A ‘great nemesis’: what is it and why is it ‘great’?

We must begin with nemesis. It is often pointed out that in Homer, nemesis is ‘indignation’ or ‘resentment’, typically felt by a human at another's misconduct (e.g. Il. 13.122), and can even be extended to mean ‘vengeance’ (Od. 22.40).Footnote ⁹ It should be stressed, though, that both gods and humans can feel this sense of indignation or anger;Footnote ¹⁰ furthermore, while rare, the divine may on occasion direct their nemesis towards humans,Footnote ¹¹ as at 1.34.1. While some readers have been inclined to translate the term at 1.34.1 as ‘punishment’, and often connect the word to a further claim that Croesus is to be punished for his hybris Footnote ¹² (a term Herodotus does not use in the passage), a consensus has developed that, in accordance with its earlier meaning in Homer, we should instead understand something like ‘indignation’ or ‘anger’.Footnote ¹³

In Herodotus, a god or gods can find human behaviour objectionable of course, but the phrasing, while similar in some ways, is different in one respect from what we find at 1.34.1. Consider 4.205: Pheretime, who has exacted a terrible revenge on her enemies, herself suffers a horrible death (being eaten alive by worms), ‘evidently because, among humans, excessive punishments are detested by the gods’.Footnote ¹⁴ Herodotus introduces a note of doubt with ὡς ἄρα,Footnote ¹⁵ as he does with at 1.34.1: evidently these moments of divine response to human error require some admission of uncertainty. Pheretime's horrible end may be understood as connected to the gods’ reaction to her behaviour, wrong because it was excessive (αἱ λίην ἰσχυραὶ τιμωρίαι),Footnote ¹⁶ and it may well be that Herodotus wants us to make that connection, but it is not stated to be so. On the other hand, at 1.34.1, nemesis is something that issues from a god and that takes hold of Croesus. Nemesis may be interpreted to imply that the god first felt indignation at Croesus’ sense of good fortune, but (again) we have to fill in the gaps: we must supply the reaction of the divine that in turn generated the nemesis that took hold of Croesus.Footnote ¹⁷ Nemesis must, to borrow Anthony Ellis’ formulation, ‘be a process or personification associated with a god, not an emotion’.Footnote ¹⁸

But if nemesis cannot be either divine ‘indignation’ or ‘punishment’, but a ‘process’, what exactly would that mean? Well, that is the point of this paper: to define what nemesis is in the Atys–Adrastus logos. To preview my main claim here: I believe that in some sense nemesis will turn out to be Adrastus himself, both as a human agent of divine anger, but also as an emblem of this ‘process’.

Another important point to consider here is the remarkable fact, routinely noted, that 1.34.1 is the only place in the entire History where Herodotus employs the term nemesis.Footnote ¹⁹ While in itself perhaps an unremarkable fact, that nemesis occurs only once in an author otherwise so preoccupied by the issue of injustice and its requital, otherwise so informed by a fundamentally moral view of the consequences of human choice and action, the extreme rarity of the term in Herodotus and his notable hedging regarding it suddenly become arresting, indeed seemingly inexplicable.Footnote ²⁰

One response is simply to group the use of nemesis at 1.34.1 with occurrences of what are taken to be rough synonyms, namely, φθόνος and τίσις,Footnote ²¹ which each occur in Herodotus with much greater frequency.Footnote ²² But this is not a satisfactory solution. These words are not in fact completely interchangeable, though they do often overlap, even in Herodotus himself.Footnote ²³ Aristotle, for one, can differentiate nemesis from phthonos, determining that ‘indignation’ is the ‘mean’ between ‘envy’ and ‘malice’ – that is, these concepts are on the same continuum of emotion, but they are not identical.Footnote ²⁴ As for tisis, it too is distinct from nemesis.Footnote ²⁵ In almost every case in Herodotus, tisis is connected to humans paying for wrongdoing to other humans, a fact that is sometimes revealed or certified by a divine communication, but that uniformly involves humans seeking vengeance.Footnote ²⁶ The one instance where tisis is not used of human action concerns the ‘retribution’ meted out to the female flying snake by her own unborn offspring for murdering the father snake during mating (3.109.2).Footnote ²⁷ Finally, that the uniqueness of nemesis at 1.34.1 is in some way meaningful is further supported by the fact that another significant and connected term, ἄτη, is only found twice in Herodotus, with both uses occurring in a related passage, a couple of chapters earlier (1.32.6).

It is important to note that it is not just any nemesis that took hold of Croesus but a great one. Nemesis is not often modified by an adjective, and when it is, the adjective often denotes speed, i.e. of the goddess Nemesis or retribution in general,Footnote ²⁸ or of the difficulty, bitterness and implacability of the goddess or identically named abstraction.Footnote ²⁹ There is a certain logic to this situation: nemesis by itself implies both large-scale wrongdoing and the reaction to it; it does not need to be described as ‘great’, because that idea is already implied in the concept. A ‘great’ nemesis seems redundant.Footnote ³⁰

But there is a set of helpful parallels for ‘a great nemesis’, one literary, the rest documentary. There is only one other place in extant Greek literature where nemesis is modified by megalē: Antimachus of Colophon (end of 5th cent.) fr. 53 Wyss = Strabo 13.1.13:Footnote ³¹

Commenting on the μεγάλη of line one, Wyss observed that, while it might seem best to dismiss the epithet as ornamental or ‘poetic’, yet, evidence was to hand that demonstrated that the pairing μεγάλη θεός was a ‘cognomen’ in the cult of the goddesses (pl.) Nemeseis.Footnote ³² I note that the passage not only has the term μεγάλη, it also features Nemesis in a distinctly cultic setting, referring to an altar and honours that celebrate her in a specific locality, and provides a byname for her, Adrasteia – obviously related to Adrastus, the name of the king who founds her cult site by the Aesepus, and of course also the name of the slayer of Atys.

The epithet ‘great’ is a common one for divinities:Footnote ³³ most memorably, from the episode in Acts, when the silversmith Demetrius whips up the Ephesians against Paul; the crowd is roused to fury and shouts μεγάλη ἡ Ἄρτεμις Ἐφεσίων – ‘Great is Artemis of the Ephesians!’ (Act. Ap. 19:28, 34).Footnote ³⁴ Nemesis, both as a god or set of gods, and as an attribute of a like divinity of justice and vengeance, in particular Adrasteia, is not infrequently identified as ‘great’, particularly in inscriptions, though these are all significantly later, but almost all do come from either Phrygia or Lydia.Footnote ³⁵

The story: repetition and stereophony

In this section I recap in detail the story of Atys and Adrastus. I highlight in particular the narrative repetitions and oddities of Herodotus’ presentation.

After the suggestion at 1.34.1 that a great nemesis from god took hold of Croesus for believing himself most blest, Herodotus states that Croesus had a dream in which his favourite son Atys was killed, struck by an iron point – his other son was ‘spoilt’ (διέφθαρτο), being unable to speak. Horrified by the dream, Croesus set about finding a wife for Atys, and he stopped sending him on military adventures abroad; he also removed all missile weapons from the men's quarters. (35) While Atys was taking in hand his marriage, a man ‘beset by misfortune’ and with unclean hands came to Sardis, identified only as Phrygian by birth (ἐὼν Φρὺξ … γενεῇ) and of the Phrygian royal family. The stranger immediately sought ritual purification from Croesus, who duly cleansed him. It is only after these formalities that Croesus asked the stranger who he was, from where in Phrygia he came, and whom had he killed. When the stranger gives his answer in response (ἀμείβετο), it is our introduction to him as well, for thus far Herodotus has suppressed his name and patronym: ‘O King, I am the son of Gordias the son of Midas. I am called Adrastus (ὀνομάζομαι … Ἄδρηστος); having murdered my own brother unintentionally I am here, driven out by my father and deprived of everything.’ Croesus recognised that Adrastus belonged to a family whose members were philoi to him and promised him aid.

Two points need stressing. First, the use of ὀνομάζομαι in the first person singular (‘I am called’) is unique in all of Herodotus (Soph. OT 8 has a similar effect); normally, names are assigned to others by the narrator, or characters use the names of others. Second, the occurrence of ἀμείβετο sets off an elaborate chain of uses of that verb, starting with Croesus’ response to Adrastus, then his discussion with his son Atys, and introducing and concluding Adrastus’ second speech to Croesus accepting the commission to accompany Atys.Footnote ³⁶ The uses of the verb by Croesus and Adrastus at 41.2 and 42.2 are linked and refer specifically to reciprocity of action.Footnote ³⁷ With both ὀνομάζομαι and the repetition of ἀμείβεσθαι linking words and actions, Herodotus brings notice to Adrastus’ name and stresses the reciprocal, give-and-take nature of the episode.

(36) In nearby Mysian Olympus, a boar had become a monstrous hazard, devastating the locals’ crops.Footnote ³⁸ The Mysians made several attempts to kill the beast on their own, but always came off the worse. Finally, messengers were sent to Croesus who brought a report of the devastation caused by the boar and the Mysians’ inability to kill it. They asked Croesus to send his son and a picked group of young men and dogs to accompany them back to Mysia.

Herodotus reports that Croesus, remembering his dream, refused to send his son, but agreed to dispatch men and dogs to get rid of the beast. (37) Although the Mysians went away satisfied, making Atys’ participation in the boar hunt unnecessary, when he became aware of what they had requested, he confronted his father with hurt and bitter recrimination: ‘I used to have the best and most noble possession – namely to distinguish myself going to war and the hunt.’ Now Croesus had shut him off from those things and had as a result lessened his standing before his fellow citizens and his new wife. He asked his father either to let him go on the hunt or (in a very oblique phrase) explain how ‘these things being done thus (ταῦτα οὕτω ποιεόμενα) were better for him’, that is, the measures Croesus took to protect Atys. The phrase has been described as ‘colorless’,Footnote ³⁹ and perhaps by itself is unremarkable. However, it is the first of several such colourless expressions in the passage – indeed, we will encounter the phrase ποιέω ταῦτα in Croesus’ response to Atys that immediately follows. Rather than ‘colorless’, I believe that the phrase ταῦτα οὕτω ποιεόμενα is deliberately vague and invites the reader to think about the significance of the events as they unfold, as generalised outcomes of any potential action.Footnote ⁴⁰

(38) Dismissing his son's suggestion that he saw cowardice or other blemish in his character, Croesus defended the ‘things I am doing’ (ποιέω ταῦτα) as precautionary against the catastrophe warned about in his dream.Footnote ⁴¹ He also remarked that he considered Atys his only son, reckoning his other son non-existent because he was ‘spoilt’ (διεφθαρμένον), a point already made by the narrator at the start.

(39) Atys forgave his father his caution but maintained that certain details of the dream had escaped him: boars do not possess iron points; if the dream had said that he would meet his end thanks to a tooth or something else suitable to a boar, then Croesus was right to do the things he was doing (again, that oblique phrase: χρῆν δή σε ποιέειν τὰ ποιέεις, 39.2).

Persuaded by Atys, Croesus sent ‘for the Phrygian Adrastus’ (τὸν Φρύγα Ἄδρηστον, 41.1), the ethnic standing out because it is a detail we already know, indeed one we were told just a few chapters before (35.1).Footnote ⁴² After his arrival, Croesus reminded him of the salient facts: ‘since you ought to repay me with benefits, I who previously did you benefits (ὀφείλεις γὰρ ἐμεῦ προποιήσαντος χρηστὰ ἐς σὲ χρηστοῖσί με ἀμείβεσθαι, 41.2), I ask you to be a protector of my son as he sets out on the hunt’.

(42) Just as Atys was troubled by what Croesus had said, so too Adrastus was somewhat resistant: ‘otherwise I would not go to such a trial of strength, for neither is it appropriate (οὔτε … οἰκός ἐστι) for one who has experienced such misfortune to go among comrades who are prospering, nor do I have the desire, but I would have for many reasons held myself back’. However, since Croesus was so insistent and Adrastus felt the need to honour his request – indeed, he expressed his sense of obligation in precisely the same terms Croesus had used (ὀφείλω γάρ σε ἀμείβεσθαι χρηστοῖσι, 42.2) – he was ready to do what Croesus had asked, or rather, ‘I am ready to do these things’ (ποιέειν εἰμὶ ἕτοιμος ταῦτα), echoing Croesus as well as Atys.

The hunting party set out and tracked down the boar, surrounded it and were throwing their spears at it;

That Herodotus wishes specially to mark this statement is in the first place signalled by the repeated use of δή, with the second after οὗτος ‘emphasiz[ing] the fact that a person has already been mentioned some little way back’.Footnote ⁴³ The person is, of course, ‘the stranger/guest’ (ξεῖνος) Adrastus, who not only is elaborately reintroduced, but whose name is flagged as a ‘speaking name’ with the participle καλεόμενος and is significantly postponed.Footnote ⁴⁴ Such a postponement makes the audience eagerly expect the name and endows it with great significance, prompting its re-examination in light of the immediate circumstances.Footnote ⁴⁵ Adrēstos is the ‘inescapable one’, nearly the familiar byname of Nemesis – ‘Adrasteia’, important especially in Phrygia and Lydia.Footnote ⁴⁶

Herodotus observes that in being struck by the spear, Atys ‘fulfilled the prophecy of the dream’ (43.3). When news reached Croesus that his son had been killed, confounded (συντεταραγμένος), he complained bitterly (ἐδεινολογέετο) that ‘he had killed him whom Croesus had himself cleansed of murder’ (ὅτι μιν ἀπέκτεινε τὸν αὐτὸς φόνου ἐκάθηρε). The confusion that Croesus experienced reminds us of Solon's dictum: ‘the divine is entirely jealous and disruptive’ (ταραχῶδες, 32.1).

In what follows, we are told that Croesus expressed this confusion specifically by calling bitterly (δεινῶς ἐκάλεε … ἐκάλεε; cf. ἐδεινολογέετο) upon Zeus under three different epithets that proved, in the event, to have been in his perspective false (44.2): Zeus of cleansing (katharsios), pointing to the things he had suffered at the hands of his guest/the stranger (τοῦ ξείνου, cf. 43.2 ξεῖνος); Zeus of the hearth (epistios), because in receiving the stranger (τὸν ξείνον) into his home he did not know he was nourishing a murderer of his own son; and Zeus of companions (hetaireios), because in sending Adrastus as a guard for his son, he had found in him instead a most hostile enemy.Footnote ⁴⁷

The passage concludes with a brief speech by Croesus, framed by action that is vividly narrated. After Croesus’ recriminating address to Zeus, the Lydians were present bearing the corpse of Atys, and following behind, ‘the murderer’ (ὁ φονεύς, 45.1), Adrastus. Standing before the corpse, Adrastus offered himself to Croesus, stretching out his hands, bidding the king to slaughter him on the corpse, speaking of his earlier misfortune and how he had ruined the man who had cleansed him. Despite his own, great grief, Croesus took pity on Adrastus: ‘I have, O stranger (ὦ ξεῖνε), all justice from you, since you condemn yourself.’

In every other exchange with Adrastus, Croesus had used Adrastus’ name; now he addresses him as ξεῖνε, the term with which the narrator had marked him before, but then had added ‘called Adrastus’ (καλεόμενος δὲ Ἄδρηστος, 43.2),Footnote ⁴⁸ stressing Adrastus’ relationship to Croesus as a guest-friend.Footnote ⁴⁹ Croesus’ reference to the admonitory dream as happening ‘long ago’ (καὶ πάλαι) is a noteworthy exaggeration; we are not told exactly how much time has elapsed from Croesus’ dream to Atys’ death, but it cannot have been that long. Divine intervention is inexorable and works over time, a point made clearer if the time in question has been made to seem longer than it was.

After Croesus had buried his son, the account closes:

The rhetorical momentum of the sentence builds to a conclusion marked by the ‘great, dragging polysyllables’ in the compounds συγγινωσκόμενος, βαρυσυμφορώτατος and ἐπικατασφάζει.Footnote ⁵⁰ The genealogy of Adrastus reminds us both of epitaphic language, and of Adrastus’ own words when introduced, and neatly ‘contrasts his ancestry with his fortune’; his suicide repeats the language he used when offering his life to Croesus just before (ἐπικατασφάζει, 45.3; cf. 45.1 ἐπικατασφάξαι).Footnote ⁵¹

Of the many noteworthy features of this last sentence, I wish to highlight three: how Herodotus places emphasis on Adrastus through repeated information; the puzzling expression he uses of Adrastus as the ‘murderer of the man who cleansed [him]’; and the Herodotean expression ‘of men he himself knew’ (ἀνθρώπων τῶν αὐτὸς ᾔδεε).

First of all, the elaborate patronym at 45.3 is repeated from 35.3. The proximity of the second deployment to the first is what is of interest. As with the unnecessary reminder earlier of Adrastus’ Phrygian heritage (41.1), the repetition of Ἄδρηστος δὲ ὁ Γορδίεω τοῦ Μίδεω has no true informational value. No one can have forgotten these details in a matter of about three pages. No, its purpose must be something else.Footnote ⁵²

Note that not only is Adrastus’ genealogy repeated, it is from the voice of the narrator this time. Information that was transmitted by a character in the narrative has now become an element in the reportage of the narrator. The world inhabited by the characters of Herodotus and the organising persona of his narrator have been linked. Moreover, in the same sentence we also have the repetition of οὗτος δὴ from 43.2, the crucial section renaming the stranger as Adrastus, and in which we are told that he misses the boar and strikes Atys instead. Recall that Herodotus uses this phrase to reintroduce already named persons from just before in the text. Thus, the repeated genealogy and dot-joining of the οὗτος δή phrases both serve the same function and bring the sharpest possible focus on Adrastus, even in a section already heavily marked by repetition.

Second, Herodotus’ use of anaphora with φονεύς articulated by μέν/δέ,Footnote ⁵³ linking Adrastus’ two identical crimes of involuntary murder, captures our attention, but also misleads, obscuring the detail that Adrastus killed Atys, not Croesus. Heinrich Stein and others have drawn parallels from tragic diction, and in particular a line spoken by Hecuba from Euripides’ Hecuba: σὺν ταῖσδε τὸν ἐμὸν φονέα τιμωρήσομαι ‘together with these [women of Troy] I will take vengeance on my murderer’ (882), where Hecuba refers to Polymestor, who had earlier killed her son Polydorus, as ‘my murderer’. Other passages are also adduced that seem to aim at the same expression of tragic ‘exaggeration’.Footnote ⁵⁴ But the parallel from Euripides is, in fact, a problematic reading in the manuscripts, and is in many texts corrected to τὸν ἐμῶν φονέα, following an emendation of Scaliger: ‘the murderer of my [children]’.Footnote ⁵⁵ Legrand draws our notice instead to 1.214.5, where Queen Tomyris of the Massagetai vaunts over the dead Cyrus: ‘you destroyed me, though I survive and conquer you in battle, having killed my son by deceit’ (σὺ μὲν ἐμὲ ζῶσάν τε καὶ νικῶσάν σε μάχῃ ἀπώλεσας παῖδα τὸν ἐμὸν ἑλῶν δόλῳ).Footnote ⁵⁶ In the speech of Adrastus to Croesus, in which he offered himself to be killed, Adrastus even says ‘he had destroyed his purifier’ (τὸν καθήραντα ἀπολωλεκὼς εἴη, 45.1),Footnote ⁵⁷ employing the same verb that Tomyris uses to speak of Cyrus ‘destroying’ her through murdering her son. And yet even Legrand's more apposite parallel suffers from the same problem as do the tragic examples (and thus from a further problem with the standard interpretation): all the passages adduced to explain 45.3 are statements made by characters in the text, in episodes of heightened dramatic and rhetorical moment, where exaggeration is to be expected, whereas the pairing ‘murderer of his own brother, murderer of his purifier’ at the end of the Adrastus logos comes from the voice of the narrator characterising Adrastus’ thinking. Furthermore, there is a real difference between ‘killing someone’ as a slayer or murderer (φονεύς), where actual, direct agency is denoted, and the verb (ἀπ)όλλυμι, ‘to destroy’ or ‘ruin’, an action that does not have to involve violent death.Footnote ⁵⁸

Third, the Herodotean expression, ‘recognising that he was the most desperately [lit. heavily] unfortunate of men whom he himself knew’ (συγγινωσκόμενος ἀνθρώπων εἶναι τῶν αὐτὸς ᾔδεε βαρυσυμφορώτατος). I call this phrase ‘Herodotean’ because the locution of superlative + relative clause and either οἶδα or ἴδμεν and the intensive pronoun αὐτός or an emphatically used first-person personal pronoun (ἐγώ or ἡμεῖς) is almost always used by Herodotus of his own experience.Footnote ⁵⁹ The phrase is found in Herodotus thirty-four times;Footnote ⁶⁰ of these, only four do not refer to the judgement of Herodotus as narrator: 5.49.5, 7.27.2, 9.78.2 and our passage. Of these, 1.45.3 is the only one where the thoughts of a character in the History are reported in indirect discourse; all the others are in the direct speech of a character. Arguably, in all the cases of direct speech, the characters involved are expressing perspectives aligned with those of Herodotus himself.Footnote ⁶¹

The three features of 45.3 I have isolated – Adrastus’ repeated genealogy, Adrastus as ‘murderer’ of Croesus, and the reported use by him of a favourite Herodotean locution – all find an explanation in the same phenomenon. The entire logos concerning Croesus, Atys and Adrastus abounds in repetitions,Footnote ⁶² some of which I have noted. The repeated use of the concept of ‘exchange’ or ‘requital’ (ἀμείβεσθαι) underscores this fact. John Gould has observed that there is an analogy between the exchanges of benefits between characters in the logos and the ‘give and take’ we see in their speeches. The whole passage, in other words, follows this same logic of reciprocity, in word and action.Footnote ⁶³

We can go further. Comments that are found first in the narrator's voice are later articulated by characters in the text (e.g. Croesus’ ‘ruined’ second son, 34.2, 38.2), and vice versa (the genealogy of Adrastus). Adrastus proclaims that ‘I am called Adrastus …’, identifying himself in a way that is otherwise only found in the voice of the narrator or in reported naming by others. Moreover, just before his suicide, he gains an understanding of his nature in language that Herodotus reserves almost exclusively for himself in other passages. In other words, there is an interpenetration of textual registers between characters’ views and the narrative. The effect that is created can be labelled ‘stereophonic’.Footnote ⁶⁴ The words, thoughts and feelings of characters in the story merge with the evaluations of the narrator, and the other way around.

Adrastus realises that he was most unfortunate of men; that is, on the authority of Herodotus as narrator, Adrastus possesses the very thing which Croesus notably lacked and that was exposed through his encounter with Solon: self-knowledge.Footnote ⁶⁵ Indeed, there is a significant difference between the verbs Herodotus gives to each man at the start and end of the logos: Croesus ‘believed himself’ the most fortunate of men (ἐνόμισε ἑωυτὸν εἶναι ἀνθρώπων ἁπάντων ὀλβιώτατον), whereas Adrastus ‘acknowledged’ that he was the most unfortunate of all those whom he knew (συγγινωσκόμενος ἀνθρώπων εἶναι τῶν αὐτὸς ᾔδεε βαρυσυμφορώτατος). The force of νομίζειν allows for error, either on the basis of faulty knowledge, or through reliance on knowledge that may or may not be true, but the interest in which is badly motivated. We see this situation most memorably in Candaules and his belief that his wife was the most beautiful by far of all women,Footnote ⁶⁶ a detail that explains his self-destructive infatuation with her. The middle of συγγιγνώσκειν (‘acknowledge, own, confess’) allows for no such uncertainty.Footnote ⁶⁷ And recall: Herodotus’ articulation of Adrastus’ realisation of his status as most unfortunate is itself cast in language that elsewhere Herodotus reserves almost exclusively for his own activity as the judge of what are the superlative achievements of humans or the remarkable features of the world.

Nemesis Adrasteia: Lydia, Phrygia and the confession inscriptions

That Adrastus can be linked to nemesis through a byname of the goddess Nemesis, namely Adrasteia, has been recognised for some time.Footnote ⁶⁸ But it has not been sufficiently stressed that the goddess Nemesis Adrasteia is specifically linked to Phrygia, also the homeland of Adrastus,Footnote ⁶⁹ as well as Lydia, the homeland of Croesus. The problem with the evidence establishing the connection between Adrastus and Adrasteia/Nemesis, however, is that it is all later in date than Herodotus, indeed considerably later (above, n. 35). Nonetheless, the Phrygian Mother Cybele, with whom Nemesis Adrasteia is connected,Footnote ⁷⁰ first appears on sculpted monuments in the early seventh century, and images associated with her are even older (eighth century); her earliest attestations in the Greek world date to the early sixth century on the west coast of Anatolia,Footnote ⁷¹ including Miletus, Cyme and Smyrna (Smyrna was also a major location for the cult of Nemesis in the form of the two Nemeseis).Footnote ⁷² Further, many see in the name ‘Atys’ a reflection of the name Attis, the consort of Cybele, and point to the similar ends of both figures (youths whose deaths are linked to wild boars).Footnote ⁷³ Sophocles in the Philoctetes (391–4) locates Cybele's origins in Lydia, and Euripides later in the Bacchae (78–9 and 126–9) connects her to Lydia and Phrygia.Footnote ⁷⁴ Most important for us here, Herodotus even refers to Cybele (Κυβήβη) as ‘a local god’ (ἐπιχωρίη θεός) of Sardis (5.102.1),Footnote ⁷⁵ and identifies the mountain where the source of the Hermus is located as ‘sacred to the mother Dindymene’ (1.80.1), another byname for Cybele.Footnote ⁷⁶

The name ‘Adrastus’ (Atrastas), probably Lydian in origin, is attested at Sardis from the late sixth century to the late fourth century on funerary stelae.Footnote ⁷⁷ Pausanias states (7.6.6) that a statue of one Adrastus, a Lydian, who died in the Lamian War (323/322), was dedicated in front of the sanctuary of ‘Persian Artemis’ (Anahita) by the Lydians, presumably at Sardis.Footnote ⁷⁸ It is certainly the case that another Adrastus at Sardis is known from an inscription dated to 306–303.Footnote ⁷⁹ This shows us that the name had a rich history at Sardis, and was associated with a female deity who was assimilated to the Phrygian Mother type. I should add that further inland, on the edge of the Maeander valley in Caria, at Attouda, three inscriptions of Roman Imperial date have been found, all attesting a divinity known as ‘the Goddess Mother of Adrastus’.Footnote ⁸⁰ As for ‘Atys’, it is a genuine Phrygian name;Footnote ⁸¹ but note also the Lydian king Atys in Herodotus, son of Manes and father of Lydus (1.7.3, 94.3; 7.74), as well as Atys, father of the Lydian tycoon Pythius (7.27.1).

If it is fair to say that Herodotus puts unusual stress on Adrastus’ Phrygian heritage, and furthermore if it is also fair to say that Nemesis was a god particularly associated in Antiquity with Phrygia, Lydia and Mysia, and further, that ‘Adrasteia’ was a common byname for Nemesis, then another line of inquiry opens up. In precisely the same area of the Atys–Adrastus logos – that is, Lydia and Phrygia – there developed over a period of three centuries (1st–3rd cent. AD) a relatively large (and ever-increasing) corpus of documents that scholars have called ‘confession inscriptions’ (Beichtinschriften).Footnote ⁸² The texts, often accompanied by a relief depicting the dedicator, begin with an acclamation of the local divinity (e.g. the Great Mother, Men, Apollo) and her/his dynamis, in not a few cases relying on the repeated use of the term ‘great’. Then follows the reason for the setting up of the stele, often in the form of an admission of guilt on the part of the dedicator who was ‘forced by the punishing intervention of the deity’ (illness or accident); the dedicator not infrequently asks the deity the reason for the punishment. The ‘confession’, punishment and eventual cure of the dedicator together appease the god, who through these actions demonstrates his/her power. The texts often end with a statement of faith in the deity, the circumstance of the publication of the text and, in Phrygia in particular, a warning: ‘I warn all mankind not to disdain the gods, for they (i.e. mankind) will have this stele as a warning.’Footnote ⁸³

It is significant that the goddesses Nemeseis and the concept nemesis (noun and verb) figure in an important sub-set of these texts.Footnote ⁸⁴ In one remarkable case, the concept is found three times (Petzl Reference Petzl1994, no. 59.18, 21 and 25). In the final lines of this text, it is clear that nemesis means, essentially, the ‘the report of sin, divine punishment, and relief from it’ – in other words, all the major points in a confession story.Footnote ⁸⁵ In other words, nemesis could in effect mean ‘an accounting’ or ‘a story of nemesis’. If anything like this practice existed in Herodotus’ day, with its attendant set of religious beliefs, it could be argued that his narrative of Croesus, Atys and Adrastus conforms to the same pattern, and in fact the practice might even help to explain some of the narrative's unusual features: a nemesis that is ‘great’; the insistence on the Phrygian background of Adrastus; the payment of the penalty through one's child, leading to the confusion as to whom Adrastus in fact ‘killed’; Croesus’ later admission of guilt after he had complained to the oracle at Delphi (ἑωυτοῦ … τὴν ἁμαρτάδα καὶ οὐ τοῦ θεοῦ, 1.91.6), where the key term is ἁμαρτάς, a concept that is common in allied terms in the confession inscriptions.Footnote ⁸⁶ Herodotus’ ‘confession’ narrative about the workings of Croesus’ nemesis becomes aligned with, or simply becomes, historical narrative. This situation can be paralleled by Herodotus’ citation of oracles and his adoption of oracular judgement in his own historiographic ‘voice’.Footnote ⁸⁷

There are major complications to this suggestion. The confession inscriptions are not only confined in space to Lydia and Phrygia, they are also confined in time to the Roman Imperial Period – the first to third centuries AD.Footnote ⁸⁸ Robert Parker, for one, has argued that these texts are unique in their sentiments and are not representative of more general Greek religious attitudes – indeed they reveal ‘a very un-Greek climate’.Footnote ⁸⁹ On the other hand, Henk Versnel and Angelos Chaniotis have stressed the need to locate the ‘confession’ texts within the larger orbit of the whole class of Greek expressions of ‘divine justice’ as found in, for example, aretalogical texts, prayers for justice and funerary curses.Footnote ⁹⁰ Furthermore, Chaniotis has linked the rituals referred to in the confession inscriptions with earlier Hittite practice,Footnote ⁹¹ suggesting a continuity of view that would then have to be understood as going well back in time in Anatolia, one that would have included Herodotus’ own period.

Athens: Nemesis at Rhamnous and Nemesis/Adrasteia on the Athenian stage

Many see in Herodotus a writer who needs to be connected to Athens.Footnote ⁹² On the basis of very slender ancient evidence, it is widely assumed that Herodotus read portions of his history out at Athens.Footnote ⁹³ Others will point to a presumed closeness with Sophocles, or an alignment with the views of Pericles.Footnote ⁹⁴ Further, there is the approach to the History that sees Herodotus as urging the Athenians to avoid the dangers of empire as represented by the Persians, one that assumes an Athenian readership.Footnote ⁹⁵ Without subscribing to any one of the biographical particulars of Herodotus, I do think that overall it is fair to say that he was sympathetic to Athens (if also critical) and convinced of its importance in repelling the Persian invasion of 480/479 (7.139.5), and further that he knew the city and its chōra well, even assuming that his audience was familiar enough with the Attic peninsula to compare a Scythian one to it (4.99.4).Footnote ⁹⁶

This information is important because, in addition to Asia Minor,Footnote ⁹⁷ a major cult centre for Nemesis was located in Attica at Rhamnous, where she was worshipped together with the goddess Themis.Footnote ⁹⁸ Furthermore, evidence suggests that the cult of Nemesis in Attica became associated in particular with the victory over the Persians in the Persian Wars. The earliest attestation for the cult is a bronze Corinthian-style helmet found in a cistern near the sanctuary with the following inscription (SEG 35, 24 = Petrakos (Reference Petrakos1999) no. 86, IG i³.522bis): Ῥαμνόσιοι hοι ἐν Λέμνο̣[ι ἀ]νέ[θεσαν Νεμ]έσει (‘the Rhamnousians dedicated [this, taken] on Lemnos, to Nemesis’, trans. Sekunda). It has been dated to 499/498, that is, from the time of Athens’ occupation of Lemnos during the Ionian Revolt; another argument has been made that the dedication should be put later, 475–450.Footnote ⁹⁹ Whatever the precise date, that the cult of Nemesis at Rhamnous received new and lasting attention after the Persian Wars is widely believed; indeed, it is asserted that the new focus on the goddess can be connected to the construction of other temples in the middle decades of the fifth century,Footnote ¹⁰⁰ which, taken together, helped to proclaim ‘that Athens was a city of craft and of military might by both land and sea, which had inflicted on the Persians a bitter nemesis’.Footnote ¹⁰¹

Pausanias reports a story that connects Nemesis at Rhamnous and the Persian Wars. At 1.33.2–3, he states that when the Persians invaded at Marathon, they brought with them a piece of Parian marble from which they intended to fashion a victory monument, but that was later used instead by Pheidias to make the cult statue of Nemesis at Rhamnous. The story is an absurdity of course and bears all the marks of an oral account generated from a physical memorial (a Monument-novelle);Footnote ¹⁰² but if this legend is datable to the fifth century, it would show that the repulse of the Persians and the Nemesis of Rhamnous were directly connected in the minds of the Athenians.Footnote ¹⁰³

Given the importance of the cult of Nemesis at Rhamnous, to judge by the building of her shrine in the second half of the fifth century and the surge of interest in her earlier in the same century,Footnote ¹⁰⁴ what, if anything, can we say about the mention of nemesis at Hdt. 1.34.1 that incorporates these facts? If Herodotus imagined that an important part of his audience were Athenians, and further if the emergence of the cult of Nemesis had associations for them with their defeat of the Persians as an act of revenge and as justification for their imperial status, then the mention of nemesis at this juncture in the narrative of Herodotus may have linked the Croesus story in their minds to the Persian Wars and its aftermath. This association may have been yet more noticeable to an Athenian audience given that the nemesis that seized Croesus was due to his resistance to the judgements of an Athenian wise man, Solon. It could be argued that the account of Croesus, which is only tangentially linked to the main narrative of the History (if at all; see below), is brought squarely into the larger narrative arc that Herodotus is developing concerning the conflict between the Greeks and Persians.

There is another connection that I wish to explore that may be even more relevant: Nemesis/Adrasteia as a concept in Athenian drama. Pride of place must go to a comedy entitled Nemesis by Cratinus. The play dealt with Zeus's pursuit of the goddess, during which he changed Nemesis into a goose and himself into a swan, and then raped her; Nemesis then produced an egg from which in turn would come Helen. It is an arresting thought that, if the widely accepted date of its production is correct (431),Footnote ¹⁰⁵ then contemporary with the later years of Herodotus’ life, and contemporary with the start of the Peloponnesian War, a play was produced that centred on Nemesis. I should add that Nemesis is depicted in an image concerning Helen from an amphoriskos dating to c. 430 (Heimarmene Painter, Berlin 30036).Footnote ¹⁰⁶

There are several places in both Attic comedy and tragedy where the goddess Adrasteia is mentioned in connection with the sudden need to propitiate her when a character either says something boastful or contemplates an action that would be offensive to the gods. A good example is Menander, Perikeiromene 304. Moschion, speaking to himself, expresses the hope that Myrrhine found him attractive when they met the night before (lines 301–4):

This locution, where a character seeks to forestall punishment from either the goddess Adrasteia or an allied abstraction (divine phthonos) by a propitiatory statement, is widespread in Athenian dramatic texts, as well as prose authors.Footnote ¹⁰⁷ Note also Moschion’s hedge ὡς ἔοικεν.

If we think again about the circumstances leading up to the Atys and Adrastus episode in Herodotus, Croesus is little different from the Moschion of Menander, except that the latter immediately self-corrects. Both men form exaggerated senses of themselves. We are even told by the narrator at 1.34.1 that this is the case for Croesus. Crucially, as at least a couple of parallels suggest, the person who earns the anger of Adrasteia need not boast; it can be an act or gesture tooFootnote ¹⁰⁸ – which is to say, it is the thoughts of the person in question that draw the notice of Adrasteia as revealed through word and/or deed.

Instructive in this connection is a passage from Theognis (278–81):

The bad (kakos) man does not boast but forms incorrect, even morally flawed, beliefs regarding correct behaviour (κακῶς τὰ δίκαια νομίζειν), without sufficiently worrying that he may generate the resentment of others (mortals, divine?); furthermore, the vile (deilos) man thinks that he can act recklessly and yet will still be able to manage all his affairs well – he will not ‘pay for it’ later. To judge by the arrangement of words in the first couplet, Theognis meant νομίζειν and νέμεσιν to be contrasted, located as they are in final position in consecutive lines. Hence, the Theognis passage suggests that a connection between nomizein and nemesis was an intuitive and real one. What Herodotus may have been attempting at 1.34.1 was to establish a similar connection between Croesus’ belief (ἐνόμισε) that he was most fortunate of men and the nemesis (νέμεσις) from a god that seized him.

That there is wordplay going on in both the lead-up to and the Atys–Adrastus episode itself has long been accepted, and I will return to this topic in earnest below. What I want to note here is the possibility that Herodotus prepares us for the linking of nomizein with nemesis at 1.34.1 by deploying yet another term that is related to them shortly before. The clearest reason why we know that Croesus believed himself the most fortunate of men, and the evidence for this mistaken opinion of himself found closest to 1.34.1, was Croesus’ rejection of Solon's views and his curt dismissal of the sage (he sends Solon away, ‘thinking him of no account’, 1.33). Croesus, while shocked (ἀποθωμάσας) by Solon's finding that Tellus the Athenian was the most blest of men (1.30.3), only became completely dismissive of Solon's views after the sage had proposed Cleobis and Biton as the second most blessed. It is at this point that Croesus verbally assaults Solon (1.32.1):

The key term is ἔνεμε ‘assigned’. νέμειν is related not only to νέμεσις, but also to νομίζειν/νόμος.Footnote ¹⁰⁹ It is thus tempting to speculate that Herodotus had in mind a suturing of sorts between the main acts in the drama of Croesus’ life before his invasion of Persia through words deriving from the same νεμ- root: stage (i) the formation of the belief that he was the most blest of men (ἐνόμισε ἑωυτὸν εἶναι ἀνθρώπων ἁπάντων ὀλβιώτατον); stage (ii) the assignment of this title by Solon to others (δευτερεῖα ἔνεμε τούτοισι), one that made explicit stage (i) in the form of Croesus’ violent reaction; stage (iii) the punishment visited upon Croesus for holding his mistaken view of himself (ἔλαβε ἐκ θεοῦ νέμεσις μεγάλη Κροῖσον).Footnote ¹¹⁰

This is a good place to revisit the idea of the influence specifically of tragedy on Herodotus, in particular on his construction of the Croesus logos. It is a commonplace to speak of Herodotus as having narratives that are somehow ‘tragic’ in conception.Footnote ¹¹¹ Both the Brill and Cambridge Companion to Herodotus contain chapters entitled ‘Herodotus and tragedy’.Footnote ¹¹² In particular, readers of the entire story of Croesus detect tragic elements, even a tragic architecture, in his logos, subdivided into ‘episodes’ of a drama: a historical prologue that mentions the distant crime of Gyges; the warnings of Solon; the ‘domestic tragedy’ of Atys and Adrastus, followed by the oracular warnings from Delphi at the height of Croesus’ prosperity; the defeat of Croesus by Cyrus, followed by his rescue from the pyre and Croesus’ recognition of his fate.Footnote ¹¹³ P.Oxy. 2382, containing a speech by Candaules’ wife to Gyges, has only added further to the conviction that in several narratives in Herodotus the influence of tragedy is felt, with the Atys–Adrastus logos serving as a leading example. It is perhaps tempting, therefore, to look for parallels in Attic drama, tragedy and comedy, for the use of significant naming and personification for what Herodotus has done. But while both tragedy and comedy do deploy personified abstractions on the stage, they are not carefully realised characters; they are not really capable of undergoing change or experiencing feelings as Adrastus and Atys clearly do. Dramatic abstractions are, in other words, essentially props.Footnote ¹¹⁴ While the story of Adrastus and Atys is undoubtedly ‘tragic’, it is so only in a general sense, and this affiliation does not explain their names and their characterisation.

Personification and allegory

To whatever degree it can be said that the Atys–Adrastus logos is tragic, it could be argued that Herodotus paid a significant cost in relating it in that way. Bernard Laurot has argued that the story of Croesus, once we get past the marking of him as ‘the first to initiate wrongful acts against the Greeks’ (1.5.3) and the notice that the Lydian king reduced the Greeks of Ionia to tribute-paying status (1.6.2), in fact detracts from the account of imperial ambition and warfare that is only resumed at 1.46.1.Footnote ¹¹⁵ The stories first of Gyges and then of Solon's visit to Croesus, followed by the king's domestic tragedy, have little to do with the larger narrative arc of the History: the conflict between Greeks and barbarians.Footnote ¹¹⁶ There is a disjunction between the proem and the first chapters in the Croesus logos. Although much is made of Croesus’ grief at the end of the episode, it is never mentioned in the History again. Instead, with disconcerting ease we move by means of μὲν … δέ from Croesus’ suffering to the formation of his plan to check the growth of Persia (1.46.1).Footnote ¹¹⁷

Furthermore, by making the story of Atys and Adrastus into a tragic account specifically, Herodotus has taken figures who were at least notionally historical – Croesus himself certainly so – and put them into a generalised or legendary past: the ‘heroic vagueness’ of tragedy, as Pat Easterling has styled it.Footnote ¹¹⁸ At some level it is important to recognise that this staging of the Atys–Adrastus logos seems to run counter to what a historian ought to be doing, that is, Herodotus is taking the specific and historical and making it general and legendary. Remember that in his remarks following the proem, Herodotus carved out a new space for himself in treating the past by ‘pass[ing] over the Heroic Age’ and moving on ‘to the present world’, thereby ‘consciously emphasiz[ing] the progress implied in this step’.Footnote ¹¹⁹ Yet, the very figures that he mentioned in this extensive praeteritio are precisely the subjects treated in Athenian tragedy: Io, Medea, Helen. And now he has reverted to this same narrative and thematic register in his own telling of the story of Atys and Adrastus!

A return to the very beginning of the Atys–Adrastus logos will help me to propose an interpretive framework through which to appreciate the episode (1.34.1–2).

The existential use of εἶναι at the start of section 2 (ἦσαν δὲ τῷ Κροίσῳ δύο παῖδες) in the first instance can be interpreted as part of a dative of possession.Footnote ¹²⁰ But the expression can also be regarded simply as existential. The naming of Atys and the focusing of the action on him are also sharply signalled: introduced in contrast to his dumb brother, he is also characterised as ‘first among his age-mates’; we then learn his name and are told that it was ‘this Atys’ about whom the dream was warning Croesus (οὔνομα δέ οἱ ἦν Ἄτυς. τοῦτον δὴ ὦν τὸν Ἄτυν σημαίνει).

The beginning ‘there (once) was a …’ is a strongly marked introductory statement,Footnote ¹²¹ and is associated with the language of fable.Footnote ¹²² Recall that we are at the start of a story whose pivotal moment is a boar hunt – the stuff of folk tale (cf. Meleager and the Calydonian boar; Heracles and the Erymanthian boar). When this formula introduces an individual, it is certain that the person in question, despite the simplicity of the statement and therefore possibly their insignificance, will be on the contrary significant in the subsequent narrative.Footnote ¹²³ The significant person is identified and then an elaborate recapitulation turns our attention back on him as a character of immediate interest. In Greek, this introduction is worked out very precisely: existential εἶναι ‘there (once) was a so-and-so’ (ἦν ὁ δεῖνα), sometimes with indefinite τις, followed by a recapitulating demonstrative οὗτος + δή + οὖν (Ionic ὦν) ‘now this so-and-so …’Footnote ¹²⁴ This is exactly the situation at 1.34.2, with the variation that a naming formula intrudes between the existential statement and the recapitulation. Compare the introduction to the connected, and similarly paradigmatic, story of Gyges, Candaules and Candaules’ wife: ἦν Κανδαύλης, τὸν οἱ Ἕλληνες Μυρσίλον ὀνομάζουσι … οὗτος δὴ ὦν Κανδαύλης ἠράσθη τῆς ἑωυτοῦ γυναικός (1.7.2, 8.1).

There is an important and related formula that is used to introduce gods and deified abstractions, one that is especially common in ‘Greek admonitory literature’: ἡ δέ τε παρθένος ἐστὶ Δίκη, Διὸς ἐκγεγαυῖα ‘and there is that maiden Right, daughter of Zeus …’ (Hes. Op. 256, trans. West).Footnote ¹²⁵ Of the examples Martin West cites as parallels to this line from Hesiod's Works and Days for the existential use of ἐστί, all have to do with Dike, Oath, Aidos, Nemesis, Adrasteia, or the vengeance of the gods – indeed two of the parallels are passages I have already brought up above (Antimachus fr. 53 and Alcman 1.36). I find it significant that while Herodotus employs a personified, divine abstraction at 1.34.1, elsewhere it is not the historian but characters in his History who use such abstractions in reported speech (e.g. Demaratus: peniē, 7.102.1, despotēs nomos, 104.4; Themistocles: Peitho and Ananke at 8.111.2; Andrians’ reply: Penie and Amechanie at 111.4).

But while Herodotus’ deployment of significant or ‘speaking’ names in Solon's encounter with Croesus and in the Atys–Adrastus logos has long been noted, it has seldom if ever been asked how these names were meant to work in the History. What is the purpose of such a deployment? How were they to be understood as functioning in the text?

The register that explains best the Atys–Adrastus logos in Herodotus is Greek wisdom literature, in particular Phoenix’ allegory of the Litai and Ate from the Iliad, Hesiod and the poet Solon (by which I mean, the poet independent of the character in Herodotus).Footnote ¹²⁶ Such an affiliation is signalled by the beginning of the story, discussed above: ‘now there were two sons to Croesus …’

Unlike Homer, for whom nemesis is not a god but a thing,Footnote ¹²⁷ Hesiod refers to the deity Nemesis twice: in the Theogony, among the offspring of Night (Th. 223); and in Works and Days, where she is paired with Shame (Aidos) (Op. 200). In general, personified elements of the natural world and abstractions such as Death, Strife, Battles, Victory and so forth populate much of Hesiod's world;Footnote ¹²⁸ and, just as important, they are capable of characterisation and the performance of meaningful action.

But perhaps more relevant to this discussion are the personifications that we find in the poetry of Solon.Footnote ¹²⁹ In particular, 4 West features several personifications of abstract concepts, all having to do with morality and justice. When ‘evils roam’ (στρέφεται κακά, line 23) among the people, not only do the poor suffer, but ‘public ill’ (δημόσιον κακόν, line 26) finds its way into the houses of everyone, even those with high walls and inner rooms (lines 28–9).Footnote ¹³⁰ Solon states that his heart commands him to instruct the Athenians that ‘Lawlessness (Δυσνομίη) provides the most evils to a city’ (line 31), whereas ‘Lawfulness (Εὐνομίη) renders all things well ordered and fitting’ (line 32).

In her acute reading of this poem, Fabienne Blaise has suggested that while the divine is very much understood to be ‘prior and necessary’, the gods are not to be found in the working out of the consequences of wrongdoing in the political world of humans.Footnote ¹³¹ Solon, Blaise argues, ‘re-works’ ‘Hesiodic theological fiction’ by associating activities attributed to Zeus, as especially found in the hymn to him which opens the Works and Days, with the abstractions of Dike and Eunomia.Footnote ¹³² Similarly, Renaud Gagné has argued that in Solon's 13 West, the notorious break in the poem and the discontinuities that are felt in it as a result can be explained by a change in perspective: until line 32, the poet, with the authority granted to him by the gods and specifically Mnemosyne (‘Memory’, a men-word), speaks through an ‘I’ voice, and the justice that is presented there is understood as divinely authorised and effected; from line 33 onwards, ‘we’ notice (with noos-words) the same topics, but from a partial, human perspective.Footnote ¹³³ Again, the influence of Hesiod is felt, but the workings of justice have been ‘depersonalized’, and the voice of the poet ‘desacralized’.Footnote ¹³⁴

It strikes me as distinctly possible that Herodotus was attempting a similar ‘reworking’ of what justice is with his story of Atys and Adrastus: Adrastus is Adrasteia/nemesis, and Atys is Ate/atē.Footnote ¹³⁵ The divine has been made human in the working out of justice in Herodotus’ story: divine Nemesis Adrasteia has become the all-too-human Adrastus, with his sense of obligation towards Croesus, his shame and remorse, and his poor aim. Like Solon, Herodotus has inherited a mechanism for the distribution of justice that he recognised must have a divine component – it must be supported by the gods in some way. But we also see in his History a commitment to the belief that the working out of issues relating to wrongdoing is a process that happens over time and involves human agency. This parallel, ‘overdetermined’ vision is precisely modelled for us in the Atys–Adrastus logos.

How is the Atys and Adrastus story to be read?

Herodotus’ account of Atys and Adrastus fits almost seamlessly with what has gone before, indeed so seamlessly that Hartmut Erbse has argued that the story was created by Herodotus for this very purpose.Footnote ¹³⁶ On the other hand, in terms of exemplarity, scholars have found in the entire episode of Solon and Croesus, as well as the follow-up story of Atys and Adrastus, the most definitive articulation of important Herodotean concerns. The Atys–Adrastus logos and the larger story of Croesus are both contrived and yet are central to what Herodotus is trying to say through his History. To borrow the thinking of Gould, the tale of Atys and Adrastus is one of the clearest cases in Herodotus of a ‘story [that] has a scale and a power and weight out of all proportion to its function as an explanatory link in the larger narrative’.Footnote ¹³⁷ This same tale is also unusual for the amount of stereophonic effect that it displays. Judgements of the characters in the story and those of the omniscient narrator echo with each other and do so to a degree that is unparalleled in the rest of the History.

So, what exactly is the Atys and Adrastus story? Do we have a sort of Renaissance morality play,Footnote ¹³⁸ with personifications performing actions with tragic consequences? Or is the passage to be understood as a long excursus built around an elaborate etymology – a massive pun? I hope I have already cast sufficient doubt on the first possibility. Regarding the latter one, I think ‘pun’ is not a useful concept in helping us understand the Atys and Adrastus story, though wordplay involving their names is a major element in the account.Footnote ¹³⁹ Herodotus is certainly capable of providing stories that feature what we would call punning on names: Crius (‘Mr Ram’) was told by Cleomenes to get his horns sheathed in metal (6.50.3); Cambyses rebukes Prexaspes, ‘this is how you carried out (Πρήξασπες … διέπρηξας) my command!’ (3.62.2); Leotychidas accepts the nomen omen when Hegesistratus (‘host-leader’) tells him his name (9.91.2). And so forth.Footnote ¹⁴⁰

But ‘pun’ implies, I think, a specific if also fortuitous connection between a name and an external, unrelated matter. Furthermore, the cases most often cited from Herodotus seem actually to feature characters in the History being reported as the ones who fabricate the pun; in those cases, the connection is not one that Herodotus makes in his voice as narrator. They are like those kings and commanders in Herodotus who recognise the true purport of divine messages, not infrequently after the fact.Footnote ¹⁴¹ Tellus the Athenian, ‘Mr End’, who demonstrates the importance of looking to the telos, may fit this pattern, inasmuch as he is reported as an exemplum by Solon to Croesus. But Atys and Adrastus, if their names are being significantly deployed, are named and their actions and motivations characterised by Herodotus himself, and their connection to atē and nemesis, if authentic, would be one that is sustained through several sections of text; it is not revealed by a quip or aside that is then forgotten.

We need to rethink what more extensive wordplay on names may have meant for Herodotus in connection with the Atys and Adrastus episode. In the first place, we need to dispense with the concept of ‘pun’, if by that we mean ‘trivial’ or ‘humorous’.Footnote ¹⁴²

Second, we need a model for wordplay that extends over a substantial amount of narrative and yet that is not to be understood as relating to the actions of personified abstractions, but, rather, to humans who are at the same time emblems of larger concepts. I think that good parallels are provided by the Old Testament.Footnote ¹⁴³ Aetiological naming abounds in the OT, and in particular in the Book of Genesis.Footnote ¹⁴⁴ One of the best-known cases is the renaming of Jacob: having wrestled with a man till daybreak, Jacob is finally overwhelmed but will not release his adversary unless he blesses Jacob: when the man asks ‘what is your name’, Jacob responds ‘Jacob’; ‘the man said, “Your name shall no longer be Jacob, but Israel, because you strove with God and with men, and prevailed”’ (Gen 32:27–8),Footnote ¹⁴⁵ with ‘Israel’ being explained as ‘God strove’. The point of the etymology, and what is especially important to stress, is that the new name for Jacob explains not just his striving with God, but, as the text suggests, that Jacob is in general a ‘striver’, one who contends with men and the divine. The name frames and explains his life.Footnote ¹⁴⁶ That is, we do not have, or not only, a scene of nomen omen; rather, several episodes and the character of the person involved are encapsulated in the name. This accounting of Jacob/Israel strikes me as similar to what we see in Adrastus, though we do not have a change-of-name scene such as we find in Genesis, but rather, a repeated naming scene.

And, in fact, there are also in the Bible other, coded, accountings for names, such as I am arguing for in Herodotus, as well as explanations relating to names that are not mentioned but are nonetheless referred to obliquely.Footnote ¹⁴⁷ Indeed, with special reference to Israel/Jacob, the one name can be explicitly mentioned while reference is also made to the other: at Amos 4:12, we read, ‘therefore thus will I do to you, O Israel; Because (^cqb) I will do this to you, Prepare to meet your God, O Israel’. As Moshe Garsiel explains, ‘the name of “Israel” appears here twice, while “Jacob”… does not – but it is to “Jacob” that … ^cqb refers’.Footnote ¹⁴⁸ The sole occurrence of nemesis in Herodotus, just before the appearance of ‘Adrastus’, could I think be said to be a similar phenomenon.

While no one biblical example exactly parallels what I take to be happening in the Atys and Adrastus story, and I would never maintain in any case that Herodotus was influenced by the OT or its antecedents, these examples of wordplay with significant names from the OT help me to imagine a way in which significant or ‘speaking’ names might work in Herodotus that goes beyond either punning or allegory, though those elements are perhaps also present. James Barr reads the story of Abigail and Nabal from 1 Sam 25 in ways similar to what I am arguing here.Footnote ¹⁴⁹

It is a mistake to see the story of Adrastus either as only that of a man cursed to live out the fate of his name, or only one of a person who is somehow a two-dimensional stand-in for nemesis. He bears an unlucky name, and he also is cursed, it seems, to commit unintentional homicide. His name is connected to other significant names, specifically Atys (atē). Adrastus can most definitely be seen as an agent of the ‘great nemesis’ that came from a god to correct Croesus’ view of himself; he hails from Phrygia – a point emphasised by Herodotus – a place that is the homeland of Nemesis, the goddess who corrects not only wrongdoing, but boasting self-importance in particular, and who sees to it that some sort of ‘confession’ is made later, when the criminal recognises his error. But Adrastus is also human, complete with hesitation, regret and shame.

What in the end does the story of Atys and Adrastus do for Herodotus? The Atys–Adrastus logos can be seen to do two, not necessarily competing, things at one and the same time. In the first instance, we must trust Herodotus’ text and see that it is not just a, but the, explanation, in a relatively short space and dramatic way, of how ‘a great nemesis from a god’ took hold of Croesus for his belief that he was the most fortunate of men. But an even larger, more generally applicable purpose is also in play. Pace the view that Herodotus did not understand causation systematically, location and uniqueness (or near-uniqueness) of narrative register for the episode argue that Herodotus wanted the Atys–Adrastus logos to function as a guide to help us see ‘why things happen’ elsewhere in the History. Such an exposure of the inner workings of the author's understanding is delicate to manage: it can be seen as heavy-handed and overly explicit, necessitating perhaps its rarity (cf. Phoenix’ allegory of the Litai, unique in Homer) and its unparalleled vocabulary – the once-occurring nemesis in all of Herodotus, a historian otherwise so preoccupied with price-paying.Footnote ¹⁵⁰