’ presentation and refutation ofthe fourth argument for the world’s impermanence is so tight that it canonly be explained on the assumption that the Epicureans were respond-ing directly to Theophrastus.

& . 526 52,7< 37BA;6?5

Moving backwards now to the third argument, we will be able to detecta similar pattern. This is the argument from ‘the dissolution of each ofthe parts of the whole’. The expansion of this at )!–$" (Aet. !%)–") con-tains some strange intrusions – not just what we have seen to be Philo’sown inserted remarks about the ‘pneumatic tension’ of stones, but alsoa long and grotesque comparison of fire to suicidal Indian snakessquashed under the fallen bodies of the elephants they have just killed.(Whether you blame this latter oddity on Theophrastus or on Philoseems arbitrary – it simply depends on which one you are prepared tohave the worse opinion of.) It can be outlined as follows. Each of the fourcosmic masses – earth, water, air and fire – consists of parts which even-tually perish: stones crumble, water becomes putrid, air becomes sickly,fire burns itself out. Therefore the universe, consisting of these mortalparts, is perishable too.

Lucretius at exactly the corresponding point – - %&(–&%&, immediatelypreceding his version of the fourth argument – similarly argues that theworld has a temporal beginning and end on the ground that each of thefour cosmic elements is unstable. Significantly he even, twice over, takesthe four elements in the same order as Theophrastus. This sequence –earth, water, air, fire – follows the cosmic strata outwards from earth to theheavens, and is not used elsewhere by Lucretius except in a passage where,unlike here, he is specifically describing the layered structure of the world(- ))"–(", )"(–#; contrast the di/erent orders of the elements adopted at, (+$, $!(, $)), - !)%–&, %)#–"). This correspondence helps confirm theimpression that he is taking his lead from the Theophrastean text.

Theophrastus’ own reply to the version of the argument which hereports is that ordered cyclical change of the elements into each otherpoints if anything towards the everlastingness of the world (cf. Aristotle,GC &&$a! / .): destructibility would follow only if all the parts perishedsimultaneously. This time the Lucretian version less obviously contains

!$) &. The imprint of Theophrastus

any direct response. But he does take up the point about cyclical change,and develops it in a di/erent direction. He describes each of the fourcosmic masses neither as perishing tout court, as in the argument reportedby Theophrastus, nor as perishing into the neighbouring element, as inTheophrastus’ reply, but as undergoing an internal cyclical change in theform of dispersal and accumulation. Earth is eroded, but replenishedwith decaying organic matter. Sea and rivers evaporate but are refilled.The air is constantly both absorbing and depositing particles shed byother things. And the sun, moon and stars are constantly both sheddinglight and renewing their supply of it.

How is this meant to prove their overall destructibility? There seemsgood reason to believe that Lucretius is reproducing, if imperfectly,Epicurus’ own revision of the argument reported by Theophrastus. Forwhere that argument had inferred the perishability of the world from theperishing of its parts, we are told that Epicurus, like Lucretius in ourpassage, inferred it from the change which the world’s parts undergo.27

Adding the clues supplied by Lucretius, we can perhaps reconstruct thefull Epicurean argument as follows. Theophrastus has argued that, pro-vided the perishing of the elements is their cyclical destruction into oneanother, there need be no implication of eventual overall destruction.But on the contrary, such radical change would itself amount to instantannihilation (see Lucretius , $#%/.), leaving the natural cycle with noenduring substrate to account for its evident regularity – the regularityinvoked by the main Epicurean arguments for atomism. In reality aprocess like, say, evaporation, which Theophrastus calls the destructionof water into air, is the dispersal of water’s unchanging atomic particlesinto the atmosphere. So too quite generally, the destructionTheophrastus speaks of takes the form of the dispersal and accumulation ofeach cosmic mass. This does, unlike Theophrastus’ theory, account forcosmic regularities. But if it is possible for a thing to be dispersed in part,it is at least possible for it to be dispersed in its entirety.28 And giveninfinite time, every possibility is sooner or later realised.29 Thereforeeach cosmic mass will eventually disperse in its entirety.

Read this way, the Epicurean argument looks like a considered reply

#. The third argument !$(

27 Scholion on Ep. Hdt. $): %1+#-/&" ) %3'$ -/&! ) ,/" 'µ/&), µ"-+2+77/" .-(. -(' . µ"#(' ..28 For a version of Theophrastus’ argument which tries to refute this inference, see Alexander,

Quaest. , %& (transl. with notes by Sharples (!""%)), which could conceivably be responding to theEpicureans.

29 For Epicurean use of the principle of plenitude, see fr. %++ Usener, and cf. Lucr. , %&%–&, ,,,#((–+!. The same principle underlies the argument of Melissus which started this entire debate(B$.%): ‘If it were to become changed by a single hair in !*,*** years, it would all perish in thewhole of time.’

to the debate constructed by Theophrastus, accepting his criticism thatthe original argument failed to allow for the cyclical character of ele-mental change, but re-analysing that cycle in atomist terms which justifythe expectation that the cosmic order must eventually disintegrate.

) . 526 @,79 5 3?< 9618?< 37BA;6?5 9

I hope by now to have made a su0ciently persuasive case for reading theentire sequence of arguments at Lucretius - %&(–&(* as a systematicEpicurean defence against Theophrastus of the third and fourth argu-ments for the world’s impermanence, as reported and criticised by him.

The first argument, from ‘the unevenness of the land’, is that, if theworld had existed from infinite time past, erosion would by now havecompletely flattened the mountains (ll. )–!$, Aet. !!#–!"). Theophrastus’reply (ll. "!–!%#, Aet. !&%–$) is that, mountains being volcanically caused,their erosion is counterbalanced by new growth. It is worth noticing thatLucretius does in passing, at the end of his version of argument &, echothis same argument. After working his way through the four elementalmasses, as we saw, he adds to his list of unstable items the two things withthe strongest claim to indissolubility: solid rock (- &*+–!$) and the vaultof heaven (&!#–%%). The fragility of solid rock is illustrated not only withthe decay of buildings but also with mountain erosion (- &!&–!$):

Do we not see rocks roll down, torn from high mountains, unable to endure themighty force of a finite timespan? For they would not suddenly be torn awayand fall if they had from infinite time past su/ered without damage all the harshtreatment of the ages.

This pointedly recalls the Theophrastean erosion argument (betweenthem, in fact, they seem to be the only recorded arguments inferring theworld’s impermanence from the timescale of mountain erosion). But thetactic has changed. The original argument had relied on the incomplete-ness of mountain erosion to show that it had started only a finite numberof years ago. Lucretius’ version instead uses the actual phenomenon ofmountain erosion to show that even the most solid objects, like lumps ofrock, are subject to dissolution within a finite timespan – hence the sameshould be true of the world as a whole. Crucially, this reworked versionis not directly vulnerable to Theophrastus’ objection. We may put itdown as yet another Epicurean rehabilitation of the arguments pilloriedby Theophrastus.30

!$+ &. The imprint of Theophrastus

30 Arnim (!#"&) notices the connexion between the Theophrastean and Lucretian arguments, butis clearly unjustified in inferring that Theophrastus was himself reproducing an Epicurean argu-

Finally, the second argument is the one from ‘the withdrawal of thesea’ (ll. !#–)*, Aet. !%*–&). It may be summarised as follows. The sea hassteadily diminished in size, as is shown (a) by islands which have withinrecorded history appeared out of the sea, (b) by the pebbles and seashellsfound in inland areas. We may infer that earth and air are likewisediminishing, and that all three elements will end up as fire.

Theophrastus objects (ll. !%"–(#, Aet. !&#–)%) that there is just as muchevidence that elsewhere the sea has spread in historical times, so that anequilibrium between land and sea is being maintained. Once again, hisreply is directly drawn from Aristotle (see Meteor. , !); cf. Theophrastus,Metaphysics !*a%#–"), but the actual evidence he cites is more widelygathered, including the Atlantis myth cited from the Timaeus.

Unlike the other three, Lucretius altogether omits this argument. Tohave adopted the premise that the sea is diminishing would clearly havemilitated against the claims he makes at - %&(–&%& when rehabilitatingthe third argument, namely that each cosmic mass, including the sea,undergoes cyclical change in the form of dispersal and counteractingreplenishment.31 For once, the Epicureans may have thoughtTheophrastus’ rebuttal well founded.32

( . 526 478-6?3?16 8@ 526842739 5A9 @7 . !#)

We can now turn to the character of the work in which Theophrastuspresented and criticised the four arguments. I have already noted twofeatures. One is that right from the outset he presents the arguments notin their own right but as materials for a dialectical confrontation. Theother is that although the rebuttals clearly accord in general terms withhis own Aristotelian views, they incorporate a good deal of materialdrawn from non-Aristotelian sources, such as the Timaeus with its thesisof periodic conflagrations and its Atlantis myth. It is as if the dialecticalconfrontation which he is engineering is meant not to be overtly onebetween Aristotelians and non-Aristotelians.

I can now add to these observations one about the heterogeneousprovenance of the four arguments themselves. It is clearly a mistake totreat them as a unified body of argument, traceable to a single philoso-pher or school. Nothing is known of the origins of the first and fourtharguments, but something can be said about the other two. The second,

$. The provenance of Theophrastus fr. !(% !$$

ment. Quite apart from the chronological problem, this overlooks the important di/erencebetween the two versions.

31 Cf. Lucretius -, +*#–&# on the equilibrium between the sea’s depletion and replenishment.32 Cf. Bignone (!"&+) ,, )"%–&.

from the sea’s withdrawal, and also Theophrastus’ own refutation of it,are both drawn directly from Aristotle, Meteorologica , !) (especially&(%a!$ / .). Aristotle in fact later (,, &, &(+b+ / .) names Democritus as itsauthor. However, we may doubt whether Theophrastus wanted to attachit exclusively to Democritus. For one thing, we know that he spoke ofAnaximander and Diogenes as proponents of the thesis that the sea willeventually dry up.33 For another, he also incorporates into the argumenta reference to seashells on dry land as testimony to the sea’s withdrawal,an observation known to stem from Xenophanes.34 Finally the furtheridea, also incorporated into the argument, that all the elements willeventually turn to fire could hardly be represented as Democritean, butmay on the other hand reflect Theophrastus’ reading of Heraclitus –that the world as a whole originates from fire and eventually returns intoit.35 It seems, in short, that Theophrastus was trying to associate theargument with a wide range of earlier philosophers, and not to recordthe views of any one person.

There is certainly no necessity to see in the second argument a specificallusion to the Stoic ‘conflagration’ (ekpyrosis) theory – even though theStoics did in due course take up the Theophrastean interpretation ofHeraclitus, as well, incidentally, as adopting a version of the secondargument itself.36 Nor need the third argument – the one from the per-ishability of the world’s parts to that of the whole – presuppose Stoicism,as many have thought, despite the fact that a Stoic version of it isrecorded.37 Zeller reasoned that, since it was Aristotle who made theworld’s eternity an issue for the first time, the objections recorded byTheophrastus must be post-Aristotelian, and that Zeno is the first suit-able proponent to assign them to. But here it may be objected that thedebate had multiple sources – witness for instance Democritus’ contribu-tion – and that it stemmed in large measure from Plato’s Timaeus and thecontroversy, which was to run thoughout the remainder of antiquity, asto whether its description of the world as 4".3-/" ) refers to a literalgenetic origin.

Speusippus and Xenocrates took creation in the Timaeus as a purely

!$# &. The imprint of Theophrastus

33 Theophrastus fr. %%!. 34 Hippolytus, Ref. , !).(.35 Fr. %%(. See e.g. Kirk (!"()), pp. &!#/., &%$/., Kerschensteiner (!"((), for the generally accepted

view that Theophrastus (i) interpreted Heraclitus this way and (ii) influenced the later Stoicisingpresentation of him. (Long (!"$(–+) disputes (ii) but accepts (i).) The evidence is not unambigu-ous, but his attribution (fr. %%(, !#–%!) to Heraclitus of a fated time-limit on cosmic change isprobably a development of Aristotle’s view (De caelo , !*, %$"b!)) that Heraclitus agrees withEmpedocles in making the world everlastingly move between alternate generation and destruc-tion. 36 DL -,, !)! fin. 37 DL -,, !)!.

expository device, but it can hardly be doubted that at least some earlyPlatonists, like their ex-colleague Aristotle38 in their own day and theirfellow-Platonists Plutarch and Atticus in the second century 3<,defended the genetic reading. It is easy to imagine that the third argu-ment’s inference, from the perishability of the world’s parts to that of thewhole, originated in the interpretation of its mirror image at Tim. %#b:the argument that the world is something ‘which becomes’ (4$4./" µ"./.)and that it therefore, as a whole, ‘has become’.

We are told (frr. %)!3–') that Theophrastus in his work On physicalopinions (!"#$! -(' . %&'$,(' . 5/@(' .), while acknowledging theSpeusippean interpretation of the Timaeus as a possible one, based hisown criticisms of Plato on the genetic reading. It is more than likely thatthis was the context from which fr. !#) is drawn. Jaap Mansfeld hasconvincingly argued that this massive work – probably in eighteenbooks, and variously called Physical opinions or On physical opinions – wasnot a history of physics, but a collection of materials for use in dialecti-cal debate about physical issues, and included Theophrastus’ owncounter-arguments to the positions supported by others.39 The dialecti-cal structure that has emerged in fr. !#) fits this description to perfec-tion.40

+ . ;656878:8BD

In Chapter (, §#, we encountered Theophrastus’ meteorological text,conjecturally identified as the Metarsiologica, an abridged form of whichis extant in Syriac and Arabic.41 Since the first partial discovery andpublication of this text in !"!#, and the subsequent work ofReitzenstein,42 it has been recognised that it bears a direct causal rela-tion to Epicurean meteorology, especially as we meet this in book -, ofLucretius. Theophrastus does not just anticipate the Epicurean methodof multiple explanations. Many of the specific explanations which he

&. Meteorology !$"

38 Aristotle (De caelo %$"b&%) attributes the non-genetic reading to ‘some people’, which is not mostobviously interpreted as signalling the unanimous view of the Academy. On Aristotle’s reading,see also Baltes (!"$+), pp. (–%%.

39 Mansfeld (!""%b). I also accept his persuasive defence of the title as (On) physical opinions, not (On)the opinions of the physicists. The significance of this, as he shows, is that the work concentrated onthe opinions themselves, not on their authors.

40 This is the work to which Diels (!#$"), pp. !*+, )#+–"!, following Usener, assigned the fragment,numbering it Phys. op. fr. !%.

41 The full text of this was published for the first time by Daiber (!""%). That it is an abridgementof the Metarsiologica (a two-book work), and not as Daiber thought the whole of it, is arguedconvincingly by Mansfeld (!""%a). 42 Bergsträsser (!"!#), Reitzenstein (!"%)).

catalogues are e/ectively identical to those in Lucretius. Moreover, therecognition of direct influence has become even more inescapable since!""%, when Hans Daiber gave us the first full edition and translation ofthe Theophrastus text. In addition to the many specific aetiologies, andthe methodological principle of multiple explanation, it turns out tocontain a theological excursus which is astonishingly close in content toan equivalent passage in Lucretius -,. Since these links have alreadybeen very e/ectively set out and explored by Jaap Mansfeld,43 I can bebrief in illustrating them.

In Metars. +.+$–#!, Theophrastus o/ers naturalistic explanations fortwo alleged facts: (!) that thunderbolts are more frequent in spring, and(%) that they are more frequent in high places. Both points are later citedin the theological excursus, where Theophrastus argues that thunder-bolts, along with the other meteorological phenomena, are not the workof god (!).!#–%():

If thunderbolts originate in god, why do they mostly occur during spring or inhigh places, but not during winter or summer or in low places? In addition, whydo thunderbolts fall on uninhabited mountains, on seas, on trees and on irra-tional living beings? God is not angry with those! Further, more astonishingwould be the fact that thunderbolts can strike the best people and those whofear god, but not those who act unjustly and propagate evil.

Lucretius, after giving an answer to question (!) – why thunderbolts aremore frequent in spring –44 which simply expands on the content ofTheophrastus’ own, turns to the same question of divine retribution (-,&$"–)%%). His comments include the following (&#$–"", )*), )%!–%):

But if it is Jupiter and the other gods who shake the gleaming precincts ofheaven with spine-chilling noise, and who hurl fire wherever it pleases each ofthem, why do they not ensure that those who have recklessly committed abom-inable crimes are struck and breathe out flames of lightning from their piercedbreast, as a grim warning to mortals? Why instead does someone with nothingshameful on his conscience, for all his innocence, get engulfed in flames andswallowed up and swept away by whirlwind and fire descending from heaven?Why, too, do they waste their e/orts aiming at deserted places? Is it that whenthey do that they are excercising their arms and hardening up their muscles?Why do they allow their father’s bolt to be blunted in the ground? And why doeshe himself let them, instead of saving it for his enemies? . . . How does he come

!#* &. The imprint of Theophrastus

43 Mansfeld (!""%a).44 -, &($–$#. As Mansfeld (!""%a), p. &%+ notes, Lucretius adds autumn. This may just be his, or

Epicurus’, addition to the explanandum, since it is in fact an awkwardness for Theophrastus thathis explanation of thunderbolts’ frequency in spring should imply that they are equally commonin autumn (which may be why he avoids mentioning autumn at all in his explanation).

to hurl it into the sea? . . . Why does he usually aim for high places, and why dowe see most traces of his fire on the tops of mountains?

Taken severally these questions might sound like stock jokes, and indeedone or two of the same digs at traditional religion are made elsewherein Lucretius’ poem.45 But the asking of these very same questions by ourtwo authors in the very same context, one in which both are to a largeextent invoking the same multiplicity of naturalistic explanations for thesame set of meteorological phenomena, rules out coincidence. Therereally is a shared body of argument here, and we are dealing with oneof those rare topics on which Epicureans and Aristotelians could see eyeto eye – the exclusion of direct divine causation from the sublunaryworld.

Before the full publication of the Theophrastus fragments, it wasusually thought that this text was a doxographical one, probably part ofthe Physical opinions, cataloguing the meteorological explanations ofearlier philosophers. It is now at least clear, as Daiber himself has shown,that the text is arguing Theophrastus’ own case. However, it would berisky to press for too sharp a contrast here. The views catalogued byTheophrastus in it are, as a matter of fact, largely collected from his pre-decessors (as Daiber’s commentary brings out), and even in this abridge-ment46 their defence is in places put in a form which implies a dialecticalconfrontation. Besides, we must remember that it is equally true of thelong fr. !#) on the world’s destructibility, which is certainly doxograph-ical and probably from the Physical opinions, that (a) Theophrastus isarguing for his own view most of the time, and (b) the views of earlierphilosophers are gathered and synthesised into a single whole, withoutany apparent emphasis on their attribution to individuals. The meteor-ological text di/ers from fr. !#) at least to the extent that Theophrastusis arguing for, not against, the position synthesised out of the work ofearlier thinkers, but we have no reason to suppose that that too did nothappen sometimes in the Physical opinions.

However, it is not my aim to close altogether the gap between theMetarsiologica and the Physical opinions. What I want to point out is howeasily the very same material as we read in the former might have beenredeployed in the latter, for di/erent but closely related purposes.

&. Meteorology !#!

45 ,, !!*!–).46 Mansfeld (!""%a), who is responsible for the proposal that it is an abridgement, points out that it

is far too short to be the whole of the Metarsiologica, which ran to at least two books. However, Iam much less confident than he is that the abridgement involves only the omission of some wholesections, with no condensation within the surviving sections (cf. Mansfeld’s n. #, p. &!+).

There is in fact very good reason to suspect that the Epicurean argu-ments in Lucretius -, are based, not on a direct reading of theMetarsiologica, but on the equivalent section of the Physical opinions.Although the actual material is largely the same in Lucretius -, and theMetarsiologica, its organisation di/ers just enough to suggest that the latteris not the direct source. For one thing, there is relatively little detailedcorrespondence between the order of explanations adopted for eachphenomenon.47 More significant, however, is the position of prester in theexposition. I remarked in Chapter ( (see p. !(" above, and Chart ) on p.!(#) that Lucretius, like the doxographers, places this phenomenon in anearly position (-, )%&–(*), before cloud and rain, thus respecting themajority ancient view that it is fiery in nature, descending from theheavens. I observed that this does not reflect Lucretius’ own Epicureanview of it, which derives it purely from cloud, and that the order followedcan therefore be seen to be determined by a doxographical source, notby considerations internal to Epicureanism. Now in the MetarsiologicaTheophrastus shows himself more or less to share the Epicurean view ofprester, which he may well have influenced. And since this is not adoxographical work, he for once departs from the doxographical orderand instead subsumes prester under the heading of winds (Metars.!&.)&–()). This discrepancy between Lucretius and the Metarsiologicaconfirms the impression that Lucretius’ material, even here, is likely tobe most directly derived not from the Metarsiologica but fromTheophrastus’ doxographical presentation of the same material in thePhysical opinions.

$. 526842739 5A9 , 64,1A7A9 3?< :A1765,A9

We have now examined two major parts of Lucretius’ exposition inbooks - and -, which incorporate or respond to Theophrastus’ work,very probably in his Physical opinions. Given our earlier findings inChapter (, §§+ and #, it hardly needs arguing here that Lucretius drewthis material, not directly from Theophrastus, but from Theophrastus’younger contemporary Epicurus, whose On nature .,–.,,, he proves tohave been following so closely throughout these two books.

It is now time to recall that a similar role for Theophrastus lookedlikely in a previous context. Both Epicurus and Lucretius, we saw (Ch.

!#% &. The imprint of Theophrastus

47 E.g. for the explanations of thunder, see Reitzenstein (!"%)), p. %+, or the Lucretian parallels citedby Daiber (!""%), pp. %$%–&. Even for the theological excursus examined above, Mansfeld (!""%a),pp. &%+–$, points out that the same arguments are used but in di/erent orders.

), §!*, Ch. (, §)), classify previous physical theories on the Aristotelianprinciple which Theophrastus made the standard doxographical one:monism, finite pluralism, infinite pluralism. They also seem reliant onthat same doxographical tradition in their actual presentation of thesedoctrines – especially in using the rubric of ‘homoiomereia’ forAnaxagoras’ theory.48 It is worth recalling that On nature .,- seems to endthe critique of the finite pluralists with Plato: this may reflect the appar-ent fact that Theophrastus took his doxography down only as far asPlato, and omitted his contemporaries, including Aristotle himself.49

Another possible case of Theophrastean influence is the following.The better-preserved parts of On nature .,- include criticisms of Platowhich, as Wolfgang Schmid pointed out, can be thought to reflectAristotle’s in De caelo ,,, $–#.50 In view of what we have already seen ofTheophrastus’ methods, it will be both plausible and economical toguess that the Aristotelian criticisms had been redeployed by him in hisreport and critique of Plato’s element theory, and were in turn taken upby Epicurus.

This leads me on to a further consideration. Epicurus in generalshows little, if any, positive influence from Aristotle’s own theories andconcepts. At least, attempts to locate such influence in his thought, forinstance in his theory of free will, have in every case proved at the veryleast controversial. And we should indeed be wary of expecting him tohave looked, for positive ideas, to a philosophical tradition so alien to hisown.51 But he does much more recognisably show a negative Aristotelianinfluence. In particular, his theory of minima involves a detailed attemptto circumvent Aristotle’s criticisms of comparable theories in Physics -,,52

and his treament of void probably responds similarly to the argumentsagainst void in Physics ,-.53 I long assumed that the Physics was oneAristotelian text which, because of the threat it posed to atomism,Epicurus did take the trouble to study.54 But in view of the pattern wehave seen, it seems much more likely that those Aristotelian criticismsreappeared in Theophrastus alongside his doxographical reports of

'. Theophrastus, Epicurus and Lucretius !#&

48 Pp. !%(>+, !)+ above; cf. Schofield (!"$().49 This is certainly true of the De sensibus, and, as far as I know, is confirmed by silence about later

philosophers in his fragments. 50 Schmid (!"&+). 51 Cf. Sedley (!"#"b), pp. !!$–!#.52 This was shown above all by Furley (!"+$). For further development of the same theme, cf.

Long/Sedley (!"#$), §§" and !!. 53 See Inwood (!"#!).54 Not much can be made of Philodemus, P.Herc. !**(, fr. !!! (Angeli (!"##), pp. !++–$), in which

a letter by a first-generation Epicurean, not necessarily Epicurus himself, is quoted as mentioningAristotle’s Analytics and !"#$! %&" '"(). Even if one assumes that this latter is the Physics, there isno indication whether he has read it, or, if he has, at what date (the immediately following letteris dated %#*/%$", which would be too late to play a part in our story).

atomism, and once again that may well be the route by which they foundtheir way to Epicurus.

Aristotle’s school treatises appear to have been relatively neglected inthe Hellenistic period.55 (It would be a curious irony if this proved to belargely due to the Hellenistic philosophers’ reliance on Theophrastus’doxography, which excluded Aristotle from its scope!) Such a conclusionis not universally shared, but when the denial of it appeals, as it some-times does, to likelihood – how could they have ignored Aristotle? – thatis uncomfortably reminiscent of the widespread, but equally unfounded,refusal to believe that Lucretius could have ignored Stoicism (see Ch. &).

Although numerous attempts have been made to discern Aristotle’sdirect influence on Epicurean texts, including the poem of Lucretius, thestrongest objection seems to me to lie in the contrast with Theophrastus.At least some of the cases where Theophrastus has influenced an argu-ment in Lucretius are quite simply unmistakable, in spite of the fact thatonly a tiny fraction of Theophrastus’ work has survived. I hope that theargument over the world’s destructibility, which has taken up the bulk ofthis chapter, will commend itself as one such case. But at any rate whatwe have seen to be the (for once positive) influence of Theophrastus’meteorology on Lucretius book -, is so transparent that as far as I knowit has never been denied. Why are there no equally unmistakableAristotelian imprints? The di/erence makes it much more believablethat it was primarily through Theophrastus, and not through the directimpact of Aristotle’s treatises, that Aristotelianism helped shape theEpicureanism which we can read in the poem of Lucretius.

It is noteworthy that Epicurus wrote a work Against Theophrastus, in atleast two books, the only surviving fragments of which concernphysics.56 Since the Physical opinions was serving Epicurus as his source-book for earlier views, not for Theophrastus’ own, On nature itself cannotalways have o/ered him adequate opportunity for direct replies to theless acceptable of Theophrastus’ own physical opinions. That he should,as an o/shoot, feel the need to write a separate treatise againstTheophrastus becomes entirely understandable.

It is surely not an accident that no corresponding treatise ‘AgainstAristotle’ is attested. There is no need to suppose that what drewEpicurus to Theophrastus was an interest in Aristotelianism as such. Wetherefore have no warrant for expecting further Aristotelian influences

!#) &. The imprint of Theophrastus

55 I here declare a large measure of agreement with Sandbach (!"#().56 Epicurus frr. %"–&* Usener, where the topic is whether colour exists in the dark.

to show through in our Epicurean texts, ones acquired through a readingof Peripatetic doctrinal writings. The Theophrastus who made his markon Epicurus, and as a result can be detected again and again in the pagesof Lucretius, was not – except incidentally – Theophrastus theAristotelian. Rather, it was Theophrastus the collector, synthesiser andcritical purveyor of doctrines from the earlier history of physics.

We have encountered a substantial quantity of Theophrastean materialin Lucretius. That Lucretius has absorbed this through Epicurus, andnot by his own direct reading of Theophrastus, seems overwhelminglyprobable. The findings of Chapter & make it intrinsically unlikely thatLucretius would strike out on his own in such a way. For Epicurus, onthe other hand, Theophrastus’ Physical opinions was a recently publishedencyclopaedia of physical argument, the first work of its kind and anideal source for someone who, in writing a global treatise on physics,sought a synoptic view of the state of play down to his own time. We inany case know enough about Epicurus’ own debt to Theophrastus inmeteorological methodology to be quite confident that here at leastTheophrastus’ influence on Lucretius was mediated by him. In thischapter I have tried to show why, in other areas too, the visible responseto Theophrastus is most likely to reflect Epicurus’ polemical concerns,rather than Lucretius’ own.

'. Theophrastus, Epicurus and Lucretius !#(

1234567 $

The transformation of book I

! . 526 18?56?5 9

I argued in Chapter ( that Lucretius, for reasons perhaps connected withthe enormous size of his Greek source text, initially drew large quanti-ties of material from it en bloc, following its order of exposition fairlymechanically. In a second phase, he set about reorganising it into thecarefully structured six-book poem that we know. But he did not live tocomplete the task. By the time of his death he had got as far as revers-ing books ,,, and ,- into their present numbered order; and he had, asfar as I can tell, fully reworked the contents of books ,–,,,. However, hehad plans for the reorganisation of books ,-–-, which can be recoveredfrom his proems, but which he did not live to put into e/ect. In theirpresent state, books ,-–-, to a large extent simply reproduce thesequence of the corresponding books of On nature. It is likely that thesame was true of the opening books in the first phase of composition.

In this chapter I want to give an idea of what the completed reorgan-isation of his material in books ,–,,, may have involved. But I shall selectfor the purpose the contents of book , only. Even here I shall largely passover the proem, which I have dealt with separately in Chapter !.

Chart ( (pp. !##>" below) shows the overall layout of the book, withequivalences to Chart % (p. !&+ above) added in italics.1

% . D R N , 3?< O N NAT U R E ,> , ,

We have seen in Chapter ( that the basic structure of this book reflectsthat of its source text, Epicurus, On nature ,–,,. This dependence can nowbe observed in one specific detail. If we bypass Epicurus’ methodolog-ical preliminaries in book ,, which are replaced in the DRN by Lucretius’

!#+

21 There is nothing innovative about this table, apart perhaps from my inclusion of $*(–!! in thefinal argument against the monists, rather than treating it as making a new beginning.

own proem, the line at which Lucretius first hooks up to his Epicureansource is precisely identifiable. At !&+–)( he declares his determination,despite the poverty of the Latin language, to illuminate Memmius’understanding of the world – to seek out the poetry ‘by which you maybe able to take a view deep into hidden things’. This line, res quibus occul-tas penitus convisere possis (!)(), which marks the transition to physicalexposition (!)#, naturae species ratioque!Epicurean %&'$/7/4$"+: see abovepp. &$>#) and specifically to the first argument, that nothing comes intobeing out of nothing, unmistakably echoes the words with whichEpicurus marked his own transition to that same opening argument.These are, in the Letter to Herodotus’ presumably condensed version,-+&' -+ 5"! 5$+7+2/" .-+) '&./#+' . [sc. 5"$' ] 3% 53 6"#$! -(' . +& 53" 7(. (Ep.Hdt. &#): ‘Having made these distinctions, it is now time for us to take aview of hidden things.’ The rare compound convisere scrupulously echoesEpicurus’ characteristic verb '&./#+' ..

Thereafter the parallelism between On nature ,–,, and Lucretius , is soclose that coincidence can be ruled out. The above table of comparison(Chart %b) borrows its contents from Chart % (p. !&+).

From (ii) onwards, Epicurus’ list contains just two items which do notoccur at precisely the same point in Lucretius’ sequence. Item (iv) is theonly significant omission, and I shall consider shortly Lucretius’ motivefor omitting it. The other, (vi), does not occur in the Letter to Herodotuseither at this point: in both Lucretius and the Letter it occurs only in itsrepetition at (viii), where it does its main work.

Of the items present in Lucretius but absent from Epicurus’ epitome,

". DRN I and On nature I–II !#$

Chart 'bNat. DRN, (i) methodological preliminaries —

(ii) nothing comes into being out of nothing , !)"–%!)(iii) nothing perishes into nothing , %!(–%+)(iv) the all never changes —

(L!) the existence of the invisible , %+(–&%#(L%) the existence of void , &%"–)!$

(v) the all consists of bodies and void , )!#–%"(vi) some bodies are compounds, others constituents —

,, (vii) nothing exists independently except bodies and void , )&*–#%(viii) bodies’ constituents (distinguished in (vi)) are atomic , )#&–+&)

Criticisms of rival theories of the elements I &#$–*"+Interlude: Lucretius’ poetic mission I *"!–$+

(ix) the all is infinite , "(!–!*(!(L&) critique of geocentric cosmology , !*(%–!!!&Closing words , !!!)–!$

!–!"

#P

roem

(!–)

& ) P

raye

r to

Ven

us()

)–")

The

det

ache

d na

ture

of

divi

nity

((*–

+!) T

opic

of

book

!: a

tom

s(+

% –$"

) Epi

curu

s as l

iber

ator

(#* –

!*!)

The

evi

ls of

relig

ion

(!*%

– &()

Wro

ng a

nd r

ight

view

s on

life

afte

r de

ath

(!&+

–)()

Luc

retiu

s’po

etic

task

! "$

–%$

"T

he

pri

nci

ple

s of

con

serv

atio

n(!

)+–#

) The

phi

loso

phic

alta

sk(!

)"–%

!)) N

othi

ng c

omes

into

bei

ng o

ut o

fno

thin

g !(ii

)!

(!("

– $&)

regu

lar

plac

e of

grow

th%

(!$)

– #&)

regu

lar

seas

on o

fgr

owth

&(!

#)–!

"!) r

egul

ar d

urat

ion

ofgr

owth

)(!

"%–#

) reg

ular

dep

ende

nce

on n

utri

tion

((!

""–%

*$) u

pper

lim

its o

n siz

e+

(%*#

–!) )

e/

ects

of

plou

ghin

g

(%!(

–+) )

Not

hing

isde

stro

yed

into

noth

ing !

(iii)

"!&

–&%

On

ly b

ody

and

voi

dh

ave

per

seex

iste

nce

!(v

ii)!

()!#

–%")

The

‘all’

is bo

dies

and

void

%()

&*–)

# ) N

o th

ird p

er se

thin

gis

conc

eiva

ble

&()

)"–(

#) A

ttrib

utes

(con

iunc

taan

d ev

enta

) do

not e

xist

per

se)

()("

–+&)

Nor

doe

s tim

e(

()+)

–#%)

Nor

do

fact

s abo

utth

e pa

st

"&

'–$

'"B

ody

exis

ts i

n t

he

form

ofat

oms

!(v

iii)

()#&

–(*%

) Thi

s is s

urpr

ising

,bu

t ...

!((

*&–!

* ) th

ere

mus

t exi

stpu

re b

ody

%((

!!–!

" ) w

hat c

onta

ins v

oid

mus

t its

elf

be so

lid&

((%*

–&")

alte

rnat

ion

ofbo

dyan

d vo

id)

(()*

–(* )

con

serv

atio

n of

mat

ter

(((

(!–+

)) im

bala

nce

ofde

stru

ctio

n an

d ge

nera

tion

+((

+(–$

+) so

ft an

d ha

rd$

(($$

–#&)

dur

abili

ty o

fst

u /s

#((

#)–"

#) re

gula

rity

of

spec

ies

"((

""–+

!)) i

mpo

ssib

ility

of

deta

chm

ent o

fm

inim

al

part

s

!($

)%–(

) om

issio

n of

void

%($

)+–(

% ) a

dmiss

ion

ofin

finite

divi

sibili

ty&

($(&

–#) s

oftn

ess o

fel

emen

ts)

($("

–+% )

mut

ual h

ostil

ity o

fel

emen

ts(

($+&

–") i

fel

emen

ts b

ecom

eot

her

stu /

s,w

hy c

all t

hem

mor

e ba

sic?

+($

$*–#

!) o

r,if

not,

they

lack

expl

anat

ory

pow

er$

($#%

–#*%

) if

ther

e is

inte

rtra

nsfo

rmat

ion

ofel

emen

ts,th

ey p

erish

#(#

*&–%

") a

tom

ism d

oes t

hesa

me

job

bette

r

(#&*

–"%*

) Ana

xago

ras !

(xxx

viii)

(#&*

–)+ )

Ana

xago

ras’

theo

ryex

plai

ned

!(#

)$–(

# ) h

is st

u/s a

repe

rish

able

%(#

("–$

)) e

very

thin

g w

illco

nsist

of

stu /

s unl

ike

itsel

f&

(#$(

–"+ )

‘the

re is

eve

ryth

ing

in e

very

thin

g’ca

n be

refu

ted

empi

rica

lly)

(#"$

–"!)

) tra

nfor

mat

ions

are

bette

r ex

plai

ned

by a

tom

ism(

("!(

–%* )

redu

ctio

ad a

bsur

dum

of‘th

e pa

rt re

sem

bles

the

who

le’

Ch

art

#

!(%

!$–%

) ) d

estr

uctio

n w

ould

be

inst

anta

neou

s%

(%%(

–&$ )

mat

ter

wou

ld h

ave

run

out b

y no

w&

(%&#

–)")

all

dest

ruct

ion

wou

ld b

e eq

ually

eas

y)

(%(*

–+))

nat

ural

cha

nge

iscy

clic

al

%$

#–'%

&T

he

exis

ten

ce o

fth

ein

visi

ble

!(L

!)!

(%$!

– "$)

win

d%

(%"#

– &*)

) odo

ur,

tem

pera

ture

,sou

nd&

(&*(

–!* )

eva

pora

tion

)(&

!!–%

! ) e

rosio

n(

(&%%

–#) g

row

th a

nd

shri

nkag

e

'%(

–"!)

Th

e ex

iste

nce

of

void

!(L

")!

(&&(

– )()

mot

ion

%(&

)+–(

$) p

erm

eatio

n&

(&(#

–+")

rela

tive

wei

ght

)(&

$*–#

&) re

dist

ribu

tion

in a

plen

um im

poss

ible

((&

#)–"

$) a

rti fi

cial

cre

atio

n of

void

(&"#

–)!$

) The

re a

re m

any

mor

e su

ch a

rgum

ents.

!*( +

!(–%

$ ) (

. ..w

hich

mus

tex

ist,t

o ex

plai

n siz

edi

/er

ence

s)!!

(+%#

–&))

sepa

rate

d m

inim

alpa

rts w

ould

lack

gen

erat

ive

pow

ers

$'#

–(%

*R

efu

tati

on o

fri

val

theo

ries

of

mat

ter

(+&(

–$!!

) Mon

ists:

Her

aclit

us!

(xxx

vi)(+

&(–)

)) H

erac

litus

isov

erra

ted

!(+

)(–(

) ) fi

re c

ould

not

expl

ain

dive

rsity

of

stu /

s%

(+((

– +))

indi

spen

sabi

lity

ofvo

id&

(++(

–#")

iffir

e be

com

esso

met

hing

else

,it p

erish

es)

(+"*

– $**

) he

turn

s the

sens

esag

ains

t the

mse

lves

(($

*!– !

!) to

cho

ose

fire,

or a

nyot

her

singl

e el

emen

t,is

arbi

trar

y

($!%

–#%"

) Plu

ralis

ts:

Em

pedo

cles

!(x

xxvii

)($

!%–!

() In

trod

uctio

n of

finite

plur

alism

($!+

–)! )

Em

pedo

cles

’mer

its,

and

failu

res .

..

(%

!–#*

Inte

rlu

de:

Lu

cret

ius’

poe

tic

task

(#!

– !*

#!T

he

infi

nit

y of

the

all

!( ix

)Sp

ace

cann

ot b

e bo

unde

d:!

("(#

–+$)

ther

e is

som

ethi

ngbe

yond

any

lim

it%

("+#

–#&)

wha

t if

you

thro

w a

spea

r pa

st th

e su

ppos

edbo

unda

ry?

&("

#)–"

$) in

a fi

nite

uni

vers

e,al

l bod

y w

ould

acc

umul

ate

at th

e bo

ttom

)("

"#–!

**!)

the

‘all’

coul

dha

ve n

othi

ng e

xter

nal t

obo

und

it

Bod

y,as

wel

l as s

pace

,mus

tbe

infin

ite:

(!**

#–(!

) oth

erw

ise it

wou

ldbe

infin

itely

diss

ipat

ed

!*#%

–!!!

'C

riti

qu

e of

geoc

entr

icco

smol

ogy

!(L

#)

!!!"

–!)

Clo

sin

g w

ord

s

some can be briefly dealt with. The interlude at "%!–(*, containingLucretius’ poetic manifesto, is Lucretius’ own contribution, and we needseek no Epicurean original. Likewise the closing words of the book. Item(L&), the criticism of geocentric cosmology, will almost certainly bedrawn from an original critique present in Epicurus, On nature but natu-rally enough omitted from the epitome, which concentrates on Epicurus’positive findings. I have argued for its provenance from Epicurus inChapter &, §+.

As for the refutation of rival theories of matter, I have argued inChapters ) and ( that Epicurus held over his critique of rival theories ofthe elements until he had completed his own physical and cosmologicalexposition in books ,–.,,, of On nature, and that in the De rerum natura thedecision to bring the critique forward to occupy a much earlier place,immediately after the initial demonstration of atomism in the first partof book ,, was Lucretius’ own. From a dialectical point of view the cri-tique now comes surprisingly early. The incapacity of the rival theoriesto explain phenomena, as atomism can explain them, is the main burdenof the critique. Yet, although by this point Lucretius has isolated atomsand void as the only independently existing things, he has done none ofhis work on their power to account for the full range of phenomena: thatis the task especially of book ,,, and to some extent of the entire remain-der of the poem. Methodologically, Epicurus’ order is sounder thanLucretius’. Lucretius nevertheless gains much by the transposition. HereI would list three main advantages achieved.

First – and this is well enough recognised2 – the critique enablesLucretius to broach the theme of how philosophy might best be written.Paradoxically, it is the prose-writer Heraclitus who is pilloried for seekingto persuade by mere innuendo (+&#–))), while the poet Empedocles,Lucretius’ proudly acknowledged literary forebear, is praised for hisgodlike lucidity of exposition.3 This contrast insinuates onto the agendathe very theme – the defence of philosophical poetry – which will moveto centre stage in the famous poetic manifesto ("%!–(*) immediately fol-lowing the critical section.

Second, one of atomism’s greatest claims to fame is its explanatoryeconomy. In the critical passage, Lucretius twice illustrates this with hisfavoured alphabetic analogy: you can account for any observed material

!"* '. The transformation of book I

22 Kollman (!"$!), Tatum (!"#)); cf. Milanese (!"#"), pp. !&)–".23 Cf. above pp. !&>!), where I argued that the comparison of Empedocles with the Delphic oracle

at , $&#–" is meant to contrast him favourably with the oracle’s proverbial ambiguity, associatedalso with Heraclitus.

change as mere rearrangement within a modest stock of atomic types,just as shifts between one word and another can be achieved by redistrib-uting a small set of alphabetic letters (#!$–%", "*$–!)). If friction in thebranches of trees produces fire, atomism e/ortlessly explains how("*$–!)):

iamne vides igitur, paulo quod diximus ante,permagni referre eadem primordia saepecum quibus et quali positura contineanturet quos inter se dent motus accipiantque, "!*atque eadem paulo inter se mutata creareignes et lignum? quo pacto verba quoque ipsainter se paulo mutatis sunt elementis,cum ligna atque ignes distincta voce notemus.

So do you now see, as I have said before, that for the very sameprimary particles it often makes a huge di/erence with whichothers and in what arrangement they are held, and whatmotions they impart and receive in relation to each other; andwood that the same ones, with a little interchange, create fire(ignes) and (lignum)? Just as the words themselves have only under-gone a small change in their elements (!‘letters’) when we usedi/erent sounds to designate ligna and ignes.

In this alphabetic shift from ligna to ignes, which parallels the combustionof actual wood through friction, the point is not the rearrangement ofan identical set of elements, since the two words are not anagrams, somuch as the exposure, by a disruption of its outer parts, of what alreadylay hidden within a structure – fire in the case of the wood, and its ana-logue the letter-group ign in the case of lignum.

Such explanatory economy is a relative matter, and it is only in thecontext of a comparison of atomism with rival physical theories that thepoint could have been so e/ectively made: how much better to positsimple atomic rearrangement than to believe, with Anaxagoras, thatwood contains fire, and indeed that everything contains everything.Hence, if this merit of atomism was to be adequately exploited, the crit-ical section could not be postponed. The sense of urgency which thisbetrays is our first indication of an important guiding factor in Lucretius’transformation of his Greek material: the need to maximise the persua-sive impact of his argument in its early stages. Epicurus had beenaddressing an already committed and presumably patient philosophicalaudience; Lucretius is only too well aware that his reader may put thepoem down as soon its impact starts to wane (, ")&–(: haec ratio plerumquevidetur | tristior esse quibus non est tractata, retroque | volgus abhorret ab hac).

". DRN I and On nature I–II !"!

Third, the transfer of the critique to this early position helps Lucretiuscreate a major structural feature for the first pair of books. By delayingthe theme of infinity until the end of the book, it generates the match-ing closures of books , and ,,, expanding our horizons beyond theboundaries of our own world. I have dealt with this already in Chapter(, §).

To sum up, the transference of the critical passage to the early posi-tion which Lucretius assigns to it can be seen as part of an elaboratelycontrived restructuring, well explained by the nature of Lucretius’ per-suasive project and the poem’s overall architectonic.

I now turn to Lucretius’ own demonstrable omission: item (iii), Epicurus’argument for the immutability of the ‘all’, which in the Letter to Herodotus(&") is summarised as follows:

Moreover, the all was always such as it is now, and always will be. For (a) thereis nothing into which it changes, and (b) beside the all there is nothing whichcould pass into it and produce the change.

This is a desperately obscure argument, and the reason for its obscur-ity was brilliantly explained by Jacques Brunschwig in a little-knownarticle first published in !"$$.4 What Epicurus actually means – or rather,what his argument will with hindsight turn out to amount to – is that (a)there can be no space beyond the universe for it to lose anything into, and(b) there can be no body beyond it to enter it and add to its contents. Butfor very sound methodological reasons, Epicurus cannot say that here.That the all consists of body and space will be demonstrated only in thenext stage of his argument; it therefore must not be presupposed here,on pain of circularity. And that is why Epicurus must content himselfwith the entirely unspecific assertion that there is simply ‘nothing’outside the universe to permit subtraction or addition.

Playing his cards this close to his chest is entirely germane to Epicurus’philosophical project, but similar reticence would be ruinous if repli-cated so early in book , of Lucretius’ poem. Hence Lucretius omits theargument here, even though he was quite happy in later books to invokeit as soon as the items in question could be unmasked as space and body(,, &*&–$, ,,, #!+–!#!- &+!–&). Whether he omitted it in book , becausehe did not understand it, or because, although he understood it, he

!"% '. The transformation of book I

24 Brunschwig (!"$$), esp. p. !# of the !"") translation. I have some residual disagreements withBrunschwig on points of detail, expressed in Long/Sedley (!"#$) vols. , and ,,, §), but they donot a/ect the present issue.

found it damaging to the lucidity of his own argument, it is hard to becertain. But I tend to the former explanation since, as I shall now bearguing, preserving Epicurus’ methodological rigour in the order of hisexposition was in any case not a consideration to which Lucretiusattached priority. Lucretius, that is, would not have thought it out of thequestion to name space and body prematurely, even where methodolog-ical rigour had prevented Epicurus himself from doing so.

& . 966<9

The proof of this greater flexibility lies in Lucretius’ readiness tosmuggle in countless advance allusions to atoms, long before their exis-tence has been demonstrated – a degree of licence which Epicuruswould never have permitted himself, but which has obvious rhetoricaladvantages to a writer eager to advertise the explanatory merits ofatomism with all speed.5

Epicurus (Ep. Hdt. &#) begins his physical argument by maintaining that

. . . nothing comes into being out of what is not. For in that case everythingwould come into being out of everything, with no need for seeds ('6""#µ+-+).

‘Seeds’ here means just that – biological seeds.6 These symbolise theunchanging regularities of natural processes, whereby a given organismcan only grow from the appropriate seeds, and not emerge out of thinair, as might have been expected if things came into being out of nothingat all.

Lucretius, at the corresponding point (stage (ii)), may appear at first toecho this argument from ‘seeds’ accurately enough (!("–+%):

nam si de nilo fierent, ex omnibu’ rebusomne genus nasci posset, nil semine egeret.e mare primum homines, e terra posset oririsquamigerum genus . . .

#. Seeds !"&

25 This theme of rhetorical anticipations in Lucretius is perceptively explored by Asmis (!"#&). Mymain di/erence from her analysis, so far as book , is concerned, is that I suppose both the DRNand the Letter to Herodotus to derive from Epicurus’ On nature, with each making its own inde-pendent selection of arguments, whereas she treats any departure from the Letter to Herodotus onLucretius’ part as a conscious one, as if it were itself a primary source for him.

26 In Greek Epicurean texts '6""#µ+ never demonstrably signifies ‘atom’. There is no foundationfor the view of Reiley (!"*"), p. &", and Vlastos (!"(*), n. !&, that Epicurus uses the word in thissense. They cite Ep. Hdt. &#, $) and Ep. Pyth. #"; but in $) the ‘seeds’ are apparently biological,and in #", although it is impossible to tell what kind of ‘makings’ of worlds or their parts theword represents, the conjoined metaphor of ‘irrigation’ makes it unlikely that we are dealingwith discrete atoms.

For if they came into being from nothing, every species wouldbe able to be born out of everything, and nothing would need aseed. To start with, human beings would be able to spring up outof the sea, and the scaly tribe out of the land . . .

The imagined flouting of environmental constraints on birth providesthe materials of the wonderful adynaton that follows, and no reader iscompelled to doubt at this stage that the required ‘seed’ is a literal bio-logical one. The singular semine certainly strengthens that impression – asingle organism needs no more than one originative seed, but manyconstituent particles.7 However, first impressions can mislead. Lookwhat follows just lines later (!+$–$!):

quippe ubi non essent genitalia corpora cuique,qui posset mater rebus consistere certa?at nunc seminibus quia certis quaeque creantur,inde enascitur atque oras in luminis exit !$*materies ubi inest cuiusque et corpora prima.

For in a situation where there were no generative bodies for eachkind, how could things have a fixed mother? But as it is, becauseeach kind is created from fixed seeds, it is born and emerges intothe realm of light from the place where each thing’s matter andprimary bodies are to be found.

The ‘seeds’ here reappear in company which radically changes theirprofile. The words which I have italicised are all Lucretian variant termsfor atoms. Precisely the same terms were conveniently announced assuch barely a hundred lines earlier, in the programmatic part of theproem ((#–+!):

quae nos materiem et genitalia corpora rebusreddunda in ratione vocare et semina rerumappellare suemus et haec eadem usurparecorpora prima, quod ex illis sunt omnia primis.

. . . which it is our practice, when giving an account, to callthings’ ‘matter’, their ‘generative bodies’ and their ‘seeds’, aswell as using the regular term ‘first bodies’, because they are theprimary things from which all things come.

The echo of these lines is too sustained and accurate to be accidental.8Hence even within his very first argument against generation ex nihilo

!") '. The transformation of book I

27 In retrospect this consideration will prove not to have been decisive: the singular of semen at !#(,%*+ and %%! undoubtedly functions as an uncountable noun designating a group of constituentparticles.

28 Lucretius’ anticipation here therefore amounts to a great deal more than what Asmis (!"#&), p.

(!+!–$&) Lucretius has appeared to shift ground: by the end of it, thefocus has proved to be not on biological regularity at the level of seedsbut on material regularity: natural species can only grow in environ-ments which provide a stock of the right constituent particles for them.Strictly speaking nothing yet asserts the atomicity (indivisibility) of thesematerial ‘seeds’, but no one need doubt that Lucretius is already hereadvertising the explanatory merits of the very particles with which, aswe have been carefully forewarned, book , will be concerned.

That he is doing so is guaranteed by the entire run of the ensuingarguments against absolute generation and annihilation (!$)–%!+),which constantly invoke the constraints imposed by the availability ofmatter, and which do so by continuing to deploy Lucretius’ pre-announced vocabulary for atoms.9 In his fourth argument, in pointingto the dependence of plant growth on the ingestion of rain, he evenmanages to contrive a first airing for the alphabetic model of atomicrearrangement (!"+–#):

ut potius multis communia corpora rebusmulta putes esse, ut verbis elementa videmus,quam sine principiis ullam rem existere posse.

Hence you can more readily believe that many things sharemany bodies in common, just as we see words share their ele-ments [!‘letters’], than that anything should exist withoutprimary particles.

Here principiis is a quite unequivocal allusion to atoms.Yet all this is done without relinquishing the advantageous biological

connotations o/ered by the atomic vocabulary. The second argumentobserves that things only come into being at times when the right seminarerum (!$+) have ‘flowed together’, a confluence immediately redescribedas the ‘creative union’ (genitali | concilio) of primordia (!#%–&). This lastword, primordia, is another standard term for atoms, already announcedas such back at , ((, and yet who can doubt that in such talk of the cre-ative union in which seeds flow together the language of biological pro-creation is being exploited to the utmost? The regularities whichLucretius is invoking are ultimately material – as they must be if the needfor an atomic substructure is to be insinuated – yet matter is being

#. Seeds !"(

(", suggests, his simple (and perhaps even uncontroversial) assumption that what things are dis-solved into are ‘bodies’.

29 semen/semina !+", !$+, !#", %*+, %%!; materies/materia !$!, !"!, %*&, %)(, %)"; corpora prima !$! (cf.!"+, %)%, %)+, %)"); primordia (along with principia, standardly used by Lucretus as its proxy in theoblique cases, for metrical reasons) !#%, !"#, %!*, %)); genitalia corpora !+$.

treated as if its creative powers were themselves virtually sexual. Theassociation is strongly encouraged by the very vocabulary of genitaliacorpora and materies – the latter derived from mater, with which it ise/ectively identified at !+$–# (quoted on p. !") above).

The same conflation is even more pronounced in the third argument,where Lucretius is appealing to the slowness of plant and animalgrowth, and has already insisted on the time required for the ‘gather-ing of seed’ (seminis ad coitum, !#() in the growth process. Although ‘seed’here must designate atomic matter, the distinction between this andbiological seed becomes even harder to maintain when, at !##–"!, weread

quorum nil fieri manifestum est, omnia quandopaulatim crescunt, ut par est, semine certo,crescentesque genus servant; ut noscere possisquidque sua de materie grandescere alique.

It is evident that no such thing [i.e. instant growth] happens,since all things grow gradually, as one would expect, from a fixedseed, and in growing preserve their kind, so that you can tell that eachgrows and is nourished from the matter proper to it.

The direct association of an organism’s dependence on ‘fixed seed’ withits unalterable membership of a species can hardly fail to put the biolog-ical connotations of the word uppermost in the reader’s mind.

In the sixth argument (%*#–!)) the need for agriculture is attributedto the presence of primordia rerum in the soil which, by use of the plough,we cimus ad ortus. These are of course particles of nutrients. But when theplough is said to ‘stimulate them to growth’ they once again threaten totake on the profile of seeds, germinating when the plough brings themto the surface.

Finally, in the further set of arguments against literal annihilation,Lucretius even manages to convey to us the characteristically atomicstructure of things, and of indestructible particles separated by void(%%!–)):10

quod nunc, aeterno quia constant semine quaeque,donec vis obiit quae res diverberet ictuaut intus penetret per inania dissoluatque,nullius exitium patitur natura videri.

But as it is, because individual things are composed of everlastingseed, until a force has impinged on them to smash them apart by

!"+ '. The transformation of book I

10 For the same point, see Asmis (!"#&), p. +*.

its impact, or to penetrate inside them through their voids and disintegratethem, nature does not allow anything to be destroyed.

I have no wish to reprimand Lucretius for his methodological transgres-sion. To start his arguments o/ by creating an image of nature in whichthe observed regularities of biological creation constitute a continuumextending all the way down to the level of primary particles is a brilliantrhetorical coup, promising the reader almost miraculous creative powersfor the atoms. Epicurus, by contrast, had started from observed naturalregularity and worked painstakingly downwards in the direction ofatomic particles. To arrive there, he had first to establish that body andspace (ordinary body and ordinary space, witnessed in our everydayexperience of moving objects) are the only things with independent exis-tence, and then to infer that at the lowest level there must be bodies alto-gether unpunctuated by empty space, and hence atomic.

In particular, Epicurus in On nature ,,, at step (viii) of his argument (seeChart %b, p. !#$), had moved to the introduction of atoms by re-invok-ing premise (vi), ‘Moreover, of bodies themselves, some are compounds,others the constituents of those compounds’ (Letter to Herodotus )!). Hewas thus able to start from the undeniable fact that there are compoundbodies, and, via the equally undeniable fact that compounds must haveconstituents, to go on to demonstrate the atomicity of the ultimateconstituents. Lucretius’ equivalent move (, )#&–+&)) gets o/ to a less aus-picious start at , )#&–):

corpora sunt porro partim primordia rerumpartim concilio quae constant principiorum.

Moreover, some bodies are the primary particles of things,others are ones which are composed from the aggregation ofprimary particles.

The editors regularly comment that Lucretius has here followedEpicurus closely, or even translated him.11 This misses the crucialdi/erence in strategy. Instead of working dialectically from an uncon-troversial notion to a controversial one, in accordance with Epicurus’methodology, Lucretius reverses the strategy and places the atoms them-selves up front as his primary exhibit. But one can see why. Havingjumped the gun by familiarising his readers from the start with theconcept of these primary particles, he has forfeited the chance to repli-cate Epicurus’ dialectical derivation of them. Instead, he must put his

#. Seeds !"$

11 ‘Followed closely’: Bailey (!")$) ad loc. ‘Translated’: Munro (!##+), Ernout/Robin (!"+%) ad loc.

best foot forward, capitalising on the notion of primary particles whichhe has been working so hard to impress on his readers.

) . <,-,?,5D

The miraculous procreative powers of atoms also provide backing for anaccompanying denial. Where Epicurus had stated his opening thesissimply as ‘Nothing comes into being out of nothing’, Lucretius in intro-ducing it adds the highly significant word divinitus,12 followed by ajustification for its addition (, !)"–(#):

principium cuius hinc nobis exordia sumet,nullam rem e nilo gigni divinitus umquam. !(*quippe ita formido mortalis continet omnis,quod multa in terris fieri caeloque tuenturquorum operum causas nulla ratione viderepossunt, ac fieri divino numine rentur.quas ob res ubi viderimus nil posse creari !(( [!(+]de nilo, tum quod sequimur iam rectius inde [!($]perspiciemus, et unde queat res quaeque creari [!(#]et quo quaeque modo fiant opera sine divom. [!((]

The starting-point [of the study of nature] will take its cue fromthis: that nothing ever comes to be out of nothing, through divinecauses. For the reason why all mortals are so gripped by fear isbecause they witness many things happening on earth and in theheaven whose causes they are quite unable to see, and supposethat they are the work of divine power. That is why, when we seethat nothing can be created out of nothing, we will then as aresult have a more accurate view of our goal – both what it isthat each thing can be created from, and how each thing comesabout without the work of the gods.

He then proceeds immediately to his first proof of the thesis. His addi-tion has sometimes been criticised as a mistake. For example, HansGottschalk writes:13

In this form, his premiss loses its universality and can no longer be inferred fromthe inductive evidence he adduces; for God is not ‘nothing’, and the fact thatparticular things are always seen to come from other particular things, does notprove that there is not a deity (instead of another particular thing) at the begin-ning of the chain. Moreover, since his avowed purpose is to disprove divinecreation (!.!("), the presence of this word in his premiss makes the argument

!"# '. The transformation of book I

12 The significance of this addition is well brought out by Asmis (!"#&), p. ($, and Gale (forthcom-ing). 13 Gottschalk (!""+), p. %&).

circular. Lucretius’ eagerness to give his demonstration an extra anti-religioustwist has betrayed him into formulating his initial postulate in a way that rendersit invalid.

I think this criticism is answerable. The word divinitus is actually beingadded to the conclusion of an argument, not to its premise. Lucretius’point is that if things came into being out of nothing – meaning from nomaterial origin whatsoever, not from no cause – there would be no phys-ical stu/ to point to as controlling how, when and where they came tobe, and the only explanation left to postulate for these regularities wouldbe god: they would come into being divinitus. Hence to argue that thingsdo not come to be out of nothing is indeed to forestall one argument forthe existence of a controlling god or gods. As he says in the lines justquoted, it is people’s inability to find a physical explanation that leadsthem to reach for a divine one instead. By proceeding to show thatgeneration invariably depends on the right material conditions,Lucretius precludes the need for a further, divine cause.

Again, however, we can see the sense of urgency which governsLucretius’ procedure and di/erentiates him from Epicurus. In intro-ducing his very first argument, he has to alert us to the long-termoutcome to which it is leading us – the exclusion of divine agency. Thatexclusion will depend, for its eventual success, on many further argu-ments, right through to the demystification of cosmic phenomena inbook -,. But in pointing out that he is already with his first argumentremoving one prop from the theological edifice, Lucretius is both quitecorrect, and quite independent of Epicurus’ own cautiously progressivemethodology. Epicurus, we may now recall, reached his own conclusionson religious belief in book .,,, of On nature, his final expository book onhis physical system (see item (xxxiv) in Chart !, p. !&&, and Chart %, p.!&+), and in Chapter ) we found no evidence of any theological passagein the opening books.

( . 85267 123?B69

In view of what we have seen, we should from now on be on the look-out for further methodological departures by Lucretius from hisEpicurean original, and especially for transpositions of arguments. Onevery clear such case (although too fine a detail to have shown up amongthe arrowed transpositions in Chart %, p. !&+ above) is the final threearguments for the atomicity of the fundamental particles (, (""–+&)!arguments "–!! in Chart (, pp. !##>" above). These all appeal in some

$. Other changes !""

way to the impossibility of dismantling an atom into its own primary or‘minimal’ parts. Now Epicurus’ demonstration of the existence of theseminimal parts (epitomised at Ep. Hdt. ((–"), which closes with the samepoint about their inseparability, came a good deal later in his expositionthan the proofs of atomicity. Lucretius, not very surprisingly, chose toomit so theoretically demanding a part of Epicurean physics from hisown exposition in book ,,; but it seems clear that in doing so he rescuedfrom it these specific points which could be cited in support of atomismin book ,, and inserted them there.

Other such transpositions are bound to exist, even if any attempt toidentify them can be no more than speculative. For example, if I am evenhalf right about Epicurus’ methodology, it would be quite surprising tofind him placing the arguments of stages (L!) and (L%) (those for, respec-tively, microscopic body and hidden vacuum, both of them omitted inhis epitome) as early as Lucretius does.14 For as presented by Lucretiusthese go beyond the self-evident existence of ordinary phenomenal bodyand space to argue that minute portions of body (%+(–&%#) and hiddenstretches of empty space (&%"–)!$) are to be found even below thethreshold of vision. This sounds like the first stage of refining familiarbody and space into the atomic particles and pure vacuum which ulti-mately underlie them at the lowest level of analysis. Given what we havealready seen of Lucretius’ willingness to sacrifice methodological order-liness for rhetorical impact, the suspicion must be entertained that he ishimself responsible for moving stages (L!) and (L%) to their present earlyposition, Epicurus having himself, I suggest, probably placed them afterstage (viii). For Lucretius to make this further move to hasten his readers’descent into the microscopic world illuminated by Epicurus would beentirely of a piece with the strategy we have already witnessed.

In the present state of our evidence this cannot be demonstrated, butit would be unwise to ignore the possibility. Is there perhaps even a tell-tale sign of the transposition in the text? At )!# Lucretius terminatesstages (L!) and (L%), the arguments for microscopic body and hiddenvacuum, and e/ects his transition to the step which I am suggestingwould in his Epicurean source have preceded it, namely that the ‘all’consists of (ordinary phenomenal) bodies and space. The words markingthe transition are sed nunc ut repetam coeptum pertexere dictis . . . – ‘But now,to return to weaving in words the task on which I embarked . . .’ In thecontext such talk of a ‘return’ to the main task is hardly anomalous,

%** '. The transformation of book I

14 What follows is an expansion of an idea I have already put forward in Sedley (!""+), p. &!(.

given that the preceding lines (&"#–)!$), in which Lucretius has beenurging that if the foregoing arguments for vacuum are insu0cient he hasplenty more with which to overwhelm Memmius, might appear anincipient digression which he now wishes to halt. Even so, it remains atempting speculation – and no more than a speculation – that thesewords may also be his acknowledgement of a return to the order oftopics as he found them in his Epicurean original.15

+ . '31E 58 6;46<81:69

Whether or not this last guess is correct, a general pattern has neverthe-less emerged. Lucretius’ strategy of persuasion does not permit the slowand painstaking process established by Epicurus, whereby the atomicityof matter would emerge only after progressive refinement of the world’sfamiliar contents. To generate the excitement of discovery Lucretiusneeds right from the outset to plunge his readers into a microscopicworld few of them can have dreamt of. Both the sequence and the styleof his opening arguments are designed to maximise this impact.

The strategy is, if anything, overdetermined. If I am right, Lucretius’procedure is self-explanatory and justified by the immediate task inhand. Yet it also reflects the power exerted on him by his principal liter-ary model, Empedocles. To appreciate this, we may return to the linesof his proem where Lucretius launches his primary elements as pro-tagonists in a natural cycle (()–#):

nam tibi de summa caeli ratione deumquedisserere incipiam, et rerum primordia pandam, ((unde omnis natura creet res auctet alatquequove eadem rursum natura perempta resolvat,quae nos . . .

For I shall embark on explaining to you about the fundamentalrationale of the heaven and the gods, and I shall expound theprimary particles of things, from which nature creates, developsand nourishes all things, and into which nature destroys and dis-solves those same things once more, which it is our practice . . .

And he continues with the list of names for these primary particles whichwe have already encountered ((#–+!). He is mimicking the opening bookof Empedocles’ On nature, which at what we now know to be lines %&&/.(see p. %# above) catalogued his four elements plus the two motive forces

&. Back to Empedocles %*!

15 See Asmis (!"#&), p. +!, for much the same suggestion about this line’s meaning.

Love and Strife as divinities engaged in a never-ending cycle of birth anddeath. Even Lucretius’ surprising reference to the gods (()) emphasisesthe link. So does his development of a set of biological metaphors for‘atoms’, which has played such an important role in this chapter: was itnot Empedocles who set the pattern, by calling his own four elements‘roots’ (#$ $>(" µ+-+)?

Reflection on this Empedoclean model may help confirm the story Ihave already told. For Empedocles certainly did not work towards hisfour elements and two motive forces by argument. He unveiled them atthe outset, as divine players in the cosmic drama which he then pro-ceeded to unfold. When Lucretius, departing from Epicurus’ methodol-ogy, places his own primary elements on centre stage from the start, andwhen in his very first set of arguments he exploits to the full the imagerywhich represents them as live procreative powers, it is the Empedocleanmodel that we should see as holding sway.

%*% '. The transformation of book I

Epilogue

The quotation from Cicero with which I began Chapter ! remains asacute an evaluation of Lucretius as any. His is indeed a poem which dis-plays many flashes of genius but also much craftsmanship. If I have hadmore to say about the craftsmanship, my excuse is that genius is best leftto speak for itself.

Lucretius’ artistry was expended upon the creation of a most remark-able poem. Its main filling is fifteen books’ worth of technical physicsfrom Epicurus’ On nature, painstakingly assembled and systematicallyreshaped into a poetic masterpiece. The upshot was a dazzlingly deliv-ered message of salvation designed to whet the intellectual appetites ofa Roman audience, and to satisfy their emotional needs, without onceasking them to compromise their own Romanness.

This filling is sandwiched between two antithetical yet curiouslycomplementary descriptive passages. One is an Empedoclean hymn tobirth and life, which, while laying Lucretius’ chosen theme of naturebefore us in all its glory, also locates the poem by reference to fixed co-ordinates on the map of Graeco-Roman poetry. The other is theThucydidean tableau of pestilence and death, which establishes afurther set of co-ordinates, this time chronological: it shows why even inAthens, the cradle of civilisation, it was only the advent of Epicurus’ phi-losophy that could successfully fortify the human spirit against every-thing that fortune might cast its way.

This complex whole is no less magnificent for the fact, which we haveseen amply confirmed, that Lucretius did not live to take his visionthrough to its final fulfilment. Disappointment at an uncompleted questwould be entirely the wrong way for us to react. Only by seeing what thequest was, and how Lucretius was working towards it, can we properlyunderstand and value the poem’s second half, and especially its ending.What I hope to have brought into clearer focus is the grand conceptionby which Lucretius, while remaining loyal to his revered Greek masters

%*&

(and I mean Empedocles no less than Epicurus), was able to transformtheir benefactions into something truly new. The fact that he has noterased all traces of earlier strata in the poem’s history – from thedoxographical imprint of Theophrastus on Epicurus, all the way downto his own abandoned first proem to book ,- – enables us the better toappreciate the extraordinary craftsmanship by which he has worked thattransformation.

%*) Epilogue

Bibliography

Adams, J., Mayer, R. (!""#), eds., Aspects of the Language of Latin Poetry (London)Alberti, A. (!"")), ed., Realtà e ragione (Florence)Algra, K. A., van der Horst, P. W., Runia, D. T. (!""+), eds., Polyhistor: Studies in

the History and Historiography of Ancient Philosophy (Leiden)Allen, R. E., Furley, D. J. (!"$(), eds., Studies in Presocratic Philosophy, ,,. The Eleatics

and Pluralists (London)André, J. M. (!"+$), Les Noms d’oiseaux en latin (Paris)Angeli, A. (!"##), Filodemo, Agli amici di scuola (Naples)Arnim, H. von (!#"&), ‘Der angebliche Streit des Zenon und Theophrastos’,

Neue Jahrbücher für Philologie und Paedagogik !)$: ))"–+$(!"*&–(), Stoicorum Veterum Fragmenta, & vols., with indexes in vol. ,- (!"%)) by

M. Adler (Leipzig)Arrighetti, G. (!"$&), Epicuro, opere, ed. % (Turin)

(!"$"), ‘Un passo dell’opera “Sulla natura” di Epicuro, Democrito, e Colote’,Cronache Ercolanesi ": (–!*

Asmis, E. (!"#%), ‘Lucretius’ Venus and Stoic Zeus’, Hermes !!*: )("–$*(!"#&), ‘Rhetoric and reason in Lucretius’, American Journal of Philology !*):

&+–++(!""*), ‘Philodemus’ Epicureanism’, in Haase/Temporini (!""*) %&+"–)*+(!""+), ‘Epicurean semiotics’, in Manetti (!""+) !((–#(

Atherton, C. (forthcoming), ed., Form and Content in Didactic Poetry, Proceedings ofthe (th Nottingham Classical Literature Symposium (Nottingham/Bari)

Auvray-Assayas, C. (!""%), ‘Le livre I du De natura deorum et le traité De signis dePhilodème: problèmes de théologie et de logique’, Revue des Etudes Latines+": (!–+%

Avotins, I. (!"#*), ‘Alexander of Aphrodisias on vision in the atomists’, ClassicalQuarterly ?9 &*: )%"–()

Bailey, C. (!"%%), Lucreti De Rerum Natura Libri Sex, ed. % (ed. !, !"**), OxfordClassical Texts (Oxford)

(!")$), Titi Lucreti Cari De Rerum Natura Libri Sex, & vols. (Oxford)Baltes, M. (!"$+), Die Weltenstehung des Platonischen Timaios nach den antiken Interpreten

(Leiden)Barigazzi, A. (!"(*), ‘La µ/.3" della terra nei frammenti ercolanesi’, Studi Italiani

di Filologia Classica %): &–!"

%*(

(!"(#), ‘Cinetica degli "$%5(7+ nel 6"#$! %&" '"() di Epicuro’, La Parola delPassato !&: %)"–$+

Barnes, J. (!"$"), The Presocratic Philosophers (London)(!"##), ‘Epicurean signs’, Oxford Studies in Ancient Philosophy suppl. vol.: "!–!&)(!"#"a), ‘Antiochus of Ascalon’, in Gri0n/Barnes (!"#") (!–"+(!"#"b), ‘The size of the sun in antiquity’, Acta Classica Univ. Scient. Debrecen %(:

%"–)!(!""$), ‘Roman Aristotle’, in Barnes/Gri0n (!""$) !–+"

Barnes, J., Brunschwig, J., Burnyeat, M., Schofield, M. (!"#%), eds., Science andSpeculation: Studies in Hellenistic Theory and Practice (Cambridge)

Barnes, J., Gri0n, M. (!""$), eds., Philosophia Togata II: Plato and Aristotle at Rome(Oxford)

Barnes, J., Mignucci, M. (!"##), eds., Matter and Metaphysics (Naples)Ben, N. van der (!"$(), The Proem of Empedocles’ !"#$! %&" '"() (Amsterdam)Bergsträsser, G. (!"!#), Neue meteorologische Fragmente des Theophrast arabisch und

deutsch (Heidelberg)Bignone, E. (!"&+), L’Aristotele perduto e la formazione filosofica di Epicuro (Florence)

(!")(), Storia della Letteratura Latina ,, (Florence)Bloch, H. (!")*), ‘L. Calpurnius Piso Caesoninus in Samothrace and

Herculaneum’, American Journal of Philology )): )#(–"&Bollack, J. (!"("), ‘Lukrez und Empedokles’, Die neue Rundschau $*: +(+–#+

(!"+(–"), Empédocle, ) vols. (Paris)Bollack, J., Laks, A. (!"$+), eds., Etudes sur l’épicurisme antique (Cahiers de

Philologie !; Lille)(!"$#), Epicure à Pythocles (Cahiers de Philologie &; Lille)

Bonitz, H. (!#$*), Index Aristotelicus (Berlin)Boyancé, P. (!"+&), Lucrèce et l’Epicurisme (Paris)Brieger, A. (!##%), Epikurs Brief an Herodot &(–(# (= Jahresbericht des

Stadtgymnasiums zu Halle !))Bright, D. F. (!"$!), ‘The plague and the structure of the De rerum natura’, Latomus

&*: +*$–&%Brunschwig, J. (!"$$), ‘L’argument d’Epicure sur l’immutabilité du tout’, in

Permanence de la philosophie (Mélanges Joseph Moreau; Neuchâtel) !%$–(*;repr. in Brunschwig (!""() !(–)%, and republished as ‘Epicurus’ argumenton the immutability of the all’ in Brunschwig (!"")) !–%*

(!"")), Papers in Hellenistic Philosophy (Cambridge)(!""(), Etudes sur les philosophes hellénistiques (Paris)(!""+), ‘Le fragment DK $* ' ! de Métrodore de Chio’, in Algra/Horst/

Runia (!""+) %!–&#Bu0ère, F. (!"(+), Les Mythes d’Homère et la pensée grecque (Paris)Burkert, W. (!"$%), Lore and Science in Ancient Pythagoreanism (Cambridge, Mass.)Burnyeat, M. F. (!"$+), ‘Protagoras and self-refutation in later Greek philoso-

phy’, Philosophical Review #(: ))–+"(!"$#), ‘The upside-down back-to-front sceptic of Lucretius ,- )$%’, Philologus

!%%: !"$–%*+

%*+ Bibliography

(!""$), ‘Antipater and self-refutation: elusive arguments in Cicero’s Academica’,in Inwood/Mansfeld (!""$) %$$>&!*

Caizzi, F. Decleva (!""%), ‘Aenesidemus and the Academy’, Classical Quarterly ?9)%: !$+–#"

Capasso, M. (!"#!), ‘‘I Problemi di filologia filosofica di Mario Untersteiner’, Elenchos%: &$(–)*)

(!"#"), ‘Primo supplemento al Catalogo dei papiri ercolanesi’, Cronache Ercolanesi!": !"&–%+)

Casson, L. (!"$)), Travel in the Ancient World (Baltimore)Castner, C. J. (!"#$), ‘De rerum natura (.!*!–!*&: Lucretius’ application of

Empedoclean language to Epicurean doctrine’, Phoenix )!: )*–"(!"##), A Prosopography of Roman Epicureans (Frankfurt am Main)

Cavallo, G. (!"#&), Libri scritture scribi a Ercolano (Cronache Ercolanesi suppl. !,Naples)

Cavini, W., Donnini Macciò, M. C., Funghi, M. S., Manetti, D. (!"#(), Studi supapiri greci di logica e medicina (Accademia Toscana di Scienze e Lettere ‘LaColombaria’, Studi $)) (Florence)

Classen, C. J. (!"#+), ed., Probleme der Lukrezforschung (Hildesheim)Clay, D. (!"#&), Lucretius and Epicurus (Ithaca)Commager, H. S., Jr (!"($), ‘Lucretius’ interpretation of the plague’, Harvard

Studies in Classical Philology +%: !*(–%!Comparetti, D., De Petra, G. (!##&), La villa ercolanese dei Pisoni. I suoi monumenti e

la sua biblioteca (Turin; repr. Naples !"$%)Courtney, E. (!"#$), ‘Quotation, interpolation and transposition’, Hermathena

!)&: $–!#Crönert, W. (!"$(), Studi ercolanesi (Naples)Daiber, H. (!""%), ‘The Meteorology of Theophrastus in Syriac and Arabic trans-

lation’, in Fortenbaugh/Gutas (!""%) !++–%"&De Lacy, P. (!")#), ‘Lucretius and the history of Epicureanism’, Transactions of

the American Philological Association $": !%–&((!""&), review of M. Capasso, Manuale di papirologia ercolanese, in American

Journal of Philology !!): !$#–#*De Lacy, P. and E. (!"$#), Philodemus, On Methods of Inference, ed. % (Naples)Delattre, D. (!""+), ‘Les mentions de titres d’œuvres dans les livres de

Philodème’, Cronache Ercolanesi %+: !)&–+#Diano, C. (!")+), Epicuri ethica (Florence)Diels, H. (!#$"), Doxographi Graeci (Berlin)

(!"*!), Poetarum Philosophorum Fragmenta (Berlin)(!"*&), Die Fragmente der Vorsokratiker (Berlin)

Diels, H., Kranz, W. (!"(!–%), Die Fragmente der Vorsokratiker, ed. + (Berlin)Dillon, J. M. (!"$$), The Middle Platonists (London)Dorandi, T. (!"#%), ‘Filodemo, Gli Stoici (PHerc. !(( e &&")’, Cronache Ercolanesi

!%: "!–!&&(!""*), ‘Gli arconti nei papiri ercolanesi’, Zeitschrift für Papyrologie und Epigraphik

#): !%!–&#

Bibliography %*$

(!""!), Ricerche sulla cronologia dei filosofi ellenistici (Stuttgart)(!""$), ‘Lucrèce et les Epicuriens de Campanie’, in Schrijvers/Koenen/

Algra (!""$) &(>)#Drexler, F. (!"&(), ‘Aporien im Prooemium IV des Lukrez’, Athenaeum !&: $&–!**Düring, I. (!"($), Aristotle in the Ancient Biographical Tradition (Göteborg)Edelstein, L. (!")*), ‘Primum Graius homo (Lucretius !.++)’, Transactions of the

American Philological Association $!: $#–"*Edelstein, L., Kidd, I. G. (!"#"), Posidonius, vol. ,: The Fragments, ed. % (ed. ! !"$%)

(Cambridge)Edwards, M. J. (!"#"), ‘Lucretius, Empedocles and Epicurean polemic’, Antike

und Abendland &(: !*)–!(Erler, M. (!"")), ‘Erstes Kapitel: Epikur’, ‘Zweites Kapitel: Die Schule Epikurs’,

‘Drittes Kapitel: Lukrez’, in Flashar (!"")) %"–)"*Ernout, A., Robin, L. (!"+%), Lucrèce, De rerum natura, commentaire exégétique et cri-

tique, ed. % (Paris)Farrell, J. (!""!), Virgil’s ‘Georgics’ and the Traditions of Ancient Epic (New

York/Oxford)Ferrari, W. (!"&$), ‘ La doppia redazione del proemio IV di Lucrezio’, Atene e

Roma &": !##–%**Ferrary, J.-L. (!"##), Philhellénisme et impérialisme (Rome)Flashar, H. (!"")), ed., Die Philosophie der Antike %: Die hellenistische Philosophie

(Basel)Fortenbaugh, W., Gutas, D. (!""%), eds., Theophrastus: His Psychological,

Doxographical and Scientific Writings (Rutgers University Studies in ClassicalHumanities (; New Brunswick)

Fortenbaugh, W. W., Huby, P. M., Sharples, R. W., Gutas, D. (!""%), Theophrastusof Eresus: Sources for his Life, Writings, Thought and Influence: cited as ‘FHS&G’(Leiden/New York/Cologne)

Fortenbaugh, W., Steinmetz, P. (!"#"), eds., Cicero’s Knowledge of the Peripatos(Rutgers University Studies in Classical Humanities, ); New Brunswick)

Fowler, P. G. and D. P. (!""+), ‘Lucretius’, in Oxford Classical Dictionary, ed. &(Oxford) ###–"*

Furley, D. J. (!"++), ‘Lucretius and the Stoics’, BICS !&: !&–&&; repr. in Classen(!"#+) $(–"(, and in Furley (!"#"a) !#&–%*(

(!"+$), Two Studies in the Greek Atomists (Princeton)(!"$*), ‘Variations on themes from Empedocles in Lucretius’ proem’, BICS

!$: ((–+); repr. in Furley (!"#"a) !$%–#%(!"$#), ‘Lucretius the Epicurean: on the history of man’, Entretiens Hardt %):

!–&$; repr. in Furley (!"#"a) %*+–%%(!"#"a), Cosmic Problems (Cambridge)(!"#"b), ‘The dynamics of the earth: Anaximander, Plato, and the centrifo-

cal theory’, in Furley (!"#"a) !)–%+Gale, M. (!"")a), Myth and Poetry in Lucretius (Cambridge)

(!"")b), ‘Lucretius ).!–%( and the proems of the De rerum natura’, Proceedings ofthe Cambridge Philological Society )*: !–!$

%*# Bibliography

(forthcoming), comments on Sedley (forthcoming), in Atherton (forthcoming)Gallo, I. (!"#*), Frammenti biografici da papiri. Volume secondo, La biografia dei filosofi

(Rome)Giancotti, F. (!"("), Il preludio di Lucrezio (Messina)Giannantoni, G., Gigante, M. (!""+), eds., Epicureismo greco e romano, & vols.

(Naples)Gigante, M. (!"$+), ‘Polemonis Academici Fragmenta’, Rendic. Archeol./Lettere e

Belle Arti Nap. (!: "!–!))(!"$"), ed., Catalogo dei papiri ercolanesi (Naples)(!"#!), Scetticismo e epicureismo (Naples)(!""(), Philodemus in Italy: the Books from Herculaneum, tr. D. Obbink (Ann Arbor)

Gigante, M., Capasso, M. (!"#"), ‘Il ritorno di Virgilio a Ercolano’, Studi Italianidi Filologia Classica ?9 $: &–+

Gilbert, W. (!+**), De Magnete (London)Giussani, C. (!#"+–#), T. Lucreti Cari De rerum natura libri sex. Revisione del testo, com-

mentario e studi, ) vols. (Turin)Glucker, J. (!"$#), Antiochus and the Late Academy (Göttingen)Godwin, J. (!"#+), Lucretius: ‘De Rerum Natura’ IV (Warminster)

(!""!), Lucretius: ‘De Rerum Natura’ VI (Warminster)Görler, W. (!"")), ‘Fünftes Kapitel: Älterer Pyrrhonismus. Jüngere Akademie.

Antiochus aus Askalon’, in Flashar (!"")) $!$–"#"Gottschalk, H. (!""+), ‘Philosophical innovation in Lucretius’, in Algra/Horst/

Runia (!""+) %&!–)*Goulet, R. (!"#*), Cléomède, théorie élémentaire (Paris)

(!"#"– ), ed., Dictionnaire des philosophes antiques (Paris)Graeser, A. (!"$(), Zenon von Kition: Positionen und Probleme (Berlin)Graham, D. W. (!"##), ‘Symmetry in the Empedoclean cycle’, Classical Quarterly

?9 &#: %"$–&!%Gri0n, M. (!"#"), ‘Philosophy, politics and politicians at Rome’, in Gri0n/

Barnes (!"#") !–&$(!""(), ‘Philosophical badinage in Cicero’s letters to his friends’, in Powell

(!""(b) &%(–)+Gri0n, M., Barnes, J. (!"#"), eds., Philosophia Togata (Oxford)Grilli, A. (!"+&), ‘5$+'-#/%3" ’, Acme !+: #$–!*!Guthrie, W. K. C. (!"+"), A History of Greek Philosophy, vol. ,, (Cambridge)Haase, W., Temporini, H. (!""*), Aufstieg und Niedergang der römischen Welt, ,, &+.)

(Berlin/New York)Habicht, C. (!""(), Athen. Die Geschichte der Stadt in hellenisticher Zeit (Munich)Hadot, P. (!"#$), ‘Théologie, exégèse, révélation, écriture, dans la philosophie

grecque’, in M. Tardieu (ed.), Les Règles de l’interprétation (Paris !"#$) !&–&)Hardie, P. (!"#+), Virgil’s Aeneid. Cosmos and Imperium (Oxford)

(!""(), ‘The speech of Pythagoras in Ovid Metamorphoses !(: Empedocleanepos’, Classical Quarterly ?9 )(: %*)–!)

Hartung, H. J. (!"$*), Ciceros Methode bei der Übersetzung griechischer philosophischerTermini (diss., Hamburg)

Bibliography %*"

Henrichs, A. (!"$)), ‘Die Kritik der stoischen Theologie im PHerc. !)%#’,Cronache Ercolanesi ): (–&%

Hinds, S. E. (!"#$), ‘Language at breaking point: Lucretius !.)(%’, ClassicalQuarterly ?9 &$: )(*–&

Hölscher, U. (!"+(), ‘Weltzeiten und Lebenszyklus’, Hermes "&: $–&&House, D. K. (!"#*), ‘The life of Sextus Empiricus’, Classical Quarterly ?9 &*:

%%$–&#Inwood, B. (!"#!), ‘The origin of Epicurus’ concept of void’, CP $+: %$&–#(

(!""%), The Poem of Empedocles (Toronto)Inwood, B., Mansfeld, J. (!""$), eds., Assent and Argument: Studies in Cicero’s Academic

Books (Leiden)Jobst, F. (!"*$), Über das Verhältnis zwischen Lukretius und Empedokles (diss., Munich)Jocelyn, H. D. (!"$$), ‘The ruling class of the Roman republic and Greek

philosophers’, Bulletin of the John Rylands University Library of Manchester (":&%&–++

Kahn, C. H. (!"+*), ‘Religion and natural philosophy in Empedocles’ doctrineof the soul’, Archiv für Geschichte der Philosophie )%: &–&(

Kenney, E. J. (!"$*), ‘Doctus Lucretius’, Mnemosyne ).%&: &++–"%(!"$!), Lucretius: De rerum natura Book III (Cambridge)(!"$$), Lucretius (= Greece and Rome New Surveys in the Classics !!; Oxford)

Kerschensteiner, J. (!"((), ‘Der Bericht des Theophrast über Heraklit’, Hermes#&: &#(–)!!

Kidd, I. G. (!""+), ‘Theophrastus fr. !#) FHS&G: some thoughts on his argu-ments’, in Algra/Horst/Runia (!""+) !&(–))

Kingsley, P. (!""(), Ancient Philosophy, Mystery, and Magic: Empedocles and PythagoreanTradition (Oxford)

(!""+), ‘Empedocles’ two poems’, Hermes !%): !*#–!!Kirk, G. S. (!"()), Heraclitus, the Cosmic Fragments (Cambridge)Kleve, K. (!"++), ‘Lukrez und Venus (De rerum natura ,, !–)")’, Symbolae Osloenses

)!: #+–")(!"$#), ‘The philosophical polemics in Lucretius’, Entretiens Hardt %): &"–$!(!"#"), ‘Lucretius in Herculaneum’, Cronache Ercolanesi !": (–%$

Kollman, E. D. (!"$!), ‘Lucretius’ criticism of the early Greek philosophers’,Studii Clasice !&: $"–"&

Kranz, W. (!"))), ‘Lukrez und Empedokles’, Philologus "+: +#–!*$Kuhrt, A. (!"#$), ‘Berossus’ Babyloniaka and the Seleucid rule in Babylonia’, in

A. Kuhrt, S. Sherwin-White (eds.), Hellenism in the East. The Interaction ofGreek and Non-Greek Civilizations from Syria to Central Asia after Alexander(London) &%–(+

Lackenbacher, H. (!"!*), ‘Zur Disposition und Quellenfrage von Lukrez ,-!–(%!’, Wiener Studien &%: %!&–&#

Langslow, D. (!""#), ‘The language of poetry and the language of science: theLatin poets and “medical Latin”’, in Adams/Mayer (!""#)

Lapidge, M. (!"$#), ‘Stoic cosmology’, in Rist (!"$#) !+!–#+Laursen, S. (!"#$), ‘Epicurus, On nature book ..-’, Cronache Ercolanesi !$: $$–#

%!* Bibliography

(!"##), ‘Epicurus, On nature ..- (Long–Sedley %*', 1, and j)’, CronacheErcolanesi !#: $–!#

(!""%), ‘The summary of Epicurus, On nature book %(’, Papyrologica Lupiensia !:!)&–()

(!""(), ‘The early parts of Epicurus, On nature, %(th book’, Cronache Ercolanesi%(: (–!*"

Leone, G. (!"#)), ‘Epicuro, Della natura, libro .,-’, Cronache Ercolanesi !): !$–!*$Lévêque, P. (!"("), Aurea Catena Homeri (Paris )Lévy, C. (!""%), ‘Cicéron créateur du vocabulaire latin de la connaissance: essai

de synthèse’, in La Langue latine, langue de la philosophie (Ecole française deRome) "!–!*+

(!""$), ‘Lucrèce avait-il lu Enésidème?’, in Schrijvers/Koenen/Algra (!""$)!!(>%)

Lewis, N. (!"("), Samothrace, the Ancient Literary Sources (London)Lloyd, G. E. R. (!"++), Polarity and Analogy (Cambridge)

(!"#$), The Revolutions of Wisdom (Berkeley/Los Angeles)Long, A. A. (!"++), ‘Thinking and sense-perception in Empedocles: mysticism

or materialism?’, Classical Quarterly ?9 !+: %(+–$+(!"$)), ‘Empedocles’ cosmic cycle in the ’sixties’, in Mourelatos (!"$)) &"$–)%((!"$(–+), ‘Heraclitus and Stoicism’, H$7/'/%$"+ (–+: !&&–(&; repr. in Long

(!""+a) &(–($(!"##a), ‘Reply to Jonathan Barnes, “Epicurean signs”’, Oxford Studies in

Ancient Philosophy suppl. vol.: !&(–))(!"##b), ‘Socrates in Hellenistic philosophy’, Classical Quarterly ?9 &#: !(*–$!;

repr. in Long (!""+a) !–&)(!""%a), review of Segal (!""*), Ancient Philosophy !%: )"&–"(!""%b), ‘Stoic readings of Homer’, in R. Lamberton, J. J. Keaney (eds.),

Homer’s Ancient Readers (Princeton !""%) )!–++; repr. in Long (!""+a) (#–#)(!""+a), Stoic Studies (Cambridge)(!""+b), ‘Theophrastus’ De sensibus on Plato’, in Algra/Horst/Runia (!""+)

&)(–+%(!""$), ‘Lucretius on nature and the Epicurean self ’, in Schrijvers/Koenen/

Algra (!""$) !%(–&"Long, A. A., Sedley, D. N. (!"#$), The Hellenistic Philosophers, % vols. (Cambridge)Longo Auricchio, F. (!"$#), ‘La scuola di Epicuro’, Cronache Ercolanesi #: %!–&$

(!"##), Ermarco, frammenti (Naples)Lück, W. (!"&%), Die Quellenfrage im $. und &. Buch des Lukrez (diss., Breslau)McDiarmid, J. B. (!")*), ‘Theophrastus on the eternity of the world’,

Transactions of the American Philological Association $!: %&"–)$MacKay, L. (!"((), ‘De rerum natura !.$!$ sqq.’, Latinitas &: %!*Manetti, G. (!""+), ed., Knowledge through Signs: Ancient Semiotic Theories and Practices

(Brussels)Mansfeld, J. (!"#"), ‘Gibt es Spuren von Theophrasts Phys. op. bei Cicero?’, in

Fortenbaugh/Steinmetz (!"#") !&&–(#(!""*), ‘Doxography and dialectic. The Sitz im Leben of the “Placita”’, in

Haase/Temporini (!""*) &*(+–%%"

Bibliography %!!

(!""%a), ‘A Theophrastean excursus on god and nature and its aftermath inHellenistic thought’, Phronesis &$: &!)–&(

(!""%b), ‘Physikai doxai and Problemata physika from Aristotle to Aëtius (andbeyond)’, in Fortenbaugh/Gutas (!""%) +&–!!!

(!"")a), ‘Epicurus Peripateticus’, in Alberti (!"")) %#–)$(!"")b), ‘A lost manuscript of Empedocles’ Katharmoi’, Mnemosyne )$: $"–#%(!""(), ‘Aenesidemus and the Academics’, in L. Ayres (ed.), The Passionate

Intellect (New Brunswick/London) %&(–)#Martin, A., Primavesi, O. (!""#), L’Empédocle de Strasbourg (P. Strasb. gr. Inv.

!&&$–!&&&). Introduction, édition et commentaire (Strasbourg/Berlin)Mewaldt, J. (!"*#), ‘Eine Dublette in Buch ,- des Lucrez’, Hermes )&: %#+–"(;

repr. in Classen (!"#+) &!–)*Milanese, G. (!"#"), Lucida carmina. Comunicazione e scrittura da Epicuro a Lucrezio

(Milan)Millot, C. (!"$$), ‘Epicure, De la nature, livre .-’, Cronache Ercolanesi $: "–&"Moraux, P. (!"(!), Les Listes anciennes des ouvrages d’Aristote (Louvain)Mourelatos, A. P. D. (!"$)), ed., The Presocratics (New York)Muehll, P. von der (!"%%), Epicurus, Epistulae Tres et Ratae Sententiae (Teubner text;

Stuttgart)Munro, H. A. J. (!##+), T. Lucreti Cari De Rerum Natura Libri Sex, ed. )

(London)Norden, E. (!#"&), ‘Über den Streit des Theophrast und Zeno 6"#$! +& %1+#'$"+)

,/" 'µ/&’, Jahrbücher für Classische Philologie, suppl. !": ))*–(%Obbink, D. (!""(), ed., Philodemus and Poetry: Poetic Theory and Practice in Lucretius,

Philodemus, and Horace (Oxford)(!""+), Philodemus On Piety, Part I (Oxford)

O’Brien, D. (!"+"), Empedocles’ Cosmic Cycle (Cambridge)(!"#!), Pour interpréter Empédocle (Paris/Leiden)(!""(), ‘Empedocles revisited’, Ancient Philosophy !(: )*&–$*

Ophuijsen, J. M. van, Raalte, M. van (!""#), Theophrastus: Reappraising theSources (Rutgers University Studies in Classical Humanities #; NewBrunswick)

Osborne, C. (!"#$), ‘Empedocles recycled’, Classical Quarterly ?9 &$: %)–(*Owen, G. E. L. (!"$&), ‘Plato on the undepictable’, in E. N. Lee et al. (eds.),

Exegesis and Argument (Assen !"$&) &)"–+!; repr. in G. E. L. Owen, Logic,Science and Dialectic (London !"#+) !&#–)$

Panchenko, D. (!"")), ‘?B?8?I and ?B?8?*MI in Thales and Anaximander’,Hyperboreus !: %#–((

Parker, R. (!"#&), Miasma (Oxford)Pearson, A. C. (!#"!), The Fragments of Zeno and Cleanthes (London)Pease, A. S. (!"((–#), M. Tulli Ciceronis De natura deorum libri III, % vols.

(Cambridge, Mass.)Pigeaud, J.-M. (!"#!), ‘La physiologie de Lucrèce’, Revue des Etudes Latines (#:

!$+–%**Pizzani, U. (!"("), Il problema del testo e della composizione del De rerum natura di

Lucrezio (Rome)

%!% Bibliography

Powell, J. G. F. (!""(a), ‘Cicero’s translations from Greek’, in Powell (!""(b)%$&–&**

(!""(b), ed., Cicero the Philosopher (Oxford)Price, D. de Solla (!"$(), Gears from the Greeks (New York)Primavesi, O. (forthcoming), Kosmos und Dämon bei Empedokles. Der Papyrus P. Strasb.

gr. Inv. !&&$-!&&& und die indirekte Überlieferung (Hypomnemata !!+;Göttingen)

Puglia, E. (!"#$), ‘PHerc. !)%*/!*(+: un volume dell’opera “Della natura” diEpicuro’, Cronache Ercolanesi !$: #!–&

(!"##), Demetrio Lacone: Aporie testuali ed esegetiche in Epicuro (Naples)Radl, A. (!"##), Der Magnetismus in der Antike. Quellen und Zusammenhänge (Stuttgart)Rawson, E. (!"#%), ‘The life and death of Asclepiades of Bithynia’, Classical

Quarterly ?9 &%: &(#–$*(!"#(), Intellectual Life in the Late Roman Republic (London)

Reiley, K. C. (!"*"), Studies in the Philosophical Terminology of Lucretius and Cicero(New York)

Reitzenstein, E. (!"%)), Theophrast bei Epikur und Lukrez (Heidelberg)Rist, J. M. (!"$#), ed., The Stoics (Berkeley/Los Angeles)Romeo, C. (!"$"), ‘Demetrio Lacone sulla grandezza del sole (PHerc. !*!%)’,

Cronache Ercolanesi ": !!–&(Rösler, W. (!"$&), ‘Lukrez und die Vorsokratiker: doxographische Probleme im

!. Buch von “De rerum natura”’, Hermes !*!: )#–+); repr. in Classen (!"#+)($–$&

Runia, D. T. (!"#+), Philo of Alexandria and the Timaeus of Plato (Leiden)(!"#"), ‘Plato, Timaeus &* ' +–1 !’, Elenchos !*: )&(–)&(!""$), ‘Lucretius and doxography’, in Schrijvers/Koenen/Algra (!""$)

"&–!*&Sambursky, S. (!"("), Physics of the Stoics (London)Sandbach, F. H. (!"#(), Aristotle and the Stoics (Cambridge Philological Society,

suppl. vol. !!, Cambridge)Schmalzriedt, E. (!"$*), Peri Physeos: zur Frühgeschichte der Buchtitel (Munich)Schmid, W. (!"&+), Epikurs Kritik der platonischen Elementenlehre (Leipzig)Schmidt, J. (!""*), Lukrez, der Kepos und die Stoiker (Frankfurt am Main)Schofield, M. (!"$(), ‘Doxographica Anaxagorea’, Hermes !*&: !–%)Schrijvers, P. H. (!"$+), ‘La pensée d’Epicure et de Lucrèce sur le sommeil

(DRN, ,-, "*$–"+!, Scolie ad Epicure Ep. ad Her. ++). Un chapitre des Parvanaturalia Epicuriens’, in Bollack/Laks (!"$+): %&!–("

(!"#&), ‘Sur quelques aspects de la critique des mythes chez Lucrèce’, inIFNM*MI8I: studi sull’epicureismo greco e latino o,erti a Marcello Gigante (Naples)&(&–$!

(!""%), ‘Philosophie et paraphrase: Lucrèce et les sceptiques’, in La Languelatine, langue de la philosophie (Ecole française de Rome) !%(–)*

(!"")), ‘Intertextualité et polémique dans le De rerum natura (- "%(–!*!*):Lucrèce vs. Dicéarque de Messène’, Philologus !&#: %##–&*)

(!""+), ‘Lucretius on the origin and development of political life (De rerumnatura (.!!*(–!!+*)’, in Algra/Horst/Runia (!""+) %%*–&*

Bibliography %!&

(!""$), ‘L’homme et l’animal dans le DRN’, in Schrijvers/Koenen/Algra(!""$) !(!–+!

Schrijvers, P. H., Koenen, M. H., Algra, K. A. (!""$), eds., Lucretius and hisIntellectual Background (Amsterdam)

Sedley, D. (!"$&), ‘Epicurus, On Nature, Book ..-,,,’, Cronache Ercolanesi &: (–#&(!"$)), ‘The structure of Epicurus, On nature’, Cronache Ercolanesi ): #"–"%(!"$+a), ‘Epicurus and his professional rivals’, in Bollack/Laks (!"$+) !!"–("(!"$+b), ‘Epicurus and the mathematicians of Cyzicus’, Cronache Ercolanesi +:

%&–()(!"#!), ‘The end of the Academy’, Phronesis %+: +$–$((!"#%), ‘On signs’, in Barnes et al. (!"#%) %&"–$%(!"#&), ‘Epicurus’ refutation of determinism’, in IFNM*MI8I: studi sull’epi-

cureismo greco e romano o,erti a Marcello Gigante (Naples) !!–(!(!"#)), ‘The character of Epicurus’ On Nature’, in Atti del XVII congresso inter-

nazionale di papirologia (Naples) &#!–$(!"##), ‘Epicurean anti-reductionism’, in Barnes/Mignucci (!"##) %"(–&%$(!"#"a), ‘The proems of Empedocles and Lucretius’, GRBS &*: %+"–"+(!"#"b), ‘Philosophical allegiance in the Greco-Roman world’, in

Gri0n/Barnes (!"#") "$–!!"(!""%a), ‘Sextus Empiricus and the atomist criteria of truth’, Elenchos !&:

%!–(+(!""%b), ‘Empedocles’ theory of vision in Theophrastus, De sensibus’, in

Fortenbaugh/Gutas (!""%) %*–&!(!""+), ‘The inferential foundations of Epicurean ethics’, in

Giannantoni/Gigante (!""+) &!&–&"; repr. in S. Everson (ed.), Ethics(Cambridge !""#)

(!""$a), ‘The ethics of Brutus and Cassius’, Journal of Roman Studies #$: )!–(&(!""$b), ‘How Lucretius composed the De rerum natura’, in Schrijvers/

Koenen/Algra (!""$) !–!"(!""#a), ‘Theophrastus and Epicurean physics’, in Ophuijsen/Raalte (!""#)

&&&–(+(!""#b), ‘Lucretius’ use and avoidance of Greek’, in Adams/Mayer (!""#)

%%$–)+(forthcoming), ‘The sequence of argument in Lucretius ,’, in Atherton (forth-

coming)Seele, A. (!""(), Römische Übersetzer, Nöte, Freiheiten, Absichten: Verfahren des liter-

arischen Übersetzens in der griechisch-römischen Antike (Darmstadt)Segal, C. (!""*), Lucretius on Death and Anxiety (Princeton)Sharples, R. W. (!""%), Alexander of Aphrodisias, Quaestiones !.!–".!$ (London)Skutsch, O. (!"#(), The Annals of Quintus Ennius (Oxford)Smith, M. F. (!"++), ‘Some Lucretian thought processes’, Hermathena !*%: $&–#&

(!"$(), Lucretius, ‘De Rerum Natura’, with an English translation by W. H. D.Rouse, revised with new text, introduction, notes and index (Loeb ClassicalLibrary; London / Cambridge, Mass.)

(!""&), Diogenes of Oinoanda, The Epicurean Inscription (Naples)

%!) Bibliography

(!""+), ‘An Epicurean priest from Apamea in Syria’, Zeitschrift für Papyrologieund Epigraphik !!%: !%*–&*

Snyder, J. M. (!"$%), ‘Lucretius’ Empedoclean Sicily’, Classical World +(: %!$–!#Solmsen, F. (!"(!), ‘Epicurus and cosmological heresies’, American Journal of

Philology $%: !–%&; repr. in Solmsen (!"+#) )+!>#&(!"(&), ‘Epicurus on the growth and decline of the cosmos’, American Journal

of Philology $): &)–(!; repr. in Solmsen (!"+#) )#)>(*!(!"+!a), ‘+$%'13'$) in Aristotelian and Epicurean thought’, Mededelingen der

Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Letterkunde %) no. #:%)!–+%; repr. in Solmsen (!"+#) +!%–&&, and in Classen (!"#+) !(!–$%

(!"+!b), ‘Greek philosophy and the discovery of the nerves’, MuseumHelveticum !#: !(*–"$; repr. in Solmsen (!"+#) (&+>#%

(!"+(), ‘Love and strife in Empedocles’ cosmology’, Phronesis !*: !*"–)#; repr.in Solmsen (!"+#) %$)–&!&, and in Allen/Furley (!"$() %%!–+)

(!"+#), Kleine Schriften vol. , (Hildesheim)Staden, H. von (!"#"), Herophilus (Cambridge)Steckel, H. (!"+#), ‘Epikuros’, RE suppl. .,: ($"–+(%Tatum, W. J. (!"#)), ‘The Presocratics in book one of Lucretius’ De rerum natura’,

Transactions of the American Philological Association !!): !$$–#"Tepedino Guerra, A. (!""!), Polieno, frammenti (Naples)Townend, G. B. (!"$"), ‘The original plan of Lucretius’ De rerum natura’, Classical

Quarterly ?9 %": !*!–!!Tyrrell, R. Y. (!##(–!"*!), The Correspondence of M. Tullius Cicero, $ vols. (Dublin)Usener, H. (!##$), Epicurea (Leipzig)

(!"$$), Glossarium Epicureum, ed. M. Gigante, W. Schmid (Rome)Vallance, J. T. (!""*), The Lost Theory of Asclepiades of Bithynia (Oxford)Vander Waerdt, P. A. (!"#(), ‘Peripatetic soul-division, Posidonius, and Middle

Platonic moral psychology’, Greek, Roman and Byzantine Studies %+: &$&–")(!"##), ‘Hermarchus and the Epicurean genealogy of morals’, Transactions of

the American Philological Association !!#: #$–!*+(!"#"), ‘Colotes and the Epicurean refutation of scepticism’, Greek, Roman and

Byzantine Studies &*: %%(–+$Vlastos, G. (!"(*), ‘The physical theory of Anaxagoras’, Philosophical Review (":

&!–($; repr. in Allen/Furley (!"$() &%&–(&, and in Vlastos (!""&) &*&–%$(!""&), Studies in Greek Philosophy, vol. ,:The Presocratics (Princeton)

Vogliano, A. (!"%#), Epicuri et Epicureorum Scripta in Herculanensibus Papyris Servata(Berlin)

(!")*), I resti dell’ XI libro del !"#$! %&" '"() di Epicuro (Cairo)(!"(&), ‘I resti del IIo libro del !"#$! %&" '"() di Epicuro’, Prolegomena %: ("–"#

Wallace, R. (!""+), ‘“Amaze your friends!” Lucretius on magnets’, Greece & Rome)&: !$#–#$

Wardy, R. (!"##), ‘Lucretius on what atoms are not’, Classical Philology #&: !!%–%#Waszink, J. H. (!"(*), ‘The proem of the Annales of Ennius’, Mnemosyne ).):

%!(–)*Weise, F. O. (!##%), Die griechischen Wörter im Latein (Leipzig)

Bibliography %!(

West, D. (!"$*), ‘Virgilian multiple-correspondence similes and their antecedents’,Philologus !!): %+%–$(

(!"$(), ‘Lucretius’ methods of argument (&. )!$–+!))’, Classical Quarterly ?9 %(:")–!!+

Wiersma, W. (!")*), ‘Der angebliche Streit des Zenon und Theophrast,Mnemosyne &.#: %&(–)&

Woltjer, J. (!#$$), Lucreti Philosophia cum Fontibus Comparata (Groningen)Woolf, G. (!"")), ‘Becoming Roman, staying Greek: culture, identity and the

civilizing process in the Roman east’, Proceedings of the Cambridge PhilologicalSociety )*: !!+–)&

Wright, M. R. (!"#!), Empedocles: the Extant Remains (New Haven)Young, D. (!"$!), Theognis (Leipzig)Zeller, E. (!#$+), ‘Der Streit Theophrasts gegen Zeno über die Ewigkeit der

Welt’, Hermes !!: )%%–"

%!+ Bibliography

365,A9, &.! $"n

$.%" $#n,, %*.) $"n

%*.!+ $"n%&.( $"n%(.!% "*n%#.! "*n%".% "*n

,- &.!! $!n%.# +"n(.+ +"n%!.) +"n

- %!.! (

3:6.3?<67 8@ 34278<,9,39De anima mantissa!&(.%)>+ )!Quaestiones, %& !$(n

3;;8?,A9 (B73;;35,1A9)De differentia adfinium vocabulorum'&.%" +*n

3?3.3B8739 (ed. H. Diels, W. Kranz)'!% !%)'!$ &%n, )(

3?3.,;3?<67 (ed. H. Diels, W. Kranz)3%+ #*n

3?8?D;A9 :8?<,?6?9,9-,,, %">&* $%n

348::8?,A9 8@ 728<69,,, #+* $n

34A:6,A9Apology%$ +

De Platone!*( !*+n

37,958423?69Frogs!*&& (

37,9585:6De caelo%$"b!) !$#n%$"b&% !$"n%"(b!*>!+ #*n,,, $>#De generatione et corruptione&&$a!ff. !$)De partibus animalium+)*a!"ff. !"nDe philosophia (fragments, ed. Ross)# !$%nMetaphysics!*$)b!*>!& !$%nMeteorologica, !) !$%, !$$, !$#&()b&%>&((a&% $"&(+b+ff. !$#Physics!"#b&% !"n!""b!*>!% !"nPoetics!)($b!& $n

37,A9 <,<D;A9fragments (ed. H. Diels)&& $"n

3526?36A9!*%' $%n

CARMEN AUREUM+$>#, $! ), $

%!$

Index locorum

135A::A9).!>% (#n

16?987,?A9De die natali!!.$ (n

1,1678Academica, + &#n

%)>" $$>#%)>+ &+>$%) $$n%#>" $#%# $#, #!%" $$n)! )&)) ))n)( #+n, #$n

,, # #$n!$ )&!# )&%# #$n%" #(n, #"n&! )&%# )&n, #$n%" #$n$& #+!%* $)n!&! $$n

Ad Att..,, (a !n.,- %*.& !nAd familiares.- !+.! &"n

!".! &"nAd Quintum fratrem,, ".) !De divinatione, "* &$nDe finibus, &! "*n

)*>! !+%,, %* !)n

!*% !)n!!" +(

,,, ( &$n!( ))n!$ ))n

- # +)nDe inventione, " )&n

&+ )&nDe oratore, !*) +)n

De natura deorum, !#>%& $&n

%* &$n, $+n%" !#n++ !)n++>$ &#n!!* &"n!!) &"n

,, )* $"n)& $"n#& $"n## #%>&") &#n!!# $"n

,,, &$ $"nDe republica,, +" )"nTimaeus%! &#%( &#%$ &#, )"Tusculan disputations, %% &#n

)! )"n- #( #"n

1:6;6?5 8@ 3:6.3?<7,3Stromateis-, &*.!>& +n

1:68;6<69De motu circulari,, ! #),, ) "*n

<3;8.6?A9 (ed. R. Kassel, C. Austin)fr. %, %">&* $%n

<6;657,A9 8@ :318?,3P. Herc. !*!%, .:,,>.:-,, $*

<6;817,5A9 (ed. H. Diels, Kranz)'$ #+>$

<,8B6?69 :3675,A9v %+ +)n-,, !)! !$#n

!)( $"n-,,, (" +n-,,, ($ %)n

$$ &, !!n. % !&*

$ !*!n, !*%!% !)n%% !+(n

%!# Index locorum

%+ !%"n%$ !*!n, !*%%$># !*!n, !)!%">&* !!*&* !*%, !)!!!" !!*n

<,8B6?69 8@ 86?83?<3 (ed. M. F. Smith)( "&n+ , !*>,, " $), !%(+ ,, ">,,, ! !%+n+ ,,>,,, !#n+ ,, )>$ !%)n!* - %>+ !)"n!* - & )!n

<,8?D9,A9 8@ 23:,137?399A9De Thucydidep. &&".!)>!+ !*&n

<8?35A9Vita Vergilii%& !((n

6;46<81:69 (ed. H. Diels, W. Kranz)3% !!n, %!n3+& %!3$% %*n3#& (n'! %#'& %%n'#>!! &%'#>" )('# !", %+n, &%'" !"'!% !", &&n'!( &%'!$ %#, %"n, &*n'!$.!>% %#'!$.%& %+n'!$.%) %)'!$.&!>( %#'!$.&)>( !"'%*.+>$ %(n'%!.!!>!% %(n'%!.!% !*n'%%.% !('%&.# !*n'&( %"n')* %(n'($ %*n'+*>! %(n'+!.% !"n'$+.!>% %(n'$$.! %(n

'#) !!, %*'#).+ %(n'#+ %*n'!!! +'!!% &>), !*n'!!& !*n'!!( #>!*, !(, &*>!n, &!'!%$ +'!%# %+'!%" %"'!&! %%n'!&$ #n, &*'!&" $n, !*, &*, &*>!n'!)* +'!)! $n'!)& $'!(! %)'!(&a (fr. !(% Wright (>+

6??,A9fr. !&$ ((n

64,1565A9Dissertationes, !%.+ !!$n

64,1A7A9fragments (ed. H. Usener))! !**n, !%!$&>"! !**n$&>#" !**n#* !!(n#% !%!#& !%!#) !**n, !%!#( !%!#$># !**n#$ !%!## !%%n%++ !$(n%"& $%n&*( !$*n&") !)"nKyriai doxai!>) !+&) !+%nLetter to Herodotus&( !**, !*$&#>)* !!&&# !#$, !"&&" (and scholion) !!&, !"%)* (and scholion) !!%>!), !%&)!>( !!))! !"$

Index locorum %!"

64,1A7A9 (cont.))+># !!!, !!(>!+, !)!)">(& !!), !!+n()>$& !!+()>( !(!((>" %**+& $!n++ $*n+#>$& !!%>!&, !!#$&>+ !%%$& (and scholion) !!%, !!+n, !%%$) (and scholia) !%!, !$(, !"&n$+>#% !%*, !%%Letter to Menoeceus!&& !+&nLetter to Pythocles#) !%*, !%##( !%*, !%###>!!* !%%##>"* !%*#" !"&n"!>& !!""! (and scholion) #)n, !**n, !!""& !%*"+ scholion !%!"#>!!+ !($"#>" !%&n"">!!! !%&n!*! %!!*)>( !("!*(>+ !!$n!!!>!( !%&n!!(>!+ !%&nOn naturebook ,,

%).!%.! Arr.% !)!n%).%".!>& Arr.% )!n%).&(>+ Arr.% !!!n%).&+>$ Arr.% !!!n, !&"%).)% Arr.% !)*n%).)&.&>)).! Arr.% )%n%).(*.!$>(!.# Arr.% !*", !!*>!!

book ,-%( Arr.% !!(

book .,%+.&$.!!>!" Arr.% !*+n%+.&#.#>" Arr.% !%*%+.)) Arr.% !)&n

book .,- (ed. G. Leone)col. ...,,, !%(cols. ...,->...,. !%(col. ...-,, !%>!$ !*+cols. ...,.>.:,,, !%(subscriptio !*%, !%"

book .-

fr. $ !%)fr. !! !%)fr. !) !%)fr. %& !%)fr. %) !%)fr. %( !%)fr. &) !%)fr. Q !%)subscriptio !*%, !%"

book ..-P. Herc. !)%*, + ,,, $ !*)n%*1! Long/Sedley !*+%*1(>!) Long/Sedley ##%*1( Long/Sedley !%#n%*1!& Long/Sedley !*$, !)%&).&*.+>$ Arr.% !*(line !+, Laursen (!""%) !*(lines )+>#, Laursen (!""%) !*)

book ..-,,, (ed. D. Sedley)# ,- ">!* !!&n!% - !*(!& ,- ! inf. – - !% sup. !*)>(!& - % inf. – -, ! sup. !*!!& -, # sup. ff. !!*n!& .,,, +>!* sup. !*), !*#>"subscriptio !*(n, !%#>"n

book ...,,&% Arr.% !%$

unnumbered books&& Arr.% !%#&+.%) Arr.% !!&

Sententiae Vaticanae&>) !+%n%" !&>!)n&! !+()* !%#

64,;6?,<69 (ed. H. Diels, W. Kranz)&3!>), # (

65D;8:8B,1A; ;3B?A;s.v. B+4.3' -$) (&n

6A7,4,<69fragments (ed. A. Nauck, R. Kannicht)#&" !+$n

6A95352,A9Commentarii ad Homeri Odysseam&!*.&&ff. %$n

B3:6?On the doctrines of Hippocrates and Plato, $.! +#n,, ( +"n

%%* Index locorum

,, (.# +#n-, %.( $*

B6::,A9- !+.& &"n.-,, !!.!!>!& $%n

26731:,5A9 (ed. H. Diels, W. Kranz)3!% $"n'"& !)n

26731:,5A9 (28;67,1A9)Allegoriae+".# %$n

267842,:A9(ed. H. von Staden)frr. +&>+ +"

269,8<Works and days!>&** +

2,448:D5A9Refutatio omnium haeresium, !).( !$#n-,, &*.&>) $

28;67Iliad.., !*$ ((n..,,, $%* (#Odyssey-, )%>+ (*-,, !**>% ((

287316Ars poetica)+&>+ !&n

:3153?5,A9Divinae institutiones,, !%>!) %&n

:,-D.:- +.& ()n

:A1,3?Pseudologista!$ +*n

:A1,:,A9 (ed. F. Marx)fr. $(& &#, &"

:A1765,A9, !>!)( !(>&)

!>)& !+!>( !$! %&>)%>" %)& %)>(+>" !$!*>%* %(%!># %(>+%">)& !$))>" !+, %+>$(*>+! %#>"()>+! &#()># %*!() %*%(( !"((#>+! !"), %**+%>!)( &%+%>$" !#n, %">&*, !)+++ ($#*>!*! !()n!*%>&( &!>%!*%>%+ !()n!!$>%+ %%!%$>&( !()n!&%>( !)"!&+>)( &%, &(, !#$!)# &$>#, !#$!)">(# !"#!(">+% !"&!+!>$& !")>(!+$># !"+!+$ !"(n!+" !"(n!$! !"(n!$)>%!+ !"(!$+ !"(!+$>$! !")!$+ !"(n!#% !"(n!#%>& !"(!#( !")n, !"+!##>"! !"+!#" !"(n!"! !"(n!"+># &"n, !"(!"+ !"(n!"# !"(n%*& !"(n%*+ !")n, !"(n%*#>!) !"+%!* !"(n%%!>) !"+%%! !")n, !"(n%&%>& !$(n%)% !"(n

Index locorum %%!

:A1765,A9 (cont.)%)) !"(n%)( !"(n%)+ !"(n%)" !"(n%+(>&%# +$n, %**%$!>"$ !!&!# )#n&%">)!$ %**&"#>)!$ %*!)*%>& !+%n)&*>#% !!%)&*>)) !)+n)+)>#% (!)$&>$ (!))"ff. )+)(% )#n)(# )+n)$+ (*)#&>+&) !"$)#&>) !"$(+$ !$)("">+&) !"">%**+&(>"%* $&, !%(, !)(+&(>)) !&n+&(>$ $&+&#>)) !"*+)!>) ++n+)$ff. !)++(! )#$*(>!! !#+n$*(>+ $&$*$>" $)$!%>!& !&n$!&>!) $&$!)>#%" %*n$!)>!( !&n$!( !$)$!+>)! !%>!($!$ !)$!#>%( !)$%( !)$%$ (&n$&%>& %!$&% !&$&)>)! !#, !)%>&$&)>( $&$&$ !&$&#>" !%, !"*n$)) !$)$(&>+% !"$$*>#! !"$#%ff. !$(#!$>%" !"!#%$ &"n

#&*ff. !%)n#&*>% )##$$># !%)#"$>"!) !%)"*$>!) &"n, !"!"%!>(* %%, )(, ((, !&), !&$n, !"*"%& (*, ((")&>( !"!")(>+ ((!*(%>!!!& $#>#%!*+#>$% #*!*$%>#% #*!*"*>! $"n!*")>!!*) #!

,, %*>+! ((%)># ((%)>( ((%( (*, ("$" ("!(& &"n!() )#!+$>#% $)n!#! $)n%!+>"& !)$%%(>&" !)$&*&>$ !"%&"& &"n&") )#n)+& &"n(**>+ (+n(%%>+# #(n(%" &"n+**>+* $(n, !&)+**>)& ()+&( (*+))>+* )(+)( ()+)+>(! !+, %$+((>+* !++##>") &"n#+(>""* !(!!*!&>%% &"n!*#!>& %(n!!*!>) !#!!!*) )#n!!!#>$) !$*!!!#>)& !$*n!!(&>+ $(n

,,, !>%% !#n& (#+># (#">!& +#!! %)n!( !&n, $!n!#>%% (*

%%% Index locorum

"#>!&( )"!&*>( )"!&!>% ()n!&# $!n!)*>% $*n%&& )#n%)( !!$>!#%+% )#n%##>&%% !!#&)& )#n&$*>"( !)"n&$! !)&)$%>& !$))#) )#n+%*>! !)#n$#)>"$ $*>!#*+>!# !)#n#!+>!# !"%#((>+! !$(n#+* )#n#$# )#n"(( (*, (""++>$ +*>!"#*>!*!% ((n!*%)>)) ((!*%(>+ ((n!*&+ (*!*&# ((

,- !>%( )(, !&)%+>)) )*>!, !&$>#%+>)! !)#>")(>(& )*, !&$>#, !)$n(! )!()>%!+ )%n, !)!()>!*" !!!($>+% )%n"( )*n"#>!*" !)!n!!*>%# !!!!!*>!( !)!n!%! !(*!&+ (*, ("!)&>$( !!!!(*>+$ !)!n!$+>%!+ !!!!$+>#( !&">)*!#!>% !)*!#&>) !)!n!"&>) )%n!"" &"n%*& )#n%)$ )#n&&% )#n&## )#n)*)>!& "*n, !&*

)+">(%! #$>")+">$% )+, #(>$(%" $%n(+% )#n($$ (&n$!*>%! "*n$!& )#n$%%>"+! !(!n$($>+$ !)"$+" (&n#%%>($ )$>##&%>& )$#&% )##(#>$+ !(!n#(#>+" $%n#$$>"*+ !(*n##$ )#n#"" &"n"*$>+! !(!n")! &"n")# )#n!!%&>&* ()!!%$ (*!!%" (*!!)" )#n!!+*>" (*, ()!%)# )"n, (*n!%(* )#n

- ">!& !#n%%>() $(n(">+& !)"+)>"* !(&$+>"* !()!!!>!% !&n, !)n!!$ (*, ("!%+>(( !(&!%#>)! $*>!!)%>& !$)!)+>(( ++n!(( !()!(+>") !(&!(+>%&) $(>#!"(>%&) $)n, !(&%&(ff. !(&n%&(>&(* !$+%&(>&%& !$), !$$%)#>" !$)%#$ )#n%"( (*, ("%"" )#n&*+>!$ !$+&!&>!$ !$+&!#>%% !$+&%)>(* !$)&%)>&) !+"

Index locorum %%&

:A1765,A9 (cont.)&&(>$ !+"n&&#>(* !$&&)( !$)&(!>$" !$&n&+!>& !"%n&#*>)!( !$&n&"$ ("&"">)** (")** (")*% (")*( ("))">(" !$))(" %)n)"(># !$)(*">$$* %!, !(%, !()(%+>&& +$n($(>#) #&>)(#(>"! #)("%>+!& !(&n+*! $!n+!* ("+%% !)&++& (&n$%*>&* "*#&$>$$ !"#&$>)! %*n#+)>+ %(n##& )#n!*%#>"* (+!*&)># (+!*&+ (*, (+!*&# ("!!&$ (*, ("!!+!>%%( ++n!%+# )#n!%#!>% !+%n!&$) )#n

-, ! !+!%">&% !+!n)&>+ !())#>$" !+&+#>$" !+%n"%>( %*n, %)"+>!!&$ !(#%*)>!% %!&&% )#n&($>$# !#*&$">)%% !#*&") )#n)%&>(* !#%)%) )#n)(+ )#n)#!>") !("(%)>&) !(#

(%$>&) !+%n($* )#n+*+ (*, ("+*#>&# !$$n$*&>!! +$n$(+ (&n#)#>" (&n")+>$ $%n!**" &"n!*))>+ (%>)!*)+ (*, (%, ()!*+& &"n!*"*>!%#+ (+>$!!&#>!%#+ !+*>(!!!) ($!!(#>" !+)!!#&>) !+)>(!!"+ ("!%*#>!% !+)!%)$>(! !+*n!%(%>( !+(!%+) )#n

;3178',A9Saturnalia, %&.% $"n

;371A9 3A76:,A9, !$.+ !!$n

;6:,99A9 (ed. H. Diels, W. Kranz)'$.% !$(n

;6578<87A9 8@ 12,89 (ed. H. Diels, W.Kranz)'! #(>+

;A936A9 (ed. H. Diels, W. Kranz)3+ (

?3A9,423?69 (ed. H. Diels, W. Kranz)') #+n

8-,<Tristia, !*.!">%% ()

4.2671. ((#!!.)>( !%+n

4.2671. !!!!fr. )).!>+ !**n, !%!, !%%n

42,:8 8@ 3:6.3?<7,3De aeternitate mundi

%%) Index locorum

( !+$n!&>!+ $(n&* !+$n!$ $(n$+>## $(n!!$ !++!!#>&! !++!!#>!" !$+!!" !+$n!%*>& !$$!%)>" !$)!%( !+$!%$ !+$n!&*>! !+#>"!&%>$ !$+!&% !+$n!&# !+$n!&#>)% !$$!)( !$*!)+>" !$!!)+ !$%!)# !$%!)" !$&

42,:8<6;A9De pietate (ed. D. Obbink)&$>)! !!)n%%(>&! !%!%%(ff. !*!%%(>$ !**n(%&>&& !%!(%&>) !**n, !*!!*)&>#" !!+>!$!*(*>) !%%n!*(*>! !**n!*((>" !%#%*))>( !&nDe signis,. !#>&# #),. &#>. !+ #)De Stoicis,.>.- +"n., )>%% !*(nAdversus [...] (P.Herc. !**().-,,, +>!% !!&nfr. !!! !#&

4:358Apology%!b)>( #$nGorgias)(&c%>) !*")()c!>% !*"Laws+$$d !$*n

$$+b ("Phaedo!*#e–!*"a #*nPhilebus!%b)>+ !*"nRepublic&+)e–&+(a (nTheaetetus!#%a &$Timaeus%% !$%%%c !$&n%&a$ !$&%#b–&)d !(%%#b !$"%"e–&*c !(&&*b–c $+&!a–b #*&!b–&%c !*+n&%b#>&&b! $#n&%c $+&&b–&)a #!&)a–)!a !(%&"b–e $+)!a ff. !(%(+e !*+n+%e!%>+&a% #*+&a%>) $"+&b%>) $"+&e&>$ $"

4:,?D 526 6:<67Naturalis historia,- !%.$& ()n

4:A53712fragments (ed. F. H. Sandbach)!"& $)nAdversus Colotem!!*#6>@ !#n, !%+n!!!!@ ff. !#n!!%&' !"nDe defectu oraculorum)%*'>1 !)")%*' "nDe exilio+*$1 #De facie quae in orbe lunae apparet")%' "nDe genio Socratis(#*1 "nDe Iside et Osiride&#%<>6 "n&#$3 "nDe poetis audiendis!)6, !(@ "n

Index locorum %%(

4:A53712 (cont.)De Pythiae oraculis)*+6 "nLife of Cicero%) +)nLife of Crassus& +)nNon posse secundum Epicurum suaviter vivi!!*+3 )!nQuaestiones convivales+)+< +n

[4:A53712]Epitome,, ).!>% $(n

48:6;8fragments (ed. M. Gigante)!%! $#n

48:D36?A9 (ed. A. Tepedino Guerra)fr. %$ !%#n

48742D7DDe abstinentia,,, %* $)n

489,<8?,A9 (ed. L. Edelstein, I. G. Kidd)@# #)@!# #)@!" #)@!!)>!+ #)@!)+ $*

4781:A9In Platonis Timaeum, %$$.# $+n

HA,?5,:,3?Institutio oratoria, ).) %&n. %.!( &"n

96?613Letters##.)& #+nNaturales quaestiones, (.! &"n

96.5A9 6;4,7,1A9Adversus mathematicos-,, %*+>!* )$n

%*" )$&#* +"n

Outlines of Pyrrhonism,,, &! $)n

&% !%+n

9,;4:,1,A9In Aristotelis Physica!+!.!)>!( %#n

STOICORUM VETERUM FRAGMENTA (ed. H. vonArnim), !)# +#n,, )*& $&n

#"$ +#n,,, Diog. %" +"n

Diog. && +"n

SUDAs.v. Empedocles !!n, %!n

5268? 8@ 9;D7?3!)>!( $!*).!>& (

5268427395A9fragments (ed. W. W. Fortenbaugh, P. M.Huby, R. W. Sharples, D. Gutas)!#) !++>$", !#!%%! !$#%%( !)+n, !$#%&* $+>$%)!3>' !$"(#)3 !$%nCharacters!+.% $De causis plantarum-, ! $+nDe sensibus(>+ $+n#&>+ $+nHistoria plantarum, +.) +Metaphysics!*a%#>" !$$Metarsiologica (ed. H. Daiber)+.+$>#! !#*!&.)&>() !#%!).!#>%( !#*

52A1D<,<69, !>## !*&n,, )$>() !+*

)".& !+))".$ !+)

-,,, )+.& !&!

-,7B,:Aeneid, )+% +*

%%+ Index locorum

VITA PHILONIDIS (ed. I. Gallo)$ !*%!& inf. !!$n

-,57A-,A9-, %.& &"n

.6?8428?Memorabilia,- ! $+

Index locorum %%$

Academics, +%>&, +$, $)n, $+>#, $">#%, #)n,#(n, "*n, !(&

Aenesidemus, $%n, #+n, #">"*, "%aether, %n, #, !(Aetius, !($>"allegory, $)>(Amafinius, &#analogy

argument by !!, +$, #&>), !*$alphabetic, &", )#n, !"*>!, !"(

Anaxagoras, )#, $&>), !%)>+, !)(>+, !#&,!"!

Anaximander, !", #*n, #%n, !$#Anaximenes, $)Andronicus, +&Antikythera mechanism, #&Antiochus, +%>), $$>#, #(n, #$n, #"nAntipater, #$nApollonius of Rhodes, "Aratus, !, (*nArcesilaus, #(>+, #$nArchimedes, #&, "!Aristarchus, "!Aristion, Epicurean tyrant, +%Aristotle, $", #+, !$#>"

re-emergence at Rome, +&>)on teleology, !"influence on doxography, %*, !%)>(, !#&,

!(!nschool treatises little known in Hellenistic

period, +&, "!>%, !#&>)Physics !*), !#+on location of mind, +#De philosophia, !+#n

Asclepiades, +"n, $%nAthenodorus, +&>)Athens

Lucretius on, ($, !+*>!, %*&as philosophical centre, +%>), +$, !%">&*

atoms, Latin vocabulary for, %#, &#>", !"&>#,%*%

Atticus, friend of Cicero, +)Atticus, Platonist philosopher, !$"Aurispa, G., had copy of Empedocles’

Katharmoi, &authority, philosophical, !#, %!, +$>#, +#>"

Berosus, "*>!Boethus, $(Bolos, "*nBrutus, +&

Caesar, Julius, +%n, +)>(Calcidius, "&nCalliope, %%, %)Carneades, $)n, #$nCassius, &", +)Catius, &", )%, !)*Cato the Younger, +&Chrysippus, +#>", !+$Cicero, xvii, ++n, #+, #"

letter about Lucretius, !>%, %%, %*&philosophical works, &+>$, )&>), +$, $$>#,

#!, #%and translation of Greek philosophical

terms, &+>", )%, )&>), )"philosophical position and contacts,

+&>(Cleanthes, $&nCleomedes, #), "&nColotes, !#n, #+n, #"ncompound adjectives, %)>(, (!, ()n, ("Crantor, $+n, $#nCrassus, +)Cratippus, +)creation, $(>#, #*>!, #%>&, !(&Critias, !%!Cyzicus, mathematical school of, #&

Demetrius of Laconia, $*, #)nDemocritus, !$#

as forerunner of scepticism, #+>$

%%#

General index

attitudes of Epicurus and Lucretius to, !#,((, $&, !)%>&, !)"n

Diagoras, !%!Dicaearchus, +)n, $%nDiogenes of Apollonia, $), !$#Diogenes of Babylon, +"nDiogenes of Oenoanda, $), "&n

elementstheory of four, %, ", !!>!(, !$>!", %), %",

&&n, ))>(, $&>), $#>#%, !*+n, !$)>+,!$$, %*!>%

Epicurean critique of rival theories of, $),!%&>+, !)(>+, !(%, !"*>%

see also aether, skyEmpedocles, !>&), !$#n

model for Lucretius, xv, %, !*>&), ))>+,!"*, %*!>%, %*&>)

Roman interest in, !>%, %%Katharmoi, %>!*, !&n, %%, &*On nature, %>&, +, #>!*, !!, !&n, !$, %*,

%!>&, %#>&)Strasbourg papyrus, !*, !*>!!n, %&, %#>"evolutionary theory, !">%*criticised by Lucretius, !#>%!, $&, !)%>&

Enniusinfluence on Lucretius, %, %%, &!>%, &&, ((,

($ninfluence of Empedocles on, %%n, &%n

Epicureanism, &+, )+as universal philosophy, xv, ("in Italy, +)>#ethics, !+, ((, !!*, !)), !+!>(evolutionary theory, !">%*on world’s destructibility, !+#>$$

Epicurusfoundation of schools, !%">&%, !+(On nature, xvi, %*n, $&, #&, #+, ##, "&,

")>!&&, !&)>+*, !+$n, !#+>"*, !"$,!"", %*&

epitomes, !**, !*+, !%&, !&%>&, !&#>)%Letter to Herodotus, !**, !*#n, !*", !!!>%&,

!%$, !&!, !&%>&, !&(>)%, !)+>$, !(!,!#$>#, !"%, !"&n

Letter to Pythocles, !**, !*#n, !*", !!">%%,!&!, !&%>&, !(&n, !($, !("

Letter to Menoeceus, !*+Great epitome, !!%, !!)n, !&(, !&#, !)!>%,

!))Kyriai doxai, !+&Against Theophrastus, !#)deathbed letter, !+(other writings, !*!, !!*, !%%nmethodology, !*", !!&, !(+, !"*>&, !"$,

!"", %**, %*%

Lucretius’ debts to, xvi, !#, &), )(, $*>%,#(>"*, "&, !&)>+(, !#%, !#(, !#+>"&,%*&>)

Lucretius’ praise of, %">&*, ((, ($>", +$>#,$(n, !(), !+!, %*&

Epimenides, (Erasistratus, +", $%nEratosthenes, "!Eudoxus, #%>&euhemerism, $%

Galen, +"geocentrism, $#>#%, !)&n, !)+, !"*gods, Epicurean, !+, %+>$, ++n, !!$, !%!>%,

!%#, !(&, !(), !($, !+&Gorgias, %%, #$nGraecisms, %), )">+!

harmonia, )", (*n, ()nheavenly bodies

size of, +$, #&>(, !!">%*, !(&nsee also moon, sun

Heracles, $(nHeraclitus, $", #"

obscurity, !&>!), !"*fire doctrine, $&>), !)+, !$#

Herculaneum papyri, ")>!&&, !)%discovery of, ")>#

Hermarchus, !#n, +#Herophilus, +"Hesiod, +, %+, %", $(nHipparchus, "!Homer, %, !!, %), &!, (!, ((

influence on Empedocles, %)

images (simulacra, eidola), &">)%, !!!>!%,!!)>!+, !&#>)!, !)#>(*, !(!n

Jerome, St, xvii

katharmoimeaning of term, )>$see also Empedocles

Lampsacus, !%">&%language

philosophy of !*(, !%$limitations of, &%>&, )(, )#>(*, !#$origin of, (+, !%%, !(&Greek and Latin, xv, &(>+!

Lucilius, &#Lucretius

life and character, xvii, (%>), +%, $%isolation from contemporary Epicureanism,

+)>$%, !)%

General index %%"

Lucretius (cont.)ignorance of recent science, xv, $*>%, #%>(ignorance of recent philosophy, xv–xvi,

+(>"&Lucullus, +&

magnets, (%>), $%n, #), !(#Melissus, %%Memmius, %%, (), !#$Metrodorus of Chios, #(>$Metrodorus of Lampsacus, +#, !*)>(, !!#,

!%$mind, see soul and mindmoon

Berosus on, "*>!halo of, !(#>"see also heavenly bodies, size of

Musaeus, (

nature (natura, physis), !+, %(>+, &$, !$*n,%*&

personified, !+, +*>!Nausiphanes, $&, #+nNigidius Figulus, +&nutrition, $%n, !(*

omissions by Lucretius, #&, !)), !)$n, !)#,!(&n, !+*, !+!, !+), !$$, !#+>$, !"%>&,%**

oracles, )>(, !%>!), !"*nOrpheus, (nOvid, %%n, ()

Palaephatus, $%nPanaetius, $(Parmenides, %&, &&Paul, St, ()Pausanias, dedicatee of Empedocles’ On nature,

+, %%, %#>", &%Peripatetics, +&>), +(, $&, "%, "&n, !#)>(phenomena, cosmic, %!, !%!>&, !((, !($>"Philo of Alexandria, !++>$, !$*n, !$)Philo of Larissa, +%>&Philodemus, +)>#, "&, !!"n, !&*

On signs, ++>$, #)>(, !*!nOn music, "+On piety, !*!, !*&nlibrary, +(, $*, "(>#, !)%

Philonides, biography of, !*!n, !*%, !!$nphysics, Greek and Latin names for, &+>#physis, see naturePiso, L. Calpurnius Caesoninus, +)>(Piso, M. Pupinius, +&>)plague passage, xvii, ($, !&), !+*>(, %*&Plato

on soul, +#>$*on Socrates, #+Republic, !*#Theaetetus, &$Timaeus, +#, +"n, $(>#, $">#%, !%(, !(!,

!(%>&, !$%, !$$>"criticised by Epicureans, $&, !*+>$, !%(>+,

!(!, !(%>&, !#&Platonists

influence in second century 3<, "%, "&nsee also Academics, creatonism, geocentrism

Plutarchon Empedocles’ proem, #>"interpreter of Plato, !$"

pneuma, $), $$n, !+$Polemo, $+>#, #*, !(&Polyaenus, +#Posidonius, +&, +">$*, #%>&Praxagoras, +#prester, )#n, (*n, !(#>", !#%Prodicus, !%!proems

to DRN, ++n, !&), !)(, !(&>), !(+>$, !#+to book ,, !(>&), ($, !)+, !)", !+&, !"),

%*&to book ,,, ()>(to book ,,,, (#, +#, $%, !+&to book ,-, &">)%, )(>+, !&$>#, !)#>(*,

%*)to book -, $%, !)", !(%>)to book -,, !(), !+!, !+&to Empedocles’ On nature, #>!*, %%>&)

prose and verse, +(>+, !)*>!, )&>+, !"*Pyrrho, #+nPythagoras, %">&*, &&, $(nPythagoreanism, $n, &&n, +&, +#, #*n, #%nPythocles, !&!

religion, criticised by Lucretius, &*, &%, )(, (),$%, $(, !(), !#*>!, !"#>"

Rome, as philosophical centre, +%>(

Sallustius, !, %%, )+nSamothrace, (%>)Satyrus, +scepticism, )+>$, #(>"*scholia, on Epicurus, #)n, "!, !**, !*",

!!%>!&, !!", !&#, !)!>%, !$(nself-refutation, )+>$, #$>"Sextus Empiricus, "%>&Sicily, Lucretius on, !%>!(, (&nsimile, multiple-correspondence, !!simulacra, see imagesSiro, +(sky, symbolising air, !)>!(, !$

%&* General index

Socrates, #(>$, !*(, !%+nsoul and mind, +#, $*, $!n, !!+>!", !%$, !)(,

!)">(%location of mind, +#>$!

Speusippus, $#n, !$#>"Staseas, +)Stoicism, !+n, &+, +&

terminology, )&>), !+$debates with Epicureans, +(, +$, $&,

#%>(on location of mind, +#>$*on elements and principles, $&>), $+>$ignored by Lucretius, $&>#(, "!>%, !#)on creation and destruction of world, $(>#,

!$#on arrangement of cosmos, $#>#%

structure of DRN, xvi-xvii, !))>+, !(%, !((>$,!+*, !#+, !"%

sundistance illusion, "*n, !&*>!as source of light, !)*, !(&nsee also heavenly bodies, size of

swerve, atomic, !%$, !)$

teleology, !#>%*, )$>#, $)n, #%>&, !(*>!,!(%>&, !$*n, see also creation

terminology, technical, &(>", )%, )&>#tetrapharmakos, !+%>&Thales, $), !%(Theophrastus

Characters, $imprint on Lucretius, xvii, #%n, !)+, !)#n,

!(!n, !($>", !++>#(, %*)

doxography, %*n, $+, #%n, !%)>+, !)(>+,!)#n, !(!n, !($>", !$$>#(, %*)

Cicero’s study of, +)nThucydides

Lucretius’ source on the plague, ($, !+*,!+&>(, %*&

book length in, !*&thunderbolts, !(#, !("n, !#*>!time, !!%, !!+, !!#>!"Timocrates, !*!tmesis, )$>#transmigration, &*, &!, &%>&transpositions by Lucretius, xvii, #&, !&+>#,

!))>#, !(*, !(%, !(&>), !((, !#+, !"*>!,!"">%**

Varro, &+>$, +&Venus, in Lucretius’ proem, !(>!$, %&>$Virgil, !!, %%, +), !((

world, destructibility of, $(>#, !(%>&, !++>$"

Xenocrates, $#n, !$#>"Xenophanes, !$#Xenophon, on Socrates, $+, #+

Zeno of Citiumfounder of Stoicism, +#>", $+on location of mind, +#>"Republic, +"non fire, $&nand Theophrastus fr. !#), !++>$, !+#n

Zeno of Sidon, +$, #), "+

General index %&!

Adams, J. N., (!nAndré, J. M., (#nAngeli, A., !!"n, !+&n, !#&nArnim, H. von, !++, !+$n, !$+>$nArrighetti, G., "$n, !*!n, !*$>", !!!n, !!%n,

!!#n, !%*, !%&n, !%$n, !&#n, !)%nAsmis, E., !+n, +(n, +$n, #)n, !&#n, !"&n,

!")>(n, !"+n, !"#n, %*!nAuvray-Assayas, C., +$nAvotins, I., )!n

Bailey, C., %)n, &"n, )!n, )$n, )">(!, (), (+,(">+!, #(n, !&#n, !(!n, !"$n

Baltes, M., !$"nBarigazzi, A., !)!n, !)+nBarnes, J., &n, +&, +$n, $$n, #&n, #)nBen, N. van der, $n, #, "n, %&nBergsträsser, G., !$"nBignone, E., !+n, !%"n, !+#n, !$&n, !$$nBloch, H., (&nBockemüller, F, !+*nBollack, J., %n, %)n, !!"nBoyancé, P., #(n, !&#n, !+#nBrieger, A., !&#nBright, D. F., !+*n, !+!n, !+)nBrunschwig, J., #(n, #+n, !"%Buffière, F., %$nBurkert, W., &*nBurnyeat, M., )+, #(n, #$n

Caizzi, F. Decleva, #"nCampos Daroca, J., "!nCapasso, M., +)n, "#n, !!&nCasson, L., (&nCastner, C. J, !*n, +(nCavallo, G., "#n, ""n, !*&n, !*(n, !!#n, !%"nCavini, W., !*&nClay, D., !(n, !+n, %*n, %%n, %(>+, &!n, +$n,

#%n, !&(, !)!n, !+*n, !+%n, !+$n, !$*nCommager, H. S., !+*n, !+)nComparetti, D., and G. De Petra, "(n

Courtney, E., %$nCrönert, W., !*%n, !%+n

Daiber, H., !("n, !$"n, !#*>!, !#%nDe Lacy, P. H., +(n, $)n; and E. A. De Lacy,

+$nDelattre, D., !*!nDiano, C., !%+nDiels, H., %, !($n, !+$n, !$!n, !$"n; and W.

Kranz, &n, (n, $nDillon, J. M., &&n, $$n, "&nDodds, E. R., !*"Dorandi, T., +%n, +(n, +"n, ""n, !&*nDrexler, F., !&#nDüring, I., +)n

Edelstein, L., %"nEdwards, M. J., !&nErler, M., !*"n, !!*n, !!%n, !&!, !&(nErnout, A., and L. Robin, $(n, #(n, !"$n

Farrell, J., ($nFerrari, W., )!n, !&#nFerrary, J.-L., +%nFortenbaugh, W. W., +)n, !++nFowler, D. P, !*n; and P. G. Fowler, !)n, !&#n,

!+*nFurley, D. J., %n, !*n, !+>%*, &*, $), $#>#*,

#!n, "!, !(!n, !#&n

Gale, M., !*n, !(n, )(n, $%n, !&$n, !"#nGallo, I., !*%n, !!$nGiancotti, F., !(nGigante, M., )$n, +)n, +(n, $#n, "#nGilbert, W., (&Giussani, C., !&#n, !(!nGlucker, J., +%n, +&nGodwin, J., )!n, (%nGörler, W., $$n, #"nGottschalk, H., $!n, !"#>"Goulet, R., +)n, "&n

%&%

Index of modern scholars

Graeser, A., !++nGraham, D. W., %nGriffin, M., +%n, +)n, +(nGrilli, A., )$nGutas, D., !++nGuthrie, W. K. C., )n

Habicht, C., +%nHadot, P., +&nHardie, P. R., %%n, &*n, &%nHartung, H. J., &+nHayter, J., "+>$Henrichs, A., +"nHinds, S. E., )#nHölscher, U., %nHouse, D. K., "%nHuby, P. M., !++n

Inwood, B., &n, +n, #n, !#&n

Jobst, F., !*nJocelyn, H. D., +%n

Kahn, C. H., &n, &&nKenney, E. J., %n, !(n, (*n, ()n, (#n, $!n,

$%n, !)#nKerschensteiner, J., !$#nKidd, I. G., !++n, !+$nKingsley, P., %n, &n, #n, #>"n, !(n, )(n, #*nKirk, G. S., !$#nKleve, K., !(n, ++, $&n, #(n, "%nKoenen, M., !)!nKollman, E. D., !&n, !"*nKranz, W., !*n; see also Diels, H.Kuhrt, A., "!n

Lachmann, K., )+Lackenbacher, H., )!n, !)!nLaks, A., !!"nLangslow, D., ($nLapidge, M., !+$nLaursen, S., "$n, !*(n, !*+n, !%+nLeone, G., "!n, "$n, !*%n, !*+n, !%(, !%"nLévêque, P., $(nLévy, C., )&n, #"n, "*nLewis, N., (%nLloyd, G. E. R., !!n, ))nLocke, J., "(nLong, A. A., %n, &n, +$n, $&n, $+n, #+n, "%n,

!+!n, !$#; and D. N. Sedley, %+n, )+n, ++n,+$n, "$n, ##, !*+n, !%$n, !+&n, !++n,!#&n, !"%n

Longo Auricchio, F., !#n, +#nLorenz, H., #*nLück, W., !&(n

Maas, P., $nMcDiarmid, J. B., !+$nMacKay, L., !)nMansfeld, J., &, +"n, $!n, #"n, !!"n, !%*n,

!%!n, !%(n, !)+n, !)#n, !($n, !("n, !$",!#*, !#!n, !#%n

Martin, A., !*>!!nn, %"n, &*>!nMarullus, &"nMayer, R., (#nMewaldt, J., !&$>#nMilanese, G., ++n, !"*nMillot, C., "$n, !*%n, !%)n, !%"nMoraux, P., +)nMuehll, P. von der, !!!, !&%, !&#nMunro, H. A. J., !"$n

Norden, E., !+#n

Obbink, D., !&n, ++n, +"n, "">!*&nn, !!$,!%%n, !%+n, !%#>"n, !&$n

O’Brien, D., %n, &n, #Osborne, C., &, +n, "n, !!nOwen, G. E. L., $+n

Panchenko, D., !)+nParker, R., )n, (n, $nPearson, A. C., !++Pease, A. S., $"nPhilippson, R., #)nPiaggio, A., "(>$Pigeaud, J.-M., +"n, $%nPizzani, U., !&$nPowell, J. G. F., ))nPrice, D. de Solla, #&nPrimavesi, O., !*>!!nn, %"n, &*>!nPuglia, E., $*n, !%$n

Radl, A., (%nRawson, E., +%n, +"nReiley, K. C., &+n, &#n, !"&nReitzenstein, E., !$", !#%nRobin, L., see Ernout, A.Romeo, C., #)nRösler, W., %*n, !%)n, !)+nRunia, D., $+n, "!n, !($n, !("n, !+#n, !$%

Sambursky, S., "%nSandbach, F. H., +&n, "!n, !#)nSchmalzriedt, E., %%nSchmid, W., !*+n, !#&Schmidt, J., +$n, $)>(, $", !&(nSchofield, M., !%+Schrijvers, P. H., +"n, $%n, #(n, #">"*, !(!nSeele, A., (*nSegal, C., ((n, !+!n

Index of modern scholars %&&

Sharples, R. W., !++n, !$!n, !$(nSkutsch, O, %%n, &!nSmith, M. F., !&>!)n, &"n, +*n, +%n, !((nSnyder, J. M., !)nSolmsen, F., %n, $%n, !(*, !+$n, !+#nSteckel, H., !*"n, !!*nSteinmetz, P., +)nStriker, G., %(n

Tatum, W. J., !*n, !&n, !"*nTownend, G. B., !()nTyrrell, R. Y., !n

Usener, H., $%n, !%!, !$"n

Vallance, J. T., +"nVander Waerdt, P. A., !#n, &&n, #(>"nn

Vlastos, G., !"&nVogliano, A., "$>#, ""n, !*!n, !*(n, !!*n,

!!!n, !&"n

Wallace, R., (%n, (&nWardy, R, )#nWaszink, J. H., &!nWeise, F. O., &$nWellmann, M., "*nWest, D., !!, )$, (#n, $!nWiersma, W., !+$nWinckelmann, J. J., "+Woltjer, J., !&#nWoolf, G., (#nWright, M. R., &n, (n, $n, #n, !*n, !!n, !&n

Zeller, E., !++>$, !$#

%&) Index of modern scholars

David N. Sedley Lucretius and the Transformation of Greek Wisdom 1998

Documents

lucretius source

transformation of book

provenance of empedocles

empedocles proem

enigma of lucretius

favourite books

conclusions lucretius

resolution appendix