OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

The Atheism of Epicurus 

Dirk Obbink 

E PICURUS, who defines himself in opposition to notorious 
atheists of the classical period, argues emphatically in extant 
works for the existence of a supremely blessed, living, and 

imperishable form of divinity ('tov 9£ov ~cpov acp9ap'tov Kat J.1a­
KUptoV). Yet there was a tradition current in many circles in 
antiquity that Epicurus was himself an atheist;! and in modern times 
his hedonism and his opposition to teleology have often been 
equated with a general godlessness and contempt for traditional 
religion.2 Scholars have long puzzled over just how this discrepancy 

1 E.g. Cic. Nat.D. 1.123 citing Posidonius (fr.22a Edelstein-Kidd=346 Theiler, 
quoted 208 infra); 1.85 (Cotta). Cf Sext. Emp. Math. 9.58 (209 infra); Pluto Mor. 
1102B, 1112D, 1119D-E, 1123E; Lact. De ira Dei 4.7 (Posidonius fr.22b E.-K.=346 
Th.); Inst.Epit. 31.3. A more complete list in M. Winiarczyk, "Wer galtimAltertum 
als Atheist?" Philologus 128 (1984) 157-83 at 168-70. 

2 A common sentiment is Clem. Alex. Protr. 5.66.5 (CCS 52 [15] p.51.6-9 Stahlin­
Treu): 'E1ttKOUPOU ~EV yap ~6vou Kal. EKrov tKh:rlO"O~at, 0<; OUOEv ~tAEtV OtE'tat 't ii> 
SEep, btu 1t<lv'tcov ao"E~rov. In modern times such representation is commonplace 
(e.g. E. Zeller, Die Philosophie der Criechen IILls [Leipzig 1923] 429f, 437 n.2), 
including a long line of German idealists, from Hegel through Marx to Sartre. 
Early protests were issued by F. Bacon (Essays no. 16), P. Gassendi (De vita et 
moribus Epicuri libri [Leiden 1647], Animadversiones in decimum librum Dioge­
nes Laertii [1649]), and later by G. Guissani (Studi Lucretiani [Torino 1845]). The 
first English translator of Lucretius, a young Puritan named Lucy Hutchinson 
who worked in the late 1640's, complains in her preface that she did not really like 
the poem and only translated it so that people might take a warning from its pre­
posterous impiety (c. A. Gordon, A Bibliography of Lucretius2, ed. E. J. Kenney 
[Suffolk 1985] 169). Even the editors of the earliest published portion of Philo­
demus' De pietate could write: "The Stoics and the Epicureans, who made lofty 
pretentions to popular applause, as the philosophical expounders of the popular 
religion, loudly accused each other of that atheism, of which both affected to be 
abhorrent, and of which both were so indubitably culpable. The atheism of the 
Epicureans seems not to have been questioned by any men of learning, though 
their exoteric doctrines were so well disguised as not to offend the multitude" (W. 
Drummond and R. Walpole, Herculanensia [London 1810] 123); cf Quarterly 
Review 3 (1810) 20, which characterized the subject of the treatise as "the piety of 
atheism." 

187 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

188 THE ATHEISM OF EPICURUS 

could have arisen in our ancient sources. 3 Now we know. New 
work on the papyri recovered from Herculaneum shows that the 
charge of atheism against Epicurus had its origin in polemics over 
theology, epistemology, and cultural history, played out in the 
philosophical schools of the Hellenistic period. More specifically, I 
shall show here that the charge stemmed partly from Epicurus' 
rejection of teleology and divine providence, partly from the con­
viction that no one was ever in a better position than Epicurus to 
dispense with the gods, once and for all. As a result, many in 
antiquity held that he that he actually did so. If Epicurus was in fact 
an atheist, his views would hold little interest for the history of 
religions or of philosophy. If he was not, how did he come to be 
regarded as such? 

Our evidence is twofold. We have to deal, on the one hand, with 
the reports of individual doxographers-among them Cicero, 
Sextus Empiricus, and Plutarch-each with his own philosophical 
aims and predelictions, each dependent upon his own selection of 
sources. Over against the doxographic record we now have, on the 
other hand, the new evidence from Herculaneum. One text in 
particular, an apologetic De pietate by the first-century Epicurean 
Philodemus, presents important information on the issue of Epi­
curus' atheism. By reconsidering the doxographical record in light 
of this evidence we may obtain a much clearer picture and chron­
ology of the arguments in the Hellenistic debate. I begin with a gen­
eral comment on the incidence of atheism in antiquity and a brief re­
view of the climate of opinion (including Epicurus' own theological 
views) in which such charges arose, then turn to new versions of 
several texts that reshape our understanding of the problem.4 

3 See e.g. Cic. Nat.D. 1.86, where the Academic speaker Cotta declares: ille (sc. 
Epicurus) vero deos esse putat, and later in the same book (1.123) says that nullos 
esse deos Epicuro videri; cf A. S. Pease, M. Tulti Ciceronis De Natura Deorum 
Libri III (Cambridge [Mass.] 1955) 36-45, esp. 43-45. A similar discrepancy exists 
in Sextus: Math. 9.58 includes Epicurus in a list of atheists, yet this is directly con­
travened at 64, and at 43-47 is related Epicurus' physical account of the gods: see 
R. Philipps on, RE Suppl. 7 A (1939) 1154; Pease 44f. 

4 A new edition of De pie tate , edd. A. Henrichs and D. Obbink, is in prepara­
tion. The standard earlier edition is Th. Gomperz, Philodem ;;.ber Frommigkeit 
(=Herculanische Studien 2 [Leipzig 1866]), to be supplemented with caution by R. 
Philippson, "Zur Philodems iiber die Frommigkeit," Hermes 55 (1920) 225-78, 
364-72; 56 (1921) 355-410; and the 1923 Konigsberg dissertation of A. Schober, 
Philodemi De Pietate pars prior, now printed in CronErcol18 (1988) 67-125, with 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 189 

I. Atheism in Antiquity 

Atheism in the ancient world was never a well-defined or 
ideologically fixed position.s But deviation from a proper attitude to­
wards the gods, particularly as recipients of cult, could result in a 
charge of impiety or in the suspicion of atheism;6 and we know of 
many atheists by name, most of them philosophers who worked in 
or around Athens in the late fifth century B.C. Their views ranged 
from explicit denial of the gods' existence to a rejection of the 
grounds for such belief. Philodemus provides a classification of 
opinion:? (1) those who say that it is unknown whether there are 

a preface by Marcello Gigante (65£). Numerous re-editions of individual fragments 
have appeared in recent years, including one substantial continuous portion: A. 
HENRICHS, "Die Kritik der stoischen Theologie im P.Hercul. 1428," CronErcol4 
(1974) 5-32 (hereafter 'Henrichs'); cf "Towards aNew Edition of Philodemus' 
Treatise On Piety," GRRS 13 (1972) 67-98. 

5 A. Henrichs, "The Atheism of Prodicus, n CronErcol 6 (1976) 15-21 at 20. 
Further on atheism: P. Decharme, La critique de traditions religieuses chez les grecs 
(Paris 1904); J. Geffcken, lwei griechische Apologeten (Leipzig/Berlin 1907); F. 
Mauthner, Der Atheismus und seine Geschichte im Abendlande (Stuttgart 1920-
23); A. B. Drachmann, Atheism in Pagan Antiquity (London 1922); W. K. C. 
Guthrie, OCD2 S.V. "Atheism"; W. Nestle, "Atheismus," RAC 1 (1950) 866-70; O. 
Gigon, "Atheismus," Lexikon der alten Welt (Zurich/Stuttgart 1965) 370f; W. Fahr, 
e£oV~ vOflil;uv. lum Problem der Anfange des Atheismus bei den Griechen (Hil­
desheim 1970); D. Babut, La religion des philosophes grecques (Paris 1974); A. Hen­
richs, "Two Doxographical Notes: Democritus and Prodicus on Religion," HSCP 
79 (1975) 93-124; K. J. Dover, "The Freedom of the Intellectual in Greek Society," 
T alanta 7 (1976) 25-54; P. A. Meijer, "Philosophers, Intellectuals and Religion in 
Hellas, n in Faith, Hope and Worship: Aspects of Religious Mentality in the 
Ancient World, ed. H. S. Versnel (Leiden 1980) 216-63, and the standard bibliog­
raphy on the major 'atheists' (including Epicurus), esp. Diagoras, Prodicus, Critias, 
Anaxagoras, Protagoras, Euhemerus, and Theodorus. 

6 E. Derrenne, Les proces d'impiite intentes aux philosophes a Athenes (=Riblio­
theque de la Faculte de Philosophie et Lettres de l'Universite de Liege 45 [Liege 
1930]); J. Rudhardt, "La definition du delit d'impiete d'apres la legislation attique, n 

MusH elv 17 (1960) 87-105; M. Ostwald, From Popular Sovereignty to the Sover­
eignty of Law (Berkeley 1986), esp. App. A-C (525-50). 

7 Philod. De piet. (P.Hercul. 1428 colI. XIY.32-xv.8) p.25 Henrichs. Sextus' divi­
sion (Math. 9.50£), by contrast, is simply into those who say aUK dval. e£O<>~, and 
those who, like good Pyrrhonist skeptics, suspend judgment on the question. I do 
not know whence Philodemus' classification derives; perhaps it is his own. It is in 
any case unlikely to be Academic (e.g. Carneades or Antiochus) because it does 
not exhibit the conflation of categories (2) and (3) apparent in Sextus' grouping, 
taken over from Carneades. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

190 THE A THEISM OF EPICURUS 

any gods or what they are like (11 tou<; liYVCOOtov et nVE<; deH. edot] 
AE'YOVta<; 111tOtOl nVE<; dow); (2) those who say openly that the gods 
do not exist (11 tou<; ~h[ a] PPllOllv on OUK d[ o]tv a1to<patVOJ1EvO'U<;); 
and (3) those who clearly imply it (11 cpavepou<; ovta<; ro<; avTIPo'Uv). 

This scheme implies that the charge of atheism could be incurred 
for something less than an outright denial of the existence of the 
gods, but the later doxographical and biographical traditions tended 
to lump Philodemus' first and third groups into the second. The 
standard rosters of atheists, for instance, regularly if falsely include 
Protagoras among "those who say that god does not exist" (Sext. 
Emp. Math. 9.56, cf. 50f). In view of explicit statements that Epi­
curus did deny the existence of the gods, I shall take such denial, or 
blatant disregard for their cult (commonly construed as tantamount 
to denial), to be the main points at issue-though Epicurus' views 
might well turn out to be in some weaker, rhetorical sense 
'atheistic' or 'heretical' by implication in the judgment of a particular 
individual.8 

II. Gnosis Theon 

Denial of the existence of the gods as a point of doctrine is not 
generally associated with any of the philosophical schools of the 
Hellenistic period. On the contrary, Sextus Empiricus reports that, 
for any philosopher who wished to put forth positive views, a 
doctrine about the gods was by far the most necessary, and that it 
seemed so to the dogmatists themselves.9 We may suppose he 
meant that any view of the nature and influence of the gods is 
especially open to suspicion and thus constitutes a kind of natural 
division between skeptical and dogmatic schools: anyone who 
purports to philosoyhize dogmatically would be faced with a chal­
lenge from skeptica quarters to produce a theology not only con­
sistent with one's own system but also capable of some measure of 

8 In the way, for instance, that a Moslem might accuse a Christian of being an 
'atheist', or when a newspaper editorial alleges that the 'bishop of Durham is an 
atheist' for taking a certain position in a church controversy. None of the persons 
engaging in such rhetoric seriously believes the other denies the existence of a 
divinity. The concept of 'heresy' on the other hand, familiar from western church 
history, requires a sense of orthodoxy which simply did not obtain in ancient 
religion. 

9 Math. 9.13: <> 1ttp\ atmv AOyOC; 1t(lVU avaY1'au)'ta'toc; dvat bOKti 'tOtC; bOYJla'tt­
K&c; CP1.A.ooexpo'xJ\v. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 191 

certainty. For Academic skeptics, as for the Pyrrhonist Sextus, be­
liefs about the gods afford no more or less truth than any other 
opinion. In every case skeptics advise suspension of judgment 
concerning certain knowledge (b[oXTt 1tEPl 1t(1v'tO>v); they maintain 
that in the balance of dispute over any given proposition, each op­
posing argument cancels the other (n toov CtVtlK£qlEVroV 1oyrov 
iaoa6tv£ta), thus making suspension of judgment a dialectical 
necessity. 10 Propositions about the gods-their diverse origins, for­
tunes, and significance among mankind, and the wide variation in 
ritual, literature, and belief fostered by Greek polytheism-are 
notoriously susceptible to such analysis. Thus they present an 
essential challenge, as Sextus suggests, to dogmatic philosophers­
especially those who, like Stoics and Epicureans, take an emphati­
cally dogmatic position on the issue of epistemology. 

Unlike those dualists who adopt a separation of the intelligible 
world from the world of physical perception (on the grounds that 
particulars are unstable and hence unknowable), both Stoics and 
Epicureans are staunch materialists. They ground all knowledge of 
the intelligible world in sense perception but allow for a com­
plementary system of inference from signs and 'universal views' 
(KOtVal EVVOtat). Thus both arrive at a position (variously argued) 
whereby certainty of knowledge is attainable, not only regarding 
mundane appearances and everyday life, but also about complex 
questions of ethics, physics, and cosmology. For the sake of co­
herence and consistency, and in order to appropriate within their 
systems a fair portion of pervasive cultural experience, the criterion 

10 See A. A. Long, "Stoa and Sceptical Academy: Origins and Growth of a Tra­
dition," LCM 5 (1978) 161-74; G. Striker, "Sceptical Strategies," in M. Schofield et 
al., edd., Doubt and Dogmatism (Oxford 1980) 54-83; D. N. Sedley, "The Moti­
vation of Greek Scepticism," in M. F. Burnyeat, ed., The Sceptical Tradition (Ber­
keley 1983) 9-23; M. Frede, "The Sceptic's Two Kinds of Assent and the Question 
of the Possibility of Knowledge," in R. Rorty et al., edd., Philosophy in History: 
Essays on the Historiography of Philosophy (Cambridge 1984) 225-54 (=Essays in 
Ancient Philosophy [Minneapolis 1987] 201-22); H. Tarrant, Scepticism or Platon­
ism? The Philosophy of the Fourth Academy (Cambridge 1985); A. M. roppolo, 
Opinione e scienze: il dibattito tra Stoici e Accademici nel III e nel II secolo a.c. 
(Naples 1986), reviewed by J. Annas, "The Heirs of Socrates," Phronesis 33 (1988) 
100-12. The approach is best documented in Pluto Adv.Col. (Arcesilaus and Middle 
Academy); Cic. Acad. 2 (Carneades); Sext. Emp. Math. 7, 9 (Pyrrhonists); for the 
intervention of Antiochus, see J. Glucker, Antiochus and the Late Academy 
(=Hypomnemata 56 [G6ttingen 1978]), with D. N. Sedley, "'The End of the 
Academy," Phronesis 26 (1981) 67-75. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

192 THE ATHEISM OF EPICURUS 

of certain knowledge can be applied to issues of theology and re­
ligion from both an anthropological (or cultural-historical) and a 
purely theoretical point of view. 11 Although they arrive at theologi­
cal positions widely divergent in detail, both Stoics and Epicureans, 
by deploying an array of sophisticated interpretive techniques, thus 
claim to have adequately explained, and therefore to have mar­
shalled the support of, the mass of commonly held beliefs about the 
gods against the erratic o6~at of deviant thinking (including, of 
course, the 'views' of their skeptical opponents).12 In regard to 
theology and religion, this position allows Stoics and Epicureans to 
accept both the existence of gods, as a necessary point of doctrine, 
and the conceptual validity of (or, in the case of the Stoics, tolerance 
for) popular religious customs and practices. 

At some point in the early history of the debate over epis­
temology, partly in response to the dogmatists' appeal to the sensus 
communis, it was pointed out 13 that an impressive body of dis­
tinguished thinkers had in fact cast doubt (either explicitly or, more 
often, implicitly) upon the validity of sense impressions and com­
mon beliefs, and upon the possibility of certain knowledge. 14 This 
posed considerable difficulty for the dogmatists, who were now 

11 M. Schofield, "Preconception, Argument, and God," in Schofield (supra n.l0) 
283-308; P. A. Brunt, "Philosophy and Religion in the Late Republic," in M. 
Griffin et aL, edd., Philosophia Togata: Essays on Philosophy and Roman Society 
(Oxford 1989) 97-119. 

12 So also Philodemus' use of the early technical term (JUVot1CtiO)(H~ in De Pie tate 
(P.Hercul. 1428 VII-IX) to criticize Chrysippus' use of allegory; A. Henrichs, "Phi­
lodems De Pietate als mythographische QueUe," CronErcol3 (1975) 5-38 at 16-18; 
F. Buffiere, Les mythes d'Homere et La pensee grecque 2 (Paris 1973) Index S.'V. "Phi­
lodeme." For Stoics: M. Schofield, "The Syllogisms of Zeno of Citium," Phronesis 
28 (1983) 31-58. 

13 By Arcesilaus, who of course wrote nothing. Our best evidence is therefore the 
contemporary attack on Arcesilaus by Epicurus' pupil Colotes (as preserved by 
Plutarch, who three centuries later still felt compelled to refute it). Colotes counters 
Arcesilaus (whom he never in fact mentions) by arguing against certain figures 
adduced by Arcesilaus as authorities-though these figures are not the only ones 
attacked by Colotes: see P. A. Vander Waerdt, "Colotes and the Epicurean Refu­
tation of Skepticism" (225-67 infra). 

14 See Cic. Acad. 2.14,72-76 (cf. 1.44-46); Pluto Mor. 1108B, 1121F-22A; Sedley 
(1983 [supra n.l0]) 15f with n.27; P. De Lacy and B. Einarson, in the Loeb edition 
of Plutarch's Moralia XIV (London 1967) 155f, 165; P. DeLacy, "Colotes' First 
Criticism of Democritus," in J. Mau and G. Schmidt, edd., Isonomia. Studien zur 
Gleichheits'Vorstellung im griechischen Denken (Berlin 1964) 67-86; Vander Waerdt 
(supra n.13) 247-53. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 193 

faced with the unhappy prospect of defending KOtVal. EVVOtat (e.g. 
about the gods) without the support of the consensus omnium. In 
response to challenges from skeptical quarters over the x:Pt'tllPtOV 
't1l<; aATl9da<;, the dogmatists counter by intensifying their dog­
matism. The Epicureans stress their claim for validity of sense per­
ception to the point of universality (Epicurus' infamous dictum "all 
perceptions are true," D.L. 10.32, 146f; Sext. Emp. Math. 7.201-10; 
Pluto Mor. 1109-10) against the introduction by Stoics of a special 
kind of unfalsifiable perception (the <pClv'tClata x:a'tClATl7t'ttx:i1).15 Epi­
cureans further develop an argument according to which any 
thinker's wholesale dismissal of ordinary perception and common­
ly held views is, if not in fact impious, on a practical level quite 
perilous because, if taken to undermine all confidence in physical 
reality, it would make it impossible not only to engage in rational 
debate and decisive action but even to live life as we know it. 16 

If not entirely valid, this self-refutation argument has a certain 
rhetorical persuasiveness. 17 It also drew into the debate the cultural 

15 G. Striker, "Epicurus on the Truth of Sense Impressions," AGPh 59 (1977) 125-
42; C. C. W. Taylor, .. All perceptions are true," in Schofield (supra n.10) 105-24; F. 
H. Sandbach, "Phantasia Kataleptike," in A. A. Long, ed., Problems in Stoicism 
(London 1971) 9-21; M. Frede, "Stoics and Skeptics on Clear and Distinct Im­
pressions," in Burnyeat (supra n.10) 65-93. 

16 The argument from 'livability' appears fully developed already in Colotes' at­
tack on Arcesilaus, and has a long history in Hellenistic philosophy. Stoics use it 
against the skeptical Academy under the name a1tpa~{a (Plut. Mor. 1122A; Cic. 
Acad. 2.25; D.L. 7.171, 9.107), or alternatively named the fOpyCilv (with a double 
entrendre characteristic of Stoic names for logical puzzles) since the argument is 
said to "'turn one's opponents to stone": Cic. Ad Jam. 9.8.1; Arr. Epict.Diss. 1.5.1-3. 
Pyrrhonists refer to it as avtvtpY11o{a (Sext. Emp. Math. 9.162) and avtvipY11'to~ 
(Pyr. 1.23f, 226; Math. 7.30). See also Striker (supra n.l0) 63 with n.25; M. F. 
Burnyeat, "'Can the Sceptic Live his Scepticism?" in Schofield (supra n.lO) 20-53 
(=The Sceptical Tradition [supra n.l0] 117-48). 

17 M. F. Burnyeat, "Protagoras and Self-refutation in Later Greek Philosophy," 
PhilRev 85 (1976) 44-69, esp. 62ff, on the earliest form, which occurs in Epicurus; 
"The Upside-down Back-to-front Sceptic of Lucretius IV 472," Philologus 122 
(1978) 197-206; D. N. Sedley, "'Epicurus' Refutation of Determinism," in I:'\)~ft't'IlOt~: 
Studi suii' Epicureismo greco e romano oJJerti a Marcello Gigante I (Naples 1983) 
1-51. On the problems with the validity of the self-refutation argument, see 
Burnyeat, who stresses the argument's rhetorical appeal (cf however Vander 
Waerdt [supra n.13] 242-45). The argument strikes many as a crude sophism. It 
must be remembered that philosophy in antiquity was conducted on a largely oral 
and agonistic basis; the ability to silence one's opponent in the public forum 
counted most. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

194 THE ATHEISM OF EPICURUS 

stigma that in the popular mind attached to a denial of the es­
tablished order of things. 18 Thus the dogmatists characterized the 
skeptical position as tantamount to suicidal destruction of both 
individual and society. The argument is perhaps best known from 
Lucretius (4.472ff), but we have earlier examples of its use against 
Academic skepticism. Epicurus' pupil Colotes, for instance, argues 
that a "universal suspension of judgment'" (£1t0Xll1t£pi 1tav't<Ov ), if 
consistently adopted, would result in the disappearance of the in­
stitutions characteristic of civilization and reduce the life of men to 
that of beasts.19 Perhaps the most striking illustration of this stage of 
the debate is Epicurus' claim (preserved by Diogenes of Oenoanda) 
that Protagoras, in saying that he knew not whether there were 
gods, meant exactly the same thing ('to au'to) as Diagoras, who said 
there were none at all. 20 Epicurus' often ridiculed conclusion about 
the 'atheism' of Protagoras (who never said OUK elvat 8£0{)<;) turns 
out to be true from the Epicureans' point of view: for Protagoras, 
like Diagoras, had precipitously destroyed the traditional founda­
tion for the beliefs upon which they based knowledge of the gods' 
existence and nature. 

III. Epicurus and the Gods 

To judge from his extant writings, Epicurus had no doubts about 
the necessity of giving an account of the gods. He seems in fact to 

18 On the motif of philosophical polemic see esp. H. Reiche, -Myth and Magic 
in Cosmological Polemics," RhM 114 (1971) 296--329, esp. 304ff. 

19 Pluto Mor. 1124D. Part of the argument is that civilization could not have de­
veloped as it did if sense perception were not reliable and basic beliefs true: the 
very existence of cities and laws, like the demonstrated ability of even the skeptic 
to live, entails that the skeptic's position is a false one. In this form the argument 
appears prominently in De pietate (P.Hercul. 1428 xIV.24-xv.13) p.25 Henrichs, but 
deployed against Stoic theology (or against Academic appeals to Stoic theology 
against Epicureans) and in the account of the origin of civilization at Philod. De 
piet. (P.Hercul. 229 I-VII) pp.142-48 Gomperz. 

20 Diog. Oen. frr.11-12 Chilton: £CPl1O'£ (sc. Protagoras) yap Ilit dB£val d 8£01 
dO'w. 'tOU'to B' EO''t1.V 'to au'to 'tii> A.iy£w dB£val on Ilit dO'w, which, as it follows in 
Diogenes' treatise directly upon Epicurus' account of the origin of language from 
n£p1. cpuO'£COC; 12, almost certainly derives from the discussion of cultural history in 
that book; cf. Philod. De piet. p.112.8 G. m[U.O\c;] (see the new text 216£ infra). Cf. 
C. W. Chilton, -An Epicurean View of Protagoras," Phronesis 7 (1962) 105-09; P. 
A. Vander Waerdt, -Hermarchus and the Epicurean Genealogy of Morals," 
TAPA 118 (1988) 87-106, who cites (89,91 n.20) the relevant texts from Epicurus' 
n£p1. cpUO'£coc; 12. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 195 

have placed it first in his system: 21 the gods are the subject22 of the 
first of the Kyriai doxai and stand first in the letter Ad Menoeceum, 
where Epicurus asserts that the gods exist, that they do so as su­
premely blessed, imperishable, yet living beings, and that ordinary 
knowledge of them, being derived from primary conceptions 
( '1 ' ) b 1 'd e" , '" \ n:p0/\,'Tl'\jlnc; , must e c ear or eVl ent: £Ot ~£V ya.p £talV' £va.p'Y'1~ 

yap au'twv fattV it yvrocrtt;. Norden cited this passage for its early use 
of yv&att; in a theological context, as implying a cognitive relation­
ship between man and the gods. 23 On the other hand, fvaprTtt; is an 
Epicurean catchword implying that such primary conceptions have 
a sound causal origin in perception. 24 Epicurus' assertion was no 
doubt intended to be provocative, much like the notorious claim 
that "all perceptions are true," which it partly recalls. 25 Epicurus 
also states explicitly that the gods are anthropomorphic (L ad KD 1), 
perfectly blessed and immortal (Ep. ad Men. 123, 'tOY ~4>ov acp8ap­
'tOY Kat ~aKaptov), and that they do not participate in the affairs of 
our world or the working of the cosmos (Ep. ad Her. 76-78), such 
activities being incompatible with their blessedness and imperish­
ability.26 Our conceptions result from a constant stream of images 
-for Epicurus the only immediate objects of perception-espe-

21 For the same reasons the Stoics placed it last. They argued from their ety­
mology of 'to .. £tO~. 'tEAE't", and the like (as allusions to the arcana philosophiae) 
that theology should come last (after physics, logic, and ethics) in philosophical 
inquiry: Pluto Mor. 1035B, 718B, 382D-E; Etym. Magn. S.'o. 'tEAE't" (SVF II 1008, cf I 
538); P. Boyance, Le cult des Muses (Paris 1937) 42f, 49 n.2; C. A. Lobeck, Aglao­
phamus I (Konigsberg 1829) 123ff. 

22 So also Lucr. 1.62ff, Philod. De dis 1, the first colon of the Epicurean four-fold 
remedy ('tE'tpa<pap~a1Co~), and first in Diogenes of Oenoanda's list of the causes of 
all evil (fr.28 col. VII Chilton). 

23 E. Norden, Agnostos Theos (Leipzig/Berlin 1913) 87ff, esp. 96£. Henrichs 
(1975 [supra n.5]) 109 compares P.HercuL1428 XVl.7-9 (Democritus 68A75 D.-K.): 'to 
E;EP'Ya~6~EvoV yv6v'ta~ (sc. 'tou~ 1tpro'tou~ avepro1tOu~) (Ji~Eoem. 

24 Cf Democritus' view that the gods exist as exceptionally large, long-lived, cir­
cumambient images, endowed with speech and prophetic powers, and capable of 
conferring benefit or harm (68A74, 77-79; B142, 166). Epicurean criticism of Democ­
ritus' theory at Diog. Oen. new fr.1.2.7-3.14 (15G Long-Sedley [nA8 infra]): M. F. 
Smith, AlA 74 (1970) 51-62; CQ N.S. 22 (1972) 159-62; D. Clay, "An Epicurean In­
terpretation of Dreams," AlP 101 (1980) 342-65. 

25 Cf supra 193. 
26 Epicurus and his followers employ 'gods' (plural) and 'the divine' (sing.) in­

discriminately as the occasion suits, as well as a range of substantives ('to ~a1Captov. 
'to (JE~v6v. 'to om~6vwv)-and a neologism, 'to (JE~vO)~a (Epic. Ep. ad Her. 77, 
P.Oxy. II 215 1.30). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

196 THE ATHEISM OF EPICURUS 

cially fine in structure, like the atoms of the soul and so perceptible 
only by thought (AOyq> 9£roPll't0i), that are naturally imprinted 
(especially during sleep) on the minds of all men. Naturally we do 
not 'see' the gods as we do material entities, but we can discern 
their characteristics, as we do those of atoms, on the analogy of or­
dinary sense-impressions. 

According to a fragment of Philodemus' De pietate ,27 Epicurus 
gave a similar account of the origin of belief in the gods in Book 12 
of his magnum opus TIept cpuo£c.oc;; (a source for DRN 5), this time 
from a cultural-historical perspective. The first men (xpiihot liv9pro-
7tOt), he says, having received images of the gods in dreams, recog­
nized that they exist and so formed an understanding28 of their 
divine nature (N 1077 xlx.23-29 [HV2 II 83; p.113 G.]):29 

KaV 'ton oro-
24 o£Ka't[roh 7t£Pt cp[u]­

o[£]ro[<; 'to]u<; 7tPro'tou<; 
CPll [OtV a ]y~ p.Ql7tou<; 
£t<; v[ ofl]~(}'t(l ('tow) [E]~ro 

28 J3a(t)vetv acp~ap'trov , 
cpu(Jerov. 

And in Book 12 of On Nature 
he says that the first men ar­
rived at conceptions of imper­
ishable external entities. 

27 Excerpted from the forthcoming edition (supra n.4). The texts amalgamate the 
readings of the Oxford and Naples copies (0 and N) with the supplements of all 
editors and commentators. For a facsimile of the Naples copy, see Herculanensium 
'()oluminum quae supersunt, Collectio altera II (Naples 1863) (= HV2 II). Sublinear 
asterisks indicate editorial corrections of putative errors in the nineteenth-century 
copy; for editorial conventions see Henrichs (supra n.4 [1972]) 72f n.15. 

28 For the expression ti~ vOTu-unu ~uivEtv (27f), cf. Philod. De piet. p.106.13-15 
G. E.[i]~ htivo\Uv uu'til~ (sc. cpUOE.CO~ 'toU 9dou) 13u~H~ov'tu; Epic. De nat. 31.21.16£ 
Arr.2 ou 13uo1.E.i['t]m btl. 'troll \j1EuOrov; Plat. Leg. 793B 13uivElV E1CtO~ 'tou KUAoU; Resp. 
380D EK13uivEtv 'til~ tuu'tou iOEU~. For cp{)OEt~ in the sense of 'existing things', as 
here, see Epic. De nat. 24.48.7, 17; 24.49.4, 8,27 Arr.2; Ep. ad Her. 40,48, 68, 71. 

29 According to another portion of this account, P.Hercul. 1098 Xlx.11-17 (De 
piet. p.133 G.), this knowledge was attained by means of corporeal sensation: 'tTtV 
1tUPUta9ftOEt OUpK[ i]Vl1t 1tE.PtAl11t'tTtV u to[~]OlV. This fragmentary account should 
be supplemented by Sext. Emp. Math. 9.25 (Epic. fr.356 Us.), which attests that 
according to Epicurus people originally derived the conception of god from pre­
sentations received in sleep (EK 'trov Ku'ta 'tou~ U1tVOU~ cpuv'tUOtrov olum 'tou~ a v -
9pOmou~ EVVOtUV t01tUKEVm 9EOU), and quotes Epicurus' explanation: "'For when 
great images of human shape impressed them during sleep, they supposed that 
some such gods of human shape really existed" (IlEYUAcoV yap tiowAcoV, CPl1oi. KUt 
av9pC01t0IlOPCPCOV Ku'ta 'tou~ U1tVOU~ 1tPOO1tt1t'tov'tcov {)1tEl •. u13ov KUt 'tui~ aA119E.iat~ 
{)1tUPXElV 'ttva~ 'tOtOO'tQ'\)~ 9EoU~ av9pC01toj.L6pcpou~), and 9.43 (both standard accounts 
of concept formation); and Lucr. 5.1168ff. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 197 

23-29 Epic. fr.84 Us., 27.1 Arr.2 23 Kay Biicheler: KPN ~ 24-26 
Biicheler 27 Obbink: Ere (vel EI:I) N[ ... ]MHTA N: E~!v[011]~'i''tct 
Usener 'tIDV add. Obbink 27f Obbink: [.]E!lIBANEIN N: ActHI~a­
VEW emendavit Usener 28 corr. Biicheler: A«I>OPT!lN N p"'o~t nov 
spat. vac. N 

Presumably at the beginning of civilization, men conceived the gods 
as anthropomorphic, blessed, imperishable, and so on; later, as in 
the nearly contemporary anthropological theories of Theophrastus 
and Dicaearchus, this conception of the gods underwent a process 
of devolution, becoming contaminated with false notions. But in 
guaranteeing a correct 1tp6A:rl"'t~ of the gods to the earliest men,30 
Epicurus effectively denies the pre-anthropomorphic stage of 
thinking about the gods that Theophrastus assigns to early man. 31 

Among false notions about the gods introduced at an early stage 
of civilization (according to the Epicurean theory of cultural his­
tory) is that they intervene in this world to reward the good, punish 
the bad, and so on.32 Epicurus states emphatically that the gods, 
being blessed and imperishable, could not conceivably reside in this 
world, where they could be neither entirely free from care nor 
immune to earthly forces of destruction. This is the main point of 
Lucretius' planned digression at 5.146f (illud item non est ut possis 
credere, sedis esse deum sanctas in mundi partibus ullis), while his 
description of the sedes quietae of the Epicurean gods at 3.18-23 (a 
rendering of the Homeric picture of a cloudless, radiant Olympus 
at Od. 6.42-46) is intended to demonstrate that mankind has always 
conceptualized the gods in much the same way, or to underscore 
poetically how deeply rooted is such a conception of the gods (as 
alien to our world) in traditional thinking-rather than to provide an 
account in dogmatic terms of their actual home. 33 Some later 

30 Or something approaching it (line 27 V0TtIlU'tu); cf. Cic. Nat.D. 1.43 antici­
pationem quandam deorum, and K. Kleve, Gnosis theon (=SymbOslo Suppl. 19 
[Oslo 1963]) 99 with n.6. 

31 Theophr. De piet., ap. Porph. Abst. 2.5; cf. Reiche (supra n.18) 331 n.SO. So also 
Epicurus' account corrects Democritus' anthropological version (68A75) of the 
origin of religious belief. 

32 Philod. De piet. pp.142-48 G. (P.HercuL 229 I-VII); Lucr. 5.1161-1240. 
33 Cf. the belief in a Homeric Olympus rationalized at Arist. Cael. 284a, and com­

pare the extra-cosmic abodes of the gods in Ar. Pax (esp. 207-09) and Aves: Z. A. 
Pokrovskaja,"Les dieux d'Epicure et Teree de Thrace," Eirene 19 (1982) 57-70. 
Later reflections: Hermocles' ithyphallic hymn (on the entry of Demetrius Po­
liorcetes into Athens in 294) ap. Ath. 6.253D-F (Powell, Coil. A lex. 173-75); Lucian 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

198 THE ATHEISM OF EPICURUS 

authors did of course occupy themselves in speculating where, in 
the Epicurean universe, such gods could exist.34 But if Epicurus 
himself ever answered the question "where do the gods live?" (and 
there is no reason to think that he did, or should have had to), he is 
likely to have said that they do not live in this world. 35 Epicurus 
chose rather to stress the persistence and regularity of individual 
conceptions of the divine, and the effect (primarily social and 
psychological) of these conceptions upon life in this world. 

Naturally Epicurus does not consider every conception of a god 

lupp.Trag. 45, probably modelled on a (Varronian) Menippean (R. Helm, Lucian 
und Menipp [Leipzig 1906] 149 n.3). 

34 Principally the intermundia (J.1E'talCooJ.ua, olao'tTtJ.1a'ta), the interstices between 
worlds in the Epicurean universe: so Cic. Nat.D. 1.18, Div. 2.40 (cf. Fin. 2.75). For a 
list of references-none earlier than Cicero and most (Seneca, Quintilian, Plutarch, 
Augustine) almost certainly derived from him-see Pease (supra n. 3) on both pas­
sages, and F. Peters, T. Lucretius et M. Cicero quo modo vocabula Graeca Epicuri 
disciplinae propria Latine verterint (diss.Miinster 1926), who includes Christian 
authors. This view was not known to the Epicurean sources used by Cicero in 
Nat.D.; thus it does not occur in Velleius' exposition in book one, where we might 
have expected it, though Cicero adds disparging references to it in his prefaces, 
perhaps in a willful misinterpretation of Lucr. 3.18-25, which he will by this time 
have known; cf esp. Dionysius episcopus nEpi CPUO'ECIl<; ap. Eus. Praep.Evang. 
14.27.8 p.782c (=Epic. fr.364 Us. in part). On the 'tTtV J.1E'talCoo[J.1lOV xropav] of 
Philod. De morte 4 col. xv.8 (p.146 Kuiper), see Philod. De dis p.24 Diels: "'Hier ist 
aber das platonische Elysium, nicht der epikureische Gotterwohnsitz gemeint." I 
take De dis 3 vIII.20-IX.27 (p.26) to be an account of how the random passage 
through the cosmos of dOCll).,a which figure in the process of conceptualizing the 
gods accidentally gives rise to the belief in catasterism and the cult of apotheosized 
rulers-and not an account of images emanating from the gods' permanent places 
of existing. Contrary to the constant assertion of commentators, Lucretius in his re­
marks on the gods does not mention or allude to the intermundia. 

35 According to an important and relatively early source-Atticus, who had 
reason to be informed on the subject-Epicurus simply locates the gods somewhere 
outside our world (ap. Eus. Praep.Evang. 15.5 p.800B [Epic. fr.362 Us.]: £~CIl 1tOU 'tou 
lCOO'J.10U lCa9{opuO'E), which is consistent with their being in the intermundia, but is 
equally likely to be an extension of an original statement to the effect that they 
cannot exist, at least as conceived by many people, anywhere in this world. Atticus 
said nothing about the J.1E'talCoO'J.1la, and it is clear from the only other source with 
a pretence to a doxographical pedigree (Hippolytus' Cj)lA.ocrocpoUJ.1EVa) that the specifi­
cation of the J.1E'talCOOJ.1ta was an inference from the belief that Epicurus locates the 
OilCll'tftPWV of the gods £~CIl 'tou lCOOJ.1OU (Diels, Dox.Graec. 572). Moreover, both 
Pluto Mor. 731 D and Philo De somn. 1.184 place the Epicurean gods either in the 
J.1E'talCOOJ.1ta or in some other world, which suggests an elaboration of an original 
£~co 1tOU 'tau lCOOJ.1OU, or OUOOJ.1Ou tV 'tCp (itJ.1&v) lCOOJ.1cp. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 199 

to be true. The gods represented by theologians and poets, together 
with such popular conceptions as engender fear, are based not on 
1tpoA:ll'1'EtC; but on false notions due to the contamination of original­
ly clear concepts. Thus Epicurus and his followers often imply an 
obligation to preserve (qmAcl't'tEtV) the integrity of the gods by en­
suring that our concepts of them are not impaired. In Ep. ad Men., 
for example, Epicurus argues that the conception of the gods held 
by many is defective; this requires (a) showing that it conflicts with 
another more fundamental (pre)conception, that of the gods as 
perfectly blessed,36 and (b) explaining how the faulty conception 
arose. Similarly a failure to preserve our conception of the gods un­
impaired effectively "abolishes" them (UVCltPEtV, tol/ere) by distort­
ing the experientially based preconceptions that guarantee genuine 
knowledge. 

Epicurus' emphasis on the role that we play in this process cannot 
be too strongly emphasized. If it was an innovation of fifth-century 
sophistic thinking that man could, by his own thought and action, 
divorce himself from the gods, it was equally a novelty of Hellenis­
tic theologies that man could by the same process align himself with 
divinity, whether or not that divinity is conceived (as e.g. in Plato) 
as exercising care for individuals. In this respect then, Epicurus 
seems to have substituted for the causal and governing roles ac­
corded to the divine in earlier Greek philosophy the notion, already 
familiar from Plato and Aristotle, of the divine as a normative object 
of moral emulation (oJloiroc:nC; eEii». Since for Epicurus the gods 
share with men pleasure as an ethical telos, and since our concep­
tions of them embody the Epicurean ideals of blessedness and 
tranquillity, the gods stand in relation to the wise virtually as 
paradigms of moral excellence. This accounts for Epicurus' in­
sistence that we conceive of the gods as anthropomorphic and 
capable of speech.J7 While the intervention of the gods in the world 
is ruled out by its incompatibility with their own blessedness, they 
do nevertheless have a real, if indirect, influence on the human 
world, inasmuch as one's own view of the gods has a major impact 

36 123f (on the constitution of the text, see most recently Vander Waerdt [supra 
n.20] 101 n.50); Ep. ad Her. 7M; Pluto Mor. 1051E, 1075E. 

37 So also Epicurus is credited with having held (following the common Hel­
lenistic trend of deification of great men) that the individual sage could achieve a 
kind of divinity: see esp. D. Clay, "The Cults of Epicurus,» CronErcol16 (1986) 
11-28 and infra n.45. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

200 THE ATHEISM OF EPICURUS 

for better or worse on one's own oui9EO'tC;, or psychosomatic 
constitution. 38 Naturally a 'true' conception of the gods is easily 
distorted by all sorts of false beliefs (e.g. that they exercise power, 
malevolence, etc.-the consequence of a literal reading of the tales 
of poets). But the wise, who preserve a correct conception of the 
gods, derive a sense of immense calm and religious awe from 
perceiving and imitating their nature. For Epicureans, the restric­
tion of divine attributes to those compatible with blessedness and 
imperishability is less a devaluation of traditional forms of piety than 
a source of a deeper psychological dimension of religious ritual. 39 
Thus traditional forms of worship are regarded as a natural response 
to the recognition of the divine nature, and are not only tolerated 
by Epicurus but recommended to his followers. Numerous acts of 
worship by Epicurus and individual Epicureans are attested, in­
cluding sacrifice,40 adoration of statues,41 dedications,42 mystery 
initiation,43 and participation in calendrical festivals 44 and rites of 

38 See esp. W. Schmid, "Gotter und Menschen in der Theologie Epikurs," RhM 
94 (1951) 97-156; A.-J. Festugiere, Epicure et ses dieux 2 (Paris 1968) 36-100. 

39 Cf Theophr. De piet., ap. Porph. Abst. 2.19; R. Parker, Miasma: Pollution and 
Purification in Early Greek Religion (Oxford 1983) 322-27; W. Burkert, Greek Re­
ligion, tr. J. Raffan (Cambridge [Mass.] 1985) 77; P.Oxy. II 215 n.2f'ttJl[&]v au'tl)v 
'tl)v OErop{av, with Burkert 311; D. Obbink, "PO xy. 215 and Epicurean Religious 
OEropla," in Aui del XVII congresso internazionale di papirologia II (Naples 1984) 
607-19. 

40 For offerings and blood sacrifice: Philod. De piet. pp.126.9f, 127.14, 146.6 G.; 
Ath. 179D (Epic. fr.S6 Us.); Pluto Mor. 1102B (Epic. fr.30 Us.); Porph. Abst. 1.7-12 
(an account and defense of the pre-historical institution of animal sacrifice)=Her­
marchus fr.24 Krohn=fr.34 in F. Longo Auricchio, Ermarco [rammenti (=La scuola 
di Epicuro 6 [Naples 1987]); cf Vander Waerdt (supra n.20); D. Obbink, CQ N.S. 

38 (1988) 428-35. 
41 Adoration of cult statutes (ayaAJla'ta): Philod. De piet. pp.l03.13ff, 124.21£f 

G.; Cic. Nat.D. 1.85; Origen c.Cels. 7.66 (Epic. fr.390 Us.); so also use of statues and 
amulets as talismans: B. Frischer, The Sculpted Word: Epicureanism and Philo­
sophical Recruitment in Ancient Greece (Berkeley 1982) 87-128. 

42 Dedication of avaOftJla'ta: D.L. 10.126; A. E. Raubitschek, "Phaedrus and his 
Roman Pupils," Hesperia 18 (1949) 96-103 (dedicatory herms to the gods of Eleu­
sis); C. J. Castner, "Epicurean Hetairai as Dedicants to Healing Deities," GRRS 23 
(1982) 51-57 (to Asclepius at Epidaurus). 

43 Initiation in the Attic (i.e., Eleusinian) mysteries: Philod. De piet. p.127.24ff G. 
44 The Attic festival of the Pitchers ('trov Xorov) on the second day of Anthesteria: 

Philod. De piet. p.127.24ff G.; c{. P.HercuL 1077 xI.25-28 (p.l05 G.) with Usener's 
restorations (=Epic. fr.lS7 Us., (r.86 Arr.2). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 201 

private and ancestral cult. 45 Their opponents, considering such 
practices incompatible with a rejection of natural teleology, divine 
providence, and divination, viewed them as insincere parodies 
designed to cultivate popular favor. 46 But they were intended rather 
to illustrate the Epicurean theory of religion and social cohesion, 
and the degree to which cultural phenomena (including false be­
liefs) can be accounted for on atomist principles, without recourse 
to the teleology of Stoics and Peripatetics: for Epicurus, like Prodi­
cus and Democritus, viewed cult as a natural outgrowth of cultural 
history. Similarly, we find Epicureans, in an attempt to rationalize 
and thereby vindicate popular belief, maintaining the proposition 
that gods are actually capable of doing men harm (i.e., the wicked, 
as a result of their own depraved conceptions of the gods).47 

While the gods exist for Epicurus fundamentally as concepts, as 
has been argued recently,48 and their effects stem primarily from 
our concepts of them, it may nevertheless be missing the point to 
insist that Epicurus' gods exist as 'mere' concepts. 49 According to 

45 For religious feasting (Ei>OlXia) especially in the context of private or ancestral 
cult, see in particular Philod. De piet. p.104 G. and the five separate calendrical rites 
stipulated by Epicurus in his will (D.L. 10.18f£), which provide for the continuing 
funeral cult ('tu £va'Yicr~a'ta) of his parents and brothers (cf De piet. p.18.20ff G.), 
the celebration of his own birthday on 10 Gamelion (together with similar rites 
for his brothers and for Polyaenus), and most importantly, the gatherings on the 
twentieth of each month ('tai.~ tl1Cacrt) to commemorate himself and his favorite 
pupil Metrodorus. For the traditionally religious character of these rites of private 
cult, see esp. D. Clay, "Individual and Community in the First Generation of the 
Epicurean School," in l:'\)~tl'tl1(Jl.~ I (supra n.17) 255-79 at 272ff and (supra n.37) 
11-28. 

46 On Epicurus' rejection of divination, see D.L. 10.135, with C. Diano, StIt 12 
(1935) 237ff. 

47 E.g. Philod. De piet. (P.Hercul. 1428 XII) pp.22f Henrichs. 
48 A. A. Long and D. N. Sedley, The Hellenistic Philosophers I (Cambridge 

1987) 144-49, an account to which I am much indebted in this section, as against 
the traditional interpretation: cf G. Schomann, Schediasma de Epicuri theologia 
IV (Gryphiswald 1864) 346; F. A. Lange, Geschichte des Materialismus F (Iserlohn 
1873) 76£; J. Bollack, La pensee du plaisir (Paris 1975) 236-38; D. N. Sedley, CR 
NS. 29 (1979) 82-84; A. A. Long, .. Epicureans and Stoics," in Classical M editer­
ranean Spirituality, ed. A. H. Armstrong (New York 1986) 135-53 at 142-45. 

49 As did many of Epicurus' ancient detractors (e.g. Cotta ap. Cic. Nat.D. 1.77: 
Epicurus' gods are no more than images). The natural, if mistaken, conclusion is 
that Epicurus 'didn't believe the gods exist', on the erroneous assumption that 
gods could not sufficiently or 'really exist' as 'mere ideas'. In the Epicurean view, 
concepts in fact have the same corporeal, atomic composition as all l>1tapxov'ta. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

202 THE A THEISM OF EPICURUS 

Epicurus, our concepts of the gods have a unique ontological status: 
our only access to the divine is through such concepts, just as our 
only access to solid objects is through the impressions they present 
to us. 50 Epicurus believes that these impressions are always true, 
insofar as they always provide reliable information: our "preconcep­
tions" of the gods are "a genuine piece of moral knowledge, an accu­
rate intuition of man's natural good."51 Epicurus' carefully orches­
trated theory of the way in which human cultures conceptualize the 
gods represents an attempt to ensure the physical possibility of the 
conditions necessary to produce these ideas. 

IV. The Atheism of Epicurus 

By the beginning of the first century B.C., when works like 
Philodemus' De pietate and Cicero's De natura deorum drew the 
attention of an educated public, we find that Epicurus, who had 
been carefully dogmatic in his own pronouncements about the 
gods, was being paraded in certain circles as an atheist. In later an­
tiquity the view that Epicurus was in fact a 'closet' atheist, whose 
statements about the gods were only intended to placate a hostile 
public, became in fact the communis opinio. 52 We know, however, 
that at the beginning of the first century B.C. it was still a sufficiently 
controversial issue that entire books were devoted to the subject. 

It might be supposed that the charge of atheism against Epicurus 
originated with an actual indictment of Epicurus (or even the threat 
of one) on charges of impiety (clcsEJ3Eta), stemming from social or 
political hostility to his theology. If so, we might reasonably expect 
that such charges would have been utilized (or even fabricated) by 
the well-known early anti-Epicurean tradition, perhaps e.g. by the 

50 According to Epicurean theory, the eternally regular pattern of cognition of 
the gods by mankind is facilitated by streams of 'similar' images (e.g. I ad K.D. 1). 
Our conceptions of them are thus said to constitute (or result from) a unique class 
of imperishable existents called in technical accounts 'similarities' (OJ.1OtOtrt'tEC;: 
Aetius P/ac. 1.7, in Diels, Dox.Graec. 307) or 'unities' (tv6'tTt'tE~: Epic. De san ct. fro 
41 Us. [fr.19A Arr.2] ap. Philod. De piet. p.ll0A-19 G.) because, unique among 
existing things, they are always formed in the same way. Thus they are dis­
tinguished from other!urely mental existents (e.g. the idea of a satyr, a cyclops, or 
an eternally long-live man) which are imaginatively or fortuitously formed by 
combination and reduction or expansion of images: see Sext. Emp. Math. 9043-47. 

51 Long-Sedley (supra nA8) 147. 
52 See e.g. Winiarczyk (supra n.1) 157-83 at 168-79 (on Epicurus and his fol­

lowers). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 203 

dissident Epicurean Timocrates, brother of Metrodorus. But while 
our sources for this anti-Epicurean tradition are given to exag­
gerated slander in other respects, they say nothing of atheism or 
impiety. 53 Philodemus, for his part, takes great pride in pointing out 
that Epicurus' detractors could cite no instance in which Epicurus 
had been satirized in comedy, a genre which, as says, was only too 
given to harassment of philosophers for their objectionable or 
outlandish views (0 1077 I center [missing in N, HV2 II; p.93 G.]): 

to those who were so diverse in their chosen styles of life as well 
(Epicurus was) inoffensive, so that not only did he not engage in any 
lawsuit against anyone, he did not even have a quarrel. Indeed, while 
some philosophers were prosecuted for their way of life and for their 
teachings, and some have been exiled from city and league and put to 
death, and all have become the butt of writers of comedy, only Epi­
curus grandly secured protection for himself together with those 
who dwelt sincerely with him, without falling prey to the virtue­
hating and all-harassing mouth of comedy. 

Philodemus' claims are supported by the record, at least insofar as 
Epicurus' own life is concerned (the passage may in fact intend a 
contrast with Socrates in this respect).54 'Epicureans' do appear oc­
casionally in Hellenistic comedy (often, for example, as cooks­
l.Hl'Y£tpot-pandering to the delicacies of taste), but in no surviving 
instance are they satirized explicitly for their theological views. 55 

So also neither the remains of Epicurus' own work nor that of his 

53 See D.L. 10.4-9, a list of 01. &uO'/ltvro~ txovtt~ 7tpO~ 'E7tl1wupov. A possible ex­
ception is that Diogenes includes in his response to this tradition (10.10) the 
defense: tile; /ltV ya.p 7tpOe; etoile; oO'1.o'tTl'toe; KUt 1tpOe; 1tu'tpiou qltA.tUe; aAtK'tOe; it 
OUletO'le; (sc . tOU 'E1ttKOl>pOU), though this might equally be taken as the language of 
conventional encomium. On Timocrates' distortion of Epicurus' views, see D. N. 
Sedley, "Epicurus and his Professional Rivals," in Etudes sur I'Epicurisme antique 
(=Cahiers de Philologie III.l [Lille 1976]) 121-59. 

54 Cf n.59 infra for attestations of social and political harassment; all are non­
Attic and postdate the lifetime of Epicurus. 

55 E.g. Damoxenus' I:UVtpo<p01. ap. Ath. 3.101F-I03B=fr.2 in PCG V; Athenion's 
I:U/loe~1Ct~ ap. Ath. 14.660E--661o=fr.l in PCG IV; Menander fr.750 Korte; E. 
Bignone, "Epicuro e i comici greci, Teocrito e Ateneo," A&R SER. VI (1938) 75-92 
(=L'Aristotele perduto II [Florence 1936] 223-47); K. Buchner, "Epikur bei Menan­
der," StIt 14 (1937) 155-66; M. Pohlenz, "Menander und Epikur," Hermes 78 
(1943) 27-75; A. Barigazzi, "Uomini e dei in Epicuro," Acme 8 (1955) 37-55 at 52ff; 
F. Wehrli, "Menander und die Philosophie," in Menandre (=Entr.Hardt 16 [Gene­
va 1969]) 147-69, with further literature cited there. It is unclear whether the Epi­
curean /layupo1. are being satirized for their attitude towards sacrifice and religious 
feasting or towards pleasure in general (the two are not mutually exclusive). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

204 THE ATHEISM OF EPICURUS 

earliest followers contain any implication of, or response to, charges 
of impiety. Rather, we know that Epicurus56 attacked notorious 
fifth-century atheists, including Prodicus, Diagoras, and Critias 
among others, for removing the divine EK 'trov OV'tCllV, which rather 
suggests an effort to distinguish his own philosophy from theoreti­
cal atheism. Epicurus' famous affirmation of the existence of gods at 
Ep. ad Men. 123 (quoted supra 195}-an emphatic inversion of the 
atheists' notorious credo OUl( dvat SE06c; and intended no doubt to 
be equally provocative in its (positive) dogmatism-indicates a de­
sire to leave no doubt on the question. But it is unlikely to have 
been a response to any actual charge of impiety, especially in light 
of what follows: OtO\)C; o· au'touc; (ot) 1tOAAoi VOJlt~o\)(nv OUl( Eiatv 
. ... aaE(3itc; 0' OUX 0 'touc; 'trov 1tOAArov SEOUC; avatprov, aAA' 0 'tac; 
'trov 1tOAArov o6~ac; (here='false beliefs') SEoic; 1tpOaU1t'tCllv ("but 
they Esc. the gods] are not such as the many believe them to be .... 
And the impious man is not he who denies the gods of the many, 
but he who attaches to the gods the beliefs of the many about 
them"), which is anything but a concession to traditional beliefs 
about the gods.57 

Moreover, public trials in Athens of philosophers for denying the 
traditional gods seem to have been confined to the second half of 
the fifth and to the fourth centuries.58 In other cities we do find epi­
sodes in which Epicureans are faced with expulsion, but in every 
case such litigation is directed at their hedonism rather than their 
theology. 59 Even the public charges of impiety promulgated at 

56 In Book 12 of his llr.p\ qn)ar.~: P.HercuL 1077 XVIII, see 216 infra. 
57 The ancient assumption that Epicurus could not have publicly denied the exis­

tence of the gods, if he so wished. seems to be totally unfounded; many others cer­
tainly did so, cf Drachmann (supra n.5) 12££. Dover (supra n.5) 25-54, Winiarczyk 
(supra n.l) 168-70. The Epicureans placed a high premium on social harmony, 
though this seems to have hindered them not at all in the development of positive 
doctrine (e.g. their theory of pleasure, for which they did incur harsh persecution). 

58 Drachmann (supra n.5) 6£; Derenne (supra n.6) 214-16 (public indictments of 
philosophers only in the fifth and fourth centuries B.C.) and 264-66 (confined to 
Athens). Cf J. Lipsius, Das attische Recht und Rechts7Jerfahren I (Leipzig 1908) 
358. 

59 For instances see H. Usener, Epicurea (Leipzig 1887) lxxii ff. These include the 
expulsion of Epicureans by the Messenians; Philiscus and Alcaeus expelled from 
Rome by L. Postumius (Ath. 547A; Ael. VH 9.12; De pr07Jid. fr.39 Herch. ap. Suda 
S.7J. 'E1t{1COUPO~, also the expulsion of unnamed Epicureans from Lyttus). For legis­
lation of Messene and Phalanna against Epicureans, P.HercuL 155/339 (Philod. De 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 205 

Athens seem to have been directed in the first instance against 
alleged disregard for religious custom, and this attitude may not 
have been widely extended to include theoretical speculation about 
the gods (Socrates is perhaps a notable exception). 60 This does not, 
of course, mean that philosophical views did not matter, but rather 
that in certain respects the community of rites and religious prac­
tices constituted the primary level of social, political, and religious 
activity. Some modern theorists, for instance, hold that it is mis­
taken to adopt structures of belief as the starting-point for analyzing 
ancient religion as a cultural phenomenon;61 the basic level of sig-

Stoieis) III.6-8 Dorandi (CronEreol12 [1982] 99): Kat OU1 'tTtV i][o]OVTtV {mo <l>a­
AaVVatCOv Kat. MEO"O"TlVtCOV EK~E~Af1O"eal 'toue; 'E1tlKO'U[p]do'Ue; (the text of VHt, 
quc.ted by Usener, should be ignored). Notices of expulsions of Epicureans from 
unspecified cities: Sext. Emp. Math. 2.25; Cleomedes 2.1; Plur. Mor. 1100D. 

60 See esp. Dover (supra n.5) 25-54; Rudhardt (supra n.6) 87-105. A possible ex­
ception would be the controversial charge of impiety for the introduction of Kalva 
O<llIlOVta, which could conceivably have been directed toward philosophical 
reinterpretations of the traditional gods. Such considerations, however, did not 
figure in the charges of impiety promulgated in the late fourth century against 
Aristotle and Theophrastus, the examples which will have been closest in time to 
Epicurus. 

61 For a review of modern approaches to the controversy see S. R. F. Price, 
Rituals of Power: the Roman Imperial Cult in Asia Minor (Cambridge 1984) 7-22, 
esp. 7-11. Brunt (supra n.l1) 183 with n.17 is skeptical to the point of anachronism. 
The value of "the mental attitudes of the vulgar" (Brunt 182) as a category of analy­
sis is difficult to assess. But the claim that "in philosophic discourse propositions of 
the kind 'gods exist' and 'gods care for mankind,' were clearly taken to be ... 
susceptible of demonstration or refutation or doubt on rational grounds, and 
philosophers clearly supposed that ordinary men understood them in the same 
way" (183 n.17) demands careful consideration. First, Cicero's motives are un­
doubtedly polemical; whatever his own view, the appeal to popular belief is an 
argumentative tactic. Second, with respect to the gods as recipients of cult, the 
status of 'belief' is particularly complex, and many in antiquity were uncomfort­
able with a literal understanding of pronouncements regarding the gods and their 
cult (see the list of counter-examples at Cic. Nat. D. 3.39-63, and the conclusion at 
64). In the schools allegoresis and Euhemerism flourished and the criticism of 
traditional religion in popular philosophy continued unabated. As for the 'vulgar', 
many rituals and myths, so far from being susceptible to rational, literal 
explanation, were popularly regarded as utterly baffling if considered in such 
terms. A standard example in antiquity was the Attic bouphonia, which already 
Aristophanes (Nubes 985ff) portrays as exemplifying a type of religious rite whose 
significance no one could understand (ef Theophr. De piet., ap. Porph. Abst. 
2.29-31; Paus. 1.24.4,28.10; Ael. VH 8.3; Suda S.'V'V. Boo<pOvla, 8a'UAOOv; Hesch. S.'ll. 

Bo'U'tu1toe;). So also Ar. Aves 960-90, Thesm. 236, 248, 331-51, Nub. 254-68, Pax 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

206 THE ATHEISM OF EPICURUS 

nificant activity in ritual need not be one that expresses a specific 
rational view, but rather one that exemplifies a conventional mode 
of conduct.62 Greek religion was in many respects as concerned 
with praxis as with belief; generally more important considerations 
are the cohesion of the social group and the evocative power of its 
symbols, the preservation of social distinctions, and the transmis­
sion of a cultural heritage. 

On this view Epicurus would have been beyond reproach (as 
later Epicureans took such delight in pointing out), not only be­
cause he could be shown to have scrupulously observed rites of tra­
ditional cult, and to have defended the use of religious language, but 
also because his theory of the origin and development of religion in 
cultural history (by which this appropriation of traditional cult was 
justified) emphasized the group-oriented functions of religion. 

On the other hand, the detractors of Epicurus, as we shall see, do 
give primacy to structures of belief: they regard religious rituals as 
expressions of specific beliefs about the workings of the divine. For 
this reason they accuse the Epicureans of inconsistency with regard 
to practice and beliefs. The Epicurean position is complex: on 
another level, of course, the Epicureans are prepared to argue 
critically and to isolate contradictions between practice and belief; 
but they choose to do so largely by emphasizing the social and 
psychological consequences of holding false beliefs about the gods. 
In contrast to the Stoics on this point, the Epicureans hold that 

1056-60; Dem. 18.258ff; Hdt. 7.34f parody religious liturgy or ritual as woefully 
inexplicable in purely literal terms: see H. Kleinknecht, Die Gebetsparodie in der 
Antike (= Tubinger Beitr. z. Altertumswiss. 28 [Stuttgart 1939]); W. Horn, Gebet 
und Gebetsparodie in den Komodien des Aristophanes [Niirnberg 1970]). Some­
one must have been laughing. At the same time, those who did feel constrained to 
argue for true belief as a basis for the gods and their cult (see Zeno's famous 
syllogism deducing the existence of the divine from honors paid in cult: Sext. Emp. 
Math. 9.133) could also condemn the use of utilitarian or literally-construed prayer 
(D.L. 7.124=Posidonius frAO Edelstein-Kidd: the wise man will pray for ta 
a.ya96., which are necessarily non-material). 

62 An alternative though related cognitive approach is the symbolic one, accord­
ing to which myths and rituals do not in fact mean what they might normally 
seem to mean. This view holds that religious customs directly entail beliefs, but 
only in so far as they embody symbolic or metaphorical truths. In antiquity this 
approach was developed to its greatest lengths by those who had recourse to al­
legoresis to explain myth, although only Epicureans (who remained rather hostile 
towards myth) seem to have extended it to an explanation of ritual practices. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 207 

participants in traditional religion really mean what they said about 
the gods; but at the same time they seek selectively to exclude such 
meanings as appear to conflict with the theoretical workings of the 
observable world according to the constraints of atomism. 

Thus we find that issues in the modern debate between the 
cognitive (or literalist) and the more public, evocative (or symbolic) 
approaches to ancient religion were being worked out already in 
the philosophical schools of the Hellenistic period. Skeptics incline 
towards the former position, viewing ritual literally as an outward 
expression of a set of beliefs, which can then be enlisted to test or 
refute criteria current in the epistemological debate. Materialists, on 
the other hand, confident of the criterial value of sense perception 
and experientially-based reasoning, deny cognitive content to ritual, 
and invoke a symbolic or psychological explanation for ritual as a 
phenomenon and its persistence through time. For them, particular 
rituals do not express any specific relationship between gods and 
mankind. But when properly understood, they are ethically signifi­
cant and tell us something about human psychology as well as our 
cultural past. In both cases, positions taken in the epistemological 
debate predetermine one's theological views, while much theologi­
cal debate is conducted primarily in pursuit of criterial questions 
and of establishing lines of demarcation in the epistemological 
debate. By Cicero's time (De natura deorum is probably the most 
obvious and misunderstood example) the two issues are inter­
dependent and inextricably intertwined. 

To be sure, Epicurean practice regarding religious ritual offers 
little reason for questioning the later defense of Epicurus on this 
charge of inconsistency. An almost identical approach, after all, was 
adopted in practice by all the Hellenistic philosophical schools 
(except perhaps the early Cynics). Even the skeptical Pyrrhonists 
advocate adherence to traditional religious customs without any 
claim to knowledge of the nature of the divine;63 and the Academic 
skepic Carneades, so far from ever being charged with impiety,64 

63 See Pyrrho Test. 55, 62 Decleva Caizzi (frr.68, 81 Diels); Sext. Emp. Pyr. 1.23f, 
226,246; 3.12; Math. 9.49; Burnyeat (supra n.16) 33 (=126); J. Barnes, "The Beliefs 
of a Pyrrhonist," PCPS 28 (1982) esp. 15f; Frede (supra n.1 0) 225-54. 

64 The contrast, I think, was first drawn by Drachmann (supra n.5) 8; cf De­
renne (supra n.6) 216 (Cicero Nat.D. 3.44 says much the same of Carneades). The 
reported indictment of Protagoras for holding virtually an identical position re­
garding the gods illustrates the apparent contrast between Hellenistic tolerance to­
wards intellectuals and the early reaction to fifth-century rationalism. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

208 THE ATHEISM OF EPICURUS 

was actually entrusted with important affairs of state as a diplomatic 
emissary.65 

Let us begin again. If contemporary social hostility is an unpromis­
ing candidate for the source of the tradition concerning Epicurus' 
atheism, we must look next to the Hellenistic philosophers 
themselves, in whose debates the charge of irreverence towards 
the established order of things was a common argumentative tactic. 
As it happens, one of the earliest attested allegations of atheism 
against Epicurus comes from philosophical quarters. The Academic 
speaker Cotta concludes the first book of Cicero's De natura de­
orum with a citation from Posidonius' TIe pi 6erov: 

verius est igitur nimirum illud quod familiaris omnium nostrum Po­
sidonius disseruit in libro quinto de natura deorum, nullos esse deos 
Epicuro videri, quaeque is de deis inmortalibus dixerit invidiae de­
testandae gratia dixisse 

It is doubtless therefore truer to say, as our good friend Posidonius 
argued in the fifth book of his On the Nature of the Gods, that Epi­
curus does not really believe in the gods at all, and that he said what 
he did about the immortal gods only for the sake of deprecating 
popular odium.66 

According to Cicero, Posidonius did not say that Epicurus ever ex­
plicitly denied the existence of gods (in fact, the fragment implies 
knowledge of statements to the contrary), nor that he was regarded 
as an atheist by his contemporaries (videri, in contrast to dixerit, is 
inferential), but only that he might have been, if his true views had 

65 To be sure, later skeptics advocate the suspension of all belief with regard to 
certain knowledge; i.e., they consider positive claims for the non-existence of gods 
on the same epistemological footing as dogmatic claims for their existence: in the 
end, they supposed, opposing positions will cancel each other and lead to the 
suspension of judgment that they advocate. In this respect they succeeded in 
epistemologically actualizing and legitimizing the intermediate position as regards 
the gods of Protagoras,!et (unlike Prodicus) without recourse to the symbolic in­
terpretation of ritual an myth; cf Henrichs (1975 [supra n.5]) 109 n.63. 

66 Cic. Nat.D. 1.123 (Posid. fr.22a Edelstein-Kidd=346 Theiler). On the context, 
see further G. Rudberg, Forschungen zu Poseidonios (=Vetenskaps-Samfundet i 
Uppsala 20.3 [1918]) 30ff. See also J. Heinemann, Poseidonius' metaphysische 
Schriften II (Breslau 1928) 153f; K. Reinhardt, Kosmos und Sympathie (Munich 
1926) 181£; M. Pohienz, GGA (1930) 143; R. Philippson, RE 7A (1939) 1154; H. Uri, 
Cicero und die epikureische Philosophie (Munich 1914) 10M; G. Pfligersdorffer, 
Studien zu Poseidonios (=SBWien 232.5 [1959]). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 209 

been popularly known/,7 Posidonius thus represents Epicurus as an 
atheist whose explicit pronouncements on the gods are merely 
specious attempts to placate a potentially hostile public. Posidonius' 
remark is well-suited to Cotta's critique of Epicurus' theology: Epi­
curus would have been better off to have denied the existence of 
the gods (they play no integral role in his physical system), and so 
might just as well have done SO.68 

Did Posidonius invent Epicurus' atheism? Our only other source 
is Sextus Empiricus Math. 9.58: KUt 'E1tiKOUpO~ OE KU't' Eviou~ ro~ 
J.1EV 1tPO~ 'tou~ 1tOAAOU~ u1toAdxEt 6EOV, oo~ OE 1tPO~ 'tllv q"U<HV 'troY 
1tPUYIl<l'tOOV ouOUJ.1&~ ("and, according to some, Epicurus in his 
popular exposition allows the existence of god, but in expounding 
the physical nature of things he does not allow it"). The latter clause 
suggests that it is Epicurus' physical system (presumably its lack of 
teleology and denial of divine providence) that seemed to conflict 
with his statements and practice concerning the gods. 69 That is, the 
philosophical charge of atheism against Epicurus by-passes Epi­
curus' explicit statements, attacking instead the premises of his 
physical system, and inferring that the gods have no necessary place 
in this system. The problem of the precise identification of the 
source of this charge has been much discussed, though Sextus' 
handling of the matter certainly suggests a skeptical origin. 70 Owing 

67 Apparently quaeque refers to Epicurus' account of the gods, in popular or tech­
nical terms. This would include his attitude towards participation in traditional 
religion and his habit of using religious language allusively, both ridiculed in the 
lines preceding the citation of Posidonius, and earlier at 1.115. 

68 See e.g. the tentative formulation at 1.87f, and the conclusion nullos esse deos 
(on the authority of Posidonius) at 1.115. 

69 Cf Lactant. De ira Dei 4.7 (Posid. fr.22b E.-K.=346 Theiler):itaque verbis illum 
deos relinquere, re autem ipsa tollere quibus nullum motum nullum tribuit offici­
um; here the last clause cannot be simply derivative from the corresponding pas­
sage in Cicero, as most commentators assume, though for the idea cf Nat.D. 1.92, 
101f; Pluto Mor. 1117B (Epic. fr.141 Us., 65 Arr.2), uqrucrtCOAO'Y'l'tov; Mor. 1102B (Epic. 
fr.30 Us.), {)1tOlcpivE'tat yap Euxa~ Kal. 1tpocr1CUvftcrEt~ ou8£v OE6J.LEVO~ ~ha cp6~ov 'to)v 
1tOAArov Kal. cp8£'Y'YE'tat <pcova~ Evav'tia~ ot~ <ptAocrO<pEl ... olhco yap 'E1tiKOt>po~ olum 
OE1.V crx:rlJ.La'ti~E<J8m Kal. J.LTt cp8oVE1.V J.L'Ilo· u1tEx8avEcr9at 't01.~ 1tOA.A.o1.~; Mor. 1119D-E, 
1123A, 1124E. 

70 See Pease on Cic. Nat.D. 1.123; R. Hirzel, Untersuchungen zu Ciceros philo­
sophischen Schriften I (Leipzig 1877) 33-37; Philippson (supra n.66); SymbOslo 20 
(1940) 25f. Hirzel (37) argues (implausibly, I think) that the source of Sextus' 
charge cannot be Posidonius alone, because Sextus asserts that this view was held 
Ka't' £viO'\)~. Nor is there the slightest evidence for Hirzel's further suggestion that 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

210 THE A THEISM OF EPICURUS 

to the uncertain state of the text of De pietate, the (relatively early) 
testimony of Philodemus has never been properly drawn into the 
debate. Now that reconstruction of the text is on firmer ground, 
the charges transmitted by Posidonius and Sextus can be seen to be 
identical to those addressed by Philodemus in that work, for here 
the papyri too support a skeptical origin. Several passages will serve 
to illustrate this, showing how context can aid in interpretation. In 
the first, the author concludes his catalogue of Epicurus' positive 
arguments about the gods in these words (N 1077 xv.3-21 [HV2 II 
79; p.109 G.]): 

Kat 1tUV-
4 "CCl~ [K]~"C' aVU)'1CCl~ 

[o't "Cau"Co]y EypmjlClV a­
[~lOU~] yE 1tPOCltPf-
a[ E<'o]~ £VEKCl eaulla-

8 ~Eaeat KCll. IlTt oux "CTtv 
(a)OuvCllllClV avo(JlO\l[~] 
vOlll~E(J8Clt· 1taV"CCl~ 
O£ av8pclmou~ ro-

12 pCl {t} AfyEtV avoalou~, e-
1tEt0111tEP ouod~ 
ElKVOUIlEVCl<; 1t£-
pt 't[ou e]~o'\)~ ,map-

-I< 

... and that of necessity 
all who have written in this 
fashion deserve to be ad­
mired because of their inten­
tions and not, on account of 
inability, to be regarded as 
impious. It would be fitting 
("it is time") to describe all 
men as impious, inasmuch as 
no one furnishes convincing 
demonstrations for the exist­
ence of the gods; neverthe­
less all men (with the excep-

the charge of impiety against Epicurus originates in the Old Stoa. With xu't' Eviou~ 
should be compared Cic. Nat.D. 1.85: quamquam video nonnullis videri Epieurum 
... verbis reliquisse deos, re sustulisse (for the purposes of argument, Cotta here 
accepts that Epicurus meant what he said about the gods, which may indicate a 
source different from 123, but Academic method certainly allows argumentation 
from one's opponent's premises [so also Sext. Emp. Math. 9.58; ef 64]). Diels, Dox. 
Graee. 225 and Sibyllinisehe Blatter (Berlin 1890) 21£, argues that the ample five 
books of Posidonius' TIEpl SECOV were filled with lists of examples of theological 
views rather than argument. Since we know that Posidonius in that work ignored 
Epicurus' explicit views on the gods, perhaps his physical views were adduced (on 
someone else's authority) as an example of atheism. In any case (on Diels' as­
sumption) Posidonius too is likely to have followed a source for the atheism of Epi­
curus. On the doxographic character of Pcsidonius' work in these areas (as borne 
out by his influence on the later doxographies), see most recently P. A. Vander 
Waerdt, "Peripatetic Soul-Division, Posidonius, and Middle Platonic Moral Psy­
chology," GRBS 26 (1985) 373-94, who defends (at 318 n.30) Diels' general for­
mulation against his subsequent detractors. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 

16 X£[ l V 't'cX<; a1to] ¢dl;£l<; 
£1,>1:t[ oPllcr ]f;:V' OIl(J}<; 
OE [cr£~]oV'tal 1tUV­
n:[<; d 1111 1tapu]K01tOl 'tl-

20 v£[<; a\)'[Ou<;, Ka]9um:p 
i)~[Et<;] . 

tion of some madmen) wor­
ship them, as do we. 

211 

3£ 1tclvl'ta.; N: 1tclvl'te.; corr. Gomperz falso (ef vv.9, 10) 5 Philipp­
son: aut ]y aut ]q:t N: [0'1. 1tEpl SE&]V Gomperz (app. crit.) sed spatio 
longius, ut vid. 6 Gomperz 7 Biicheler 9 a- add. Biicheler 
avocrLov[,;] Obbink: -LIOI [. N: avocrwt Biicheler, edd. omnes llf copa 
(sc. copa{t}) Biicheler: QIPAI N: (Il)ropta Gomperz 15 't[ou] Philipp­
son: 't[ou.; S]~ou.; Gomperz: n ... ]COYE N 15£ lmapxE[tv] Anon. A: 
imclpxdt] Gomperz 16 ['ta.;] Philippson [a1to]Qd~ft.; Gomperz 17 
E1)~[6p"cr]~v Philippson post ~v spat. vac. N 18-20 [crE~]ov'tat ... 
'tlVe[.;] Gomperz (app. crit.) 20f Philippson 

Here Philodemus complains that his opponents comrletely ignore 
the kind of detailed argumentation about the nature 0 the gods that 
he has delineated in the preceding columns. (Previous editors mis­
undertood this passage because, by reading the columns of P.Her­
cuI. 1077 in the wrong order, they thought this text came before the 
series of arguments on the gods.)71 In this respect the Epicurean 
authors cited are not lacking in ouvalll<;,72 They are perfectly 
capable of producing detailed discussions on the nature of the gods; 
but the validity of their conclusions had never been called into ques­
tion. Similarly, we know that in the charges of Posidonius and 
Sextus, the Epicureans' theological doctrines are not attacked 
specifically; instead, they are regarded on more general grounds as 
insincere. But the adversaries' claim must have been a more 
general, perhaps skeptical one (presumably directed towards the 
premises of these accounts), since Philodemus offers in response 
the dogmatic claim (on the basis of consensus omnium) for the 
existence of gods. Here he brings into sharp focus the implications 

71 D. Obbink, "Philodemus, De pietate I" (diss.Stanford 1986) 24-43. 

72 For the meaning of aOUVUIl1.U in 9 (,inability' due to defect of intelligence) see 
L. Taran, Aeademiea: Plato, Philip of Opus, and the Pseudo-Platonic Epinomis 
(Philadelphia 1975) 292, on [Plat.] Epin. 985E3--4; if. Leg. 821A8-B2, Hip.Ma. 295E9-
10; Arist. Poet. 1460b 17. Possibly aOuVUlltU here refers to lack of power or influence, 
as at Marcus Aurelius 2.2.3 (with Farquharson's note, II 520). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

212 THE ATHEISM OF EPICURUS 

of his opponents' criticism:73 if Epicurus is to be regarded as im­
pious because his physical system does not offer logical proofs 
(a1to8d~Et~) for the gods' existence, so also must virtually all man­
kind, though this hardly keeps them from worshipping gods as 
existing beings.l4 Epicurus and his followers do so too, as the author 
proceeds to document in detail. 

The papyri help further. Later in the treatise, at the conclusion of 
his catalogue of instances of religious participation by Epicurus and 
his followers, Philodemus reviews the charges of atheism and 
impiety promulgated by his opponents (0 1077 II left and right 
(missing in N, HVZ II; pp.93f G.): 

Col. 46 
[ ... Kal'tT)v fUEP'Yf]­
otav oi~ 0 ~tO~ ft[ll]wv 
8taOw~f'tat, q>[A]U­
aptav 8£ XOlli~Eo-

4 Sat Kat 'tU Ilu[o'tllpt]-
a Kat 'tU~ £op'tu~ a-
'tf 'twv 8(l)' OU~ 'Y~lv2[ v]­
'tart] A[fhOIl£VcrV 1l[1l]-

8 8fVO~ E1ttc:hp£q>do]­
Sat. Kat q>aot 'tOY 
'E1tlKOUP9V EK1tfq>fU­
"([£V]at'tov'A't't[t]K[OV] 

12 §[XA]OV OUX on d .. J 
THOfECONTQ[ ... ft't]-

[Our opponents object that 
Epicurus accepts] benefaction 
(from the gods), by which 
things our life is preserved, 
but that the mysteries and the 
festivals are to be regarded as 
foolishness, since those (i. e. 
the gods) for whose sake they 
(i. e. mysteries and festivals) 
take place are said to pay no at­
tention to anything. They also 
claim that Epicurus escaped 
from the Athenian masses not 
because [ - - - - - ] he had less 

73 In lines 11f a more literal readering of wpa is: "it is fitting time to call all men 
impious" (i.e., if, on the opponents' view, one were to regard the Epicureans de­
scribed in the previous lines as impious, it would be necessary to call all men 
impious). For wpa sc. eoti=tempus est, see Philod. De mus. p.10S.1S Kemke (p.216 
Van Krevelen): (if one is to follow the common man's value judgments, many 
unworthy practices have been honored), Kat <jltA.ooo<piav oi: Ot<l to 11ft tE't£UXEVU[ l 
t]o'i<; OA.ol<; bta8ACOV atet[l1]aISCtV wpa. 

74 Philodemus contrasts this rhetorical argument (he later calls it a 1ttCHt<;) for 
their existence with U1toO£t!;t<;. On the argument from consensus omnium, see esp. 
Cic. Nat.D. 1.43, Div. 1.1, with Pease's notes on both; Arist. Eth.Nic. 1172b36-
1173a1, Cae!. 270bS-6; Sext. Emp. Math. 9.61-65; R. Schian, Untersuchungen uber 
das 'argumentum e consensu omnium' (Hildesheim 1973); Schofield (supra n.1l) 
283-308. The Epicureans combine the consensus argument with their doctrine (for 
which see Epic. Ep. ad Men. 123; Cic. Nat.D. 1.44) that it is part of human nature 
to form 1tPOAlJ'Vet<; of the gods. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 213 

'tOY ao"E~El<; dX[EV] 
i)1t6All'l'El~, aA[Aa 'tOlL] 

16 OLaAEA1l8€va[ l 1tOA]­
AOU~ av8[p ]O)1tou[~] 
['t]tlv q>lAOO"Oq>i[a]v a[u]­
[ 't ]ou. 1t pOo"E1tlq>€ pou-

20 o"lV*O£ Kat Ol6't[l] 
';lWJV aya8rov avoprov 
Kat OtKairov 1tapat­
povJlE8a 'ta[~] KaAa~ 

24 tA1tioa~ &~ tv 'tOt~ 
[8]£Ot~ EXOUQ"l Kat Ol­
[a] 'tOU'tO JlE'ta 'ta~ 
KaAAlEp"o"El~ Kat 

Col. 47 .. 'f 

'ta~ o"1tou[oaia~ KEU]­
q>rovou~ ~[l'ta~ avat]­
p'O'(U)JlEV ['tllv 1tapa]-

4 Jlu8iav ['troY EUro]­
XlroV. 

ImpIOUS VIews, but because 
his philosophy had escaped 
the the notice of many men. 
And they charge in addition 
that we deprive just and good 
men of the fine hopes which 
they have in the gods, and 
that therefore, after pro­
pitious offerings and sincere 
and sonorous prayers, we do 
away with the excuse for 
sacrificial meals. 

Col. 46: 0 1077 II laev. (p.94 G.; desunt N, HV2 II) + 0 1077 I (p.93 
G.; desunt N, HV2 II) 1 ['to. roq,EA,\lfla Kat 'ti1v £u£py£]laiav coni. 
Obbink 3 corr. Gomperz: XOfl- 0 4£ Gomperz 6f (vel 
ntnl'ta(\]) Henrichs (ef e.g. Philod. Rhet. I 104.31-34, 191.24f}: 
L\OycrCINE[ ]TA[.] N (de hoc loco desperavit Gomperz) 7 Obbink: 
A [.]TOMENQN 0 11 y[ supplenda ex 0 1077 1.10 (sottoposto): 
£K7tE<pEUI[YEv]a\ Gomperz 12 Philippson: S[(?) probavit 0 1077 1.11 
(sottoposto): [o11fl]ov Gomperz: [Il-roll-]ov Diels 12£ E[ ... ]1 0 1077 II 
+ THOfECONTQ[ ..... ]1 0 1077 L12 (sottoposto)+O 1077 II: 

['E7ti]~o\>[p]~S ov'tcp[S Philippson (sed nomen Epicuri iam v.l0 supra) 
13-26 Gomperz 17 init. supplenda ex 0 1077 I: [.]OUS 0 1077 II 18 
[.]1JV ex 0 1077 I: [ .. ]C 0 1077 II 19 corr. Gomperz: EY[.] 0 27 
Philipps on: KAANEP 0: KaA,~S plJaw; Gomperz haesitanter 27f: 
columnas 46-47 coniunxit 0 1077 I 

Col. 47: 0 1077 II dext. (p.94 G; desunt N, HV2 II) 1 a7tou[oaias] 
Gomperz 1£ [nu]l<provouS Philippson M\'ta.c;] Obbink: A[ 0: 
~UXa.S] Philippson 2£ Gomperz: p[[n]], O'MEN[ 0 3£ Gomperz: 4£ 
l'trov eu]IXtrov Philipps on 

As often in Hellenistic philosophical polemic, Philodemus does 
not here respond to his opponents by name. But the criticisms of 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

214 THE ATHEISM OF EPICURUS 

Epicurus outlined here agree with those attested by Posidonius and 
Sextus, augmenting them in some detail. Philodemus reports the 
basis for his opponents' attack on Epicurus: they acknowledge that 
Epicurus professes to accept certain features of traditional religion 
(46.1£: we get good things from the gods)Js But they allege that 
Epicurus (implicitly) rejects the literal understanding upon which 
traditional religion is based. Accepting the Epicurean premise that 
the gods are removed from this world and unmoved by human 
appeals, the opponents proceed to argue that such gods, for whom 
rites of worship as mysteries and festivals are conducted, must 
necessarily take no notice of them (3-9). Therefore, they conclude, 
Epicurus' practice of observing religious measures is insincere: in 
other words, his practice contracts his beliefJ6 This is made 
especially clear by the link between the two columns, which now 
afford some continuous sense: 77 the Epicureans are said to engage in 
prayer and other rites of traditional religion, while in effect «doing 
away with the consolation" (47.3f [1t(lp(l]~uei(lv) offered by cult 
that constitutes the theoretical basis for such acts (namely, that we 
are rewarded in return by the gods ).78 

Philodemus' opponents further allege (9-21) that Epicurus' true 
views (on the gods) were in fact unknown to people generally, as a 
result of his flight (following the doctrine of Miee ~lci>a(lC;) from the 
masses. Several columns later, Philodemus capitalizes upon the 

75 That Epicurus professes to accept some features of traditional religion as valid 
seems clear from the implied contrary of 46.2f cp[A]uaplaV Q£ lC'tA. which came in 
the lacuna immediately preceding the beginning of the column. I tentatively sug­
gest -benefits" (i.e .• from the gods) on the basis of the restoration [ti)EpyE]oiav in 
1. preceded by 'tel mcpEAtlla (or something similar) since ote; in 1 requires another 
(masculine or neuter) antecedent. Cf Epic. Ep. ad Men. 124 be 9Erov £1tayoV'tal 
leat WcpEA.£lat.; Philod. De piet. (P.HercuL 1428 xII.22-25) p.23 Henrichs £~ au'trov 
(sc. 9 EroV) AEYOIlEV 1tapaleoAo'U9ElV lea[t] 'trov aya9rov 'tel IlEyto'ta; Epic. ap. 
Philod. De piet. p.124 G. (on the text. see D. N. Sedley. CronErcol4 [1974] 89-92). 
and p.145.11-21 G. On Epicurean - PMiPll-rocptA.ua-Lehre" see especially Schmid 
(supra n.38) 115-26. 

7' Epicurus' detractors (who assume that specific rituals entail specific beliefs) con­
clude that the Epicurean practice of engaging in ritual is insincere (46.19-47.5). 
Philodemus has already demonstrated in considerable detail that Epicurus partici­
pated in the Attic (i.e. Eleusinian) mysteries (p.127.24ff G.) and in specific calen­
drical festivals (p.127.24ff G.; cf 105 G.=Epic. fr.157 Us .• 86 Arr.2). 

n See Obbink (supra n.71) 116-19. 160-62. 
78 At 47.2f. [avat]po(u)IlEV is clearly inferential; so also Philod. De piet. (P.Her­

cuL 1428 col. xv.5-8) p.2S Henrichs. quoted supra 189f. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 215 

premise that Epicurus' views went unknown at Athens in order to 
refute the allegation that they caused social harm; his opponents' 
charge of atheism itself betrays the fact that it had no historical basis, 
but was rather a rhetorical ploy derived from later philosophical 
polemics. Philodemus further suggests an alternative explanation 
(quoted in part supra 203) for Epicurus' flight from the masses and 
argues that by conceptually preserving the philosophical integrity 
of the Greek gods as free from carc or blame, Epicurus actually 
holds views more pious than those of the masses. 79 

The (unnamed) detractors of Epicurus thus capitalize on his own 
premises and allege that participation in religious ritual, if sincere, 
would conflict with basic Epicurean doctrines (e.g., their denial of 
cosmic teleology and divine providence, insofar as these would 
"destroy the theoretical basis"--do ut des-for such ritual). 
Epicurus' opponents regard ritual primarily as the outward ex­
pression of a set of literal beliefs about the gods (i.e. , that the gods 
confer benefits on mankind in return for rites conducted in their 
honor), an approach rejected (at least implicitly) by the Epi­
cureans.80 

As in the charges reported by Posidonius and Sextus, the op­
ponents of Philodemus represent Epicurus as an atheist whose true 
views on the gods were screened from the public by the allegedly 
contradictory practice of religious observance. If we can identify 
these opponents with Posidonius and Sextus' source on this point 
(supra 208f), we can go one step further towards isolating the origin 
of this tradition. 

V. Gods and Madmen 

Sextus' testimony is particularly important because it suggests that 
the charge of atheism against Epicurus has a skeptical origin. In Sex­
tus' doxographical account of Ot 1tEpi SEiov A6yot, Epicurus is in­
cluded at the end of a long list of notorious atheists, which Sextus 
presents to show that belief in the gods' existence is by no means 

79 Cf Epic. Ep. ad Men. 123. On the psychology of the Epicurean 'flight' from 
the masses, see esp. Frischer (supra n.41). 

80 In fact, Philodemus is careful in the preceding section listing instances of Epi­
curean participation in cult to delineate its theoretical justification (i.e., based not 
on a literal understanding of pronouncements and procedures regarding the di­
vine, but on considerations of personal psychology and social justice). 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

216 THE ATHEISM OF EPICURUS 

universal; in this way the skeptics argue that the dogmatists' appeal 
to universally accepted opinion about the gods has no claim to 
validity. 81 This list of atheists had in fact a long history in Hellenistic 
doxography, though it is unknown at what juncture the name of 
Epicurus became incorporated into the catalogue. 82 This must have 
occurred at some point earlier than the work of Posidonius and 
Philo demus, though there is no evidence that it goes back to Epi­
curus' own day. 83 The inclusion of Epicurus, whose dogmatic pro­
nouncements on the gods were well-known, would be curious 
enough; but in fact the earliest known ancestor of Sextus' list of 
atheists appeared in Epicurus' own work (Book 12 of Ih:pt qn)cn:ro<;;) 
and is known, as it happens, only from a fragment preserved by 
Philodemus in De pietate. I offer here a text of the fragment (87 
Us., 27.2 Arr. 2 ), substantially revised after examination of the 
original (N 1077 XVIII.1-12 [HV2 II 82; p.112 G.]): 

Kat 7tClOaV Jl[aviav 'E]-
7tiKOUpOC;; E)J.[EJl'l'a]-
to tote;; to [8ElOV E]-

4 K trov ovtmv [aval]­
POUOlV, roc;; Ka[v trot] 
OmOEKUtm[t TIpo]-

Similarly, for their com­
plete madness Epicurus 
reproached those who re­
move the divine from the 
sum of things, as in Book 
12 (of On Nature) he re-

81 For the introduction of atheists (usually philosophers) by Academic skeptics 
as a counter-argument, see Cic. Nat.D. 1.62-64; 3.8, 11, 17 with Pease on 1.43; by 
Pyrrhonists: Sext. Emp. Math. 9.50-59 (cf 191£). At Nat.D. 1.63 Cotta suspects 
there must be some, but cannot name any; so also Pluto Mor. 1075 A; Diod. 3.9.2 
and Sext. Emp. Pyr. 3.234 name only OA.tYOt 'troy Aiau)1tcov. One dogmatic alterna­
tive was to posit a stage in cultural history when atheism had developed (the 'first 
men' having universally believed in gods): so Theophr. De piet., ap. Porph. Abst. 
2.8, and Simpl. In Epict. Ench. 222c-23A; Epic. De nat. 12, ap. Philod. De piet. 
(P.HercuL 229 I-VII) pp.113, 142-48 G. 

82 On the catalogue of atheists and its transmission in antiquity, see M. Winiar­
czyk, "'Der erste Atheistenkatalog des Kleitomachos," Philologus 120 (1976) 32-46 
(who provides a stemma), esp. 37 and 44££ on Epicurus and Philodemus (cf T. 
Thalheim, RE 2 [1896] 1529; F. Zucker, Philologus 64 [1905] 460-69). 

83 Constraints of anachronism require that the composition of Posidonius' TI£pl 
a£rov antedate the dramatic date of Cicero's dialogue, i.e., before 77-75 B.C. Cicero 
heard Posidonius lecture in Rhodes in 78; all the speakers in the dialogue could 
have heard him (as ambassador from Rhodes) at Rome in 86. The treatise De 
pietate could have been written as early as this date, or as late as the 50's-if 
Philodemus wrote it; if not, it dates from the floruit of Phaedrus at the turn of the 
century. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 217 

ob::rot Kat Ata[yopat] 
8 Kat 'K' pttiat K(i[llotC;] 

~£J:!.cp[E'tat] Cf~C; Ka[i 1tapa]­
K6~i[Et]v Kat [llatVEO]-
Sat Kat (3aKXEuou-

12 CJtV aU1:0UC; [d]1Ca[Cn]. 

proaches Prodicus, Diago­
ras, and Critias among 
others, saying that they 
rave like lunatics and 
likens them to Bacchant 
revelers. 

1-40bbink 1: ~[ov~v] Philipps on 2f: EA[ N: E~ [dA.E]ho Philipp­
son 3: 7to[A.A.Ot<;] Philippson 4f: [aA.A.' E]lpouALV Usener (unde 
Philipps~n, Symb Oslo 19 [1939] 24): [cpA.ua] Ipouotv dubitanter 
Gomperz (unde Philippson 1921) 5-8 Biicheler 6 OmOEK(l:tm[1.], sc. 
ITEpt cpuoEm<;. ut titulum libri Epicuri colI. frr.82-87 Us. + P.HercuL 
1111 fr.44 8' K' pt'tta1. corr. ex 'tpt'tta1. N Kli[A.A.o1.<;] Usener 9 
gt~<p[E'ta1.] <£~<; ,!a[pa] Usener: MENC[. .]8~CKA N Ka[t 7tapa] 
Gomperz 10f Gomperz 12 Usener 

This fragment was not included by Diels-Kranz in Die Fragmente 
der Vorsokratiker, presumably because in the form in which it was 
presented by the editors of the day, the fragment contained no 
clear representation of the views of the thinkers named. 84 The para­
tactic character of the early catalogue style (exemplified more fully 
by the lists of poets and philosophers in P.H ercul. 1428) is apparent 
even from this brief citation. Cicero says that Epicureans when 
reeling off such lists sounded like the censor calling off the roll of 
the senate. 85 Nestle first recognized that this list of atheists was later 
taken over and considerably expanded by Clitomachus, the adher­
ent of Carneades. 86 Thanks to Nestle's observation, we know that 

84 The use of UiPE1.V (4f) and its compounds to denote the (negative) implications 
of philosophical views is common parlance. For the locution, see De piet. (P.Her­
cuL 1428 xIl.llff) p.21 Henrichs, 'trov (l1tA.ro~ 'to 9[E1.0]V n[v]mpouV'tcov; Pluto Mor. 
1124D, and often; with the dative participle: Sext. Emp. Math. 9.123; cf Epic. Ep. 
ad Her. 71, OUK £~EA.a't€OV Ex: 'tou ov'to~ 'tuu'tTlV 'ti1v hupyduv. Previous re­
constructions were based on a misreading from the Naples apograph (in line 2). 
Winiarczyk (supra n.82) 37, 44f, relying on Usener's UA.A.' e]pou01.v, wrongly 
concludes that Philodemus is dependent upon an Academic report for Epicurus' 
list of atheists. 

85 Cic.Nat.D. 1.94: cum tamquam senatum philosophorum recitares, summos vi­
ros desipere delirare dementis esse dicebas. For earlier examples of the paratactic 
catalogue-style in lists, see Gorgias 82Bl1.6ff D.-K.; Antiphon Soph. 87B44 II.26-
Iv.22 D.-K.; [Xen.] Ath.Pol. 2.11ff. 

86 Theophil. Ad Autol. 3.7; W. Nestle, Yom Mythos zum Logos (Stuttgart 1942) 
416 and (supra n.5) 869, with Winiarczyk's corrections (supra n.82) 35f. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

218 THE ATHEISM OF EPICURUS 

Sextus' list is a direct descendent of Clitomachus'. 87 The catalogue 
of atheists, a version of which appears in Cicero (Nat.D. 1.63), was 
also a popular argumentative device among members of the later 
skeptical Academy. It was repeatedly augmented because the more 
historical instances of atheism that could be adduced, the more 
damaging was the objection to the argument from consensus 
omnium and the validity of KOtval. EvVOtat. Placed side by side with 
some of the absurdities about the gods put forth by dogmatic 
philosophers, the catalogue of atheists provides a striking illustra­
tion of the contraposition and equipollence of arguments, on the 
basis of which the skeptic recommends suspension of judgment. 
Carneades himself was fond of this line of attack;88 and since 
Carneades (through Clitomachus) had a hand in the expansion of 
the catalogue, he (or Clitomachus) may have included Epicurus in 
the list. If so, Carneades may be the ultimate source for the atheism 
of Epicurus,89 if not identical with the opponents of Philodemus in 
the passages from De pietate discussed above, as well as the source 
of the charges of Posidonius and Sextus.90 

In the absence of more specific testimony, this remains of course 

87 Sext. Emp. Math. 9.1, 140, 182, 190, and see Winiarczyk (supra n.82) 32--46; 
Diels, Dox. Graec. 58f; Henrichs (1975 [supra n.5]) 113 n.72. 

88 Cf Cic. Acad. 2.117 (from Carneades): est inter magnos homines summa dis­
sentio, and the catalogue of philosophers that follows. 

89 Zeno the Epicurean, who had a prominent influence on De pietate, heard Car­
neades at Athens (Cic. Acad. 1.96, who notes that Zeno ab eo Esc. Carneade] 
plurimum dissentiret). For the indebtedness of portions of De piet. to Zeno Epi­
cureus (who wrote a nEp\ EUO'EPdae;: Philod. npOe; 'toile; [O'O<p1.O''tae;] [P.Hercul.l005] 
col. VII.20 p.93 Sbordone=A. Angeli and M. Colaizzo, -I frammenti di Zeno Si­
donio," CronErcol9 [1979] 75 fr.12), see Hirzel, Untersuchungen III (1883) 26ff; R. 
Philippson, Hermes 56 (1921) 364; Diels, Dox. Graec. 12M. The thesis that most of 
the polemic in De pietate dates from the floruit of Carneades and Zeno Epicureus 
is supported by the fact that the chronological catalogue of Stoic philosophers 
criticized in P.Hercul. 1428 stops with Diogenes of Babylon, a contemporary of 
Zeno and Carneades. 

90 Even if the author of De piet. were refuting the account of Posidonius (on the 
chronology, supra n.83), he must have directed the argument toward his earlier 
Academic source: at P.Hercul. 1077 xv (supra 210) he adduces the argument from 
consensus omnium against his opponent-which would hardly be effective against 
a Stoic, since Stoics and Epicureans were in complete agreement on this point (they 
differ as to the precise content of the 1tpOA.TI'I'1.e; of the divine). The Stoic Balbus, e.g., 
appe~ls to this argument (Cic. Nat.D. 2.5, 12)-against Cotta and the Academic 
skeptics. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 219 

a mere possibility. Have we any reason to suppose that Epicurus' 
theology was ever represented in this light by Carneades? Plutarch 
does in fact preserve a remnant of Carneades' own allegations 
against Epicurus (Mor. l089c=Epic. fr.436 Us.). Carneades is said to 
have mocked «(aKro7t'u:) Epicurus for making too much of 
pleasures derived from the cultic observance of the ancestral 
EiKUOEC;;: 

oihE 'tOUS llE'tpi01'>S Kat crro<ppovas eiKOS EvOta'tpi~Elv 't11 bttvoi~ 'trov 
'tOtou'tcov ouoE aXEP EcrKCOX'tE'tOV 'EXtKOUpOV KapvEuol1<;91 xpunov'tas 
otov E~ E<P' TtIlEpiocov avaAEYEcr8at '"xocrUKtS 'Hod~ Kat AEOV'ticp 
cruvil A8ov;" i\ '"xo-\.> 8ucrtov E1ttOV;" i\ '"xoias eiKuooS EodxvT\cra 
XOAU'tEAEcr'ta'ta; " 

It is quite unlikely that persons of moderation and temperance 
should dwell on such thoughts and do the sort of things with which 
Carneades mocked Epicurus, and gather as from an official record­
book statistics about '"How often I got together with Hedeia or 
Leontion," '"Where I drank Thasian wine," or '"At which of our 
school's festival celebrations I had the most sumptuous dinner. "92 

Carneades' specific object of attack in this instance is probably the 
doctrine of pleasure. Yet his criticism clearly implies that Epicurus 
had no greater regard for, or better grounds for observing, rites of 
ancestral cult than he did for drinking fine wine or having inter­
course with one of the Epicurean hetairai. Thus the charge that Epi­
curus' observance of traditional cult was insincere is present already 
in Carneades, a view that corresponds fairly closely to the first and 
third charges delineated by Philodemus in the passage discussed 
above. If the charge of insincerity can be linked with Carneades, 
the other allegations (that Epicurus does not really mean what he 
says about the gods) seem to follow necessarily. Certainty may be 
unattainable, but Carneades (unlike e.g. Posidonius) had everything 
to gain by representing Epicurus-an avowed dogmatist and critic 
of notorious atheists-as himself an unbeliever; and it would be 

91 'E1th.:o'Upov add. Bern.; KupveaoT\~ Wyttenbach: lCupvEaOT\v X2g: KOpulOT}V X 1a: 
lCUPVtU c. 

92 See supra nn.37, 45. The £l.1Cao£~ were celebrations held by the Epicureans on 
the twentieth of each month on the model of Athenian ancestral and private cult, 
first in honor of Metrodorus, and after Epicurus' death, in his honor. They in­
cluded a cultic dinner, libations of wine, symposiastic activities, and other elements 
of private, ancestral, and hero cult: D.L. 10.18; Clay (supra n.37) 11-28. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

220 THE ATHEISM OF EPICURUS 

characteristic of Carneades' method to hoist Epicurus by his own 
petard, as it were, by including Epicurus in his list of atheists. 93 

Thus the skeptical Academic invention of the 'atheism' of Epi­
curus was, at least originally, a cunningly thought out argument in­
tended to expose the uncompromisingly anti-teleological focus of 
the Epicurean system, as well as some peculiarities in their theology 
-in particular with regard to the cultural appropriation of religious 
forms. Philodemus' treatise documents a further stage of the con­
troversy, insofar as it shows how the later Epicureans answer the 
Academic argument point for point, clarifying Epicurus' views on 
the role of religion in an hedonistic ethics and an atomist view of 
cultural history. They succeed at least in exposing the historical and 
biographical weaknesses of the skeptical argument. Moreover, they 
do in fact propose a counter-argument to the rather damaging objec­
tion the Academics raise-on the basis of the historical incidence of 
atheism-to the dogmatists' appeal to universal opinion about the 
gods. Although well-documented in De pietate, this argument has 
gone relatively unrecognized. As a strategy for preserving the 
alleged universality of belief in gods, the Epicureans counter by 
imputing insanity to such thinkers (including skeptics), considering 
them not merely philosophical deviants, but actual madmen (1tapa­
K01tOt),94 whose abnormal views do not figure in the determination 
of universally accepted and naturally formed opinion. 95 In this way 

93 For the possibility that Epicurus compiled this list in a self-refutation argu­
ment, reducing the figures named to a state of insanity and thus to a1tpa~ia, see 
infra; this move would closely parallel Carneades' method, which employs his 
interlocutors' premises (but never his own) to attack their doctrines: see Striker 
(supra n.l0) 54-83; on the arguments ad hominem in the skeptical Academy, see J. 
Annas and J. Barnes, The Modes of Scepticism (Cambridge 1985) 14, 45, 49f. The 
recurrent argument in Cotta's speech in De natura deorum that Epicurus' theology 
is unnecessary, and therefore better dispensed with, certainly has a Carneadean 
ring to it. Carneades' criticism of the Epicurean clinamen (Cic. Fat. 23-25) takes 
exactly the same form; see Sedley (supra n.17) 49f and n.70. 

94 For the argument (especially in theological polemic) to which the Epicureans 
were particularly but by no means exclusively addicted, see Xen. Mem. 1.10-15 (of 
sophists); Pease on Cicero Nat.D. 1.37 (quasi delirans), 1.34 (puerilibus fabulis), 1.94 
(desipere delirare dementis esse); Long-Sedley (supra nA8) I 23H with II 151£. For 
an example of its re-use by an Academic skeptic against a Stoic: Cic. Nat.D. 3.62. 

95 Arist. Eth.Eud. 1214b28ff excludes children and the insane from the range of 
those whose ethical opinions would be surveyed: they are in need not of argu­
ments (Myot) but of either political or medical chastisement; so also Eth.Nic. 1179 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK ZZI 

Philodemus is content to exclude known atheists from his argu­
ment from consensus omnium: 7tuv'tE[<; d I.rn 7tapa]Ko7toi 'ttVE<;).96 
Now that we know the original order of columns in De pietate, we 
can understand why. In the new reconstruction, this passage fol­
lows closely upon the citation (discussed above, from Book 12 of 
ilEpt CPUo"E(o~) that reports that the "first men" (7tpro'tOt liv8pco7tot) 
originally formed proper conceptions of divine nature. 97 In the 
succeeding citation (also from TIEpi CPUCJEO><; 12) on the madness of 
Prodicus, Diagoras, and Critias, atheism-together with the the­
ories of these writers, which ascribe the development of belief in 
the gods to fear of natural phenomena or fictions invented for 
political control (and thus implicitly "remove the divine from the 
sum of things")-is explained as a later, deviant stage in cultural 
history.98 

The Epicurean counter-argument has a certain rhetorical appeal, 
particularly in the sphere of epistemology. A skeptical opponent, 
for instance, could be placed in a position of a1tpa~ia, and thus 
silenced or forced to concede the possibility of his irrationality (and 
hence to sacrifice the credibility of his position), or to advance a 
positive defense of his skepticism in violation of E1toxil 1tEpl 1t(xv't(t)v 
or at least upon grounds that would play well into the hands of ma­
terialists appealing to sense impression and common notions. 

The result of the Academic argument was the misrepresentation 
of Epicurus as an atheist that gained abiding credibility in the minds 
of the educated public, to judge from its ready acceptance by 
Cicero. At the same time, there is simply no reason to suppose that 

b28f on the irrational persons whose 1ta.eo~ responds not to A.6yo~ but to ~ia. : M. 
Nussbaum, "Therapeutic Arguments: Epicurus and Aristotle," in The Norms of 
Nature: Studies in Hellenistic Ethics, edd. M. Schofield and G. Striker (Cambridge 
1986) 31-74 at 58f; cf contra P. A. Vander Waerdt, CQ NS. 37 (1987) 412 nAO. 

96 Supra 210f. The argument is also present in Cicero: see Nat.D. 1.37 (quasi 
delirans) with Pease ad loc., and Philippson (supra 217) 28f, for lists of correspon­
ding expressions in Philodemus and Cicero. 

97 Philod. De piet. p.l13 G. (P.Hercul. 1077 XIx.23-29): edited supra 196. 

98 One portion of this account appears in Philod. De piet. (P.Hercul. 229 I-VII) pp. 
142-48 G., and another at Lucr. 5.1168ff. Of course knowing the precise order of 
citations in De pietate tells us nothing certain about the original order of these frag­
ments in llep1. <Pi>(JeCl)~ 12 (Epicurus may have begun the section on early religion 
by attacking the theories of Prodicus et al.). But Epicurus' language in the present 
fragment suggests an attempt to associate their views with the secondary, deviant 
stage of thinking about the gods. 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

222 THE ATHEISM OF EPICURUS 

any ancient writer had at his disposal, as a basis for the charge, any­
thing that we would dignify by the name of 'evidence'. To be sure, 
Epicurus' theology, like that of Prodicus, represents a complex bor­
derline case between orthodox belief and heresy, insofar as Epi­
curus fully accepts the gods' existence but, in denying that they 
care for the affairs of men, represents them as essentially different 
from the traditional gods. 99 Yet this can hardly justify his skeptical 
opponents' charge, for it centers not on Epicurus' rrincipal claim 
(namely, the possibility of the physical existence 0 gods without 
recourse to cosmic teleology and divine providence), but rather on 
its sincerity. At the same time, one of Epicurus' important argu­
ments-that collective atheism is unknown, while isolated cases of 
atheism are easily accounted for through impaired concepts-is 
almost totally obscured in the Hellenistic debate. In this sense the 
entire history of the debate can be regarded as corroborating a 
theory popular a generation ago among French historians of 
philosophy and now somewhat out of favor: namely, that the 
Hellenistic controversy over skepticism (or, more accurately, its 
methods of argumentation), with its emphasis on playing off one 
philosophical school or set of doctrines against another, has the 
effect of distorting the distinctions of doctrine between dogmatic 
thinkers.loo The lasting result of the rolemical representation of Epi­
curus as an atheist was a distortion 0 historical fact. 

To be sure, the Epicureans themselves play no small role in this 
process, and Epicurus' representation (perpetuated not only by the 
skeptics but by Philodemus) of Pro dicus, Protagoras, and Critias as 
"atheists" (because they implicitly "did away with the divine") may 
be cited in this respect. Epicurus, however, has a justifiable dog­
matic point of contention with these figures over the role played 

99 Henrichs (1976 [supra n.S]) 21 (on Prodicus). quoting Drachrnann (supra n.S) 
3. 

100 See A.-J. Festugiere. La revelation d'Hermes Trismegiste II (Paris 1949) 3S9. 
362-69 ("Le skepticisme et les doxographies"). esp. 366 and n.l. Festugiere. char­
acteristic of his era. was searching for ·les origines de l'eclectisme et syncretisme." 
Following L. Robin, Pyrrhon et Ie skepticisme grec (Paris 1944) 129-34. he argues 
(367-69) that the skepticism of the New Academy before Antiochus. with its ob­
session for playing off the doctrines of one philosophical school against another. 
had the effect of • effacer les traits caracteristiques qui distinguent chaque doctrine" 
and led ultimately to ·un 'syncretisme' qui. sacrifiant ce qu' en chacune il y a 
d'original. les brouille toutes et finalment. fausse l'histoire." Even the Stoicizing 
dogmatism resurrected by Antiochus is viewed as ·un dogmatisme eclectique." 



OBBINK, DIRK, The Atheism of Epicurus , Greek, Roman and Byzantine Studies, 30:2 (1989) 
p.187 

DIRK OBBINK 223 

by religion in the origin and development of civilization. In Epi­
curean terms, Prodicus and Critias actually had denied the role of 
the gods themselves in the origin of religion (and thus in the order 
of things), and Epicurus would have liked to have had no part of it.101 

Hellenistic philosophy was itself in many respects a philosophy of 
doxography, and the latter was as easily distorted as the former. 102 

BARNARD COLLEGE, COLUMBIA UNIVERSITY 
N O'Vember, 1989 

101 For a valiant attempt to save Prodicus on the point of having done away with 
the traditional gods, see Henrichs (1976 [supra n.5]) 15-21, who argues that 
Prodicus denied their divinity, but granted them at least a past human existence; 
cf HSCP 88 (1984) 139-58. Epicurus would still disagree. 

102 The 'scissors-and-paste' method characteristic of many doxographers was 
adopted by philosophers and could, on occasion, direct the course of inquiry. Cf 
the assessment of the relation between source criticism and philosophy in antiquity 
by A. E. Douglas, Cicero (=G&R New Surueys in the Classics 2 [Oxford 1968]) 28, 
with the reaction of J. Mejer, Diogenes Laertius and his Hellenistic Background 
(=Hermes Einzelschr. 40 [Wiesbaden 1978]) 10ff. 

Ancestors of this paper were read at Columbia, the University of Texas at 
Austin, Tulane, UCLA, and Duke. I am happy to thank the members of the 
audiences on those occasions for criticisms and lively discussion. Special thanks to 
Albert Henrichs, A. A. Long, Peter Parsons, David Sedley, Susan Stephens, Zeph 
Stewart, Leonardo Taran, and Michael Wigodsky.