Much of the medieval Persian epic tradition consists of retellings of the stories of major epic heroes such as Alexander the Great and Bahrām Gur. One such story is Bahrām’s hunting expedition, in which the monarch kills his favourite concubine when she fails to sufficiently appreciate his skill in archery The moral of the locus classicus of the story in Abu al-Qāsem Ferdowsi’s Shāhnāmeh (Epic of Kings) is that social hierarchy must be maintained, and women have no place in matters of war (Meisami 1989, 41–45). Both the moral and content of the story were transformed in later retellings (Meisami 1989; Bernardini 1992; Gabbay 2009). Nezāmi (d. 1209), Amir Khosrow (d. 1325), and Hātefi (d. 1521) each retold the story, renaming the concubine and changing her from a cautionary example into an increasingly magnificent protagonist, whose feats of strength, skill, or craftsmanship match or—for Amir Khosrow and Hātefi—outshine the martial power of the king himself (Gabbay 2009; Bernardini 1992). Nezāmi has her carry a bull upstairs, while Amir Khosrow has her put animals to sleep with her music (Gabbay 2009, 683, 689). Hātefi has her (and her daughter) build a palace (Bernardini 1992, 40). By the fifteenth century, these tales had become part of a political discourse critical of how rulers treated architects and artisans, and by extension, urban populations (Bernardini 1992, 36–38). The story as retold by later poets undermined social hierarchy, making it a later expression of a ‘pious egalitarianism’ rooted in the early Islamic period and associated with resistance to political authorities (Marlow 1997, 35–36). These retellings are also meditations on authorship: the concubine’s position relative to her royal master is analogous to that of authors’ positions relative to their royal patrons, and her wondrous feats are, rather self-evidently, the authors’ inventions. Thus, the authorial presence merges with the figure of the concubine in their shared creative acts. In short, Nezāmi and his successors changed the story’s meaning not only by changing the ending, but by changing its epistemic status, from a historical legend demonstrating the character of a well-known Iranian king to a fiction with a protagonist whose name and deeds changed in each retelling.

I argue here that the Kushnāmeh (Epic of Kush), written by Irānshāh ebn Abu al-Khayr between 1108 and 1111 CE, and dedicated to the Seljuq ruler Mohammad ebn Malekshāh, represents an earlier example of such fictionalisation of the epic tradition.Footnote 1 The Kushnāmeh is the tale of a monstrous king, with an overtly fictional frame tale and other parodic elements that signal its fictionality. These parodic elements challenge the authority of the epic tradition, especially Ferdowsi’s Shāhnāmeh, as the basis of a communal narrative—a narrative defining a community in terms of its origins and cultural norms. Many scholars maintain that fictionality in medieval Islamic texts was not recognised as a distinct mode of expression. Thus, fantastical stories such as those of the Thousand and One Nights were classified as ‘absurdities’ (khorāfāt), (Bonebakker 1992, 23). Such disregard did not extend to didactic parables, and Bo Utas associates Nezāmi’s free inventiveness with a tradition of Sufi literature in which parables expressed intuitive truth (Utas 2014, 171, 174–76). However, as the example of the Bahrām Gur story demonstrates, the fictionalisation effected by Nezāmi and his successors is not simply a matter of telling parables in the service of a higher truth. Their message and medium are not so distinct. In contrast to the free inventiveness of these later metaphysical romances, the Kushnāmeh’s fictionality is authorised by selective verisimilitude and by appeals to reason and piety. The present discussion of its fictionality and metatextuality offers insight into the textual and social dynamics at work in later Persian literature.

Literary and social context

The Kushnāmeh belongs to a genre of epics and epic romances about the kings and heroes of Persian legend, the most famous literary progenitor of which is the Shāhnāmeh of Ferdowsi, completed in 1010 CE. While the Shāhnāmeh is celebrated as a monument of Persian literature, its content is also historical. Its kings are named in Arabic histories alongside those of other peoples such as the Romans and Arabs. As a Persian epic about pre-Islamic Iranian kings, the Shāhnāmeh is a nostalgic monument to the Sasanian imperial tradition. As such, it fed into a larger political trend. While the Sasanian Empire was destroyed by the Arab conquests in the mid-seventh century, the empire of the ‘Abbasid caliphs (750–942) revived aspects of Sasanian administrative and courtly traditions, and its successor states in Iran and Central Asia extended this revival to the Persian language—New Persian, a form of Persian written in the Arabic script. The Shāhnāmeh is one of the oldest extant works in New Persian, and a communal narrative at whose core is a cycle of the rise and fall of Iranian world-kings, culminating in the tragedy of the Arab conquests.

A principle central to this narrative is nezhād, a unitary concept comprising both ethnicity and aristocratic lineage, which manifests as an individual inheritance that largely determines a character’s conduct and fate (Hayes 2015, 370–76). Thus, in the Shāhnāmeh, the Arab conquests are perceived by the Iranians as a tragedy because they will allow the ascendancy of unworthy people: witnessing the Sasanians’ defeat, an Iranian general predicts that ‘“a talentless vassal will become king”’ (Hayes 2015, 370). Nezhād partly defined the community of Persian-speaking elites, especially the dehqān (rural gentry) class to which Ferdowsi belonged, who justified their social position with claims to descent from pre-Islamic kings and heroes. Their oral storytelling tradition was the direct or indirect basis for much of the content of the Shāhnāmeh (Davis 1996; Hayes 2015, 369–75). The Shāhnāmeh is ‘a work produced by an entire community’ (Hayes 2015, 370–71), thus, there was a strong mutual dependence between this narrative tradition and nezhād as a living social principle. Ferdowsi places the principle of nezhād in tension with more universalistic ideals, suggesting his own writing was constrained by his project of preserving traditional narratives (Hayes 2015, 370–71; Gabbay 2021).

The Iranian kings’ authority in the epics derives also from their dynastic creed, which in the Shāhnāmeh manifests as a generic monotheism. In the Kushnāmeh, this takes the concrete form of a set of ethical teachings, a credal inheritance that mirrors the Shāhnāmeh’s assertions of authenticity. Ferdowsi claims that his text was based on a compendium of ancient history compiled by a dehqān who sought ‘elder sages from every land,’ asking them ‘how first kings ruled the world, and how they passed it on to us in such a sorry state’ (Ferdowsi 1987, 12).Footnote 2 Irānshāh likewise claims that the Kushnāmeh was based on a prose story that he put into verse (130); however, throughout the text, he refers to his narrative as being the words of an elder storyteller (jahāndideh, lit. ‘experienced one’) with phrases such as ‘the experienced one relates that…’ (2399, 5345, 7332, 7903, 8735, 10010, 10111).Footnote 3 These references do not demarcate sections of the text, but function as ambient reminders of its putative origins. While such references to a storyteller are formulaic, and both texts undoubtedly draw on written sources, authentic knowledge of an oral tradition is clearly important to their authoritativeness and to their status as communal narratives.Footnote 4 Thus, credal lineages in both epics mirror their own claims to authenticity.

The Kushnāmeh: plot and parody

The main narrative of the Kushnāmeh follows the exploits of Kush (referred to here as Kush the Tusked when it is necessary to distinguish him from several other characters in the epic named Kush), an evil, monstrous king with tusks and ears like an elephant, who rules China during the reign of his uncle, the arch-tyrant Zahhāk.Footnote 5 Kush’s enemy in the first half of the epic is Ābtin, a descendant of Jamshid, the greatest world-king in the Iranian epic tradition. Ābtin and the other Jamshidians (descendants of Jamshid) hide in the forests of China, until they are given refuge in Besila (a corruption of Silla, an older dynastic name for Korea) where Ābtin marries the king’s daughter. Upon their return to Irānzamin (present-day Iran, Iraq, and Afghanistan), they have a son, Faridun. Faridun overthrows Zahhāk, goes to war with Kush, and captures him. Faridun later releases Kush to fight invaders from Beja and Nubia, and Kush becomes king of North Africa and Iberia. The cruel and violent Kush is physically and morally transformed near the end of the epic. No details are given about his death, but he lives for one thousand and five hundred years, very long even for an epic hero. The Kushnāmeh’s main narrative runs parallel to that of the Shāhnāmeh and largely avoids retelling events from it, except where Kush’s story intersects with those of the main figures of the epic tradition. The timeline of the Shāhnāmeh serves as a background against which the story of Kush takes place.

Following a prefatory invocation typical of Islamicate texts (1–229), the Kushnāmeh’s narrative begins with the dictation of a forged prophecy, whose words begin the first of two frame tales preceding the main narrative (230–45). The abruptness of the opening has been explained as the result of a lacuna in the manuscript. As I shall argue, it may instead be read as an in medias res opening that signals the fictionality of the frame story. Treating this and other apparent seams in the text as interpretable, intentional features is, admittedly, a risky interpretive gambit, especially considering that the Kushnāmeh survives in only a single manuscript—though other versions of the tale of Kush do survive in brief summaries (‘Awliyā and Bashiri 2018). However, as we shall see, some of these features cannot be explained away as scribal errors, and there is other evidence of Irānshāh’s critical attitude towards the epic tradition.

The question of the Kushnāmeh’s fictionality has been addressed by Behruz and Behzād Atuni, who argue that the Kushnāmeh differs from other epics because of the excision or rationalisation of fantastical elements that were part of the narrative tradition, and which are preserved in the Shāhnāmeh and later epics. Such omissions, and the Kushnāmeh’s references to the more widely-accepted Qur’anic historical tradition establishing correspondences with authoritative historical narratives, give it the character of a historical novel (Atuni and Atuni 2019, 116). Irānshāh expresses this rationalising ethos in his retelling of the Shāhnāmeh’s story of Faridun being nursed by Barmāyeh, the peacock-hued cow: ‘People say that [Faridun] drank the milk of the cow, Barmāyeh, but if you hear the story from ordinary people, you won’t hear it told as it happened. That “milk” secretly signifies knowledge.’ (4675–77) The miraculous cow is replaced with Barmāyon, an elderly minister, who educates Faridun. The story of Zahhāk’s illness (in the Shāhnāmeh, serpents spring from his shoulders) is likewise explained as cancer, using Galenic medical theory (2870–80).

Irānshāh did indeed depart greatly from his literary models in rationalising fantastical elements of the narrative tradition. However, the Kushnāmeh does contain a few fantastical elements—most notably, Kush’s own monstrous visage—which present themselves as symbols to be decoded. Surely, if Irānshāh’s primary concern was realism, a monstrous king with tusks and elephant ears is an odd choice of protagonist. To what degree the epic was received as a historical narrative presents another puzzle. The figure of Kush the Tusked has a minimal presence in later Persian histories, with the notable exception of the Mojmal al-Tavārikh (the only text to identify the Kushnāmeh’s author), which also adopts key elements of its narrative of Ābtin (Mojmal al-Tavārikh 2020). While the Kushnāmeh positions its title character as a major rival of the Iranian world-kings, this view finds no precedent in earlier histories, and was evidently not adopted by later historians. Later geographies’ association of Korea with Kangdez suggest the Kushnāmeh was influential in other ways—though it is also possible both Irānshāh and the geographers drew on the same lore about Korea (Vosooghi 2018, 71).

One solution to these puzzles is to recognise a parodic dimension in the Kushnāmeh, following Linda Hutcheon’s broad conception of parody as imitation that engenders critical distance but that is not necessarily comic (Hutcheon 1985). Kush is a parodic counterpart of Jamshid, who in the Shāhnāmeh symbolises the preservation of life, royal charisma, class hierarchy, and Iranian communal identity, as well as kings’ blasphemous claims to divinity. The Kushnāmeh offers an ironic performance of realism and authoritativeness by centering a monstrous protagonist in a story otherwise mostly devoid of fantastic elements. We may thus consider it to be a metafictional text that reframes the epic tradition as a whole as fiction and whose own truth status is accordingly ambiguous. This parody of the epic tradition and its claims to authenticity have a clear social meaning. Parodic rewritings of communal narratives can attempt to codify memory of social change by telling us who we no longer are. The Kushnāmeh positioned itself against the narrative tradition preserved in the Shāhnāmeh, and against an imagined pre-Islamic past, as if to tell its audience that, just as their community no longer worshipped idols, revered emperors as divine, or believed stories of a king who nursed from a peacock-hued cow, neither should they believe that a few noble lineages were uniquely virtuous and fit to rule. The Kushnāmeh told its readers how they once, but no longer, knew their place and their history, and how they might know them and write them anew.

Paradox in the frame tales

Each of the Kushnāmeh’s two frame tales corresponds to a moment of cultural contact: the Sasanian and Islamic empires’ acquisition of Greco-Roman knowledge and Alexander’s legendary conquests. The epic’s main narrative is introduced through a frame tale of Alexander’s travels to the eastern sea, where he fights vicious dog-headed men and then sees a statue of Kush the Tusked bearing an inscription on its hand: ‘This is a statue of Kush, of twisted character, the elephant-tusked king and commander of China… Three times five hundred [years] I walked the earth… All that I ruined was made right again… Now my organs have rotted: the fate of all men’ (575–90). Alexander then finds peaceful horse-headed people and asks them about Kush, but is unable to understand their speech. He prays and becomes able to understand them (600). They lead him to a Jamshidian sage on a mountaintop, who gives him a book containing the epic’s main narrative. This sage, preserver of ancient lore, mirrors the ‘experienced’ storyteller invoked throughout the text as Irānshāh’s source, noted above.

The frame story of Alexander is itself introduced by another frame tale, of a vaguely monstrous Kush, king of ‘Irānzamin’ ruling from Baghdad—a name also used for the Sasanian capital, Ctesiphon. This outermost frame tale opens with this Kush (henceforth, Kush of Baghdad) preparing to plant a false prophecy to fool Mānush, Caesar of Rome. Its first diegetic words recount Kush dictating the counterfeit prophecy to be planted outside the gate of Constantinople:

See, now, what Kush did upon the earth, leader of steel-clad heroes, two eyes like the sky and a face like boiling blood, his visage worse than an elephant’s:

  ‘Three hundred and some years after today, from Iran, a man will come like no other. Every king who follows his command will hold onto his crown and his life. If he leads an army to Rome, my descendants will no longer rule there. Ruin will befall anyone who goes to war with him. Give him whatever treasure he wants, so that he leaves without causing injury or hardship.’

  Once Kush had recorded this message, he rolled it up and placed his seal on it. That devious king drew the face of [Mānush’s ancestor] Dārnush upon that seal, gave the writer a sum of dinars, and bound his tongue with an oath… (230–45)

Mānush finds the jug, believes the prophecy, and surrenders treasure to Kush, including giving his brother as a hostage and renovating mosques in Roman domains. The brother has Mānush send Kush books, including a story of Alexander, which is the inner frame tale, described above.

Kush of Baghdad, with ‘a face like boiling blood,’ ‘eyes like the sky,’ and ‘a visage worse than an elephant’s’ at first appears to be the title character of the epic. Yet this same character, through the window of the next frame tale, reads about a statue of ‘Kush... the elephant-tusked king of China… whose organs have rotted.’ The paradox is heightened by the frame stories’ strong temporal structure, each subsequent frame belonging to an earlier and wholly distinct epoch. The reference to mosques unambiguously places the outermost frame story in the Islamic era. The first frame shift takes us back to the time of Alexander. The main narrative, of Kush the Tusked, takes place in the time of the Pishdādiān and Kayānid kings, with references to the future deeds of Alexander near the end of the epic indicating that this narrative terminates before Alexander’s time (10082). This temporal progression is reinforced by the thematic and narrative continuity between the frame stories and the main narrative, giving the whole text a mobius-strip-like quality.

Many details of the outermost frame story draw attention to its post-Christian chronotope and to old texts as sources for knowledge of the past. Indeed, most of the description of the booty Kush obtains from Mānush is devoted to the books, which are strongly Greco-Roman and Christian in character: ‘the sixteenth book of Galen,’ the writings of Hippocrates and Appolonius of Tyana, histories of Christian kings, the story of the Seven Sleepers of Ephesus, and a story of two disciples of Jesus, Andreas and Simon the Zealot. The story of one Christian king, Stalinas, and the stories of the Seven Sleepers and Andreas and Simon are summarised as interpolated tales. The last book described becomes the frame story of Alexander which, even in its strange, remote setting, centres the transmission of knowledge, as Alexander communicates with the horse-headed people and receives a book from the Iranian hermit, containing the story of Kush the Tusked.

Saghi Gazerani argues that Irānshāh added the frame tales to make the main narrative, originally derived from Sassanian conflicts with the Kushan empire and thus obsolete in Irānshāh’s own time, relevant to his audience (Gazerani 2019, 864). Noting that the period in which the Kushnāmeh was written saw the production of influential encyclopedic works, Gazerani argues that ‘Irānshāh, too, wanted to cast the story... as part of a large encyclopedia of knowledge bequeathed by the Romans’ (Gazerani 2019, 866). That Irānshāh wrote the frame stories seems certain. However, if associating the content of the Kushnāmeh with the catalogue of Greco-Roman books in the frame story could invest the epic with intellectual authority, attributing the recovery of this ‘encyclopedia’ to a king resembling the epic’s monstrous title character hardly reassures us about its provenance. Furthermore, Muslims believed that Christians distorted their scripture, and that philosophy was a fallible, human invention. Accordingly, Irānshāh’s prefatory invocation, which surveys the cosmos from roses to constellations, refutes astronomers’ and philosophers’ authority to explain natural phenomena: ‘Neither the astronomer nor the naturalist (tabaye‘-pazir) offer due thanks to the Lord, for what does a substance know of good and bad?’ (26) and ‘Philosophy has drawn you toward exile in damnation...’ (29). Greco-Roman knowledge was valuable, but not authoritative. By invoking it, the frame stories do not assert authority but emphasise mediation and contingency—not only over time but across the frontiers of rival empires.

References to Kush the Tusked’s descendants in the main narrative might explain the paradox of Kush seeming to read about his own death: Kush of Baghdad may be Kush the Tusked’s descendant through Nimrod. However, this is never stated directly, and readers would not become aware of the relationship until mid-way through the epic. While it thus may be tempting to posit that missing verses at the beginning of the frame story identified Kush of Baghdad in this way, there remains the larger problem that no king named Kush ruled Irānzamin in the Sasanian or Islamic periods. Even the Mojmal al-Tavārikh, which does mention Kush as king of China, mentions no post-Alexandrine Kush (Mojmal al-Tavārikh 2020, 104, 123, 309, 313). The name ‘Kush’ would have seemed unduly strange for a king of Irānzamin, and likely read as intentionally fictional, much as we might react to a story set in New York under a President Ebenezer Prendergast. Such a picturesque name, evidently not an actual president, evokes either an alternative past in the manner of counterfactual historical fiction, or a more distant, unfamiliar era. The frame story thus immediately demands suspension of disbelief.

However, the name ‘Mānush’ appears to reference Romanus IV Diogenes, the emperor defeated by Irānshāh’s own patrons, the Seljuqs, at the Battle of Manzikert in 1071. Names in the Kushnāmeh derive largely from the Persian epic tradition, and there is no Roman emperor Mānush in the Shāhnāmeh or other familiar Persian epics. Arabic- and Persian-speakers may have read the name ‘Romanus’ as a portmanteau of ‘Rum’ (Rome) and ‘Manus’; in Persian, -ush is preferred over -us as a final syllable. Thus, a name that read as Rome-Manus became Mānush, Caesar of Rome.Footnote 6 Romanus did not have a great presence in Persian or Arabic historiography, so this might also signify a generic Roman emperor. The strange figure of Kush, king of Irānzamin, and the use of the name Mānush make the chronotope of the outermost frame story ambiguous. It has the form and verisimilitude of a historical tale, but its protagonist cannot be historical.

Kush appearing to read about his own death in the frame stories also contrasts with the strict linearity of the main narrative, which focuses on relatively few characters, and in which almost every situation is the direct outcome of their past interactions. This differs from the more episodically-structured epics of Garshāsp or Alexander, in which the order of episodes could, hypothetically, be changed without major inconsistencies. In spite of this tightly-structured linear plot, events in the epic’s main narrative are repetitive: Kush fights three armies sent by Faridun before his capture, and afterwards, three invasions from Nubia and Beja, and two more armies led by Salm and Manuchehr. Both as ruler of China, and later as ruler of the west, Kush builds a great city, attempts to marry his daughters, demands hundreds of young women be brought to him, and commands his subjects to worship him and fashion idols.

The progression of the plot and passage of epochal time makes these repetitions conspicuous, signs of Kush’s monstrous longevity. Near the mid-point of the epic, Irānshāh makes a comment expressing disagreement with the elder storyteller’s view that Kush the Tusked died in prison and that the Kush freed by Faridun was actually his son, who resembled the father (7530–7536). In one verse, during Kush’s reign over the West, he refers to him as the grandson of Kush the Tusked, an inconsistency Gazerani identifies as evidence of Irānshāh’s having combined texts into one epic (Gazerani 2019, 885–86). If Irānshāh forged a grand antiheroic narrative from multiple sources, he did not, in the end, combine them into a seamlessly-integrated plot. Not only is Kush’s lifespan preternaturally long, as we are told by the inscription on Kush’s statue, but by acknowledging disagreement over whether the same Kush ruled both China and the west, Irānshāh draws attention to the seam joining the segments.

Nezhād in the Kushnāmeh

With its narrative centering conflict between descendants of Jamshid and Zahhāk, respectively the most glorious and most evil figures in the epic tradition, the Kushnāmeh is even more structured by nezhād than other epics. However, in the epic’s Seljuq political context, the opposition between a monstrous king of the East and descendants of the true king in hiding resembled the conflict between the Turkic Seljuqs and their Shi‘i opponents, especially the Isma‘ilis. And as noted above, the name Mānush resembles that of the Seljuqs’ enemy and thus connects Kush of Baghdad with Irānshāh’s own Seljuq patron. The Seljuqs’ Central Asian origins connected them to Turān, the eastern part of the world in the Shāhnāmeh’s geographical schema, encompassing Central Asia as well as China and other lands to the East—all (except Besila) ruled by Kush in the first part of the Kushnāmeh. In many stories of the Shāhnāmeh, the Turānians are the principal enemies of the kings of Irānzamin. While Turks were portrayed as possessing seductive beauty (as well as violent power), some Islamicate texts in this period depicted eastern steppe peoples as physically grotesque (Frenkel 2005; Brookshaw 2009). This political context makes it difficult to accept that the epic’s moral sensibility is as simple as the moral polarity between the Jamshidians and Zahhākians.

The Kushnāmeh orients itself against the ethnocentrism inherent in the principle of nezhād by centering the bonds of fealty between Ābtin and Tayhur, king of Besila. The narrative of Ābtin, and the strongly favourable characterisation of both Korea and the Iberian Peninsula, valourise hybridity and travel to the distant corners of the world (Hemmat 2019, 37). The messianic Faridun is born to a princess of Besila, not an Iranian mother—in contrast, Kush twice attempts incest (4800–4910, 9560). Ābtin is associated throughout the epic with travel and hybridity. His story closely prefigures that of Siyāvash, perhaps the single most sympathetic character in the Shāhnāmeh, the tragedy of whose death resulted from the failure of the kings of Iran and Turān to overcome mutual enmity. Here, Ferdowsi already challenged the ethnocentrism of the narrative tradition (Gabbay 2021).

The Kushnāmeh’s treatment of class hierarchy is more ambiguous. The Shāhnāmeh followed the Avestas (Old Persian Zoroastrian scriptures) in attributing to Jamshid the division of humanity into a hierarchy dominated by priests and warriors (Ferdowsi 1987, 42). Kush and his allies mock Faridun’s greatest general, Qāren, son of Kāveh the Blacksmith, about his humble origins (‘Has blacksmithing made your brain swell with pride?’), asserting this principle of hierarchy (9002; see also 8376, 8929, 9182). Qāren is victorious in these encounters, but makes no pretense of becoming king. Kush invokes his connection to the royal lineage of Zahhāk, as do Faridun’s sons Salm and Tur, whose mother is Zahhāk’s descendant (8706, 9010, 9345). Enforcement of class hierarchy becomes a sign of Kush’s uncouthness.

Curiously, the last great offence of Kush the Tusked as king of China, the one which provokes a final, effective response from Faridun (who had already attempted to subdue him twice, and given up) was confiscating wealth to make the land prosper: ‘wherever was ruined, the experienced one used his wealth to make it prosperous,’ and ‘having cleaned out his treasury in this manner, he took many valuable things from people… anyone with wealth, he left with neither wealth nor peace. He made the world prosper from them and their treasure’ (6799–6806). The phrase ‘made the world prosper’ generally indicates the building of cities, roads, fortifications, or agricultural improvements. Irānshāh elsewhere graphically describes the misery resulting from sieges and invasions but scarcely elaborates on the effects of Kush’s final act of misrule. Was this more self-evidently abhorrent than his gruesome mistreatment of his children, which elicits sympathy even from Zahhāk (4800–4948)? We may read a subtle irony into this laconic description of Kush’s misrule and its consequences. Rather than a full-fledged tragedy in its own right, the main significance of this tersely described event is to mark Kush’s confiscation of wealth as his ultimate point of disagreement with the Jamshidians.

Kush the Relentless: a parody of Jamshid

As an antihero, Kush is an inherently parodic character. His monstrous longevity is both formally and thematically central to the parody. He plays all the roles taken on by epic heroes and kings: fighting on the battlefield, concocting stratagems, ruling, and (mis)guiding his subjects in matters of religion (4778, 6795, 8445, 9570, 9675). He even takes on the roles of lover and parent, albeit with horrible consequences. Kush’s exploits are thus a dark reflection of the epic tradition as a whole. In the end, Kush’s longevity and immunity to illness are turned against him, making his entrapment in the forest a potentially eternal torment. His position as a Methuselan paramount antihero connects him specifically to Jamshid, who was associated with the preservation of life. In the Shāhnāmeh, there was no death or disease during Jamshid’s reign. While Jamshid’s claim to divinity brings about his downfall, Kush continues claiming to be Creator, citing his own longevity as evidence, until the moment he repents.

Some important elements from the Jamshid cycle also occur in the Kushnāmeh’s story of Ābtin and the Shāhnāmeh’s story of Siyāvash, which share core thematic components. In the Avestas, Jamshid creates a sanctuary which preserves people and specimens of useful plants and animals through a cataclysmic winter (Skaervo 2012). This feat is absent from the Shāhnāmeh, but it may be related to a brief episode in its account of Alexander, who immediately after giving up his quest to find the water of life, encounters a crystalline palace containing a light-emitting gem and the corpse of a boar-headed man seated upon a magnificent throne (Ferdowsi 2005, 100–101). This episode may be related to the Kushnāmeh‘s Alexander frame story and closely resembles a brief episode near the end of the epic: Kush builds a dome with a wondrous crystalline chandelier and a crystalline statue of himself in the city of ‘Zorosh by the sea’ (perhaps Torrox, Spain); Alexander destroyed the statue but left the dome (1075–85). This last part of the epic, when Kush is a just king, also describes other wondrous monuments that outlast their creators.

The Avestan Jamshid’s vault has important parallels with, and likely represents an older form of, Kangdez, a city founded by Siyāvash in the Shāhnāmeh. Irānshāh explicitly connects Kangdez with Besila, adding a cryptic remark about his holding an unorthodox view that the earth was inhabited before Adam (2399–2413). Al-Bīrūnī identifies Kangdez with a place Indians call ‘Yamkoti,’ adding: ‘Koṭi means “castle” and Yama “the angel of death,”’ and that ‘Kangdez… according to the Persians, had been built by Kaikāvus or Jam[shid] in the most remote east, behind the Sea’ (Vosooghi 2018, 74). Yama is a Sanskritic cognate of Jamshid, though al-Bīrūnī does not mention this. Thus, narratives of Jamshid and Siyāvash (in the Shāhnāmeh) and Kush and Ābtin (in the Kushnāmeh) share a set of core elements that were essential to their meaning, which identify them as variants on a common form. By centring these themes in the stories of Kush and Ābtin, Irānshāh was writing in the symbolic grammar of ancient Iranian myth.

In contrast to the death-defying power of Jamshid and Kush, who represent preservation and longevity, Siyāvash and Ābtin symbolise renewal and change. While Jamshid in the Shāhnāmeh is a tragic figure who falls from grace by claiming divinity, precipitating struggles to restore the Pishdādian lineage and to maintain rightful succession within it, Kush is tragicomic, living a life of unending sin until his abrupt transformation. We have thus arrived at an explanation for the frame story: the punchline, as it were, is that Kush always returns, even after death, escaping the boundaries of the text and projecting his sinister presence into the reader’s world.

Roughing up Jamshid’s creed

Just as this parodic interpretation of the epic supports reading its abrupt beginning with the dictation of a false prophecy as an intentional in medias res opening, it also invites us to reconsider its two expositions of the Jamshidians’ creed. Like the subsequent frame transitions, which occur with the opening of books, and emphasise the characters’ and the reader’s dependence on ancient texts for knowledge of the past, the in medias res opening colours the narratives to come by reminding us that received narratives may be false, even malicious forgeries. Such ironic accounts of provenance are attested in earlier texts, such as al-Isbahānī’s (d. 971) preface to the story of Laylā and Majnūn (Beck 2018).

The epic’s abrupt, cautionary opening is mirrored by its equally abrupt, Sufism-inflected ending, which is unique to this version of the tale of Kush and thus almost certainly authored by Irānshāh (‘Awliyā and Bashīrī 2018). In his confrontation with the hermit, Kush claims to be Creator. The hermit then asks why Kush cannot change his own face; surely, these monstrous tusks and elephant ears must be his own will. This provokes a heartfelt, if brief, reflection: ‘Since my fundamental nature is this’—or, to translate the phrase more literally, ‘with a nature that emerged in this way from its fundaments’—‘how could I not follow a foul creed? I would be satisfied if I had no treasure and my heart did not suffer from this ugliness!’ (9799). But Kush immediately repents and submits to the guidance of the hermit, who imposes an ascetic regimen on him and shrinks his tusks and ears. Kush then returns to his throne and is henceforth a just king. Having invoked nezhād as the rationale for his own conduct, Kush immediately proves it to be an arbitrary distinction, and in so doing, falsifies the main structuring principle of the narrative. Kush was fundamentally a monster, at war with the fundamentally good Jamshidians—but suddenly, we learn, a monster need not be a monster.

The Jamshidians’ authority derives partly from their dynastic theology, which the hermit teaches to Kush. And yet, the Kushnāmeh’s two expositions of the Jamshidians’ creed contain discrepancies that give them the appearance of being rendered carelessly. These expositions are catalogues of virtues and vices, including a list of Pahlavi (Middle Persian) terms with New Persian glosses, all delivered through dialogue. The first set of dialogues are derived from an extant Pahlavi text, known as the ‘Testament of Bozorgmehr’ (yādgār-e bozorgmehr), a smaller part of which is also translated in the Shāhnāmeh (Matini 1986). Ābtin’s minister, Kāmdād, describes ‘seven treasures’: ‘wisdom,’ ‘good nature,’ ‘hope,’ ‘skill,’ ‘religion [din] itself,’ ‘contentment,’ and ‘an enlightened creed [rāy-e dānā]’ (4380). Soon after, the learned elder Barmāyon states that he will name ten demons (i.e. vices), then names eleven, and when Ābtin’s minister, Kāmdād, responds by asking him about them, he asks about only eight, including one not in Barmāyon’s original list (4340–4600). In the second exposition, the Jamshidian hermit tells Kush of ‘seven treasures—and I do not know an eighth’: ‘wisdom,’ ‘learning,’ ‘cleverness,’ ‘righteousness,’ ‘purity,’ ‘kindness,’ with the seventh being ‘justice and nobility’ (dād va āzādegi), which are two distinct concepts, not typically paired (9849). That is, he contradicts himself by saying ‘I do not know an eighth’ and immediately listing eight virtues. It is also significant that his list has only the first item, wisdom, in common with the ‘seven treasures’ previously given by Kāmdād.

Apart from the two missing ‘demons,’ these discrepancies cannot be explained as scribal errors, which would have broken the metre. The hermit’s catalogue of vices and virtues appears to be a pastiche of the earlier dialogue between Barmāyon and Kāmdād, which has an authentic Pahlavi source. The contrast between the characters’ reverent insistence that the specific virtues and vices they list are uniquely important, and Irānshāh’s casual treatment of this material, is stark. Irānshāh treats this ostensibly ancient Pahlavi lore more as a discursive style that gives historical flavour to his characters than a creed worthy of preservation. He demonstrates his own mastery of this style by his pastiche in the second dialogue. Bearing in mind the Qur’an’s claim to inimitability, Irānshāh’s pastiche recognises the ‘Testament’ as scriptural in function, while rejecting its authority. The Jamshidian hermit, like the one in the Alexander frame story, and the sage Barmāyon are metatextual figures who mirror the storytellers Irānshāh repeatedly names as his own sources. Irānshāh thus demonstrates a determination both to master ancient lore, and to freely rewrite it.

Conclusion: a Brutalist epic?

Kush’s seemingly post-mortem appearance in the outermost frame story, his long lifespan and other parodic elements of the anti-hero narrative, and the twist ending are simultaneously diegetic and metatextual, dramatising implicit criticisms of the epic tradition. The most striking such dramatisation is the start of the outermost frame story, in which Kush of Baghdad obtains knowledge by means of a forgery. Taking a cue from this story, we may read the Kushnāmeh in its entirety—notwithstanding the authenticity of Irānshāh’s own sources, for which Gazerani makes a plausible case (Gazerani 2019)—as a pastiche of the Shāhnāmeh and other tales of more familiar heroes that, by the act of imitation, denies their authority as communal narratives.

I have also argued that, rather than seeing the various contradictions and repetitions in the epic merely as forensic evidence of Irānshāh having combined multiple narrative traditions, some of these rough edges in the narrative can be understood as stylistic choices. Readers must have recognised that Irānshāh did not view the ‘seven treasures’ as a creed that he needed to faithfully render in verse. This imprecision is all the more striking given the simplicity of his language: his renderings were not constrained by a need to deliver his narrative through complex imagery and word-play. His style is thus a literary analog of Brutalist architecture: its bare, grotesque form looms over us unconstrained by any need to conceal its raw materials, revealing its inner workings to the public while mocking its predecessors’ pretensions to authenticity.

Similarly, the Kushnāmeh’s rough simplicity and its use and decoding of symbols opens up the epic tradition—not only as a source of entertaining stories, but specifically as a vehicle of moral and political wisdom—to a wider audience as authoritative interpreters and potential writers. While verisimilitude authorises fictionality in the modern novel, Irānshāh grounds his narrative in appeals to reason and piety, juxtaposing historical realism with self-evident inventions. But unlike the metaphysical romances of Nezāmi and his successors, which built elaborate fictions upon such familiar narrative ground as the stories of Alexander and Bahrām Gur, Irānshāh’s narrative of Kush is a distorted replica of history. Its selective verisimilitude keeps Kush close to history’s world, and so places the monster right behind us, in our own lineage.