• 1 The various scholars who have conceived of Democritus’ gods not as real entities, but only as mere (...)
• 2  See McGibbon 1965; Eisenberger 1970; Taylor 1999 p. 211-216; Rechenauer 2013 p. 904-906. A status (...)
• 3  On the search for the reasons behind (physical and human) phenomena as the main feature of Democri (...)
• 4  Dodds 1959 p. 116-117 and p. 149-150.
• 5  Jaeger 1961 p. 281-286.
• 6  Before Plato’s Timaeus (44c) the noun πρόνοια often acquired the meaning of anticipation or previs (...)
• 7  Cf. DK 68 B 17-18 (= fr. 574 Luria 2007 [1970] [= Luria]); B 21 (= fr. 575 Luria); B 112 (= fr. 57 (...)
• 8  For this reason, the interpretations that have reduced Democritus’ theology to a mere anthropologi (...)

1So far, all the studies on the theology of the early Atomists have almost unanimously found it impossible to consider Democritus an ‘atheist’ in the literal sense of the term, viz. a radical denier of the existence of the gods.1 Most recently, it has been shown that Democritus had specific theological interests, i.e. that he sought to determine the real nature of the gods,2 from both an ontological and causal point of view.3 The perspective of this new research represents an important historico-philosophical shift in the study of the thought of the early Atomists. It should suffice to recall that even Eric R. Dodds limited his considerations on Democritus’ ‘irrational’ to some notes on his concept of ἐνθουσιασμός and his theory of dreams.4 Werner Jaeger, in his turn, devoted only a few pages to Democritus in his major work on the theology of the first Greek thinkers. As a matter of fact, he was still convinced that, unlike Anaxagoras and Diogenes of Apollonia, Democritus had not established his own theology, such as could be systematically framed within his naturalistic philosophy.5 By contrast, in accordance with the new trend mentioned above, we are now allowed to consider Democritus’ inquiries on the relation between gods and men — ranging from the problem of ‘providence’6 to that of poetic inspiration7 and the origin and development of religious ideas in the history of mankind — as ‘theological’ interests.8 Therefore, the causal approach to these tricky questions cannot overshadow the scientific-descriptive approach. On the other hand, the sources at our disposal, while not easy to interpret, seem to attest that Democritus considered the gods real beings, albeit mortal beings who are not concerned with the vicissitudes of mankind.

• 11  Cicero, ND, II [5] 13-15 (= SVF I 528 [I]; Cf. also Cicero, ND, III [7] 16 = SVF I 528 [II]; the s (...)
• 12  David N. Sedley per litteras. On Sextus Empiricus’ use of Democritus, see, among others, Decleva C (...)

4Indeed, it is controversial to what extent the first part of this testimonium can be considered a genuine piece of evidence for Democritus’ theology. The notion of ἔννοια of the gods does not properly belong to the lexicon of the early Atomists and should rather be considered a Stoic concept, later reused in sceptical doxography. In this regard, for instance, one could refer to Cleanthes’ four reasons to explain why in animis hominum informatas deorum esse notiones.11 Anyway, one should consider that the verb ἐννοεῖν is already ubiquitous in Plato and that, provided it is rightly contextualized, Sextus’ source is, if anything, more likely to be Epicurean than Stoic. For this reason, the consideration that the noun ἔννοια is probably post-Democritean “does not disqualify the passage from being a testimonium, just from being a verbatim fragment (although there are other reasons for denying that it is a testimonium).”12 The second part of the passage in question, focused on primitive humans’ feelings of fear and frustration towards the forces of nature, deserves in my opinion careful consideration, if for no other reason but the presence of parallel or similar sources (earlier than the 2^nd century AD) that help prove in some way the degree of reliability of Sextus’ account.

• 13  Lucretius, V 1183 ff. (= partim fr. 581 [III] Luria = DK 68 A 75 [III]). See Gigante 1957, who ide (...)
• 14  Lucretius, V, 1183-1184.
• 15  fr. 12a Ross = fr. 947 Gigon (= Sextus, M. IX 20-23), which is most likely the incipit of Book 3 o (...)
• 16  Norden 2002 p. 515-516.
• 17  Luria 2007 p. 1252-1253. The Russian scholar still relied on Th. Gomperz’s old edition of De pieta (...)
• 18  On the structure and content of this papyrus, see Vassallo 2017. The passage belongs to the so-cal (...)
• 19  Henrichs 1975 p. 96-106.
• 20  Conspectus siglorum: Bücheler = Bücheler 1865 p. 529; Crönert = Crönert 1965 p. 130, n. 542; Gigan (...)
• 21  PHerc. 1428, cr. 3, col. 329 (= fr. 16) et superp. ibid., sin. sup. col. 328 (= fr. 15) collocatum (...)

5To begin with, Democritus’ religious anthropology — as sketched out by the sceptic philosopher — could be compared to Lucretius’ Epicurean observations on the origin of civilization.13 But, upon a closer inspection, in De rerum natura the list of cosmic and natural phenomena that led men to believe that the gods govern the world is more varied. Lucretius mentions not only lightning, thunder, storm, and earthquakes, but also, more generally (and at the very beginning of his list), the rational order of the movement of the stars and the regular succession of the seasons (praeterea caeli rationes ordine certo / et varia annorum cernebant tempora verti).14 Therefore, in Lucretius, the cornerstone of the psychological mechanism that leads primitive men to worship the gods seems to lie not in fear tout court, but in a more general sense of wonder towards those things that are inexplicably greater and stronger than human nature. Such things, mutatis mutandis, were the subject of what, many centuries later, Immanuel Kant considered the dynamisch-Erhabene expressed by nature. On the other hand, in Aristotle’s On Philosophy, primitive faith in the gods was said to depend, among other things (viz. the soul and its mechanisms of inspiration and divination), on the contemplation of celestial phenomena, namely the beauty of the sky and the ordered movement of the stars (ἀπὸ τῶν μετεώρων· θεασάμενοι γὰρ μεθ’ ἡμέραν μὲν ἥλιον περιπολοῦντα, νύκτωρ δὲ τὴν εὔτακτον τῶν ἄλλων ἀστέρων κίνησιν, κτλ.).15 All these facts testify that Sextus’ account on Democritus was, at least on this point, rather partial and incomplete. Eduard Norden supposed that the sceptic philosopher took from his source — Posidonius’ treatise On Gods — only the negative aspect of Democritean δόξα, that is to say, the idea that the terrible and awe-inspiring meteorological phenomena induce men to worship gods (only) out of fear. Democritus, instead, would have also referred to phenomena positive and beneficial for human life in order to explain the origin of divine worship not only on the basis of the feeling of fear, but also out of a sense of gratitude (later we will see whether and in what sense his exposition was a polemical one).16 This thesis was already accepted by Salomon Luria,17 who found a confirmation of it in the aforementioned testimonium by Philodemus. This testimonium is handed down by the treatise On Piety, especially in the section of that work devoted to the criticism of ancient theological conceptions, from the Presocratics to the early Stoics. This passage can be read in col. 329 Vassallo (= fr. 16 Schober) of PHerc. 1428, a papyrus that transmits the final columns of the large roll containing the work.18 Albert Henrichs has argued convincingly that this entire column of PHerc. 1428 refers only to Democritus and that there are good reasons to believe that it transmits not just a witness, but an actual fragment (ipsissima verba) concerning his theological conception.19 The new reconstruction which I offer here provides further food for thought on this testimonium, based on a doxographical and, above all, philosophical point of view.20 21

• 24  See Vooys 1934-1941, I, p. 34, where three occurences of the adverb, with the Latin meaning of ab (...)
• 25  On the semantic and linguistic complexity of this term, see Humbach 1967, along with the commentar (...)
• 26  See Vooys 1934-1941, I, p. 112. Also Henrichs 1975 p. 102 with n. 42 acknowledged the irrelevance (...)
• 27  DK 68 B 182 (= frs. 35 and 774 Luria) and B 264 (= fr. 604 [III] Luria).

6Up until now, only lines 21-31 of this column had been published. As mentioned above, Henrichs tried to demonstrate how those lines contained a quotation of a Democritean fragment rather than simply a paraphrase. The scholar founded his hypothesis above all on a lexical analysis of the passage, and in particular on the use of the following nouns: a) the adverb ἄνωθεν, meaning from on high or from the sky, is not attested in Philodemus, judging from the current research;24 b) the adjective διειπετής, which, in this more archaic form meaning fallen/come from Zeus (viz. from the sky), is a Philodemean hapax legomenon, therefore ascribable to the lexicon of the philosopher about whom Philodemus is speaking;25 c) the verb ἐξεργάζομαι, which Henrichs wrongly considered not to be attested in Philodemus,26 while it is undoubtedly present in two ‘ethical’ fragments of Democritus.27

• 28  Cicero, ND I [12] 29 (= DK 68 A 74 [III]): “(…) Then in what a maze of error is Democritus involve (...)
• 29  In contrast, Piergiacomi (forthcoming) sees in the πά[ν]|τα ταῦτα of lines 24-25 a probably (indi (...)

7Personally, I do not believe that these linguistic and lexical elements are enough to demonstrate that we are dealing with a word-for-word quotation of a Democritean fragment. But it is highly probable that, as is the case elsewhere in On Piety and in other Philodemean treatises, the Epicurean philosopher provides the reader with a paraphrase that intermittently quotes words originally used by the author on (or against) whom he is writing. In this specific case, as will be explained shortly, this supposition is proved by the opening lines of the Herculanean passage in question. Before analyzing those lines, we ought to observe that, as a whole, the new reconstruction of the column leads us to make a better informed comparison with the parallel text of Cicero’s On the Nature of the Gods, where the Democritean theory of divine simulacra is openly criticized.28 Philodemus’ testimonium is totally different from Cicero’s. It deals with the genesis of religious belief among men as an anthropological, and not merely epistemological, process, according to Democritus. Indeed, there is no trace of the εἴδωλα theory in it, and on the basis of our current knowledge it seems that in no Herculanean text an explicit reference to that theory is to be found.29 But, as I have mentioned above, it is rather the first (and thus far unpublished) part of the column that provides a further contribution to our understanding of Democritus’ conception of the worship of the gods. There we can read the name of Zeus and the context allows us to think that here Philodemus partially quotes Democritus’ fr. 30 DK (= fr. 580 Luria). It is the earliest evidence for this fragment in the history of ancient philosophy. In its extended version, it is only known to us through Clement of Alexandria and Eusebius of Caesarea:

• 32  Reinhardt 1912, p.  510-511, who ascribed the fragment to the Little Cosmology.
• 33  Diels & Kranz 1951-1952 II, p. 151 n. See also Jaeger 1961 p. 285-286 and Janko 2008 p. 38. As Alf (...)
• 34  The same opinion is expressed, among others, by Luria 2007 p. 722, who refers to Philodemus, Piet.(...)
• 35  Piergiacomi 2017 p. 21, who, albeit cautiously, brings the Democritean work entitled Αἰτίαι ἀέριοι (...)
• 36  I adopt this intepretation thanks to the supplement [οὐ at the end of line 10 by David N. Sedley p (...)
• 37  This thesis is a much debated one and the sources concerning the question are not so clear. Cf. Ci (...)
• 38  Cf. e.g. Aristoteles, De resp. IV 471b30 (= DK 68 A 106 = frs. 446 and 452 Luria, where other test (...)
• 39  See Norden 2002 p. 515; Gigante & Indelli 1980 p. 452, n. 4; Montano 1984 p. 456-457; Dragona-Mona (...)
• 40  DK 88 B 25 (= Sextus, M. IX 54; Aëtius I 7, 2, Dox.Graec., p. 298).

8Karl Reinhardt was the first to consider this fragment an elucidation of Democritus’ conception of the origin of religion.32 But this interpretation of the passage has never been unanimously accepted. Some scholars have used it to prove that Democritus saw the human worship of gods in a favourable light. Within this trend, there have also been those who have highlighted the positive connotation of the adjective λόγιοι, which according to Hermann Diels should be considered “die Weisen der vorhellenischen Urzeit.”33 Other scholars (the majority) have instead interpreted the fragment as a Democritean attack on the popular worship of the gods, reading the reference to λόγιοι ironically.34 Most recently, there have been some scholars who, while denying that the adjective is ironic, see the fragment as reflecting Democritus’ polemics against the conception of God/ἀήρ held not only by primitive humans, but above all by Diogenes of Apollonia: if so, Democritus would be criticizing those who uphold, philosophically as well as religiously, the divine origin of cosmic phenomena.35 Indeed, Philodemus’ account seems to me to belong to the tradition that interprets the fragment as a ‘rationalist’ attack against the worship of the gods. The Epicurean philosopher maintains that the definition of Zeus as the “king of all things” (ll. 5-6: πά[ντων | βασιλε]ύ̣ς) has been the consequence of how (according to Democritus) the ancient people began to worship the gods, and in particular Zeus. Based on the observation of natural phenomena, they concluded that Zeus did not treat humans with contempt and finally came to worship him as the source of their sustenance. Therefore, there would be a strong relationship between the human worship of the gods and the satisfaction of the material needs of men.36 The unveiling of the real origin of worship is an attack on the common (viz. unphilosophical) conception of divinity, not against the notion of divinity in itself and even less against the divine qua the ‘rational principle’ of beings and essence or substance (θεία οὐσία) endowed with the same fiery nature as the atoms that compose individual souls.37 On the other hand, such fiery (or psychical) atoms, which would seem to have been ignored by ancient people, are called by Democritus ‘divine’ and considered by him to be synonymous with the soul and intelligence.38 We can infer that the similarity between Democritus’ fr. 30 DK and Critias’ Sisyphus, posited by some scholars, is rather unlikely.39 In the Sisyphus40 the gods are portrayed as the invention of a shrewd person (πυκνός τις καὶ σοφός [...] ἀνήρ) in order to guarantee social order and citizens’ compliance with the laws. Moreover, the invention in question plays only on the fears (φόβους) of men, although Critias mentions both disadvantageous natural phenomena (e.g. lightning and thunder) and favourable ones (e.g. the beauty of the starry sky and the order of time that gives rhythm to the cosmos). This is the only position in antiquity that can be considered truly or potentially ‘atheistic’: it combines an anthropological inquiry into the origin of divine worship with the firm belief that the ‘divine’ is only the outcome of a false tale (ψευδεῖ καλύψας τὴν ἀλήθειαν λόγῳ).

• 41  Further literary references on the relation between theology and the cycle of the seasons, especia (...)
• 42  This fact seems to be all the more true in the light of the new readings made of the first part of (...)

9Democritus’ criticism against the popular conception of divinity does not follow an ‘atheistic’ position. According to this perspective, one should relate, in my opinion, Philodemus’ testimonium to Prodicus transmitted by PHerc. 1428, col. 333 Vassallo (= fr. 19 Schober) as well. This testimonium presents a structure analogous to that of the second part of the column under consideration. According to Philodemus, Democritus maintained that the worship of the gods was born from the human belief that the cycle of the seasons and all that is useful to human life derive from the sky (namely from Zeus and the other gods);41 however, he would have not inferred the non-existence of the gods from this psychological mechanism. In the same way, Philodemus maintains that Prodicus attacked exclusively the gods that are the object of human faith, not the gods in themselves: the human worship of the gods amounted to nothing but the worshipping of the fruits of the earth and of all that is useful to the material aspects of human life. Therefore, there is good reason to believe that Prodicus, as well as Democritus, identified the origin of the worship of the gods in the natural phenomena not in order to deny the existence of the gods, but rather in order to denounce the inability of ordinary men to recognize the true God and his main characteristics.42

• 43  Cf. the sources collected in fr. 472a Luria (= DK 68 A 33; A 74; A 78-79; B 10a; B 166).
• 44  As maintained by the Peripatetic doxography: Cf., for instance, Aristoteles, Div.Som. 2, 464a5 (= (...)
• 45  See A. Capizzi in Zeller & Mondolfo 1969 p. 279-282.

11Humans’ perception of the gods, according to Democritus, derive from the stream of images (εἴδωλα) that emanate from them. Covering great distances, these images move until they reach the human senses of sight and hearing. On this complex theory we have a large amount of testimonia,43 which are difficult to order and to interpret unambiguously. Some scholars have suggested that Democritus’ εἴδωλα theory, instead of having an epistemological function (i.e. to define the nature of the senses),44 only has a theological function, because it aims to define atomistic ‘demonology.’45 Also in this case, Sextus Empiricus provides some useful information, which however should be approached with a degree of caution. In Democritus’ well-known fr. 166 DK (= fr. 472a [II] Luria) we read as follows:

• 49  Cf., for instance, Themistius, in Div. Som. 464a5, p. 43, 1 (= fr. 472a [I] Luria = deest DK). For (...)
• 50  On the criticism against the Stoics in the Oenoanda Inscription, see now Gourinat 2017.
• 51 Diogenes Oenoandensis, fr. 10 (HK fr. 52 + NF I) IV 10-V 6 Smith. See Smith 1993 p. 450-451; Güreme (...)

12If we exclude the reference to φαντασία, which seems to be another re-use of an Epicurean or Stoic concept, other parallel texts49 make this testimonium reliable in its account of Democritus’ (images of) gods as extremely large and hard to destroy. The two features in question are the hallmarks of these images, which, on the other hand, also share the common features of all other images. The most complete list of these features can be found in fr. 10 Smith of Diogenes of Oenoanda, which deals with visions (φάσματα) in dreams and criticizes both the Stoics and Democritus on this topic: the Stoics for depriving the visions of something they have and for reducing them to “empty illusions of the mind” (κενὰ μὲν οὖ̣ν [σ]κ̣ι̣[α]|γραφήματα τῆς δια|νοίας);50 Democritus for endowing them with something which they do not have: 51

• 53  Cf. Aëtius (Ps.-Plut.) V 2, 1 (= DK 68 A 136 = fr. 473 Luria): Δημόκριτος τοὺς ὀνείρους γίνεσθαι κ (...)

13If — as commonly stated — Democritus’ sleep-visions are produced by images,53 we can guess that the former must have the same features as the latter. This fact is already attested in an important testimonium that we find in Plutarch’s Convivial Questions, where Democritus is said to have maintained:

• 57  Cambiano 1980 p. 442-444.
• 58  See Sassi 1978 p. 70-76. In this regard, also Diogenes of Oenoanda fr. 43 (NF 12 + 12) Smith is to (...)
• 59  Cf. Clemens, Strom. V 88, II, p. 383, 25 Stählin (= DK 68 A 79 = fr. 572 [II] Luria) and Aëtius, I (...)
• 60  See Reinhardt 1912.
• 61  Hermippus, De astrologia (Iohannes Catrares) II 1, 13, p. 33, 15 Kroll & Viereck 1895 (= DK 60 B 5 (...)
• 62  Luria 2007 p. 1247.
• 63  Cf. PCG, I 240 (= frs. 255-257 Kaibel = DK 23 B 56-57): ὁ βίος ἀνθρώποις λογισμοῦ κἀριθμοῦ δεῖται (...)
• 64  Here I am using the term “nature” according to its more extended atomistic meaning. For an in-dept (...)
• 65  Cf. frs. 288-323 Luria (with their correspondences with the Diels & Kranz collection).
• 66  Plutarchus, Quaest. conv. IV 2, 4, 665F (= DK 68 B 152 = frs. 281 and 415 [II] Luria). But one sho (...)
• 67  Hermippus, De astrologia I 16, 122, p. 26, 13 Kroll & Viereck 1895 (= DK 68 A 78 = fr. 472a [VI] L (...)
• 68  Cf. Leucippus’ fr. 2 DK (= Aëtius I 25, 4, Dox. Graec., p. 321), according to which “nothing happe (...)

14In this piece of evidence, as Giuseppe Cambiano observes, the likeness of an image to its object concerns the object in its entirety: this fact implies an absolute correspondence between the (qualitative) properties of an animated object and those of the image that it produces.57 Now, combining Plutarch’s and Diogenes’ pieces of evidence, we may infer that rationality (along with sense perception58 and reality) is one of the main features of the simulacra of the gods as well, and consequently of the gods themselves, even if irrational living beings too are reached and hit by these simulacra.59 Rationality is therefore a common quality of gods, men, and nature. In this regard, it is meaningful that at the end of the Democritean Entstehungsgeschichte of mankind sketched out by Hermippus,60 a clear connection between divine and human reason is affirmed: ἐρρύη δέ τι καὶ θειότερον εἰς αὐτόν, καθ’ ὃ νοῦ καὶ λόγου καὶ διανοίας μετέσχε καὶ τὰ ὄντα ἀνηρευνήσατο (italics mine).61 As Luria observes,62 intelligence, reason, and intellect represent the ‘divine principle’ (θεῖον) of each human being. Such a principle is strictly linked to human rational supremacy over nature through the crafts (τέχναι), as already attested in a famous fragment by Ps.-Epicharmus.63 But the human control of nature ensured by technical crafts necessarily implies the human knowledge of nature’s principles and causes. And this is made possible, in its turn, only by the common structure of φύσις and human reason. This means that natural (and cosmic) phenomena conceal rational causes that a philosopher is called to discover. 64 In this sense, and only in this sense, one could speak of Democritus’ nature (and cosmos) as something vivified by a kind of ‘divine principle.’ Without a doubt, this last point is not directly inferable from the sources at our disposal. Furthermore, it could be theoretically questioned in the light of the sources concerning Democritus’ cosmology. Indeed, Democritus does not describe the cosmogonic process as ‘divine,’ but makes it the outcome of the mechanical necessity for the atomic whirl that obeys the mere laws of physics: namely, the attraction between similar (animate and inanimate) things based on the timeless whirling motion of the atoms.65 However, there are some elements that lead us to postulate a relationship between the action of fiery (or psychic) atoms and celestial phenomena, at least insofar as Democritus seems to insist on the fiery nature of the sky and of the heavenly regions. In this regard, one could recall at least another piece of evidence by Plutarch, where the brightness of aether (παρ’ αἰθρίης [...] σέλας) is described as being διόβλητον (‘sent by Zeus’) and εὐαγές (‘bright’ , ‘clear’, ‘pure’, perhaps also ‘holy’);66 and another testimonium by Hermippus, where the gods/daemons coincide with their images and the air (qua physical principle) is said to be full of them (μεστόν τε εἶναι τὸν ἀέρα τούτων).67 If so, the principle that we have previously considered ‘divine’ would be nothing but a ‘physical’ one. In other words, it would be the starting point in any inquiry free of the superstitious prejudices of popular imagery and absolutely in agreement with the structure of the cosmos it attempts to understand. According to this perspective, we might suppose that the λόγιοι of Democritus’ fr. 30 DK are neither ‘ancient sages’ nor (ironically) ‘suckers.’ On the contrary, these λόγιοι would be more literally men provided ‘with reason’, in the same way as the gods: that is to say, those who, within their rational investigation of the causes behind physical phenomena, are not frightened by what seems to be incomprehensible and do not lose ‘faith’ in the rationality and necessity68 of the eternal order of all things that derive from those causes.