The single functions of the tale can be separated as follows:

1. Blank promise of the male lover
2. Demand of the female lover to gain a position equal to the wife (with the robe as a symbol of marriage or power)
3. Fruitless attempt of the male lover to avert peril
4. Jealousy of the wife
5. Intrigue of the wife against the (presumed) female lover
6. Forced participation of the male lover in the wife’s intrigue
7. Death of the (presumed) female lover

Semele is impregnated by Jupiter, which has Juno rage with jealousy (function 4). Taking the shape of Semele’s nurse Beroë, she talks the girl into doubting the true identity of her lover and wanting to see the god in his original form as a bolt of lightning (function 5: intrigue). Semele makes Jupiter grant her whatever she may wish (function 1) and then explicitly demands to be made equal to his wife:

Jupiter tries to avert peril (‘the god would have stopped her lips as she spoke’, 295–296) and, when this proves to be unsuccessful, to use a ‘lighter dart’, thereby being forced to participate in the intrigue (function 6)—but this is still too much for Semele who burns to death (function 7).

Herodotus’ adaption of the mythical basic structure is evident. The order of functions, however, is changed, the mythical tale beginning with Juno’s jealousy and intrigue (functions 4 and 5), whereas Amestris only enters the scene after the blank promise and Artaÿnte’s foolish wish. Apart from that, Herodotus’ duplication of both the female lover and the blank promise is surprising—the second promise becomes part of function 5, the wife’s intrigue. Also, some parts of the traditional structure are missing: Artaÿnte is not pregnant; her mother is not killed immediately but ‘only’ mutilated (probably to the same effect, though). I will later discuss these changes.

The historicity of the story seems virtually impossible—too striking is the fact that this is not the first time a link between Xerxes and Zeus functions as a means of characterisation in the Histories. Xerxes’ hubris is a basic trait of his character, and it is frequently connected in some way with the father of the gods: Xerxes steals Zeus’ effigy in Babylon and kills the priests, for example (1.183.3). But a lack of respect is not the only striking feature of his relationship with the god; he also engages in some kind of imitatio Iovis. For his ambitious foreign policies he explicitly states the goal of making ‘the borders of Persian territory and of the firmament of Zeus’ heaven be the same’ (7.8.γ.1). When crossing the Hellespont, Xerxes’ carriage rides directly behind Zeus’ carriage of honour while both vehicles’ drivers walk next to them (7.40.4). In the context of his greatest sacrilege, the desecration of the Hellespont and the forced joining of two continents (cf. below chapter II.2.1), an unnamed Hellespontian approaches him with the following words:

There are suggestive details in the story of Masistes’ wife and daughter as well: Xerxes grants his wife’s wish by nodding, and Herodotus uses the Homeric verb κατανεύειν that is exclusively reserved for the gods. The incestuous flavour of Xerxes’ relationship with his niece and daughter-in-law, too, hints at Zeus’ general amorous conduct.

The effect of the episode’s likeness with the myth of Zeus and Semele is obvious: Xerxes’ actions do not bode well for his future, as can be seen by parallels to the numerous other Greek literary or mythical characters who overestimate themselves or contend with the gods—and never go unpunished. The biographical piece of information mentioned before—Xerxes will die in a ‘private’ palace intrigue—adds another level to the story: the failed ruler of the world tries to play Zeus at least and fails even in his own house.

Considering the general tendency of the story, a few remarks on another mythical motif should not seem over-interpreted: the motif of the deathly robe. Heracles dies by a robe sent by a woman; Euripides’ Medeia, too, uses this apparently typically female weapon—against her rival Glauke/Creüsa. In Herodotus’ story, the addressee of the robe does not die; it is, however, possible that the robe anticipates the end of the king as a symbol of the ill fate Xerxes calls for by his hubris even within his own family. If he is thus interpreted as the actual victim of the robe, it becomes clear why Amestris does not give it to her enemy but to her husband.

However, the main reference here is to the mythical tale of Semele, the complex structure of which is changed but still recognizable in its main characteristics. The question remains why the story is not more closely modelled after the structure of the Semele-tale. Especially the duplication of the—at any rate: potential—female lover and the blank promise is puzzling. The answer results partially from the logic of the story itself: the second promise is ‘dramaturgically’ necessary as Amestris does not have Juno’s powers and needs to be authorised by Xerxes. This would also mean a certain respect for realism.

The duplication of the lover, too, could indicate such a respect for realism, even for the attempt of preserving certain historical data inserted into the mythical scheme—by Herodotus himself or by his sources. It might e.g. have been the case that Xerxes had an affair with his niece Artaÿnte, and that his nameless sister-in-law, the wife of Masistes, was cruelly mutilated. This material might now be gathered in the known narrative structure without exactly reproducing it. Such a partial fidelity to historical reality could also be the reason for omitting the motif of the female lover’s pregnancy: it would not have fit in with the rest.

However, the contrary is possible, too: Vivienne Gray observes that the story is an “incremental triple series of crises” (2002:311) typical for Herodotus’ style of narration, the three crises being the dilemmas of the king after the two blank promises and his unsuccessful attempt to marry Masistes to his daughter (seeing his wife is doomed to be killed by Amestris, 9.111). The narrative requirement for a triple structure would also be a good reason for doubling the first crisis. Although this kind of literary strategy would mean the exact opposite of the aforementioned “respect for realism”, both are plausible for Herodotus.

Having said that, a structural analysis of the episode will still tell us more about the duplications than all reflections on its origins. If we follow the paradigmatic axis of the story, the most striking effect of the duplication is a certain intensification: the motif of the blank promise, and with it the parallel with Zeus, are much emphasised in connection with Xerxes’ character—the megalomaniac Persian king not only acts in ways traditionally associated with the father of the Gods but does so excessively, with disastrous consequences seeming even more extreme as a completely innocent person is concerned.

An explicit categorisation of Xerxes’ actions as ‘sacrilege’ is unnecessary. In his dealings with the daughter of Masistes, the king is clearly imitating Zeus. That this is a bad omen need not be explained to the Greek recipient—and could not, as Herodotus does not use an unambiguous terminology. As in the following examples, it is the mythical parallel that shows what is so wrong about the actions of Herodotus’ protagonists.

Several of the Persian kings portrayed by Herodotus desecrate rivers or even the sea. These actions are always punished, if not always immediately, and usually consist of several or all of the following functions:

1. Implied personification of the water as a personal opponent
2. Verbal desecration of the water
3. Conceit (verbally expressed by the characters or stated by the narrator)
4. Change of natural environment or crossing of a natural border

The first instance of this pattern is the story of king Cyrus feeling insulted by the river Gyndes, where one of his sacred white horses has drowned (1.189–190; 5.52). He verbally threatens the river to make it so weak that even a woman can cross it without so much as wetting her knees, then has it parted into 360 canals, simply to punish it and without any strategic reason, according to Herodotus. The story contains all elements mentioned above, with the one exception of an explicit verbalisation of Cyrus’ conceit; but this third function is more than prominent in another act of desecration performed by the king at the river Araxes leading immediately to his downfall (1.205–208). Here, even the mere crossing of a waterway seems to be considered a moral transgression—an imperialist attempt to expand the area of human authority. The river Araxes that Cyrus travels across to fight the tribe of the Massagetes is not personified, nor does the king verbally desecrate it—but the crossing of the natural border seems to be of great significance in the context of Cyrus’ ensuing death. Also, Herodotus describes Cyrus’ motivation as seeming/believing (δοκεῖν) to be ‘something more than mortal’, 1.204.2)—function 3—and states that it would easily have been possible to fight the Massagetes without overstepping the river, which, in fact, several people had suggested to the king (1.206).

The peak of this traditional desecrating of waters is reached by Xerxes, who first changes the peninsula of Mount Athos into an island by digging a canal (7.22–24; 37) out of sheer megalomania (7.24)—making Herodotus’ readers think of the Cnidians who had been forbidden to perform the same on their own land by the Delphic oracle, as ‘Zeus would have given you an island, if he had wanted to’ (1.174). Second, he tries to bridge the Hellespont (7.33–36).

When his first attempt fails (7.34), he combines personification and verbal desecration of the sea (functions 1 and 2) by scourging the sea and sinking fetters into it (7.35). Raging with anger, he has the Hellespont branded and all the workers beheaded. While the sea is being scourged, the executioners have to speak ‘words outlandish and presumptuous’:

The vainglorious terminology of the δεσπότης obviously demonstrates his conceit (function 3). As for his calling the sea a ‘river’, this may be both mistake or aggressive downgrading on Xerxes’ side, but it also calls to mind Herodotus’ statement that the Persians usually worship rivers greatly (1.138.2), which makes Xerxes’ action outrageous even in the case that he should believe the Hellepont to be a river.

The fourth function is especially prominent as again, two different continents are yoked together; the violation of Nature turns into rebellion against divine order. Herodotus has Themistocles voice it when he talks about Gods and Heroes, ‘who deemed Asia and Europe too great a realm for one man to rule’ (8.109.3), and therefore made the Greeks victorious.

Before Achilleus kills Lycaon, he demonstrates his own self-confidence saying

Even though Achilleus talks about his own mortality in the ensuing verses, the conceit of the Herodotean figures is prefigured here. Having said that, it is significant that this function is more emphasised in the Histories, as we shall see in a moment.

In Homer, the personified (function 1) river god Scamander gets angry and complains about the obstruction of his stream which Achilleus has blocked with corpses (219–220)—the emphasis clearly lies with Achilleus’ hindrance of the natural order of things (function 4).

Why do Herodotus’ kings become guilty of such Achillean desecration? Firstly, both in the Iliad and in the Histories, there is always a connection between the deed and an important decision of the protagonist with regard to overstepping a boundary. But we can also see a significant difference between the Herodotean kings and Homeric Achilleus here: causes and consequences for the battles with waters are fundamentally different. The comparison with the Iliad example makes the acts of the Persian kings look decidedly more excessive.

Let’s first take a look at the matter of decisions. The Iliad’s plot is entirely dominated by the question whether Achilleus will fight again after having been insulted by Agamemnon. After the hero decides to join the forces again, things are delayed by his lack of armour; when he finally fights again, Scamander will be the last obstacle separating him from his main opponent, Hector.

Darius and Xerxes both make important—and false—decisions in the context of their crossing the sea: Darius is on his way to his doomed Scythian campaign, Xerxes to his failed attack on Greece.

Accordingly, the battle against a body of water in Herodotus corresponds often with the decision to cross a border illicitly, to enter a forbidden space—a motif central to the Histories, as has often been noticed. Pericles Georges calls Herodotus’ work “the story of the Persians’ confinement to their continent … The Persians are able to coexist neither with the land, nor with the people, nor with the gods of the world beyond Asia” (1994, 203).

But the Persian kings’ illegitimate crossing of borders is different from Achilleus’ battle against Scamander as well. The Homeric hero is supported by the gods Hera and Hephaestus (21.327–384), and his battle against the river ends in a battle of the gods (21.385–611). His superhuman status is therefore clearly marked, whereby he decidedly differs from the Persian kings. That being so, he still shows more humility—here speaking, again, to his victim Lycaon:

Consequently, the scene of Achilleus and Scamander constitutes not just a parallel, but also a contrast to the Persian desecrations of waters. A mortal must not be conceited when confronted with a natural phenomenon that is the object of cultic reverence, especially not out of seeming / believing to be ‘something more than mortal’—if even the demi-god Achilleus is conscious of his mortality at the moment of battle. The mortal kings Cyrus, Darius and Xerxes enter a geographic and moral space where their existence loses all justification.

In the context of king Croesus’ attack on the Persians (1.75), Herodotus offers another instance of desecrating water, which fundamentally illuminates the Histories’ way of using traditional structures.

In the beginning, Herodotus emphasises that Croesus had led his army over the river Halys using the bridges that were already there, ‘as I maintain’; there is, however, an alternative story (1.75.3): Thales of Miletus had diverted the stream by digging a crescent from the river passing the rear of the camp and entering the original bed again, thus making the Halys in front of the army shallow, or even emptying it. ‘But I do not believe this’, says Herodotus (75.6), and the reader, having been informed that Croesus actually used existing bridges, must assume Herodotus does not believe anything about the Thales story. Surprisingly, though, his distrust only refers to the very last possibility: the complete drying up of the river, which would have made Thales’ trench in the back of the army as deep as the original river, and ‘in that case, how did they pass the river when they were returning? (75.6)’

This kind of reasoning seems absurd. Herodotus presents three versions, only one of which is true, as he states explicitly at the beginning. Then the other two are discussed and only one of them is rejected, although the other one is not compatible with the first, ‘true’ version, either.

Instead of brooding further on the logical problems of Herodotus’ reasoning, it is worthwhile to ask for the actual meaning of this ‘original sin’ of oriental river-crossing. It has been noted by Rosaria Munson (1986:97) and Stewart Flory (1987:55–69) that the untrue versions serve a better illustration of Croesus’ illegitimate violation of the natural boundary and therefore deserve to be told along with the factual truth. But the coexistence of versions seems to result from other narrative strategies, too. Not only does the fictional story contain more truth than the factual version, but together they arrive at another level of truth.

Three degrees of sacrilege are described in the three versions:

1. Croesus has crossed the river on existing bridges; his ‘transgression’ therefore consists of the crossing of a natural boundary, but not in its removal.
2. Croesus has had the river diverted, making it shallow and walkable. This exacerbates his sacrilege to the extent of Cyrus’ diverting the Gyndes.
3. Croesus has entirely diverted the river, emptying the old river-bed, and has thus committed the sacrilegious act of changing Nature.

Herodotus may most incline towards the first version and least towards the third, but what Croesus does is morally questionable in any case, and by the presentation of different versions, an unambiguous verdict is made impossible. Munson and Flory may be right in that the factually ‘wrong’ versions contain more of the truth Herodotus wants to tell, but the false versions do not replace the correct one—all are placed side by side.

This course of action is very much in tune with the general characterisation of Croesus in the Histories. The Lydian king is portrayed in a very complex way especially from a moral angle. Positive and negative actions alternate constantly, starting with his introduction as ‘the one who I myself know first did the Greeks unjust deeds’ (1.5.3). Next, he is shown as the wise ruler who refrains from attacking the Ionian islanders because he listens to his adviser who tells him to respect the geographical facts, or what results from them: Greeks are good seafarers, whereas Persians are better on land. In contrast, in the lengthy account of his meeting with Solon, the wise Athenian (1.30–33), Croesus is shown as completely unreasonable—Herodotus even interprets the death of Croesus’ son Atys as divine revenge, because the king believes himself to be the happiest of men (1.34.1).

Later, Herodotus shows Croesus’ humility, when he tells poor Adrastos, who has involuntarily killed his son Atys, that Apollo was responsible for the deed (45.2). Immediately afterwards, however, Herodotus recounts the king’s dubious oracle test (46–49) proving his demanding attitude towards the divine, as are the immensely valuable and numerous dedications listed by Herodotus (50–52). They demonstrate Croesus’ piety, but also his status as a ‘good customer’ which he will insist on after his downfall.

Croesus’ lack of attention towards the divine is also corrobated by his complacent misinterpretation of the infamous Delphic oracle saying ‘he would destroy a great empire’ (53.3), which he takes as a clear indication of victory, as if the god could not foretell anything negative to his ‘sponsor’. On the other hand, he does show some serious generosity towards the Delphians in the very next chapter (54), then again misinterprets another oracle (55–56) and does not listen to advice that might have saved him (71).

After being captured by Cyrus, Croesus’ realises he should have listened to the wise Solon, and Apollo saves his life by sending rain to put out the pyre (86–87). He is therefore perceived by Cyrus as ‘dear to god and a good man’ (87.2) and becomes the Persian king’s own wise adviser (88–89)—only ever failing in his very last suggestion: that Cyrus should cross the river Araxes … (206–207).

A key scene is Croesus’ complaint to the Delphic priests, reproaching the ‘ungrateful’ god for having been deceived despite his generosity and reverence for oracles (90.2–4). The obvious impertinence goes unpunished; Apollo’s answer even proves that Croesus is loved by the gods: the king has to atone for the sin of his ancestor Gyges, but Apollo has personally secured him three additional years. As for the oracles, he tells the king their true meanings, and finally, Croesus humbly accepts his own responsibility (91.6).

Croesus is portrayed as a character permanently wavering between good and bad, piety and hubris, whose very nature is ambivalence. Consequently, the manner of narrating his encounter with the river Halys seems like a metaphor for his general character: it is clear he has overstepped boundaries, yet not only are we ignorant of the extent, but Herodotus presents us with clearly incompatible versions, which illustrate the contradictions in Croesus’ character.

The use of different, incompatible versions is not unusual in the Histories, as will be discussed again in chapter VI.2.1.