The Behistun Inscription (DB)

High on a limestone cliff on the old road from Babylon to Ecbatana, some sixty-six metres above the springs on the plain, Darius I had cut a great relief and a long inscription in three languages and scripts. The place was already holy: its Old Persian name, recoverable from the Greek Bagistanon that Ctesias reports, was Bagastāna, "the place, or stand, of the god(s)", and Darius set his monument of victory where the god's own presence was felt.^[1] The setting was as public as the cliff-face was unreachable. The mountain is the last peak of a long, narrow range skirting the plain of Kermanshah, rising abruptly where a number of springs feed a clear pool at its foot; the caravan road passes between the springs and the rock, and, in the words of the 1907 British Museum edition, "Here from time immemorial caravans have halted and watered their beasts, and every army which has marched from Persia into Northern Babylonia has drunk at these famous springs."^[2] The route itself was one of the great arteries of the ancient world, running on eastward into what later ages would call the Silk Road.^[3] Sacred mountain, water, and highway met in one place, and the king chose it.

There was likely a fourth reason. The fortress of Sikayauvatiš in the Median district of Nisāya, where Darius killed Gaumāta on 29 September 522, has been sought in this very region, and Lecoq suggested that the fortress lay near Bisotun itself, which would make the monument stand almost above the spot where the reign began.^[4] Darius may well have passed the cliff in the spring of 521 as he marched from Babylon to Media. The choice of a triumphal relief high on a Zagros cliff also had a precedent in stone: about 150 kilometres to the west along the same road, at Sar-e Pol-e Zohāb, a ruler of the third millennium, Anubanini of the Lullubi, had carved a victory relief in which the king stands with his foot on a prostrate enemy while the goddess Ishtar leads bound captives toward him on a rope.^[5] The visual scheme at Behistun answers it point for point, and certain turns of formula recall the Urartian royal inscriptions as well.^[1]

When the work was finished the king had the access stairway removed and the path below sheared away, so that no hand could deface or amend what he had set there; for two thousand years the text could be reached only by a hard and dangerous climb.^[6]

The sculpture is the centrepiece; the great inscriptions are arranged in relation to the relief and not to one another.^[6] The panel is modest in size for its fame, about three metres high and five and a half wide, called by local people the Nine Dervishes for its row of standing figures; the measurements taken by L. W. King in 1904 give the figure of Darius five feet eight inches, his two attendants each four feet ten, and each captive three feet ten, so that the king overtops every other human figure in the scene.^[2] Darius stands at the left, his left foot planted on the chest of a prostrate man who raises his hands in entreaty, his bow held before him, one hand lifted toward a figure in a winged disc hovering above the scene. Behind the king stand two of his own men, one bearing a bow and a quiver, the other a spear; the king and his attendants are carved with elaborately curled hair and beards, in greater detail than the god or the prisoners, and the bow's ends are worked into ducks' heads.^[2] Before him, roped together at the neck and with their hands bound behind them, comes a line of nine defeated rebels, each labelled with a small caption naming him and his lie. The man under Darius's foot is Gaumāta. King and Thompson concluded from close examination that the captives' faces are portraits, and their dress and headgear distinguish them people by people.^[2]

Above them all floats the winged emblem, read for a century as a portrait of the god Ahura Mazdā himself, a view still defended, though others take the figure for the royal glory or fortune (khvarnah) that alights on the rightful king; the identification remains genuinely unresolved (see the winged symbol).^[6] Eleven minor inscriptions, cut as close as possible to each figure, identify them: DBa the king, DBb Gaumāta, and DBc–k the nine pretenders.^[6]

What kind of image this is matters as much as what it shows. Margaret Cool Root's study of Achaemenid royal iconography observes that "Scenes depicting the king triumphant over enemies in any explicit 'historical' sense are extremely rare in Achaemenid art. The Behistun relief of Darius is the only known sculpture from this period which can be so described."^[5] Garrison has called Bisotun "the only Achaemenid monument that unambiguously seeks to commemorate known and specific historical events".^[7] The later official art of Persepolis would show the empire as harmonious procession and the king as serene master; only here, at the beginning, is the king shown standing on a beaten enemy. The composition looks back rather than forward: to Anubanini's investiture-victory at Sar-e Pol, which Root reads as "primarily a mutation of the common Mesopotamian investiture scene", and behind it to the Assyrian palace reliefs of Nimrud and Nineveh, the latest of them barely a century older.^[5][7]

The monument as it stands took shape through repeated returns to the cliff. Study of the rock has shown that the master text was engraved in stages between the end of Darius's first regnal year, about March 520, and some time after his third, in 518.^[6] The first main version was in Elamite, the language of the old chancellery of the region and the working tongue of the royal court until about 520. The Babylonian (Akkadian) version was added next, on a projecting slope of rock that looks like a huge clay tablet leaning against the cliff. The Old Persian version was added last of all, directly beneath the relief.^[6][4]

Then came a further alteration. When a fresh campaign in the king's third year, against the Sakā "who wear the pointed hat", gave him a further enemy to display, the rebel queue was lengthened: "And (he) who was their chief, called Skunkha – him they took prisoner, led to me in fetters. There I made another their chief, as was my desire."^[8] Skunkha's figure, in his tall pointed cap, taller in silhouette than the king himself, had to be cut into space already occupied by text: King measured a strip of the first Elamite version just over three feet wide that was chiselled away to make room for him.^[2] The mutilated Elamite text was then re-carved in full in a new position below and to the left. As Waters puts it, the first version was "reinscribed after the addition of the last rebel king to the relief necessitated the defacement of part of the first version".^[9] The king, in other words, sacrificed part of his own inscription to keep the picture of triumph current. A short postscript in the fifth Old Persian column alone, added later and never translated into the other languages, records the campaigns of the second and third years against the Elamites under Atamaita and the Scythians under Skunkha.^[4]

The inscriptions around and below the relief cover, all told, an area some 7.8 metres high and 22 metres long.^[1] The Old Persian runs to five columns of 96, 98, 92, 92 and 36 lines, 414 lines in all and roughly 3,600 words; the moved Elamite copy runs to 260 lines, and the Babylonian to 112 exceptionally long ones.^[6] Every paragraph opens with the same drum-beat, "Darius the king proclaims", so that the whole reads as the king's own speech, and the god's favour rolls through it like a refrain: "By the favour of Auramazda, I am king; Auramazda bestowed kingship upon me."^[8]

The versions are not mechanical copies of one another. The Babylonian text adds, battle by battle, the numbers of enemy dead and prisoners, figures running from the hundreds to the tens of thousands, and the fragmentary Aramaic copy carries the same apparatus with divergent totals; no independent check on any of these numbers exists.^[4][9] The Babylonian version also describes Gaumāta as a Mede and omits his status as a magus, and it renders the dynastic claim without the ambiguity of the Persian.^[4] Elamite came first for a reason: it had been, as Stolper puts it, "the first language used by the Achaemenids for formal inscriptions", the script of the region's chancellery centuries before Persian was ever written.^[10] Because the Old Persian script was new, the words could not simply be written down: the king dictated, bilingual Elamite scribes turned the sentences into Elamite, and the chancellery text became the model for all that followed. The Old Persian is nonetheless the version Darius seems to have regarded as definitive; it was read out to him for approval and shows small edits and omissions that betray a royal hand at work.^[6]

None of this makes the translations subordinate. Kuhrt underlines that "all forms of the inscriptions represent the king's word equally – the Old Persian version does not have priority or carry greater authority than one of the others"; each audience received the royal voice whole, in its own script.^[11] This bears on method: a variant in the Babylonian or Elamite is not a scribe's corruption of a Persian original but a deliberate act of the same chancellery, which is why the casualty figures, the Marduk substitution at Babylon, and the differing ethnic labels for Gaumāta are evidence about the empire's communication rather than noise in it.

The Old Persian version makes its most consequential claim in its final paragraph. On the long-standard reading, Old Persian cuneiform did not exist before this monument: the script was devised expressly to carve this text, and the king says as much, in a passage whose details are disputed but whose plain sense is, in Kuhrt's words, an interpretation "widely, albeit not universally, accepted":^[8]

By the favour of Auramazda, this (is) the form of writing, which I have made, besides in Aryan. Both on clay tablets and on parchment it has been placed. Besides, I also made the signature (?); besides, I also made the lineage. And it was written down and read aloud before me. Afterwards, I sent off this form of writing everywhere into the countries. The people strove (to use it? abide by it?).

Darius I, DB §70, trans. Kuhrt^[8]

The Elamite version is blunter still: "this is my text, which I made, in Aryan, as previously was not."^[8] The physical evidence of the rock supports the claim of novelty, since the Elamite version was carved first and the Old Persian squeezed in afterward; and the short Old Persian inscriptions at Pasargadae in Cyrus's name, which would otherwise be older, are now widely regarded as later additions made under Darius himself.^[4][12] Yet every clause of §70 is contested. Schmitt, the editor of the standard Old Persian text, wrote of his own treatment of the paragraph that "I would like to add that I myself regard neither the restoration nor the translation given here for this …paragraph as final."^[4] Rossi's survey of the current state of the question is more guarded still: it is, he writes, "commonly assumed – with no sure linguistic evidence" that the Old Persian script was not used before Darius.^[10]

A dissenting reading goes further. Rezai Baghbidi has argued that the prevailing opinion, on which "Aryan" here means the Old Persian cuneiform script, has no firm basis: on his reading Darius speaks not of inventing a script but of adding a version in his own language, and the two disputed words at the paragraph's end (which Kuhrt renders "the signature (?)" and "the lineage") are nāmanāfa-, "genealogy", and uvādā-, "personality", the two supplements sent out with the circulated copies:^[13]

Besides, I also made (my) lineage; besides, I made (a manifesto of my) personality. And it was written down and was read aloud before me. Afterwards I sent this text everywhere into the countries.

DB §70, trans. Rezai Baghbidi^[13]

On this view the copies that went out into the provinces carried, appended to the narrative, exactly the kind of genealogical and personal material found in the tomb inscriptions DNa and DNb, and the Aramaic copy from Elephantine, which splices a passage of DNb onto the Behistun text, is the argument's chief exhibit.^[13] The question remains open; what no reading disputes is that the paragraph describes an empire-wide publication programme, on clay and on parchment, in more languages than the rock itself carries.

The text opens with the king's name, his titles, his descent, and the god who gave him the kingdom, then names the twenty-three lands that obeyed him, the oldest surviving catalogue of the empire's provinces, rendered as a claim of obedience and tribute reaching to the edge of a world:

These are the peoples who obeyed me; by the favour of Auramazda, they were my faithful subjects; they brought me 'tribute'; what was said to them by me, whether by night or by day, that they did.

Darius I, DB §7, trans. Kuhrt^[8]

Then follows the account of the crisis. Cambyses, the reigning king, had secretly killed his own brother Bardiya; while Cambyses was away in Egypt, a Magus named Gaumāta rose in rebellion claiming to be that dead Bardiya, son of Cyrus; Cambyses died; and the impostor held the throne until Darius, with a handful of loyal men, killed him. In the king's telling, the whole convulsion is the work of a single cosmic enemy. Of the Magus: "He lied thus to the people: 'I am Bardiya, son of Cyrus, brother of Cambyses.'"^[8] And of the empire's sickness: "the people became disloyal; and the Lie grew among the people, both in Persia and Media and among the other peoples."^[8]

The narrative is precise where precision serves and silent where it does not. Gaumāta's rebellion and his seizure of the kingship are dated to the day (11 March and 1 July 522); the death of Cambyses, which falls between them in the telling, is dated not at all, and is disposed of in five words: "After that, Cambyses died his own death."^[8] What that phrase means has been argued over since the nineteenth century. The old inference of suicide is not supported by any of the three versions; the modern consensus, surveyed by Stolper, is that it means a natural death, though the wording may deliberately contrast the king's own death with the killing he had ordered.^[14] The usurpation itself is painted as a reign of terror: "The people were very much afraid of him, (thinking that) he would kill many people who had known Bardiya previously."^[8] Then the deliverance, dated to the day: "Ten days of the month Bagayadi were past (29 September 522), then I, with a few men, killed that Gaumata the magus", at the fortress of Sikayauvatiš in Media.^[8] The six men are named near the end of the text, and the naming carries an obligation for every successor: "You, who shall be king hereafter, take good care of the family of these men" (see the Seven).^[8] And the restoration is total: "The kingdom which had been taken away from our family, I re-established it, I put it back in its place."^[8]

The one glimpse of what the interregnum actually changed comes in the same paragraph, and it is famously obscure. Darius claims to have rebuilt the āyadanā which Gaumāta destroyed (the word occurs nowhere else in Old Persian; "cult-centres" is a guess steadied by the Babylonian's "sanctuaries, temples"), and continues, "I restored to the people, the pastures and herds, the household slaves and the houses/domains" that the Magus had taken away.^[8][4] Every term in the sentence is debated, but the shape of the claim is clear enough: the usurper is charged with an assault on cult and property together, and the new king poses as the restorer of both. What social programme, if any, lay behind the charge, whether Bardiya's regime had actually remitted tribute and service, as Herodotus's popular Smerdis suggests, is one of the standing puzzles of the reign.^[4]

What follows is a chronicle of revolt. Pretender after pretender rose, each named, each with his province and his fate: Āçina and Nidintu-Bēl in Elam and Babylon, Fravartiš in Media, Martiya, Ciçantaxma, Vahyazdāta the second false Bardiya in Persis, Araxa in Babylon, Frāda in Margiana, and more, some crushed by Darius in person, some by his generals and his father Hystaspes. Each episode is stamped with the same seal: the rebel lied, Auramazda helped, the army of the Lie was utterly defeated, and the day is given. The summary gathers it into a single boast:

This (is) what I have done, by the favour of Auramazda, in one and the same year, after I became king. I have fought nineteen battles. By the favour of Auramazda, I defeated them and took nine kings prisoner. One called Gaumata, a magus, lied, saying: 'I am Bardiya, son of Cyrus'; he made Persia rebellious.

Darius I, DB §52, trans. Kuhrt^[8]

The punishments are set down with the same exactness as the battles. Of the Median pretender Fravartiš the king records: "I cut off his nose, ears and tongue, and tore out one eye."^[8] The mutilated man was exhibited, "He was held in fetters at my palace entrance; all the people saw him", before being impaled at Ecbatana, his chief followers hanged in the fortress.^[8] The display was the point: rebellion against the rightful king was to be seen ending this way.

Five times the king insists he did it all in one and the same year.^[8] The claim was long doubted, and the internal chronology has real problems: the dated battles, when set in order, straddle the accession year and the first regnal year, and the relation of Gaumāta's two dated acts to the undated death of Cambyses is exactly the kind of detail the king left vague.^[11] Borger's study of the monument's chronology showed the boast to be defensible on its own terms: if the killing of Gaumāta is set aside, the run of victories from the first defeat of the Babylonians in December 522 to the overthrow of Frāda a year later did fall within a single year, which happened to carry an intercalary month.^[15] The king himself seems to have felt the strain of his own arithmetic, for he swears to it: he takes the god's anger on himself "that I did this truly, not falsely", a formula Schmitt renders simply "I swear by Auramazda".^[8][4] And he anticipates the reader's disbelief with an admission, that the record has been trimmed for credibility's sake:

By the favour of Auramazda, much else has also been done by me, that has not been written in this inscription; it has not been written down for this reason: for fear that, whoever should read this inscription hereafter, it should seem too much to him, (and so) it should not convince him, (but) he think it false.

Darius I, DB §58, trans. Kuhrt^[8]

No other ancient royal inscription reasons with its audience in quite this way; the sentence assumes a reader weighing credibility, and answers him.

The organising idea of the whole monument is the war of the truth against the drauga, the Lie. Every rebel is a liar; rebellion itself is falsehood made into war; and the king holds the throne because he keeps faith with arta, the right order of things, under the hand of Ahura Mazdā. The theology has its own history (see Religion and the Lie); in the inscription itself it works as accusation. The nine denials in the summary all take the same grammatical shape, a man, a name claimed, a province made rebellious, and the verdict on the whole is delivered as cosmology:

These (are) the countries which became rebellious. The Lie made them rebellious, because these (men) lied to the people. After that Auramazda gave them into my hand; as was my desire, so I did unto them.

Darius I, DB §54, trans. Kuhrt^[8]

Darius says of himself that Auramazda and "the other gods who are" helped him "because I was not disloyal, I was not a follower of the Lie, I was not an evil-doer", neither he nor his family; "I acted according to righteousness."^[8] And he presses the lesson on those who come after, first as a warning, "You, who shall be king hereafter, be firmly on your guard against the Lie",^[8] then as a plea to believe him: "You, who shall read this inscription hereafter, let what (has been) done by me convince you; do not think it a lie."^[8]

The plea is pointed: the king who insists his record is true is the same king whose victory left him the only survivor able to tell the tale, so that the theology of truth against the Lie doubles as the charge-sheet against everyone he defeated.

The dynastic claim is staked in the opening paragraphs and grounds the king's right in his blood: "Eight of our family were kings before; I am the ninth; nine kings are we in succession."^[8] Even that closing phrase is argued over: the Old Persian word duvitāparanam has been read as "in succession", as "in two lines" (the two branches of the Achaemenid house), and, by Schmitt, as "now as ever", so that the very grammar of the dynastic claim is uncertain; the Babylonian version smooths it into "eight among my family administered the kingship before me. I am the ninth. Nine kings of an eternal lineage are we".^[8][4] The genealogy runs back through Hystaspes, Arsames, Ariaramnes and Teispes to Achaemenes, and Teispes is the splice: he appears in Cyrus's own genealogy on the Cyrus Cylinder, but Cyrus never names Achaemenes, and Kuhrt's judgement is that "Darius' genealogy has been manipulated to underpin his claim to have an indisputable family-right to the throne".^[4] The claim's construction, and what it built, the Achaemenid house itself, have their own accounts (see the Achaemenid Dynasty and Darius I); for the document, the genealogy is not a preserved family record but the inscription's opening argument.^[12]

Behistun is a document, but also a made thing, a work of imperial art. Its acknowledged models are the Assyrian Royal Annals, above all those of Assurbanipal, which likewise recount a king's deeds; Waters observes that Darius "recorded his exploits on the model of the Assyrian kings' annals, with a repetitive precision that highlighted his inexorable victory".^[6][9] Even the title "king of kings" points away from Babylon: it is rare among Mesopotamian rulers but frequent among the kings of Urartu.^[4] Yet the repetitive, formulaic rhythm, the recurring "Darius the king proclaims" and the drum-roll of the god's favour, owes something to the models of oral poetry.^[6] The construction is subtler than it first seems: each paragraph is framed as the king's direct speech, introduced as a quotation, so that the text as a whole reads as if a narrator were relaying the royal voice. The Old Persian version, the one Darius held to be standard, is terse and lapidary, its language plain and unbombastic, its sentences sometimes awkward but weighed with evident care.^[6]

The nearest thing to it in the ancient world came five centuries later. Wiesehöfer's judgement has become standard: "The inscription is doubtless a form of royal self-portrayal" and propaganda, "and in this respect it has been rightly compared with the res gestae of Augustus".^[3] Both texts are first-person official accounts, displayed at a charged site and distributed in copies across an empire; both present a contested seizure of power as the rescue of the state from chaos. The difference is timing: Augustus summed up a long reign at its end, while Behistun was composed, translated and broadcast within about two years of the events it describes, a victor's account issued while the blood was barely dry.^[3]

Darius did not leave his history on one unreachable cliff. He commands its publication in the text itself, with a blessing and a curse attached:

Now, let what (has been) done by me convince you; proclaim it thus to the people, do not conceal it! If you do not conceal this record, (but) proclaim (it) to the people, may Auramazda be a friend to you, may you have offspring in great number and may you live long!

Darius I, DB §60, trans. Kuhrt^[8]

The concealer is cursed in mirror image: "If you conceal this record and do not proclaim it to the people, may Auramazda be your destroyer and may you have no offspring!"^[8] The final paragraph states that the text was sent "everywhere into the countries", and the ground has borne this out. Fragments of a Babylonian version, carved on stone and accompanied by a version of the relief, have been found at Babylon itself, set up against the palace wall facing the processional street; in it, wherever the rock names Auramazda, the Babylon version seems to have substituted the city's own god, Bēl-Marduk.^[4][16]

Far more striking are the remains of an Aramaic translation on papyrus from the Jewish military colony at Elephantine in Egypt: a roll originally of twenty-four glued sheets carrying eleven columns of text, of which five columns, in dozens of battered fragments, survive. By the reckoning of Greenfield and Porten, some 79 lines survive, in part, of an original of about 190.^[16] The roll dates from the late fifth century: the memoranda on the reused back of the papyrus begin in 417, and Kuhrt observes that "the Bisitun text and tomb inscription were recopied close on 100 years after Darius I's victory, perhaps to commemorate its centenary", or, on another view, that the text was deliberately recirculated when Darius II had just won his own contested succession.^[4] That an official Aramaic rendering of the founder's res gestae was still being copied and read in a garrison town on the Nile, four generations after the events, shows how the inscription served the empire as its authorised history, carried in the chancery's own lingua franca to peoples who would never see the rock. The roll even folds in a passage from the address to posterity on Darius's tomb (DNb), grafting one royal text onto another.^[16]

The paradox of Behistun is that the empire's most public text is physically illegible. The relief, large and high-cut, could be made out from the road below; the inscription could not. Waters states the position plainly: "The relief would have been visible from the road, but not the inscription."^[9] And once the stairway was cut away, even a determined climber could no longer stand before the text.^[6] The rock version was therefore never the working copy. The working copies were the tablets and parchments of §70, the Babylon stele, the Elephantine papyrus, and behind them the oral publication the text itself demands: it was "read aloud" before the king, and the reader is commanded to "proclaim it to the people", a command to publish, with preservation only the lesser duty.^[8] Publication in the ancient empire meant proclamation, in every province and in whatever language served.

The rock itself served other audiences. In part it addressed the gods and posterity, the reader "who hereafter shall behold this inscription", whom the king blesses and curses; in part the monument worked exactly as a monument, by being seen rather than read, the king triumphant over the roped liars visible to every caravan on the road below. For the small literate audience that could engage the versions together, the interplay itself carried meaning: Stolper notes that "Darius and Xerxes expected the audience for their words not only to read and hear but also to think in Persian and Elamite together".^[17] The monument, in short, addressed every audience the empire had, gods, subjects, scribes and successors, each through a different channel.

The monument was famous in antiquity, and the ancients misread it. Ctesias, reported by Diodorus, took the sheer cliff "sacred to Zeus" (that is, to Ahura Mazdā) and its relief and cuneiform ("Syrian letters") for a monument of the legendary Babylonian queen Semiramis, complete with a park she was said to have laid out below; in this telling the queen ascended the cliff on a ramp built from the packs and saddles of her baggage animals, and the royal figure with his nine captives became Semiramis with a hundred lance-bearers.^[18][2] Diodorus adds that Alexander visited the rock on his march from Susa to Ecbatana.^[2] Isidore of Charax records the village as Baptana, without a word for the sculptures; the Arabic geographer Yāqūt, whose form of the name Rawlinson later borrowed, noted the great height of the rock and its carvings.^[2] That the king's own res gestae could be so thoroughly forgotten and reattributed within a few centuries is a caution in miniature about how quickly the meaning of even a colossal public text can be lost.

Early modern travellers did no better, for the sound reason that no one could read a word of it. Ambrogio Bembo described the sculptures comparatively accurately in the seventeenth century; Otter took the winged figure for a heraldic device; Olivier drew Darius seated on a throne; Gardanne saw a cross above twelve apostles.^[2] Sir Robert Ker Porter, whose 1822 account included the best drawing yet published, read the scene as Assyrian: "In the royal figure, I see Shalmaneser, the son of the renowned Arbaces", with the ten captives as the ten tribes of Israel.^[2] Every age had projected its own past onto the rock; what was missing was the key.

Behistun's second life began in the 1830s. Georg Friedrich Grotefend had already broken into the values of several Old Persian signs using the short royal inscriptions copied at Persepolis, and Rask, Saint-Martin, Burnouf and Lassen had refined the results; but the texts were too short for real progress on the language.^[2] What decipherment needed was a long text, and the long text was some two hundred feet up a precipice.

Henry Creswicke Rawlinson, a young officer in the Persian service stationed at Kermanshah from 1835, copied the trilingual inscriptions at Mount Elvend on his way out and worked out key sign-values independently; from 1835 to 1837 he gave his leisure to Behistun, and by the end of 1837 had copied about half the Old Persian text.^[2] The physical circumstances were extraordinary, and Rawlinson described them himself. The ledge below the Persian text is eighteen inches wide; "with a ladder long enough to reach the sculptures sufficient slope cannot be given to enable a person to ascend", so the upper lines had to be copied standing on the top step, body steadied against the rock: "In this position I copied all the upper inscriptions, and the interest of the occupation entirely did away with any sense of danger."^[2] Reaching the Elamite version meant bridging a chasm; his shortened ladder, laid across it, shed its rungs and "went crashing down over the precipice" while he clung to the surviving side rail.^[2] He returned in 1844 with helpers to complete the Old Persian and Elamite copies, and in 1847 came back for the Babylonian version, cut on an overhang that the local cragsmen declared unclimbable. The solution was found by a volunteer: "At length, however, a wild Kurdish boy, who had come from a distance, volunteered to make the attempt, and I promised him a considerable reward if he succeeded."^[2] The boy crossed twenty feet of nearly smooth rock on his fingers and toes, "in a manner which to a looker-on appeared quite miraculous", rigged a rope traverse, and then, in Rawlinson's words, "Here with a short ladder he formed a swinging seat, like a painter's cradle, and, fixed upon this seat, he took under my direction the paper cast of the Babylonian translation of the records of Darius."^[2]

The rest followed from the paper. Rawlinson published the Old Persian text and translation in 1846–49, the first correct translation of a long Old Persian text ever made; he handed his Elamite materials to Edwin Norris, whose translation appeared in 1855, and edited the Babylonian himself in 1851.^[2][6] Because Behistun repeated one text in three scripts of graduated difficulty, and because the readable Old Persian, a near-alphabetic syllabary of some three dozen signs, could be broken first through the royal names, it did for Mesopotamia what the Rosetta Stone did for Egypt: it gave scholars a bilingual key, and from the Old Persian the decipherers moved outward to the Elamite and then to the far harder Babylonian.^[6][9] Waters calls it "the fundamental document in the foundation of the modern discipline of Assyriology", and notes that its translation "opened the door to a vast catalog of lost Near Eastern literature", the Epic of Gilgamesh among the rest.^[9] Scholars regularly call this inscription the most important single document of the ancient Near East.

Rawlinson's paper squeezes, presented to the British Museum, were handled to pieces within half a century, and the rock itself was faring no better; already in 1847 Rawlinson judged the overhanging Babylonian block "doomed to a speedy destruction" by the water seeping behind it.^[2] A. V. Williams Jackson reached the ledge in 1903 to collate doubtful passages and took the first photographs; in 1904 the British Museum sent L. W. King and R. C. Thompson, who examined every sign from cradles slung on ropes from a ledge two hundred feet above, and their 1907 edition, with its collations, measurements and photographs, put the text on a modern footing.^[2] George G. Cameron came twice more, in 1948 and again in 1957, taking casts in latex rather than paper and reading for the first time the abandoned first Elamite version.^[15] Erosion, seepage and drifting sand have damaged all three versions, especially the Babylonian, and in the last century soldiers on the road below used the carved figures for target practice.^[6] The site was inscribed on the UNESCO World Heritage List in 2006.^[7]

The king had foreseen the danger, if not the artillery. His address to posterity asks, on pain of the god's curse, exactly what conservators now attempt: "You who hereafter shall behold this inscription, which I have inscribed, or these sculptures, do not destroy (them); as long as you shall have strength, care for them!"^[8] Of the empire's other rock inscriptions only Ganjnameh, where Darius and Xerxes carved trilingual praises of Ahura Mazdā on Mount Alvand near Ecbatana, keeps Behistun any kind of company, and it records no events at all.^[7]

Behistun is the ground beneath a great deal else in Achaemenid history. It is the earliest and fullest statement of Achaemenid royal ideology, "the blueprint for subsequent Persian royal ideology" in Waters's phrase, the source of the province-list and the dynastic claim, the charter of the arta-versus-Lie theology that runs through the reign, and, on the prevailing view, the occasion for the Old Persian script itself.^[9] The province-lists, the truth-against-the-Lie language and the first-person royal voice all recur in the later inscriptions, from Darius's own tomb text at Naqsh-e Rostam to the proclamations of Xerxes, for which Behistun set the pattern.