Al-ʿUṣūr al-Wusṭā 23 (2015): 87-125 Ibn Aʿtham and His History * Editor’s Introduction The editors of al-ʿUṣūr al-Wusṭā are delighted to publish this long-awaited piece by Lawrence I. Conrad on Ibn Aʿtham al-Kūfī and his Kitāb al-futūḥ. The article was written on the basis of two papers presented in 1992 (see initial note) and subsequently prepared for publication. It has circulated among colleagues, but, for various reasons, never appeared in print. Professor Conrad, with characteristic generosity, has given us per- mission to publish the text. It stands as a monumental piece of scholarship and the most comprehensive study on the subject to date. By way of introduction, a few historiographical comments are in order. Limited attention has been devoted to Ibn Aʿtham since the early 1990s. Conrad himself wrote a brief entry for The Routledge Encyclopedia of Arabic Literature [London and New York: Routledge, 1998, 314], summarizing his findings and arguing that Ibn Aʿtham flourished in the early third/ninth century. There he rejects Ibn Aʿtham’s conventional death date of 314/926-7 as “an old Orientalist error.” Conrad went on to advocate for the earlier date in subsequent publications (e.g., “Heraclius in Early Islamic Kerygma,” in G.J. Reinink and B.H. Stolte (eds.), The Reign of Heraclius (610-641): Crisis and Confrontation [Leuven: Peeters, 2002], 132). This view was adopted by several scholars and corroborated on the basis of the content of the work. (See in particular A. Borrut, Entre mémoire et pouvoir: l’espace syrien sous les derniers Omeyyades et les premiers Abbassides (v. 72-193/692-809) [Leiden: Brill, 2011], index; E. Daniel, “Ketāb al- Fotūḥ,” Encyclopaedia Iranica Online, 2012 [http://www.iranicaonline.org/articles/ketab-al-fotuh]; J. Schein- er, “Writing the History of the Futūḥ: The Futūḥ-Works by al-Azdī, Ibn Aʿtham, and al-Wāqidī,” in P.M. Cobb (ed.), The Lineaments of Islam: Studies in Honor of Fred McGraw Donner [Leiden: Brill, 2012], 151-176). Conrad’s early dating of Ibn Aʿtham has been challenged recently by Ilkka Lindstedt (“Al-Madāʾinī’s Kitāb al-Dawla and the Death of Ibrāhīm al-Imām,” in I. Lindstedt et al. (eds.), Case Studies in Transmission [Mün- ster: Ugarit-Verlag, 2014], esp. 118-123; and “Sources for the Biography of the Historian Ibn Aʿtham al-Kūfī,” in Jaakko Hämeen-Anttila, Petteri Koskikallio, and Ilkka Lindstedt (eds.), Proceedings of Union Européenne des Arabisants et Islamisants 27, Helsinki, June 2nd-6th, 2014 [Leuven: Peeters, forthcoming]). On the basis of new biographical evidence, Lindstedt argues that Ibn Aʿtham actually flourished in the late third/ninth-early fourth/tenth century. Mónika Schönléber, a doctoral candidate at Pázmány Péter Catholic University (Budapest), is preparing a critical edition of the first portion of the Kitāb al-futūḥ, and her work will help clarify the complex histo- ry of the text (see, for now, her “Notes on the Textual Tradition of Ibn Aʿtham’s Kitāb al-Futūḥ,” in Jaakko Hämeen-Anttila, Petteri Koskikallio, and Ilkka Lindstedt (eds.), Proceedings of Union Européenne des Ara- bisants et Islamisants 27, Helsinki, June 2nd-6th, 2014 [Leuven: Peeters, forthcoming]). Regardless of whether one accepts it as an early third/ninth-century text or a product of the late third/ ninth-early fourth/tenth century, the Kitāb al-futūḥ stands as an invaluable source. It is hoped that the publi- cation of Conrad’s meticulous and elegant study will foster more research on what remains a much-neglected text. We publish the text below in its original form. — Antoine Borrut Lawrence I. conrad University of Hamburg 88 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) It is probably a general rule of thumb that the larger and earlier an Islamic historical text is, the more likely it is to attract the attention of modern scholars. If this is so, then the rule’s most glaring exception is the Kitāb al-futūḥ of Abū Muḥammad Aḥmad ibn Aʿtham al-Kūfī. Though a work of considerable bulk, running to over 2700 pages in the Hyderabad edition,1 a text which covers many aspects of the first 250 years of Islamic history, and one which has been known since the mid-nineteenth century, at least in its Persian translation, the Kitāb al-futūḥ has never enjoyed the attention one might have expected it to receive. One reason for this is surely that Ibn Aʿtham has had, since the days of Brockelmann, a bad reputation as a purveyor of—to use his phrasing—“a fanciful history” written from a Shīʿī viewpoint.2 This tends to invite the conclusion that a careful reading of the Kitāb al-futūḥ would be a waste of time; but to this one might easily reply that regardless of whether a work strikes modern observers as good or bad history, it may reveal much about its cultural tradition and thus—for that reason alone—prove to be eminently worthy of investigation. In passing it must be said that irrespective of the extent to which it can or cannot be made to give up “historical facts”, this fascinating text has much to tell us about how history was perceived and transmitted in early Islamic times. In my remarks here, however, I will address only a limited number of points central to further work on the text. On some questions, including that of who Ibn Aʿtham himself was, the complexities of the extant material allow details to emerge only in rather piecemeal fashion, and an attempt will be made at the end of this study to summarize conclusions that have been drawn at various earlier points. It must be conceded from the outset that the basis for historiographical study of this history is not ideal. As with the Taʾrīkh al-rusul wa-al-mulūk of al-Ṭabarī (wr. 303/915), the textual tradition of the Kitāb al-futūḥ of Ibn Aʿtham consists of a number of incomplete Arabic MSS and a later Persian translation which sometimes manifests important discrepancies from the wording of the Arabic. Coverage of the text, as presented in the Hyderabad edition, can be summarized as follows:3 * This study arises from two different papers presented at the annual meeting of the American Oriental Society, Cambridge, Mass., on 29 March 1992, and at Leiden University on 20 May 1992. I am grateful to the participants in those sessions for their valuable discussion, and especially to Professors Fred M. Donner and Wadād al-Qāḍī for their comments and suggestions. 1. Ibn Aʿtham al-Kūfī, Kitāb al-futūḥ, ed. Muḥammad ʿAbd al-Muʿīd Khān et al. (Hyderabad: Dāʾirat al-maʿārif al-ʿuthmānīya, 1388-95/1968-75) in eight vol umes. The recent three-volume edition by Suhayl Zakkār (Damascus: Dār al-fikr, 1412/1992) appeared too late to be taken into consideration here, but does not, in any case, replace the Hyderabad edition. Zakkār’s work does not use the Gotha MS, and so is missing the first 485 pages of the Hyderabad text; it also takes no account of the Persian translation, and thus fails to notice many lacunae. The apparatus criticus cites Qurʾānic quotations, draws attention to significant passages in a few parallel works, and provides some useful explanations of terms, but is very weak where consideration of variant readings is concerned. 2. GAL, SI, 220; EI1, II, 364b. 3. In addition to these MSS, Ambrosiana H-129, copied in 627/1230 and not used by the Hyderabad editors, covers the text from the conquest of al-Rayy and al-Dastabā (II, 62:12) to the murder of ʿAlī ibn Abī Ṭālib (IV, 147ult). See Eugenio Griffini, “Nuovi testi arabo—siculi”, in Centenaria della nascita di Michele Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 89 MS siglum Text covered Gotha 1592 al-aṣl I, 1:1-II, 146ult. Ahmet III 2956 al-aṣl II, 147:1-VIII, 354:7 (end) Chester Beatty 3272 d II, 147:1-VI, 100:3 Mingana 572 br III, 108ult-VI, 97:11 Persian translation al-tarjama I, 1:1-V, 251:34 It can immediately be seen that the first part of the book, extending to almost 500 pages, is fully attested only by the Gotha MS; at the end, only the Ahmet III codex extends past the first third of volume VI. When one adds to this the fact that the Chester Beatty MS is clearly a descendent of the Ahmet III exemplar, it becomes clear that through the majority of the book, the manuscript tradition provides rather thin testimony for the fixing of the text. This problem is rendered more serious by other difficulties. Loss of single or multiple folios, and even of entire signatures, has resulted in a number of major gaps in the Arabic text,5 and other shorter lacunae are numerous. Quite often one encounters passages where an erasure, probably to delete an incorrectly copied word or phrase, has been left unfilled. Passages in verse have perhaps suffered worst: poems surviving in the Persian translation are in the Arabic often dropped entirely, or represented only by the maṭlaʿ or some other illustrative verse. Though some clarification of this problem can be proposed, it is still not entirely clear how the Persian text can be used to check the Arabic, since there seem to exist multiple versions of this Persian rendering. Some of these and other difficulties will return to our attention below. At this point it will suffice to observe that while the Hyderabad edition usually draws the reader’s attention to such problems, it seldom resolves them in a way conducive to a critical historiographical assessment of the Arabic text. Amari (Palermo: Virzì, 1910), I, 402-15. The Bankipore MS Khuda Bakhsh 1042, copied in 1278/1861, contains an ʿAlid version of Saqīfat Banī Sāʿida and the election of Abū Bakr, an account of the ridda wars, and a few pages on the conquest of Iraq; the MS has recently been described by Muḥammad Ḥamīd Allāh as “the unique manuscript” of the Kitāb al-ridda of al-Wāqidī (d. 207/823), in the recension of Ibn Aʿtham, and published as such in his Kitāb al-ridda wa-nubdha min futūḥ al-ʿIraq (Paris: Editions Tougui, 1409/1989). But a decade earlier two other scholars had already independently noticed that this was nothing more than an extract from Ibn Aʿtham’s own history (= Hyderabad ed. I, 2:5-96:6, ending in the midst of a long lacuna in the Gotha MS); see Fred M. Donner, “The Bakr b. Wāʾil Tribes and Politics in Northeastern Arabia on the Eve of Islam,” Studia Islamica 51 (1980), 16 n. 2; and Miklos Muranyi’s publication of the section on the election of Abū Bakr in his “Ein neuer Bericht über die Wahl des ersten Kalifen Abū Bakr,” Arabica 25 (1978), 233-60. Ḥamīd Allāh’s publication is nevertheless useful, for reasons which will emerge below, and here it will be referred to as “Ibn Aʿtham, Bankipore Text”. 4. The recension of the Persian translation available to me (see n. 42 below) begins somewhat differently than the Arabic, but this discrepancy is not noticed in the Hyderabad edition, which usually does comment on such anomalies, but uses a different edition of the Persian text. 5. The most serious of these are at I, 5:4-5, 91:2-100:1, 318:7-324:1, 334:2-349:1; II, 95:2-107:1, 193:3-208:1; IV, 206:6-209:1. The first of these lacunae, and part of the second, have been filled by Ibn Aʿtham, Bankipore Text, 22:10-42:2 (cf. Muranyi, “Ein neuer Bericht,” 239-47), 128:9-137ult. 90 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) Date of Composition A fundamental point of departure is that of when the author lived and when he wrote his history. The difficulty here is that as a historical personality Ibn Aʿtham was almost entirely unknown to later writers. Yāqūt (d. 626/1229), the only medieval biographer who has original information on him, will return to our attention below. Here we may simply note that he knows nothing about Ibn Aʿtham’s life or date of death, and can offer little information beyond what might be gained by perusing his works (e.g. knowledge of Ibn Aʿtham’s Shīʿī sympathies) or by consulting a rijāl al-sanad compendium (i.e. his reputation among ḥadīth transmitters as ḍaʿīf).6 Ibn Ṭāwūs (d. 664/1266) refers to him by name and quotes from the Kitāb al-futūḥ, but seems to know nothing about him personally.7 Al-Ṣafadī (d. 764/1363) and Ibn Ḥajar (d. 852/1449) both have entries for Ibn Aʿtham, but all of their information comes from Yāqūt.8 The copyist of the Ahmet III MS, writing in 873/1468- 69, refers to our author as Ibn Aʿtham “al-Kindī”, thus suggesting his membership of the southern tribe of Kinda, but this is almost certainly a misreading of “al-Kūfī”.9 Ḥājjī Khalīfa (d. 1067/1657) mentions Ibn Aʿtham twice in his Kashf al-ẓunūn, but he has no personal details about him and simply describes him as the author of a futūḥ book translated by al-Mustawfī, to whom we shall return below.10 Here we have to do with conclusions reached only on the basis of access to a subject’s book, in this case the Persian translation of the Kitāb al-futūḥ. Al-Majlisī (d. 1110/1697) also made use of the work in his vast compendium of Shīʿī traditions, but seems not to have known anything about its author.11 This dearth of information has not deterred modern scholarship from offering a range of possibilities for the period to which Ibn Aʿtham belongs. An early attempt to establish the identity of Ibn Aʿtham was made by William Nassau Lees, one of the first Western editors of futūḥ texts. In the introduction to his editio princeps of the pseudo-Wāqidī Futūḥ al-Shām, Lees proposed that Ibn Aʿtham was to be identified as Abū Muḥammad Aḥmad ibn ʿĀṣim al-Balkhī a muḥaddith who died in 227/841-42.12 But for several reasons this argument, such as it is, must be rejected. First, it is at least curious, if Aḥmad ibn ʿĀṣim is our author, that none of the many accounts of him mentions that this man was the author 6. Yāqūt, Irshād al-arīb ilā maʿrifat al-adīb, ed. D.S. Margoliouth, 2nd ed. (Leiden: E.J. Brill, 1923-31), I, 379:1-8, no. 104. 7. Ibn Ṭāwūs, Kashf al-maḥajja li-thamarat al-muhja (Najaf, 1370/1950), 57, cited in Etan Kohlberg, A Medieval Muslim Scholar at Work: Ibn Ṭāwūs and His Library (Leiden: E.J. Brill, 1992), 358-59, with the observation that this passage is not to be found in the Arabic text we have today. 8. Al-Ṣafadī, Al-Wāfī bi-l-wafayāt, VI, ed. Sven Dedering (Wiesbaden: Franz Steiner Verlag, 1972), 256:7-11 no. 2740; Ibn Ḥajar al-ʿAsqalānī, Lisān al-mīzān (Hyderabad: Dāʾirat al-maʿārif al-niẓāmīya, AH 1329-31), I, 138:16-18 no. 433. 9. Ibn Aʿtham, Kitāb al-futūḥ,VI, 100 n. 4; VIII, 354 n. 7. 10. Ḥājjī Khalīfa, Kashf al-ẓunūn ʿan asāmī al-kutub wa-al-funūn, ed. Şerefettin Yaltkaya and Kilisi Rifat Bilge (Istanbul: Maarıf Matbaası, 1941-47), II, 1237:15, 1239:27-29. 11. Al-Majlisī, Biḥār al-anwār (Beirut: Muʾassasat al-wafāʾ, 1403/1983), 1, 25:9. 12. The Conquest of Syria Commonly Ascribed to Aboo ʿAbd Allah Mohammad b. ʿOmar al-Wáqidí, ed. W. Nassau Lees (Calcutta: F. Carbery, 1854-60), I, vii. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 91 of a book—of any description. Such information is routinely given in the various types of biographical compendia. Second, while the name Aḥmad was not yet common in the second and third centuries ah, the kunya Abū Muḥammad certainly was, and the fact that two Aḥmads shared the same kunya in no way suggests, much less proves (as Lees seemed to believe), that they were one and the same person. Indeed, the case for the opposite conclusion is compelling. Aḥmad ibn ʿĀṣim al-Balkhī is the subject of numerous notices in rijāl al-sanad compendia and is named as one of the authorities cited by al-Bukhārī (d. 256/870),13 but nowhere is there any hint of a father or grandfather named Aʿtham, i.e. some ancestor who would account for why the Aḥmad ibn ʿĀṣim of the rijāl compendia would be called Aḥmad ibn Aʿtham in the Kitāb al-futūḥ. Similarly, no one with any information on Ibn Aʿtham mentions an ancestor named ʿĀṣim. As the two names are not orthographically similar, this discrepancy clearly establishes that no case can be made for the argument that the two names refer, as Lees thought, to the same historical figure. In fact, such an identification is precluded by the fact that Aḥmad ibn ʿĀṣim, as an informant of al-Bukhārī, must have been a Sunnī muḥaddith. As we shall see below, however, the author of the Kitāb al-futūḥ was a strident Shīʿī; when he cites ḥadīth, he almost exclusively quotes ʿAlid legitimist, Shīʿī, and virulently anti-Umayyad traditions from the Prophet and the Imams. While one must guard against the temptation to project back into early Islamic times Sunnī/Shīʿī differences which only emerged later,14 most of Ibn Aʿtham’s traditions clearly comprise material which no authority of al-Bukhārī would have taken seriously, much less transmitted. In his work on Arabic historians, Wüstenfeld gives the date of Ibn Aʿtham’s death as ah 1003 (= AD 1594-95),15 which is the date cited in Flügel’s edition of Ḥājjī Khalīfa.16 But in the more recent and far superior Istanbul edition of the Kashf al-ẓunūn, based on the author’s autograph, the space for the date is left blank; the date in Flügel’s edition may well have been erroneously carried up from the next entry below it, where the text in question is also by an author said to have died in ah 1003. Further, such a date is impossible since, as we shall see momentarily, Ibn Aʿtham’s Futūḥ had already been translated into Persian four centuries earlier. 13. See, for example, al-Bukhārī (d. 256/870), Al-Taʾrīkh al-kabīr (Hyderabad: Dāʾirat al-maʿārif al-ʿuthmānīya, ah 1360-64), 1.2,6:3-4 no. 1500; Ibn Abī Ḥātim (d. 327/938), Al-Jarḥ wa-al-taʿdīl (Hyderabad: Dāʾirat al-maʿārif al-ʿuthmānīya, 1371-73/1952-53), I.1, 66: 10-11 no. 118; Ibn Ḥibbān al-Bustī (d. 354/965), Kitāb al-thiqāt (Hyderabad: Dāʾirat al-maʿārif al-ʿuthmānīya, 1393-1403/1973-83), VIII, 12: 3-4; al Mizzī (d. 742/1341), Tahdhīb al-kamāl fī asmāʾ al-rijāl, ed. Bashshār ʿAwwād Maʿrūf (Beirut: Muʾassasat al-risāla, 1985/1306-proceeding), I, 363: 2-0 no. 55; al-Dhahabī (d. 748/1348), Mizān al-iʿtidāl, ed. ʿAlī Muḥammad al-Bijāwī (Cairo: Īsā al-Bābī al-Ḥalabī, 1382/1963), I, 106: 2-4; Ibn Ḥajar, Tahdhīb al-tahdhīb (Hyderabad: Dāʾirat al-maʿārif al-niẓāmīya, ah 1325-27), I, 46: 4-11 no. 76. 14. On this problem, see Andrew Rippin, Muslims: Their Religious Beliefs and Practices, I: The Formative Period (London: Routledge, 1990), 103-16. 15. Ferdinand Wüstenfeld, Die Geschichtschreiber der Araber und ihre Werke (Göttingen: Dieterische Verlags-buchhandlung, 1882), 253 no. 541. 16. Ḥājjī Khalīfa, Kashf al-ẓunūn ʿan asāmī al-kutub wa-al-funūn, ed. Gustav Flügel (London: Oriental Translation Fund, 1835-58), IV, 380: 5-6 no. 8907. 92 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) This date is in any case not the one usually cited. Most modern scholarship gives the year of Ibn Aʿtham’s death as ca. 314/926-27: this is the date one finds not only on the title page of the Hyderabad edition itself, but also in studies pertaining to Ibn Aʿtham by, for example, Rieu,17 Brockelmann,18 Griffini,19 Storey,”20 Massé,21 al-Amīn,22 al-Ṭihrānī,23 Cahen,24 Togan,25 Fuat Sezgin,26 Zirikli,27 Muranyi,28 and Ursula Sezgin.29 The apparent security of this death date is reflected in the comments of Brockelmann, who asserts that it is the only information we know about Ibn Aʿtham,30 and Massé, who refers to Ibn Aʿtham as a contemporary of al-Ṭabarī and observes that “il est généralement admis que l’historien arabe Ibn Aʿtham composa ses ouvrages sous le règne du calife Moqtadir et qu’il mourut en 314/926”.31 Here too, however, the ascription is entirely baseless. All scholarship after the publication of Brockelmann’s monumental Geschichte der arabischen Literatur quite naturally takes the date from him, but Brockelmann himself, as well as Rieu and Storey, have it not from any medieval authority, but from a curious bibliography of medieval Islamic texts compiled in St. Petersburg in 1845 by C.M. Frähn.32 As is well-known, Russia in this period was beginning to harbor imperial designs on territories in Central Asia, and 17. Charles Rieu, Catalogue of the Persian Manuscripts in the British Museum (London: British Museum, 1879-83), I, 151a. 18. GAL, SI, 220; EI1, II, 364b. 19. Griffini, “Nuovi testi arabo-siculi,” 407; idem, “Die jüngste ambrosianische Sammlung arabischer Handschriften,” ZDMG 69 (1915), 77. 20. See C.A. Storey, Persian Literature: a Bio-Bibliographical Survey (London: Royal Asiatic Society, 1927-proceeding), I.1, 207 no. 261. 21. Henri Massé, “La chronique d’Ibn Aʿtham et la conquête de l’Ifriqiya,”, in William Marçais, ed., Mélanges offerts à Gaudefroy-Demombynes par ses amis et anciens élèves (Cairo: Institut francais d’archéologie orientale, 1935-45), 85. 22. Muḥsin al-Amīn, Aʿyān al-shīʿa (Damascus: Maṭbaʿat Ibn Zaydūn, 1353-65/1935-46), VII, 428-29. 23. Āghā Buzurg al-Ṭihrānī, Al-Dharīʿa ila taṣānīf al-shīʿa (Najaf: Maṭbaʿat al-Ghazzī, 1355-98/1936-78), III, 220. 24. Claude Cahen, “Les chroniques arabes concernant la Syrie, l’Egypte et la Mésopotamie de la conquête arabe à la conquête ottomane dans les bibliothèques d’Istanbul,” REI 10 (1936), 335. 25. Zeki Velidi Togan, art. “Ibn Aʾsemülkûfî” in Islam Ansiklopedisi, ed. A. Adivar et al. (Istanbul: Maarif matbaasi, 1940-86), V, 702a. 26. GAS I, 329. 27. Khayr al-Dīn al-Ziriklī, Al-Aʿlām, 3rd ed. (Beirut: Dar al-ʿilm li-al-malāyīn, 1969), I, 96b. 28. Muranyi, “Ein neuer Bericht,” 234. 29. Ursula Sezgin, “Abū Mikhnaf, Ibrāhīm b. Hilāl aṯ-Ṯaqafī und Muḥammad b. Aʿtam al-Kūfī über ġārāt,” ZDMG 131 (1981), Wissenschaftliche Nachrichten, *1. 30. EI1, II, 364b. 31. Massé, “La chronique d’Ibn Aʿtham,” 85. 32. C.-M Frähn, Indications bibliographiques relatives pour la plupart à la littérature historico— géographique des arabes, des persans et des turcs (St. Petersburg: Académie impériale des sciences, 1845), 16 no. 53. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 93 in expectation of the usual fruits of conquest, Frähn compiled for the Russian Academy of Sciences what amounted to a wish-list of desirable historical and geographical texts. The work is addressed to “nos employés et voyageurs en Asie” on the assumption that important manuscript treasures could be gained for the Academy by watchful officials and travelers.33 Frähn’s inventory was essentially derived from the Kashf al-ẓunūn,34 and most of the books he lists are lost. As would be expected for a work of this period, Frähn’s list is full of mistakes and erroneous conjectures. Where Ibn Aʿtham is concerned, the death date of 314/926-27 is proposed as a guess—with a question mark after it—and no corroborating evidence is cited. In fact, it seems that no such evidence exists. Here the point of importance is that all modern scholarship citing this date has it ultimately—and only—from Frähn: it has no foundation in the primary source material relevant to the subject of our inquiry. A third date was first noticed independently by C.A. Storey35 and ʿAbd Allāh Mukhliṣ,36 was subsequently rejected by Āghā Buzurg al-Ṭihrānī,37 and has more recently been upheld by M.A. Shaban in his Encyclopaedia of Islam article on Ibn Aʿtham38 and in further detail in his introduction to his book on the ʿAbbāsid revolution.39 The source for this date is the introduction to the Persian translation of the Kitāb al-futūḥ, extant in numerous manuscripts40 and printed in India several times in the nineteenth century. The translator was Muḥammad ibn Aḥmad al-Mustawfī al-Harawī, and in his eloquent but verbose introduction he provides some details important to the background for his work. These may be summarized as follows: Having spent his career serving great men, he says, he had hoped to retire to a life of pious seclusion; but as he had no secure source of income, this proved impossible. Then a powerful but unnamed political figure (referred to. as ṣāḥib al-sayf wa-al-qalam, in Arabic, plus many other honorific titles) took him in, and al-Mustawfī enjoyed some years of esteem and wealth. In ah 596 (= ad 1199-1200) this patron summoned him to Tāybād,41 where al-Mustawfī was honored with further generous patronage and was welcomed into the circle of seven most learned (but again unnamed) scholars. One day, when his patron was present, a member of the assembled company recited some anecdotes from the Kitāb-i futūḥ of khavāja Ibn Aʿtham, who had written this book in ah 204 (= AD 819-20); the patron was so impressed that he asked al-Mustawfī to 33. Ibid., xxvii. 34. Ibid., xxxvii—xxxix. 35. Storey, Persian Literature, 1.2, 1260, in the corrections to his main text. 36. ʿAbd Allāh Mukhliṣ, “Taʾrīkh Ibn Aʿtham al-Kūfī,” Majallat al-majmaʿ al-ʿilmī al-ʿarabī 6 (1926), 142-43. 37. Āghā Buzurg al-Ṭihrānī, Al-Dharīʿa ilā taṣānīf al-shīʿa, III, 221. His argument is the fairly obvious one that a historian who wrote a history in ah 204 could not still have been active more than 100 years later, in the reign of al-Muqtadir. See below. 38. M.A. Shaban, art. “Ibn Aʿtham al-Kūfī” in EI2, III (Leiden: E.J. Brill, 1971), 723a. 39. Ibid., The ʿAbbāsid Revolution (Cambridge: Cambridge University Press, 1970), xviii. 40. See Storey, Persian Literature, I.1, 208-209. 41. I.e. Tāyābādh in the region of Herat. See Yāqūt, Muʿjam al-buldān (Beirut: Dār Ṣādir, 1374-76/1955-57), II, 9b. 94 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) translate the entire work into Persian. Though elderly, pressed with family responsibilities, and troubled with the cares of difficult times, the latter took into consideration the spectacular merits of the book and thus agreed to undertake the translation.42 Other information indicates that he died before he could finish the task, and that the work was completed by a colleague, Muḥammad ibn Aḥmad ibn Abī Bakr al-Kātib al-Mābarnābādī.43 Among the currently extant Persian manuscripts, the date of ah 204 seems to appear in very few codices,44 which may raise the question of whether or not this information is to be trusted. But in al-Mustawfī’s day no useful purpose would have been served by forging it: in ah 596 there would have been nothing remarkable about knowing (or claiming) that Ibn Aʿtham had written his Kitāb al-futūḥ in ah 204, and someone inventing a date would not have done so without some further purpose in mind—for example, to establish some specific connection with one of the Shīʿī Imāms. But in al-Mustawfī’s introduction the date is simply stated in passing, without being pursued to some further point. It is also worth asking how this information came to be known to him and no one else. One can never be absolutely certain on such matters, of course, but the most likely explanation is that this detail was mentioned in the colophon of the Arabic MS from which al-Mustawfī worked. In any case, there is no immediate reason for doubting that this information comes from al-Mustawfī, or for suspecting a priori that such a date for the composition of the Kitāb al-futūḥ is spurious. Support for this date may be found in Yaqūt’s tarjama of Ibn Aʿtham, in which a certain Abū ʿAlī al-Ḥusayn ibn Aḥmad al-Sallāmī al-Bayhaqī quotes two lines of verse which he says were recited to him by “Ibn Aʿtham al-Kūfī”.45 Unfortunately, there appear to be several al-Sallāmīs with very similar names, who were variously quoted by al-Thaʿālibī (d. 429/1038), al-Gardizī (wr. ca. 442/1050), Ibn Mākūlā (d. 473/1081), Ibn al-Athīr (d. 630/1233), al-Juwaynī (wr. 658/1260), Ibn Khallikān (d. 681/1282), and al-Yāfiʿī (d. 738/1367). One of these al-Sallāmīs was the well-known historian of Khurāsān;46 little personal information is available concerning him, but on the basis of details provided by al-Thaʿālibī his date of death must be placed after 365/975.47 42. Al-Mustawfī, Tarjama-i Kitāb al-futūḥ (Bombay: Chāpkhānē Muḥammad-i, ah 1305), 1:4-2:15. 43. See Massé, “La chronique d’Ibn Aʿtham,” 85; Togan, “Ibn Aʿsemülkûfî,” 702b. 44. It is worth noting that while a number of Persian manuscripts were catalogued prior to the appearance of Storey’s Persian Literature, no date but that suggested by Frähn was given for the composition of the Kitāb al-futūḥ, until Storey (I.2, 1260) noted the date of ah 204 in a catalogue of Mashhad Persian MSS which had just come to his attention. Several Bombay lithographs, however, include this date in their texts of the introduction, and do not seem to be copying one from the other, which suggests that several MSS available in Bombay also bore the date of ah 204 for the composition of the text. 45. Yāqūt, Irshād al-arīb, I, 379:5-8. These verses celebrate the value of a forgiving friend. 46. See W. Barthold, Turkestan down to the Mongol Invasion, 3rd ed. (London: E.J.W. Gibb Memorial Trust, 1968), 10-11; Franz Rosenthal, A History of Muslim Historiography, 2nd ed. (Leiden: E.J. Brill, 1968), 321 n. 7. 47. Al-Thaʿālibī, Yatīmat al-dahr, ed. Muḥyī al-Dīn ʿAbd al-Ḥamīd (Cairo: Maktabat al-Ḥusayn al-tijārīya, 1366/1947), IV, 95:8-16. It does not seem to have been noticed that at the end of this notice, al-Thaʿālibī refers to two verses by al-Sallāmī and then says: “I did not hear the two verses from him, but rather only found them in a copy of his [book]”. The implication of this statement is clearly that al-Thaʿālibī anticipated that his audience would suppose that he had heard the verses from the author himself; this in turn suggests that he Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 95 This does not seem to connect with anything else which is known about Ibn Aʿtham or his history. Another al-Sallāmī (or al-Salāmī), however, was an obscure faqīh in Baghdad whose career may be assigned to the first half of the third/ninth century.48 A scholar of this period could easily have heard, in his student days, poetry from an author who finished a history in ah 204; and on the assumption that this history was not necessarily written in the last years of its author’s life, it is possible that the two men were colleagues in Baghdad. In terms of genre formation, the compilation of such as text as the Kitāb al-futūḥ, reflects one of the well-known features of early Arabic historiography: topical monographs of the second century ah providing the building blocks for, and ultimately giving way to, the comprehensive histories of the third. Ibn Aʿtham’s book was a Shīʿī manifestation of the sort of work one often encounters in this period, and it comes as no surprise to find such a text appearing at the beginning of the third century ah. Once largely limited to Medina and al-Kūfa, the Shīʿa had by this time established a significant presence for themselves in Baghdad,49 where such developments as the Shuʿūbīya controversy, the rise of the Muʿtazila, the miḥna, and the foundation of the Bayt al-Ḥikma would in the very near future demonstrate the depth, range, and intensity of the cultural foment that prevailed in the capital in this formative era.50 Ibn Aʿtham’s history represented his effort to set before Muslims at large his own growing community’s views on the live historical issues under discussion in his day, and to do so with an extended account of the Islamic past. A composition date of 204/819-20 also finds at least some direct support in the Arabic text. At the beginning of one of his sections, Ibn Aʿtham says: “Jaʿfar ibn Muḥammad used to say to my father…”51 As this Jaʿfar figures in isnāds in the text, and in them occupies key positions where the Imāms would be quoted in Shīʿī ḥadīth,52 he can be none other than the sixth Imam, Jaʿfar al-Ṣādiq (d. 148/765); it is perfectly plausible that the son of one of his students or tradents should have written a historical work 54 years after the Imām’s could have done so—i.e. that al-Sallāmī was his older contemporary. As al-Thaʿālibī was born in 350/961 (GAL, I, 284), it is unlikely that he would have been hearing poetry from al-Sallāmī before about 365/975. This year can thus be taken as approximating the earliest possible death date for this al-Sallāmī. 48. Abū al-Ḥusayn al-Rāzī (d. 347/958) reports details about a certain Maḥmūd al-Miṣrī who was a student of Ibn Hishām (d. 218/834), saw al-Shāfiʿī (d. 804/820) as a boy, and heard a story about al-Shāfiʿī majlis from one of his students (Yāqūt, Irshād al-arīb, IV, 379:14-380:4). This Maḥmūd was thus probably born ca. 195/810, and engaged in studies through ca. 225/840. He refers to hearing al-Sallāmī speak about al-Aṣmaʿī (d. 213/828) at second remove, so a floruit of ca. 220-40/835-55 may be set for al-Sallāmī himself. This would also fit a report (ibid., I, 392:14-393:1) of al-Sallāmī reciting poetry to the poet Jaḥẓa (224-326/839-938), on the one hand, and having information about the wazīr Aḥmad ibn Abī Khālid (d. 211/827) at second remove (ibid., I, 118:14- 119:4), on the other. 49. See Etan Kohlberg, “Imam and Community in the Pre-Ghayba Period,” in Said Amir Arjomand, ed., Authority and Political Culture in Shiʿism (Albany: State University of New York Press, 1988), 37. 50. For further discussion of the response of literature to controversies prevailing in society at large, see Lawrence I. Conrad, “Arab-Islamic Medicine,” in W.F. Bynum and Roy Porter, eds., Companion Encyclopaedia in the History of Medicine (London: Routledge, 1993), 686-93; and more generally, M. Rekaya, art. “al-Maʾmūn” in EI2, VI (Leiden: E.J. Brill, 1991), 331-39. 51. Ibn Aʿtham, Kitāb al-futūḥ, II, 92ult. 52. Cf. ibid., II, 390:3. 96 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) death. This line of investigation leads into the difficult issue of Ibn Aʿtham’s informants, however, and this problem requires some prior consideration of the structure of the work as a whole. The Structure of the Kitāb al-Futūḥ A read through Ibn Aʿtham’s history will leave no doubt that he was a fervent supporter of the Shīʿa, not only in their legitimist claims to the caliphate, but also in their early doctrines concerning the religious knowledge of the Imāms, and in their highly emotional focus on the sufferings and travails of the ʿAlid line under the Umayyads.ʿAlī ibn Abī Ṭālib is upheld as the Prophet’s paternal cousin, the first male convert to Islam, a brave warrior, and an upright man; along more religious lines, he is described as free from error, passion, or fault, and as Muḥammad’s waṣī and the heir to his knowledge.53 He was the candidate most deserving of the caliphate after Muḥammad’s death, and was deprived of his right on entirely specious grounds.54 Of al-Ḥusayn, it is stated that he was “the most excellent of the progeny of the prophets” and the bearer of Muḥammad’s staff (qaḍīb), and that the rendering of support to him was as much a personal religious duty as were prayer and almsgiving.55 Foreknowledge of his death is bestowed upon Muḥammad, Fāṭima, and ʿAlī through vivid dreams, visions, and visitations by angels, and is linked with the events of the Apocalypse.56 Supernatural phenomena and eschatological predictions are routinely evoked. Even the stars in the heavens and the plants on the earth weep at Karbalāʾ, for example, and a Jewish soothsayer pours abuse on the Umayyads when al-Ḥusayn is killed: had Moses left one of his descendants among the present-day Jews, he says, they would have worshipped him rather than God, but the Prophet had no sooner departed from the Arabs than they pounced upon his son (sic.) and killed him; he warns that the Torah decrees that anyone who kills the progeny of a prophet will forever after meet with defeat and upon his death will be roasted in the flames of hell.57 It is important to bear in mind that the Shīʿī emphasis of the text is not a matter of overtones or coloring, but rather of intense emotional involvement on the part of the author, and no small degree of polemic. Ibn Aʿtham himself was concerned about how his work would be received, and expressed anxiety to his patron (on whom more will be said below) over the possibility that his work would be mistaken for a rāfiḍī tract, and so bring them both into difficulty.58 In light of his Shīʿī emphasis, it is quite striking to see how frequently this perspective is directly contradicted elsewhere in the text. In the first volume, on the ridda wars and the 53. Ibid., 11, 466:11-18; III, 57:3, 74:1-12, 264:3-5. Many other examples of this kind could of course be adduced. 54. Ibn Aʿtham, Bankipore Text, 28:21-30:4 (= Muranyi, “Ein neuer Bericht,” 246- 47, lines 166-203 of the Arabic text). 55. Ibn Aʿtham, Kitāb al-futūḥ, V, 13:2, 16:17, 39:10-13. 56. Ibid., II, 4650.4-466:10; IV, 210:15-224:10. 57. Ibid., IV, 222:10-223:5; V, 246:7-247:6. 58. Ibn Aʿtham, Bankipore Text, 30:5-8. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 97 early conquests, Abū Bakr is on almost 80 occasions referred to as al-ṣiddīq or khalīfat rasūl Allāh. In one report, a tribesman of Tamīm argues that the Prophet gave no one knowledge for the sake of which others might follow him, and recites a verse pointing out that while Muḥammad deserved obedience, he appointed no successor to whom this obedience should then be transferred. These ideal openings for advancing Shīʿī or ʿAlid counterclaims are all missed, however, and the report ends with the thoroughly Sunnī argument that rejection of Abū Bakr’s caliphate is tantamount to kufr.59 Elsewhere, a conversation between Abū Bakr and ʿUmar concedes that ʿAlī is “a fair man acceptable to most of the people in view of his virtue, courage, close relationship to the Prophet, learning, sagacity, and the gentleness he shows in endeavors he undertakes”; but at the same time, it concludes that his gentleness makes him unsuited to military leadership.60 Obedience to ʿUmar is obligatory, even if one doubts his judgment, because he is amīr al-muʾminīn, and ʿAlī himself exalts ʿUmar’s merits, heaps praises upon him, calls him al-fārūq, and takes charge of his burial arrangements.61 In a poem in which a Meccan comments on the failure of Ibn al-Zubayr to practice what he preaches, the poet upholds the conduct of ʿUmar as al-fārūq and aligns himself with the sunna of Abū Bakr, whom he calls ṣiddīq al-nabī.62 The phenomenon of a history which speaks with numerous voices is absolutely typical of early Arabic historiography, as Noth has conclusively shown, and betrays the origins of such texts. These were not original essays composed by single authors, whose own personal conceptions of the past would then be reflected in them, but rather were compilations based ultimately on large numbers of short reports set into circulation, transmitted, and recast by many people over long periods of time. It is this essentially compilatory character which accounts for the contradictions and discrepancies, even on fundamental issues, which one repeatedly encounters in these works.63 The Kitāb al-futūḥ is in many ways typical of these patterns of compilation, but whereas authors often wove their source materials together in such a way that signs of the process of compilation were rendered fairly subtle, Ibn Aʿtham made no effort to produce a history which would read as a unitary whole. The arrangement of material (especially in the first two thirds of the book) is, largely the product of selecting monographs on various subjects and linking them end-to-end. Breaks marking the transition from one source to another are not smoothed out or disguised, but overtly signaled. In a few cases this is done with collective isnāds (to which we shall return below), but most frequently it take the form of headings, some of which announce recourse to a new source with the word ibtidāʾ followed by the new subject. 59. Ibid., 1,60:8-61:17. 60. Ibid., I, 72:1-11. 61. Ibid., I, 218:3-6; II, 92ult-93:11. 62. Ibid., V, 288:10. 63. See Albrecht Noth, “Der Charakter der ersten grössen Sammlungen von Nachrichten zur frühen Kalifenzeit,” Der Islam 47 (1971), 168-99; idem, Quellenkritische Studien zu Themen, Formen and Tendenzen frühislamischer Geschichtsüberlieferung, I. Themen und Formen (Bonn: Orientalische Seminar der Universität Bonn, 1973), 10-28; Stefan Leder, Das Korpus al-Haiṯam ibn ʿAdī (st. 207/822). Herkunft, Überlieferung, Gestalt früher Texte der Aḫbar Literatur (Frankfurt: Vittorio Klostermann, 1991). 98 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) The main sources for the text appear to consist of a limited number of monographs of the type usually ascribed to the akhbārīs of the second century ah.64 Ibn Aʿtham’s account of the election of Abū Bakr, for example, seems to be based on one earlier ʿAlid Kitāb al-saqīfa, which he refers to as riwāyat al-ʿulamāʾ,65 and terminates with remarks suggesting that he has reached a point where his source also ends.66 His narrative on the ridda also appears to be a summary from a single source;67 it ends with a doxology which can only have come from a written monograph source, and which typifies Ibn Aʿtham’s disinterest in smoothing out the rough edges as he shifted to a new subject to be covered from a new source: inqaḍat akhbār al-ridda ʿan ākhirihī bi-ḥamd Allāh wa-mannihi wa-ḥusn taysīrihi wa-bi-ʿawnihi wa-ṣallā Allāh ʿalā sayyidinā Muḥammad wa-ʿalā ālihi wa-ṣaḥbihi wa-sallama taslīman kathīran.68 His treatment of the early conquests, which immediately follows, seems to have involved the interweaving of two texts: a Futūḥ al-Shām textually related to the Futūḥ al-Shām of al-Azdī (fl. ca. 180/796),69 and a Futūḥ al-ʿIrāq.70 Other futūḥ works are also in evidence for later periods, for example, concerning the conquest of Khurāsān, Armenia,71 the Mediterranean islands,72 and probably also Egypt.73 64. On the themes of interest to these akhbārīs, see Noth, Quellenkritische Studien, 29-58. The term akhbārī is a convenient substitute for the perhaps inappropriate term “historian”, but it must be borne in mind that the authorities in question are not known to have called themselves akhbārīyūn, and that this term is first attested in the Fihrist of al-Nadīm (wr. ca. 377/987). See Stefan Leder, “The Literary Use of the Khabar: a Basic Form of Historical Writing,” in Averil Cameron and Lawrence I. Conrad, eds., The Byzantine and Early Islamic Near East, I: Problems in. the Literary Source Material (Princeton: Darwin Press, 1992), 314 n. 165. 65. See Ibn Aʿtham, Kitāb al-futūḥ, I, 2:3-5:4, with the lacuna filled by the Bankipore Text, 20:16-30:8 (= Muranyi, “Ein neuer Bericht,” 239-47). The title for this narrative is typical: Dhikr ibtidāʾ saqīfat Banī Sāʿida wa-ma kana min al-muhājirīn wa-al-anṣār (the Bankipore Text, 21:1, simply has Akhbār saqīfa Banī Sāʿida). 66. See below, p. XX (near note 134). 67. Ibn Aʿtham, Kitāb al-futūḥ, 5:5-89:17 is defective; for the complete text, see the Bankipore Text, 30:9- 125ult. 68. Kitāb al-futūḥ, I, 89:16-17; = Bankipore Text, 125:7-8. 69. See al-Azdī, Futūḥ al-Shām, ed. William Nassau Lees (Calcutta: Baptist Mission Press, 1854). On this work, see my “Al-Azdī’s History of the Arab Conquests in Bilād al-Shām: Some Historiographical Observations,” in Muḥammad ʿAdnān al-Bakhīt, ed., Proceedings of the Second Symposium on the History of Bilād al-Shām during the Early Islamic Period up to 40 ah/640 ad (Amman: University of Jordan, 1987), I, 28-62. 70. On the early futūḥ monographs, see Noth, Quellenkritische Studien, 32-34. 71. Ibn Aʿtham, Kitāb al-futūḥ, II, 108:1-116:12. 72. Ibid., II, 117:14-146:11, with some interpolations. On this material, see Griffini, “Nuovi testi arabo- siculi,” 402-15, especially on Sicily; Lawrence I. Conrad, “The Conquest of Arwād: a Source-Critical Study in the Historiography of the Early Medieval Near East,” in Cameron and Conrad, eds., The Byzantine and Early Islamic Near East, I, 317-401. Note the curious way in which Ibn Aʿtham attempts to make the transition to this work from the preceding account of campaigns in Armenia by inserting a brief description of an Ethiopian maritime raid on baʿḍ sawāḥil al-muslimīn and resulting Muslim deliberations on how to respond (II, 116:13- 117:13), as if the maritime campaigns in the Mediterranean could somehow be seen as the repercussions of this raid. 73. There seems to be a major lacuna where an account of Egypt would have stood. Volume I, most of which is attested only by the Gotha MS, suddenly breaks off as ʿUmar is about to write to ʿIyāḍ ibn Ghanm Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 99 In later volumes, accounts of the murder of ʿUthmān, the battle of Ṣiffīn, and the uprising of al-Ḥusayn are all prefaced with isnāds indicating that for these important events Ibn Aʿtham collected a number of works and drew on all of them to produce a single narrative covering the issues and details he wished to include: “I have combined what have heard of their accounts, despite their differences in wording, and have compiled [this material] uniformly into a single narrative”.74 There are many other areas, however, where important events appear to have been treated on the basis of either one or a very few monograph sources: the murder of ʿUmar ibn al-Khaṭṭāb,75 the ghārāt,76 and the advent of the ʿAbbāsids,77 for example. But even in such cases as these, the task of harmonizing information from sources was not one to which Ibn Aʿtham paid much attention. For his account of the rebellion of Zayd ibn ʿAlī (d. 122/740), for example, he seems to have had two sources. Setting out on the basis of one of them, he begins with a heading: dhikr wilāyat Yūsuf ibn ʿUmar al-Thaqafī al-ʿIrāq wa-ibtidāʾ amr Zayd ibn ʿAlī ibn al-Ḥusayn wa-maqtalihi.78 But within three pages he finds that he needs to use material from the other source; he thus begins again from a somewhat different approach, complete with a new heading on exactly the same subject: ibtidāʾ khabar Zayd ibn ʿAlī ibn al-Ḥusayn raḍiya, Allāh ʿanhum.79 To this string of only superficially integrated sources Ibn Aʿtham has added numerous “interpolations”. This term is used advisedly, since there is again nothing subtle about these additions, which often represent significant digressions. A heading or an isnād announces the beginning of the interpolation, and the end is frequently signaled with a phrase advising the reader that Ibn Aʿtham will now return to his main source or subject: thumma rajaʿna ilā ḥadīth..., thumma rajaʿna ilā al-ḥadīth al-awwal, thumma rajaʿna ilā al-khabar, and so forth.80 On one occasion, it could hardly be made clearer that an account is being interpolated into the main narrative from some other source: wa-hādhā dākhil fī (p. 334:1), and resumes with ʿAmr ibn al-ʿĀṣ about to march against the Berbers (p. 349:1). The Persian text provided by the Hyderabad editors includes some details relevant to Egypt (pp. 346:16-349:11), but it is unlikely that this is all Ibn Aʿtham could say or wished to say about this important subject. 74. Ibid., II, 149:2-3, 345:7-9. Cf. IV, 210:13-14. 75. Ibid., II, 83:4-95:1, ending in a major lacuna. 76. Ibid., IV, 36:10-37:2. The section is entitled Ibtidāʾ dhikr al-ghārāt baʿda Siffīn, and opens with an isnād identifying this material as taken from the work of Abū Mikhnaf (d. 157/774) on the subject. Cf. Ursula Sezgin, Abū Miḫnaf. Ein Beitrag zur Historiographie der umaiyadischen Zeit (Leiden: E.J. Brill, 1971), 56-58; idem, “Abū Miḫnaf… über ġārāt,” 445-46. 77. Ibn Aʿtham, Kitāb al-futūḥ, VIII, 153pu-211pu. This section begins with the heading: wa-hādhā ibtidāʾ khabar Abī Muslim min awwalihi; no source other than the akhbārī al-Madāʾinī is mentioned, but he is named twelve times (pp. 159:9-10, 160:9-10, 190:4, 17, 192:4, 14, 195:7, 196:7, 202:3, 205:6, 206:12, 207pu), and Ibn Aʿtham’s source here was probably a history by this writer. 78. Ibid., VIII, 108:3-4. 79. Ibid., VIII, 110:15. 80. Ibid., I, 114:6, 271:9; 11, 12:16, 18:9, 81:2, 467:1, 470:10, 472pu, 487:11, 493:11; III, 85:6, 93pu, 105:8, 135:11, 145:12, 169:12, 207:12, 317:4; IV, 224:11; V, 269:9; VI, 158:5; VII, 51:4, 107:11, 231:1, for some of the more obvious examples. 100 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) ḥadīth al-azāriqa.81 That sources and new information should be so roughly integrated suggests no particular skill as an akhbārī. And if we examine the interpolations to see what it was that Ibn Aʿtham sought to add to his sources, this conclusion is quickly confirmed. In many cases, his major interpolations are the stuff of popular folklore and pious legend. In his account of the conquest of Syria, for example, he intervenes with a long aside on al-Hilqām ibn al-Ḥārith, a warrior in Yemen in Jāhilīya times who bests the most outstanding Arab champions and proves to be a better fighter than a thousand men; eventually he converts to Islam and fights on the Muslim side in Syria.82 There are extraordinary stories of leading Muslim warriors debating with Byzantine generals, and even Heraclius himself; one has Muslims going to Antioch, where they confront Heraclius and Jabala ibn al-Ayham, find that their conquests are predicted in the New Testament, and discover that the Emperor has in his possession a casket (tābūt) containing pictures of the prophets, including Muhammad.83 There are also late Umayyad accounts encouraging the jihād against Byzantium—for example, relating at length how “the ten penitant youths of Medina” gave up the joys of their jawārī to march off to fight the Rūm when they heard that the caliph ʿAbd al-Malik (r. 65-86/685-705) was organizing an expedition.84 Iraq receives less attention of this kind, but also attracts some remarkable tales. In one, Yazdagird goes out to hunt and pursues an onager into the desert; when the onager has led him beyond earshot of his retinue, it turns to him and, “with God’s permission”, warns him to believe in his Lord and to refrain from kufr, otherwise he will lose his kingdom. The terrified ruler flees back to his palace and reports what has happened to his mōbadhs and his asāwira, who straightaway conclude that the doom foretold by the onager could only befall him at the hands of the Arabs currently active in his domains.85 Historical accounts are sometimes interrupted with faḍāʾil material on, for example, the congregational mosque of al-Kūfa, the province of Khurāsān, and even ʿUmar ibn al-Khaṭṭāb.86 The supernatural element is often prominent: encounters with hawātif are described,87 and where Shīʿī foci of piety and devotion are concerned there are frequent evocations of angelic visitations.88 The Shīʿī tenor, of course, also arises in other ways in Ibn Aʿtham’s interpolations. Traditions of the Prophet have it that Muḥammad forbade that any candidate of the Sufyānid line should assume the caliphate, cautioned the believers to separate Muʿāwiya and ʿAmr ibn al-ʿĀṣ anytime they are seen together (“they will not be sitting together pondering anything good”), and commanded that if they see Muʿāwiya 81. Ibid., VII, 52:5-7. 82. Ibid., I, 104:12-114:6. On the “thousandman”, the hazārmard of Persian tradition, see Noth, Quellenkritische Studien, 152. 83. Ibn Aʿtham, Kitāb al-futūḥ, I, 126:1-132:5. 84. See ibid., VII, 171:1-184:1, referred to in the heading as a khabar ḥasan. 85. Ibid., I, 161:13-162:6. 86. Ibid., 1, 286:17-288:11; II, 78:1-81:1, 92:16-94:8. 87. E.g. ibid., I, 249ult-253:5, two especially interesting cases. 88. E.g. ibid., IV, 210ult-224:10, a series of stories on such matters. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 101 “on my minbar”, they should slit him open from belly to spine.89 Pious narratives describe al-Ḥusayn’s distress as he bends over his mortally wounded father, weeping and calling down curses on Ibn Muljam, while ʿAlī himself tries to calm his son and assures him that “what is ordained will come”.90 Muʿāwiya is the subject of numerous moralizing or entertaining anecdotes promoting ʿAlid or Shīʿī positions;91 and Zaynab, “so eloquent that it was as if she were speaking through the mouth of her father,” upbraids the Kūfans after Karbalāʾ.92 All this was, of course, the stock and trade of the early Muslim qāṣṣ, and there can be little doubt that Ibn Aʿtham was just such a pious storyteller, in this case from a Shīʿī perspective. As such, his interest was not so much in the final shape of his history, or the extent to which it did or did not hold together as a whole, as it was in the various discrete contents of the work and the themes they could be used to illustrate. Sources were selected for their “qiṣaṣ-appeal” and didactic merit, and to the resulting mélange were added other reports and tales which he happened to know. In fact, it is likely that the transitional phrases and headings which strike the modern reader as crude and indicative of poor integration in many cases reflect a subtler purpose: as these transition points were so obvious, the reader could not fail to distinguish stories introduced by Ibn Aʿtham, and thus to be credited to his talents as a qāṣṣ, from those which were already present in his main monograph sources. Further, stories from such a loosely assembled text could easily be extracted and related separately. To judge from his book, Ibn Aʿtham must have done this many times himself with his own sources and materials, and it is from the recitation of precisely such excerpts that his Kitāb al-futūḥ came to the attention of the later unknown figure who commissioned al-Mustawfī’s translation.93 Once Ibn Aʿtham is recognized as a qāṣṣ, and of the Shīʿa into the bargain, the question of why he is such an obscure figure immediately becomes clearer. He was not a scholar of Sunnī or Shīʿī ḥadīth, and did not pursue a line of studies which would have attracted students to himself. And in his own day his work was probably not esteemed as much more than what it really was, a loose compendium of material which, while including historical works among its sources, was assembled with popular preaching and storytelling in mind. With no great work to preserve the memory of his name, or students to cite him in their silsilas, he quickly faded to anonymity and did not attract the attention of later compilers of biographical literature. Even among Sunnī muḥaddithūn, who predictably dismissed him as ḍaʿīf, he gained so little notice that he appears in none of the extant rijal al-sanad or 89. Ibid., II, 390:3-8; V, 24:12-13. 90. Ibid., II 466:11-18. The medieval reader would of course have realized instantly the powerful import of this statement—it applied not only to ʿAlī, but to al-Ḥusayn as well. 91. Ibid., III, 89:3-93:9, 101:4-105:7, 134:1-135:10, 142:9-145:11, 204:11-207:10. The same basic narrative form prevails in these tales: “after that”, as Muʿāwiya and his courtiers sit in his majlis, someone asks leave to enter and is admitted; a repartee follows, usually with liberal citation of poetry. 92. Ibid., V, 222:4-226:2. 93. See al-Mustawfī, Tarjama-i Kitāb al-futūḥ, 2:3. 102 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) ḍuʿafāʾ works.94 The transparent way in which Ibn Aʿtham uses sources to compile his history invites the conclusion that it would be an easy matter to recover these sources from the Kitāb al-futūḥ. But recent research has shown that the works of the akhbārīs betray a significant creative dimension; compilers not only collected and assembled material, but also reshaped and revised it to suit their own needs and interests.95 As a result, blocks of text attributed to a certain author do not necessarily represent the text exactly as that author left it, and any effort to recover a lost source thus becomes a most painstaking and difficult task. A qāṣṣ like Ibn Aʿtham would have been no less likely to have engaged in such revision, and there are in fact obvious signs of this in his history. A useful illustration is his account of the “thousandman” al-Hilqām ibn al-Ḥārith.96 The story begins by describing how the Arabs in days of yore used to raid and kill one another, their greatest warriors being ʿĀmir ibn Ṭufayl al-ʿĀmirī, ʿAntara ibn Shaddād al-ʿAbsī, and al-ʿAbbās ibn Mirdās al-Sulamī. On one occasion, these three, accompanied by a thousand of the finest warriors of Qays, set out on an expedition in which they wreaked great slaughter, defeated every foe they encountered, and won much booty. They then decided to return home, and when they arrived, they each in turn recited verse in which they boasted of their exploits to the people. In the original story, the poetry would of course have been cited at length, but here not a line of it appears; Ibn Aʿtham simply states the order in which the three warriors spoke, betraying with repeated recourse to an introductory qāla the fact that he has dropped all of the verses.97 Another qāla then introduces the statement that “they continued on with the booty and goods until they came to a wadi near the land of Yemen…”, which marks another gap, since we have just been told that the intent of the warriors had been to return home.98 When they confront al-Hilqām, the combatants are all said to have recited rajaz verses (wa-huwa yartajizu) as they came forth to fight, but whereas the original story would surely have cited these verses, Ibn Aʿtham again drops them entirely.99 Close analysis of his history would provide a sharper picture of how Ibn Aʿtham handled his material, but for present purposes it is already clear that he did not simply copy out what was available to him. Like other authors of his day, he considered it entirely legitimate to engage in revision. For modern historians, this means that the Kitāb al-futūḥ must be regarded not only in terms of numerous major sections com prised of older sources and interrupted by various interpolations and asides, but also with a view to the possibility of changes and revisions by Ibn Aʿtham to both types of material. And as will be seen below, it is further possible that revision was undertaken again, once the first two thirds of 94. Our only indication that he was noticed at all appears in a negative comment on his reliability in Yāqūt, Irshād al-arīb, 1, 379:2: wa-huwa ʿinda aṣḥāb al-ḥadīth ḍaʿīf. Yāqūt’s source for this observation is unknown. 95. See Leder, Korpus al-Haiṯam ibn ʿAdī, 8-14; Conrad, “The Conquest of Arwād,” 391-95. 96. Ibn Aʿtham, Kitāb al-futūḥ, I, 104:12-114:6. 97. Ibid., I, 105:6-9. 98. Ibid., I, 105:10-11. 99. Ibid., I,108pu-109pu. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 103 the text had already been finished. Continuations of the Text in the Third/Ninth Century The abrupt transitions, digressions, and discontinuities in the text, together with the formulae used to mark them, highlight some very important aspects of the structure of the work as a whole. But at the same time, they have served to obscure the most important transition of all. In his account of the caliphate of al-Rashīd (r. 170-93/786-809), Ibn Aʿtham provides only three paragraphs on this ruler before the appearance of the terminating sentence: tamma Kitāb al-futūḥ.100 That is, the text as composed by Ibn Aʿtham ends at this point, and the rest of the work as we have it today comprises a continuation, or dhayl. Confirmation of this comes from the account of Ibn Aʿtham by Yāqūt, who describes as follows the material available to him: He wrote... a Kitāb al-futūḥ, a well-known work in which he discusses [events] to the days of al-Rashīd, and a Kitāb al-taʾrīkh [extending] to the end of the days of al-Muqtadir and beginning with the days of al-Maʾmūn, such that it is practically a continuation (dhayl) of the former. I have seen both books.101 This suggestion of two histories, one continuing the other, points to a common phenomenon in Arabic literature,102 but it is very unlikely that Ibn Aʿtham intended that the main text should terminate the way it does. He provides a domestic anecdote, refers to the size and complexity of the ʿAbbāsid court and bureaucracy under al-Rashīd, and describes the immense wealth gained by this caliph, and with that the text just stops. There are no concluding eulogies or praises of God and the Prophet, as one often finds at the end of an Islamic text, and there is no apparent reason for why the book should terminate at this point. One may thus conclude that Ibn Aʿtham was suddenly unable to proceed any further, and although we cannot “know” what it was that cut short his work, his death would of course be one plausible explanation. If the Kitāb al-futūḥ ended at this point, then the material following must belong to some other work, and there immediately arises the question of whether this last section is the Kitāb al-taʾrīkh seen by Yāqūt. In all likelihood it is. This new section devotes 99 pages to the ʿAbbāsid caliphate, beginning in the reign of al-Rashīd, in much the same way that the Kitāb al-futūḥ had covered, at much greater length, the history of earlier times. Its function is precisely that of a dhayl, as Yāqūt observed, although it is uncertain whether the title he gives it was the original one (assuming that there was an original one). Yāqūt’s reference to seeing “both books” (al-kitābayn) could be taken as meaning texts in two separate MSS, but it is at least as likely, and perhaps more so, that what he had was very similar to what survives today: a history with its dhayl continuing on in the same MS, but with a title provided to announce the beginning of the new work. 100. Ibid., VIII, 244ult. 101. Yāqūt, Irshād al-arīb, I, 379:2-5. 102. See Caesar E. Farah, The Dhayl in Medieval Arabic Historiography (New Haven: American Oriental Society, 1967). 104 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) One point on which Yāqūt errs, however, is his assumption that the Kitāb al-taʾrīkh (as the dhayl will henceforth be called here) was the work of Ibn Aʿtham. It is immediately clear how he arrived at this conclusion: the dhayl opens with an isnād which begins ḥaddathanī Abū Muḥammad, and Abū Muḥammad was the kunya of Ibn Aʿtham. Further, the continuator followed the example set by Ibn Aʿtham in offering only loosely integrated materials, making extensive use of headings or isnāds to mark separate narratives, and continuing the popular tenor of the original in his dhayl. It was thus an easy matter to conclude that both parts of the text had been composed by Ibn Aʿtham. There are, however, a number of clear indications that the dhayl cannot be the work of Ibn Aʿtham. This is, of course, already the working hypothesis with which we must begin: if Ibn Aʿtham was unable to complete the Kitāb al-futūḥ, then the material following on where it breaks off is not likely to be his. The reference to “Abū Muḥammad” in the isnād opening the dhayl of course proves nothing, since this kunya was a very common one. Direct indication of a change in authorship arises in the fact that as one moves to the Kitāb al-taʾrīkh, the interest in Shīʿī issues disappears. Ibn Aʿtham had pursued such matters not just to the time of Karbalāʾ, but beyond this, if with much decreased intensity, to later affairs of special concern to the Shīʿa. The pro-ʿAlid poet al-Kumayt (d. 126/743), for example, receives considerable attention,103 as do the risings of Zayd ibn ʿAlī (d. 122/740) and his son Yaḥyā (d. 125/743).104 This stands in sharp contrast to the situation in the dhayl, which has not a word to say about any of the persecutions suffered by the ʿAlids and their supporters under the early ʿAbbāsids, nor of the bayʿa sworn to ʿAlī al-Riḍā in 201/816, or of his death under obscure circumstances in 203/818. It is true that no historian would have failed to recognize such subjects as sensitive areas of discussion, but while this would explain a lack of any effort to lay blame at the door of the ruling house, it does not account for the way in which the dhayl entirely ignores the ʿAlids and the Shīʿa.105 Also revealing is the fact that while the Kitāb al-futūḥ occasionally betrays its use of a source or sources written according to some basic annalistic principle,106 it more usually relies, as we have already seen, upon the sort of akhbārī-style topical monographs that were in circulation in the late second century ah. The Kitāb al-taʾrīkh, on the other hand, is based on materials which reflect a much more developed stage in the evolution of Arabic historical writing, organized according to reigns of caliphs or annalistic chronology. The author of the Kitāb al-taʾrīkh routinely cites the dates of important events to the day, uses such introductory formulae of the annalistic tradition as fa-lamma dakhalat sana...,107 ends the section on each caliph with sīrat al-khulafāʾ material setting forth the ruler’s physical 103. Ibn Aʿtham, Kitāb al-futūḥ, VIII, 82:6-97:13. 104. Ibid., VIII, 108:3-129ult. 105. On these matters, more will be said below. 106. See, e.g., ibid., VIII, 82:4, stating “and in that year Kumayt ibn Zayd al-Asadī was imprisoned”, although the year in question has not been mentioned earlier. 107. On the annalistic organization of historical texts according to the hijra reckoning as a secondary development, see Noth, Quellenkritische Studien, 40-44. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 105 appearance, moral demeanor, and culture,108 and sometimes shows concern for identifying the leader of the annual pilgrimage.109 After the passage announcing the end of the Kitāb al-futūḥ, the text continues with twenty pages on the reign of al-Rashīd, almost half of them dealing with the caliph’s relations with al-Shāfiʿī.110 This material on al-Shāfiʿī is introduced by isnāds citing as their immediate informant “Abū Muḥammad”, which at first glance, as we have seen, may seem to refer to Ibn Aʿtham; in fact, al-Majlisī took this to indicate that Ibn Aʿtham was himself a Shāfiʿī.111 But this is certainly not the case, nor is it possible that these reports could even have been known to our author, or to anyone else of his time. Al-Shāfiʿī is described as al-imām, the sunna of the Prophet is treated as an already established keystone in some “Shāfiʿī” system, and the master’s death is described as an occasion for much grief among a large throng of followers. While it may be conceded that al-Shāfiʿī enjoyed prestige and influence in his own lifetime, and that the collection and dissemination of his teachings began very soon after his death,112 the material here clearly presupposes the existence of a Shāfiʿī madhhab in a form sufficiently coherent to make the master the subject of considerable veneration. Now, as we shall see below, Ibn Aʿtham was probably working on the Kitāb al-futūḥ after ah 204, which is both the date given by al-Mustawfī for the completion of the Arabic text and the year of al-Shāfiʿī’s death. But as his father had been a student of Jaʿfar al-Ṣādiq, it is unlikely that Ibn Aʿtham lived long enough past ah 204 for accounts referring to al-Shāfiʿī in this way to have been in circulation in his day.113 If there be any doubts about this, they are dispelled by the fact that one of the two akhbār on al-Shāfiʿī is cited on the authority of al-Mubarrad,114 who died in 285/898, almost eighty years after the benchmark date of ah 204 for Ibn Aʿtham’s work on the Kitāb al-futūḥ. The isnād citing him begins with the name of “Abū Muḥammad”, who has the account of al-Mubarrad through “one of the men of learning”, which indicates that the kunya “Abū Muḥammad” here, and probably also in the isnād at the beginning of the dhayl, refers to someone who lived at least a decade or so after al-Mubarrad. The text which Yāqūt knew as the Kitāb al-taʾrīkh is thus a dhayl composed no earlier 108. Noth (ibid., 37-38) regards the theme of sirāt al-khulafāʾ as primary, in that it does not in any manifest way derive from some other theme, but while this may be the case, the presumptions (e.g. the caliph as the center of political authority) and articulation (e.g. knowledge of minute personal details) of the theme suggest a perhaps relatively late development. 109. Ibn Aʿtham, Kitāb al-futūḥ, VIII, 253:1-2, 272ult-273:1, 275:7-276:9, 298:4, 300:5, 307pu-308:2, 317:12, 321:1, 322:11-12, 323:9, 13, 325:4-5, 330:15-16, 18, 339ult-343:11, 346:13, 352:13-14, 354:14-15. 110. Ibid., VIII, 245:1-263:10. 111. Al-Amīn, Aʿyān al-shīʿa, VII, 429. I have not seen the passage in the Biḥār al-anwār to which al-Amin refers. 112. Al-Rabīʿ ibn Sulaymān al-Murādī (d. 270/883-84) was already transmitting the Kitāb al-umm in Egypt in 207/822-23, only three years after the master’s death. See al-Shāfiʿī, Kitāb al-umm (Cairo: Al-Maṭbaʿa al-amīrīya al-kubrā, ah 1321), II, 93:19. 113. On the rise of the Shāfiʿī madhdhab, see Heinz Halm, Die Ausbreitung der šāfiʿitischen Rechtsschule von den Anfängen bis zum 8./14. Jahrhundert (Wiesbaden: Ludwig Reichert, 1974), 15-31. 114. Ibn Aʿtham, Kitāb al-futūḥ, VIII, 252:8-9. 106 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) than the very end of the third/ninth century, which is far too late to have be written by Ibn Aʿtham. Further, it is not the work of a single continuator. Having just related some developments pertaining to al-Rashīd’s joint nomination of his sons Muḥammad (the future caliph al-Amīn) and ʿAbd Allāh (al-Maʾmūn) to the caliphate, the text again confronts us with an abrupt and unexpected turn of direction: These are some fine narratives concerning al-Rashīd which I wrote down on the authority of a certain litterateur and added them in your [copy of the] book (wa-alḥaqtuhā bi-kitābika) so that you might peruse them, for they really are choice tales.115 This is followed by four akhbār, all anecdotes focusing on the impressive education and overall worthiness of al-Rashīd’s sons (especially Muḥammad),116 and concluding with the heading: thumma rajaʿna ilā al-khabar al-awwal min amr al-Rashīd wa-ibnayhi Muḥammad wa-ʿAbd Allāh,117 indicating a return to his point of departure in the basic text of the dhayl. Upon initial reflection the reference to “your book” may seem to be addressed to the unknown author of the Kitāb al-taʾrīkh, i.e. by a student or younger protégé. But a closer look will reveal that this is unlikely. The language, suggesting that the writer has taken the liberty of adding material from someone else so that the person addressed might thereby learn something, would be outrageous presumption if addressed by a student to his teacher. On the other hand, it is absolutely typical of how writers of the third century ah and later would posture before a patron. The phrase bi-kitābika, literally “in your book”, would thus mean “in your [copy of the] book”, an entirely acceptable sense for such a phrase. The material introduced by this heading thus marks the beginning of an interpolation by some scribe copying the text for a patron or client. This interpolation clearly extends only to the end of the fourth anecdote, as the scribe is at pains to advise the reader—to whit, his patron—that he is now returning the text to its original subject, the prelude to the conflict between al-Amīn and al-Maʾmūn. In introducing this section, he follows Ibn Aʿtham’s own method in the main body of the book, and in closing it he uses the same 115. Ibid., VIII, 263:11-43. 116. The anecdotes consist of the following tales: 1) ʿAlī ibn Ḥamza al-Kisāʾī (d. 189/865) reports on how, in 183/799, he was asked by al-Rashīd to examine his sons to see how well they had been educated. The examination is followed by praises for the caliph and his son, and interspersed with verses of poetry and comments on grammar. 2) Khalaf al-Aḥmar (d. ca. 180/796) tells how he was charged by al-Rashīd to tutor Muḥammad. As the caliph’s demands were quite stern, the instruction was very demanding. Muḥammad complained to Khāliṣa, his mother’s slave attendant; she asked Khalaf to relent, but he refused. 3) This links with the second anecdote, and here Khāliṣa tells Khalaf how Zubayda, Muḥammad’s mother, had an ominous dream about him. Despite the reassurances of astrologers and dream interpreters, she continues to be anxious about the dream’s meaning and its import for her son. 4) The section closes with an anecdote related by the future ḥājib of al-Amīn, al-ʿAbbās ibn al-Faḍl ibn al-Rabīʿ, on the prince and his educational training. The tale stresses that as Muḥammad shares the Prophet’s name and his epithet al-Amīn (Quraysh, he says, called the Prophet by this name before the mabʿath), he may be the amīr whom the ʿulamāʾ say will come to spread justice, revive the sunna, and stamp out oppression. 117. Ibid., VIII, 272:15-16. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 107 technique (a heading) and wording (thumma rajaʿna ilā...). That is, having recognized how Ibn Aʿtham had worked interpolations into the framework of his sources in the Kitāb al-futūḥ, the scribe proceeded to add material to the dhayl in the very same way. It is also possible that this same scribe (or some other one, for that matter) made similar additions elsewhere in the text, but in such a way that the interpolation does not draw immediate notice. Such activity, of course, would not necessarily be limited to the dhayl. In the main body of the Kitāb al-futūḥ (i.e. before the beginning of the dhayl), one of al-Mansūr’s daughters tells a tale of how her grandmother, pregnant with the future caliph, dreamed that a lion came forth from her and received the homage of all the other predatory beasts.118 As it happens, the immediate informant for this story is al-Ḥasan ibn al-Ḥubāb al-Muqriʾ al-Baghdādī, who died in Baghdād in 301/914.119 Assuming that this figure was an informant of the scribe, this latter person’s interpolations into the book could be dated roughly to the first half of the fourth century ah. The problem with this proposition, however, is that the Arabic text is clearly defective right where the interpolation from al-Ḥasan ibn al-Ḥubāb begins, and this anomalous passage may well have been just a marginal note in the MS which was copied into the main body of the text by mistake.120 If this was an interpolation by the scribe, it seems to have been a exceptional case; there are no other similarly obvious instances of such additions within the main body of the Kitāb al-futūḥ. Once the dhayl returns to its original author, it continues for 82 pages and covers the death of al-Rashīd, the caliphates of al-Amīn (r. 193-98/809-13) and al-Maʾmūn (r. 198-218/813-33), and the first half of the caliphate of al-Muʿtaṣim (r. 218-27/833-42).121 This material includes narratives for numerous events of this period, but again in a highly incidental fashion. For the reign of al-Muʿtaṣim it provides only brief references to the foundation of Sāmarrāʾ in 220/835 and two versions of the defeat and execution of Bābak in 222/837. At this point the text suddenly states: The length of his caliphate was the same as that of Shīrawayh, son of Kisrā, murderer of his father. He lived to the age of 24, and his death took place in Sāmarrāʾ in Al-Qaṣr al-Muhadhdhab (sic.).122 This of course can have nothing to do with al-Muʿtaṣim, who died after a reign of eight and a half years at the age of 46 or 47.123 The comparison is rather the well-known one between the six-month reign of the Sasanian ruler Shīrawayh and the six-month reign 118. Ibid., VIII, 211ult-212:4. 119. See al-Khaṭīb al-Baghdādī (d. 463/1071), Taʾrīkh Baghdād (Cairo: Maktabat al-Khānjī, 1349/1931), VIII, 301:4-302:2, no. 3813. 120. There are, in fact, a number of marginal notes in MSS of this work, some of them quite long and providing supplementary material relevant to the topics under discussion in the main text. 121. Ibn Aʿtham, Kitāb al-futūḥ, VIII, 263:11-353ult. 122. Ibid., VIII, 353:1-3. 123. See al-Ṭabarī (d. 310/923), Taʾrīkh al-rusul wa-al-mulūk, ed. M.J. de Goeje et al. (Leiden: E.J. Brill, 1879- 1901), III, 1323pu-1324:4. 108 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) of al-Muntaṣir 247-48/861-62), who did in fact die at the Al-Qaṣr al-Muḥdath palace in Sāmarrāʾ at the age of 24 or 25.124 Ibn Aʿtham’s identification of the caliph’s deathplace as Al-Qaṣr al-Muhadhdhab may easily be dismissed as a manifest error by the scribe or modern editor.125 For present purposes the import of all this is that the Kitāb al-taʾrīkh fails to say a word about the caliphates of al-Wāthiq (r. 227-32/842-47) and al-Mutawakkil (r. 232-47/847-61), and this seems to mark a further break and a new stage in the elaboration of the text. That a different hand is at work where the narrative resumes is also indicated by the fact that while the earlier material consisted of detailed narrative, this new stage comprises only a brief summary of caliphal chronology, providing nothing but accession and death dates and ending with the abdication of al-Mustaʿīn in 252/866. As nothing is said about the end of the three-year reign of his successor al-Muʿtazz (r. 252-55/866-69), it would at first seem that this final stage was the work of someone writing in the brief reign of this caliph. But this is of course impossible. If the author of the Kitāb al-taʾrīkh was writing late enough to cite an isnād in which al-Mubarrad (d. 285/898) figures at third remove, which would mean that al-Mubarrad was probably long since deceased by that author’s time, then in the party responsible for extending the dhayl even further we cannot be dealing with someone who could have been active in the 250s/860s. Here we may return to Yāqūt’s comment that the manuscript he saw extended to the reign of al-Muqtadir (r. 95-320/908-32). This suggests that the text as we have it is defective at the end. The extent of the lost material is difficult to judge, and would depend on how long into the caliphate of al-Muʿtaṣim the detailed content of Ibn Aʿtham’s continuator, the author of the Kitāb al-taʾrīkh, extended. An attractive hypothesis would be that as so often happened with medieval MSS, only the last folio was damaged, with loss of text to both recto and verso, most likely to the lower half of the page. If this was the case, then only some lines of text would have been affected. Circumstantial support for this explanation may be seen in the fact that the text at this point offers only a few key dates, and so would not have required more than a few lines to reach the reign of al-Muqtadir. For present purposes the important point is that what stands at the end of the extant text is not really its proper end, but rather a fragment probably representing the only legible part of a damaged terminus. Had this damage not occurred, our text would probably accord with Yāqūt’s description of a text extending to the time of al-Muqtadir. The gap may have existed only in the textual transmission underlying the Ahmet III MS, but as there is no other manuscript material for this part of the text, it is impossible to pursue this matter further. Development within Ibn Aʿtham’s Text We may now turn our attention to a major problem within the original Kitāb al-futūḥ. As we have seen, the text for which Ibn Aʿtham himself was responsible extends only to 124. Ibid., III, 1498:8-13. 125. On Al-Qaṣr al-Muḥdath, see Ernst Herzfeld, Geschichte der Stadt Samarra (Hamburg: Verlag Von Eckardt und Messtorff, 1948), 216, 227. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 109 the opening passages concerning the caliphate of al-Rashīd. But this does not tally with the Persian translation, which ends with the immediate aftermath of Karbalāʾ. The discrepancy cannot be attributed to al-Mustawfī’s use of an incomplete Arabic MS, since he knows that the Arabic text was written in ah 204. As this sort of information would almost certainly have been provided in a terminal colophon, his Arabic MS must have been complete up to and including this colophon. Nor can an explanation be sought in an incomplete Persian translation, since, as we have seen above, al-Mustawfī’s rendering was finished after his death by al-Mābarnābādī. One must conclude, then, that an Arabic MS of the Kitāb al futūḥ, complete to a terminal colophon dated ah 204, was translated in its entirely into Persian; and this, in turn, suggests that at first Ibn Aʿtham brought his text down only as far as Karbalāʾ. Turning to the Arabic text as we have it today, the factors at work here may be explained in terms of the author’s motives and aims in compiling his book. It is amply clear that while Ibn Aʿtham may have brought no particular skill as a compiler to his task, he did have some overarching agenda in mind. This is hinted at in several passages in the book itself. In volume VIII, at the end of his account of a Khārijite rebellion against the Umayyad caliph Marwān ibn Muḥammad, Ibn Aʿtham observes that the demise of the Umayyad regime was close at hand and then suddenly states: This then—may God honor you—is the last of the futūḥ, and after this we begin with akhbār on Naṣr ibn Sayyār, al-Kirmānī, and Abū Muslim al-Khawlānī al-Khurāsānī.126 This is followed by a major heading: Ibtidaʾ khabar Khurāsān maʿa Naṣr ibn Sayyār wa-Judayʿ ibn ʿAlī al-Kirmanī wa-Abī Muslim ʿAbd al-Raḥmān ibn Muslim, which introduces the continuation of the text from the point where Ibn Aʿtham had just broken off. From this it would seem that he considered it difficult to carry the theme of futūḥ past the campaigns and expeditions of the later Umayyads, and hence felt a bit self-conscious at continuing his Kitāb al-futūḥ into an era in which the specific theme of futūḥ could no longer be the primary concern. This solicitude for the integrity of some notion of futūḥ emerges again in his account of the reign of al-Mahdī (r. 158-69/775-85), where the text advises the reader that “concerning al-Mahdī there are narratives (akhbār) and fireside tales (asmār) which are not relevant to the subject of futūḥ”.127 That is, Ibn Aʿtham considers that he is still writing on the subject of futūḥ and the irrelevance of the accounts in question to this topic is the reason why Ibn Aʿtham is not going to cite them here. Exactly what this notion of futūḥ was is difficult to judge, but may be viewed in relation to the fact that by the dawn of the third century ah, Muslim audiences were accustomed to the presentation of futūḥ within the framework of Islamic salvation history: military conflict was a means through which the will and plan of God were realized on earth, with the outcome establishing the divinely ordained order, and, at the same time, rewarding the righteous and God-fearing and punishing their enemies and opponents.128 The archetypical 126. Ibn Aʿtham, Kitāb al-futūḥ, VIII, 145:17-18. 127. Ibid., VIII, 239:8-9. 128. For the general background to such writing, see John Wansbrough, The Sectarian. Milieu: Content 110 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) paradigm for this was the conquest of Syria, which not only established true religion in a new land, but also, on the one hand, rewarded the Muslims for responding to God’s summons to believe in Him and abandon their old pagan ways, and on the other, punished the Rūm for their tyranny, injustice, and above all, disbelief.129 To an audience already familiar with such paradigms, Ibn Aʿtham offered a popular history which situated the Shīʿī case against a backdrop of military conflict: just as God’s will had been worked out in the conquests which achieved the expansion of Islam, so also it would be in the strife which marked the course of the Umayyad caliphate, continued to plague the ʿAbbāsids, and repeatedly had dire consequences for the Shīʿa and the ʿAlid line. Pursuing such a comprehensive view of history in terms of futūḥ, would only be meaningful, of course, if it could be brought to a satisfactory conclusion: that is, where in Ibn Aʿtham’s scheme of things was the al-Yarmūk required to mark the fruition of divine design? This was surely not to be seen in the debacle at Karbalāʾ, where the Persian translation ends, much less in the reign of al-Rashīd, where the author’s original Arabic terminates. If we attach primary significance to the year ah 204 itself, rather than to the point reached in the text by that time, a very attractive hypothesis immediately arises for our consideration. Only six weeks into this year (Ṣafar 204/August 819), the triumphant entry of al-Maʾmūn into Baghdad marked the end of a decade of terrible civil war which had brought much destruction and suffering to the capital itself. The question of the greater meaning and import of a communal history marked by continual military strife was thus one that must have been on the minds of many as the war entered its final stages and then gave way to recovery and the re-establishment of order. But at a key point in the conflict, an event of particular importance to the Shīʿa also occurred. In 201/816-17, al-Maʾmūn had the eighth Imam, ʿAlī ibn Mūsā, taken to his residence at Marw, and there proclaimed him his successor to the caliphate with the title of al-Riḍā. The Imam was married to one of al-Maʾmūm’s daughters, and the black banners of the ʿAbbāsid house were replaced by the green ones of the line of the Prophet. To the expanding Shīʿī community back in Baghdad, this move must have come as a complete surprise: al-Maʾmūn’s ʿAlid proclivities were not unknown, but ʿAlī ibn Mūsā was far older than the caliph, and hitherto he had been living a secluded life of quiet devotion to scholarship in Medina.130 The impact of the announcement would in any case have been enormous; after more than 150 years of rule by usurpers, the rightful reunion of political and religious authority in the person of the and Composition of Islamic Salvation History (Oxford: Oxford University Press, 1978), esp. 1-49; and more generally, Bernd Radtke, Weltgeschichte und Weltbeschreibung im mittelalterlichen Islam (Stuttgart: Franz Steiner Verlag, 1992), 160-68. 129. See Conrad, “Al-Azdī’s History of the Arab Conquests,” 39-40, esp. n. 46; idem, “Conquest of Arwād,” 369-70. 130. See Francesco Gabrieli, Al-Maʾmūn e gli ʿAlidi (Leipzig: Eduard Pfeiffer, 1929), 29-47; Dominique Sourdel, “La politique religieuse du calife ‘abbâside al-Ma’mûn,” Revue des études islamiques 30 (1962), 27-48; Tilman Nagel, Staat und Glaubensgemeinschaft im Islam. Geschichte der politischen Ordnungsvorstellungen der Muslime (Zurich and Munich: Artemis Verlag, 1981), I, 170-84; Hugh Kennedy, The Early Abbasid Caliphate: a Political History (London: Groom Helm, 1981), 157-61. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 111 Imam could at last be realistically anticipated. To a qāṣṣ and aspiring author like Ibn Aʿtham, the prospect of the accession of the Imam to the caliphate would have been especially significant. The violent repression which had periodically been brought to bear against advocates of Shīʿī claims under earlier ʿAbbāsid caliphs131 would not have encouraged the production of a Shīʿī view of Islamic history, however crudely pieced together it may have been. This is not to suggest that pro-Shīʿī literature had not been produced in earlier years—it certainly had, and much of it was in fact used by Ibn Aʿtham. But the invective in such literature had been reserved for the Umayyads, who had been overthrown by the ʿAbbāsids and could easily be vilified without consideration for the consequences. A comprehensive history, however, would carry the narrative into the ʿAbbāsid caliphate and Ibn Aʿtham’s own contemporary period, where the prevailing mood of the times would not have encouraged the composition of a history focusing on the ʿAlids and the Shīʿa, which by al-Maʾmūn’s reign had already suffered major repression. The proclamation of ʿAlī al-Riḍā as walī al-ʿahd, however, not only signaled that the way was clear for a general exposition of the history which had brought the umma to the brink of this great event, but also provided a culminating point with which a narrative could most appropriately end: the theme of futūḥ, articulated from the ridda wars through the early Islamic conquests, the travails of the ʿAlid family, and the further expansion of Islam under the Umayyads, and ending with the great civil war between al-Maʾmūn and al-Amīn, would climax in the dramatic fulfillment of divine plan with the promise of a caliphate which would bring Shīʿī aspirations to fruition.132 To whom would such a history have been directed? Any number of possibilities could be advanced, but an especially revealing passage at the end of Ibn Aʿtham’s discussion of the election of Abū Bakr at the Saqīfa Banī Sāʿida narrows the options down significantly. Here our author concludes the section as follows: This, may God honor you, is what happened at the Saqīfat Banī Sāʿida. This is the recension of the religious scholars, and here I have not wished to write down anything of the additions [introduced by] the Rāfiḍa; for were this book to fall into the hands of someone other than yourself, it could have certain implications even for you, may God preserve you.133 The first thing this passage establishes is that the Kitāb al-futūḥ was a commissioned work: Ibn Aʿtham did not proceed at his own initiative, but was working for a patron. But who was the patron? Ḥamīd Allāh, who thought the Bankipore Text was the Kitāb al-ridda of al-Wāqidī, suggested that this passage might have been addressed to the caliph al-Maʾmūn.134 This could as easily be proposed with respect to Ibn Aʿtham, but cannot 131. For a summary, see Bernard Lewis, art. “ʿAlids” in EI2, I (Leiden: E.J. Brill, 1960), I, 402b. 132. It goes without saying, of course, that many would have observed that ʿAlī al-Riḍā, being older than al-Maʾmūn in the first place, might never accede to the throne, and that even if he did, no commitments had been made to the legitimacy of continuing ʿAlid claims after his death. 133. Ibn Aʿtham, Bankipore Text, 31:5-8; = Muranyi, “Ein neuer Bericht,” 247:204 206. 134. Ibn Aʿtham, Bankipore Text, 30 n. 2. 112 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) be the case where a work finished in ah 204 is concerned, since Ibn Aʿtham must have begun work on the text much earlier, i.e. when al-Maʾmūn was far away to the east and preoccupied with much more important matters. On the other hand, the passage could have been addressed to a high-ranking official among the caliph’s supporters in Baghdad. Such an official, whose identity seems beyond reach, would have merited the honorifics which Ibn Aʿtham addresses to him, and at the same time would have shared the author’s concern lest they both come to be associated with a text taken for a rāfiḍī tract. In a circumstantial fashion, the possibility of such patronage is supported by the fact that Ibn Aʿtham did, as we shall see, have close contacts with numerous personalities who had been members of the imperial entourage under earlier ʿAbbāsid caliphs. These concerns soon became moot, however, for Ibn Aʿtham’s enterprise to fashion a popular history promoting a Shīʿī vision of the Islamic past would have suffered a devastating blow in Shaʿbān 203/September—October 818, when ʿAlī al-Riḍā suddenly died under suspicious circumstances in Ṭūs. The arrival of the news in Baghdad some weeks later would have rendered any history conceived along these lines pointless, and it would thus come as no surprise to find the author of such a work abandoning his task, at least for the time being. If one searches for a telltale caesura in the Kitāb al-futūḥ, it clearly appears after Karbalāʾ. The text up to this point reflects all the zeal and fervor which one would expect from a qāṣṣ writing in the aftermath of ʿAlī al-Riḍā’s appointment as walī al-ʿahd, and the fact that this ends with Karbalāʾ, and that the Persian translation also ends there, simply indicates the point at which the dramatic setback represented by the death of ʿAlī al-Riḍā compelled Ibn Aʿtham to suspend work on his book. That is, the text available to aI-Mustawfīi 400 years later was a full copy of the book as Ibn Aʿtham left it in ah 204—a first recension, as it were. If this hypothesis is valid, then the remainder of the text, up to the reign of al-Rashīd, must represent later work by Ibn Aʿtham, and in it we should expect to see signs of the difficulties encountered in continuing a work when its original plan and aim had been irretrievably compromised. This is plainly in evidence in the remainder of the Kitāb al-futūḥ after Karbalāʾ. The former zeal is gone, and while developments relevant to the Shīʿa continue to be discussed, they suggest no particular interpretation; the Imāms themselves seem deliberately to be avoided, the oppressive measures taken against the Shīʿa by al-Manṣūr (r. 136-58/754-75) and al-Hādī (r. 169-70/785-S6) go unnoticed, and ʿAlid rebellions against the ʿAbbāsids are passed over in silence. One might readily see why Ibn Aʿtham, writing at the seat of the ʿAbbāsid caliphate, might hesitate to treat such events with the zeal with which he had taken up earlier developments, but it is nevertheless noteworthy that his attitude toward the history of his own community becomes so ambivalent that al-Majlisī, using the Kitāb al-futūḥ 900 years later, took him for a Sunnī and included him among the mukhālifūn, whom he says he will cite in order to refute them.135 And as the passages cited above clearly show, even the theme of futūḥ itself seems to have become difficult for Ibn Aʿtham to sustain. 135. Al-Majlisī, Biḥār al-anwār, I, 24:13, 25:9. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 113 It is also noteworthy that in several significant ways Karbalāʾ marks a shift in Ibn Aʿtham’s technique as a historical writer. As observed above, the Kitāb al-futūḥ is a compilation largely achieved by copying earlier monographs on major subjects one after the other, literally end to end. While this tendency may be seen both before and after Karbalāʾ, it is most pronounced in the first part of the book, where almost all of the text has obviously come directly from topical monographs: works on the Saqīfa Banī Sāʿida, the futūḥ in various regions, the murder of ʿUthmān, Ṣiffīn, Nahrawān, the abdication of al-Ḥasan, and the events leading up to Karbalāʾ. Aside from Ibn Aʿtham’s own interpolations, “filler” on matters of lesser concern, taken from other written sources, is very limited—hardly more than ten percent of the text. After Karbalāʾ the material becomes far more varied, and specialized monographs, while still prominent, are nowhere near as dominant in their role as sources. In part this reflects the fact that in terms of the developing historical consciousness of the Shīʿa, such events as Ṣiffīn and Karbalāʾ were far more important than anything which was to follow. But the shift after Karbalāʾ is not just away from extended quotation from long monographs on issues relevant to the Shīʿa, but away from extended quotation from long monographs in general, and so suggests the changed working method of a writer returning to a task he had set aside for some time. Related to this is Ibn Aʿtham’s use of the isnād. This question will be pursued below, but here it is worth observing that Karbalāʾ marks a dramatic shift in our author’s method of citing authorities. Prior to this benchmark in the text, he cites long collective isnāds for the most important extended narratives taken from his monograph sources, but hardly ever gives isnāds for brief individual akhbār. After Karbalāʾ this pattern is reversed: the collective isnād is never used, while the number of isnāds for individual reports, though still modest in absolute terms, rises dramatically in comparison to the number given earlier. This interpretation of the extant textual evidence and its historical context has a number of important implications. First, and most obviously, the composition date of ah 204 refers only to the Arabic text down to the account of Karbalāʾ and its immediate aftermath; the rest was composed at some later time. Unfortunately, the dearth of personal information about Ibn Aʿtham allows us minimal grounds for estimating how much later this continued work could have occurred. As has been observed several times already, our author’s father was a student of Jaʿfar al-Ṣādiq, who died in 148/765. If one takes into consideration Bulliet’s argument that medieval Islamic education largely involved the teaching of the very young by the very old,136 then it must be conceded that Ibn Aʿtham may still have been active in the 220s and 230s ah and that work on his history could have continued as late as this. Second, if Ibn Aʿtham abandoned work on his history in ah 204, once he had reached Karbalāʾ, and then resumed work later, the question arises of whether his extension of 136. Richard W. Bulliet, “The Age Structure of Medieval Islamic Education,” Studia Islamica 57 (1983), 105-17. 114 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) the text was accompanied by revision of the part already completed. This is what one would expect in any case, and later revision of the Arabic text up to volume V, 251 of the Hyderabad edition would explain why, for example, the Persian translation by al-Mustawfī contains so much material, especially Arabic verse, which is lacking in the Arabic text. In such a situation the Persian translation becomes extremely important, as the sole surviving comprehensive witness to the first recension of the Arabic text as it stood in ah 204. A critical edition of this Persian text is thus to be encouraged as a contribution of considerable potential value; until one is available, the question of possible revision of the first Arabic recension cannot be addressed in any serious way. Use of the Isnād The Kitāb al-futūḥ poses serious problems where proper names are concerned. Throughout the book, both in the text and in the isnāds, names are often badly garbled or completely different from what one finds in parallel passages in other works, and the Hyderabad edition often compounds the confusion by adding its own mistakes or engaging in hypercorrection, on the assumption that the forms of names in other printed texts must be the “correct” ones: e.g. Bishr ibn Ḥarīm in the MSS is “corrected” to Khuzayma al-Asadī in the edition, al-Raqqa becomes al-Ruṣāfa, Mūsā al-Hāshimī is replaced by ʿAlī ibn ʿĪsā ibn Māhān, and Ibrāhīm ibn Muḥammad al-Ghassānī appears as al-Sarī ibn Manṣūr al-Shaybānī.137 The isnāds in the text are often confused, and while some of the errors can be corrected fairly easily, others pose very difficult problems indeed. And rather than assist with such difficulties, the Persian translation often compounds them; where the Arabic has Asīd ibn ʿAlqama, for example, the Persian has Rashīd ibn ʿAbd Allāh al-Azdī.138 Some of the confusion may be put down to the process of textual transmission, or perhaps to Ibn Aʿtham’s revision of his first recension; but from what we have already seen above, it would be a mistake to presume that Ibn Aʿtham took the isnād any more seriously than he did other aspects of the formal akhbārī’s craft. As a qāṣṣ, he legitimated his work in the eyes of his audience not by proofs of ability as a textual critic, but through the power of his stories to moralize, entertain, or teach didactic points. The question of Ibn Aʿtham’s use of the isnād thus becomes very complicated when studied in detail, especially where investigation of his sources is concerned. This topic is being pursued elsewhere,139 however, and here discussion will be limited to those areas which can inform us on matters already raised above. Ibn Aʿtham does not deploy the isnād in any consistent fashion in his text, and it is certainly not the case that he “belongs to the classical school of Islamic history writing, basing himself on akhbār introduced by their isnads”.140 Indeed, isnāds are rarely given 137. Ibn Aʿtham, Kitāb al-futūḥ, V, 222:5; VIII, 217:3, 259:11-12, 312:3. 138. Ibid., I, 249ult. 139. See n. 64 above. 140. See Hugh Kennedy, The Prophet and the Age of the Caliphates: the Islamic Near East from the Sixth to the Eleventh Century (London: Longman, 1986), 362-63. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 115 through the first five volumes of the Kitāb al-futūḥ to Karbalāʾ, and collecting them does not in itself offer a conspectus of Ibn Aʿthām’s sources. His usage of the isnād may best be assessed in terms of the two types he offers in the two recensions of his text, as identified above, and as these attestations of authority serve very different purposes, they may be discussed separately. Collective Isnāds There are four collective isnāds supporting long sections of text on major topics which would have been covered in early akhbārī monographs, and these name authorities for extended blocks of text on the election of Abu Bakr,141 the caliphate of ʿUthmān ibn ʿAffān,142 the battle of Ṣiffīn,143 and the events leading up to the death of al-Ḥusayn at Karbalāʾ.144 A fifth isnād cites a single chain of informants for the ghārāt.145 It will immediately be seen that these isnāds all support material of special importance to the Shīʿa, and that all fall within the first recension of the text. This would indicate that here, at least, Ibn Aʿtham felt the need for some formal verification of his authorities. Unfortunately, these isnāds are in varying states of disarray. At the cost of considerable time and effort, one can often put such matters right, but here the problem is compounded by the fact that Ibn Aʿtham’s chains of authorities include so many obscure or unknown persons for whom external evidence allows us to propose no floruit. At this point, all that can be said is that even when Ibn Aʿtham does cite authorities, he is highly erratic and shows no concern for the for mal criteria of isnād criticism which were well-established by the third century ah. Nuʿaym ibn Muzāḥim al-Minqarī, presumably the brother of the better-known Naṣr ibn Muzāḥim al-Minqarī (d. 212/827),146 is twice cited by Ibn Aʿtham as a direct oral informant (ḥaddathanī…),147 but, in the other two collective isnāds, another informant stands between him and our author.148 Hishām ibn al-Kalbī (d. 204/819) is cited once directly,149 but twice through Abū Yaʿqūb Isḥāq ibn Yūsuf al-Fazārī.150 Materials from al-Madāʾinī (d. 228/843) are handled in a particularly inconsistent fashion. 141. See Bankipore Text, 19:3-11; cf. also Muranyi, “Ein neuer Bericht,” 236. 142. Ibn Aʿtham, Kitāb al-futūḥ, II, 147:3-149:3. The Ahmet III and Chester Beatty MSS open with this isnād, and Shaban (ʿAbbāsid Revolution, xviii) thus took it as identifying the sources for the Kitāb al-futūh as a whole. 143. Ibid., II, 344:10-345:9. 144. Ibid., IV, 209:4-210:14. 145. Ibid., IV, 36ult-37:2, following immediately on after the heading: ibtidāʾ dhikr al-ghārāt baʿda Ṣiffīn. 146. Muranyi (“Ein neuer Bericht,” 237) considers that where Nuʿaym’s name is given, it is actually Naṣr who is meant. This is unlikely. The form Nuʿaym consistently appears as such in the text (see the next two notes), with no discrepancies among the MSS, and in one case the two brothers and Naṣr’s son al-Ḥasan all appear in the same collective isnād (Ibn Aʿtham, Kitāb al-futūḥ, II, 344:2, 345:4). 147. Ibn Aʿtham, Kitāb al-futūḥ, II, 147ult, 344:12. 148. Ibn Aʿtham, Bankipore Text, 19:5-6 (= Muranyi, “Ein neuer Bericht,” 236); idem, Kitāb al-futūḥ, IV, 209:7-8. 149. Ibn Aʿtham, Kitāb al-futūḥ, II, 344ult-345:1. 150. Ibid., II, 147ult-148:1, 342:4-5. 116 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) In a collective isnād for the caliphate of ʿUthmān, he is named as a direct oral informant and referred to as Abū al-Ḥasan ʿAlī ibn Muḥammad al-Qurashī,151 while in a second- recension isnād for the uprising of Muṣʿab ibn al-Zubayr in al-Baṣra during the Second Civil War, Ibn Aʿtham cites al-Madāʾinī’s material through ʿAbd Allāh ibn Muḥammad al-Balawī.152 Elsewhere, however, our author is satisfied to quote, as we have already seen, from one of al-Madāʾinī’s books.153 Examples of such patterns could be pursued further, but it is already clear that while Ibn Aʿtham makes use of collective isnāds, even these betray his disinterest in the critical considerations which isnāds were used to address in the first place. To have unnecessary links in his isnāds, or to quote from a book or second-hand informant when the author was personally known and accessible to him, did not seem to trouble him. He was willing to cite anyone who was available and who had interesting material to offer; indeed, a list of his immediate informants makes sense only if one recognizes it not as a group of teachers or authorities of the generation prior to his, but rather as a general collection of informants active at the time the Kitāb al-futūḥ was written. It is true, of course, that matters of isnād criticism were far more important in the field of ḥadīth, where the transmission of the words, deeds, and sanctions of the Prophet were at stake, than they were in akhbār. But this is not the point at issue here. The features discussed above demonstrate that Ibn Aʿtham did not handle isnāds with critical considerations in mind, and consequently, that one cannot assess them in terms of the formal critical principles which we know prevailed in his day. When we add to this problem his frequent citation (as in isnāds for individual reports) of unknown informants, his references to names which could refer to numerous persons,154 and the highly defective editorial state of many of the chains, it becomes amply clear that at present it is difficult to do much with these isnāds. Two rather limited conclusions, however, can be drawn from them at this time. First, the death dates of the identifiable informants with whom he had direct personal contact range from 201/816 for ʿAlī ibn ʿĀṣim ibn Suhayb155 to 228/843 for al-Madāʾinī. A first recension completed in 204/819-20 could easily have made use of information from 151. Ibid., II, 147:3-4. The Ahmet III and Chester Beatty MSS read Abū al-Ḥusayn for Abū al-Ḥasan, but the Chester Beatty text is based on that of the Ahmet III MS, and as Shaban (ʿAbbāsid Revolution, xviii) argues, this reading may be dismissed as a scribal error. Al-Madāʾinī’s correct kunya is given elsewhere in the text (VI, 253ult-254:1), where he is again called “al-Qurashī”, as he is also, for example, in Ibn Saʿd (d. 230/844), Kitāb al-ṭabaqāt al-kabīr, ed. Eduard Sachau et al. (Leiden: E.J. Brill, 1904-40), 1.2, 30ult. As al-Madāʾinī was in fact a mawlā of Quraysh (also as observed by Shaban), it is not unusual that some tradents should have referred to him by the laqab al-Qurashī rather than al-Madāʾinī. 152. Ibn Aʿtham, Kitāb al-futūḥ, VI, 253ult-254:1. 153. See n. 76 above. 154. Cf. Leder, Korpus al-Haiṯam ibn ʿAdī, 41-42; G.H.A. Juynboll, Muslim Tradition: Studies in Chronology, Provenance and Authorship of Early Ḥadīth (Cambridge: Cambridge University Press, 1983), 146-59. 155. See al-Khaṭīb al-Baghdādī, Taʾrīkh Baghdād, XI, 446:6-458:5, no. 6348; also GAS, I, 97. This tradent was born in 105/723, and so was a very old man when he died; his transmission of material to Ibn Aʿtham could have occurred almost anytime within the latter’s career. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 117 all these authorities, but of particular interest is the fact that Ibn Aʿtham appears to have relied upon both older contemporaries who had already died by the time he began the Kitāb al-futūḥ, and younger colleagues who were to remain active for many more years. Second, he quotes from numerous Sunni authorities, although the collective isnāds are all used to support long texts promoting the Shīʿī view of important historical events. Indeed, the part of the long collective isnād for Karbalāʾ which cites the Shīʿī Imams156 does so with no special honorifics, and appears only after four other more mainstream chains of authorities have been given, and with others yet to come. This appears to comprise an attempt to present distinctly Shīʿī material as representative of some broader perspective on the early decades of Islamic history, and addresses the question of why Ibn Aʿtham provides these collective isnāds in the first place. For him, these were devices through which he could propose that the emerging Shīʿī view of key events was an entirely legitimate Islamic view with which various non-Shīʿī authorities—scholars whom he knew personally—agreed on numerous points. An investigation of the extent to which he actu ally used material from the various authorities he names could prove most revealing. In his account of Karbalāʾ, for example, the complex collective isnād introducing the section cites some of the most famous akhbārīs of his day, including authors known to have written on Karbalāʾ; and as their narratives on this subject were used by such later historians as al-Ṭabarī, it is possible to check the extent to which Ibn Aʿtham really made use of their works. What follows this isnād, however, is an account quite unlike what one finds in al-Ṭabarī, but textually very similar to (and perhaps the source of?) the later Maqtal al-Ḥusayn of al-Khwārizmī (d. 568/1172).157 Isnāds for Individual Akhbār Where individual akhbār are concerned, the frequency with which Ibn Aʿtham uses the isnād is most interesting. There are only nineteen isnāds for individual reports in the part of the text covered by the first recension, and in some places one can read for hundreds of pages without encountering an isnād. In part this can be explained by the fact that he was using the sources already named in a collective isnād to construct an extended account of a single major event, and so considered it unnecessary to name the same authorities again for individual reports within that extended account. But in numerous places this explanation cannot be invoked, and here the interpolations are illustrative. Of the many opportunities where Ibn Aʿtham at least could have used an isnād to claim specific and unequivocal credit himself for a particular story or piece of information, i.e. by stating qāla Abū Muḥammad, he takes advantage of only one.158 Considering that this pattern prevails through more than 1600 pages of Arabic text, it may be taken as, first, indicating that Ibn Aʿtham did not see the isnād as a means to legitimate individual reports or add prestige or authority to their contents, and second, further confirming that not all that many 156. Ibn Aʿtham, Kitāb al-futūḥ, IV, 209ult-210:1. 157. Muwaffaq ibn Aḥmad al-Bakrī al-Khwārizmī, Maqtal al-Ḥusayn, ed. Muḥammad al-Samāwī (Najaf: Maṭbaʿat al-zahrāʾ, 1367/1947). See GAL, SI, 549, and the relevant Nachtrag (SI, 967). 158. Ibn Aʿtham, Kitāb al-futūḥ, III, 304ult. 118 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) individual reports were being incorporated into this part of the book in the first place. In the post-Karbalāʾ part of the text, however, individual isnāds suddenly become more frequent. There are sixteen in volumes VI and VII (i.e. none in the concluding parts of volume V): one is a multiple-link silsila from al-Balawī through al-Madāʾinī and two prior authorities to al-Shaʿbī (d. 103/721),159 one cites al-Madāʾinī on his own,160 two refer to the general category of ahl al-ʿilm,161 and the others name al-Haytham ibn ʿAdī (d. 207/822) or earlier tradents who normally figure in al-Haytham’s isnāds in the Kitāb al-futūḥ.162 In volume VIII this number rises to 26 (up to the point where Ibn Aʿtham’s own text ends); and with the exception of thirteen references to al-Madāʾinī,163 these isnāds never refer to the same informant more than once. While this is a marked increase over the rate of citation evident in the first part of the book, 42 isnāds through over 700 pages of text still reflects an attitude in which the device counts for very little. To this one could object, of course, that some of the early akhbārīs who compiled very worthy historical works also showed little or no concern for the isnād. Ibn Aʿtham’s indifference in this matter could thus be regarded as following a pattern quite common among these early akhbārīs, and manifest in such works as the Ayyām al-ʿarab of Abū ʿUbayda (d. 210/825)164 and the Futūḥ Khurāsān of al-Madāʾinī.165 But such a comparison is misleading, and to see why we need only consider the materials which Ibn Aʿtham uses an isnād to support. The kinds of reports for which isnāds are given at first seem quite diverse. In some cases, the structure of the narrative requires one: in first-person accounts, for example, or in accounts in which an informant states something like “I asked NN about...”, to name an informant is to identify a character in the story, and an isnād is accordingly provided for that purpose.166 In a few cases, an isnād is used to alert the reader to the fact that the information comes from the Shīʿī Imāms,167 or to name an authority for a precise piece of information, e.g. the exact date for the murder of ʿUthmān and his age at the time,168 or the 159. Ibid., VI, 253ult-254:1. 160. Ibid., VII, 278:11. 161. Ibid., VI, 161:2, 279:11. 162. Ibid., VII, 52:8, 107:11-13, 109:3, 9, 11, 110:5, 111:3, 124:2, 131:13, 138:13, 145:10-11, 171:2, 7. 163. Ibid., VIII, 159:9-10, 160:9-10, 190:4, 17, 192:4, 14, 195:7-8, 196:7, 202:3, 205:6, 206:12, 207pu, 218:10. 164. The extensive fragments quoted from this book by later authors have been collected and studied in an excellent two-volume work by ʿĀdil Jāsim al-Bayātī, Kitāb ayyām al-ʿarāb qabla al-Islām (Beirut: ʿĀlam al-kutub and Maktabat al-nahḍa al-ʿarabīya, 1407/1987). 165. See Gernot Rotter, “Zur Überlieferung einiger historischer Werke Madāʾinīs in Ṭabarīs Annalen,” Oriens 23-24 (1974), 103-33; Lawrence I. Conrad, “Notes on al-Ṭabarī’s History of the Caliphate of Hishām ibn ʿAbd al-Malik,” Journal of the Royal Asiatic Society, Third Series, 3 (1993), 1-2. 166. Ibn Aʿtham, Kitāb al-futūḥ, I, 249ult-250:1, 252:4, 286ult; II, 342:4-6, 390:3, 466:11; IV, 210ult-211:3, 212:6, 217:11 (returning to the narrative begun at 212:6), 222:10; V, 222:5; VI, 253ult-254:1; VIII, 94:5, 95:10, 96:7-8. 167. Ibid., II, 92ult, 390:3. 168. Ibid., II, 241:5. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 119 number of those killed at al-Jamal.169 In some places, informants are named for a cluster of reports on a particular subject: for instance, heavenly predictions of Karbalāʾ,170 the rebellion of Ibn al-Ashʿath,171 the travails of al-Kumayt,172 and the affairs of al-Saffāḥ.173 The impression of diversity continues if one considers the personalities cited and the way Ibn Aʿtham quotes them. Of the 68 authorities named in individual isnāds in the Kitāb al-futūḥ, 56 (i.e. over 80 percent) are cited only once through the entire length of the book. Beyond this, what Ibn Aʿtham most frequently offers is not a proper “chain” of authorities, but rather a single name (qāla fulān) which serves to introduce a report. But there appears to be no coherent pattern for the selection of individuals to be named in such isnāds. On occasion, the authority is someone from whom Ibn Aʿtham may in fact have heard the report, but most often the person named proves to have lived long before Ibn Aʿtham’s lifetime, or at least too early to have passed information on to him personally. Also, it is difficult to explain the isnāds in terms of the importance of supported material: not even citations of Shīʿī ḥadīth are consistently introduced by isnāds. The key to understanding the deployment of these isnāds lies in recognizing them as, for the most part, devices used by Ibn Aʿtham to mark interpolations, as observed above. In some cases this is obvious. The removal of the fourteen pages of reports introduced by isnāds at the beginning of the account of Karbalāʾ, for example, simply brings the reader to the real beginning of the account in Ibn Aʿtham’s main source; and lest there be any doubt, Ibn Aʿtham announces the fact: thumma rajaʿna ilā al-khabar al-awwal.174 Here the character of his heading as a mere cliché is readily apparent: he obviously cannot be “returning” to his “first account” when that “first account” has not even begun yet; the heading simply marks the end of a series of interpolated anecdotes. In many cases the persons cited are utterly obscure individuals, known to us only because their names also appear in some other work. Here again it would seem that Ibn Aʿtham was simply using isnāds as markers, and not to appeal to his audience’s sense of authority or to serve some critical scholarly purposes. It is certainly clear that he had no intention of authoring a book in which systematic consideration of the authority for specific accounts would be a task taken seriously, and this fact sets him far apart from the more serious historical akhbārīs, irrespective of whether or not they too used the isnād. But why, we might ask, should there be a sudden increase in the use of the isnād in volumes VI-VIII? At least a partial answer suggests itself once it is understood that this is all material added in the course of the second recension of the text. Collective isnāds 169. Ibid., II, 342:4-6. 170. Ibid., IV, 210ult-211:3, 212:6, 213:7, 215:6, 217:11, 222pu. 171. Ibid., VII, 124:2, 131:13, 138:13, 145:10-11. These reports all come from al-Haytham ibn ʿAdī, and as this author is not known to have written any separate work on Ibn al-Ashʿath, these citations probably indicate access to one of al-Haytham’s more comprehensive histories. 172. Ibid., VIII, 94:5, 95:10, 96:8-9. 173. Ibid., VIII, 190:4, 17, 192:4, 14, 195:7-8, 196:7, 202:3, 205:6, 206:12, 207pu, all from al-Madāʾinī. As indicated above (see n. 75), Ibn Aʿtham seems to have used a monograph by al-Madāʾinī which dealt with the reign of the first ʿAbbāsid caliph in detail. 174. Ibid., IV, 210ult-224:11. 120 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) are entirely absent here, and large monographs are used for fewer extended narratives. Larger numbers of individual accounts from a variety of sources were thus being used, and as collective isnāds were no longer being used to specify sources, occasions where doing so for individual reports were very much more numerous. But as the text had now lost its vital sense of purpose, Ibn Aʿtham shifted to a sporadic pattern of naming authorities, only doing so in such cases where it was a matter of some interest to him. His motives in this regard appear to relate to the fact that as the text approaches his own lifetime, the number of isnāds dramatically increases: Ibn Aʿtham had more comments of his own to inject, and thus more interest in citing authorities. Here the situation becomes clear if one looks at the persons from whom he takes information at this point. A number of these personalities were mawālī or companions of the caliphs al-Mahdī and al-Manṣūr,175 which suggests that Ibn Aʿtham himself moved in Baghdadi circles which had been close to the center of power in the second half of the second century ah. His répertoire of imperial anecdotes about the early ʿAbbāsids may thus reflect material actually in circulation in court circles in the late second century, and his connections with figures who had known al-Manṣūr and al-Mahdī further strengthens the case for accepting al-Mustawfī’s date of ah 204 for the completion of the first recension of the text: any number of persons who had been court figures during the reigns of these two caliphs would, in their old age, have been accessible to an author active at the turn of the century or shortly thereafter, and who subsequently returned to his work some years later. It is also worth noting that the dhayl continues this citation of court figures,176 which suggests that this part of the text was also written by an author in Baghdad with close ties to the ʿAbbāsid court before its transfer to Sāmarrāʾ in 220/835. Another interesting question is why Ibn Aʿtham marks some interpolations with isnāds, and others only with descriptive headings. While it is impossible to speak with certainty on such a subjective matter, the distinction here may to some extent be one between written and oral sources. The difference between the two is not so simple as has often been thought, and so must be regarded with caution.177 Still, it can be said that reports in the Kitāb al-futūḥ which are supported by individual isnāds tend to be short akhbār, and can often be linked with known literary works. The accounts introduced with descriptive headings, on the other hand, are more often long popular tales full of imaginary and supernatural elements and usually very moralizing, and absolutely typical of old qiṣaṣ lore which one would expect to have circulated orally. 175. Ibid., VIII, 212:5 (mawlā of al-Manṣūr), 238ult (ṣāḥib of al-Mansur), 239pu, 240:8 (companion of al-Mahdī), 242:4-5 (two mawālī of al-Mahdī). 176. Ibid., VIII, 263:14-15 (the tutor of al-Rashīd’s sons), 266ult (the tutor of Muḥammad al-Amīn), 270:6 (the future ḥājib of al-Amīn), 275:10:41 (a mutawallī amr al-sūq under al-Rashīd), 277:6-7 (a chess partner of al-Rashīd), 295:1-2 (a sub-attendant of al-Amīn, waṣīf khādim al-Amīn). 177. A seminal series of studies on this question has recently been published by Gregor Schoeler. See his “Die Frage der schriftlichen oder mündlichen Überlieferung der Wissenschaften im frühen Islam,” Der Islam 62 (1985), 201-230; “Weiteres zur Frage der schriftlichen oder müindlichen Überlieferung der Wissenschaften im Islam,” Der Islam 66 (1989), 38-67; “Mündliche Thora und Ḥadīṯ,” Der Islam 66 (l989), 213-251; “Schreiben und Veröffentlichen. Zu Verwendung und Funktion der Schrift in den ersten islamischen Jahrhundert,” Der Islam 69 (1992), 1-43. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 121 Ibn Aʿtham and His History Some important features of Ibn Aʿtham’s life and work have been discussed in the pages above, and before addressing a concluding question it may be well to summarize what has emerged so far. Ibn Aʿtham was the son of one of the students or tradents of the sixth Imam, Jaʿfar al-Ṣādiq, and grew up in the mid-second century ah. He composed some poetry, as did many in his day, but his special interest lay in popular preaching and storytelling; many of the accounts he related in his capacity as a qāṣṣ were of general interest to Muslims, but his perspective on key issues was specifically Shīʿī. He had connections with a number of tradents and compilers who already were or would later become well-known for their literary accomplishments in historical studies, and with court figures who had stories to tell about the reigns of past ʿAbbāsid caliphs. Early in the caliphate of al-Maʾmūn, and with the support of an unknown but highly placed patron, he assembled a history by cobbling together a number of existing monographs by other authors, revising as he saw fit and adding numerous interpolations which he had both from other written sources and from oral informants. One can with no particular difficulty harmonize al-Mustawfī’s use of a Kitāb al-futūḥ extending to Karbalāʾ and written in 204/819-20, an extant text continuing to the abdication of al-Mustaʿīn in 252/866, and Yāqūt’s reference to two histories ending, respectively, with the reigns of al-Rashīd and al-Muqtadir. First, al-Mustawfī’s statement that his translation was based on an Arabic text composed in ah 204 refers to a first recension of the book, one which had proceeded as far as Karbalāʾ when work was abruptly suspended. A hypothesis which fits the available evidence, and perhaps best clarifies a number of other questions, is that Ibn Aʿtham, working during the new stage of disorder which followed the overthrow and execution of al-Amīn, had set out to compile a history which would see in the suddenly presented prospects of an ʿAlid caliphate the fulfillment of divine promise and the climax of futūḥ itself. But with the death of ʿAlī al-Riḍā in 203/818, the raison d’être of such a book vanished, and Ibn Aʿtham’s work on it thus temporarily ceased shortly thereafter, in 204/819. It was a copy of this first recension that eventually made its way to Tāyābādh in the east, where a session featuring readings from it led an unknown political figure to commission al-Mustawfī to begin a Persian translation in 596/1199-1200. This work was still incomplete at the time of al-Mustawfī’s death, and was finished by a colleague. At some unknown point, Ibn Aʿtham resumed work on his history, but without the zealous sense of purpose that had inspired him earlier. This second recension was brought down to the caliphate of al-Rashīd, where it stops in a decidedly unsatisfactory fashion. Whether this was due to the death of the author, or the simple abandonment of an enterprise which no longer inspired him, is impossible to say. It is also unclear to what extent Ibn Aʿtham took this as an opportunity to revise what he had already completed in ah 204, although at least some such revision seems very likely. There is nothing in the second recension to indicate when work on it ceased, but allowing for the possibility that 122 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) Jaʿfar al-Ṣādiq was very old when he taught Ibn Aʿtham’s father, who may then have been young, and positing the same in the father’s transmission to Ibn Aʿtham himself, it is conceivable that our author was still alive in the 220s or even 230s ah. Shortly after the end of the third/ninth century, this second recension of the Kitāb al-futūḥ came to the attention of a later Sunnī writer, who continued the text at least as far as the defeat of Bābak in 222/837 and some uncertain distance further into the caliphate of al-Muʿtaṣim. This was the work which Yāqūt called the Kitāb al-taʾrīkh and also attributed to Ibn Aʿtham. If the proposal made above for damage limited to a final folio is correct, this continuation could not have extended more than a page beyond its present terminus. If the proposal is wrong—that is, if there were numerous folios missing at the end of the dhayl— then the continuator could have written a great deal covering events up through the brief reign of al-Muntaṣir in 247-48/861-62. This continuation was then itself continued by a brief chronology from the death of al-Muntaṣir to the reign of al-Muqtadir. The same damage which affected the end of the Kitāb al-taʾrīkh also affected the end of the final chronology, hence our suspicion that these damaged sections were on the recto and verso of the same folio, and thus that the lost text is in both cases less than a page. In the case of the terminal chronology, the lost material probably consisted only of a few dates from the abdication of al-Mustaʿīn to the reign of al-Muqtadir. At some point a scribe also copied a series of anecdotes into a patron’s copy; this same scribe may also have made additions to the main body of the Kitāb al-futūḥ, although evidence for this is very limited and can easily be accounted for otherwise. This would explain Yāqūt’s reference to a text coming down to the reign of al-Muqtadir and to two books which were so similar that one seemed to be the extension of the other. What we now have represents a text damaged at the very end, but otherwise identical to what Yāqūt saw, and an extended version of the Kitāb al-futūḥ as Ibn Aʿtham had originally left it. This original text may itself be viewed as representing two stages of work by the author. The first recension extended to Karbalāʾ and is now accessible through the Persian translation by al-Mustawfī; the second recension, which involved the revision of the first and its extension to the reign of al-Rashīd, is what we have today in at least most of the extant Arabic MSS and the Hyderabad edition. In closing this study, it may be asked how the conclusions reached above affect the usefulness of the Kitāb al-futūḥ to modern scholarship. Viewed from a historiographical perspective, Ibn Aʿtham’s place in the generation of the akhbārīs of the late second and early third centuries ah establishes his Kitāb al-futūḥ as a source of valuable insights on Arabic historical writing in this period. There are many lines of investigation which might profitably be pursued in future research, and, by way of illustration, attention may here be drawn to a particularly important one—the role of qiṣaṣ and other popular lore. It has long been known that some of this material is of very early origin, but it has often been assumed, and argued, that from the beginning it comprised a literary category separate from history and looked down upon by the “serious historians” of the second half of the second century ah.178 But these authors are in turn known to us almost exclusively through 178. See, e.g., Rosenthal, A History of Muslim Historiography, 186-93; A.A. Duri, The Rise of Historical Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 123 the even more “serious” historians of the third century, and it begs important questions to observe the relatively minor role of qiṣaṣ quoted from the early authorities in such later works, and from this to conclude that historical writing per se was always as critical as these works seem to indicate. The Kitāb al-futūḥ demonstrates how easily a gag could enter the field in early ʿAbbāsid times, and with clear expectations of public acceptance: Ibn Aʿtham would have not compiled his history the way he did if the public conception in his day of what history was all about would have resulted in the rejection and repudiation of his work. The ultimate obscurity of his book thus has less to do with his shortcomings as viewed in his own times, than with the major changes in attitudes toward historical writing which occurred in the course of the third century, as well as other factors which have little to with whether or not he wrote “good history”. By comparing Ibn Aʿtham to other early sources, which bear some of the same popular tales, it can easily be seen that this material was not distinct and separate from historical writing in the second century, but rather, closely intertwined and bound up with it.179 While Ibn Aʿtham’s work may embody a more popular folkloric element than that which is discernible among other akhbārīs whose historical works survive only in later quotations, he was an akhbārī all the same,180 and his history offers a unique opportunity for exploration of the ways in which folkloric elements contributed to early Arabic historiography, and then were gradually marginalized.181 At a broader level, this is precisely the sort of process one must expect. An emerging political, social, and religious community does not possess a sophisticated sense of history and historical writing from the beginning, any more than it possesses a fully developed theology from the beginning. Both evolve gradually, as more mature thinking replaces older formulations which, however satisfactory they may have been in the past, eventually come to be regarded as primitive and inappropriate. It has recently been argued that while it is certainly possible to define and study the genre of writing subsumed under the rubric of qiṣaṣ, which refers in particular to legends and myths of ancient prophets, it is problematic to extend this category to include other accounts which also bear this kind of “popular” imprint, and then to suppose that such an exercise in terminology tells us anything about the origins of the reports or addresses the question of their factual truth. Accounts regarded as qiṣaṣ may contain authentic historical information, while ostensibly sensible akhbār may contain sheer inventions.182 The Kitāb al-futūḥ provides innumerable illustrations of the importance of this observation, and Writing among the Arabs, ed. and trans. Lawrence I. Conrad (Princeton: Princeton University Press, 1983), 122-35. 179. See, for example, some of the tales in al-Azdī’s Futūḥ al-Shām. As many of these also appear in Ibn Aʿtham’s text, which is related to that of al-Azdī, but not taken from it, one must conclude that these tales were already present in the source common to both authors, and so must already have found a place in the futūḥ tradition by the mid-second century ah. 180. Yāqūt, who saw his work, concedes him not only this title, but also that of muʾarrikh; see Irshād al-arīb, 1, 379:1-2. 181. For the context of such a process, see Conrad, “The Conquest of Arwād,” 386-99. 182. Leder, “Literary Use of the Khabar,” 311-12. 124 • Lawrence I. conrad Al-ʿUṣūr al-Wusṭā 23 (2015) while it is true that Ibn Aʿtham had embellished his history with great amounts of baseless popular lore, this does not disqualify him as a historical source. In the first instance, his reports, even where manifestly untrue, are often important in ways untouched by their basis in fact (or lack thereof). Massé, for example, devoted a study to Ibn Aʿtham’s account of the conquest of Ifrīqiya, and arrived at the conclusion that here our author is probably not to be believed.183 But for historiographical purposes the same text reveals much about how topoi and narrative schema were deployed in historical writing, and for the cultural historian it highlights the lively interest in futūḥ which clearly prevailed in Ibn Aʿtham’s day. That this interest encompassed a broad range of material, and not just what modern scholars would regard as sober factual narrative, is surely a matter of crucial concern to any effort to establish the historical course of the Islamic conquests in North Africa. Of special interest in this regard are Ibn Aʿtham’s tales about dialogues, debates, and disputes between Byzantine dignitaries and early Muslims. Some of these tales are likely to be inventions of the early ʿAbbāsid period itself, when large-scale summer raids into Byzantine territory were undertaken on a regular basis, but others appear to be much older. The account (referred to above) of an encounter with Heraclius himself in Antioch has as its climax the discovery that the Emperor’s casket, full of pictures of the prophets, includes a picture of Muḥammad.184 Such a report, innocent of even the slightest iconoclastic sensitivities, would seem to substantiate King’s argument that traditional scholarly views on the iconoclastic tendencies of the early Muslims have been exaggerated.185 It also needs to be said that for establishing historical fact the Kitāb al-futūḥ is still a source of some importance. Two examples may serve to illustrate this point. In Ibn Aʿtham’s account of the early Islamic conquests, the familiar topological paradigm of the futūḥ tradition is violated in startling fashion by a novel explanation for the onset of Arab campaigns in Iraq. As Ibn Aʿtham’s source has it, the tribe of Rabīʿa, of the Banū Shaybān, was obliged by drought in Arabia to migrate to Iraqi territory, where the local Sasanian authorities granted them permission to graze their herds on promise of their good behavior. But the presence of these tribal elements eventually led to friction, which the Rabīʿa quite naturally interpreted as unwarranted reneging on an agreed arrangement. When they called on their kinsmen elsewhere for support, the crisis quickly escalated.186 This report is innocent of any awareness of the decisive role of great generals, or of a central authority directing all operations from far-off Medina. Nor does it comprise tribal fakhr, since it does not go on to award Rabīʿa special credit for success in Iraq. It may well represent the survival of an accurate account of how tribal movements along the Sasanian frontier gradually led to violent confrontation, with no role played by the caliph ʿUmar ibn 183. Massé, “La chronique d’Ibn Aʿtham,” esp. 89-90. 184. Ibn Aʿtham, Kitāb al-futūḥ, I, 130:9-131ult. 185. See G.R.D. King, “Islam, Iconoclasm, and the Declaration of Doctrine,” Bulletin of the School of Oriental and African Studies 48 (1985), 267-77. 186. Ibn Aʿtham, Kitāb al-futūḥ, 1, 88:7-89:6. Al-ʿUṣūr al-Wusṭā 23 (2015) Ibn Aʿtham and His History • 125 al-Khaṭṭāb, or even by an eminent Muslim commander. Similarly, it is well-known that in the tense first year of the caliphate of Yazīd ibn Muʿāwiya, the unfolding political crisis focused on Yazīd’s efforts to compel a small circle of leading Muslims to pledge their allegiance to him. But the religious eminence of these individuals notwithstanding, it is not clear why this should have been so important. The key seems to be provided by Ibn Aʿtham’s version of the terms under which al-Ḥasan ibn ʿAlī had earlier renounced his claim to the caliphate: one of the provisions mentioned by Ibn Aʿtham,187 but not by al-Ṭabarī, was that Muʿāwiya agreed that he would not himself appoint a successor to the caliphate, but rather would leave this decision to a shūrā of leading Muslims. The formation of such a committee would have been reminiscent of that convened by ʿUmar, and had it ever met, it would have included precisely the personalities whom Yazīd now sought to pressure into acknowledging him; the new caliph probably wished to convene the shūrā as a means of legitimating his rule, but knew that left to its own devices it was unlikely to name an Umayyad—and certainly not him—as caliph. The provision for a shūrā is also mentioned by al-Balādhurī (d. 279/892)188 and Ibn Abī al-Ḥadīd (d. 656/1258),189 both of whom take their information from al-Madāʾinī; Ibn Aʿtham also makes frequent reference to al-Madāʾinī, and was in any case his contemporary. The shūrā stipulation was thus commonly known a century before al-Ṭabarī wrote, and offers a cogent explanation for an issue crucial to our understanding of the crisis that arose on Yazīd’s succession.190 It is to be observed that here, as in many other places, Ibn Aʿtham used sources identical or similar to those available to such later historians as al-Balādhurī and al-Ṭabarī. If there is any single compelling argument for closer attention to the Kitāb al-futūḥ, it lies in the simple fact that all of our historical sources for early Islam are of essentially compilatory origin. Ibn Aʿtham offers a valuable opportunity to observe the variety and scope of the second-century compilations upon which all of our knowledge ultimately rests; and while some of the problems posed by these compilations are particularly easy to discern in his text, the implications of these difficulties are relevant not just to his history alone, but more generally to the entire range of later works for which the early compilations comprised almost exclusive sources of information. No other history as broad in scope as the Kitāb al futūḥ has survived from the dawn of the third century ah, and for both historical and historiographical questions its testimony is of importance throughout the range of the topics it covers. 187. Ibid., IV, 159pu-160:1. 188. Al-Balādhurī, Ansāb al-ashrāf, II, ed. Muḥammad Bāqir al-Maḥmūdī (Beirut: Dār al-taʿāruf, 1397/1977), 42:2-3. 189. Ibn Abi al-Ḥadīd, Sharḥ nahj al-balāgha, ed. Muḥammad Abū al-Faḍl Ibrāhīm (Cairo: ʿĪsā al-Bābī al-Ḥalabī, 1959-64), XVI, 22ult. 190. Cf. S. Husain M. Jafri, Origins and Early Development of Shiʿa Islam (London: Longman’s, 1979), 152-53.