Jump to ContentJump to Main Navigation
Myth, Truth, and Narrative in Herodotus$

Emily Baragwanath and Mathieu de Bakker

Print publication date: 2012

Print ISBN-13: 9780199693979

Published to Oxford Scholarship Online: September 2012

DOI: 10.1093/acprof:oso/9780199693979.001.0001

‘Strangers are from Zeus’: Homeric Xenia at the Courts of Proteus and Croesus

(p.143) 5 ‘Strangers are from Zeus’: Homeric Xenia at the Courts of Proteus and Croesus
Myth, Truth, and Narrative in Herodotus

Elizabeth Vandiver

Oxford University Press

Abstract and Keywords

This chapter discusses Herodotus' use of the Homeric concept of xenia in the Histories. It argues that the appearance of xenia in key passages reflects the importance of Homeric epic and of the Greek legendary and mythic tradition for Herodotus' historiography. Herodotus foregrounds xenia in two logoi: Croesus' acceptance of Adrastus as a xenos (I.35–45) and Proteus' rebuke of Paris for wronging Menelaus (II.114–117). These logoi culminate in the death of Croesus' son Atys and in Herodotus' statement of his own opinion about the reason for Troy's destruction. The terminology of xenia establishes a Homeric tone that highlights these passages' significance for one of the overarching themes of the Histories: from the earliest encounters of Greeks and Asians onward, the gods made it clear that great transgressions by Eastern rulers would be punished. Paris in the remote past and Croesus at the cusp of humanly verifiable memory are guilty of the same transgression as was Xerxes within living memory; they overstep their bounds and claim more than is their right. Herodotus' inclusion of recognizably Homeric xenia in these logoi underscores the inevitability of divinely-sanctioned nemesis against such transgressions.

Keywords:   Adrastus, Croesus, Homeric epic, Menelaus, Memesis, Paris, Proteus, Troy, Xenia, Xerxes

In order to understand Herodotus we must cease to regard him as a historian, and see him as a narrator, whose narrative art is related to that of his sources. Herodotus should be accepted as the creator of a new generic form which only later became identified as history.

(Oswyn Murray)1

Herodotus’ approach to narrating the past provided a crucial foundation on which later historians could build their discipline, but he himself was a transitional figure, for whose ‘new generic form’ we do not even now have an adequate term. To view Herodotus in the way Murray suggests, as a narrator and a master of his art, allows an appreciation of precisely those elements that are most problematic when Herodotus is judged purely as a historian.2 Luraghi reminds us that, when authors no longer claimed inspiration by the Muse, ‘establishing a new way of dealing with the past implied the creation of a new kind of authority’.3 Herodotus’ ‘creation of authority’ was complex and multifaceted, but one element in it was his use of concepts familiar from epic, oral tradition, and the rich background of traditional tales (‘what is commonly meant, in ordinary usage, by (p.144) both “myth” or “legend” ’, to use Finley’s formulation4) to endow particularly important scenes or characters with an aura that would recall the tone, assumptions, and authority of epic. To understand Herodotus on his own terms, we should consider his use of narrative patterns and concepts inherited from epic and from the oral traditions about gods and heroes not as blemishes on his historiography or as evidence of his bad faith as a historian, but rather for what they accomplish in his own form of narrative.

This chapter examines Herodotus’ use, in two key passages, of one such traditional concept, xenia, usually translated into English with the rather cumbersome terms ‘guest-friendship’ or ‘guest–host relationship’.5 The reception and proper treatment of guests, their reciprocal treatment of their hosts, and the hereditary obligations incurred by the establishment of xenia between two men (and hence two families) are prominent topics in many traditional narratives about gods and heroes and are especially important in the Homeric epics. Particularly in the Odyssey, xenia ranks among the moral imperatives that humans cannot violate without bringing down divine vengeance upon themselves; as Nausicaa and Eumaeus both put it, ‘all strangers are from Zeus’ (6.207–8 = 14.57–8).6 The concept of xenia is, of course, by no means limited to traditional tales nor to the Homeric epics. Herman has shown that xenia was a powerful social concept and a method of forming extra-familial alliances in the fifth-century polis; it included formal political bonds between one polis and another or between monarchs as well as the relationship (p.145) between two individuals (and by extension their families) so familiar from Homer.7 This extended range of xenia is evident in Herodotus’ work; most of his uses of xenia terminology either refer to a clearly political alliance between two poleis or rulers, or mean simply ‘stranger’ or ‘foreigner’ without any implication of a complicated, binding, and inheritable relationship of mutual trust and benefit.8 But there are also occasions when Herodotus portrays xenia as a relationship between two individuals, whose obligations to one another are conceptualized within a framework that would be recognizable to any reader of Homer. This chapter examines two key logoi where xenia between individuals figures prominently and colours the entire logos by giving it a recognizably ‘Homeric’ cast: Croesus’ acceptance of Adrastus as a xenos (1.35–45) and Proteus’ rebuke of Paris for wronging Menelaus (2.114–117).9

Scholars have long recognized Herodotus’ debt to Homeric epic. This debt should not surprise or disconcert us; it would be astonishing had Herodotus not conceptualized his work in broadly Homeric terms. To recognize epic’s pervasive influence on the Histories’ organization, themes, concepts, and even its influence on the lexical level is not to question Herodotus’ accuracy or his evidentiary value for events of the fifth century, but rather to recognize that (obviously) he wrote from within his own culture and that his work is, among other things, a ‘combination of the Homeric and the new’.10 Direct and oblique references to Homer allowed Herodotus to provide his narrative not only with an implied chronology (however broad or vague it might be) and with an authoritative voice,11 but also with recognized reference points and key concepts with which he could stress particularly important episodes or figures in his narrative. In the two logoi we shall consider here, Herodotus uses xenia between individuals in precisely this way, to highlight the passages’ significance by (p.146) establishing a Homeric tone for them. The logoi about Croesus and Adrastus, and about Proteus and Paris, thus have important implications for our understanding both of Herodotus’ literary technique and of his conception of historiography. Both stories are far removed temporally from the Persian Wars that form the main subject of Herodotus’ work, and both feature recognizably Homeric xenia operating among non-Greeks (Lydians and Egyptians). I will discuss these two passages in reverse order, starting with Proteus, Helen, and Paris, since that logos directly reworks Homeric material.

1. Paris, Helen, Proteus, and Menelaus

Herodotus introduces the Helen-in-Egypt logos (2.112–20) to explain a certain temple’s name. The temple in question is in honour of ‘Aphrodite the Guest (or Stranger, or Foreigner)’ in Egypt (xeinēs Aphroditēs).12 Herodotus conjectures that in fact this temple honours Helen, not Aphrodite herself, and grounds his conjecture in two pieces of evidence; first, he has heard that Helen spent some time in the court of the Egyptian king Proteus, and, second, no other temple gives Aphrodite the title xenē (2.112).13

The relationship of Herodotus’ logos about Helen in Egypt to Stesichorus’ ‘Palinode’, according to which the gods sent an image (eidōlon) of Helen to Troy while she herself remained elsewhere, is a vexed question, as is the possibility of other sources for Herodotus’ (p.147) Egyptian version of Helen’s story.14 In Herodotus’ version, whatever its sources may have been, Helen remains in Egypt because of the intervention of the good Egyptian king, Proteus. A storm drives Paris (whom Herodotus consistently calls Alexander) and Helen ashore in Egypt near a temple to Heracles that has a tradition of offering sanctuary to runaway slaves.15 Several of Paris’ slaves escape to the temple and tell its attendant, Thonis, about Paris’ abduction of Helen. What happens next establishes the importance of xenia in this logos:

Thonis immediately sent a message to Proteus at Memphis, saying ‘a xenos, Trojan by birth, has come here after committing a sacrilegious deed [ergon…anosion16] in Greece. He seduced the wife of his own xenos and then carried her off along with a great deal of money, and now he has been forced by the winds to your land.’ (2.114)17

The messenger asks Proteus if Thonis should allow Paris to sail away again unharmed or if he should confiscate the goods with which Paris arrived (ta ekhōn ēlthe, presumably including Helen). In his response, Proteus too stresses xenia: ‘Seize this man, whoever he is who has committed sacrilegious acts against his own xenos [anosia exergasmenos xeinon ton heōutou], and bring him to me so I may see him and hear what he says’ (2.114). When he is brought before Proteus, Paris gives truthful answers to questions about who he is and where he is from, but when Proteus asks where he got Helen, Paris begins to circumvent the truth. The runaway slaves (p.148) inform Proteus of the true events, and Proteus’ reaction continues the foregrounding of xenia that has run throughout the scene:

‘If I did not think it of the utmost importance to kill no xenos [mēdena xeinōn kteinein] of those who have been driven off course by the winds and have come to my country, I would take vengeance against you on behalf of the Greek, you most evil of men, who when you had met with hospitality [xeiniōn tukhōn] performed the most sacrilegious deed [ergon anosiōtaton]: you approached the wife of your own xenos [para tou seōutou xeinou tēn gunaika ēlthe] and even that wasn’t enough for you, but once you had stirred her up, you went off with her. And still you weren’t satisfied, but you also plundered the household of your xenos.’ (2.115.4–5)

Proteus repeats that, since he thinks it crucial not to kill a xenos (here using the compound verb xeinoktoneein), he will himself take charge of the woman and the money and will guard them for ‘the Greek xenos’ until he arrives to claim them. Paris and his companions, however, must sail away from Proteus’ land within three days or be treated as enemies (2.115).18

The frequency of the xenia terms indicates how central the concept is for this logos. There are nine uses of xenos, xenia, or a compound in rapid succession; nowhere else in the Histories do xenia-words cluster so thickly together. Proteus makes it clear that the main source of his outrage and horror at Paris’ actions is precisely the violation of xenia that those actions entail.19 Proteus does not merely express disgust at the thought of one man abducting (or seducing) another’s wife, nor at the thought of property theft per se; rather, he hammers home that Paris’ deeds are utterly beyond the pale because they were perpetrated against a xenos.20 This point is highlighted by the emphatic repetitions of the reflexive pronouns: xeinou gar tou heōutou (of his own xenos), xeinon ton heōutou (his own xenos), tou seōutou xeinou (of your own xenos). The text thus foregrounds the point that Paris’ crime goes beyond seduction and adultery to violate not only (p.149) marriage norms and sexual mores but the assumptions of xenia as well. Xenia functions in many ways as an institution that parallels marriage, and thus Paris has transgressed against both of the systems whereby extra-familial alliances are formed, so that his crime constitutes a double outrage against the structures of society.21 Proteus also claims a kind of proleptic xenia with Menelaus; when the ‘Greek xenos’ arrives, Proteus says, he will return Helen and the treasure to him. Xenos could here be read only with respect to Paris, in which case Proteus means ‘when the Greek man who is your xenos arrives’. However, the later segment of the logos in which Menelaus receives xenia from Proteus strongly suggests that Proteus’ words here form an anticipatory claim of xenia between himself and Menelaus as well, to imply ‘when my xenos arrives’ and thus already to cast himself in the role of one who must protect Menelaus’ interests.

It is especially striking that, even so, Proteus considers himself bound not to harm Paris. However much Paris has shown himself to be utterly contemptuous of the obligations of xenia, Proteus cannot likewise disregard those obligations. Paris has arrived at Proteus’ shores as a xenos—here the word must mean ‘stranger’ first and foremost, since neither Proteus nor Paris has suggested any formal entry into ritualized friendship—and therefore must be hospitably received and protected.22 Proteus’ treatment of Paris (and later of Menelaus) is strikingly reminiscent of Alcinous’ treatment of Odysseus in the Odyssey; the wandering stranger accidentally driven ashore is automatically owed unconditional hospitality by his host, no matter what crimes he may have committed or who he may be.23 The underlying assumptions here are very much like those of the epic poet, for whom, as Herman says, it was ‘a part of the natural order of things’ that ‘the obligations of guest-friendship should be set above all other obligations’.24

(p.150) At the close of 2.120, the end of his account of Paris, Helen, Proteus, and Troy, Herodotus gives a statement of his own opinion about why the Greeks did not believe the Trojans’ assertions that Helen was not in Troy and why, more broadly, Troy fell. Herodotus is certain Homer knew the story that Helen was in Egypt, and he is also sure that it is the correct version, since it is unthinkable that Priam would not have given Helen back had she been there to give; no king would have countenanced the devastation of his city and the loss of his kingdom merely to let one son, and that one not even the eldest, live in adulterous bliss with someone else’s wife.25 But the Greeks did not believe the Trojans, because Troy had to fall; Herodotus says that, in his opinion, Troy’s destruction came about to make it evident to all that great wrongdoings receive great punishments from the gods:

But they did not have Helen to give back, nor did the Greeks believe them when they told the truth, because—I reveal my own opinion here—the divine26 was contriving matters so that the Trojans, through their utter destruction [panōlethriēi], would make it absolutely clear among human beings that the punishments the gods impose on great wrongdoings [megalōn adikēmatōn] are also great. (2.120)

Herodotus does not, in this concluding paragraph, state what the great wrongdoing of the Trojans was, but the earlier part of the logos makes it clear that it was Paris’ violation of xenia, identified by Proteus as an ergon anosiōtaton (2.115.4). The prince’s individual wrongdoing implicates his whole society, and Herodotus here reiterates the idea that Aeschylus puts into the mouth of the chorus in Agamemnon (ll. 60–2), that Zeus in his role as Xenios, the god who oversees xenia, required the destruction of Troy.27

The significance of this passage for our understanding of Herodotus’ purpose and methodology can hardly be overstated.28 As (p.151) Waters remarks, Herodotus’ stress on divine vengeance for the violation of xenia parallels ‘the orthodox Homeric view of the Trojan War’.29 While scholars have rightly noted the passage’s stress on the importance of argumentation from likelihood, it is no less noteworthy that Herodotus here combines that rational analysis of sources with the traditional moral framework reflected in Homeric epic. Furthermore, Herodotus shows not the faintest shadow of doubt that the Trojan War occurred and that Helen, Menelaus, Priam, Paris, and the rest were actual people whose actual deeds led to the war and its consequences. While in his argument from likelihood concerning Helen’s whereabouts Herodotus does indeed ‘come down on the side of pragmatism’,30 his conclusion that Helen could not have been in Troy is pressed into service to support the deeply traditional statement about divine vengeance; Helen’s absence from Troy was the means whereby the justice of the gods was worked out, to prove that great wrongdoings obtain great punishments.

Herodotus’ stress on Paris’ violation of xenia throughout this logos and his explanation in 2.120 of the consequences of that violation for Troy form the strongest possible contrast with the rationalized version of the abductions of women that opens Book One, a version that Herodotus attributes to Persian logioi.31 The emphasis of 2.120, where Herodotus specifically says that he is revealing his own opinion (gnōmē), could scarcely be more different from the emphasis of 1.5. Wise men do not make a fuss about such abductions, say the Persian logioi; divinity decreed that Troy must fall to show that great wrongdoings encounter great punishments, says Herodotus speaking in propria persona.32 Far from being a merely sordid and trivial personal (p.152) matter, as the Persian logioi would have it, Paris’ theft of Helen is a wrong that engages divine vengeance and requires the destruction of an entire kingdom. In short, 2.120 provides very strong evidence that Herodotus does not endorse the rationalization of abduction myths that he attributes to the Persians and Phoenicians, and that the rationalized version does not represent his own view of those stories.33

The Proteus logos complicates its presentation of divine vengeance, however, by including a further violation of xenia, enacted not by Asian against Greek but by Greek against Egyptian, that surprisingly goes unpunished. When the Greeks took Troy and found that Helen was not there, they sent Menelaus to Proteus in Egypt. Menelaus, received as a guest just as Proteus had predicted he would be, engages in a crime even more horrifying than Paris’—he sacrifices two Egyptian children:

When Menelaus arrived in Egypt, sailed up to Memphis, and told the truth about what had happened, he both received great hospitality [xeiniōn ēntēse megalōn] and got Helen back completely unharmed, as well as all of his goods. But even though he was treated so well, Menelaus became an unjust man [anēr adikos] to the Egyptians…He performed an unholy deed [prēgma ouk hosion]. Taking two children of local men, he made them sacrificial victims. (2.119)

(p.153) The text here downplays Proteus’ role as Menelaus’ host and instead presents Menelaus as a xenos of the Egyptian people in general. In fact, Waterfield translates: ‘The Egyptians looked after him magnificently, returned Helen to him completely unhurt, and gave him back all his property as well.’34 The implication that Menelaus was effectively a xenos not just of Proteus but of the whole Egyptian people casts his sacrifice of the children as a violation of xenia. As Munson says: ‘Adapted to the xenie theme of the narrative, this reenactment of the sacrifice of the Greeks at Aulis produces a host-sacrifice that reverses the alleged guest-sacrifices of the Egyptian Busiris.’35 In this reading, Menelaus’ crime becomes that of xenoktoneein, the murder of a xenos, precisely the crime that Proteus has twice articulated as most to be avoided. Agamemnon’s sacrifice of Iphigenia at Aulis, a killing that has no direct bearing on xenia, is displaced onto Menelaus in Egypt, and both Trojan and Greek participants in the war are shown to be guilty of xenia-violations.36 Paris’ violation of xenia occurred before the war began and was in fact the cause of the war, which Herodotus portrays as a means of divine vengeance against Paris. Menelaus’ violation occurred as he journeyed home after a victory that, as the discovery of Helen in Egypt proved, was empty. Proteus, transformed from the Odyssean Old Man of the Sea into a wise and righteous king,37 provides the locus for these two violations of xenia to be juxtaposed with one another. As Munson (p.154) observes: ‘Menelaus’ “impious” deed in Egypt, clearly evoking the human sacrifice at Aulis in one of the Greek traditions, gives a hint that not all may be morally right with the aggressor against Troy.’38 The account of Menelaus’ crime reminds the reader that the Achaeans were hardly blameless or wholly just in their conduct during the Trojan War and that their transgressions during the Sack of Troy led to misfortune for many of them on their journeys homeward, so that ultimately neither Trojan nor Greek avoided the retribution of the gods for their wrongdoings. But Herodotus does not mention any retribution that befell Menelaus himself after his sojourn in Egypt; this parallels Helen’s immunity from retribution in the Homeric presentation of the story.39 While Agamemnon dies and Odysseus wanders, Helen and Menelaus arrive home unharmed and, apparently, unscathed.40

Within the immediate context of the Egyptian logos, Herodotus’ crafting of the story stresses the opposition between the just Egyptian king Proteus and the barbaric, wandering foreigners, both Greek and Trojan; the Egyptian king’s moral superiority to either xenos is clear. Vasunia has argued that ‘Egypt occupies an interesting intermediate position between European Greece and barbarian Asia in Herodotus’ narrative’;41 here, Egypt under Proteus provides a template of true civilization against which the misdeeds and violations of both Trojan and Greek can be evaluated. Proteus and the Egyptians protect and obey the obligations of xenia, which is expanded from its essentially Greek status to appear as a moral value of the much older Egyptian civilization; Herodotus backreads Greek constructs into Egyptian culture in the case of xenia, much as he does in the case of religious beliefs and ceremonies.42

Within the Egyptian logos, then, the story of Proteus, Paris, Helen, and Menelaus emphasizes the position of Egyptian civilization as a source and guarantor of Greek values;43 the Egyptian Proteus (with (p.155) his Greek name) understands and practises the values of Greek civilization when not only Trojan but even Greek violate them.44 In the wider context of Herodotus’ entire work, however, the story emphasizes Trojan culpability over Greek, as Herodotus’ final expression of his own gnōmē in 2.120 indicates. This sentence, with its assumption that Paris’ theft of Helen was a crime deserving of divine punishment, foregrounds Trojan wrongdoing over Greek, and especially foregrounds the tendency of eastern potentates—here, in the person of Paris—to grasp more than belongs to them. The reason for this is to be found, I think, not just in Herodotus’ view of the Trojan War but also in the overall construction of his entire narrative.45 Divine retribution for Greek misdeeds in and after the Trojan War was not his main topic, but a demonstration of the inevitability of divine retribution for the transgressions of non-Greek monarchs was. Thus, the text can allow Menelaus’ sacrifice of two Egyptian children to go unpunished, while Paris’ theft of Helen brings down divine wrath upon his city and leads to its total destruction.

2. Croesus, Atys, and Adrastus

The stress on Homeric xenia in the Proteus, Paris, and Helen logos could in part be due to that story’s clear relationship, in both subject matter and characters, to Homeric epic. But, in the second logos I want to consider, the story of the death of Croesus’ son Atys (1.34–45), the characters are far removed from Homer’s epics in both time and place, yet Homeric xenia plays a crucial role there as well, even to the extent of including verbal echoes of both epics.46 (p.156) Indeed, Gomme calls this episode ‘the most moving and most Homeric story in Herodotos’.47

Herodotus begins his account of Croesus by marking the Lydian king out as a crucial figure who is, in some sense, the starting point for his enquiry into the causes of conflict between Greeks and Asians.48 The Croesus logos falls into several parts: background on Croesus’ ancestor Gyges; Croesus’ interaction with Solon; Croesus’ loss of his son Atys; his defeat by Cyrus and miraculous rescue from execution by Apollo; his later role as wise adviser to Cyrus and Cambyses. The segment of the story that concerns us here recounts the circumstances of the death of Atys. Herodotus’ narrative directly links the youth’s death to Croesus’ interaction with Solon and his inability to understand the point of Solon’s warning against assuming that one is truly fortunate: ‘After Solon left, great retribution [nemesis] from a god49 seized Croesus, very likely because he considered himself to be the most fortunate [olbiōtaton] of all human beings (1.34.1).’50 This divine nemesis consisted of the death of Atys, Croesus’ son and (p.157) heir. Croesus was fated eventually to lose his kingdom (and thus to illustrate the truth of Solon’s teachings about human prosperity) as payment for the crime of his ancestor Gyges (see 1.91.1). Herodotus strongly implies, however, that the misfortune of Atys’ untimely death was not part of this moira-decreed punishment for Gyges’ wrongdoing but was instead direct divine retribution for Croesus’ own behaviour.

Croesus dreamt that Atys would be killed by the iron point of a weapon.51 Alarmed by this dream, he forebade the youth to go on any expeditions or to be anywhere near weapons. Soon after Croesus had his dream, a Phrygian, Adrastus, came to Croesus’ court, begging to be cleansed of blood guilt because he had accidently killed his brother and had been driven away from his home. Croesus purified Adrastus and received him hospitably into his own court. Later, when Atys asked his father if he might participate in a boar-hunt, Croesus sent Adrastus along to protect the young man from robbers on the road. During the hunt, Adrastus threw a spear, which missed the boar, hit Atys, and killed him. Croesus thus lost his son, as the dream had foretold, and lost him to the very man whom he had received as a xenos.52

The similarities of the Croesus logos to tragedy have long been recognized.53 However, the logos also includes many echoes of Homeric epic, both conceptually and on the verbal level. The verbal similarities begin at the moment of Croesus’ and Adrastus’ meeting, when, after performing the needful purification ceremonies, the king questions the fugitive in words that recall the Homeric phrase tis pothen eis andrōn (‘who are you of men, and from where’, Od. 1.170).54 (p.158) Herodotus echoes this phrase in his introduction of Croesus’ questions and again, immediately, in Croesus’ own words:

When Croesus had performed the customary purification rites, he asked where the man was from and who he might be [hokothen te kai tis eiē], saying the following: ‘Sir, who are you, and from what part of Phrygia have you come [tis te eōn kai kothen tēs Phrugiēs hēkōn], to arrive at my hearth?’ (1.35.2–3)55

This is the first of several Homeric echoes in the Adrastus story; Croesus’ conversation with Adrastus at 1.41–2 and especially his words at 1.45 also show strong parallels to scenes in Homer. At 1.41–2, Croesus asks Adrastus to go out to the hunt as a protector of Atys, and Adrastus reluctantly agrees.56 Croesus begins by reminding Adrastus that he has purified him of murder and received him into the royal household. Therefore, he says, Adrastus should reciprocate with deeds that will benefit Croesus, and should accede to Croesus’ specific request that he act as Atys’ guardian on the proposed boar hunt.57 In addition to Adrastus’ duty to return benefits for benefits, Croesus further says that it would also be appropriate for Adrastus to distinguish himself in his deeds (kai se toi khreon esti ienai entha apolampruneai toisi ergoisi, 1.41.3). Adrastus responds less than eagerly:

Sire, otherwise I would not, for my part, go into this contest [allōs men egōge an ouk ēia es aethlon toionde]; it is not right that someone caught in such misfortune associate with his peers who are doing well, nor do I wish to do so…But as it is, since you are eager for this and it is necessary for me to please you (for I should recompense you by beneficial deeds), I am ready to do these things. (1.42)

(p.159) Croesus’ request and Adrastus’ response recall the exchange between Laodamas and the disguised Odysseus on Scheria (Od. 8.145–57), as the emphasized words show:58

‘Come you also now, father stranger [xeine pater], and try these contests [aethlōn],

if you have skill in any. It beseems you to know athletics [aethlous],

for there is no greater glory that can befall a man living

than what he achieves by speed of his feet or strength of his hands…’

Then resourceful Odysseus spoke in turn and answered him:

‘Laodamas, why do you all urge me on in mockery

to do these things? Cares are more in my mind than games [aethloi] are,

who before this have suffered much and had many hardships,

and sit here now in the middle of your assembly, longing

to go home, entreating your king for this, and all his people.’

(Od. 8.145–57)59

In both texts, a host or host’s son asks a xenos to take part in an aethlos, a show of strength and skill, stating that such participation is particularly fitting for the xenos. In both, the xenos cites his troubles as a reason for his reluctance to take part in aethloi and says that his misfortunes make him unfit for normal company, and in both cases the xenos is persuaded against his original inclination (Odysseus when he is taunted by Euryalus, and Adrastus by his sense of obligation to his xenos).

So far in the Adrastus story, however, the specific vocabulary of xenia has been absent. When Croesus hears Adrastus’ account of his ‘misfortune’ in killing his brother, he responds: ‘You are from a family of friends and have come among friends’ (1.35.4), but the word he uses is philos, not xenos.60 But, as the story moves to its terrible (p.160) conclusion, the importance of xenia is increasingly foregrounded.61 The shift in focus to xenia is directly signalled in the sentence that describes Adrastus’ slaying of Atys: ‘Then indeed the xenos, the same man, indeed, who had been purified for murder, the one named Adrastus, threw his javelin at the boar. He missed the beast, however, and instead hit Croesus’ son’ (1.43.2). This is the first use of the word xenos to describe Adrastus, and it could hardly be more emphatic or more horrifying in the context.62 For the second time, Adrastus (whose name means something like ‘he who cannot flee’ and is cognate with adrasteia, ‘necessity’, a title of the goddess Nemesis63) has inadvertently killed someone whom he was bound by all social codes to protect; already the murderer of his own brother, he is now guilty of xenoktoniaxenos-slaughter—as well.64 Indeed, Croesus’ entrusting of his son to Adrastus in the first place can be read as a reflection in miniature of the practice of sending children to a xenos for fostering; the idea that one’s xenos is the proper person to protect and guard one’s son is deeply embedded in the structure of the xenia relationship.65 The horror felt by both Croesus and Adrastus thus involves much more than a sense that it is particularly terrible for Adrastus to have killed the son of a man who had been his benefactor (p.161) and had purified him; the specific crime of xenoktonia must not be overlooked here.

When Croesus hears the news of his son’s death, he calls on Zeus by three names: Katharsios (‘of purification’), Epistios (‘of the hearth’), and Hetairēios (‘of comradeship’). Croesus does not directly invoke Zeus Xenios, but the concept of xenia is kept in the foreground of the passage by the narrator’s repetition of the word xenos to describe Adrastus, as Croesus calls Zeus to witness what he has suffered:66

Croesus, greatly distressed [suntetaragmenos] by the death of his son, took it even harder [edeinologeeto] that the one who killed the youth was the man he himself had purified for murder. Chafing terribly under his misfortune he called upon Zeus of Purification, testifying to the things that he had suffered at the hands of his xenos. He invoked the same god as Zeus of the Hearth and Zeus of Comradeship: invoking Zeus of the Hearth, because when he received the xenos into his household he was nurturing the murderer of his son, although he did not know it; and he invoked Zeus of Comradeship, because the man he sent out as guardian of his son he found to be his most bitter enemy. (1.44)67

Adrastus enters at this point and begs Croesus to kill him, saying that life is not bearable to him now that he has slain his benefactor’s son. The drama culminates in Croesus’ final words to Adrastus. Despite his own misfortune, Croesus is moved to pity, addresses his son’s killer directly as his xenos, and absolves him of blame:

Xenos, I have full justice from you, since you sentence yourself to death. You are not blameworthy to me for this evil [eis de ou su moi toude tou kakou aitios], except in so far as you enacted it unwillingly, but some one of the gods is blameworthy [theōn…tis], who long ago foretold [proesēmaine] to me the things that were going to occur [ta mellonta].’ (1.45.2)

(p.162) Croesus’ words here include one of the Histories’ most striking verbal echoes of Homer, since his second sentence is almost a quotation of Priam’s haunting words to Helen in the Iliad:

  • ‘I am not blaming you: to me the gods are blameworthy
  • [ou tu moi aitiē essi, theoi…moi aitioi eisin],
  • who drove [ephōrmēsan] upon me this
  • sorrowful war [polemon] against the Achaeans.’
  • (Il.3.164–5)

The verbal parallelism here is very marked. Both speakers, directly addressing another person, say ‘you are not aitios’ (blameworthy), using second-person singular forms of the verb ‘to be’; both attribute the blame to the gods (theoi); in both speeches the next clause is introduced by the nominative of the relative pronoun (hos and hoi, ‘who’) followed by the dative of the first-person singular pronoun (moi, ‘to me’). Furthermore, in each instance the next main element of the clause is a past-tense verb (proesēmaine, ‘foretold’; ephōrmēsan, ‘drove’) followed immediately by a substantive in the accusative case (ta mellonta, ‘the coming things’; polemon, ‘war’), even though this word order is by no means fixed in Greek.68 In short, there is a very strong echo here, in Croesus’ words, of one of the most memorable scenes in the Iliad; and in both cases, a powerful eastern monarch forgives a foreigner, whose acceptance into the monarch’s house has (p.163) caused disaster for that monarch.69 Earlier in the logos, Croesus has addressed Solon as ‘Athenian xenos’; it requires his interaction with two xenoi, one of whom warns him against overconfidence in his good fortune and the other of whom enacts the nemesis that punishes that overconfidence, for Croesus to be able to understand that his son’s terrible death is indeed ek theou.70

Croesus’ full realization of Solon’s wisdom will not come, of course, until he is on the funeral pyre set for him by Cyrus (1.86.3). Pelling notes that Croesus’ conversation with Cyrus at 1.87–8 echoes the scene between Achilles and Priam in Iliad 24 both thematically and verbally. He connects this scene with Croesus’ earlier words to Adrastus; ‘Croesus’ final pity for Adrastus…has something in common with Achilles’ for Priam…and if it is the killer in Homer and the bereaved father in Herodotus who pities, and the bereaved father in Homer and the killer in Herodotus who is so marked out by disaster, that reflects the unity of loss and suffering that both parties share.’71 Thus, at two key points in the Croesus logos, Herodotus uses echoes of Homer, and specifically of scenes featuring the Asian king Priam, to underscore the pathos and the inevitability of Croesus’ downfall. Pelling does not discuss the echo of Priam’s words to Helen, however, or the role of xenia in the Adrastus story. The text’s association of Croesus with Priam is even richer than Pelling’s analysis recognizes, since Herodotus’ description of Croesus’ pity for Adrastus and the words he puts into Croesus’ mouth simultaneously evoke Priam as ruler and host forgiving a foreign guest for bringing disaster upon him and Priam as suppliant and bereaved father, a guest in Achilles’ tent, begging for the dead body of his son.72

(p.164) 3. Conclusion

In the two passages we have considered, as nowhere else in the Histories, Herodotus portrays a form of xenia that the characters of Homeric epic would recognize and understand. He uses this concept to ground his narrative in Homeric tradition and thus to stress the scenes’ importance and narrative authority. I have argued elsewhere that Herodotus uses the heroes of Greek tradition as one organizing principle of his work, as a way of pointing out the resonances and parallels between the Greek–Asian conflicts of old and those of his own time.73 Essentially the same methodology is at work in his incorporation of recognizably Homeric xenia into logoi focusing on earlier eastern transgressors who did unjust deeds (adika erga) against the Greeks. Croesus, as the ‘first foreigner Herodotus knows’ to have wronged the Greeks, is highlighted as especially significant by having his story cast in recognizably Homeric terms. In Book Two, the gravity of Paris’ wrongdoing is forcefully reiterated by terms that hark directly back to Homer. Paris in the remote past of legend and Croesus at the cusp of humanly verifiable memory both exhibit what Herodotus seems to consider a besetting fault of Asian monarchs—that of overstepping their bounds and assuming more than is their right—and thus they set up a pattern of wrongdoing committed by Asian royalty against Hellas and the Hellenes that will find its culmination in Xerxes. Paris’ abduction of Helen is the most concrete example; he assumes that, if he wants this particular woman, he may take her, no matter what societal or divine mores he has to violate to do so. Croesus’ transgression is less direct, since he is guilty not of active wrongdoing but of an attitude that predisposes an autocrat towards wrongdoing; he fails to remember that, as a human being, he is liable to reversals of fortune, to pain, and to loss, despite his vast riches and apparently secure kingdom—and so Herodotus says that he was struck by a ‘nemesis from a god’, because he considered himself the most fortunate of all men who had ever lived.74

(p.165) Pelling notes that ‘overconfidence about the future’ is dangerous because ‘like a tyrant’s transgressive behavior [it] lays claim to an immunity that no mortal can assume’.75 This is the key characteristic that Paris and Croesus share; the assumption of superiority to normal limitations leads to action in Paris’ case and not in Croesus’, but the underlying disposition is the same, and it is the disposition that will lead Xerxes to think that he, too, can claim ‘immunity’ to divine retribution.76 Thus Paris and Croesus function, in the Histories, as forerunners of Xerxes in their arrogance, their wrongdoings against Greece, and their violations of boundaries established by the gods. Paris, Croesus, and Xerxes epitomize Asian transgressions in, respectively, the heroic age, the remote-recent past, and the past of living memory. Each of these men overstepped his bounds and brought down upon himself, as Herodotus specifically says about Croesus, a great nemesis ek theou. From the earliest encounters of Greeks and Asians, the gods were concerned to make it clear that great transgressions on the barbarians’ parts would be punished. Throughout the Histories Herodotus uses shared concepts and patterns of activity to link the past with the present. In the logoi we have considered here, the remote past of Troy and the almost-legendary past of Croesus and Solon are made paradigmatic for the recent past of the Persian Wars through the traditional Homeric concept of xenia.

Herodotus, of course, was not alone in this use of Homer as the comparandum par excellence, as a glance at the tragic stage will remind us. Aeschylus, we are told, referred to his own works as ‘crumbs from Homer’s table’.77 Herodotus too gathered many Homeric crumbs, but he seasoned those crumbs with a new and wholly (p.166) individual salt. As he explores the patterns and causes of events, Herodotus uses traditional concepts, legends, and vocabulary to form a narrative and a style of explanation that are utterly his own, bringing heroic past and human present face to face and showing how the same narrative constructs could be used to discuss them both.


(1) Murray (2001b: 322). See also Boedeker (2000: esp. 113–14).

(2) See Calame (1995: 76–91) and also the Introduction to this volume, pp. 2–10, on the Herodotean ‘paradox’.

(3) Luraghi (2006: 86–7).

(4) Finley (1975a: 13). For an overview of some of the problems inherent in assuming that Greek thought moved from ‘myth’ to ‘reason’, and that ‘myth’ corresponds to a recognizable ancient category of thought, see the introduction of Buxton (1999). Herodotus himself uses the word muthos only twice: see the Introduction to this volume, pp. 11–13. In both instances he clearly considers the story in question to be inaccurate, but it is far from clear that he intended any sharp differentiation in kind between stories that he identified as muthos and other forms of narrative; see Vandiver (1991: 7–8).

(5) The words xenia and xenos appear in several different dialect forms in our extant sources. For the sake of familiarity, I use the Attic forms in this essay, even though Herodotus himself used the Ionic xeinos and xeiniē.

(6) See Reece’s cogent discussion (1993: 182) of Odysseus’ slaughter of the suitors, which he correctly reads not as purely personal vengeance but rather as an indication of Odysseus’ ‘role as a guardian of society and an instrument of divine justice dispensing…punishment on those who have subverted the basic institutions that define civilization: marriage, inheritance, property rights, the agora, sacrifice, suppliancy, and, most pertinent here, xenia’.

(7) Herman (1987).

(8) Herodotus uses xenia eleven times and xenos eighty-four times. There are also numerous occurrences of cognate words such as xenizō and xenikos. For a discussion of political xenia in Herodotus, especially between tyrants, see Fisher (2002: 209–14).

(9) For an analysis of the Proteus figure, see de Bakker, this volume, Ch. 3. On the Helen logos, see also de Jong, this volume, Ch. 4.

(10) Thomas (2000: 268). For Herodotus’ reception of Homer, see inter alia Jacoby (1913: cols. 502–4), Huber (1965), Strasburger (1972), Erbse (1992), Boedeker (2002), Marincola (2006, with suggestions for further reading, and 2007a), and Pelling (2006a). Further bibliography is cited in the Introduction to this volume, n. 1.

(11) De Jong (2004b).

(12) Austin (1994: 122–3) makes a strong case for translating xeinēs here as ‘guest’ rather than ‘foreigner’. A. B. Lloyd (1988b: 45), with bibliography, identifies the temple as that of Astarte.

(13) Austin (1994: 119) refers to the ‘suspicious speed’ with which Herodotus makes the identificatory leap to Helen of Sparta. Fehling (1989: 65) notes the oddity of the identification and comments that it gives us ‘a picture of the author hearing of the existence of a temple of Astarte in the Phoenician quarter, instantly divining a link with Helen, and hurrying off to secure confirmation in the form of a few uncomprehending nods to be extracted from guides devoid of any knowledge of Greek’. Even if one does not assume with Fehling that Herodotus’ Egyptian guides were ignorant of Greek, the rapidity of Herodotus’ identification of Helen remains a problem, probably best solved by assuming that he ‘has arrived in Memphis with this story already in hand…Whether following the trail of his predecessor Hekataios or in someone else’s footsteps, Herodotus knew what he expected to find when he walked into the sanctuary’ (Austin 1994: 122–3).

(14) See Vandiver (1991: 125 nn. 1–2) for partial bibliography, and, in this volume, de Bakker, Ch. 3, n. 6, and de Jong, Ch. 4, n. 3. A. B. Lloyd (1988b: 47) notes that one of the most striking elements that differentiates Herodotus’ account of Helen from the version of her story preserved in Stesichorus’ ‘Palinode’ is the ‘removal of the eidōlon’; he thinks that the version Herodotus recounts must have been developed by someone ‘active before H.’s dialogue with the priests and someone with a relatively sophisticated psychological awareness as well as a liberal dose of rationalism’, and comments that the ‘obvious candidate’ is Hecataeus. See also Austin (1994: 127–8, and his whole discussion of Helen in Egypt, 118–36).

(15) On the possible significance of Herodotus’ use of the name Alexander, see de Jong, this volume, Ch. 4, pp. 132–3. For possible identifications of which Egyptian god is meant by ‘Heracles’ here, see A. B. Lloyd (2007: 323). For sanctuary in Egypt, see Lloyd (1988a: 48), who argues against assuming that this is a Greek element in the story; Fehling’s claim (1989: 64) that ‘all scholars agree’ that asylum ‘has no parallel in Egypt’ is greatly overstated. For the function of the temple as a Herodotean piece of opsis in the overall story, see de Jong, this volume, Ch. 4, pp. 130,132.

(16) On anosion, A. B. Lloyd (2007: 323) comments: ‘it was worse than an ἄδιϰϰον; for he had committed an offence against Zeus Xenios.’

(17) Translations of Herodotus are my own.

(18) On Proteus in this story as a foreign educator of the Greeks on the subject of xenia, see de Bakker, this volume, Ch. 3, pp. 113–17.

(19) A. B. Lloyd (2007: 324): ‘Proteus’ respect for the obligations of ξενία is made to contrast forcefully with Alexander’s disregard for them.’ See also A. B. Lloyd (1988a: 45).

(20) A. B. Lloyd (2007: 324) points out that the switch to oratio recta in Proteus’ denunciation of Paris ‘has the effect of throwing the dénouement into sharper and more dramatic relief ’. For alternative explanations of the switch, see, in this volume, de Bakker, Ch. 3, pp. 124–5, and de Jong, Ch. 4, pp. 131–2.

(21) On xenia and marriage as the foremost means for forming alliances with those outside one’s own community, see Herman (1987: 36). See also Finley’s statement (19783b: 99) about xenia in Homeric epic: ‘guest-friendship was a very serious institution, the alternative to marriage in forging bonds between rulers.’

(22) On formal ceremonies for beginning a xenia relationship, see Herman (1987: 44–54).

(23) I owe this comparison to Emily Baragwanath and Mathieu de Bakker. As we shall see below, the Croesus/Adrastus logos similarly foregrounds the host’s obligation to help his guest before knowing what has brought that guest to his country.

(24) Herman (1987: 2).

(25) On Herodotus’ implicit criticism of Homer here, see de Jong, this volume, Ch. 4, pp. 139–40.

(26) In translating tou daimoniou as ‘the divine’, I follow T. Harrison (2000a: 105). For a discussion of daimon’s range of meanings in the Histories, see T. Harrison (2000a: 164–9).

(27) De Jong, this volume, Ch. 4, pp. 140–1, reaches a similar conclusion.

(28) Aly (19692: 67) sees Herodotus’ concluding sentence with its claim of divine vengeance as evidence of his particular interest in this logos: ‘Es kommt hinzu, dass er sich für diesen Logos besonders interessiert, den er ausdrücklich für zutreffend erklärt und emphatisch mit dem allgemeinen Satze beschliesst’ (‘There is also the fact that he is particularly interested in this logos, which he expressly declares true and emphatically concludes with the general maxim’).

(29) K.H. Waters (1985: 101); see also Vandiver (1991:128–30). But see de Jong, this volume, Ch. 4, pp. 140–1, for an argument that Herodotus’ ‘moral model’ is a departure from the Homeric version. Austin (1994: 135) severely downplays the importance of the gods here: ‘With Helen securely hidden in Egypt…the whole Trojan War became then a human catastrophe’, although he does note that ‘Herodotus still interprets the Trojan War as the grand spectacle designed by those whom he designates in the traditional fashion as the gods’ (emphasis added).

(30) Baragwanath (2008: 73).

(31) The tone and significance of this ‘Persian version’ of Greek myths is hotly disputed; for brief discussion and bibliography, see Chiasson (2003: 17–18). Cf. also, in this volume, Dewald, Ch. 1, pp. 61–7, and Saïd, Ch. 2, pp. 101–5.

(32) See T. Harrison’s comment (2000a: 109–10): ‘The divine response to Alexander’s theft of Helen, the destruction of Troy, is (whatever we might think) envisaged apparently by Herodotus as a proportional response.’ Elsewhere Harrison is, I think, overly restrictive in contrasting Herodotus’ statement of divine retribution at 2.120 only with Paris’ assumption in 1.5 that he would receive no punishment for abducting Helen (2000: 104–5). The portrayal of divine punishment in 2.120 contrasts with the assumptions of the Persian logioi in general that such abductions were trivial, not just with Paris’ false expectation for his own case. Pelliccia (1992: 82) surprisingly omits any mention of Herodotus’ statement about divine punishment from his discussion of Herodotus’ ‘good sophistic argument from probability’ that Helen could not have been in Troy.

(33) Arieti (1995: 9–11) argues that Herodotus’ presentation of the accounts of the Persian logioi is intended as a negative characterization of the Persians, not as an indication of Herodotus’ own tendency to rationalize ‘myth’. I agree with this point and would add that Herodotus’ presentation shows us learned Persians who are contemptuous and dismissive of Greek mythic traditions, concerned to destroy the kleos of Greek accounts of events rather than to preserve it; thus, the Persians and the Greeks are in conflict even in their understanding of the most remote past and the views of the Persian logioi are in direct opposition to Herodotus’ stated purpose for writing ‘so that great and marvellous accomplishments would not lose their glory’ (mēte erga megala te kai thōmasta…aklea genētai, proem). See further Vandiver (1991: 114–24). For very different interpretations of the abduction narratives, see, e.g., Pelliccia (1992: 76–80) and Moles (1993: 95–6).

(34) Waterfield (1998: 140).

(35) Munson (2001: 144).

(36) Baragwanath (2008: 112) notes the parallel terminology of Paris’ anosion ergon (unholy deed) and Menelaus’ prēgma ouk hosion (an action not holy), although she does not comment on the role of xenia. On Menelaus’ sacrifice of Egyptian children, A. B. Lloyd (1988b: 51) comments that ‘whatever the origin of the tale…it should be noted that the act would be equally offensive to Gks. and Egs.—to the former as a violation of xenia, to the latter because of their repugnance for human sacrifice’. This overlooks the purely Greek view of xenia that Herodotus attributes to Proteus; whatever actual Egyptians’ primary reasons for horror at human sacrifice might have been, in the context of Herodotus’ logos the xenia violation surely looms as large for the ‘Egyptian’ characters (wholly Herodotean constructs in their language, viewpoints, and actions) as for the Greek ones. ‘The language and the social values are unequivocally Greek’ (A. B. Lloyd 2007: 323). On the Herodotean ‘fingerprint’ throughout the Helen logos, see de Jong, this volume, Ch. 4.

(37) According to A. B. Lloyd (1988a: 32), the name Proteus ‘has no Egyptian ancestor [and]…is clearly that of the Homeric sea god’. This is the only name in Herodotus’ list of Egyptian kings that ‘has demonstrably been injected into the tradition from Greek sources’ (A. B. Lloyd 1988a: 33). On the transformation, see de Bakker, this volume, Ch. 3, pp. 109–13.

(38) Munson (2001: 186).

(39) See also de Jong, this volume, Ch. 4, pp. 136–40 and nn. 38–9; for a different view, see Saïd, this volume, Ch. 2, p. 101.

(40) It is worth noting that in the Odyssey it is Proteus (the Old Man of the Sea) who tells Menelaus that he will end his life in the Elysian fields.

(41) Vasunia (2001: 125).

(42) ‘Herodotus’ text both Hellenizes and mythologizes Egypt’ (Vasunia 2001: 125; see his whole discussion of Helen in Egypt, pp. 124–6). Cf. Haziza (2009).

(43) De Bakker, this volume, Ch. 3, pp. 113–22, reaches a similar conclusion and argues that Proteus is staged as a methodological example for Herodotus as well.

(44) See Vasunia’s comment (2001: 188) that Herodotus ‘reverses the ethnic logic of the [Busiris] story, not only by repudiating the likelihood of human sacrifice in the socioreligious context of Egypt, but also by imputing to a Greek hero the very actions that the canonical narrative, despite his history, represents as characteristically Egyptian’.

(45) See Cartledge and Greenwood’s statement (2002: 356) that, ‘by concluding the episode in this way, Herodotus appropriates oracular authority for himself ’.

(46) Cf. also Dewald’s observation, this volume, Ch. 1, p. 71, that several names in this story have mythic resonances. For my purposes here, the question of how much of the Croesus logos was traditional is irrelevant; the crafting of the story and its emphases were Herodotus’ own. For an argument that the Atys/Adrastus story was traditional, see Miller (1963: 90–2); for a discussion of the difficulties involved in trying to determine what material in Herodotus’ unattributed logoi was traditional and what was his own invention, see de Jong (1999: 244–5). Concerning the Atys/Adrastos logos, de Jong (1999: 245) says: ‘je pars de l’hypothèse que les narrataires ne connaissaient pas l’histoire et que le narrateur devait lui-même créer la tension’ (‘I start from the assumption that the narratees did not know the story and that the narrator had to create the suspense himself ’).

(47) Gomme (1954: 80).

(48) Herodotus’ introduction of Croesus as ton de oida autos prōton huparxanta adikōn ergōn es tous Hellēnas (‘the first whom I myself know began unjust actions against the Greeks’, 1.5.3) of course has complex chronological and historiographical implications; for recent discussions of Herodotus’ chronology, see Thomas (2001a), Vannicelli (2001: esp. 230–5), and Cobet (2002: 405–11).

(49) This is Herodotus’ only use of the term nemesis: Asheri (2007: 105). Since Greek has no indefinite article, it is unclear whether ek theou means ‘from a god’ or ‘from god’ (presumably Apollo, here); I have chosen the more indefinite translation.

(50) I take hōs eikasai as limiting the clause that follows it, not the entire sentence. Neither How and Wells (1928) nor Asheri (2007) comments on this use of the phrase. Barbour (1964: 226) identifies it as ‘limiting the following clause’; so also Rieks (1975: 38) and T. Harrison (2000a: 36). The translations by Godley (1981), Grene (1987), Blanco (1992), Waterfield (1998), and de Sélincourt (2003) all agree; e.g.: ‘But after Solon’s departure, the divine anger fell heavily on Croesus: as I guess, because he supposed himself to be blest beyond all other men’ (Godley 1981: i. 41). For readings that take hōs eikasai as limiting the first clause, see, e.g., Gould (1989: 79), and Fisher (2002: 218). However, Gould (1994: 95) appears to reverse his reading in his later article, when he refers to ‘the explicit admission that the identification of the motive for the (unidentified) god’s anger is a “guess” or probable inference’ (emphasis added). If it is correct to take hōs eikasai as limiting the second clause, then, although Herodotus asserts only likelihood for the reason for the nemesis, he does not question that what happened to Croesus was indeed nemesis ek theou. In other words, Herodotus allows for the possibility that he is mistaken in his identification of the cause of the divine retribution, but he does not doubt that nemesis ek theou is what these events in fact are. In this reading Herodotus as narrator agrees with the assessment of the events given by Croesus as character, when he says to Adrastus that the gods are responsible; see further below.

(51) On premonitory dreams in Herodotus, see Lévy (1995).

(52) For a discussion of the many folkloric motifs in the Atys–Adrastus logos, especially the boar-hunt, see Aly (19692: 38–40).

(53) See, e.g., Rieks (1975), Laurot (1995), Saïd (2002: 132–7, with bibliography), and Chiasson (2003).

(54) ‘Croesus first takes care of the stranger and then asks for his name—a Homeric form of courtesy’ (Asheri 2007: 105); Asheri does not note the Homeric phrasing of the questions. See also Barbour (1964: 227) and Arieti (1995: 57). Long (1987: 84) says that the scene is ‘conducted on the same terms as the conversations of the noble heroes before Troy in the Homeric epics’ but does not mention the specific verbal parallels.

(55) Proteus asks similar questions of Paris, but these are reported only by the narrator and not in direct speech: eirōta ton Alexandron ho Prōteus tis eiē kai hokothen pleoi (‘Proteus asked Alexander who he was and where he had sailed from’, 2.115.2).

(56) Pace Long (1987: 93), who says ‘Adrastus leaps at the opportunity’.

(57) Gould (2001a: 287–8) discusses Croesus’ assumption that his good treatment of Adrastus requires good in exchange as an example of reciprocity in gift-giving, but does not locate the scene within the framework of xenia.

(58) Rieks (1975: 31) notes the similarity: ‘Die Weigerung Adrasts, an der Jagd teilzunehmen, dürfte von der entsprechenden Odysseus-Szene beim Wettkampf der Phaiaken beeinflusst sein’ (‘Adrastus’ refusal to participate in the hunt may have been influenced by the corresponding scene of Odysseus at the contest of the Phaeacians’); see also p. 36 n. 60. Long (1987: 82) and de Jong (1999: 250) both note that aethlos is an unusual word to apply to a hunt, since it would normally be used of contests between human beings. Neither comments on the connection with Odysseus’ words to the Phaeacians.

(59) Translations from Homer are Lattimore’s (1951, 1965).

(60) Herman (1987: e.g. 106–8) repeatedly demonstrates that philia terminology could be used interchangeably with xenia terminology at times.

(61) Recent scholarship has tended to overlook xenia here and to stress instead the supplication and purification elements in the logos. For instance, Saïd (2002: 135) says that Croesus ‘had received [Adrastus] as a friend and ritually purified [him] of a murder’. T. Harrison (2000a: 41) similarly elides the xenia out of the scene by referring to Adrastus as ‘the suppliant whom Croesus had received into his house’. Chiasson (2003: 9) says that ‘the arrival of Adrastus, with his request for ritual purification, sets into motion the common tragic motif of supplication’. Adrastus was, of course, a suppliant in need of ritual purification, but he was also a xenos and was thus bound to Croesus by double ties of obligation, both of supplication and of xenia. Arieti (1995: 62) refers to the pre-existing xenia between the two royal families but not to Adrastus’ status as Croesus’ current xenos: ‘Having performed the ritual purification for Adrastus, which established a very close bond between them, Croesus has every reason to expect Atys [sic] to be a devoted servant. There is also the additional guest–friendship (xenia) enjoyed between the two families.’

(62) Long (1987: 95) does not comment on the word xenos in his otherwise close analysis of Herodotus’ vocabulary here.

(63) How and Wells (1928: i. 71). Fisher (2002: 205) says that Adrasteia is ‘a Greek term for the personification of Fate’.

(64) Herman (1987: 124) refers to ‘the horror inherent in xenoktonia—a loaded word, the connotations of which can never fully be rendered by the phrase “murder of a guest–friend” ’.

(65) On foster-parenthood as an aspect of xenia, see Herman (1987: 22–6).

(66) In fact, the stress on xenia is so powerful that de Sélincourt (2003: 19) translates ‘Zeus Hetairēios’ as ‘God of guest–friendship,’ as though Herodotus had indeed written Xenios. See also Long’s reference (1987: 97) to ‘Croesus’ complaint against Zeus as the god of hospitality’. Long (1987: 97) seems to conflate Epistios with Xenios, since he translates Hetairēios as ‘god of companionship’.

(67) I assume, with the standard translations, that kaleō here means ‘invoke’. Powell (1960 2: 181) cites this passage as an instance of the meaning ‘invoke a god’ (s.v. καλέω). Contra Long (1987: 75), who appears to read kaleō as meaning ‘summon to justice’, since he says that Croesus’ words are ‘a blasphemous imprecation against Zeus’. See also Immerwahr’s statement (1966: 70) that Croesus ‘accuses Zeus’.

(68) Aly (1969 2: 267) identifies Croesus’ words as one of several ‘äusserlich nicht erkennbare Zitate an bemerkenswerter Stelle’ (‘not outwardly recognizable quotations in a remarkable place’). Huber (1965: 30–1) comments that Herodotus’ adaptation of Priam’s words foregrounds the role of fate, arguing that ‘Kroisos gegenüber Adrast die Worte des Priamos an Helena…nicht nur wiederholt, sondern die “Schuld” des Adrast durch den Zusatz εἰ μὴ ὅσον ἀέκων ἐξεργάσαο differenziert und mit ἀλλὰ θεῶν κού τις, ὅς μοι…προεσήμαινε τὰ μέλλοντα ἔσεσθαι das Schicksalhafte noch stärker hervorhebt’ (‘Croesus not only repeats to Adrastus the words of Priam to Helen, but differentiates the “guilt” of Adrastus through the sentence “except in so far as you enacted it unwillingly”, and with “but some one of the gods is blameworthy, who long ago foretold to me the things that were going to occur” he emphasizes even more the role of fate’). Huber does not discuss the close formal parallelism between the second clauses of Croesus’ and Priam’s remarks. Long (1987: 100) thinks that Croesus’ words to Adrastus show a notably ‘natural’ tone; he further comments (1987: 103) that Croesus speaks ‘with a freedom and directness that we have otherwise heard from him only…with his own son. Herodotus allows the mask and tone of office to slip.’ This misses the fact that Croesus is here all but quoting Priam, the eastern monarch par excellence; these markedly Homeric words are unlikely to indicate that Croesus is stepping outside his royal role.

(69) Although Croesus forgave Adrastus, Adrastus did not forgive himself; he committed suicide on Atys’ grave (1.45.3).

(70) Obviously I disagree with Long’s interpretation (1987: 102) that xenos here is an ‘alienating’ term to be translated ‘stranger’, which ‘marks Adrastus as an outcast, once more an outsider, now certainly an intruder who was once tolerated in the palace but is henceforth denied purification, shelter, reception’.

(71) Pelling (2006a: 86 n. 30); see also Pelling (2006b: 160). De Jong (1999: 250) identifies a Homeric echo in the description of Adrastus’ fatal javelin throw: ‘l’expression “il manque A, mais atteint B” est une réminiscence épique,’ citing Il. 4.491–2 as a comparandum. She does not mention the other Homeric echoes in the passage.

(72) For further Homeric resonances and echoes in the end of the Croesus logos, especially concerning blame, see Huber (1965: 34–6). See also Chiasson (2003: 27 n. 78).

(73) Vandiver (1991: passim).

(74) For a discussion of whether Croesus’ assumption that he is ‘fortunate’ equals fully-fledged hubris, see Pelling (2006b: 150, with bibliography in nn. 34–6). Asheri (2007: 68) assumes that Croesus’ mental disposition is in fact equivalent to hubris: ‘the moral link between ὕβρις and ἄτη connects the dialogue of Croesus and Solon to the drama of Atys and Adrastus, as well as to Croesus’ words on the pyre and afterwards.’ See also A. B. Lloyd (1988a: 28). For arguments that Croesus’ assumption does not constitute hubris, see, e.g., Gould (1989: 80–1) and Fisher (2002: 218).

(75) Pelling (2006b: 150).

(76) In Croesus’ famous misinterpretation of the oracle about destroying a great empire by attacking Persia (1.53.3), his assumptions about his own status as a fortunate king do in fact lead to an action, the campaign against Cyrus, that destroys the very fortune that tempted Croesus into misjudgement of the oracle in the first place. Here I disagree with Pelling’s assertion (2006b: 159 n. 66) that ‘the only reasonable interpretation of Delphi’s response was one of encouragement’. It seems to me, on the contrary, that any enquirer not blinded by the assumption of his own fortunate status would immediately have recognized the ambiguity of the Delphic response; on this point see Lévy (1997: 363). Dewald, this volume, Ch. 1, pp. 75–80, 82 discusses the (Iliadic) ‘thoughtless ruler’ theme in the Croesus and Cyrus logoi.

(77) temakhētōn Homērou megalōn deipnōn, literally, ‘slices from Homer’s great banquets’ (Athenaeus 8, 347E).