GREEK ATHEISM: A MIRAGE

Abstract

Taking its start from an argument of H. S. Versnel, that Greek expressions of disbelief in the existence of the gods are evidence of the possibility of belief, this article reviews the evidence of such expressions, and of ascriptions of atheism in Greek sources, and suggests that there was a difference of type, not only of degree, between Greek ‘atheism’ and our understanding of the term today. Atheist discourse in Greek sources is characterized by frequent slippages: for example, between the charge of ‘existential atheism’ and the failure to give the gods due acknowledgement; between introducing new gods and disrespecting the old. Ascriptions of atheism to third parties are commonly based on inferences from an individual's actions, lifestyle or presumed disposition – which in turn are rooted in a network of theological assumptions. The phenomenon of ‘Greek atheism’ is, fundamentally, a scholarly mirage.

In the climactic fourth appendix to his Coping with the Gods, Henk Versnel makes a striking argument: that Greek expressions of disbelief in the existence of the gods are evidence of a more general belief. ‘How can one person deny (the existence of) gods unless (all) others do believe that they exist?’¹ The examples of atheism that Versnel cites are what we might think of as the usual suspects:² Diagoras of Melos’ reputation for denial of the existence of the gods,³ Protagoras’ profession of ignorance about the nature of the gods,⁴ the Sisyphus fragment,⁵ Plato's description of a generation of atheists in Book 10 of his Laws, and some passing references in Aristophanic comedy – for example, the passage near the beginning of the Knights (30–4) in which one slave asks another whether he ‘thinks the gods true’ (ἐτεὸν ἡγεῖ γὰρ θεούς;), and if so on the basis of what proof. ‘These variant testimonies of critical reflection on both the nature and existence of gods’, Versnel concludes, ‘starting in the fifth and coming to blossom in the fourth centuries, strongly confirm that, with the exception of these agnostic or atheistic critics, “the Greeks” did believe in the gods.’

I yield to no one in my admiration for Henk Versnel's sprawling magnum opus. And yet, even while cheering along to his arguments for the validity of the search for ‘belief’ in the Greek world, this passage causes me unease. This unease has two focuses. The first is the intellectual manoeuvre itself: the argument that atheism clinches the case for ‘belief’. The second focus for unease is the premise for this argument: that there was uncomplicatedly such a thing as atheism in the Greek world. On this second point, I am surprised that Versnel has sold the pass so cheaply.

My default position has always been to emphasize the context of such expressions of religious scepticism, or their narrow reference: to argue, for example, that scepticism on the validity of one form of divination does not necessarily reflect doubt over all forms (but may even provide the necessary flexibility to support belief). In short, I have worked to chip away at the evidence of atheism rather than explicitly to deny it.⁶ This approach might seem to chime with Tim Whitmarsh's recent characterization of classical scholarship in his 2015 Battling the Gods. Whitmarsh there conceives of atheism as having been marginalized and suggests that this is indicative of a wider religious presumption:⁷

A pungent reminder, this, that Classics is not always the secularised discipline we have been led to think it is since the nineteenth century. But even when the anxiety is not rooted in (apparently) confessional concerns, it is there all the same. Classicists have been conditioned to think of Greek culture as saturated with a religiosity that is ‘embedded’ so deeply that any resistance to the divine would be unthinkable. Rooted in the anthropological and ‘ritual turn’ of the late nineteenth century, this assumption is designed to resist Christianising projections (e.g. of the distinction between sacred and secular space, of belief as a privileged cognitive state, etc.); but in fact it ends up simply replicating a different kind of Christianocentrism, by defining antiquity in terms solely of what Christianity lacks.

Whitmarsh's account of how the attempt to resist ‘Christianising assumptions’ ends up privileging Christianity as a point of reference for classical religion is one, in particular, with which we might have sympathy.⁸ The confessional concerns of some scholars have undoubtedly also shaped the field. When Robertson Smith, for example, distinguished between religion (in his own world) as a ‘system of belief’ and (in antiquity) as ‘a body of fixed traditional practices, to which every member of society conformed as a matter of course’, his own distinctively Protestant faith was clearly a driving factor.⁹ The way, however, in which individuals’ religious backgrounds have played out in their scholarship since that point is perhaps more varied and more complex than any catch-all formulation can hope to capture. (Simon Price, the son of a Church of England bishop, but who ‘seems to find [Christianising assumptions] under every bed’, is the figure above all whose shadow lies behind Whitmarsh's words here.¹⁰) Alongside the drive to see religion (at least in the form of ritual) as omnipresent and embedded, there has also been a contrary tendency: to latch on to isolated expressions of scepticism and to represent them – I would say, misleadingly – as reflective of a more general scepticism towards ‘Greek religion’.¹¹

This is a tendency which has been particularly manifest in response to literary sources for ancient religion. It is in significant part a legacy of the elision of belief in studies of Greek religion. (By contrast, the more complex, relational map of Greek religious beliefs that we lay out, the more resilient Greek belief becomes: the clearer it is that some parts can move without the whole structure giving way.¹²) It is a tendency to which Whitmarsh himself – with his overt celebration of ancient atheists – is arguably the heir.¹³

Lucien Febvre and the problem of unbelief

My own perspective, for what it is worth, has been framed not by any direct confessional concerns (though a mixed religious background has undoubtedly helped to inform my interest) but by an exposure to later periods of history: in particular, through saturation in the work of the medievalist R. W. Southern, and an early introduction to Lucien Febvre's famous thesis of the impossibility of unbelief in the sixteenth century.¹⁴ For Febvre, the pattern whereby writers of the sixteenth century are ‘casually assumed to have swung at will from aggressive unbelief to the most traditional kind of belief’ is a problem that has ‘been brought into being by us, and us alone’.¹⁵ (As Versnel would say, it is our problem, not theirs.¹⁶) Febvre's focus, by contrast, is on how Rabelais’ audience read his work, what it was possible for him to have meant in that specific time and context.¹⁷ His much-quoted conclusion runs as follows:

Today we make a choice to be a Christian or not. There was no choice in the sixteenth century. One was a Christian in fact. One's thoughts could wander far from Christ, but these were plays of fancy, without the living support of reality. One could not even abstain from observance. Whether one wanted to or not, whether one clearly understood or not, one found oneself immersed from birth in a bath of Christianity from which one did not emerge even at death.¹⁸

The account that follows this passage, of the penetration of ritual in all aspects of private and public life, is one that could easily have been written of the classical Greek world. (Again, it is very reminiscent of the work of Simon Price and others.) And, indeed, Jan Bremmer has gone some way in developing that argument in his essays on Greek atheism.¹⁹

Inevitably, some aspects of Febvre's account of the impossibility of unbelief have been challenged. Scholars of the Middle Ages and of the early modern period have lined up to show that there were in fact some individuals ‘mentally capable of thinking outside the accepted framework of religion’²⁰ – even if such treatments reveal a tendency to conflate atheism or unbelief with the merely heterodox.²¹ ‘My Lord, if any heere can proove there is a God, I will beleeve it’, asserted one man to the Bishop of London in 1599.²² But there are some lessons drawn from Febvre's work that are still powerfully relevant for our context, as well as for the medieval world.

Perhaps the most central lesson of Febvre's account, from which all others hang, is his conviction that atheism is not a timeless condition that can be diagnosed across the centuries – or his parallel suspicion of a teleological approach that sees earlier writers as foreshadowing later ones (what Quentin Skinner has termed the ‘mythology of prolepsis’):²³

unbelief changes with the period…It is absurd and puerile, therefore, to think that the unbelief of men in the sixteenth century, insofar as it was a reality, was in any way comparable to our own. It is absurd, and it is anachronistic. And it is utter madness to make Rabelais the first name in a linear series at the tail end of which we put the ‘freethinkers’ of the twentieth century (supposing, moreover, that they are a single bloc and do not differ profoundly from each other in turn of mind, scientific experience, and particular arguments). This whole book has shown this, or else it is worth nothing.

These are dangers to which classicists – because of the lure of the Classical world's primal status – are always especially vulnerable, and never more so perhaps than in relation to this topic. By contrast to Febvre, for Whitmarsh, ‘Disbelief in the supernatural is as old as the hills.’ ‘There have been many throughout history and across all cultures who have resisted the divine.’²⁴ Similarly, when David Sedley deploys Protagoras’ famous statement of agnosticism as the basis for setting a terminus ante quem of 420 bc for the development of atheism, it is clear that we are tracing the genealogy of our modern atheism: ‘Before someone can explicitly suspend judgement as to whether or not x exists, cases both for and against x's existence are likely already to have some currency.’²⁵ Even for Versnel, it seems, atheism is atheism is atheism. Versnel is concerned to defend the validity of using etic categories in the study of Greek religion. Even if we were to bracket ‘atheism’ with ‘belief’, ‘ritual’, and other terms for which there is no ready equivalent in the Greek context (but which we deploy nonetheless in its study), we might at least concede that the atheism of any period is differently constituted and bears the imprint of the particular form of theism against which it is reacting. ‘For atheism will always be a rejection, negation, or denial of a particular form of theism.’²⁶ Should we think of a distinctively polytheistic atheism, for example? How does a polytheistic background alter what it means to deny the existence of god(s)?

An added difficulty we face is that the sources on which we are reliant for our knowledge of classical Greek atheism indulge wilfully in the kind of teleological fallacy which Febvre denounced. From Epicurus, or Theophrastus, onwards, a canonical list of hardened atheists became entrenched.²⁷ This then requires us to work against the grain of this retrospective ancient tradition. Scratch the surface of the traditions of any one of the rogues’ gallery, and their credentials begin to seem more doubtful.

Plato's picture of a generation of atheists in the Greek world, and his diagnosis of three forms of atheism, is arguably much the most solid evidence for classical atheism – even if, as Robert Parker puts it nicely, Plato is ‘too imaginative and too emotional to be a very careful reporter of other people's views’.²⁸ For Sedley, despite the possibility of some ‘McCarthyite exaggeration’,²⁹ it is clear that Plato is ‘describing an intellectual trend which he has directly encountered at Athens’. This depends, however, upon a particular interpretation of the lack of names given by Plato to the atheist prose and verse texts to which he alludes: that they were circulated anonymously for fear of impiety prosecutions, and so represent ‘the cumulative voice of an authentic atheist underground’, in Sedley's phrase.³⁰ An alternative interpretation (to which I will return) is that his picture of a generation of atheists depends upon the tendentious interpretation of very identifiable authors and works – an interpretation which would have been harder to sustain if the tracts in question were identified.

The conditions for unbelief

Another central plank of Febvre's argument was the need for suitable conditions for atheism to take root and flourish. ‘In all periods’, Febvre wrote, ‘there have been heroes, or hotheads, who were unconcerned about difficulty – and in the sixteenth century such hotheads often had heat applied to their bodies.’³¹ In the absence, however, of developments in philosophy and science that made it possible to defend unbelief – a scientific culture, an idea of truth as democratic, ‘the common property of all men’, without some sense of teamwork – an atheistic position was unable to gain any traction; one needed ‘good reasons to shake off the common yoke’.³²

Were such conditions in place in antiquity? For Sedley, atheism's precisely dated emergence in or before 420 bc.³³

required the coincidence of two independent breakthroughs: first, a physics that could account for the existence of some cosmic structures without some kind of divine causation; and second, an anthropology that could explain the origins of religious belief by analogy with the rise of nomos in other cultural domains.

His necessary conditions are intellectual, not social. What of Febvre's criteria? We might detect sufficient evidence perhaps for a sense of a scientific culture, even – if we accept the case made for an ‘atheist underground’ in Athens – arguably for a sense of teamwork. A widely disseminated concept of truth? Probably not. Except as an object of mockery and distortion in passages such as those from the Knights, it is evident that ‘atheist’ ideas do not gain wider currency but are exclusive to a narrow elite. Even in the fullest account of classical Greek atheism, that of Plato in the Laws, we might point out that atheism is seen as an aberration: in its first, purest form as a feature of youth that invariably evaporates as an individual matures, and as a disease (Pl. Leg. 888a–b). For Clinias, one of the proofs of the existence of gods is precisely ‘that all Greeks and barbarians believe there to be gods (ὅτι πάντες Ἕλληνές τε καὶ βάρβαροι νομίζουσιν εἶναι θεούς, Pl. Leg. 886a)’. (And Epicurus similarly condemned Prodicus, Diagoras, and Critias as raving madmen, likening them to Bacchants.³⁴)

We could also make the point that ‘atheistic’ positions are rarely followed through in terms of any systematic withdrawal from religious observance. (If, in some philosophical positions, beliefs are only meaningful if they provide a guide to action,³⁵ the same should apply to atheist beliefs.) We have evidence for wilful displays of contempt for traditional practice such as the kakodaimonistai, deliberately feasting on inauspicious days.³⁶ However, as Versnel has pointed out, ‘With the exception of a few isolated cases of ostentatious atheism, the explicit refusal of worship is an unknown phenomenon in the archaic and classical periods.’³⁷ In short then, one possible Febvre-inspired response would be to concede that atheism was in some sense thinkable but to deny that it had sufficiently wide currency to be significant.

Atheism was also, in the phrase of Stephen Greenblatt (again writing in the context of a later period), ‘almost always thinkable only as the thought of another’.³⁸ In Febvre's characterization of the sixteenth century, accusations of atheism were slung in all directions, with each scandalized by the atheism of their rivals:³⁹ sometimes these accusations were a demonstration of learning, on other occasions they were ‘no more than ‘a kind of obscenity meant to cause a shudder in an audience of the faithful’.⁴⁰

We arguably sense something similar to the world described by Febvre from Plato's Apology and Socrates’ suggestion that his critics merely pump out stock accusations because they have been riled by his pupils,⁴¹ or more clearly perhaps when the author of the Hippocratic On the Sacred Disease makes the gambit of suggesting that the quacks who prescribe purifications and promise the impossible with their spells cannot themselves believe that the gods exist, and that they are impious (Hipp. Morb. sacr. 3–4).⁴²

Whoever by purifications and magic (περικαθαίρων…καὶ μαγεύων) can take away such a condition can also by similar techniques bring it on, so that by this argument the divine ceases to exist (ἐν τούτῳ τῷ λόγῳ τὸ θεῖον ἀπόλλυται). By saying such things and using these methods they make out that they know more and they deceive men by prescribing for them purifications and cleansings, and most of their talk turns on the divine and the daimonic (τὸ θεῖον…καὶ τὸ δαιμόνιον). They seem to me indeed to be impious (δυσσεβεῖν ἔμοιγε δοκέουσι) and neither to acknowledge the gods’ existence nor their strength (θεοὺς οὔτε εἶναι νομίζειν οὔτε ἰσχύειν οὐδὲν). And their arguments seem to me to be concerned not with piety (περὶ εὐσεβείης), as they think, but rather with impiety (περὶ δυσσεβείης), that the gods do not exist (ὡς οἱ θεοὶ οὐκ εἰσὶ), and that what they call piety and the divine is, as I shall teach, irreverent and unholy (τό τε εὐσεβὲς καὶ θεῖον αὐτῶν ἀσεβὲς καὶ ἀνόσιόν ἐστιν). For if they claim to know how to bring down the moon, to make the sun disappear, to make storm and sunshine, rain and drought, the sea impassable and the earth barren, and all such wonders, whether it be by rites or by some knowledge or practice that those who practise such things say is possible, to me at least they seem to be irreverent and either to consider that the gods do not exist or that, though they exist, they have no strength (καὶ θεοὺς οὔτε εἶναι νομίζειν οὔτ' ἐόντας ἰσχύειν οὐδὲν), and that they would not hold back from the most extreme actions (οὔτε εἴργεσθαι ἂν οὐδενὸς τῶν ἐσχάτων).

In the context of such tendentious argument, it seems self-evident, testimonia to atheism cannot simply be taken at face value. We need to distinguish between the narrow meaning of such ascriptions of atheism and their intended force, ‘the illocutionary act’.⁴³ Two ostensibly similar statements that the ‘gods do not exist’ may in fact be fundamentally different in meaning.

‘The fool hath said in his heart’

To illustrate the point, it might be useful to step aside from the classical world. When the Psalmist cites the Fool as saying ‘there is no God’ (Ps. 53.1–4; cf. Ps. 14),⁴⁴ how should this be interpreted?

1 The fool hath said in his heart, There is no God. Corrupt are they, and have done abominable iniquity: there is none that doeth good.

2 God looked down from heaven upon the children of men, to see if there were any that did understand, that did seek God.

3 Every one of them is gone back: they are altogether become filthy; there is none that doeth good, no, not one.

4 Have the workers of iniquity no knowledge? who eat up my people as they eat bread: they have not called upon God.

Even if the Fool is, of course, distanced and refuted – as with the mouthpieces of atheist phrases in Aristophanes, say – can we not still deploy such passages as evidence that atheism is in some sense thinkable? For Whitmarsh, such passages of the Psalms reveal an ‘awareness that not everyone commits equally to belief in Yahweh’.⁴⁵ For the Biblical scholar John Barton similarly, we ‘cannot rule out atheism as a possibility in the world of the Old Testament’:⁴⁶

Many Old Testament scholars have strenuously denied that anyone in ancient Israel, or perhaps in the ancient world as a whole, could conceivably have believed there was no God, and so have argued that this means practical atheism, as it seems to in Psalm 10 – God has forgotten, is not paying attention, does not care. But on the face of it the Hebrew ‘eyn ‘elohim means ‘there is a lack of god’, that is, God does not exist. Maybe this is the Psalmists’ logical inference from the fool's behaviour, rather than something the fool actually asserted, but even in that case there was at least one person, namely the psalmist, who thought that it was possible to formulate the belief that there is no god in existence. So we cannot rule out atheism as a possibility in the world of the Old Testament, and we can be sure that if it did occur, the psalmists regarded it as sinful, not simply mistaken.

Aside from the concession that the ascription of atheism might be the psalmist's inference, this all seems rather literal-minded. Rather than seeing belief as a phenomenon to be measured on a sliding scale, measuring the level of a believer's commitment, in the manner of Whitmarsh, or hesitantly ascribing a quasi-modern atheism to the world of the Hebrew Bible, in the manner of Barton, it might be fruitful to look at the line in context. In this and in two comparable settings, the surrounding passages show clearly that the denial of God's existence is associated both with bad actions and a failure to understand or to ‘seek for’ God. The implicit suggestion then (to paraphrase) is as follows: to act in this way is to turn away from God; it is tantamount to denying him, to denying his existence.

To return to our immediate Greek context, it is striking how, in Plato's characterization of three types of atheism, it is judged that ‘no one who believes the gods to exist kata nomous (‘according to the laws’) has ever willingly performed an impious act or let out a lawless word’ (θεοὺς ἡγούμενος εἶναι κατὰ νόμους οὐδεὶς πώποτε οὔτε ἔργον ἀσεβὲς εἰργάσατο ἑκὼν οὔτε λόγον ἀφῆκεν ἄνομον, Pl. Leg. 885b). (And, conversely, for the author of the Sacred Disease someone who holds the gods not to exist or not to have power would be capable of any impious action: Hipp. Morb. sacr. 4.) How very easy, then, would it be to characterize the impious man as by definition not believing in the existence of the gods? Elsewhere in the same passage, atheism is similarly envisaged as much as a lifestyle as an intellectual stance. One cause of atheism, attributed to Clinias by the Athenian, is that ‘it is solely by their incontinence in regard to pleasures and desires that their souls are impelled to their impious life’ (ἀκρατείᾳ μόνον ἡδονῶν τε καὶ ἐπιθυμιῶν ἐπὶ τὸν ἀσεβῆ βίον ὁρμᾶσθαι τὰς ψυχὰς αὐτῶν, Pl. Leg. 886a). In similar fashion, we might explain the semantic shift in the meaning of atheos in Greek, from godless or deserted by the gods to denying the existence of gods, or the conjunction of atheos with other terms (anomos (‘lawless’), acharis (‘graceless’), deinos (‘fearful’), apistos (‘faithless’), adikos (‘unjust’), prodotis (‘treacherous’) in a number of passages of Euripides.⁴⁷ Atheism here, in short, is less a philosophical position than an inference from atheist behaviour. (We might also observe that, in a reverse pattern, barbaric rites paid to a foreign deity seem to be the central factor that persuade Herodotus of that deity's merely ‘local’ status.⁴⁸)

The ascription of atheism appears to involve a similar slippage in other ways also. For Sedley, the ‘expression “not recognising (the) gods” was the favoured way of referring to what we today call atheism’; ‘the explicitly existential aspect, namely denial of the gods’ existence, is usually less emphasized than the cultic one: failure to take part in worship’.⁴⁹ It is striking, however, how often Socrates in the Apology seems to alternate between the ‘existential’ phrase (nomizein tous theous einai) and the non-existential, the mere acknowledgement of the gods (nomizein tous theous).⁵⁰ The phrase nomizein tous theous then makes no distinction between positive atheism and a mere lack of belief – or, in our terms, between agnosticism and atheism. How easy would it be again then for a Protagoras to be cast – misleadingly – as a positive atheist? (And indeed Diogenes of Oenoanda makes that very manoeuvre, on the basis that agnosticism is tantamount to denial of the gods.⁵¹) There is also a slippage between the introduction of new gods and the positive rejection of traditional deities. ‘“[K]ainotheism” is not an alternative to atheism but the form it takes’, in the words of Parker: speaking of one's own personal daimon will lead to the charge of rejecting (the existence of) traditional deities.⁵² And, as the ventriloquizing of Socrates’ accusers in the Apology makes clear, any inquiry into nature was liable to be presented as a rejection of the gods (Pl. Ap. 18b–c):

there is a certain Socrates, a wise man, who as a thinker has thoroughly investigated the heavens and everything under the earth and makes the weaker argument the stronger. These men, fellow Athenians, who spread this report against me are my terrible accusers. For those who hear such things consider that those who make such inquiries do not acknowledge the gods (οἱ γὰρ ἀκούοντες ἡγοῦνται τοὺς ταῦτα ζητοῦντας οὐδὲ θεοὺς νομίζειν).

Finally, we can see ample potential for slippage implicit in Plato's three-way definition of atheism. There are three possible explanations, for Plato, for why an individual might perform an impious act: that he does not believe in the gods’ existence; or that he does but believes also that they have no care for men (ὄντας οὐ φροντίζειν ἀνθρώπων); or that he believes that the gods are easily won over when induced by offerings and prayers (εὐπαραμυθήτους εἶναι θυσίαις τε καὶ εὐχαῖς παραγομένους). How easy is it to conflate those who believe that the gods can be ‘bribed’ or induced – that the relationship of worshipper and deity is a purely mechanical, monetary one – with those that do not believe the gods to exist? We have already indeed seen a similar manoeuvre performed by the author of the Sacred Disease: the claim that those who sell purifications or cleansings – or make other impossible claims – are in fact impious towards the gods, do not consider the gods to exist or to have any strength. The conflation of charges defies any rationalisation. (How can one believe that gods have no strength – no care for men – if you do not consider that they exist in the first place?) But, of course, the charlatans in the line of the Hippocratic author's fire themselves credit the ‘divine and the daimonic’ for their cures (τὸ θεῖον…καὶ τὸ δαιμόνιον). It is only that their exaggerated claims are tantamount to impiety; that – since they credit themselves as so powerful – they must really believe the gods to be impotent or worse.

Against the background of so much tendentious posturing – the potential for ‘slippage’ in ascriptions of atheism, to put it more neutrally – what does Plato's description of a generation of atheists really amount to? Do the Laws reveal an underground of hardened, self-identifying atheists, circulating their doctrinal opinions in samizdat pamphlets? Or is Plato, instead, gathering together a much more ragtag group of figures, each of them throwing out – in a context that was, doubtless, intensely sensitive – defensive accusations against one another? Given, especially, the absence of confirmatory evidence for an ‘atheist underground’, my money would be on something closer to the second alternative.

To return to the Fool of the Psalms, there is another reason, of course, why he cannot easily be adduced as evidence of an historical atheism in David's Israel. The Fool is a kind of cautionary figure within the grammar of the Psalms. Within the long afterlife of the Psalms, for example in the context of Anselm's Ontological Argument in the Proslogion,⁵³ the Fool emerges as an Everyman, the projection of the individual believer's own propensity for sin. The Fool's atheism is not so much a real-life alternative position – one of a spectrum of readily available religious choices – so much as a mirage generated out of a religious discourse. Something similar can be said of the unbeliever Apistos of the Epidaurian iamata inscriptions:⁵⁴

A man who had no strength in any of the fingers of his hand except one came as a suppliant to the god. Contemplating the tablets in the sanctuary he did not believe the cures and gently mocked the inscriptions. It seemed to him that, as he was playing knucklebones close by the temple and was about to throw the knucklebone, the god appeared to him, seized his hand and stretched out his fingers. When the god moved away, he seemed to bend his hand and then stretch out his fingers one by one. When he had straightened them all out, the god asked him if he still did not believe the inscriptions on the memorials in the sanctuary, and he said that he no longer disbelieved. ‘Well, since you once disbelieved things that are not incredible’, he said, ‘in future let your name be Disbeliever (Apistos)’.

For Whitmarsh, Apistos ‘[provides] precious evidence for religious skepticism in practice, as espoused by a regular, everyday Greek’.⁵⁵ Of course, there will have been those who scoffed at the possibility of miraculous intervention at Epidaurus. To that extent, religious scepticism is ‘thinkable’. That level and degree of scepticism, however, is hardly a surprise, but is inherent to belief in miraculous intervention. Importantly, moreover, such scoffing has been shaped into the form of an idealized unbeliever whose essential role is to be confounded, and to affirm belief in intervention. So far from scepticism leaking, as it were, into other areas of religious life, the sceptic is boxed into a narrative that is implicitly proselytizing. Indeed, though he may reflect the common-sense perspectives of each and every potential believer in Asclepius’ cures, Apistos is little more than a phantom generated for that single purpose.

Clearly there is a step-up from Apistos or the Fool of the Psalms to more sophisticated expressions of alternative religious viewpoints such as the Sisyphus fragment. We must still, however, ask the question, well formulated by Robert Parker in his Athenian Religion. A History, about whether instances of scepticism are ‘attacks from without’, or whether they form part of an ongoing religious discourse, one which readily accommodates scepticism.⁵⁶ The principle that the nature of the gods is unknowable – common to a range of authors – may, in some forms (and within limits), be a manifestation of a sceptical trend of thought – but at the same time it also serves to make traditional conceptions of divinity impervious to challenge, allowing the ‘compromise’, in Parker's phrase, ‘by which traditional forms of cult…are accepted by the educated as the proper way to honour a divine principle that is intellectually quite differently conceived’.⁵⁷ If the divine is unknowable, the standard response is, in fact, to proceed as if on the basis of secure knowledge.⁵⁸ Similarly, if the claim that gods do not exist, or that there is no cause for their worship, is made in the context of complaints of a lack of divine justice (that a bad action by x man or y city has gone unpunished),⁵⁹ such protests are easily answered. As soon as that bad action finds its recompense, we resume a steady course (E. fr. 913 N):

‘See, all of you who think the gods do not exist, how you are doubly in error with your poor judgement! They indeed exist, they do! (ὁρᾶθ᾿, ὅσοι νομίζετ᾿ οὐκ εἶναι θεούς, δὶς ἐξαμαρτάνοντες οὐκ εὐγνωμόνως. εἰσὶν γὰρ εἰσίν). And if anyone evil is prospering, let him enjoy the time of his prosperity; for in due course he will pay the penalty (δώσει δίκην).’

A unitary religious perspective?

One final point concerns the consistency or internal coherence of sceptical or atheist thought. As mentioned above, David Sedley deploys Protagoras’ well-known statement of ‘agnosticism’ concerning the existence of the gods (Laks-Most D10 = DK B4) to establish a ‘likely terminus ante quem for the emergence of atheism’, on the basis that Protagoras’ statement ‘implies a context in which a case for the non-existence of the gods had already been propounded’.⁶⁰ ‘Before someone can explicitly suspend judgement as to whether or not x exists, cases both for and against x's existence are likely already to have some currency’.⁶¹ Sedley's crisp argumentation is seductive, and yet – if our concern is a cultural history rather than philosophical exposition – there are perhaps dangers inherent in this very precision. Can we establish the origins of atheism in such a clear-cut analytical fashion? ‘Religious expression’, Versnel writes in a characteristic passage, ‘is mostly unreflective, very much gnomic, and with no deep interest in logical consistency.’⁶² Why do we credit atheistic expressions with a distinctive – might we say doctrinal or credal? – clarity,⁶³ or suppose that ancient writers had a neat, unitary perspective on traditional religion (that they were, alternatively, hostile, sceptical etc., or that they subscribed to traditional religious values)? Ancient testimonies make clear the confusion of ideas of individual philosophers. (Democritus, Cicero judged, ‘seems to waver in his opinion concerning the nature of the gods’.⁶⁴) They also, in many cases, emphasize that those who maintained controversial theological positions led blamelessly pious lifestyles. Despite Diagoras’ reputation for atheism, for example, Philodemus is clear that in other contexts he ‘spoke of divinity with a poet's reverence’.⁶⁵ (It is tempting to imagine that stories of how he turned to atheism because of the experience of being plagiarized are a clumsy attempt to rationalize that contradiction in biographical terms.⁶⁶)

Where – exceptionally – authors survive in more than rare fragments, this helps to reveal the complexity of religious positions that could be maintained. One might adduce here the complexities of position revealed in Thucydides’ History – which confound any characterization as straightforwardly atheist or sceptical.⁶⁷ But instead this article ends – where all early Greek history should begin and end – with Herodotus.

Passages, in particular, of Herodotus’ account of the origins of Greek religion in the Egyptian logos – his description of the Pelasgians, praying to the gods without using names, calling them theoi because they had established (thentes) all things in the cosmos, his assertion that it was Homer and Hesiod ‘who taught the Greeks the descent of the gods, and gave the gods their names, and determined their spheres and functions, and described their outward forms’ (Hdt. 2.53), or his statement at the opening of the Egyptian logos that ‘all men understand equally about such things’ [the divine] – present clear parallels with some of our fifth-century ‘atheists’: with the opening of Protagoras’ Peri Theôn, or Xenophanes’ declaration of the impossibility of clear knowledge about the gods;⁶⁸ with the presocratics’ critique of the centrality of Homer and Hesiod;⁶⁹ or with the various accounts of the origins of human religion developed by Democritus, Prodicus, and others.⁷⁰ If those passages (and, in addition, his remarks on the non-anthropomorphic form of Persian deities in the Persian ethnography of Book 1, 1.131, with their possible echo of Xenophanes) were all that survived of the Histories, Herodotus might well be taking his place alongside Critias, Diagoras, and others in Whitmarsh's Atheist Hall of Fame.⁷¹ Clearly his work overlaps with the ‘atheist’ discourse of the late fifth century. And yet there are also striking differences: Herodotus does not, in fact, follow Protagoras in supposing that the gods’ existence is open to question.⁷² The Histories, moreover, are replete with other passages that would give the lie to his simply being aligned with other sceptical or ‘atheist’ voices: passages that seem to suggest a continuing belief in the possibility of individual divinities’ intervention in the human sphere.

How should we respond? One route is to seek to argue away all those contrary instances.⁷³ Alternatively, we might seek to read the ‘sceptical’ passages of Book 2 as less profoundly sceptical: to suggest that Herodotus’ theorizing on the origins of the gods is covered, as it were, by the principle that, since the gods are unknowable, one should proceed on the basis of the traditional conceptions of the gods; that the Pelasgians’ intuition (like that of the logioi [‘wise men’] who divined the existence of Zeus in Democritus’ account⁷⁴) was a wise one; and that the thesis of the human construction of the gods is not in fact destructive of traditional religion.⁷⁵ But equally it may be – given the confusion of argument in Book 2 more than any part of the Histories – that we should settle for inconsistency. At any rate, where we have sufficient evidence to put sceptical or atheist expressions in a wider context to a significant degree, it emerges that there were borders to such scepticism. Strikingly, this is a conclusion that David Sedley himself reaches in a passing moment in discussion of Democritus. Why, Sedley asks, did Democritus not make the final move and remove the gods? One possible explanation given is the fear of consequences. ‘Part of the answer’, however,⁷⁶

no doubt lies in the ubiquity of religious experiences, such as divine epiphanies in dreams in a culture saturated in cultic practices, mythological narratives and divine images. It may never even have occurred to him that these divine images were in fact illusory.

We are back, it seems, in Lucien Febvre's bath: atheism was indeed scarcely thinkable.⁷⁷

The case for belief from atheism

I return now finally, and briefly, to the point from which I began: Versnel's argument for belief from atheism. My central conclusion here is a negative one: that within the wider argument for the legitimacy of searching for ‘belief’ in Greek religion, what evidence survives for classical Greek atheism cannot serve as a silver bullet. The narrow currency of ‘atheism’, its lack of underpinning, and its lack of clarity (to put the case minimally) are partial factors here, but I would make the further suggestion that what Sedley terms existential statements of atheism (nomizein tous theous einai; there is no God) are in reality scarcely that at all. If conceptions of atheism are always determined by their context, it seems clear that the atheism of our contemporary ‘Western’ context – and the atheism too often projected onto antiquity – is fundamentally that of the disengaged non-believer: an abstracted, existential atheism. By contrast, when the Greeks – or the Psalmist – ascribe this position to a third party, it is an inference from that individual's actions, lifestyle or (presumed) disposition – an inference predicated, in fact, on a network of theological assumptions. The focus is less on the existential status of God as on the Fool's (or our own) position in relation to god(s). There is, in short, a difference of type, not only of degree, between the atheism that could be expressed in the fifth century bc and the atheism that we envisage today.

To the extent that Versnel's argument for the validity of belief depends upon a distinction between an emic and etic perspective, and on the distinction between a low-intensity and a high-intensity belief (or between weak belief and strong⁷⁸), this argument from atheism is anyway perhaps not necessary in the first place. The (relative) absence of the ‘credal’ in Greek religion – the absence of emphasis on the importance of assent to key propositions – does not affect the case for our looking for ‘beliefs’ in the Greek world – because belief does not have to be defined in such narrow terms. And we can also deploy a reverse manoeuvre and suggest that Christian credalism is not the whole truth either, that Christian belief also is predominantly low-intensity in practice.⁷⁹ What expressions of atheism prove is that, not only did the Greeks have at their disposal a panoply of (broadly, low-intensity) beliefs about the gods and their intervention in the world, but that they also had a discourse about belief.⁸⁰ But then we perhaps knew – or should have known – that already: from the example of Apistos at Epidaurus; from the Derveni papyrus’ imploring question ‘why do they disbelieve?’ (τί ἀ[πισ]τοῦσι;);⁸¹ from Xenophon's repeatedly enjoining his readers to ‘work with god’ (e.g. Xen. Hipp. 9.8–9); or from the implicit proselytizing of Herodotean miracle stories.