Ibn Tulun's Jihad: The Damascus Assembly of 269/883






Publication: Journal of the American Oriental Society
Author: Bonner, Michael
Date published: October 1, 2010

Ahmad ibn Tulun is famous as the first ruler of Egypt in Islamic times who became, to a large degree, independent of the central caliphate. Between his appointment as governor in 254/868 and his death in 270/884, Egypt emerged as a serious power in the Islamic and Mediterranean worlds, after centuries of subordination to the Roman, Byzantine, Persian, and caliphal empires. In particular, Egyptian-based Tulunid power extended - if only for a while - across Palestine and Syria to the borderlands facing Byzantium in southern Anatolia. Meanwhile, Islamic Egypt expressed itself in an increasingly mature and distinctive voice, using the Arabic language, in the religious sciences, secular literature, and the visual arts.1

This article will focus on an episode that is already known but has not received special attention. Toward the end of his life, Ibn Tulun convened an assembly of jurists and notables at Damascus in order to thwart the designs of al-Muwaffaq, brother of the reigning cAbbasid caliph, al-Muctamid. Not only did Ibn Tulun seek approval in declaring al-Muwaffaq deposed from his position as heir apparent (wall l-(ahd), but he tried to cajole or intimidate the assembly into subscribing to a declaration of jihad against this powerful prince. This unusual and (arguably) unprecedented move came as the culmination of a long involvement, on Ibn Tulun's part, with the Arab-Byzantine frontier district and the various practices of jihad.

Before I narrate this episode, I wish to describe its context, which is, broadly speaking, the frontiers of Islam, where we find, not surprisingly, a distinctive ethos and code of behavior for participants in the war against the adversaries of Islam. At the same time, we find an environment that is conducive to political creativity. In an unsettled era, when the central power and authority of the caliphate have grown fragmented and weak, many political and military entrepreneurs emerge. These tend to be men of low origins, often beginning as the client (mawla) or freedman (ghulam) of some great commander or prince. But they rise, compete, and then, in a few rare cases, become the founders of successful dynastic enterprises. Ibn Tulun is one of these. And like others, he participates in and deploys an ideology that we can identify - again, broadly speaking - as that of jihad. On the one hand, this jihad carries a backward-looking message: restoration of a long-gone social order, adherence to an obsolete system of exchanges and rewards, participation in wars of conquest in environments where conquest is no longer really possible. On the other hand, jihad carries new possibilities and meanings, especially when deployed creatively by the likes of Ibn Tulun. For, in fact, jihad is a core element in Ibn Tulun's project of founding a new, stable dynastic state (dawla). And with jihad comes, necessarily, the frontier, which I began by describing as the context of all this activity. As it happens, Ibn Tulun has the good fortune of becoming the ruler of Egypt, a sedentary and wealthy (if not entirely peaceful) country. But throughout his career he also remains attached to the wilder terrain of the Arab-Byzantine frontier. This attachment comes not only from pious devotion, but also from political ambition and calculation. For, again, the frontier is the breeding ground where entrepreneurial experiments can turn into flourishing Islamic states.

The episode in question took place as follows.2 In 269/883, Ibn Tulun marched with his army from Egypt to Syria, in order to deal with the northern limits of his domains. He had made a similar journey only a few years before (see below). This time, however, he had a new list of problems to deal with, which included the recent defection of Lu5Iu5, whom Ibn Tulun had previously appointed as his deputy in northern Syria and (according to some accounts) the northwestern frontier region known as the Thughur. Ibn Tulun also had to deal with the refusal of a certain Yazman, who had recently become the governor or strongman of Tarsus, capital of the Thughur, to recognize his authority at all. Most importantly, either before or during this march to Syria,3 Ibn Tulun received a message from the cAbbasid caliph al-Muctamid saying that he (al-Muctamid) had slipped away from S amarra5 in Iraq and was on his way to meet Ibn Tulun in al-Raqqa, in northeastern Syria. Al-Muctamid had held the title of caliph ever since his accession in 256/870, but had been compelled to share power with his brother, Abu Ahmad al-Muwaffaq. A talented and forceful military commander, al-Muwaffaq held the title of wall l-(ahd, or heir apparent.4 As time went by and as al-Muwaffaq defended the caliphate against its most dangerous adversaries, especially Yacqub the Coppersmith in the East5 and the Zanj in southern Iraq,6 his power and ambition grew, as did also, not surprisingly, the tension between the two brothers. In this quarrel, Ahmad b. Tulun in Egypt sided more and more with the caliph al-Muctamid in Iraq. Thus when al-Muctamid wrote to Ibn Tulun to say he was coming to meet him at al-Raqqa, it was because Ibn Tulun had already invited him to do this.

Ibn Tulun now waited in Syria for al-Muctamid to arrive, expecting to escort him to Egypt; we are told that in this way Ibn Tulun hoped to make his own capital, al-Qata'ic (near Fustat), the seat of a restored cAbbasid caliphate.7 However, al-Muwaffaq's agents got wind of the scheme and a commander loyal to him, Ishaq b. Kundaj, then governor of Mosul, encountered al-Muctamid at Haditha. According to some accounts, a leading administrator (katib or wazir) of al-Muwaffaq's named Sacid b. Makhlad also had a leading role, as he allayed the suspicions of al-Muctamid's commanders and then, once their guard was down, clapped them into irons.8 One way or another, al-Muctamid returned to Samarra' in virtual confinement. Al-Muwaffaq then awarded Ibn Kundaj the governorship of the western provinces, including Egypt, although Ibn Kundaj was not so foolhardy as to try to seize these from the powerful Ibn Tulun.9

In Damascus, meanwhile, Ibn Tulun convened an assembly of judges, jurisprudents, and notables (al-qudat wa-l-fuqahai wa-1-ashraf) from all his territories, which by now had come to include Egypt, Syria, Palestine, Diyar Mudar in the Jazira, 10 and the frontier region known as the Thughur, facing the Byzantine Empire. Once they had gathered, the participants made speeches denouncing al-Muctamid's confinement and al-Muwaffaq's "betrayal." However, Ibn Tulun wanted more. He demanded that the assembly approve of a document (kitab) that removed al-Muwaffaq from his position as wall l-cahd, or heir apparent, because he had usurped and denied the lawful caliph's rights. Ibn Tulun also demanded assent to the following proposition: "jihad against [al-Muwaffaq] is incumbent upon the [entire] Muslim community."

The document received the approval of most participants in the assembly, including the three chief qadis (judges) who held jurisdiction, respectively, over northern Syria and the Thughur, central Syria and Palestine, and Diyar Mudar. These three judges subscribed, in their own handwriting, to this effect. However, three Egyptian jurists, led by Bakkar b. Qutayba, chief qadi of Egypt for well over twenty years, balked and refused to subscribe to the document in its entirety. Furious, Ibn Tulun proceeded to issue his declaration nonetheless. From Iraq, al-Muwaffaq then replied with a counter-declaration that pronounced a curse against Ibn Tulun. This was read out from the pulpits of places still under direct cAbbasid rule, and most notably in S amarra5, where it was pronounced by none other than the caliph al-Muctamid and his son Jacfar al-Mufawwad. n Ibn Tulun was now stymied in his contest with al-Muwaffaq, but he proceeded with his northern campaign and his goal of dislodging Yazman from his position in Tarsus. This campaign went badly for Ibn Tulun as the people of Tarsus, who had already had an unfriendly encounter with him only a few years before (see below), now put up stiff resistance in dismal winter conditions. As Ibn Tulun and his army besieged the town, its defenders diverted the flow of the river Baradan so that it inundated them and forced them to abandon their camp and everything in it. In the midst of all this Ibn Tulun fell ill and returned to Egypt, where he died several months later, in 270/884.

The assembly that Ibn Tulun convened at Damascus was unusual for its time.12 Nearly a century earlier, cAbd al-Malik b. Salih, who held the governorship of Syria intermittently during the reigns of al-Rashid and al-Amin (170-198/786-813), solicited the opinions of several leading jurists regarding the fiscal and political status of the island of Cyprus. However, this consultation took place through correspondence (rasa'il) and not by a meeting.13 We may also consider the conclave of jurists convened years afterward, in Baghdad in 309/922 by the cAbbasid vizier Hamid b. al-cAbbas, with the goal of destroying the ecstatic mystic al-Hallaj. There, after much prodding, the jurists signed a death warrant, the caliph al-Muqtadir approved it, and al-Hallaj was gibbeted, tortured, and dispatched. This death warrant contained signatures from representatives of the different madhahib, or schools of law. 14 It is also noteworthy that both these proceedings - the consultation about Cyprus and the kangaroo court for al-Hallaj - were not instigated by caliphs, but rather by a governor and a vizier seeking legal cover for their undertakings. Both instances also generated some publicity. In these ways they had a certain amount in common with Ibn Tulun's assembly of 269/883.

Otherwise the early caliphs, both Umayyad and cAbbasid, sought advice from legal experts from time to time, but using discretion and out of the public eye. The caliph al-Ma'mun (198-218/813-833) may have been an exception, since he liked to listen to the disputations of jurists, as well as theologians. However, he apparently did not make his ruling decisions on the basis of these sessions. 15 It is not until considerably later that we find caliphs seeking advice or approbation through open, well-publicized consultations with experts in religion and law.16

What makes the Damascus episode of 269 especially interesting is the declaration of jihad against al-Muwaffaq. By this time several caliphs had been deposed 17 and a smaller number of them had been murdered, especially during the so-called "anarchy of S amarra5" (247-256/861-870). In the course of all this, quite a few holders of the office of wall l-cahd had also been set aside. Declarations of jihad against them, however, were unheard of. This declaration against al-Muwaffaq seems all the odder when we consider that at the time al-Muwaffaq was actually only second in line to the succession. 18 Of course, we can explain the anomaly by al-Muwaffaq's unprecedented position, as wall l-cahd with powers surpassing those of the caliph himself. His other attributions, moreover, were new, vague, and difficult to construe juridically. In other words, al-Muwaffaq's office of wall l-(ahd was the only one that made much formal sense and could therefore provide a target for Ibn Tulun's attack. Nonetheless, the use of "jihad" in this context seems strange. What precisely did Ibn Tulun (and his chancery officials and legal advisors) have in mind when they used this term?

Another point to consider at the outset is that, among our historical sources for these events, the Egyptians al-Kindi and al-Balawi - writing roughly a half-century later - provide us with the text of Ibn Tulun's document denouncing al-Muwaffaq, together with the story of the Damascus assembly and its aftermath. Otherwise, whenever this episode comes up in Arabic books of history and in biographies of the various jurists involved, the discussion hinges on Ibn Tulun's attempt to depose al-Muwaffaq, without any mention of his declaration of jihad, and often without any mention of the assembly at Damascus. And yet, in the two versions of the document that we have from al-Kindi and al-Balawi, this declaration of jihad is central to the logic and rhetoric of the entire argument. We will try to explain this disparity later on, but for now we may note that it underlies Thierry Bianquis's perplexity about the affair:

In fact, even despite the texts' silence on this issue, the obligation of jihad could not be invoked in a conflict among Sunnis, since the dispute did not weaken the capacity of the Dar al-Islam to confront non-Muslims or heretics.19

We can see that from the perspective of the doctrine of jihad as we know it Ibn Tulun's declaration does not actually make much sense. However, we know that jihad, both as doctrine and as practice, was at this time still at an early stage in its development and growth.20

In what follows, I will seek answers for some basic questions. Did this Damascus assembly and declaration of jihad really happen? If they did, then why do we know about them only from al-Kindi and al-Balawi and almost no one else? Moreover, al-Kindi and al-Balawi seem to transcribe the text of Ibn Tulun's declaration directly from official documents drawn from the archives. If this is so, then why did this text not trump the other evidence that was available to later historians? And what do we know, in more general terms, about archival and documentary evidence for the history of the Tulunids in Egypt and Syria?

Another question may be expressed as follows. There is general agreement that Ibn Tulun did much to make Egypt into an independent power, with a political and cultural clout that began to match the economic, demographic, and strategic importance that the country already had. But in this case, why did Ibn Tulun consume so much of what we would now call his "political capital," together with so much of Egypt's wealth, in order to gain authority over the Arab-Byzantine frontier region of the Thughur? After all, this chaotic and underpopulated region posed no threat to him, but it also offered few realistic prospects for conquest and occupation. No modern historian has explained this involvement of Ibn Tulun with this frontier region, beyond general considerations such as political ambition, prestige, and religious requirements regarding the conduct of warfare.

All these things come together in the episode of Ibn Tulun's Damascus assembly and declaration of jihad. Accordingly, this article will seek an improved understanding of that episode. After a brief discussion of the Arabic sources for this episode and for the life and career of Ibn Tulun, I will undertake a somewhat detailed account of Ibn Tulun's involvement with the Thughur. Then I will come back to the Damascus assembly, with a focus on the two extant versions of the document that Ibn Tulun prepared for that occasion. At that point we may ask what Ibn Tulun was trying to accomplish, both with his invocation of jihad and his policy regarding the Byantine frontier. We will also examine the reaction of his adversaries, especially when they responded more or less in kind. Along the way we will bring up questions regarding Tulunid historiography and the views that Arabic historians have taken, over the long term, of these twin themes of frontier and jihad.

THE SOURCES

We begin with a look at the major sources for the life and career of Ahmad b. Tulun. For our purposes here, the most important of these date from not long after the events in question, namely, the late third/ninth and the early-to-mid fourth/tenth centuries. Three of these sources in particular are richly detailed, laudatory, and often partial. Like other Arabic histories and biographies of this period, these books offer vivid portraits of individuals, and we may begin by asking how they characterize this particular hero. Ibn Tulun's outstanding qualities in these books include honesty, piety, refinement, and, above all, astuteness. The more detailed narratives ascribe to him an uncanny ability to intuit motives and to anticipate and outwit his opponents on the political chessboard. They also ascribe to him a thin-skinned ruthlessness: if he felt that you weren't expressing enough hayba ('awe') in his presence, you were likely to receive both a flogging and an assignment to the dungeons of Fustat, which overflowed with political victims of this kind.21 In the course of what follows, we should keep in mind these coexisting qualities of Machiavellian refinement and Stalinist paranoia.

We begin with Sirat Ibn Tulun ("the exemplary biography of Ibn Tulun") written by Ahmad b. Yusuf b. Ibrahim, known as Ibn al-Daya and as al-katib ('the scribe'). A man of wide-ranging literary and scientific interests, Ibn al-Daya died sometime between 330 and 340/941-95 1.22 During his childhood, his father Yusuf (also known as Ibn al-Daya) was held prisoner for a while in Ibn Tulun's palace, obtaining his release through the intervention of a group of devoted and courageous friends, led by the well-known Muhammad b. cAbd Allah b. cAbd al-Hakam.23 Afterward Ahmad Ibn al-Daya had his own brush with Ibn Tulun's prison.24 Nonetheless, his biography of Ibn Tulun is entirely laudatory. He also wrote biographies of the later Tulunid rulers and of commanders in the Tulunid service (ghilman bani Tulun), which may have actually been part of Sirat Ibn Tulun, but in any case have not survived.25 Meanwhile, Ibn al-Daya's biography of Ibn Tulun has survived only in an abridged version, which Ibn Sacid al-Andalusi (d. 685/1286) prepared and included in his al-Mughrib fl hula l-Maghrib.26 Here Ibn Sacid does not announce the title of Ibn al-Daya's work as Sirat Ahmad Ibn Tulun, but rather as al-Mustahsan min akhbar Ahmad b. Tulun ("The pleasing/appropriate selection from the reports of Ahmad b. Tulun"). 27 However, at the end of Ibn Sacid's abridgement, the work is called "the sira of Ibn Tulun,"28 and this is how it has usually been known.

Another book on the Tulunids, this one clearly titled Sirat Al Tulun ("The exemplary biography of the Tulunid family") was composed, apparently not long after Ibn al-Daya's death, by cAbd Allah b. Muhammad al-Balawi. This work, which survived in a unicum manuscript in Damascus,29 is considerably longer than what we have of Ibn al-Daya's biography of Ibn TOlOn. However, we do not know a great deal about its author, al-Balawi. The Fihrist of Ibn al-Nadim describes him as an Egyptian, a learned jurist and preacher, an author of books on religious law, and a Shicite of the Ismacili persuasion.30 Whatever we may think of this identification of al-Balawi as an Ismacili,31 there seems to be nothing in his Sirat Al Tulun to support it. Afterward Twelver Shicite tradition claimed al-Balawi as one of its own, though with little enthusiasm.32 The Sunnis, for their part, viewed al-Balawi as a falsifier of hadith and as the author of a work known as Rihlat al-Shafi(i ("the journey of al-Shafici" [to Egypt]) or alternatively Mihnat al-Shafici ("the trial or interrogation of al-Shafici"), either way, according to Ibn Hajar, a notorious fabrication.33 Oddly enough, the memory of al-Balawi preserved in bibliographical and biographical works, among both Sunnis and Shicites, does not include his book on the Tulunids. Equally oddly, few if any of the later Arabic historians seem to have made use of this book.34

Al-Balawi begins by saying that a certain (unnamed) patron has asked him to write a more detailed and better-organized biography of Ibn Tulun than that of Ibn al-Daya.35 Though he criticizes Ibn al-Daya's method and style, al-Balawi does not say precisely how he has used his predecessor's text and how he has acquired additional information. He certainly adds details to and elaborates on stories that he has taken from Ibn al-Daya. According to the modern editor, al-Balawi took over approximately fifty narratives or stories (qisas) from Ibn al-Daya, to which he added around forty more. 36 However, since the only extant text of Ibn al-Daya's Sira is an abridgement, we cannot really say if those forty-odd extra narratives were originally in Ibn al-Daya or not. Moreover, some modern scholars have thought that al-Balawi made more use of archival documents than his predecessor Ibn al-Daya did.37 However, this is impossible to prove for the same reason, namely, that we have only an abridged version of Ibn al-Daya. This matter of the archive will come up again later in this article.

We have seen that there is some doubt as to whether Ibn al-Daya's biography of Ibn Tulun actually had the title of Sira, but this does not matter much, and since Ibn al-Daya and al-Balawi's works are closely linked in purpose, content, and style, we will refer to them here as "the two Siras." It is clear, at any rate, that the literary genre of exemplary biography of a political ruler or religious scholar - with or without the title of Sira - was immensely popular in Egypt at the time,38 as well as in Iraq.39

A third source from this period is the famous Book of Governors [of Egypt] by Abu cUmar Muhammad al-Kindi (d. 350/96 1).40 Although this work is roughly contemporary with the two Siras and shares their positive attitude toward Ibn Tulun, it does not cite them directly and, at least at first glance, does not seem to share their sources. Here again, however, it is difficult to prove anything, since al-Kindi does not name his sources for Tulunid history.41 He gives plenty of information on the appointment of governors and bureaucrats, and seems on several occasions to quote from the state archives. Again, this matter of archives will come up later on.

Among other nearly contemporary historians, al-Tabari provides information on Ibn Tulun, but less than the three Egyptians just mentioned. Al-Yacqbi has little to say, al-Mascdi somewhat more.42 Later Arabic historians who relate these events rely heavily, though not solely, on these earlier sources, especially Ibn al-Daya, al-Kindi, and al-Tabari. Among these later sources, the ones of most interest to us here are Ibn al-Athir, Ibn Khaldun, al-Maqrizi, and Ibn Taghribirdi,43 as well as the Syrian urban histories of Ibn cAsakir and Ibn al-cAdim.44

IBN TULUN, JIHAD, AND THE THUGHUR

This section will set out what these literary sources tell us about Ibn Tulun's life and career, but with a strict thematic limitation. That is, I will not recount the life and career of Ibn Tulun, but instead will relate that life and career to the history of the Thughur, the Arab-Byzantine frontier district. Accordingly, I will leave out many basic aspects of Tulunid history. However, since Syria and Palestine lie between Egypt and the Thughur, and since Ibn Tulun's activities and ambitions in the frontier district were intertwined with his Syrian policies, Syria and Palestine will receive some notice.

The chronology of these events is rather difficult to establish. One reason is that the two Siras, with all their anecdotal detail, are not primarily interested in matters of dating and chronology, although they do pay some attention to them. Another reason is that we have relatively little information on these events from "mainstream" chroniclers such as al-Tabari. Fortunately we have al-Kindi, a specialist in Egyptian history who scrupulously reports dates, often down to the day of the month. However, al-Kindi tends to be laconic and does not cover everything. In the end, there are several unresolved matters here for which historians, both premodern and modern, have arrived at different solutions.

The two Siras say that when he was a young man in S amarra5, Ibn Tulun became upset at the behavior of the Turks around him. Apparently motivated by a desire to get away from them, he asked his (Turkish) friend Ahmad b. Muhammad b. Khaqan to put in a request to his uncle, the (Turkish) vizier cUbayd Allah b. Yahya b. Khaqan, for a posting to the Byzantine frontier district, known as the Thughur. The vizier granted this request. When Ibn Tulun arrived at the frontier, he received instruction from the hadith scholars and ascetics (zuhhad) of Tarsus. He admired their insistence on "upholding the good and abstaining from the wrong," and spent so much time attending their sessions that he used to arrive home late at night.45 We find no mention here of any military activity against the Byzantines. The only martial exploit credited to Ibn Tulun is a dash against Bedouin marauders during his return journey from the Thughur to Iraq. We are also told that he returned to Iraq earlier than he would have liked, because his mother missed him and feared for his safety. Had it not been for this filial loyalty, he would have happily remained in the Thughur.46

These anecdotes connecting the young Ibn TOlOn to the Thughur appear only in the two Siras and in later works that depend on them. We cannot say much about their literal truthvalue, but, of course, they fit in with the hagiographical portrait of the two Siras. The same applies to certain other themes in the Siras. For example, we have Ibn Tulun's aversion to the Turks, as when we are told that he saw them as uncouth and unfit to hold high office.47 In reality, of course, Ibn TOlOn was himself of Turkish descent. The point is, however, that the Siras portray Ibn Tulun as someone who, though a member of the Turkish military elite, is a civilized, pious man, associated with the Hanaf (or Shafici) school of law,48 the possessor of a thorough literary and religious education, and a participant in the holy war.

In 254/868, when Ibn Tulun first went to Egypt, his appointment as governor or viceroy extended only over that country, and not even all of it. From the beginning, however, he had a keen interest in "the Syrias" (al-Shamat).49 His first involvement with this region came in 256, when Isa Ibn al-Shaykh al-Shaybani was governor of Ramla and al-Urdunn (i.e., central and southern Palestine). With the caliphal government in S amarra5 apparently floundering, Ibn al-Shaykh refused to swear allegiance to the new caliph, al-Muctamid, and seized the tax revenues of Egypt in transit to Iraq. Ibn Tulun assembled an army and marched against Ibn al-Shaykh. Once the opposing armies made contact in southern Palestine, however, Ibn Tulun received a message "from Iraq" ordering him to return to Egypt. A force from Iraq then defeated Ibn al-Shaykh's troops, killing his son. Ibn al-Shaykh fled to Armenia, which he later managed to subdue. Al-Muctamid now appointed Amajur (or Majur) as governor of Damascus, in 257.50

Having missed this chance at intervening in Syrian politics, Ibn Tulun assembled his army in Alexandria in 258. 51 It was there that he seems to have made his first gesture toward the Arab-Byzantine frontier district. Ibn Tulun's brother Musa, who felt he was owed a high position in the government,52 requested the governorship of Alexandria, "because it is a frontier post (thaghr)" and thus, presumably, a place of high prestige. Ahmad promised him this but then vacillated, since he did not wish to alienate the city's current governor. Alexandria, after all, had come under Ibn Tulun's authority only a year before, when his patron and father-in-law Yarjukh granted it to him from S amarra5.53 Ahmad now proposed instead to assign Musa the governorship of Tarsus, chief city of the Thughur. The problem here, of course, was that Ibn Tulun had no authority over this region, even though, we are told, he expected to receive it soon, "because it was of all places dearest to his heart."54 The two brothers then fell out, with Musa accusing Ahmad of denying him his "due" (haqq) and of preferring his slave-officers (ghilman) over his own flesh and blood. Ibn Tulun replied by administering twenty lashes to his brother, reportedly with his own hand. He then sent him off to, of all places, Tarsus, the governorship of which (according to al-Kindi) he refused when Ahmad offered it to him a few years later.55

Meanwhile Amajur, the governor of Syria, felt threatened by his neighbor in Egypt, and warned the caliph of Ibn Tulun's growing might. Al-Muctamid then instructed Ibn Tulun to report to S amarra5 for reassignment. Ibn Tulun avoided the trap, however, by sending an agent who distributed lavish presents in S amarra5 and by mobilizing the efforts there of his patron and father-in-law, Yarjukh.56

In all this we see the cAbbasid caliphate of S amarra5 reasserting itself after its precipitous decline after the murder of al-Mutawakkil in 247/860. Beginning with the accession of al-Muctamid in 257/870, the caliphal government achieved some impressive successes. Five years into al-Muctamid's reign, his brother al-Muwaffaq became heir to the throne, wall l-cahd, second in line after al-Muctamid's son Jacfar al-Mufawwad. At the same time al-Muctamid granthosely divided what remained of the caliphate into two vice-regencies, with the western part (including Egypt) to be governed by al-Mufawwad, and the eastern part by al-Muwaffaq. Al-Muwaffaq, however, had the lion's share of power and prestige, and acquired more of these as time went by. As the two brothers grew more estranged, Ibn TOlOn supported al-Muctamid. This alliance received a boost in 262 or 263, when Ibn Tulun complained to the caliph that he could not govern Egypt without control over the country's finances. Al-Muctamid responded by appointing, as fiscal administrator in Egypt, Nafis al-Khadim, who became one of Ibn Tulun's trusted advisors. Nafis arrived in Egypt accompanied by several jurists, including Salih b. Ahmad Ibn Hanbal, who had previously served, notably, as qadi of the Thughur. We are also told that around this time, al-Muctamid gave Ibn Tulun the assignment he coveted most, the governorship of the Thughur.57

To understand the circumstances of this appointment, we need to examine the turbulent politics of the Thughur in these years. Here, however, the events and chronology are especially confusing. One historian who made a concerted effort at ordering this material is no less than Ibn Khaldun, and we may begin by relying on him. 58 In the early 260s the Thughur had as their commander a certain Sima al-Tawil, who had his capital at Tarsus, as was usual for the governors of the frontier district. At some point or another, Sima went east to Antioch (Antakya), intending to take this city, with its excellent fortifications, for himself. Here the governor on behalf of the caliphate, Muhammad b. cAli al-Armani, put up resistance, but Sima contrived to enter the city and to kill al-Armani. Faced with this insubordination, al-Muwaffaq appointed a Turkish officer named Arjun (or Arkhuz) b. Awlagh Tarkhun as governor of the Thughur and ordered him to seize Sima. However, Arjn left Sima unmolested in Antioch and moved on to Tarsus, where he mistreated the local population. We are told that Arjn delayed or seized people's stipends (arzaq) and rations (mira). Matters came to a head at a fortress directly facing Byzantine territory called Lu'lu'a. Here the garrison was low on supplies and funds, and urgently asked "the people of Tarsus" for help; otherwise, they said, they would have no choice but to abandon their outpost. The TarsQsis responded as best they could, collecting 15,000 dinars for the garrison. However, Arjun seized this money for himself. The garrison then abandoned Lu'lu'a and the Thughur fell into turmoil.59

Faced with this crisis in the border region, al-Muctamid urged his brother al-Muwaffaq to assign responsibility for it to Ibn Tulun. Understandably reluctant, al-Muwaffaq replied that "the Thughur need the presence of someone who will go on campaign (ghazw) together with the residents [of the district]," and that Ibn TOlOn was likely to appoint agents there who would not take up this cause personally. 60 However, al-Muwaffaq acquiesced, probably because he could not spare any resources from his struggle against the Zanj, whereas Ibn Tulun now had a powerful military establishment and could provide plenty of soldiers and commanders. Once he received this appointment, Ibn Tulun named one of these commanders, Tukhshi b. Balin, as his deputy in Tarsus, though it is not clear precisely when Tukhshi took up this position.61

Ibn Tulun thus acquired control over the frontier district, most likely in 262, 62 on condition that he supply 400,000 dinars a year. He owed this appointment to al-Muctamid, who at this time could still act on his own authority, though apparently he needed at least the acquiescence of his brother al-Muwaffaq. All the while, however, al-Muwaffaq was growing alarmed at Ibn Tulun's growing power. Shortly afterward, in 263, we are told that he decided to remove Ibn Tulun by conferring the governorship of Egypt on Amajur, the governor in Damascus. Amajur, however, declined, claiming (quite reasonably) that he lacked the strength to oppose Ibn Tulun. Al-Muwaffaq then sent a letter to Ibn Tulun demanding his resignation. Ibn Tulun replied with a letter that was resolute and taunting. Al-Muwaffaq then sent Musa b. Bugha to Syria to oust Ibn Tulun. Musa, however, conducted a long, inconclusive campaign in the Jazira, which ended with his ignominious return to Iraq, followed shortly by his death in 264.63

Nonetheless, the realities on the ground were still not favorable for Ibn Tulun, with Sima al-Tawil ensconced in Antioch and with Amajur still lord of Syria. In 264, however, Ibn Tulun had a stroke of luck, when Amajur died after seven years as governor in Damascus. Pretending that he was on his way to the Thughur to conduct an expedition against Byzantium, Ibn Tulun proceeded to Syria, where the governorship had passed to Amajur' s son cAli. No match for Ibn Tulun, this young man quickly yielded. Ibn Tulun then sent his agents to occupy the major cities of Syria and Palestine. He also besieged Sima in Antioch, and defeated and killed him in 265. M In this way Ibn Tulun became master of all of greater Syria, without any appointment from the caliphal government.

Instead of remaining in Syria to consolidate his position there, Ibn Tulun moved on to Tarsus and the Thughur, where he did have a previous caliphal appointment. According to the two Siras, he wished to return to his youthful activity of raiding and campaigning (ghazw), and once again he enjoyed the company of the ascetics of the borderland.65 However, some of his followers found this move useless and extravagant. ^ And the local population became so unhappy with the presence of many foreign troops and with the rising prices in their markets that they demanded that Ibn Tulun reduce his forces or simply leave. Amazingly, he did just that, pulling back from Tarsus.67 Meanwhile, Ibn Tulun received word that his son and heir al-c Abbas, whom he had left as his deputy in Egypt, had rebelled against him. Ibn Tulun returned to Egypt to deal with this situation in 266/879-80. During the next few years several military conflicts took place against the Byzantines, but Ibn Tulun was not personally involved in them, and in general his writ does not seem to have extended very far.68 Before leaving, he appointed a freedman (ghulam) of his and an officer in his service, Lu5Iu5, as his deputy over the frontier district and northern Syria and Diyar Mudar; or, according to other accounts, he left Lu5Iu' in charge of Diyar Mudar alone, with its capital of al-Raqqa.69 Again by some accounts Lu5Iu5 continued to campaign against Byzantium, at one point refusing the emperor's request for a truce. He also "repaired the Thughur and their funding (arzaq)."10

Not long afterward, Lu5Iu5 rebelled against his lord and master.71 He cut off his revenue payments to Ibn Tulun, seized Balis and Qarqisiyya from other commanders loyal to Ibn Tulun, and contacted al-Muwaffaq, then in the midst of battling the Zanj. Agreeably surprised, al-Muwaffaq invited Lu5Iu5 to Baghdad, where he showed him honor and used his troops in fighting the Zanj.72 In later years Lu5Iu5 must have regretted this move, because around three years after Ibn Tulun's death, al-Muwaffaq seized Lu5Iu5, imprisoned him, and despoiled him of 400,000 dinars.73 At the time, however, this defection of Ibn Tulun's lieutenant changed the situation in northern Syria and the frontier district, and ratcheted up the tensions between Ibn TOlOn and al-Muwaffaq. It was at this point that Ibn Tulun began to communicate directly with al-Muctamid, inviting him to take refuge in Egypt.74

Ibn Tulun now seemed poised for success, with the cAbbasid caliphate about to come under his protection. This time, however, his luck failed him. After years of desperate struggle, his adversary al-Muwaffaq was achieving major victories against the Zanj, enhancing his prestige and allowing him to divert some resources to Syria. Accordingly, as we have seen, when al-Muctamid left Samarra5 to join Ibn Tulun, he was foiled by al-Muwaffaq's agents and troops. Ibn TOlQn vented his frustration and fury in several ways; the one that concerns us here is his convening the assembly in Damascus, which I will take up again in the following section.

First, however, we need to look briefly at the situation in the frontier capital of Tarsus. Before leaving in 266, Ibn TOlQn had left there as his deputy Tukhshi (or Takhshi) b. Balburd, who upon his death was replaced by a certain Khalaf al-Farghani. 75 In 268 or 269, after the defection of Lu5Iu5, Ibn TQlQn instructed Khalaf to take action against Yazman, a eunuch formerly in the service of al-Fath b. Khaqan. 76 Khalaf seized Yazman and held him prisoner in Tarsus. There, however, the fighters (fund) freed Yazman of their own accord. Khalaf then fled to Damascus, as Yazman took his place as governor or leader of Tarsus. By at least one account, the TarsQsis were so glad to have rid themselves of Ibn TQlQn and his henchmen that they now cursed Ibn Tulun from their pulpit, anticipating the curse that al-Muwaffaq would pronounce against him soon afterward.77 In this way, as Ibn Tulun undertook what proved to be his final journey to Syria and the Thughur, he had to face not only the defection of Lu5Iu5, but also the bitter opposition of his beloved Tarsus.

Like other frontier districts, the Arab-Byzantine frontier district of the Tulunid era constitutes a kind of political laboratory or testing-ground. Its protagonists, such as Sima al-Tawil, Arjun, Tukhshi, Lu'lu', Yazman, and even Ibn al-Shaykh,78 can be seen as military and political entrepreneurs, with at least two outstanding characteristics.

The first of these is shifting loyalty. Each of these men begins his career as someone's mawla (freedman, client, protg). They rise to high rank through tenacious loyalty to a master such as al-Muwaffaq or Ibn Tulun. But then, paradoxically, they can only continue to rise - and arguably, they can only survive in power - if at some point they renounce this loyalty and take up the cause of a new master. An example is Lu5Iu5 and his betrayal of Ibn Tulun in favor of al-Muwaffaq. The career of Yazman in Tarsus is somewhat unusual: Yazman comes to prominence as the mawla of someone who is long gone (al-Fath b. Khaqan) and receives backing from an unusually independent-minded group, the "people of Tarsus," regarding whom I will have more to say later on.

Seen in this perspective, Ibn Tulun was a political and military entrepreneur just like the others, except that he had more refined political skills, a stronger network of companions (partly inherited from his father), and better luck. After all, it was largely through sheer luck that he managed to build such a brilliant career without ever betraying a benefactor - with the glaring exception of his failure, in Iraq in 252/866, to protect his charge, the deposed caliph al-Mustacin, from assassination, an event that the two Siras take some pains to justify and to explain away.79 Ibn Tulun was first installed, in 254/868, as governor of Egypt by Bakbak (or Bayakbak), who was put to death in S amarra5 a few years later. He then benefited from the patronage of Yarjukh in S amarra5, but never had to betray him. Then, when it came to his confrontation with al-Muwaffaq, Ibn Tulun had no previous links of clientage or exchange of benefits with this adversary. During this confrontation Ibn Tulun portrayed himself as the upholder of loyalty, especially toward the person and office of the caliph. This may seem ironic, in view of Ibn Tulun's earlier behavior toward al-Mustacin. But in any case it is to these entrepreneurs of the frontier that Ibn Tulun is best compared.

The second characteristic shared by most of these entrepreneurs of the frontier is that of jihad, an obvious component of frontier life, and nowhere more so than in this frontier facing Islam's oldest and greatest enemy. The following sections may give us a closer idea of what this characteristic actually meant in this rough-and-tumble world.

THE DAMASCUS ASSEMBLY

At the beginning of this article we saw that al-Muwaffaq's agents - including the governor of Mosul, Ishaq b. Kundaj, and the bureaucrat Sacid b. Makhlad - thwarted al-Muctamid's attempt at joining Ibn Tulun in 269 and led the caliph into semi-captivity in Iraq. Ibn Tulun then withdrew to Damascus (unless he was already there to begin with) and convened an assembly of judges, jurisconsults, and nobles (or notables, ashraf). It is likely that this group was large: we have the names of ten participants who came, "among others," from Egypt.80 During the assembly Ibn Tulun urged the participants to endorse a document in which he declared jihad against al-Muwaffaq and called for his removal from the office of heir apparent or wall l-(ahd. Except for three jurists from Egypt, all the participants complied.

Among the nearly contemporary Arabic literary sources that are still extant, only al-Kindi and al-Balawi report on the assembly.81 These two writers also furnish texts of all or part of the document that Ibn TQlQn produced for the occasion. Al-Balawi, who gives the fuller version of the two, calls this a "document of deposition," kitab al-khal(, which is what I will call it here. There are, however, quite a few differences between al-Kindi's and al-Balawi's versions of this text. Al-Kindi's account is rather short and includes only a few direct quotes from the document.

Ibn Tulun found out what Abu Ahmad [al-Muwaffaq] and Ishaq b. Kundaj had done. He then returned to Damascus and wrote to his administrator (camil), instructing him to convene the judges, jurisprudents, and nobles (al-qudat wa-l-fuqaha' wa-1-ashrf). He also wrote out [a document] regarding al-Muctamid's situation and what was to be done about it. This document of his arrived in Misr [al-Fustat, capital of Egypt], where it was read aloud to its people, to the effect that (bi-anna) Abu Ahmad [al-Muwaffaq] had violated his oath of allegiance to al-Muctamid, had imprisoned him and incited [people] against him (wa-harrasha calayhi) in the palace of [AbQ] Ahmad b. al-Khasib;82 and that al-Muctamid had thereby been reduced to an unmentionable condition and was weeping bitterly. Then, [when] the preacher in Misr preached his Friday sermon, he relayed the information that had been received regarding al-Muctamid. He added [these words] to his sermon: "O God, requite him against those who have imprisoned and oppressed him." [The Egyptian delegation of jurists then proceeded to Damascus, where] the people of the Syrias and the frontier region (ahi al-Shamat wa-l-Thughur) were also present. When they had [all] gathered, Ibn Tulun ordered that a document (kitab) be produced in which he declared AbO Ahmad al-Muwaffaq deposed from his position as heir apparent because he had acted in opposition to al-Muctamid and was holding him in confinement. And he wrote in [this document] "that Abu Ahmad has cast off his obethence and has forfeited [God's] protection; and that accordingly, jihad against him is incumbent upon the community" (fa-lamma jtamacu amara Ahmad b. Tulun bi-kitab khalacafihi Aba Ahmad al-Muwaffaq min wilayat al-cahd li-mukhalafatihi li-l-Muctamid wa-hasrihi iyyahu wa-kataba fihi anna Aba Ahmad khalaca l-taca wa-bari'a min al-dhimma fa-wajaba jihaduhu cala l-umma).83

Al-Kindi writes here of two documents: one sent back to Fustat before the Damascus assembly and a second document produced during the assembly and then promulgated throughout Ibn Tulun's domains. Again, al-Kindi provides only a few direct, verbatim quotes from both of these.

Al-Balawi's version is considerably longer and apparently gives the entire text of the kitab al-khalc or "document of deposition."

. . . Ibn Tulun established himself in Damascus and sent for the judges of his various provinces. . . . Each [of them] advised him to depose [al-Muwaffaq] (fa-kullun aftahu bi-khalcihi), except for Bakkar b. Qutayba. . . . [Ibn Tulun] wrote out the document of deposition, having multiple copies made (wa-kataba kitab al-khalc cala nusakh). He sent a copy to each of his provincial governors, [with instructions for it] to be read aloud from the pulpits of all their cities and preserved (wa-tukhallada). Included in all this was the following: "In the name of the merciful and compassionate God. This has been agreed upon by the qadis, the pious men (al-awliya'), and the notables of the cities (wujuh ahi al-amsar). . . . [Then comes a long, dramatic account of al-Muwaffaq's mistreatment of al-Muctamid, including Ibn Kundaj 's trickery and violation of his word.] . . . Al-Muctamid adjured [al-Muwaffaq] to respect the oath of allegiance. But instead he performed [his outrages]. He [al-Muwaffaq] has withdrawn from his obethence; he has forfeited [God's] protection; and jihad against him has become an obligation upon the community (wa-innama qadima wa-qadfaraqa l-taca wa-bari'a l-dhimma wa-wajaba jihaduhu cala l-umma). . . . Because the community [of Muslims] has refrained from assisting [al-Muctamid], it is in a state of sin (fi haraj). And because its pious men (awliya') have failed to observe their oath of allegiance, they are in a state of perjury (fi hinth). ... All those in attendance [at the assembly] are of the view that [al-Muwaffaq] must be stripped of the office of heir apparent, which the Commander of the Faithful previously conferred upon him; that no one must have any further association with him; and that he must be fought (fa-ra'a kullu man hadara khalcahu mimma kana amr al-mu'minin battahu lahu min wilayat cahdihi wa-l-tabarri>84 minhu wa-ljihad lahu). For he has thwarted (reneged on) three obligations (idh kana qad manaba huquqan thalathatari): first, the obligation imposed by the imamate; second, the obligation imposed by brotherhood; and third, the obligation imposed by benefitfs] previously conferred upon him (haqq al-nicma calayhi). . . .85

The two versions (al-Kindi and al-Balawi) converge at several points, especially where they say: "[al-Muwaffaq] has withdrawn from his obethence, he has forfeited [God's] protection; and jihad against him has become an obligation upon the community." In both versions this is a culminating moment. In other words the declaration of jihad is not a mere rhetorical flourish, but is instead the very heart of the matter. I will return to this in the following section.

First, however, we must note that remarkably few of the later Arabic historians report this episode of the Damascus assembly, and even fewer of them mention this "document of deposition" at all. (An exception is the industrious al-Maqrizi. ?) Does this mean that we should question the historicity of the assembly, the jihad, the kitab al-khal(, and the rest of this story?

We may recall that al-Kindi, Ibn al-Daya, and al-Balawi are our most detailed (surviving) sources for Tulunid history. It is difficult to say to what extent al-Kindi worked independently from the two Siras. But in any case, all three have the merit of being Egyptian, unlike other historians of the time (such as al-Tabari and al-Mascdi) who have less information on Egypt (and also Syria). Furthermore, most of the historians of Egypt and Syria who wrote afterward, during the Ayyubid and Mamluk eras (ca. 1180-1520 ce.), relied heavily on al-Kindi and the Siras for their Tulunid information, although these were not their sole sources.87

Another peculiarity of this situation is that both extant versions of the episode - those of al-Kindi and of al-Balawi - are built around the text of the kitab al-khalc that Ibn Tulun imposed on the assembly. On the basis of these and similar passages modern scholars have thought that al-Kindi and al-Balawi, like some other Arabic historians of the time, acquired much of their material by transcribing documents that they found in official archives. 88 To our way of thinking, this sort of information - transcribed from official documents, directly into a historian's book - deserves our special interest and respect. Once again, however, we run into trouble. If al-Kindi and al-Balawi both copied this document directly from the archives, why do their texts diverge, at least in their wording if not in their basic sense?

As a first step, we can look to a recent article on al-Kindi by Wadad al-Qadi. 89 Al-Kindi often seems to cite or to use documents, but how does he get hold of these? It happens, but only rarely, that he announces that an archival document is the textual (verbatim) basis of a report that he is including in his own text. 90 Even there, however, it turns out more often than not that al-Kindi is quoting a document that was seen by one of his own sources, a scholar who worked a generation or more before himself, but that al-Kindi has not seen the original. It can happen, exceptionally, that al-Kindi examines a document with his own eyes and transcribes it directly into his own text: an example is a papyrus from the year 131/749, which al-Kindi says he found in the Umayyad-era archive (diwan Boni Umayya) and recorded in his Book of Judges, and which provides the starting point for the article by Wadad al-Qadi. But even then, al-Kindi does not have the attitude toward documentary evidence that we have today. He may even think of these archival discoveries as interesting curiosities and not much more.91

For al-Kindi, as for other Muslim scholars of the time, historical information must be transmitted according to proper form and technique, whereby teachers, students, and other transmitters meet face-to-face in rather formal sessions, using their voices at least as much as their eyes. As for reports derived from official documents, these do not seem to have been handled very differently from information that circulated orally before being written down. Now, it is likely that when an official document was first issued by a caliph, governor, bureaucrat, or military commander, the earliest transmitters recorded it accurately. But then, as the text passed from tradent to tradent and from teacher to student, what guaranteed its soundness was not its "documentary" status, but rather the reputation of the transmitters who relayed it. In this way texts that had originated as official documents continued to circulate as "reports," akhbar, subject to the same alterations and changes in emphasis as any other akhbar.

Afterward historians who composed new works on these topics could include these reports based on archival documents, or they could omit them. They might make this choice because of the constraints of literary genre and form, thematic priorities, ideological biases, the demands of patronage and raison d'tat, or a lack of comprehension of details from a long-gone era. In any case, if a historian had to choose between two contradictory reports, one of which was based on an official document while the other was not, there was no guarantee that he would privilege the "documented" report over its rival, as we can see already in al-Kindi's Book of Judges.92 Having said this, however, we need to remember that it remained perfectly possible for historians to transcribe and to use official documents. Let us now look more closely at the two surviving versions of the "document of deposition" that Ibn Tulun produced at Damascus and then promulgated throughout his domains.

I begin with al-Balawi' s version. First of all, we must note that the episode of the Damascus assembly and the text of the kitab al-khal( occur in al-Balawi but not in Ibn Sacid al-Andalusi's abridgement of Ibn al-Daya. There are two possible explanations for this disparity. The first is that Ibn al-Daya did include this material, and Ibn Sacid subsequently omitted it when he abridged Ibn al-Daya's book for inclusion in his al-Mughrib. The other is that Ibn al-Daya never had this material and that al-Balawi located it on his own (whether from an archive or somewhere else) and added it to his longer, fuller Sira.

There are several reasons for arguing that al-Balawi's account of the assembly and the document of deposition actually goes back to Ibn al-Daya. One is that in the abridged version of Ibn al-Daya that we have, this episode seems truncated and does not, on its own, entirely make sense. Another reason has to do with the availability of archives in Egypt. In the early third/ninth century, an archive containing first /seventh-century materials was apparently still standing in Fustat, behind the Mosque of cAmr,93 and we have seen that al-Kindi had access to archival materials from the late Umayyad era. But what about the more recent Tulunid archives? In 292/905, when forces commanded by Muhammad b. Sulayman overthrew the Tulunid regime and (re)conquered Egypt for the cAbbasid caliphate, they made prisoners of the surviving members of the Tulunid family, butchered whatever soldiers they could find, pillaged and raped civilians, and destroyed much of Fustat and al-Qata5ic, including the famous Tulunid buildings (with the exception of the great mosque). I have not found any direct mention of the Tulunid archives in the descriptions of this horror,94 but it is easy to imagine that these were either burned on the spot, or carted off to Baghdad with the prisoners and other trophies. Now, if these archives were destroyed, then al-Balawi and al-Kindi never saw them, since they came to adulthood after 292/905. Ibn al-Daya, however, belonged to the previous generation and was, already in his youth, a bureaucrat (katib) in the Tulunid service.

The main argument in favor of Ibn al-Daya as the source of this information, however, comes from al-Balawi's Sira itself. As we have seen, this book includes a transcription of the original document of the kitab al-khalc. At the conclusion of this, we read

Ten copies [of this document] were accordingly written out, in a single series, with no variation among them. These also contained [the confirming statements] in the handwriting of the qadis, [set out here] according to my transcription (fa-kutiba bi-dhalika cashr nusakh nasaqan wahidan la yughayiru bacduha bacdan wa-fiha khutut al-qudat bi-ma nasakhtuhu).95

Then follows the text of the subscribing statements made by three qadis (to be discussed below). From this passage it seems most likely that either al-Balawi or Ibn al-Daya transcribed not only the judges' statements, but the entire document. But how do we know if the original speaker and transcriber, speaking here in the first person, is al-Balawi or Ibn al-Daya?

Throughout his Sira, al-Balawi occasionally cites his sources, which often coincide with those of his main source, Ibn al-Daya.96 When he arrives at the Damascus assembly, however, he does not do this, and we have to look back for several pages before we find anything like a listing of sources. When we do find one, the attribution is ambiguous: qala mu'allif hadha l-kitab "the author of this book says."97 This in itself does not point convincingly to either al-Balawi or Ibn al-Daya as the source of the document. What does point to Ibn al-Daya, in addition to the other arguments already made, is what we find in the later author al-Maqrizi (d. 845/1442). Unusually for a Mamluk-era historian, al-Maqrizi reports the episode in full. Here he not only makes use of al-Kindi,98 but also gives substantially the same text of the Damascus assembly and the document of deposition as the one in al-Balawi' s Sira (with some variations, mostly minor).99 However, the source he cites here is "Ahmad b. Yusuf al-katib," i.e., Ibn al-Daya. 10 Al-Maqrizi does not mention al-Balawi in any place that I have found.101 However, al-Maqrzi's version of the text corresponds to the version that we know from al-Balawi and that is lacking in Ibn Sacid's abridgement of Ibn al-Daya. From this we may conclude, first, that al-Maqrizi had access to an unabridged version of Ibn al-Daya's biography of Ibn TQlQn; and second, that Ibn al-Daya is the transcriber of the "document of deposition," and with that, the original author of the account of the Damascus assembly that we now have in the Sira of al-Balawi.

I now return to al-Kindi. He gives no indication of his sources for the governorship of Ibn Tulun. This is his usual procedure for later periods of Egyptian history, close to his own time, and unlike his treatment of earlier periods. Accordingly, we do not know if al-Kindi' s Tulunid sources included Ibn al-Daya, but in any case he no doubt used other sources as well. Here, as al-Kindi relays the document of deposition, he quotes a piece of it, and gives a general sense of the rest ("to the effect that," bi-anna). In other words, al-Kindi relays information that he has received from a scholarly source, but not through examination of the original documents, which in his day had probably already ceased to exist.

We may conclude that the "document of deposition" that we have in al-Balawi and al-Maqrizi is what it purports to be, namely, the verbatim text of the document that Ibn TQlQn sent out, in ten identical copies, to the major provinces of his realm. Moreover, it was Ibn al-Daya who copied this text from one of the original ten documents, preserved in the archive at al-Qata'iS for inclusion in his biography of Ibn TQlQn. Later on, al-Balawi included this material from Ibn al-Daya in his Sira, in the form that we now have. Al-Kindi, meanwhile, probably did not have the original document, and in any case did not quote it in full. Three centuries later, when Ibn Sacid al-Andalusi edited Ibn al-Daya's work for inclusion in his al-Mughrib, he omitted the "document of deposition" and the story of the assembly. Most later Arabic historians who dealt with these matters did the same, whether through ignorance or by choice.

Finally, al-Kindi tells us that everyone at the assembly testified in favor of the document, with three exceptions (see the following section). 102 Al-Balawi and al-Maqrizi give the text of the statement that three senior judges appended to the document, in their own hand.

'Ubayd Allah b. Muhammad al-cUmari, qdi of the Junds of Qinnasrin and the cAwasim, of the Syrian Thughur, and of the Junds of Hims and Antioch, says the following. This document has been read to me, and it correctly represents my position and the truth as I see it, and my legal opinion. This is because of what I have ascertained regarding the treachery of the False One (al-nakith) known as Abu Ahmad, and his hostility and breach of obethence toward the Commander of the Faithful, may God assist him; and furthermore, because I have ascertained that [Abu Ahmad] merits this [treatment] because of his past behavior. [All this is] in accordance with what has been stated and described in this document, namely, that his name should be excised, he should be deposed, and he should no longer be named in the [Friday] prayer. [I have further ascertained and agree] that he does not deserve to hold the office of Imam of the Muslims, that he cannot be entrusted with authority over them, and that he cannot be relied upon [in that office]. I call as witnesses those who have subscribed their testimony to this effect in this document, in accordance with my pronouncement and legal opinions (wa-ashhadtu cala qawli wa-fatayaya man kataba shahadatahu fi hadha l-kitab).103

In addition to al-cUmari, 104 the exact same statement was written out by cAbd al-Hamid b. cAbd al-????, qadi of Damascus, al-Urdunn, and Palestine; 105 and Ahmad b. Abi l-cAla\ qadi of Diy ar Mudar. 106 Of these three, only the Hanafi cAbd al-Hamid had much of a reputation as a jurist; the other two had no clear affiliation with any madhhab. All three, however, were quite visibly in their seventies or eighties (cAbd al-Hamid, we are told, did not dye his hair107).

Having obtained the signature of three of his chief judges, Ibn TQlQn needed the fourth. This was Bakkar b. Qutayba,108 chief judge of Egypt, a respected Hanafi jurist who had also reached an advanced age. Bakkar did not actually refuse to subscribe (as was and is often said). Instead, he wrote out a statement that seemed to give Ibn TQlQn most of what he wanted.

Bakkar b. Qutayba held back from confirming his testimony (tawaqqafafi shahadatihi). Ahmad b. Tulun then grew angry, because [Bakkar] did not set out the legal arguments that the others had provided and did not testify as they had done. The reason why [Bakkar] held back was his insistence on piety and religion. Accordingly, he wrote: "Bakkar b. Qutayba, qadi of Misr (Fustat) and Alexandria and their districts, testifies to what has been set out and described in this document, 109 from its beginning to its end, regarding the kindness and favor that the Commander of the Faithful, may God assist him, has shown to the False One (al-nakith) Abu Ahmad b. Jacf ar al-Mutawakkil cala llah, and regarding [Abu Ahmad's] hostility toward the Commander of the Faithful. [Bakkar further testifies] that the False One, Abu Ahmad, because of his actions, merits [the treatment meted out to him, sc] being deposed and left out of the [Friday] prayer. Bakkar b. Qutayba has written [this] in his own hand."110

This statement from Bakkar seems to contradict the entire story of his conflict with Ibn Tulun, as it is usually told, because he names al-Muwaffaq as al-nakith "The False One," and actually endorses his dismissal from the succession. U1 However, al-Balawi and al-Maqrzi - each independently following Ibn al-Daya, as we have seen - also say here that Bakkar "held back." Perhaps all we can say is that Bakkar's statement is less strong than the other three, and that Bakkar was seen afterward as Ibn Tulun's chief opponent in this matter. For by all accounts Ibn TQlQn was not satisfied with whatever it was that Bakkar said and wrote, and he persecuted him for what little remained of both their lives. Meanwhile, the document was read out from the pulpits, and al-Muwaffaq's name was ostentatiously omitted from the Friday prayer.

JIHAD AND PROPAGANDA

The "document of deposition" transmitted by al-Balawi and al-Maqrizi from Ibn al-Daya's transcription of the original does not provide a theoretical basis for invoking jihad, although it uses the word jihad twice. It does, however, give us some idea of what the author(s) of the document had in mind. To begin with, it describes, at some length, the deception and mistreatment that al-Muwaffaq and his henchman, Ishaq b. Kundaj, have inflicted on the caliph. Then it says that al-Muwaffaq has breached (or 'impeded', mana(a) three obligations or duties (huqUq). The first of these is the obligation of the caliphate, which is the primary concern of the document. More specifically, al-Muwaffaq has violated the solemn oath of allegiance (bay(a) that he made to the caliph upon his accession. The second obligation or duty, somewhat surprisingly (for us), is conferred by brotherhood. Third, and even more surprisingly, we find the obligation imposed by "benefit[s] previously conferred upon him" (haqq al-ni(ma calayhi), here referring to al-Muctamid's acts of kindness toward his brother.112

Here we have a triad that covers the entire universe of these duties and obligations, as we find them so often in works of history and belles-lettres from this and later periods.113 The oath of allegiance to a caliph is the most solemn and theatrical of the commitments that people make, more or less voluntarily and contractually, as they create and confirm what Roy Mottahedeh has called their "acquired loyalties." Loyalty to one's brother is, in Mottahedeh's language, a "loyalty of category," an obligation that we incur simply by being who we are. The third kind of obligation, "gratitude for benefits previously bestowed," though somewhat more nebulous than the other two, is a universally shared value in the late cAbbasid world. The Siras, in fact, take some pains to show that Ibn TQlQn was scrupulous and demonstrative in his gratitude for gifts and other benefits that he received.114

In more general terms, we might say that society consists of an intertwined fabric of loyalties and associations, all in a constant process of refashioning and renegotiation. Oaths and vows are the way to make binding arrangements for the long and short term, whenever such arrangements cannot be guaranteed by already existing ties of kinship and patronage. Thus, al-Muwaffaq has violated the principle around which society coheres: loyal observance of one's obligations, both when these are guaranteed by oath and when they are not.

Everyone could agree, then as now, that ingratitude, rejection of kinship ties, and violation of oaths are terrible things, but they seem unusual as casus belli and a basis for jihad. This provides us with another explanation for why later writers, including Ibn Sacid al-Andalusi, omitted the story of the declaration and the Damascus assembly: they found Ibn TQlQn' s declaration of jihad unorthodox, or simply distasteful. Thus, whatever Ibn TQlQn had in mind here, it was ultimately misunderstood or rejected. To put it differently, the doctrine of jihad, and the practices associated with it, all became identified more rigidly in the following generations; in Ibn Tulun's day there was more latitude in this matter.

The kitab al-khal( - or perhaps the pair of documents 115 - was broadcast widely, and thus may fairly be described as a piece of propaganda. But, of course, the Tulunid propaganda machine had other elements, in addition to the chancery. In particular, legions of poets clustered at the palace in Fustat, though none of these was remembered long afterward. 116 Two of their compositions for this occasion have survived, plus a fragment of a third, preserved in al-Kindi's Governors of Egypt. The first is by Qacdan b. cAmr, an otherwise unknown poet with a connection to the Tulunid court.117

Divine guidance is long-lasting through the amir Ibn Tulun, just as

Religion flourishes through him, even beyond [normal] religion and Islam.

He has led the armies from Fustat, at whose head,

On his behalf, is one118 who proceeds fearlessly, never flinching,

Accompanied by a host [gathered against] the Fates, its cavalrymen riding

In squadrons, concealed among banners and standards.

In it are whites, lofty nobles from the sons of Shem,

And blacks, lions from the sons of Ham.

Even if the spirit of the Sons of Kundaj were suspended

From Jupiter or Mars, he still would not escape him. 119

Our caliph has encompassed the caliphate and the world

With the mightiest of God's sharp-edged swords.

O you people! Hasten to aid him,120

And join up with the amir and his dusky mounts, at [the sign of] the lam. 121

When you pray, your prayer is not valid,

Nor is your fasting acceptable when you fast,

Until the blessed chief sees you defending the imam

With the points of your bloody lances.

The poem begins as a panegyric (madlh) of Ibn Tulun, with tropes conventionally used in praise of caliphs. Then comes a martial setting, as Qacdan provides an image of the army on the march, though not of the amir himself. In the last four verses, the focus shifts to the quarrel with Ibn Kundaj and then to the caliph al-Muctamid. The poet includes no juridical language regarding the obligation of jihad, but summons everyone (presumably both civilians and soldiers) to the fight. And just as the kitab al-khal( described al-Muctamid's sufferings while calling for redress and revenge, here again the fight is on behalf of the caliph or imam. Everyday religious observance - prayer and fasting - is invalid so long as this caliph/imam is held captive. The remedy is for everyone to fight under the banner of Ibn Tulun, through whom "religion flourishes, [even] beyond the [normal] requirements of religion and Islam."

These two claims - that jihad must be conducted to defend the person of the imam, and that Ibn Tulun has the ability, through his jihad, to restore the world to its normal condition - may have struck contemporaries as bizarre, though we do not know for certain. We can surmise, however, that later generations found these ideas strange or repugnant, and we can see a hint of this in the poem's transmission history. In the British Museum manuscript that Guest used for his edition of al-Kindi's Book of Governors, the final verse reads:

Until the blessed chief [Ibn Tulun] sees you defending Islam

With the points of your bloody lances.

hatta yarn l-sayyidu 1-maymUnu dhabbakumu

'ani l-islami bi-atrafi l-qana l-dami

This verse contains a basic error in its meter. Guest, the editor, suggested in a note that the correct reading might be dhabbakumu I (ani l-imami "your defending the imam." This conjecture must be correct, because it fits both the meter and the sense. What happened, then, was that at some point this text fell into the hands of a reader (and/or scribe) who found this verse outlandish - too Shici? - and proceeded to correct it, substituting "Islam" for "imam," even though this resulted in an elementary metrical error. Jihad, in this reader's mind, was conducted for God and Islam. It may be conducted at the behest of the caliph or imam, but not for his sake. Over time, this little episode of textual editing was echoed, in a variety of formats and venues, when what I am calling "Ibn Tulun's jihad" became lost to historical memory.

The second composition is by the equally obscure Munsif b. Khalifa al-Hudhali, who clearly had ties to the Tulunid court and army.122

O most excellent man of this world, you whose deeds

Set outstanding examples to which all humanity adheres,

You are the amir over Syria and its frontier district (thaghr)

And the two Raqqas 123 and whatever the East encompasses.

For you, Egypt and Barqa, and [their] Hijaz, 124

For you the heart of everyone is longing.

An upstart125 has murdered the caliphate, with his crony Ishaq,

Showing contempt - and there is no greater fool than an envious man.

Our swords are death-bringing silver blades; would only that they

Be perfumed with the blood of those who have betrayed the imam.

Night and day, you126 strike blows in his defense,

With a well-honed [sword], from which destruction and death are dispersed.

Your followers include Sacd and the commander Taytak,

And al-Ladhiqi, with DhO 1-Hafiza127 tagging along.

This poem also calls for warfare on behalf of the caliph, whose chief defender, and the guarantor of stability, is once again Ibn Tulun. Here the poet speaks to army regulars (mentioning, affectionately and perhaps humorously, their names and nicknames), and does not call upon everyone (i.e., civilians) to join the struggle. However, the poet evokes Ibn Tulun's vast domains and claims that within them "everyone's heart longs for you."

These two poems, 128 together with the declaration, constitute the sum total of our firsthand evidence for Ibn Tulun's propaganda in this contest, or, if you like, for his jihad. None of this is recondite or obscure. The emphasis is, first of all, on the sacred person of the caliph or imam, whose captivity violates the order of the world. There is also emphasis on loyalty to the caliph's rescuer and avenger, the amir Ibn Tulun, whose actions restore order and allow religious observance to become valid once again. The message is pitched to Ibn Tulun's professional soldiers, although others are welcome to join. These soldiers are a diverse group, which includes Central Asian Turkish and European whites as well as African blacks. They show to Ibn Tulun the same loyalty that he shows to his master (and protg) the caliph. Their struggle may fairly be described as jihad (though of a rather Shici-flavored variety), because its purpose is to restore Islam and its order in the world, by defending and avenging the imam. Entirely lacking, however, are the adversaries that we expect to encounter in situations of jihad. The Byzantine emperor, the "tyrant of the Romans" (taghiyat al-Rum), never appears, but neither, for all that, does the Muslim heretic or rebel. This figure was reserved for Ibn Tulun himself, when his adversaries responded with their counter-propaganda.

OPPOSITION

During and immediately after the Damascus assembly, Ibn Tulun faced opposition from three quarters: his main adversary al-Muwaffaq; a few jurists, led by Bakkr b. Qutayba, who defied him in Syria and then in Egypt; and the residents of Tarsus and the Thughur, who had their own ideas about what constituted jihad and its proper performance.

1. al-Muwaffaq. Al-Balawi gives us the text of al-Muwaffaq's reply to Ibn Tulun, including an "anathema" (barala). m This text speaks in the name of the captive caliph al-Muctamid, but everyone can see al-Muwaffaq's agency in all this.

When al-Muwaffaq learned that Ibn Tulun was omitting his [al-Muwaffaq's] name and had stopped offering the prayer on his behalf, he ordered that he be cursed from the pulpits. A document of anathema (bara'a) was issued, containing [this] curse, to all the cities. Its text read as follows: "God, may He be exalted, has joined obethence to Himself together with obethence to His Messenger and to those who hold authority. . . . Now the enemy of God, who has deviated from the community of the Muslims (al-mubayin li-jamacat al-muslimin), the man known as Ahmad ibn Tulun, has openly committed acts of rebellion and heresy (azhara ma kana min macsiyya wa-shiqaq) everywhere between the regions of the farthest West and the [near] flanks of Iraq. He has renounced the true religion and has opposed the Commander of the Faithful. He has devastated the frontier regions of the Muslims (wa-akhraba thughur al-muslimin) and has fought the fighters (mujahidun) in those regions, in concert with [his own] corrupt, heretical people (maca ahi al-fisq al-mulhidin), treating their [the mujahidun's] womenfolk as licit, and shedding their blood. Accordingly, when the Commander of the Faithful ascertained the truth of the matter [concerning Ibn Tulun], and knew of his disbelief, he dissociated himself from him before Almighty God, and ordered that he be cursed. . . ."130

Al-Muwaffaq does not reply directly to Ibn Tulun's declaration of jihad against him, nor does he challenge Ibn Tulun's authority to make such a declaration. Instead he brands him as a rebel performing his nefarious work with corrupt and heretical confederates (maca ahi alfisq al-mulhidin). On this basis, we might expect al-Muwaffaq to proceed to declare his own jihad against Ibn Tulun. However, he does not, perhaps because such a move would lay him open to the charge to which Ibn Tulun is already vulnerable, namely, his lack of authority for such an action.131 At this point it is also within al-Muwaffaq's power to obtain a declaration of jihad from the hapless caliph who, after all, has just been compelled to curse his own protector and ally. 132 And yet he does not.

What al-Muwaffaq does instead is to portray the encounter with Ibn Tulun as a drama of the frontiers. Ibn Tulun has been active "everywhere between the farthest regions of the West" to the borders of Iraq, the heart of the empire. He and his confederates have laid waste to the frontier regions, violating the women and killing the men. Of course, this is rhetorical invention, but the reference to the frontier and its mujahidun is effective: we need only recall Ibn Tulun's difficulties with Tarsus and the hatred that he and his men inspired there.

Al-Muwaffaq was no doubt the originator of this response to Ibn Tulun, but the cAbbasid establishment had its part in its planning and orchestration. And like its Tulunid counterpart, this cAbbasid establishment included a legion of court poets. Here the cAbbasids had an edge over their rivals: instead of obscure poets whom we cannot identify nowadays, they called upon the famous al-Walid al-Buhturi (d. 284/897), an experienced veteran of the propaganda wars, as well as upon some lesser lights.133

2. The Egyptian jurists. If Ibn Tulun's declaration of jihad ever included a coherent line of juridical reasoning, we cannot see it now. It is likely that one of his motivations for convening the assembly was to develop precisely such an argument. But as so often happens, things did not go according to plan. Instead, Ibn Tulun encountered dogged opposition from the man who, more than anyone else, represented Egypt's learned, religious elite at the time. This was the chief qadi of Egypt, Bakkar b. Qutayba, who, along with two other Egyptian judges (of whom we know), refused to assent to the document of deposition (or at least to all of it).

Bakkar was a highly respected jurist who came originally from Basra, in Iraq. 134 A member of the Hanafi school of law, he received his appointment as qadi of Egypt from the caliph al-Mutawakkil in 246, and spent the rest of his long life in that office. As Ibn Tulun established himself in Egypt, he went out of his way to show respect to Bakkar, occasionally attending his public recitations of hadlth. In 266, when Ibn Tulun confronted the revolt of his son al-cAbbas, he sent Bakkar to him as a personal envoy. Even though Bakkar showed some independence on this occasion, he later came down in favor of Ibn Tulun, by ruling that the aman (safe-conduct) that had been offered to al-cAbbas's companions, the supposed fomenters of the rebellion, was not valid. In this way Ibn Tulun had the green light for putting these unfortunates to death, despite the promises that had been made to them.135

Bakkar's opposition to Ibn Tulun began at Damascus in 269 and then continued in Fustat, in what turned out to be the final year in the life of both adversaries. According to most accounts (though with variations in the details), Ibn Tulun had Bakkar exposed 136 and paraded. He was then held under house arrest for a time, before being thrown into the infamous matbaq or dungeon. Various charges were brought against Bakkar, some in relation to his judicial decisions and some in relation to his personal conduct, but he defended himself vigorously and refused to yield. By some accounts, though not all, he died in prison, soon after or before the death of Ibn Tulun himself. 137 Throughout all this, Bakkar refused to assent to the deposition of al-Muwaffaq because, as he put it, he saw no written authorization from the caliph for this act. Ibn Tulun replied that al-Muctamid could not provide any such proof (hujja) because he was confined and under constraint. Ibn Tulun then went to considerable trouble to demonstrate the precise meaning of confinement and constraint to Bakkar, but to no avail.138

We have the names of two other jurists who, like Bakkar, did not assent to Ibn Tulun's document (whether in its entirety or in part). These were Ibn al-Mawwaz (Muhammad b. Ibrahim b. Ziyad al-Iskandarani), a leading Maliki jurist in Egypt at the time,139 and Fahd b. Musa b. Abi Rabah al-Iskandarani, qadi of Alexandria, also an affiliate of the Maliki school. 140 There is no record that any of these three ever mentioned jihad during these arguments. Accordingly, it does not seem that the refusal of Bakkar, a Hanafi, and Ibn al-Mawwaz, a leading Maliki, to assent to the deposition of al-Muwaffaq had anything to do with those schools' positions with regard to jihad. Indeed, Christopher Melchert has shown that the jurists' opposition to Ibn Tulun did not break down along these lines: the majority who subscribed, like the minority who refused, were all of different madhahib.141

It is likely that when Ibn Tulun convened the assembly at Damascus, he wished to obtain the backing of the jurists and men of religion. But here, as in his other projects at the end of his life, he failed.

3. Tarsus and the Thughur. Throughout his career, Ibn Tulun paid close attention to the frontier region abutting the Byzantine Empire. We have seen this in the encomiastic poems composed for him during his last expedition to Syria, in hagiographical accounts composed later (the two Siras), and even in historical accounts that are more neutral in tone. The failure of these efforts did not prevent his son and successor, Khumarawayh, from following a similar policy. 142 Here we will look again briefly at the internal situation of the Thughur at this time, and especially its capital city, Tarsus.

Our knowledge of Tarsus and the Thughur in these years is both varied and limited. The historical chronicles (Arabic, Greek, and Syriac) allow us to reconstruct the broad outlines of political and military events. For internal conditions, we have biographical information on many individuals who lived there, whether as permanent residents or temporary visitors, during the two and a half centuries in which the Thughur comprised a distinct region or province of the caliphate. 143 These biographical sources tell us about the individual merits, accomplishments and quirks of scholars and ascetics, but show little interest in the political and military circumstances of life in the Thughur. This is true even for those individuals who held the highly politicized position of qadi of the Thughur. 144 As a result, the internal politics of the place, its habits and patterns regarding leadership, remain obscure to us. In a number of ways, however, the internal life of the Thughur and, in particular, Tarsus seems to have been different from that of other cities and provinces under Tulunid rule.

Before Ibn Tulun arrived as governor or viceroy of the region, the caliphal authorities in S amarra' had only a semblance of control over it. Governors or commanders such as Sima al-Tawil and Arjun operated basically on their own. When Ibn Tulun entered the scene, he enjoyed the loyalty of Tukhshi and Khalaf al-Farghani, but not for long. Then came the defection of Lu5Iu5 and the rise of Yazman, who was beholden neither to the caliphal government nor to Ibn Tulun. We have seen that Ibn Tulun tried mightily to oust Yazman but failed. Yazman's subsequent career marks the beginning of an increased independence for the frontier region, together with a renewed intensity of the wars against Byzantium.145

Most striking in the history of the Thughur in these years is the readiness of the "people of Tarsus" to defy their political leadership, or any political leadership whatsoever. When al-cUmari, chief qadi for northern Syria and the Thughur, approved of the deposition of alMuwaffaq, he clearly was not speaking directly for the residents of the frontier. For all that, we do not know if the Tarssis favored either al-Muwaffaq or Ibn Tulun in their quarrel; what we do know is that they freed Yazman, an outsider, from Khalaf al-Farghani's prison and obstinately defended him, together with their city, from Ibn Tulun's army. All in all, among the many political and military opponents who faced Ibn Tulun over the years, there were none quite like the Tarssis. Despite Ibn Tulun's fierce temper and his well-attested willingness to throw political adversaries into prison, he yielded meekly when the Tarssis resisted him.146

The Arabic sources that describe the Thughur during these years occasionally mention arzaq ('provisions, grants').147 When Ibn Tln as a young man received his assignment to the frontier from the vizier, he was allocated arzaq (or its singular, rizq), presumably a kind of salary. 148 When Arjun b. Awlagh Tarkhun became governor of the Thughur on behalf of al-Muwaffaq and the caliphal government, it was a sign of his arrogance that he delayed or confiscated people's arzaq. 149 Later, when Lu5Iu5 took over the region on behalf of Ibn Tulun, he repaired or restored these same arzaq. 15 More needs to be done to establish what was going on here. However, these arzaq are descended from the stipend (fata*) that used to be granted to fighters in the early Islamic armies. By contrast, in Ibn Tulun's Egypt and Syria, the armies consisted largely of ghilman, slave-soldiers coming from the Turkish steppes and black Africa. The Thughur, however, brought together not only old-fashioned army regulars (jund), sustained from local tax revenues,151 but also volunteers (muttawwi(a, mujahidun) who arrived from the entire Islamic world and received support from waqfs, pious foundations established for this purpose in their countries of origin.152 The frontier was thus, on the one hand, a place that stuck - symbolically and even really - to an obsolete military and social order. On the other hand, it was a place where these different groups could enter into new and strange alliances. Thus we have Yazman, apparently the leader of a group of ghilman, slave-soldiers, joined by the Tarsusis - apparently both army regulars and civilian volunteers - all of them united in their rejection of Ibn Tulun and his granthose claims.

CONCLUSION

If jihad is the ideology of the frontier, then Ibn Tulun's invocation of it, throughout his career and especially at its end, may seem like an attempt at restoring an obsolete, antiquated order of society. Such an attempt, of course, was bound to fail. At the same time, however, Ibn Tulun's jihad represents something new in the political language and landscape of the time. My conclusions regarding this adventure fall under four headings.

1. A political project. Ibn Tulun's campaign of 269/883 was part of a longer sequence of events. The critical moment actually came four years earlier, when he seized control of Syria and Palestine without any appointment or permission from the cAbbasid government. This action, though technically illegal, made geopolitical and strategic sense. After all, the rulers of Egypt, from the Old Kingdom to the Free Officers, have always concentrated on their most vulnerable point, which is their northeastern flank, and have always tried to exert influence or control in Syria and Palestine.

By moving unilaterally into Syria in 265, Ibn Tulun guaranteed that his relations with al-Muwaffaq would deteriorate even further than they had already done. Soon Ibn TOlOn was withholding revenue from the cAbbasid treasury, and an armed clash between cAbbasid Iraq and Tulunid Egypt seemed on the horizon. However, Ibn Tulun in 265 did not do what we might now consider the rational thing, namely, to consolidate his position in Syria and to prepare for conflict with the cAbbasids. Instead he went to Tarsus, capital of the frontier district, where he actually had a caliphal appointment. Once he arrived there, however, Ibn Tulun accomplished little and then allowed the residents of the place to kick him out in an especially humiliating way. Why did he do this? One answer comes from the Siras: he wanted to participate in the region's activities, which included extreme ascetic practices and the transmission of religious learning, and also, somewhat marginally, warfare against the Byzantine enemy. But beyond this, the Arab-Byzantine frontier was not only the object of Ibn Tulun's pious desire, but the focus of his political efforts, a conscious, deliberate choice.

As the contest between Ibn Tulun and al-Muwaffaq heated up in 269/883, it was more than a war of words. During the pilgrimage season, rival forces from Egypt and Iraq clashed in a pitched battle at Mecca. The Tulunid side had the worst of it, suffering 200 dead. 153 Fighting also took place in northern Syria between Ibn Kundaj and pro-Tulunid forces. 154 All this raises a question posed by some earlier scholars: why did Ibn Tulun not simply march his powerful army against al-Muwaffaq and settle the quarrel once and for all? 155 This, of course, would have ignited a civil war. But again, Ibn Tulun desired Tarsus, not S amarra5 and Baghdad. Occupation of the frontier, and residence in it, were goals he shared not only with his followers in the military elite, but also - or so he hoped - with the civilian elite that monopolized the transmission of religious learning and the holy law. At Damascus, Ibn Tulun attempted to make this goal the basis of a shared political project. However, his heavy-handed tactics alienated much of his authence. Then the opposition of a minority of the jurists helped to doom this political project.

2. Jihad, with and without caliphs. How could Ibn Tulun, a mere delegate or appointee of the caliph, presume to declare jihad in 269/883? Early Islamic legal texts that dealt with these matters recognized the conduct of warfare against external enemies as a duty and prerogative of the caliph or imam, and many historical episodes reported in the chronicles show that this principle was observed in reality. However, while these lawbooks speak of the imam's duty to conduct expeditions and to organize defense against enemy attack, they do not discuss "declaration of jihad" as such, and certainly do not envisage the sort of project that Ibn Tulun tried to accomplish at Damascus in 269.

What Ibn Tulun did there was to seek the approval of the experts in this matter. The very act of convening such an assembly was unusual, as we have seen. When the assembly took place, however, the jurists did not wish to enter this territory. They dealt with the question at hand - the deposition of al-Muwaffaq - on its own while studiously avoiding this matter of declaring jihad. Their reluctance is understandable, because if they had endorsed Ibn Tulun's call for jihad against al-Muwaffaq - as they endorsed his call for removing al-Muwaffaq from the succession - they would have declared themselves, as a body, the competent legal authority in this matter. After all, the "document of deposition" does not identify anyone, including Ibn Tulun himself, as competent to make this declaration, other than the undersigning jurists.

We have seen that another characteristic of Ibn Tulun's jihad is that it was conducted on behalf of the caliph or imam. The document of deposition, together with the poems by Qacdan and Munsif, describe al-Muctamid's sufferings at some length, expecting its hearers to take these sufferings to heart. Here we need to remember that the notion of jihad conducted in defense of the caliph/imam had lots of resonance in early Islamic historiography. An example is The Battle ofSiffin by Nasr b. Muzahim al-Minqari (d. 212/827). This book describes the battle between Mucawiya and ??? at Siffin on the upper Euphrates in 37/657. Here, as war begins to brew, Mucawiya approaches his future ally, cAmr b. al-cAs:

Mucawiya said: "Abu cAbd Allah, I summon you to fight against [or: to perform jihad against] this man, who has rebelled against his Lord, killed the caliph [cUthman], fomented dissension, split the community, and sundered his ties [of kinship and loyalty]" (inniadcuka ila jihad hadha l-rajul alladhl casa rabbahu wa-qatala l-khalifa wa-azhara l-fitna wa-faraqa l-jamaca wa-qataca l-rahim). cAmr asked: "Against whom?" Mucawiya answered: "To perform jihad against cAli" (ila jihad cAli). cAmr then said, "Mucawiya, you and cAli are not equal in status. 156 You don't have his hijra [emigration with the Prophet], nor his precedence [in Islam], nor his companionship [with the Prophet], nor his jihad" (ma laka hijratuhu wa-la sabiqatuhu wa-la suhbatuhu wa-la jihaduhu).157

Other versions of the same story then follow, which make no mention of jihad. In this particular narration, however, Mucawiya's idea resembles Ibn Tulun's: jihad is something that we exercise against the flagrantly disloyal, the sunderers of ties, and the killers (or oppressors) of caliphs. What cAmr means here by jihad is not immediately clear.

Otherwise, when we find the word jihd in narratives about Siffin, it is in the mouths of cAli's partisans, as when they demand that he let them loose against Mucawiya's men, because they "prefer jihad against them over jihad against the Turks or Byzantines." 158 Nasr b. Muzahim has been called an early Shici historian, and while not everyone agrees with this, it is fair to describe him as friendly to the cAlid cause. 159 The point here, however, is that we can identify, in early Islamic historical writing, a jihad that involves warfare for the sacred person of the imam. In a very broad sense we might speak of this sort of jihad as "Shici," keeping in mind, however, that it need not have been practiced or advocated exclusively by actual Shicis. More specifically, we have no indication that Ibn Tulun had any particular fondness for the Shica and its ideas. In his day, however, a certain kind of jihad was still imaginable, which to later generations of Sunnis appeared unacceptable, probably because of its "Shici" character. This, of course, provides one more reason why, with the passage of time, Sunni historians ignored Ibn Tulun's assembly and his declaration of jihad.

3. Historiography. We have seen that the reports of the assembly and declaration in al-Balawi and al-Maqrizi (and possibly also in al-Kindi) are based on a longish document that Ibn al-Daya probably copied from an original in the Tulunid archive. From this I have concluded that the assembly did happen, along the lines that these writers described. However, we have also seen that the majority of Arabic historians, not only those who came after al-Kindi and al-Balawi, but several who were their contemporaries or even earlier, omitted this episode, whether by choice or through lack of awareness. This shows, not surprisingly, that for those historians, the "documentary" quality of a report did not have the same power to convince as it does for us.

Reports of the assembly and declaration were suppressed or went unnoticed for other reasons as well. In the fourth/tenth century, Islamic historical writing was still - though decreasingly - based primarily in Iraq. Historians from the "mainstream," such as al-Tabari, reported these events - to which they were nearly contemporary - using information available to them in Baghdad, much of it purveyed by the authorities. So, for example, "mainstream" historical writing reports the declaration issued in Iraq against Ibn Tulun, but not the declaration that Ibn Tulun issued in Syria against al-Muwaffaq. But, of course, this "mainstream" was not everything. Egypt already had its own school of historical writing,160 an enterprise that, like other branches of learning and the arts, received a jolt from the presence and patronage of the Tulunid dynasty. It is still difficult to tell precisely how other Egyptian historical writers of the time, apart from al-Kindi, Ibn al-Daya, and al-Balawi, viewed these matters, but these three are enough to give us an idea.

With the passing of time we can detect a process of editing and revising, in both Egyptian and non-Egyptian historical writing. In retrospect, the most important moment turns out to be when Ibn Sacid al-Andalusi takes Ibn al-Daya's biography of Ibn Tulun, composed around three centuries earlier, and incorporates it into his al-Mughrib, omitting the story of the assembly and the declaration. We do not know Ibn Sard's reasons for this, but we can surmise that they included the Shici flavor (at least in retrospect) of Tulunid jihad; what must have appeared to be Ibn Tulun's misunderstanding of the doctrine; and the unfavorable light that this entire story casts on Ibn Tulun. After all, the story pits an obsessive (not to say paranoid) tyrant against the pious Bakkr b. Qutayba, who becomes a martyr by, as we say now, "speaking truth to power." Syro-Egyptian historians of the Mamluk era may or may not have viewed Ibn Tulun as a founding figure and political exemplar, but in any case this story seems to have made them uncomfortable. 161 Thus, while the story remained available to hard-working scholars such as al-Maqrizi, for the most part it passed out of historical memory.

4. The frontier. A generation before al-Maqrizi, when Ibn Khaldun wrote a prcis of Tulunid history as part of his world history, the Kitab allibar, he relied on Ibn Sacid's redaction of Ibn al-Daya. 162 Ibn Khaldun probably composed this in the Maghrib, years before he took up residence in Egypt. But one way or another he did not relate the story of the assembly and the declaration. Perhaps he would have had he had more information about it. Far more interesting is what Ibn Khaldun actually does. More than any historian before or since, he considers Tulunid history as a tale of the Arab-Byzantine frontier. He often has more to say, and certainly does more extensive research, 163 about Tarsus and the Thughur than he does about Egypt, with the possible exception of its capital Fustat/al-QataV.

Broadly speaking, the Tulunids fit in with Ibn Khaldun's view of history, which we (think we) know from his Muqaddima. 164 This is not only because they follow the familiar pattern of descending, over three or four generations, from vigorous conquest and state-building into drunken decadence and political collapse. Ibn Tulun, the founder, arrives in Egypt at a moment when the first great expansion of Islam is flickering and sputtering. Yet the original impulse toward expansion is not dead, and the Islamic empire has not collapsed. New actors have arrived, especially Turks from Central Asia, bringing the "group feeling" ((asabiyya) of tribal peoples from the steppe. How Ibn Tulun creates and maintains such loyalty among his armed men, many of whom are black Africans, is at the heart of the story, and perhaps still not understood. Meanwhile, Ibn Tulun adheres to the scenario of the Muqaddima, by taking over a flourishing metropolis (Fustat) and building himself a new capital (al-Qata'i0) nearby. Like other state-building dynasts in the Muqaddima, he faces a problem from his close relatives, who feel entitled to a share of his power. And like those other new dynasts, he tries to solve the problem by sending his own (remarkably dysfunctional) family off to the frontiers, or by allowing them to migrate there: thus we have the movement of his brother Musa to Tarsus, and of his son, the doomed al-c Abbas, to Barqa (west of Alexandria). In short, Ibn Khaldun sees the frontier as the crucible of the Tulunid dawla, or dynastic state.

It is generally recognized that Ibn Tulun's main accomplishment, in the broad scale of things, was to enable Egypt to emerge as a major political and economic power, with a distinctive cultural voice of its own. It therefore seems strange that he should have directed so much energy and effort toward the remote, fractious district of the Thughur. However, Ibn Tulun did this, and in the decades after his death, the Thughur remained a bone of contention among the caliphate and several of its Muslim rivals, including the later Tulunids.

For Ibn Tulun the Thughur proved an unobtainable object of desire, just as his declaration of jihad against al-Muwaffaq came to naught. In a larger perspective, however, the venture made sense. The cAbbasid caliphate had recently been in a shambles and was now recovering its position with great difficulty. For Egypt, subordination to Iraq meant high costs and ineffective government, at once less desirable and more avoidable than before. However, if Ibn Tulun looked about for a model to emulate, he did not choose the example of the earlier, powerful, autonomous rulers of Egypt, about whom he probably knew very little and who had not been Muslims in any case. Instead, what he had before him was the example of his contemporaries the Saffarids, who at the time dominated most of the eastern Islamic lands.

Deborah Tor has argued that Yacqub the Coppersmith, and the Saffarid dynasty as a whole, lived and breathed for one purpose only, which was performance of holy war against infidels and heretics. 165 In this view, the bands of Sistani (ayyarun ('toughs') were really muttawwi(a and ghazis, volunteer warriors for the faith; by remaining true to the ascetic ideals of these bands, Yacqub and his successor, his brother cAmr, won justification for their wars of conquest. In particular, Tor says, they won the support of religious learned groups in the cities in this way. This view of the Saffarids provides welcome relief from the views that have often prevailed of them, as well as of other amirs of the "Eastern dynasties," as either overambitious soldiers of fortune or as Iranian patriots and nationalists avant la lettre. On the other hand, it may be too much to ascribe the entire Saffarid enterprise to a single motivating principle of holy war. 166 Be that as it may, we can see in the rise and expansion of the Saffarids a drama of state formation, with stage and backdrop provided by the eastern frontier. Closer to home, Ibn Tulun had the example of the Aghlabids of North Africa, with their ongoing project of conquest in Sicily and continental Italy.

Seen from this perspective of the frontiers, the experience of the Tulunids in Egypt was also quite different from that of the Saffarids in Iran and the Aghlabids in North Africa. For one thing, the Tulunid scenario allowed only a limited role for civilian roughnecks (cayyarun) or volunteers for the faith (mujahidun, muttawwica, etc.). Moreover, right up until Ibn Tulun's arrival, the frontiers of Islamic Egypt had been relatively close to home, in Nubia in the south and Ifriqiya in the west. 167 Some may even have thought of early Islamic Egypt as a kind of frontier province in itself, with its urban, Arab Muslim elite surrounded by a rural, Coptic Christian majority. 168 But now, if Ibn Tulun wanted to emulate the political success of the Saffarids and Aghlabids, he needed to operate in a frontier zone where armed, non-Muslim adversaries were eager for combat and where new territories were open, at least in theory, for conquest and occupation. Accordingly, Ibn Tulun staked his claim to the Thughur of northern Syria and southern Anatolia, the frontier of Islam par excellence. He did this for general reasons of legitimacy and prestige, as everyone could see then and now. More specifically, a viable, vigorous dynastic state operating on the fringes of the crisis-ridden 'Abbasid caliphate had no choice but to be, very literally, an enterprise of frontier and jihad.

1. I delivered an earlier version of this paper at the AOS meeting in Albuquerque in March 2009, and I am grateful to the authence there for several ideas and suggestions. In particular, I wish to thank Matthew Gordon, author of a forthcoming biography of Ibn Tulun (see below, n. 2), for his detailed, expert comments; Peri Bearman, JAOS Editor for the Islamic Near East, for her advice and encouragement; an anonymous reader for the Journal; Paul Cobb; and Daniela Gobetti.

2. The following summary of the episode is based on Ibn al-Daya's biography of Ahmad b. Tulun, as preserved in Ibn Sacid al- Andalusi, al-Mughrib fi hula l-Maghrib, in two editions: ed. K. Vllers, in Semitistische Stuthen 1 (Erganzungsheft zur Zeitschrift fur Assyriologie, Berlin, 1894), henceforth ISV; and ed. Z. M. Hasan (Cairo: Matbacat Jamicat Fu'ad al-Awwal, 1953), vol. 1, henceforth ISH; in addition al-Balawi, Sirat Ahmad ibn Tulun, ed. M. Kurd cAli (Damascus: Matba'at al-Taraqqi, 1939); al-Kindi, The Governors and Judges of Egypt, ed. R. Guest O^ondon and Leiden: Luzon and E. J. Brill, 1912); al-Tabari, Ta>rikh al-rusul wa-l-muluk, ed. M. J. de Goeje and others (Leiden: E. J. Brill, 1879-1901); and Ibn al-cAdim, Bughyat al-talabfi ta'rikh Halab, ed. S. Zakkar (Damascus, 1988). For modern discussions of Ibn Tulun and the Tulunid dynasty, see Matthew S. Gordon, "Tulunids," EP, 10: 616-18; idem, The Breaking of a Thousand Swords: A History of the Turkish Military of Samarra0 (Albany: State Univ. of New York Press, 2001), esp. 62-70, 116-18; T. Bianquis, "Autonomous Egypt from Ibn Tulun to Kafur," in Cambridge History of Egypt, vol. 1, ed. C. F. Petry (Cambridge: Cambridge Univ. Press, 1998), 86-119; O. Grabar, The Coinage of the Tulunids (New York: American Numismatic Society, 1957); Z. M. Hassan (Hasan), Les Tulunides: Etudes de l'Egypte musulmane iafin du IXe sicle, 868-905 (Paris: Busson, 1933). Matthew Gordon is now preparing a biography of Ibn Tulun for the Makers of the Muslim World series, to appear with Oneworld Publications.

3. According to al-Kindi, 225, this happened while Ibn Tulun was already en route.

4. Al-Muwaffaq was actually second in succession, after al-Muctamid's son, Jacfar al-Mufawwad, in an arrangement dating back to 261/875. However, everyone could see that if al-Muctamid should predecease al-Muwaffaq, the latter would have no difficulty in pushing al-Mufawwad aside. In the event, al-Muctamid and al-Muwaffaq died nearly simultaneously, in 278-79/891-92, and al-Muwaffaq's son Abu l-cAbbas al-Muctadid quickly removed his cousin al-Mufawwad and assumed the caliphate. See al-Tabari, Ta'rikh, 3: 2119-23, 2131-33; H. Kennedy, arts. "al-Muctamid cala 'llah," EP, 7: 765-66; "al-Muwaffaq," EP, 7: 801.

5. C E. Bosworth, The History of the Saffarids ofSistn and the Maliks ofNimruz (Costa Mesa and New York: Mazda Publishers, 1994); idem, "Saffarids," EP, 8: 795-98.

6. A. Popovic, The Revolt of African Slaves in the 3rd/9th Century (Princeton: Markus Wiener, 1998); idem, "al-Zandj," EI^sup 2^, 11: 444-46.

7. Ibn al-Daya, at ISV, 69 and ISH, 127; al-Balawi, 281.

8. al-Tabari, 3: 2137-39. On Ishq b. Kundj, see al-Tabari, 3:1877, 1879-80, 1930-31, 1939, 1992, 2037-49, 2137; on Sacid b. Makhlad, idem, 3: 1930, 1988, 201 1, 2037-40, 2048-49, 2079-80, 2083, 2086, 2106, 2122, 2144, 2146.

9. al-Kindi, 225.

10. al-Balawi, 296-97. The Jazira, or upper Mesopotamia, corresponds in modern terms to the borderlands shared by Iraq, Syria, and Turkey, where the upper reaches of the Tigris and Euphrates form an "island" (jazira).

11. al-Tabari, 3: 2048; Ibn al-cAdim, 2: 834.

12. Bianquis ("Autonomous Egypt," 101-2) says that comparable assemblies had met in Damascus during the reign of Mu'awiya (41-60/661-680), when there were consultations between the caliph and his advisors on the occasion of the weekly jumca. I have not found reference to this, though see Hassan, Les Tulunides, 85. In any case, it seems unlikely that memories of those sessions would still have been alive two centuries later.

13. Abu 'Ubayd al-Qasim b. Sallam, Kitab al-Amwal (Beirut: Dar al-Kutub al-Tlmiyya, 1406/1986), 184-90, nos. 468-76; al-Baladhuri, Futuh al-buldn (Leiden: E. J. Brill, 1866), 155-58.

14. Miskawayh, Tajarib al-umam, ed. A. Emami (Tehran: Soroush Press, 2001), 5: 137-39.

15. See, however, al-Ma'mun's statement in his third letter at the outset of the Mihna, in al-Tabari, 3: 1129, trans. C. E. Bosworth (The History of al-Tabari, vol. 32 [Albany: State Univ. of New York Press, 1987], 218): "If the Commander of the Faithful had sought legal justification in combatting [certain named adversaries] for the sake of God and launching a holy war against them solely on the grounds of their taking usury and that which has been revealed in the Quran regarding their likes, he would have found it lawful. How will it be with them now, when they have added polytheism to their practice of usury and have become just like the Christians?" Here al-Ma'mun is willing to take expert advice, while reserving final say in the matter for himself.

16. Bianquis ("Autonomous Egypt," 102) cites the denunciation of the Faumids by al-Qadir (422-467/10311075) and the efforts by al-Nasir (575-622/1180-1225) to conciliate jurists and men of religion. This question may take on a broader dimension if we consider it to include the mazalim sessions. A vivid example is the session over which al-Muhtadi presided, apparently to denounce the conduct of a Turkish officer, and apparently with jurists in attendance. See al-Sl, Awraq, ed. and trans. V. I. Beliaev and A. B. Khalidov (St. Petersburg: Tsentr Peterburgskoe Vostokovedeniia, 1998), 366-67 (no. 465). I owe special thanks to Matthew Gordon for this idea and this reference.

17. Franz-Christoph Muth, "'Entsetze' Kalifen: Depositionsverfahren im mittelalterlichen Islam," Der Islam 75 (1998): 104-23.

18. See above, n. 4.

19. Bianquis, "Autonomous Egypt," 101-2.

20. M. Bonner, Jihad in Islamic History (Princeton: Princeton Univ. Press, 2006, 2008).

21. 18,000 people (!) are said to have died in Ibn Tulun's prisons (Ibn al-cAdim, Bughya, 2: 829).

22. Yaqut, Mucjam al-udaba' (Cairo: Maktabat cIsa al-Babi al-Halabi, 1936-1938), 5: 154-60, calling this book Sirat Ahmad b. Tulun. See also al-Safadi, al-Wafifi l-wafayat (Istanbul: Matbacat al-Dawla, etc., 1933-1998), 8: 282-83; Rosenthal, "Ibn al-Daya," EP, 3: 745-46. The name Ibn al-Daya ('son of the wet-nurse') belonged originally to his father Yusuf, a foster-brother of either the caliph al-Muctasim or the (counter-) caliph Ibrahim b. al-Mahdi.

23. Yaqut, Mu'jam al-udaba', 5: 155-57. Earlier on, Muhammad b. cAbd Allah had been carried off to Baghdad for questioning regarding the createdness of the Qur'an; he refused to answer and was sent back to Egypt, apparently unscathed. He also had another near-escape from Ibn Tulun's wrath, see al-Safadi, Wfi, 3: 338-39.

24. When the elder Ibn al-Daya (Yusuf) died, Ibn Tulun confiscated his books and writings and hauled in his two sons - one of whom was our Ahmad b. Yusuf - for investigation, seeking to tie them to "people from Baghdad." When no incriminating evidence emerged, Ibn Tulun ordered the brothers released. Yaqut, Udaba', 5: 155f.

25. Yaqut, Udaba', 5: 159; Rosenthal, EP art. "Ibn al-Daya."

26. For the two published editions, by Vollers (ISV) and Hasan (ISH), see above, n. 2.

27. ISV, 3; ISH, 73. Ibn Khaldun (Kitab allibar [Beirut: Dar al-Kitab al-Lubnani, 1958], 4: 236) refers to it as Akhbar Boni Tulun. The ornate title at the head of Ibn Sard's chapter on the Tulunids (only at ISH, 73), al-Durr al-maknun f hula dawlat Boni Tulun ("The concealed pearl amidst the finery of the Tulunid dynasty") is clearly Ibn Sard's invention.

28. Kamila kitab Ibn al-Daya ft sirat Ahmad ibn Tulun. ISV, 77; ISH, 133.

29. The work was edited and published in 1939 by Muhammad Kurd cAli, as Sirat Ahmad b. Tulun (see above, n. 2). However, the manuscript's title was Sirat Al Tulun, as Kurd cAli says himself.

30. Ibn al-Nadim, Fihrist, ed. G. Flugel (Leipzig: F C. W Vogel, 1871), 1: 193, where al-Balawi appears in the list of Ismacili authors.

31. The book titles that Ibn al-Nadim ascribes to al-Balawi, Kitab al-Abwab, Kitab al-Macrifa, and Kitab ai-Din wa-fara'idihi, are compatible with an Ismacili affiHiation. In his introduction to al-Balawi's Sira, Kurd cAli accepts this identification, but G. E. Shayyal, in his art. "al-Balawi," EP, 1: 990, dismisses it.

32. al-Najashi, Rijal (Qom: Mu'assasat al-Nashr al-Islami, 1416/1995), 324-25, mentions cAbd Allah b. Muhammad al-Balawi, describing him as "weak" (dacif) and citing him in a single isnad; al-Tusi, Fihrist asma' musannifi l-Shfa (Cairo: Mu'assasat Nashr al-Faqaha, 1417/1996), 169, repeats Ibn al-Nadim's description.

33. al-Dhahabi, Mizan al-iHidal (Beirut: Dar al-Macrifa, 1382/1962-63), 2: 491; Ibn Hajar, Usan al-mizan (Beirut: Mu'assasat al-Aclami li-1-Manshurat, 1970), 3: 338, 4: 238.

34. In his introduction to al-Balawi's Sira, Kurd cAli says (without references) that it was used by al-Maqrizi and by Ibn Tulun's namesake, Shams al-Din Ibn Tulun. However, I have not been able to ascertain this. See below, nn. 99, 100.

35. al-Balawi, 31-32.

36. Kurd cAli (introduction to al-Balawi, 10-11) thinks that the forty "new" stories actually did come from Ibn al-Daya, either from his biography of Ibn Tulun, or from his Kitab al-Mukafa'a wa-husn al-cuqba (Beirut: Dar al-Wahda, 1975).

37. Kurd 'Ali, introduction to al-Balawi, 11; G. Shayyal in EP, 1: 990.

38. The earliest and most famous of these books is Sirat cUmar b. cAbd al-cAziz by cAbd Allah b. cAbd al-Hakam (Cairo, 1954), on the Umayyad caliph ^Urnar ?. The author of this work was the father of that Muhammad b. cAbd Allah b. cAbd al-Hakam (d. 268 or 269/882-3), who obtained Yusuf Ibn al-Daya's release from Ibn Tulun's prison, at considerable risk to himself, (see above, n. 23). In addition, several siras are credited to Ibn Zulaq al-Sadafi (al-Misri), including biographies of major Ikhshidid figures (see Guest's introduction to al-Kindi's Governors, 45), as well as a biography of Ibn Tulun himself (see Ibn al-cAdim, Bughya, 2: 834, "I have read in the sira of Ahmad b. Tulun by Ibn Zulaq al-Misn . . ."). Al-Kindi (see below, n. 40) wrote a sira of al-Sari b. al-Hakam, an earlier governor of Egypt who achieved some autonomy during the Fourth Fitna and its aftermath, during the 810s and 820s C.E.; see Guest's introduction to al-Kindi, 9-10.

39. As in Sirat al-Imm Ahmad Ibn Hanbal by Salih b. Ahmad Ibn Hanbal, d. F. M. Ahmad (Alexandria: Mu'assasat Shabab al-Jami'a, 1981).

40. al-Kindi, The Governors and Judges of Egypt (above, n. 2), esp. 212-33. See also F. Rosenthal, EP, 5: 121-22; Brockelmann, GAL, 1: 155f., Sl: 229f.; Sezgin, GAS, 1: 358.

41. This is discussed more fully below, n. 89.

42. al Yacqubi, Ta'rikh (Beirut: Dar Sadir, 1960); al-Mascudi, Muruj al-dhahab wa-macadin al-jawhar, ed. C. Pellat (Beirut: Manshurat al-Jamica al-Lubnaniyya, 1965-1979).

43. Ibn al-Athir, al-Kamilfi ?-ta'rikh (Beirut: Dar Sadir and Dar Bayrut, 1965); Ibn Khaldun, Hbar, 4: 636-52; al-Maqrizi, Kitab al-Khitat (Beirut: Dar Ihya' al-cUlum, 1959), 2: 90-103; idem, al-Muqajfa (Beirut: Dar al-Gharb al-Islami, 1991), 2: 446-51; Ibn Taghribirdi, al-Nujum al-zahira (Cairo: al-Mu'assasa al-Misriyya al-cmma Ii-ITa'lif wa-1-Tarjama wa-l-Tibaca wa-1-Nashr, 1929-1972), 3: 1-49.

44. Ibn <Asakir, Ta'rikh madinat Dimashq (Beirut: Dar al-Fikr, 1995-2001); Ibn al-cAdim, Bughyat al-talab fi ta'rikh Halab, ed. S. Zakkar (Damascus, 1988), esp. 2: 826-35. In addition to the primary sources and modern works already listed, see A. A. Vasiliev and Marius Canard, Byzance et les arabes, vol. 2.1: La dynastie macdonienne (867-959) (Brussels: Fondation Byzantine, 1968), and vol. 2.2: La dynastie macdonienne (867-959): Extraits des sources arabes (Brussels: Editions de l'Institut de Philologie et d'Histoire Orientales et Slaves, 1950).

45. ISV, 4-5; ISH, 74; al-Balawi, 34-35; Ibn Khaldun, 4: 637; Ibn Taghribirdi, 3: 4-5; Bianquis, "Autonomous Egypt," 91. On the (problematic) Turkish identity of Ahmad b. Muhammad and cUbayd Allah b. Yahya, see Matthew Gordon, "The Khaqanid Families of the Early 'Abbasid Period," JAOS 121 (2001): 236-55.

46. ISV, 5; ISH, 74. Here an assembly (jama'a) rules on a point of siyar, or law of war, to the effect that whoever has a father or mother who disapproves of his staying on campaign must return home.

47. ISV, 4; ISH, 74; Ibn Khaldun, 4: 237; Ibn Taghribirdi, 3: 4.

48. Some historians, such as Ibn Taghribirdi (3: 4), emphasize the Hanafi association, but Ibn al-Athir (7: 409) says that Ibn Tulun inclined to Shaficism. Biographers and readers can choose according to their preference.

49. ISV, 7; ISH, 76; Hassan, Les Tulunides, 34-35.

50. al-Kindi, 214-15.

51. ISV, 11-12; ISH, 80-81; al-Balawi, 50-51; Ibn Khaldun, 4: 639; Bianquis, "Autonomous Egypt," 94.

52. al-Kindi, 215: Ahmad repeatedly appointed Musa chief of the security force (shurta) and then dismissed him.

53. ISV, 11; ISH, 80; al-Balawi, 46; al-Kindi, 216; Hassan, Les Tulunides, 47.

54. ISV, 12; ISH, 81; al-Balawi, 47.

55. ISV, 12-13; ISH, 81-82; al-Balawi, 51; al-Kindi, 217; Hassan, Les Tulunides, 178-80. There are indications that Musa may also have gone to Iraq, which seems to mean that he went over to al-Muwaffaq.

56. ISV, 14-15; ISH, 83-84; al-Balawi, 57-58; Hassan, Les Tulunides, 50. This episode does not occur in al-Kindi.

57. ISV, 24; ISH, 84; al-Balawi, 73; al-Kindi, 217; Ibn al-'Adim, 2: 826, 828-29; Ibn Khaldun, 4: 639; Hassan, Les Tulunides, 53.

58. Ibn Khaldun, 4: 642f. Of course, we need to be careful, in case Ibn Khaldun imposes his own ideas on the material: see below, nn. 162-64.

59. ISV, 24-25; ISH, 90-91; al-Balawi, 90; al-Mascudi, Muruj, 8: 68f.; Ibn al- AtMr, 7: 308-9; Hassan, Les Tulunides, 66; Vasiliev and Canard, Byzance et les arabes, 2.1: 79-81. The inhabitants of Lulu'a were Slavs who had gone over to the Muslim side.

60. ISV, 24-25; ISH, 91; al-Balawi, 91.

61. al-Kindi, 217.

62. Most of the sources assign no date to this event, but Ibn al-cAdim (2: 826, 828-29) names the year 262, which makes sense, given what happens next.

63. al-Kindi, 218; Ibn Khaldun, 4: 641; Hassan, Les Tulunides, 59-63; Bianquis, "Autonomous Egypt," 95.

64. ISV, 55; ISH, 115-17; al-Balawi, 91-92; al-Kindi, 219-20; Bianquis, "Autonomous Egypt," 95.

65. ISV, 56; ISH, 117-18; al-Balawi, 98-99. The encounter here is with one of "the devotees of hard practice" (al-cubbad al-khushn); see M. Bonner, Aristocratic Violence and Holy War: Studies in the Jihad and the ArabByzantine Frontier (New Haven: American Oriental Society, 1996), 124-30, 157-84.

66. BV, 56-57; ISH, 118; Hassan, Les Tulunides, 66.

67. al-Kindi, 220; Hassan, Les Tulunides, 66. Al-Balawi seems to describe this departure as a staged capitulation to the city's demands (as Matthew Gordon has reminded me). This is all difficult to understand (see following note), but in any case it looks like a humiliating setback for Ibn Tulun.

68. al-Tabari, 3: 1929; Ibn al-Athir, 7: 316-17, where Ibn Tulun withdraws from Tarsus in order to strike more fear into the Byzantines than he does into the Tarssis (?). The warfare and campaigns of these years are discussed in Vasiliev and Canard, Byzance et les arabes, 2. 1 : 82-94. Canard notes (p. 87) that the sources for this period are difficult to reconcile and do not yield a unified picture. The main extent of Ibn Tulun's involvement is to appoint some of the Muslim commanders in the field, such as Sima (not to be confused with Sima al-Tawil) (al-Tabari, 3: 1937, 1942, under the year 266). Campaigns and sieges farther north, as at Marcash, were clearly beyond Ibn Tulun's purview.

69. al-Tabari, 3: 1939-40. According to al-Balawi (101), Ibn Tulun sent Lu'lu' to al-Raqqa. Al-Kindi (224) has a different version, whereby Ibn Tulun sent Lu'lu' to Syria with an army, at which point he entered into contact with al-Muwaffaq, in 268. See also Hassan, Les Tulunides, 72; Bianquis, "Autonomous Egypt," 96.

70. Ibn Khaldun, 4: 642-43. Some of this may be tendentious or even fanciful.

71. BV, 67-68; al-Kindi, 224; al-Tabari, 3: 2025, 2028-29; Ibn al-Athir, 7: 393; Hassan, Les Tulunides, 77-79. Hassan cites a possible reason for LuWs defection, namely Ibn Tulun's assigning Diyar Mudar to his own brotherin-law Muhammad b. al-Fath b. Khaqan; see ISV, 64; al-Balawi, 173.

72. al-Tabari, 3: 2080-82.

73. al-Tabari, 3: 2112, under the year 273; Ibn Khaldun, 4: 648; Ibn Taghribirdi, 3: 44. Lu'lu' died in poverty in Egypt toward the end of the Tulunid era.

74. ISV, 68-69; ISH, 127; al-Balawi, 281; Hassan, Us Tulunides, 81-82.

75. al-Kindi, 224; Hassan, Les Tulunides, 82. In some places there is confusion between Tukhshi and Khalaf. At al-Tabari, 3: 2026, Khalaf conducts a large-scale raid against the Byzantines in 268.

76. This al-Fath is the famous political advisor, littrateur, and patron of the arts who was murdered together with the caliph al-Mutawakkil in Samarra' in 247/860. On Yazman, see al-Tabari 3: 2028; al-Kindi, 224-25; Ibn Taghribirdi, 3: 45; Hassan, Les Tulunides, 82; Vasiliev and Canard, Byzance et les arabes, 2.1: 100-101.

77. Ibn Taghribirdi, 3:45.

78. Ibn al-Shaykh may be seen as a frontier figure, because once he loses out in southern Syria, he re-establishes himself in the northern frontier of Armenia and the Caucasus.

79. ISV, 6; ISH, 75-76; al-Balawi, 39-41; Hassan, Us Tulunides, 32-33. Al-Tabari (3: 1670-71) does not implicate Ibn Tulun in the murder. On the importance of networks of companions for the rise of Ibn Tulun, see Gordon, The Breaking of a Thousand Swords, esp. 62-70.

80. Al-Kindi, 226, lists these as Bakkr b. Qutayba, Minhal b. Habib, Ishaq b. Muhammad b. Macmar, Qays b. Hafs, cAbd Allah b. Bashir, Hawthara b. cAbd al-Rahman, Sacid b. Sacdun, Fahd b. Musa, cAli b. Muhammad b. 'Abd al-Hakam, "and others" (wa-ghayruhum). A tenth Egyptian participant is Ibn al-Mawwaz (see below, n. 139), who, together with Bakkar and Fand, refused to subscribe to the document in its entirety.

81. Al-Tabari, 3: 2048, laconically reports a message sent by Ibn Tulun to "the people of Syria," but not an assembly of judges and jurists.

82. When al-Muctamid was forced to return to Samarra', at first he was held in isolation in "the palace of Abu Ahmad b. al-Khasib" (al-Balawi, 293).

83. al-Kindi, 226.

84. Al-Maqrizi (Muqajfa, 2: 445) has, somewhat more correctly, tabarru>.

85. al-Balawi, 294-96.

86. al-Maqrizi, Khitat, 2: 103; idem, Muqajfa (Beirut, 1991), 2: 442-54, esp. 444f.

87. Ibn al-'Adim, in particular, has a rich array of sources for this period, well beyond al-Kindi and the Siras.

88. Above, n. 37.

89. W. al-Qadi, "An Umayyad Papyrus in al-Kindi's Kitab al-Qudatl" Der Islam 84 (2008): 200-245.

90. Ibid., 231-33: only five instances in all of al-Kindi's extant writings.

91. Ibid., 242.

92. Ibid., 235 f. For the use and function of documents in early Islamic historiography, see A. Noth and L. I. Conrad, The Early Arabic Historical Tradition: A Source-Critical Study (Princeton: Darwin Press, 1994); C. Robinson, Islamic Historiography (Cambridge: Cambridge Univ. Press, 2003); G. Schoeler, The Oral and the Written in Early Islam (London: Routledge, 2006). Here the first two centuries of Islam have received far more attention than the third and the fourth.

93. al-Qadi, "An Umayyad Papyrus," 235.

94. Hassan, Les Tulunides, 150-52; P. M. Sijpesteijn, "The Archival Mind in Early Islamic Egypt: Two Arabic Papyri," in From al-Andalus to Khurasan: Documents from the Medieval Muslim World, ed. eadem et al. (Leiden: Brill, 2007), 163-86, esp. 164.

95. al-Balawi, 296.

96. Above, n. 35.

97. al-Balawi, 289. This happens elsewhere in al-Balawi's Sira, as at 102, 310.

98. Especially in Khitat, 2: 9Of.

99. al-Muqaffa, 2: 442-54, esp. 449f.

100. Ibid., 445. Al-Maqrizi does not explicitly name Ibn al-Daya as his source for the document, but rather as his source for Bakkar - but these two clearly go together.

101. Above, n. 34.

102. al-Kindi, 226: wa-shahida 'ala dhalika iam' man hadara illa Bakkar. . . .

103. al-Balawi, 296-97; al-Maqrizi, Muqajfa, 2: 449-50. I have used al-Maqrizi's text for the last sentence, because al-Balawi here makes no sense wa-ashhadtu calayya wa-fatayaya man kataba shahadatahu fi hadha l-kitab).

104. On him, see Ibn cAsakir, 38: 102-4. Al-cUmari was Meccan in origin and a descendant of the caliph cUmar I. After these events he became chief qadi in Damascus, when Khumarawayh dismissed the well-known Abu Zurca al-Dimashqi from that position. Then al-cUman was appointed to al-Urdunn and Palestine, while Abu Zurca was restored to Damascus. Al-cUmari had a poor reputation as a traditionist, and apparently died in 293.

105. Ibn cAsakir, 34: 78-86; al-Khatib, Ta'rikh Baghdad, 11: 62; al-Safadi, Wafi, 18: 72; Wakic, Muhammad b. Khalaf, Akhbar al-qudat, Cairo: Matbacat al-Istiqama, 3: 34; al-Dhahabi, Siyar aHam al-nubala', 13: 539. cAbd al-Hamid was a Hanafi from Basra, having studied with Hilal b. Yahya al-Ra'y. He was appointed qadi of Damascus in 264. Afterward, when al-Muctadid became caliph, cAbd al-Hamid accompanied him to Baghdad, where al-Muctadid (apparently forgiving his earlier stance against al-Muctadid's father, al-Muwaffaq) appointed him judge of the Sharqiyya quarter in 183. He died in Baghdad in 292, at the age of 95, which means he was in his early seventies at the time of Ibn Tulun's assembly in Damascus.

106. Ibn cAsakir, 5: 120-26. He was born in 192, apparently in Jazira (his nisba is al-Raqqi), and died as qadi of al-Raqqa in 274 or 276.

107. Ibn cAsakir, 34: 86.

108. Ibn cAsakir, 10: 368-74; see also above, n. 99 (al-Maqrizi's Muqajfa), and below, n. 134.

109. Thus al-Maqrizi; al-Balawi: "in these documents," fi hadhihi l-kutub.

110. al-Balawi, 297-98; al-Maqrizi, 2: 450.

111. C. Melchert The Formation of the Sunni Schools of Law [Leiden: Brill, 1997], 121) says that al-Maqrizi's text proves that Bakkar did, in fact, comply with Ibn Tulun's demand.

112. This is how Bakkar b. Qutayba understood the matter; see above, n. 110.

113. R. P. Mottahedeh, Loyalty and Leadership in an Early Islamic Society (Princeton: Princeton Univ. Press, 1980, rpr. 2001).

114. For instance, al-Balawi, 194f., where Ibn Tulun's gratitude for gifts and benefits received is coupled with generosity toward the poor and scholars.

115. Not the two versions that we have in al-Kindi and al-Balawi, but rather (1) the text that, according to al-Kindi, Ibn Tulun sent back to be read out at Fustat, and (2) the text that Ibn Tulun presented at Damascus and then promulgated.

116. Guest, introduction to al-Kindi's Governors, 42; Hassan, Les Tulunides, 266-78.

117. al-Kindi, 227. I have not found any information about this poet, apart from Hassan, Les Tulunides, 269. The measure of the poem is basit:

tala l-huda bi-bni tuluna l-amiri kama

yazhu bihi l-dinu 'un dinin wa-islami

qada l-juyasha mina l-fustati yaqdumuha

minhu cala l-hawli madin ghayru mihjami

fijahfalin li-l-manaya fi maqanibihi

makaminun bayna rayatin wa-aclami

yasmu bihi min bani samin ghatarifatun

bidun wa-sudun usadun min bani hami

law anna ruha bani kundaja muHaqatun

bi-1-mushtari lam yafuthu aw bi-bahrami

hata l-khilafata wa-l-dunya khalifatuna

bi-sarimin min suyufi llahi samsami

ya ayyuha l-nasu hubbu nasirina lahu

maca l-amiri bi-duhmi l-khaylifi l-lami

laysat salatu musallikum bi-ja'izatin

wa-la l-siyamu bi-maqbulin li-sayyami

hatta yara l-sayyidu l-maymunu dhabbakumu

cani l-imami bi-atrafi l-qana l-dami

118. This apparently refers to a commander in Ibn Tulun's army. Accounts of the march to Syria, as at al-Balawi, 286f., do not name such a commander, other than Ibn Tulun himself. They do all mention, famously, that Ibn Tulun brought along his son al-cAbbas, in chains.

119. Ishaq b. Kundaj cannot escape from Ibn Tulun. The word for "Mars" here is Persian bahram.

120. The caliph al-Muctamid.

121. Apparently referring to the letter lam on Ibn Tulun's banners.

122. al-Kindi, 228. Measure: kamil. See also Hassan, Les Tulunides, 269-70.

ya ghurrata l-dunya lladhi afcaluhu

ghurarun biha kullu l-wara tatacallaqu

anta l-amiru cala l-sha'ami wa-thaghriha

wa-l-raqqatayni wa-ma hawahu l-mashriqu

wa-ilayka misru wa-barqatun wa-hijazuha

kullun ilayka fu'aduhu mutashawwiqu

hataka l-khilafata saHdun wa-khattluhu

ishaqu licban wa-l-hasudu l-akhlaqu

asyafuna bidu l-manuni fa-laytaha

bi-najfi man khadhala l-imama tukhallaqu

tumshi wa-tusbihu dariban min dunihi

bi-muhannadin minhu l-hutufu tufarraqu

yatluka sacdun wa-l-muqaddamu taytaku

wa-l-ladhiqiyyu wa-dha l-hafizati yalhaqu

123. The "two Raqqas" refers to the twin cities of al-Raqqa, on the banks of the Euphrates, and al-Rafiqa, founded next to it by the cAbbsid caliph al-Mansur. See Yaqut, Mu'jam al-buldan (Beirut: Dar Sadir, 1955-1957), 3: 15-16, 58-60.

124. The Hijaz was contested, with fighting at Mecca between Tulunid and cAbbasid forces, see below, n. 153.

125. Literally sa'id, a punning reference to Sacid b. Makhlad, who took the lead in deceiving al-Muctamid and his commanders and then received him in Samarr'; see al-Balawi, 293; Ibn al-Athir, 7: 374.

126. Ibn Tulun.

127. "He who nurtures a grudge." I am unable to identify any of these officers.

128. Al-Kindi (228) also preserves a fragment of a third poem, also by Qacdn b. cAmr. The beginning of this one, "Who will inform Syrian Mudar, and all that Egypt contains . . . ," is echoed by poets on the opposing side; see below, n. 133.

129. On bara'a in the sense of 'quittance' in papyrus documents, see al-Qadi, "An Umayyad Papyrus," 238.

130. al-Balawi, 299-300.

131. The detailed accounts in al-Tabari of al-Muwaffaq's campaigns against the Zanj are also lacking in such language, to a remarkable extent.

132. al-Tabari, 3: 2048; Ibn al-cAdim, 2: 834. Al-Muctamid pronounced the curse against Ibn Tulun from the Dar al-cmma, and his son al-Mufawwad then did the same.

133. Al-Kindi (229-31) records an anti-Tulunid poem by al-Buhturi, and poems by a certain Muhammad b. Da'ud, likely to be the same as the Muhammad b. Da'ud b. al-Jarrah al-Isfahani (d. 296/908), who was the author of Kitab al-Waraqa, Kitab al-Zahra, and other anthologies, though not known otherwise as a poet in his own right. See Sezgin, GAS, 2: 75, 100, 439, and index. Hassan Us Tulunides, 270) thinks this poet lived in Egypt, but it is difficult to see how this could be so. Both these poems, especially al-Buhturi's, engage in taunting repartee with the pro-Tulunid poems just mentioned.

134. Above, n. 108. For the following episode, see also al-Kindi, 215, 221, 224, 226, 231, 275-79; al-Balawi, 296f., 316-19, 331-33; al-Maqrizi, Muqajfa, 2: 442-54; Ibn Hajar, Raf al-isr, in Guest's edition of al-Kindi's Governors, 505-14.

135. At al-Maqrizi, Muqajfa, 443, Ibn Tulun demands a written certification mahdar) for this decision.

136. According to Ibn Hajar Rafc al-isr, 512-13), Bakkar was driven around on foot, wearing nothing but his trousers sarawil), slippers, and headgear qalansuwa), with his other clothing stripped away.

137. According to al-Balawi (318), Bakkar died in prison.

138. al-Balawi, 316-17; al-Maqrzi, Muqajfa, 451f.

139. al-Kindi, 226; Ibn cAsakir, 51: 197-98; al-Safadi, Wafi, 1: 335; al-Dhahabi, Siyar, 13: 6. Ibn al-Mawwaz returned to Alexandria, but there is no indication that Ibn Tulun persecuted him as he did Bakkar. A decade later he fled from Egypt to Syria, where he died in hiding in 181.

140. al-Kindi, 226; Ibn <Asakir, 48: 460-61; al-Dhahabi, Siyar, 16: 388. Fahd died soon afterward, in 270. His Maliki affiliation appears in the fact that he related traditions from Salih, the scribe of al-Layth.

141. Melchert, Formation of the Sunni Schools, 79, 1 19-22. Melchert also notes that the subscribers included the Syrian Abu Zurca al-Dimashqi, who is not named in the biographies of Ibn Tulun: Ibn Hajar, Raf al-isr, 520.

142. Vasiliev and Canard, Byzance et les arabes, 2.1, 12Of.

143. For the collected biographies of these individuals, see M. Bonner, Aristocratic Violence and Holy War, esp. 157-84; idem, Jihad in Islamic History, 97-1 17; A. Noth, "Les 'ulama' en qualit de guerriers," in Saber religioso y poder poltico en el Islam, ed. M. Marin (Madrid: Agencia Espaola de Cooperacin Internacional, 1994), 175-95.

144. These included the famous Abu cUbayd al-Qasim b. Sallam, Musa b. Dawud al-Dabbi (d. 216 or 217), and al-Husayn b. al-Husayn b. cAbd Allah al-Antaki (d. 319) (see Bonner, Aristocratie Violence and Holy War, 170-71), Salih b. Ahmad ibn Hanbal (see above, ?. 57), and 'Ubayd Allah b. Muhammad al-cUmari, chief qadi of northern Syria and the Thughur, who signed the document deposing al-Muwaffaq (see above, n. 104).

145. Vasiliev and Canard, Byzance et les arabes, 2.1, 101.

146. Above, n. 67.

147. On these, see al-Qadi, "An Umayyad Papyrus," 216-17, and bibliography listed there.

148. ISV, 4; ISH, 74; al-Balawi, 34.

149. Above, n. 59.

150. Above, n. 70.

151. Qudama b. Jacfar, Kitab al-Kharaj wa-sinacat al-kitaba, ed. M. al-Zubaydi (Baghdad: Dar al-Huriyya li-l-Tibaca, 1981), 186-88.

152. Ibn Hawqal, Surat al-ard, ed. J. H. Kramers (Leiden: E. J. Brill, 1938), 184.

153. al-Tabari, 3: 2083-84; Ibn al-Athir, 7: 395.

154. al-Tabari, 3: 2083-84. Ibn Kundaj fought against Ibn Da'ash, described here as governor of al-Raqqa and the Thughur (?) on behalf of Ibn Tulun.

155. C. Becker, Beitrge zur Geschichte gyptens (Strassburg: K. J. Trbner, 1902-1903); 2: 177; Hassan, Les Tulunides, 85.

156. Ma anta wa-cAli bi-cikmay bacir, literally "you and CAH are not loads of equal weight placed on either side of the camel" (see the editor's note).

157. Nasr b. Muzahim, Waqcat Siffin, ed. Harun (Cairo: Maktabat Muhammad cAli Subayh wa-awladih, 1382/1962-63), 27-28. Cf. S. Leder, "The Literary Uses of the Khobar" in The Byzantine and Islamic Near East, vol. 1: Problems in the Literary Source Material, ed. A. Cameron and L. Conrad (Princeton: Darwin Press, 1992), 296.

158. Nasr b. Muzahim, Waqcat Siffin, 92-93; see also 104. See also al-Tabari, 1: 3304-6, 3316-22.

159. F. Sezgin, at GAS, 1: 313; Leder, "Literary Uses of the Khobar," 296, citing C. Brockelmann, "Nasr ibn Muzahim, der alteste Geschichtschreiber der Schia," Zeitschrift fur Semitistik 4 (1926): 15.

160. Egyptian historical writing from this period is becoming better known. See, for instance, Ta'rikh Ibn Yunus al-Sadafi, ed. ?. F. F. cAbd al-Fattah (Beirut: Dar al-Kutub al-cIlmiyya, 2000); Ibn Zulaq, Akhbar Sibawayh al-Misri, ed. H. Dib (Cairo: Maktabat al-Abhath al-Tlmiyya li-Nashr Ulum al-'Arabiyya, 1410/1989, rpr. of 1933 ed.); idem, Fada>il Misr wa-akhbaruha wa-khawassuha, ed. CA. M. cUmar (Cairo: Maktabat al-Khanji, 2000).

161. Biographical treatments of Bakkar and the other jurists can emphasize this confrontation, but here the focus is on these individuals and their exemplary role; Ibn Tulun becomes part of the background.

162. Ibn Khaldun, 4: 636-52. Ibn Khaldun used other sources as well, but these will require further study.

163. Above, n. 58.

164. In the following I am indebted to Gabriel Martinez-Gros, Ibn Khaldun et les sept vies de l'Islam (Paris: Sindbad, 2006).

165. D. Tor, Violent Order: Religious Warfare, Chivalry, and the cAyyar Phenomenon in the Medieval Islamic World (Wurzburg: Ergon Verlag, 2007).

166. As was done by Paul Wittek, The Rise of the Ottoman Empire (London: Royal Asiatic Society, 1938). By advancing jihad - or, in Wittek's case, ghaza - as an explanatory principle or cause, we risk falling into a circular argument, just as when we advance Islam itself.

167. As expressed in the classic work of Egyptian historical writing, The Conquest of Egypt and the Maghrib by <Abd al-Rahman b. cAbd al-Hakam (d. 257/871).

168. Melchert, Formation of the Sunni Schools, 119. This image certainly applies to the earliest period of Islamic Egyptian history, but by the Tulunid era, conversion to Islam was well underway and the ruling elite was at once less compact and more diverse.

Author affiliation:

MICHAEL BONNER

UNIVERSITY OF MICHIGAN, ANN ARBOR

The use of this website is subject to the following Terms of Use