Printer Friendly

The succession to the caliph Musa al-Hadi.

Harun al-Rashid succeeded his elder brother Musa al-Hadi to the caliphate in confused circumstances. It is clear that Musa had not wanted his brother to succeed him, but not whether in his short caliphate Musa had formally replaced Harun as heir apparent with his own young son. The open question is whether Harun succeeded to the caliphate by virtue of being heir apparent, or owing to a military intervention after Musa's sudden death. Al-Tabari clearly manipulates his textual evidence to prove the former case, and is greatly helped by an influential Barmakid tradition making the same point. If al-Tabari's editorial intervention can be discounted, it is arguable on the basis of both textual and numismatic evidence that Harun's caliphate was as much the creation of the army as was that of the first Abbasid caliph, [Abu'l-.sup.[subset]]Abbas.

THE MANNER IN WHICH Harun al-Rashid succeeded his brother Musa al-Hadi in the caliphate in 170/786 has been of great interest to both medieval and modern historians. The evidence of the medieval texts has been studied in detail by Sabatino Moscati, Nabia Abbott, and Hugh Kennedy, and important contributions based on numismatic evidence have been made by Michael Bonner. (1) Also now relevant in the light of Bonner's work is a discussion by Patricia Crone of the slogan al-rida min al Muhammad. (2) This present article is, in the main, a re-examination of the textual evidence, with the benefit of these later studies. It focuses especially on one question, whether al-Hadi in his short caliphate did or did not formally replace his brother Harun as heir apparent with his own young son [Ja.sup.[subset]]far. The question is important for our understanding of exactly how Harun eventually became caliph, but it has not, to my knowledge, been clearly addressed before. It has been obscured in our principal textual source by the entire ly synthetic issue of how al-Hadi died.

Moscati's account of the succession crisis in al-Hadi's caliphate is mainly an accurate, though uncritical, reading of al-Tabari. It is al-Tabari, however, who obscures the question of Harun's replacement as heir apparent, and Moscati does not raise the question either. (3) Abbott had the advantage over Moscati of knowing a report provided by Ibn Abi [Usaybi.sup.[subset]a that claims Harun actually was replaced. She accepts this evidence, but does not comment on al-Tabari's efforts to create the strong impression that he was not. (4)

Kennedy relates the succession crisis to the wider issue of a power struggle under al-Hadi between the state bureaucracy and the military establishment. Al-Hadi is seen as favoring the military at the expense of the bureaucracy, and the Barmakid Yayhya ibn Khalid appears as the leading representative of the latter. Whether al-Hadi's eventual successor would be his brother Harun or his son [Ja.sup.[subset]]far was a question of which faction would prevail--Yahya ibn Khalid who supported Harun or the military who wanted [Ja.sup.[subset]]far. Kennedy too relies mainly on al-Tabari, and he believes that the formal replacement of Harun with Ja[subset]far never actually took place. (5)

Bonner is substantially in agreement with Kennedy's overall analysis, although he sees a danger of reading into this early conflict a pattern of the later Abbasid period. (6) He notes that not all the army commanders were opposed to Harun or, as both he and Kennedy see it, to the Barmakids. (7) In his view, court intrigue has received more than its fair share of attention and he emphasizes instead that the struggle over the succession was not confined to Baghdad and that Harun's accession was more than "a neatly executed coup d'etat pulled off in Baghdad by the Barmakids and al-Khayzuran." (8) Harun--or again as he sees it, the Barmakids--had a provincial power base in the north and west of the empire, where they had the support of several military commanders including the leading Khurasani officer Khuzayma ibn Khazim. It was this Khuzayma who intervened forcefully and decisively in the capital when alhadi died, to ensure that it was Harun who succeeded him. (9) As to whether Harun had been formally replaced as heir apparent, Bonner argues that he had, and he adduces the correspondence of textual and numismatic evidence on this point. He does not, however, deal with it as an issue within the textual sources. (10) In order to do so now, we shall begin with an account of how both Musaal-Hadi and his brother Harun al-Rashid acquired title to the succession from their father Muhammad al-Mahdi.

Once al-Mahdi succeeded his father Abu [Ja.sup.[subset]]far al-Mansur to the caliphate in 158/775, he moved quickly to appoint his own son Musa as his heir apparent. Musa was already just about old enough to succeed to the caliphate if necessary and the oath of allegiance was sworn to him as heir apparent (wali [al-.sup.[subset]]ahd) in 160/776. (11) Dirhams minted at Basra from 164/780-81 onwards and at al-Muhammadiyya (al-Rayy) in 167/783- 84 and 168/784-85 bear the inscription mimma amara bihi Musa wali [ahd.sup.[subset]] al-muslimin. (12) It is not recorded specifically that Musa received the title al-Hadi together with the oath of allegiance in 160/776, but al-Tabari clearly associates this title with Musa's new status. (13) Occasional reports give the impression that al-Mahdi had the oath of allegiance sworn to both Musa and his younger brother Harun at the same time, but this is almost certainly a retrospective telescoping of events. (14) Even so, that it was already al-Mahdi's intention in 160/776-77 to have the oath of a llegiance sworn eventually to Harun is suggested by the appointment in that year of a secretary and vizier to look after Harun's affairs. (15) Also a dirham issued in Ifriqiya in 160/776-77 bears the inscription mimma amara bihi Harun ibn amir almu[contains]minin. (16) This inscription appears regularly on the dirhams of Ifriqiya from 164/780-81 onwards. (17)

In 163/779-80 al-Mahdi appointed Harun to a large governorship referred to as "the whole of the West" (almaghrib kulla-hu), together with Armenia and Azerbaijan. (18) This gave Harun nominal responsibility for the Muslims' three main border regions in the western half of their empire--North Africa, the Byzantine frontier, and the Caucasus. (19) Al-Mahdi was renewing a policy begun by his father towards the end of his caliphate. In 154/770-71 Abu [Ja.sup.[subset]]far had dispatched a large army to restore order in Ifriqiya and decided in the same year to build a new base at al-Raqqa for the war with Byzantium. Construction of the base began in 155/771-72 under the auspices of his son al-Mahdi, who was then his heir apparent. This was a practical demonstration of the new dynasty's commitment to the defense of the western as well as the eastern empire, and Abu [Ja.sup.[subset]]far traveled in person to Jerusalem to inaugurate the new arrangement. (20) Al-Mahdi as caliph did the same in 163/780, after taking Harun to the Byzan tine frontier and sending him across the frontier to raid. (21)

In 165/782 Harun led a second and unusually bold incursion into Byzantine territory, after which both sides agreed to a three-year suspension of hostilities. This was seen by the Muslims as a great success. In 166/782-83 al-Mahdi had his commanders swear the oath of allegiance to Harun "after Musa," and gave him the title al-Rashid. (22) Dirhams issued in Ifriqiya continue to bear the inscription mimma amara bihi Harun ibn amir [al-mu.sup.[contains]]minin, and the same inscription also appears on dirhams issued by the mint of Harunabad/al-Haruniyya on the Byzantine frontier in 168/784-85 and 169/785-86. (23)

The numismatic evidence and some of the textual evidence suggests that, despite the oath of allegiance to Harun, the office and title of wall [al-.sup.[subset]]ahd remained exclusively Musa al-Hadi's. Wali [al-.sup.[subset]]ahd, in other words, remained strictly the one individual who would automatically become caliph in the event of the actual caliph's death. (24) This interpretation of al-Mahdi's policy has not held up well in the Arabic sources. (25) A series of highly tendentious reports in al-Tabari is clearly designed to represent Harun specifically as wall [a1-.sup.[subset]]ahd under al-Mahdi and, moreover, from at least as early as l63/779-80. (26) Other sources commonly make no distinction between Musa and Harun in respect of the wilayat [al-.sup.[subset]]ahd. (27)

Al-Mahdi now turned his attention to the heir apparent, Musa al-Hadi, and to the east. In 167/783-84 Musa was sent to lead a large military expedition against the strategically important and still unsubdued mountain principality of Tabaristan. Substantial reinforcements were sent in the following year, l68/784-85. (28) The importance of this expedition is somewhat obscured by its premature abandonment when al-Mahdi died soon afterwards, but it was in fact a high-profile military response to a serious uprising in which many Muslims had been killed. (29) Al-Mahdi was again following a precedent set by his father Abu [Ja.sup.[subset]]far, who had sent al-Mahdi himself to take charge of a similar invasion of Tabaristan in 141/758-59. (30) Al-Mahdi had stayed on in the east as governor of Khurasan, and Musa might well have done the same if his father's caliphate had not been cut short. (31) As a further distraction from Musa's campaign, the truce that Hakrun had agreed upon with the Byzantines in 165\782 broke down and t he western border war that was still nominally his responsibility began again to intensify. (32)

This was the situation in 169/785 when al-Mahdi died suddenly in Mgsabadhan in western Persia, most probably in a hunting accident. (33) Al-Tabari reports that al-Mahdi was just about to promote Harun as heir apparent ahead of Musa when he died, and adds by way of corroboration another report that al-Mahdi set off for Masabadhan in a great hurry. (34) However, it may be doubted that al-Mahdi at the time shared the reporter's subsequent knowledge of his imminent demise there, and none of the other reported circumstances of his death suggest that he was in a hurry to go anywhere. On the contrary, the sources in general make it clear that he had gone to Mgsabadhan for recreation, and they occasionally say so explicitly. (35)

With Musa far away in the east, it fell to [al-Rabi.sup.[subset]] ibn Yunus, al-Mahdi's mawld whom he had left in charge of Baghdad, to try to obtain the oath of allegiance to the new caliph in the capital. The attempt provoked a riot by the troops, who saw it as an opportunity to demand pay, and order was restored only after an offer of pay was made. (36) It is not certain whether the troops in the capital then actually gave the oath of allegiance. (37) Al-Tabari does say explicitly that envoys were sent to the provinces, where they obtained the oath of allegiance not only to al-Hadi's as caliph but also to Harun as heir apparent (wall [al-.sup.[subset]]ahd). (38) This was probably the first occasion on which Harun was so acknowledged. (39) Harun himself, with the advice of [al-Rabi.sup.[subset]], sent out these envoys, and all of this must have been presented to his brother on his return as a fait accompli. (40)

These events set the scene for the succession crisis of al-Hadi's caliphate. Al-Tabari offers a lengthy account of the crisis in which he draws together much of the available textual evidence. Yet, remarkably, he does not deal with the succession as a topic in its own right. He says nothing at all about it until he comes to the matter of al-Hadi's death, and only there does he finally raise the question of the succession. (41) But now he can also raise the enduringly interesting question of whether al-Hadi was murdered by his mother. The idea that he was so murdered, or even that he might have been, is likely, as we shall see, to predispose the reader to a particular understanding of the succession crisis.

Al-Tabari begins his account of al-Hadi's death with his usual simple statement of the fact. In 170/786-87 Musa al-Hadi died at [Isabadh.sup.[subset]]. How he died is disputed. Some say he died of an internal ulcer, others that his mother, al-Khayzuran, had him murdered. What follows is now represented as an explanation of why his mother wanted to kill him. (42) The explanation takes the form of a compilation of reports, most of which are attributed to a named source. The first of these is attributed to Yahya ibn al-Hasan (ibn [Abd.sup.[subset]] al-Khaliq), and all feature al-Khayzuran, whom they show consistently in a harsh and unfavorable light. (43) When her son Musa became caliph, she assumed that she could exercise his authority. Musa forbade her from interfering in the affairs of state but assured her of his filial obedience in all her legitimate requirements. This was a privilege she exploited to the full, and after her death she was found to have accumulated a vast quantity of clothing. More objectionable to al-Had i was the use she made of her privileged access to him in order to extend her own patronage to her many connections. He would never refuse her a personal request, and she had only to disguise her friends' desires as her own in order to obtain what they wanted. This she did so freely as to generate a considerable traffic of petitioners to her gate and eventually to provoke a confrontation with her son. Mother and son both became enraged, the caliph threatened death to anyone who dared call on her henceforth, and al-Khayzuran stormed out in a blind fury. This first narrative concludes with the categorical statement that she never spoke a word to him again, a point that will be important later. (44)

From here Yahya ibn al-Hasan continues with a report that al-Hadi tried unsuccessfully to poison his mother. On its surface the report is implausible, but it is clearly intended to lend credibility to the allegation that follows. This next report is discreetly attributed to an anonymous Hashemite, presumably an Abbasid and perhaps Muhammad ibn Sulayman ibn [Ali.sup.[subset]], whom Yahya ibn al-Hasan will later quote by name. (45) It says that al-Hadi fell sick and that al-Khayzuran sent her slave women to suffocate him. Her motive was the caliph's determination to remove her other son, Harun, from the succession and to replace him with his own son Ja[subset]far. AlKhayzuran is said to have feared for Harun as he came under pressure from his brother. There is no reference in this report to al-Khayzuran's own quarrel with al-Hadi, but the context in which Yahya ibn al-Hasan has placed it suggests that her concern for Harun was founded on a practical calculation of where her own interests now lay. (46)

Yahya ibn al-Hasan's material ends here for the time being, and al-Tabari continues with a short report attributed to al-Fadl ibn [Sa.sup.[subset]]id. (47) It tells a story similar to Yahya ibn al-Hasan's, though with less of the drama and no obvious hostility towards al-Khayzuran. She did want to control al-Hadi, but he would now allow her to do so. Her patronage of military officers is again an irritant to al-Hadi, but this time he puts a stop to it by confronting the officers rather than al-Khayzuran. Instead of Yahya ibn al-Hasan's lively scene between mother and son, there is now a stylized and entertaining dialogue between the caliph and the assembled officers. No explicit threat is made, but the officers stop calling on al-Khayzuran. This, we are told, made her swear she would never speak to al-Hadi again, and in fact she never did call on him again until he was dying. (48)

Al-Fadl ibn [Sa.sup.[subset]]id's narrative ends here for the time being. On the face of it, his account might seem to corroborate in a general way what has just been seen of Yahya ibn al-Hasan's, although there has been no mention of the succession or the circumstances of al-Hadi's death. Al-Fadi ibn [Sa.sup.[subset]]id's assertion that al-Khayzuran swore she would never speak to al-Hadi again might also seem close enough to Yahya ibn al-Hasan's that she never did. The difference, however, is crucial and amounts, as we shall see, to a specific refutation of the charge of murder. (49)

To introduce his next reports, al-Tabari picks up the reference to the question of the succession at the end of Yahya ibn al-Hasan's previous compilation. For Yahya ibn al-Hasan this issue was clearly secondary to the main theme of al-Khayzuran's quarrel with al-Hadi and the manner of his death. The next series of reports, by contrast, are all concerned with the succession. Only one of them mentions al-Khyzuran at all and none of them mentions her quarrel with al-Hadi or the manner of his death. The central figure in every one of them is the Barmakid Yahya ibn Khalid. Like Yahya ibn al-Hasan's compilation, this present series is skillfully strung together in an ordered progression leading to a dramatic climax. It amounts to the Barmakid version of the crisis. [50]

The first report explains why Yahya ibn Khalid was involved in the succession in the first place. Al-Hadi on his accession is said to have confirmed Yahya's responsibility for Harun's governorship of the West, and the caliph's apparent confidence in Yahya is an important theme of what follows. (51) Yahya's troubles began not so much with the caliph as with the army. Al-Hadi did want to remove Harun from the succession to make way for his son [Ja.sup.[subset]]far, but the pace was set by a number of leading army commanders. On their own initiative they actually renounced Harun, gave their oath of allegiance to [Ja.sup.[subset]]far, and began agitating against Harun within the military establishment. Al-Hadi gave a limited measure of symbolic endorsement to their campaign and, apart from Yahya, Harun soon found himself friendless at court. (52)

Harun himself hardly appears in these reports. The problem for both the caliph and the army in disposing of Harun was entirely a matter of dealing with Yahya and he, rather than Harun, became the object of hostile intrigue. It was whispered to the caliph that only Yahya stood in his way and that he should threaten to kill him. When in the next report Yahya is summoned to the caliph in the night, it is no surprise to the reader that he prepares himself for the worst. (53) Yet what this and the next three reports go on to show is that Yahya with his personal qualities, political skill, and powers of persuasion, was more than a match for his enemies. In Yahya's first, nocturnal interview with al-Hadi, he persuades the caliph that his only ambition is to serve him in the office to which he himself has appointed him, and he denies encouraging Harun to resist him. We are told, however, that Yahya was doing precisely that, Harun being completely infatuated with his wife Umm [Ja.sup.[subset]]far and indifferent to his polit ical future. Yahya, it seems, pursued his lonely and dangerous course without any help or encouragement from even the chief beneficiary of his efforts. (54)

A second summons in the night is initially as alarming as the first, but again turns out well for Yahya. (55) Then a rather petty attempt by the caliph to humiliate Yaya in public has an even more surprising outcome. To the amazement of the assembled court, al-Hadi graciously and publicly apologizes to Yahya, who duly responds at his most charmingly deferential. More substantial than these pleasantries is the political lesson that follows. The occasion is not now specified, but al-Hadi and Yahya are said to have discussed seriously the whole question of the succession. As might be expected against a background of military indiscipline of which he is the likely victim, Yahya's main concern is with controlling the army. The caliph's only hold over the soldiers is their oath and he should not encourage them to think they could break it. Since they had sworn allegiance to Harun the caliph should hold them to it, but he could still make them swear allegiance to [Ja.sup.[subset]]far as well. In the long run, that would be safer for [Ja.sup.[subset]]far too, a point that al-Hadi now gratefully acknowledges. The intended lesson for the reader is that a caliph can control his army if only he will listen to his vizier. (56)

Al-Hadi's gratitude was evidently short-lived. In the next report he imprisons Yahya to force him to agree to what he wants, providing for the reader yet further evidence of Yahya's remarkable ability to hold things up. Even from jail he persuades al-Hadi to give him a private audience and this time raises the question of his son [Ja.sup.[subset]]far's minority. If, God forbid, al-Hadi were to die before [Ja.sup.[subset]]far came of age, the army would refuse to have a child as their caliph. Pretenders would spring up from within, or even without, the Abbasid family and the succession in al-Mahdi's line would be lost. Once again al-Hadi is grateful for the warning.

Yahya presses home his advantage. His warning of a threat to al-Mahdi's line is the key to his argument. Given the danger of not having an heir apparent who could actually assume the caliphate if necessary, al-Hadi would have been well advised himself to nominate his brother Harun. Since his father had already done so for him, it was now unthinkable to undo this provision. Instead he should simply wait for [Ja.sup.[subset]]far to come of age, by which time Harun would gladly resign his position in favor of his nephew. This was enough to persuade al-Hadi, and Yahya was released from prison. (57)

What Yahya clearly wants is a postponement. (58) This idea is now firmly planted in the mind of the reader, who knows that al-Hadi is going to die very soon. The mood of the reports now changes. Hope gives way to fear and the narrative thread leads into a deepening crisis. With the encouragement of his mawali as well as his generals, al-Hadi soon forgot Yahya's advice and began to put pressure on his brother. Yahya, again by way of postponement, advised Harun to leave town on a pretext and to delay his return as long as possible. He did so, but his failure to return on al-Hadi's repeated instruction proved dangerous to his cause. The caliph made his displeasure known and those hostile to Harun were encouraged to speak out openly against him. They were already urging al-Hadi to change the succession without Harun's consent if necessary. Harun was kept informed of all this and decided to return. It is clear enough, though not said explicitly, that this was to prevent his immediate removal from the succession. A lso implied, though more subtly, is that his return actually did prevent it for the time being. (59)

The sense of imminent danger is considerably heightened by the next report, in which al-Khayzuran makes her first appearance in the Barmakid version. No longer the power-hungry and ruthless matriarch of Yahya ibn al-Hasan's version, she is here the distraught mother whose only concern is for her younger son. She accuses Yahya of risking her son's life and begs him to let Harun give in to his brother. Behind the emotional charge of this report lie now familiar themes. So far as the succession is concerned, everything still depends on Yahya. Not only Harun but also his mother would have given in to al-Hadi. Yahya was entirely alone. For his selfless and dangerous devotion to duty, his only reward is a desperate mother's reproach. (60) The crisis is now at its height. Inducements having failed to shift Yahya, al-Hadi now threatens him with death. Yahya is with Harun constantly, day and night. As if incidentally, almost the last thing we are told at this late stage is that Harun was still heir apparent (wall [al-.sup.[subset]]ahd). (61)

Everything is now in place for the resolution of the crisis, but first al-Tabari affords the reader a measure of relief. The single long report that follows is entirely self-sufficient, with no relation to any of the other reports in al-Tabari's compilation. It does not mention al-Khayzuran, or Yahya ibn Khalid, or any attempt by al-Hadi to change the succession. The story is that one day, early in his caliphate, al-Hadi received his brother coldly, and in the presence of others told him that, despite the dream they both knew about, he had no chance of becoming caliph. Harun replies with courage and dignity, warning his brother that the way to true greatness is through humility and not tyranny, and admitting that he did hope to be caliph one day. He hoped above all to restore justice to those whom his brother had wronged and yet still do his duty to their father by honoring his brother's sons above his own, and marrying them to his daughters.

This impressive performance immediately dispels the fit of petty jealousy that had seized al-Hadi, He now insists that his brother take the place of honor in his presence and makes him a huge grant of funds from the treasury. By now, of course, the reader wants to know about the dream, and the story proceeds to tell him. Al-Mahdi had once dreamed that he gave a staff to each of his sons, Musa and Harun. The top of Musa's staff had put out a few leaves, but Harun's had put out leaves from end to end. A perspicacious interpreter of dreams had predicted at the time that both brothers would rule, but al-Hadi's reign would be short, while Harun's would be a long and golden age. And so it was. A few days later al-Hadi fell ill and died. Harun became caliph and his was indeed a golden age. Whatever little unpleasantness there may have been over the succession in al-Hadi's short caliphate, the reader can be sure that all turned out well in the end. (62)

Al-Tabari now offers the isolated snippet of information that al-Hadi went out to al-Haditha near Mosul, fell seriously ill there, and returned. (63) What follows next is a rapid and skillful literary resolution of the succession crisis as al-Tabari has built it up. Three short reports wrap up quickly, neatly, and in reverse order all three main versions of events that al-Tabari has until now left unfinished. The first and most urgent tells how Yahya ibn Khalid was saved at the last minute from the imminent threat of death. The second, by al-Fadl ibn [Sa.sup.[subset]]id, tells his version of the end of al-Hadi's and al-Khayzuran's quarrel. The third and last, which Yahya ibn al-Hasan now attributes to the Abbasid Muhammad ibn Sulayman ibn [Ali.sup.[subset]], is a sinister conclusion to the tale of murder with which al-Tabari's whole compilation began.

The first of these three concluding reports refers to al-Hadi's return from al-Haditha, which is why al-Tabari has previously provided the information that he went there. (64) The report adds that al-Hadi had also written to all the provincial governors and ordered them to present themselves. He was now gravely ill, and the soldiers who had already sworn their allegiance to his son [Ja.sup.[subset]]far found themselves in a dilemma. Al-Hadi had still not given in wholly to their pressure by formally removing Harun from the succession. If, as this report makes the soldiers say amongst themselves, Yahya ibn Khalid were to come to power, he would kill them. They thought of killing Yahya on the pretence that al-Hadi had ordered them to do so, but had to face the dangerous possibility that the caliph might recover and hold them to account.

At this point Yahya gains a crucial advantage over his military opponents. Al-Khayzuran, who here makes a second intervention in the Barmakid version, let Yahya know that al-Hadi was definitely dying. With time to prepare, Yahya was able as soon as the latter did die to send out letters to the provincial governors in al-Rashid's name, confirming them all in their offices. This was clearly in order to cancel the instructions that at-Hadi had just sent them to present themselves. What is also certainly to be inferred is that al-Hadi;s unusual general summons to the governors had been in order to announce a change in the succession. (65) But Yahya's policy of postponement had worked and al-Hadi had left it till too late. Yahya's bureaucratic intervention at the very end was not in any sense a coup d'etat, but simply a confirmation of Harun's legitimate succession to the caliphate. (66)

Next to finish is al-Fadl ibn [Sa.sup.[subset]]id He reminds us that al-Khayzuran had sworn she would never speak to her son again and would not go near him. From this crucial point of apparent agreement with Yahya ibn al-Hasan, he goes on, in effect, to deny the charge of murder. (67) When first told that al-Hadi was dying, al-Khayzuran was cold and indifferent, but she relented and made ready to go and see him. With mother and son now reconciled, the story ends with another fulfilled prediction, this time based on the slightly interesting coincidence that Harun's son [Abd.sup.[subset]] Allah, who himself would be caliph one day, was born on the night of his father's accession to the office. (68)

Yahya ibn al-Hasan has the last word. The end of his account has al-Khayzuran at [Isabadh.sup.[subset]] in the company of four Abbasid matrons. (69) Her maid comes in and al-Khayzuran asks as casually as she can if anything has happened. The maid answers that the caliph has died and been buried. The reply, "If Musa's dead, that leaves Harun" is the extent of this mother's grief. Entirely unconcerned that she has missed seeing her dying son, she calls for drinks for herself and her four companions and presents each of them with a large gift of money. (70) With their silence bought, she sets off to join Harun in Baghdad. (71)

This whole compilation of reports has been put together with considerable editorial skill and deliberation. The result is an outstanding literary success, a vivid, exciting and credible drama, a realistic account of human players, their motives, and interactions in the field of high politics. However, it is unlikely that al-Tabari saw himself as writing historical fiction and the question arises what, if any, historical conclusions this compilation was intended to promote. Two simple observations make a starting point. Firstly, al-Tabari says nothing at all about the succession to al-Hadi until he comes to the matter of al-Hadi's death, even though the succession crisis as he represents it must have been the dominant political issue of al-Hadi's short caliphate. Secondly, most of the material he brings in under the heading of al-Hadi's death is actually about the succession. The issue of how al-Hadi died, which al-Tabari represents as the matter of dispute, is a red herring. True, there is a story that the ou tcome of al-Hadi's final illness was not left to nature and that his mother had him finished off. But this story depends on one source clearly hostile to al-Khayzuran and on the somewhat improbable notion that the dying caliph, alone and unprotected, could be smothered by his mother's womenservants without anyone raising the alarm. It is hard to see a real contest between this claim and the alternative that al-Hadi simply died of an internal abscess or ulcer. So why does al-Tabari pretend to take it seriously?

Most of what al-Tabari has to say about the succession is in the account of Yahya ibn Khalid's struggle with al-Hadi and with the army. All the reports in this Barmakid version place Yahya himself at the very center of the prolonged succession crisis. The reader is left in no doubt that it was Yahya who almost single-handedly saved Harun's caliphate. For the Barmakids' admirers that is undoubtedly the main point. For al-Tabari, what is probably important is not who saved Harun's caliphate but how he saved it: by postponing Harun's deposition from the office of heir apparent until, for al-Hadi and the army, it was too late. Apart from Yahya's personal triumph, what the Barmakid version of the crisis proves is that the succession to al-Hadi was legitimate and not, as [Ja.sup.[subset]]far's would have been, the result of a dangerous agitation by a turbulent soldiery. In the context of the Abbasid caliphate, this is a point of real historical importance.

So far as the matter of timing is concerned, this point is reinforced by the allegation that al-Hadi was murdered. His death was not an accident but the result of a deliberate intervention to prevent him from changing the succession. This is the key to al-Tabari's compilation, the reason why he says nothing about the succession before the matter of al-Hadi's death, and why he then surrounds his main account of the succession crisis with Yahya ibn al-Hasan's tale of murder. However, there is a danger in this scheme. In the story of Yahya ibn Khalid's successful struggle, there is no suggestion that al-Hadi's death was in any way suspicious. But Yahya does play a minor part in the story of al-Hadi's murder and al-Khayzuran is instrumental in the story of Yahya's triumph. As long as these two stories are kept apart, there is nothing in either to cast any suspicion on Yahya, but once put together they almost inevitably add up to a conspiracy. (72) If the whole point is to demonstrate the legitimacy of Harun's suc cession, this is obviously counterproductive. It is presumably for this reason that al-Tabari has not, in fact, immediately framed Yahya's story with Yahya ibn al-Hasan's, but has first enclosed it in the two parts of al-Fad ibn [Sa.sup.[subset]]ids. This ultimately harmless account of al-Hadi's and al-Khayzuran's quarrel forms an insulating layer between the two important components of al-Tabari's compilation, a barrier to suspicion that Harun's accession to the caliphate was due to anything more sinister than Yahya's political skill and the guiding hand of providence.

What al-Tabari has presented so far is a careful selection from the material available to him on the succession crisis and the death of al-Hadi. His skillful use of this material creates the strong impression of a single coherent narrative of events, but in fact the picture is a good deal less tidy than this narrative suggests. Al-Tabari himself knew other material that fits badly or not at all into this picture and we shall see that he has postponed using this material until both the succession crisis and al-Hadi have been laid to rest. Even within the material he has used so far, there is some uncertainty on apparently simple matters of circumstantial fact, in particular where Harun and Yahya ibn Khalid actually were at the crucial moment. The Barmakid version says explicitly that Harun and Yahya were both staying at the Khuld palace in Baghdad. (73) Both appear to have been at liberty on the night of al-Hadi's death, in the case of Yahya necessarily so since he had to organize secretaries to write letters to the provincial governors in time for the moment of the caliph's death. (74) Yahya ibn al-Hasan's version also has Harun and Yahya apparently at liberty, but at [Isabadh.sup.[subset]], from where al-Khayzuran follows Harun to Baghdad as soon as the caliphate is his. (75) This contradiction is detectable but not obvious in al-Tabari's compilation concerning al-Hadi's death. It becomes obvious later, when al-Tabari discloses a remaining part of Yahya ibn al-Hasan's narrative. (76)

Other reports that al-Tabari does not use at all only add to this uncertainty. According to al-Jahshiyari, after al-Hadi fell ill he again summoned Yahya in the night and declared that now he was going to kill him. He was persuaded to wait until the morning and Yahya was imprisoned to await his fate. This was presumably at [Isabadh.sup.[subset]], though it is not said so explicitly. Yahya himself is made to tell the story of his prison vigil and of what he took to be the inevitable summons to his death when his cell door was opened. But in fact it was a summons to appear before al-Khayzuran and hear that al-Hadi was dead. Having seen the body, Yahya then made his way to the Khuld palace in Baghdad to wake Harun and to send letters out to the provinces with the news." [Al-Ya.sup.[subset]]qubi has basically the same story, with one particular difference. Al-Hadi was threatening to kill Harun as well as Yahya and they were both in prison. Yahya again tells his story. This time his imprisonment lasted for several days before h is door was opened in the night. As before, the summons was not to his death but to appear before al-Khayzuran and learn that, al-Hadi was dead. Instead of waking Harun at the Khuld palace, Yahya now has to release him from prison before he can assume the caliphate. [Al-Ya.sup.[subset]]qubi does not say where any of this happened, though he does mention later that al-Hadi was buried at [Isabadh.sup.[subset]] and that Harun performed the prayer at his funeral. (78)

Clearly both these versions are based on the same story, attributed to Yahya ibn Khalid himself, of relief after adversity. The difference in circumstantial detail is of no real importance to the story. For al-Tabari, neither version would have added much to his account of Yahya's triumph, whereas either would have spoiled his neat denouement in which Yahya is free in the night to cancel al-Hadi's summons to the provincial governors. But both have in common with al-Tabari's Barmakid material the same unquestioned assumption that, short of killing him, Yahya ibn Khalid was an insuperable barrier to al-Hadi's plans for the succession. We shall see, however, that this constantly reiterated theme of Yahya's central and solitary role in these events is open to question.

Al-[Ya.sup.[subset]]qubi reports that, when al-Hadi first floated the idea of replacing Harun with [Ja.sup.[subset]]far as heir apparent, the majority of the army commanders advised him against it. A certain number, however, supported and encouraged the idea and expressed their opposition to Harun. One of these was Abu Hurayra Muhammad ibn Farrukh a veteran of the revolution who had previously been instrumental in removing [Isa.sup.[subset]] ibn Musa from the succession to al-Mahdi in al-Hadi's favor. (79) This officer was sent off with a numerous army to the provinces of al-Jazira, Syria, Egypt, and Ifriqiya to secure assent from the men based there to the removal of Harun, by use of armed force if necessary. As it turned out, he got only as far as al-Raqqa before he heard the news of al-Hadi's death. (80) Al-Tabari has none of this, although he does mention without any explanation that Harun as caliph had Abu Hurayra brought back to Baghdad and executed in 171/787-88. (81)

At this point al-[Ya.sup.[subset]]qubi tells the story of Yahya ibn Khalid's arrest and imprisonment and it would not in fact be too difficult to harmonize his version of the succession crisis with al-Tabari's Barmakid version. The Barmakid version acknowledges no support for Harun from any part of the army, but it could be argued that the majority had been intimidated into silence by the minority who, after all, had the caliph on their side. The provinces to which Abu Hurayra was sent were all part of the great northern and western governorship to which Harun had been appointed by his father al-Mahdi in 163/779-80. (82) According to al-Tabari's Barmakid version, al-Hadi on succeeding his father had confirmed Yahya in this position. (83) It is possible therefore, as Bonner has argued, that these provinces constituted a military power base for Yahya ibn Khalid, which al-Hadi in his struggle with Yahya then had to destroy. (84) But if we accept this argument, it is easy to lose sight of the fact that al-[Ya.sup.[subset]]qubi 's account of the succession crisis and of Abu Hurayra's mission says nothing of Yahya ibn Khalid apart from a self-contained report that is itself a common literary type. Al-[Ya.sup.[subset]]qubi evidence is that a significant part of the army in the capital and in the western provinces supported Harun, not Yahya. To take this as support for Yahya is to read the Barmakid version even into evidence that goes against it. Where al-[Ya.sup.[subset]]qubi does agree with the Barmakid version is in his contention that al-Hadi's attempt to replace Harun with his son [Ja.sup.[subset]]far as wall [al-.sup.[subset]]ahd was ultimately unsuccessful. His use of the story of Yahya's imprisonment is to reinforce precisely that point.

The suggestion that Harun and Yahya were both in prison and awaiting execution at the time of al-Hadi's death also appears in al-Tabari, but only when he comes eventually to the matter of Harun's accession. According to his first report in this connection, it was not Yahya ibn Khalid but the soldier Harthama ibn [A.sup.[subset]]yan who released Harun in the night when al-Hadi died and it was Harun who then released Yahya. (85) Yahya was immediately appointed to the vizierate and gave instructions to a secretary to draw up letters. Here, however, it is not said what the purpose of these letters was. (86)

Despite the unexplained appearance of Harthama in this report, it preserves what, for al-Tabari, is the main point of the Barmakid version. Harun and Yahya can only have been in jail because al-Hadi had failed to get from them what he wanted, which was Harun's removal from the succession. The same is true of a long and rather improbable account of Harthama's involvement in the succession, in the anonymous [al-.sup.[subset]]Uyun wa-'l-[hada.sup.[contains]]iq. According to this account, Harthama was summoned by al-Hadi and ordered to kill Harun, though it was still Yahya ibn Khalid against whom the caliph's anger was chiefly aroused. Harthama was appalled and tried to dissuade the caliph, in the process apparently innocently reminding the reader that Harun was still heir apparent at this late stage. For his hesitation in obeying the caliph, Harthama soon found himself awaiting execution. When the expected summons came, he like Yahya discovered to his great relief that it was not to his death that he was called but to al-Khay zuran and a dead caliph. From there he went off to inform Yahya ibn Khalid and to have the oath of allegiance sworn to Harun, neither of whom seems to have been in jail at the time. (87)

It is hard to see this account as much more than a crude adaptation of the Barmakid version. The change of protagonist might be due simply to Harthama's undoubted prominence in the latter part of al-Rashid's caliphate. We shall see later another and radically different version of events in which Harthama plays the leading part and to which this present version could even be a counter. (88) But to return to al-Tabari, the only reason for his report that happens to mention Harthama is to introduce the text, which he now gives in full, of an address that is said to have been made to the army commanders on the morning after al-Hadi died. This text will be discussed later. (89)

Next in al-Tabari is the last part of Yahya ibn al-Hasan's narrative. Here, at [Isabadh.sup.[subset]] as is now made clear, Yahya wakes up Harun in the night with the news of his brother's death and his own accession to the caliphate. Yahya ibn al-Hasan's Harun is confident and imperious in his new role as caliph, in strong contrast to the feeble and infatuated youth of the Barmakid version. One of his first acts is to execute an unfortunate Abu [Isma.sup.[subset]], who is said to have insulted Harun when al-Hadi was caliph. In a stereotyped situation, he had forced Harun to give way at a narrow bridge to al-Hadi's young son [far.sup.[subset]], describing [Ja.sup.[subset]]far as he did so as "the heir apparent." (90)

This is the first such reference in al-Tabari to [Ja.sup.[subset]]far ibn al-Hadi. By itself it does not prove anything. We have been told before that several army commanders had sworn allegiance to [Ja.sup.[subset]]far without al-Hadi's explicit authorization and it is entirely possible that Abu [Isma.sup.[subset]] was one of them. The ambiguity of this evidence is repeated in the conclusion to the narrative. Al-Mahdi had given Harun a ring, which al-Hadi had demanded from him when he became caliph. Rather than give it up, Harun had thrown it into the river. Now, as caliph after all, he ordered divers to look for it, and of course they found it. (91) Had he lost the ring and recovered it against all odds, or had it, in the providential scheme of things, never really been lost?

These is no ambiguity about what comes next. Al-Tabari's last report on Harun's accession states baldly that al-Hadi as caliph had stripped his brother of his right to the succession. He had then had the oath of allegiance sworn to his own son [Ja.sup.[subset]]far. When al-Hadi subsequently died, it was a senior army commander, Khuzayma ibn Khazim, who took the initiative. Backed by five thousand of his own armed mawali, he seized [Ja.sup.[subset]]far from his bed in the night and forced him to renounce his title. When the rest of the army arrived in the morning, they found [Ja.sup.[subset]]far waiting to tell them he had abdicated in favor of his uncle, Harun. This fait accompli was apparently accepted without protest, though one senior officer, [Abd.sup.[subset]] Allah ibn Malik, found it prudent to walk to Mecca to absolve himself of his oath of allegiance to al-Hadi's young son. (92)

After all al-Tabari's previous efforts to prove otherwise, it is startling to be told in the end that al-Hadi had removed Harun from the succession and that Harun's caliphate was due to a straightforward military coup. But al-Tabari has indeed shifted gradually, almost imperceptibly, from the solid Barmakid line that Harun remained heir apparent until the moment of al-Hadi's death, through Yahya ibn al-Hasan's less committed position that he may have done so but that it made no difference in the end, to this final revelation that he was definitely not heir apparent when his brother died but that it still made no real difference. Al-Tabari will, it seems, eventually report inconvenient information, but only when the reader has been fortified against the undesirable implications of the story as it emerges.

The anonymous author of [al-.sup.[subset]]Uyun wa'l-hada iq was not prepared to take the same risk with his readers. His account of al-Rashid's accession is a summary of al-Tabari's, except that he carefully smoothes away the potentially disruptive elements. In the story of Harun and Abu Isma, he adds a gloss to the effect that [Ja.sup.[subset]]far, whom Abu [Isma.sup.[subset]] calls wali [al-.sup.[subset]]ahd, was the son whom al-Hadi wanted to make his heir apparent. In his summary of Khuzayama ibn Khazim's intervention, he omits to mention that al-Hadi had removed Harun from the status of heir apparent and explains, with another gloss, the need to force [Ja.sup.[subset]]far to renounce it, this time to the effect that al-Hadi had ordered an unspecified group of people (a [jama.sup.[subset]]a) to swear allegiance to him. This comes almost certainly from the Barmakid version, in which, as we have seen, a group of officers had sworn an unauthorized oath of allegiance to [Ja.sup.[subset]]far. The message is the same, that al-Hadi came close to formally chang ing the succession, but never quite did so. (93)

As only Abbott seems previously to have noticed, Ibn Abi [Usaybi.sup.[subset]]a has a long description of an occasion on which al-Hadi apparently did change the succession. The caliph is said to have summoned the whole of the court to renounce their allegiance to his brother Harun as heir apparent and to swear allegiance to his son [Ja.sup.[subset]]far. All did so, except for the now not unexpected figure of Harthama ibn [A.sup.[subset]]yan. Insisting that he had only the caliph's best interests at heart, he warned him boldly that men who had just broken their oath of allegiance to Harun could as easily break their oath to [Ja.sup.[subset]]far. This unwelcome intervention infuriated al-Hadi, but Harthama was unshakable. In the end al-Hadi admitted that Harthama was right and denounced as traitors those who had just cast off their allegiance to Harun. As the assembly broke up, many now realized that what they had just done had been done too soon, and they saw the danger of what might happen if al-Hadi died. This of course he did soon af terwards. (94)

As the lonely hero of this account, Harthama seems once again to have taken on the part of Yahya ibn Khalid. His warning to al-Hadi not to encourage the breaking of oaths is the same as Yahya's, as are his selfless concern for the interests of the state, his courage and determination in the face of danger, and his remarkable justification in the presence of the caliph and the assembled court. What is clearly missing from this account is the Barmakid version's careful insistence that al-Hadi never did change the succession and that Harun's eventual accession was a victory for Yahya, on the one side, over the army on the other. As an account of Harthama's heroism it is no more inherently credible than the Barmakid version of Yahyas, but it highlights the intimate and necessary connection in the Barmakid version between the crucial delaying role claimed for Yahya in saving al-Rashid's caliphate and its insistence that al-Hadi failed to act decisively until it was too late. There is no attempt here in Ibn Abi Usa [ybi.sup.[subset]]a to connect this same idea of successful postponement with Harthama.

In all of this textual evidence there is clearly no doubt that al-Hadi would have made his son [Ja.sup.[subset]]far his heir apparent if he had lived longer, and nearly all of it agrees that he at least began the process of doing so. Where the evidence disagrees is on whether that process was completed before al-Hadi died. There is clearly a strong textual tradition that it was not, the foundation of which view is the Barmakid version. But the very artfulness of that version is what tends to undermine its credibility. Its key component is that al-Hadi did not decide to act until he knew he was dying. But why, in that case, would al-Hadi make such a last-minute change in favor of a minor son, a son whom he knew to be incapable of assuming the caliphate in the immediate future? If al-Hadi really did take steps to change the succession, it is surely more likely that he did so earlier, when he himself was in good health, and in the natural expectation for a man in his early twenties that he would live to see his young s on come of age. He would certainly have been taking a risk, but not making an obviously futile gesture.

Al-Hadi may have had occasion to make an early change. It has been seen that the oath of allegiance to Harun as heir apparent was probably first sworn on al-Hadi's accession to the caliphate, but without his brother's knowledge or involvement. Apart from Harun himself, the key figure involved in procuring this oath of allegiance was [al-Rabi.sup.[subset]], on whose advice Harun sent envoys out to the provinces for this purpose. (95) AI-Tabari reports that for this reason in particular al-Hadi was extremely displeased with [al-Rabi.sup.[subset]], but when he returned to the capital he forgave him and appointed him vizier. (96) It seems that [al-Rabi.sup.[subset]], rather than Yahya ibn Khalid, may have initially persuaded al-Hadi that making Harun his heir apparent was the right thing to do under the circumstances. But within a year [al-Rabi.sup.[subset]] was dead. Al-Tabari reports specifically that al-Hadi did not attend his funeral, and that Harun performed the prayer. Harun, he points out, was still heir apparent at the time. (97) If al-Hadi still harbored resentment and suspicion at the choice by others of his heir apparent, this was perhaps the moment when he decided to do something about it.

In fact, as we have seen, it remains doubtful whether the oath of allegiance to Harun as heir apparent had ever been sworn in the capital. (98) The situation that al-[Rabi.sup.[subset]] persuaded al-Hadi to accept on his return may have been no more than an uncertain status for Harun, heir apparent so far as the provinces knew, but not acknowledged by anyone in Baghdad. The insistence of the Barmakid version that al-Hadi never removed Harun from the office of wall [al-.sup.[subset]]ahd may simply disguise the fact that, so far as the army were concerned, Harun had never been wall [al-.sup.[subset]]ahd. They were free in any case to recognize [Ja.sup.[subset]]far as heir apparent with the caliph's blessing. All that was left unfinished before al-Hadi died was Abu Hurayra's military mission to the western provinces, to enforce there what had already been settled in the capital.

The coins of al-Hadi's caliphate offer some support for this interpretation of the texts. Dirhams issued in Ifriqiya in both 169/785-86 and 170/786-87 have the inscription mimma amara bihi Harun wall [ahd.sup.[subset]] al-muslimin. (99) This confirms that the oath of allegiance must have been sworn to Harun as heir apparent in at least one western province. The fact that only the dirhams of Ifriqiya carry this inscription suggests that distance from the capital and possible political support for Harun in the province may account for this plain acknowledgement of his status as wall [al-.sup.[subset]]ahd in both years of al-Hadi's caliphate. (100) These coins do not necessarily tell us what was happening in Baghdad.

There is also a dirham issued by the North African mint of [al-.sup.[subset]]Abbasiyya in 170/786-87 with the inscription mimma amara bihi Harun ibn amir al-[mu.sup.[contains]]minin. (101) The same inscription appears on dirhams issued by the mint of al-Haruniyya on the Byzantine frontier in both years of al-Hadi's caliphate. (102) The formula Harun On ibn amir al-[mu.sup.[contains]]minin was used consistently to refer to Harun on the dirhams of al-Mabdi's caliphate, while his brother Musa had been referred to equally consistently as Musa wall [ahd.sup.[subset]] al-muslimin. (103) With al-Hadi as caliph, the formula Harun ibn amir al-[mu.sup.[contains]]minin is now curiously anachronistic and seems to require some explanation. On the one hand, it fails to recognize Harun as wall [ahd.sup.[subset]] al-muslimin. On the other hand, it suggests that Harun still had a potential claim to the succession and a claim moreover derived from his late father, al-Mahdi, rather than his brother the caliph, al-Hadi. Jere Bacharach has argued against this interpretat ion and has suggested that of itself the formula (fulan) ibn amir al-[mu.sup.[contains]]minin has no necessary relevance to the succession. (104) He uses two examples to make his point, the first of which is a dirham minted at al-Muhammadiyya in 171/787-88 and bearing the inscription mimma amara bihi Muhammad ibn amir al-[mu.sup.[contains]]minin. At this point in al-Rashid's caliphate, as Bacharach points out, the future Muhammad al-Amin was just one year old and hardly a credible successor to his father. But Muhammad was entrusted at a very early age to the care of a senior military officer, [Ja.sup.[subset]]far ibn Muhammad ibn al-[Ash.sup.[subset]]ath, who in 171/787-88 was appointed governor of Khurasan. (105) This connection with the governorship of Khurasan suggests that al-Rashid was already preparing for his son's eventual nomination as his successor. (106) The second example is a dirham minted in Armenia in 172/788-89 and bearing the inscription mimma amara bihi [Ubayd.sup.[subset]] Allah ibn amir al-[mu.sup.[contains]]minin. Bacharach argue s that [Ubayd.sup.[subset]] Allah was simply al-Rashid's brother who happened to serve as governor of Armenia and that the inscription has again no bearing on the succession. But according to the Armenian historian Lewond, [Ubayd.sup.[subset]] Allah was a rival to Harun and was given Armenia as part of a territorial division between the two. (107) What this coin suggests is that [Ubayd.sup.[subset]] Allah may have attempted to use Armenia as a territorial base from which to challenge his brother, claiming a title for himself from their father with exactly the same numismatic anachronism as we see being used for Harun during the caliphate of his brother Musa. In fact it seems that [Ubayd.sup.[subset]] Allah went further than that, as Bacharach notes, since his coin does not even acknowledge his brother Harun as caliph.

In the case of Harun and his brother al-Hadi, the use of the anachronistic mimma amara bihi Harun ibn amir al-[mu.sup.[contains]]minin suggests a similar assertion of a potential title to the succession. Its use both in Ifriqiya and at al-Haruniyya might have been a precautionary response to al-Hadi's presumed or known unwillingness to acknowledge Harun as his heir apparent, or a more positive if still cautious assertion of Harun's title in response to attempts by al-Hadi or others to suppress it. The point of the anachronism is that Harun held his title from his father, just as al-Hadi held his. (108)

Of the two dirham issues of al-Haruniyya naming al-Hadi as caliph and bearing the legend mimma amara bihi Harun ibn amir al-[mu.sup.[contains]]minin, that of 169/785-86 also names Khuzayma ibn Khazim, presumably as the governor of Armenia with its associated mint at al-Haruniyya. As Bonner has pointed out, this fits in well with the textual evidence of Khuzayma's support for Harun in the capital on his brother's death. (109) There are four other dirham issues of al-Haruniyya naming al-Hadi as caliph, two from 169/785-86 and two from 170/786-87, but without any mention of Harun. Two of these, one from each year, have instead the formula mimma amara Yazid ibn Mazyad. (110) Yazid had now evidently replaced Khuzayma ibn Khazim as governor and, as Bonner points out, it is unusual for a governor's name to appear in this formula on a dirham, where it is usually reserved for the caliph and his heir or potential heir apparent. Bonner suggests that the new governor had now deliberately removed Harun's name from the coinage. T his also fits in with textual evidence of Yazid's support for removing Harun from the succession and leads to Bonner's conclusion that Harun was in fact removed around the turn of the year. (111) But against this clear-cut conclusion is one of the other dirham issues of al-Haruniyya mentioned above, which in 170/786-87 still has the formula mimma amara bihi Harun ibn amir al-[mu.sup.[contains]]minin, and without naming Khuzayma ibn Khazim as governor. (112) The picture from the coinage of Ifriqiya and al-Haruniyya is one of uncertainty, with evidence of conflict over the succession but no clear outcome. A coin naming [Ja.sup.[subset]]far ibn al-Hadi as wali [ahd.sup.[subset]] al-muslimin would settle the manner, but, so far as I know, none has appeared. (113) The coins that we do have fit well with the evidence of Abu Hurayra's unfinished mission to the western provinces. [114]

The most striking feature of the coinage of al-Haruniyya, and the main focus of Bonner's important contribution, is the appearance on three issues of 170/786-87 and 171/787-88 of the new caliphal designation alkhalifa al-mardi, combined on one issue of 170/786-87 with the formula mimma amara bihi Harun amir [almu.sup.[contains]]minin. (115) This designation of Harun as al-khalifa al-mardi is not known from textual sources and disappears from the coinage after 171/787-88. Bonner has connected the title with the old revolutionary slogan of the Abbasids' Khurasani partisans, al-rida min al Muhammad. The importance of this connection is clearer in the light of Crone's subsequent study of this same slogan." (116) She argues that the point of the call to al-rida was not, as the sources claim and is still often believed, to disguise the identity of a specific candidate whom the revolutionaries had already secretly adopted, but to invoke the principle of communal choice or shura. (117) The rider min al Muhammad was in that case not a legitimist slogan in the sense that it asserted the right to rule of a particular candidate entitled by descent, but only insofar as it excluded the Umayyads as non-Hashemites. Which particular Hashemite should replace them was a matter for the revolutionaries themselves to decide." (118)

For our present purpose, it is important to note that shura or rida as a legitimizing slogan is a two-edged weapon. On the one hand, the call for shura was used by rebels and opponents of the Umayyads as a counter to Umayyad dynastic pretensions." (119) On the other hand, the shura that had actually taken place on the death of [Umar.sup.[subset]] ibn al-Khattab was invoked by the Umayyads themselves as the basis of their dynastic legitimacy. (120) So far as they were concerned, the acknowledged leaders of the community had long ago weighed the respective merits of [Ali.sup.[subset]] and [Uthman.sup.[subset]], and had chosen the Umayyad. The Khurasani call to al-rida was a call to re-open that early shura, and the rider min al Muhammad a guarantee that this time the Umayyads would not reap the benefit.

Once installed as caliphs by their Khurasani partisans, the Abbasids were just as able as the Umayyads to appropriate the slogan of al-rida, and transform it from a revolutionary call for election to a conservative claim of institutional legitimacy. (121) When Abu [Ja.sup.[subset]]far al-Mansur sent the Khurasani Abu Humayd to cajole-and if necessary intimidate-Abu Muslim into surrendering to him, it was first as a fellow revolutionary that Abu Humayd tried to appeal to his former chief: "Abu Muslim, you are still amin al Muhammad, the troops recognize that." A moment later the tone has changed and Abu Humayd's talk is all of the miraculous unifying effect on the previously scattered Khurasanis of ahl bayt al-nabi bani ['l-.sup.[subset]]Abbas, and of obedience to this family whom God has appointed to rule. As Abu Muslim remarks dryly to his companions, "This isn't him talking." (122) What Abu Muslim hears is the voice of AbEl Jacfar telling him that the revolution is over.

Within a few years of AbEl Muslim's demise, Abu [Ja.sup.[subset]]far was boldly turning the rhetoric of revolution in precisely the opposite direction, against the still tenuous legitimacy of the Abbasid succession. Abu Humayd's old script was dusted off in 147/764-65 and reworked for use against [Isa.sup.[subset]] ibn Musa, hose title to the succession Abu [Ja.sup.[subset]]far now wished to set aside in favor of al-Mahdi. (123) As [Isa.sup.[subset]] was not slow to point out, his title was actually no different from Abu [Ja.sup.[subset]]far's both of them having been nominated by the previous caliph Abbas. (124) But where Abu Humayd had lectured Abu Muslim on how it took the Abbasids to turn the Khurasanis into an irresistible force, Abu [Ja.sup.[subset]]far could now lecture [Isa.sup.[subset]] ibn Musa on how it took the Khurasanis to rescue the Abbasids from political and social humiliation under the Umayyads. The Khurasanis were obviously an agent of providence; they had now been inspired with a determination to have al-Mahdi as their next caliph and wh o was he, Abu [Ja.sup.[subset]]far gainsay them? Abu [Ja.sup.[subset]]far does not go so far as to invoke explicitly the call to al-rida, though the theme of [ijtima.sup.[subset]] al-kalima, ma ['jtama.sup.[subset]]at [alayhi.sup.[subset]] ['l-.sup.[subset]]amma, alladhi ['jtama.sup.[subset]]a [alayhi.sup.[subset]] [ra.sup.[contains]]yu [ra.sup.[subset]]iyyat (amir [al-mu.sup.[contains]]minin) is a pretty close substitute. (125) A broader hint is Abu [Ja.sup.[subset]]far's application to his own situation of Zachariah's [Qur.sup.[contains]]anic prayer for an heir, [wa-'j.sup.[subset]]al-hu rabbi radiyyan. (126)

Al-Mahdi became caliph with the continuing problem of [Isa.sup.[subset]] ibn Musa's residual title to the succession. If he had been able to wait, the problem would have solved itself soon enough when [Isa.sup.[subset]] died in 167/784; but the army had scented power under the old caliph and were determined to have it under the new one. In 147/764-65 they had abused and threatened the caliph's nephew and heir apparent, but only when Abu [Ja.sup.[subset]]far had encouraged them to do so. (127) In 160/776 al-Mahdi was barely able to restrain them and the rhetoric of al-rida is correspondingly explicit. (128) Al-Mahdi himself declares from the pulpit that the Khurasanis had now agreed to have his son Musa as the new heir apparent li-'khtiyari-him lahu wa-rida-hum bihi, and [Isa.sup.[subset]] ibn Musa for his part is made to write in his deed of abdication, wa-dakhaltu fi-ma dakhala fihi 'l-muslimuna mina 'l-rida bi-Musa 'bni amiri 'l-[mu.sup.[contains]]minin. (129) Not only that, but al-Mahdi promises on his son's behalf that he will act as calip h, bi-kitabi 'l-lahi wa-sunnati nabiyyi-hi, a slogan whose revolutionary content is not only a claim to have principles, but an imputation that the present regime, or in this case the unwanted alternative to Musa's future regime, has none and is therefore illegitimate. (130)

Whatever principles may once have inspired the Khurasani revolutionaries, by the time of al-Mahdi's death in 169/785, it was money that motivated the soldiery of the Abbasid capital. (131) The right to elect the caliph had boiled down in practice to the right to be paid, even for officers who fled the caliph's enemies. (132) The fact that al-Mahdi had an approved successor in his son Musa al-Hadi did not prevent the military riot in Baghdad when the news of his death arrived. The Abbasids who were present at the time were quick to offer the troops what they wanted ([ra.sup.[contains]]a ['l-.sup.[subset]]Abbas (ibn Muhammad ibn [Ali.sup.[subset]]) an yurdaw), and what they wanted was pay. Even promises of pay were at first not good enough (lam yardaw wa-lam yathiqu bi-ma dumina lahum), but eventually a guaranteed lump sum of eighteen months' pay was a sufficient concession to the soldiers' principles to restore order. (133)

Under al-Hadi's new caliphate the whole process began again. "We won't have him (la narda bihi)," the military agitators against Harun as heir apparent are said to have declared. (134) From the other side of the argument, Yahya ibn Khalid asks al-Hadi whether the army would accept his minor son [Ja.sup.[subset]]far as their caliph if the worst happened (a-tazunnu anna 'l-nasa... yardawna bihi). (135) For al-Hadi, as we know, the worst did happen; the army realized that Yahya (or Harthama) had been right all along and decided they had better have Harun after all. He duly appears on the coinage of al-Haruniyya as al-khalifa al-mardi, a reminder to the army that Harun was not only the caliph they themselves had chosen but also, for some of them, the caliph they had been forced to choose for want of their originally preferred alternative. (136)

In al-Mawardi's theory of the caliphate, the concept of al-rida is still recognizable as the elective principle of ikhtiyar, a right, however, now institutionalized and restricted to the competent ahl al-ikhtiyar. The right of ikhtiyar is not only a restricted but also a residual right, to be exercised only in default of an [ahd.sup.[subset]] by which the community is bound to acknowledge as caliph a properly appointed wali [al-.sup.[subset]]-ahd on the death of his predecessor. (137) This later theoretical formulation suggests strongly that al-khalifa al-mardi was the caliph chosen without benefit of [ahd.sup.[subset]]. Harun, we can conclude, was not in fact wali [ahd.sup.[subset]] al-muslimin at the time of his accession, either because al-Hadi had formally replaced him with his son [Ja.sup.[subset]]far, or, more likely, because the army of the capital had never recognized him as such. It must be for the same reason that [Abu'l-.sup.[subset]]Abbas al-Saffah is remembered as having been called al-murtada. (138)

So, in fact, was Harun. A poem attributed to the Kaysani poet al-Sayyid al-Himyari praises the Hashemite caliphate of Abu [Ja.sup.[subset]]far and his descendants. The poem is said to have been recited to al-Mahdi after he had the oath of allegiance sworn to his two sons Musa and Harun and acknowledges the duty of obedience ([ta.sup.[subset]]a) to both al-Mahdi and his son Musa. Harun, too, whom the poet now calls al-murtada, has a right to something, but to what is not specified (wa-li-'l-rashidi ['l-rabi-.sup.[subset]]i 'l-murtada muftaradun min haqqi-hi 'l-lazimi). If this apparent distinction between al-Mahdi and Musa on the one hand and Harun on the other is significant, it tends to confirm the suggestion made earlier that the wilayat [al-.sup.[subset]]ahd was seen, like the caliphate itself, as fundamentally indivisible. The poet, in effect, makes a distinction between two levels of allegiance, one to the caliph and his heir apparent, which once acknowledged confers an absolute right to rule, and a lesser or provisional allegiance in which the element of consent still persists. The distinction neatly relegates the principle of al-rida to a secondary status below that of institutionalized succession. A Kaysani who had decided to flatter the Abbasids might well do just that. (139) The poet Salm al-Khasir, an adherent of the Barmakids, may have expressed a similar preference for [ahd.sup.[subset]] when he congratulated al-Hadi on assuming the caliphate without the involvement of shura. (140)

The title al-khalifa al-mardi has another echo in the literature, as Bonner has pointed out. (141) Here we return to al-Tabari's text of the formal address that is said to have been delivered to the army on the morning of Harun's accession. The text reminds the Abbasids and their partisans of their combined victory over the wicked Umayyads, the evils they had all suffered at their hands, and God's extraordinary favor in bringing them together in this sacred cause. The troops must not now put all this at risk by dissent. God had appointed Harun as caliph, rashidan mardiyyan, in place of the deceased al-Hadi. The text stresses Harun's beneficent intentions towards the army, and especially his determination to see them well paid. An immediate payment, equivalent to several months' salary, is promised. With these inducements, the troops are finally called on to give their oath of allegiance. (142)

This text is clearly anachronistic. It belongs in both language and content to the immediate aftermath of the Khurasani revolution, nearly forty years before. What it conspicuously fails to mention in 170/786 is any previous oath of allegiance to Harun, whether "after Musa" in 166/782-83 or at the time of al-Hadi's accession in 169/785. This suggests that on neither occasion had the army been committed to Harun as wali [al-.sup.[subset]]ahd. What is now invoked in place of an [ahd.sup.[subset]] is the old revolutionary rhetoric of al-rida, though it may be doubted whether it still counted for much apart from the offer of pay that went with it. The soldiers may have liked the idea of al-khalifa al-mardi on their dirhams, but a dirham is a dirham for all that. It was left to Harun to try to rebuild, if he could, the institutional legitimacy that his predecessors had put at risk. The rapid disappearance from the coinage of the title al-khalifa al-mardi was a step in that direction. Executing Abu Hurayra Muhammad ibn Farrukh w as probably another. It may be significant that the man sent to arrest the latter was Abu Hanifa Harb ibn Qays, one of the men whom Abu [Ja.sup.[subset]]far had used over thirty years earlier to murder Abu Muslim. (143)

If Harun was not heir apparent when he succeeded to the caliphate, why would al-Tabari want to prove that he was? It seems that, for al-Tabari, the principle of al-rida was at best a debased and at worst a dangerous political slogan. After the establishment of the Abbasid caliphate, the slogan first reappears in his chronicle in 169/786, when the Talibid al-Husayn ibn [Ali.sup.[subset]] ibn Hasan raised a small rebellion in Medina. (144) Al-Tabari treats the episode as an entirely futile descent into lawlessness, from bloodshed and violation of property to desecration of the Prophet's mosque, and all this on the basis of an oath of allegiance to al-Husayn as al-murtada min al Muharnmad. (145) His promise to stand down if his behavior in office fell short of expectations suggests not so much a laudable idealism as a reckless disregard for the realities and responsibilities of government. (146) In contrast to this is al-Tabari's treatment of the installation of [Abu'l-.sup.[subset]]Abbas as caliph in Kufa in 132/749. Despite his own account of a military rebellion and civil war in which the physical safety, let alone the political future, of the Abbasids depended entirely on the intentions of the rebels, al-Tabari is determined that the elevation of [Abu'l-.sup.[subset]]Abbas to the caliphate had been a forgone conclusion from the day when the Prophet Muhammad had first tipped off his uncle [Abbas.sup.[subset]] to that effect. (147) He is clearly aware of the claim that the Abbasids acquired their status as Hashemite pretenders from the Talibids, and specifically from a line of Talibids in whose interest the slogan al-rida min al Muhammad had been raised in the past, and he is quick to refute it. According to his version of the story, there had indeed been a meeting between the Abbasid Muhammad ibn [Ali.sup.[subset]] and the Talibid Abu Hashim ibn Muhammad ibn al-Hanafiyya, but Abu Hashim had done no more than try to tell his cousin what he already knew, that the longed for Hashemite caliphate was actually going to be an Abbasid caliphate. There is no mention here of any testament (wasiyya), nor even any suggestion that Abu Hashim was dying at the time. (148) When al-Tabari proceeds to events in Kufa in 132/749 he makes no reference to al-rida, and there is no hint that the choice of [Abu'l-.sup.[subset]]Abbas as caliph had anything to do with the revolutionaries. Amongst the several versions of the discovery of [Abul-.sup.[subset]]Abbas and his relatives in Kufa by officers of the Khurasani army, al-Tabari has a rare variant in which the soldiers' first question to the Abbasids is not which of them is [Abu'l-.sup.[subset]]Abbas, or Abd Allah ibn Muhammad, or Ibn al Harithiyya, but which of them is the caliph. (149)

Although in theory a call to legitimate revolution, the slogan of al-rida was too easily abused in practice as a license for rebellion in the provinces, for military insubordination in the capital, or even, as al-[Ma.sup.[contains]]mun was to show, for unwelcome initiatives by the caliphs. (150) As a principle of legitimate government, it was worse than useless. Legitimate government could never be a matter of choice by this or that faction and, even in the days before open faction, the caliph [Umar.sup.[subset]] ibn al-Khattab had taken the precaution of restricting the exercise of shura by means of an [ahd.sup.[subset]] to only six of the Prophet's companions. The continuing legitimacy of Abbasid government depended even more on succession by [ahd.sup.[subset]] and not shura. The [ahd.sup.[subset]] was a safeguard not only against revolution but also against arbitrary government. Behind al-Tabari's concern for Harun's status as wali [al.sup.[subset]] lies this issue of constitutional principle. On both occasions when [Isa.sup.[subset]] ibn Musa was rem oved from the office of wali [al-.sup.[subset]]ahd, al-Tabari takes some care to demonstrate that he gave his consent. (151) The point of principle is clearly that an [ahd.sup.[subset]], once contracted, binds not only the Muslims but also the caliph, and cannot be lightly cast aside with a self-serving appeal by either party to the slogan of al-rida. In Harun's case, since there is no evidence that he ever gave his consent to being removed from the wilayat [al-.sup.[subset]]ahd, it can only mean for al-Tabari that he never was removed. As the Barmakid version reassures him, even though al-Hadi was urged by his impatient officers to remove Harun without his consent if necessary, he fortunately hesitated to do so until providence made sure he could not. (152) Al-Hadi, like al-Mansur and al-Mahdi before him, recognized at least that degree of restraint over his own prerogative and the demands of his unruly faction, and for al-Tabari the legitimacy of Abbasid government before the great fitna of 193-204/809-19 was upheld.

(1.) N. Abbott, Two Queens of Baghdad: Mother and Wife of Harun al-Rashid (Chicago, 1946); M. Bonner, "AI-Khalifa AlMardi: The Accession of Harun Al-Rashid," Journal of the American Oriental Society 108 (1988): 79-91; idem, "The Mint of Harunabad and al-Haruniyya, 168-171 H.," American Journal of Numismatics 1, 2nd ser. (1989): 171-93; idem, Aristocratic Violence and Holy War: Studies in the Jihad and the Arab-Byzantine Frontier (New Haven, 1996); H. Kennedy, The Early Abbasid Caliphate (London, 1981); idem, "Succession Disputes in the Early Abbasid Caliphate (132/749-193/809)," in Union Europeenne des Arabisants et Islamisants, 10th Congress, Edinburgh 1980: Proceedings, ed. R. Hillenbrand (Edinburgh, 1982), 29-33; S. Moscati, "Le Califat d'al-Hadi," Studia Orientalia 13.4 (1946): 3-28.

(2.) P. Crone, "On the Meaning of the [Abbasid.sup.[subset]] Call to al-Rida," in The Islamic World from Classical to Modern Times: Essays in Honor of Bernard Lewis, ed. C. E. Bosworth et al. (Princeton, 1989), 95-111.

(3.) Moscati, "Califat d'al-Hadi," 18-22.

(4.) Abbott, Two Queens, 94-111.

(5.) Kennedy, Early Abbasid Caliphate. 101-2, 106-7, 109,110-13.

(6.) Disagreement with Kennedy's analysis has been expressed by A. Elad, "Aspects of the Transition from the Umayyad to the [Abbasid.sup.[contains]] Caliphate," Jerusalem Studies in Arabic and Islam 19 (1995): 115-17.

(7.) Bonner, "Al-Khalifa Al-Mardi," 84; idem, Aristocratic Violence, 78.

(8.) Bonner, "Al-Khalifa Al-Mardi:' 89; idem, Aristocratic Violence, 80.

(9.) Bonner, "A1-Khalifa Al-Mardi," 83-89; idem, Aristocratic Violence, 80-85; on Khuzayma, see P. Crone, Slaves on Horses: The Evolution of the Islamic Polity (Cambridge, 1980), 180.

(10.) Bonner, "AI-Khalifa Al-Mardi," 84, 87; idem. Aristocratic Violence, 80, 82.

(11.) Anon., al [Uyiin.sup.[contains]] [wa-'l-hada-.sup.[contains]]iq fi akhbar [al-haqa-.sup.[contains]]iq, vol. 3, ed. M. J. de Goeje & P. de Jong (Leiden, 1869), 269, 271; Yazid ibn Muhammad al-Azdi, [Ta.sup.[contains]]rikh al-Mawsil, ed. A. Habiba (Cairo, 1967), 238; Muhammad ibn [Abdus.sup.[subset] al-Jahshi-yari, [al-Wuzara-.sup.[contains]] wa-'l-kuttab, ed. M. [al-Saqqa-.sup.[contains]] et al. (Cairo, 1938), 145-46; Abu [Ja.sup.[supset]]far Muhammad ibn Jarir al-Tabari, [Ta.sup.[contains]]rikh al-rusul wa-'l-muluk, ed. M. J. de Goeje et al. (Leiden, 1879-1901), III, 471-77; Ahmad ibn Abi [Ya.sup.[supset]]qub [a1-Ya.sup.[supset]]qubi, [Ta.sup.[contains]]rikh, ed. M.Th. Houtsma (Leiden, 1883), II, 476. It is variously reported that Musa a1-Hadi was 21, 23, 24, 25, or 26 when he died in A.H. 170; Anon., [Uyun.sup.[subset], III, 289; al-Azdi, Mawsil, 259; Ahmad ibn Yahya al-Baladhuri, Ansab al-ashraf, vol. 3. ed. A. A. al-Duri (Wiesbaden & Beirut, 1978), 278; Abu [Umar.sup.[supset]] Ahmad ibn Muhammad lbn [Abd.sup.[supset]] Rabbi-hi, [al-.sup.[supset]]Iqd al-farid, ed. A. Amin et al. (Cairo, 1948-62), V, 116; Abu Muhammad [Abd.sup.[supset]] Allah ibn Muslim Ibn Qutayba, [al-Ma.sup.[supset]]arif, ed. Th. [Ukasha.sup.[supset]] (Cairo, 1960), 381; Khalifa ibn Khayyat, [Ta.sup.[contains]]rikh, ed. A. D. [al-.sup.[supset]]Umari (Najaf, 1967), II, 478; Abu Bakr Ahmad ibn [Ali.sup.[supset]] al-Khatib al-Baghdadi, [Ta.sup.[contains]]rtkh Baghdad (Beirut, n.d.), XIII, 22; Abu '1-Hasan [Ali.sup.[supset]] ibn al-Husayn [al-Mas.sup.[supset]]udi, Muruj al-dhahab [wa-ma.sup.[supset]]udin al-jawhar (Beirut, 1966), III, 324; idem, al-Tanbih wa-'l-ishraf, ed. M. J. de Goeje (Leiden, 1894), 344; al-Tabari, [[Ta.sup.[contains]]rikh, III, 579-80; [a1-Ya.sup.[supset]]qubi, [Ta.sup.[contains]]rikh, II, 491; two sources give the year of Musa's birth as A.H. 146, making him 23 or 24 when he died; Khalifa, [Ta.sup.[contains]]rikh, II, 478; Yaqut ibn [Abd.sup.[supset]] Allah al-Hamawi, [Mu.sup.[supset]]jam al-buldan (Beirut, 1957), III, 297, s.v. Sirawan; see Abbott, Two Queens, 23. Discounting the one report that Musa was twenty-one when he died (Khalifa, [Ta.sup.[contains]]rikh, II, 478), he must have been at least twelve, and may have been sixteen, when he w as made heir apparent in 160/776. Musa's own son, [Ja.sup.[supset]]far, was born in 163/779-80; al-Azdi, Mawsil, 243.

(12.) S. Lane-Poole, Catalogue of Oriental Coins in the British Museum, vol. 1: The Coins of the Eastern Khaleefehs in the British Museum (London, 1875), 53-54, nos. 97(?), 98-100; G. C. Miles, The Numismatic History of Rayy (New York, 1938), 46-47, nos. 67C, 68D; N. D. Nicol, "Early [supset]Abbasid Administration in the Central and Eastern Provinces 132-218 A.H./ 750-833 A.D." (Ph.D. thesis, University of Washington, 1979), 303, 328; W. al-Qazzaz, "al-Dirham al[supset]Abbasi fi zaman al-khalifatayn al-Mahdi wa-'1-Hadi' Sumer 20 (1964): 266-67, nos. 20-23; V. G. Tizengauzen, Monety vostochnago khalifata (St. Petersburg, 1873), 106, 110, 113, 282, nos. 968, 1015, 1047, 2772.

(13.) al-Tabari, [Ta.sup.[contains]]rikh, III, 492, 517.

(14.) al-Baladhuri, Ansab, III, 255; Ahmad ibn [Da.sup.[contains]]ud al-Dinawari, al-Akhbar al-tiwal, ed. V. Guirgass (Leiden, 1888), 382; Abu 'l-Faraj [Ali.sup.[supset]] ibn al-Husayn al-Isfahani, al-Aghani (Bulaq, 1868-69), VII, 14 = ed. I. al-Abyari (Cairo, 1969-1979), 7: 2695; al-Tabari, [Ta.sup.[contains]]rikh, III, 468.

(15.) R. A. Kimber, "The Early Abbasid Vizierate," Journal of Semitic Studies 37 (1992): 70-72.

(16.) al-Qazzaz, "al-Dirham al-abbasi," 262, no. 6.

(17.) Bonner, "Al-Khalifa Al-Mardi," 84, no. 26; Lane-Poole, Catalogue, 1: 52, nos. 93-94; al-Qazzaz, "al-Dirham al-'abbasi," 263, 265, 272, nos. 9, 10, 15, 16, 40, 42; Tizengauzen, Monety, 103-4, 106, 109, 111, 113, nos. 957, 977, 999, 1018, 1046. Jere Bacharach has argued that with neither a laqab nor the phrase wali [ahd.sup.[supset]] al-muslimin, the formula mimma amara bihi (fulan) ibn amir [al-mu.sup.[contains]]minin is of no significance for the succession: J. L. Bacharach, "Laqab for a Future Caliph: the Case of the Abbasid al-Mahdi," JAOS 113 (1993): 271-74; idem, "Al-Amin's Designated Successor: the Limitations of Numismatic Evidence," JAOS 116 (1996): 108-13. The fact that the formula is used here, before the oath of allegiance had been sworn to Harun, might seem to support this argument, but the point is discussed further below. See also the opinion of Fraehn, quoted by Tizengauzen, Monety, 104.

(18.) al-Tabari, [Ta.sup.[contains]]rikh, III, 500. Al-Tabari places his record of this appointment in the year A.H. 163, in which al-Mahdi sent Harun across the Byzantine frontier to raid. But this was probably late in A.H. 163, in the summer of A.D. 780, and al-Azdi is probably right to place the appointment early in A.H. 164 after Harun had returned from his raid to al-Raqqa: al-Azdi, Mawsil, 245; Michael the Syrian, Chronique de Michel le Syrien, ed. and trans. J.-B. Chabot (Paris, 1899-1910), 3: 1; H. Turtledove, The Chronicle of Theophanes (Philadelphia, 1982), 138-39.

(19.) According to al-Azdi, Harun was appointed to Armenia, Azerbaijan, al-Jazira (including Mosul), Syria, and Ifriqiya: al-Azdi, Mawsil, 245.

(20.) Anon., [Uyun.sup.[supset]], III, 264-65; al-Azdi, Mawsil, 218, 223-24; Ahmad ibn Yahya al-Baladhuri, Futuh al-buldan, ed. S D. al-Munajjid (Cairo, 1956-60), 213; Abu ['l-Abbas.sup.[subset]] Ahmad ibn Muhammad al-Marrakushi Ibn [Idhari.sup.[subset]], al-Bayan al-mughrib fi akhbar al-Andalus wa-'l-Maghrib, ed. G. S. Colin and E. Levi-Provencal (Leiden, 1948), I, 77-79; al-Tabari, [Ta.sup.[contains]]rikh, III, 372, 373; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 465; Michael, Chronique, 2: 526; Turtledove, Theophanes, 133; Bonner, Aristocratic Violence, 65, 67-68; Encyclopaedia of Islam, vol. 8 (Leiden, 1995), 410 (s.v. al-Rakka).

(21.) Anon., [Uyun.sup.[subset], III, 278; al-Azdi, Mawsil, 243-44; al-Tabari, [Ta.sup.[contains]]rikh, III, 500; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 480; Michael, Chronique, 3: 1; Turtledove, Theophanes, 138-39; Bonner, Aristocratic Violence, 73. Reports in which the Barmakids figure prominently define Harun's governorship as extending "from al-Anbar to Ifriqiya," without mentioning Armenia and Azerbaijan. They also say that Yahya ibn Khalid ibn Barmak was given full authority over the new governorship in Harun's name: al-Jahshiyari, [Wuzara.sup.[contains]], 150 (where the reference to Khalid ibn Barmak, line 9, must be a mistake for Yahya ibn Khalid; see lines 2-3 and p. 151 II. 17-18); al-Tabari, [Ta.sup.[contains]]rikh, III, 545, 571; [Umar.sup.[subset]] ibn Ahmad Ibn [al-Adim.sup.[subset]], Zubdat al-halab min [ta.sup.[contains]]rikh Halab, vol. 1, ed. S. al-Dahhan (Damascus, 1951), 61-62. Harun's appointment now appears as a mainly administrative measure reflecting the growing power of the Barmakids and the bureaucracy, but such assertions of th e Barmakids' importance deserve a more cautious appraisal than they usually receive. Al-Tabari is well aware of the formula "from al-Anbar to Ifriqiya," but does not mention it in his primary chronological record of Harun's appointment in A.H. 163. He does mention Yahya ibn Khalid, but only to say that he became responsible for Harun's official correspondence in connection with his new governorate: al-Tabari, [Ta.sup.[contains]]rikh, III, 500; see Kimber, "Early Abbasid Vizierate," 71-72.

(22.) Anon., [Uyun.sup.[subset]], III, 278-79; al-Azdi, Mawsil, 246, 247; al-Baladhuri, Futuh, 199; Khalifa, [Ta.sup.[contains]]rikh, II, 469; al-Tabari, [Ta.sup.[contains]]rikh, III, 503-6; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 476, 478, 486; Michael, Chronique, 3: 2; Turtledove, Theophanes, 141-42; Banner, Aristocratic Violence, 74.

(23.) See above, note 17; Bonner, "Harunabad and al-Haruniyya," 182-84, nos. 1-3 (with full references).

(24.) al-Tabart, [Ta.sup.[contains]]rikh, III, 492, 506. By contrast, the Umayyad al-Walid ibn Yazid refers explicitly to his "two heirs apparent," waliyyay [ahdi-hi.sup.[subset]] in the document announcing the simultaneous appointment of his two sons to that office: al-Tabari, [Ta.sup.[contains]]rikh, II, 1764.

(25.) al-Mawardi recognizes in theory the possibility of a nomination to the succession that falls short of an [ahd.sup.[subset]]; Abu 'l-Hasan [Ali.sup.[subset]] ibn Muhammad al-Mawardi, al-Ahkam al-sultaniyya wa-'l-wilayat al-diniyya (Cairo, 1960), 14-15; see also Abu [Ya.sup.[subset]]la Muhammad ibn al-Husayn [al-Farra.sup.[contains]], al-Ahkam al-sultaniyya, ed. M. H. al-Fiqi (Cairo, 1938), 11.

(26.) al-Tabari, [Ta.sup.[contains]]rikh, III, 495-98, especially 496 line 7, 497 II. 11-12.

(27.) Anon., [Uyun.sup.[subset]], III 279; al-Azdi, Mawsil, 247; al-Baladhuri, Ansab, III, 255; al-Dinawari, Akhbar, 382; al-Jahshiyari, [Wuzara.sup.[contains]], 150; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 476.

(28.) al-Tabari, [Ta.sup.[contains]]rikh, III, 518-19, 521. The strategic importance of Tabaristan was its proximity to the main artery of communication between Baghdad and Khurasan.

(29.) E. G. Browne, An Abridged Translation of the History of Tabaristan Compiled about A.H. 613 (A.D. 1216) by Muhammad ibn al-Hasan iba Isfandiyar (Leiden, 1905), 125-31; S. Moscati, "Studi storici sul califfato di al-Mahdi," Orientalia 14 (1945): 347-50; W. Madelung, "The Minor Dynasties of Northern Iran," in The Cambridge History of Iran, vol. 4: The Period from the Arab Invasion to the Saljuqs, ed. R. N. Frye (Cambridge, 1975), 202. In some modern Studies Musa's dispatch "to distant Jurjan," where he was based for the campaign, can appear as a minor provincial governorship, or even as evidence of his father's disfavor: Abbott, Two Queens, 69; Kennedy, Early Abbasid Caliphate, 107, but cf. idem, "Succession Disputes," 32; Bonner, "Al-Khalifa Al-Mardi," 84; idem, Aristocratic Violence, 77; D. Sourdel, Le Vizirat [abba-side.sup.[subset]] de 749 a 936 (132 a 324 de l'Hegire) (Damascus, 1959-60), 1:117.

(30.) Anon., [Uyun.sup.[subset]], III, 229; al-Tabari, [Ta.sup.[contains]]rikh, III, 133-34, 136-37, 139-40; al-Ya[subset]qubi, [Ta.sup.[contains]]rikh, II, 447; Browne, History of Tabaristan, 118-22; Madelung, "Minor Dynasties," 200. Al-Tabari already refers to al-Mahdi as wali [ahd.sup.[subset]] (sic., without the article) in connection with this mission, even though the formal removal of his cousin [Isa.sup.[subset]] ibn Musa from the office was not until A.H. 147; [Ta.sup.[contains]]rikh, III, 133-34. There is also numismatic evidence that al-Mahdi was being officially identified as heir apparent well before A.H. 147: Bacharach, "Laqab for a Future Caliph," 272-74; Miles, Rayy, 26-27, nos. 47B, 47C; Tizengauzen, Monety, no. 724 (a fals struck in Bukhara in A.H. 143, two years earlier than Miles' dirham and fals of A.H. 145, and already referring to al-Mahdi by his laqab).

(31.) When Abu [Ja.sup.[subset]]far appointed [Isa.sup.[subset]] ibn Musa to the governorship of Kufa, he was warned to expect removal from the office of wali [al-.sup.[subset]]ahd. If Abu [Ja.sup.[subset]]far had intended to keep him as heir apparent, he would have appointed him governor of Khurasan, balad [shi.sup.[subset]]ati-ka; Ahmad ibn al-Qasim Ibn Abi [Usaybi.sup.[subset]]a, [Uyun.sup.[subset]] [al-anba.sup.[contains]] fi tabaqat [al-atibba.sup.[contains]], ed. N. Rida (Beirut, 1965), 231.

(32.) al-Azdi, Mawsil, 252; a1-Baladhuri, Futuh, 226; al-Tabari, [Ta.sup.[contains]]rikh, III, 521. The evidence for Harun's continuing responsibility for the border is the appearance from A.H. 168 onwards of coins from the new mint of Harunabad/al-Haruniayya, bearing the inscription mimmd amara bihi Harun ibn amir [al.-mu.sup.[contains]]minin. The evidence is set out in full in Boner. "Harunabad and al-Haruniyya," 176-79

(33.) Anon., [Uyun.sup.[subset]], III, 279-80; al-Azdi, Mawsil, 253-54; al-Baladhuri, Ansab, III, 275; al-Dinawari, Akhbar, 382; Ibn Qutayba, [Ma.sup.[subset]]arif, 380; al-Jahshiyari, [Wuzara.sup.[contains]], 168 Khalifa, [Ta.sup.[contains]], II, 471; [al-Mas.sup.[subset]]udi, Muruj, III, 309, 323; idem, Tanbih, 343; al-Tabari, [Ta.sup.[contains]]rikh, III, 523-26; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rish, II 484-85; Michael, Chronique, 3:2: Turtledove, Theophanes, 143; H. Zonteberg, Clronique de Abou-Diafar-[Mo.sup.[subset]]hammed-Ben Djarir-Ben-Yezid Tabari, traduite sur la version persane [d'Abou-.sup.[subset]]Ali [Mo.sup.[subset]]hammed [Bel.sup.[subset]]ami (Paris, 1867-74), 4: 439-40.

(34.) al-Tabari [Ta.sup.[contains]]rish, III, 523; Anon., [Uyun.sup.[subset]], III, 279. Kennedy refers cautiously to this report as a rumor, but other have simply accepted it: Kennedy, Early Abbasid Caliphate, 108; cf. Idem, "al-Mahdi," Encyclopaedia of Islam, vol. 5 (Leiden, 1986), 1239; S. Moscati, "Nuovi studi storici sul califfato di al-Mahdi," Orientalia 15 (1946): 160-61, 171; idem, "Califat d'al-Hadi," 5; Abbott, Two Queens, 70; Sourdel, Vizirat, 1: 118; Bonner, "Al-Khalifa Al-Mardi," 84; idem, Aristocratic Violence, 77 (with a qualification).

(35.) See above, note 33; Kennedy, Early Abbasid Caliphate, 108. The district of Masabadhan was not even on the way to Jurjan, which is perhaps why both Moscati and Sourdel have al-Mahdi die "near Masabadhan," rather than in it: Moscati, "Nuovi studi," 171; idem, "Califat d'al-Hadi," 5; Sourdel, Vizirat, 1: 118. The persistent rumor of a late change to the succession may have arisen as a story of relief after adversity, in which the idea plays a minor part: al-Jahshiyari, [Wuzara.sup.[contains]], 167-68.

(36.) Anon., [Uyun.sup.[subset]], III, 283; al-Tabari, [Ta.sup.[contains]]rikh, III, 545, 547; Zotenberg, Chronique, 4: 446.

(37.) It seems certain that Hurun was with al-Mahdi when he died in Masabadhan and obtained the oath of allegiance to al-Hadi there: al-Azdi, Mawsil, 257; [al-Mas.sup.[subset]]udi, Muraj, III, 324; idem, Tanbih, 343-44; al-Tabari, [Ta.sup.[contains]]rikh, III, 544; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 487. Al-Dinawari and Ibn Qutayba say that HurOn obtained the oath of allegiance to al-Hadi in Baghdad, but al-Dinawari also says that this was on 22 Muharram, which others say was the day of al-Mahdi's death: al-Dinawari, Akhbdr, 382; Ibn Qutayba, [Ma.sup.[subset]]drif, 380; Khalifa, [Ta.sup.[contains]]rikh, II, 471; [al-Mas.sup.[subset]]udi, MurOj, III, 324; al-Tabari, [Th.sup.[contains]]rikh, III, 526; [al-Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 484. [Al-.sup.[subset]]Uyun wa-'l-[hadd.sup.[contains]]iq says that Harun obtained the oath of allegiance to al-Hudi in a context that suggests it was in Baghdad: Anon, [Uyrun.sup.[subset]], III, 283. Al-Jahshiyari says vaguely that [al-Rabi.sup.[subset] took charge of the oath of allegiance in Baghdad (qama bi-amri 'l-[bay.sup.[subset]a) a nd al-Baladhurt that Harun With [al-Rabi.sup.[subset]] kept things under control (dabata 'l-umur): al-Jahshiyari, [Wuzara.sup.[contians]], 167; al-Baladhuri, Ansab, III, 278. Harun's role in obtaining the oath of allegiance in Masabadhan could easily have been extended retrospectively to Baghdad without any positive evidence of what actually happened there; cf. Moscati, "Califat d'al-Hadi," 5-7; Abbott, Two Queens, 77-79; Kennedy, Early Abbasid Caliphate, 108.

(38.) al-Tabari, [Ta.sup.[contains]]rikh, III, 547.

(39.) See above.

(40.) This account of events in Baghdad is one of two different versions. The other is a typically slick creation designed to contrast the political savoir faire of Yahya ibn Khulid with the lesser skills of [al-Rabi.sup.[subset]], whom Yahya has to rescue from the new caliph's ire: Anon., [subset]Uyun, III, 282-83; al-Tabari, Ta[contains]rikh, III, 545-47. The latter version tends to be believed: C. E. Bosworth, The History of al-Tabari, vol. 30: The [Abbasid.sup.[subset]] caliphate in equilibrium (Albany, 1989), 7 n. 22. See also below, note 96.

(41.) al-Tabari, [Ta.sup.[contains]]rikh, III, 569-79.

(42.) al-Tabari, [Ta.sup.[contains]]rikh, III, 569.

(43.) Yahya ibn al-Hasan was the maternal uncle of al-Fadl ibn [al-Rabi.sup.[subset]] and he reports elsewhere how al-Khayzuran blocked his nephew's promotion until the day she died: al-Tabari, [Ta.sup.[contains]]rikh, 111, 332, 598, 608-9. This might explain his hostility to al-Khayzuran in what follows.

(44.) al-Tabari, [Ta.sup.[contains]]rikh, III, 569-70; [al-Mas.sup.[subset]]udi, Muruj, III, 327-28.

(45.) See below.

(46.) al-Tabari, [Ta.sup.[contains]]rikh, III, 570-71.

(47.) He may have been a son of [Sa.sup.[subset]]id ibn Salm ibn Qutayba, a boon companion of al-Hadi: Bosworth, History of al-Tabari, 44 n. 179, 59 n. 243.

(48.) al-Tabari, [Ta.sup.[contains]]rikh, III, 571.

(49.) See below.

(50.) al-Tabari, [Ta.sup.[contains]]rikh, III, 571-76.

(51.) al-Tabari, [Ta.sup.[contains]]rikh, III, 571, 573 (also 546); cf. al-Jahshiyari, [Wuzara.sup.[contains]], 169, 174.

(52.) al-Tabari, [Ta.sup.[contains]]rikh, III, 571-72; Anon., [Uyun.sup.[subset]], III, 285; cf. al-Jahshiyari, [Wuzara.sup.[contains]], 174, where the suggestion is avoided that steps were actually taken towards removing Harun from the succession. This is the one passage above all in al-Tabari where, as Bonner points out, the reader gains the impression that the generals were solidly behind al-Hadi with only the Barmakids backing Harun, and it forms an important part of Kennedy's argument: Bonner, "Al-Khalifa Al-Mardit," 84; Kennedy, Early Abbasid Caliphate, 110-11. It shows how much that argument depends on the central figure of Yahya ibn Khalid and the tendency of what is here called the Barmakid version to magnify his importance in the succession crisis. Bonnet also uses this passage, and specifically the removal from Harun of the symbolic privilege of having a lance carried before him, as evidence that Harun was actually removed from the office of heir apparent: Bonner, "Al-Khalifa Al-Mardi," 87; idem, Aristocratic Violenc e, 82-83. But the point of the passage is to show that al-Hadi's response to the military agitation fell short of formally changing the succession.

(53.) Nor is it any surprise that he does so in accordance with literary convention; cf. al-Tabari, [Ta.sup.[contains]]ritkh, III, 117, 583.

(54.) al-Tabari, [Ta.sup.[contains]]rikh, III, 572-73; Anon., [Uyun.sup.[subset]], III, 285-86; al-Jahshiyari, [Wuzard.sup.[contains]], 169-70.

(55.) al-Tabari, [Ta.sup.[contains]]rikh, III, 573.

(56.) al-Tabari, [Ta.sup.[contains]]rikh, III, 573-74; al-Jahshiyari [Wuzara.sup.[contains]], 170.

(57.) al-Tabari, [Ta.sup.[contains]]rikh, III, 574-75' al-Jahshiyari, [Wuzara.sup.[contains]], 170; al-[Mas.sup.[subset]]udi, Muruj, III, 333.

(58.) Two sources say explicitly that Yahya's policy was to play for time; Anon., [Uyun.sup.[subset]], III, 285; al-Jahshiyari, [Wuzara.sup.[contains]], 174. This makes sense only because the reader knows that al-Hadi is not going to live long. Yahya could hardly have known this, and the whole idea of playing for time must be a purely literary creation.

(59.) al-Tabari, [Ta.sup.[contains]]rikh, III, 575; al-[Mas.sup.[subset]]udi Muruj, III, 333.

(60.) al-Tabari, [Ta.sup.[contains]]rikh, III, 575.

(61.) al-Tabari, [Ta.sup.[contains]]rikh, III, 575-76.

(62.) al-Tabari, [Ta.sup.[contains]]rikh, III, 576-78; al-[Mas.sup.[subset]]udi, Muruj, III:334-35.

(63.) al-Tabari, [Ta.sup.[contains]]rikh, III, 578.

(64.) al-Tabari gives no date for al-Hadi's excursion to al-Haditha, and the reader quite naturally assumes, as the Barmakid version here requires, that it was in A.H. 170 just before al-Hadi died: cf. al-[Mas.sup.[subset]]udi, Muruj, III, 333. However, [al-.sup.[subset]]Uyun wa-'l-hada'iq and al-Azdi both have the same report that al-Hadi returned from al-Haditha not only because he was ill, but also because he had heard of the rebellion in the Hejaz of the Talibid al-Husayn ibn [Ali.sup.[subset]] who was later killed at Fakhkh. If true, this means that the excursion to al-Haditha was in A.H. 169 and not in A.H. 170, and therefore several months before al-Hadi's death: Anon., [Uyun.sup.[subset]], III, 284, 288; al-Azdi, Mawsil, 258, 259. Al-Jahshiyari just has al-Hadi fall ill without sending him off to al-Haditha first: al-Jahshiyari, [Wuzara.sup.[subset]], 174.

(65.) This inference is explicitly drawn in al-Azdi, Mawsil, 259.

(66.) al-Tabari, [Ta.sup.[contains]]rikh, III, 578; Anon., [Uyun.sup.[subset]], III, 288-89.

(67.) This apparent order of the two accounts is the way al-Tabari has arranged them and not necessarily their original order. It is equally possible that Yahya ibn al-Hasan picked up the same crucial point in al-Fadl ibn [Sa.sup.[subset]]id's account and twisted it to his own sinister purpose. Elsewhere he uses a similar technique to refute a (probably Barmakid) allegation that his brother-in-law al-[Rabi.sup.[subset]] ibn Yunus was murdered by al-Hadi:al-Tabari, [Ta.sup.[subset]]rikh, III, 597-98, 598-99.

(68.) al-Tabari, [Ta.sup.[contains]]rikh, III, 578; Anon., [Uyun.sup.[subset]], III, 289. Al-[Mas.sup.[subset]]udi uses essentially this same ending to wrap up both al-Khayzuran's quarrel with al-Hadi, which he relates in Yahya b. al-Hasan's version but without any allegations of murder, and the question of the succession, for which he gives the Barmakid version. His ending goes further than al-Tabari's, with a deathbed scene in which al-Hadi expires with his mother's hand held to his breast: al-[Mas.sup.[subset]]udi, Muruj, III, 327-28, 333-34. Al-Tabari, mercifully, knew when he had said enough to make his point. Such a transparent attempt as al-[Mas.sup.[subset]]udi's to refute an allegation of murder does not, however, prove that the original allegation was true, but only that it was thought likely to be believed, which it still is: cf. Bosworth, History of al-Tabari, 42 n. 171.

(69.) The published text of al-Tabari looks wrong here. It says "we were four women" but then appears to identify five. Instead of ana wa-ukhti wa-umm al-Hasan [wa-.sup.[subset]][A.sup.[contains]]isha bunayyat Sulayman wa-[ma.sup.[subset]]ana Rayta umm [Ali.sup.[subset]], I think the correct reading must be ana wa-ukhti umm al-Hasan [wa-.sup.[subset]][A.sup.[contains]]isha etc. or even ana wa-ukhtaya umm al-Hasan [wa-.sup.[subset]][A.sup.[contains]]isha, etc. This would mean: "We were four women, myself, my sister Umm al-Hasan, and [[blank].sup.[subset]][A.sup.[contains]]isha (or myself and my two sisters Umm al-Hasan and [[blank].sup.[subset]][A.sup.[contains]]isha), (we three being) daughters of Sulayman, together with Rayta Umm [Ali.sup.[subset]] (making four). Zaynab (the narrator), Umm al-Hasan and [[blank].sup.[subset]][A.sup.[contains]]isha were indeed three full sisters, which makes it unlikely that "my sister" refers to a fourth, unnamed sister: Ibn Qutayba, [Ma.sup.[subset]]arif, 375; cf. Bosworth, History of al-Tabari, 57; J. A. Williams, Al-Tabari: the Early [Abbasi.sup.[subset]] Empire (Cambridge, 1988-89), 2: 165.

(70.) Presumably the sum of 400,000 dinars is specified so that each of the four women present, including Rayta Umm [Ali.sup.[subset]], would receive 100,000. The implication is that al-Khayzuran's four companions knew as well as she what had really happened. Sadati here must surely refer to these four women, and not, as Bosworth says, to "my masters"; Bosworth, History of al-Tabari, 57.

(71.) al-Tabari, [Ta.sup.[contains]]rikh, III, 579.

(72.) The two versions are actually fused in Anon., [Uyun.sup.[subset]], III, 288-89; also in Moscati, "Califat d'al-Hadi," 23.

(73.) al-Tabari, [Ta.sup.[contains]]rikh, III, 576.

(74.) The version adds that the secretaries were brought to the house of Yahya's son, al-Fadl, who has previously been mentioned as representing his father at the court. This presumably indicates that the secretaries were based at the court at [Isabadh.sup.[subset]], that al-Fadl's house was there too, and that it was more convenient to bring the secretaries there than to the Khuld palace in Baghdad: al-Tabari, [Ta.sup.[contains]]rikh, III, 575, 578. Abbott takes this reference to al-Fadl's house to mean that Yahya must have been in prison at this point, but I do not think that this is the implication here: Abbott, Two Queens, 107.

(75.) al-Tabari, [Ta.sup.[contains]]rikh, III, 571, 579.

(76.) al-Tabari, [Ta.sup.[contains]]rikh, III, 601-2; see below.

(77.) al-Jahshiyari, [Wuzara.sup.[contains]], 174-75.

(78.) al-[Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 490, 491; Moscati, "Califat d'al-Hadi," 20 nn. 2, 3; Bosworth, History of al-Tabari, 51 n. 199.

(79.) al-Azdi, Mawsil, 236, 238; al-Tabari, [Ta.sup.[contains]]rikh III, 36, 469, 471. Al-Jahshiyari lists him amongst the officers who are said to have urged al-Hadi to remove Harun from the succession, although al-Tabari does not mention him in this context: al-Jahshiyari, [Wuzara.sup.[contains]], 174; al-Tabari, [Ta.sup.[contains]]rikh, III, 572.

(80.) al-[Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 489-90; Khalifa, [Ta.sup.[contains]]rikh, II, 479.

(81.) al-Tabari, [Th.sup.[contains]]rikh, III, 606; al-Azdi, Mawsil, 267.

(82.) See above, and note 19.

(83.) See above.

(84.) Bonner, "AI-Khalifa Al-Mardi," 87-88; idem, "Haruna-bad and al-Haruniyya," 180; idem, Aristocratic Violence, 80-81, 83-85.

(85.) In Yahya's own story according to al-[Ya.sup.[subset]]qubi, he says that he was let out first, though not by whom, and that he then released Harun.

(86.) al-Tabari, [Ta.sup.[contains]]rikh, III, 599-600. The text does not say explicitly that Harun was in prison, only that Harthama "brought him out."

(87.) Anon., [Uyun.sup.[subset]], III, 286-88; also Abu [Ali.sup.[subset]] al-Muhassin ibn [Ali.sup.[subset]] al-Tanukhi, al-Faraj [ba.sup.[subset]]da 'l-shidda, ed. [Abbud.sup.[subset]] al-Shaliji (Beirut, 1978), III, 19-22.

(88.) See below.

(89.) See below.

(90.) al-Tabari, [Ta.sup.[contains]]rikh, III, 601-2. For the situation, see al-Isfahani, al-Aghani, IX, 65 = X, 3590; al-Tabari, [Ta.sup.[contains]]rikh, III, 720.

(91.) al-Tabari, [Ta.sup.[contains]]rikh, III, 602. Al-Jahshiyari has the same story, though in his version Yahya ibn Khalid plays a part: [Wuzara.sup.[contains]], 174.

(92.) al-Tabari, [Ta.sup.[contains]]rikh, III, 602-3. subset]Abd Allah's perhaps Ostentatious show of commitment to his oath was duly rewarded with command of the new caliph's haras, and the governorship of Mosul: al-Tabari, [Ta.sup.[contains]]rikh, III, 603; al-Azdi, Mawsil, 269, 271, 273, 275; Khalifa, II, 502; al-[Ya.sup.[subset]]qubi, [Ta.sup.[contains]]rikh, II, 520; Crone, Slaves, 181.

(93.) Anon., [Uyun.sup.[subset]], III, 290-91. [Bal.sup.[subset]]ami's version of al-Tabari is a similar harmonization. Al-Hadi had wanted to make [Ja.sup.[subset]]far his heir apparent and some of the army commanders had sworn allegiance to him. But al-Hadi then gave up the idea, and [Ja.sup.[subset]]far was obliged when Harun became caliph to release these officers from their oaths of allegiance: Zotenberg, Chronique, 4: 456.

(94.) Ibn Abi [Usaybi.sup.[subset]]a, Tabaqat al-[atibba.sup.[contains]], 221-22; Abbott, Two Queens, 105-6.

(95.) See above.

(96.) al-Tabari, [Ta.sup.[contains]]rikh, III, 547-48; al-[Mas.sup.[subset]]udi, Tanbih, 344; Kimber, "Early Abbasid Vizierate," 75-76, 78-79. This explanation of al-[Rabi.sup.[subset]]'s temporary disgrace, which implies that he rather than Yahya ibn Khalid was Harun's political mentor at this stage, is obscured by a rival version in which al-Rabi? but not the prudent Yahya, blunders in to see al-Khayzuran before al-Hadi gets back and arouses the caliph's jealous anger on that account: al-Tabari, [Ta.sup.[contains]]rikh, III: 546; Moscati, "Califat d'al-Hadi," 6 n. 2; Abbott, Two Queens, 78; Sourdel, Vizirat, 1: 119; Kennedy, Early Abbasid Caliphate, 108-9; see above, note 40.

(97.) al-Tabari, [Ta.sup.[contains]]rikh], III, 548, 598; al-Jahshiyari, [Wuzara.sup.[contains]], 167; Encyclopaedia of Islam, 8: 351 (s.v. al-[Rabi.sup.[subset]] b. Yunus).

(98.) See above.

(99.) al-Qazzaz, "al-Dirham [al-.sup.[subset]]Abbasi," 279-80, nos. 69, 71, 73-74;p Tizengauzen, Monety, 119, 121, nos. 1086, 1100.

(100.) Bonner, "Al-Khalifa Al-Mardi," 87.

(101.) Tizengauzen, Monety, 121, no. 1103, possibly also no. 1107.

(102.) Bonner, "Harunabad and al-Haruniyya," 184-85, 187-88, nos. 4 & 10 (with full references to these and other issues of al-Haruniyya mentioned below).

(103.) See above, notes 17, 23.

(104.) Bacharach, "Laqab qab for a future caliph," 272; idem, "Al-Amin's Designated Successor," 109; see above, note 17.

(105.) Hamza ibn al-Hasan al-Isfahani, [Ta.sup.[contains]]rikh sini muluk alara wa-'l-[anbiya.sup.[contains]] (Beirut, 1961), 164; [Izz.sup.[subset]] al-Din [Ali.sup.[subset]] ibn Muhammad Ibn al-Athir, al-Kamil fi'l-[ta.sup.[contains]]rikh (Beirut, 1965-66), VI, 114; al-Jahshiyari, [Wuzara.sup.[contains]], 193; Khalifa, [Ta.sup.[contains]]rikh, II, 498; Ahmad ibn Abi [Ya.sup.[subset]]qub al-[Ya.sup.[subset]]qubi, al-Buldan, ed. A. W. T. Juynboll (Leiden, 1892), 304; Crone, Slaves, 185.

(106.) AI-Jahshiyari also notes that Muhammad had his own secretary at this early stage; [Wuzara.sup.[contains]], 193; sec above.

(107.) Z. Arzoumanian, History of Lewond: The Eminent Vardapet of the Armenians (Philadelphia, 1982), 147-48 (but cf. Bonner, "AI-Khalifa Al-Mardi," 88-89, with further references).

(108.) al-[Ya.sup.[subset]]qubi has a story illustrating the principle that, until a dead caliph is actually buried, the brother of his son and heir takes precedence over the heir's son: al-[Ya.sup.[subset]]qubi [Ta.sup.[contains]]rikh, II, 476. The anachronism Harun ibn amir al-[mu.sup.[contains]]minin is saying in effect that al-Mahdi may be dead, but is not yet buried.

(109.) Bonner, "Al-Khalifa Al-Mardi," 84-85; idem, Aristocratic Violence, 80-81.

(110.) Bonner, "Harunabad and al-Haruniyya," 186-88, nos. 6, 7,8, 9.

(111.) Bonner, "AI-Khalifa Al-Mardi," 85-87; idem, Aristocratic Violence, 81-83.

(112.) Banner, "Harunabad and al-Hartuniyya," 187-88, no. 10. There is also an issue of A.H. 169 that has Khuzayma as governor but does not mention Harun: ibid., 186, no. 6.

(113.) Three dinar issues of A.H. 170 have the novel addition of a personal name below the customary reverse legend Muhammad rasul Allah. One of the three different names is [Ja.sup.[subset]]far. It has been proposed that [Ja.sup.[subset]]far is [Ja.sup.[subset]]far ibn al-Hadi, but this is not certain. It is not even certain that the coins belong to al-Hadi's rather than al-Rashid's caliphate. Various personal names, including [Ja.sup.[subset]]far, continue to appear in this position on dinars for the first several years of al-Rashid's caliphate: N. A. R. Daftar, "Early Arab Islamic Coinage," Sumer 38 (1982): 155; T. EI-Hibri, "Coinage Reform under the [Abbasid.sup.[subset]] Caliph al-[Ma.sup.[subset]]mun," Journal of the Economic and Social History of the Orient 36 (1993): 69; Lane-Poole, Catalogue, 1: 61, 64-66, nos. 135, 141-42, 148, 150-52; idem, Additions to the Oriental Collection, 1876-1888, part 1 (London, 1889), 47-48, nos. 135a, 144b, 145a, 152a, 152c, 154a; N. S. M. al-Naqshabandi, al-Dinar al-islami fi 'l-mathaf al-[Iraqi.sup.[subset]] (Baghdad , 1953), 39- 40, 96, 99.

(114.) Bonner, "Al-Khalifa Al-Mardi," 88; idem, Aristocratic Violence, 83-84.

(115.) Bonner, "Harunabad and al-Haruniyya," 188-89, nos. 11, 12, 13.

(116.) Banner, "Al-Khalifa Al-Mardi," 89-91.

(117.) Crone, "The [Abbasid.sup.[subset]] Call to al-Rida," 95-99.

(118.) Crone, "The [Abbasid.sup.[subset]] Call to al-Rida," 100-102, 104; P. Crone and M. Hinds, God's Caliph: Religious Authority in the First Centuries of Islam (Cambridge, 1986), 32-33.

(119.) Crone, "The [Abbasid.sup.[subset]] Call to al-Rida," 96-98.

(120.) Crone & Hinds, God's Caliph, 3 1-32.

(121.) Crone does not entirely distinguish between the call to al-rida as used by the Abbasids themselves and as used by the Khurasani revolutionaries who actually overthrew the Umayyads. Her argument applies mainly to the Khurasanis, who she argues were in any case the real leaders of the revolution, but for the Abbasids the call to al-rida may well have served as a useful cover for their dynastic ambitions: "The [Abbasid.sup.[subset]] Call to

(122.) al-Tabari, [Ta.sup.[contains]]rikh, III, 106-7.

(123.) al-Tabari, [Ta.sup.[contains]]rikh, III, 338-41. Compare especially 111, 339-40 with III, 106.

(124.) al-Tabari, [Ta.sup.[contains]]rikh, III, 342-43.

(125.) All of which [Isa.sup.[subset]] ibn Musa reduces bluntly to ma [ajma.sup.[subset]]ta [alayhi.sup.[subset]], with the contrary advice that Abu [Ja.sup.[subset]]far be content (wa'-rda) with what God, through [Abu'l-.sup.[subset]]Abbas, had already provided for him: al-Tabari, [Ta.sup.[contains]]rikh, III, 342, 343.

(126.) al-Tabari, [Ta.sup.[contains]]rikh, III, 341. Abu [Ja.sup.[subset]]far was even moved to reward a poet who obligingly declared fa-qad radina bi-'l-ghulami 'l-amrad: al-Tabari, [Ta.sup.[contains]]rikh, III, 347.

(127.) al-Tabari, [Ta.sup.[contains]]rikh, III, 338, 344-45; Elad, "Aspects of the Transition," 100-101.

(128.) al-Tabari, [Ta.sup.[contains]]rikh, III, 471.

(129.) al-Tabari, [Ta.sup.[contains]]rikh, III, 472, 474.

(130.) al-Tabari, [Ta.sup.[contains]]rikh, III, 473; see Crone and Hinds, God's Caliph, 62-63, where it is pointed out that the Umayyads did not normally apply this slogan to their own rule, but that the Umayyad rebel Yazid HI did precisely that.

(131.) As Abu [Ja.sup.[subset]]far had already warned his son, "These are people whose love for you will never leave their hearts, provided that you treat them well, turn a blind eye when one of them does wrong, and reward them for what they have done": al-Tabari, [Ta.sup.[contains]]rikh, III, 444.

(132.) al-Tabari, [Ta.sup.[contains]]rikh, III, 527.

(133.) al-Tabari, [Ta.sup.[contains]]rikh, III, 547. When Abu Muslim was killed, his men were likewise offered money to ease their principles (hatta radu). They took it, but had a bad conscience about it afterwards: al-Tabari, [Ta.sup.[contains]]rikh, III: 117.

(134.) al-Tabari, [Ta.sup.[contains]]rikh, III, 572.

(135.) al-Tabari, [Ta.sup.[contains]]rikh, III, 574.

(136.) Stickel makes the interesting suggestion that coins with the inscription al-khalifa al-mardi could have been issued by supporters of [Ja.sup.[subset]]far ibn Musa who were holding out against the caliphate of Harun. He does not see his idea as very probable, but it does help to explain why it was in Harun's interest to claim the title for himself: J. G. Stickel, Handbuch zur morgenlandischen Munzkunde: Das Gro[beta]herzogliche orientalische Munzkabinet zu Jena (Leipzig, 1845), 88.

(137.) al-Mawardi, Ahkam, 6-15; for the collocation of rida and ikhtiyar see ibidem, 8; al-Tabari, [Ta.sup.[contains]]rikh, III, 472.

(138.) al-Khatib, [Ta.sup.[contains]]rikh Baghdad, X, 46, 47; also al-Baladhuri, Ansab, III, 162 (anta mahdiyyu hashimin wa-rida-ha). Mansur ibn al-Mahdi, who was urged to claim the caliphate in opposition to al-[Ma.sup.[contains]]mun after the death of al-Amin, was also called al-murtada, presumably by the Baghdadis who were urging him. He for his part prudently declined the title and all it implied: al-Khatib, [Ta.sup.[contains]]rikh Baghdad, XIII, 82; cf. al-Tabari, [Ta.sup.[contains]]rikh, III, 1001, 1005-6 (la narda bi-'l-majusiyyi 'bni 'l-majusiyyi 'l-hasani 'bni sahl). I am grateful to Judith Ahola for these references.

(139.) al-Isfahani, Aghani, VII, 14 = VII, 2696; Encyclopaedia of Islam, vol. 9 (Leiden, 1997), 116-17 (s.v. al-Sayyid al-Himyari).

(140.) al-Isfahani, Aghani, XXI, 128 = XXII, 7590.

(141.) Bonner, "Al-Khalifa Al-Mardi," 90.

(142.) al-Tabari, [Ta.sup.[contains]]rikh, III, 600-601.

(143.) al-Tabari, [Ta.sup.[contains]]rikh, III, 111-12, 114, 606.

(144.) al-Tabari, [Ta.sup.[contains]]rikh, III, 551-68.

(145.) al-Tabari, [Ta.sup.[contains]]rikh, III, 554.

(146.) al-Tabari, [Ta.sup.[contains]]rikh, III, 564. Al-Tabari's hostility to al-Husayn ibn [Ali's.sup.[subset]] rebellion is in marked contrast to al-Isfahani's account: Abu 'l-Faraj [Ali.sup.[subset]] b. al-Husayn al-Isfahani, Maqatil al-tabiiyyin, ed. A. Saqr (Beirut, 1987), 371-82; see also M. Q. Zaman, Religion and Politics under the Early [Abbasids:.sup.[subset]] the Emergence of the Proto-Sunni Elite (Leiden, 1997), 76 n. 29.

(147.) al-Tabari, [Ta.sup.[contains]]rikh, III, 23-24. What al-Tabari says here is essentially what Akhbar [al-.sup.[subset]] Abbas says more circumstantially. with the additional polemical edge that [Ali.sup.[subset]] is made to Witness the tip-off and the dying Abu Hashim ibn Muhammad ibn al-Hanafiyya is made to tell the story: Anon., Akhbar al-dawla ['l-.sup.[subset]]abbusiyya wa-fihi akhbar [al-.sup.[subset]]Abbas wa-waladi-hi, ed. A. A. al-Duri and A. J. al-Muttalibi (Beirut. 1997), 186-87. AlTabari as usual knows how much is enough to make his point.

(148.) al-Tabari, [Ta.sup.[contains]]rikh, III, 24; cf. M. Sharon, Black Banners from the East: the Establishment of the [Abbasid.sup.[subset]] State-incubation of a Revolt (Jerusalem, 1983), 121-40.

(149.) al-Tabari, [Ta.sup.[contains]]rikh, III, 35; al-[Mas.sup.[subset]]udi, Muruj, III, 255. Cf. Anon., [Uyun.sup.[subset]], III, 197, 199; al-Azdi, Mawsil, 121; al-Baljidhuri, Ansab, III, 139; al-Jahshiyarl, [Wuzara.sup.[contains]], 87; al-Tabari, [Ta.sup.[contains]]rikh, III, 28, 36.

(150.) When a1-[Ma.sup.[contains]]mun chose a Talibid as heir to the Abbasid caliphate, he could name him al-Rida because "he is rida in in the view of the Commander of the Faithful": Crone & Hinds, God's caliph, 138 n. 38.

(151.) [Isa's.sup.[subset]] consent is clearly seen in the heresiographical literature as a safeguard against extremism: Abu Salaf [Sa.sup.[subset]]d ibn [Abd.sup.[subset]] Allah al-[Ash.sup.[subset]]ari al-Qummi, al-Maqalat wa-'l-fraq, ed. M. J. Mashkur (Tehran, 1963), 67-68; Abu Muhammad al-Hasan ibn Musa al-Nawbakhti, Firaq al-[shi.sup.[subset]]a, ed. H. Ritter (Istanbul, 1931), 44-45; Zaman, Religion and Politics, 41-42. Al-Mawardi, by contrast, has no problem with a caliph removing a wali [ahd.sup.[subset]] appointed by his predecessor. He suggests very plausibly that Abu [Ja.sup.[subset]]far wanted [Isa's.sup.[subset]] consent only for political reasons, to maintain the unity of the Abbasid family: al-Mawardi, Ahkam, 13-14. See also above, note 30.

(152.) al-Tabari, [Ta.sup.[contains]]rikh, III, 575.
COPYRIGHT 2001 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Harun al-Rashid
Author:Kimber, Richard
Publication:The Journal of the American Oriental Society
Geographic Code:70MID
Date:Jul 1, 2001
Previous Article:The Kitab al-kasb attributed to al-Shaybani: poverty, surplus, and the circulation of wealth.
Next Article:Diptych in verse: gender hybridity, language consciousness, and national identity in Nirala's "Jago Phir Ek Bar".

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |