BILINGUALISM AND GREEK IDENTITY IN THE FIFTH CENTURY B . C . E .*

The study of bi-and multilingualism in the ancient Mediterranean has come into its own in recent decades. The evidence is far greater for the Hellenistic and Roman periods than the Classical, so naturally scholarly attention has focussed less on the earlier era. This has led to some enduring notions about bilingualism in the fifth century B . C . E . which are yet to be fully scrutinized, including the idea that a Greek ’ s speaking another tongue was inherently transgressive. What did it mean for a Greek to speak a second language? This article re-evaluates the evidence for individual bilingualism in Herodotus and Thucydides in their fifth-century context, focussed on our two best-documented examples of bilingual Greek individuals (Histiaeus of Miletus and Themistocles of Athens). Close reading of Herodotus and Thucydides suggests that not only does the notion of an inherently transgressive bilingualism hold little water for this period, but bilingualism may even be a sign of μ ῆ τις .

INTRODUCTION 1 The study of bi-and multilingualism in the ancient Mediterranean has come into its own in recent decades, in line with increasing emphasis on ethnic and cultural identity more broadly.Adams's work on bilingualism in the Roman world has been a particular source The polarity still has relevance, however, and the focus has moved to defining the exact limits of these shifting boundaries between Greek and non-Greek. 7In the case of linguistic identity, although less attention has been paid to the Classical period, scholarship has approached the literary evidence for bilingualism, especially Herodotus and Athenian drama. 8In some quarters, however, there persists a notion of bilingualism as inherently transgressive: that speaking another tongue was either a fundamental betrayal of one's Greek identity, or a cultural transgression that invites, encourages, or even leads inevitably to personal or political betrayal. 9Such views may well derive their logic from the many passages in ancient Greek literature where non-Greek language is 'almost invariably an index of primitivism, uncouthness, intellectual or cultural inferiority, irrationality, or madness'. 10It does not necessarily follow, however, that such attitudes automatically apply to Greeks speaking non-Greek languages; we might think of certain modern environments in which respect for bilingual members of one's own community coexists with disdain for native speakers of the secondary language. 11One reason for these underexamined assumptions may be the simple lack of extant examples for analysis, since the vast majority of intercultural interactions in ancient Greek literature are written without acknowledgement of the multilingual circumstances. 12Only rarely are we made aware of the presence of interpreters, or the multilingual abilities of an interlocutor.It is possible that individual bilingualism was more common among Greeks-even elite Greeks-than the historical record implies; 13 however, few are M.C. Miller, Athens and Persia in the Fifth Century BC: A Study in Cultural Receptivity (Cambridge, 1997); M.L. West, The East Face of Helicon: West Asiatic Elements in Greek Poetry and Myth (Oxford, 1997). 7See the survey in T. Harrison, 'Reinventing the barbarian', CPh 115 (2020), 139-63. 8e.g.Hall (n.1); J. Campos Daroca, Experiencias del lenguaje en las "Historias" de Heródoto (Almería, 1992); Harrison (n.1); S. Colvin, Dialect in Aristophanes: The Politics of Language in Ancient Greek Literature (Oxford, 1999); Munson (n.1); L. Miletti, Linguaggio e metalinguaggio in Erodoto (Pisa and Rome, 2008); Brandwood (n.1); Figueira (n.4); E. Nolan, 'Athenians, Amazons, and solecisms: language contact in Herodotus', AJPh 142 (2021), 571-96. 9e.g.Gera (n.1), 454: 'it seems as if bilingual Greeks are treacherous Greeks virtually by definition: language differences must remain in place'; Clackson (n.1), 31: 'until the peculiar bilingual phenomenon of the Roman Empire … bilingualism of any sort was treated as evidence of divided loyalties'; Brandwood (n.1), 15: 'suspicion of bilingualism is a familiar enough trope that was echoed in antiquity after the Persian Wars and even later'.(On Clackson and Brandwood, see further n.26 below).Beyond Hartog's comments on the Scythians (below, pp.7-8 and n. 40), the only clear example I have found from non-Anglophone scholarship is G. Mosconi, 'La lingua creola del demos.Sul Vecchio Oligarca (Ath.resp.2,8) e sui significati sociopolitici del plurilinguismo nel pensiero greco', Historika 11 (2021), 43-82, perhaps clearest at 56: 'L'individuo che parla più lingue ha (è considerato avere) un'identità incerta e fluttuante; per i Greci, chi parla più lingue manifesta una ridotta solidarietà verso i membri della propria comunità, o addirittura mostra una vera e propria propensione al tradimento'.Cf. also Dubuisson (n.1), 211, who notes briefly-in discussing several anecdotes from Diodorus and Plutarch about the negative aspects of bilingualism-that 'une telle hostilité risque évidemment de rejaillir sur tous les Grecs capables de s'exprimer en langue étrangére, et de jeter le discrédit sur les bilingues et le bilinguisme'; the rest of the article, however, is more sceptical than some Anglophone scholarship of reading treachery into the term δίγλωσσος, for example (cf.Brandwood [n.1], 32 n. 4).
10 Munson (n.1), 2; cf.Lejeune (n.explicitly described in our sources as learning or using non-Greek tongues in the Classical period. 14his article will focus on the only two examples of historical Greek individuals from the fifth century B.C.E.who are explicitly described as bilingual and whose linguistic ability is first attested in extant fifth-century sources: Histiaeus of Miletus and Themistocles of Athens.These are the key figures on whom any argument about Greek bilingualism in the period must be based. 15Both men are said to speak 'Persian', the language of the Achaemenids, who were still major players in late fifth-century Greek politics; thus, we are concerned with a more politically charged bilingualism than, say, Greek-Lydian or Greek-Egyptian, and so one more likely to favour tropes of dual loyalty or treason. 16This leads us to the fact that these two figures are often paired as exemplifying Odysseus-like μῆτις ('intelligent cunning') and I will consider these implications in the conclusion. 17My discussion of sources will focus primarily on material from the fifth century B.C.E.For Histiaeus, there is little beyond Herodotus anyway, but a long tradition dealing with Themistocles' bilingualism stretches well into the Roman period and beyond. 18This complex reception history has contributed to distorting the analysis of Themistocles' bilingualism by removing it from its fifth-century context, with scholarship often retrojecting Roman-era attitudes onto Thucydides. 19Histiaeus and Themistocles are both culturally and politically complex, liminal figures, who fortunately receive relatively thorough characterization from Herodotus and Thucydides.Through close reading of these texts, I argue that the 14 Especially compared with widespread bilingualism among Greeks of the Roman era; see e.g.B. Rochette, Le latin dans le monde grec (Brussels, 1997). 15See above (n.5) on defining 'bilingual'.We might ask whether other prominent Greeks from Classical sources were bilingual-especially those who spent periods in the Persian court-but we cannot assume this when not explicitly stated, particularly given the prevalence of non-Greek, Greek-speaking interpreters through whom they may have communicated.To take an example from the fourth-century Anabasis of Xenophon, we cannot assume that Phalinus or Heracleidesboth of whom are employed by non-Greek rulers (2.1.7-23;7.6.41)-are'bilingual' ( pace Gera [n.1], 454); it is just as likely they spoke to their bosses through the same interpreters as Xenophon's Greeks (interpreters for Tissaphernes: 2.3.17-18; for Seuthes: 7.3.25,7.6.8-9,7.6.43).Ctesias may have learned Persian, but the evidence is not clear (Miller [n.6], 132; E. Almagor, 'Ctesias and the importance of his writings revisited', Electrum 19 [2012], 9-40, at 21-2).Alcibiades is also alleged to have 'learned Persian just as Themistocles had done' (τὴν Περσικὴν ἔμαθε φωνήν, καθάπερ καὶ Θεμιστοκλῆς, Athen.Deipn.12.535e).While this is not implausible, Miller (n.6), 131 is right that it is most likely to be 'a later fiction resulting from a perceived parallel with the life of Themistokles'.Pythagoras' supposed knowledge of Egyptian (Diog.Laert.8.1.3,citing Antiphon) is a more interesting case, but the evidence is fragmentary and harder to analyse in context.Other passages hint at wider bilingualism or at least use of interpreters or translators (Thuc.4.50; the Old Oligarch's comments at [Xen.] Ath.Pol.2.8 reflect something different from individual bilingualism, pace Mosconi [n.9], 56 and 62).In the fourth century, Alexander's retinue included Peucestas and Laomedon who knew non-Greek languages (Arr.Anab.3.6.6;6.30.3; 7.6.3). 16The actual language may have been Old Persian, Aramaic, or even both: C. Nylander, 'Assyria grammata: remarks on the 21 st "Letter of Themistocles"', Opuscula Atheniensia 8 (1968), 119-36; Gera (n.1), 448 n. 13. 17 O.Murray, 'The Ionian Revolt', CAH 2 , 4.461-90, at 486; D. Lateiner, 'Deceptions and delusions in Herodotus', ClAnt 9 (1990), 230-46, at 231. 18 Gera (n.1); K. Mayer, 'Themistocles, Plutarch, and the voice of the Other', in C. Schrader, V. Ramón and J. Vela (edd.),Plutarco y la Historia (Zaragoza, 1997), 297-304. 19See n. 9 above.
evidence does not permit us to assert that individual bilingualism signified cultural betrayal for Greeks of the fifth century B.C.E.; indeed, it may even have been a marker of μῆτις.

BILINGUAL INDIVIDUALS IN FIFTH-CENTURY CONTEXT
Let us first contextualize our study of Histiaeus and Themistocles with a brief overview of the wider representation of bilingual individuals in Herodotus and Thucydides in their fifth-century context.We will cover bilingual figures of lower status who acted as intermediaries (interpreters, messengers), as well as other elite bilinguals-such distinctions of class and status are often overlooked in assessing bilingualism.I will conclude that the two major fifth-century historians evince no inherent distrust in bilingual individuals from across the class spectrum; this will set the scene for a more detailed analysis of our two elite Greek bilingual men.

Bilingual intermediaries (interpreters/messengers)
Many assumptions about bilingualism in the fifth century B.C.E.can arguably be blamed on a famous anecdote from a work written five centuries later-Plutarch's Themistocles. 20The story concerns Themistocles' arrest of a bilingual messenger/ interpreter (described as both δίγλωσσος 'bilingual' and ἑρμηνεύς 'interpreter'), eventually having him executed 'because he dared to employ the Greek language for barbarian commands' (ὅτι φωνὴν Ἑλληνίδα βαρβάροις προστάγμασιν ἐτόλμησε χρῆσαι, Plut.Them.6.2). 21Herodotus mentions no such incident; indeed, he notes specifically that heralds were not sent to Athens and Sparta in 481 because of the treatment they had received a decade earlier when the Athenians had thrown the herald into the βάραθρον. 22Plutarch's story is likely a tall tale based on a confusion of several Persian embassies, aligning neatly with later Greek perspectives on the Persian Wars (including Plutarch's De malignitate Herodoti) in depicting a more united Greek front against the enemy than is evident from fifth-century accounts, and even presenting Themistocles as 'saviour of Greece'. 23Moreover, Plutarch's comments on the use of Greek versus non-Greek tongues parallel others in his oeuvre and tally with a certain Second Sophistic emphasis on Greek linguistic purity. 24The anecdote should thus not be taken uncritically as evidence of fifth-century attitudes towards Greek or bilingual expression. 25ne underappreciated reason for this is the need to distinguish between attitudes towards bilingual intermediaries (lower status figures who work for others) and elite individuals who adopt another tongue in their own interest (like Histiaeus and Themistocles); neglecting such differences of class and status can lead to misinterpretation. 26ilingual intermediaries certainly could be a source of suspicion in antiquity, perhaps inevitably so given their role as political agents in whom much diplomatic trust is placed-but that is why we must not lose sight of other reasons for negative characterisation beyond bilingual ability.Most evidence for suspicion of bilingual intermediaries comes from the Roman era, 27 and few would argue that elite bilingualism was not widely accepted in that period.In the fifth-century Greek world, however, there is little unequivocal evidence for such a perspective on bilingual intermediaries. 28his brings us to a key point: distrust in bilingual intermediaries is arguably absent from Herodotus and Thucydides.Herodotus includes several references to these figures, including his own engagement with a ἑρμηνεύς at the pyramids (2.125.6)and other references to the Egyptian interpreter class (2.154.2, 164.1); interpreters in Scythia (4.24); and those interpreters and bilingual envoys in the service of the Persian King. 29Although these intermediaries sometimes do not understand the specific cultural content (as with Croesus' words to Cyrus), there is no suggestion that they cannot understand the language or that they are untrustworthy. 30Even Cambyses' 'Fish-Eaters'-sent to the Ethiopian king as 'spies' (κατάσκοποι)-are portrayed as 'effectively faithful transmitters of linguistic and ethnographic information, despite Cambyses' malign intentions in sending them'. 31When Mys understands a Boeotian oracle spoken in Carian (8.133-5), Herodotus focusses on the phenomenon as a 'great marvel' (θῶμά μοι μέγιστον); there is no trace of Plutarch's interpretation that the god did not wish the Greek language to be in service to barbarians 26 Clackson's comments (n.9) on 'bilingualism of any sort' being 'evidence of divided loyalties' occur in a discussion of interpreters specifically.With Brandwood's remarks (n.9) on the 'suspicion of bilingualism', the relevant footnote in his chapter (p.33 n. 6) specifies 'suspicion of interpreters'.
27 Mairs (n. 1) deals with this topic admirably.See also C. Wiotte-Franz, Hermeneus und Interpres: zum Dolmetscherwesen in der Antike (Saarbrücken, 2001), 158; Gera (n.1), 452-4 argues for the untrustworthiness of interpreters using Roman-era examples. 28We might read 'suspicion' into the treatment of the Athenian ambassador in Aristophanes' Acharnians (91-114), as Brandwood ([n.1], 33 n. 6) suggests, but the focus of the skewering is the deception and vain promises of Athenian officials rather than interpreters more generally.Similarly, the link drawn by Mosconi (n.9), 65 n.85 between 'alloglossia e inganno' for Sophocles' use of ἀλλόθροος at Phil. 540 is tenuous at best; see De Luna (n.1), 117.A passage from Gorgias (Pal.7 = DK 82B 11A.7) is prima facie more compelling, when Palamedes describes the idea of an interpreter as a 'third person becoming a witness to things which need to be hidden' (τρίτος ἄρα μάρτυς γίνεται τῶν κρύπτεσθαι δεομένων, text at DK II, 296).Out of context, this passage might suggest that interpreters inherently possess the capacity to betray confidences; in context, however, Palamedes is noting that if he had met with Priam to betray the Greeks they would have required an interpreter to translate, and thus a legal witness to Palamedes' treason-but none exists, central to his defence.One further possible hint of this trope comes from Aeschylus' Agamemnon, when the chorus states that Cassandra needs a 'sharp/clear interpreter' (1062-3: ἑρμηνέως … τοροῦ).The qualifying adjective might imply the existence of less reliable (trustworthy?)interpreters, perhaps as fidus interpres implies the opposite (see Mairs)-but this is not self-evident, and complicated by Aeschylus' exploitation of the ambiguity of ἑρμηνεύς as interpreter of foreign language and oracles (Hall [n.1], 117-18). 291.86, 3.19.1,3.38.4,3.140.3. 30Pace Brandwood (n.1), 30-1; although the interpreter at the pyramid likely offered a creative mistranslation, Herodotus does not doubt the interpreter, only his own memory; cf. also 2.28 (reservations about the truth of the content, not about what he had heard-presumably through an interpreter). 31Brandwood, 32 n. 4.
(Mor.412A). 32None of these references to bilingual intermediaries in Herodotus evince distrust or betrayal; when included in historical narrative-rather than ethnographic observation-these figures primarily represent a gap in cultural understanding who disappear once that gap is bridged. 33As Munson notes, in the Histories language 'represents a particularly unproblematic area of difference' since interpreters could override language barriers. 34This is perhaps partly due to Herodotus' own interventional role in the text as 'arch-translator', dipping into the narrative intermittently from on high to provide translations of foreign words. 35hucydides mentions no ἑρμηνεῖς at all, and his sole reference to a bilingual messenger, the Carian δίγλωσσος Gaulites (8.85.1-2), carries no detectable distrust or suspicion.Indeed, Thucydides notes that Tissaphernes sends him on the embassy because he was his confidant (ξυνέπεμψε δὲ καὶ Τισσαφέρνης αὐτῷ πρεσβευτὴν τῶν παρ᾽ ἑαυτοῦ), signalling the gravity of the embassy for Tissaphernes.It might perhaps be argued that regular interpreters could not be trusted for such an important mission; 36 however, that is hardly evident from the text.The reference to Gaulites as δίγλωσσος lacks any notes of treachery and simply explains his presence and function on the embassy as a bilingual man. 37n Herodotus and Thucydides, then, bilingual intermediaries are not treated as objects of suspicion as often in Roman-era historiography.One reason for this neutral depiction could well be that, as far as we can tell, such figures were non-Greek individuals-the Carian Gaulites, Fish-Eaters from Elephantine, the Egyptian interpreter class-and thus might have been held to a different standard by these Greek historians.We can explore this further by examining the depiction of elite non-Greek figures, before assessing the Greek historians' treatment of our two elite Greek bilinguals.

Elite non-Greek bilingual figures
Several arguments about bilingualism in Herodotus, or even the fifth century more broadly, are based on non-Greek figures.Scholars often cite three episodes: Scyles the Scythian king (taught Greek by his mother, killed by Scythians: 4.76-80); the Median king Cyaxares and his Scythian suppliants (who teach Scythian to Median children; they kill the children because of a perceived slight from Cyaxares: 1.73); and the Pelasgians' 'Lemnian deeds' (the half-Pelasgian/half-Athenian boys on Lemnos are taught Attic dialect and customs by their Athenian mothers, then killed by the Pelasgians as a perceived threat: 6.138). 38Brandwood claims these episodes together represent 'bilingualism and its accompanying dangers'. 39One suspects the influence of Hartog's claim that 'to travel [referring to Anacharsis] and to be bilingual [Scyles] come down to the same thing; both are dangerous, for they lead to forgetting the frontier and thus to transgression'. 40Yet it is debatable to what extent these tales and their results are based on bilingualism specifically rather than broader issues of cultural transgression (Scyles) or revenge and the value of children (Cyaxares, Pelasgians); at best, language is one factor among several.
Scyles, as an elite Scythian, is an illustrative example.His knowledge of Greek has even been labelled 'the cause of his death'. 41But this is not true: Scyles is killed for involvement in foreign cult.Herodotus is explicit (4.79.1): 'But when matters had to turn out badly for him, they did so for the following reason (ἐγίνετο ἀπὸ προφάσιος τοιῆσδε)'; he then recounts Scyles' desire to be initiated into the Dionysian mysteries, where he is caught playing the Bacchant (4.79.5).Scyles may have become interested in Dionysian cult because he had learned Greek (4.78.1), but that linguistic knowledge alone is not why his brother murdered him. 42Indeed, Anacharsis comes to a similar fate without learning Greek; he is likewise killed for practising Greek rites (4.76). 43e will return to the significance of these interpretations later, but for now it is enough to state that bilingualism is not the cause of their deaths or even the specified reason for transgression.Herodotus' focus is on the Scythians' exceptional contempt for foreignespecially Greek-customs, rather than language. 44eyond these episodes, the Persian noble at the Theban banquet in Herodotus' Book 9 offers another example of an elite non-Greek performing bilingual acts without direct negative consequences.Just as Histiaeus is the only Greek in the Histories to speak Persian, the unnamed man seated beside Thersander at the banquet for Mardonius before Plataea (9.16) is the only Persian to speak Greek. 45He is said to address Thersander in Greek (Ἑλλάδα γλῶσσαν ἱέντα), offering a beautiful speech about the brevity of life, the Persians' imminent death in battle, and the inconsequentiality of human knowledge versus divine will.These words, rich in Homeric pathos, 'contain many Greek-indeed Herodotean-thoughts that make the two peoples here seem more similar than dissimilar'. 46Any bad end to which this Persian came (we never hear of him again) had little to do with his tragic expression in another tongue.Amidst the Theban medizing, it seems, a Persian's Hellenizing was more than appropriate.
To sum up.The distrust of lower-status bilingual intermediaries, so evident in Plutarch and other Roman-era authors, is not clearly attested in fifth-century literature and is absent from Herodotus and Thucydides.The bilingualism of elite non-Greek individuals (Scyles, the Persian noble) does not seem to be an inherent source of suspicion or cultural betrayal either; when negative consequences do occur, other aspects of cultural or political behaviour are in play.With this relevant background, let us now address our two primary case studies of elite Greek bilinguals.

HISTIAEUS IN HERODOTUS
Herodotus mentions only one Greek who is explicitly said to know another language-Histiaeus of Miletus. 47Unlike Themistocles, Histiaeus' linguistic abilities do not seem to have captured the imaginations of subsequent writers. 48This is despite his central role in Herodotus' detailed account of the Ionian Revolt of which Histiaeus was allegedly a co-instigator (along with Aristagoras of Miletus). 49Histiaeus has been considered an Ionian Greek patriot; 50 consistently pro-Persian; 51 and everything in between.But Herodotus' own verdict is less than clear. 52Some argue that Herodotus criticizes Histiaeus for acting in self-interest, 53 although such a perspective is complicated by fifth-century views of 'trickster' figures like Odysseus, whom historiography tended to view positively. 54Indeed, Murray saw Histiaeus (like Themistocles) as an 'Odyssean' figure, arguing that his portrait is based on a tradition centred around stories of Greek μῆτις. 55μῆτις is a multifaceted virtue, but certainly not exclusively negative in a fifth-century Greek context: as Detienne and Vernant argue, 'Christianizing ideals as well as the concept of Platonic truth have led to negative associations […] alien to fifth-century Greek thought'. 56Herodotus' views are complex but, like many contemporary Greeks, he generally seems to have appreciated 'artful deception and quick-thinking acts that promote self-preservation'. 57With this ambiguity in mind, let us examine Herodotus' depiction of Histiaeus' bilingualism.in order to escape.Defending himself against Darius' accusations of instigating the revolt, he denies that he would ever do anything, 'great or small', to harm the King (ἐμὲ βουλεῦσαι πρῆγμα ἐκ τοῦ σοί τι ἢ μέγα ἢ σμικρὸν ἔμελλε λυπηρὸν ἀνασχήσειν, 5.106).Hornblower suggests that Herodotus saw the expression 'great or small' as 'specially suited to Greek spoken in a Persian context'. 58This overlooks Herodotus' other use of the expression, however, which suggests a conversation in the Persian language: Prexaspes to Cambyses (3.62.3: ἢ μέγα ἢ σμικρόν), likewise defending himself against a kingly accusation of treachery. 59A similarly Persianizing expression of contrast-'by day and by night' (καὶ ἡμέρης καὶ νυκτός)-is used by Megabazus with respect to Histiaeus (5.23.2) and is found on the Bisitun (Behistun) inscription ( §7) and in Xerxes' letter in Thucydides (1.129.3). 60Such binary contrasts (great/small, day/night) occur often in Persian royal inscriptions. 61Hornblower is right to say that 'we are not meant to ask what language [Histiaeus and Darius] really spoke in, or if they used interpreters'-at least not yet, since we as readers have not been told that Histiaeus spoke Persian (if indeed he had learned it by this point).Such details are rarely cited in ancient historiography, which is why Herodotus' inclusion of Histiaeus' Persian ability is so worthy of examination.Rather, Hornblower's suggestion that Herodotus is 'attempting a Greco-Persian idiolect' is more attractive-we are meant to understand the 'Persian-ness' of Histiaeus' speech here. 62ut to what end?Hornblower argues for 'ethnic characterisation or stereotyping: the clever Greek adopts elaborate and stylized 'orientalising' courtesies'. 63That is, Herodotus depicts Histiaeus as employing these linguistic tactics to 'deceive' Darius, for the historian has already told us Histiaeus' true plans and motivations for escape and revolt (5.35).(Histiaeus is simply returning the favour, since Darius had used similar cunning in persuading him to come to court in the first place: 5.24). 64Hornblower's interpretation must be right, although we might expect more than a singular 'orientalising courtesy'.And, indeed, in the same speech, Histiaeus says (5.106.5-6): the expression seems to derive from a Greek understanding of a genuine Persian formulation. 66Burkert suggests that 'Histiaios, about to become a traitor, swears by the Gods of the Great King in order to break the oath at once'. 67Lateiner adds that 'had [Histiaeus] sworn by his Hellenic gods, those deities would have punished the forsworn oath'. 68So Histiaeus is here again portrayed as linguistically Persian-but cunningly Greek-in his successful attempt to 'deceive' Darius (5.107: διέβαλλε).We are meant to understand this behaviour as the clever means to his own ends.
When the Greeks were routed, Histiaeus, expecting that the king would not put him to death for this current misdemeanour, took up the following life-saving stratagem. 69He was overtaken in his flight by a Persian man, and when he was seized and about to be run through, he cried out in the Persian language and revealed that he was Histiaeus the Milesian.
We have seen Histiaeus using Persian linguistic affectations before, and if we again remember the complexity of Greek views of cunning, we might even see Herodotus' presentation of those earlier incidents as 'positive' (for want of a better word)-the clever Greek, Odysseus-like, deceiving the Persian king with bilingual dexterity.This capture scene, however, appears prima facie more complicated.Herodotus labels Histiaeus' behaviour with the term φιλοψυχίη ('love of life', 'clinging to life'), a hapax in the Histories.Hornblower and Pelling claim 'the word has a negative ring', citing Tyrtaeus 10.17-18 W 2 ('[exhorting] young soldiers to take up a great and brave spirit in their hearts, "and do not φιλοψυχεῖν when fighting against men"'). 70he word often appears in the sense of 'cowardice' in contrast to 'fighting', and certainly Histiaeus is now 'fleeing' (φεύγων). 71If pejorative here, 'it is the only place where Herodotus or the biography criticizes Histiaeus'. 72The term might recall Herodotus' assessment of Histiaeus' associate Aristagoras, 'a man of little courage' (ψυχὴν οὐκ ἄκρος, 5.124.1).
As Scott notes, however, these are 'rather different contexts' from Histiaeus' predicament. 73He and his band of Greeks have already fought the Persians at Malene 'for a long time' (χρόνον ἐπὶ πολλόν, 6.29.1), so Histiaeus can hardly be accused of choosing cowardice over fighting.There is at least some difference between negative images of φιλοψυχία prior to an engagement 74 and a ruse after being routed and on the point of being killed.Powell's proposed sense of 'way of saving his life'-followed by many translators (including, essentially, Hornblower and Pelling)-perhaps rightly minimizes, while not excluding entirely, negative aspects and focusses on the stratagem itself: an attempt at further cunning, a degraded instance of μῆτις. 75ven if taken as a purely cowardly act, however, there is little focus on Histiaeus' speaking of Persian here anyway; the emphasis is on his appeal to his own name and status with Darius, against whom he is ostensibly now fighting. 76The reference to Persian in this final scene is important because it reflects Histiaeus' cultural liminality, his ability to switch back and forth between these spheres when it suits him. 77But it also recalls his previous deception of Darius via his own brand of Greco-Persian bilingualism -something which, it now becomes even clearer, had been greatly successful in Herodotus' eyes.First, the Persians who capture Histiaeus soon kill him partly for fear that he will escape and 'immediately become great by Darius' side' (αὖτις μέγας παρὰ βασιλέι γένηται, 6.30.1), something that Themistocles will eventually accomplish (Thuc.1.138.2). 78 Second, Herodotus states his belief that Darius would have spared him (6.30.1), something Histiaeus himself believes (6.29.1) as he recalls the supposed Persian kingly custom of weighing up one's deeds (which Herodotus explicitly praises, 1.137.1). 79Third, Darius, upset, buries Histiaeus' remains with all ceremony as a 'great benefactor' to Persia (6.30.2).Histiaeus' bilingual ability had been part of his success with Darius, and saves his life here once more-even if that success then provokes the envy that leads to his murder. 80istiaeus' bilingualism, then, does not appear inherently transgressive of his Greekness-on the contrary, it plays a key part in his characterization as cunning Greek throughout the Histories.There are indeed questions to be asked about Histiaeus' cultural identity in Herodotus-his tattooing of a slave's head (5.35) certainly registers as 'barbarian' behaviour in the Herodotean imagination-but bilingualism does not appear to form part of this calculation for the Greek historian.If Herodotus passes final judgment, it is on Histiaeus' revolving allegiances rather than his linguistic abilities.

THEMISTOCLES IN THUCYDIDES
Let us now move on to our most famous example.Our earliest and only fifth-century source for Themistocles' bilingualism is Thucydides' account of his Persian exile (1.138.1).Themistocles' depiction must be considered alongside that of Pausanias, the two portraits forming a pair in a long 'flashback', informing each other and sharing structural similarities. 81Thucydides presents Themistocles' medism through comparison with Pausanias, and I argue that his description of Themistocles' adoption of Persian highlights the man's positive qualities.As with Histiaeus, we can see Themistocles' clever use of language adapted to suit his Persian royal audience even before we are told explicitly of his bilingual abilities.

'Persianisms' in Themistocles' letter to Artaxerxes
Themistocles in both Herodotus and Thucydides is the quintessential figure of μῆτις, sensitive to his audiences and adapting his language accordingly. 82Famously, he interpreted the 'wooden wall' oracle as referring to the Athenian navy (Hdt.7.140-4) and attempted to persuade the Ionians on the Persian side to defect via an inscribed Greek message (Hdt.8.22).It is hardly surprising, then, to find in Thucydides that Themistocles is motivated to study Persian before meeting Artaxerxes.First, however, from Ephesus he sends a letter to the King containing language that displays awareness of Persian royal customs (1.137.4).We will examine these 'Persianisms' below.

'Your house'
When Themistocles describes the Persian empire as 'your house' (τὸν ὑμέτερον οἶκον), his words reflect official Persian royal language on the conception of their imperial role.We can see this elsewhere in Thucydides (Xerxes' letter to Pausanias, 1.129.3),Herodotus, and a fragment of Philochorus. 83More significantly, we find it in texts composed by Persian kings: Darius' letter to Gadatas (GHI 12.15-17) includes the 81 There is a vast bibliography on the Pausanias-Themistocles excursus, including whether it is based on a written source from the Ionian tradition (H.D. Westlake, 'Thucydides on Pausanias and Themistocles-a written source?',CQ 27 [1977], 95-110; W. Blösel, 'Thucydides on Themistocles: a Herodotean narrator?', in E. Foster and D. Lateiner (edd.),Thucydides and Herodotus [Oxford, 2012], 215-40) or simply composed in a 'Herodotean' narrative style (S.Hornblower, A Commentary on Thucydides.Volume I: Books I-III [Oxford, 1991], 211-12; R.V. Munson, 'Persians in Thucydides', in E. Foster and D. Lateiner (edd.),Thucydides and Herodotus [Oxford, 2012], 241-77, at 251-6).In terms of Themistocles' letter, these questions are mostly irrelevant to my argument, since from Herodotus alone Thucydides could have recognised the (Greek interpretation of) Persian cultural knowledge contained therein and still found little to criticise in terms of Themistocles' Persianisms and bilingualism.Still, if Thucydides' account is based on an Ionian source, then it raises the possibility that both references to the bilingualism of Histiaeus and Themistocles come from the Graeco-Persian spaces of Ionia and Caria.
term 'King's house' (ἐμ βασιλέως οἴκωι), 84 while the Bisitun inscription contains several references: 'I strove until I had restored our house […] did not make our house destitute' ( §14); 'Him who strove for my house, him I treated well' ( §63).The Old Persian term vith-used in these passages means both 'house/domain' and 'family/clan', and was evidently translated into Greek as οἶκος. 85This term lacked the same semantic range as the Old Persian, so 'your house' used in this context likely came across as rather 'Persian' in Greek. 86Themistocles' letter thus appears appropriately 'culturally literate'.

'Benefaction'
Themistocles also highlights his own role as benefactor to the King, a significant one in Persian royal society.He states his actions to Xerxes' benefit, claiming that 'there is an obligation owed to me' (καί μοι εὐεργεσία ὀφείλεται). 87While benefaction concerning one's polis was common in the Greek world, the unique feature of Persian benefaction is that it was directed to the house of the King. 88Its importance is well documented, including in the Bisitun inscription: 'The man who was loyal, him I rewarded' ( §8). 89Darius' letter to Gadatas states that 'great favour will be laid up for you in the King's house' (σοι κείσεται μεγάλη χάρις ἐμ βασιλέως οἴκωι). 90erxes' letter to Pausanias states that 'an obligation is laid up for you in our house, recorded forever' (Thuc.1.129.3:κείσεταί σοι εὐεργεσία ἐν τῷ ἡμετέρῳ οἴκῳ ἐς αἰεὶ ἀνάγραπτος). 91In Herodotus, the Persian Zopyrus contrived a plan to become a benefactor to Darius, 'for good service (αἱ ἀγαθοεργίαι) among the Persians is highly honoured' (Hdt.3.154.1). 92Histiaeus is designated εὐεργέτης for his service to 'Darius and the Persians'. 93Certainly the terminology of εὐεργεσίη in Herodotus is almost exclusively used with respect to the Persians, as well as being a motif in Thucydides' account of Themistocles. 94Historian, subject and audience would have known well the significance of benefaction to the Great King.

'Weighing up deeds'
Relatedly, Themistocles' letter also displays awareness of the supposed Persian custom of 'weighing up someone's good deeds against the evil ones' 95 when Themistocles claims that the harm he did to the royal house was surpassed by the good (κακὰ μὲν πλεῖστα … πολὺ δ' ἔτι πλείω ἀγαθά).Herodotus describes the custom in his 'Persian ethnography', praising the nomos by which the king cannot kill anyone for a single cause, and that not until a calculation finds that someone's 'transgressions are more and greater than his services may a man give in to anger' (πλέω τε καὶ μέζω τὰ ἀδικήματα ἐόντα τῶν ὑπουργημάτων, οὕτω τῷ θυμῷ χρᾶται, 1.137.1). 96As Thomas notes, however, Herodotean Persian kings do not always behave accordingly, and Herodotus' description seems a Greek 'idealization'-albeit one based on certain Persian ideas of regal justice. 97till, this letter assumes knowledge of this supposed Persian custom-just as Histiaeus' capture scene in Herodotus-and Greek readers of the fifth century would no doubt have understood Themistocles' Persian 'cultural awareness' here.Just as Herodotus did with Histiaeus' speech to Darius, in this letter Thucydides demonstrates Themistocles' 'familiarity' with the language of Persian custom, creating the impression that Themistocles is adapting his expression to his royal audience.Alongside this Persianizing cultural intelligence, however, we also see the traditionally Greek Odysseus-figure in action-again, like Histiaeus-as Themistocles lies to the King about his role in holding back the Greeks from cutting the Hellespontine bridges (1.137.4).Both facets, perhaps, evince the μῆτις inherent to Themistocles' character.
'I am here, pursued by the Greeks because of my friendship with you.But let me have a year; I want to declare in person to you the reasons for my coming.'It is said that the King marvelled at Themistocles' intention and ordered him to go ahead with this plan.Themistocles spent this period learning the Persian tongue-as much as he was able-as well as the habits of the region.
According to Thucydides (1.138.2),upon arrival after this year Themistocles 'becomes great alongside the King' (γίγνεται παρ᾽ αὐτῷ μέγας), more than any other Greek up to this point (πω), primarily because of his intelligence (ξυνετός; cf.ξύνεσις at 1.138.3).The terminology of 'greatness alongside' an object of influence occurs often in Greek, including with respect to the Persian king; in the Classical period, however, only here and once in Herodotus do we see the specific combination of μέγας + παρὰ 9.18.3;cf.1.69.3 (Lydian), 3.47.1 (Spartan).For benefaction as a 'Herodotean' theme in Thucydides' excursus: Hornblower (n.81), 220. 95 [βασιλέι] + γίγνομαι. 98The Herodotean passage was mentioned earlier: Histiaeus' Persian rivals feared he would 'become great alongside the King' (6.30.1).Both passages immediately follow the explicit references to the bilingualism of Histiaeus and Themistocles.Modern commentators have not connected the two. 99Perhaps there is an implication that ability to speak directly to the King in his own tongue gave Greeks an advantage in gaining influence. 100If alluding to Herodotus here, Thucydides may be suggesting that Themistocles succeeded where even his bilingual predecessor Histiaeus could not. 101he rest of Themistocles' 'obituary' in Thucydides is overwhelmingly praising of the Athenian's brilliance.How does such a portrait square with the fact of Themistocles' exile half a century earlier by Thucydides' own citizenry?The intervening decades saw an apparent rehabilitation, and by the Peloponnesian War Athenians were claiming Themistocles as a local hero once more.102 This seems reflected in Thucydides' narrative structuring of the Pausanias-Themistocles excursus: 'The basic difference is that Pausanias' travel was a voluntary one caused by his own ambition, while Themistocles' travel was one forced by his circumstance.'103 Thucydides creates the impression that, in contrast to Pausanias, Themistocles was forced into medism.Given the encomiastic nature of Thucydides' general portrait of Themistocles, then, it is difficult to read his adoption of Persian as anything other than an additional marker of his excellence.104 While this conclusion has been stated before, what has been overlooked is how Thucydides' description of Themistocles' medism is important for our assessment.Themistocles is depicted in control of the medizing process: in his language learning, the qualifier 'as much as he was able/as much as was in his power' (ὅσα ἐδύνατο) offers a contrast to Pausanias' inability to control his adoption of Persian customs.105 Indeed, Thucydides uses the same verb twice to describe the Spartan's lack of control (1.130.1): After the receipt of this letter, Pausanias became prouder than ever, and was no longer able (οὐκέτι ἐδύνατο) to live in a customary manner, but went out of Byzantium in Persian dress, was attended on his march through Thrace by a bodyguard of Persians and Egyptians, kept Persian dining habits, and was quite unable to contain his intentions (κατέχειν τὴν διάνοιαν οὐκ ἐδύνατο) … Thus, while Pausanias' medism was initially premeditated, his consequent adoption of Persian ways is presented as irrational and beyond his own control, in tyrannical-or even stereotypically 'barbarian'-fashion. 106 Indeed, in the whole of the excursus (1.128-138), culturally loaded βάρβαρος-language is only used with reference to Pausanias' intriguing with, and imitation of, 'the barbarian(s)' and never concerns Themistocles, whose dealings are exclusively with 'Persians' and 'the King'. 107oreover, while Themistocles is said to learn 'habits' (ἐπιτηδεύματα) as well as language, only Pausanias is shown to perform Persian customs (1.130). 108As we have seen, Themistocles certainly displays awareness of Persian customs in his letter-but his performance remains restricted to language.Overall, then, in contrast with Pausanias, Themistocles' initial medism is involuntary, but after that he directs its progression; learning Persian language and customs appears as a chiefly practical, rational decision. 109Thus Themistocles' calculated delay and study of Persian is simply more evidence for Thucydides' praise of his intelligence and reason.Now, one might argue that Thucydides protests rather too much-that he presents Themistocles in this way as a polemic against rival narratives critical of his medism.So, for example, he might be underplaying Themistocles' willingness to learn Persian or the extent of his knowledge ('as much as he was able' could perhaps be read differently, for instance).However, this would still be evidence that individual bilingualism was not the inherently transgressive act it is often painted as for contemporary Greeks.On the contrary, it would suggest that one could 'do it right'.Thucydides does not attempt to deny or dismiss Themistocles' bilingualism, nor does he really dwell on it at all.Learning Persian language and ἐπιτηδεύματα are just two smart things for him to do.

CONCLUSION
What did it mean to be a Greek who spoke another language in the fifth century B.C.E.? It is now clear that individual bilingualism was not inherently a sign of divided loyalties or a betrayal of one's Greek identity.The texts of Herodotus and Thucydides evince no innate distrust in bilingualism across the class spectrum-from non-elite intermediaries to kings, tyrants, and generals-and from Greek to non-Greek.The fifth century provides us with only two examples of bilingual Greeks to analyse, both elite figures, and while it may have been that bilingual non-elite Greeks were viewed differently, we cannot know this.Nor do the portraits of Histiaeus and Themistocles differ significantly from those of non-Greek figures on the matter of individual bilingualism.None of this means that Histiaeus and Themistocles (or Scyles) are without criticism in these textsonly that such criticism is not a direct result of their bilingual abilities.The burden of proof should now lie with those who argue for transgression.
Analysing the bilingualism of Histaeus and Themistocles within the context of their broader character portraits has provided a wider cultural lens and brought out the 'Persianisms' in their communication long before each is explicitly said to speak the language.These results suggest that bilingual ability fits into a broader matrix of μῆτις and emerges as a culturally code-switching form of cleverness.Odysseus' 'multilingual' dexterity in epic and tragedy provided the perfect pre-existing model into which Histiaeus and Themistocles could fit. 110That these two were the most notable contemporary representatives of μῆτις might even suggest that bilingualism in this era was an excellent tool in the arsenal of cunning political players.This view is strengthened by the fact that our two case studies possess the form of bilingualism most open to tropes of dual loyalty or betrayal-as Plutarch's later anecdote shows us-and yet still these fifth-century accounts of their Persian-speaking abilities show no evidence of such a perspective.
Why, then, does the notion of transgressive bilingualism still linger?First, the conflation of attitudes towards bilingual intermediaries and bilingualism more broadly, which, as we have seen, neglects key differences in class and status.While not a major factor for the fifth century, nevertheless these differences in the Roman era have resulted in wider confusion.
Second, the retrojection of Roman-era linguistic debates onto fifth-century evidence.The different cultural politics of (especially Greek-Latin) bilingualism in the Roman period have been the focus of bilingualism scholarship, and so have been privileged when approaching the Classical evidence without sufficient attention to the differences.The spectre of Second Sophistic Greek perspectives on language especially still haunts these parts.
Finally, the conflation of language and customs, two aspects of cultural identity often linked but nevertheless distinct.Both Histiaeus and Themistocles employ or learn Persian customs as well as language, and separating these aspects shows that there could be transgression in adopting foreign customs-as the cases of Pausanias and Anacharsis, for example, make clear.Evaluation of an individual's 'cultural characterisation' must be made on a case-by-case basis, taking into account the full literary and cultural complexity of these representations of identity. 111From such analysis we can conclude that individual bilingualism was not inherently transgressive for Greeks of the fifth century-indeed, it was likely associated with the quintessentially Greek virtue of μῆτις.

DYLAN JAMES
Center for Hellenic Studies, Harvard University dylan.s.james@gmail.com 110The first reference to bilingualism in Greek literature is a lie told by Aphrodite in the context of establishing a persuasive false identity (Hymn.Hom.Ven.113-16); I owe this point to one of CQ's readers.Cf. also Iphigeneia's deception of the Taurians through 'barbarian' song (Eur.IT 1337-8) and Orestes' cunning use of Phocian (Aesch.Cho.563-4).