The Elamites

The Elamites did not call themselves Elamites. The name is, in Potts's formulation, "an artificial construct, a name coined by Mesopotamian scribes, gazing across the alluvium towards the Iranian plateau, who imposed it from without on the disparate regions of highland southwest Iran and its peoples".^[4] The Sumerian writing was the sign NIM, "high", usually with the determinative for "land": to the scribes of the flat alluvium, the easterners were simply the people of the high country. Akkadian rendered this elamtum; the Hebrew Bible adopted the name, and from the Bible it reached modern scholarship through the Greek and Latin traditions.^[5] The people so labelled knew their own country as Haltamti (later Hatamti), a name first attested in the eighteenth century BCE; its etymology is debated, but a plausible reading combines hal "land" with temti "lord", "land of the Lord".^[5]

Behind the borrowed name lies a deeper artificiality. The highlanders of south-western Iran never described themselves by any single ethnic label: in Potts's list, "They were Anshanites, Marhashians, Shimashkians, Zabshalians, Sherihumians, Awanites, etc."; it was only from the Mesopotamian vantage that the region's "easterners – lowlanders and highlanders alike – were all Elamites".^[4] At some point the outsiders' label was nonetheless taken up from within, since Haltamti itself appears to be an Elamite adaptation of the Mesopotamian idea, and the kings of Susa and Anshan ruled a state that modern convention calls Elamite without embarrassment. Potts draws the methodological conclusion for the whole subject: what the sources offer is "not merely the artificiality of the construct 'Elam', but the notion that many Elams were constructed over time, no two of which were probably coterminous culturally, politically or geographically with each other", and the story of the people is "a palimpsest of ever changing definitions of what it meant to be Elamite".^[3] Even the Greeks, who fought Elamite soldiers at Thermopylae, had no word for Elam: they knew the lowlanders as Cissians or Susians, and the name Elam survived in the West only in the Hebrew scriptures.^[6]

Elam's geography explains much of its history and most of its resilience. The country had two unequal halves: the lowland of Susiana, the plain of modern Khuzistan around Susa, an eastward extension of the Mesopotamian alluvium, hot, irrigable and open to Babylonia; and the Zagros highlands to its east and south-east, above all the valleys of Fārs around Anshan, a world of herdsmen, intermontane farmland and hard winters. Gorris and Wicks describe Elam as "remaining as it had been throughout its history a dual highland-lowland cultural entity", and the pairing of the two zones under one kingship, whenever it could be achieved, is what the classic royal title "king of Anshan and Susa" proclaimed (see Anshan).^[7]

Ancient observers already sensed that Elam was more than Susiana. Strabo set apart "the countries of the Elymaei and the Paraetaceni, who are predatory peoples and rely on the ruggedness of their mountains", distinguishing the highland Elymaeans from the Susian plain, and the Talmud carefully distinguished lowland Khuzistan from Elam proper.^[8][6] Modern scholarship nonetheless located Elam at Susa for a century, following the Book of Daniel, which sets Shushan "in the province of Elam"; the discovery in 1973 that the great mound of Tall-e Malyan in Fārs was the ancient city of Anshan forced the revision, argued above all by Vallat, that relocated Elam's centre of gravity to the highlands (see Anshan).^[6] The consequence for Persian history is direct: the highland half of Elam is the country that became Persis. The Persians did not settle next to the Elamite world; they grew up inside it.

The Achaemenid chapter is the last of a very long book, and the earlier chapters can be given in a paragraph. Elamite-speaking polities enter history at the end of the third millennium BCE, with the dynasties Mesopotamian scribes called Awan and Shimashki and the short-lived "empire" of Puzur-Inshushinak; through the first half of the second millennium the sukkalmah regents of Elam traded, intermarried and warred with the great Mesopotamian kingdoms as an equal.^[9] The apogee came late: "From the 14th century BC, the kingdom of Elam knew its apogee under the reigns of the sovereigns of the Igihalkid dynasty and then, in the 12th century BC, Shutrukids", a period Bridey calls "a golden age for the arts of fire", of glassmaking, faience and bronze casting on a monumental scale.^[10] Untash-Napirisha built the great ziggurat city of Chogha Zanbil (Dur-Untaš) in the thirteenth century; Shutruk-Nahhunte and his sons overran Kassite Babylonia and carried the stele of Naram-Sin and the law code of Hammurabi home to Susa, and so the masterworks of Babylonian art and law came to be dug from an Iranian tell (see Susa). The Shutrukid dynasty's fall to Babylon's Nebuchadnezzar I, around 1100 BCE, closes the Middle Elamite age and opens the long Neo-Elamite period in which Elam and Persia would meet.^[7]

The Neo-Elamite kingdom that faced Assyria was a diminished thing. Potts's assessment is that, under Assyrian pressure, "Elam as it had existed in the Middle Elamite period was no longer a unified state linking the highlands of Fars and the lowlands of Khuzistan", and that cities such as Hidalu and Madaktu "were no longer bound by the authority of a single Elamite king at any one time".^[11] Fragmented or not, the kingdom fought Assyria for a century, usually in alliance with the rebels of southern Babylonia, and the war ended in two catastrophes. In 653 BCE, at Til-Tuba on the river Ulai, Assurbanipal's army destroyed the host of king Teumman, who was beheaded with his son on the field; the scene filled the palace reliefs of Nineveh, and Assurbanipal boasted of the slaughter: "I dammed up the River Ulai with the bodies of the warriors and people of Elam. For three days I made that stream flow full of bodies instead of water."^[12][13] Seven years later, after further Elamite turmoil, the Assyrians reached Susa itself. The sack of 646 was an act of deliberate erasure, temples demolished, royal graves opened, the city, in the annals' claim, wiped from the face of the earth (the full inventory of the destruction is given at Susa).^[12]

Two details of the catastrophe carry the next act of the story. First, in the same Assyrian record, the victory prompted a number of Iranian chiefs to send tribute to Nineveh, among them, possibly, the grandfather of Cyrus the Great; in Potts's words, "with it the Persians are introduced into our story for the first time".^[11] Second, the destruction did not reach the hills. "Rock reliefs at Kul-e Farah and Shikaft-e Salman", Potts notes, "seem to relate to petty kings in the highlands who wrote in Elamite but were not necessarily under the authority of the Elamite kings mentioned in Assyrian and Babylonian texts."^[11] Lowland kingship itself revived: kings with the old royal names, among them Shutur-Nahhunte son of Indada and the Atta-hamiti-Inshushinak whom Gorris and Wicks call "the last great Neo-Elamite king", reigned at Susa deep into the sixth century, and Vallat and Tavernier have dated the reign of Tepti-Huban-Inshushinak II to about 550–530 BCE, "which would make him a vassal king of the Teispid king Cyrus II the Great".^[14] An Elamite kingdom, reduced and perhaps dependent, was still alive through the whole reign of Cyrus.

When did Elam end? The older literature answered with the Assyrian sack: 646 as "the final chapter", the rise of Persia as "a new dawn", a dead civilisation succeeded by a young one.^[3] The reading rested on real evidence, the silence of the sources after Assurbanipal and the physical devastation of Susiana, and it had the convenience of a clean period boundary. Potts's synthesis of 1999 was written against it:

I can see no justification for terminating the story of the Elamites with the campaigns of Assyria and the emergence of the Achaemenid Persian empire … I hope that others will see, perhaps for the first time, that the story of Elam and the Elamites does not end with Assurbanipal or the coming of Cyrus the Great.

Potts, The Archaeology of Elam, p. xx^[3]

His argument is historical as much as archaeological. "Elam was no less an entity with a particular linguistic and cultural character in the post-Assyrian period than it had been in more remote antiquity", and its absorption by Persia "no more signalled its demise than had its suppression by the Old Akkadian or Ur III empires" a millennium and a half earlier; the pattern across three thousand years is "the consistent reappearance of Elam following periods of political reversal", an "essential independence – linguistically as well as culturally – of Elam and the Elamites" that he calls "a phenomenon of incredible longevity".^[15]

The excavated evidence has since moved decisively his way. Gorris and Wicks report the state of the question at Susa itself:

from now until the reign of Darius, a continuity in material culture attests to the city's quick recovery after 646 … More than merely surviving, Susa evidently thrived during these late years of Elamite history with religious institutions, administrative systems and artistic traditions that would be inherited by the Persian Empire.

Gorris & Wicks, 'The Last Centuries of Elam'^[16]

Even the old cult city of Chogha Zanbil was still important enough in the seventh century to figure among Assurbanipal's plundered "royal cities".^[16] And the last Elamite century is precisely where Persia becomes visible inside Elam: the Neo-Elamite III period preserves "a particular group of cylinder seals in which Elamite and Persian aspects are blended", and its written Elamite "shows close links to that which was later employed during the reign of Darius at the great Achaemenid city of Persepolis".^[11] The current synthesis therefore dates the Neo-Elamite period's end neither to 646 nor to Cyrus's accession but to 520 BCE, and describes the era not as a death-throe but as "a time of rapid development in southwest Iran characterized by an increasing cultural diversity and political vitality".^[7] What still limits the argument is the spade: the highlands, half of the dual kingdom, remain archaeologically thin for these centuries, and the sequence of the last kings is contested, so the transformation thesis is better documented at Susa than anywhere else.^[7]

Between the surviving kingdom and the Persian empire no war is recorded. The silence has been read positively: Potts suggests that it "indicates that the Elamites and Persians were much more closely bound than otherwise thought and should not be treated as opponents", and that when Cyrus's power spread over the lowlands on his way to Babylon, "Susa was already part of the Elamo-Persian world and hardly a hostile entity".^[1] De Miroschedji went further: to a lowland Elamite, the arrival of a king of Anshan may have looked like the old double kingdom restored rather than a foreign occupation.^[1]

Behind these judgements stands the acculturation argument, the finding that centuries of shared life in Fārs had merged the two populations before either entered imperial history, and that the exchange ran in both directions:

The symbiotic existence of Persians and Elamites, whether in Susiana or in the highlands, had acculturated the Persians to Elamite and, via the Elamites, Babylonian and Assyrian culture as much as an acculturation of Elamites was effected by their contact with the Persians.

Potts, The Archaeology of Elam, p. 307^[1]

The Persian side of that merger, the ethnogenesis of the Persians out of Elamite and Iranian elements and Henkelman's conclusion that Persian culture was a locally grown, Elamite-schooled product, is set out at the Persians; the Elamite side is administrative fact. Tavernier states it from the Persepolis archives: their texts are drafted in Elamite, "proving that the administration set up by Cyrus II in the Anshanite region was originally an Elamite one".^[17] The Teispid state, on this evidence, was less a foreign power that annexed Elam than a fragment of the Elamite world that outgrew it.

Political Elam nonetheless had a last decade, and it is documented by its destroyer. When the empire convulsed at Darius's accession in 522 (see the Accession of Darius), Elam rebelled before any other land. The Behistun inscription gives the sequence: "After I had killed Gaumata the magus, a man called Acina, son of Upadarma", the king proclaims, "rebelled in Elam. To the people, he declared this: 'I am king in Elam.' After that, the Elamites became rebellious; they went over to that Acina; he became king in Elam."^[2] This first kingdom of Açina collapsed at a summons; Darius sent to Elam, and the rising ended in a single line: "That Acina was led to me as a prisoner; I killed him."^[2] The second revolt is the more revealing. Its leader Martiya was, by Darius's own account, a Persian from Kuganaka in Persia, and his name is Iranian; yet he proclaimed himself king in Elam as "Imanish", the Elamite royal name Ummanunu, borne by a king attested at Susa a generation earlier.^[2][18] A Persian could still reach for the Elamite crown under an Elamite throne-name, which is the acculturation of the heartland caught in the act of rebellion. Darius did not need to fight him either: "At that time, I was near Elam. Then the Elamites were afraid of me; they seized that Martiya, who was their chief, and killed him."^[2]

The third revolt alone required an army, and it stands apart in the inscription, added in the fifth column with the campaigns of the king's second and third years. There, the king records, "a country called Elam became rebellious; a man called Athamaita, an Elamite", was raised to lead it, and this time no summons sufficed:

After that, I sent an army; a man called Gaubaruva, my servant – him I made its chief. Then Gaubaruva went to Elam with the army; he joined battle with the Elamites. Afterwards, Gaubaruva defeated the Elamites and decimated them; and he took their leader prisoner, brought him to me. After that I killed him. After that the country became mine.

Darius I, DB §71 (trans. Kuhrt)^[19]

Athamaita is the Old Persian rendering of Atta-hamiti-Inshushinak, and Gorris and Wicks hold that the rebel "probably assumed the royal name of the last great Neo-Elamite king Atta-hamiti-Inšušinak to enforce his claim on the Elamite crown"; the sixth-century king and the rebel of 520 are two different men joined by a deliberately revived name.^[14] With his defeat the independence of Elam ended, on the current reading, for good: the three revolts show that the lowland "was not yet under complete Achaemenid domination" before Darius, and "Upon his defeat Elam was incorporated into the Achaemenid Empire".^[14] The fifth column adds a condemnation with no parallel in the main text, applied afterwards only to the Sakā: "Those Elamites were disloyal, and by them Auramazda was not worshipped."^[19] And the paragraphs recording all this are the one part of the monument that was never translated: unlike the rest of Behistun, the fifth column exists in Old Persian alone, so the end of Elam's independence is the single episode of the inscription that Elamite readers were never given in their own language.^[19]

What the empire took from the kingdom it had extinguished begins with the language. Elamite "was spoken in the southwest of Iran between at least the 23rd and the 4th century BCE", a documented life of nearly two millennia, and it is, in Tavernier's description, "completely isolated": it has no demonstrated relative, living or dead, unless McAlpin's reconstruction of a remote "Proto-Elamo-Dravidian" ancestry, which remains unproven, is accepted.^[5] It was written in cuneiform borrowed from Mesopotamia, in a syllabary progressively simplified until, in the Achaemenid period, it used little more than a hundred signs. Far from dying with the kingdom, the language entered its best-documented age under the Persian kings:

When the Achaemenid king Darius I decided to have his glorious reign eternalized in a large rock inscription at Bisotun, he originally chose to have it recorded in Elamite. Only later were an Old Persian and a Babylonian version added. As his successors took over this habit, various Elamite versions of Achaemenid royal inscriptions are now known.

Tavernier, 'The Elamite Language', in The Elamite World^[20]

The king's first great proclamation to the world was thus an Elamite text, worked by scribes trained in the Neo-Elamite chancellery tradition, and Elamite versions accompanied the royal inscriptions to the end: the youngest known is a text of Artaxerxes III, carved within a generation of Alexander's conquest.^[20] Below the monumental register sat the working mass: the Persepolis Fortification and Treasury archives, thousands of administrative documents in Elamite, with smaller finds at Susa, at Qasr-i Abu Nasr and as far away as Kandahar in Afghanistan indicating parallel Elamite-writing bureaus across the empire's heartlands.^[20] Achaemenid Elamite is saturated with Old Persian influence, in loanwords, names and even word order, and Gershevitch pressed the observation to the claim that it was no living language at all but Old Persian in Elamite disguise; Tavernier's judgement, which reflects the general view, is that "one need not go so far as Gershevitch", the influence being real but the language its own (the alloglottography debate is treated with the archive at the Persepolis Fortification tablets).^[20]

The language survived because its speakers did, and in the empire's most sensitive employment. Briant's account of the heartland administration notices "the activity of Elamite scribes in the chancelleries" as the ordinary explanation for the spread of Elamite cult across the Persepolis region, and identifies the north-western part of the archive's territory as "a veritable Elamite enclave, as is confirmed by the use of Elamite month names for dating documents".^[21] The clerks who wrote the king's accounts, the month-names by which they dated them, and the scribal conventions they worked in were all inheritances of the kingdom that Assurbanipal had supposedly annihilated two generations before Darius.

Within the ration lists themselves the Elamites are at once everywhere and nowhere. Elamites appear among the kurtaš, the dependent workers of the crown economy, in a roster that spans the whole empire.^[22] But as an ethnic label "Elamite" is almost absent from the heartland's documents, exactly as "Persian" and "Mede" are: the Persepolis administrators tagged imported outsiders, Egyptians, Lycians, Thracians, and left the heartland's own peoples unmarked, a silence Tavernier and others read as meaning that the distinction no longer carried administrative weight (see the peoples of the empire).^[23] In the province where Elamites and Persians had merged, the ledger had no more use for "Elamite" than for "Persian"; the empire's own paperwork treats the two heartland peoples as one population.

Greek writers preserve an outside view of the Elamites under the empire, though under other names. Herodotus knows the lowland as Cissia: in the tribute catalogue "the eighth was Susa and the rest of the Cissian country, paying three hundred talents" (see tribute and taxation).^[24] In Xerxes' army the Cissians marched beside the best troops of the empire: "The Cissians in the army were equipped like the Persians", differing, in Herodotus's catalogue, only in that they wore "turbans and not caps".^[24] At Thermopylae they were spent first: Xerxes "sent the Medes and Cissians against them" before committing the Immortals, which places Elamite soldiers in the vanguard of the empire's most famous battle.^[24] And Cissia received the empire's human traffic: the Eretrians deported after 490 were settled by Darius at "a domain of his own called Ardericca in the Cissian land", some forty kilometres from Susa.^[24] The equipment notice is the point. To Greek eyes the Cissians were distinguishable from Persians by headgear alone; Persian and Elamite had become, from the outside as from within, all but indistinguishable.

Religion is where the death of Elam was argued longest, and where the documents have answered most sharply. Koch, one of the few scholars to work through the Persepolis material in the twentieth century, stated the death thesis in its fullest form, taking Inshushinak, "lord of Susa" and patron of Elamite kingship, as her test:

The power and importance of the god Inshushinak for all inhabitants of Elam, from the most humble up to the king, is quite clear from the earliest times onward. So it is a striking fact that he disappeared completely in Achaemenid times. It is understandable, however, since as protector against enemies from outside, he had failed. … Thus with the decline of the Elamite kingdom, Inshushinak lost all of his influence.

Koch, in Civilizations of the Ancient Near East (1995)^[25]

Henkelman's rebuttal turns on the shape of the evidence. The argument presumes an antithesis of Elamite and Iranian, and "Such an approach obscures the historical realities. Sources from the very latest, post-Assyrian part of the Neo-Elamite period still mention Inšušinak"; no class of document survives from Achaemenid Susa that could have recorded his cult, and his absence from the Fārs archive proves nothing about the lowlands. His conclusion inverts Koch's: "Inšušinak's cult at Susa is not only likely to have existed still in the Persian period, but it was, in all probability, actively supported by the Achaemenid rulers", and the god's profile as dynastic protector is "in some respects very similar to that of Auramazdā".^[26]

For the highlands the archive speaks directly. Elamite religion was, in Henkelman's phrase, "very much a patchwork of local traditions", and the gods of its eastern, highland zone, Humban and Napirisha, "are indeed found in the Fortification texts (Humban is even the most popular of the gods mentioned)".^[26] Humban's primacy in the state-funded offerings of Darius's own heartland, where he outranks Ahura Mazdā himself in recorded allocations, is quantified at Ahura Mazdā; Napirisha, the "Great God" of the highlands, whose cult is amply attested to the end of the Neo-Elamite period, drew offerings under his old name and his old logogram alike.^[27] One journal entry shows the fused cult in a single act, an Iranian priest serving the Elamite high god beside mountains and the lan rite:

242 qts. of wine Kammargina, the makuš, received: 20 (as) offering (gal) for/at Mt. Irkamma, 20 from? the King for (the) Mišebaka, 20 (as) offering for lan, 20 (as) offering for Napiriša, 20 (as) offering for/at Mt. Kamarziya, 20 as offering for/at Mt. Battinaša

Persepolis Fortification text NN 2265, trans. Henkelman^[28]

A makuš is a magus; the god he served with the king's wine was the "Great God" of old Elam. Briant fills out the roster: "The attendants of Elamite gods (Humban, Napirisa, Simut, Napazapa) and Babylonian gods (Adad, ki) also received rations", and rations went jointly to attendants of Iranian and non-Iranian gods, which "may indicate that at some locations there were sanctuaries for both".^[21] The Elamite pantheon did not survive on sufferance at the margin of a Persian state cult; it was part of the state cult, funded from the same storehouses, and its endurance in the empire's own heartland is among the strongest single proofs that the Elamites had not been replaced but absorbed.

The kingdom's art ended as its gods did, by flowing into the successor's. The clearest late Elamite monuments are the two open-air sanctuaries of the Izeh-Malamir valley in the eastern highlands, "Kul-e Farah and Šekaft-e Salman, distinguished by their rock-carved reliefs", venues, on the current reading, for "ceremonies involving prayer, animal sacrifice, feasting and musical processions".^[29] Six reliefs spanning perhaps the ninth to sixth centuries cover the cliffs and boulders of Kul-e Farah; the most monumental, KF IV, arranges 141 figures around a seated central figure beside "two tables laden with food and drink".^[29] Only KF I gives names: its chief figure is Hanne, kutur ("leader") of Ayapir and vassal of a king Shutur-Nahhunte son of Indada, who in the accompanying twenty-four-line Elamite inscription (EKI 75) invokes the Elamite gods, records his victories and pieties, and curses whoever should harm his image.^[29] These highland reliefs, carved by exactly the petty Elamite dynasts among whom the Persians rose, "attest to the ritual use and political exploitation of the area until the end of Elamite history".^[29]

Persian art answered them directly. Henkelman calls the Kul-e Farah reliefs "a true gold mine" of parallels for early Persian material culture, and instances the two most pointed borrowings: "The use of the distinctive 'Elamite dagger' by the Persians and the depiction of the Elamite court robe in the relief of the four-winged genius at Pasargadae both reflect an awareness and active appreciation of the Elamite heritage."^[30] The founder's own dynastic monument at Pasargadae, that is, dresses its guardian spirit in the ceremonial robe of the Elamite court. The same continuity runs through humbler media: the blended Elamite-Persian seals of the last Neo-Elamite century, and the glazed-brick craft that Elamite Susa had practised for five hundred years before Darius's builders revived it for his palace friezes (see Susa).^[11]

Written Elamite ended with the empire that had employed it. The administrative archives cease under Artaxerxes I; the last royal inscription with an Elamite version belongs to Artaxerxes III; after Alexander, no Elamite text of any kind is known.^[20] The spoken language is another matter, and Tavernier assembles a long trail of afterlife. The kings of Elymais, the Elamites' Hellenistic and Parthian successor state in the old highlands, bore the name Kamnaskires, "certainly of Elamite origin (the Elamite form being kapniškir 'treasurer')", and his conclusion is that "The vernacular language in Elam may still have been Elamite" centuries after it ceased to be written.^[31] The Elymaeans whom Strabo describes raiding from their mountains are the Elamites under the last transformation of their name, and "Elam" itself lived on as a province of the Nestorian church into the Islamic era.^[31][15]

Three late witnesses close the trail. In the New Testament's Pentecost catalogue, "Parthians, and Medes, and Elamites, and the dwellers in Mesopotamia" hear the apostles in their own tongues: for a first-century CE writer, Elamite was still "a separate and autonomous language".^[32][31] The Talmud records that the Book of Esther was read each year to the Jews of Susa in the proper language of the region, possibly Elamite still, in the Sasanian centuries.^[31] And Arab geographers of the ninth and tenth centuries CE describe an "awkward", "incomprehensible" tongue in Khuzistan called Khūzī; al-Iṣṭakhrī reports that "The common people of Khuzistān speak Persian and Arabic, but they have also another language, Khūzī", a speech "neither Hebrew nor Syriac or Persian". Tavernier's cautious verdict: "Possibly we are dealing here with a late variant of Elamite."^[31] If so, the language of Haltamti was heard on the Susiana plain some fifteen centuries after the last cuneiform tablet, and the people who had no ethnic label in the Persepolis ledgers kept their speech longer than the empire, and the empires after it, kept their names. Potts's summary can stand as the epitaph the Elamites never received: "Elam and the Elamites periodically underwent a process of transformation until the disappearance of the name from Nestorian ecclesiastical sources well after the Islamic conquest."^[15]