Darius the slave

One of the more curious statements made by Greco-Roman authors about Darius III, the last of the Great Kings of Achaemenid Persia, is that he had been a slave before his accession. Briant has relatively little new to say about this matter in the latest edition of his monograph on Darius, other than that the accusation was made ‘inaccurately’.Footnote ⁵ It is therefore opportune to dig somewhat more deeply into why this charge surfaced in the Greco-Roman sources. Building on Badian and Briant's thoughts, we attempt here to determine more precisely the rationale behind the source tradition of ‘Darius the slave’, and its implications for the presentation of Darius in the various source traditions underpinning the extant Greco-Roman accounts of his accession and brief reign.Footnote ⁶ Of particular interest is that none of the major narratives that we have for Darius’ reign – which, of course, focus primarily on the eastern conquests of Alexander the Great – introduces a servile origin for him. Arrian, whose work owes much to the ‘eyewitness’ accounts of Aristobulus and Ptolemy, certainly says nothing of the sort. He does not even delve into Darius’ past, which coheres with Payen's view that Darius is a ‘figure très marginale’ in the extant narratives of Alexander's accomplishments.Footnote ⁷ Likewise, Diodorus Siculus and Curtius Rufus, whose narratives owe more to the so-called ‘Alexander Vulgate’ – a problematic term used here more for the sake of convenience than anything else – fail to touch on the charge of slavery.Footnote ⁸

Plutarch, who does not really belong to either of the two main source traditions, as he appears to have utilised a wide variety of sources, alludes to a servile state for Darius in his De fortuna Alexandri (Mor. 326e). Here, Plutarch has Alexander, in a hypothetical discussion with Fortune (Τύχη), state that ‘Darius was your handiwork, he who was a slave and courier of the king’ (Δαρεῖος ἦν σὸν ἔργον, ὃν ἐκ δούλου καὶ ἀστάνδου βασιλέως …), and that it was Fortune who made Darius master of Persia.Footnote ⁹ Plutarch elaborates more on the same theme in his Vita Alexandri. At Alex. 18.4, which broadly seems to follow the Alexander Vulgate tradition given that a similar story is presented by Curtius (3.3.2–5), Plutarch describes a dream of Darius. In this dream, he saw the Macedonian phalanx in flames, and Alexander dressed in the Persian king's former attire, that of a ‘royal courier’ (ἀστάνδης ὢν βασιλέως).Footnote ¹⁰ All this was interpreted by the Magi as a positive sign for Darius, while Plutarch observes that the dream prophesied that Alexander was upon the same road to mastery of Asia, with the burning phalanx signifying the brilliance of the Macedonian military exploits to come. Now the word ἀστάνδης, repeated by Plutarch at Alex. 18.5, appears to be of Persian origin, as the Suda suggests,Footnote ¹¹ yet its usual translation as ‘courier’ could well be misleading: Badian's contention that Darius held ‘an appointment in the King's “postal service”’ is probably closer to the mark.Footnote ¹² Aelian (VH 12.43) follows the same thread as Plutarch and states, in the context of naming people who attained high status from humble origin, that ‘the last Darius’ (ὁ … τελευταῖος Δαρεῖος), defeated by Alexander, was a ‘slave’ (δοῦλος).

Αnother source tradition omits the notion of Darius having once been a slave or a royal courier, but claims that he was not of the royal family, or at least the immediate royal family. In an extremely brief summary of Persian history, Strabo (15.3.24) states that Darius was made king by Bagoas, described as a eunuch, upon the murder of Arses (Artaxerxes IV). But he was ‘not of the royal family’ (οὐκ … τοῦ γένους τῶν βασιλέων). Diodorus provides a fuller account. With Bagoas’ assassination of Arses, Diodorus (17.5.4–5) writes that there was no one left alive in the direct line of ascent, and we are told that Bagoas was forced to look elsewhere for a royal puppet. He settled on Darius, designated by Diodorus (17.5.5) as one of the ‘friends’ of the king (ἕνα τῶν φίλων), i.e. a member of court, as Bradford Welles’ translation suggests.Footnote ¹³ Darius is described as the son of Arsanes, and grandson of a brother of Artaxerxes, one Ostanes (17.5.5). This would mean that Darius was, in a way, descended from Persian royal blood. Badian thinks it more likely that this genealogy was a fiction, and that Darius was an Achaemenid on his father's side, but was perhaps of other stock – possibly Semitic – on his mother's side. This helps to explain why he was not murdered by Ochus (Artaxerxes III), who allegedly killed numerous family members.Footnote ¹⁴ Darius, or Codomannus as he was supposedly known before his accession, at least according to Justin (10.3.4), was simply not perceived as a threat to the king's power.Footnote ¹⁵ And so he survived. The new king also appears to have been a seasoned campaigner, with Diodorus (17.6.1) noting that, in the reign of Artaxerxes III, he single-handedly killed a fearsome champion of the Cadusians in mortal combat. So with this hardly obscure origin in mind, let us now turn to explaining the origin of ‘Darius the slave’.

Badian proposes that δοῦλος is ‘the standard Greek translation of OP [Old Persian] bandaka, one “bound” to a superior, especially the King’, a view supported by Briant.Footnote ¹⁶ He also observes that Darius I uses the term in many places in the Bisitun inscription to describe his senior military officers, including a member of one of the other six families that had participated in overthrowing the false Smerdis/Bardiya and installing Darius on the Persian throne. In fact, Darius uses the term bandaka for Gobryas (DB 5.71), Intaphernes (DB 3.50) and Hydarnes (DB 2.25), all part of the Seven. In addition, he uses the term for other elite persons, such as Dardaršiš, satrap of Bactria (DB 3.38) and Vivana, satrap of Arachosa (DB 3.45).Footnote ¹⁷ In another text, Darius calls Gadates, his satrap in Ionia, his ‘slave’.Footnote ¹⁸ It therefore seems clear that the word does not mean ‘slave’ in the same sense as the Greek δοῦλος. Rather, the Old Persian bandaka arguably entails honourable service, possibly with undertones of piety given the semi-divine status of the Persian king, and not outright abasement and servitude. In her translation of the text, Brosius simply uses ‘subject’, which seems apt enough.Footnote ¹⁹ But Greek writers who came across the term in the context of Darius, presumably in the context of early propaganda material attempting to legitimise Alexander's rule and delegitimise that of Darius, used it indiscriminately to describe the last Achaemenid king.Footnote ²⁰ By the time of Plutarch and Aelian, the original sense of the word had clearly been lost. It is therefore necessary to explore further how the term was used by these authors, and presumably by the sources underpinning their brief accounts of Darius, since it should seem reasonably clear that they are using δοῦλος for reasons that are more rhetorical than historical.

It might well be suggested that, in the source tradition privileging the use of δοῦλος to describe Darius, he was characterised as a slave, or at least a low-born Persian, in order to add a greater legitimacy to Alexander's succession. In short, Darius, put on the throne by a puppet-maker, was cast as being ‘not a proper king’, presumably because he was not of the immediate royal family, even if Alexander did see to it that Darius was buried in the royal tomb (Arr. Anab. 3.22.1, with 3.22.6). But this does not necessarily explain why Darius was specifically called a slave in the source tradition of interest to us. In Greek and presumably also Macedonian eyes, all subjects of the Great King were regarded as slaves. This was not just in the sense of one ‘bound’ to a superior, as the original Old Persian term suggests, with its connotation of honourable service; rather, the term was used in the sense of a completely subordinated person prepared to abase him or herself before a master; witness the opposition of many Greeks to the performance of proskynesis in Alexander's day, as is exemplified by Callisthenes of Olynthus’ objection to the practice (Arr. Anab. 4.10.1; Plut. Alex. 54.1–2).

By contrast, freeborn Greeks believed, rightly or wrongly, that they were not subjected to the same level of domination by their political leadership. For example, Isocrates (Paneg. 150) writes that Persians, as a whole, are a mob ‘trained more effectively for servitude than slaves in our country’ (πρὸς δὲ τὴν δουλείαν ἄμεινον τῶν παρ’ ἡμῖν οἰκετῶν πεπαιδευμένος).Footnote ²¹ Even the Macedonians of Alexander, who obviously lived under a royal regime, with a king at its head, might not have equated their condition with that of the Persians ruled by the Great King. Plutarch (Alex. 51.3), by way of example, refers to the Macedonians not being slaves like the Persians. In addition, we witness numerous instances of the Macedonians – even the common soldiery – arguing with Alexander, such as when they refused to go any further into India in 326 BC (Arr. Anab. 5.25.2).Footnote ²² This ability to question the king's orders seems to have been regarded as a fundamental right, yet it was something which would not have been countenanced when dealing, as a subject, with a Persian king.

Thus Darius III was characterised as a slave in one delegitimising source tradition, simply because he was not an immediate descendant of a Persian king, even if he was clearly a relative of kings. The absurdity of the accusation is magnified by the fact that even a brother of a king was also technically a slave, as Xenophon (An. 1.9.29, 2.5.38) suggests: he twice refers to Cyrus the Younger as the ‘slave’ (δοῦλος) of Artaxerxes II, his brother. Every Persian king had therefore been a bandaka at some point. But to the Greeks, Darius, not having been born into a position conducive to becoming king, would have had to abase himself on a regular basis. Could such a man really be a king? In the eyes of the delegitimising source tradition argued for here, presumably not. Of course, Alexander himself – at least according to our main sources for the period – seems to have had rather different ideas on this matter, with Darius’ regicide at the hands of Bessus and Nabarzanes allowing him be portrayed as an avenger of Achaemenid honour, particularly through his castigation of the ‘usurper’ and ‘parricide’ Bessus.Footnote ²³ Alexander was thus able to position himself as a restorer of the proper order of things, rather than one who had swept away an unworthy ‘slave-king’ who had been placed on the throne, as a puppet, by the intrigues of Bagoas.Footnote ²⁴ Curtius (5.9.9) seems to have followed this way of thinking, for he makes Darius call the treacherous Nabarzanes, when the latter had proposed to transfer the kingship to Bessus, satrap of Bactria, ‘basest of slaves’ (pessimum mancipium) – and this without any hint of irony.Footnote ²⁵ That Curtius could make Darius call Bessus, a satrap and a man possibly also of distant Achaemenid blood, manceps again reinforces the view that all Persians aside from the king could be regarded as slaves. Since Curtius does not dwell on Darius’ origin, calling Bessus a slave should be no cause for alarm – indeed, it makes perfect sense.

To return to the source tradition of interest to us, describing Darius as a slave, in the sense of someone who had once been in what Greek authors might regard as a form of bondage, was always a rather specious claim, and probably a manifestation of early Macedonian attempts to delegitimise Darius’ rule before Alexander's final victory – after which Darius was accorded a more legitimate royal status, with the tag of ‘usurper’ being transferred to Bessus, thus allowing Alexander to be seen not as a deposer of an unfit monarch of base origin, but as a restorer of order and a champion of some sort of continuity. As Briant has continually sought to show us, the character of Darius is used as a cipher through which we are meant to form an opinion on Alexander. The facts underpinning the narrative, including Darius’ former (honourable in Persian eyes) position as a bandaka, were therefore manipulated to advance a larger narrative project – even if we are now only left with fragments of this endeavour, such as those of Plutarch and Aelian. Briant's description of Darius being ‘inaccurately’ called a former slave in the normal Greek sense of the term is therefore quite correct. But this error arguably stems from an inability on the part of our extant sources, seemingly not sufficiently cognizant of the intentions of the sources underpinning their assertions regarding Darius’ background, to provide the correct nuance of the Persian term bandaka, which word could really be applied to any Persian.Footnote ²⁶ As a result, the more familiar sources referring to Alexander's castigation of the ‘parricide’ Bessus and his burial of Darius with the usual royal honours, such as Arrian, Curtius and Diodorus, arguably embrace a different narrative thread compared to those preserving the view that Darius was low-born, and indeed a slave. In the end, it is not necessary to regard the use of ‘slave’ as an outright fabrication on the part of the immediate sources underlying Plutarch and Aelian; rather, this inaccurate use of a Persian term, misappropriated by those writing towards the end of the Achaemenid period or shortly thereafter, fed into one interpretation or even literary appropriation of the Persian monarchy, and how those living under such a regime were perceived by the Greeks and Macedonians.

The arrest of Darius

Much has been written about the arrest and deposition of Darius, particularly on the motives of the conspirators themselves, with most authors, such as Briant, merely pointing to the highly ranked nobles Bessus and Nabarzanes as the instigators, while Payen, in his review of the first edition of Briant's work, simply blames ‘ses proches’.Footnote ²⁷ But little has been done to establish which troops, in particular, were responsible for carrying out these orders. An intriguing opportunity to establish the precise agents of Darius’ capture and downfall is presented by Justin's epitome of the history of Pompeius Trogus, composed at some uncertain point but probably in the early third century AD.Footnote ²⁸ He makes the following curious statement at 11.15.1: interea Darius in gratiam uictoris a cognatis suis aureis conpedibus catenisque in uico Parthorum Thara uincitur. This is translated by Yardley as: ‘Meanwhile, in the Parthian village of Thara, Darius was bound in fetters and chains of gold by his own kinsmen, who wished to conciliate the victor.’Footnote ²⁹ This has occasioned some rather confused commentary. In the same volume, Heckel observes, in his commentary on the passage, that ‘Bessus is not called a kinsman of the king (cf. Arr. Anab. 3.21.5, 30.4), though it is not clear whether he [Bessus] was related by blood or marriage, as Badian and Bosworth contend, or if, in fact, he belonged to the group known as the King's kinsmen (syngeneis = cognati).’Footnote ³⁰ Chambrey and Thély-Chambry translate cognati as ‘ses propres parents’, though no commentary is offered on the matter.Footnote ³¹

Now Heckel clearly assumes that Justin, at the locus in question, is referring, in very specific terms, to the men responsible for putting Darius in chains. In his treatment of the problem, Heckel holds that the answer to the question of what Justin means by cognati must lie in determining if Bessus (a) was indeed a blood relative of the king,Footnote ³² (b) was related to him by marriage, or (c) was one of the mysterious group described, in more figurative terms, as the ‘kinsmen’ of the king, i.e. men of a certain exalted social rank.Footnote ³³ It would appear that these are the supposedly fifteen thousand cognati regis (‘kinsmen of the king’) described by Curtius at 3.3.14, in the context of a rather fanciful procession of Persia's military forces before the battle of Issus (333 BC). Yet there remains another possibility: the so-called ‘Kinsmen Cavalry’ or συγγενεῖς ἱππεῖς, a group which could presumably also be presented by Latin writers as cognati. These particular συγγενεῖς were organised into a 1,000-strong unit, if we follow Diodorus’ description (17.59.2) of them at Gaugamela in 331 BC, and constituted the Great King's cavalry bodyguard, with the foot guard being the so-called ‘Apple Bearers’ or μηλοφόροι, also presumably a thousand strong.Footnote ³⁴

Let us see if the involvement of the συγγενεῖς ( = cognati) in Darius’ capture represents a genuine possibility. In Gaugamela's aftermath, we are told by Arrian (Anab. 3.16.1) that Darius fled, accompanied by many (πολλοί) of his μηλοφόροι (the rest having presumably fallen), the Bactrian cavalry, which had been posted close to Darius in the battle, and the συγγενεῖς.Footnote ³⁵ In his summary of Darius’ life, Arrian (Anab. 3.22.5) described the Persian king being ‘at the end a fugitive from his own kingdom and a wanderer, betrayed by his own escort to the worst of fates, to be at once a king and prisoner carried off in dishonour’ (φυγάς τε ἐκ τούτου τῆς αὑτοῦ ἀρχῆς πλανώμενος καὶ τελευτῶν πρὸς τῶν ἀμφ’ αὑτὸν ἐς τὰ ἔσχατα προδοθείς, βασιλεύς τε ἐν τῷ αὐτῷ καὶ δεσμώτης ξὺν ἀτιμίᾳ ἀγόμενος).Footnote ³⁶ Here, τῶν ἀμφ’ αὑτόν could conceivably refer to those troops who still accompanied Darius, and his remaining guard troops in particular, the greatest part of which, according to Arr. Anab. 3.16.1, were the συγγενεῖς = cognati.Footnote ³⁷ It is unfortunate that Arrian (Anab. 3.21.1) does not describe Darius’ imprisonment, but merely reports that he had been arrested – curiously in the following order: by Nabarzanes, a Persian military commander, Bessus, satrap of Bactria, and Barsaentes, satrap of the Arachotians and the Drangians (or Zarangians), even though it was Bessus who supposedly claimed the kingship (Arr. Anab. 3.21.4).

Of clear relevance, too, is the precise role of the conspirator Nabarzanes in the arrest and subsequent murder of Darius. This Nabarzanes, although not specifically described in the context of Gaugamela itself, is variously described by Arrian as ‘chiliarch of the cavalry which had taken flight with Darius’ (χιλιάρχης τῶν ξὺν Δαρείῳ φευγόντων ἱππέων, Anab. 3.21.1), and ‘Darius’ chiliarch’ (ὁ Δαρείου χιλιάρχης, Anab. 3.23.4). Although some have seen Nabarzanes solely as a commander of the thousand-strong Apple Bearers, or – less likely – the commander of the Bactrian cavalry, there is reasonable cause to hold that he was, at the very least, associated with the συγγενεῖς ἱππεῖς.Footnote ³⁸ It is even possible that, given the unusual set of circumstances at the end of the Achaemenid era, Nabarzanes had nominal command of all the guard units, which would explain his high level of influence in the subsequent machinations against Darius.Footnote ³⁹ If so, we have further cause to hold that Darius’ own cavalry guard was in some way involved in the king's actual arrest, after he had been betrayed by their likely commander Nabarzanes.Footnote ⁴⁰

So where does the story told by Trogus through Justin originate? We turn next to Diodorus (17.73.2). But he only writes that Darius made his way to Bactria, supposedly accompanied by 30,000 Persians and Greek mercenaries, and was arrested and thereafter murdered by Bessus, the satrap.Footnote ⁴¹ There is no mention of co-conspirators. The next stop must therefore be Curtius, who based his work largely on the late fourth- or early third-century BC Cleitarchus, yet also had recourse to the autobiography of Ptolemy and – more importantly for our purposes – the history of Trogus.Footnote ⁴² This has significant implications, for Diodorus, although he generally follows the Alexander Vulgate tradition rather than the apologetic tradition of Arrian and the Itinerarium Alexandri, says next to nothing about Darius’ arrest, while Curtius is our main source for it. Given that, as we shall see, there are marked similarities between (a) the little that Justin says about the matter, and (b) what Curtius has to say, one might reasonably hypothesise that they share a basic underlying narrative, which one might presume would have been found in Trogus’ history. Atkinson is firmly of this view, and discounts the possibility that Justin ‘departed from Trogus’ text to incorporate passages from Curtius’.Footnote ⁴³ Let us keep that in mind as we explore Curtius’ description.

At 5.9.2, Curtius writes that Nabarzanes, together with Bessus, had formed a conspiracy to arrest the king ‘by means of the forces which they both commanded and putting him [i.e. the king] in fetters’ (regem suum per milites quibus ambo praeerant comprehendere et uincire decreuerant). We then read (5.10.5–7) that the two conspirators resolved to have the Bactrian troops, being the most likely to carry out the order, arrest the king, since ‘so many thousands of the Persians would bear aid to their king’ (tot Persarum milibus laturis opem regi). In the end, the conspirators resolved to carry out the deed, and ‘the men accustomed to form his bodyguard, who ought to have protected the king's safety even at the peril of their lives’ (custodiae eius assueti, quos regis salutem uel periculis uitae tueri oportebat, 5.12.9), left the king to what seem to be the Bactrians. If so, the traitorous troops to whom Curtius refers must be the Apple Bearers and the presumably more numerous συγγενεῖς. If further evidence is required, one might well note that, in the context of Issus, Curtius (3.9.4) describes Darius’ cavalry guard, supposedly composed of 3,000 elite horsemen (tria milia delectorum equitum), as ‘his customary bodyguard’ (assueta corporis custodia). The similarity of the language (compare custodiae eius assueti with assueta corporis custodia) could mean that it is possible that Curtius (or rather his underlying source, since Curtius does not seem to refer to cognati anywhere in his narrative as the cavalry guard) had the συγγενεῖς = cognati in mind on both occasions.

With his own guard having turned a blind eye, Darius was ‘seized and bound’ (comprehendi uincirique, 5.12.15), again presumably by the Bactrians. A short time thereafter, certain unnamed men ‘bound Darius in fetters of gold’ (aureis compedibus Dareum uinciunt, 5.12.20). Rolfe believes that the subject of uinciunt ‘is, of course, not Persae, as the connection might suggest, but Bactriani, or better, Bessus et Nabarzanes. This is clearly shown by the following ludibria, and by Justin xi. 5 a cognatis suis … uincitur.’Footnote ⁴⁴ Yet there is no certainty that Curtius means that it was the Bactrians and the two named conspirators who bound the king in golden fetters. Furthermore, Rolfe appears to ignore that, in the context of the Achaemenid military, cognati = συγγενεῖς could also refer to elite cavalry.Footnote ⁴⁵ In addition, the preceding sentence at 5.12.19 describes the Persians being persuaded by Bessus to fall in with the designs of the conspirators, and joining themselves with the Bactrians (coniunxere se Bactrianis). But the sentence of interest to us, namely 5.12.20, suggests a certain degree of remorse on the part of the perpetrators – especially that they applied golden fetters out of respect for the king's honour (ne tamen honos regi non haberetur) – which arguably fits in with the guilt-burdened Persians better than the shameless Bactrians. One might well reflect on 5.12.13, where the Persians were paralysed by fear (attoniti metu) as the conspiracy got under way, and did not know whether to take up arms, or simply let matters run their course (nec quiescere audebant, ne impie deserere regem uiderentur). Finally, pace Rolfe, it is unclear why the subject of uinciunt should be anyone other than the Persae of the preceding sentence, since it was they who ‘joined themselves with the Bactrians’, who are in the dative case:

One might naturally extract from this that Persae coniunxere se Bactrianis … [et] aureis compedibus Dareum uinciunt. In short, it is more likely that it was the Persian troops rather than the Bactrians who ‘honoured’ the king by tying him up with golden fetters, and that a good proportion of these men would have been the cavalry guard, the cognati = συγγενεῖς, even if they are not specifically named as such in the vignette presented by Curtius. What is more important for us is the possibility that Curtius’ source did identify them.Footnote ⁴⁷

To reflect back on Justin's problematic interea Darius in gratiam uictoris a cognatis suis aureis conpedibus catenisque in uico Parthorum Thara uincitur at 11.15.1, it is possible to see those whom he describes as binding Darius in fetters of gold as members of the king's cavalry guard – or at least the guard in general in a collective sense. Although we clearly cannot be sure without having recourse to Trogus’ original text, which might have provided more details about which actors precisely were on stage at this part of the tragedy, it is unwise to state categorically that Justin's cognati can refer only to Bessus and Nabarzanes. The discussion above shows that the word could also refer to the king's guard cavalry or – to extrapolate further, and feed into the tragic rhetoric surrounding Darius’ downfall and his subjects’ treacherous conduct – those who were known to him and were meant to protect him. Such an interpretation would obviously include Bessus and Nabarzanes, but would also encompass his Persian guard, including the cognati = συγγενεῖς. If so, we have further evidence that the fall of Darius was presented in a way that underlined the supposed barbarism and cowardice of the Persians, an outcome which, in a way, contributes to the portrayal of Darius as a rather pathetic figure, abandoned even by his closest followers, with only Alexander and his army being interested in keeping him alive (Curt. 5.13.3–5, with Arr. Anab. 3.21.1; Diod. Sic. 17.72.3) – all of which helps to justify the Macedonian intervention into Persian affairs, and the end of Achaemenid rule.