The Persians

The name comes first, and it comes from Assyria. From the ninth century BCE the annals of Shalmaneser III and his successors record raiding and tribute-taking in a land the scribes called Parsua, and by the middle of the eighth century Assyria had organised a province of that name; Tiglath-pileser III installed a governor over it in 744, and Sargon II marched through Parsuash among the Median principalities in 714.^[2][3] Every one of these attestations places Parsua in the central Zagros, near Media and far to the north-west of Fārs, the country later called Persis. The classic solution held that the Persians migrated southwards along the Zagros chain during the eighth and seventh centuries, carrying their name with them until it settled on Fārs. That reconstruction is no longer secure. Wiesehöfer notes that "cogent doubts have since been cast" on the assumed transfer,^[4] and Waters states the evidential position plainly: if such a migration continued into the mid-first millennium, "it is not evident in the available sources, nor have clear traces of such migration been found archaeologically", and the pottery argument sometimes offered in its support "has not found widespread acceptance".^[3]

The debate now runs between three positions. Briant retains a form of the migration narrative, placing it much earlier: "It is generally agreed that, having come from the northern Zagros, rather than directly from the Iranian Plateau, the Iranians moved gradually into Ansan toward the end of the second millennium."^[5] Waters keeps the question open with an explicit alternative: "We thus have Persians attested separately in the central Zagros and in Fars, with the former attested somewhat earlier", and it may be "that two separate groups of Persians migrated into the central Zagros and into Fars, respectively, during roughly the same period", two Persias rather than one people on the move.^[3] Even the prosopographic bridge between the Assyrian record and the dynasty has weakened: the Kurash, king of Parsumash, who submitted to Ashurbanipal in the 640s was long identified with Cyrus I, grandfather of Cyrus the Great, but, as Briant puts it, "this interpretation is now being questioned", since "the proposed equivalence between Parsumas and Persia is uncertain, and Parsumas is probably distinct from Ansan (though this point is still under discussion)".^[5] Behind all three positions stands a shared caution: the earliest "Persians" are Assyrian exonyms, labels applied from outside to Iranian-speaking groups whose own self-understanding is unrecorded, and the route by which Iranian speakers reached the plateau at all, through the Caucasus or around the eastern flank, is itself debated.^[3] What would settle the question, archaeological traces of a moving population or texts naming one, has not been found.

What the migration debate leaves unexplained, the archaeology of Fārs supplies from another direction. The highland later called Persis was, through the second millennium, part of the Elamite world, with the great city of Anshan at its centre; and the dynasty that produced Cyrus titled itself accordingly. "It is now generally accepted", Kuhrt writes, "that, by at least the late second millennium, Iranian people had infiltrated this territory and mingled with the local population, a process leading to the emergence, or better 'ethnogenesis', of the people we call Persians."^[2] Teispes, Cyrus's great-grandfather, "arrogated to himself the title 'king of Ansan'", staking a claim, on Briant's reading, to the heritage of the Elamite kings in the highland that would take the name Persis,^[5] and Cyrus still carried that Elamite royal title in the Babylonian record (see Anshan and the Cyrus Cylinder). Rollinger describes the result as a secondary state formed on Elam's periphery around 600 BCE:

a new political power emerged in the modern Iranian province of Fārs, which clearly exhibited an Elamite sociopolitical and cultural background, but at the same time involved political agents and elites with Iranian roots ... The new political entity labeled itself Anshan

Rollinger, 'From Assurbanipal to Cambyses'^[6]

On this reading the Persians were not a tribe of newcomers who displaced an older people; they were the older and the newer populations fused. Cyrus's own name is generally taken as Elamite, and Potts has even argued that the founder himself should be counted an Elamite (see Cyrus the Great); Kuhrt's conclusion is that "the two ethnic elements were so interwoven that it is probably meaningless to try to draw hard and fast dividing lines".^[2] Potts draws the political consequence: the rise of the Achaemenids "was more comparable to a change in political leadership via an ethno-classe dominante ... than it was to the ascendancy of a 'new' tribal group over an 'exhausted' civilization", the new leadership emerging in a land long accustomed to kingship and statehood.^[7] Briant reaches the same verdict from the institutions: "we now tend to think that the Persians of the time of Cyrus comprised a population descended from a blending of Iranians and Elamites", and "far from being a 'nomadic and primitive' state, the kingdom of Cyrus, based on the Elamite model, was forged with administrative devices that evoke and presage the organization seen fully in operation in Fars at the time of Darius".^[5] How deep the fusion ran in daily life is best seen in the heartland's documents; its terms remain poorly known, because the archaeology of early Persis is thin, and "ethnogenesis" names a process the sources show only at its end.

The country that bore the people's name is a highland wedge between the Zagros and the Gulf, and the ancient geographers described it by its climates. Strabo's account is the fullest:

its seaboard is burning hot, sandy, and stinted of fruits except dates ... secondly, the portion above the seaboard produces everything, is level, and is excellent for the rearing of cattle, and also abounds with rivers and lakes; the third portion, that on the north, is wintry and mountainous; and it is on the borders of this portion that the camel-breeders live.

Strabo, Geography 15.3.1, trans. Jones^[8]

The habitable heart of this threefold country is the middle zone, the valleys around the Pulvar and Kur rivers where Pasargadae and Persepolis were built and where the dynasty buried its kings (see Naqsh-e Rostam). Darius prayed for the land as a possession of his god: "This country Persia which Ahuramazda bestowed upon me, good, possessed of good horses, possessed of good men—by the favor of Ahuramazda and of me, Darius the King, does not feel fear of (any) other."^[9]

The Greeks preferred a harder picture. In Herodotus, the Lydian sage Sandanis warns Croesus against attacking

men who wear breeches of leather and their other garments of the same, and whose fare is not what they desire but what they have; for their land is stony. Further they use no wine ... nor have they figs to eat, nor aught else that is good.

Herodotus, Histories 1.71, trans. Godley^[10]

Herodotus adds in his own voice that "the Persians, before they subdued the Lydians, had no luxury and no comforts".^[10] The stony-land portrait served a Greek argument, poverty breeding hardiness, and it belongs to the same moral scheme as the decadence topos; but the geography beneath it is real enough, and the royal ideology itself made the homeland's men a distinct boast. Persis was a country of villages, herds and estates, without great cities before the kings built them, and its central valleys carried the names of its leading tribe: Pasargadae, as Briant notes, "was both ethnonym and toponym", the capital of Cyrus planted in the tribal territory of his own people.^[5]

The one surviving map of Persian society from the outside is Herodotus's tribal catalogue, delivered in passing as Cyrus raises the nation against the Medes:

There are many tribes in Persia: those of them whom Cyrus assembled and persuaded to revolt from the Medes were the Pasargadae, the Maraphii, and the Maspii. On these hang all the other Persians. The chief tribe is that of the Pasargadae; to them belongs the clan of the Achaemenidae, the royal house of Persia. The other Persian tribes are the Panthialaei, the Derusiaei, and the Germanii, all tillers of the soil, and the Dai, the Mardi, the Dropici, the Sagartii, all wandering herdsmen.

Herodotus, Histories 1.125, trans. Godley^[10]

The list has a tidy internal logic: three leading tribes, three of farmers, four of nomads, with the royal clan (see the Achaemenid dynasty) nested inside the first of the first. That tidiness is part of what makes it suspect. The standard commentary observes that the Maraphii cannot be located and the Maspii, Panthialaei and Derusiaei are otherwise unknown; that Herodotus's Greek kinship vocabulary of genos and phratria is applied in a way no Greek would recognise, the "clan" here being a subdivision of the tribe; and that the genealogy he elsewhere attaches to the royal house, descent from Perses son of Perseus, is "pure Greek invention".^[11] Some entries nonetheless touch attested ground: "The Sagartii (Asagarta in Old Persian) were a Persian-speaking nomadic tribe", and their country appears in Darius's Behistun inscription among the rebel lands of 522–521.^[11] The Germanii are usually taken as the people of Carmania, on the fringes of Fārs rather than within it.^[11]

The schema, then, is a Greek ordering of real but imperfectly understood material, and the second Greek witness disagrees with the first: Xenophon's idealising Cyropaedia divides the Persians into twelve tribes, not ten, a discrepancy that shows how loosely the Greeks held Persian internal divisions.^[12] The Persian evidence neither confirms nor refutes the catalogue. The royal inscriptions never enumerate tribes; the Fortification archive administers estates, storehouses and work-gangs, and its clerks had no more use for tribal labels than for the ethnic label "Persian" itself. What survives of the schema on Persian ground is a name: the tribe of the Pasargadae in the place-name of Cyrus's capital, and the Achaemenid clan in the dynasty's own genealogy.^[5] A ten-fold or twelve-fold nation of neatly sorted farmers and herdsmen is best read as a Greek simplification of a society that its own records show ordered by estate, office and rank.

Rank is where the Persian record does speak. Darius opens Behistun with a pedigree and a claim:

Saith Darius the King: My father was Hystaspes; Hystaspes' father was Arsames; Arsames' father was Ariaramnes; Ariaramnes' father was Teispes; Teispes' father was Achaemenes. Saith Darius the King: For this reason we are called Achaemenians. From long ago we have been noble. From long ago our family had been kings.

Darius I, DB §§2–3, trans. Kent^[9]

Nobility from long ago: the phrase assumes a society in which some families were noble and others were not, and everything else known about Persian society confirms the assumption. (That the pedigree itself may be a construction of Darius's is a separate question; see the Accession of Darius.) Briant assembles the evidence for a steep internal hierarchy: Greek writers grade individual Persians as nobler or better born than one another, and Alexander's historians distinguish "Persians of distinction" from the mass. His conclusion: "The imperial triumph of the dominant ethnic group clearly did not wipe out class differences. The dominant socioethnic class held the preeminent position not just in the Empire but in Persian society itself."^[13] From the Greek vocabulary of homotimoi and isotimoi, "peers" and "equals in honour", he reconstructs a threefold order of great nobility, lesser gentry and commons, with landholding as the material base of aristocratic power and a real, though hard to measure, distinction between families of ancestral estate and families raised by royal favour.^[13]

The hierarchy ran further down than the Greek sources cared to look. The Fortification tablets record Persians among the dependent workforce of the crown, including "Persian boys (puhu) [who] copy texts" on rations lower than a man's; Briant infers "that some Persians were reduced to the status of kurtas under circumstances we do not understand".^[14] Between the boy-scribe and the King of Kings stretched the whole ladder of Persian society, and the empire added a rung above every old one: proximity to the king, cemented by the personal bond the inscriptions call bandaka, came to outrank ancestry alone (see the King of Kings).^[13]

The richest ancient description of the Persians as a people is Herodotus's survey of their nomoi, their customs (1.131–140). Herodotus writes as an outsider ordering foreign practice for Greek readers; he generalises, moralises and occasionally garbles. He also preserves detail no other source offers, some of it corroborated from the Persian side.

On cult he begins with what the Persians lack: statues, temples and altars, whose makers they deem foolish, since they never conceived the gods as manlike. Instead they sacrifice to the sky on the mountain-tops, and "also to the sun and moon and earth and fire and water and winds".^[10] The report of aniconic, open-air worship agrees broadly with the archaeology of the early empire, though his identifications of Persian gods run through Greek names and include at least one slip over the borrowed goddess he calls Mitra (see Anahita and Mithra). At the sacrifice itself, he reports, no altar or fire is needed, but a Magus must chant over the offering, "for no sacrifice can be offered without a Magian" (see the Magi); and the prayer is communal by law: "To pray for blessings for himself alone is not lawful for the sacrificer", who prays instead for the king and all the Persians, "for he reckons himself among them".^[10] That last clause is the sharpest ancient statement of Persian collective identity: the individual's good is folded into the good of people and king together.

On manners and the table:

The day which every man most honours is his own birthday. On this he thinks it right to serve a more abundant meal than on other days; before the rich are set oxen or horses or camels or asses, roasted whole in ovens ... They are greatly given to wine ... Moreover it is their custom to deliberate about the gravest matters when they are drunk; and what they approve in their counsels is proposed to them the next day by the master of the house where they deliberate, when they are now sober ... And when they have taken counsel about a matter when sober, they decide upon it when they are drunk.

Herodotus, Histories 1.133, trans. Godley^[10]

The double vetting of decisions, drunk and sober, has become the most famous of Persian customs; whatever its literal truth, it presents wine as an instrument of counsel rather than of dissipation, a nuance the later luxury literature would erase. Greeting, Herodotus continues, is graded by rank, from a kiss on the lips between equals down to prostration before a superior (the practice the Greeks called proskynesis; see proskynesis), and esteem is graded by distance: "they deem themselves to be in all regards by far the best of all men, the rest to have but a proportionate claim to merit, till those who dwell farthest away have least merit of all".^[10] Persian ethnocentrism, in other words, was concentric, an ordering of the world from a Persian centre that matches exactly the shape of the empire's own ideology. Yet the same people, he insists, were the ancient world's great borrowers: "But of all men the Persians most welcome foreign customs", wearing the Median dress as more beautiful than their own and the Egyptian corslet in war.^[10] The pairing is one of Herodotus's acutest observations: self-regard at the centre, openness at the edges, both at once.

On values, the passage every later account rests on:

After valour in battle it is most reckoned as manly merit to show the greatest number of sons: the king sends gifts yearly to him who can show most. Numbers, they hold, are strength.

Herodotus, Histories 1.136, trans. Godley^[10]

And their sons, from the fifth year to the twentieth, are taught "three things only, riding and archery and truth-telling".^[10] Justice, he reports, is proportional: a praiseworthy law "suffers not the king himself to slay any man for one offence", nor a master to maim a servant for one fault; only when "the offender's wrongful acts are more and greater than his services may a man give vent to his anger".^[10] Lying is the foulest thing of all and debt next after it, "that the debtor must needs (so they say) speak some falsehood"; the truth–lie polarity of this report is the one point where Greek ethnography and the royal inscriptions' war on drauga, the Lie, visibly describe the same value-world (see the drauga and arta).^[10] Rivers are held sacred: "they will neither make water nor spit nor wash their hands therein, nor suffer anyone so to do".^[10] Even onomastics gets a notice, the observation that Persian "names, which agree with the nature of their persons and their nobility, all end in the same letter", true of the Greek transcriptions rather than the Old Persian.^[10] On death he distinguishes carefully what later tradition flattened: exposure of the corpse to bird or dog "is the way of the Magians", who practise it openly, while of the Persians at large he is certain only "that before the Persians bury the body in earth they embalm it in wax".^[10] Read whole, the survey is neither a marvel-list nor a slander: it is a coherent portrait of a hierarchical, pious, family-centred warrior society, drawn by an observer who mostly admired what he described.

What the Persians said of themselves survives above all in one formula. On his tomb at Naqsh-e Rostam Darius names himself "son of Hystaspes, an Achaemenian, a Persian, son of a Persian, an Aryan, having Aryan lineage".^[9] The same self-description, with the patronymic changed, was carved for Xerxes in the daiva inscription and recurs at Susa, reproduced almost word for word across reigns: a fixed dynastic identity-statement, deliberately transmitted (see the daiva inscription).^[15] Kuhrt reads the sons' verbatim repetition of their fathers' texts as a claim of seamless dynastic and ideological continuity.^[2]

The formula is built as a ladder, each rung wider than the last: family (son of Hystaspes), clan (Achaemenid), people (Persian, of Persian stock), and outermost the ariya, the Aryan. An old philological suggestion, recorded by Potts, held that the sequence directly reflects "Avestan social terminology, from family/house (nmana), village/clan (vis-), and tribe (zantu-) to land (dahyu)", the nested categories of Iranian patriarchal society; Benveniste and Christensen took the quadripartite scheme as essentially territorial.^[16] Whatever the exact correspondence, the inscriptions leave no doubt about the outermost term's range: ariya embraces more than the Persians alone. The kings called their god's worshippers, their language and their script "Aryan", and the Avestan cognate names the Iranian-speaking peoples at large; "Aryan" in the Achaemenid texts is an ethno-linguistic and religious self-designation, the widest circle of belonging within which "Persian" was the privileged inner ring.^[4][17] Wiesehöfer observes that under Darius "membership in the Persian or Aryan ethnocultural community now became particularly advantageous", and that the kings "stressed their particularly close connection to an 'Aryan' God, Auramazda" (see Ahura Mazdā).^[17] The word's later career is another matter: the racial sense fastened on "Aryan" in nineteenth- and twentieth-century Europe, and exploited by National Socialism, has no basis in the ancient usage, and scholarship since has insisted, in Wiesehöfer's formulation, that the word "has meaning only as a linguistic term denoting 'Indo-Iranian' ... and that it should be used exclusively in this sense".^[4]

Beside the pedigree stands the boast. The tomb inscription tells the doubter to count the throne-bearing peoples carved beneath the king: "the spear of a Persian man has gone forth far".^[9] Here the ethnos itself, not the dynasty, is the subject of the sentence: the martiya, the Persian man, as conqueror. Ethnic pride has become imperial ideology, and the same reliefs that parade the subject peoples in their variety set the Persian at their head (see the peoples of the empire and Naqsh-e Rostam).

How the Persians actually ruled their empire is the question Pierre Briant's synthesis placed at the centre of the field. His starting observation is prosopographic:

Considered as a whole, the composition of the political staff under the first kings demonstrates the grip on power held by the representatives of the Persian aristocracy in the government of the Empire. It was exclusively Persians who held the command and policy positions ... Local dignitaries were associated with the government of the Empire as auxiliaries to a new ethnically and socially homogenous ruling group. This group would henceforth become the dominant socioethnic class, consisting for the most part of representatives of the great aristocratic families of Persia.

Briant, From Cyrus to Alexander, p. 82^[13]

"Dominant socioethnic class" is the standard English rendering of Briant's French coinage ethno-classe dominante, introduced in his work of the late 1980s and adopted across the field.^[4] The class was held together from above. Each aristocratic family had privileged access to office; each aristocrat stood to the king in the unequal relation of bandaka, bound by exchanged favour and service; and the whole structure rested on what Briant calls the dynastic pact: "the gifts/services exchange was included in the 'dynastic pact,' by which the king undertook ... to protect and favor his Faithful (bandaka)".^[13] Gifts flowed in both directions, tribute and presents upward, estates, honours, precedence and daughters downward; the kings' endogamous marriage policy kept the great families bound into the dynasty itself.^[17] Office, land and honour thus circulated inside a closed circle of Persian houses, and the subject elites, however rich or indispensable, were, in Briant's summary, collaborators of the class rather than co-holders of its power.^[13]

The monuments draw the same line with precision. On the reliefs the Persians appear as throne-bearers of the king, never among the gift-bringing delegations; in the building inscriptions of Susa the subject peoples supply materials and craft while the Persians are absent from the labour lists. Briant's gloss: "even though the Persians are special, they are still subjects of the king".^[13] The ruling people stood above every other people and beneath one man, and the ideology took care to say both things at once.

Briant built the qualifications into the model himself, and the debate since has mostly widened them. First, "Persian" was not a purely ethnic term. Non-Persians reached command positions, the Babylonian Belshunu (Belesys), satrap of Ebir-nāri, being the standard example, and of such men Briant writes that they "were always considered Persians and considered themselves Persians in the political sense of the word".^[18] In his conclusion the point becomes general: the governing elite of the late empire "tended to create a dominant imperial class that was not limited to its Persian component; at the same time, whatever their ethnic origins had been, the leaders always considered themselves Persian in the political sense", and even the sons of mixed marriages could be granted full Persian status by the king.^[18] Persianness, on this evidence, was at its edges a status conferred and performed, entered by service and royal favour as well as by birth.

Second, the evidence itself blurs the boundary the ideology drew. Personal names are the usual test of ethnicity in the documents, and Briant warns that "we must remain circumspect about the ethnic background of people, especially since it is not impossible that non-Iranians took Iranian names when they were incorporated (in one way or another) into the dominant socioethnic class"; Medes and other Iranians stood around the king in any case, and the sources sometimes use "Persian" for any king's man.^[13] Third, the exceptions were real but rare. Wiesehöfer, endorsing the general picture of reserved office, counts the non-Persian satraps and commanders, the Carian dynasts, Belesys, the Greek brothers Mentor and Memnon, and concludes that "such examples are few and far between".^[4] And beneath the elite entirely, the Persepolis archive shows a heartland administration to which the ethnicity of Persians was invisible because it was assumed. The model that survives these qualifications is looser than its name suggests: a ruling class Persian in core and culture, porous at its margins, defined finally by relation to the king.

The ruling people did not stay home. "The settling of a veritable imperial Persian diaspora in the provinces", Briant writes, "demonstrates the royal intent to create conditions for effective authority over the conquered territories and populations."^[19] The instrument was land. From the first reigns, estates in the conquered countries were assigned to Persian families on condition of service, above all the obligation to furnish trained cavalry at the satrap's demand; in Babylonia the bow, chariot and horse estates of the land-for-service system, well documented in the later Murašû archive, tied allotments to military obligation, and royal grants could be revoked from the disloyal and reassigned.^[19][1] The Persians of the diaspora were mustered like the garrisons: "they were summoned (with other troops) for annual reviews in 'the places of assembly.' In this way the Persians of the diaspora were included in the organization of the territorial occupation troops".^[19] Below the great estate-holders, garrison colonies of soldiers settled on allotted land, like the one whose archive survives at Elephantine in Egypt, held the routes and fortresses (see the Elephantine papyri).^[19]

The best-documented Persian abroad is Arshama, the fifth-century satrap of Egypt whose Aramaic letters manage estates scattered across provinces from an absentee's distance; the best-described estate is that of Asidates in Mysia, rich in dependants and slaves, which Xenophon's Greeks tried to plunder about 400.^[19] Dispersal raised the problem of remaining Persian, and the crown seems to have addressed it. When a later writer reports that Artaxerxes II ordered cult-statues of Anahita erected in centres across the empire, from Bactra to Sardis, Kuhrt suggests "we should perhaps understand this as a way of uniting the Persians of the imperial diaspora": a portable, king-sponsored focus of Persian religious identity for communities a thousand miles from Persis (see Anahita and Artaxerxes II).^[2] Estates, reviews, garrisons and cult together made the diaspora an armature of empire: a Persian presence planted in every province, financed by the conquests it guarded.

Ancient writers kept returning to how the Persians raised their sons, and their reports agree on a core. To Herodotus's triad of riding, archery and truth-telling, Strabo, drawing on the Persica literature, extends the term to the twenty-fourth year, adds the javelin to the bow, the horse and the truth, and fills in the curriculum around the martial skills:

they use as teachers of science their wisest men, who also interweave their teachings with the mythical element ... And these teachers wake the boys up before dawn by the sound of brazen instruments, and assemble them in one place, as though for arming themselves or for a hunt

Strabo, Geography 15.3.18, trans. Jones^[8]

The boys are then divided into companies of fifty, each led by a son of the king or of a satrap, and set to race after their leader over marked courses.^[8] The regime is spare on purpose: "Their daily food after their gymnastic exercises consists of bread, barley-cake, cardamum, grains of salt, and roasted or boiled meat; but their drink is water."^[8] Xenophon's Cyropaedia systematises the same material into a philosophical constitution: a "Free Square" by the government buildings from which "the hucksters with their wares, their cries, and their vulgarities are excluded", four age-classes passing through it in order, and a curriculum whose primary subject is character itself, for "the boys go to school and spend their time in learning justice".^[12] In the Anabasis he shows the system's product: the education of nobles at the royal court, where "all the sons of the noblest Persians are educated", where boys watch "the spectacle of men honoured by the King and of others dishonoured", and where "from earliest boyhood they are learning how to rule and how to submit to rule" (see Xenophon and Cyrus the Younger).^[20]

Two caveats are built into this tradition, one by Xenophon himself. The famous education was an aristocratic institution: "all the Persians may send their children to the common schools of justice ... only those do send them who are in a position to maintain their children without work; and those who are not so situated do not", a restriction his own commentator glossed as confining the system to the leisured "peers".^[12] The herdsmen and ploughmen of Herodotus's tribal list were never drilled in companies of fifty. The second caveat is literary. The Cyropaedia is a didactic fiction, and Briant states the difficulty exactly: "It is rather difficult to separate what is truly Persian in the Cyropaedia from Xenophon's implicit but significant transposition from Spartan institutions, which he has described in another work."^[21] The pre-dawn drills and boy-companies may owe as much to the Spartan agoge as to Persis. What survives the caveats is the triad itself, verbatim in Herodotus and Strabo and embodied in Xenophon, and its likely Persian core: Briant suggests that behind the Greek words for truth and justice "corresponded the Persian word arta", so that "what we hear Herodotus and Xenophon saying is that the youths were educated in the duty of dynastic loyalty", trained into the truth-order the inscriptions oppose to the Lie (see arta).^[21]

No institution of the ruling people impressed the Greeks more than the king's dinner. The fullest fragment is from Heracleides of Cumae, preserved by Athenaeus:

some dine inside with the king: and even these do not actually eat with him; but there are two rooms opposite to one another, in one of which the king eats his meal, and in the other the guests eat theirs. And the king beholds them through the curtain which is at the door; but they cannot see him.

Heracleides of Cumae, in Athenaeus 4.145, trans. Yonge^[22]

Heracleides also supplies the numbers that became proverbial, a thousand animals slaughtered daily, and the fourth-century Persica writers Ctesias and Dinon put fifteen thousand diners at the royal table on a single day (see Ctesias).^[22] The Greeks filed these figures under gluttony. Heracleides himself drew the opposite conclusion, and his correction is the key to the institution. The king's supper, he says, "will turn out to be economically and carefully managed": "the greater part of these victims and of this meat is carried out into the court" for the spear-bearers and light-armed troops the king maintains, and as Greek mercenaries take coin, "so do these men receive food from the king, on account, as if it were money".^[22] The Table was, in effect, a payroll and a redistributive engine: the produce of the empire flowed in, Dinon says the fruits of every province were set on it as first-fruits of the king's dominion, and flowed out again as rations, favour and precedence, the same gift-economy that bound the aristocracy graded down to the last guardsman.^[23][24] Dining near the king, behind his curtain, at his expense, was rank made visible and edible, and the administrative tablets of Persepolis show the same redistributive machinery from the inside (see the Persepolis Fortification tablets).

Greek literature of the fourth century built from these materials a theory of Persian history: that luxury had rotted the hardy conquerors of Cyrus's day into the effeminate subjects of the last kings. Athenaeus preserves the topos at its baldest: "But of all nations the Persians were the first to become notorious for their luxury".^[23] The proofs were the wonders of the court. Three hundred women guard the king, and "they sleep all day, that they may watch all night; and they pass the whole night in singing and playing, with lights burning".^[23] Chares of Mytilene supplied the image of the royal bedchamber:

The Persian kings had come to such a pitch of luxury, that at the head of the royal couch there was a supper-room laid with five couches, in which there were always kept five thousand talents of gold; and this was called the king's pillow.

Chares of Mytilene, in Athenaeus 12.514, trans. Yonge^[23]

Clearchus of Soli made the moral mechanism explicit: "the unseasonable and superfluous luxury of their daily life could make even the men who are armed with spears, mere mountebanks".^[23] Plato gave the theory its philosophical form. In the Laws, Cyrus's sons receive "a womanish education, conducted by the royal harem": "women and eunuchs had given his sons the education of a Mede", each heir reared in "unrestrained luxury", so that after Darius the kingship never recovered, and hardly a Persian king thereafter was great except in name.^[25] The epilogue attached to Xenophon's Cyropaedia runs the same argument through every Persian institution, down to the cuisine: "they keep on always inventing something new besides; and it is the same way with meats; for in both branches of cookery they actually have artists to invent new dishes".^[26]

The scheme was not a fourth-century invention. Herodotus had already framed his whole work with it. Sandanis's hard Persians in leather breeches open the Histories; at Plataea, Pausanias sets Mardonius's captured banquet beside a Spartan dinner and invites the generals to see "the foolishness of the leader of the Medes; who, with such provision for life as you see, came hither to take away from us ours, that is so pitiful";^[10] and the work's closing chapter has Cyrus himself warn the Persians against migrating to richer land, since soft lands make for soft men, and if they insist they must "make ready to be no longer rulers, but subjects".^[10] The hardship–luxury arc is structural in Greek writing about Persia from its first great monument to its last.

Modern scholarship long repeated the Greek theory as history: an empire in decline from Xerxes onward, ripe for Alexander. Its demolition, begun in the 1980s and now the standard position, proceeded on three fronts. The first was source-criticism of the key text. The decline-epilogue of the Cyropaedia has been doubted as Xenophon's since the nineteenth century, and Sancisi-Weerdenburg noted that "recently it has been argued by Hirsch, and on good grounds, that authorship by Xenophon of Cyropaedia VIII viii is unlikely"; more damagingly, she showed that its before-and-after structure manufactures decline by setting a prescriptive ideal past against cherry-picked present anecdotes, so that "although the epilogue to the Cyropaedia undeniably contains reliable historical information, it says more about the Greek vision of, than the situation prevailing in, the Achaemenid empire in the fourth century". In such sources, she observed, "through simple comparison with the present situation, the past tends to appear as golden".^[27]

The second front was the topos's ideological function. Briant's 1989 study argued that the decadence literature is about Greece: Xenophon, Isocrates and Plato were not engaged in historical analysis, for "leur projet est avant tout idéologique" [their project is above all ideological], and "les discours de Platon et de Xénophon ont pour sujet réel moins la Perse que Sparte et Athènes" [the real subject of Plato's and Xenophon's discourses is less Persia than Sparta and Athens].^[28] In his synthesis he pressed the argument that the sources' very agreement condemns them, since the thesis was already fully formed among the fourth-century authors and "it is this very unanimity that casts grave doubt on every one of its constituent elements"; the same orators applied the same stereotypes to Macedon, whose army no historian would dream of assessing from Demosthenes' harangues, and Briant asks "how could any historian be so credulous as to portray the Persian Empire on the basis of the words of the Athenian orators and polemicists?"^[29] The conclusion severs defeat from decline: "il n'y a aucune raison de postuler que la disparition d'un Empire soit une preuve d'une décadence antérieure" [there is no reason to postulate that the disappearance of an empire is proof of a prior decadence], and everything now shows, he concludes, "que l'Empire achéménide n'était pas en 334 ce moribond complaisamment décrit par Platon, Xénophon, Isocrate et quelques autres" [that the Achaemenid empire of 334 was not the moribund thing complacently described by Plato, Xenophon, Isocrates and a few others].^[28] The fuller verdict on the empire's condition under its last kings belongs with the dynasty's story (see the Achaemenid dynasty and Darius III).

The third front re-read the luxury itself. Antiquity already contained the counter-voice: Heraclides of Pontus held that those who "have deliberately chosen to live a life of luxury, have been magnanimous and magnificent people", naming the Medes and Persians as the bravest and most magnanimous of the barbarians.^[23] The modern re-reading follows him: the court's splendour was the choreography of majesty and redistribution, and, in Briant's formulation, "tryphe is not reducible to decadent luxury; it is, rather, a striking sign of royal power".^[29] The demolition has itself been qualified. Llewellyn-Jones and Robson, revisiting Ctesias, whom Sancisi-Weerdenburg had charged with inventing the effeminate "Orient", answer that not all the ugliness can be laid at his feet: "Ctesias does indeed create the 'Orient', but an Orient on the Orient's terms", a picture built in part from genuine Persian traditions.^[30] The consensus that survives is narrower and firmer than either extreme: the luxury was real, the decline it was said to prove was not.

One privilege set the ruling people apart on paper. Closing his catalogue of Darius's twenty tribute districts, Herodotus records the Persian country alone as non-tributary, the one land whose people he cannot record as taxed.^[10] That exemption has to be read carefully, and against the documents. The Persepolis archive shows the heartland anything other than fiscally empty: Persis was criss-crossed by assessments, storehouses and payments in kind, a taxed and administered economy in full working order (see tribute and taxation and the satrapy system). Wiesehöfer states the apparent contradiction: "Something that has puzzled scholars to this day is the exemption from duties (ateleia) of Persis, as attested by Herodotus", while the tablets show a profusion of payments from its provincials. His resolution is that "Herodotus, however, merely maintains that Persis was free from the payment of phoros (i.e. tribute)", the assessed levy laid on conquered, dependent countries.^[4]

What Persis rendered instead was gifts. Aelian describes the custom, nomos persikos, still alive under Artaxerxes II:

A custom most carefully maintained by the Persians (nomos persikos), when the king drives to Persepolis, is that each and every one of them, according to his means ... brings an offering. Since they are engaged in farming and toil on the land, living by what they produce (autourgoi), they bring no pretentious or unduly expensive gifts, but rather oxen, sheep, or corn, or in other cases wine. As the king passes and drives on his way these objects are laid out by each man, and are termed gifts (doron); the king treats them as such.

Aelian, Varia Historia 1.31, as given in Briant^[24]

On the reading Briant's account of the gift economy supports, the distinction between tribute and gift was the distinction between subjection and belonging: tribute marked a conquered people, and the Persians, in the empire's own conceptual grammar, were the conquering people, givers of gifts to a king who was one of their own.^[24] Herodotus complicates any neat equation, for in the same catalogue the Ethiopians, Colchians and Arabians also render gifts rather than tribute, as autonomous peoples at the empire's fringe rather than co-rulers at its centre (see the satrapy system). For Persis itself, though, the exemption from phoros reads as a statement of political identity rather than of fiscal immunity, and it matches everything else in the record: the absence of Persia from the later royal lands-lists, the Persians' absence from the gift-bearing delegations on the reliefs, the prayer that folds the sacrificer's good into that of king and people together.

Beneath the political nation laboured a population the Greek sources never mention. The Fortification tablets are dominated by the kurtaš (Old Persian *garda-), the dependent workers of the crown economy, and Kuhrt's summary conveys both the scholarly puzzlement and the answer: "Were they slaves? Serfs? Deportees? Prisoners-of-war? Forced labour gangs? In fact ... it means workers in the widest sense", specialist craftsmen and elite servants included, all those the administration moved, rationed and paid.^[2] Their origins spanned the empire: "the kurtas represented a sampling of nearly all of the peoples of the Empire, including Persians in apparently limited number. Among the kurtas were Bactrians, Sogdians, Babylonians, Assyrians, Elamites, Arabs, Syrians, Egyptians, Lycians, Carians, Ionians, Sardians, Cappadocians, and Thracians."^[14] The gangs were large and mobile, hundreds strong, transferred between Susa, Persepolis and the Fārs countryside; men, women and children worked and were rationed on graded scales, with bonus rations for mothers, doubled at the birth of a boy, a quiet administrative echo of the value Herodotus reports the Persians set on many sons.^[14][10]

For the history of the Persians as a people, the workforce carries two lessons. The heartland of the ruling ethnos was worked to a striking degree by foreigners, the display-lists' subject peoples turned into ration-drawing labour; and the boundary of privilege ran through the Persian people itself, since Persians too appear among the kurtaš, down to the boy-scribes copying texts for less than a man's ration. Persian ethnicity by itself guaranteed neither office nor ease; what it opened, for those born or raised into the right families, was the ladder that led through estate and court to the king.^[14][13]

The archive that ignores Persian ethnicity also shows, more clearly than any other source, what that ethnicity was made of. The language of the heartland's administration was Elamite: "Elamite, not Old Persian or Aramaic, was the prime administrative language in Persepolis" and other centres, continuing the scribal practice of the old Elamite kingdom, and the original Behistun monument itself was carved with an Elamite text before the Old Persian version was added. Henkelman quotes Stolper's aphorism: "Elamite was how Iranians communicated in writing".^[31] The gods of Persis tell the same story. The state issued offerings to Iranian and Elamite deities side by side and, in the surviving sample, the Elamite god Humban outranks the god of the royal inscriptions: "With 26 occurrences in the available Fortification sample, Humban is by far the most popular god", drawing offerings worth more than three times those recorded for Ahuramazdā (see Ahura Mazdā, where the tablet evidence is treated in full).^[31] Henkelman's reading of this pantheon closes the circle opened by the ethnogenesis debate:

just as the native population consisted, in the eyes of the administration, of Persians and Persians only, the Persepolis pantheon was built up from Elamite and Iranian deities that were treated indiscriminately. If 'Persian' means the result of Elamite-Iranian integration, the gods of the Achaemenid heartland had become truly Persian.

Henkelman, The Other Gods Who Are, p. 351^[31]

Herodotus, without knowing it, had described the same people: his Persians who most welcome foreign customs, who wear another nation's dress and adopt another's armour, are the ethnographic face of a culture that had been formed by absorption from its beginning. The conclusion of a generation of archive scholarship is Henkelman's:

And most important, it emphatically defines Persian culture as a locally-grown product. With due recognition of the Indo-Iranian heritage, I think scholars, notably students of ancient Iranian religion, cannot avoid this simple maxim: Persia was made in Elam.

Henkelman, The Other Gods Who Are, p. 49^[31]

The question of where the Persians came from thus receives its current answer in a changed form: they did not, in the sense that mattered, come from anywhere. Iranian speech and Elamite statecraft, highland herding and lowland scribal craft, an Aryan pedigree and an Anshanite throne fused, over centuries in Fārs, into the people who then carried the result to the Aegean and the Indus. The Persians of the empire were younger than most of the peoples they ruled, and the identity their kings carved as immemorial, Persian, son of Persian, Aryan of Aryan lineage, was the newest thing about them.