In the May 2025 issue of Greece & Rome, Thomas Harrison argued that atheism in classical Greece is something which has not merely been exaggerated by modern scholars but profoundly misunderstood. In his opinion, atheism is ‘not a timeless condition that can be diagnosed across the centuries’, and in the age of Plato this view ‘was indeed scarcely thinkable’, and even concludes in his article’s abstract, provocatively enough, that Greek atheism is ‘fundamentally, a scholarly mirage’.Footnote 1 There is no doubt that Harrison’s contribution is inspiring, packed with argument, and amounts to the most compelling challenge to the historical reality of Greek atheism to date. He revives the spirit of the French historian Lucien Febvre for ancient Greece and makes the negative case against Greek atheism with a sophisticated handle on the workings of Greek religion and its best theoretical paradigms. Yet from another point of view, Harrison’s article is quite open to challenge on two fronts: firstly, his readings of the best ancient evidence for classical atheism, above all from Plato; and secondly his methodological approach to the evidence in his choice of theory and its execution. This present article will argue on the contrary that Greek atheism did exist on the margins of fifth-and fourth-century Athens: it is not a ‘scholarly mirage’ that belongs only to the present, but was a reality, however rare. This is a response to Harrison which also expands on Plato’s ambivalent relationship with atheism and how classical atheism can be gauged as an idea and a phenomenon in a dominantly religious society.
Before proceeding, however, the meaning of ‘atheism’ in this debate should be clarified. In ordinary language use, ‘atheism’ does not necessarily mean the same thing to everyone, and in recent scholarly parlance it can often describe non-religious beliefs such as agnosticism.Footnote 2 This broader definition can have its uses, but not in this ongoing controversy about the existence of Greek atheism, which does not concern terminology but a specific world phenomenon that is only common in modern societies.Footnote 3 The question is not whether the Greeks challenged popular understandings of the gods, like Xenophanes and his denial that the gods had a human appearance (who, it should be noted, was never called an atheist in antiquity). Some of the philosophers ventured even further than Xenophanes and made surprisingly few attempts to synthesize their theologies with more traditional views. But we live at a time where secular thought is established or creeping into many countries, non-Western ones included, and historically there have never been more atheists who deny the existence of any god than the current day. This is ‘strong atheism’, and a key source of controversy is whether this type of unbelief had any following and significance in pre-modern history. More broadly, this discussion has implications for how to apply modern concepts in our descriptions of the past, and there continues to be mixed messaging around this problem in the study of atheism as a discipline. A good example comes from the Cambridge History of Atheism (2021), a capstone of scholarship with chapters in the first volume which marshal reasonable evidence for atheism in ancient Greece and India. But by the time one gets to the second volume, one contributor advances the more historicist claim that ‘there is a strong case to be made that atheism is an inherently modern outlook … [since] atheism is made possible and intelligible by certain specifically modern epistemological assumptions’.Footnote 4
Classical scholarship on ancient Greece is often sceptical about whether atheism had any currency at that time. However, anyone who has read Battling the Gods: Atheism in the Ancient World (2015) will know that Tim Whitmarsh has spiritedly attacked this thesis, presenting his book as ‘a kind of archaeology of religious skepticism’ which recovers (and speculatively reconstructs in some cases) the views of ancient freethinkers found in the archaic period and up to the age of early Christianity. In the book, Harrison’s evocative imagery of the ‘mirage’ of atheism is already used in the service of the opposite thesis, as Whitmarsh rejects the notion that atheism was ‘unthinkable’ in premodern periods as an ‘illusion’ and concludes at the end of his volume that it is not a ‘historical anomaly’ exclusive to the modern, secularized West.Footnote 5 But Harrison is not so impressed with this digging into the past. He argues that it is an anachronism to project this understanding of atheism onto the Greeks, and inadmissible to try and legitimize atheism by imposing a secular concept onto an intrinsically religious period of history, classical antiquity.Footnote 6 Similar historicist statements to these are threaded throughout Harrison’s earlier work, including his first book Divinity and History: The Religion of Herodotus (2000), where he challenges attempts of twentieth-century scholarship to modernize Herodotus, such as projecting onto him an urbane distrust of miracles or a freethinking scepticism about how much humans can understand about the gods.Footnote 7 In both studies Harrison gives credit to the argument from Lucien Febvre’s book The Problem of Unbelief in the Sixteenth Century, originally published in French in 1942, a dense and monumental work of scholarship, which argued that atheism was simply inconceivable in the time of Rabelais and historically ‘without significance’.Footnote 8
As David Wootton has rightly pointed out, Febvre did not argue that there were no individual unbelievers at all in the sixteenth century, but that their words would have been so out of step with the social omnipresence of religion and the intellectual climate of the times that they would have been irrelevant ramblers to whom no one would have seriously listened, comparable to a drunken beggar in tenth-century Trier shouting about the reality of jet planes or quantum mechanics.Footnote 9 In other words, the Febvrian thesis on sixteenth-century unbelief was not primarily existential but axiological: he argued that even if this view was accepted by the occasional ‘precursor’, any expressions of pre-modern atheism were ‘without significance’ and ‘inconsequential’ on a social and intellectual plane, and ‘did not matter, historically speaking’.Footnote 10 In his argument Harrison takes on some of this framing of atheism from The Problem of Unbelief, but it is a survival of the Febvrian tradition that I will argue should be permanently retired not merely in Classical scholarship but the study of unbelief in general.
Febvre argued for the historical insignificance of atheism in the sixteenth century on the basis of three general observations. Because Harrison repurposes all of these for classical Greece, but accords more weight to some observations over others, as will be seen, each point will be considered to gauge whether Febvre’s perspective is evidentially and theoretically appropriate for Greece. The first section will test Febvre’s contention that atheism was only a slur that could signify whatever pre-modern people wanted it to mean; the second will assess Febvre’s position that there were no intellectual tools for atheism; and the third will conclude on his thesis that atheism was ‘insignificant’ in a religious society. On the flipside, these same views, once reversed, can be used to affirm the reality of Greek atheism.
What the accusation of atheism meant in the fifth and fourth centuries
In his revised essay ‘Of Atheism’ (1625), Francis Bacon pointed out that ‘[t]he contemplative atheist is rare’, ‘yet they seem to be more than they are; for that all that impugn a received religion, or superstition, are by the adverse part branded with the name of atheists’. Be that as it may, twenty years before Febvre published The Problem of Unbelief, the French historian Abel Lefranc was encouraged to view Rabelais as ‘a modern Lucian’ because he had a reputation for atheism and free thought among other literary men. In his book Febvre pushed back against this, laying down the law that we should never take past historical statements at face value. As he put it, in the form of a Nietzschean aphorism, ‘let us be suspicious of the words of the past’.Footnote 11 Calling someone an atheist in the sixteenth century was a ‘supreme insult’, and even if Rabelais was accused of atheism, these charges assume a different character when we realize that he accused contemporaries of the same thing, and on the most specious of grounds: ‘it was used in whatever sense one wanted to give it’.Footnote 12
The rules of the game were not very different in classical antiquity, in a period when atheism was at the epicentre of a ‘moral panic’ (as recent scholarship has argued).Footnote 13 To be an unbeliever in the fifth and fourth centuries was to be implicated with immoral attitudes and behaviours such as hedonism and oath-breaking, making it a useful slur to hurl at enemies rather than an objective philosophical label for self-identification.Footnote 14 Greek intellectuals like Prodicus were parodied as pretentious freethinkers in Greek comedy and accused of atheism in later philosophical sources, and the odds that they were genuine atheists are usually low. Socrates certainly believed in gods, attended festivals, recognized the authority of the oracle at Delphi, and followed the mandate of his inner spiritual voice, but he was accused of not believing in the gods of the state and was slandered as ‘godless’ (ἄθϵος) on a number of occasions, and his impiety trial caused a series of aftershocks in early Socratic authors such as Plato and Aeschines the Socratic. Unsympathetic to Socrates’ religious views and seeking to discredit him in court, Plato’s Meletus told Socrates ‘that you do not recognize any gods at all’ (ὡς τὸ παράπαν οὐ νομίζϵις θϵούς).Footnote 15 Febvre would have surely liked Alcibiades by Aeschines the Socratic, too, which was written around the time of Plato’s Apology. In it, Socrates makes a similar, recriminating chess move to Rabelais and his literary rivals when he anticipates the charge of having an impious view, but retorts that ‘those who believe the opposite to me are the godless ones rather than myself’ (τοὺς τἀναντία ἐμοῖ δοξάζοντας ἀθέως ἔχοντας ἢ ’κϵῖνοι ἐμέ).Footnote 16 Tu quoque, in other words; the charge of atheism or ‘godlessness’ was a philosophical hot potato, and this has vanishingly little to do with modern unbelief. The ‘testimonia to atheism’, Harrison concludes, ‘cannot simply be taken at face value’.Footnote 17
Yet this material cannot be rejected out of court, either, as there are moments in early modern history when unbelievable accusations of atheism co-existed alongside real atheists. In the seventeenth century, philosophers such as Thomas Hobbes (1588–1679) and Spinoza (1632–1677) were slandered by their enemies as unbelievers, and this has led some historians of European intellectual history to claim that atheism could have only existed in this form. For Alan C. Kors, all atheists in literature before the eighteenth century were a phantasmic creation of religious authors, creating a phenomenon where there was ‘atheism without atheists’.Footnote 18 On the other hand, this rather neat claim has been criticized for sweeping the real examples of atheism under the rug. It is enough to note that not long after Hobbes published his Leviathan in 1651, an anonymous atheist in France also finished Theophrastus redivivus (1659), a Latin manifesto with over 300 pages arguing against the existence of any idea of god.Footnote 19 For this reason, although Harrison’s warnings against taking the Greek spectre of atheism for granted are salutary, this is hardly the whole story. Aside from indirect references to atheism in fifth-century drama, the best testimony for classical atheism is Plato, and assessing whether this worldview had any proponents depends on how we read him, and whether his description of atheism fits the intellectual and social conditions of his time.Footnote 20
Plato’s relationship with atheism was doubtless coloured by the trial of Socrates, along with a sense that the real atheists in Athens escaped punishment but not his pious teacher. He only namedrops atheism again in the Republic during the ‘immoralist challenge’ sketched in Book 2, and in a conventionally negative way. When Plato’s brothers prod Socrates on the question of why an unjust man should not commit injustice undetected, Adeimantus warns Socrates that he cannot drag the gods into his response, since according to prevailing Athenian social mores, the unjust man need not fear punishment from them. He frames his response as a trilemma where every possibility supports the conclusion that undetected injustice is profitable: ‘if the gods do not exist (ϵἰ μὲν μή ϵἰσιν) or they are indifferent towards human affairs, why should we worry about evading them? But if they do care for human affairs … the gods can be won over, persuaded by friendly prayers and offerings’.Footnote 21 With these sacred barriers safely out of the way, apparently, the unjust man in Athens could scheme with his friends in their conspiratorial cabals (συνωμοσίαι) and capitalize on what he has learned from the sophists (‘teachers of persuasion’, πϵιθοῦς διδάσκαλοι) to evade punishment in the Athenian lawcourts.Footnote 22 Because he aimed for relative theological neutrality in meeting the ‘immoralist challenge’, Plato nowhere answered Adeimantus’ trilemma in the Republic, but one suspects he was aware that a theological rebuttal was more than possible, and developed responses to the trilemma in parallel to his composition of the Republic. The fruits of this meditation are found in the Laws, a work composed in Plato’s old age. In that dialogue Plato reframed the views in Adeimantus’ trilemma as the three causes of impious action and speech,Footnote 23 and refuted each one in sequence with arguments of varying lengths. In the Laws Plato went so far as to recommend outlawing these views to stifle the spread of impiety (ἀσέβϵια), and in his second-best utopia to the Republic, the constitution for Magnesia in Crete, imprisonment awaited anyone who held these impious views, and execution for those who dared to reoffend.Footnote 24 Even Plato’s philosophical avatar, the Athenian Visitor, could not help but get furious at the prospect of arguing for the gods’ existence with the wilful atheists:
Well then, how can someone argue that the gods exist without getting heated (θυμῷ) about it? For surely one must get furious and hate (χαλϵπῶς φέρϵιν καὶ μισϵῖν) those men who have been to blame and are now to blame for the fact that we make these arguments.Footnote 25
Surely, all this counts against Plato as a reliable testimony for atheism, this bilious, cantankerous old man. By the time he got to the Laws, he was ‘too emotional’.Footnote 26 Thomas Harrison certainly thinks so, and argues that Plato’s Laws does not respond to any bona fide atheists but emotively infers atheism on the basis of their terrible behaviour. Just as the Fool in the Bible who ‘says in his heart, “there is no God”’ (Ps. 53.1) was not a modern atheist but a perceived wrongdoer who was assumed to resist (= deny the existence of) God, atheism in Plato’s Laws ‘is less a philosophical position than an inference from atheist behaviour’.Footnote 27
Does this reading put the report about atheism in Plato into its full context? There are, in my view, some limitations involved in comparing a thorny philosophical work such as the Laws to the simpler moral message found in the Psalms. Certainly, their concepts of atheism should not be conflated too lightly, since Plato’s polemic targeted a group of impious men, which also included outspoken philosophical atheists. The view of the non-philosopher Cleinias that atheism was caused by hedonistic perversity is in fact rejected by Plato’s avatar as too simplistic,Footnote 28 and it is interesting that Plato’s alternative explanation for atheism is intellectual (hedonistic and impious actions, for him, are more so an expression of atheism and other impious views). For the Athenian Visitor, atheism is rooted in serious philosophical ignorance (ἀμαθία), and if the atheist detained in Magnesia could be persuaded to recognize the existence of the gods, he would come closer to rehabilitation.Footnote 29 Perhaps one could reply that Plato still moralizes this ignorance, and just as the plays of Euripides represent ἀμαθία about the hidden world of the gods as a sort of hubristic insensibility, Plato’s aetiology of impiety was hardly a value-free or purely intellectual enterprise.Footnote 30 But this too is complicated at the close of Book 10 of the Laws when Plato returns to the three types of impiety introduced in the Republic and subdivides them into two further categories: those who hold an impious opinion openly or in secret, concluding the discussion with six classes of impiety.Footnote 31 The Athenian Visitor presents open atheism as a philosophical life option that does not always guarantee immorality, and he only justifies this revised division by appealing to atheists in particular, not the other groups of the impious. Is this simply explained by the logical order of his discourse, tackling atheism first, or did Plato feel compelled to make an exception for them?
As for the impiety (πϵρὶ ἀσέβϵιαν) of these [people], there are three causes (αἰτίαις μὲν τρισίν) which we have already examined; and among these there are two for each cause, which should form six categories worth differentiating for people who sin against the gods, and all of these [categories] do not require the same or similar penalty. After all, the moral character of someone who does not believe in any gods at all can naturally incline towards justice (ᾧ γὰρ ἂν μὴ νομίζοντι θϵοὺς ϵἶναι τὸ παράπαν ἦθος φύσϵι προσγένηται δίκαιον), with such people despising the very existence of wrongdoers, and unable to tolerate injustice or to commit deeds of such a kind; and they flee from the unjust among men and have affection for the just (τούς τϵ μὴ δικαίους τῶν ἀνθρώπων φϵύγουσι καὶ τοὺς δικαίους στέργουσιν).Footnote 32
This is a limited concession, but a pointed one, and this atheism has a different character to the ‘mind-reading’ unbelief in the Psalms, since Plato seems to react to a philosophical life option rather than describe an immoral person who ‘resists’ God. After this passage he adds that candid atheists are open about their unbelief in the gods, openly mock sacrifices and oaths, and, with their scornful self-assurance, are even capable of persuading others to atheism:
But for those who hold the view that ‘everything is empty of gods’ (πρὸς τῇ δόξῃ τῇ θϵῶν ἔρημα ϵἶναι πάντα) and are beset by a lack of self-control about pleasure and pain, and have a strong memory and sharp faculties of learning at their command (μνῆμαί τϵ ἰσχυραὶ καὶ μαθήσϵις ὀξϵῖαι παρῶσι), the disbelief in the gods is a condition which proves to be instinctively common to both [types of atheist]; but in terms of their harm towards other men, one is less of a threat, but the other who practises wicked deeds is more so (τὸ δὲ πλϵίω κακὰ ἐργάζοιτ’ ἄν). For one is filled with candour when he talks about the gods and sacrifices and oaths (ὁ μὲν γὰρ λόγῳ τϵ ἂν πϵρὶ θϵοὺς παρρησίας ϵἴη μϵστὸς καὶ πϵρὶ θυσίας τϵ καὶ ὅρκους), and given his mockery towards his opponents, he could quickly win over others to his cause (καὶ ὡς τῶν ἄλλων καταγϵλῶν τάχ’ ἂν ἑτέρους τοιούτους ἀπϵργάζοιτο), provided that he goes unpunished; but the other who holds the same beliefs as the former is a so-called man of genius (ϵὐφυής), and filled with deceit and treachery.Footnote 33
It is difficult to imagine the mental gymnastics required for inventing an idea, condemning and refuting that idea over a long essay, and finally admitting reluctantly that some exponents of that invented idea are better people than might have been expected, and it is more likely that Plato knew a handful of atheists in Athens and stereotyped them as two stock characters, the philosopher and the dissembler. Perhaps for Harrison, Plato’s ‘open atheist’ could still have a proselytizing function for an intellectual who was trying to strengthen his belief in the gods, and the theological discourse in the Laws was really medicine for a momentary doubter in the Academy, not for hardened atheists. But is that the same thing as reducing this character to that function and divesting the figure of any historical relevance, of any currency particular to his time? Do we turn to the Fool of the Psalms, a character ‘boxed into a narrative that is implicitly proselytizing’, and ‘a phantom generated for that single purpose’?Footnote 34 The much closer parallel to Plato’s just atheist is the proponent of the ‘might is right’ doctrine in the Gorgias, the young Athenian gentleman Callicles, whom Plato viewed with a similar mix of horror and admiration. As he did the open atheist, he praised Callicles for his candour (παρρησία),Footnote 35 and E. R. Dodds mused that Plato may have seen Callicles as his potential self, had he not met Socrates.Footnote 36 But no scholars will deny the parallels between Callicles’ Realpolitik and the wartime conditions of the late fifth century, so memorably reflected in Thucydides, and it would be a curious hermeneutic method to read the Gorgias and reduce Callicles to one protreptic function. For Plato, the force of Callicles consisted in his immediate historical vitality: he was part of the harsh cultural landscape of Athens, and his menace disappears if he was completely a literary construct.Footnote 37 There is no reason to take the philosophical atheist in Plato as different in this regard, and a better guide than Febvre for weighing up the accusations of atheism in the classical period can be found in the scholarship of Michael Hunter, who cautions against simplistic readings of the stock figure of the atheist in early modern England: ‘[a]s with more recent “moral panics”, hostility to a specific stock figure could encapsulate a wider range of anxieties, and only a sensitive appraisal can do justice to the mixture of real and exaggerated phenomena involved’.Footnote 38 Given the awkwardness of his concession that there are naturally good atheists, only after expressing his hatred and hurling a series of arguments against them, the reservations of modern scholars that Plato is ‘too emotional’ to be reliable are misplaced. Arguably, he was too emotional to be wrong.
Inventorying the Outillage Mental
More work, however, remains to be done. In his polemic, Plato sounds like he is responding to real atheism, and not merely creating it out of thin air, but this reading must be tested against the wider intellectual and social conditions of Athens. Could these circumstances have produced genuine atheists for Plato to react against? If this Febvrian criteria is met, I submit, we have reasonable grounds for suspecting that atheism was not just a spectre.
In The Problem of Unbelief, Febvre claimed that every culture had its own ‘mental toolkit’ (outillage mental) which limited what the inhabitants of that culture could believe and embody in their everyday lives, and he made a point of contrasting the tools available in the sixteenth century with those accessible to Europeans after the scientific breakthroughs of the nineteenth century: ‘not as ours’ was Febvre’s historicist mantra across the book.Footnote 39 For Febvre this toolset included language, the availability of concepts and values, and broad thought-patterns, such as the willingness to embrace contradictions and ambiguity, and none of the tools in sixteenth-century Europe were suitable for atheism. The (French) language in this period was verbose, missing scores of abstract terms required for analytic thought, ‘philosophy then was only opinions, a chaos of contradictory and wavering opinions’, and the scientist was not a worker in a rational, collaborative programme but an eccentric tinkerer: ‘a cobbler in his shop’.Footnote 40 Georges Minois came to a similar conclusion for Graeco-Roman atheism, arguing that the ancient understanding of atheism was too diffusely polemical, and their intellectual conditions for atheism insufficiently polished, for the view to realistically gain any ground.Footnote 41
Febvre was right that atheism only surfaces in specific contexts, and in many cultures the gods are often believed in relational terms, as objects of hope and expectation rather than an idea presented in a scholastic argument. When the Greenlandic-Danish anthropologist Knud Rasmussen (1879–1933) asked Aua (an Inuit elder) about his religious beliefs, Aua denied that his tribe ‘believed’ or ‘explained’ anything about the gods, but they ‘feared’ the gods and the spirits of the dead.Footnote 42 In some sense, however, fifth-century Greek thought had the basic conceptual toolkit for unbelief, and Harrison does not push against this too much. On the basis of Plato’s presentation of the atheistic worldview in the Laws, David Sedley has shown that the intellectual conditions for atheism were already in place by the late fifth century when the Greeks could access a scientific worldview which did not require any gods, alongside a conception of culture which could present the gods as a human invention.Footnote 43 None of the scientists in Socrates’ day were outspoken atheists, but natural philosophy after Anaxagoras could often take a reductive turn when it eliminated divine causation in the creation of the universe (in atomism) or downplayed it considerably (as did Anaxagoras’ Athenian pupil Archelaus).Footnote 44 The sophistic view of norms and social institutions was also essential for unbelief, and in times of institutional instability it could be threatening. In the Clouds, after Pheidippides is radicalized by Socrates’ modern education, he attacks the taboo against beating one’s own father by appealing to the arbitrariness of social norms: ‘wasn’t it just a man like you and me who set down this custom (νόμον) in the first place, and persuaded the ancients with his speech (καὶ λέγων ἔπϵιθϵ τοὺς παλαιούς)?’.Footnote 45 As a part of tradition and not a distinct domain of Greek life, religion was not immune to this prying into the cultural past, which could suggest the dangerous idea that the gods, even divinity as a generic concept, were the invention of lawgivers.Footnote 46 In addition to these intellectual ingredients, which are well known, the debating method of the sophists was surely another tool for atheism, especially as a thought pattern. Protagoras taught his pupils to argue about any topic, presenting every dispute as one between two opposing sides: there were ‘two arguments opposed to one another’ (δύο λόγοι ἀντικϵίμϵνοι ἀλλήλοις) ‘for every subject’ (πϵρὶ παντὸς πράγματος).Footnote 47 This reductive propositional attitude could easily go against the grain of how the Greeks usually believed in their gods. In Henk S. Versnel’s view, ‘in everyday religious practice individual Greek gods were practically never conceived of as powers, let alone as cultural products, but were in the first place envisaged as persons with individual characters and personalities’.Footnote 48 Protagoras, by contrast, recast the gods into an abstract notion, a proposition reduced to a binary dispute. ‘Do the gods exist or not?’, he asked, to no avail, but posing the question was an invitation for his students and readers to think in a way which was hardly natural or self-evident. There is reason to believe that his shadow looms over Plato’s atheists when they demanded convincing proofs (τϵκμήρια ἱκανά) for the existence of the gods.Footnote 49
All these intellectual developments are woven into Plato’s account of the atheistic worldview, but the source of these views proves to be more controversial. Sedley, who argues that atheists in ancient Athens had to work undercover, suggests that Plato’s Laws was targeted at a literary corpus of anonymous authors in Athens, a series of ‘prose and verse texts’ which fronted an ‘anonymous literary defence of atheism’.Footnote 50 Harrison, by contrast, replies that Plato was probably responding to a coalition of strawmen, presenting a ‘tendentious interpretation of very identifiable authors and works’ which included theistic writers such as Anaxagoras, who had already been maligned as a denier of the gods.Footnote 51 But neither of these readings, in my view, adequately account for the polemical tenor of Plato in the Laws, who was writing as an educator and a moralist. Both Sedley and Harrison seem to interpret Plato’s Athenian Visitor as addressing his literary opponents on his same level, but to my mind the polemic is triangular, with Plato’s avatar addressing young atheists directly, but in reference to the earlier sources of the atheists’ ideas. Just as he was concerned with how poets and prose authors distorted contemporary morality in the Republic, Plato was focused on the net social effect of literature in the Laws, which makes the divide between the source of an idea and its expression less rigid for his purposes.Footnote 52 He complains that when young atheists argue that the natural world is lifeless matter, they are inspired by the reasonings of earlier natural philosophers:
Well then, the arguments of such [clever] men have this effect (τόδϵ οὖν οἱ τῶν τοιούτων ἐξϵργάζονται λόγοι): whenever we present proofs that the gods exist, pointing to same phenomena, namely that the sun, moon, and the stars and the earth are divine beings and gods, those who have been convinced by clever men would reply that these same things are made of earth and stones, with no capacity whatsoever to care for human affairs.Footnote 53
Strong causal verbs get the point across. For Plato’s Athenian, the spread of philosophical atheism comes from authorities in natural philosophy whose works bring it about (ἐξϵργάζϵσθαι, ἀπϵργάζϵσθαι) that young men will accept atheism, and in turn the honest and outspoken atheists in this cohort can mockingly change (ἀπϵργάζϵσθαι) others into atheists as well.Footnote 54 Plato evidently thought there was something suggestible in these works which fostered atheism in young minds, and it only took a reader with intellectual interests and a dash of scornful charisma to take this reading out of Archelaus or the sophists before the idea could spread like a virus. I suggest that Plato is like the pastor in Victorian England who is worried that his congregation is reading Darwin; the effect of atheism is both real and imaginary, and the text is only the catalyst.Footnote 55 This would suggest that atheism was not promoted by famous Greek philosophers, as Whitmarsh proposes at various points in Battling the Gods, but it was an artefact from the wider reception of their views. If someone read a book which showed that the natural world could be explained without the gods, or a treatise which only suggested that the concept of god came from humans, it is not too hard to see how atheism could surface as a fringe counter-philosophy under specific historical conditions.
In the course of his Febvrian argument, Harrison relies more on identifying the absence of other kinds of mental tools for atheism, suggesting that the Greeks lacked fixed linguistic expressions for the concept and the rigid thought patterns needed to pin down a position like it. He rightly points out that the phrase the Greeks used for believing in the gods’ existence, θϵοὺς νομίζϵιν, could mean other things unrelated to atheism, and this could create ‘slippages’ in meaning. So θϵοὺς νομίζϵιν in the fifth century can often mean ‘acknowledging’ the gods in the sense of ‘esteeming’ them, both on the level of attitudes and practical conduct, and this use was still normal around the trial of Socrates, as shown in Lysias’ prosecution speech Against Eratosthenes (403 bce). In the speech Lysias insults one of the Thirty Tyrants who was associated with the accused, saying he ‘neither regards gods nor people’ (οὔτϵ θϵοὺς οὔτ’ ἀνθρώπους νομίζϵι), and the ‘axiological’ meaning of this phrase is clear: the Tyrant cannot deny the existence of humans so much as he thinks little of them, and does not act rightly by them.Footnote 56 On the other hand, this issue was already settled by a number of twentieth-century scholars who identified unambiguous cases in later fifth-century and fourth-century sources where the phrase only has an existential meaning, especially in our disputed text, Plato’s tenth book of the Laws.Footnote 57 We have already seen an example of this earlier, when Plato describes the just atheist who openly argues that no gods exist: ‘he who does not recognize the existence of any gods at all’ (μὴ νομίζοντι θϵοὺς ϵἶναι τὸ παράπαν).Footnote 58 Probably, this usage was not common in everyday discourse, and Harrison is right that the expression for ‘acknowledging the gods’ (θϵοὺς νομίζϵιν) was popularly used in an equivocal way, but Plato shows that the phrase could be used in an existential sense.Footnote 59
Protagoras and the debating methods of the sophists have already been highlighted above as a part of the mental toolbox for atheism, but the question of whether the theological views of Protagoras can be taken as an intellectual barometer for other new ways of thinking about the gods has been another point of difference between Sedley and Harrison. Famously, notoriously, Protagoras announced to the world that he was an agnostic:
On the gods, I am unable to know whether they exist or not (οὔθ’ ὡς ϵἰσίν, οὔθ’ ὡς οὐκ ϵἰσίν), or what they are like in regards to form; for many things get in the way of knowing — both the obscurity of this matter and the shortness of human life.Footnote 60
He could not tell whether there were gods, or not. David Sedley sees this as an indirect testimony for fifth-century atheism, pointing out that if Protagoras saw the existence and non-existence of gods as equally uncertain possibilities, this ‘implies a context in which a case for the non-existence of the gods had already been propounded’.Footnote 61 However, Thomas Harrison suggests that Protagoras’ statement need not signal an ongoing debate between two logically articulated positions. As he puts it, ‘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?’Footnote 62 Fusing together the theories of Febvre and Versnel, Harrison is suggesting here that the thought pattern of the ancient Greeks could not allow a view like atheism to take root in their minds, and when they made sceptical (or even atheistic-sounding) statements about the gods, these do not really imply a fixed commitment to a non-religious creed. After all, as Versnel put it, the Greeks usually adopted ‘a multiperspective view’ about the gods which allowed them ‘to cope with the ambiguity by shifting from one point of view to the other, depending on what the context, focus, discourse or frame of mind required’.Footnote 63 Harrison weaponizes this paradigm for thinking about Greek religion against the existence of atheism: ‘[w]hy do we credit atheistic expressions with a distinctive – might we say doctrinal or credal? – clarity, or suppose that ancient writers had a neat, unitary perspective on traditional religion?’Footnote 64
But this form of putting the question risks pushing Versnel’s theory to its absolute limits, which contrasts with the approach of the original author. In Coping with the Gods, Versnel said that the belief that the gods exist was ‘obvious’ to the Greeks and part of their ‘social knowledge’, but he allows for an occasional Diagoras to surface as a disbeliever.Footnote 65 By contrast, Harrison reasons that even lone individuals were not immune to the fluid religious mentality of the Greeks; that the closest thing to a Greek atheist would be a character in Euripides who makes a rhetorical statement that sounds atheistic, but affirms that the gods exist in other contexts. But the problems with this approach are evident if we are reliant on a paradigm designed to accommodate a wide array of cases, and not all of them. What would happen, for example, if the same thinking was applied to the religious views of Protagoras? If we tried to make the sophist from Abdera adopt the ‘multiperspective view’ of someone like Herodotus, his verdict on the gods’ existence would dissolve into a rhetorical statement, and we would expect to find other fragments where he presented their existence as uncontroversial. Not only that. When he explained his uncertainty by appealing to ‘the shortness of human life’, a hyper-Versnel would have to imagine Protagoras was being frivolous and was not saying anything definite about the human condition and the insurmountable barrier between gods and humans. I do not believe this is a good reading of Protagoras, and it seems unlikely this reading would gain purchase if more of his work survived; indeed Plato says that Protagoras avoided tackling the gods in all his writings (Tht. 162d3–e2). It is therefore questionable to chip away at the recalcitrant evidence where the Greeks do not seem to think fluidly about the gods from an internally religious point of view. Marek Winiarczyk has cautioned against the temptation to ‘absolutize certain aspects of Greek thought’ in discussions about ancient atheism, and the same caution should apply in how we read Greek religious mentalities.Footnote 66
We have inventoried the mental toolkit of Socrates’ time, and all the mental tools were there for atheism: the concepts, the language and the (Protagorean) thought patterns. But there is one last test that Plato’s invective against philosophical atheism must be put through. Were the social conditions of late fifth century and fourth century Athens favourable enough for some of the Greeks to not merely hear about, but accept philosophical atheism? Was the religious life of Athens simply too inescapable, too pervasive, to think differently?
Irrelevant, or unreal … or just real?
Febvre presented the omnipresence of religion in the sixteenth century in strong terms, concluding in his book that it was ‘the century that wanted to believe’.Footnote 67 Wherever you were, it could not be escaped: ‘Christianity was the very air one breathed’.Footnote 68 Since the 1980s, Jan Bremmer has argued that the ‘main findings’ of Febvre apply to ancient Greece, which Harrison accepts.Footnote 69 Since, for Bremmer, the Greek way of life was so involved in attending festivals, praying and sacrificing at temples, and swearing oaths to the gods, ‘atheism never developed into a popular ideology with a recognizable following’.Footnote 70 Although antiquity allowed for the rare ‘exceptional individual who dared to voice his disbelief’, these figures would have been speculative types who still participated in daily religious practice.Footnote 71 It is for this reason that Bremmer falls short of calling any ancient unbelievers ‘atheists’.Footnote 72
But the difficulties with inviting Lucien Febvre into classical Athens are manifold, and there are problems of consistency to consider. If an ‘exceptional individual’ in antiquity (such as Plato’s atheist) could disbelieve in the existence of gods, as Bremmer seems to allow, this would not justify describing atheism as a ‘scholarly mirage’, in Harrison’s terms; a rare phenomenon is not a non-existent one. The Derveni papyrus is a unique document, but not a forgery; the critics who condemned slavery as unnatural in Aristotle (Pol. 1253b14–22) were marginal voices in antiquity, but not characters in a historical novel. Nor would it be intuitive to deny that there were ancient atheists if they attended festivals, and the like; in common usage, an atheist is someone, for the Oxford English Dictionary, ‘who denies or disbelieves the existence of a God’.Footnote 73 Provided that one pluralizes that definition into ‘one who denies or disbelieves the existence of any gods’, Plato described people who fit that description to a tee. But these are preliminary observations, and more needs to be said.
The social conditions of classical Athens do not suggest that Plato or the Attic dramatists were only reacting to a spectre of atheism. If anything, they tip the scales further: the late fifth century was a tumultuous time for the civic institutions of Athens, which, like any Greek polis, did not separate religion from politics.Footnote 74 After the Peloponnesian War erupted, Athens was faced with two alarming and uncontrollable plague outbreaks (in 430–426 bce), and in their last run of the war with Sparta (in 413–404 bce) the Athenian democracy was overturned by radical oligarchs among the upper classes (twice, in 411 and 404 bce) after a series of brutal military defeats. It is generally thought, however, that the anxieties around traditional religion peaked in Athens a couple of years before the democracy was seriously under threat. Shortly before the Sicilian Expedition in 415 bce, the street statues of Hermes were vandalized overnight in Athens, and in the course of an official investigation, scores of young upper-class Athenians were accused of parodying the Eleusinian Mysteries, the most sacrosanct local cult in Attica, in their homes.Footnote 75 But for all its revelatory power, ‘Hermsgate’ was no isolated incident. As exponents of the democratic ideology of Athens, orators such as Lysias and Demosthenes were hired to discredit opponents of their clients by drawing attention to their upper-class immorality, and this can shine a light on other irreligious activities which the aristocratic Athenians got up to in the fifth and fourth centuries.Footnote 76 In a prosecution speech from the fourth century, a middle-aged Athenian named Ariston charged his adversary Conon with assault and accused him and his sons (presented as archetypically upper-class, ‘the sons of noble gentlemen’, κἀγαθῶν ἀνδρῶν υἱϵῖς) of belonging to a club called the Ithyphalloi (‘Hard-Cocked Men’), darkly hinting to the jury that they had vulgar hazing rituals.Footnote 77 Such practices seemed to have run in the family, as Ariston continued that Conon was once a member of a drinking club as a young man which feasted on stolen offerings to Hecate and discarded offerings left after assembly meetings.Footnote 78 But there were more pretentious and studied cases of impiety than this. Diagoras of Melos, a dithyrambic poet of modern tastes, was accused in 415 bce of deliberately revealing the Eleusinian Mysteries, possibly (and amusingly) disclosing them too early in a procession hymn.Footnote 79 The ill-starred poet Cinesias was accused of organizing an Evil Spirit Club (κακοδαιμονισταί) with his friends, who deliberately met on an unlucky day and travestied the norms of dining and sympotic ritual by toasting a wicked spirit (κακὸς δαίμων) after a meal.Footnote 80 All this, in turn, points to a wider social malaise in Athens. In the Laws Plato addresses the typical atheist as ‘my boy’ (ὦ παῖ), a phrase he never used for males older than their early twenties,Footnote 81 and in exactly the period Plato was writing, Isocrates complained (in 354 bce) that male youths were constantly wasting their time drinking, partying, and gambling.Footnote 82 Part of this profile can be put down to the condescension of Plato, and perhaps there were maturer atheists here and there, but the pattern is suggestive. Both in thought, and in action, the Athenian youth did not always follow the rules, and this fuelled the stereotype which was weaponized by Plato and Isocrates in their old age.
This picture corroborates the presentation of atheism in Plato’s Republic and Laws, according to which atheism was accepted in upper-class and even reactionary circles. Two caveats, however, are necessary. In Athenian public life, anyone who spoke in a law court or the popular assembly had to appear religious before their audience, and this ‘public speech’, in Robert Parker’s words, would have excluded endorsements of atheism.Footnote 83 As a second point, Athenian religion showed remarkable ‘resilience’ during the social and religious crisis of the fifth century, ruling out the possibility that atheism could have emerged as a widespread phenomenon at the time.Footnote 84 The polytheism of the Greeks was always adaptable, and a polis could always rely on different interpretations of oracles to address a social crisis, and even introduce new gods into the city to remedy any gaps in religious observance. (We can still see this happening today, too; during the worst phase of the COVID-19 pandemic in 2020–2021, the Hindu priests in Tamil Nadu did not turn to atheism, but introduced a goddess for the virus named Corona Devi).Footnote 85 These powerful countervailing forces meant that atheism was still stigmatized, inexpedient for political life, and could never have competed with a dominant religious mentality. Esther Eidinow is right, in this sense, to suggest that a simple opposition like ‘belief vs. atheism’ is not applicable to ancient Greece as a tool of analysis.Footnote 86
These opposing conditions would suggest something like the following. In scenarios where atheism did find acceptance in Athens it must have largely been behind closed doors,Footnote 87 lending a kernel of truth to the story where the philosopher Bion of Borysthenes (c. 335–c. 245 bce) was asked if gods exist (ϵἰ θϵοί ϵἰσιν) and, out of circumspection, he quoted a line of epic poetry telling the crowd in the street to go away.Footnote 88 One suspects, on this point, that atheists did not talk in homogeneous groups so much as there were settings where some atheists could be listened to, which finds a reasonably close parallel in Enlightenment France when the arch-atheist Baron D’Holbach freely discussed his unbelief during his dinner parties, but most of his company were not atheists.Footnote 89 For this reason we are probably dealing not with a ‘movement’ of atheists but a sliver of a climate of opinion among the educated upper classes of Athens, where the idea was more often talked about than believed.Footnote 90 But Adeimantus’ trilemma shows that atheism belonged to a sea of options in the city, even if it was extremely unpopular.
Despite these caveats, this overall scheme for how atheism existed in Athenian society chimes with Sedley’s idea that Greek atheism was ‘underground’. Usefully, Harrison has already objected to this characterization on two fronts. He argues, as a first point, that atheists who conceal their unbelief in public are not meaningfully atheists, suggesting, or raising the possibility, that ‘beliefs are only meaningful if they provide a guide to action’.Footnote 91 This relies on a controversial understanding of belief, but it is an understanding worth testing against our evidence. Let us consider an example where a Greek could conceal his unbelief in a god, such as Pentheus in Euripides’ Bacchae. In a conversation outside the palace, he tells two Theban elders that Dionysus is not a god, but one of them, Cadmus, tells him to quit his anti-social behaviour towards Dionysus and participate in the ‘customs’ (νόμοι) of the city: ‘and even if he is not a god, as you claim, just say that he is’ (κϵἰ μὴ γὰρ ἔστιν ὁ θϵὸς οὗτος, ὡς σὺ φήις, | παρὰ σοὶ λϵγέσθω).Footnote 92 Suffice to say, Pentheus does not go down this simple path of dissimulation, as he should have done. But let us say he did, and went to the festivals for Dionysus without any complaints, only rejecting his existence in learned company. Even then, how could Pentheus not disbelieve, in this scenario? If there is a genuine contradiction between his actions and his views, there must be some sense that Pentheus does not accept Dionysus. Long ago, it is true, the British philosopher R. B. Braithwaite argued that belief was a disposition to act: ‘I have a disposition to act as if p were true’.Footnote 93 But it seems untrue to disregard the inner workings of our experience as though we only live hypothetically (‘as if x’) and never hold definitive views about the world in our minds (‘I hold p to be true’). Surely beliefs include convictions in the indicative: ‘Dionysus is not a god’, and if not, our theory should change and not our convictions. Even then, it is not self-evident that a Braithwaite-inspired response would automatically disqualify Greek atheism as a ‘mirage’. The unbelief of Pentheus towards Dionysus still has effects on the world, which shows he has a disposition to act as if the god is unreal, even if he is constrained by his social context: his confidants know he does not believe, and perhaps Pentheus shows less fear of the Mad God’s powers during festivals. It seems, on the whole, too neat to assume that covert beliefs are not genuine or are ineffective. Some atheists in modern Islamic countries pretend to be religious but their actions are influenced by their atheism, often presenting as believers who are neglectful of ritual.Footnote 94
Harrison’s final objection arrives at a core problem with Febvre’s legacy in the historiography of atheism. For the sake of argument, Harrison suggests, let us grant that the intellectual conditions for complete unbelief were in place. Let us concede the Sedleyan idea that these views could have been diffused in clandestine settings, and that the anonymous atheists who met in Athens worked together. Even then, says Harrison, it would rule out ‘[a] widely disseminated concept of truth’ for atheism, and ‘“atheist” ideas’ would remain ‘exclusive to a narrow elite’.Footnote 95 Although Harrison denies that Greek atheism existed on an individual level, he attempts to cover his bases with the argument that ‘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’.Footnote 96
It is this italicized term, inherited from Febvre, which needs to be singled out and examined. If we concede that atheism was ‘thinkable’, how could it have been ‘insignificant’ to Plato when he tried to refute it, in the sense of ‘meaningless’ or ‘irrelevant’ from a historical point of view? Was it ‘insignificant’ in the same way for any Greeks who accepted, or entertained it? Finally, in a modern context, is it not worth investigating any ancient phenomenon that could be genuine, however rarely it was believed? For this reason, one objection to this Febvrian expression is that it is too polemically vague and suggests more than it can possibly mean. But there is a still more fundamental problem at stake with this wording. When Febvre says that atheism in the sixteenth century was ‘insignificant’, an intelligent scholar can easily read his words as an existential thesis: atheism in this period did not exist. This can be credited to the fact that language about value or meaning can easily be confused with existential statements, and this type of linguistic confusion is already apparent in Aristophanes’ Clouds. In the comedy, Socrates tells his aspiring pupil Strepsiades that Zeus is not real, emphasizing he is both non-existent and nonsensical to entertain: ‘what “Zeus”? Don’t talk nonsense, there is no Zeus’ (ποῖος Ζϵύς; οὐ μὴ ληρήσϵις· οὐδ’ ἔστι Ζϵύς, 367). Nonetheless, Strepsiades takes Socrates to mean ‘Zeus means nothing’, in the sense that he has been made worthless and redundant now that another god has taken his place:Footnote 97
Strepsiades: But who drives [the clouds] to move, if not Zeus?
Socrates: Not him, unless you mean the whirl of the aether.
Strepsiades: Whirl? I had never realized that Zeus is nothing (ὁ Ζϵὺς οὐκ ὤν), and instead of him, now Whirl is in power (βασιλϵύων).Footnote 98
It stands to reason that another comic intellectual like Socrates who heard Strepsiades could read an existential claim back into Strepsiades’ value statement, and, in much the same way, a Febvrian statement of value like ‘atheism is not significant’ can easily blur into ‘atheism is not real’. Surely, the tradition we have inherited from Febvre is too porous to allow these enormously different claims to appear to be the same. If, however, we retire this unhelpful and value-laden terminology, indebted to the polemical excesses of Febvre, we can rely on clearer wording for expressing the marginality of a viewpoint like Greek atheism, such as ‘unpopular’, ‘inconceivable for a majority’, ‘rarely accepted’.
Without this ambiguity, I submit, we are in a better position to continue this debate without overstating the fundamental difference between modern and classical atheism. And it is also easier to see, in my opinion, that Greek atheism passes Febvre’s criteria for unbelief, ticking off every box. In the period of Greek history we have been exploring, Plato appears to be critiquing a phenomenon which is neither a strawman nor an inference from immoral behaviour, and the social and intellectual conditions of his time do not cast any material doubt on his testimony. However rare it was, Greek atheism was a reality.
Open access
