Artaxerxes IV (Arses)

The king's personal name was Arses, a Greek rendering of an Old Persian short-name formed on aršan-, "male, manly", a stem well attested in Persian compound names.^[2] Babylonian documents write the name Aršu (Arshu). No text in his own voice survives in any language, so the forms of his name are known entirely from documents written about him or dated by him.

His throne-name is an inference, though a strong one. A fragmentary Babylonian tablet, BM 36613, preserves the equation directly: "of Arshu, the son of Umasu … who [is called] Artakshats[u", that is, Arses son of Ochus (Umasu = Artaxerxes III), called Artaxerxes.^[3] The same broken text later names "[Aleksan]darris, the great king", so the surviving copy looks back from at least Alexander's time, and Kuhrt cautions that its nature is "completely uncertain": perhaps a chronicle fragment, perhaps a later historical letter.^[3] It is strong evidence for the throne-name and weak evidence for anything else. Rüdiger Schmitt accordingly keeps a question mark in the dynastic sequence, listing the late kings by birth-name as "Ochus–Arses–Ochus–Arses" corresponding to "Darius II, Artaxerxes II and III, and Arses [= Artaxerxes IV?]".^[4] Briant states the prevailing view: "In all probability this king", like his predecessors, "took the reign name" Artaxerxes, the fourth of the line.^[5]

The repetition of birth-names across the dynasty's last century is itself part of the evidence. Arses bore the name his grandfather Artaxerxes II had carried before accession, in the variant forms Arsaces or Oarses; his father Ochus repeated the birth-name of Darius II (Babylonian Umakush).^[4]

Few Achaemenid names invite confusion so readily, and the scholarship polices the boundaries with care. Eran Almagor notes that "Some of the MSS of the Artaxerxes (GR) have the form ὁ Ἄρσης for the alternative of Artaxerxes' given name", and that the reading "may have been the original reading of Plutarch's text, corresponding to the name of Artaxerxes II's grandchild, the Persian king Arses (338/7–336/5 BCE)", also found "in the Babylonian astronomical documents" as Ar-šu.^[6] An "Arses" in a discussion of Artaxerxes II may therefore be a manuscript variant of that king's own birth-name or a reference to his grandson. Distinct again are Arsanes (Oarsanes), the father of Darius III; Arsites, the satrap of Hellespontine Phrygia in 334; and the Parthian dynastic name Arsaces. None of these men is Artaxerxes IV.

A possible Egyptian attestation shows how thin the direct record is. Jan Tavernier discusses a demotic writing I-r-s-s-s that could transcribe Arses, but sets out the objections, among them that "Arses is the king's real name and not his throne name. One would expect the throne name (Artaxsaca-) to be written here", and concludes that "A final solution of this problem cannot yet be offered."^[7] The spelling may belong to a private individual, or to the prince before accession; it cannot be listed as a secure royal attestation.

The one contemporary notice of the succession is Babylonian. A tablet concerned with a solar eclipse, BM 71537, records that in "Month Ululu" Umakush, that is Ochus, "(went to his) fate", and that Arshu, "his son, sat on the throne".^[8] Month VI of that year ran from 26 August to 25 September 338, which fixes the accession within a month.^[8] The laconic line carries a further implication. "Went to his fate" is the conventional Akkadian formula for an ordinary death, in contrast to the explicit murder notice the Babylonian tradition gives for Xerxes; Kuhrt reads it as evidence that Artaxerxes III died naturally, against the Greek story of poisoning.^[9] The tablet cannot logically exclude a concealed murder, and the death of the father remains disputed (see Artaxerxes III); what it does establish is that the succession itself passed to a son, in the ordinary formula, with no Babylonian hint of crisis.

The Greek tradition tells a different accession. In Diodorus, Bagoas the chiliarch, having "killed Ochus with poison", "promoted Arses, the king's youngest son" to the throne, and secured his creature by a further crime:

He also killed the king's brothers, who had just reached adulthood, in order to isolate the youth and make him easier to control. But the lad was angry at these outrages and made it clear that he would punish the perpetrator of the crimes

Diodorus Siculus 17.5.4–5, trans. Kuhrt^[10]

The late reference tradition adds that his mother was Atossa, a daughter of Artaxerxes II.^[2] Nothing else of his life before the throne is recorded: no age, no earlier office, no marriage. That a "youngest son" was raised over his elders is itself part of the hostile narrative of manipulation, and cannot be checked.

How long Arses reigned was long answered from the Greek narrative alone: roughly two years, ending in the summer of 336, a chronology stated in classic form by Olmstead as "less than two full years (November, 338, to June, 336)".^[11] The dated documents have overturned the endpoint. An Aramaic slave-sale papyrus from Samaria, the best preserved of those found in the Wadi Daliyeh caves, opens: "The twentieth day of the month of Adar, second year, accession year of Darius the king".^[12] The composite formula equates Arses's year 2 with the accession year of Darius III, and so dates itself, in Briant's words, "to the very moment of the transition" between the two reigns; he converts the day to 19 March 335, Kuhrt to 20 March, a one-day difference that belongs to the modern conversions and should be left standing.^[12][13]

A group of Idumean barley chits may narrow the death further. One ostracon is dated "30 Shebat, 2 [Arses]"; the king's name is an editorial restoration resting on scribal hands and prosopography, and the chits are unprovenanced market acquisitions, which Porten and Yardeni report reached them "as a group from the collector Shlomo Moussaieff".^[14][15] If the attribution holds, the latest chit and the Samaria papyrus bracket the change of reign: "Arses would have died between February 27 and March 19, 335."^[15]

Diodorus places the murder "in his third year of reign",^[10] and the late chronographic tradition rounds the reign to three years: Syncellus's tables, in Cory's old collection, give "Arses, or Arses Ochus, (or Artaxerxes), reigned / 3 years".^[16] The counts are compatible with the Babylonian year 2. In Babylonian reckoning the months from his accession to the next new year (to 18 April 337) counted as an accession year, with year 2 running to spring 335; a Greek writer counting inclusively from 338 reaches a "third year" over the same span.^[13] The reign is therefore best stated as 338 to 335, from an accession between late August and late September 338 to a documented transition in March 335, with the conventional "338–336" surviving in older literature.

One monument may show the empire's ordinary machinery at work under Arses. The trilingual stele from the Letoon sanctuary near Xanthos in Lycia records, in Lycian, Greek and Aramaic, the establishment of a local cult; the Aramaic summary is dated:

In the month Siwan of year one of Artaxerxes the king, in the fortress of Orna … Pixodaro, son of Katomno, (being) the satrap who (governs) in Caria and (in) Lycia.

The Xanthos trilingual, Aramaic text, trans. Kuhrt^[17]

Which Artaxerxes had a "year one" with Pixodarus in office? Since Badian's study of 1977 the prevailing answer has been Artaxerxes IV, an attribution Kuhrt calls "virtually certain";^[18] on that dating the text belongs to the early summer of 337 and is, as Briant puts it, agreed to be the first official mention of the king. It gives the reign what Briant calls an unexpected administrative reality: "the life of the empire continued, even during a period that, to believe the classical sources, was entirely taken up with sordid and bloody palace conspiracies".^[5] The attribution is influential rather than settled. Christian Marek states the dissent plainly: "Although widely accepted, there are strong arguments against this", and the rival dating to year one of Artaxerxes III (359) remains defended, at the cost of an otherwise unexplained early command, or a co-regency, for Pixodarus.^[19] The Aramaic is also only a short official endorsement of a longer Lycian enactment (for the elastic reach of the satrap's title in this region, see the Satrapy System); it does not show that every local provision required royal approval.

A fourth-century Aramaic document from Bactria, C1 in the Khalili collection, opens: "In the month of Kislev, year 1 of Artax[erxes] the King", and records "Provisions in Maithanaka for Ba[yasa], when … he passed from Bactra to Varnu".^[20] Which Artaxerxes is again the question. The first editors, Naveh and Shaked, reasoned that "As Arses was named Artaxerxes [IV] … Bessus, who adopted the name Artaxerxes, should be regarded as Artaxerxes V", and assigned the document to Bessus's rebel kingship in 330/329.^[21] Shaked has since changed his mind: as Hyland reports, he now "prefers to identify the King as Artaxerxes IV and the year as 338", which would make C1 a record of the eastern administration provisioning an official's journey in Arses's first winter.^[22] The dossier's provenance imposes caution of its own; the editors state that "The provenance of these documents … is not known. They were acquired from several dealers in London and elsewhere around the year 2000."^[23] C1 is a possible witness to the reign, and cannot be called a secure one.

The only connected account of the reign's end is Diodorus's. The young king, angered by the murder of his brothers, moved against his maker, "so Bagoas anticipated his plan and killed Arses together with his children".^[10] With that, Diodorus declares, "The royal house was now bereft", and Bagoas set up Darius, a man of a collateral line.^[10] The claim of extinction is false in the letter: at least one son of Artaxerxes III, Bisthanes, was alive to meet Alexander in 330, and Darius III himself was an Achaemenid of a cadet branch, a grandson of a brother of Artaxerxes II and so a kinsman of the king he replaced.^[24] The exaggeration serves the story's shape, in which a monstrous servant empties the royal house and hands the crown to his last creature.

A further version made the successor the murderer. In the letter Arrian gives Alexander before Issus, Darius is charged with taking the crown "after slaying Arses, as well as Bagoas" and "unjustly seizing the throne" against "the law of the Persians".^[25] The letter belongs to a purported exchange that scholarship treats as spurious composition rather than preserved diplomacy,^[26] and its accusation is evidence for what Alexander's side found useful to allege, that Darius III's title was tainted, rather than for the deed. The charge shows how Alexander's literary case represented Darius's accession; it does not establish how Arses died.

Modern criticism has narrowed what the Bagoas narrative can prove. Kuhrt observes that "Making Bagoas the protagonist of these murders almost certainly belongs to the popular topos of the 'treacherous eunuch'", and that the tale "should be taken with a pinch of salt".^[27] Binder draws the surviving line: "The historicity of Bagoas as a powerful figure at court cannot be doubted", yet "That he alone, however, staged the elevation of kings at his discretion is inconceivable."^[28] A real and powerful chiliarch operating within wider court factions is recoverable; the lone kingmaker with his poisoned cups is a literary construction, and with him goes any secure knowledge of what Arses did or intended in his own court. Even the shape of the tradition is unstable: where Diodorus builds the succession around Bagoas's crimes, Justin's epitome passes over the reign entirely and has Darius chosen "because of his outstanding bravery", two accounts of the same transition that cannot both be describing a remembered court politics.^[27]

The stakes of the criticism reach past one king's reputation. The puppet-years of Arses long served as a set piece of the decadence narrative, the empire rotting at its centre while Macedon rose, and it is this use of the court stories that the current scholarship most firmly rejects: for Llewellyn-Jones, "To read these stories … as evidence for the decay of the Empire … is not only futile but completely wrong." His verdict on the period is the honest summary of what the evidence permits: "All that can be said with any certainty is that evidence for the period is both confusing and frustrating."^[29]

The kingmaker must also be kept apart from his namesakes. The chiliarch of the 330s is a different man from Bagohi (Bagavahya), the governor of Judah addressed by the Jewish garrison of Elephantine in the late fifth century, and from the young Bagoas who was later Alexander's favourite. The name was common; the men were not connected.

A damaged Babylonian text has been used to support the murder narrative, but it is not independent confirmation. The so-called Dynastic Prophecy, a Hellenistic-era composition that recounts past successions in the future tense, reaches a ruler identified as Arses:

For two (?) years [he will exercise the kingship]. That king a eunuch [will murder]. Any prince […] will attack and [seize] the thro[ne].

The Dynastic Prophecy, iii.2–8, trans. Kuhrt^[30]

The genre is what Boiy calls "divinatio ex eventu", prophecy written after the event.^[31] Only one upright of the opening numeral survives; restoring "two" depends on identifying the next ruler as Darius III. More decisively, the verb translated "[will murder]" is restored on the assumption that Diodorus's Bagoas story was widely known.^[30] The passage can therefore be aligned with Arses and Darius III, whose five Babylonian regnal years fit the following number, but it cannot independently confirm either a two-year figure or a violent death.^[32] Its phrase "any prince" may suggest uncertainty about Darius III's claim, but the broken text cannot supply a court narrative.

Arses's reign coincided with the decisive turn in the west: Philip II's victory at Chaeronea in 338, the formation of the League of Corinth, and the launching of the war against Persia. What the Persian court did about it under Arses is unrecorded. Arrian's Alexander-letter asserts a grievance, "when Arses, son of Artaxerxes, ascended the throne, Philip was the first to practise" injustice "towards him, though he had suffered no injury from the Persians",^[25] but the offence is unnamed and the letter is the same suspect composition as the regicide charge. The letter's own framing is nonetheless revealing: it concedes that "friendship and alliance had subsisted between Philip and Artaxerxes" the father, so that the case for war had to turn on a wrong done to the short-lived son.^[25] Whatever passed between the courts in 338–336 was remembered, on the Macedonian side, as the moment the relationship broke. The older reference literature filled the gap by reconstruction: Lecoq's Iranica article has Philip demand compensation for Persian aid to Perinthus and Arses refuse.^[2] The refusal is nowhere attested; the aid to Perinthus in 340/339 belongs to Artaxerxes III's reign.^[33]

Waters states the honest position: for all the Greek material charting Philip's activities, "we cannot track Persian reactions" to his rise, closely though the court will have watched him.^[33] What can be dated is Macedonian action and its check. Philip sent an advance force under Parmenion and Attalus across the Hellespont in 336; during 336 and 335 the Persian counter-effort associated with Memnon of Rhodes recovered ground in western Asia Minor.^[33] Whether any of this reflects decisions taken by Arses, by his court, or only by the western satraps cannot be said, and the attribution of particular measures to the king personally is not possible.

At some point in these decades a native claimant, Khababash, ruled in Egypt long enough to be remembered as king. The chief evidence, the Ptolemaic Satrap Stela, recalls him confronting a hostile Persian king it calls "Xerxes", a name no fourth-century king bore. Briant records the attempts "to place him around 342–338, or between 338 and 336", and concludes that the revolt fell "at some date that we cannot establish (Artaxerxes III, Artaxerxes IV, or Darius III)"; beyond that, "we can say nothing about the magnitude of the revolt".^[34] He restates the deadlock in the later book: "there are no truly solid grounds for choosing one of the three dates over another."^[35] Colburn adds the caution that governs the stela itself: "The Satrap Stela is thus not a straightforward historical text"; "it makes ideological claims that are meant to depict Ptolemy as the successor to Alexander and a restorer of order in Egypt."^[36] No secure reconstruction makes Arses either the king Khababash expelled or the king who restored Persian control, and a reign already stripped to documents cannot carry an Egyptian war on this evidence.

A single line in a single text raises the possibility of trouble in Babylon at the reign's end. The Hellenistic king-list from Uruk, in its damaged reverse, appears to interpose a ruler between Arses and Darius III, one whose "other name is Nidin-B[el?]".^[37] Boiy sets the stakes: "After 482 BC no rebellion against the Persian authority is reported. … Only if Nidin-Bel from the Uruk king list really was a Babylonian usurper, there was one rebellion" in the late Achaemenid period.^[38] Tolini, who entertains a local claimant seizing recognition in the months after Arses's death, concedes the isolation of the witness: "À ce jour, seule la Liste royale d'Uruk le mentionne" (to date, only the Uruk royal list mentions him).^[39] The rival reading, favoured by Kuhrt and Boiy, sees a scribal displacement: Nidintu-Bel was the Babylonian rebel of 522 against Darius I, and a copyist may have slipped the famous name in before the wrong Darius (for the genuine rebellions of 522–521, see the accession of Darius).^[38] No dated tablet recognises any such king, and a Babylonian revolt in 335 remains a debated reading of one king-list entry, no more.

Nothing royal survives that Arses made. No building inscription, no relief, no tomb and no coinage can be securely assigned to him. Briant notes that "neither Arses/Artaxerxes IV nor Darius III seems to have been supplied with an individual sepulchre", and allows that "the short duration of Artaxerxes IV's reign and the tragic conditions surrounding his physical elimination by Bagoas may constitute plausible (though not fully convincing) explanations".^[40] The unfinished royal tombs at Persepolis have attracted attributions to the last kings, but none is demonstrable.

The silence asks for interpretation and forbids most of it. Read one way, it is the material trace of obliteration after a short reign whose end later writers made violent; the contemporary record does not state the cause of death. Justin's silence has even prompted the suggestion of a formal damnatio memoriae, which Kuhrt weighs without endorsing, observing that the Babylonian record of his ordinary accession argues against a deliberate erasure.^[27] Read another way, it is what two years of any reign would leave if the reign ended without a tomb: the dated documents from Samaria, Idumea, Lycia and perhaps Bactria show routine administration proceeding, and argue against turning the missing monuments into evidence of paralysis.^[5] The limits of the material record are specific. No tomb or royal inscription can be assigned to Arses, while his probable throne-name survives only in a broken later copy.^[3] Dated administrative texts establish accession and continued government in some regions; Diodorus supplies the murder narrative, and Arrian's literary letter turns the dead predecessor into an accusation against Darius III. None of those later uses confirms the cause of Arses's death.