Bardiya and Gaumata

Bardiya was the second son of Cyrus the Great and, by the agreement of the sources, the full brother of Cambyses: both were the children of Cyrus and Cassandane, daughter of Pharnaspes, an Achaemenid.^[1] The Behistun inscription is emphatic that the two were sons of the same father and the same mother, a point Darius needs, since it makes Cambyses the murderer of his own brother.^[2] Herodotus agrees that they were full brothers, and against the divergent genealogy of Ctesias, who made an Amytis their mother, the agreement of Behistun and Herodotus is decisive.^[3][4]

When Cyrus arranged his succession he named Cambyses heir, and had already marked the choice by giving Cambyses the title king of Babylon in 539. To the younger son he gave a great territorial command in the east, with an unusual privilege attached. In the version Ctesias preserved, Cyrus on his deathbed appointed Bardiya governor of the Bactrians, Chorasmians, Parthians and Carmanians, to be held free of the obligation to pass tribute up to the central treasury; Xenophon, telling the same scene, makes him satrap of the Medes, Armenians and Cadusians.^[1] The exemption reads as compensation, a domain rich enough to soothe the resentment of a son passed over for the throne.^[5] It also left him, for all his wealth, a dependant of his brother the king, a position Briant judged could only sour between two ambitious men.^[5]

Few figures of the period are called so many things, and the tangle is more than a curiosity of transmission: it is part of the historical problem. The Old Persian of Behistun names him Bardiya; the Elamite version writes Pirtiya, and the Babylonian version and the dated business tablets give the Median form Barziya.^[1][6] The philology is now settled. Jan Tavernier treats Old Persian Bardiya as the Persian equivalent of Median *Barziya, a *-ya- extension of a stem *brzi-, "high", the Old Persian -rd- being the native development of the Median -rz-, so that the prince's own name meant something like "the exalted one".^[6] The Greeks and their successors multiplied the forms further: Smerdis in Herodotus, Mardos in Aeschylus, Mergis in Justin, Merphis in Hellanicus. Ctesias alone calls Cyrus's younger son Tanyoxarces, and Xenophon Tanaoxares.^[1] Dandamayev reads these last as Old Persian *tanu-wazraka, "large-bodied", and notes that Bardiya itself, "high" or "exalted", carries the same suggestion of exceptional physical size, so that Tanyoxarces looks like a byname of the one prince rather than the name of a second.^[1][4]

The proliferation is itself a measure of how far the man had dissolved into competing stories within a century. A figure known under half a dozen names in mutually contradictory accounts is a figure whose very identity the tradition had failed to fix, and the confusion runs deep enough that even in antiquity the same person could be split in two: Hellanicus gave Cambyses two brothers, Maraphis and Merphis, which Dandamayev takes for two garbled variants of the single name Bardiya.^[4] The onomastic scatter is the first sign of the problem the whole subject poses, that the sources cannot agree on who the man was even at the level of what to call him.

The sources agree that Bardiya died at his brother's hand, and disagree about everything else concerning the death. Behistun places it before Cambyses left for Egypt in 525, done in secret so that the people did not know the prince was dead.^[2] Herodotus places it during the Egyptian campaign and supplies a motive and a mechanism. In his telling Cambyses grew jealous of Bardiya, whom he calls Smerdis, because Smerdis alone among the Persians could draw, if only a little, the great bow sent as a challenge by the king of the Ethiopians; Cambyses sent him home to Persia, then dreamed that his brother sat on the royal throne, and despatched his trusted Prexaspes to kill him.^[3] Herodotus offers two variants of the murder itself, that Smerdis was killed on a hunt near Susa, or drowned in the Erythraean Sea.^[3] Ctesias tells it differently again: a Magian named Sphendadates, flogged by Bardiya for some offence, denounced the prince to Cambyses, who had him put to death by making him drink bull's blood, and dated the killing some five years before the usurpation.^[7] Justin puts the death after Cambyses's own, not before.^[5]

Already in antiquity, then, the sources contradicted one another on the time, the place and the manner of the killing, and that disagreement is one of the oldest arguments against the official version: a secret so well kept that its own reporters cannot agree on when or how it happened is a secret hard to credit. What Herodotus's bow story does convey, whatever its historical worth, is a relationship between the brothers under strain: Briant read the anecdote as a metaphor, in a culture that measured a man by his prowess as a warrior and bowman, for Bardiya's own aspiration to the throne his brother held.^[5] The killing itself, whether it happened as Darius says or at all, belongs to the case against Cambyses; it is also the central claim on which the whole official story rests, and the one that no source can confirm.

Darius's account is the earliest and the fullest, carved high on the Bisotun cliff a few years after the crisis it describes, in three languages at once. It turns the concealed murder into the opening for a fraud. While Cambyses was away in Egypt, Darius says, the empire sickened and a Magian rose to exploit the secret:

The people became hostile, and the Lie multiplied in the land, even in Persia and Media, and in the other provinces.

Darius I, DB §10^[2]

The Magian was Gaumāta, who rose on 11 March 522 from a place called Paisiyauvada, by a mountain named Arakadri, and claimed to be the dead prince. Darius marks him with the one accusation he returns to against every pretender of that year:

He lied; he said: I am Bardiya, the son of Cyrus.

Darius I, DB §11^[2]

The people went over to him, Darius records, both Persia and Media and the other provinces; Gaumāta seized the kingship on 1 July 522, and Cambyses died soon after. On this account the impostor held power by terror, killing any who might expose him, and Darius makes him say as much, that he slew people "lest they know me, that I am not Bardiya the son of Cyrus".^[2] On 29 September 522, Darius says, he and six noble Persians killed Gaumāta at the fortress Sikayauvatiš, in the Median district of Nisaya, and by the grace of Ahura Mazdā he took the throne and restored the sanctuaries the usurper had destroyed and the property he had seized.^[2] The name Gaumāta belongs to this official version alone: it is found only in the inscriptions and, in the form Cometes, in Justin.^[8] Darius's handling of the man's name points the same way. Of the other rebels of the year the inscription gives extensive information about descent and origin, while the most dangerous adversary of all is named throughout only "Gaumāta the Magian", without father or house. Dandamayev takes the omission as deliberate rather than accidental, a mark of contempt, and it would follow just as well if the man had no separate identity to give.^[4] The killing itself, and the coup of the six who became the Seven, belong to the accession of Darius and the Seven; the man they struck down is the subject here.

Whoever he was, the king called Bardiya reigned, and left a mark on the record independent of Darius. In Babylonia, where every legal contract was dated to the regnal year of whoever held the throne, documents begin in April 522 to name a king Barziya, styled king of Babylon and king of the lands; the invocation formula of one contract calls upon "Bel, Nabu and Barziya, the king of Babylon, and the king of the lands".^[4] This is the one external, non-propagandistic evidence for the reign. It fixes that such a king was recognised across Babylonia within a month of his rising, and honoured with the full royal titulary and the divine invocation; it cannot say who he was, and the clay is silent on the single point in dispute.^[9] The reign was remembered as a generous one. Herodotus reports that the king remitted, across the whole empire, three years of tribute and military service, and that the peoples mourned him:

In this time he greatly benefited all his subjects, in so much that after his death all the Asiatics except the Persians wished him back; for he sent hither and thither to every nation of his dominions and proclaimed them for three years freed from service in arms and from tribute.

Herodotus 3.67, on the reign of the false Smerdis, trans. Godley^[3]

The remission can be read two ways, and the reading depends on which theory of the king one holds. As the act of an insecure impostor it looks like a bid to buy loyalty; as the policy of a legitimate king with a base in the provinces it looks like ordinary statecraft. The mourning Herodotus reports, from subjects who had nothing to gain by praising a dead fraud, sits awkwardly with the official picture of a hated usurper. Briant took the confiscations Darius denounces at Behistun not as an attack on the aristocracy at large but as a move against the specific noble houses conspiring to remove the king, their conditional land-grants revoked and redistributed to loyal supporters, which is the conduct of a king defending a throne, not of a magus terrorising the realm.^[5] On this reading the one non-Persian record of the reign, the Babylonian clay, and the one Greek notice of its policy both describe an ordinary king going about the ordinary business of rule, and it is only Darius's monument that turns that king into a monster.

The potency of the dead prince's name outlived his reign, and did so in a way that bears directly on the identity question. When the empire rose against Darius in the year that followed, one of the most dangerous revolts was raised in Persia itself, the Achaemenid heartland, by a man named Vahyazdāta. He too took up the name of Bardiya son of Cyrus, so that Darius's record carries not one false Bardiya but two.^[2] Vahyazdāta gathered an army, held out through two pitched battles, and was captured and impaled with his leading followers, having drawn support in the very country where the true line was best known.

That a pretender could rally the Persian homeland under that name, months after the first Bardiya's death, is a fact both theories of the affair must accommodate, and it points in two directions at once. If the name was worth impersonating twice within a single year, and worth impersonating in the one region where a fraud was least likely to pass, that argues the name of a genuinely legitimate prince whose survival people were ready to believe. But it argues, equally, that "Bardiya son of Cyrus" had become a free-floating token of legitimacy that any bold claimant might seize, which is precisely how the first Gaumāta could have done the same. Serial impersonation cannot decide the case; what it shows past doubt is how much political value the name carried, and why Darius had so strong a stake in fixing who could lay claim to it. A king who has killed one man for using a name, and must then kill a second for the same offence, has an interest in fixing forever that the name was never anyone's to use.

The classical accounts share the broad shape of Darius's story and diverge in the details, and the divergences are as instructive as the agreements. Herodotus makes the usurpation the work of two Magian brothers: Patizeithes, left by Cambyses in charge of the palace, and his brother, who both resembled the murdered Smerdis and happened to bear the same name, and whom Patizeithes set on the throne.^[3] Ctesias replaces the two Magi with the single Sphendadates and the courtiers who back him, and makes the resemblance so exact that those closest to the prince could not tell the two men apart.^[7] The variety runs deep enough that Briant judged the classical writers had mingled several oral versions in circulation among the Persian nobility and in Asia Minor, none of them an independent check on Darius.^[5]

The most awkward witness for the official version is the earliest Greek one. Aeschylus, writing the Persians in 472, only fifty years after the events, put a roll of Persian kings in the mouth of the ghost of Darius, and in it the man Darius killed appears not as an impostor but as a legitimate monarch, fifth in the line:

Fifth in the list, Mardus came to power, a disgrace to his native land and to the ancient throne; but he was slain in his palace by the guile of gallant Artaphrenes, with the help of friends whose part this was.

Aeschylus, Persae 774–776, the ghost of Darius names the kings, trans. Smyth^[10]

Mardos is a disgrace, but he is a king, holding "the ancient throne" in due succession, and he is killed not by Darius but by one of the conspirators. That the first Greek treatment of the affair remembered the man as a monarch rather than a magus, and put the memory in Darius's own mouth, is a thread the sceptical case has always pulled hard.^[11][4]

Two women stand at the edges of the story, and each turns the identity question in a different direction. The first is Phaidyme, daughter of Otanes, who had been a wife of Cambyses and passed, as royal wives did to a successor, into the household of the man ruling as Bardiya. Herodotus makes her the instrument of the fraud's detection. Otanes, suspecting the king to be an impostor, sends word to his daughter to test him as he sleeps, for it was known that Cyrus had once cropped the ears of a certain Magian named Smerdis as a punishment:

do as I bid you and feel for his ears; if you see that he has ears, then you may think that it is Smerdis son of Cyrus who is your lord; but if he has none, it is Smerdis the Magian.

Herodotus 3.69, trans. Godley^[3]

Phaidyme finds the ears gone and the fraud is exposed. As history the scene is worth little: the cropped ears, the sleeping test, the near-perfect double are the furniture of folktale, the kind of recognition motif that attaches itself to any story of a hidden impostor. As evidence it cuts the other way from its intent. The sceptics have long pressed the difficulty the story is built to answer, that the royal women, Cambyses's own sister Atossa and his wives among them, had been in the palace from the start of the new reign and had every chance to know whether their lord was the prince they had lived beside; Dandamayev notes that the harem was plainly not the sealed place Herodotus's source needs it to be, since Phaidyme herself communicates freely with the outside, and that a genuine impostor should have been unmasked at once by the people who knew the real Bardiya best.^[4] The Phaidyme episode is thus less a proof of the imposture than an admission of how hard the imposture was to believe.

The second woman is Parmys, and Darius put her to use. Herodotus lists among the new king's marriages, beside Cyrus's daughters Atossa and Artystone, a marriage to Parmys, daughter of Bardiya son of Cyrus.^[3] On the sceptical reading these unions are the tell of a usurper knitting himself into the family whose throne he had taken, marrying the daughters and the niece of the line he had displaced to father heirs of unimpeachable descent. Olmstead counted the marriage to the dead prince's own daughter among the legitimising matches Darius needed.^[11]

The doubt begins from Darius's own place in the dynasty. He belonged to a collateral branch: his father Hystaspes was a satrap, not a king, and the crown had descended from Cyrus to Cambyses without ever passing through Hystaspes's line. A man so placed needed exactly the account Behistun supplies, that Cyrus's real son had died in secret long before, leaving the throne of 522 in the hands of a fraud whose killing was not regicide but a rescue. The suspicion has a pedigree almost as long as the modern study of the subject.

The maximal sceptical case was first put by Paul Rost in 1897 and Hugo Winckler in 1898, and though it found little favour at first the doubters grew steadily more numerous through the twentieth century.^[12] Its classic English statement is A. T. Olmstead's, in a chapter pointedly titled "Usurper Darius". Olmstead pressed every seam in the official story: the sources disagree flatly on the time, place and manner of the alleged secret murder; the recantation and suicide of Prexaspes are "frequent devices of the propagandist"; the impostor's resemblance was supposedly so perfect that even the prince's mother and sisters were deceived, which he called the height of absurdity; Aeschylus remembered Mardos as a legitimate king; the run of legitimising marriages Darius made, to Cyrus's daughters and to Bardiya's own daughter Parmys, are what a usurper needs; and Darius so insists on his enemies' lying that "we are convinced 'he doth protest too much'".^[11] Olmstead drove the point to its edge by noting that the "false" Smerdis was false only in claiming Cyrus for a father: "his actual name was Smerdis!"^[11] Muhammad Dandamayev developed the argument at book length in his political history, and pressed the practical difficulty hardest: a prominent satrap, son of Cyrus, could not plausibly have vanished for four or more years while his mother, sisters, daughter, servants and friends were told nothing and suspected nothing, only to learn the truth at last from the man who had just killed his successor and seized the crown. Dandamayev's verdict there is blunt: "All circumstances which are described above force us to doubt the truth of Darius' statement that he killed a false Smerdis", and he concludes that Bardiya, Mardos, Smerdis, Tanyoxarces and the Gaumāta of the inscription were most likely one and the same man, the younger son of Cyrus.^[4]

The reference articles are more cautious, and one scholar is cautious in two voices. Dandamayev, writing the Bardiya article for the Encyclopaedia Iranica, lays the sceptical thesis out fairly and then declines to endorse it:

Some modern scholars (e.g., Olmstead) believe that the man who revolted against Cambyses was his true brother and lawful heir and that Darius killed him, calling him Gaumāta and inventing the story of the false Bardiya in order to justify his own seizure of the kingship.

Dandamayev, Bardiya, Encyclopaedia Iranica^[1]

His published verdict there is that "this must remain hypothetical", the same scholar holding back in the reference article the conclusion he had argued for in his monograph, a fair measure of how far the evidence permits certainty in either direction.^[1] Briant, in the companion article on Gaumāta and at length in From Cyrus to Alexander, goes further toward the sceptics on the narrow question and then away from them on the wider one. He takes it as evident that Darius had every reason "to invent a character (Gaumāta) condemned for his acts against gods and men", and speaks plainly of "the pseudo-Gaumāta invented by Darius".^[8] But he rejects the older romance that grew up around the sceptical thesis, which made the genuine Bardiya a social reformer and his overthrow an aristocratic restoration. That reading, Briant showed, rests on leaning too hard on one obscure paragraph of the inscription (taking the Old Persian kāra, which means the people or the host at large, to signify "the nobility") and on trusting Herodotus's charter of privileges for the Seven, whereas the great houses in truth got no such grant and the image of a king first among seven mighty families is "the two sides of the same historiographical myth".^[8] On Briant's reconstruction the genuine Bardiya drew on the Persian nobility just as his father and brother had done, so that his months on the throne marked no break in the character of the dynasty; with the childless Cambyses dead, "Bardiya was the sole male descendant of the royal family", a king entitled to rule by birth alone.^[5] Darius, Briant argued, "was motivated to date Bardiya's official assumption of power before the death of Cambyses", so as "to transform into a usurper a king who could legitimate his authority by his filiation".^[5] Seen this way the coup was nothing rarer than the seizure of a throne by a kinsman, the sort of thing that happens repeatedly in the dynasty's history, "rather banal" except that this time the victor cut his account into a cliff.^[8]

Recent scholarship divides along these lines without resolving them. The 2021 Companion sets the two possibilities squarely side by side, that the official version is true, or that "the false Smerdis is an invention by Darius to motivate his usurpation" and "Darius in fact had killed the real Smerdis, the legitimate heir", and reports that the balance of opinion has moved toward doubt.^[12] None of this, though, amounts to proof in either direction, and that is the settled state of the question. Nothing outside Darius's own texts speaks to whether the magus ever existed; the clay from Babylon establishes a reigning Barziya and nothing about his person; Herodotus and Ctesias wrote generations later, and Herodotus most likely inherited a version already shaped by Darius's broadcast, so his agreement lends weight to the official story without independently testing it. In Darius's favour there is only this: his is the sole eyewitness testimony to survive, nothing in it collapses under its own weight, and even a man who stood to gain by lying is not for that reason lying. Certainty is available to neither side. The evidence that would break the deadlock, a mention of Gaumāta the Magian in some text Darius did not compose, or a contemporary word on who the Babylonians' Barziya really was, is exactly what the record withholds and will most likely never surrender. Recent opinion inclines toward mistrust, the doubting voices having grown more numerous, though views still diverge widely and no reading commands assent; the man of 522 cannot be identified on the evidence that survives, and Behistun has to be read for what it claims and for the interest its author had in the claim together.

Whatever the truth of his identity, the dead prince served Darius's legitimacy even in death, through the marriages already noted and through the way the killing was remembered. Herodotus records that the Persians kept the anniversary of the deed as a festival, the Magophonia or "slaughter of the Magi", on which no Magian might appear in public, a memory of the affair as the purge of a Magian usurpation.^[3] The festival preserved Darius's version in the calendar itself, an annual public re-enactment of the claim that what had fallen in 522 was a magus and not a king.

Gaumāta himself, if he existed, survives as the figure carved beneath Darius's foot on the Behistun relief, the small prostrate man over whom the king stands with his bow. Briant's summing-up catches the paradox of a figure who may be an invention and is certainly a monument: Gaumāta "has become a historical figure only because on the rock in Bīsotūn and in the countries of his empire Darius tampered with timeless memories".^[8] The man Darius killed, whether prince or magus, is remembered chiefly because the man who killed him needed him remembered a particular way.