1. Fatimid Historiography: A Lost Corpus?

Modern scholarship on medieval Islamic history, historiography and book culture has not as yet unburdened itself of the historically and historiographically problematic notion that the last Fatimid vizier and first Ayyubid sultan Ṣalāh al-Dīn al-Ayyūbī (the Fatimid vizier from 1169–71 and the Ayyubid sultan from 1171–1193) destroyed the Fatimid royal libraries in the early 1170s, resulting in the loss of the primary sources documenting the Fatimid era (909–1171).Footnote ¹ The idea has persisted, pace Lewis, Daftary, HalmFootnote ² and others, despite evidence to the contrary in a variety of studies that examine the topic directly or indirectly, in the contexts of both Fatimid historiography and of medieval learning in general, such as those by Walker, Brett, Sayyid and Hirschler.Footnote ³ That troublesome assumption is itself part and parcel of a wider series of modern debates engendered by the tendentious narratives of medieval Arabic chronicles. Virtually all of these are preoccupied with pushing forward, wittingly or not, sectarian or political agendas: pro-Fatimid, anti-Fatimid, anti-Ayyubid, pro-Zengid, pro-Ayyubid and so forth, in a variety of permutations, allied with some instances, particularly in court chronicles, of the production of historiography as hegemonic narratives, expressive of a will-to-power, or of post facto rationalisation and legitimisation of political sovereignty.Footnote ⁴ These agendas then intersect with another grand trope in Islamic history, that of library destruction, which is itself a subset of the decline paradigm of medieval Islamic intellectual history, to produce a set of assumptions of which the constituent elements are somewhat problematic to untangle and differentiate; this awaits deeper and broader analysis.Footnote ⁵ To this mix we may add the early-generated and persistent ‘Saladin myths’, which in recent years have resulted in a raft of biographical works in European languages that attempt to separate ‘fact’ from legend in the depiction of the sultan.Footnote ⁶ The loss of historiographical records, the destruction of libraries, the assertion of a new political/sectarian affiliation on the part of political states, alongside a debunking of the legitimacy of those deposed, and the partisan portrayals of Ṣalāḥ al-Dīn's religious and political character in Islamic Arabic literature: these familiar themes form the ideological and narrative background against which the present discussion is framed.

Within this multifaceted context, the principal topic of this article is the evidence from one particular narrative strand of Arabic historiography on the issue of how Ṣalāḥ al-Dīn dealt with the Fatimids’ books upon his assumption of the Fatimid vizierate and subsequently the new Ayyubid sultanate in Egypt (and later Syria). Was there an anti-Isma‘ili ideological aftermath of (to use a modern term) book trashing that directly resulted in the relative paucity of Fatimid-era chronicles which beleaguers our understanding of the period today? For this is the import of the idea of ‘destruction’ propounded by Daftary and others, who argue not just that the Ayyubids dismantled the literary institutions of their predecessors and dispersed their contents, but extirpated them and thereby removed the native Fatimid corpus from circulation altogether, in the vein of the Fatimids’ own putative destruction of an Aleppo library in 1076 on account of anti-Isma‘ili contents within its walls.Footnote ⁷ How is it, indeed, possible to piece together a view of the events of Ṣalāḥ al-Dīn's first years in power in Egypt, given the array of various affiliations and vantage points expressed by the handful of Arabic chronicles we may draw upon to throw light on the matter?

A secondary issue, following on from this main set of questions regarding Ṣalāḥ al-Dīn and his dealings with the Fatimid royal library collections, is the broader question of how much Fatimid-era historiography, a sub-corpus of the Fatimids’ literary legacy, survives into later chronicles, and what this might hope to tell us about the ‘destruction’ trope with which this article begins. These are the themes I address here, where I juxtapose reports from historiographical sources in order to complement broader parallel or prior studies showing similar results from widely different sources.Footnote ⁸ I use the comparative approach to Arabic historiographical narratives deployed by Little and then Massoud, less in order to fulfil their threefold self-reflexive aims of elucidating Mamluk historiography in general, identifying borrowings and building a ‘repertorium of sources’, in Cahen's wording, and more in order to throw light on an historical episode itself, risky as this may seem in view of the proclivities of the chroniclers.Footnote ⁹ Other sources, narrative and documentary, such as Ibn al-Furāt's (d. 1405) inventory of Fatimid-era sources (on which I will say more later), the documentary encyclopaedia of al-Qalqashandī (d. 1418), the annalistic and topographical works of al-Maqrīzī (d. 1442) and documentary specimens such as the Damascene Ashrafiyya Turba's catalogue examined by Hirschler,Footnote ¹⁰ provide irrefutable evidence of the healthy survival of Fatimid chronicles, documents and books in general beyond the duration of the dynasty. Collectively, these studies pave the way for a reconsideration of the still pervasive assumption that the Fatimids’ library holdings perished alongside the ruling family; the findings of this article fall into that context and add a textual contribution to that re-evaluation. Against this backdrop, we can now examine the main question that arises in this enquiry, namely, the role of Ṣalāḥ al-Dīn's policies in the destruction or survival of the Fatimid historiographical corpus, a subject necessarily addressed in preliminary fashion here in view of the vastness of the source base for the crossover period from Fatimid to Ayyubid rule, both documentary and narrative.Footnote ¹¹

2. Fatimid Libraries

As a dynasty whose love of learning entailed a knowledge of books within and outside their own hotly contested traditions, the Egyptian Fatimids from earliest times were known to collect books and treatises in Arabic on a wide range of subjects, particularly confessional literature of a variety of hues.Footnote ¹² The famed literary collections of the Fatimids are attested from as early as al-‘Azīz's reign (975–996), were by then already sizeable,Footnote ¹³ and were housed in several major libraries established in Cairo in the Fatimid era, including the well-known and publicly accessible Dār al-Ḥikma founded by al-Ḥākim (996–1021) in 1005, where the collection was much wider than simply a set of tools for the da‘wa, and was not limited to religious subjects.Footnote ¹⁴ Private libraries were no doubt also a significant aspect of this literary landscape, but the data in this regard appears to be sparse.Footnote ¹⁵

The royal Fatimid libraries did not enjoy a stable existence, however, and were on several occasions plundered for their treasures by force of financial and political exigencies.Footnote ¹⁶ The most serious of these incidents took place during the financial and social crisis of the 1060sFootnote ¹⁷ cited in the sources as the Shidda (calamity), when, in 1068, soldiers and officials who had been unpaid by the state for some unspecified period of time plundered the Fatimid palaces, including the royal library complex, which comprised two collections: one each in the inner and outer palace libraries. The latter was ransacked while the former remained intact.Footnote ¹⁸ The soldiers looted books and other treasures in order to sell them, thus dispersing a significant portion of the Fatimid artistic and literary collections.Footnote ¹⁹ Ironically, this act of destruction is also indicative of the high regard in which books were held in Fatimid times, as they were through much of the medieval Arabic world: they were objects of immense value, supplanting monetary wealth when the latter dried up or was unavailable.Footnote ²⁰

The plunder of the Fatimid book collections mid-way through Fatimid rule in Egypt was to be repeated in a markedly different manner from 1174, when Ṣalāḥ al-Dīn (d. 1193) returned Egypt to the ‘Abbasid caliphate under al-Mustaḍī’ bi Amrillāh (d. 1180) in Baghdad, and appointed a Shafi‘ite chief judge.Footnote ²¹ He subsequently broke up, mainly through sale or donation, the Fatimid treasures, including the book collections at the caliphal palace and at al-Azhar. For by the end of Fatimid rule, the royal palace library complex was the main one left intact,Footnote ²² and it is this collection that features in the present article. Medieval sources, too, seem concerned solely or mainly with the royal collections. The precise nature of this latter episode of book dispersal is discussed presently, but what is not in doubt is that Ṣalāḥ al-Dīn's liquidation of Fatimid assets, viewed alongside his reassignment of government, military and religious posts, of state wealth and of iqṭā‘s, as examined by Lev and Frenkel amongst others, was a profound actualisation of the end of Fatimid rule in Egypt, with far-reaching symbolic as well as economic significance.Footnote ²³

3. Ṣalāḥ al-Dīn and the Fatimids’ Books

While the limits of space constrain this article from enumerating fully all the Ayyubid and Mamluk histories reporting Ṣalāḥ al-Dīn's assignation of the books he found in the caliphal palace in Cairo, a range of reports on the issue gathered from the main medieval Arabic chronicles indexed by Little's study of Ayyubid and Mamluk historiography, supplemented by Mackensen's work on medieval Muslim libraries,Footnote ²⁴ reveals several strands of information, some contradictory, on how the books were dispersed. Among the authors and works whose reports describe the sequence of events in question are: ‘Imād al-Dīn al-Iṣfahānī (d. 1201) and Ibn Abī Ṭayy (d. 1232) (both as received via Abū Shāma's Rawḍatayn, d. 1267, and the former also via al-Bundārī's summary of his lost Barq al-Shāmī, d. 1245); Ibn al-Athīr's al-Kāmil (d. 1233); Ibn Shaddād (d. 1234) in his al-Nawādir al-sulṭāniyya,; the lost Naẓm al-sulūk fī ta’rīkh al-khulafā’ wa ’l-mulūk by Nāṣir al-Dīn Shāfi‘ b. ‘Alī (1251–1330) (via Ibn al-Furāt's Ta’rīkh al –duwal wa ’l-mulūk); Abū Shāma's Rawḍatayn (d. 1268); Ibn Khallikān's Wafayāt al-a‘yān (d. 1282); Ibn al-Furāt's Ta’rīkh (d. 1405); Ibn Khaldūn's Kitāb al-‘ibar (d. 1406); al-Maqrīzī's Khiṭaṭ (d. 1442) and Ibn Taghrībirdī's al-Nujūm al-zāhira (d. 1470).Footnote ²⁵ All agree that Ṣalāḥ al-Dīn found an enormous literary collection, amongst other treasures at the palace, but its size (where mentioned), or his methods of disposal, are differed upon widely. I do not deal here with the number of books or volumes reported as such, which are often tropological, symbolic or figuratively ‘rounded up’ in any case, but with the question of disposal.Footnote ²⁶ The earliest extant description of the collection is provided by Ibn Ṭuwayr, a Fatimid official whose truncated account, echoed in most of the later sources, is taken from his lost chronicle of the Fatimid and Ayyubid dynasties, Nuzhat al–muqlatayn fī akhbār al–dawlatayn (the Fatimid section of which has now been reconstituted from citations in later sources), and does not mention the fate of the Fatimids’ books, though he witnessed the period of transition from Fatimid to Ayyubid rule, and indeed served both administrations.Footnote ²⁷

The earliest of the authors to describe the vicissitudes that befell the Fatimids’ literary assets, ‘Imād al-Dīn al-Iṣfahānī (d. 1201), Ṣalāḥ al-Dīn's private secretary and one of his three chief advisors and ‘spokesmen’,Footnote ²⁸ describes in detailed reports extant in Abū Shāma's extract and in al-Bundārī's summary, his personal experience of acquiring books from those collections, and he refers to books purchased from the Fatimids’ royal collections in 572/1176, a fact supporting the assertion that Ṣalāḥ al-Dīn sold the books over a period of time, ten years according to Abū Shāma, and sold twice weekly in at least the six years after the Ayyubid takeover in ‘Imād's report; these accounts undermine the theory of immediate wholesale disposal or even destruction once the collection fell into Ayyubid hands.Footnote ²⁹ The next earliest of these authors, Ibn Abī Ṭayy, also comes to us through later quotations; in this instance, his account, extracted from his lost chronicle on Ṣalāḥ al-Dīn, is related at length by Abū Shāma in the first volume of his Kitāb al-Rawḍatayn, and briefly in al-Maqrīzī's Khiṭaṭ. Footnote ³⁰

According to the sources above, the following possibilities explain the fate of the Fatimids’ royal book collections:

Of these narratives, those earliest and closest to the historical events of the 1170s are collected by both Ibn al-Furāt and – more so – by his predecessor Abū Shāma, who relates reports from eye-witnesses to Ṣalāḥ al-Dīn's early policies. One such eyewitness is an emir named Shams al-Khilāfa, who appears in the early volumes of Ibn al-Furāt's chronicles (volumes 1–3 of the 9-volume autograph MS) in the latter's account of late Fatimid rule. This emir's father lived under late Fatimid rule and relayed accounts of it to his son; Shams al-Khilāfa's accounts were transmitted to both Abū Shāma and Ibn al-Furāt by the intermediary figure of the Aleppan Shi‘i chronicler Ibn Abī Ṭayy.Footnote ³¹ Alongside other unspecified ‘Egyptian sources’ [al-Miṣriyyūn], the emir informs Abū Shāma that some of the wealth left by the Fatimids was distributed by Ṣalāḥ al-Dīn among his coterie of men, and that he sold much of the rest, including the contents of the reputedly greatest library then known in the Islamic world, the collection at the Fatimids’ caliphal palace, which was, incidentally, the only major library complex in Cairo to have partially survived the above-mentioned year of looting and inter-regimental violence that had erupted over a hundred years earlier in 461/1068–9.Footnote ³² As well as iterating the well-known descriptions of these book collections as containing well over a million books and many hundreds of copies (though perhaps ‘volumes’ is meant here) of al-Ṭabarī's history alone,Footnote ³³ Shams al-Khilāfa's report provides several unique titbits of information that bring clarity to the vague outline of Ṣalāḥ al-Dīn's disposal of the Fatimids’ assets, including their books, supplied by other sources, and the more substantial textual evidence of their survival embedded within the composition of chronicles.

Abū Shāma's informant describes the collections that fell into Ṣalāḥ al-Dīn's hands in 1171,Footnote ³⁴ and then reconstructs related events as follows:

1. 1. Qāḍī al-Fāḍil, a lover of books, was allowed to survey the collections and take ‘a great quantity’ of them.

2. 2. Ṣalāḥ al-Dīn cut off the bindings from many of the rest that were in a good condition, to sell or re-use the leather, and threw the unbound contents into a lake.

3. 3. Other volumes were sold to the public,Footnote ³⁵ over a period of ten years.

4. 4. When news of the books in the lake became common knowledge, ‘people’ recovered those, too, because ‘[books] are revered items’.

5. 5. Much of the corpus that was initially discarded was thereby preserved, and this is how ‘that which we have [today] was attained’.

The sequence described here suggests, and permits tentative conclusions, that:

1. (i) Ṣalāḥ al-Dīn's method of dispersing the Fatimids’ book collections – those of the palace libraries in particular, which are mentioned by name in many sources – was ad hoc rather than planned;

2. (ii) that he freely allowed Qāḍī al-Fāḍil to take what he could for his own use;

3. (iii) that he permitted monetary value to be derived from the rest, even if this entailed tearing valuable covers from the books and discarding their contents.

The recovery of this last mentioned portion of books by ‘the people’ (which could refer to ordinary, socially unspecified members of the public, or, more likely, middle-class trading/mercantile-class and perhaps even scholarly-class Cairenes, in Maqrīzī's classification, who would appreciate the value of the contents of these books) is significant: though we have little palaeographic evidence that water-spoiled books from this era were re-copied and their contents thus saved, the description certainly fits well with the evidence from the historiographic sources of some earlier texts surviving into the Mamluk era partially intact and others whole.Footnote ³⁶ Historiographical texts might only have formed a fraction of the contents of the Fatimid libraries; but their survival into the Mamluk period suggests that Ṣalāḥ al-Dīn did not expunge all or even most vestiges of the Fatimids’ literary corpus, historiographic and otherwise. This line of investigation resolves the paradox expressed by al-Maqrīzī, who suggests in the Itti‘āẓ that Egyptian books on the Fatimid caliphate are hard to find; meanwhile, both his Itti‘āẓ and his Khiṭaṭ subvert his argument and bear witness to the richness of the source base for the Fatimids available to the historians of their era and earlier.Footnote ³⁷

The testimony of Shams al-Khilāfa is, alongside ‘Imād al-Dīn's report cited by Abū Shāma, the earliest surviving first-hand account of this episode available. His mention of Qāḍī al-Fāḍil's role in the transfer of books from the royal palaces, together with that of (minimally) three other chroniclers, is a significant point that requires further expansion, and I will turn to Qāḍī al-Fāḍil's role in this context, as well as his influence on Ṣalāḥ al-Dīn's policy in general, presently.Footnote ³⁸

5. Fatimid Historiography in Mamluk Chronicles

In addition to these transparent discussions of the fate of the Fatimids’ books in the Arabic chronicles, can we turn to the general content of the chronicles themselves to seek indications of how much Fatimid-era historiography (though largely not works in other genres) survived the demise of the dynasty? An early phase of (thus far unpublished) historiographical research, in which the present author had an opportunity to edit and discuss the account of early Fatimid history produced by a Fatimid-era Shafi‘ite state official, al-Quḍā‘ī (d. 1062), in the surviving mukhtaṣar (digest) of his work,Footnote ⁴⁸ revealed immediately the inadequacy of the idea that the Fatimids’ history books fell into oblivion after 1171. It did so by demonstrating that while al-Quḍā‘ī's universal chronicle, ‘Uyūn al-ma‘ārif, is no longer extant, some of the main threads of information in the mukhtaṣar of that work still present (in at least two redactions) were successfully preserved and used in their more complete forms by at least nine major Ayyubid and Mamluk authors. In other words, this particular early Fatimid historiographical account continued to circulate widely after Ṣalāḥ al-Dīn's supposed destruction of the Fatimids’ book collections in the interval between the death of the last Fatimid caliph al-‘Aḍid in 1171 and the inception of the new Ayyubid sultanate in 1174. It became clear to me at that juncture that further scrutiny of Ayyubid and Mamluk chronicles covering Fatimid rule might yield not dissimilar results as those provided by al-Quḍā‘ī's epitome.

Subsequent research by the present author into the later phase of Fatimid historiography (c. 1100–1170) as represented in Mamluk chronicles, especially in the first three unpublished volumes of the Cairo-based Hanafite historian Ibn al-Furat's (d. 1405) multi-volume universal history mentioned above, the Ta’rīkh al-duwal wa ’l-mulūk, a holographic unicum overwhelmingly composed of ascribed reports from earlier textual witnesses, also sheds light on the range, extent and nature of Fatimid-era historical reports as they were preserved by the historians of Mamluk Egypt. Ibn al-Furāt's chronicle offers proof from a text additional to and earlier than al-Maqrīzī's (d. 1442) more seamless Itti‘āẓ al-ḥunafā’, which is similarly founded upon native Fatimid-era sources, that Fatimid historiography was a robust, creative genre that enjoyed authority as well as longevity, despite the seeming paucity of Fatimid chronicles as a record of, if not a justificatory vehicle, for Fatimid rule – a paucity both in terms of what was produced and what survives. Like so many Mamluk chronicles, the Ta’rīkh al-duwal reveals that the dramatic motif of library destruction has become in this instance a truism, and an image that abides, despite an array of indications to the contrary, for its seems to capture the ideological flavour of a period of Sunni resurgence within Egypt.Footnote ⁴⁹

Yet the chronicles overall tell a nuanced story, laying bare the superficiality of the view that the Sunni ‘hero’ Ṣalāḥ al-Dīn entered Fatimid Egypt and wiped out its ‘heretical’ literary legacy; or its obverse, that the anti-Isma‘ili ‘villain’ Ṣalāḥ al-Dīn systematically jettisoned or destroyed the Fatimids’ valuable intellectual heritage. The diverse accounts of his takeover in Egypt, examined in detail by Lev, reveal something of his general circumspection in shoring up power for himself, and that his approach to his new political roles was far from straightforward, characterised by his “feeling his way” into them, free from either sincere service to the waning Fatimid house or the overlordship of Nūr al-Dīn.Footnote ⁵⁰ The portrayal of Ṣalāḥ al-Dīn's policy in regard to the vast stores of books left behind by the Fatimid ruling house, in the dual contexts of late Fatimid and early Ayyubid Egypt narrative history, and the depiction of Ṣalāḥ al-Dīn more generally by authors from eras ranging from the late Fatimid period to the late Mamluk (c. 1150–1500): these are themes that require further elucidation, and I address them next, alongside an appraisal of the textual evidence we currently possess of substantial works of Fatimid-era historiography that clearly did survive well into the Mamluk period, but perhaps not after the fifteenth century. After that stage, we have no record of their continued presence, either as extant archaeological objects or as bodies of reports transmitted to and within later works.

To return to one particularly fulsome Mamluk account of late Fatimid and early Ayyubid rule, Ibn al-Furāt's narrative of that era (c. 1130–1180), like those of the Kitāb al-Rawḍatayn of Abū Shāma (d. 1267) before him and Itti‘āẓ and Khiṭaṭ of al-Maqrīzī (d. 1442) after him, collects and preserves a portion of what circulated in his times of the main strands of Fatimid-era historiographic material, without which the dynasty's last years would be poorly understood. For Ṣalāḥ al-Dīn's rule proper (1174–1193), he uses conventional sources: al-Muntaẓam of Ibn al-Jawzī (d 1201), ‘Imād al-Dīn al-Iṣfahānī's reports (d. 1201) (apparently via the Rawḍatayn of Abū Shāma) and al-Kāmil fi ’l-ta’rīkh of Ibn al-Athīr (d. 1233) amongst others.Footnote ⁵¹ Where his own voice comes through, that is to say where he offers an opinion rather than a quotation, the Hanafite Ibn al-Furāt expresses a rather muted admiration for Ṣalāḥ al-Dīn and his new policies, such as the establishment of new Sunni institutions of learning in Cairo in 1170, where none had existed previously (while Alexandria had had Sunni madrasas since the vizierate of Riḍwān in the 1130s.)Footnote ⁵² At the same time, the chronicler apparently evinces no ill will towards the Fatimid imams for this state of affairs, merely suggesting that while their rule endured, their creed prevented them from fully catering to the needs of the Sunni population: a position characterised by an absence of sectarian language or attitude.Footnote ⁵³

Do these three above-mentioned works, by Abū Shāma, Ibn al-Furāt and al-Maqrīzī, encompass the full range of Mamluk perspectives on late Fatimid history, however? A rudimentary but reliable litmus test of how well the Fatimid period is documented in Mamluk historiography is whether or not we have a detailed and multi-sourced chronological account of Fatimid rule: judging by the outline provided in the Mamluk record, which can be positively tested against limited corroboration or ‘control’ by external sources of evidence such as Geniza papers, chancery records and archaeological evidence, it appears that we have, and a rich account of the Fatimid era in its entirety at that, from a wide range of authors.Footnote ⁵⁴ The areas where mystery or confusion persist, for instance the cacophony of discordant views on the succession to the caliph al-Āmir in 1128–9 registered by Ibn al-Furāt and fellow Mamluk annnalists,Footnote ⁵⁵ are but familiar variations on the political succession problems of Islamic political life recorded across Islamic historiography as a whole, as Kennedy describes in relation to the chronicles of the late ‘Abbasids.Footnote ⁵⁶ A multiplicity of voices reveals the comprehensiveness of the Mamluk record of Fatimid rule as it survives to the present day in as much as it reflects the conflicting interests and perspectives that prevailed in the late Fatimid political milieu.Footnote ⁵⁷

The conundrum regarding the Mamluk preservation of earlier historical reports is expressed well by Brett when he notes that while most Fatimid-era documents and chronicles are no longer extant, it is clear that they did survive, in substantial quantities, into the Mamluk period, and their subsequent loss is the real question we have insufficient evidence to answer.Footnote ⁵⁸ While that issue is not addressed directly by the contents of Ibn al-Furāt's chronicle, or indeed by this article, as it requires an examination of the working methods of the medieval Arab historian – in the vein of Hallaq's investigation of the working practices of the medieval qāḍī Footnote ⁵⁹ – the myth of the disappeared Fatimid corpus is certainly one that the Ta’rīkh al-duwal exposes, as would further research into other substantial Mamluk-era works that preserve earlier reports or documents, such as al-Maqrīzī's Itti‘āẓ and Khiṭaṭ, and the Ṣubḥ al-a‘shā of al-Qalqashandī. This is especially true in cases where we can examine ‘first and final versions of one and the same work side by side’,Footnote ⁶⁰ such as the Khiṭaṭ, which allow original Fatimid-era sources to be identified, and the precise deployment of earlier material to be traced.

6. Fatimid Historiography and Ṣalāḥ al-Dīn's Early Policies

Does the survival of Fatimid historiography shed further light on early Ayyubid policy, to supplement the data explicitly offered by the chronicles, and exploited in the wide-ranging corpus of ‘Ṣalāḥ al-Dīn literature’ in English and other European languages? How, indeed, has modern scholarship on early Ayyubid rule dealt with the issue of the ‘regime change’ of the 1170s, and its ideological implications as expressed in the remaining textual sources that firstly record the events of that historical juncture, and secondly reveal how much Fatimid historiography was accessible to Ayyubid and Mamluk authors? The reportage in Fatimid, Ayyubid and Mamluk sources of Ṣalāḥ al-Dīn's treatment of the Fatimids’ books is evidently varied and not without contradictory elements; it is, however, consistent with the narrative variations that characterise the historiography of Ṣalāḥ al-Dīn's first years in Egypt as a whole (1169–74), a period prior to the better-attested remainder of his time as sultan in Egypt (1174–1193), when his official biographers presented more detailed, continuous and politically-motivated accounts of his policies.Footnote ⁶¹ In other words, the person of Ṣalāḥ al-Dīn, an immensely important figure in history, one reputed to have changed the course of the Crusades, to have ‘returned’ Egypt to Sunnism, and to have inaugurated a mode of leadership that was to pave the way for the more long-standing military-political regime of the Mamluks, is not a figure who is readily understood upon reading the medieval accounts of his life. The ubiquitous reports on his early policies are similarly problematic, being simultaneously facile and opaque by dint of the alignments of the chroniclers.

Against this textual and historical background, Ehrenkreutz attempted to reconstruct an account of Ṣalāḥ al-Dīn's planning and execution of a coup d’état in Cairo in the late-560s/early-1170s (which brought him to power in Egypt, albeit in ‘stages’), based on this confusion of discordant early materials, and to thereby revise and indeed overturn previously-held benign views of him as an ‘accidental’ or even an ‘incidental’ hero who resurrected effective leadership in Egypt once the Fatimid caliphs lost every vestige of sovereignty and charismatic appeal, and their viziers any semblance of political control.Footnote ⁶² However, the selective use of early historical reports for Ṣalāḥ al-Dīn by Ehrenkreutz, noted elsewhere,Footnote ⁶³ leaves the issue of the nature of his ‘step into the breach’ in Egypt unresolved, and Ehrenkreutz is as much at the mercy of the regional, confessional, political and personal biases of the medieval chroniclers as anyone; these biases are prominent in standard works such as those of Ibn al-Athīr or Abū Shāma, even if the latter avows an ‘eirenic purpose’ in his dual biography of Nūr al-Dīn and Ṣalāḥ al-Dīn.Footnote ⁶⁴ In contrast, Lev's study of Ṣalāḥ al-Dīn's re-Sunnification of Egypt is highly conscious of the axes that chroniclers would have wished to grind.Footnote ⁶⁵ Yet he, too, brings to the fore the ambiguity of Ṣalāḥ al-Dīn's position in relation to both the Fatimid house and Nūr al-Dīn Zengī, and the duality (as distinct from duplicity) of his policy towards both. The matters of Ṣalāḥ al-Dīn's attitude towards the Fatimids he initially served, and the assessment of his policies in general, remain delicate yet controversial issues, though recent efforts, which explicitly aim to re-examine the ‘Saladin legend’ have presented rounded, well-referenced and nuanced views.Footnote ⁶⁶

On the specific question of the Fatimids’ books and Ṣalāḥ al-Dīn's way of dealing with them, Ibn Abī Ṭayy transmits one of the earliest accounts of this episode, as set out earlier. As an Imami Shi‘i, he might be expected to attach opprobrium to Ṣalāḥ al-Dīn's single-minded pursuit of the re-Sunnification of Egypt. This could help to explain his characterisation of Ṣalāḥ al-Dīn as high-handed in his insistence that from the year 1171, the Friday sermon be read out in the name of the ‘Abbasid caliph, first at the foundational mosque of ‘Amr, and then elsewhere in the capital. Yet, overall, it is far from clear that Ṣalāḥ al-Dīn planned his takeover in Egypt, in defiance of both the few remaining Fatimids and their loyalists, and his former overlord Nūr al-Dīn Zengī, very tightly or early on after his arrival there. The chronicles present conflicting evidence on this score. Advocates of both Ṣalāḥ al-Dīn as a zealous anti-Isma‘ili awaiting his chance to stage a takeover, and Ṣalāḥ al-Dīn as a cautious politician who acted out ambiguities towards the Isma‘ilism of the Fatimids ever-present on the ground in Egypt, can find historiographical reports in the medieval sources that would appear to support their interpretations.Footnote ⁶⁷

The difficulty with even the earlier testimonies adduced here is that one might succeed in identifying the bias of the chroniclers, but filtering out the ‘facts’ of the matter is a problematic exercise; a decisive view of the events of the 1170s may remain elusive. This is especially evident in view of the narrative refractions that characterise Islamic historiography – especially Fatimid-era, as discussed by M. Brett – in which primary factual elements are re-interpreted and embellished in successive source-types so as to render them more figurative than factual, and then later reintegrated into ‘factual’ historiographical contexts.Footnote ⁶⁸ This micro-genesis of historical reports is the more pronounced in relation to Ṣalāḥ al-Dīn and the Fatimid libraries in view of Lewis’ suggestion that the overplaying of Ṣalāḥ al-Dīn's role, in some medieval sources, as the destroyer of the Fatimids’ legacy, fits in with a wider use of library destruction as a triumphalist motif in Islamic historiography, in this context on the part of the mainly Sunni authors who recount his first years in office in Egypt, first as the vizier to the Fatimids and then as the Ayyubid sultan.Footnote ⁶⁹

These caveats notwithstanding, the twin facets of textual evidence examined here, namely, the reports describing the dispersal of the royal book collections of the Fatimids, and the demonstrable historiographic continuity from Fatimid through to Ayyubid and Mamluk times, collectively serve to bring into focus significant aspects of Ṣalāḥ al-Dīn's attitudes towards the Isma‘ilis and their legacy in Egypt. A Sunni-centred narrative of victory might rejoice in the dispersal of the Fatimids’ treasures and books. In practice, however, Ayyubid and Mamluk sources more often than not lament the necessity of selling these assets to raise funds, and, in the case of Abū Shāma, express relief that so many books were nonetheless saved by one means or another.Footnote ⁷⁰ No distinction is made, in the discussions of the Fatimids’ books cited here, between ‘heretical’ Isma‘ili books or other literary genres; some Shi‘i books from the Fatimid collections were certainly part of the Ashrafiyya endowed library mentioned earlier;Footnote ⁷¹ though we do see a more clearly pronounced sectarian angle on this corpus on the part of al-Ṣuyūṭī later in the Mamluk period.Footnote ⁷² Many Ayyubid and Mamluk historians, men of learning who, for all their Sunni commitments, valued learning for its own sake, clearly rejoiced in the recovery of the Fatimid-era corpus, much of which was far from exclusively ‘Shi‘ite’, as an eyewitness description from Ibn Ṭuwayr describes: even the fiqh books spanned all the major Sunni as well as Shi‘i schools of jurisprudence.Footnote ⁷³

For many of these preserved or reclaimed sources, Abū Shāma is the conduit, and his inventory of late Fatimid-era sources including eyewitness accounts both reveals the scope of surviving works and presents an example of first-hand sourcing that later authors followed. For if we see echoes of Ibn al-Furāt's attitudes, method and sources in al-Maqrīzī's works of history, as several modern scholars have done,Footnote ⁷⁴ so we see the influence of the Rawḍatayn on the Ta’rīkh al-duwal, which constitutes a chain of textual witnesses, from Abū Shāma through to Ibn al-Furāt through to later authors including al-Maqrīzī, that allows us to appreciate the robust survival of Fatimid-era historical reports in substantive Mamluk-era works.Footnote ⁷⁵ This methodological influence inheres in several aspects of the Ta’rīkh al-duwal: the layout of the reportage, with clear attribution for the most part (though Ibn al-Furāt is more assiduous in the practice than Abū Shāma), and in the use of Ibn Abī Ṭayy's Ma‘ādin al-dhahab, which, true to its name, is a goldmine of unique reports on late Fatimid as well as early Ayyubid Egypt, from which Ibn al-Furāt provides reports that are earlier than Abū Shāma's since his history begins some decades earlier. In addition to the sources used by Abū Shāma and other historians who mediate Fatimid-era sources for Ibn al-Furāt (for example Ibn Khallikān [d. 1282], from whose Wafayāt al-a‘yān Ibn al-Furāt copies reports from Ibn Ẓāfir's [d. 1216] Akhbār al-duwal al-munqaṭi‘a), there are those Fatimid-era sources that Ibn al-Furāt apparently had independent access to, notably Ibn al-Ṣayrafī's al-Ishāra ilā man nāla ’l-wizāra, Ibn Ṭuwayr's Nuzhat al-muqlatayn, and, uniquely, the aforementioned treatise he calls Akhbār al-dawla al-Miṣriyya wa mā jarā bayn al-mulūk wa ’l-khulafā’ min al-fitan wa ’l-ḥurūb min ayyām al-Āmir ilā ayyām Shīrkūh [Account of the Egyptian dynasty, and the tribulations and wars that afflicted the kings and caliphs, from the era of al-Āmir to that of Shīrkūh, c. 1130–1170],Footnote ⁷⁶ an eyewitness account with detail and local colour not available elsewhere in the Ayyubid or Mamluk historiography of late Fatimid Egypt. These sources evidently did endure well into Mamluk rule and appear clearly in the historographical record.

One might quibble that Ibn al-Furāt's chronicle, on which the argument of Fatimid historiography's survival in this article has largely though not exclusively relied, is the exception that proves the rule, that other Ayyubid and Mamluk chroniclers did not adduce the breadth of evidence he contrived to use and exploit. Yet the ubiquity of Fatimid-era reports in later sources (not least in least Abū Shāma's and al-Maqrīzī's accounts) – even if not all of them cite as much original or unique material as Ibn al-Furāt – proves that as far as the sources indicate, the historiography produced in the Fatimid era did not wholly or even substantially fall victim to the change of regime. Indeed, works in other genres cited as lost, in the medieval sources, were later found to be a part of the corpus that survived the period of looting in 1068, and in fact, late Fatimid sources such as Ibn Ṭuwayr's Nuzhat indicate that the royal collections had been restored since that time, with holdings as vast as or greater than any others in the Islamic world.Footnote ⁷⁷ The works of Ayyubid and Mamluk historians do indeed shed intriguing light on the political worlds they describe, and not just through their explicit accounts of history. Their very makeup tells a political story of sorts.

That said, and to further the argument I made earlier, the politico-religious policies of Ṣalāḥ al-Dīn are well-known but described in a complex web of often tendentious historiographical sources that are ipso facto slippery and subversive of the prospects for unequivocal interpretation. Should we read the establishment of waqfs as a way for Ṣalāḥ al-Dīn to re-Sunnify EgyptFootnote ⁷⁸ when Gary Leiser finds, in his in-depth study of Egypt's return to Sunnism, little reason to suppose that he built madrasas to counter Shi‘ism as such, and that a papable reason behind the establishment of new legal colleges by Ṣalāḥ al-Dīn and several emirs of the Ayyubid state, as expressed in the sources, was to bolster their power and influence in the new regime?Footnote ⁷⁹

These two facets of Ṣalāḥ al-Dīn's religious policies are compatible, for each set of indications in the texts presumably points to an aspect of reality; nonetheless, the lack of definition in the picture, as projected by primary and secondary sources (as elaborated by Lev, drawing on the works of Maqrīzī, Ibn Wāṣil, the Bustān al-jāmi‘,Footnote ⁸⁰ Ibn Abī Ṭayy via Abū Shāma and a host of others) is hardly helped by the multifarious threads of detail surrounding Ṣalāḥ al-Dīn's attitudes towards those who threatened his personal hegemony and that of the new state: within Cairo, the main opposition to him does not appear to have come chiefly from Isma‘ili or even Shi‘i quarters, despite his heavy-handed response to the supposed plot of 1174 alluded to above;Footnote ⁸¹ had he intended to quash Shi‘ism at all, he would have tried to suppress it in its most populous region of Upper Egypt, especially in Qus, but, to all accounts, did not.Footnote ⁸² The Niẓārī ‘Assassins’, on the other hand, who attempted to murder him on several occasions, are described by him, in statesmanly rhetorical mode, as one of the three ‘enemies’ against whom he is trying to defend Islam (the other two being the Zengid rulers in Mosul and the Franks).Footnote ⁸³ The Fatimids do not feature in Ṣalāḥ al-Dīn's reported formulation; the Niẓārīs are a force to be reckoned with not because of their religious claim or because they are a residual Shi‘i group but because of their attempts on his life and their political ambition. Several later sources cite another putative statement of his, namely, that one of his two main missions in life was to rid the Nile Valley of the Isma‘ili heresy (the other being the removal of the ‘infidels’ from Jerusalem). In practice, however, his fiscal and endowment policies appear to be designed to, amongst other things, wipe out the influence of Armenians who had prospered in Egypt during the tenure of several powerful Armenian viziers (chief among them Badr al-Jamālī, d. 1095 and his son al-Afḍal, d. 1121) rather than an erasing of Fatimid influence per se.Footnote ⁸⁴ On the other hand, his fiscal reforms pursued the agenda of diverting wealth away from the structures designed by the Fatimids to remunerate their elite, towards new financial instruments through which wealth could be released into a more socially equitable system, in keeping with his avowed aim of establishing principles of social justice, and secondly of allowing him to repay the loyalty of the soldiery that sustained his power. The waqfs he established over the course of the years following his new sultanate all point to these aims.Footnote ⁸⁵

Thus the assumption that Ṣalāḥ al-Dīn worked to wipe out the Fatimid stamp upon Egypt reveals its fragility on deeper scrutiny. One might in fact deduce from the sources that by the time Ṣalāḥ al-Dīn took up the sultanate, having seemed acquiescent to treading the admittedly precarious tightrope of vizier for the last two years of Fatimid rule, when Nūr al-Dīn's expectation that he would hand Egypt over to him persisted, Isma‘ilism was so attenuated a force in Egyptian life that he hardly needed to take a pronounced stance against it. His policies would do the necessary work for him over the course of time.Footnote ⁸⁶ The evidence examined by Frenkel suggests that he did indeed attempt to liquidate the assets of the Fatimids as he did those of all his adversaries, current and former, but his motives were as much economic as ideological, and hence the sale and dispersal of the Fatimids’ books collections rather than their annihilation.

As for his religious tolerance in general, this is an equally many-sided, equally inconclusive area: medieval Arabic sources are far from consistent on the matter. Some offer straightforward panegyrics, while others describe how dreadful a foe he was, and how cruel he could be, apparently as one mode of eulogy amongst others.Footnote ⁸⁷ Aside from the official biographers, whose descriptions form the backbone of Mamluk accounts of Ṣalāḥ al-Dīn's life and rule, alternative voices also abound. Usāma b. Munqidh, whose autobiography, Kitāb al-i‘tibār, recounts many incidents from late Fatimid and early Ayyubid history as he himself experienced them, and who probably harboured some pro-Isma‘ili sympathies as strongly implied in a passage of his quoted by al-Dhahabī, presents Ṣalāḥ al-Dīn as an awe-inspiring and uncompromising ruler who strikes fear into his opponents.Footnote ⁸⁸ Conversely, William of Tyre portrays him as ruler given to leniency towards his captives and great generosity,Footnote ⁸⁹ while Geniza documents offer the impression that his stances towards religious minorities, in particular the Jews, was tolerant, especially so in contrast to the policies of the ‘zealous’ Mamluks.Footnote ⁹⁰ One is left noting these conflicting perspectives without the opportunity of reaching a firm conclusion, dissatisfying as this may be, since the primary and some secondary sources have personal stances to postulate, not merely in relation to the outline of history or the character of Ṣalāḥ al-Dīn himself, but also in relation to the narrative modes and historiography that has portrayed him over the centuries. A range of narrative and more limited documentary evidence attests that several important Fatimid-era works survived the fall of the dynasty; if Ṣalāḥ al-Dīn had intended to efface the Fatimid written record (a dubious assumption, as numerous sources indicate), his attempts to do so were half-hearted if not indifferent, and need to be viewed against the wider context of his trying to keep in balance a variety of political, religious and financial aims.