Sources & Historiography

Ctesias is the classical source with the greatest apparent authority and the least demonstrated reliability, and both halves of that verdict must be held together. AUTHORITY: he lived for years at the court of Artaxerxes II as the King's physician (after c. 405 BCE; Diodorus' 'seventeen years' is probably inflated, Schmitt), witnessed the battle of Cunaxa in 401 and healed the King's wound there (Xenophon, Anabasis 1.8.26, naming Ctesias as his own source for the Persian side), enjoyed the confidence of the queen-mother Parysatis, and claimed to have used the 'royal parchments' (basilikai diphtherai), so that on the court he is the one insider among outsiders. UNRELIABILITY: that claim to the archive will not bear weight: a second Diodorus passage (2.22.5) gives the 'royal records' at second hand, 'according to what the barbarians say,' and it is doubtful he read any documents himself (Schmitt); his chronology is badly confused (he passed from Thermopylae straight to Plataea and shunted Salamis, out of order, into an appendix, errors that are his own, not Photius', Bigwood 1978 in Schmitt); and where the Iranian evidence controls him he is wrong against it: his list of Darius' six fellow-conspirators (Onaphas, Idernes, Norondabates, Mardonios, Barissos, Ataphernes) does not match the Behistun roster (DB 4.83–86), whereas Herodotus' list does (Schmitt). His whole treatment is disproportioned: the reign of Darius I and Xerxes, the age of the Persian Wars, got only two of twenty-three books, while four or five went to the eight years he had seen and five to the already-legendary Cyrus. BIAS: he wrote for Greek readers, in open and often unjust polemic against Herodotus; Plutarch called him philolakōn, a friend of the Spartans, and the Persica shows a pro-Spartan colour (Schmitt); and his taste ran to sensation: harem intrigue, poison, eunuchs, sentimental romance, closer to Hellenistic popular literature than to Herodotus or Thucydides, so that he is fairly called a father of the historical romance (Jacoby, in Schmitt). The transmission compounds every fault: the Persica is lost, and later writers preferred to quote his most incredible stories, so the surviving Ctesias is a residue selected for strangeness (Schmitt). GOOD FOR: the interior life of the late Achaemenid court, above all under Artaxerxes II: the royal women (Parysatis pre-eminent), the eunuchs, the ceremonial of the King's table, palace faction and intrigue, matter largely absent from Herodotus; the reign of Artaxerxes II and the events around Cunaxa (as a check on and complement to Xenophon and behind Plutarch's Artaxerxes); and, at one remove, evidence for how the Persians of c. 400 BCE imagined their own Assyrian and Median past. BAD FOR: chronology and the sequence of events of any kind; the Persian Wars (all but untold, and garbled where told); Assyrian and Median history (largely myth: Ninus, Semiramis, Behistun refathered on a legendary queen); and any single fact taken on his word where no Iranian or independent control exists. THE RULE: prefer the Iranian evidence and Herodotus wherever they bear on the same matter; use Ctesias for court-texture and for the reign of Artaxerxes II, always corroborated, never alone; and treat his palace melodrama as evidence for the court's own tales and for the Greek image of Persia at least as much as for hard fact. The proverbial judgement stands with a caveat: his unreliability is real enough to make Herodotus look accurate by comparison, but the modern reassessment (flagged below) is right that he is a witness, not merely a liar.

For the Achaemenid empire Herodotus is the single richest classical source and, for long stretches, the only continuous narrative we have; but he is an outsider's source and must be used as one. RELIABILITY divides sharply by kind of matter. The institutional and factual skeleton is often sound and repeatedly confirmed by the Iranian evidence: the Cyrus Cylinder bore out his same-named grandfather of Cyrus; the Persepolis and Babylonian tablets answer to his Magi, his open sacrifice, his provisioned Royal Road and his fiscal geography; and his account of Darius's accession runs strikingly close to the Behistun inscription, so close that he probably knew a circulated version of that royal text (Rollinger, Encyclopaedia Iranica). Against this, the STORIES built over the facts, the motives, the court intrigues, the set speeches (the Constitutional Debate of 3.80-82), the moralising on hubris, are heavily shaped by Greek narrative and dramatic convention, and the famous individual portraits (a vain, impious Xerxes) are Greek types at odds with the inscriptions. His known GROSS ERRORS are the army numbers (over 1.5 million fighting men for Xerxes, a physical impossibility; modern estimates run to a few hundred thousand at most), the confusions that follow from his interpretatio graeca (rendering Ahura Mazdā as 'Zeus'; the outright error at 1.131 that makes 'Mitra' a goddess of love, most likely a Mithra/Anāhitā confusion, Rollinger), and the Greek-drama shape of court-scenes he could not have witnessed. BIAS: he is a sympathetic outsider (ancient critics called him philobarbaros), writing from the western fringe of the empire in the generation after the events, in Greek and for Greeks, and cannot read the Iranian languages, so the inner life of the court reaches him at second hand through Greek and Hellenized informants and through the traditions of the Greek cities of Asia. His SOURCE-CITATIONS are themselves contested: honest eyewitness notes (Jacoby, 1913) or in part a literary fiction of citation (Fehling, 1989), with a middle position (Fowler, 1996). GOOD FOR: the shape of the Persian Wars; the ethnography of Persian religion and custom (1.131-140), which agrees well with Iranian sources; the satrapy and tribute lists (3.89-97); the accession narrative (as a check on and complement to Behistun); and, not least, an empathetic Greek imagining of the many-peopled empire from the inside. BAD FOR: exact numbers of any kind; the interior psychology and verbatim speech of the Persian court; the precise theology of Persian gods under their Greek masks; and any single detail taken on his word alone where no Iranian control exists. The rule is Rollinger's caveat: not to accept his report as history without serious discussion, and to prefer the Iranian evidence, wherever it survives, as the control on the Greek. Cicero's double name still fits: father of history and father of lies at once.

Behistun is at once indispensable and treacherous, and it must be read as both. Indispensable: it is the only Achaemenid royal text that narrates historical events, composed within about two years of them, publicly displayed and empire-wide in circulation, and it corroborates in real detail against the Greek tradition. Herodotus (3.70) names six fellow conspirators, and the inscription (DB §68) names six; Herodotus (3.30) says the murdered Smerdis was Cambyses's full brother, and so does DB: correspondences close enough that some scholars think Herodotus knew a version of the text. Treacherous: it is royal apologia in its purest form, a res gestae whose author chose the medium precisely for self-justification, written by the victor to legitimise a violent and contested accession. It frames every rival as a follower of the Lie, so that rebellion becomes cosmic falsehood and Darius's own killing of the sitting king becomes the restoration of truth; the king even insists, in the text itself, that the reader must not think it false. Two things above all cannot be independently confirmed. The first is the central claim: that the man Darius killed was the Magus Gaumāta impersonating a Bardiya already secretly dead, and not the true Bardiya son of Cyrus. A strand of modern scholarship, from A. T. Olmstead onward, suspects the reverse: that the sitting king was the real Bardiya and Darius the usurper, the impostor-story his necessary cover; the suspicious pattern is that the wave of revolts erupted only after Darius seized power, not during Bardiya's quiet months, and that Darius's is our only source for the imposture. The second is the arithmetic of triumph: the count of nine rebel kings and nineteen battles, and the insistence that all fell in one year, are the king's own boast (though the one-year claim, at least, survives scrutiny if Gaumāta is set aside). The most honest reading is that Behistun is the single most authoritative statement of how Darius wished the seizure of the throne to be understood, not a neutral chronicle of what happened; where it can be checked it is often reliable in particulars, and where it cannot, its interest is precisely served by the version it gives. That it was translated into Aramaic and Babylonian and read aloud in the provinces for generations only sharpens the point: this was the empire's official history, and it worked.

As a historical source the cylinder must be weighed on two quite separate questions, and confusion between them is the root of most of what is wrongly said about it. The first question is factual reliability: does it report truly? On the bare event, the fall of Babylon to Cyrus in 539 BCE, it is a genuinely contemporary witness, buried within a year or two of the conquest, and its central claims about the manner of the takeover, the peaceful entry, the sparing of the city, the change of masters willed by Babylon's own priesthood, are corroborated by the independent and far drier Nabonidus Chronicle and, for the restoration of cult-statues, by administrative tablets. That convergence is worth much: it is rare to be able to check a royal proclamation against a temple chronicle and a business record, and here the outline holds. But everything in the cylinder is refracted through a purpose. It is a Babylonian priestly composition made for a conqueror, and its evaluative frame, that Nabonidus was an impious tyrant hated by gods and people, and that Cyrus is the righteous instrument of Marduk restoring the proper order, is precisely the frame that served both the incoming Persian king and the Marduk priesthood he was courting. Nabonidus's alleged crimes are the standard charges a Mesopotamian text lays against a fallen predecessor; his real religious policy, a promotion of the moon-god Sin over Marduk, gave the Babylon priesthood its grievance, and the cylinder is that grievance turned into a legitimation of conquest. It cannot be read as a neutral account of why Babylon fell or of what Nabonidus was; on those it is an interested party. The second question is genre, and it is here that the modern myth collapses. The cylinder is not a decree, a charter, a law, or a policy statement addressed to the empire or to mankind; it is a foundation deposit in the traditional form of a Mesopotamian building inscription, written in a literary idiom modelled on the inscriptions of Ashurbanipal, sealed into a wall, and speaking in the borrowed voice of Babylon's temple. To call it the first declaration of human rights is a category error twice over, anachronistic in concept (the idea of universal human rights is entirely foreign to the sixth century BCE) and mistaken in kind (the object is a pious wall-deposit, not a proclamation of rights). The honest reading, and the settled scholarly one since Kuhrt 1983, is that the cylinder is superb evidence for two real things, the contemporary Babylonian legitimation of Cyrus's conquest, and, behind the rhetoric, a genuine Achaemenid policy of ruling through local cults and reversing the older practice of deportation, the tolerationist statecraft Darius inherited. It is poor evidence, indeed no evidence at all, for a Persian philosophy of freedom or a charter of rights; that reading was manufactured in 1971 for a modern monarchy's self-image and has been sustained by a forged expanded 'translation' that no scholar credits. The cylinder is thus a lesson in source-criticism in miniature: a document can be authentic, contemporary, and largely reliable on its bare facts, and at the same time be, in its shape and its silences, a work of interested propaganda whose most quoted lines mean something narrower and more conventional than a later age wished them to mean.

The Fortification Archive is, for the study of how the Achaemenid empire actually worked, the most valuable primary source in existence, and its value rests almost entirely on what it is not. It is not propaganda: no royal inscription's self-justification shapes it. It is not literature: no Greek author's love of the dramatic set-piece or the round number colours it. It is not retrospective: it was written in the middle years of Darius I, at the moment, by the office that did the thing. It never expected a reader beyond the next clerk, and so it is candid in a way no consciously composed source can be, booking the wine for the queen and the grain for the god in the same flat hand as the fodder for the mules. Because of this it is the single best control on the whole tradition: where Herodotus or the inscriptions can be checked against it, the checkable claims about administration, roads, women, and cult are often confirmed in detail, and several Greek stereotypes are quietly overturned, the harem (the royal women are landholders and employers), the notion of enforced Mazdā-monotheism (the state subsidised Iranian, Elamite, and Babylonian gods alike, the Elamite ones in fact more often), and the picture of a simple slave empire (the kurtaš are rationed dependent labour, and true chattel slavery is scarcely visible). No source does more to let the empire be seen from the inside. But its candour is not omniscience, and three limits must be held firmly. First, coverage: the archive is a snapshot of one region administered from Persepolis, over about sixteen years, in the specialised vocabulary of a food-and-travel bureau; it is nearly silent on high politics, warfare, law, and the inner lives of the people it enumerates, and an argument from its silence is worth very little, since its silence usually means only that a matter fell outside a ration-clerk's remit. Second, language: the texts are in Elamite, and the Iranian names, terms, and realities they preserve reach us only through Elamite spelling and administrative convention, so that reconstruction always works back through a screen (the whole enterprise of recovering Old Iranian from the tablets, most fully in Tavernier 2007, is the measure of that screen). Third, provenance-of-view: the archive was produced by the very institution whose reach and competence it documents, so it shows the heartland as the bureau saw and ordered it, not as a neutral eye would. And it is, still, only fractionally published: perhaps fifteen to eighteen thousand legible documents exist, of which Hallock's classic edition treated about two thousand, and though the Persepolis Fortification Archive Project has recorded thousands more since 2006, any statistical generalisation from the archive is provisional on a corpus still being read. The honest verdict is that the Fortification tablets are the empire's own books, incomparably trustworthy for what they record and treacherous only if pressed to answer questions they were never kept to answer; used as a control on the narrative sources rather than as a narrative in themselves, they are the firmest ground in the whole field.

For the Achaemenid empire Xenophon is the one classical writer who had actually travelled and fought inside it, and his value is high but sharply divided by work and by kind of matter. RELIABILITY: the ANABASIS is a first-hand military memoir of the 401 BCE expedition of Cyrus the Younger and the retreat of the Ten Thousand, and is the most directly usable of Xenophon's works, an eyewitness on the roads and the reckoning of distance in parasangs, the satraps and their households, the land and its peoples, the King's mixed host, and the one full narrative of a royal battle (Cunaxa) we possess for the whole span between the Persian Wars and Alexander; its old reputation for topographic and logistical accuracy is well founded, though a first-hand witness leaves as many puzzles as clean answers (Tuplin, Encyclopaedia Iranica). The OECONOMICUS (esp. book 4) is grounded and cautiously usable for institutions, being the earliest Greek testimony to the royal paradeisoi (the walled pleasure-garden, whence 'paradise'), the ideology of the cultivating, productive king, the King's review and reward-and-punishment of his governors, and the Lysander-at-Sardis anecdote of Cyrus the Younger planting his own garden; because Xenophon there adduces Persia only from a belief that it was true and pertinent, the passage is a strong measure of his confidence in his own Persian knowledge (Tuplin). The AGESILAUS and HELLENICA carry the Persian dimension of Greek affairs incidentally and in a pro-Spartan frame, with valuable vignettes of satraps (Pharnabazus, Tissaphernes) but a moralising portrait of the Great King as luxurious and secluded. THE CARDINAL CAVEAT is the CYROPAEDIA: cast as the life of Cyrus the Great, it is a philosophical-didactic ROMANCE, a 'mirror for princes', NOT history, its narrative frame 'radically un-Herodotean' (Cyrus allies with a Median Cyaxares, gains Media by marriage, dies peacefully in bed) and regarded by few as a serious alternative to Herodotus; its Cyrus is an idealised Greek construct of leadership, and even its long institutional set-piece (Book 8, on the court, the satraps' imitatio regis, the King's Eyes and Ears, the eunuch guards, the meritocratic ruling class, the King as law) is shaped by its didactic purpose, so that the line between selective truth and agenda-driven invention will be drawn differently by different historians (Tuplin). BIAS: Xenophon writes as a Greek aristocrat and moralist, a disciple of Socrates and a partisan of Sparta, an Athenian exile who had served and admired the King's own rebel (Cyrus the Younger); his Panhellenism colours the Anabasis (though qualified by a sense of Greek weakness), and his judgements on Persian 'decline' and royal luxury are Greek moral types, not Persian facts. GOOD FOR: the empire's interior seen first-hand (roads, distances, satraps, land, army) in the Anabasis; the royal paradise-garden and the ideology of productive rule in the Oeconomicus; the texture of satrapal courts in the Hellenica; and the fullest Greek portrait of Persian KINGSHIP as an ideal (the Cyropaedia), invaluable for the history of Greek ideas of Persia. BAD FOR: the life and reign of Cyrus the Great (use the Cylinder and the Babylonian record instead); any event or institution taken from the Cyropaedia's narrative without independent corroboration; and Persian interior psychology, court ceremonial theology, or 'decadence' taken on his moralising word. The working rules are those set out in the compendium's source-criticism briefs: never let the Cyropaedia's narrative back a fact, use its institutional colour only where corroborated, and prefer the contemporary Iranian evidence wherever a Greek report and the Near Eastern record conflict.