5.1 Rustam in Arabic and Persian Literature

The greatest hero of Firdawsī’s Shāhnāme, Rustam, is sparsely documented from pre-and early Islamic times,¹ but there can be little serious doubt as to his importance in at least the East Iranian world. From the tenth century onwards he became in a short time a national hero, as not only shown by Firdawsī’s Shāhnāme, and its tenth-century sources, but also by the proliferating genre of later epics, largely centred on Rustam and the other Sistanian heroes, much of the material going back to times before Firdawsī (see Chapter 4.7).² The scarcity of extant Middle Persian references to Rustam³ is clearly due to the lack of preserved sources in Pahlavi and/or the fact that Rustam stories continued to circulate in oral transmission as part of the repertoire of storytellers (cf. Chapter 1.4).

Most of the stories of Rustam are linked to Persian national history and are, at least tangentially, related to the material in the Khwadāynāmag. As Ibn al-Muqaffaʿ’s translation has been lost, its contents have to be deduced from later quotations and references and Arabic and Classical Persian sources. One of the open questions is whether and to what extent Rustam and the other Sistanians had a place in the Khwadāynāmag. Another question is when have the two traditions been joined together to form one continuous narrative. These two questions will be discussed in this chapter.

Although Ibn al-Muqaffaʿ’s translation has later disappeared, it was influential in its own time and several centuries thereafter. In order to discuss whether it contained material on Rustam, we have to go through early Arabic sources, or sources that contain early material. Firdawsī became influential in Iran especially in the twelfth century, and also Arabic works written later than that are always open to doubt as to whether or not they have been influenced by material derived directly or indirectly from Firdawsī’s work. Sources earlier than this, in both Arabic and Classical Persian, mainly derive their material from the now lost earlier sources and often differ in details from Firdawsī. Arabic and Persian historical works remained largely untouched by the epic tradition even later, though, and, especially on the Arabic side, Firdawsī’s influence was limited, despite his overwhelming influence on Persian belles lettres from the twelfth century onward. Arabic sources usually circulate material derived from earlier historical works and show only limited marks of borrowings from Firdawsī’s epic, presumably through Classical Persian historical works. On the Persian side, Firdawsī’s influence is stronger, but here, too, many sources prefer the “historical” tradition to Firdawsī’s “epic” tradition.⁴

When going through first-millennium Arabic texts, the first thing that strikes one is how rarely Rustam is mentioned and how little the Arabs seem to have known about him. The list of Arabic sources that completely ignore Rustam is long. To take but a few examples, al-Jāḥiẓ, who is usually well informed about everything, does not even mention him in his main works (Bayān; Ḥayawān; Rasāʾil), and we search in vain for him in al-Iṣfahānī’s Kitāb al-Aghānī. Likewise, Ibn Qutayba, mentions him neither in his Maʿārif, which contains a chapter on Persian kings (pp. 652–667), deriving its material from kutub siyar mulūk al-ʿajam, nor in the ʿUyūn, and al-Thaʿālibī is equally ignorant of him in his Thimār and has little to say about him in his other works. In his Iʿjāz, pp. 32–33, there are some maxims attributed to Rustam (and others to Zāl), but one can hardly recognize Firdawsī’s Rustam from these rather stereotyped sayings that have nothing heroic in them.⁵ Ibn Ḥamdūn, Tadhkira i: 278 (no. 733), only gives a brief saying by an unidentified Rustam (“when you want to be obeyed, ask what can be done” idhā aradta an tuṭāʿ fa-sal mā yustaṭāʿ). Al-Zamakhsharī, Rabīʿ ii: 792, gives the same saying, but attributes it to Isfandiyār.⁶ Al-Thaʿālibī’s Iʿjāz, p. 33, gives us a clue as to how this confusion was generated: there the saying is implicitly attributed to Rustam, who has been identified as the speaker of the previous saying and who gives this piece of advice to Isfandiyār (wa-qāla [i.e., Rustam] li-Isfandiyār).

When one does encounter the name Rustam, it is usually the general of al-Qādisiyya who is being referred to. Zāl, Sām, and the other members of the Sistanian family are equally unknown in these sources. On the Christian Arabic side, the situation is similar: e.g., Eutychius does not even mention the name Rustam.

It is often, but erroneously, stated that Rustam and his deeds were already known on the Arabian Peninsula in the early seventh century and that stories about him were brought there by al-Naḍr ibn al-Ḥārith, who had learned them in al-Ḥīra.⁷ In modern studies, Theodor Nöldeke (1920): 11, n. 5, seems to be the first to mention this, twice referring to Ibn Hishām’s (d. 218/833) Sīrat rasūl Allāh. In Sīra i: 246, Ibn Hishām tells that al-Naḍr ibn al-Ḥārith learned in al-Ḥīra tales of Persian kings and “aḥādīth Rustam wa-Isfandiyār”. In Sīra i: 294, he says that al-Naḍr related stories about the mighty Rustam and Isfandiyār (wa-ḥaddathahum ʿan Rustam al-Sindīd – read: al-shadīd – wa-ʿan Isfandiyār) and the kings of Persia.⁸

In Nöldeke’s time, Ibn Hishām’s Sīra was mainly taken at face value, miracles excluded. Over the last few decades, it has become increasingly clear that historians’ reports on early Islam and the life of the Prophet should not be taken as faithfully reflecting the conditions of the early seventh century, but should be considered products of their authors’ time or, at most, of the eighth century.⁹ Hence, the passages only prove the obvious, namely that Arab scholars of the late eighth, early ninth century knew about Rustam.

How vaguely even later authors probably did this is shown by al-Suhaylī’s (d. 581/1185) commentary on Ibn Hishām’s Sīra, al-Rawḍ al-unuf. The main passage on Rustam comes in Rawḍ iii: 157–160, commenting on Ibn Hishām’s mention of al-Naḍr. In iii: 158, al-Suhaylī writes: “Rustam, who is called the Lord of Banū *Dastān,¹⁰ was a Turkish (sic) king”. Some lines later he adds: “There is also another Rustam who has earlier been mentioned in the stories about Kay Qubād. He lived before the time of Solomon. After Kay Qubād, Rustam was Vizier to his son Kay Qāwūs”. A page later he has this to say (iii: 159–160): “and I do not know whether the Rustam whom (sic) Isfandiyādh killed was the same as the Rustam who accompanied Kay Qāwūs, or someone else (… wa-lā adrī hal Rustam alladhī qatalahu Isfandiyādh¹¹ huwa Rustam ṣāḥib Kay Qāwūs am ghayruhu), but it would seem that he was not, because the period between Kay Qāwūs and Kay Yastāsb¹² is very long. We have already mentioned that he was a Turk”. If anything, these passages show how ignorant the writer was about Rustam.

In Qurʾānic commentaries, Q 31: 6 is understood to refer to this al-Naḍr, and more or less the same scanty information is given in almost all tafsīrs. In some, such as that of al-Bayḍāwī (late seventh/thirteenth century) (Anwār iv: 150), it is further stated that al-Naḍr found the story of Rustam and Isfandiyār and bought it. While seemingly an interesting reference to the story existing in a buyable, and hence written, form, the verb is unfortunately derived from the formulation of the Qurʾān, which is here taken in a literal sense: wa-min al-nāsi man yashtarī lahwa l-ḥadīth (literally: “among people there are some who buy diverting stories”).¹³ The verb is merely copied from the Qurʾān into al-Bayḍāwī’s narrative and the exegetical tradition in general.

It should be emphasized that the fact that Ibn Hishām and the authors of the commentaries knew Rustam and that they connected him to al-Naḍr and the asāṭīr al-awwalīn only shows that they were aware that there were some stories about Rustam circulating in Persian lore. It does not follow that they would have known these stories in any detail.¹⁴ That Rustam was the hero of long stories of the Persians was common knowledge by the end of the eighth century, cf. Chapter 2.2.1 and below.

When we come to historical sources, we find some information about Rustam, but it is still meagre and sometimes disquietingly different from what we might expect on the basis of Firdawsī.

In his al-Akhbār al-ṭiwāl, al-Dīnawarī (d. not later than 290/902–3) first, p. 6, mentions that the Indian King Porus (familiar from the Alexander Romance and other Alexander narratives)¹⁵ and, according to some, Rustam were descended from Ghānim ibn ʿAlwān. On pp. 27–28, he tells that Rustam was the governor of Sistan and Khurasan for Bishtāsb. He was in the service of Kay Qubād and grew furious because Bishtāsb had converted to Zarathustra’s (new) religion and for this reason rebelled. Bishtāsb sent his son Isfandiyādh against him. Isfandiyādh challenged Rustam but was killed by him, and “Persians tell a lot about this” (fa-yaqūlu l-ʿajam fī dhālika qawlan kathīran). The author adds that Rustam died soon after, but gives no details concerning his death. On p. 29, he tells that later Bahman killed those he could of his offspring and family, but again gives no names. Much later, p. 82 (in the story of Bahrām Chūbīn), he lets Bahrām briefly refer to Rustam having saved Qābūs when the latter was imprisoned, but does not mention his role in extracting revenge on Siyāwush’s account. This is all this historian from Dīnawar, in Western Iran, has to tell about Rustam.

Except for a few stray notes on Rustam, al-Dīnawarī concentrates on the battle between Rustam and Isfandiyār, which is typical of most early Arabic historians, as will be seen. Another theme that should be pointed out is the conversion of Bishtāsb to Zoroastrianism, contrasted with Rustam’s refusal to leave his ancestral religion, an event used to explain the falling out of Bishtāsb and Rustam. Later Arabic and Classical Persian sources often elaborate on this and, either implicitly or explicitly, identify this ancestral religion with monotheism.¹⁶

The anonymous Nihāyat al-arab seems to share the same sources with al-Dīnawarī’s Akhbār, but the mutual relations of the two are still unclear.¹⁷ It is evident, however, that they represent traditions that circulated in Arabic before al-Ṭabarī, who, in general, derives much material from the same tradition.

The Nihāya shows that its author was intimately familiar with the battle between Rustam and Isfandiyār. On p. 26, he briefly mentions that Rustam the Mighty (text: rqtm al-Shadīd) fought against Isfandiyār, but on pp. 82–85, he elaborates on this under the heading Ḥadīth Rustam wa-Isfandiyār “The Story of Rustam and Isfandiyār”, given on the purported authority of Ibn al-Muqaffaʿ.¹⁸ The story starts with a clear indication of source, put in the mouth of Ibn al-Muqaffaʿ: “I found in/among the books of the Persians (the story of) the war between Rustam and Isfandiyār” (wajadtu fī kutub al-ʿajam ḥarb Rustam wa-Isfandiyār), as if this were a separate story, as it probably was, cf. Chapter 2.2.1. It should be noted that “the books of the Persians” is an often-used formulation and does not imply that the source was in Classical or Middle Persian. More probably, the expression here refers to books in Arabic by Persian authors. There is no indication that the author would have known Persian.

The story is related in a more extensive form than in al-Dīnawarī’s version, but in a similar fashion. According to this version, some learned Persians claim that Rustam lived in Sistan and was descended from Ṭasm ibn Nūḥ, while others (still Persians?) say that his mother was a Ṭasmī, but his father descended from Nimrod. Bishtāsf converted to Zarathustra’s religion. Earlier he had been imprisoned by a king descended from Ḥām and had been freed by Rustam. Bishtāsf had given Rustam Khurasan and Sistan to rule and had crowned him. But when Rustam heard about the conversion, he became furious and rebelled. Bishtāsf sent his son Isfandiyār against Rustam. Rustam told him that he would fight until Bishtāsf left Zoroastrianism.¹⁹ They fought for 40 days. Rustam made a trick and led his army, against the agreement, into battle against Isfandiyār’s army, but to no avail. Again they fought a duel, in which Isfandiyār shot a thousand arrows at Rustam and all hit their mark. Isfandiyār called to him and suggested they stop for that day.

His horse Rakhsh could not take him over a deep river, so Rustam dismounted. Back home, he attended to his wounds and called for a kāhin. The kāhin predicted that Rustam would kill Isfandiyār, but would himself die soon thereafter. He further told that he would be able to kill Isfandiyār with arrows made of the tamarisk which grew on the island of Kāzarūn. Rustam sent a message to Isfandiyār and asked for a longer respite. Isfandiyār consented to this, and Rustam sailed to an island near Ṭabaristān and got the wood for his arrows. (There is no mention of Sīmurgh, usually called al-ʿAnqāʾ in Arabic sources,²⁰ in the story, nor in the whole book). On the following day Rustam shot three arrows and killed Isfandiyār, whose army returned to report to Bishtāsf. The king died of sorrow, and Bahman ascended the throne. Soon after Rustam had a hunting accident and died in a pit, but it is also said that he died of the wounds caused by Isfandiyār. The killing of his family is not mentioned.

These two sources lead us to the greatest historian of the first millennium, al-Ṭabarī (d. 310/923). The information we receive about Rustam is marginal and strictly centred on the episode of Rustam and Siyāwukhsh.²¹ Ta⁠ʾrīkh i: 598–604//iv: 2–7, is the longest passage on Rustam and it only narrates the episode of Siyāwukhsh (also giving Rustam’s full name with four forefathers between Dastān, i.e., Zāl, and Sahm, i.e., Sām), with reference to “a long story” told about him. Then the text continues with the attempt of Kay Kāwūs to fly and relates how he was imprisoned in Yemen and saved by Rustam. This is partly narrated on the authority of Hishām (ibn Muḥammad al-Kalbī, d. 204/819).²²

The other mentions of Rustam are marginal. Ta⁠ʾrīkh i: 681//iv: 76, tells, on the authority of Ibn al-Kalbī, that Isfandiyār was killed by Rustam, and Ta⁠ʾrīkh i: 687//iv: 81–82, that Bahman slayed Rustam, Dastān, Azwāra, and Farāmarz. The only remaining reference to Rustam in the whole Ta⁠ʾrīkh comes in ii: 1154//xxiii: 98, where a mighty warrior is first compared to Satan and then to Rustam.

The Persian translation/reworking of al-Ṭabarī’s Ta⁠ʾrīkh by Balʿamī (d. towards the end of the tenth century) is hardly more informative, even though its author had at his disposal Persian works belonging to the Book of Kings tradition (Chapter 3.6). His unwillingness to provide more material on Rustam hardly depends on his wish to follow al-Ṭabarī here more closely – elsewhere, he freely adds material from Persian and other sources – but on the fact that he had little additional material at hand. Whatever the reason, it proves that in Bukhārā, where Balʿamī wrote (or partly commissioned) his work, Rustam was not the central character of national history: Balʿamī’s Sāmānid patron Manṣūr ibn Nūḥ obviously did not expect him to deal any more extensively with Rustam.

Balʿamī concisely narrates the following episodes related to Rustam’s life: Siyāwukhsh (pp. 419–421); Kay Kāwūs in Yemen (pp. 422–423); Rustam kills Isfandiyār (pp. 468–469); and finally, with explicit reference to al-Ṭabarī (p. 482), Bahman’s killing of Rustam’s father and brother. A couple of lines earlier, based on Kitāb-e Akhbār-e ʿajam, Balʿamī had told that Rustam had already been killed by a brother of his, which, unsurprisingly, shows that Firdawsī did not invent this motif but that it was already in circulation in the tenth century.

Other early Arabic historians also indicate that Rustam was strongly present only in the episodes concerning Siyāwush and Isfandiyār. Al-Maqdisī (d. after 355/966), a very well-informed historian, who used native sources (Chapter 3.6),²³ is only slightly more informative. In his Badʾ iii: 147–148, under the title “The story of how Rustam saved Kay Kāwūs”, he tells how the latter was imprisoned by the Ḥimyar. Rustam came from Sistan with a great army and asked al-ʿAnqāʾ (i.e., Sīmurgh) for help. The bird gave him one of his own feathers and promised to come if Rustam were to burn it. The Ḥimyarī king had, by magic, suspended his town between heaven and earth. Rustam called al-ʿAnqāʾ to help him and the bird took his horse in his claws and let Rustam ride on his back. Thus, he took Rustam to the town, where Rustam rescued Kay Kāwūs from the pit, taking also Suʿdā (Arabicized for Sūdābe) back to Babylon. Then the author briefly refers (Badʾ iii: 148–149) to the story of Siyāwush and Suʿdā, which, he says, is like that of Joseph and Zulaykhā. Siyāwush is imprisoned, and Rustam comes to kill Suʿdā. (There is no mention of the Turkish adventures of Siyāwush, except that he was killed in the land of the Turks.) The passage ends by throwing doubt on the credibility of the story of al-ʿAnqāʾ, wa-llāhu aʿlam.

Even the best authority on pre-Islamic Iran, Ḥamza al-Iṣfahānī (d. 350/961 or 360/971), almost completely ignores Rustam in his Ta⁠ʾrīkh sinī l-mulūk, which was written on the basis of several versions of the Arabic translations of the Khwadāynāmag and other historical works (Chapter 3.6). In the chapter on the South Arabian kings (not the Persians), Ḥamza only mentions (p. 101) that the South Arabian Shammar-Yarʿash was, according to some, killed by Rustam ibn Dastān. It is indicative that the focus here is on the South Arabian king, not Rustam. This absolute paucity of Rustam material is significant since Ḥamza seems to have followed very closely the Arabic translation(s) of the Khwadāynāmag, on which he is our most reliable and best-informed authority.

Another usually well-informed author is Miskawayh (d. 421/1030), whose Tajārib again provides meagre results concerning Rustam. Tajārib i: 70–72, resumes the story of Kay Qābūs, Siyāwukhsh and Rustam: Rustam educates Siyāwukhsh (i: 70). Siyāwukhsh implores Rustam to ask Kay Qābūs to send him to fight against Afrāsiyāb (i: 71, as in Firdawsī, but this detail is lacking from al-Ṭabarī, one of Miskawayh’s sources). When Bīb (= Gīw) brings Kay Khusraw to Iran, Rustam comes with an army to welcome him and in several battles defeats the Turkish forces that had followed the fugitives (i: 72). Finally, Rustam saves Kay Qābūs from Yemen. This is the longest passage on Rustam in Miskawayh’s work, but there is also a reference to the Persians telling stories about Rustam’s strength (i: 72). Miskawayh (i: 72), presents a manumission letter to Rustam, a Persian version of which is found in Ibn al-Balkhī’s Fārsnāme, p. 43.²⁴ He provides no further references to Rustam in the Kayanid history and has nothing on him in the chapter on Kay Khusraw.

Other early Arabic historical and geographical sources, excepting al-Masʿūdī and al-Thaʿālibī, Ghurar, to be discussed later, provide only negligible references to Rustam or follow one of the above-discussed sources. Al-Balādhurī (d. 279/892), Futūḥ, p. 394, Ibn al-Faqīh (wrote in 290/903 or soon after), Mukhtaṣar, p. 208, and Ibn al-Athīr (d. 637/1239), Kāmil iii: 128, mention “Rustam’s Stable” in connection with the Arab-Islamic conquest of Sistan, which has been taken²⁵ as an indication that Rustam was already famous at that time. As the passage concerns Iran and more specifically Sistan, he was obviously famous, but again one should keep in mind the historiographical difficulties: what in a historical source is set at the time of the conquests, need not, and very often does not, date from that far back.

In his Āthār, al-Bīrūnī (d. about 442/1050), mentions in one sentence (p. 121/104//112) how Rustam ibn Dastān ibn Karshāsb al-malik rescued Kay Kāwūs when Shammar-Yarʿash of Yemen had imprisoned him, deriving this information from Ḥamza (in whose Ta⁠ʾrīkh this detail is, however, not given or preserved). Some pages later, on p. 151,²⁶ Rustam is said to have killed Shammar-Yarʿash, which does come from Ḥamza. In this book, al-Bīrūnī seems almost completely unaware of Rustam’s heroic deeds. It should be noted that al-Bīrūnī is one of the rare Arabic authors who had Abū ʿAlī al-Balkhī’s al-Shāhnāma at their disposal (Āthār, p. 114/99//107–108, cf. Chapter 4.1.2), and al-Balkhī had been able to use both Ibn al-Muqaffaʿ’s Siyar and possibly other Arabic translations of the Khwadāynāmag. Hence, the almost complete lack of Rustam material is highly significant when assessing what Ibn al-Muqaffaʿ’s translation of the Khwadāynāmag contained and what it did not.

The situation does not change much in al-Bīrūnī’s other books. In Kitāb Mā li’l-Hind, there is only one mention of Rustam at the very end of the book, p. 547 (trans. Sachau 1910, ii: 246). The rainbow, qaws-Quzaḥ, is attached by Indians to the name of a hero of theirs “just as our common people attach it to the name of Rustam”.²⁷ Al-Bīrūnī, who is usually extremely well informed about matters Persian, seems to know surprisingly little about Rustam (although the short note in Kitāb Mā li’l-Hind is interesting in itself).

Later geographical works are equally sparse when it comes to Rustam. Yāqūt (d. 626/1229), Muʿjam, mentions him twice.²⁸ In an article on Zābulistān (iii: 125), he explains that the toponym derives from an eponymous Zābul (cf. Zāl), the grandfather (sic) of Rustam ibn Dastān. The second mention comes in an article on Sistan (iii: 191) and, on the authority of Ibn al-Faqīh, defines it as the kingdom of Rustam the Mighty, who had been made king over it by Kay Qāwūs.

Finally, we come to al-Masʿūdī (d. 345/956), one of our main sources on pre-Islamic Persia. In his Tanbīh, p. 94//136, there is an extremely important passage on the wars between the Persians and the Turks:

At the end of the seventh part of Kitāb Murūj al-dhahab we have mentioned the reason why Persians exaggerate the [regnal] years of these kings, their secrets concerning this, and their wars against the kings of the Turks – these wars are called Baykār, which means “battle” – and other nations, as well as the battles between Rustam ibn Dastān and Isfandiyār in Khurasan, Sistan, and Zābulistān.

The term baykār would seem primarily to refer to the battles between the Persians and the Turks, where Rustam plays a major role.

Al-Masʿūdī’s Murūj, §§541 and 543 (on Farāsiyāb), gives the key to our understanding of the place of Rustam in pre-Islamic and early Islamic sources. The passages read:

The Persians tell a lot about Afrāsiyāb’s death and his battles, the battles and raids between the Persians and the Turks, the death of Siyāwush, and the story of Rustam ibn Dastān. All this is found explained in the book titled Kitāb al-Sakīsarān, which was translated by Ibn al-Muqaffaʿ from Ancient Persian into Arabic. The story of Isfandiyār (…) and how Rustam ibn Dastān killed him is narrated there, as well as how Bahman ibn Isfandiyār killed Rustam and other wonders and tales of the Ancient Persians. Persians think highly of this book because it contains stories about their ancestors and their kings’ histories. Thank God, we have been able to narrate many of their histories in our earlier books.

Murūj §541

According to what is told in the Book of al-Sakīsarān the Persians say that his paternal grandfather Kay Qāwūs was the king before Kay Khusraw and that Kay Khusraw had no offspring, so he gave the kingship to Luhrāsb.

Murūj §543

Thus, this Kitāb al-Sakīsarān seems to have concentrated on the Turkish wars, Siyāwush, Isfandiyār, and Rustam. It also shows that the story of Rustam was already integrated with royal matter in the Kitāb al-Sakīsarān.²⁹

In another passage, al-Masʿūdī seems to derive partly the same information from Kitāb al-Baykār, also translated by Ibn al-Muqaffaʿ:

This fortress was built by an Ancient Persian king of old times, called Isbandiyār ibn Bistāsf (…). This is one of the fortresses in the world that are described as impenetrable. The Persians mention it in their poems (ashʿārihā) and tell how Isbandiyār ibn Bistāsf built it. Isbandiyār waged many wars in the East against various peoples. He was the one who travelled to the farthest parts of the Turkish lands and destroyed the City of Brass. The deeds of Isbandiyār and all the things we have told are mentioned in the book known as Kitāb al-Baykār, which Ibn al-Muqaffaʿ translated into Arabic.

Murūj §§479–480

What the passages clearly tell is that there was a vivid tradition of historical books, other than the Khwadāynāmag, and some of these came to be translated into Arabic, whether by Ibn al-Muqaffaʿ or others (see Chapter 2.2.1). At least two such books, Kitāb al-Sakīsarān and Kitāb al-Baykār, contained Rustam material, and it is specifically this material that we find quoted, or referred to, in early Arabic works. The Khwadāynāmag, or its Arabic translation, the Siyar al-mulūk, on the contrary, is not mentioned by al-Masʿūdī, and may have contained next to no mentions of Rustam, which would not be surprising, as the refractory vassal would not have fitted in easily into a royal chronicle. The two books, as described by al-Masʿūdī, cover virtually all the material that may be found in early Arabic sources, and it is probable that they were the sources the other authors tapped, too, for this material, not the Khwadāynāmag and its translations. It should be emphasized that no source of ours, excepting the problematic Nihāya, claims to derive Rustam material from the Khwadāynāmag or its Arabic translations. To speculate about this without tangible evidence is rather futile.

In Murūj §542, the unlucky Yemenite excursion of Kay Qāwūs is referred to, and the Yemenite king is identified as Shammar-Yarʿash, and his daughter is Suʿdā, the Sūdābe (Sūdāwe) of the Iranian tradition. Al-Masʿūdī briefly tells how Rustam ibn Dastān marched to Yemen with 4,000 men, killed Shammar-Yarʿash, and saved Kay Qāwūs, together with Suʿdā, which led to the scene between Suʿdā and Siyāwukhsh “until what famously happened to him with Afrāsiyāb the Turk, how he sought asylum with him, and married his daughter”, how Kay Khusraw was born, how Siyāwukhsh was killed by Afrāsiyāb, and how Rustam killed Suʿdā and took revenge for Siyāwukhsh’s death by killing noble Turks.

According to Murūj §550, it was Bahman who, after several battles, killed Rustam.³⁰ The conversion of Bishtāsb to Zoroastrianism is mentioned in the same paragraph, but the two incidents are not explicitly connected.

Al-Masʿūdī is not alone in giving us information about separate translations of Rustam stories into Arabic. Ibn al-Nadīm (d. in 380s/990s), Fihrist, p. 364/305//716, mentions a Kitāb Rustam wa-Isfandiyār, translated by Jabala ibn Sālim (late second/eighth century) (cf. Chapter 2.2.1).³¹

Al-Jāḥiẓ (d. 255/868–869), Risālat al-Ḥanīn (Rasāʾil ii: 408) may refer to this book’s Middle Persian original: “the Mōbad has told that he has read in the Life of Isfandiyār (…), written in Persian,³² that when Isfandiyār raided the land of the Khazars in order to save his sister³³ from captivity (…)”. This quotation explicitly comes from a written Persian, most probably Middle Persian, source, not its Arabic translation. If it refers to the original text of the Rustam wa-Isfandiyār mentioned by Ibn al-Nadīm, then the focus of this book may have been on Isfandiyār rather than Rustam.

The only case where the Khwadāynāmag, in its Arabic translation, would seemingly be the source for an episode related to Rustam and his family is Nihāya, p. 82, quoted above. In addition, on p. 85, it is told, again on the authority of Ibn al-Muqaffaʿ, that Bahman married the great-granddaughter of Solomon, Ūmīdh-dukht: “I have found in Siyar mulūk al-ʿajam in the story of Bahman ibn Isfandiyār (…)”. At first sight, this would seem to locate at least these episodes in an Arabic Book of Kings. The Nihāya, however, is a highly problematic source, which attributes materials in a blatantly anachronistic way to eminent authorities to gain prestige for its tales (cf. Chapter 3.4). The latter passage is also problematic because it makes Ibn al-Muqaffaʿ refer to his own translation as his source.

Thus, reading extant early Arabic sources only, one receives the impression that, with the exception of the story of Isfandiyār, Rustam is a minor hero, on a par with other Persian generals. It is significant that none of the stories about him are attributed to Ibn al-Muqaffaʿ’s translation of the Khwadāynāmag (except for the dubious case of the Nihāya), and the information is probably derived from other, independent works, either translated from Middle Persian or written in Arabic on the basis of (Middle) Persian sources, either written or oral.

Kitāb al-Sakīsarān, Kitāb al-Baykār, and Kitāb Rustam wa-Isfandiyār (perhaps translated from the Persian Sīrat Isfandiyār), as far as we can deduce their contents, actually cover all the material that was transmitted in other Arabic sources, which means that there is no reason to attribute any of it to the Middle Persian Khwadāynāmag where, moreover, Rustam would have been out of character if we assume, as is usually, and with good reason, done that the Khwadāynāmag was a royal chronicle.³⁴ A subaltern prince would not too easily have been shown superior to the kings in such a source, so one would expect this to be the situation: the Rustam stories’ mise-en-scène could more easily be expected to be separate narratives of perhaps more popular origin than a royal chronicle.

On the other hand, there is reason to assume that many such stories were not translated from Middle Persian but were first composed in Arabic, although based on Persian lore. In some cases, such as that of Bahrām Gūr, it would be difficult to explain how the Arabs could have played such a major role in a book authored by Persians in Sasanian times or even soon after. If, on the other hand, the Arabs are removed from this story, very little remains, which makes it rather obvious that the story was first composed in an Arab context and probably in Arabic.

Once we turn to Classical Persian sources of the sixth/twelfth century and thereafter, the picture dramatically changes. The anonymous Mujmal (written 520/1126) shows both the influence of Firdawsī (explicitly mentioned), of other tenth-century versions of the Classical Persian Book of Kings, and of various other nāmes (Chapter 4.7), some of the last mentioned probably not in the form they have been preserved to us, but as earlier versions. The author also used the historical works of Ḥamza and al-Ṭabarī, thus combining various lines of traditions. The “official” Islamic version of history, as presented by al-Ṭabarī, does not, however, push the Persian tradition aside. On the contrary, on, e.g., p. 71/89, the author explicitly prefers these ancient sources to al-Ṭabarī.

The difference to Arabic sources is huge. The anonymous author summarizes virtually everything Firdawsī narrates about Rustam, but it must be kept in mind that the author is also partly using the same sources as Firdawsī, so we cannot be sure whether in a particular case he is summarizing Firdawsī or his other sources. The Mujmal lists the family members of Rustam, both ancestors and descendants, with genealogical details (pp. 23–24/25–26) and synchronizes or equates them with Biblical figures: Narīmān is identified with Noah and Rustam is given an alternative Arab genealogy (p. 32/38).³⁵ Isfandiyār fled from Rustam to Turkistān, but Rustam followed him there to kill him. “This is utter nonsense,” concludes the author, “but we mention it because it is found in (the Persians’) tall tales (khurāfāt) and decrepit (dāris) books, which we have seen” (p. 34/38).

The marriage of Zāl to Mihrāb, Rustam’s mother, is mentioned on pp. 36–37/42–43, and the following page (p. 37/43–44) summarizes the deeds of Sām. On p. 38/45, we come to Rustam’s story: Zāl sends him to bring Kay Qubād to be crowned. Rustam’s first battle (p. 38/45) is told in the same way as in Firdawsī: Rustam almost captures Afrāsiyāb, but Afrāsiyāb’s belt breaks and he gets away.

Mujmal, p. 39/45–46, narrates how Rustam saved Kay Kāwūs and killed the White Demon and the King of Māzandarān. Rustam and Afrāsiyāb fought in the Sawād of Baghdad or, according to another version, Rustam followed the Turkish King into Turkistān and fought him there. On the same page, it is told how Rustam freed Kay Kāwūs from Hāmāwarān. Brief mentions of Rustam’s new battles against Afrāsiyāb follow and then we are told the story of Suhrāb with all the details familiar from Firdawsī’s Shāhnāme, starting with Rakhsh having gone missing and ending with Rustam tragically killing his own son.

After this, the Mujmal moves on to narrate the story of Siyāwush. Rustam rears Siyāwush, whom Sūdāwe later attempts to seduce, although to no avail. Finally, Rustam slays the scheming stepmother and brings Kay Khusraw to Iran. Rustam fights in Turkistān for seven years (p. 40/46).

In Kay Khusraw’s time Rustam intercedes for Ṭūs, kills Fūlādwand, and fights against Afrāsiyāb. This is followed by “the story of Akwān Dēw”. Then Rustam frees Bīzhan by disguising himself and his men as merchants and attacking Afrāsiyāb by night. All this is told on p. 41/48. Farāmarz is sent to India, and Rustam takes part in renewed battles against Afrāsiyāb (pp. 41–42/49). Later, p. 44/52, it is told how Gustāsf sent Isfandiyār to fight Rustam and bring him to Iran in chains. Isfandiyār was mortally wounded (no mention of Sīmurgh is made) and left Bahman to be reared by Rustam. Later Gustāsf demanded Bahman back. Shaghād managed to kill Rustam and Zawāre (p. 44/53), and later Bahman marched to Sistan to take revenge on the remaining family members (pp. 44–45/53–54).

There are also a few scattered mentions of Rustam elsewhere in the book, which further testify to Rustam’s fame at the time (cf. the Index of the Mujmal). Zarathustra’s sleight-of-hand in Balkh is mentioned on p. 72/92, but Rustam plays no role in this context.

Keeping in mind that the author wished to present a concise historical work and hence condensed his material, it can be said that the whole Rustam material found in Firdawsī’s epic, and some other episodes, is contained in this work. The additional pieces certainly came from the group of narratives known as the Sistanian Cycle, i.e., independent epics on the family of Rustam (see Chapter 4.7). We do know from the Mujmal’s Preface (p. 2/2) that the author used several Sistanian books as his sources (Chapter 3.6). These early versions should not be confused with later epics with the same titles.

Although the Mujmal is the clearest example of Rustam’s importance in early Persian sources (excluding Firdawsī’s Shāhnāme), many other works give a similar impression of his fame. Ibn al-Balkhī’s Fārsnāme (written before 510/1116) is largely dependent on Arabic sources, but the author has augmented these with Persian ones. In this book, the main passage on Rustam comes in the chapter on Kay Kāwūs, pp. 40–43.³⁶ The passage relates how Rustam educated Siyāwūsh (sic, elsewhere in the Fārsnāme Siyāwush) in Zāwulistān; how with his troops he brought Kay Khusraw to Iran and slew the army of the pursuers (no other generals are mentioned: Rustam is the sole hero); and how he freed Kay Kāwūs from Yemen. Two versions of this are given, one according to Persian and the other according to (South) Arab historians, but both come from Arabic sources. The passage ends with Kay Kāwūs’ manumission of Rustam, and the manumission letter (āzādnāme) is given in full (cf. above).

In addition, there is on p. 53 a short mention of how Wishtāsf sent Isfandiyār to fight (paykār) Rustam-e Dastān “as is well known” and Isfandiyār was killed. Although this is only a brief mention, it shows how this particular episode was considered to be generally known. The use of the word paykār is again worthy of attention.

In Gardīzī’s Zayn (written in the early 440s/1050s), the influence of Firdawsī, or his source, explains Rustam’s strong presence.³⁷ Rustam frees Kay Kāwūs from “Māzandarān, which is called Yemen”. Kay Kāwūs rewards him by giving him Sistan and other fiefs (p. 74, no manumission letter is mentioned). In the Siyāwush episode, Rustam marches to Turkistān to take revenge on Afrāsiyāb for the death of Siyāwush and fights many battles there, finally killing Afrāsiyāb (p. 76). When he grew tired of worldly life, Kay Khusraw gave presents and fiefs, giving Rustam Sistan (again) and other provinces, as well as his personal clothes and gardens. Rustam and the other nobles followed him on his last mysterious trip (pp. 76–77). On pp. 77–78, Gardīzī tells how at the time of Kay Gushtāsp, Zarathustra introduced a new religion. No mention of Rustam’s reaction is given. On pp. 78–79, it is told how Gushtāsp sent Isfandiyār against Rustam, and Isfandiyār gave him the choice either to convert, to fight, or to be bound in chains and brought to the court of the king (the demand of conversion was not mentioned on p. 78). Rustam chose to fight. Sīmurgh is not mentioned, otherwise the fight follows (in an abbreviated form) the version of Firdawsī (or his source). The dying Isfandiyār left Bahman for Rustam to rear. Finally, on p. 80, it is told that when Bahman took his revenge, Rustam was already dead.

Ḥamdallāh Mustawfī (d. 750/1349), Tārīkh-e guzīde, follows the model of Firdawsī. Kay Qubād freed Iran from the hands of Afrāsiyāb by the aid of Zāl-e zar and his son Rustam and made Rustam the champion (jahān-pahlawān, p. 86). In the chapter on Kay Kāwūs’ reign Rustam’s heroic deeds, the haft-khān, are referred to but not related, and later he frees the King in Hāmāwarān, and Kāwūs gives him his sister Mihrnāz as wife (p. 87). This is followed by Rustam’s hunt in Samangān and the episode of Rustam and Suhrāb, told in five lines, under the indubitable influence of Firdawsī (p. 88). Next, Rustam, the atābak of the king, kills Sūdāwe, and later destroys Turkistān, taking part in the war against the Turks, to revenge Siyāwush’s death (pp. 88–89). The story of Bīzhan and Manīzhe is briefly told in Firdawsī’s version (pp. 89–90). Then Gushtāsf marches against Arjāsf, but Rustam remains behind. Later, Isfandiyār is sent against Rustam and is killed. Finally, Bahman kills Farāmarz in his war against Rustam’s family (one manuscript mentions that Rustam had already been killed by a brother of his) (pp. 93–94).

In the anonymous Tārīkh-e Sīstān (the main part of which was probably written soon after 448/1062) the whole Sistanian family is prominent.³⁸ In this book, Rustam’s story starts during Kay Qubād’s rule when the hero is fourteen and fights in Turkistān, taking revenge for Siyāwukhsh (p. 53, trans. Browne 1905: 5). The anonymous author refers to Farāmarz’s deeds, which he knows in an edition of twelve volumes.³⁹ As the deeds of Narīmān, Sām, and Dastān are told in the Shāhnāme (but it remains open to whose Shāhnāme the author is referring) they need not be repeated here, the author says. He also knows that the ḥadīth-e Rustam has been versified by Bū l-Qāsim Firdawsī and repeats the legend that Maḥmūd of Ghazna said that the Shāhnāme was nothing, except for the story of Rustam, and that he had in his army a thousand Rustams. All the heroes of the Sistanian family are well known, the author adds, and it is not possible to repeat all their deeds. He even mentions the Bakhtiyārnāme, thus bringing the story of Rustam’s family up to the fifth generation, counting from Rustam’s grandfather, Sām.⁴⁰ All this is told within the limits of one page, p. 53 (trans. Browne 1905: 5). On the next page, p. 54 (trans. Browne 1905: 6), the genealogy of the author’s patron is taken up to Rustam and the Sistanian heroes.

The author also knows Bū l-Muʾayyad’s Kitāb-e Garshāsb (p. 75, trans. Browne 1905: 24).⁴¹ He emphasizes that the Sistanian family, up to Farāmarz, kept their aboriginal religion, which they derived from Adam (p. 73, trans. Browne 1905: 23). The battle, paykār (note again the word), between Isfandiyār and Rustam was caused by the new religion of Zartusht (pp. 73–74, trans. Browne 1905: 23).

To end the section of Persian authors, Ṭūsī’s ʿAjāʾib is a valuable, but all too little studied book. It takes us to a different tradition, which is sparsely documented. Ṭūsī’s ʿAjāʾib taps sources, oral or written, which are more popular than those used by historians of the time and gives us a glimpse of what went on outside learned circles. It is not surprising that Ṭūsī includes references to stories which later surface in popular epics.

Ṭūsī’s ʿAjāʾib was written soon after the last date mentioned in the text, 562/1166 (p. 300, cf. Preface, p. xvi)⁴² and it uses a lot of material familiar from later epics, but little from Firdawsī’s Shāhnāme. However, the author highly respects Ḥasan-e (sic) Firdawsī of Ṭūs (p. 246).⁴³ On p. 473, there is the earliest attestation of a story, another version of which is found in, e.g., the Bāysunqurī Shāhnāme.⁴⁴ According to it, Firdawsī became rich after having seen Rustam-e Zāl in a dream and been told about a treasure in Ṭūs. The book also contains dozens (if not hundreds) of references to Alexander, largely familiar from the various versions of the Alexander Romance,⁴⁵ and also to Anūshirwān’s miraculous deeds and journeys. Afrīdūn, Ḍaḥḥāk, and Bahrām Chūbīn also often appear.

Ṭūsī mentions Narīmān’s conquest of China (pp. 191, 419) and tells an interesting variant concerning the reason why Zāl was abandoned by Sām (p. 418): it was the blackness of Zāl’s body, not the whiteness of his hair that was the cause of shame. Also otherwise the story differs from Firdawsī: the author knew Firdawsī and respected him, but he either did not know the contents of the Shāhnāme too well or did not care to offer the version told there, but preferred other narratives that, as in this case, directly contradict what Firdawsī wrote.

Ṭūsī uses Rustam’s standard Arabic epithet al-Shadīd, the Mighty (pp. 263, 419),⁴⁶ which may indicate that at least sometimes he used, either directly or indirectly, Arabic sources for Persian national history. The author tells that Rustam and Zāl’s tombs are in Samanjūr and that Rustam’s palace lies in ruins outside of Zāwulistān (p. 230). He also tells that the descendants of Rustam still rule bwls, which lies on the coast of daryā-ye Maghrib, only six parasangs from al-Andalus (p. 190).

Like many other sources, Ṭūsī tells (p. 420) how Rustam liberated Kay Kāwūs. The story of Rustam and Akwān Dēw is mainly told on the lines of Firdawsī, but with some significant differences (pp. 493–494). The source is given as “it is told in books” (dar kutubhā āwurde-and) and Firdawsī is not mentioned. On p. 510, Ṭūsī briefly relates the story of Rustam and the White Demon. The most interesting passage comes on p. 75, where it is told why Rustam did not believe in Zarathustra: in his early career Zarathustra had practised jugglery (ḥuqqa-bāzī) in the court of Rustam, who had given him a small reward. When Zarathustra later claimed to be a prophet, Rustam did not believe in him.⁴⁷

The Sistanian material of the book in the main differs from that in the earlier Arabic and Persian sources, including Firdawsī. Most probably it comes from the separate epic stories about the Sistanians (Chapter 4.7) and thus through a line separate from that of the Khwadāynāmag and its Nachleben. Note that very few traces of these traditions are found in early Arabic literature, as shown above, which further supports the by now rather obvious conclusions that there was no such material in the Khwadāynāmag.

This selection of Persian sources shows that the image of Rustam was much more central in the Persian than in the Arabic tradition. Yet even though all early Persian historical sources, except Balʿamī, are later than Firdawsī they do not slavishly follow his version of Rustam’s adventures. In some, the influence of Firdawsī is clear, and some mention him as one of their sources, but even these add incidents known neither from Firdawsī’s Shāhnāme nor from the Arabic tradition. When narrating the same episodes, they may also have significant differences to Firdawsī, which implies that they also had other sources at hand and sometimes preferred these to Firdawsī.

It is clear that in early Islamic Iran a wide range of Rustam narratives was in circulation. Some may have been oral, but references to separate books, where Rustam played a role (Kitāb al-Sakīsarān, Kitāb al-Baykār, Kitāb Rustam wa-Isfandiyār, Sīrat Isfandiyār) and which were not integrated into the Khwadāynāmag, or its Arabic translation(s), imply that written Middle Persian versions were also available. Some of these separate stories may first have been written down in Arabic, while others may have circulated in written Middle Persian texts, and yet others may have been set down in early Classical Persian in the tenth century directly from oral tradition.

Al-Thaʿālibī’s Ghurar stands out among early Arabic sources (cf. Chapter 4.4). The difference to earlier Arabic sources is considerable. For al-Thaʿālibī – and one should keep in mind that he may, or may not, be the same al-Thaʿālibī as the famous author of the Thimār and the Iʿjāz – Rustam is a figure of central importance and there are few stories of him in Firdawsī’s Shāhnāme that are not paralleled in the Ghurar. Thus, one finds there the story of Rustam’s birth and youth (pp. 104–106), his finding a horse, Rakhsh (pp. 140–145), his first fight against Afrāsiyāb (pp. 145–147), his freeing Kay Kāwūs from the King of Yemen, Dhū l-Adhʿār (pp. 161–163), a brief mention of Rustam being made the iṣbahbadh of Iran by Kay Kāwūs, who also renews his vice-regency (tawliya) in Nīmrūz, Zābulistān, and India (p. 165), the story of Siyāwush, including Rustam rearing him (pp. 168–170), Siyāwush going to war against Afrāsiyāb with Rustam and their making peace with the Turkish King (pp. 187–198), the killing of Suʿdā alias Sūdāne (sic)⁴⁸ by Rustam, and the revenge for Siyāwush (pp. 216–218), Rustam and others welcoming the returning Kay Khusraw (p. 221), his receiving a legacy from Kay Khusraw, and the new King, Luhrāsf, giving an audience to him (p. 238), and the haft-khān of Isfandiyādh, which ties up with the story of Rustam (pp. 301ff.).

The conflict between Isfandiyādh and Rustam is discussed in detail on pp. 341–375. The story is very similar to that given by Firdawsī (and, presumably, the source common to both), but it contains some interesting differences, the most remarkable of which is the mention of a raven that guided Bahman, the son of Isfandiyādh, to where Rustam was hunting. This detail is attributed to khurāfāt al-Furs, which, again, implies that al-Thaʿālibī is using other (oral or written) sources to complement his main source. Finally, on pp. 379–385, it is told how a brother of Rustam, Shaghāy,⁴⁹ killed him by a ruse, and how Bahman later took his revenge on the other members of Rustam’s family (pp. 386–388). The same passage, p. 388, also mentions that according to Masʿūdī-ye Marwazī’s Persian muzdawija, Bahman also killed Zāl during this expedition, a detail running contrary to the main story of al-Thaʿālibī (and Firdawsī). What it shows is that al-Masʿūdī al-Marwazī had already interwoven the fates of the dynasty of the Sistanians with national history, which, of course, we also know on the Arabic side from the other al-Masʿūdī, the author of the Murūj and Tanbīh, onward.

On pp. 301–302, al-Thaʿālibī refers to Isfandiyādh’s haft-khān as irrational and says that he repeats the story only because it is famous, and kings and ordinary people like it, and because it is found on ṣuḥuf (separate, short manuscripts?) as well as in pictorial representations.⁵⁰

The version of al-Thaʿālibī gives Rustam the central place he also has in Firdawsī’s epic, and it seems obvious that the Prose Shāhnāme is the origin of the Rustam stories that are common to both Firdawsī and al-Thaʿālibī. Episodes found in Firdawsī and lacking in al-Thaʿālibī and presumably in the Prose Shāhnāme are few, the most important being the story of Bīzhan and Manīzhe; Rustam’s haft-khān; Akwān Dēw; and the tragic story of Rustam and Suhrāb.⁵¹ These were probably lacking in the common source of al-Thaʿālibī and Firdawsī, as al-Thaʿālibī does not usually drop whole scenes, and only the dropping of Rustam’s haft-khān and his encounter with Akwān Dēw could be explained by al-Thaʿālibī’s negative attitude towards the khurāfāt al-Furs. More probably they were added to the whole story by Firdawsī, the haft-khān probably on the basis of Isfandiyār’s similar deeds.⁵² However, they cannot be used as binding evidence for Firdawsī having invented these episodes or having been the first to insert them into national history. What does strike one, though, is that these particular episodes stand out as rather separate stories, not quite as clearly linked to the main story as most other episodes are.⁵³

The inspection of early Arabic and Classical Persian sources enables us to assess the position of Rustam before Firdawsī. Our sources on Rustam in pre-Islamic times are meagre, but there is no reason to doubt that he was a major character in the Eastern Iranian world, that stories about him were told or sung in some Iranian language(s), and that he was known at least by name also in the Western parts of Iran and in Armenia.

In the mid-eighth century some of these stories reached the Arabic world through the translation by Ibn al-Muqaffaʿ of Kitāb al-Baykār and Kitāb al-Sakīsarān and Jabala’s Rustam wa-Isfandiyār. It is not clear whether it was Rustam or Isfandiyār who was the main focus in the last-mentioned book: the title Sīrat Isfandiyār, used by al-Jāḥiẓ and possibly referring to the same work, would imply that it may well have been Isfandiyār, who, despite his final defeat at the hands of Rustam, was the work’s main character.⁵⁴ In the first two books, Rustam was clearly present but again it remains uncertain whether or not he was their main character.

The Rustam episodes of these separate books influenced only a small part of Arabic historical literature. Ibn al-Muqaffaʿ’s translation of the Khwadāynāmag was, on the other hand, extremely influential and many later Arabic historical works seem to tap it for materials. Thus, we have no dearth of material on mythological figures such as al-Ḍaḥḥāk (cf. Chapter 5.2) or Jamshīd and later kings in Arabic sources that discuss pre-Islamic Iran. Yet, Rustam is almost ignored in the Arabic tradition before al-Thaʿālibī, except for the matter covered by the separate translations by Ibn al-Muqaffaʿ and Jabala and quoted only in a few books. Had Rustam been strongly present in Ibn al-Muqaffaʿ’s translation of the Khwadāynāmag, it would be difficult to explain why certain early sources, such as al-Yaʿqūbī’s Ta⁠ʾrīkh and Ibn Qutayba’s Maʿārif, have nothing on Rustam, though they have plenty of material on other figures of Persian national history.

This seems to leave but one explanation. Ibn al-Muqaffaʿ’s widely-known translation of the Khwadāynāmag contained little material on Rustam. Further, although it is not impossible that Ibn al-Muqaffaʿ could have left out such material on purpose, no obvious reason for this can be seen. More probably, Ibn al-Muqaffaʿ’s translation of the Khwadāynāmag had little to tell about Rustam because its Middle Persian original did not have much on Rustam either.

This is actually what we might expect. If the Khwadāynāmag was, as it seems to have been, a royal chronicle, the counterweight to the kings had little to do in it: the Sasanian kings were hardly enthusiastic about a hero who is often shown to be superior to his overlords in a moral sense. Hence, a priori, one expects Rustam not to have been given much place in such a work and Ibn al-Muqaffaʿ’s lack of Rustam stories corroborates this. The Arabic evidence makes it hard to claim Rustam had more than a marginal role to play in the Khwadāynāmag, if even that.

The Arabic translations of some separate episodes of Persian national history (Kitāb al-Sakīsarān, Kitāb al-Baykār, Rustam wa-Isfandiyār, perhaps the same as Sīrat Isfandiyār) show that by the mid-eighth century Rustam had to some extent been integrated into the history of the kings, but this does not mean that he would have found a place in the Khwadāynāmag itself. The integration took place through independent books that have nothing to do with the Khwadāynāmag.

Tenth-century evidence shows that at that time Rustam was fully integrated into the storyline of national history and had found a place in works that related this history. This should not be taken to mean that the Khwadāynāmag would later have been revised in its Middle Persian form.⁵⁵ When tenth-century kings patronized the writing of Persian history, Middle Persian texts were not what they were after. They wanted to have texts in their own literary language, the emerging Classical Persian, such as the translation/re-working of al-Ṭabarī’s Ta⁠ʾrīkh by Balʿamī. The story about the compilation of the Prose Shāhnāme does not indicate that the scholars involved would have written their work in Middle Persian and it is not even clear to what extent they used Middle Persian works as their sources (Chapter 4.2). They probably did use whatever Middle Persian material they had at hand (Chapter 4.6), but they will also have used earlier texts written in Persian or Arabic, as well as oral information, whether epic songs or prose stories. To claim that these scholars, or anyone else, wrote new Middle Persian versions of the Khwadāynāmag – or any new Middle Persian works – is speculative and unwarranted. We have no evidence for this, and it would run counter to the currents of the tenth century, which favoured translations from Middle Persian into Classical Persian, not new secular works in Pahlavi.

From the point of view of Firdawsī, it seems that he received most of the Rustam material already integrated into national history in the Prose Shāhnāme.⁵⁶ In addition, he may well have found other separate stories involving Rustam in a variety of roles, such as that of Bīzhan and Manīzhe or Rustam and Suhrāb, which first surface in his Shāhnāme. Whether they derived from Āzādsarw⁵⁷ we cannot know, but it is possible. Some of these stories may already have been added to the Prose Shāhnāme or the other Shāhnāmes of the tenth century, even though the evidence from al-Thaʿālibī’s Ghurar would seem to speak against this.

A separate origin for at least some of Firdawsī’s Rustam stories finds some evidence in his habit of referring to old dihqāns and other authorities when he comes to such passages. It seems that when versifying his main source, the Prose Shāhnāme, Firdawsī does not bother to give proofs for the authority of his stories – he was resuming well-known material and hence was not in need of further authorization. When adding separate incidents, on the contrary, he was stepping outside the limits of the authoritative history of Iran and had to defend his additions by referring to authorities. Only when he was being innovative did he feel the need to refer to venerable sources. This is also seen in the fact that references to “ancient sources” start with the Rustam cycle, as if Firdawsī wanted to emphasize that these stories, too, were worthy of inclusion into national history. Other orphan stories, which are marked by such references and thus probably originally come from outside the established tradition, seem mainly to include stories inappropriate for a Middle Persian Khwadāynāmag (e.g., especially, Dārāb’s fight against the Arab army led by Shuʿayb, perhaps modelled after stories about Abū Muslim, d. 137/755).⁵⁸

To resume, we have next to no indication that Rustam would have been known to the Arabs before Ibn al-Muqaffaʿ in the mid-eighth century. Up until the mid-tenth century, sources seem to concentrate on a limited number of scenes in Rustam’s life and these particular scenes were the subject of separate texts on Rustam, known to have existed in the mid-eighth and the ninth century and nowhere claimed to have constituted part of the Middle Persian Khwadāynāmag or any of its translations into Arabic. They were only integrated in the tenth century into the Shāhnāmes written in early Classical Persian. The Middle Persian Khwadāynāmag, as we can see from the Arabic books that used it, may not perhaps even have mentioned Rustam and if it did, he was probably on a par with other heroes, and was not the central character of the narrative. The separate Arabic texts, on the other hand, show that the stories of Rustam were interwoven into the lives of some Persian kings (especially Kay Qubād, Kay Kāwūs, and Kay Khusraw), which proves that the process of intermingling the two traditions had begun by the mid-eighth century.

In the tenth century, as shown by Firdawsī’s epic, other Shāhnāmes, and al-Thaʿālibī’s Ghurar, the process had been finalized and Rustam had become the greatest hero of Persian national history, but there is no tangible evidence that this would have found form in any rewritten version that would have been titled Khwadāynāmag or would have been in Middle Persian. What is clear, though, is that the various Shāhnāmes of the tenth century had produced a storyline mainly in harmony with the later work of Firdawsī.

The existence of a voluminous repertoire of stories about the Sistanian heroes is proven by the later epics which contain individual details that can be corroborated by sources earlier than Firdawsī and have, hence, to tap sources (oral or otherwise) that existed before him. This also makes it probable that instead of inventing new episodes, Firdawsī, as most contemporary authors were wont to, received the stories from older tradition and merely versified them. Later, he inserted them into his magisterial epic. It is possible that he himself conceived the concept of a unified narrative only after he had begun his career as an epic poet by composing separate stories.

5.2 Armāyīl and Garmāyīl: The Formation of an Episode in Firdawsī’s Shāhnāme

In the previous chapter (5.1), I endeavoured to approach the question of the contents of the Khwadāynāmag through an analysis of several works that may derive their material partly from the Khwadāynāmag, although mainly indirectly. This chapter (5.2) turns the focus on Firdawsī and his Shāhnāme and studies one specific episode to show what may have been Firdawsī’s part in developing the text he versified (cf. also Chapter 4.6).

Among the many impressive episodes in Firdawsī’s Shāhnāme is the feeding of human brains to the snakes which grow out of Ḍaḥḥāk’s shoulders. It is a favourite passage in Arabic and Classical Persian literature and the concomitant aetiological myth of the origin of the Kurds is told in perhaps more sources than almost any other passage of the Shāhnāme.

Told in brief, Iblīs, who earlier had incited Ḍaḥḥāk to parricide, reappears to him in the shape of a cook and accustoms him, now the King of Yemen, to eating meat instead of his earlier, mainly vegetarian dishes (J125–146).⁵⁹ In J147–155 Iblīs, as a reward for his gastronomic prowess, asks permission to kiss Ḍaḥḥāk’s shoulders. Receiving the permission he kisses him and instantaneously disappears, as if the ground had swallowed him up. Two black snakes grow out of Ḍaḥḥāk’s shoulders. Whenever cut down the snakes grow again, and physicians are unable to help the king (J156–160). Iblīs again reappears, now in the shape of a doctor, and tells what to do: the snakes have to be fed with human brains (J161–166). Firdawsī does not explicitly say that the snakes annoyed Ḍaḥḥāk, but, evidently, they would have done so.⁶⁰

After telling this, Firdawsī drops the subject for some forty verses, to return to it in Z12–37. Here he tells how two pious men, Armāyil and Garmāyil,⁶¹ discussed the iniquities of Ḍaḥḥāk, now also known as Bīwarasp, who had meanwhile become the King of Iran. They infiltrate his service as cooks in order to save at least one of the two men daily slaughtered for the snakes and each day start letting one of the two intended victims free and replacing his brains with those of a sheep. When two hundred (or, according to a variant, twenty) men have been rescued, they give them some sheep and some goats and send them off, telling them to keep out of towns. This, says the narrator, is the origin of the Kurds. After this, the narrator goes on to relate the revolt of Kāwe and the uprising of Ferīdūn.

Some of the themes in this episode go back to Indo-Iranian mythology. From Avestic times myths about the man-eating Azhi Dahāka had hovered between him being a humanized dragon or a dragonized mythic hero.⁶² The episode as a whole, though, is much more recent and the purpose of this chapter is to delineate the development of the episode in Early Islamic times, focusing on the figure of Armāyīl.

The oldest testimony for Armāyīl is Shahrestānīhā ī Ērānshahr §28, where an Armāyīl is mentioned in connection with Azhi Dahāg:

Twenty-one cities that were built in Padishkhwārgar were either built by Armāyīl or, following his order, by the mountaineers, who had acquired from Aži Dahāg the mountains as their dominion.⁶³

The passage tells us little more than that Ḍaḥḥāk and Armāyīl were somehow connected at the time of this text, the final redaction of which seems to date to the eighth century, although much of the material is considerably earlier.⁶⁴ It uses the term kōfyārān “mountaineers”, which we will meet again in later sources as kōhyār (Arab. kūhiyar [written kūhbār] in al-Dīnawarī, Akhbār, p. 10; al-Ṭabarī, Ta⁠ʾrīkh i: 229//ii: 26, has qūhiyār).⁶⁵ It also presents Armāyīl as a person who was important enough to have twenty-one cities built by or for him.

The name Armāyīl has been explained by Markwart (1931: 68) as the Middle Persian Armāyēl with a Georgian ethnic suffix (“the Aramean”), and it is attested in Armenian sources.⁶⁶ The etymology is less than certain, though, and one might equally well see it as an invented name.

It seems that the next reference to the episode comes from Ibn Qutayba (d. 276/889), who in his Maʿārif, p. 618, mentions that the Kurds are the leftovers of Bīwarasf’s food. He also tells that Bīwarasf ordered two persons to be slaughtered every day, but that his Vizier Armāʾīl pitied the victims and let one of them live. It is noteworthy that Ibn Qutayba does not speak about the brains of the men, merely saying that Bīwarasf ate their flesh.⁶⁷ Ibn Qutayba does not mention snakes, but sees Bīwarasf in his archaic role as a cannibalistic monster, Ḍaḥḥāk-e mardās.⁶⁸

This seems to be the original scenario of the episode: one nobleman, Armāyīl, feeds Ḍaḥḥāk. Only late sources mention two persons and make them cooks, and even they make it clear that they were no ordinary cooks but noblemen disguised as such. There is no evidence that the other character, Garmāyīl (or Karmāyīl) would have been invented before the mid-tenth century and, taking into consideration the large number of texts that do contain this episode, it is improbable that a variant version with two cooks would have left no traces, had the second cook been an early addition. The second name seems to have been created as a Schallwort to echo the first.⁶⁹

The earliest source to speak specifically of brains and at the same time the first to mention the snakes on the King’s shoulders is al-Yaʿqūbī (d. 284/897), Ta⁠ʾrīkh i: 158. Al-Yaʿqūbī is very concise, criticizing the irrationality of these stories.⁷⁰ He does not identify this Persian king by name but, obviously, he is speaking of Ḍaḥḥāk. His knowledge of this mythical material was, though, not intimate, as can be seen from the list in which he claims that one of the kings had several mouths and eyes and another had snakes on his shoulders and ate men’s brains. Anyone familiar with Persian mythology would have seen he was speaking of the one and the same monstrous king.

The dislike of khurāfāt may have been behind the rationalizing explanation for the snakes growing on Ḍaḥḥāk’s shoulders. The first to explain away the unnatural was the contemporary of al-Yaʿqūbī, al-Dīnawarī (d. not later than 290/902), who offers this explanation in al-Akhbār al-ṭiwāl, pp. 6–7. He uses the word silʿa “sebaceous cyst” for the things that grew on Ḍaḥḥāk’s shoulders. This remains the standard expression in rationalistic descriptions, although laḥma⁷¹ and faḍla, or gūsht-faḍla,⁷² are also occasionally used. In late versions this rationalization is taken a step further by speaking of wounds.

From a medical diagnosis, there was only a short step to a medical cure. Al-Dīnawarī, however, does not take this step. Contradicting himself he says that the brains were fed to the silʿas. Al-Ṭabarī (d. 310/923), Ta⁠ʾrīkh i: 204–205//ii: 3–4, is the first to speak of anointing the silʿas with brains to alleviate the pain. Al-Ṭabarī claims that this passage, as well as much else he tells about Ḍaḥḥāk in his Ta⁠ʾrīkh, derives from Ibn al-Kalbī (d. 204/819). The first to mention that the snakes grew after Iblīs had kissed Ḍaḥḥāk is al-Maqdisī (d. after 355/966), Badʾ iii: 141, and the first to explain this as a reward for his gastronomic feats is Firdawsī, followed by al-Thaʿālibī (wrote around 412/1022), Ghurar, p. 18.

The archaic version of the story seems to have been that Ḍaḥḥāk’s vizier fed his master, or the snakes growing out of his shoulders, with human flesh, or brains. This is amply documented in early sources. The following list contains the most important early (pre-1200) attestations of the theme, as well as one later one which is of particular interest. Most later sources merely repeat what Firdawsī or the historical tradition have already said. The contents which are related to this episode in each work are briefly described after each item.

Ibn Qutayba (d. 276/889), Maʿārif, p. 618: Persians say that the Kurds are the leftovers of Bīwarasf’s food. Every day he ordered two people to be slaughtered and ate of their flesh. He had a Vizier, called Armāʾīl, who slaughtered one of the intended victims but let the other live, sending him⁷³ off to the mountains of Fārs, where they multiplied.

al-Yaʿqūbī (d. 284/897), Ta⁠ʾrīkh, i: 158: upon the shoulders of a king there were two snakes that ate men’s brains.

al-Dīnawarī (d. not later than 290/902), Akhbār, pp. 6–7: Persians call al-Ḍaḥḥāk by the name Bīwarasf. Two silʿas grew out of his shoulders in the shape of snakes. They pained him until they were fed (sic) with human brains. Four bulky men were daily brought to be slaughtered. He had a Vizier, Armiyāyīl, who let two of them live, substituting their brains with those of two rams, and told them to go where no one could find them. They went to the mountains. People say that this is the origin of the Kurds.

Ibn al-Faqīh (wrote in 290/903 or soon after), Mukhtaṣar, pp. 275–276: Afrīdhūn brought al-Bīwarasf to Mt. Demavend and put Armāʾīl in charge of him and his nourishment. Every day he used to slaughter for him two people with whose brains al-Bīwarasf nourished himself. Armāʾīl thought it a sin to slaughter people and managed to save (some of) them.⁷⁴

al-Ṭabarī (d. 310/923), Ta⁠ʾrīkh i: 204–205//ii: 3–4 (← Ibn al-Kalbī, d. 204/819): Two silʿas grew out of al-Ḍaḥḥāk’s, alias Bīwarasb’s, shoulders and pained him until they were anointed with human brains. Every day, two men were slaughtered. i: 206/ii: 6: Many people say that they were pieces of swollen flesh, shaped like a viper’s head, while others say that they were snakes.

al-Masʿūdī (d. 345/956), Murūj §§1115–1116: Two snakes grew out of al-Ḍaḥḥāk’s shoulders and fed on human brains. This led to the death of many until people rose against him. Afrīdūn chained him in a cave in Mt. Demavend, as has been mentioned (§538). Every day the Vizier of al-Ḍaḥḥāk had slaughtered (qad kāna … yadhbaḥ) a man and a ram, mixing their brains for the snakes to eat. He drove the other man to the mountains where the freed men grew numerous. This is the origin of the Kurds.⁷⁵

al-Masʿūdī (d. 345/956), Tanbīh, pp. 85–86//123–124: Persians exaggerate about al-Bīwarasb, alias al-Ḍaḥḥāk, telling how two snakes grew out of his shoulders and were only pacified by human brains. More about this has been told in the Murūj.⁷⁶

Ḥamza al-Iṣfahānī (d. 350/961 or 360/971), Commentary to Abū Nuwās (d. c. 198/813), Dīwān ii: 2: Persians claim that al-Ḍaḥḥāk is still alive on Mt. Demavend. On his back there are two snakes, which daily feed on flesh. If flesh is not given to them, they bite him.⁷⁷

Balʿamī (wrote in 352/963–4), Tārīkh, pp. 97–99; Tārīkhnāme i: 102–103: two long pieces of flesh (gūsht) grew on Ḍaḥḥāk’s, or as Magians say, Bīwarasb’s, shoulders and after 700 years of his rule these became wounds and started to ache. No one knew how to cure them until Ḍaḥḥāk had a dream, wherein a voice said to him that he should cure the wounds with human brains. After this he daily slaughtered two people and put some of their brains upon the wounds. This went on for 200 years. He had a cook (khwān-sālār) who took care of this. Every day he killed one man, but let the other one go, mixing lamb’s brains with one of the victim’s. When some time had gone by, he smuggled the saved people by night out of the town. This is the origin of the Kurds.

al-Maqdisī (d. after 355/966), Badʾ iii: 141–143: Iblīs, in the shape of a young man, came to al-Ḍaḥḥāk, i.e., Bīwarasb, and kissed his shoulders. Two snakes grew out of them and fed on human brains. Every day al-Ḍaḥḥāk slaughtered two men. Bīwarasb had a cook, called Azmāyil. When young men were brought to him to be slaughtered, he let one of the two live and sent him out into desert. The Kurds derive from these men.⁷⁸

Firdawsī (d. 411/1019–20), Shāhnāme (see above).

Miskawayhi (d. 421/1030), Tajārib i: 62: al-Ḍaḥḥāk, alias Bīwarasb, had on his shoulders two silʿas, which he moved to frighten people, claiming they were snakes. (No mention of slaughtering anyone, except for the general one that al-Ḍaḥḥāk, alias Bīwarasf, killed and crucified people, but this is not connected with the silʿas).

al-Thaʿālibī (wrote around 412/1022), Ghurar, pp. 20–25: Two snakes grew out of al-Ḍaḥḥāk’s, alias Bīwarasf’s, shoulders when Iblīs kissed them and blew on them. Some say that they were silʿas, merely resembling snakes. Iblīs appeared to him and told him that the snakes will never be separated from him, but that they can be pacified by feeding them with human brains. Al-Ṭabarī has said that most people of the Book say that they were lengthy pieces of flesh, like the head of a viper. Two men were slaughtered every day and their brains were fed to the snakes. Al-Ḍaḥḥāk had two cooks, Armāyīl and Karmāyīl, who one day decided to set free one of the two men and to substitute a sheep’s brains for those of his, feeding the (freed) one with the (rest of the) sheep. Set free, they became the origin of the Kurds.

al-Thaʿālibī (429/1038), Thimār, p. 284: the two “horns” (qarn) of al-Ḍaḥḥāk, alias Bīwarasf, were two silʿas, which people call snakes.⁷⁹

al-Bīrūnī (d. about 442/1050), Āthār, pp. 282–283/227//213–214: Bīwarasf ordered two men every day to be slaughtered to feed his two snakes with their brains. Azmāʾīl was commissioned to take care of this, but he freed one of the two, replacing the brains of the freed one with those of a ram. Others say that they were two silʿas, which were anointed with the brains.

Gardīzī (wrote in early 440s/1050s), Zayn, p. 67: two snakes, some say two wounds, grew on the shoulders of Ḍaḥḥāk, alias Bīwarasp. Every day two men were killed and their brains were given to the snakes or put upon the wounds. P. 70: after Ḍaḥḥāk was imprisoned, Afrīdhūn thanked the Vizier of Ḍaḥḥāk, Armāyīl, who had set the men free. They became the Kurds of the West of Kūhistān. P. 354: Bīwarasb, i.e., Ḍaḥḥāk, wanted two men to be slaughtered daily, but his Vizier Armāyīl set one of them free.

Ibn al-Balkhī (wrote before 510/1116), Fārsnāme, p. 35: Upon the shoulders of Bīwarasf, alias Ḍaḥḥāk, there were two silʿas, i.e., gūsht-faḍlas. To frighten them he let people think they were snakes. Finally they became painful, but the pain was alleviated when they were anointed with brains. The killing of young men continued until the rebellion of Kāwe.

Mujmal al-tawārīkh (written 520/1126), pp. 34–35/40–41: there was on Bīwarasb’s, alias Ḍaḥḥāk’s, shoulders a sickness (ʿillat), which people called snakes. The world was depopulated as people’s brains were extracted to feed the snakes. After 700 years Armāyil and Karmāyil came into his service and slaughtered one of the two men but let the other one free and sent him off into the desert. The Kurds are the offspring of the freed men.

Muḥammad Ṭūsī (wrote in the late sixth/twelfth c.), ʿAjāʾib, pp. 130–131: Ḍaḥḥāk was a tyrant who used to give human flesh to feed the snakes which grew out of his shoulders. After imprisoning Ḍaḥḥāk in a pit in Mt. Demavend, Afrīdūn ordered Armiyāyīl to provide him daily with two human brains. Some time went by. Finally, Armiyāyīl repented and started giving him the brains of two sheep and let the men go. P. 236: Sarakhs is a city built during the time of Ḍaḥḥāk, who ate people. People were fed to the snakes which grew out of his shoulders, but some of these people escaped.

Yāqūt (d. 626/1229), Muʿjam ii: 475 (← Ibn al-Kalbī): Armāʾīl, the Nabatean from al-Zāb, supervised al-Ḍaḥḥāk’s, alias Bīwarasf’s, kitchen. He used to slay one young man and let the other free, mixing the flesh of a ram with that of the other. After having imprisoned al-Ḍaḥḥāk, Afrīdūn wanted to kill Armāʾīl.

When this act of cannibalism took place is somewhat obscure. Early sources give two possibilities. Either Ḍaḥḥāk, or his snakes, ate the victims while he was ruling as the King or he did this when imprisoned in Mt. Demavend.⁸⁰ The latter option is slightly surprising, as this evil act is difficult to explain when the monster is in chains. This unmotivated act might yet be the earlier, for two reasons. Firstly, we may take this as a lectio difficilior of sorts: it is easier to understand why the eating would have been retrojected from the imprisonment period back to Ḍaḥḥāk’s rule than vice versa. Secondly, the freed men, the forefathers of the Kurds, are in many early versions said to live around Mt. Demavend, which is understandable if they were set free there. However, Ḍaḥḥāk is also otherwise connected with Mt. Demavend, so this is by no means decisive.⁸¹

Eating people during Ḍaḥḥāk’s reign is attested earlier in our sources than the other option, being implicitly mentioned by Ibn Qutayba in his Maʿārif, p. 618, where Bīwarasf is said to have ordered two men to be slaughtered. The earliest source to date this habit to the period of Ḍaḥḥāk’s imprisonment in Mt. Demavend is, though, not much later, as the detail turns up in Ibn al-Faqīh’s⁸² Mukhtaṣar, pp. 275–276. Here the one to feed the beast is Armāyīl, set by Ferīdūn to guard the prisoner.⁸³ Whether the tradition which derived the Zoroastrian dynasty of Maṣmughān from the descendants of Armāyīl is ancient or not is uncertain, but it, too, is already found in Ibn al-Faqīh.

Balʿamī (wrote in 352/963–4) is the first to mention that the habit of eating brains only began after 700 years of Ḍaḥḥāk’s reign (Tārīkh, p. 98, Tārīkhnāme i: 102).⁸⁴ This is in contradiction to Firdawsī’s version because in his Shāhnāme the snakes grow out of Ḍaḥḥāk’s shoulders and Iblīs gives his nefarious advice before Ḍaḥḥāk’s victory on Jamshīd. Balʿamī, who does not mention Iblīs at all, also has the curious detail of Ḍaḥḥāk seeing in a dream the cure for his pains, whereas all other sources attribute this advice to Iblīs. Balʿamī’s version cannot be easily brushed aside because he has remarkably archaic features in his narrative. Implicitly, and rather surprisingly, this is supported by ps.-ʿUmar-e Khayyāmī,⁸⁵ Nawrūznāme, p. 9, which tells that in the beginning of his rule Ḍaḥḥāk ruled justly, which is directly contrary to the main tradition.⁸⁶

At whatever time Ḍaḥḥāk adopted his, or his snakes’, unnatural diet, all early sources agree, if they mention the matter at all, that he was fed by only one man, Armāyīl.⁸⁷ For the entrance of the second nobleman/cook we have to wait until Firdawsī himself. But was he the inventor of the second cook?

Most of the earlier sources have disappeared, but Arabic and Persian texts that derive their material from the lost sources help us partially to reconstruct the material in circulation before Firdawsī.

The earliest source, after Firdawsī and al-Thaʿālibī, to have two cooks is the anonymous Mujmal, which mentions them by name (p. 35/40–41). In the same passage the author quotes a verse (Z309) by Firdawsī. By the 13th century Firdawsī had attained great fame and it is easy to find sources following his version of the story, but it should be emphasized that until the 13th century the existence of two cooks is rarely mentioned and the scene with one cook, or vizier, remains standard throughout the twelfth century.

So far, Firdawsī seems the obvious inventor of the second cook, but the question is not as simple as one might think. Al-Thaʿālibī’s Ghurar closely agrees with Firdawsī in this episode, as well as in many others, though also using al-Ṭabarī and other sources. Al-Thaʿālibī, too, has two cooks with these names. Did he use Firdawsī as one of his sources or do both authors derive the second cook from the lost common source, the Prose Shāhnāme?

As we have seen, the Prose Shāhnāme was the source of both Firdawsī and al-Thaʿālibī (Chapters 4.4–6). In this episode, there is one significant detail that strengthens the case and shows that al-Thaʿālibī did not base his translation on Firdawsī’s text. Firdawsī mentions (Z35) that whenever a group of two hundred (duwīst) men, rescued from the kitchen, had been collected, or in a variant twenty (bīst), they were sent off to the wilderness. The rhyme (kīst) fixes the possible readings to 200 or 20. Al-Thaʿālibī, however, speaks of groups of ten (Ghurar, p. 25). When he wants to embellish his text al-Thaʿālibī freely elaborates his source by adding maxims or using rhymed prose, but he rarely invents unnecessary details. Moreover, the number of the men does not seem to be an issue in any early source and one wonders why in his prose he should have changed the original number.⁸⁸ Firdawsī, on the other hand, has a possible reason for doing so because of the rhyme, although one has to admit that he would have mastered rhymes well enough to keep the number had he wanted to do so. But as the exact number is of no great importance he may well have changed the original “ten” to “twenty” for an easy rhyme.

On the other hand, we come across certain difficulties with Gardīzī’s Zayn, which contains a version with only one cook. Gardīzī’s version is very similar to both Firdawsī’s and al-Thaʿālibī’s, though there are significant differences, which show that the author cannot be dependent, or solely dependent, on Firdawsī. The combat scene between Ḍaḥḥāk and Afrīdūn (pp. 69–70) is firm proof that Gardīzī used another source or other sources. The scene is full of seemingly archaic magic, Ḍaḥḥāk taking the shape of a sparrowhawk to get on the roof of the pavilion, kūshk, whereas in Firdawsī’s version he does the same prosaically with the help of his lariat.⁸⁹ On the other hand, there are detailed lexical links between Gardīzī and Firdawsī, including the very significant use of the word maḥḍar (Gardīzī, p. 68; four occurrences in Z210–215), which cannot be a coincidence. In Firdawsī, this manifestly Arabic word calls attention to itself. The scene in which it is used is rarely found in other sources and it is even lacking in al-Thaʿālibī. If we assume it was invented by Firdawsī, we encounter two difficulties. Why did Firdawsī break his habits and use a manifestly Arabic word where Persian words would easily have been available?⁹⁰ And secondly, how does Gardīzī end up using that particular word? In short, we seem to be in a situation where we have to assume that both authors are here making use of the same source, which, however, only had one cook, or vizier.⁹¹

Incidentally, also in another case Gardīzī and Firdawsī agree with each other as opposed to the Prose Shāhnāme, as documented in its Older Preface. The Preface mentions that Afrīdūn stopped with his foot the stone his envious brothers had set rolling down upon him.⁹² Firdawsī makes him use magic (Z291)⁹³ and Gardīzī implies the same by making (Zayn, pp. 68–69) him stop it with his word. Al-Thaʿālibī does not have this scene.

There may also be a third significant similarity between Firdawsī and Gardīzī. In several early versions of the story, Armāyīl is the Vizier of Ḍaḥḥāk, but in Firdawsī a mysterious character called Kundraw takes this role and also warns his master of the unwelcome guests that had stormed his harem, shabistān. In the notes (on Z369) to his edition, Khaleghi-Motlagh (2001): 98, takes up the possibility that the name is a corrupt form of Gandarw, another pre-Islamic dragon, but provides no evidence for this.

It seems that the name is attested, besides Firdawsī, in only two sources, both intriguing in their ways. Mujmal, p. 71/89, refers to a certain Kundrawaq. The author of the Mujmal often follows Firdawsī, even quoting his verses and mentioning him by name, so there is a proven dependency of the work in general on Firdawsī.

But there are difficulties. The Mujmal does not place the character into the story, merely mentioning him at the end of Ḍaḥḥāk’s rule as his wakīl,⁹⁴ in a way that closely resembles the style of early chronicles. But how do we explain the form Kundrawaq? The prosaic author of the Mujmal had no need to change the name, but again the reverse is true: for a poet writing in the mutaqārib, the name Kundrawaq is difficult, as it should regularly become Kundĕrawaq, with two short syllables following each other, and other options are equally unpersuasive. The only way to solve the problem is to posit another source (possibly with a chronicle structure) using the name Kundrawaq, which the author of the Mujmal has kept, while Firdawsī has changed it to fit the metre. The final Q would speak for an Arabic source, as the expected Persian form would be *Kundrawag, which is not attested in any of the preserved sources.

As the author of the Mujmal knew Firdawsī, we cannot know from which source he took the second cook. One should, though, note that the detail of the two cooks comes in the middle of a passage quite unrelated to what is given in the Shāhnāme of Firdawsī, and the two cooks are said to have come to Ḍaḥḥāk’s service after he had ruled for 700 years, a detail lacking in Firdawsī, but supported by Balʿamī, Tārīkh, p. 98; Tārīkhnāme, p. 102. Without this detail, the passage on the cooks would evidently be an unmarked interpolation from Firdawsī, as it comes somewhat abruptly and interrupts the narrative. It is quite possible, perhaps even probable, that the author of the Mujmal has throughout his book used Firdawsī only as a secondary source, excerpting him whenever convenient but basing his narrative on other sources. Thus, he could well have changed his main source’s Armāyīl to Armāyil and Karmāyil by inspiration from Firdawsī.

The name Kundraw/Kundrawaq is also found, albeit in a somewhat garbled form, in Gardīzī, Zayn, pp. 69–70, which mentions a treasurer (ganjwar), who performs more or less the same function as Firdawsī’s Kundraw. It is probable that ganjwar is a corruption of either Kundraw, which it rather closely resembles, or Kundrawaq (or *Kundrawaj), which is not far off either.⁹⁵ Thus, it does not help in deciding which of the forms is the older, but again it shows the dependence of Gardīzī on either Firdawsī or their common source.

The Mujmal is a compilation that uses various interwoven sources⁹⁶ and it is quite possible its author took the passage on Armāyil from another source, but added the second cook from Firdawsī. In a similar fashion, he has added much material from the Garsāsbnāme, which he mentions among his sources on p. 2/2, and has woven this into his narrative, which otherwise follows other sources, Firdawsī and Gardīzī virtually ignoring Garsāsb.

If we focus on the word maḥḍar and claim that Firdawsī and Gardīzī used the same source, we come across the difficulty that Gardīzī (Zayn, p. 70) only knows one Vizier, Azmāʾīl (p. 70 – p. 354 reads Armāyīl). If the common source of al-Thaʿālibī and Firdawsī already had two cooks, then Gardīzī should agree with them if he, too, used the very same source as Firdawsī as the use of the word maḥḍar would imply. Hence, it is easier to assume that this common source only had one cook and the second cook was added by Firdawsī.

That the second cook was not present in Firdawsī’s source might further be supported by some linguistic evidence in Firdawsī’s Shāhnāme. In Z30–32 and 35–36, Firdawsī lapses into the use of the singular when speaking of the cooks. Such use of singular forms for plurals is not unknown in early Persian and without any supporting evidence, one might take this as an admissible linguistic lapse in marking the plural. Considering, however, all the evidence, the verses may well echo a text where there was only one cook, Armāyīl. Firdawsī would have added another character but not been consistent when versifying his source and making the necessary changes.

This, however, would mean that al-Thaʿālibī, who also has two cooks, must have used Firdawsī, besides their common source. Thus, it seems that the only way out of this labyrinth is to posit that al-Thaʿālibī did occasionally use Firdawsī as a source, as already suggested in Chapter 4.4.

Why was the second cook invented out of thin air? Whoever did this, and I believe it was Firdawsī, probably did it in order to heighten the dramatic effect of the narrative by letting the two discuss the matter with each other and also perhaps to parallel the two victims. One is also tempted to see the mirroring scene of Iblīs as a cook due to an acute and conscious literary mind. However, we know this scene to have been invented before Firdawsī and to originate with another author: creative literary minds had been working with the material even before Firdawsī. It seems, though, that the “vegetarian” scene may well have developed hand in hand with turning a vizier into a cook. Both episodes are dramatic and thematically tied together. Both have cooks that are not what they seem and serve the king only to drive through their own agenda and manage to do so without arousing the king’s suspicions. Obviously, they are the creation of a fine literary mind, or several fine literary minds.

What has this little study on Armāyīl taught us? Any analysis of Firdawsī should be based on a detailed study of both the epic and the early testimonies. To understand the working of the literary mind we have to know what materials the author had at hand to build on. An analysis of the text of the Shāhnāme which does not look at its sources may become seriously flawed. We all too easily think of Firdawsī as handling raw material to forge his unique epic, whereas in reality he may be closely following earlier sources of some literary value. This does not, however, diminish the value of Firdawsī’s Shāhnāme. Whatever the author’s relation to his sources, the final result is a superb piece of epic poetry. This, though, should not blind us to the fact that the Book of Kings tradition can boast of more than one creative mind.

Secondly, we should not draw a line between history and literature. The Shāhnāme belongs to world literature, but it tapped historical sources and was itself used as a serious source for Persian history. Even today it provides materials for the study of Sasanian history. Similarly, not all historical sources are devoid of literary interest.⁹⁷ We have no way of clearly determining what kind of work the Prose Shāhnāme was, but undoubtedly it was a valuable literary work in its own right.