Achaemenid Art and the Imperial Style

The term Achaemenid can name a date, a political setting, a patron, or a recognisable vocabulary of forms, and the four do not coincide. An image made in Persian-ruled Egypt may follow Egyptian conventions exactly and carry nothing Persian; an object showing a Persian court robe could be made, owned and buried well beyond the frontier. Neither a findspot nor a single decorative feature establishes who commissioned a thing.^[4]

Royal inscriptions give the firmest anchor. A monument that names an Achaemenid king and states what he built can be tied to the dynasty far more securely than an uninscribed vessel acquired on the antiquities market. Architecture, excavation context, and repeated combinations of costume and imagery extend the comparison outward from that anchor. Even so, court production is internally varied: a statue designed for an Egyptian sanctuary need not repeat the arrangement of a doorway at Persepolis, and it does not.^[2][8]

An “imperial style”, then, denotes a family of related choices rather than a single template. The long pleated robe, the patterned beard, the dentate crown, the armed hero, winged creatures, and orderly approaches to the ruler recur across stone, brick, metal and sealstone; their combinations and their execution vary widely. Mark Garrison separates the iconography associated with the court from particular carving styles, since a provincial engraver could take up a court subject and render it in an entirely local manner. That separation allows local participation to be described precisely, instead of being written off as diluted copying of a heartland original.^[9]

The administrative reach of the monarchy also ran well beyond the surviving distribution of monumental sculpture. Taxes, governors and military obligations operated in regions where no royal relief has ever been found. Absence of a relief is no argument that a region stood outside Persian rule, and the presence of one calls for an explanation of the particular circumstances in which it was cut.^[6]

At Pasargadae, the residence founded by Cyrus II, the architectural sculpture already draws on several traditions at once. Guardian bull colossi on the Neo-Assyrian model stood facing outward at the north-west and south-east entrances of the gatehouse, Palace R, and survive only as traces. In Palace S the paired jambs of each door repeat a single composition in mirror image, the figures preserved only to knee height and all moving out of the main hall: at the north-west door a bare-footed, bare-legged human followed by a monster with birds’ talons; at the south-east door a bare-footed human in fish garb followed by a bovine figure holding a massive staff or standard; at the south-west door three robed humans leading a walking bull. Root reads the first two as strongly indebted to Neo-Assyrian prototypes of cultic and mythological figures, and the third as a procession of gift-giving of the kind elaborated far more fully on the Persepolis Apadana. All of it was set among columned halls, porticoes and a formal garden, a setting that belonged to none of the sources quoted.^[10][1]

One door jamb relief, on the north-east side of the gatehouse, is still standing. Its figure walks into the building, wears a long fringed robe and two pairs of outspread wings, and carries on its head what Root calls a faithful quotation of the Egyptian ḥm-ḥm war crown. The conventional label “genius” borrows Assyrian terminology and settles nothing. Root proposes an allegorical representation of the royal person, assembled from Elamite, Assyrian, Babylonian and Egypto-Levantine cult imagery. Whether the crown is an oblique echo of an Egyptianising Phoenician bronze or ivory, or a direct allusion to Egyptian kingship and Cyrus’s imperial ambitions there, is itself disputed; Root finds the second reading the more compelling.^[10]

Above the figure stood a trilingual inscription until the mid-nineteenth century. It was copied before it was lost, and Kent prints the Old Persian version complete: “I am Cyrus the King, an Achaemenian.”^[11] The relation between that text and the sculpture beneath it has moved in the scholarship. In 1979 Root treated the inscription as securing the gatehouse for Cyrus and relegated to a footnote the suggestion that the Old Persian texts were added by Darius. By 2021 she had adopted the later-addition view and read it as a deliberate gesture: the sculptures were certainly created before Cyrus died, the inscriptions naming Cyrus appear to have been added by Darius I, and the addition was “a pious and also a politically expedient act” affirming the earlier ruler and Darius’s link to him. Dating the carving and dating the text are two questions, and the second has a different answer from the one long assumed.^[12][10]

In Palace P the doorjambs survive to just above waist level. They show a royal figure at about life size followed by a smaller attendant who originally held a parasol over him, and the costume already forecasts the stiffly pleated court robe that would be the hallmark of Persepolis and Susa two generations later. Slots cut in the stone held metal attachments on the robe and the shoes. Even here the surviving stone was only part of the finished object, and the coordination of images with movement through doorways was already in place: the figures move out of the hall into the plain, as they do wherever a king is shown leaving a building at Persepolis.^[10]

The building inscription DSf, from Susa, is the single richest ancient statement about how an Achaemenid royal building was made. It is also not a monument text in the ordinary sense. Fragments have been recovered across the site on clay tablets, barrels, stone tablets and glazed bricks: thirteen pieces of the Old Persian version, one of them almost complete, twelve Elamite and twenty-seven Akkadian. The text read today is a reconstruction from those pieces, and the versions differ. DSf closes by asking Auramazda to protect the king, his father Hystaspes and his people; the two parallel foundation charters from the Apadana terrace, DSz and DSaa, close without naming Hystaspes at all. Amélie Kuhrt writes that this could be taken to indicate that Hystaspes was still alive when that version of the building text was written, and adds at once that there are so many fragments and versions of the inscription that it is difficult to connect them directly with historical circumstances.^[13]

What the charter asserts is a geography of supply organised by royal command. Cedar was brought from Lebanon, carried by Assyrians as far as Babylon and from Babylon to Susa by Carians and Ionians; yaka-timber came from Gandara and Carmania, gold from Sardis and Bactria, lapis lazuli and carnelian from Sogdiana, turquoise from Chorasmia, silver and ebony from Egypt, wall ornament from Ionia, and ivory from Ethiopia, Sind and Arachosia.^[14]

The list is a claim about reach as much as a record of logistics, and its details resist tidy reading. Kuhrt observes that Egypt is not usually regarded as a silver source, that the ebony will have come through Egypt from further south, and that the Ionian “decoration” may be the glazed relief bricks recovered at Susa, which is possible but not certain.^[13] Material origin and worker identity are separate categories inside the text itself: Egypt supplies silver and ebony and also supplies goldsmiths and woodworkers, while Ionians appear transporting cedar, cutting stone, and as the source of a decorative element. No table can be built in which each province yields one commodity and one style.

The craft clauses are the most quoted lines in Achaemenid art history, and the most often compressed:

The stone-cutters who wrought the stone, those were Ionians and Sardians. The goldsmiths who wrought the gold, those were Medes and Egyptians. The men who wrought the wood, those were Sardians and Egyptians. The men who wrought the baked brick, those were Babylonians. The men who adorned the wall, those were Medes and Egyptians.

Darius I, DSf §§3j–3k, trans. Kent^[14]

Two separate statements are frequently merged into one. The wall ornament that came “from Ionia” is a thing brought to Susa; the men “who adorned the wall” were Medes and Egyptians. The charter does not say that Ionians decorated the walls at Susa, and reading it that way turns an imported material into an imported workforce.^[14]

The categories are also thinner than they look. “Stone-cutters” cannot be expanded into the designers of sculptural compositions, and the assignment of gold or brick to a named people does not produce a single ethnic style for everything those people made. Against the charter stands the Persepolis administrative evidence that Root assembled in 1979. Ionians appear in the Fortification texts only a few times, and are never singled out for an artistic contribution; the ration groups identified by nationality are Sardian blacksmiths, Sardian makers of a commodity called ḫalapzi, Babylonian scribes writing on parchment, female Ionian irrigation workers. Where the texts speak of “artisans” and “makers of stone”, they attach no nationality at all.^[15]

This bears directly on an older thesis. That Ionian and Sardian workers were present is secure on the king’s own testimony. Several twentieth-century attributions turned those craft clauses into claims about individual artists, which Root places among the extremes to which G. M. A. Richter’s reading of Achaemenid sculpture as Greek work had been carried. She records J. M. Cook’s identification of Telephanes of Phokaia as the “grand master of the Persepolis friezes” and Heinz Luschey’s attribution of the Behistun relief to Theodorus of Samos. Her objection is not that the attributions cannot be proved but that they would settle nothing: if Telephanes, known only through a few lines of Pliny, was the grand master of Persepolis, then he “clearly allowed his art to be so prescribed by the court that his Greek nationality had no real significance in determining the nature of the art he supposedly executed”, and the same criticism falls on Luschey’s attribution.^[16] Erich Schmidt’s formulation of 1957 marks the midpoint of the retreat: the architectural sculpture was “executed most probably with the help of foreign — principally Ionian and Egyptian — artists and artisans”, borrowed most of its themes from the conquered west, and “nevertheless represents in its formal, often monotonous, dignity the typical art of the Achaemenid-Persian court”.^[17] Root’s rejoinder was to move the decisive activity to the commissioning court, where designs were selected, combined and approved, and to treat the mixture of hands as a condition of production rather than a dilution of authorship.^[5]

Repetition in Achaemenid relief preserves evidence about how the work was organised. Carl Nylander’s and Michael Roaf’s studies of masons’ marks on column bases and reliefs at Persepolis suggest that much of the sculpture was carved by crews working section by section, and analysis of parts left unfinished suggests that an individual carver was often responsible only for isolated details of a figure rather than for a figure, let alone a composition. Root accordingly suggested a workshop system divided among specialists responsible for details such as hair, jewellery, scabbards and ears, all working from the same or similar models to secure the stylistic homogeneity of the result. Under such a system, she argued, Babylonians may have carved the “typically Greek” omega hem and Greeks the guardian bull colossi.^[18]

An unpublished sculptor’s model of a single ear, two to three times life size, recovered at Persepolis, belongs to the same picture of standardised parts. So does an ancient case that makes the point from the other end: a painted relief of the late fourth century in the tomb of Petosiris in Egypt shows Egyptians manufacturing characteristically Achaemenid Persian vessels. The Petosiris comparison undercuts a simple inference from a work’s style to the ethnicity of its maker. Without independent evidence, resemblance to a Greek, Egyptian, Babylonian or Persian idiom does not identify the craftsman’s origin.^[18]

The technical line of enquiry has widened since. Systems of proportion and the production techniques of baked-brick architectural sculpture have been analysed; and close scrutiny of the Persepolis surfaces recovered preliminary sketches for the polychrome treatment of the reliefs, incised into the stone before painting.^[5] These findings do not decide what the finished scenes meant. They do allow questions about teams, sequence and pattern-use to be asked separately from the interpretation of the finished image, which is the main methodological gain of the last half-century.

Achaemenid reliefs were encountered as parts of buildings. Stone columns, jambs and stairways belonged to structures that also held brick walls, timber roofs, plaster, hangings and furniture. The loss of that fabric has opened sightlines across the Persepolis terrace that no ancient visitor had, and the modern ability to survey several monuments at once reproduces nothing about the ancient route through them.^[1]

A classical description of the complex survives in the account of its destruction. Diodorus, writing in the first century BCE from the Alexander historians, describes a citadel enclosed by three walls, the first sixteen cubits high and set with sumptuous buildings and towers, the second as high again, and the third as a foursquare enclosure sixty cubits high, built of hard marble and described as extraordinarily durable.^[19] He places the royal tombs in a rock-cut mount to the east, reached by hoisting the coffins up and lowering them in, and puts inside the walls lodgings for the king and his soldiers together with chambers for the treasure. The passage belongs to a later Greek literary account of a sacked palace, and its measurements are not usable as survey data. Its emphasis falls on enclosure, buildings, tombs and treasury chambers rather than on the sculpture.^[19]

Within that enclosure, image and movement were coordinated. Reliefs accompanied the ascent of a stairway, the passage of a gate and the crossing between buildings; a visitor stepping through a doorway moved beside a carved king shown advancing with his attendants, and the processions ranged along a stairway façade developed the approach to a hall alongside the represented approach to the ruler. Susa deployed related subjects in a different medium, assembling glazed bricks into large coloured fields in which colour articulated wall and figure together. Similarity of subject between a Susa brick panel and a Persepolis relief implies nothing about shared technique or comparable viewing conditions, and the two media survive so differently that they preserve different parts of the original programme.^[20][1]

Access was uneven throughout. Palace interiors and elevated precincts were not open civic spaces; some inscriptions were buried as foundation deposits, others set too high to read from the ground. Jacobs uses these conditions against interpretations that assume a broad public for every monument, and against the notion that imperial art was a medium of communication with subjects at large. Root places the weight instead on the experience of those who did move through the ceremonial spaces. Both positions require the original location of a work to be established before its audience is described.^[6][5]

Royal texts occasionally address the person looking at the monument, which gives the audience question a foothold in the evidence. At Behistun, Darius turns from narrative to instruction and makes the survival of the image a condition of divine favour:

Darius the king proclaims: If you shall behold this inscription or these sculptures, (and) shall not destroy them and, as long as you have strength, care for them, may Auramazda be friendly to you, may you have offspring in great number, may you live long, and what you shall do, may Auramazda make it successful for you!

Darius I, DB §66, trans. Kuhrt^[21]

The clause that follows inverts each blessing for anyone who destroys them. A few paragraphs later the same text describes its own circulation: the new form of writing was placed “Both on clay tablets and on parchment”, it was written down and read aloud before the king, and afterwards “I sent off this form of writing everywhere into the countries”.^[21] The relief at Behistun is set high on a cliff and is effectively illegible from below; the text imagines a beholder anyway, and records a programme of copying that made the words, though not the sculpture, portable. The imagined viewer and the practical viewer are not the same person, and that gap is the substance of the dispute between Root and Jacobs rather than an objection to either.^[21][6]

Two relief panels, each 6.27 by 3.15 metres, present the clearest surviving image of Achaemenid kingship in operation. A king sits on a high-backed throne with his feet on a stool; behind him stands a second royal figure, marked as the crown prince by his near-identical rendering, his place on the royal podium, and his height; behind them a servant with a towel and a noble in military dress with dagger, bow-case and battle-axe. Facing the king across two incense altars, a dignitary raises his hand in the formal gesture of respect. Standard-bearers stand beyond the canopy, which carries lions flanking a winged disc, with a larger winged disc above enclosing a figure who holds a ring.^[22][20]

The panels were excavated in the Persepolis Treasury and were long described from that findspot. During the Apadana restoration work supervised by Ann Britt Tilia and Giuseppe Tilia, respectively collaborator and director of the Persepolis restoration team of the Istituto Italiano per il Medio ed Estremo Oriente, it was discovered that the central fields of the north and east stairway façades as they now stand are replacements.^[23][24] Ann Britt Tilia then demonstrated what Schmidt had suspected but could not prove: the two so-called Treasury reliefs were the original central panels of those façades. The replacement panels show two confronting files of four guards dressed alternately in the Persian robe and the riding tunic and trousers, separated by a blank inscription panel. Of the two original panels, Root recorded in 1979 that the one from the east façade remained in the Treasury while the one from the north façade had gone to the archaeological museum in Tehran;^[23] the museum Root named has since been renamed the National Museum of Iran (see the Apadana Reliefs).^[25]

The move to the Treasury has been read as the demotion of a disgraced image. Kuhrt records that the panels were carefully installed and positioned inside the Treasury building rather than destroyed, which weakens that reading, and adds that they may well have been generic representations of a king and his designated successor rather than portraits at all.^[22] The identity of the two royal figures is genuinely open: Darius with Xerxes, and Xerxes with his son Darius, who was later implicated in his murder and executed, have both been proposed, and the sculpture supplies no caption.^[22]

The bowing official has attracted four distinct identifications, which Root sets out and weighs: a dignitary performing the act of homage the Greeks called proskynesis; the head of the royal bodyguard, the chiliarch of the Greek sources; the chief treasurer reporting to the king; and the grand marshal of court ceremonies with an ambassadorial function. The third is instructive about method, since Root describes it as an interpretation born of the assumption that the audience reliefs had always stood in the Treasury — an identification generated by a findspot that turned out not to be the original location. Root prefers the fourth reading in view of the newly established original position of the panels. Kuhrt’s caption, by contrast, takes the staff in the man’s hand and the embassies waiting behind him to suggest the chiliarch, in charge of managing access to the king. The figure’s identification therefore depends on which reconstruction is adopted.^[26][22]

On Root’s reading the panel captures a threshold rather than an event: the reliefs of the façade “represent a unified pictorial vision of the moment before the commencement of the presentation of gifts to the king by delegates from the subject nations”, with the waiting delegations drawn up on one wing and the court halted in processional formation on the other.^[27] That reading explains the relation between centre and wings without requiring the whole composition to record a particular occasion, and it leaves room for the ceremony itself to have varied in its participants and arrangements.

On each stairway façade three registers carry a court entourage advancing toward the central panel, and three registers on the opposite wing carry, in Root’s count, twenty-three groups of tribute bearers, each led toward the centre by an usher. Cypress trees separate the groups; clothing, hair, equipment, animals and carried objects distinguish them; the leading delegate of each group is taken by the hand. Repetition supplies the common order, and the detail preserves difference within it.^[23][20]

The two façades are mirror images in scheme and not duplicates in content. Corresponding delegations do not always bring the same gifts: the third delegation carries Median costume and vessels on the north stair and a horse with a single large vessel on the east. The ushers hold their batons differently on the two sides, and one turns his head back toward the delegate he leads. Details of drapery found on some north-stair court robes are absent from the east stair and are paralleled instead on the statue of Darius from Susa. Root reads these as possible signs that the façades were not carved at the same time, while warning that so little is understood about the organisation of labour at Persepolis that stylistic difference there may equally reflect different teams working simultaneously or successive generations of craftsmen on one programme. Taken with certain technical considerations, she nonetheless finds it probable that the north stair was planned first and that its early execution was a priority, given its immediate impact on a visitor approaching from the terrace stair on the north-west.^[28]

Identifying every delegation is not equally secure. The captioned throne-bearers on the royal tombs help, and so do geographical associations of dress and object, but a garment may be shared across regions and a gift need not be exclusive to one country. A further hazard is visible in the Egyptian tradition of foreign gift-bearing scenes that lies behind this kind of composition. Ann Gunter, surveying the New Kingdom tomb reliefs, reports the observation that the foreigners’ appearance, dress and gifts are occasionally “hybrid” constructions, in which artisans “apparently combined elements taken from multiple diverse sources or displaced features from one stock scene to another”; a Syrian gift-giver in the tomb of Sobekhotep carries a lotiform chalice alongside a griffin-shaped rhyton apparently transferred from a stock scene of Aegean bearers. Gunter also records Panagiotopoulos’s alternative explanation: the coexistence of nonlocal and Egyptian products may be deliberate “leveling”, making the act of giving more important in the Egyptian view than the object or its origin.^[29] Identifications proposed for the Apadana groups accordingly carry differing degrees of confidence, which repetition on museum labels does not raise (see the Apadana Reliefs).

What the delegates are doing has been debated for a century. Herodotus reports an annual assessment paid largely in coined and bullion silver and gold, from which Persia alone was exempt, and separately reports peoples such as the Colchians who paid no annual tax but undertook to render a regular gift. Root observes that Herodotus describes no ceremony corresponding to what the Apadana shows, that men from Persia itself are conspicuously absent from the file, and that the objects carried are suggestive of a symbolic offering rather than a fiscal levy.^[30] Her conclusion is that the kind of tribute brought by the delegates “suggests the concept of an encomium rather than the concept of an annually levied tax”.^[31]

The reconstruction of the scene as an illustration of the New Year festival has fared worse still. Root sets out the objections under that heading: there is no contemporary Achaemenid description of any Nō Rūz ceremony, so the reconstruction rests on Sasanian celebrations recorded by Arab scholars, including a fragmentary anecdote from an Arabic source of about the eleventh century CE; and the absence of a Persian delegation bearing gifts on the Apadana façade sits badly with the later festival on which the analogy depends.^[31] Her objections turn chiefly on the dates of the comparanda, which are Sasanian and later.

Rhyne King’s published abstract frames journeys to the king in the Fortification archive, their seasonal distribution and their relation to taxation as evidence bearing on the New Year question.^[32] Dated administrative records can test hypotheses about seasonal gatherings without making any particular relief a literal illustration of them.

At Naqš-e Rostam the tomb façades combine a palace front with an elevated scene of kingship. Representatives of the lands of the empire support a platform on which the king stands facing a fire altar, with a winged figure above; the tomb entrance occupies the architectural register below. Darius’s inscription on the façade turns the viewer toward the supporting figures and tells him what to conclude:

If now you should think: ‘How many are the countries which King Darius held?’, look at the sculptures (of those) who bear the throne, then shall you know, then shall it become known to you: the spear of the Persian man has gone forth far; then shall it become known to you: the Persian man has delivered battle far indeed from Persia.

Darius I, DNa §4, trans. Kuhrt^[33]

The passage supplies the key to the image and limits what may be read from it. The figures are counted and differentiated in order to demonstrate reach, and the sentence that follows converts that reach into military action at distance. Coordinated support expresses the king’s preferred arrangement of the empire; it says nothing about the terms on which any particular country was acquired or continued to submit, and the monument’s success as a statement depends on presenting that arrangement as legitimate.^[33][6]

Two side attendants on Darius’s tomb carry captions. DNc identifies “Gobryas, the Patischorean, spear/lance-bearer of Darius the king”; DNd reads “Aspathines, the ‘garment-bearer’ (i.e. chamberlain), holds Darius the king’s bow-and-arrow case.”^[34] Named individuals here do not convert every comparable figure elsewhere into a portrait. In Kuhrt’s discussion, the corresponding side figures on the surviving later royal tombs are unnamed, while Darius’s tomb supplied the layout and decorative pattern followed by the subsequent Achaemenid royal tomb façades.^[34]

The Behistun relief is exceptional in the royal repertoire for its tie to a detailed narrative of political conflict. Darius stands with his foot on the fallen Gaumāta while bound claimants file before him, and the inscriptions name them and describe the suppression of their revolts. The scene condenses the king’s claims into a single ordered tableau instead of unfolding them as episodes.^[35]

Its formal ancestry is traceable to a specific monument. About a hundred kilometres west of Behistun, near Sar-i Pul-i Zohab, the relief of Anubanini shows the king of the Lullubi with his left foot on the chest of a prostrate victim, bow and arrow in his left hand and a crooked staff in his right, facing the goddess Ištar, who extends a ring while dragging a bound, naked prisoner by a rope; six small standing captives occupy the narrow zone below. Root reads the Sar-i Pul relief as primarily a mutation of the common Mesopotamian investiture scene, with the king appearing as victor rather than petitioner, and she also records what its own inscription says it depicts: “Anubanini, the most powerful king, king of the Lullubi, has commanded a representation of himself and of the goddess Istar to be carved on Mount Batu.” The captives are not mentioned at all. Her discussion places the monument within the third millennium and reports Dietz’s proposal, argued from grammatical features of the inscription, of a date after Naram-Sin and before Gungunum of Larsa.^[36] Potts dates the relief to the Isin-Larsa period.^[37]

The comparison establishes the availability of a long-standing vocabulary of domination in exactly that landscape, and it supplies a caution about ancient descriptions of images: a text accompanying a relief may name only part of what the relief shows. The Achaemenid commission reselected and rearranged those elements in answer to Darius’s particular need to justify an accession that was open to challenge (see the Accession of Darius).^[35]

Other royal images show a heroic figure in court dress overcoming a lion, a bull or a composite creature. These convey controlled strength through a single legible encounter; their religious and political reading varies, and a figure in royal dress is not a record of the king performing a literal act. The catalogue of the Fortification seals, whose first volume is devoted entirely to images of heroic encounter, shows how widely the motif circulated below the level of royal monuments and in how many carving traditions.^[38][9]

The near-absence of battle narrative from the palace reliefs has encouraged a reading of Persian art as a programme of imperial peace, often by contrast with the siege and deportation scenes of Assyrian palaces. Jacobs instead argues that the images are undramatic rather than peaceable, and that they omit testing situations such as war or hunting because those activities were not the basis of royal legitimation in this visual programme.^[6] A programme that gives little space to warfare can belong to a conquering monarchy without contradiction, and DNa says so in the same breath as it points to the throne-bearers.^[33] The interpretative problem is to account for the selection of subjects without mistaking a monument’s preferred image of order for a description of government.^[6]

Costume is the principal means by which these images sort people. The pleated court robe differs from the tunic, trousers and outer coat suited to riding; a crown, a weapon, a parasol or the arrangement of attendants further specifies rank. The conventions are economical enough to work at small scale and in restricted spaces.^[10][9]

A Greek description of the king in procession supplies the detail that stone cannot. Curtius describes Darius III in a purple tunic with white woven at its centre, a gold-embroidered cloak patterned with confronted hawks, a gold belt, a dagger in a jewelled scabbard and the royal kidaris, its blue ribbon flecked with white.^[39]

Curtius wrote in the Roman period about Darius III’s army marching from Babylon, and his account is a set piece of eastern luxury as well as a description. It nonetheless registers colour, pattern and applied ornament on a royal costume that survives in stone only as pleats and outlines, and it offers a check on how much of the original appearance the bare relief withholds.^[39]

The two main forms of elite dress at Persepolis have conventionally been labelled “Persian” and “Median”. The label misleads. Kuhrt records the philological correction plainly: “Median dress” was the military garb of the Persians.^[40] Herodotus, describing Marathon, says the Athenians were “the first to face the sight of Median dress and the men who wore it”, which shows the Greek usage forming as an ethnic marker at the very moment it described Persians.^[40] The alternation of the two costumes along the Apadana façades has accordingly been read in more than one way: Ernst Herzfeld proposed that figures were shown in the two garbs alternately to create a regular rhythm of folded and plain material, while Root holds that the alternation was intended to carry iconographic value.^[41] Neither reading turns the two files into two ethnic populations.

The same restraint applies to provincial objects. A seal owner might adopt a court image as an expression of office, ambition or affiliation, alongside a local language, local kin and other forms of self-presentation. Identifying such an owner as an immigrant Persian from the image alone, then using that inferred identity to argue that the image circulated only among Persians, would be circular.^[9] The carved body itself is highly selective: coiffure, beard, garment and gesture take the forms appropriate to rank, and the resulting regularity has sometimes been read as an inability to individualise. Root’s production-oriented reading is compatible with that regularity: these programmes standardised rank-bearing traits, while captions supplied names in the cases where named identity was required.^[5][34]

The present appearance of the monuments understates colour heavily. Early excavators recorded pigment on reliefs that faded soon after their contact with light, and traces of black, brownish-red, green, blue, white and yellow were recorded on the east face of the Apadana, the Central Building, the Hall of a Hundred Columns, the Treasury and the Palace of Darius. Analyses identify asphalt for black, opaque red glass, Egyptian blue, ochre for yellow, and a red floor in the Treasury made of lime plaster with red ochre; the wall paintings of the Chaour palace at Susa yielded Egyptian blue, hematite and cinnabar. Painted plaster, coloured floors, textiles and metal fittings all belonged to interiors whose fragile components have gone.^[7]

Where a pigment sits on an object can change what the result means. Janet Ambers and St John Simpson identified hematite on the trimmed right vertical edge of the British Museum’s chariot-horse fragment BM 118851, from the uppermost register of the east wing of the Apadana north stairway; the coloured area had been concealed by a nineteenth-century stone mount. The presence of colour on that edge, they write, supports an earlier observation by Ann Britt Tilia that red paint was applied to the smoothed anathyrosis edges of sculpted blocks at Persepolis, that is, to joint surfaces invisible once the masonry was assembled, the purpose presumably being to make the vertical edges fit precisely.^[42] Their analysis of BM 2000-3-28,10, by contrast, identified hematite in a small fragment of red-surfaced floor plaster found at the site. They report it as confirming a suspicion recorded by Weld in 1892 and as agreeing with Matson’s analyses of 1953, which had found red ochre on a separate but similar fragment of lime plaster from Room 73 of the Treasury. Floors differed by building: red plaster in the Palace of Darius and in one or more buildings south of the terrace, greyish-green gypsum plaster in the Apadana and the so-called Harem, plain white in the Hall of a Hundred Columns. The same mineral thus supplied evidence for two quite different uses.^[42]

The third specimen turned out not to be ancient at all. On the seated bearded sphinx BM 129381, green flecks trapped within a carved earring, noticed by Shahrokh Razmjou and extending downwards as though washed down, suggested malachite. Raman analysis of a few grains produced a complex spectrum including anatase, a form of titanium oxide which can occur as a natural mineral but is not common and, the analysts write, “does not seem to have been used as a painting pigment until the discovery of an artificial manufacturing process in around 1920 AD made it readily available”. They interpret the spectrum as a modern pigment based on titanium white, the green resulting from a further unidentified component. Ambers and Simpson draw the moral themselves: “The identification of this modern paint fleck highlights the extreme care with which preliminary observations on the possible colouration of sculptures ought to be treated prior to scientific analysis.”^[43] The same fragment carries a second, unrelated uncertainty: published as probably from Palace H, it may instead derive from the south façade of Palace G, part of which was reused in a later reconstruction of the north façade of Palace H, in which case the present partial restoration of that façade is wrong.^[43]

Work in situ has added specimens of a different kind. Alexander Nagel’s examination of the western bull head in the north portico of the Hall of a Hundred Columns found the uppermost layer of the limestone preserving a fine layer of blue paint, in some areas still up to a millimetre thick, now shiny and almost glassy, crystalline and flaky; he judges it likely that a semi-vitreous substance such as Egyptian blue was employed. Red appears in the nostrils, and abundant traces of a red pigment, most likely ochre, survive on the left eyeball, with the same pigment in the tear duct and on the eyelids of other stone bulls. Nagel couples the finding with a caution of his own: whether the coating once covered the whole of the body is undetermined, and Tilia had argued that the ochre served as a ground for a finishing layer rather than as a final colour. The companion head in Chicago preserves no original paint at all, and must have been cleaned at some point between excavation and installation. Nagel presses the historiographic question that follows: why Herzfeld, the conservator Bastiani and Tilia documented no paint on these heads, Tilia despite a long article on colour at Persepolis.^[44]

Razmjou’s short note on the statue of Darius from Susa shows the limits of an observation made before laboratory analysis. Razmjou found paint on the statue: “The side folds are painted in a brown-redish colour… On the folds with inscription, however, no traces of colour pigments have been found as yet.” Samples were taken for laboratory analysis, and the note ends with the questions still open, whether the statue was painted throughout and whether the pigments are Persian or Egyptian in origin. The result is not reported there. What the record holds is an observation with analysis pending, which is a different thing from an identified pigment.^[45]

A study published in 2024 re-examines a winged-figure fragment from the Hall of a Hundred Columns now in the Harvard Art Museums, reporting Egyptian blue, malachite, cinnabar and other pigments, including evidence interpreted as blue hair and beard. Several features remain uncertain, and detected particles do not always establish an original painted boundary; the authors reassess earlier reconstructions rather than substituting a finished new colour scheme.^[46] Earlier records are consistent with coloured winged figures of this kind, since Ambers and Simpson record colour traces on winged figures in the Fogg Art Museum and the Persepolis Museum, and Nagel attests pigment on the doorjambs of the Hundred Column Hall.^[7][44] Consistency of that sort is not confirmation, and the 2024 authors present their identifications as the result of analysing one fragment rather than as a scheme for winged figures generally.^[46]

The three British Museum specimens keep the evidential levels apart. Identified pigment, colour inferred from a comparable object, and hypothetical completion are distinct categories, and a coloured rendering that merges them loses the distinction: an ancient pigment need not belong to the visible decorative surface, and a coloured trace need not be ancient at all.^[42][43]

Susa preserves the other half of the colour evidence, in a medium where colour can be structural rather than merely applied. Harper, Aruz and Tallon divide Achaemenid brick decoration there into two broad categories: relief decoration made from uncoloured clay bricks cast in moulds, and coloured decoration made from siliceous bricks with coloured glazes. Polychrome decoration at Susa was overwhelmingly siliceous: each brick was shaped in a mould, with or without relief, and fired up to three times. The craftsmen outlined their designs with threads of thick glaze to create separate compartments holding the different coloured enamels, a technique the catalogue calls “cloisonné”, and they set the bricks edgewise with the decorated side up to apply the glazes, as drips down the wide undecorated sides show. X-ray diffraction of the glazes identifies lead-antimony for yellows, copper for blue-greens, ferrous manganese for blacks and browns, tin to make the glaze opaque and white, cobalt for a bright blue. True red is absent from Susian faience and glasswork, although it was known to Assyrian and Egyptian glassmakers and frequently used in mural painting in Mesopotamia and Susa. The glazed bricks run about 8.5 by 33 centimetres, slightly wedge-shaped so the decorated faces fit together, with corner pieces glazed on two sides. Some clay bricks were exceptionally given a coloured glaze, carrying geometrical motifs or elements of Persian dress; Harper, Aruz and Tallon suggest these may have been repair pieces made to match broken elements in the larger siliceous series.^[47]

The reconstruction of these panels is far more conjectural than their museum display suggests. Almost all the decorated bricks were recovered from later levels where they had been reused haphazardly, and Harper, Aruz and Tallon stress that how the elements fitted into overall designs can only be conjectured. The most famous panel is the least secure. The catalogue explains that modern restoration used the lion-frieze arrangement as a model, while proposing that the archer frieze probably belonged in two or more superposed registers, with left- and right-facing archers converging on a central element bearing a trilingual inscription.^[47]

The same account proposes that the frieze adorned a façade of projecting pilasters alternating with recessed niches rather than a flat wall, which is why some bricks, the corner pieces, are glazed on two faces.^[47] The restoration history runs back further than the present arrangement suggests. The archers and the lion frieze recovered by the Dieulafoy excavations were exhibited at the Louvre in the late 1880s, after a first restoration campaign, in what the catalogue calls their newly restored splendor; the archer frieze is reported to have been restored with the assistance of the ceramic factory at Sèvres.^[48]

One assembled panel can be described from the museum record, and it is worth keeping distinct from the frieze as a whole. The Louvre catalogues SB 3317 as the Frise des archers, in glazed siliceous brick, 183 centimetres high, 68 wide and 13 deep, and connects the excavated fragments with the eastern gate of the Apadana complex, recording finds in 1908 and 1913.^[49] That catalogue history belongs to that panel. It does not extend to every archer fragment, and it does not make the modern arrangement of the archers a recovered ancient design (see Susa).

Seals set image-making inside repeated transactions. A cylinder seal made its image by being rolled across a prepared surface, a stamp seal by pressure; the resulting mark could identify a person or an office, authenticate a document, or secure a container. The engraved object and its impression are separate kinds of evidence, and for the Persepolis material the impressions are what survive.^[50]

The sealing praxis of the Fortification archive is defined by document class rather than by a general rule. In the catalogue volume of the Fortification seals, Garrison and Root set out two categories where the practice is clear. T texts are letters, and here the addresser alone appears to seal the tablet. Q texts record disbursements relating to travel, and give access to a wide range of offices and travelling individuals: “These tablets are sealed with the seal of the supplying agent or his office on the left edge. The receiver of commodities seals the tablet on the reverse (and often, but not always, on other available surfaces).”^[3] A third class, J texts, covers goods dispensed before, or on behalf of, the king and other members of the royal family, and only a small group of seals represents oversight of that activity: PFS 7*, a royal name seal of Darius; PFS 66*a, PFS 66*b and PFS 66*c, three almost identical seals clearly meant to replicate each other, each representing a royal office of flour supply; and PFS 93*, a royal name seal of Cyrus of Anšan. Even here the protocols are less straightforward than the authors would wish, and PFS 859* occurs where PFS 93* would be expected. Personal seals of principal figures can also be identified, such as PFS 38 for Irtašduna and PFS 51 for Irdabama.^[3]

The triplet of near-identical flour-office seals is a strong argument that a seal could represent an office rather than a person, and the same conclusion follows from the movement of PFS 7* through the system. Garrison illustrates the two-seal pattern with a single document: the left edge of PF 702 carries PFS 66a*, the reverse PFS 7*.^[51] The orthography varies between publications, the catalogue volume writing PFS 66*a and the later synthesis PFS 66a*, for the same seals. No two-seal rule covers the archive as a whole; the pattern belongs to particular classes of transaction.

The people who appear in these transactions are not confined to the court. Garrison and Root run their list from the top of the hierarchy down: members of the royal family such as Irtašduna, daughter of Cyrus and wife of Darius, using PFS 38; exalted nobility such as Gobryas, the king’s spear-bearer, using PFS 857s; a camel driver using PFS 1532s; and a man far from home, from Sardis, using PFS 1321s. Their own summary is that although the tablets are appropriately called administrative, they are by no means limited in the range of participants to officials of the court.^[52] The images on those seals therefore document choices made across a wide social range.

A single device can occasionally be followed beyond Persepolis. An Achaemenid Elamite tablet from Susa, “unfortunately without any specific archaeological context”, is definitely related to the Fortification archive system and is sealed with PFS 7*; a tablet from Qasr-i Abu Nasr, probably ancient Shiraz, may well belong to the Fortification system and carries a cylinder similar to but not identical with others in the corpus.^[53] The Susa tablet is dated in the later synthesis to 500/499 BCE, a date the catalogue volume does not give.^[50] A shared impression is far firmer evidence of connection than a stylistic resemblance, and Garrison relates such links to officials who travelled with the king as he moved between capitals. That remains an interpretation of a distribution; the tablet records no itinerary.^[51]

Style within the archive is plural, and its taxonomy is defined within a stated corpus. Within the first catalogue volume, the Fortification Style accounts for 158 seals, just over half; alongside it stand the Modeled Style, Mixed Styles I and II, and the Court Style, which accounts for only twelve seals in that volume. The term Court Style was popularised by John Boardman and has conventionally been defined through a set of iconographic features — the court robe, date palms, figures emerging from winged symbols — rather than through strict stylistic criteria. Even within those twelve, Garrison and Root distinguish two apparently contemporary stylistic expressions, one hard and linear as on PFS 7*, the other softer and more plastic as on PFS 11*. PFS 9*, Parnaka’s first seal, is an archaising adaptation of Assyrian imagery carved in the local Fortification Style.^[38] Garrison and Root give these proportions for the corpus of that volume, which is one published instalment of one archive.

What the images meant to their users is harder to recover than how they were used. The choice of a precious material, the wearing of a seal visibly, and the selection of an elaborate design could all express standing or affiliation, and the text of a document may identify a user without explaining his choice of scene.^[50] One case shows the court image travelling on its own. A cylinder seal known from twelve impressions at Daskyleion in Hellespontine Phrygia reproduces the Persepolis audience scene almost element for element — royal figure, throne, official, spear-bearer, servant with towel, bow-bearer with battle-axe, winged disc, incense altars — and is inscribed with the name of Artaxerxes in Old Persian. Kuhrt calls the differences minimal: no crown prince, no baldachin, and a fan held by a differently dressed servant. Her 2007 caption states that no such audience scene had so far been found among the seal impressions in the Fortification archive.^[54]

One object is invoked in almost every account of Achaemenid glyptic, and it carries more uncertainty than its fame suggests. BM 89132 is a cylinder of green and grey-brown clouded chalcedony, 3.7 centimetres high and 1.7 in diameter, acquired from the Salt Collection in 1835. The king stands in a chariot driven to the right and aims a third arrow at a rampant wounded lion; the horses leap over the body of a fallen lion; a figure in a winged disc hovers above; palm trees flank the scene, and between them runs a trilingual cuneiform inscription in Old Persian, Elamite and Babylonian, SDa: “I [am] Darius, great king”, the word “great” occurring only in the Babylonian version. The catalogue of the 2005 British Museum exhibition describes it as probably a seal of office, and records that it is “Said to have been found in Lower Egypt”.^[55]

That last phrase is the whole of the provenance. An Egyptian origin has been proposed rather than established, and the proposal has been critically reviewed. Garrison holds that the seal shares little stylistically with what he calls the central Achaemenid court style, describes it as remaining an enigma, and suggests instead a western Anatolian origin. A fragmentary impression on the obverse of uninscribed Fortification tablet 1673-201 shows a scene very similar to the London cylinder, but the poor and fragmentary preservation of that impression, catalogued as seal impression 0603, makes it impossible to determine whether the seal was the London Darius cylinder or a different seal carrying a very similar scene.^[56] Sajjad Alibaigi has since argued that the closest compositional parallel is an excavated cylinder from the Iron Age sanctuary of Surkh Dom-e Luri in Luristan, which the editors of the excavation report assign to the Iranian provincial Neo-Assyrian style.^[56]

Alibaigi’s argument concerns iconographic ancestry rather than the identity of the engraver. A pattern can travel without its carver, as the Petosiris relief demonstrates from the opposite direction, and a seal whose composition descends from a Zagros workshop tradition tells nothing about the ethnicity or the location of the craftsman who cut this stone. What BM 89132 does establish is that royal imagery and a royal name could appear together on an object whose owner cannot be identified: the inscription does not show that Darius personally carried or used it. Its evidential value is therefore quite different from that of an impression on a dated administrative document.^[55][3]

Metal vessels, jewellery, stoneware, glass, ivory and textiles extended the court environment beyond architecture. They were used, worn, displayed, exchanged and carried, and the surviving corpus is not a sample of what the royal centres produced or consumed. Precious metal was especially vulnerable: an object could be melted, altered or passed between owners without retaining any trace of its earlier setting, and a great many known pieces reached collections through the antiquities trade. Objects from burials are better placed, since deposition gives a date before which they existed, but not where or how long before they were made.^[4]

Garrison accordingly questions classifications that distribute metalwork confidently among royal, satrapal and local workshops on stylistic grounds. Such schemes describe degrees of resemblance to court imagery; the archaeological foundation for the proposed workshop geography is generally weak, and the possibility of forgery complicates any unprovenanced gold or silver further.^[4] Depictions of vessels on the Apadana offer comparisons rather than certificates: a vessel shown as a delegation’s gift may belong to a widely distributed type, and a provincial imitation may reproduce the form of a prestigious object without its material or technique.^[20][4]

For the grandest lost objects the only evidence is Greek. A golden plane-tree and a jewelled vine recur in classical writing about the Persian court for centuries. Herodotus has the Persians answer Xerxes’ question about the wealthy Lydian Pythius by saying that “he is the man who presented your father Darius with the golden plane-tree and vine”.^[57] Xenophon preserves a sceptic: reporting back from the Persian court in 367, Antiochus of Arcadia said that the vast wealth of the king “was all vain boasting, because even the golden plane-tree, always being harped upon, was not big enough to provide shade for a grasshopper”.^[58] Diodorus records that when Antigonus took over the citadel at Susa in 316 “he found in it the golden climbing vine and a wealth of other finely crafted objects, to a total of fifteen thousand talents”.^[59] Athenaeus, compiling in the Roman period and reportedly drawing on Alexander’s chamberlain Chares, places the vine indoors: it extended over the royal bed and was set with gems. He also preserves Amyntas’s report that its clusters were made from the costliest jewels and mentions a gold mixing goblet standing not far off, crafted by Theodorus of Samos.^[60]

These notices agree on the existence of the objects, and Kuhrt notes that the Oxford Classical Dictionary attribution of the golden plane-tree to Theodorus of Samos, a metalworker active in the late sixth century, is geographically and chronologically possible, though the Athenaeus passage names him only for the mixing goblet and does not settle that attribution. They do not agree on setting or function, and Kuhrt’s verdict is that the precise significance and use of these objects remain unknown.^[60] An ancient sceptic in the dossier is a reminder that Greek accounts of Persian magnificence were contested when they were new.

A second Athenaeus notice describes a ceremonial object that does survive in representation. In Deinon’s account the king neither jumped from his chariot nor leaned on another person’s arm; instead he descended by placing his foot on a golden stool set out for him, whose bearer accompanied him for that purpose.^[61]

A courtier carrying the royal seat or footstool appears on the eastern Apadana stairway, and the Greek title for the office corresponds to an Elamite term attested in the Fortification texts, a case in which classical description, relief and archive converge on a single court function.^[61]

The over-life-size statue of Darius found at Susa in 1972 is a royal commission made in Egypt and later installed in an Achaemenid gateway at Susa. Colburn identifies the stone as Egyptian greywacke and gives its preserved height as 2.46 metres; the body combines Egyptian standing-statuary conventions with Persian court dress and imperial imagery.^[62] Root’s earlier account records its Achaemenid find position against the west wall of a gateway through the fortification, where it had gradually been buried by debris. It is carved in Egyptian stone, with the back pillar and forward left leg of Egyptian standing statuary, a short Egyptian-type stick in the right fist, and what appears to be a flower in the left; the upper third was not recovered, so whether the king wore a Persian or an Egyptian crown is unknown. The body wears Persian court dress with strapless royal boots, bracelets and a dagger. A trilingual cuneiform text runs down the pleats on the right; a hieroglyphic text on the left links Darius to Atum of Heliopolis, and cartouches appear on the ends of the belt. On the base, Nile gods tie together the plants of Upper and Lower Egypt, and twelve fortress rings on each long side carry kneeling personifications of the countries.^[63]

Root in 1979 took the Egyptian manufacture as secure and proposed Heliopolis as a possible original temple setting; she left the circumstances of transfer to Susa uncertain and discussed both an original later moved east and an Egyptian-made copy intended for Persia.^[63] Later work has shifted the proposed original setting. Colburn records that the hieroglyphic inscription repeatedly invokes Atum of Tjeku, another name for Pithom at Tell el-Maskhuta, and treats this as evidence that the statue originally stood there in a temple of Atum. The date and circumstances of its removal to Susa remain disputed.^[62]

The combination suited more than one requirement at once. The king appears in unmistakable Persian dress and in a monumental form intelligible in an Egyptian religious setting, addressed in both scripts to both audiences. The statue joins local Egyptian royal form to Persian dress and imperial imagery within one commission. Colburn connects that dual design to Pithom’s position on Darius’s Red Sea canal, while the later transfer to Susa gave the same object a new architectural setting.^[62][8] The statue also carries the surface evidence Razmjou recorded: brown-red paint on the side folds, no pigment yet found on the inscribed folds, and the samples still unanalysed when the note was published.^[45]

The contents of an Achaemenid palace are not the same thing as its production. Schmidt states the point about Persepolis directly: the Treasury, which held most of the objects he published, “was the depository of many pre-Achaemenid treasures looted from Egypt and Mesopotamia, as indicated by votive objects and vessels inscribed with the names of kings who ruled those countries before Persepolis was founded”, so that many objects in the collection probably antedate the site where they were found.^[17]

One of them has generated a dispute worth following, because the disagreement is about the same stone. A marble torso of the type known as the Penelope was found in the Treasury, its right hand recovered near the north-east corner of Hall 38 and the torso in a corridor of the same building. Cleta Margaret Olmstead assigned it to the mid-fifth century, took it for an Ionian statue of Penelope type and perhaps the original of the numerous later copies, and related its white marble, speckled with glittering crystals, to Aegean island quarries; Paul Jacobsthal, agreeing with John Beazley, told Schmidt in 1947 that in his view too the Persepolis statue was the original, or one of the originals, of the type. Ernst Langlotz proposed instead that it is a copy, carved hardly before the end of the fifth century, of a statue wrought presumably in bronze about 460 BCE, and identified the original as a town goddess, the nymph Larisa, by Telephanes of Phokaia. Schmidt, printing both, expected further and variant interpretations.^[17]

Schmidt presented the identification and date as unresolved between the positions he quoted. The object is instructive under either reconstruction. A Greek statue in a Persian treasury shows that gift, booty, transfer and collecting are all possible routes into a royal building, and that the archaeological find alone selects none of them. The route determines what the object can be used to argue: an original Ionian work carried east is evidence about acquisition, a late copy is evidence about continued traffic in Greek sculpture, and neither is evidence about the sculptors of the terrace.^[17]

Loss has shaped the modern account of Achaemenid art as much as survival has. Roofs, plaster, textiles, wooden objects and most metal fittings have gone or survive in fragments, so that the stone of Persepolis occupies a far larger place in present understanding than it did in the complete environment; Susa’s glazed brick preserves a different portion of the same distortion. Reliefs were prised from their settings, fragments entered collections separately, and details recorded at the moment of excavation were sometimes lost soon afterwards, as happened repeatedly with pigment.^[1][7]

The recovery methods are correspondingly documentary. Joins, excavation records, early drawings, restoration archives and conservation histories can restore architectural relationships that the objects themselves no longer carry, as the identification of the Treasury panels as the Apadana central fields shows. A modern photograph establishes present condition and omits what was recorded when a surface first emerged; the reverse is true of an early drawing, which may preserve what has since faded and may also silently improve it.^[23][7]

Any synthesis of the imperial style depends on the character of its surviving sample. Monumental stone, archive sealings and unprovenanced metalwork have different strengths and different silences: the first may be dated by inscriptions, archaeological context and style in varying combinations; the second is dated and located but rarely preserves the seal object itself; the third can be visually rich while lacking recoverable history. Treated as interchangeable illustrations they produce a composite Persian court whose details never coexisted. Where the dates, locations and conditions of use can actually be established, comparison across the three classes rests on something; where they cannot, the composite is a modern artefact.^[50][4]