Mages and Ionians

2017.06.26 | By Gregory Nagy

This text is the basis for a lecture I presented in Budapest, Hungary, 2017.06.26, on the occasion of my induction as a corresponding member of the Hungarian Academy of Sciences (Magyar Tudományos Akadémia). The lecture, which I had composed in Hungarian, was an abridged version of what I present here in English.

Preliminary Note 1:
A video recording of the original lecture, 2017.06.26, which I had read out in Hungarian after a brief introduction spoken in English, is available here:

Preliminary Note 2, written 2018.11.23:
A printed version of the original online text of 2017.06.26 appeared over a year later, in 2018: Antichi Persiani: Storia e Rappresentazione, edited by Clelia Mora and Cesare Zizza (with an introduction by Antonio Panaino), Biblioteca di Athenaeum 60, 2018 Edipuglia (Redazione: Valentina Natali) pp. 97–121. The page-breaks of that printed version have been inserted into the online text here. Also, the printed version included some addenda to the orignal online version, as in the case of footnote 17 at p. 112 (this footnote 17 has been added in the online version as well, at §54).

(Corrigenda for the printed version:

—§5, delete “Persianate”

—§40 number 1, in “Lydian Empire, which had dominated Ionia”, insert “mainland” before “Ionia”

—§41, not “obliteration” but “capture”

—§53 number 3, not “the navy of this city was” but “the forces of this city were”.)

"Persiani" (19th c.), artist unknown. Image via The British Museum.
“Persiani” (19th c.), artist unknown. Image via The British Museum.


§0. {|97} There is a problem with the meaning of the word ‘mage’, mágos (μάγος) in Greek, which was borrowed into Greek from a language known to linguists as Old Persian. In the Bisotun Inscription, which celebrates the accession of Darius the First in 522 BCE and his subsequent successes as the new king of the Persian Empire, the corresponding Old Persian word is maguš, referring to a man named Gaumāta whom Darius accused of usurping the Persian kingship (Column 1 lines 35-43, etc.). In the History of Herodotus (3.61–79), there is a convergent narrative about a mágos who usurped the kingship, though his name is given there as Smerdis.[1] In any case, here is the problem: aside from the idea that, once upon a time, a maguš / mágos usurped the kingship of the Persian Empire, the contexts of mágos in Greek sources do not match neatly the contexts of Old Persian maguš and of its cognates as found in other Iranian languages. Aiming at a solution to this problem, I argue that the Greek and the Iranian contexts of the word can best be understood by taking a closer look at historical evidence for interactions between Greeks and Persians in the sixth, fifth, and fourth centuries BCE.[2] On the basis of such evidence, I also argue that the original mediators for the word mágos and for the ideas underlying this word were the Ionians.[3]

A problem with the word mágos

§1. On the surface, in popular cultures that flourish in what is vaguely understood to be the “West,” the word mágos, which I translated as ‘mage’ in the title of this {|98} presentation, does not seem to be all that problematic. We know that this word is somehow related to the word magic. So, a ‘mage’ must be some kind of a practitioner of magic, no? The meaning seems esoteric or even exotic, but understandable. So, what is the problem here?

§2. The problem has to do with origins. If we look up the word in any encyclopedia, the basic fact emerges: this word mágos originates from Iranian civilization. But the meaning of such a word in its Iranian contexts does not square with our superficial “Western” understanding of the word. I say Iranian contexts, not context, since the historical background of the word mágos in Iranian civilization is multiple—and complicated. By contrast with the modern “Western” understanding, which is superficially simple, there is no unified “Eastern” meaning of the word. The Iranian contexts of mágos are quite diversified: the word seems to mean different things at different times and in different places within the vastly diverse world of Iranian civilization.[4] There are clearly multiple understandings, and a unified explanation has up to now eluded any consensus—even among experts in Iranian civilization.[5]

Another problem with the word mágos

§3. Now we come to a second problem. Adding to the complications that I have already sketched is the fact that the word mágos crosses into the “Western” world by way of the ancient Greeks, and that their uses of the word, like the “Eastern” uses, are problematic in their own right. In this case, the problem is not a divergence of meanings, but an evolving convergence, eventually centering on the idea, already noted, that a mágos (μάγος) is some kind of a practitioner of magic. It is Greek civilization, which we consider to be the fountainhead of Western civilization, that has given to the modern world the seemingly unified meaning of this word—by contrast with Iranian civilization, representing the “mysterious” East, where the real meaning is multiple and complicated.

§4. But there is more than one way of thinking about Greek civilization. Here the problem is hidden, because Greek civilization as articulated by the Greek language seems to be on the surface so unified, so monolithic. After all, there is and always has {|99} been only one Greek language, and it has survived to this day as one single Greek language. There is evidence to show that this language has lasted continuously for over three thousand years of world history. By contrast, when we consider this same period of three or so millennia, we find evidence for many different Iranian languages. If we track Iranian languages as they existed during this stretch of time, we find many different languages conveying many different civilizations that were or are in many ways discontinuous from each other. The only Iranian continuum that is comparable to the Greek language is represented by the languages known to linguists as Old Persian, Middle Persian, and New Persian.

§5. There is no time for me to go into details here about Iranian discontinuities, and I content myself with just one obvious example that goes back to the earliest recoverable times: when we consider the sacred texts of what is generally known as the Avesta, it is obvious that the language of these texts cannot simply be traced forward in time to match what is generally described as Old Persian, which was the official Iranian language of the Achaemenid dynasty.

§6. To resume my main argument… By comparison, as I said a moment ago, with the Iranian languages, the Greek language tends to be viewed as some kind of a miraculously ongoing continuum—as an exceptionally unified vehicle of Greek civilization. It can even be said, and I did say it already, that the Greek language is something of a monolith. Granted, there were always Greek dialects, but such subgroupings clearly belonged to a single language that we now recognize as one single Greek language. So, if the teleology of the Iranian word mágos as used in Greek points to an eventual meaning that we may tentatively define as a practitioner of magic, then such an outcome may seem to be good enough for us. But it is not at all good enough.

§7. The appearances of a monolithic Greek language can be deceptive, and this deceptiveness can become an overall problem for me as I look further into the meaning of the Greek word mágos as borrowed from the Old Persian word maguš. This overall problem can be broken down into two very specific problems with reference to the Greek word mágos:

1. It is only in relatively later Greek texts, stemming from the Hellenistic and later eras—so, in texts dated after the death of Alexander in 323 BCE—that we find unambiguous attestations of the word mágos in the sense of magician or practitioner of magic.

2. In earlier Greek texts, the word mágos is used in a different and more complicated sense. And this different sense, as we will see, operates not in the Greek language considered as a whole but only in one dialect of the Greek language. That dialect, as we will also see, is Ionic, as spoken by Ionians. And here we will be looking at a situation where the history of the Greek language cannot to be viewed monolithically: one dialect will reflect a world view that is significantly different from the world view reflected by the other dialects. And the different sense of the {|100} word mágos as used in the Ionic dialect can be viewed as a most telling sign of such differences in world views.

§8. We are about to explore, then, a glaring example of discontinuities as well as continuities in the Greek language, which turns out to be not as monolithic as it might appear to be at first sight. The example focuses on the positive as well as the negative uses of the word mágos in Greek.

Some positive and negative views of the word mágos

§9. I start with negative uses. In later Greek texts, a notorious example of the word mágos in a negative sense is the narrative about one Simon the Mage in Acts of the Apostles 8:9–24. This mágos named Simon is viewed as a magician who threatens to invalidate the authority of the apostle Peter himself in a set of confrontations. Even more vivid confrontations between Simon the Mage and Peter can be found in the Apocrypha, as in Acts of Peter 32, where we see Simon in the act of captivating his onlookers by performing for them magic tricks like levitating high in the air, in imitation of Christ ascending to heaven. Seeing Simon levitate, Peter feels threatened, and he reacts by praying that Simon should crash. Peter’s prayers are answered. Unlike Christ, Simon now comes crashing down to earth, humiliated and disempowered.

§10. In such narratives about Simon the Mage, we can see that mágoi have the potential to authorize themselves as authorities in their own right, and this potential is seen as esoteric and even exotic. Viewed negatively, such a potential for esoterism and exoticism can be demonized as magic.

§11. Viewed positively, however, the esoterism and exoticism of mágoi can be exalted as the power of authority, which can validate authority itself. Such a positive view is evident in the Gospel according to Matthew 2:1–12, where the wise men who visit the infant Jesus are mágoi from the exotic ‘East’ whose role is to recognize and thus to validate the kingly authority of the divine child.[6]

§12. From such relatively later Greek contexts, then, we can see that mágoi have the potential to validate as well as to invalidate authority, and that only their potential to invalidate could be viewed negatively as magic. By contrast, if this same potential of the mágoi is viewed positively, then it becomes an exalted and even sacred power.

§13. That said, what can we say about the relatively earlier Greek contexts of the same word mágos? Here we will need to be all the more careful in avoiding the assumption that this word consistently conveys the idea of magic, which as we have seen is only a negative view of the word as used in its later Greek contexts. The positive view, as we have also seen, is that the potential of the mágos to validate authority must be a sacred power. Further, as we will soon see from the evidence of earlier {|101} Greek contexts, such a positive view must have allowed for the power of the mágos to invalidate an authority that is supposedly bad, not only the power to validate an authority that is supposedly good.

§14. When the authority is supposed to be good, it is visualized as a form of kingship that is sacralized by a divinity who protects kingship. In the case of the Bisotun Inscription, for example, the god Ahuramazdā protects the kingship of Darius. In this case, however, as we read in the text of the Bisotun Inscription, the rebellious maguš is supposedly bad, not good, and only the king Darius is good. Correspondingly, in the matching Greek text of Herodotus, the mágos there too is supposedly bad, not good. In this case, then, it is presumed that a good king has invalidated a bad mágos. In an alternative universe, however, it could have been a good mágos who invalidated a bad king. Either way, the point remains this: in the relatively early Greek context where Herodotus narrates events that match what is narrated in the Old Persian text of the Bisotun Inscription, the bad mágos is viewed not as a magician. Rather, in terms of the negative view as shaped in the Greek text of Herodotus, the supposedly bad mágos is a false king, a usurper. And, similarly in the Old Persian text of the Bisotun Inscription, the negative view of the supposedly bad maguš presents him not as some magician but as a false king, a usurper.

§15. Still, lexicographers tend to interpret mágos as conveying the idea of a magician even when the word is found in earlier Greek contexts. My favorite example comes from Sophocles, Oedipus Tyrannus 387, where we see one of the earlier Greek attestations of the word: here the king Oedipus refers to the seer Teiresias insultingly as a mágos. It is commonly thought that the king in this context is accusing the seer of acting like some kind of a magician. But is Teiresias here really a practitioner of magic, as we think of magicians? Or is he more of a usurper? Here I return once again to the Old Persian word maguš, as attested in the Bisotun Inscription: there, as we saw from the beginning, the word refers to a man who was accused by the king Darius of usurping the kingship of the Persian Empire (Column 1 lines 35–43, etc.). In that context, the maguš is viewed as an evil usurper, not as an evil magician, and his success in usurping the kingship is not linked with magic. A similar formulation applies to the mágos in the History of Herodotus (3.61–79): here too the role of this mágos is evidently the same, a usurper. And a similar formulation applies even to Teiresias in the Oedipus Tyrannus of Sophocles: when the king Oedipus calls Teiresias a mágos at verse 387, he is in effect accusing the seer of threatening the king’s kingship. And what exactly is so threatening about the role of Teiresias as a would-be mágos? My answer is this: the knowledgeability of Teiresias about the cosmos—both physical and political—is viewed here as a potential threat, not only as a potential aid, to kingship. And, as we will now see, such knowledgeability is exactly what can seem to be so threatening about the role of the mágoi in the overall context of the Persian Empire. {|102}

Ionian views of the word mágos

§16. So far, we have found nothing inherently magical about mágoi in early Greek sources as represented by Herodotus. But there is, as we will now see, something inherently priestly about them. Such a view of the mágoi, as priests of some kind, was transmitted into the Greek language by way of one dialect in particular, and that dialect was Ionic as spoken by Ionians. As I have already indicated, I will be arguing that speakers of the Ionic dialect were transmitters of a world view that was different from other world views as reflected by speakers of other Greek dialects, and that this difference in world views is exemplified by the use of the word mágos by Ionians during a specific period of their history. In what follows, I will concentrate on the historical context of that period as I proceed to analyze the special use of the word mágos by Ionians.

§17. In the course of my analysis, I will re-examine the relevant evidence of two major Greek sources: one is the History of Herodotus, composed in the Ionic dialect, and the other is the Derveni Papyrus. More needs to be said about the second source. Though the writing-down of the text that we read in the Derveni Papyrus is generally dated to the fourth century BCE, the author of the text itself can be dated further back, to the fifth century BCE, and so he may be roughly contemporaneous with Herodotus himself.[7] Moreover, just as the text of Herodotus was composed in Ionic, that is, in the dialect of the Ionians, so too the language used by the author of the Derveni Papyrus is heavily influenced by the Ionic dialect.[8]

§18. Both of these two fifth-century Greek sources provide essential pieces of information about the role of mágoi in the era of the Achaemenid dynasty, which ruled the Persian Empire from around 550 to 330 BCE.

§19. According to Herodotus (1.101), the Mágoi (Μάγοι) were one of six social subdivisions (éthnea) of the Iranian people known as the Medes (Μῆδοι). Even in the Bisotun Inscription, the Babylonian portion of the trilingual text refers to the maguš there as a ‘Mede’ (line 15: ma-da-a-a … ma-gu-šu, /madaya magušu/). Accordingly, we may at first sight think of the mágoi as a priestly class within the overall framework of the Persian Empire. A closer look, however, suggests that the mágoi were already a priestly class in the empire of the Medes, which preceded the more multiethnic Persian Empire, and that the status of mágoi as imperial priests in this later empire was no longer so strictly linked to their original status as an ethnic grouping that was also sacerdotal within the empire of the Medes themselves.[9] {|103}

§20. I concentrate here on what we read in Herodotus (1.132.2–3) about the role of the mágos (μάγος … μάγου) in authorizing thusíai (θυσίας) ‘sacrifices’ by way of epaeídein (ἐπαείδει) ‘singing incantations’, where the epōidḗ (ἐπαοιδήν) ‘incantation’ is described as a theogoníē (θεογονίην) ‘theogony’.

§21. Keeping in mind such a role for the mágoi as sacrificers, I now highlight what we read in column VI of the Derveni Papyrus. Here it is said that the mágoi (μάγων at line 2, μά̣[γο]ι̣ at line 5, μάγοις at line 9) are models for the performance of rituals by initiands (lines 8–9), where the word I am translating as ‘initiands’ is mústai (μύσται at line 9). Here is the way I translate the relevant wording (at lines 8–9): ‘the mústai perform-preliminaries-of-sacrifice [pro-thúein] to the Eumenides in the same way as the mágoi perform-preliminaries-of-sacrifice’ (μύσται | Εὐμεν̣ίσι προθύουσι κ[ατὰ τὰ] α̣ὐτὰ μάγοις). Within this overall context, we find three most relevant words that seem to be connected with the activities of the mágoi:

1. eukhḗ ‘prayer’ ([εὐ]χ̣α̣ί at line 1, to be compared with εὐχη̣̃ς at column I line 6)[10]

2. thusíā ‘sacrifice’ (θυσ[ί]α̣ι correlated with ([εὐ]χ̣α̣ί, at line 1; also θυσ[ία]ν̣ π̣[οιοῦσ]ι̣[ν] … | οἱ μά̣[γο]ι̣ at lines 4–5)

3. epōidḗ ‘incantation’ (ἐπ̣[ωιδὴ δ]ὲ̣ μάγων at line 2).

§22. The first and the third of these words are in regrettably fragmentary condition, but the majority of editors agree on the restorations that I have given here. All three of these words are most relevant to what I cited just a moment ago from Herodotus (1.132.2–3), who speaks about the role of the mágos (… μάγος … μάγου …) in authorizing thusíai (θυσίας) ‘sacrifices’ by way of epaeídein (ἐπαείδει) ‘singing incantations’, where the epōidḗ (ἐπαοιδήν) ‘incantation’ is described as a theogoníē (θεογονίην) ‘theogony’.[11] Here we see the idea of singing about the cosmos—even singing the cosmos—linked with the idea of authorizing a sacrifice. From a comparative point of view, as we know from surveying a wide variety of cross-cultural evidence, the performance of a theogony can function as an authorization of kingship. We read similar ideas about the correlation of cosmos and kingship in the work of Xenophon known as the Education of Cyrus (8.1.23) with specific reference to the ritual activities of the mágoi: two relevant words used there are humneîn (ὑμνεῖν) ‘performing hymns’ and thúein (θύειν) ‘sacrificing’. Here too, mágoi are viewed as coefficient with the king of the Persian Empire in sustaining cosmic order by way of sacrifice and incantation.

§23. Even more relevant is what we read in Herodotus (7.191.2) about mágoi who are described as ‘sacrificing’ (éntoma poieûntes) and ‘singing incantations’ (kataeídontes {|104} boêisi) in response to a violent wind that had at that time seriously damaged the fleet of the Persian Empire by destroying many of its ships anchored in the shallows at Cape Sepias. The context here makes it clear that the mágoi were performing these rituals in order to salvage the royal project of Xerxes, successor to Darius as king of the Persian Empire, who was attempting to conquer the Greeks inhabiting the mainland situated on the European side of the Aegean Sea. In the same narrative of Herodotus (7.191.2), these mágoi are described as also sacrificing (thúontes) to the goddess Thetis, mother of Achilles, and to her sister goddesses, the Nereids. Herodotus says more (again, 7.191.2): the mágoi sacrificed (éthuon) to Thetis because of what they had heard ‘from the Ionians’ (parà tôn Iṓnōn). What they heard, as Herodotus recounts (again, 7.191.2), is a sacred narrative about Cape Sepias, the place situated on the European coast where the fleet of the Persian Empire was damaged by the violent wind: it was at this same place, the Ionians told the mágoi, that Achilles was conceived when his immortal mother Thetis was impregnated by his mortal father Peleus.

§24. Given the fact that the native Greek narratives attributed cosmic powers to the goddess Thetis, I argue that we see here another example of a link between the idea that the mágoi have the power to sing the cosmos and the idea of their authorizing a sacrifice by way of their incantations.

§25. But who were these Ionians who had narrated for the mágoi the sacred narrative that led to the magian sacrifice to divine forces worshipped by the Greeks? It is clear from what we read in Herodotus (7.191–192) that these Ionians were Asiatic Greeks fighting on the side of the Persians and against the Greeks of Europe who were defending their homeland from the invasion initiated by Xerxes. These Ionians were serving in the invading navy of the multiethnic Persian Empire.

§26. By studying the political and cultural identity of these Greeks who called themselves Ionians, I argue, we can develop a holistic explanation for the meaning of the word mágos in Greek. And that is because these Greeks, as Ionians, were the primary transmitters of Persian civilization beyond the Persian-speaking world.

Ionians and the Persian Empire

§27. The point that I just made about Ionians as the primary transmitters of Persian civilization beyond the Persians is I think the most consequential aspect of my overall argumentation. When I say Persian civilization in this context, I use the term in an inclusive rather than exclusive sense. Such a civilization was not at all exclusively Persian in its ethnicity, since the Persian Empire promoted a multiethnic world {|105} view that prided itself on its Greek-speaking constituency as a singularly prestigious aspect of its overall civilization.[12]

§28. I am arguing that the civilization of the Ionians was a jewel in the crown for the Persian Empire—at least, from an Ionian perspective. And such an argument requires a rethinking of four subjects, three of which are ideas and one of which is a historical fact:

1. the idea of ‘Ionians’ as defined by the so-called Ionian Dodecapolis, a confederation of twelve states, which took shape in the late eighth and early seventh centuries BCE

2. the idea of a Debate of the Constitutions, as dramatized by Herodotus (3.80–84), which supposedly took place among Persian elites in 522 BCE

3. the historical fact of an Ionian Revolt against the Persian Empire in 499–494 BCE

4. the idea of ‘Ionians’ as viewed after the sea battle at Salamis in 480 BCE.

§29. I have listed these four subjects in chronological order here, but the order of my analysis will be 1-4-3-2.

The idea of ‘Ionians’ and the Ionian Dodecapolis

§30. The name of the region Ionia (Iōníē), and the name of the people, Ionians (Íōnes), are hardly even recognized in the popular cultures of the “West.” Nor are they all that well understood by experts in the history of Greek civilization. The name Ionian is ordinarily used today with reference to populations that spoke Ionic, which was a dialect of the unified language that was Greek. This Ionic dialect, as linguists can tell us, was closely related to another dialect, Attic, spoken in Attica, which was the name given to a region dominated by the city-state of Athens in the classical period of the fifth century BCE and beyond. Linguists think of Attic as a dialect spoken in one part of a larger dialect-area that they describe as Attic-Ionic, while Ionic is understood to be the dialect spoken in the other part. But how to define that other part? Whereas Attica, the original area where Attic was spoken, can be defined easily, since it is situated on the mainland of what we know as Europe, it is much more difficult to define the areas where Ionic was spoken.

§31. The fact that I need to say plural areas and not singular area is already an indication of the difficulties that await us. Contributing to these difficulties is the fact that the Ionians who were most active and influential in defining Ionian identity were exclusivist in using the name Ionian. As we are about to see, these exclusivist Ionians were the proud inhabitants of a region they called Ionia, situated in what we know as Asia Minor, on the east side of the Aegean Sea. They mythologized themselves as {|106} descendants of adventurers who had followed kingly leaders stemming from Athens, capital city of Attica in Europe, in crossing over from the west side of the Aegean Sea and establishing on the east side twelve states at sites located along the central coastline of mainland Asia Minor and on two important offshore islands.

§32. These twelve states were consolidated into a confederation known as the Ionian Dodecapolis. Herodotus (1.142.3) lists all twelve states: of these twelve, the ten Asiatic mainland city-states were Miletus, Myous, Priene, Ephesus, Colophon, Lebedos, Teos, Klazomenai, Phocaea, and Erythrai, while the two Asiatic island-states were Samos and Chios.

§33. The self-mythologizing of Ionian states as offshoots of settlements originating from Athens can be seen as the most prominent feature of a much broader set of myths that can be categorized under the heading Ionian Migration, where the Greek word that I translate as ‘migration’ is apoikiā (as used in Strabo 13.1.3 C582 and elsewhere).[13] Although the mythmaking of the Ionian Dodecapolis must have screened out or at least minimized alternative myths that would have aetiologized the founding of other Ionian states that were not members of the twelve-state confederation, we have evidence for the survival of such alternative myths, as we see for example from a passing remark made by Pausanias (7.3.3) concerning an Ionian migration from Athens to Naxos—which was one of the most important islands in a chain of islands known as the Cyclades. Evidently, the Ionian populations of the Cyclades were excluded from the confederation of the Ionian Dodecapolis. Another example of exclusion is Euboea, a large and most important Ionian island situated offshore to the east of Attica.

§34. Even if we define the Ionians narrowly, restricting the field of vision to the Asiatic Greeks of the Ionian Dodecapolis, I maintain that their importance was still paramount in the history of Greek civilization. But the problem is, the achievements of these Ionians, especially in science and in philosophy, are today generally seen as merely an aspect of Greek civilization viewed as a unified whole. It is almost as if the localization of such achievements by Ionians inhabiting the Asiatic region called Ionia were simply a historical accident. Or, to put it another way, it is as if the Asiatic Greeks of the Ionian Dodecapolis were no different from European Greeks.

§35. But the fact is, the Asiatic Greeks of the Ionian Dodecapolis had their own distinct Greek identity—or, as I would prefer to say, their own self-defined ethnic identity. For these Ionians of the Ionian Dodecapolis, this ethnic identity was formalized by way of celebrating a seasonally recurring festival named the Paniōnia, the venue for which was a place named the Paniōnion. As Douglas Frame has shown in an important book about the myth of Nestor, king of Pylos (Frame 2009), the apogee {|107} of this Ionian confederation—and of the festival of the Paniōnia as celebrated at the Paniōnion—can be dated to the late eighth and early seventh centuries BCE, and there survives a specific reference already in Homeric poetry to the festival that formalized the identity of the celebrants (Iliad 20.403–405).

§36. As Frame (2009: ch.10) has also shown, the original leader of this twelve-state confederation was the city-state of Miletus, and the rulers of this state claimed that they were descended from Neleus father of Periklymenos, who was a brother of Nestor, and that Periklymenos was an ancestor of Kodros, a king of Athens, who in turn was the father of a second Neleus. The Athenian Neleus founded not only Miletus but also, ultimately, the entire confederation of the Ionian Dodecapolis—according to the mythology propagated by Miletus about the Ionian Migration, as we read in a fifth-century source, Hellanicus (FGH 323a F 11; see also a later source, Parian Marble FGH 239F A27; see also Suda under the entry Πανύασις, where the early-fifth-century poet Panyasis is cited as another source).

§37. Alternatively, however, as we read in another fifth-century source, Pherecydes (FGH 3 F 155, via Strabo 14.1.3 C632–633), the leader of the Ionian Migration from Athens to Ionia (ἄρξαι δέ φησιν ῎Ανδροκλον τῆς τῶν Ἰώνων ἀποικίας) was another son of Kodros the king, Androklos of Athens, who was known as the ‘legitimate’ son of Kodros (υἱὸν γνήσιον Κόδρου τοῦ Ἀθηνῶν βασιλέως) and who was the founder of the city-state of Ephesus (γενέσθαι δὲ τοῦτον Ἐφέσου κτίστην), establishing in that city a basíleion ‘royal center’ for all the Ionians (διόπερ τὸ βασίλειον τῶν Ἰώνων ἐκεῖ συστῆναί φασι)—which is why, says this source, the descendants of Androklos in Ephesus retained the title of basileîs ‘kings’ (καὶ ἔτι νῦν οἱ ἐκ τοῦ γένους ὀνομάζονται βασιλεῖς). So, the city-state of Ephesus, which as we have already seen is listed by Herodotus as a member of the Ionian Dodecapolis, was a rival of Miletus in claiming to be the city-state that dominated the new homeland of the Ionians in Asia Minor.

§38. What we see here are two mutually contradictory versions of an overall myth about the Ionian Migration. According to one version, the state of Miletus was clearly in the forefront of a grand movement of Ionians traveling from Athens to their new home in Asiatic Ionia, while in the other version it was the state of Ephesus that took the lead—and became the dominant city in Ionia.

§39. An explanation is needed for the emergence of Ephesus, at the expense of Miletus, as the dominant city of the Ionian Dodecapolis. As I will argue, this dominance took shape in an era that postdated by around two centuries the apogee of the Ionian Dodecapolis, which as I already said can be dated to the late eighth and early seventh centuries BCE. During those two centuries, as I will also argue, it was still Miletus and not Ephesus that maintained a position of dominance over the Ionian Dodecapolis.

§40. The question for now is this: how did Ephesus emerge as a serious rival to Miletus? The answer, as we will see, has to do with the fact that the entire region of {|108} Ionia was for a long time under the domination of the Persian Empire, which was exerting its control over the cities of Ionia through the primary agency of the satrap of Lydia, stationed in the inland city of Sardis, the capital of what had once been the Lydian Empire. For further background, here are two basic historical facts about the era of Persian domination over Ionia:

1. The Persian domination of Ionia got underway in 546 BCE, when the Persian Empire overwhelmed the Lydian Empire, which had dominated mainland Ionia beforehand.

2. During the era when the Ionian Dodecapolis was dominated by the Persian Empire, it was secondarily dominated by the city-state of Miletus—and Miletus must have maintained this secondary domination of Ionia till the end—about which I will have more to say in §41. [[The wording toward the end of this sentence was rewritten 2018.11.23.]]

§41. The end for Miletus came with the Ionian Revolt, which lasted from 499 to 494 BCE. This revolt was started by Miletus and ended with the capture of this city by the Persian Empire in 494 BCE. This disaster, as narrated by Herodotus (6.18), was commemorated in a tragedy produced soon thereafter in Athens, composed and directed by the dramatist Phrynichus (6.21.2).

§42. I will postpone till later my analysis of the Ionian Revolt, except to say this much already now: before the end came, Miletus could still claim to be the primary originator of the Ionian Migration. Most relevant is what we read in Herodotus: when Aristagoras of Miletus went to Athens for the purpose of persuading the Athenians to come to the aid of the Ionian Revolt (5.97.1), he referred to the myth of the Ionian Migration: or, to put it in terms of the rhetoric paraphrased by Herodotus, Aristagoras reminded the Athenians of the claim that the people of Miletus were ápoikoi ‘colonists’ originating from the city of Athens (5.97.2)—in other words, that Athens was the originator of the apoikíā or ‘migration’ of Ionians to the East.

The idea of ‘Ionians’ as viewed after the sea battle at Salamis

§43. The ethnic self-definition of the Asiatic Greeks who called themselves Ionians (Íōnes) must be contrasted with a rival ethnic self-definition of European Greeks who called themselves Hellenes (Héllēnes). To drive this point home, I start by asking a hypothetical question. This question centers on a major event in world history, namely, a sea battle that took place at the island of Salamis in 480 BCE, where the naval forces of Athens, combined with smaller naval forces sent by the island state of Aegina and by other allied states, decisively defeated the combined naval forces of the Persian Empire. The question is, what would have happened if the Persian Empire had won this naval battle, and if Athens and its allies had lost? Well, in terms of my argumentation, the Persian Empire would have conquered European Greece, and there would have been, after that, no more Hellenes in Europe. There would continue to be Greek-speaking populations there, yes, and they would even be the same populations as before, but they would not be calling themselves Hellenes any more. {|109} Instead of Hellenes, there would now be Ionians in Europe—that is, in the region of Athens known as Attica and in such outlying islands as nearby Euboea— just as there were Ionians already in Asia Minor. To put it more broadly, Greekness would have been reconfigured in terms of the Eastern Greeks, with the obliteration of Greekness as configured by the Western Greeks of Europe who had up to now called themselves Hellenes.[14]

§44. This formulation is relevant not only to the European Greeks who were Ionians but also to other European Greeks who were Aeolians and Dorians. If the forces of the Persian Empire had been victorious in the sea battle at Salamis, then there would have been a more general re-alignment affecting

1. the Ionians of Europe, like the populations inhabiting Attica and the outlying island of Euboea

2. the Aeolians of Europe, like the populations inhabiting Boeotia and Thessaly

3. the Dorians of Europe, like the populations inhabiting most of the land-masses situated to the north as also to the south of the Corinthian Gulf.

§45. These populations of European Greeks, who had all been calling themselves Hellenes at this moment in history when the sea battle took place, would thereafter have been re-aligned respectively with the following populations of Asiatic Greeks:

1. the Ionians in the central coastlands of Asia Minor as also in the outlying islands of Chios and Samos

2. the Aeolians in the northern coastlands of Asia Minor as also in the outlying islands of Lesbos and Tenedos

3. the Dorians in the southern coastlands of Asia Minor as also in the outlying islands of Rhodes and beyond.

§46. But none of this happened. In the end, after the defeat of the Persian Empire’s naval forces in the battle of Salamis and, later, after the defeat of its land forces at Plataea, the Greekness of the Hellenes in the West prevailed over the Greekness of the Ionians and the Aeolians and the Dorians in the East. And, hurrah, the democracy that had been the new way of life for Greeks of the West as represented primarily by Athens could now prevail over the tyranny that was the old way of life for the subjugated Greeks of the East as represented primarily by the Ionians.

§47. What I have just formulated is a common contemporary understanding of what happened after the sea battle at Salamis in 480 BCE—followed by the land battle at Plataea in 479 BCE. But this formulation is oversimplified. In the case of Ionia, for example, we are about to see that the cities of that region were at this same moment in history governed by regimes that were considered to be democracies, and {|110} such democratic regimes had been installed at the initiative of the Persian Empire. As I am about to argue, the victory of the European Greeks at Salamis and at Plataea in 480 BCE and 479 BCE may have been seen as a victory for democracy as we know it, yes, but the fact remains that an alternative victory for the Persian Empire could have been likewise ideologized as a victory for democracy—at least, for those Greeks who were Ionians. And such an ideology would have been promoted by the Persian Empire itself. In terms of the multiethnic ideologies of this empire, I argue, such a victory would have been equated with a victory not for Persians but rather for Ionians—that is, for the Asiatic Greeks who populated Ionia. At this point in history, as we are about to see, these Ionians could have promoted an alternative version of democracy that supposedly rivaled the democracy as configured in European Athens.

§48. Before I can proceed with my argument, I need to highlight a basic fact of history. As a consequence of the sea battle of Salamis in 480 BCE, where the navies of the Persian Empire were on the losing side while the navy of Athens was on the winning side, the dominion over Ionia and over other Greek-speaking regions in the East was lost by the Persian Empire and was won by Athens, which thereafter evolved into what we know retrospectively as the Athenian Empire. To say it more bluntly, Athenians became the super-Greeks of the East. In the Greek-speaking world of the East, the Athenians could now dominate the Ionians in the form of the Delian League, but eventually they dominated also the Aeolians to the north of the Ionians and the Dorians to the south. And the domination extended even further, to the regions of the Hellespont and the Propontis and beyond. An architectural landmark of this new dominion of Athens over the Asiatic Greeks was the Odeum of Pericles, built to replicate the Grand Tent of Xerxes, which had been confiscated by the Athenians after their victory in the naval battle at Salamis. Plutarch (Life of Pericles 13.9–11) gives a most vivid description of this new imperial building, adjacent to the Theater of Dionysus: here was a People’s Palace for all Asiatic Greeks to admire as an orientalizing stupor mundi, situated in the heart of a European city that could now also play the rôle of an Asiatic metropolis.[15]

§49. The new domination of Eastern Greeks by Athens was viewed by most Greeks as tyranny—a new form of tyranny that superseded the old domination of these same regions by the Persian Empire. Yes, Athens could still be considered a democracy on the inside, but it had now become a tyranny on the outside (Pericles is quoted as saying this much by Thucydides 2.63.1–2).[16] Ionia could now be controlled from Athens in Europe, not from Sardis in Asia Minor. And a parallel formula of control could now be extended to the Aeolians and to the Dorians situated respectively to the north and to the south of the Ionians in Asia. {|111}

§50. By contrast, if the forces of the Persian Empire had won at Salamis, the new super-Greeks would have been not the Athenians as controllers of Ionians and beyond but the Ionians as controllers of Aeolians to the north and of Dorians to the south in Asia Minor. And, in fact, the prototype for such a smaller-scale empire was already visible in the context of what historians today call the Ionian Revolt. As I will now argue, such a smaller-scale empire could be described as an Ionian Empire in the making. But such an empire was only in the making, since the Ionian Revolt ultimately failed.

Ionian Revolt

§51. Most relevant to the historical contingencies of Ionian identity was the Ionian Revolt, which lasted from 499 to 494 BCE. This revolt of Ionians from the Persian Empire—along with its grim aftermath—is documented primarily and in fact almost exclusively by Herodotus (5.23.1-6.42.2), whose narrative highlights the many successes and failures experienced by both sides in the conflict, which in any case culminated in the decisive victory of the Persian Empire.

§52. In undertaking my analysis of the historical background, I will focus on four Ionian states as the primary settings for various events that I will be foregrounding:

1. the city-state of Miletus, situated on the mainland in Asia Minor

2. the city-state of Ephesus, situated on the mainland in Asia Minor

3. the outlying island-state of Samos

4. the city-state of Athens, situated on the mainland on the other side of the Aegean Sea, in Europe.

§53. And I will focus on four relevant historical facts:

1. The city-states of Miletus and Ephesus, as well as the island-state of Samos, were all members of the Ionian Dodecapolis.

2. The city-state of Miletus claimed Athens as its mother-city or mētropolis. Such a claim was actualized, as we have seen, in the Milesian version of myths about an Ionian Migration.

3. The Ionian Revolt, led by the state of Miletus, failed partly because the state of Samos defected at a critical moment in a decisive sea battle at Lade in the year 494 BCE. As for Ephesus, which like Samos was becoming a most serious rival of Miletus in claiming leadership of the Ionian Dodecapolis, the forces of this city were not even present at the sea battle of Lade. Our primary source for such telling details is Herodotus (6.14.2–3).

4. The failure of the Ionian Revolt led to the capture of Miletus and to its obliteration as a viable state in 494 BCE, followed by other harsh forms of retribution inflicted on other Ionian states by the victorious forces of the Persian Empire in 494/3 BCE. Again, our prime source is Herodotus (6.18 and 31–32). At this point, as I infer {|112} from the account of Herodotus (6.33), much of the punitive naval action against the Asiatic Greeks seems to be blamed on the Phoenicians.

§54. Finally, I will focus on four most relevant dramatis personae in the story as narrated by Herodotus:

1. Histiaios. He was tyrant of Miletus and the foremost ally of Darius, king of the Persian Empire, in an imperial expedition against the nomadic Scythians in 513 BCE. Our primary source here is Herodotus (4.137–138). In this context, it is made clear that the Ionian city-states, including Miletus, were at that time ruled by tyrants who had been installed by the Persian Empire, and the relevant word used here by Herodotus is turanneúein ‘rule as a tyrant’ (4.137.2). As tyrant of Miletus, Histiaios assumed the role of chief spokesman for the Ionians, and Herodotus quotes him as saying that all the tyrants of the Ionians owe their political power to Darius, without whose domination the Ionians would rather ‘have democracies’ than ‘have tyrannies’—and the relevant words used here are dēmokrateîsthai and turanneúesthai (again, 4.137.2). Although Histiaios collaborated with Darius, he had so much personal power in Ionia that he was eventually relocated to the Persian capital (Herodotus mentions only the capital at Susa, not the other capital at Persepolis), supposedly as the honored guest of Darius (Herodotus 5.25 and 30).

2. Aristagoras. He became the next tyrant of Miletus. At the time of his rule, Miletus maintained its status as the most eminent and successful state among all the states in Ionia. Again, our primary source is Herodotus, who describes Miletus as ‘the pride [proskhēma] of Ionia’ (5.28: τῆς Ἰωνίης … πρόσχημα). In this context, as Herodotus reports further (5.38.1), Aristagoras became the main instigator of the Ionian Revolt, and he arranged for the tyrants of the various Ionian cities, as formerly installed by the Persian Empire, to be removed. I interpret these actions as indications of an overriding motive: Aristagoras and his co-conspirators were attempting to form an Ionian Empire, dominated by Miletus, and this empire would be independent or quasi-independent from the Persian Empire.[17] To support my argument that the motive of Aristagoras was to form some kind of a breakaway empire, I cite the reportage of {|113} Herodotus (5.109.3) concerning the expression used by the delegates of the Ionians in referring to their constituency at this time: tò koinòn tōn Iṓnōn ‘the commune of the Ionians’. This same term koinón ‘commune’, as I have argued in another project, was used decades later by the Athenians in referring to their own empire (as we see for example by way of Thucydides 2.60.4, 2.61.4).[18] I will have more to say presently about the eventual Athenian Empire, but for now I must continue to concentrate on the would-be Ionian Empire. Although the motive of Aristagoras in instigating the Ionian Revolt was imperialistic, as I argue, he claimed that his motive was democratic: as Herodotus puts it, Aristagoras renounced his own turannis ‘tyranny’ in Miletus and proclaimed as its replacement the principle of isonomíē ‘equitable participation’ for the city (5.37.2). Here I must return to a passage of Herodotus that I have already cited earlier (3.80–84), about the Debate of the Constitutions, the dramatic date for which is 522 BCE. In the context of this Debate as dramatized by Herodotus, the same word isonomíē is used to express the idea of democracy, advocated by one of the three debating Persian nobles, Otanes, as the best form of government for the Persian Empire (3.80, 3.83). At a later point in the narrative, Herodotus will cross-refer to this passage, and, in the context of his cross-reference, the word he uses for the idea of democracy is now dēmokratíē (6.43.3), which is clearly meant here as a synonym of isonomíē. I should add that Herodotus uses the same word dēmokratíē (6.131.1) in referring to the establishment of democracy in Athens by Cleisthenes in 508/7 BCE. I should also add here a detail that I have already noted: Herodotus uses the word dēmokrateîsthai ‘have democracy’ (4.137.2) in referring to the negative attitude of Histiaios, tyrant of Miletus, toward the very idea of democracy. I will soon consider further the context of the reference made by Herodotus to dēmokratíē in his cross-reference to the Debate of the Constitutions (6.43) but for now I will stick to the present context of his reference to the isonomíē (5.37.2) declared by Aristagoras for the state of Miletus in 499 BCE. It is relevant to such a context that the city-state of Athens is linked with this declaration of Aristagoras. Recruited as supporters of the Ionian Revolt were two Ionian city-states on the European side of the Aegean Sea. According to Herodotus (5.77 and 5.99) one of these states was Eretria, an Ionian city situated on the island of Euboea and an old ally of Miletus; and the other state was, not coincidentally, Athens, the notional mother city of Miletus. The government of Athens at this time had been a dēmokratíē ever since this form of government was established there by Cleisthenes in 508/7 BCE, and this democratic identity would have been compatible with the ideology newly declared by Aristagoras for Miletus and, by extension, for all the states of Ionia. After the Ionians were joined by the Athenians and the Eretrians, they undertook an ultimately unsuccessful attack on Sardis in the year 498 BCE. After this serious provocation against the Persian Empire, {|114} there was no turning back for the Ionians participating in the Revolt, as Herodotus darkly observes (5.103). And these Ionians now included not only the states of the Ionian Dodecapolis but also the states of Athens and Eretria. But now let me finish with Aristagoras: during the final phases of the Ionian Revolt, which was already collapsing, Aristagoras had to leave Miletus and was killed in the course of a military side-adventure in Thrace (5.126).

3. Artaphernes. He was a brother of Darius and ruled Ionia by virtue of his royal appointment as satrap of Lydia, with headquarters in the inland city of Sardis.[19] As we read in Herodotus (5.70–74), Artaphernes was involved in a political crisis that took place in the city of Athens after 510 BCE, which was the year when Hippias, of the lineage of the Peisistratidai, was overthrown and exiled as tyrant there: at around 507 BCE, in the context of an ongoing effort by proponents of democracy as led by Cleisthenes to seek external allies in their political struggle against proponents of oligarchy as led by Isagoras, ambassadors were sent from Athens to Artaphernes, satrap at Sardis, and they agreed to offer tokens of earth and fire, thus signaling submission to the Persian Empire (5.74).[20] Around eight years later, as we read in Herodotus (5.96), Artaphernes sent a threatening message to the Athenians, ordering them to undo their prevailing democratic regime and to restore to power the lineage of the Peisistratidai, who remained exiled from Athens. It was at this same point, says Herodotus (5.97.1), that Aristagoras of Miletus happened to arrive in Athens for the purpose of asking the Athenians to come to the aid of the Ionian Revolt. As I have already noted, one of the arguments presented by Aristagoras to persuade the Athenians to join the Ionians in their revolt was his reference to the myth of the Ionian Migration (5.97.2). After noting this new involvement of Artaphernes in the politics of the Athenians, Herodotus has no more to say about this Persian until his narrative reaches the final phases of the Ionian Revolt. It was near the time of the bitter end, as we follow the narrative of Herodotus (6.30), that this same Artaphernes got involved in the capture of Histiaios, whom he promptly executed. Herodotus (6.32) goes on to narrate the horrific obliteration of Miletus in 494 BCE together with the ghastly retributions suffered by other Ionian states in the grim aftermath, and he concludes this part of the narrative by gravely observing that the Ionians, now that their revolt was utterly defeated, became once again what they had been before, that is, slaves of the Persians. In the course of narrating this bitter end, Herodotus keeps Artaphernes out of his narrative. But then, Herodotus (6.42) starts to narrate what seems to be a {|115} new beginning of sorts, and, at this point, Artaphernes is reintroduced into the narrative: the satrap now convenes representatives of the remaining Ionian states and proceeds to set up new treaties with them—arrangements that are described in most positive terms by Herodotus. There is another reference to these new arrangements by Diodorus of Sicily (20.25), and we learn from this later source that a chief negotiator for the Ionians in their dealings with Artaphernes was Hecataeus of Miletus. As I argue in another project, Herodotus thought of this Hecataeus as one of the lógioi ‘word-masters’ of the Persians; in other words, the Persian “party line” about the wars between Hellenes and Persians was represented not by Persians but by Ionians like Hecataeus who spoke as collaborationists promoting the agenda of the Persian Empire.[21] As we know from Herodotus (5.36.2–4), Hecataeus had been an outspoken opponent of the Ionian Revolt.

4. Mardonios. A son-in-law of Darius, he was put in charge as the leader of a military expedition sent to Europe in 493 BCE for the purpose of punishing Athens and Eretria, which had been the two Ionian city-states in Europe that had supported the revolt of the Asiatic Ionians. That is what Herodotus (6.43) reports, adding that the real motive was to extend the power of the Persian Empire to the European side of the Aegean Sea. At this point, the narrative of Herodotus (again, at 6.43) reveals a most interesting detail about the actions of Mardonios as the generalissimo of the military forces that had been sent off to subdue the European Greeks: advancing from the south of Asia Minor toward the north, where he would make his eventual crossing into Europe at the Hellespont, Mardonios separated from his infantry and sailed with his navy along the entire coastland of Ionia, traveling from south to north, and what he proceeded to do upon arrival at each Ionian city was to dismiss each túrannos ‘tyrant’ that had been appointed by the Persian Empire, establishing dēmokratíai ‘democracies’ in their place (δημοκρατίας κατίστα). We see a double-headed strategy here: while Artaphernes as satrap at Sardis was establishing new treaties between the Persian Empire and the states of Ionia, Mardonios as generalissimo of the empire’s armed forces was ensuring a democratic base for all these Asiatic Ionians. At this point, Herodotus (again, at 6.43) makes a most ostentatious remark: the Greeks of his own generation, he says, hearing more than a half-a-century later what he has just said about things done by a general of the Persian Empire (in 493 BCE), would find it unbelievable that Mardonios could have done such things in Ionia, namely, that ‘he was establishing democracies in the cities’ (δημοκρατίας κατίστα ἐς τὰς πόλις)—just as it may seem unbelievable for these Greeks who are living in the time of Herodotus to hear that there had once taken place, at an even earlier time (522 {|116} BCE), the so-called Debate of the Constitutions, on which occasion some of the Persian elites actually argued that they needed to establish a democracy for themselves (ὡς χρεὸν εἴη δημοκρατέεσθαι Πέρσας). In this context, then, Herodotus is ostentatiously cross-referring to an earlier moment in his History where he narrates this Debate of the Constitutions (3.80–84)—and where he says that one of the three debating Persian nobles, Otanes, actually advocated democracy and not monarchy or oligarchy as the best form of government for the Persian Empire. But Herodotus insists (again, at 6.43) that Mardonios did in fact establish democracies throughout Ionia before he proceeded with his navy to the Hellespont, where he would be joined by his infantry. I will not proceed to review what Herodotus narrates (again, at 6.43) about the misfortunes that awaited Mardonios and his combined armed forces after they all crossed the Hellespont—except to note that they got bogged down in the region of Mount Athos, and that Mardonios was then relieved of his command as generalissimo (6.94.2). Thus the subsequent combination of (A) success for the Persian forces at Eretria and (B) failure for them at Marathon in 490 BCE must be attributed to military leaders other than Mardonios. We will have to wait 13 years for Mardonios to re-emerge as a dramatis persona, when he becomes once again the generalissimo of the Persian Empire in the expedition of 480–479 against the Hellenes of Europe, culminating in the defeats at Salamis and Plataea.

Debate of the Constitutions

§55. The so called Debate of the Constitutions, dramatized by Herodotus (3.80–84) as happening in 522 BCE, is often dismissed as pure invention: many have gone on record to claim that Persian elites at this point in history could not have been thinking of three forms of government as conventionally described in ancient Greek traditions: monarchy, oligarchy, democracy.[22] In my own work on the Debate passage, by contrast, I have argued against such claims of invention, emphasizing (1) the accuracy of Herodotus in his use of wording to describe the three forms of government and (2) the applicability of this wording to forms of government as they actually existed in the Greek-speaking world at this time.[23] My argument is, these three forms of government—or at least the idea of these forms—already existed in the Greek-speaking world that belonged to the Persian Empire at this time, 522 BCE. My argument goes further: the Persian elites as pictured by Herodotus in the Debate of the Constitutions would have been debating in real life not the ideal form of government for the Persians as Persians but for the Asiatic Greeks who inhabited the westernmost part of their empire, especially for the Ionians, who were considered to be the dominant {|117} culture of the Asiatic Greek-speaking world by contrast with the Aeolians to the north and the Dorians to the south.

§56. And here I return to what Herodotus says (6.43) about the actions of Mardonios in 493 BCE: as we have already seen, Mardonios established democracies in the cities of Ionia, deposing the Ionian tyrants who had been installed by the Persian Empire to rule these cities. I argue that the Debate of the Constitutions, as narrated in the History of Herodotus (3.80–84), makes sense only retrospectively. In other words, the things that were supposedly being said in 522 BCE, which was the dramatic date of the Debate, could only be understood in terms of the things that were being done by Mardonios in 493 BCE. In the Debate dramatized by Herodotus, as we have already seen, the word isonomíē is used to express the idea of democracy, advocated by Otanes, one of the debating nobles, as the best form of government for the Persian Empire (3.80, 3.83). Later, when Herodotus cross-refers to this passage in the context of narrating the establishment of democracies by Mardonios, he uses the word dēmokratíē (6.43), which as we have already seen is clearly meant there as a synonym of isonomíē. And, as we have also seen already, Herodotus uses the same word dēmokratíē (6.131.1) in referring to the establishment of democracy in Athens by Cleisthenes in 508/7 BCE. Conversely, Herodotus uses the word isonomíē (5.37.2) in referring to the democracy proclaimed in Miletus by Aristagoras in 499 BCE. And there were even earlier prototypes of such democracy, at least conceptually. Herodotus (3.142.3) mentions a striking example: he says that isonomíē was proclaimed in Samos by Maiandrios, the new tyrant of that island state, successor to Polycrates, who died not long after 522 BCE. In this case, however, Herodotus goes on to narrate how the proclamation failed (3.142.4–3.143.2). Failure or no failure, however, democracy was an option to be reckoned with in the Greek-speaking regions of the Persian Empire.

§57. But now I move fast-forward in time from 493 to 480 BCE. The cities of the Ionians who were fighting on the side of the Persian Empire had presumably still retained democratic forms of government as established by Mardonios thirteen years earlier.[24]

§58. In any case, the question remains: what would have happened if Mardonios had been victorious in the grand expedition that culminated in the naval battle at Salamis? Well, if we follow the reportage of Herodotus (7.6.1), Mardonios would have become the satrap of the entire Greek-speaking world of Europe.[25] But there is more {|118} to it. I think that the Ionians who were his collaborators would have played a major role in the political reorganization of the European Hellenes. Besides Asiatic Ionia, there could now be a European Ionia as well, with the annexation of Attica together with such outlying islands as Euboea, and the capital city of such a reconfigured Greater Ionia could have remained Athens, which was after all the original mother city or metropolis of the mythical Ionian Migration. And this venerable metropolis could even have remained a democracy of sorts—at least in name.[26]

One last time, back to the mágoi

§59. It can be debated whether the Ionians would really have become new continuators of democracy if Mardonios as generalissimo of Xerxes had defeated the Athenians. Speaking for myself, I am not certain. But I would be more certain of something else: if the Persian Empire had won at Salamis, the Ionians would have become far more important politically as well as culturally. And, leaving aside such hypothetical questions, I am even more certain about a simple fact that has emerged all too clearly in the course of this presentation: the Ionians were and always had been very important for the Persian Empire. Here we come back one last time to the mágoi.

§60. Because these mágoi, like the Ionians, were very important for the Empire, it stands to reason that the Ionians understood well the mágoi just as the mágoi understood the Ionians. I close this presentation with two examples that illustrate the point I just made.

§61. The first example centers on a most famous Ionian intellectual, Heraclitus, the dating of whose lifespan stretches from the middle of the sixth century BCE down to a few years beyond the year of the naval battle at Salamis, 480 BCE. This Ionian was a native of one of the greatest cities of the Ionian Dodecapolis, Ephesus, which had stayed under the domination of the Persian Empire during practically the entire extent {|119} of his lifetime, except perhaps for a few years after the events of 480 BCE. In Heraclitus Fragment 14 (= B 14 DK), as mediated by Clement of Alexandria (Protrepticus 2.22.2–3), Heraclitus is said to be ‘speaking as a seer’ (manteúein) to various kinds of occultists, described by Clement as nuktipóloi ‘those who go roaming at night’ / mágoi ‘mages’ / bákkhoi ‘devotees of Dionysus’ / lênai ‘she-devotees of Dionysus’ / mústai ‘initiands’, and he warns them about various possible punishments, including fire, in a negative kind of afterlife—supposedly because they conducted rituals of initiation ‘in a way that is not sacred [an-hierōstí]’ (Τίσι δὴ μαντεύεται Ἡράκλειτος ὁ Ἐφέσιος; “Νυκτιπόλοις, μάγοις, βάκχοις, λήναις, μύσταις”, τούτοις ἀπειλεῖ τὰ μετὰ θάνατον, τούτοις μαντεύεται τὸ πῦρ· “τὰ γὰρ νομιζόμενα κατὰ ἀνθρώπους μυστήρια ἀνιερωστὶ μυοῦνται”). The wording of Clement strikes me as a composite of many different contexts where Heraclitus was referring to many different kinds of initiations and initiands.[27] Most relevant to this fragment of Heraclitus is Column VI of the Derveni Papyrus: here, as we have already seen, the author mentions mágoi (lines 2, 5, 9) as models for mústai ‘initiands’ (line 9). Also relevant, with respect to two other fragments of Heraclitus (22 B 3 and B 94 DK), is Column IV of the Derveni Papyrus: there, as Franco Ferrari argues, the author of the Papyrus refers to Heraclitus (Ἡρ̣άκλ̣ε̣ιτος, line 5) in the context of this thinker’s thinking about ritual practices, and among those practices are rituals of sacrifice as practiced by the Persians, as we can see from the restored wording [Π]έ̣ρ̣σ̣α̣ι̣ θύο̣υ̣[σι] ‘the Persians sacrifice [thuein]’ (line 11).[28] The Persians who are sacrificing in this context, I think, are the mágoi.[29]

§62. My second example centers on the Ionians who advise the mágoi to sacrifice and sing incantations in order to appease the wind of the Hellespont and the mother of Achilles, as we read in the narrative that I cited earlier on from Herodotus (7.191.2). To me, that narrative says it all—in its very own microcosm.



Asheri, D. 1983. Fra Ellenismo e Iranismo, Bologna.

Benveniste, E. 1929. The Persian Religion according to the chief Greek Texts. Paris.

Benveniste, E. 1938. Les Mages dans l’ancien Iran. Paris.

Bernabé, A. 2014. “On the Rites Described and Commented Upon in the Derveni Papyrus, Columns I–VI.” Poetry as Initiation: The Center for Hellenic Studies Symposium on the Derveni Papyrus (ed. I. Papadopoulou and L. Muellner) 19–52. Hellenic Studies 63. Cambridge, MA, and Washington, DC.

Betegh. G. 2004. The Derveni Papyrus: Cosmology, Theology and Interpretation. Cambridge.

Briant, P. 1996. Histoire de l’Empire Perse. De Cyrus à Alexandre. Paris.

Briant, P. 2002. From Cyrus to Alexander: A History of the Persian Empire (translated by P. T. Daniels). Winona Lake, IN.

Davidson, O. M. 2013. Poet and Hero in the Persian Book of Kings. 3rd ed. Cambridge, MA.

Ferrari, F. 2011. “Rites without Ordeals: Magi and Mystae in the Derveni Papyrus.” Zeitschrift für Papyrologie und Epigraphik 179:71–83.

Ferrari, F. 2014. “Democritus, Heraclitus, and the Dead Souls: Reconstructing Columns I–VI of the Derveni Papyrus.” Poetry as Initiation: The Center for Hellenic Studies Symposium on the Derveni Papyrus (ed. I. Papadopoulou and L. Muellner) 53–66. Hellenic Studies 63. Cambridge, MA, and Washington, DC.

Frame, D. 2009. Hippota Nestor. Hellenic Studies 34. Cambridge, MA, and Washington, DC.

Haubold, J. 2007. “Xerxes’s Homer.” Cultural Responses to the Persian Wars: Antiquity to the Third Millennium (ed. E. Bridges, E. Hall, and P. J. Rhodes) 47–63. Oxford.

Horky, P. S. 2009. “Persian Cosmos and Greek Philosophy: Plato’s Associates and the Zoroastrian magoi.” Oxford Studies in Ancient Philosophy 37:47–103.

Nagy, G. 1990. Pindar’s Homer: The Lyric Possession of an Epic Past. Baltimore.

Nagy, G. 2009|2008. Homer the Classic. Printed | Online version. Hellenic Studies 36. Cambridge, MA, and Washington, DC.

Nagy, G. 2010|2009. Homer the Preclassic. Printed | Online version. Berkeley and Los Angeles.

Nagy, G. 2011. “The Aeolic Component of Homeric Diction.” Proceedings of the 22nd Annual UCLA Indo-European Conference (ed. S. W. Jamison, H. C. Melchert, and B. Vine) 133–179. Bremen. In Nagy 2012 v1.

Nagy, G. 2013. The Ancient Greek Hero in 24 Hours. Cambridge, MA.

Nagy, G. 2014. “Herodotus and the Logioi of the Persians.” No Tapping around Philology: A Festschrift in Honor of Wheeler McIntosh Thackston Jr.’s 70th Birthday (ed. A. Korangy and D. J. Sheffield) 185–191. Wiesbaden.

Nagy, G. 2016. “The Idea of an Archetype in Texts Stemming from the Empire Founded by Cyrus the Great.” The Archaeology of Greece and Rome: Studies in Honour of Anthony Snodgrass (ed. J. Bintliff and K. Rutter) ch. 14 = 337–357.

Nagy, G. 2017.04.19. “Thinking Iranian, Rethinking Greek.” Classical Inquiries.

Panaino, A. 2001. “Greci e Iranici: confronto e conflitti.” I Greci. Storia, Cultura, Arte, Società. Vol. 3. I Greci oltre la Grecia vol 3 (ed. S. Settis) 79–136. To­rino.

Panaino, A. 2009. “Aspetti della complessità degli influssi interculturali tra Grecia ed Iran.” Grecia Maggiore: Intrecci culturali con l’Asia nel periodo arcaico / Graecia Maior: Kulturaustausch mit Asien in der archaischen Periode. Atti del simposio in occasione del 75o anniversario di Walter Burkert / Akten des Symposions aus Anlass des 75. Geburtstages von Walter Burkert (ed. Ch. Riedweg) 19–53. (Bibliotheca Helvetica Romana 30). Basel.

Panaino, A. 2011. “Erodoto, i Magi e la Storia Religiosa Iranica.” Herodot und das persische Weltreich (ed. R. Rollinger, B. Truschnegg, Reinhold Bichler) 343–370. Wiesbaden.

Panaino, A. 2015.  “Jesus’ trimorphisms and tetramorphisms in the meeting with the Magi.” From Aṣl to Zā’id: Essays in Honour of Éva M. Jeremiás (ed. I. Szántó) 167–209. Piliscsaba.

Papadopoulou, I., and Muellner, L., eds. 2014. Poetry as Initiation: The Center for Hellenic Studies Symposium on the Derveni Papyrus. Hellenic Studies 63. Cambridge, MA, and Washington, DC.

Shayegan, M. R. 2012. Aspects of History and Epic in Ancient Iran: From Gaumāta to Wahnām. Hellenic Studies 52. Cambridge, MA, and Washington, DC.

Shayegan, M. R. 2017. “Persianism: Or Achaemenid Reminiscences in the Iranian and Iranicate World(s) of Antiquity.” Persianism in Antiquity (ed. R. Strootman and M. J. Versluys) 401–455. Stuttgart.

Sissa, G. 2012. “Democracy: A Persian Invention?” Mètis: Anthropologie des mondes grecs anciens 10:227–261.

Skjærvø, P. O. 2015. “The Gāthās as Myth and Ritual.” The Wiley Blackwell Companion to Zoroastrianism (ed. M. Strausberg and Y. S.-D. Vevaina, with the assistance of A. Tessmann) 59–67. Chichester, West Sussex.

Tsantsanoglou, K. 2014. “Some Desiderata in the Study of the Derveni Papyrus.” Poetry as Initiation: The Center for Hellenic Studies Symposium on the Derveni Papyrus (ed. I. Papadopoulou and L. Muellner) 1–18. Hellenic Studies 63. Cambridge, MA, and Washington, DC.



[1] For more on Smerdis / Gaumāta, I strongly recommend the work of Shayegan 2012; I follow up on his interpretations in Nagy 2016.

[2] On such Greek/Persian interactions, I refer in general to the pioneering work of Benveniste 1929; also Benveniste 1938. A most incisive set of observations on Greek /Persian intereactions can be found in the work of Panaino 2009, especially at p. 25 with reference to the word mageíā in the Platonic Alcibiades I (22a), which is defined there as theōn therapeíā ‘caring for the gods’ (ἔστι δὲ τοῦτο θεῶν θεραπεία). (In Nagy 2013 6§54, I survey Greek examples of therapeúein in the specialized sense of ‘ritually care for, take care of’ with reference to whatever is considered to be sacred.)

[3] In general, what I will argue is inspired by the remarks of Tsantsanoglou 2014:16, who emphasizes the importance of an Ionian reception for the mágoi.

[4] A particularly significant attestation of Iranian magu- is Pahlavi mowbad, reconstructed as *magu-pat- and meaning something like ‘master of the magu-s’: see Panaino 2009:24. On the authority of the generic mōbad in classical Persian poetics, as represented in the Shāhnāma of Ferdowsi, I follow the interpretation of Davidson 2013:43.

[5] For background on the concept of mágos in Iranian contexts, along with ample bibliography, I recommend the admirable analysis of Panaino 2011, where he also considers the rôle of Zoroaster / Zarathuštra (especially pp. 344–347). Likewise relevant is what Panaino 2009 has to say (especially at p. 25) with reference to the use of the word mageíā in the Platonic Alcibiades I (22a), since the author there connects this Greek word with ‘Zōroastēr son of Ōromazos’ (μαγείαν τε διδάσκει τὴν Ζωροάστρου τοῦ Ὡρομάζου – ἔστι δὲ τοῦτο θεῶν θεραπεία). As for the presentation I offer here, what I will be arguing with reference to the mágoi is relevant to what I have to say about Zoroaster / Zarathuštra in Nagy 2017.04.19 §§15–17.

[6] Panaino 2015, with further analysis. More in Panaino 2011:361–363.

[7] On the Derveni Papyrus in general, I always learn something new whenever I read Betegh 2004. Especially useful for me in the present context is his interpretation of Diogenes Laertius 1.7, who refers to the practices of mágoi in interpreting eidōla ‘images’—in the sense of insubstantial images of the dead.

[8] On this point, see Ferrari 2014:57.

[9] Most informative here is the analysis of Panaino 2011:347–361.

[10] This word eukhḗ ‘prayer’ is relevant to what I have to say about Zoroaster / Zarathuštra in Nagy 2017.04.19 §§15–17. A work on Zoroaster that I find singularly useful is the article of Skjærvø 2015.

[11] I interpret theogoníē ‘theogony’ as a kind of eukhḗ ‘prayer’: here I invoke the context of the word eukhḗ as used in the Derveni Papyrus. See §21 above.

[12] An authority on the subject of the multiethnic nature of the Persian Empire is Briant 1996/2002.

[13] In Nagy 2011, I survey the myths about an Ionian Migration and compare them with the rival myths about the Aeolian Migration.

[14] On this point, I offer further argumentation in Nagy 2014. Also, as Rahim Shayegan points out to me, the so-called “Daiva Inscription” of Xerxes, found at Persepolis (XPh 23–25), already refers to the Asiatic Greeks and the European Greeks together as ‘the Ionians who live by the sea and who live across the sea’: yaunā taya drayahiyā dā|rayatiy utā tayaiy paradraya dārayat|iy.

[15] Nagy 2009|2008 4§§174–175, §§177–180

[16] Nagy 2009|2008 4§24.

[17] Addendum, taken from the printed version of 2018, p. 112. I do not have enough space here to examine more fully the events reported by Herodotus (5.30–38) concerning the failed military expedition of Aristagoras to capture the Ionian island of Naxos. I confine myself here to noting two arguments I plan to develop in a separate project. One, with reference to the two hundred triremes that were authorized by Darius himself for Aristagoras to use in his expedition (5.30–33), I will argue that this contribution undertaken by the Persian Empire was more directly a contribution fulfilled by the cities of the Asiatic Greek world, which were ruled at the time by tyrants appointed by the empire—just as Miletus was ruled by such an appointee in the person of Aristagoras himself. Two, I will argue that the narrative of Herodotus about the double-crossing of these Greek tyrants by Aristagoras (5.36–38) leaves room for interpretations that differ from the one that the narrator himself prefers, which is, that Aristagoras feared retribution from Darius for the failure of the expedition. In terms of my own interpretation, Aristagoras seized an opportunity to capitalize on this failure by attempting to turn it into a successful revolt against his Persian sponsors. In any case, as Herodotus notes (5.35–36), Aristagoras may well have been secretly aided and abetted by his predecessor Histiaios, who as I have noted had been detained as a guest of the king in the capital of the Persian Empire.

[18] Nagy 2009|2008 4 §25.

[19] Addendum, taken from the printed version of 2018, p. 112. Artaphernes was involved in the events reported by Herodotus (5.30–38) concerning the failed military expedition of Aristagoras to capture the Ionian island of Naxos. In my earlier footnote about this expedition, I already noted that Herodotus speaks of two hundred triremes that were authorized by Darius himself for Aristagoras to use in his expedition (5.30–33); in that same context, I must now add, Herodotus also speaks of the role of Artaphernes himself in persuading Darius to make such an authorization.

[20] I do not have enough space here to delve into the complexities of this political crisis in Athens.

[21] Nagy 2014. I would now add this observation: the use of the word Héllēnes by Hecataeus (FGH 1 F 1) in referring to Greeks as makers of multiple and laughably unreliable lógoi ‘words’ was meant to be a negative reference to the mythmaking of European Greeks in his era, to be contrasted with the supposedly reliable mythmaking of Asiatic Greeks.

[22] Sissa 2012 resists such negative claims (with bibliography at p. 230n7). For further analysis (with further bibliography), see Panaino 2001.

[23] Nagy 1990:181–192, 265–266.

[24] One exception at this time, and there may have been more, was the island-state of Chios: within the time-frame of the year 479 BCE, Herodotus (8.132.2) mentions in passing that a tyrant by the name of Strattis was in charge there at that point in time. Elsewhere, Herodotus (4.137.2) mentions that Strattis was in charge of Chios already during the events narrated for 513 BCE. In this case, then, the tyrant may have been deposed in 499 BCE and then may have made a comeback, as it were, even after the democratic deals made by Mardonios with the cities of Ionia in 493 BCE.

[25] As Rahim Shayegan points out to me, there existed an Old Persian word for such a ‘super-satrap’, to be reconstructed as *kārana-, which is attested in Greek as karanos (κάρανος). For example, Cyrus the Younger was such a karanos (κάρανος), as we read in Xenophon Hellenica 1.4.3–4. He was appointed the ruler of three satrapies in Asia Minor: Lydia, Greater Phrygia, and Cappadocia. See Shayegan 2017.

[26] I can see a possible objection to my raising this possibility. In Herodotus 7.6.1–5, where Mardonios is portrayed as agitating at the court of Xerxes for war against the European Greeks, and where his motive is said to be his ambition to rule over all the Greeks (7.6.1), it is also said that the Athenian family of the Peisistratidai and the Thessalian family of the Aleuadai were also present at the court of Xerxes—and were also agitating for war. Text and comments in Nagy 2010|2009:348–349 (see also Haubold 2007:52–54). Still, we cannot assume that Mardonios and the other agitators were all on the same political side. Even if the Peisistratidai, as former ‘tyrants’ of Athens, would not have been advocating a democracy for Athens, the same kind of advocacy cannot be assumed for Mardonios, who must have had his own political agenda. As for what happened 13 years earlier at Marathon, when the Persian Empire was ready to reinstate the Peisistratidai as tyrants of Athens by restoring Hippias to power, I need to recall the fact that Mardonios had already been relieved of his command of the forces on the Persian side, as we see from the reportage of Herodotus (6.94.2). If Mardonios had still been in command in 490 BCE, I think that Hippias would not have been given another chance to attempt a political comeback in Athens.

[27] I think that Plato in Theaetetus 179e is making references to comparable contexts of initiation as interpreted by latter-day followers of Heraclitus in Ephesus (Nagy 2009|2008 2§137 and 3§96).

[28] Ferrari 2014:63–65. I am grateful to Ioanna Papadopoulou for reminding me of this restoration.

[29] Such an argument is actually made by Ferrari 2011, who also cites further sources, with bibliography (especially valuable are his observations at p. 71, with reference to the important work of Horky 2009 and others).

Annotations loading . . .

We're trying out a new look. 🎉 Let us know what you think! Hide.