Academic Article

The Initial Destination of the Fatimid caliphate: The Yemen or The Maghrib?


Da‘wainfo-icon in Yemen, Da‘wa in Maghribinfo-icon, Fatimid caliphate, al-Mahdiinfo-icon

Download PDF version of article (273 KB)

Table of contents



The year 297/909 witnessed the declaration of ‘Abd Allah al-Mahdi as the first Fatimid imaminfo-icon/caliphinfo-icon in the Maghrib (North-West Africa) - a momentous event in Ismaili history. A study of the politico-military events that led to the establishment of the Fatimid state in the Maghrib poses a serious question: Had al-Mahdi’s original destination been the Yemen? If indeed that was the case, as a reading of the primary sources indicate, and had al-Mahdi declared his zuhurinfo-icon (‘emergence’) from Yemen, the entire course of subsequent Fatimid history and indeed of the general history of that region as well as that epoch would have moved in a different direction. The question is important for any attempt to comprehend and assess the objectives the Fatimidsinfo-icon had set themselves as well as to gain an insight into the policies they implemented for the achievement of the goals.

A study wholly focusing on this crucial issue has been undertaken by no contemporary scholar in the field, but it has been touched upon by historians such as Ivanow, Gateau, T. Sharaf and H. I. Hasan, Canard, H. Hamdani and more recently, Dachraoui. Most of these historians are in agreement that al-Mahdi was planning to go to the Yemen upon his departure from Salamiyyainfo-icon in Syria in 289/902. They do differ considerably, however, on the factors that may have been responsible for the change of venue - if that was the case. The paper aims to examine this decisive aspect of Fatimid history.

A pitfall into which present-day historians have generally fallen when examining al-Mahdi's decision is that they have not conducted an in-depth study of the da’wa in the Yemen and the Maghrib. Nonetheless, the need for a study of the genesis of the da’wa in these two areas has been expressed by Bernard Lewis, who observes: ‘The history of the da’wa in Yemen and North Africa is yet to be written’ -- a point that finds echo in the words of the well-known authority on Ismaili studies, W. Ivanow: ‘…perhaps a proper study of the history of Yemen will reveal … why the plan of the move to the Yemen should have been altered and when this change took place.’ A background of the kind suggested will be extremely useful and, in fact, necessary not only for identifying all the possible factors that may have influenced al-Mahdi’s reversal of decision, but also for determining the relative importance of each of those factors. It is with this in mind that I propose to offer a brief survey of the politico-military activities of the Ismaili da’wa in the Yemen and in the Maghrib prior to the establishment of the Fatimid ‘state’. But first a survey of our sources.

The Sources

If the paucity of historical information contained in Ismaili sources is a frequent course for lament, the politico-military aspect of the Ismaili da’wa movement in the 2nd-3rd/8th-9th centuries in the 2 major centres of its activity - the Yemen and the Maghrib - is particularly well served by primary Ismaili works. Ismaili historical sources offer more information than their non-Ismaili counterparts. That is partly due to the fact that several branches of the Ismaili da’wa, although active in the various regions of the Islamic world by the late 9th century, do not appear to have been viewed by the non-Ismaili historians of the time as different manifestations of a single entity, directed from a centre. They were, instead, seen as local and isolated bids for political power and military supremacy by individual and independent personalities. The Ismaili writers, on the other hand, had access to more information on the matter, and had a greater vested interest in reporting at length on the affairs of the Ismaili Imam and the da‘wa. This was particularly the case after the establishment of the Fatimid caliphateinfo-icon, when the fear of persecution by the ‘Abbasidsinfo-icon was no longer felt, so that the shrouding of the movement’s activities in secrecy was no longer warranted.

Ismaili Writings

An indispensable source for the da‘wa activities in the Yemen as well as the Maghrib is al-Qadiinfo-icon al-Nu‘man’s well-known historical work Iftitah al-da‘wa wa-ibtida’ al-dawlainfo-icon. Completed in 346/957, the Iftitah has been acclaimed by esteemed present-day historians such as W. Ivanow and S. M. Stern as the earliest extant source on the da’wa activities. The two editors of the Iftitah, W. al-Qadi and F. Dachraoui, as well as Stern, have painstakingly shown how it became the almost unique source for later historians and how it was absorbed not only into the Ismaili flow of information but also into the mainstream of general Islamic history.

The second valuable Ismaili source for the period under investigation is the Sirat Ja‘far al-Hajib composed by Muhammad b. Muhammad al-Yamani at the beginning of the reign of the Fatimid imam/caliph al-‘Aziz bi-llah (d. 386/996). The Sirainfo-icon is a most useful and detailed eye-witness account of al-Mahdi’s journey from Salamiyya in Syria to the Maghrib. Moreover, it sheds light on the crucial issue of al-Mahdi’s initial dar al-hijrainfo-icon and the subsequent change of venue.

The information in the Iftitah and the Sira is corroborated by accounts preserved in the ‘Uyuninfo-icon al-akhbar wa-funun al-athar of the 9th/15th century Yemeni da‘iinfo-icon, Idris ‘Imad al-Dininfo-icon. Although this is a work of a later period, ‘Imad al-Din, as da‘i al-du‘atinfo-icon, had access to Ismaili works which have survived only as excerpts in the ‘Uyun. Further references are also found in Ahmad b. Muhammad al-Naysaburi’s Istitar al-imam.

Non-Ismaili Writings

In addition to the Ismaili sources, there are a number of general historical works that have proved particularly pertinent to our purpose. Besides supplementing Ismaili works, they serve a most valuable purpose in that they enable us to control the information reported by Ismaili authors. As there are some historians who report exclusively either on the da‘wa in the Yemen or the Maghrib, they will be considered under separate headings as follows:

The Yemen

Among the local Yemeni historians who have paid considerable attention to the Ismaili da‘wa movement in the region are Muhammad b. Malik, a 5th/11th century jurist who joined the Ismaili fold for a short period but subsequently withdrew from it. Notwithstanding the polemics that he conducted against the Ismailis in his Kashf asrar al-batiniyyainfo-icon, Ibn Malik provides a useful insight into the politico-military operations of Ibn Hawshab and ‘Ali b. al-Fadl, the two pioneering da‘is of the Yemen. The Kashf’s most significant contribution, however, is the information it offers on Ibn al-Fadl’s activities in the Yemen, a subject on which the Ismaili sources remain almost silent. Supplementary information is also found in Kitab al-suluk fi tabaqat al-‘ulamainfo-icon wa-l-muluk, by Baha al-Din al-Janadi (d. 732/1331). References to the da‘wa are also contained in al-‘asjad al-masbuk fi man waliya al-Yaman min al-muluk by Ali b. al-Hasan al-Khazraji (d. 812/1409).

It must be pointed out that the Zaydi work Sirat al-hadi ila al-haqqinfo-icon Yahya b. al-husayn, being a biography of the first Zaydi imam of the Yemen, is a source of paramount importance for gauging not only the Zaydi involvement in the politico-military scene in the Yemen at the time, but also for ascertaining Ismaili-Zaydi relations in the region.

The Maghrib

Among the major works of non-Ismaili historians that refer to the da‘wa activities in the Maghrib is al-Bayan al-mughrib fi akhbar muluk al-Andalus wa-l-Maghrib by Ibn ‘Idhari. The indispensability of his treatise lies in the fact that, besides his reliance on the Iftitah, he incorporates detailed reports from chronicles that were contemporary to the da‘wa activities in the Maghrib, many of which have now been lost. Apart from the apparent bias of the author, a Maliki jurist, the only major flaw with this source is that its writer reports year by year (annals), thus presenting the historical events as a unit. This method of narrating information militates against the emergence of a sequence of events.

Other sources that touch, albeit briefly, on da‘wa activities include Ibn al-Athir’s al-Kamil fi al tarikh, al-Durra al-mudiyya fi akhbar al-dawla al-Fatimiyyainfo-icon by Ibn al-Dawadari, Ibn Khaldun’s al-’Ibar wa-diwaninfo-icon al-mubtada’wa-l-khabar and Itti’az al-Hunafa bi-akhbar al-A’imma al-Fatimiyyin al-Khulafa by al-Maqrizi.

The Genesis of the Ismaili da‘wa in the Yemen

By way of prefatory comment it is important to make the point that in the eighth and ninth centuries CE there was a great flurry of Ismaili da‘wa activity in various parts of the Muslim world. Whatever the activity and wherever located, it was directed centrally by the Ismaili Imam, who at that time resided at Salamiyya in Syria. One of the prominent political aims of this da‘wa was to establish a political realm for the imam.

As far as the Yemen is concerned, the earliest dated record of the sending of Ismaili da‘is or ‘emissaries’ to the country relates to the dispatch of Ibn Hawshab and ‘Ali b. Fadl in 268/881. However, Qadi al-Nu‘man reports on Ibn Hawshab’s authority that upon the latter’s arrival at ‘Adan La’a in the north of the Yemen a man called Ahmad b. ‘Abd Allah b. Khali was conducting da‘wa activities and was awaiting the arrival of Ibn Hawshab from the central headquarters at Salamiyya. Ibn Hawshab also came across some Shi‘ainfo-icon (qawm min al-shi‘ainfo-icon) known as the Banu Musainfo-icon in the south of the country, who were in close contact with the Shi‘a in the north. Thus it would appear that da‘wa activities had already taken root in the Yemen at a relatively early date.

Subsequent to their arrival in the country, Ibn Hawshab and ‘Ali b. al-Fadl settled in the two extremes of the country, the former in the north and the latter in the south. Consequently, within a couple of years of their activities they were in a position openly to declare that they were preaching for the Ismaili Imam ‘Abd Allah al-Mahdi. These two Ismaili da‘is had to contend for politico-military supremacy with the Yu‘firid dynasty, which represented ‘Abbasid interests in the area; and the Zaydisinfo-icon, who formed another branch of Shi‘isminfo-icon. Within two and a half decades of their arrival in Yemen, the two da‘is had gained virtual mastery over the country. In 293/905 they met each other for the first time since the inception of their activities in the country. They had disputes over matters of policy as well as a clash of personalities, as will be discussed later.

Consequently, in 296/909 'Ali b. al-Fadl broke ties with the Ismaili da‘wa altogether. This took place in the very year in which al-Mahdi declared himself the first Fatimid imam/caliph in the Maghrib.

The Genesis and Dissemination of the Ismaili da’wa Activity in the Maghrib

Chronologically, the first known instance of da‘is being sent to the Maghrib is to be found in al-Nu‘man’s Iftitah al-da‘wa. He states that the two da‘is al-Hulwani and Abu Sufyan were sent by the 5th Shi‘'i imam in the Husayni line, Ja‘far al-Sadiq (d. 148/765) in 145/762. These two da‘is settled in the Kabyliainfo-icon region of what is today Algeria.

With the arrival of da‘i Abu ‘Abd Allah al-Shi‘i in the Maghrib in 280/893 Ismaili da‘wa activities gained impetus. Al-Shi‘i took up residence with the Kutama Berbers, who were ranked among the most numerous and strong Berberinfo-icon tribes in the region. They also possessed some of the best lands in the Maghrib. Moreover, they were largely independent of the Aghlabid amirs - the vassals of Abbasid authority in the region. Of all the dynastic states in the 3rd/9th century Maghrib, the Aghlabidsinfo-icon proved to be the most vulnerable to the emergent Ismaili state under al-Shi‘i. In fact, it was to the latter that Aghlabid rule succumbed in 296/908. By 289/901-2 al-Shi‘i had secured his first major victory against them, by wresting the strategic fortress of Milainfo-icon from their hands. Another quarter from which al-Shi‘i had to face opposition were the various Kharijite principalities that had become a part of the Maghribi scene since the late 7th/early 8th century.

Following the conquest of Mila, al-Shi‘i invited the imam ‘Abd Allah al-Mahdi to come to the north. In the meantime, an untimely revolt of some ex-da‘is to gain political authority for the imam in Syria, coupled with a warrant of arrest issued against the imam by the ‘Abbasid caliph, made al-Mahdi leave Salamiyya secretly. He arrived in Egypt in 291/903-4, and it was in Egypt that he was faced with the momentous decision of choosing either the Maghrib or Yemen as the centre of his zuhur or ‘emergence’, and therefore the seat of the future Fatimid caliphate.

Al-Madhi’s Change of Proposed Destination

The two major Ismaili works reporting on the issue of al-Mahdi’s destination, the Iftitah and the Sira, make explicit mention of al-Mahdi’s intention to travel to the Yemen.

Ja‘far al-Hajib, one of the companions of al-Mahdi on the latter’s journey from Salamiyya and therefore an original source on the travel plans of the imam, states that on leaving Salamiyya, al-Mahdi declared that he was going to the Yemen. He admits that until the night before the departure from Egypt to the Maghrib, ‘al-Mahdi never revealed his real intention of going to the Maghrib.’ There are a number of statements in the Iftitah that may be adduced to corroborate this fact. Most non-Ismaili sources state that al-Mahdi’s initial destination was the Yemen. In general, an examination of the sources reveals that the Yemen was indeed the original setting for the establishment of the Fatimid caliphate and the zuhur of al-Mahdi. The question then arises: What was it that caused al-Mahdi to reverse such a crucial and considered decision?

It is difficult, in fact well nigh impossible, to determine all the possible factors that may have contributed to this change of decision. This is partly due to the remoteness in time of the issue in question. However, this difficulty arises mainly from the very nature of the information that is sought. It was a decision taken by a certain individual under specific circumstances, and the only manner in which a complete answer can be ascertained is by probing into the mind of the individual who resorted to such a judgement. Matters would perhaps have been much more straightforward if there had survived a biographical work on the decision maker (that is, al-Mahdi), where perhaps the factors which were taken into consideration by him for resolving such a momentous issue would have been laid out. Unfortunately, this is not the case. What can be attempted, however, is an examination of the factors, suggested by the sources, which al-Mahdi may have taken into account in changing his decision.

The Iftitah is the earliest extant Ismaili source shedding any light on this important issue. It states that one of al-Mahdi's da’is preceded him to the Yemen, where he incited the da’i Ibn al-Fadl to revolt. Consequently, the da’i ruined al-Mahdi's plan of entering the Yemen, who therefore ‘disliked entering the Yemen under such a circumstance.’ Further, what is implicitly evident throughout the Iftitah, though not stated explicitly by al-Nu’man is that the politico-military success achieved by the da’i al-Shi’i in the Maghrib was a key factor in al-Mahdi’s preference for it.

A detailed examination of Ismaili da’wa activities in Yemen and the Maghrib at the time suggests many other potential causes for the change of al-Mahdi’s plans. In order to afford a clearer insight into what these may have been I will now consider them under the three following heads:

‘External Factors’, ‘Yemenite Factors’, ‘Maghribi Factors’.

Some Possible Factors Influencing al-Mahdi’s Change of Plan

External Factors

‘Abbasid persecution

The fact that the Abbasids were on a constant lookout for any political activities in which the Shi’a may have been involved at that time is too well known a matter to need further elucidation. As can be attested from the sources, a particularly vigilant watch was maintained on Ismaili activities during the last years of the dawr al-satrinfo-icon (period of concealment). Indeed, one of the major causes cited as motivating al-Mahdi's departure from Salamiyya in 289/901 was the warrant of arrest issued against him by the ‘Abbasid caliph al-Muktafi (d. 295/907).

It is quite likely that ‘Abbasid agents were aware of al-Mahdi’s plan to go to the Yemen. Consequently, a large number of them had been dispatched there to await al-Mahdi’s arrival. This concentration of ‘Abbasid power in the Yemen may indeed have dampened al-Mahdi’s enthusiasm to proceed to the Yemen.

The defection of da‘i al-du‘at Firuz

The defection of the chief da‘i, Firuz, at a crucial moment in al-Mahdi’s brief stay in Egypt is singled out by al-Nu’man as the chief factor responsible for al-Mahdi’s abandoning the Yemen. Although the Iftitah does not mention Firuz by name, the activities that it ascribed to the man in question are exactly the same as those conducted by Firuz, according to the eye-witness account of Sirat Ja‘far. Significantly, the Sira, which provides by far the most detailed account on the defection of Firuz, explicitly states that it was only after ‘Firuz became quite certain that al-Mahdi really intended to go the the Maghrib .. [that] he decided to deceive the Imam’. ‘Imad al-Din Idris substantiates the information provided by the Sira. Thus, whatever the reasons for the defection of Firuz - an issue that needs to be researched at some length, but which falls beyond the scope of this study - his role in seriously harming the Ismaili movement in the Yemen and subsequently influencing al-Mahdi to abandon it must be viewed, if not with outright rejection, at least with some scepticism.

Yemeni Factors

Conflict between Ibn Hawshab and ‘Ali b. al-Fadl

Ironically, the single most important factor that sapped the strength of the political and military achievements of the Ismaili movement in the 3rd/9th century Yemen was the worsening internal dissension between the 2 pioneering da’is, Ibn Hawshab and ‘Ali b. al-Fadl. At their first meeting since their arrival in the Yemen in 292/904-5, the very year in which al-Mahdi was on the verge of taking the vital decision whether to proceed to the Yemen or not - the two da’is differed on matters of policy. Ali b. al-Fadl, whose military successes had outweighed those of Ibn Hawshab, was of the opinion that a policy of further military conquests should be embarked upon. Ibn Hawshab, on the other hand, sought a policy of consolidation in existing territories, particularly since the two da’is were by that time operating in the same territories. But this was not all; there was a clash of personalities between the two da’is as well. Thus dissension between them was inevitable. Al-Mahdi, who was undoubtedly watching the activities of the two da’is in the Yemen, can only have been deterred by these discouraging developments within the Ismaili movement in the country.

Yu‘firid and Zaydi rivalry

As stated earlier, the Ismaili cause in the Yemen was challenged politically by the Yu’firids - the young but powerful dynasty that represented ‘Abbasid interests in the area - but more significantly, the Ismaili da’is had to cope with another Shi’a power, namely, the Zaydis, whose cause had been considerably strengthened by the arrival of their Imam al-Hadi ila al-haqq in 280/893. The choice for al-Mahdi was, in Ivanow's words ‘… between the quiet and certain position of an ordinary Imam in the Yemen and the ambitious programme of the Mahdi …’ in the Maghrib. This statement may be taken as meaning that in the Yemen al-Mahdi faced the prospect of being just another Shi'i Imam, while in the Maghrib he would be the awaited Imam. Moreover the Ismaili da’is were at the time involved in an as yet unresolved struggle for politico-military supremacy in the Yemen and had achieved limited successes only. Meanwhile, politico-military developments in the Maghrib seemed more promising for the establishment of an Ismaili dar al-hijra.

Maghribi Factors

The politico-military success of al-Shi’i

The politico-military successes achieved by the da’i Abu ‘Abd Allah al-Shi’i in the Maghrib are cited by the sources and faithfully repeated by present-day historians as a major force that tipped the scales in favour of the Maghrib as the venue for al-Mahdi's state. While these successes were of major consequence, their extent and significance in the Maghrib circa 290-2/903-5 have been over-estimated by scholars. A careful reading of the source material shows that the conquests conducted by al-Shi’i up to that time had not been decisive enough to compel al-Mahdi, solely on that basis, to take positive steps towards the Maghrib. The only military success of which al-Mahdi would have been aware before his decision to head for the Maghrib was the conquest of Mila in 289/901-2.

The weaknesses of Aghlabid realm

Unlike the case in the Yemen, where ‘Abbasid interests had been reinvigorated by the powerful Yu’firid dynasty, the Abbasid cause in the Maghrib had suffered a major setback with the accession of the indolent and pleasure-loving Ziyadat Allah III as the Aghlabid ruler in 290/902-3. Moreover, the inherent weakness within the Aghlabid realm as well as external aggression sapped even further the authority of the amirinfo-icon over his territories. It seems logical to assume, given the sharp political acumen of al-Mahdi, that he would have had the foresight to sense the imminent politico-military achievements of the Ismaili da‘wa in the area, and the weaknesses of the Aghabid state would have provided him with one such indication.

The Idrisidsinfo-icon

Finally, even though the Zaydi Idrisids had managed to establish themselves as rulers in Morocco as early as 172/788 - contrary to the position in the Yemen, where the Ismailis and Zaydis were vying for political and military authority in the same territories - the two movements in the Maghrib were operating in areas that were independent of each other.


In conclusion, a more detailed reading of the primary Ismaili and non-Ismaili sources reveals that the Yemen was indeed the initial destination that al-Mahdi had in mind for his dar al-hijra. The analysis and interpretation of the available evidence presented in this paper suggest that no single factor can be held to have prompted al-Mahdi's decision to switch his plans away from the Yemen in favour of the Maghrib. In general, it is permissible to conclude that prospects of establishing an Ismaili state in the Yemen must have seemed much more bleak than those offered by the optimistic scene that had developed in the Maghrib. It would appear that this is why al-Mahdi, in the final analysis, opted for the latter as suitable for making his zuhur. What also comes to light is the fact that al-Mahdi, like his predecessors at Salamiyya, was the central directing force within the da’wa. The evidence of the sources, furthermore, suggests that al-Mahdi’s decision was exclusively his own, for he does not appear to have discussed the change of destination with his da‘is.

British Society of Middle Eastern Studies Bulletin 1, 13, 1986, pp. 15-26.