In what follows I will return to the problem of the Derveni author—not in order to give him a name, but to understand who he could have been, and what his world and his background might have been. To this purpose, I will start from the crucial column xx, and then look at other hints in the text. The picture that {67|68} emerges will not be radically new, but I am confident that it will help to situate the text somewhat better.

The author regrets that the clients of these private mystery initiators did not obtain the knowledge they could have gained if they only had asked their initiator. He concedes that such a demand would have been out of place during the rites in the cities: “It is impossible to hear the ritual words and at the same time to understand them.” In this reading, then, the key to knowledge is not the ritual gestures or the objects shown, despite the focus on seeing (εἶδον), but the words spoken during the ritual, and these words are not immediately understood. In order to obtain the knowledge they are entitled to, the initiates would need not only to hear the legomena, but to have them explained: interpretation has to follow initiation. The words uttered during the rituals (prayers, hymns, invocations) have a surface meaning that the participants could easily understand during the ritual and thus think they had gained knowledge; but the deeper meaning has eluded them, because they did not ask for elucidation after the end of the ritual.

There is, of course, a text that immediately qualifies for exactly this sort of legomenon: it is the theogony of Orpheus, whose surface meaning of a rather alarming theogony the Derveni author explains in physical terms. In column vii 2, he introduces the text he is interpreting as a “hymn” ([ὕ]μνον). In the following characterization of this text as “a hymn saying things that are sound and sanctioned by divine law” ([ὕ]μνον [ὑγ]ιῆ καὶ θεμιτὰ λέγοντα), the author makes a strong claim that the following surface reading does not justify at all. The author is quick to point this out: the poem appears as “strange and a riddle for humans” (ξ[ένη τις ἡ ποίησις [κ]αὶ ἀνθρώ[ποις] αἰνι[γμ]ατώδης, vii 4): the characterization thus draws attention to the necessity for the allegorical {70|71} reading that the author advocates, in a well-established move that takes an apparent scandalous statement in the text as the starting-off point for an allegorical reading. Perhaps he justified this claim in the sentence that immediately followed: one possible way of filling the lacuna at the end of vii 2 is Janko’s [ἱερολογεῖ]το γὰρ [τῆ]ι ποιήσει, “he said sacred things in his poem”: thus the strange and offending surface meaning must be misleading. But another supplement is more likely.

The papyrus contains another case of an auctorial digression on methodology that is framed by two paragraphoi and thus singled out as interrupting the flow of the allegorization, and important enough to be underlined. It is to be found in column xiii 5 and 6:

The dichotomy Synesius uses goes back to Plato himself. Plato, however, used it differently, to describe in mystery language how a philosopher could arrive at truth. As Christoph Riedweg showed long ago, the Platonic tradition uses a three-step access to divine revelation as performed in the mystery cults as a metaphor for the philosopher’s trajectory to truth: after an initial purification rite, the second step consists of teaching and learning (παράδοσις, traditio), and the final step is pure vision, ἐποπτεία. This is never explicitly spelled out in Plato’s extant writings, but it is present in later Platonists, starting with his close student, Aristotle: it might well be a part of Plato’s oral teachings that did not surface in his own written dialogues, but only in the writings of his students and followers.

The same discourse resonates in yet another late fifth-century Athenian text, Aristophanes’s Clouds. Socrates is about to initiate the reluctant Strepsiades, and he lures him with a question:

The means to obtain this knowledge, at least in Socrates’ promise, is a meeting and discussion with the divinities themselves: {77|78}

There can be little doubt that the entire list that I have highlighted, from νυκτιπόλοις to μύσταις, is Heraclitean, and that the rest is Clement’s summary of the Heraclitean passage. In addition, νυκτιπόλοι is, in all likelihood, not a noun but an adjective that is applied to the following nouns: thus, four groups of people are active during the night, three groups of performers of Bacchic mystery rites (the βάκχοι being perhaps a special group among the initiates {79|80} of Dionysus, more exalted than the simple “bearers of narthex,” the λῆναι and μύσται), and a group of religious specialists, the μάγοι, who must belong to the same contemporary Ionian world and, more narrowly, to the same religious background as the three Bacchic groups. Like them, the term μάγος is not intrinsically negative, but is as descriptive as the other three are, despite its Persian origin—but unlike those the term is not Greek but Persian. We thus learn (and I confess my own surprise) of priests in Bacchic mystery cults in Heraclitus’ world during the Persian occupation who were called or called themselves μάγοι.

This forces me to rethink the semantics of the three occurrences of μάγος in the fifth century outside Herodotus’ ethnographical text. Two cases come from tragedy, Sophocles Oedipus Rex 384–386 and Euripides Orestes 1494. In the Orestes, a servant describes the sudden disappearance of Helen, “through sorcery, the art of the magoi, or the secret attack of the gods” (ἤτοι φαρμάκοις | ἢ μάγων τέχναις ἢ θεῶν κλοπαῖς). This passage is fully descriptive; the μάγοι wield supernatural power which humans cannot resist; there is nothing intrinsically negative in the term. The Sophoclean passage is less easy to gauge. Oedipus is angry at Tiresias and abuses him as a false prophet, bought by Creon—“this wizard hatcher of plots, this crafty beggar who has sight only when it comes to profit, but in his art is blind” (μάγον τοιόνδε μηχανορράφον, | δόλιον ἀγύρτην, ὅστις ἐν τοῖς κέρδεσιν | μόνον δέδορκε, τὴν τέχνην δ’ ἔφυ τυφλός). Hugh Lloyd-Jones’ translation “wizard” begs the question: it might well be that the term is descriptive only and gets its negative force from the adjectives. At any rate, even if negative in itself, the English “wizard” is somewhat beyond the point: Tiresias is a seer, and it is his divinatory profession that is expressed by μάγος. Ἀγύρτης, the second noun, is in itself descriptive as well: he is the priest who “collects contributions.” Such priests, however, never belonged to established polis cults but to marginal and often foreign cults; unlike the polis priest, they were itinerant professionals, not citizens serving their community. The contrast with the citizen priesthood must have been enough to give the term a somewhat negative connotation. But again, as with the μάγος, it is the adjective that carries the main weight of Oedipus’ revilement.

Let me end with a final suggestion. Once again, Heraclitus has cropped up, this time as the main witness to the μάγοι as they appear in column vi. Furthermore, Heraclitus’s νυκτιπόλοι μάγοι, βάκχοι, λήναι and μύσται—Bacchic initiates and their initiation priests—must have taken a lively interest in the afterlife: if Heraclitus predicts to them fire after their death, he must somehow turn the tables on them. Clement’s patchwork quote seems to make at least this clear, despite its opaqueness: Heraclitus does not just attack mystery rites, he connects them with eschatological beliefs that he himself rejects. Fire must either have been significantly absent from their post-mortem hopes, or (much more likely) it must have played a negative and punitive role in their eschatology.

But there is at least one other, presumably earlier, and certainly very Greek attestation. It appears in a strangely suggestive context: {84|85}

One cannot but wonder, on several levels. Since the Derveni Papyrus did not simply serve to kindle the pyre, its text was supposed to bring some comfort to the nobleman burnt and buried there. Sarah Iles Johnston treats some of the fears he might have had in her contribution to this volume. There might have been others: I wonder whether the dead Macedonian was as afraid of the fire in the beyond as was the Athenian Axiochus, Alcibiades’ uncle and thus his social equal. If so, he had found comfort in the rites performed by a Bacchic μάγος and a book written by one of them with an ambitious intellectual outlook, destined to convey this sort of comfort through the correct explanation of the rituals and of Orpheus’ theogonic hymn. {85|}