Among books that all know and none has seen, the Middle Persian Khwadāynāmag presents a towering figure. Not only has this “official history of the Sasanids” been lost so that not one sentence in Middle Persian can safely be attributed to it, but its Arabic translation(s) have also vanished into air almost as thin.

Yet every scholar in the field seems to know the book. For some, Firdawsī’s Shāhnāme virtually equals the Khwadāynāmag, to others some existing Arabic work, or a combination of quotations from existing sources, can with little hesitation be used as indicative of the contents of the Khwadāynāmag. Most do not even stop to ask themselves what relation a certain text may actually have to the Khwadāynāmag, but speak summarily of “the Khwadāynāmag tradition” with, or usually without, defining the term.

The lack of critical discussion about the Khwadāynāmag is surprising, granted its importance for Late Antique and Early Islamic historiography. Not only is it important as part of the rather scanty non-religious Pahlavi literature, it is also crucial for the reconstruction of the historical events of the Sasanid period and for understanding the genesis of Arabic historical writing and the relation of Firdawsī to his sources. All are major questions in their various fields.

Let us take but two examples, one on the Arabic and the other on the Persian side. The question of the genesis of the Arabic historiographical tradition is almost without exception addressed from an Islamic viewpoint, through ḥadīths and akhbār, and it has become commonplace to claim that historical books started being written by Ibn Isḥāq and his generation on the basis of information preserved orally or in brief notes concerning the Prophet Muḥammad and the birth of Islam. Such comments ignore the fact that the Khwadāynāmag was translated into Arabic as a complete book some decades before the death of Ibn Isḥāq. As the text was well known to early historians who wrote in Arabic, it cannot be separated from the main tradition of Arabic historiography or implied to have been influential only within the sphere of the translation movement, but not among historians themselves. For the Persian parts of their works, Ibn Qutayba, al-Dīnawarī, and al-Ṭabarī are to a large extent ultimately dependent on Middle Persian material, and they must unavoidably have been influenced by Middle Persian ways of writing history.

On the Persian side, the question of Firdawsī’s sources may be taken as an example of the range of Khwadāynāmag studies. In her book The Oral Background of Persian Epics (2003), Kumiko Yamamoto opines (p. xix) that the study of the sources of Firdawsī’s Shāhnāme has come to a dead end and other viewpoints are needed. While most certainly right when it comes to a need for fresh viewpoints, Yamamoto does not quite do justice to source studies. It is true that there have been tedious repetitions in the field of Firdawsī’s source studies, but this is not due to the question itself, but to the restricted use of source material for such studies. A fuller analysis of the Khwadāynāmag and the Arabic and Persian literature dependent on it helps to settle this key text to its rightful place, after which we will be able to approach Firdawsī’s Shāhnāme from a fresh viewpoint.

This book, however, is neither about Arabic historiography nor Firdawsī’s Shāhnāme. Its focus is on the lost Middle Persian Khwadāynāmag and its translations and reverberations in later literature. However, while trying to clear the ground by showing what there was between the eighth to tenth centuries and what the relations of individual texts were I also hope to be able to offer some freshness to both these fields.

I discuss the Khwadāynāmag from various viewpoints. Chapter 1 clarifies the terminology and introduces the pre-Islamic sources that are relevant for the study of the Khwadāynāmag. Some of the Pahlavi texts analysed in this chapter are from the Islamic period but they tap older sources. Chapter 2 describes the translation culture in the centuries when the Khwadāynāmag was translated into Arabic and gives an overview of what was, in general, translated from Middle Persian into Arabic.

Chapter 3 moves on to the Arabic translations of the Khwadāynāmag, and Chapter 4 discusses the various narratives of Persian national history (Books of Kings, Shāhnāmes) in Persian until Firdawsī and even slightly later. Chapter 5 consists of two case studies, where the potential content of the Khwadāynāmag is studied through an analysis of the works that, in one way or another, have a relation to the Khwadāynāmag.

Chapter 6 comes back to the questions laid out in the first chapter and sums up the discussion in this book, which is concluded by Chapter 7, where the most important passages from Arabic and Persian sources are translated for the benefit of a reader who does not readily have at hand the various editions from which they have been culled or has not enough fluency in either Arabic or Persian.

Technical Notes

This book uses materials in mainly three languages, Pahlavi, Arabic, and Classical Persian, most of them coming from a range of 700 years (500–1200 ad). In transliterating Pahlavi, I have used the system of David MacKenzie (1971): x–xv, which seems to have become the standard in Middle Persian studies, except for using kh instead of x and sh instead of š.

The case of Arabic and Persian is slightly more complicated. As many terms, personal names, and book titles are used in both Arabic and Persian sources, it would be confusing to use transliterations separately for the two languages. The majority of the material comes from Arabic sources, and I have adopted the standard transliteration of Arabic for both languages (ei-Three with some slight modifications), thus writing Firdawsī and al-Thaʿālibī, irrespective of whether I am quoting an Arabic or a Persian source. Likewise, I use three short vowels, a, i, u, and three long ones, ā, ī, ū, thus ignoring the majhūl vowels ē and ō, while well aware of the fact that when the majority of the Persian texts used for this study were written, they were still pronounced separately from ī and ū. Likewise, postvocalic D is written in Persian texts usually as D, even though it was pronounced as dh. For consistency’s sake, I use D for Dh even in the rare cases where the editor of a text has opted for a Dh, thus writing būd instead of the more correct būdh. In order to distinguish between Z+H and Zh (as well as T, D, K, G, S + H), I write the former combination with an apostrophe (z’h), e.g., Nuz’hatnāme.

However, I have made some exceptions, mainly to comply with the prevailing usages in the field. Thus, the ezafet is transliterated as -e (or -ye), the final vowel written in the Persian script with -H as -e, and the conjunction as o-. Likewise, I write mōbad and hērbad (even in transliterated Arabic passages), instead of mūbad and hīrbad, as they are used as technical terms in scholarly literature. Persian verses are transliterated grammatically, i.e., without taking into account the changes in vowel lengths and other metrical exigencies.

The names of characters playing a role in Persian national history have usually been given in the form in which they appear in each source. Thus, Isfandiyār and Isfandiyād refer to the same person, as do Wishtāsp, Bishtāsf, Gushtāsp, Gushtāsb, and Gushtāsf. Usually, the variants should be understandable in the context, and when not they have been explained. The reason for keeping the name forms as they are attested in the texts is that they may be helpful in understanding the relations of the texts and detecting an author’s sources.

The term Classical Persian is used when there is a possibility of confusion between Middle and Classical Persian. The term Archaic Persian is not normally used. The term Pahlavi specifically refers to the so-called Book Pahlavi of Zoroastrian literature, whereas Middle Persian is a larger term, covering also other forms of contemporary language. Both are used in this book.

When referring to the Islamic period, I usually give both Hijri and ad dates, in that order. Thus, e.g., 350/961 refers to the year 350 ah = 961 ad.

In the case of certain important books which are available in two critical editions and/or a commonly used translation, I give references to both editions (and the translation) to help the reader find the passage in his copy. The editions are separated by a slash, and the translation is separated by a double slash. The following works are referenced in this way:

Ibn al-Nadīm, Fihrist: ed. Tajaddud/ed. Flügel//trans. Dodge (1970).

al-Bīrūnī, Āthār: ed. Adhkāʾī/ed. Sachau//trans. Sachau (1923).

al-Masʿūdī, Tanbīh: ed. de Goeje//trans. Carra de Vaux (1896).

Mujmal al-tawārīkh: ed. Najmabadi–Weber/ed. Bahār.

al-Ṭabarī, Ta⁠ʾrīkh: ed. de Goeje et al.//trans. Rosenthal et al. (1987–2007).

Thus, quoting, e.g., the Fihrist, I will primarily use the Tajaddud edition, in more important cases also supplying references to Flügel’s edition and Dodge’s translation. Thus, e.g., Fihrist, p. 305/245//589 refers to ed. Tajaddud, p. 305, ed. Flügel, p. 245, and trans. Dodge (1970): 589. (The edition of Fuʾād Sayyid seriously suffers from unindicated emendations and will not be used in this study, except on rare occasions. Dodge’s translation is often faulty. Both should be used with care.)

If not otherwise indicated, all translations are mine, even in the cases where I give a reference to the standard translation.

I am grateful to the publishers for permission to reuse materials that have previously been published in the following articles:

“al-Maqdisī and His Sources,” Orientalia Lovaniensia Analecta 207 (2012): 151–163; “al-Kisrawī and the Arabic translations of the Khwadāynāmag,” Studia Orientalia 114 (2013): 65–92; “Armāyīl and Garmāyīl: the Formation of an Episode in Firdawsī’s Shāhnāme,” wzkm 104 (2014): 87–103; “Ibn al-Muqaffaʿ and the Middle Persian Book of Kings,” Orientalia Lovaniensia Analecta 254 (2017): 171–184; “Rustam in Arabic Literature and the Middle Persian Khwadāynāmag,” wzkm 107 (2017).

I am also grateful to Dr Ilkka Lindstedt (Helsinki), who kindly read through a draft of this book and gave many valuable suggestions that I have been able to make use of in the final version.