In Search of the Lost Testament of Alexander the Great

by David Grant

  But Heracles’ story began some twenty-four years earlier and in far less tragic times for Macedonia. Following the Persian defeat at Issus in November 333 BCE, the royal family of Darius III was captured and it included his wife Stateira, his mother (Sisygambis), two adultae virgines (his daughters Stateira and Drypetis), and a ‘not yet past his sixth year’ son (Ochus).164 Immediately after Issus, Parmenio captured other notable Persian women at Damascus, including the wife and the three ‘maiden’ daughters of the former Great King known as Artaxerxes III Ochus, whose son (Arses, Artaxerxes IV) Darius had deposed with the help of the eunuch Bagoas. We may assume one of these daughters was Parysatis, Alexander’s later bride at Susa, though the other two remain anonymous. Also listed in this set of captives was the daughter (probably Amastris) of Oxyathres the brother of Darius III. To the tally of captured women we need to add the Great King’s concubines, allegedly 329 in number.

  Additionally apprehended were the wife and son of Pharnabazus (the son of Artabazus) who had been given supreme Persian command of the Aegean coast by Darius, along with the three unnamed daughters of Mentor, the some-years-dead Rhodian mercenary who had fought both against, and then for, Artaxerxes III. Next were listed the ‘wife and son of the renowned general Memnon’; he was Mentor’s brother and Alexander’s most talented opponent in Asia Minor who had died of illness much more recently. We are led to believe that this captured wife, Barsine, became Alexander’s mistress; we are told she was a daughter of Artabazus, the ‘chief of courtiers’ and the former Persian satrap of Hellespontine Phrygia (grandson of Artaxerxes II and nephew of Artaxerxes III). Artabazus’ son (Ilioneus) and wife were also in the group; she was possibly the Rhodian sister of Mentor and Memnon who had borne to Artabazus eleven sons and ten daughters, that is if a later bride was not being referred to here as might be suggested by Ilioneus’ immaturity. Alongside this royal haul at Damascus, hostile Spartans and Athenians had been rounded up along with a useful cache of 2,600 talents of coined money, 500 pounds of wrought silver and 7,000 loaded pack animals.165

  Artabazus had initially refused to recognise his uncle, Artaxerxes III, whose pogrom wiped out his brothers’ rival lines (when eighty of Sisygambis’ brothers by various concubines were killed in a single day) as he ascended the throne in 358 BCE; the new Great King then issued a royal edict that the mercenary armies of the satraps were now to be disbanded. Although assisted by his Rhodian brothers-in-law in the so-called Great Revolt of the Satraps, Artabazus was finally defeated and took refuge with his family at the court of Philip II in Macedonia in 349/348 BCE, along with Memnon.166 So he, along with his large clutch of offspring and his talented son-in-law were well known to (the then young) Alexander;167 his ‘honourable’ surrender to the Macedonians in Hyrcania some three years after Issus, and the reported warmth between him and his captor, stemmed from this former ‘guest exile’, whereupon Alexander appointed Artabazus as his new satrap in Bactria. Memnon had fared worse; he had died when still in opposition to the Macedonian invasion of Asia.168

  The dense and intertwined branches of the line of Artabazus with the Rhodian brothers and sister, the inter-related Achaemenids and the captive lists from Issus and Damascus with their female anonymae, were bound to provide latitude for confusion to historians.169 For here we have three daughters of Artaxerxes III Ochus, three of Mentor, two of Darius, and ten of Artabazus; most were unnamed and all of them could be referred to as ‘royal’.170 To add further scope for misidentification, Justin described the allure of these regal women: ‘He [Alexander] fell in love with his captive Barsine for her beauty, by whom he had afterwards a son that he called Heracles’, though Justin later linked Heracles to Roxane in his careless epitomising form, whilst Porphyry erroneously stated Roxane was the daughter of Darius.171 But it seems all prominent royal women were described as visually prepossessing: Darius III’s two daughters were also complimented this way, whilst his wife was voted ‘the most beautiful woman in all Asia’, with Roxane a runner up; Plutarch additionally stated that Darius’ daughters resembled their handsome parents.172

  Plutarch’s rundown of Barsine’s qualities included her ‘agreeable disposition’, and thus ‘Alexander determined (at Parmenio’s instigation, claimed Aristobulus) to attach himself to a woman of such high birth and beauty.’173 But, once again, this description could fit the daughters of both Great Kings, and, in particular, the daughter of Darius who was curiously named ‘Barsine’ by Arrian.174 Tarn convincingly argued that the ‘beautiful captive’ Parmenio urged Alexander to marry was one of Darius’ two daughters, for that would clearly provide the legitimacy the Macedonian king sought in Asia.175 Darius had reportedly proffered his daughter to Alexander previously, along with a portion of his empire, and Parmenio urged him to accept the olive branch.176

  The bold rejection we read of may not in truth have been so confidently drafted, as the marriage would have been attractive to a man who had now conquered the western satrapies of the Persian Empire and who now sought legitimacy in the East. Alexander allegedly replied with: ‘That which was offered was already his.’ The result was that he had Darius III’s daughters with him for perhaps two years after Issus and may have come to know them well, despite Plutarch’s claim that the women continued ‘to live as though guarded in sacred and inviolable virgins’ chambers instead of in an enemy’s camp, apart from the speech and sight of men.’177 Leaving them at Susa in 331 BCE, Alexander ordered that the princesses be given a Greek education, which in itself suggests he had unique future plans for the girls: he and Hephaestion married them when they returned to Susa some seven years on.178

  It was Plutarch (uniquely it seems) who claimed Heracles’ mother was one of the daughters of Artabazus, who was already in his late sixties when Barsine was captured post-Issus in 333 BCE (though Curtius implied he was around ninety).179 To truly entangle matters, Plutarch stated Barsine had been the wife of Memnon (Tarn thought he had followed the inaccurate Duris in this; Curtius and Diodorus never named Memnon’s widow), but Arrian stated that Barsine had a daughter by Mentor; this would suggest (Tarn called it ‘a modern invention’) that Barsine had married her dead husband’s brother, unless he, or later scribes, were confused by their similar names (Strabo may have made the same mistake). Plutarch’s statement reads:

  But Alexander, as it would seem, considering the mastery of himself a more kingly thing than the conquest of his enemies, neither laid hands upon these women, nor did he know any other before marriage, except Barsine. This woman, Memnon’s widow, was taken prisoner at Damascus. And since she had received a Greek education, and was of an agreeable disposition, and since her father, Artabazus, was son of a king’s daughter…180

  Here the Greek education rears its head again. Plutarch believed Barsine was the only woman Alexander had consorted with before his marriage to Roxane in 327 BCE; this is a romantic, and yet highly unlikely, proposition, and more so if we give any credence to Diodorus’ claim that Alexander consorted with concubines from Darius’ harem; he also detailed his (less likely) thirteen-day tryst with the Amazon queen, Thalestris.181 There is no further evidence that the widowed Barsine remarried Memnon; only a nameless ‘widow’ of that Rhodian commander was mentioned in the captive list by Curtius and Diodorus.182 Perhaps because Mentor’s wife was not listed, Plutarch (or his source) concluded Memnon was caring for his brother’s three orphaned daughters and thus he had married Barsine for the practical application of that. This may have further led Plutarch to conclude Memnon’s young son was born from the new union.183

  Clearly, judging by these associations, Barsine already had a number of children by the time she was captured. Taking this into account, Tarn quite reasonably concluded she was a woman of a different (older) generation to Alexander who was aged twenty-three when he captured her. The reasoning is further supported by evidence that a son of Mentor (if Barsine was his mother too) may have been mature enough for battle in 327/6 BCE; Memnon also had sons fighting beside him at the Gran
icus River in 334 BCE, though Plutarch must have assumed these were from a previous marriage (if indeed his above text originally referred to ‘Memnon’ and not ‘Mentor’).184 Moreover, there is no mention that this Barsine accompanied Alexander for the next six years, or longer, until 327-326 BCE, the time Heracles would have been born.185 We also know from Arrian, who stated that he was taking some of his detail from Aristobulus at this point (and probably from Ptolemy too) that one of Barsine’s daughters had attained marriageable age by 325 BCE at the Susa weddings:186

  … to Ptolemy the Bodyguard and Eumenes the Royal Secretary he gave two daughters of Artabazus, Artacama to Ptolemy and Artonis to Eumenes; to Nearchus the daughter of Barsine and Mentor…187

  The inference given by Plutarch when describing the same Susa weddings was that the Barsine who consorted with Alexander had only two sisters, and not two of ten, if this is not a misleading translation:

  For Barsine the daughter of Artabazus, the first woman whom Alexander knew in Asia, and by whom he had a son, Heracles, had two sisters; of these Alexander gave one, Apame, to Ptolemy, and the other, also called Barsine, to Eumenes.188

  The names Plutarch provided for the brides of Ptolemy and Eumenes (Apame and Barsine) are different from those given by Arrian (Artacama and Artonis) and it is unlikely (though not impossible) that Artabazus named two daughters ‘Barsine’. However, ‘Apame’ was a well-established Achaemenid name and one also attached to Seleucus’ bride.189 How could such confusion arise when eyewitness historians presumably passed down their identities? Ptolemy was a bridegroom at Susa and Aristobulus may well have been. Tarn, who paradoxically concluded Ptolemy’s detail came from the Eumenes-compiled court Journal, simply quipped: ‘One may suppose that they knew their wives’ names.’190

  Identifications within the interwoven royal lines of Persia are perilous; a read of Plutarch’s Artaxerxes (III) reveals the recurring use, through the generations (though not for two coeval daughters), of these traditional Achaemenid names, including Oxyathres, Apame, Stateira and Parysatis. Darius III himself had married a ‘Stateira’, his second wife, who was in fact his sister (or half-sister).191 He had been previously married to a sister of Pharnaces (possibly related to Artabazus), the Persian commander who died in battle at the Granicus River, and he probably had a daughter by her.192 And as it has been pointed out, when considering wives beside concubines, Darius III could have had many more than the three children who were named, and his mother, Sisygambis, may have had as many as seven children herself.

  A seemingly unexploited conclusion, surrounding the identity of Barsine, is that we are dealing with sisters from more than one generation of Artabazus’ line, with the confusion surely arising from what were ambiguous original references to ‘nieces’, ‘sisters-in-law’, ‘aunts’ and ‘uncles’, and probably ‘cousins’ too. For if Mentor had married a daughter of Artabazus, and recalling that Artabazus had in turn married Mentor’s Rhodian sister, then Mentor married his niece. This exhibits a true strategic bonding of the two families, and judging by the attested children from both sides it had proved a highly successfully union.

  This dynastic interweaving would make it quite probable that Memnon had also married another of Artabazus’ ten daughters. Furthermore, Mentor’s own three daughters by Barsine would have also been termed Memnon’s ‘nieces’, and vice versa, as well as them being the nieces of Artabazus’ wife (their sister). Barsine’s three daughters (assuming Mentor’s girls were by her) were additionally nieces to the remaining nine of Artabazus’ girls, whose own daughters (not mentioned but surely existing) were in turn the nieces of Barsine. In fact, the two unnamed daughters of Artaxerxes III Ochus (Parysatis was the third) were cousins of Artabazus.

  The potential age ranges of the captives, both the youngest daughters of Artabazus and the oldest of his granddaughters, could well have made them suitable for intimacy with their captors between Damascus in 333 BCE and Susa in 325 BCE, moreover, marriages between mature men and far younger women were commonplace then.193 It seems doubtful in this environment that Alexander would have chosen an older woman who was already the mother of numerous children as his mistress when far younger, equally royal and illustrious virgines filias were available to him.194

  It is quite possible that more of Artabazus’ ten daughters, who now depended upon Alexander for their wellbeing in the new Macedonian-ruled empire, were present at the Susa weddings. If ‘Barsine’ was a traditional family name, or title, it could have proliferated through their offspring. So a creditable explanation for two same-named girls is to conclude one of the three daughters by Barsine and Mentor was named after her mother, in which case she is a more credible candidate as the mother of Heracles. Considering the many corruptions that crept into manuscript transmission over two millennia or more, could a single archetypal sentence, stripped of its precision, have led to the different names given by Arrian and Plutarch? Well, the wording below, unpunctuated to highlight its potential ambiguity, could (along with other permutations) achieve just that:

  … to Ptolemy the Bodyguard and Eumenes the Royal Secretary he gave sisters of Barsine the daughter of Artabazus and to Nearchus Alexander gave a daughter of Mentor and Barsine who had two sisters including Alexander’s mistress also named Barsine to Eumenes Artonis Artacama to Ptolemy Apame to Seleucus the daughter of Spitamenes…195

  Without punctuation, the double references to Eumenes and Ptolemy look awkward, that until a break after the second mention of ‘Barsine’ would suggest the source was proceeding to name the aforementioned anonymous women. This wording would justify Plutarch’s belief that Heracles’ mother had two sisters (so accounting for the three daughters of Mentor, one named Barsine after her mother) and it overcomes Tarn’s objection that ‘… these women all belonged to an older generation.’ It would explain the presence of two Barsines, and the corruption (or loss) of punctuation would explain how both Nearchus and Eumenes were linked to a Barsine as well. Plutarch would then have believed the ‘stranded’ names of Artonis and Artacama were the remaining two (non-paired) sisters of Nearchus’ wife.

  The same ambiguity could additionally indicate how Ptolemy and Seleucus were linked in marriage to Apame, and explain why Strabo thought Apame was a daughter of Artabazus, where Arrian believed Seleucus’ unnamed wife was the daughter of Spitamenes.196 Strabo frequently used Aristobulus as a source and Aristobulus was the only source named by Arrian for detail of the Susa weddings, so perhaps it was he who provided the original unifying text.

  In this nexus of Persians, progeny and ‘beautiful’ prisoners, it seems a spurious bridge or two was built, with the mother of Heracles sitting somewhere in its span. What does appear to have happened, either through marriages to the daughters or granddaughters of Artabazus, is that Ptolemy, Nearchus and Eumenes became related through Heracles. Unfortunately, as events made clear, they thought it of little significance. Nearchus’ alleged speech at the Assembly in Babylon in which he made a case for the boy, certainly supports the connection, if it was not a back-construct of Cleitarchus. Eumenes also appears to have gained family loyalty, for Pharnabazus fought for him in 321/320 against Craterus.197 But we must recall that Heracles himself never featured in the Pamphlet Will, and this would appear counterproductive to Eumenes in the circumstances, that is until we recall Antigonus was in control of the boy. In which case Heracles’ absence from the testament was an attempt to neutralise exactly that, while it promoted the cause of Alexander IV who was in Olympias’ custody in Pella.

  The claim that Heracles was based at Pergamum in 323 BCE (T11), rather than Susa, for example, further supports his descent from Artabazus, for the family estates did reside in Hellespontine Phrygia with its southern border falling just to the north of the city.198 The family condottieri, Mentor and Memnon, had also been granted substantial estates in the Troad, probably for helping Artabazus to reclaim his father’s lands from Autophradates.199 And we can assume that following Artabazus’ voluntary retirement from ser
vice under Alexander, he and his family returned to their lands. It is tempting to credit his retirement and Barsine’s departure from the court and the king’s bed, to Alexander’s marrying Roxane, for according to Curtius, Roxane’s father, Oxyartes, was granted the governance of Bactria, the very satrapy Alexander had previously bestowed on Artabazus. But this appears to have taken place a little later, though much chronological confusion exists at this point in the extant texts.200

  It is universally assumed that Alexander never married his mistress, Barsine, so Heracles remained an unrecognised bastard son; this would of course go some way to explaining his rejection at Babylon. But at Susa, Alexander was demonstrably ‘marriage-minded’ in the face of whatever objections Roxane may have thrown up or the fears she harboured, for there he took the hand of Stateira the daughter of Darius III, and Parysatis the daughter of Artaxerxes III Ochus.201 If Heracles’ mother was descended from an Achaemenid line, there remains the question of why he did not legitimise his (then) only son.202 We cannot in fact be sure he did not, for a slight amendment to Arrian’s statement (or Aristobulus’ before him) changes the context of the marriages altogether, and explains why Arrian named Darius’ daughter ‘Barsine’ when the Vulgate texts gave us the more convincing ‘Stateira’. Consider the text as it stands relating to the weddings at Susa:

  He [Alexander] himself married Darius’ eldest daughter Barsine, and according to Aristobulus, another wife as well, Parysatis, the youngest daughter of Ochus (he was already married to Roxane, the daughter of Oxyartes of Bactria).

  Although this represents a modern translation descended from Arrian’s Greek, it once again requires no more than a punctuation change, with a comma or break to be placed after ‘Darius’ eldest daughter’ to suggest that three marriages took place including one to Barsine whose identity had already been established in texts above relating to the hetairoi marriages.203 The loss of punctuation (or spacing) led to Arrian’s confusion. When dealing with this grand event, whether Aristobulus was the sole source for Plutarch, Arrian, and Cleitarchus before them, or simply an auxiliary source for extra detail (as the above text could imply), we have shown that a single archetypal statement could in fact have provided the different conclusions we read.