The Prayers of Abū Muslim and al-maʾmūn. An Exercise in Dating Ḥadīth

Transcription

1 brill.com/jas The Prayers of Abū Muslim and al-maʾmūn. An Exercise in Dating Ḥadīth Stijn Aerts University of Leuven, Belgium Abstract In using ḥadīth narratives as historical sources, the challenge is to determine their origin. Different methods have been devised to trace the provenance of ḥadīth reports, but all these approaches lack an external frame of reference. Data retrieved from other sources, such as the historiographical genres, can be used to validate the results of ḥadīth-analysis under certain conditions. The method proposed here is exclusively applicable to traditions that describe events or support a rule that can be linked to a historical event described in the chronicles. Two case studies concerning the festival prayer and the pronouncement of a final takbīr in the regular prayers demonstrate that the results of the proposed method of dating ḥadīths correspond to the dates ascribed to the events in the chronicles. Keywords Ḥadīth isnād analysis Islamic ritual Islamic prayer Introduction The Ḥadīth, the disjointed narratives about actions and sayings of the Prophet that were initially transmitted orally and put to writing only generations later, continue to fascinate historians of the formative period of Islam. In using them as historical sources, the challenge is to determine whether they originated * The research for this article was supported by a fellowship of the Research Foundation Flanders (FWO). koninklijke brill nv, leiden, 2014 doi /

2 The Prayers of Abū Muslim and al-maʾmūn 67 with the Prophet (as purported), the compilers (the earliest of which are from the middle of the second/eighth century), or somewhere in between. Different methods have been devised to trace the provenance of ḥadīth reports, focusing on the chain of (alleged) transmitters (isnād) that acts as lead-in to every ḥadīth, the report itself (matn), or a combination of both. The most wellknown method centering on the transmission chains is the Common Link (CL) theory. The concept of the CL was coined by Joseph Schacht,1 but it was Gautier Juynboll who brought the theory to fruition.2 More recently, several scholars have shifted or extended the focus to the matn. Pursuing the consequences of the realization that the Ḥadīth is as much a literary as a historiographical enterprise, Sebastian Günther subjected traditions to literary (narratological) analysis and was able to unearth traces of the authorial processes that constitute their origin.3 Harald Motzki, one of Juynboll s fiercest critics, adopted the latter s CL method but insisted that the matn as well should be taken into account, which resulted in what he calls the isnād-cum-matn analysis.4 Irene Schneider, too, combined isnād and matn analysis, yet unlike Motzki she does not impute textual variation to accidents in the oral transmission; she sees them, much like Günther, as remnants of the development of the narrative.5 What all these approaches have in common is that they lack an external frame of reference. Whether focusing on the isnād, the matn, or both, it is not an essential part of these methods to assess whether the achieved results fit in with the wider historical background. This weakness was identified by Michael Cook, who showed that for three ḥadīths which he externally dated on the basis of the eschatological elements they contained, the CL theory did not 1 Schacht, The Origins, The principles of the method and the jargon involved are best explained in the introduction to Juynboll, Encyclopedia of Canonical Ḥadīth; also idem, Some Isnād-Analytical Methods. Before Juynboll, Josef van Ess (Zwischen Ḥadīṯ und Theologie) had already explored the consequences of Schacht s CL theory. For an alternative method centering on isnāds, see Behnam Sadeghi, The Traveling Tradition Test. 3 Günther, Fictional Narration and Imagination; idem, Modern Literary Theory. Literary theories and theories of narrative were first applied to historical akhbār by Stefan Leder, The Literary Use of the Khabar, and Daniel Beaumont, Hard-boiled: Narrative Discourse. 4 For the outlines of the method, see Motzki, Dating Muslim Traditions; for the method at work, see idem, Quo Vadis, Ḥadīṯ-Forschung? The same principle was applied to the sīra literature by Gregor Schoeler, Charakter und Authentie. 5 Schneider, Kinderverkauf und Schuldknechtschaft. Her approach provoked a debate with Motzki who reviewed her work in Der Prophet und die Schuldner; Schneider replied and Motzki reacted in Ar-Radd ʿalā r-radd (English transl. in Analysing Muslim traditions).

3 68 aerts yield a satisfactory result.6 It is my contention, however, that the CL theory has a future and that its results can sometimes be validated by data retrieved from other sources, such as the historiographical genres. It should be noted, however, that this cannot be established as a general rule within the confines of this article. The method proposed here is applicable exclusively to traditions that describe events or establish (or support) a rule that can be linked to a historical event described in the chronicles. For the majority of the ḥadīths, especially those dealing with religious matters, this is simply not the case. This article deals with two such traditions, both of which pertain to the history of the obligatory, ritual prayer (ṣalāt). In the past decades, the history of Islamic ritual has received relatively little attention. This is undoubtedly due to the problematic nature of the source material, which consists primarily of ḥadīths.7 Traditionally, this problem has been evaded by resorting to what I call the influence paradigm. The question whether the ṣalāt-prayer as well as other Islamic rituals are the product of Jewish, Christian, Pagan, or Zoroastrian influence has dominated the study of Islamic ritual for more than a century, but has not been particularly instructive. Influence is well-nigh impossible to measure and recourse is almost always taken to similarities, an approach that has produced contradictory conclusions.8 More importantly, the results of such undertakings tend to be conflated in time. The thesis that the ṣalāt-prayer has its origin in Jewish prayer, for instance, does not say much about the diachronic process of such adoption.9 On this point, an approach that is vested in the Ḥadīth, however difficult that may turn out to be, proves to be of great use, 6 Cook, Eschatology and the Dating of Traditions; see Görke, Eschatology, History, and the Common Link (especially ) for a critical reconsideration of Cook s findings. 7 Very little rules for the performance of ritual prayer, for instance, have a qurʾanic basis. The majority of the legal rules concerning ritual are based on the practice of the Prophet (Sunna), which is preserved for future generations in the form of Ḥadīth. 8 This is amply illustrated in the works of two German scholars published one year apart: C.H. Becker (Zur Geschichte des islamischen Kultus) compared the ṣalāt -prayer to eastern Christian mass, identified a number of (obvious) similarities between the Islamic and Christian rites, and concluded that the ṣalāt -prayer was borrowed from eastern Christianity, while Eugen Mittwoch (Zur Entstehungsgeschichte) argued on similar grounds that Islamic prayer derived from Jewish prayer. 9 Consider for instance the prayer times, which have been explained as being the result of Christian, Jewish, and Zoroastrian influence by, respectively, Goldziher, Die Bedeutung der Nachmittagszeit; Mittwoch, Zur Entstehungsgeschichte, 11-13; and Goldziher, Islamisme et parsisme, 133. S.D. Goitein (Studies in Islamic history, 85) has gone so far as to reconcile two of the three views by suggesting that the number of prayers (five) is the result of the Prophet deliberately averaging the three prayers of Judaism and the seven prayers of (monastic) Christianity.

4 The Prayers of Abū Muslim and al-maʾmūn 69 for it allows the creation of an image of the development, over time, of Islamic ritual.10 The first tradition deals with the festival prayer (ṣalāt al-ʿīdayn), a prayer that follows a different pattern from the regular prayers and which is performed on the Day of Sacrifice and the Breaking of the Fast. The second tradition concerns the pronouncement of a final takbīr (Allāhu akbar) after the taslīm (al-salāmu ʿalaykum wa-raḥmat Allāh) in the regular compulsory prayers. Both the festival prayer and the concluding takbīr are aspects of the Islamic ritual that, as I argue, were subject to change until the early Abbasid period.11 Two episodes preserved in the historiographical literature serve as point of departure; subsequently, the ḥadīth literature as well as the relevant reports are subjected to a CL analysis in order to demonstrate that the dating obtained by applying the CL analysis to the ḥadīths corresponds to the dates ascribed to the events in the chronicles. Two Episodes from the Historiographical Literature In 129/747, the Umayyad Caliphate was approaching its end: in Khurasan, Abū Muslim al-khurāsānī openly preached the Abbasid Revolution. The chroniclers (e.g., al-ṭabarī s Tārīkh) have reported that within a week after his public proclamation of the revolution,12 on the day of the Breaking of the Fast (ʿīd al-fiṭr), Abū Muslim asked Sulaymān b. Kathīr to lead him and his partisans (shīʿa) in prayer.13 Abū Muslim told his imam to observe the following rules in the festival prayer: (1) the actual ṣalāt should be performed before the sermon 10 Although Goldziher is often pictured as the pioneer of revisionism, I find my view to be in line with his basic assumption that the ḥadīth will not serve as a document for the history of the infancy of Islam, but rather as a reflection of the tendencies which appeared in the community during the mature stages of its development (Goldziher, Muslim Studies, II, 19). Even if the sources are of a later date than they purport to be and do not originate from the Prophet but instead originate from debates on socio-political, theological, or legal issues from later times, they still contain valuable information about these times and about these debates. 11 The performance of the taslīm itself was subject to change until the middle of the second/ eighth century. The original practice, preserved among Mālikites, Twelver Shiites and Ibāḍites, of pronouncing it only once was gradually superseded by the Sunnite practice of saying it twice; see Dutton, An innovation from the time of the Banī Hāshim. 12 This can be deduced from the dates mentioned in Ṭabarī, Tārīkh, II, (the reference is to the marginal pagination which corresponds to the older Leiden edition). 13 Ibid., II,

5 70 aerts (khuṭba); (2) the call to prayer (adhān) and its repetition in the mosque at the beginning of prayer (iqāma) ought to be omitted; (3) in the first rakʿa, six takbīrs (Allāhu akbar, God is great ) should be pronounced followed by a seventh, paired with a prostration, after the Qurʾan recitation; in the second rakʿa, five takbīrs should be pronounced, followed by a recitation and a sixth takbīr with a prostration; (4) the sermon would have to begin with a takbīr and end with a qurʾanic recitation. Abū l-khaṭṭāb, al-ṭabarī s informant, clarifies that this constituted a departure from the prayer practices of the ruling Umayyad dynasty.14 On feast days they used to issue a call to prayer, begin the service with a sermon, pronounce the iqāma, and only then proceed to the actual prayer.15 In other words, they followed the pattern of the congregational Friday prayer.16 Only four takbīrs were pronounced in the first rakʿa, and three in the second. According to reports of the awāʾil genre, these prayer practices had been introduced for the first time by the Umayyads, and hence implied a deviation from the Prophetic Sunna.17 Both ʿUthmān b. ʿAffān (d. 35/656) and Marwān b. al-ḥakam (d. 65/685) are said to have been the first to deliver the khuṭba before the prayer on feast days.18 Muʿāwiya (d. 60/680) and his governor in Iraq, Ziyād b. Abīhi (d. 53/673), and sometimes more generally the Umayyads or Marwānids are credited with being the first to sit down when delivering the khuṭba, allowing also the people to remain seated, as well as introducing the iqāma and adhān on the feast prayers, although the latter custom may also have been initiated by the anti-caliph Ibn al-zubayr (d. 73/692).19 Muʿāwiya and Ziyād are also charged with dropping (some of) the takbīrs that had to be pronounced during the bodily gestures of the ṣalāt, such as prostrating and standing up, 14 Ibid., II, The ambiguity about the order of the parts of prayer that could arise from al-ṭabarī s use of wa- in al-khuṭba wa-l-adhān has been noticed and resolved by later chroniclers, who used bi- when reverse chronology was implied (cf. Ibn al-jawzī, Muntaẓam, VII, 271; Miskawayh, Tajārib al-umam, II, 551). 16 In the translation of al-ṭabarī, John A. Williams interpreted the statement on the similarity to the Friday prayer (... bi-l-iqāma ʿalā ṣalāt yawm al-jumʿa) differently: the iqāma was added on Friday (Ṭabarī, The history, XXVII, 67). Later versions of the report which have ʿalā rasm ṣalāt yawm al-jumʿa support my interpretation (e.g., Miskawayh, Tajārib al-umam, II, 551). 17 Cf. Ibn Kathīr, Bidāya wa-nihāya, XIII, Ibn Abī Shayba, Muṣannaf, VII, 248:35753, 270:35987; Abū Zurʿa al-dimashqī, Tārīkh, 645:1907; Suyūṭī, Awāʾil, 34; Sayyid, Muʿjam, Ibn Abī Shayba, Muṣannaf, VII, 247: , 248:35744, 249: , 252:35792, 257:35844, 261:35892; Abū ʿArūba al-ḥarrānī, Awāʾil, 156:141, 157-8:144-5; Suyūṭī, Awāʾil, 23,

6 The Prayers of Abū Muslim and al-maʾmūn 71 after ʿUthmān had already started to pronounce them with a weaker voice.20 It is furthermore said that Marwān or his son Bishr were the first to take the pulpit outside for the two feast prayers, which were often performed outside in an open space (muṣallā).21 In view of these awāʾil reports, it is possible to come to an understanding of Abū Muslim s motivation. It is implied that the Umayyads started to use communal prayers as a means to spread political propaganda and augmented the role of the imam in prayer. Opponents of the Umayyad regime felt that the call to prayer de facto invited the believers to the sermon rather than the prayer, and that by putting the actual prayer last, the Umayyads made sure everyone effectively attended the sermon. In addition, the (Abbasid) sources reveal that the sermon had lost some of its religious relevance, a situation that Abū Muslim wished to rectify by putting the khuṭba between a takbīr and a qurʾanic recitation. The Umayyad reduction of the number of takbīrs if there was in fact a reduction is framed as the product of the Umayyads general indifference to the correct performance of prayer. Although the Umayyad prayer practice played at the most a minor role in the complex of feelings and motivations that lie at the basis of the Abbasid Revolution, the symbolic importance had not escaped Abū Muslim s attention and, according to this report, one of the first things he did was to set this straight. The second event to be dealt with here took place on Friday 16 Ramaḍān 216/27 October Caliph al-maʾmūn (d. 218/833) ordered the Baghdadi governor Isḥāq b. Ibrāhīm al-muṣʿabī to make sure that the troops and understandably also other believers pronounced a threefold takbīr while standing upright at the end of every ṣalāt. The practice was initiated in Medina and Ruṣāfa. These takbīrs ought not to be confused with the additional takbīrs that al-maʾmūn allegedly wanted to have pronounced in his own funeral service an altogether different question nor with any voluntary supplications that can be performed after the ṣalāt-prayer and which may or may not include takbīrs.23 The principal difference with the other event is that while the changes to the festival prayer ascribed to Abū Muslim eventually made it into the Classical Sunnite rite, al-maʾmūn s takbīrs did not. 20 Ibn Abī Shayba, Muṣannaf, VII, 248:35742; Abū ʿArūba al-ḥarrānī, Awāʾil, 157:143; Suyūṭī, Awāʾil, Ibn Abī Shayba, Muṣannaf, VII, 247:35734; Abū ʿArūba al-ḥarrānī, Awāʾil, 160:148; Suyūṭī, Awāʾil, 34; cf. Ibn Rusta, Aʿlāq, Ibn Abī Ṭāhir Ṭayfūr, Kitāb Baghdād, 265; Azdī, Tārīkh, 405; Ṭabarī, Tārīkh, III, Cf. Sourdel, La politique religieuse, 41-42; Nawas, Reexamination,

7 72 aerts Isnād Analysis of Reports on the Two Events In the Ḥadīth literature we find reports that either sanction or condemn the practices described in these two (supposedly) historical events. With respect to the position of the khuṭba in the festival prayers, we find for instance three very simple but oft-transmitted traditions that supply the practice of putting the prayer before the khuṭba on feast days, as Abū Muslim wanted it, with Prophetic legitimization: On the Day of Slaughtering and the Breaking of the Fast, the Prophet used to pray, then preach. On the feast days, the Prophet, Abū Bakr, and ʿUmar used to begin with the ṣalāt before the khuṭba. The Prophet, Abū Bakr, and ʿUmar used to pray the feast [prayer] before the khuṭba.24 To subject a tradition to the common link theory, all isnāds that support it in the different ḥadīth collections non-canonical as well as canonical need to be gathered and combined in one stemmatic diagram (which Juynboll calls an isnād bundle ). If the diagram allows the identification of a key figure (CL) from whom most or, ideally, all transmission lines branch out, it is thought that this person is responsible for the wording of the matn and its distribution. His or her position becomes furthermore exceedingly credible as he or she is supported in the next generation (i.e., upwards in the diagram) by other key figures from whom again multiple transmission paths fan out. Such transmitters are called partial common links (PCLs). In applying the CL theory to the traditions under scrutiny, I have treated them as three variants of a single tradition rather than as three separate traditions. In other words, their isnāds were gathered in a single diagram, which from then on served as the unit of analysis. This approach allows to determine when and with whom certain variants originated and branched off from the main transmission tree, whereas the common ancestry of the variants is fully accounted for. Whether or not similar ḥadīths are textual variants of the same tradition or altogether different traditions can only be established through trial and error. In this case, the 24 Ibn Ḥanbal, Musnad, IV, 325:4602, 466:4963; V, 170:5663; Bukhārī, Ṣaḥīḥ, I, ; Muslim, Ṣaḥīḥ, II, 605:8; Ibn Māja, Sunan, I, 407:1276; Tirmidhī, Ṣaḥīḥ (Sunan), I, 393:531; Bazzār, Baḥr (Musnad), XII, 136:5710; Nasāʾī, Sunan, III, 183; Ibn Ḥibbān, Ṣaḥīḥ, III, 417:2822; Ṭabarānī, Muʿjam awsaṭ, II, 111:1416; cf. Mizzī, VI, 122:7805, 126:7823, 159:8045.

8 The Prayers of Abū Muslim and al-maʾmūn al-bazzār Ibn Ḥibbān al-ṭabarānī 73 al-bukhārī al-tirmidhī Ibn Māja Muslim al-nasāʾī Muḥammad b. al-muthannā Ibn Abī Shayba Ibn Ḥanbal Ḥammād b. Usāma al-kūfī ʿAbda b. Sulaymān Ḥammād b. Masʿada ʿUbayd Allāh b. ʿUmar b. Ḥafṣ Nāfijiʿ ʿAbd Allāh b. ʿUmar Prophet Diagram 1 Isnād Bundle of... the Prophet used to pray, then preach superimposition of the three isnād diagrams is permissible because it gives the same result (i.e., the same CL) as all three separate diagrams. ʿUbayd Allāh b. ʿUmar b. Ḥafṣ is the CL of this tradition in favor of putting the actual ṣalāt before the sermon on feast days (see Diagram 1),25 and of which the extant Ḥadīth collections have recorded three variants. ʿUbayd Allāh, a great-great-grandson of ʿUmar b. al-khaṭṭāb, was a Medinan jurist and Ḥadīth transmitter who died in 147/ He is sometimes called one of the so-called Seven Jurists of Medina,27 although, as Juynboll noticed, the first and only biographer to list him among the Seven Jurists appears to be Ibn Ḥajar.28 His position as CL is confirmed by three PCLs: Ḥammād b. Usāma (d. 201/817) and ʿAbda b. Sulaymān (d. 188/804) from Kufa, and Ḥammād b. Masʿada (d. 202/817) 25 All isnād diagrams are simplified: whenever transmitters have been omitted, this is indicated by the use of an arrow on the transmission line. 26 Mizzī, Tahdhīb, V, 54: See, for instance, Madelung, Succession to Muḥammad, 83 n Juynboll, Encyclopedia, 400 n. 1.

9 74 aerts from Basra. The single strand from the Prophet to al-ṭabarānī is, in all likelihood, forged by one of the later transmitters along this path, who copied the ḥadīth but furnished it with a new isnād in order to gain prestige, a phenomenon which has been called the spread of isnāds by Schacht and Cook, and diving by Juynboll.29 The single strand from the Prophet to ʿUbayd Allāh is difficult if not impossible to verify. Whatever the case, even if the CL faithfully transmitted the tradition from the person he named as his informant, the fact remains that the tradition under scrutiny suddenly gained popularity in the early Abbasid period, when ʿUbayd Allāh started to disseminate it. Iraqi students especially included it in their repertoire of Prophetic traditions, and spread it to ever more traditionists in the subsequent generations. This is in correspondence with the date of the episode involving Abū Muslim as retrieved from the chronicles. Moreover, other traditions that endorse this sequence of ṣalāt and khuṭba on feast days, although as a side note, date to the same period according to the CL theory: Juynboll has already demonstrated that Ibn Jurayj (d. 150/767) is the CL of two traditions in which the Prophet orders women to perform dhikr and give alms respectively after the feast prayer.30 The second ḥadīth concerns, as related above, al-maʾmūn s interference with the regular compulsory prayer. Relatively few traditions exist which refer to the concluding takbīr. On the other hand, in some traditions it is unambiguously stated that the ritual prayer begins with a takbīr, and is concluded by the taslīm: The Prophet said: The key to the prayer is purification, the takbīr constitutes its consecration/beginning, and the taslīm is its desecration/end. 31 From Diagram 2, a schematic representation of all the transmission paths of this ḥadīth, again one transmitter emerges as an undisputable CL: Sufyān 29 Schacht, Origins, ; Cook, Early Muslim dogma, ; Juynboll, Encyclopedia, xxii-xxiii. 30 Juynboll, Encyclopedia, 214, Sindī, Musnad al-shāfiʿī, I, 70:206; ʿAbd al-razzāq, Muṣannaf, II, 72:2539 (see also 2340); Ibn Abī Shayba, Muṣannaf, I, 208: ; Ibn Ḥanbal, Musnad, II, 39:1006, 60:1072; Dārimī, Musnad (Sunan), I, 539:714; Abū Dāwūd, Sunan, I, 56:61, 208:618; Tirmidhī, Sunan, I, 13:3; Ibn Māja, Sunan, I, 101: ; Bazzār, Musnad, II, 236:633; Abū Yaʿlā al-mawṣilī, Musnad, I, 271:612, 476:1120; Ṭabarānī, al-muʿjam al-awsaṭ, III, 36:2390; Dāraquṭnī, Sunan, II, 178:1356, 179: , 216:1421; Abū Nuʿaym, Ḥilyat al-awliyāʾ, VII, 124; VIII, 372; Bayhaqī, al-sunan al-kubrā, I, 140:358, 531:3971; cf. Mizzī, Tuḥfa, VII, 442:10265; Juynboll, Encyclopedia, 639.

10 The Prayers of Abū Muslim and al-maʾmūn al-tirmidhī Ibn Ḥanbal ʿAbd al-raḥmān Abū Dāwūd Abū Yaʿlā Ibn Māja Wakiʿ Ibn Abī Shayba al-dāraquṭnī al-bazzār al-dārimī al-shāfijiʿī 75 al-dāraquṭnī Abū Yaʿlā Ibn Abī Shayba Ibn Māja Sufyān al-thawrī ʿAbd Allāh b. Muḥammad b. ʿUqayl Muḥammad Ibn al-ḥanafijiyya ʿAlī Prophet Diagram 2 Isnād Bundle of... and the taslīm is [the prayer s] end al-thawrī (d. 161/778). The CL theory dates this tradition to the middle of the second/eighth century. A rare ḥadīth that does mention a takbīr after the taslīm is the following: Ibn ʿAbbās reported: We used to know that the Prophet had completed the ṣalāt by [hearing] the takbīr. 32 From the isnād bundle of this tradition, presented in Diagram 3, Sufyān b. ʿUyayna (d. 198/814) emerges unambiguously as key figure. This tradition admits the possibility that the final takbīr was an inherent part of the ṣalāt. 32 Sindī, Musnad al-shāfiʿī, I, 99:287; Ḥumaydī, Musnad, I, 225:480; Ibn Ḥanbal, Musnad, II, 454:1933; Abū Yaʿlā al-mawṣilī, Musnad, II:401:2388; Bukhārī, Ṣaḥīḥ, I, 216; Muslim, Ṣaḥīḥ, I, 410: ; Abū Dāwūd, Sunan, I, 306:1002; Nasāʾī, Sunan, III, 67; Nasāʾī, al-sunan al-kubrā, I, 204:1259; Ibn Khuzayma, Ṣaḥīḥ, III, 102:1706; Abū ʿAwwāna, Musnad (al-mustakhraj), I, 553: ; Ibn Ḥibbān, Ṣaḥīḥ, III, 254:2231; Ṭabarānī, al-muʿjam al-awsaṭ, II, 186:1669; Ṭabarānī, al-muʿjam al-kabīr, IX, 424:12200; Bayhaqī, al-sunan al-kubrā, II, 262: ; cf. Mizzī, Tuḥfa, V, 256:6512; Juynboll, Encyclopedia, 594.

11 76 aerts Ibn Ḥibbān ʿAbū ʿAwwāna al-bayhagī Ibn Khuzayma Abū Dāwūd al-bukhārī Muslim al-nasāʾī Abū Yaʿlā al-ṭabarānī ʿAbd al-jabbār b. al-ʿulāʾ al Ḥumaydī Ibn Ḥanbal al-shāfijiʿī Sufyān b. ʿUyayna ʿAmr b. Dīnār Nāfijidh Abū Maʿbad Ibn ʿAbbās Diagram 3 Isnād Bundle of We used to know that the Prophet had completed the ṣalāt by [hearing] the takbīr Perhaps therefore it is mentioned in the Sunnite Ḥadīth collections almost invariably in conjunction with another tradition that clarifies that the takbīr is not actually part of the prayer, but instead of the supererogatory supplications that follow the prayer: Ibn ʿAbbās said: In the time of the Prophet, people said dhikr aloud when they left the compulsory prayer. He also said: I used to know [that prayer was done] when they left like that and when I heard that. 33 This tradition, again ascribed to Ibn ʿAbbās and notably non-prophetic, is pretty much meaningless without reference to the previous one. Its function appears to be to take away any doubts about the interpretation of the other ḥadīth ascribed to Ibn ʿAbbās. The CL is slightly more difficult to identify: both Muḥammad b. Bakr al-bursānī (d. 204/819f) and ʿAbd al-razzāq (d. 211/827) qualify.34 Whoever it is, this ḥadīth stems from the end of the second/ beginning of the ninth century. In the few extant ḥadīth- and fiqh-works which predate 33 ʿAbd al-razzāq, Muṣannaf, II, 245:3225; Ibn Ḥanbal, Musnad, III, 456:3478; Bukhārī, Ṣaḥīḥ, I, 216; Muslim, Ṣaḥīḥ, I, 410:122; Abū Dāwūd, Sunan, I, 306:1003; Ibn Khuzayma, Ṣaḥīḥ, III, 102:1707; Abū ʿAwwāna, Musnad, I, 552:2065; cf. Mizzī, Tuḥfa, V, 257: Cf. Juynboll, Encyclopedia, 221.

12 The Prayers of Abū Muslim and al-maʾmūn 77 al-maʾmūn s caliphate, this ḥadīth is not included.35 In all of the thematically arranged (muṣannaf ) collections in which it occurs, it immediately precedes or follows the other ḥadīth of Ibn ʿAbbās ( We used to know that the Prophet.... ), except in the earliest such collection, that of ʿAbd al-razzāq, where it occurs in isolation. Furthermore, ʿAbd al-razzāq s version is the only one where the isnād of the first part of the tradition reaches back only to Abū Maʿbad (d. 104/722-23), a mawlā of Ibn ʿAbbās, whereas in the later collections it is reported by Ibn ʿAbbās himself. The isnād of the second part extends to Ibn ʿAbbās in all instances. In its totality the ḥadīth is non-prophetic. The actors are the Prophet s Companions; at best a passive and tacit approval of the Prophet, who is not explicitly present, is implied by setting the event in the time of the Prophet (ʿalā ʿahd al-nabī), the very presence of this stipulation betraying the later origin of this ḥadīth. The most plausible explanation for this anomaly is that the ḥadīth under scrutiny existed before (i.e., earlier than the CL), but was converted from a Successor tradition to a Companion tradition by our CL.36 This was done because its attribution to Ibn ʿAbbās facilitated its collocation and association with the other ḥadīth of Ibn ʿAbbās, i.e., the tradition on the takbīr ending prayer. From now on, it would be interpreted in light of this tradition with which it forms a pair, and which, in its own turn, was supposed to be clarified by this tradition.37 This must have happened after ʿAbd al-razzāq composed his Muṣannaf, and Muḥammad b. Bakr is the more probable CL. So even though the gist of the ḥadīth is likely much older, it was suddenly transformed and more vigorously circulated at the end of the second/eight or the beginning of the third/ninth century. At the same time, this tradition became meaningful in the context of the discussion on takbīr at the end of the ṣalāt, which was not its original intent.38 This is reflected not only in its adjacency to the other tradition, but also in the chapter (bāb) titles under which they appear in the collections. With ʿAbd al-razzāq, that title is Bāb makth al-imām baʿdamā yusallim, introducing a section in which he was concerned with acts to be performed by the imam, after the taslīm, while the people are leaving. This is indeed very 35 In the works of Mālik b. Anas (d. 179/795) (al-muwaṭṭaʾ), al-shaybānī (d. 189/805) (al-aṣl fī l-furūʿ and al-āthār), al-shāfiʿī (d. 204/820) (Sindī, Musnad al-shāfiʿī), and the traditions ascribed to Abū Ḥanīfa (d. 150/767) (Abū Nuʿaym, Musnad Abī Ḥanīfa; Khwārazmī, Jāmiʿ), the takbīr is only mentioned in connection with the opening of the ṣalāt, the bodily gestures of the ṣalāt, the funeral service, the ḥajj rituals, and the adhān. 36 Schacht (Origins, 30-33) first formulated the hypothesis that Successor traditions predate Companion traditions, which in turn are earlier than Prophetic traditions. 37 Cf. Ibn Ḥajar, Fatḥ al-bārī, II, In fact, the takbīr is not even mentioned, only the dhikr is.

13 78 aerts compatible with the wording of the tradition under scrutiny. In some later collections, however, the tradition is brought to bear on the takbīr. With Abū Dāwūd, we find Bāb al-takbīr baʿda l-ṣalāt, with Ibn Khuzayma Bāb rafʿ al-ṣawt bi-l-takbīr wa-l-dhikr ʿinda qaḍāʾ al-imām al-ṣalāt. Al-Bukhārī and Muslim stick to the more conservative Bāb al-dhikr baʿda l-ṣalāt, but they too leave out the important reference to the imam. Conclusion For all four traditions the dating arrived at by applying the CL theory corresponds with the chronology of related episodes in the chronicles, which describe the (attempted) institution of rules that are provided with Prophetic legitimization in the ḥadīths. It has been shown that shortly after the Abbasids had changed the order of the sermon and the prayer proper on feast days, the first traditions turned up to legitimize the new practice, and they were quickly popularized. Likewise, a number of ḥadīths correspond to al-maʾmūns attempt to put a takbīr at the end of every ṣalāt-prayer. The ḥadīth of Sufyān al-thawrī is representative of the mid-second/eighth century: the takbīr was the established opening of the ṣalāt-prayer and was never considered a possible concluding part of prayer. At the time of al-maʾmūn, perhaps slightly earlier, traditions surfaced which could be interpreted as sanctioning a takbīr at the end of the ṣalāt; others were quick to have the same Companion clarify that this interpretation could not possibly be right, and that no takbīr should be pronounced at the end of prayer. By this time, a discussion had begun about a threefold takbīr at the end of the ṣalāt.39 In this debate, the Caliph took a stance, but one that would not make it into Sunnite orthopraxy. More than so far has been the case, existing theories of dating ḥadīths should be put to the test; the ability to use with greater confidence the Ḥadīth as a historical source would vastly improve our understanding of the formative period of Islam (until the third/ninth century). Data gleaned from historiographical sources can validate the results of the CL theory, and a better under- 39 Traces of the debate can be found elsewhere: the Mālikite jurist ʿAbd al-malik b. Ḥabīb (d. 238/852) strongly encouraged a threefold takbīr especially among soldiers (quoted in Ibn Baṭṭāl, Sharḥ, II, 458; not present in the partial edition by Ossendorf-Conrad, Das K. al-wāḍiḥa ). Another Mālikite scholar, Muḥammad b. Aḥmad al-ʿutbī (d. 255/869), declared it a novelty of the Abbasids (al-musawwida; Ibn Baṭṭāl, Sharḥ, II, 458; the so-called al-ʿutbiyya is also partially edited as ʿUtbī, Kitāb al-ḥajj).

14 The Prayers of Abū Muslim and al-maʾmūn 79 standing of the historical context can shed light on the motives of the CL to bring a certain ḥadīth into circulation, i.e., to support a legal opinion that he or she considered correct. In connection with the history of the ṣalāt-prayer, this exercise has indicated that Muslim historiographers either did not try to rewrite the history of the Muslim prayer ritual so as to ascribe all aspects of it to the Prophet, or did a very poor job in doing so, as the literature makes no secret of the different stages of the evolution of the ritual prayer, which lasted until the Abbasid period. It may be so that major concepts of the ṣalāt-prayer are borrowed from the Judeo-Christian tradition, but genuine Muslim struggle and decision-making deserve our attention as well, for instance in the case of the reversal of the khuṭba and the ṣalāt in festival prayers and the establishing of the closing rituals of the compulsory prayer. References Primary Sources ʿAbd al-razzāq al-ṣanʿānī (d. 211/827), al-muṣannaf, ed. by Ḥabīb al-raḥmān al-aʿẓamī, 12 vols, Beirut: al-maktab al-islāmī, 1983 Abū ʿArūba al-ḥarrānī, al-ḥusayn b. Abī Maʿshar (d. 318/931), Kitāb al-awāʾil, ed. by Mashʿal b. Bānī l-jibrīn al-muṭīrī, Beirut: Dār Ibn Ḥazm, 2002 Abū ʿAwwāna (d. 316/928f.), Musnad (al-mustakhraj), ed. by Amīn b. ʿĀrif al-dimashqī, 5 vols, Beirut: Dār al-maʿrifa, 1998 Abū Dāwūd (d. 275/889), Sunan, ed. by Muḥammad ʿAbd al-ʿazīz al-khālidī, 3 vols, Beirut: Dār al-kutub al-ʿilmiyya, 2007 Abū Nuʿaym al-iṣfahānī (d. 430/1038), Ḥilyat al-awliyāʾ wa-ṭabaqāt al-aṣfiyāʾ, 10 vols, Beirut: Dār al-kutub al-ʿilmiyya, 1988, Musnad al-imām Abī Ḥanīfa, ed. by Naẓar Muḥammad al-fāryābī, Riyadh: Maktabat al-kawthar, Abū Yaʿlā al-mawṣilī (d. 307/919), Musnad, ed. by Muṣṭafā ʿAbd al-qādir ʿAṭā, 7 vols, Beirut: Dār al-kutub al-ʿilmiyya, 1998 Abū Zurʿa al-dimashqī (d. 281/895), Tārīkh, ed. by Shukr Allāh b. Niʿmat Allāh al-qawjānī, Damascus: Majmaʿ al-lugha al-ʿarabiyya, 1980 Aḥmad b. Ḥanbal (d. 241/855), Musnad, ed. by Aḥmad Muḥammad Shākir and Ḥamza Aḥmad al-zayn, 20 vols, Cairo: Dār al-ḥadīth, 1995 Al-Azdī, Abū Zakariyyāʾ Yazīd b. Muḥammad (d. 334/945), Tārīkh al-mawṣil, ed. by Alī Ḥabība, Cairo: Dār al-taḥrīr, 1967 Al-Bayhaqī, Abū Bakr Aḥmad b. al-ḥusayn (d. 458/1066), al-sunan al-kubrā, ed. by Muḥammad ʿAbd al-qādir ʿAṭā, 11 vols, Beirut: Dār al-kutub al-ʿilmiyya, 2003

15 80 aerts Al-Bazzār, Aḥmad b. ʿAmr (d. 292/904f.), al-baḥr al-zakhkhār (Musnad), ed. by Maḥfūẓ al-raḥmān Zayn Allāh, ʿĀdil b. Ṣaʿd and Ṣabrī ʿAbd al-khāliq al-shāfiʿī, 18 vols, Medina: Maktabat al-ʿulūm wa-l-ḥukm, Al-Bukhārī, Muḥammad b. Ismāʿīl (d. 256/870), al-jāmiʿ al-ṣaḥīḥ, ed. by Ludolf Krehl and Th.W. Juynboll, 4 vols, Leiden: Brill, Al-Dāraquṭnī (d. 385/995), Sunan, ed. by Shuʿayb al-arnaʾūṭ, 6 vols, Beirut: Muʾassasat al-risāla, 2004 Al-Dārimī (d. 255/869), Musnad (Sunan), ed. by Ḥusayn Salīm Asad al-dārānī, Riyadh: Dār al-mughnī li-l-nashr wa-l-tawzīʿ, 2000 Al-Ḥumaydī, ʿAbd Allāh b. al-zubayr (d. 219/834), Musnad, ed. by Ḥabīb al-raḥmān al-aʿẓamī, 2 vols, Beirut: ʿĀlam al-kutub, 1381 Ibn Abī Shayba (d. 235/849), Muṣannaf, ed. by Kamāl Yūsuf al-ḥūt, 7 vols, Beirut: Dār al-tāj, 1989 Ibn Abī Ṭāhir Ṭayfūr (d. 280/893), Kitāb Baghdād, ed. by H. Keller, Leipzig: Harrassowitz, 1908 Ibn al-jawzī (d. 597/1200), al-muntaẓam fī tārīkh al-mulūk wa-l-umam, ed. by Muḥammad ʿAbd al-qādir ʿAṭā and Muṣṭafā ʿAbd al-qādir ʿAṭā, 18+1 vols, Beirut: Dār al-kutub al-ʿilmiyya, 1992 Ibn Baṭṭāl, Abū l-ḥusayn ʿAlī b. Khalaf (d. 449/1057), Sharḥ Ṣaḥīḥ al-bukhārī, ed. by Abū Tamīm Yāsir b. Ibrāhīm, 10 vols, Riyadh: Maktabat al-rushd, 2003 Ibn Ḥajar al-ʿasqalānī (d. 852/1449), Fatḥ al-bārī bi-sharḥ Ṣaḥīḥ al-bukhārī, ed. by ʿAbd al-ʿazīz b. ʿAbd Allāh b. Bāz, Muḥammad Fuʾād ʿAbd al-bāqī and Muḥibb al-dīn al-khaṭīb, 13 vols, Beirut: Dār al-maʿrifa, 1379 Ibn Ḥibbān al-bustī (d. 354/965), Ṣaḥīḥ, ed. by Maktab al-buḥūth wa-l-dirāsāt, 6 vols, Beirut: Dār al-fikr, 1996 Ibn Kathīr (d. 774/1373), al-bidāya wa-l-nihāya, ed. by ʿAbd Allāh b. ʿAbd al-muḥsin al-turkī, 21 vols, Cairo: Dār Hajr, 1997 Ibn Khuzayma (d. 311/923), Ṣaḥīḥ, ed. by Muḥammad Muṣṭafā al-aʿẓamī, 4 vols, Beirut: al-maktab al-islāmī, 1980 Ibn Māja (d. 273/887), Sunan, ed. by Muḥammad Fuʾād ʿAbd al-bāqī, 2 vols, Cairo: Dār Iḥyāʾ al-kutub al-ʿarabiyya, Ibn Rusta, Abū ʿAlī Aḥmad b. ʿUmar (d. 300/912), al-aʿlāq al-nafīsa, ed. by M.J. de Goeje, Leiden: Brill, 1892 Al-Khwārazmī, Muḥammad b. Maḥmūd (d. 665/1266f.), Jāmiʿ masānīd al-imām alaʿẓam, 2 vols, Hyderabad: Majlis Dāʾirat al-maʿārif al-niẓāmiyya, 1332 Mālik b. Anas (d. 179/796), al-muwaṭṭaʾ, ed. by Muḥammad Fuʾād ʿAbd al-bāqī, 2 vols, Cairo: Dār Iḥyāʾ al-kutub al-ʿarabiyya, 1951 Miskawayh, Aḥmad b. Muḥammad (d. 421/1030), Tajārib al-umam wa-taʿāqub alhimam, ed. by Sayyid Kasrawī Ḥasan, 7 vols, Beirut: Dār al-kutub al-ʿilmiyya, 2003

16 The Prayers of Abū Muslim and al-maʾmūn 81 Al-Mizzī, Jamāl al-dīn Yūsuf b. ʿAbd al-raḥmān (d. 742/1341), Tuḥfat al-ashrāf bi-maʿrifat al-aṭrāf, ed. by ʿAbd al-ṣamad Sharaf al-dīn, 16 vols, Beirut: Dār al-kutub al-ʿilmiyya, 1999, Tahdhīb al-kamāl fī asmāʾ al-rijāl, ed. by Bashshār ʿAwwād Maʿrūf, 8 vols, Beirut: Muʾassasat al-risāla, 2010 Muslim (d. 261/875), Ṣaḥīḥ, ed. by Fuʾād ʿAbd al-bāqī, 5 vols, Cairo: Dār Iḥyāʾ al-kutub al-ʿarabiyya, Al-Nasāʾī (d. 303/915), al-mujtabā (Sunan) bi-sharḥ al-ḥāfiẓ Jalāl al-dīn al-suyūṭī wa-ḥāshiyat al-imām al-sindī, 8 vols, Cairo, 1348, al-sunan al-kubrā, ed. by Abū Anas Jād Allāh b. Ḥasan al-khaddāsh, 3 vols, Beirut: Maktabat al-rushd, 2006 Al-Sayyid, Fuʾād Ṣāliḥ, Muʿjam al-awāʾil fī tārīkh al-ʿarab wa-l-muslimīn, Beirut: Dār al-manāhil, 1992 Al-Shaybānī (d. 189/805), Kitāb al-aṣl fī l-furūʿ (al-mabsūṭ), ed. by Abū l-wafāʾ al-afghānī, 5 vols, Beirut: ʿĀlam al-kutub, 1990, Kitāb al-āthār, ed. by Abū l-wafāʾ al-afghānī, 2 vols, Beirut: Dār al-kutub al-ʿilmiyya, 1993 Al-Sindī, Muḥammad ʿĀbid, Tartīb Musnad al-imām Abī ʿAbd Allāh Muḥammad b. Idrīs al-shāfiʿī, ed. by Yūsuf ʿAlī al-zawāwī al-ḥasanī and ʿIzzat al-ʿaṭṭār al-ḥusaynī, 2 vols, Beirut: Dār al-kutub al-ʿilmiyya, 1951 Al-Suyūṭī, Jalāl al-dīn (d. 911/1505), al-wasāʾil ilā maʿrifat al-awāʾil, ed. by ʿAbd al-raḥmān al-jūzū, Beirut: Dār Maktabat al-ḥayā, 1988 Al-Ṭabarānī, Abū l-qāsim Sulaymān b. Aḥmad (d. 360/971), al-muʿjam al-kabīr, ed. by Ḥamdī b. ʿAbd al-majīd al-salafī, 25 vols, Cairo: Maktabat Ibn Taymiyya, 1983, al-muʿjam al-awsaṭ, ed. by Ṭāriq b. ʿIwaḍ Allāh b. Muḥammad and ʿAbd al-muḥsin b. Ibrāhīm al-ḥusaynī, 10 vols, Cairo: Dār al-ḥaramayn, 1995 Al-Ṭabarī, Tārīkh al-rusul wa-l-mulūk, ed. by Muḥammad Abū l-faḍl Ibrāhīm, Cairo: Dār al-maʿārif, 1969, The history of al-ṭabarī, ed. by Ehsan Yar-Shater, 40 vols, Albany NY: State University of New York Press, Al-Tirmidhī (d. 279/892), al-jāmiʿ al-ṣaḥīḥ (Sunan), ed. by Maḥmūd Muḥammad Naṣṣār, 5 vols, Beirut: Dār al-kutub al-ʿilmiyya, 2007 Al-ʿUtbī, Muḥammad b. Aḥmad (d. 254/868), Kitāb al-ḥajj min al-masāʼil al-mustakhraja min al-asmiʿa mimmā laysa fī l-mudawwana, ed. by Miklos Muranyi, Beirut: Dār Ibn Ḥazm, 2007 Secondary Sources Beaumont, Daniel, Hard-boiled: Narrative Discourse in Early Muslim Traditions, in: Studia Islamica 83 (1996), 5-31

17 82 aerts Becker, C.H., Zur Geschichte des islamischen Kultus, in: Der Islam 3 (1912), Cook, Michael, Early Muslim Dogma: A Source-critical Study, Cambridge: Cambridge University Press, 1981, Eschatology and the Dating of Traditions, in: Princeton Papers in Near Eastern Studies 1 (1992), Dutton, Yasin, An Innovation from the Time of the Banī Hāshim : Some Reflections on the Taslīm at the End of the Prayer, in Journal of Islamic Studies 16/2 (2005), Goitein, S.D., Studies in Islamic History and Institutions, Leiden: Brill, 1968 Goldziher, Ignaz, Islamisme et parsisme, in: Actes du premier Congrès international d histoire des religions, Paris: E. Leroux, 1900, , Die Bedeutung der Nachmittagszeit im Islam, in: Archiv für Religionswissenschaft 9 (1906), , Muslim Studies, 2 vols, London: Allen & Unwin, Görke, Andreas, Eschatology, History, and the Common Link: A Study in Methodology, in: Herbert Berg (ed.), Method and Theory in the Study of Islamic Origins, Leiden: Brill, 2003, Günther, Sebastian, Fictional Narration and Imagination Within an Authoritative Framework: Towards a New Understanding of Ḥadīth, in: Stefan Leder (ed.), Storytelling in the Framework of Non-fictional Arabic Literature, Wiesbaden: Harrassowitz, 1998, , Modern Literary Theory Applied to Classical Arabic Texts: Ḥadīth revisited, in: Verena Klemm and Beatrice Gruendler (eds.), Understanding Near Eastern Literatures: A Spectrum of Interdisciplinary Approaches, Wiesbaden: Reichert, 2000, Juynboll, G.H.A., Some Isnād-Analytical Methods Illustrated on the Basis of Several Women-Demeaning Sayings from Ḥadīth Literature, in: Al-Qanṭara 10/2 (1989), , Encyclopedia of Canonical Ḥadīth, Leiden: Brill, 2007 Leder, Stefan, 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, Princeton, NJ: Darwin Press, 1992, Madelung, Wilferd, The Succession to Muḥammad: A Study of the Early Caliphate, Cambridge: Cambridge University Press, 1997 Mittwoch, Eugen, Zur Entstehungsgeschichte des islamischen Gebets und Kultus, Berlin: Reimer, 1913 Motzki, Harald, Quo Vadis, Ḥadīṯ-Forschung? Eine kritische Untersuchung von G.H.A. Juynboll, Nāfiʿ the mawlā of Ibn ʿUmar, and his position in Muslim ḥadīṯ literature, in: Der Islam 73 (1996), 40-80; ; English transl. in Harald Motzki et al.,

18 The Prayers of Abū Muslim and al-maʾmūn 83 Analysing Muslim Traditions: Studies in Legal, Exegetical and Maghāzī Ḥadīth, Leiden: Brill, 2010, , Der Prophet und die Schuldner: Eine ḥadīṯ-untersuchung auf dem Prüfstand, in: Der Islam 77 (2000), 1-83, Ar-Radd ʿalā r-radd: Zur Methodik der Ḥadīṯ-Analyse, in: Der Islam 78 (2001), ; English transl. in Harald Motzki et al., Analysing Muslim Traditions: Studies in Legal, Exegetical and Maghāzī Ḥadīth, Leiden: Brill, 2010, , Dating Muslim Traditions: A Survey, in: Arabica 52/2 (2005), Motzki, Harald, Nicolet Boekhoff-van der Voort and Sean W. Anthony, Analysing Muslim Traditions: Studies in Legal, Exegetical and Maghāzī Ḥadīth, Leiden: Brill, 2010 Nawas, John A., A Reexamination of Three Current Explanations for al-mamunʼs Introduction of the Mihna, in: International Journal of Middle East Studies 26/4 (1994), Ossendorf-Conrad, Beatrix, Das K. al-wāḍiḥa des ʿAbd al-malik b. Ḥabīb. Edition und Kommentar zu Ms. Qarawiyyīn 809/40 (Abwāb al-ṭahāra), Beirut/Stuttgart: Franz Steiner, 1994 Sadeghi, Behnam, The Traveling Tradition Test: A Method for Dating Traditions, in: Der Islam 85 (2010), Schacht, Joseph, The Origins of Muhammadan Jurisprudence, Oxford: Clarendon Press, 1959 Schneider, Irene, Kinderverkauf und Schuldknechtschaft, Stuttgart: Harrassowitz, 1999 Schoeler, Gregor, Charakter und Authentie der muslimischen U berlieferung über das Leben Mohammeds, Berlin: De Gruyter, 1996; English transl. by Uwe Vagelpohl, ed. by James Montgomery, The Biography of Muḥammad: Nature and Authenticity, Abingdon/New York: Routledge, 2011 Sourdel, Dominique, La politique religieuse du calife ʿabbāside al-maʾmun, in: Revue des Études Islamiques 30/1 (1962), Van Ess, Josef, Zwischen Ḥadīṯ und Theologie: Studien zum Entstehen prädestinatianischer U berlieferung, Berlin: de Gruyter, 1975