Documentary Constellations in Late-Mamlūk Cairo: Property- and Waqf-Related Archiving on the Eve of the Ottoman Conquest of Egypt

Abstract The scholarly discussion of archives in the premodern Islamicate world is beset by problematic generalisations. Such a view to some degree stems from a top-down view of archiving that focuses on state archives at the expense of practices of archiving occurring outside a chancery context. This article challenges the assumptions that support an enduring narrative of paucity, by examining non-chancery archival practices in Mamlūk Cairo on the eve of the Ottoman conquest in 922/1517. In doing this, it looks to some of the surviving original documentary material: legal property deeds with connections to waqf endowments whose potential to shed light on archival history has largely remained untapped. Surviving in large numbers in modern collections in Cairo, these documents contain abundant traces of their own archival histories. By presenting a micro-scale case study drawn from this material, this article shows the energetic and meticulous documentary and archival practices that surrounded property transactions in late-Mamlūk Cairo.


Introduction
The predominant scholarly narrative surrounding the archives of the premodern Islamicate world is one of paucity. The overwhelming non-survival of state archives from before the advent of the Ottoman Empire means that surviving documentary traces are often presented as rare exceptions to a general rule of loss. The acclaimed archival prolificacy of the Ottomans and their extensive state bureaucracy is thus contrasted with the putatively slim and fragmentary extant documentary record of earlier periods. 1 In the Arabic-speaking lands it is the Mamlūk Sultanate which bears the brunt of such a narrative, its own chanceries having furnished little original archival material. 2 The upheavals of the Ottoman conquest of the Mamlūk capital of Cairo in 922/1517 are, in fact, in part credited for this paucity of surviving original Mamlūk-era material, with chancery archives assumed to have been destroyed in the process of violent regime change. While such a catastrophic view of the transition period is increasingly challenged, 3 the contrast between early modern Ottoman abundance and medieval Mamlūk paucity still endures. 4 In spite of this, research on the archives of the pre-modern Islamicate world has now moved beyond this narrative of paucity. Recent contributions have advanced two strong contentions: that in fact abundant documentary material does survive from prior to the Ottoman period, and that by broadening our analytical focus from the scrutiny of archives as fixed sites or institutions to a broader appreciation of archival practices, we can access a whole variety of fascinating archival histories. 5 To continue to broach this history with a line of questioning that repeatedly asks "why did the archives not survive?" is, thus, to do a disservice to such scholarship. The unsatisfactory nature of the narrative of paucity is especially noticeable when one moves beyond the chancery, a shift which the participants of the 2018 conference Beyond the Islamicate Chancery: Archives, Paperwork, and Textual Encounters across Eurasia (Fifteenth-Twentieth Centuries) were in various ways exhorted to make. Though extant chancery manuals offer abundant evidence of complex documentary and archival practices existing within the Mamlūk chancery, this in some ways serves to accentuate just how much original material we appear to have "lost." 6 On the other hand, the large majority of documents that do survive originate from outside the chancery, containing extensive evidence for archival practices which can be obscured if we focus too exclusively on the archiving activities of the central state apparatus. The Mamlūk material thus immediately sets itself apart from the Ottoman archival landscape, where state bureaucracies inevitably dominate scholarly discussion of surviving documentation. For those working on the documentary record from the pre-Ottoman Middle East, moving away from the chancery presents an opportunity to fully scrutinise extant original material and, in so doing, to work further towards challenging received ideas about the nature of premodern Islamicate archival and documentary practice. This, in turn, can contribute to a broader and more inclusive picture of the cultures of documentation that distinguish Islamicate history, but which do not always resemble received models of what archives, or bureaucracies, "should" be. 7 In this article, I endeavour to do this by presenting a snapshot of some of the archival practices that prevailed in late-Mamlūk Cairo. In this I rely overwhelmingly on the testimony of surviving original documents: documents which were not the product of the Mamlūk chancery. Specifically, this article offers a case study of the archival practices that emerge from legal documents related to waqf endowments made by the Mamlūk elites. Two substantial collections of such documents survive today in Cairo's modern archives: the National Archive (Dār al-Wathāʾiq al-Qawmīya) and the archive of the Ministry of Religious Endowments (Wizārat al-Awqāf). 8 This waqf-related material has earned considerable fame within the field of Mamlūk history, serving as one of the most important sources for the reconstruction of the urban sociopolitical and architectural history of Mamlūk Cairo. 9 Its scrutiny from the point of view of archival history is, however, only in its infancy. 10 This article offers insights into some of the archival practices that are visible through examination of this material. The examples I present are selected to reflect an historical moment immediately prior to the Ottoman conquest of Cairo, after which the city lost its political and administrative pre-eminence, becoming merely a provincial capital in the expanding Ottoman Empire. In light of the presumed archival disjuncture between the Mamlūk and Ottoman periods, this is a moment in archival history which merits examination.
When exploring archival practices in the premodern Islamicate world, waqf-related material represents a pertinent case study. Waqf endowments involved the immobilisation of private property, the revenues of which were theoretically dedicated in perpetuity to the upkeep of pious and educational institutions, charitable causes, or to the family and future descendants of the founder (wāqif/wāqifa). These endowments sat at an important point of intersection in Mamlūk society, between the political, financial, and charitable interests of the Mamlūk elites, the religious, educational, and juridical milieus of scholars and legal practitioners, and the wider commercial, spiritual, and communal life of the city. Waqf therefore represents an institution in which a large cross section of society held a stake, and whose effective management was of common interest. 11 Even more significantly, waqf has also been seen as having a particularly close link to processes of archiving, with waqf endowments considered particularly conducive to practices of long-term archival preservation. This supposition hangs in a large part on the legal perpetuity of waqf endowments, expected to give documents generated during the endowment process longer legal, and therefore archival, life-spans than those connected with other document-generating processes. 12 In fact, as we shall see, such hypotheses are complicated by the nature of the documents preserved in Cairo's collections, which should not be seen solely through the lens of waqf, but must be connected to a wider range of property transactions. It is clear to me that legal theory is not alone sufficient to explain archival practices, which depend on an enormous range of historically contingent variables, from the political, social, economic, and legal, to the simply practical. Indeed, scholarship on waqf is increasingly shifting away from a narrowly legalistic view of this institution, recognising the diverse social practices it generated in different historical settings. 13 Such questions do not ultimately lie at the core of this article, which instead focuses on examining the various archival practices that waqf-related documents can reveal. 14 The wider significance of waqf and the ubiquity of waqf-related documentation in the Islamicate world at large means that questions arising from the consideration of the late-Mamlūk material may have a significance that reaches well beyond the geographical and temporal boundaries of the Mamlūk Sultanate.
In what follows, I rely upon a sample of documents preserved today in the Wizārat al-Awqāf, but many of whose characteristics are shared by documents housed in other collections. I begin by introducing the corpus, outlining the limited previous scholarly endeavours to examine the archival practices it exhibits. Moving beyond this, I highlight the importance when dealing with this corpus of determining the long-term property histories the documents reveal to us, through the lens of which we can access their own archival lives. In the second and third parts, I elucidate this connection by presenting a micro-scale study of a small group of documents linked to each other by their common treatment of a small pocket of commercial property in Mamlūk Cairo. I use these documents to trace the ownership and endowment trajectory of this property, moving on in the third part to examine the various aspects of archival practice that emerge: the deliberate rapid generation of written records, the varied conceptual configuration of groups of documents in archives of different types, and practical tools devised to aid their navigation. This case study offers us some precise historical context, which allows us to connect the archival practices the documents reveal to more specific phenomena: namely, the approach of the penultimate Mamlūk sultan, Qāniṣawh al-Ghawrī, towards building up his massive waqf endowments. Defining these documents as a micro-scale property archive, I argue that they show evidence of a dynamic series of archival practices surrounding the legal status of private and waqf property on the eve of the Ottoman conquest. Such evidence can, in its own small way, contribute to nuance the received narrative regarding archiving in the Islamicate world.

The Archival Lives of Waqfīyāt and Waqf-Related Deeds
The waqf-related documents housed in the Cairo collections, with very few exceptions, represent legal genres, mostly deeds concerning property. Such documents were drawn up with heavy oversight from professional legal practitioners: qadis and their notarial staff. The traces of this close supervision are ubiquitous on the documents themselves, which consist of lengthy paper scrolls and contain numerous witness statements and attestations, as well as the verification of qadis. 15 Many of the deeds are also attested with isjāls, records that authenticate the judicial proceedings detailed on the documents, thus providing legal evidence that the qadi recognised the transaction's validity. 16 Particularly important documents, such as the waqf deeds of sultans, were attested by qadis representing each of the four madhhabs, presenting a long chain of isjāls which added further legal weight to the transactions they authenticated. 17 Such features facilitated the use of these documents as evidence in the event of a legal challenge by enhancing their ability to stand scrutiny in a sharīʿa court. 18 These deeds, their format and content, were thus heavily shaped by the formal conventions of contemporary legal documentary practice. They did not, however, belong to a qadi's archive. They were, instead, drawn up by legal professionals at the behest of the individuals carrying out the transactions, and are normally understood to have belonged to a waqf or household archive. 19 Their design as scrolls probably reflects this, being intended for preservation by diverse households and in various locales, rather than constituting part of a centralised judicial archive. 20 The production, use, and archiving of these documents thus drew together a whole range of individual interests: the buyers, sellers, and wāqif/wāqifas, as well as qadis, notaries, and scribes.
As one the largest and best-known extant corpora of pre-Ottoman Arabic documentary material, these documents have regularly been referred to in discussions of archiving. 21 Nonetheless, such mentions have seldom been accompanied by close study of the original documents themselves, nor with detailed discussions of the nature of the material these collections contain. One major exception to this can be found in the publications of Julien Loiseau, who, having worked very closely with these collections, highlighted some of the archival phenomena that emerge from scrutiny of waqf endowment deeds (singular: waqfīya, plural: waqfīyāt). In his observations, he emphasised aspects of document design. He commented on the archival nature of the documents themselves: long scrolls upon which records of multiple transactions were written, and which were carried out and recorded on the same scroll on different dates. Looking at such documents, one can perceive an extended archival life-span, sometimes stretching over more than a century. The first deed written on the support constituted its "documentary matrix" (matrice documentaire), to borrow Loiseau's terminology, lending the scroll its logic, and becoming the axis around which future records added to the scroll would revolve. The materiality of the scrolls also reflected this logic, with the documentary matrix taking up most of the width of the scroll, and subsequent additions placed in the large right-hand margin, or in empty space below the initial deed. Waqf deeds were later supplemented with additional marginal records of various types: adding property to the original endowment with further waqf deeds, extracting property from the endowment by means of istibdāl (exchange), 22 laying out changes made to the legal stipulations of an endowment, or detailing matters of litigation concerning the waqf. In light of such features, Loiseau concluded that each waqf deed represents a "veritable archive" in itself, collating the relevant legal information for the endowment into an "integral form" or summary. 23 Such a document could thus serve as a complete record of the waqf endowment, from the time of its foundation and onwards into a perpetual future.
The document-as-archive is not, as Loiseau is undoubtedly aware, a feature unique to waqf deeds, though his analysis focuses on these particularly striking scrolls. Physically, waqf deeds represent some of the most impressive extant documents. The deed for sultan al-Ghawrī's large funerary complex, for instance, reaches some thirty metres in length. 24 The predominance of waqf deeds in the Cairo collections is, however, overstated. These collections contain many different documentary genres, though the distinction between these different document types is often muddied by the way the collections are referred to in the scholarly literature: as "a collection of endowment deeds," "waqfīyāt/waqfīyas," or more ambiguous formulations such as "waqf documents." 25 Descriptions such as these can exaggerate the centrality of the waqf institution for the history of this material, thus colouring how we perceive its archival history. While waqf is an essential key to understanding the historical and archival logic of documents such as these, it must be seen as one facet within the broader range of property transactions that prevailed in late-Mamlūk Cairo.
In the Wizārat al-Awqāf collection, for instance, just 20 percent of the documents are actual waqfīyāt. 26 Even so, of the remaining 80 percent, the vast majority do exhibit clear connections to a waqf endowment. The extended life-spans of these documents, flagged by Loiseau, mean that a document that began its life apparently unrelated to a waqf endowment can become connected to waqf over time. These connections often emerge towards the end of a much longer documentary life-span: for instance, a scroll whose initial documentary matrix is a sale deed eventually recording the later incorporation of the property into a large waqf endowment. There are many such documents in the Wizārat al-Awqāf (49 percent of the collection), concerning discrete pockets of real estate in Cairo which, at the end of an extended transaction history, were immobilised (mawqūf) in waqf endowments. These deeds often follow the progression of the property in question through the hands of multiple owners before its final immobilisation in waqf, and each transaction is recorded in an addition to the original scroll. 27 Other deeds were initially produced to record istibdāl procedures (24 percent of the Wizārat al-Awqāf collection). In such cases, property that had previously been immobilised in a waqf endowment was extracted and passed on to a new owner, who could then sell it on as private property (milk). Many of these documents also show the properties ending up immobilised in another waqf endowment, thus tracing an arc in the property's history, from one waqf endowment to another. It is to accommodate these complex documentary life-spans that I use the precise, if cumbersome, "waqf-related documents." With this terminology I acknowledge that the relationships of the documents in the collections to waqf endowments can be complex. To carry out a full investigation of the archival practices that characterise such a corpus we must take the entire documentary life-span into consideration.
What the documents in the Cairo collections, whether waqfīyāt or other genres, certainly reveal is just how closely their archival lives mirror the transaction histories of the properties they deal with. The trajectory of a property or endowment over an extended period of time can be traced in the additions made to the scrolls, which themselves represent the culmination of an array of documentary and archival practices. With additions to older scrolls, for instance, we are alerted to processes of document accumulation and the copying of details from older documents, and thus to the continuous multiplication of documents, as well as their assembly and rationalisation in archives. The connection of the documents' archival logic to property histories, in turn, shows them to be deeply intertwined with the activities and strategies of numerous individuals: from the Mamlūk sultans and other elites involved in property transactions to the legal practitioners and notaries charged with producing the documentary record. As we shall see, this has profound archival implications for a corpus rooted in an historical period marked by an intense interest in property accumulation and endowment.

A Tale of Four Shops
In what remains of this article, I present a case study drawn from the corpus of waqf-related documents housed in the Wizārat al-Awqāf. The documents I discuss detail the sale and endowment trajectory of a small pocket of urban real estate, four shops in the papermaker's market in Cairo (arbaʿ ḥawānīt al-kāʾin bi-sūq al-warrāqīn) which were purchased and sold on together during the early decades of the tenth/sixteenth century. During this period the sultanate was ruled by its penultimate sultan, Qāniṣawh al-Ghawrī, who, as we shall see, played some role in determining the fate of these four shops. The documentary material discussed here takes us to a micro-level where we can uncover the shifts in ownership and legal status that this property underwent over the course of just over eight years. By following the way this property history is recorded in the documents, we can in turn trace the archival lives of the documents themselves, which I will address in the next section. This case study thus offers a concrete example of the practices flagged above, revealing the ways in which they can become visible to us through close scrutiny of the extant documentary record.
Our first encounter with these four particular shops comes in a modest-sized paper scroll of about a metre in length, initially drawn up to record an istibdāl procedure which was completed on 28 Dhū 1-Ḥijja 913/29 April 1508. 28 The shops had previously been endowed to support the waqf of Jamāl al-Dīn al-Ustādār (d. 812/1409), a senior amir who served as the manager of the sultan's household (ustādār) from the reign of sultan al-Nāṣir Faraj (801-8/1399-1405). 29 Around a century after the establishment of his waqf, then, the extraction of these four shops from the endowment via istibdāl was finalised, and they passed into the private ownership of another individual, al-Zaynī Sallāma ibn ʿUmar al-Minshāwī. By the time this document was being drawn up, istibdāl procedures were widespread in the Mamlūk capital. In spite of its dubious legality, this mechanism had the advantage of allowing individuals to continue buying and selling property in a city where available real estate was increasingly permanently immobilised in perpetual waqf endowments. Derelict or badly maintained buildings could, by means of istibdāl, be brought back onto the market, their demolishment providing the space for new structures, including the imposing architectural complexes that benefitted from sultanic and amiral waqfs. 30 Istibdāl documentation could perhaps be seen as the archetype of "waqf-related" material, its very existence dependent upon, and indicative of, the prevalence of waqf endowments in Mamlūk society.
The istibdāl procedure was itself a document-heavy process. It was initiated by the submission of a petition to the qadi, who could then delegate deputies to take charge of the process, inspecting the property concerned for evidence that it was no longer of tangible benefit to the waqf. 31 It is difficult to tell from the content of extant documents, whose texts are formulaic and follow standard patterns, the exact physical conditions of the particular properties they concern. Nonetheless, for istibdāl to be permissible, it was theoretically necessary to prove that the property was no longer yielding sufficient revenues, or otherwise that it was in a dangerous state of disrepair. 32 Some extant istibdāl documentation thus bears extensive traces of these inspection and approval processes, revealing how these successive stages were recorded and certified in writing, sometimes over a period of several months. At times, the records of a single istibdāl procedure could spread across multiple separate scrolls, furnishing an extremely thorough and comprehensive documentary witness to proceedings. 33 The documentary manifestations of several of these stages of istibdāl procedure are visible on the scroll examined here. While the documentary matrix is a deed of istibdāl, this is not the first text that one sees upon unrolling the scroll. At the top, we can find the original petition that initiated the istibdāl process, glued to the scroll. 34 The deed of istibdāl is then written below this, its formula outlining the whole lengthy procedure, and making a clear reference to the petition "attached above" (mulṣaq bi-aʿālīhi). The attaching of the petition to the legal deed offers us some insight into the way the producers and custodians of these kinds of documents attempted to rationalise the multiple records of such complex procedures. The petition was of a different documentary genre to most of the other texts on these scrolls. While these other texts represent a range of deeds and legal procedural elements written and notarised within the tightly regulated setting that surrounded a qadi and his staff, the petition at least theoretically stemmed from outside this legal notarial world. Such a text was not subject to the extensive witnessing and verification procedures required of the legal material and served an essentially communicative purpose. This difference in function is also reflected in the materiality of the petition, a short rectangular document which followed its own set of norms for its wording and layout. 35 The gluing of this petition onto the same support as the istibdāl deed, then, points to a calculated attempt to synthesise the various documentary traces of this complex, multistage legal procedure. Physically attaching the petition to the same material support served to generate a comprehensive and integrated record of the procedure, and one which incorporated the various different kinds of documentation that the procedure generated. Such practices present a glimpse of some of the pressures shaping prevailing archival practice. One can, for instance, imagine that this was a judicious method of preserving a small petition in archival spaces geared more towards the preservation of scrolls than loose paper leaves. This one small instance of istibdāl thus already offers a glimpse of the ways in which the archival requirements that such documents raised were dealt with in practical terms.
This istibdāl was not, however, the end of the story for these four shops, nor even the last legal transaction that they underwent on this one day. Additional records added to the same scroll inform us that the property was sold on again the very same day in 913/1508. Over the following year, the shops changed hands a further three times, finally landing, on 3 Jumādā I 915/19 August 1509, in the private ownership of the then sultan, Qāniṣawh al- Ghawrī (r. 906-22/1501-16). The final record, added to the scroll seven years later, shows these four shops being endowed in waqf to the benefit of al-Ghawrī's large funerary complex, which comprised a mosque, madrasa, Sufi hospice (khānqāh), primary school, and drinking fountain, and still stands today in Cairo. 36 Notably, this occurred just months before the sultan's death in battle against the Ottoman forces in Syria. The full ownership trajectory of these four shops is outlined below in Figure 1. As I am, at the time of writing, not in possession of images of this scroll, its layout is schematically represented in Figure 2, illustrating the way in which records of subsequent transactions were progressively added to the original deed. 37 The eventual resting place of these four shops, in the waqf of sultan al-Ghawrī, is not exceptional. The importance of this particular sultan for the material housed in the Wizārat al-Awqāf is well known. 38 As many as 54 percent of the documents preserved here reveal explicit connections to this sultan, a high survival rate certainly linked to the proximity of the end of his reign to the advent of Ottoman rule in previously Mamlūk territories, and the introduction of different documentary and archival practices. 39 These documents either record the details of al-Ghawrī's own waqf endowments, 40 or concern discrete properties that passed through his hands, and usually subsequently into his waqf. 41 Al-Ghawrī's waqf strategies thus play a profound role in shaping this corpus. This is significant, as al-Ghawrī's approach to waqf property has earned some attention. Though al-Ghawrī was the penultimate Mamlūk sultan, he was the last to establish any significant waqf endowments. 42 Relying in part on the enormous documentary record left by these waqfs, Carl Petry argued that this sultan tread new ground in his financial strategies, undertaking a mass expropriation of property on a scale not attempted by previous sultans in order to augment his own private wealth. 43 Certainly, Ottoman-period chroniclers subjected him to a damning critique, accusing him of corruption and embezzlement, and of ultimately precipitating the decline of Egypt. 44 The comparative scale of his property accumulation may be difficult to glean from the sample of documentary material that survives, which tells us so much more about his strategy than it does about those of his predecessors. Nonetheless, the traces of the process endure in the hundreds of documents that survive in the Wizārat al-Awqāf, including the istibdāl scroll described above.
Returning to the four shops in the papermakers' market, it becomes evident that sultan al-Ghawrī's designs stretch further back into their ownership trajectory than the moment when the properties actually fell into his hands. The identity of their second owner, a certain ʿAbd al-Barr ibn al-Shiḥna, offers a reliable indication that the final destination of these shops, in the waqf of the sultan, was already anticipated when the istibdāl took place. Ibn al-Shiḥna was the Ḥanafī qadi of Cairo during the first decade of the tenth/sixteenth century, and appears with some regularity in the surviving documentation, though usually as the qadi overseeing transactions or a notarial witness, rather than a contracting party. As Petry pointed out, Ibn al-Shiḥna looms large as one of the primary masterminds of the sultan's waqf enterprises, using his considerable skills of legal manipulation to facilitate a project that may not have fallen quite in line with waqf regulations as they were outlined in the law books. 45 The appearance of Ibn al-Shiḥna early on in this chain of transactions suggests that the entire property trajectory outlined above might be considered as one instance of al-Ghawrī's notorious property expropriations, on this occasion using the mechanism of istibdāl to get hold of property previously tied up in waqf. 46 In such an enterprise Ibn al-Shiḥna's legal expertise was undoubtedly a boon, giving a legal veneer to property transactions that might, by some, have been considered questionable.
Our impression of the calculated nature of this project is further augmented by the penultimate stage in the property's trajectory: its sale to al-Ghawrī. At this moment too, the identity of the previous owner is of significance. Al-Naṣirī Muḥammad ibn Taghrībirmish was a son of one of al-Ghawrī's close associates, who served as vizier for much of his reign. 47 It is also possible that al-Sayfī Kasbāy, who sits between Ibn al-Shiḥna and Ibn Taghrībirmish in the above ownership trajectory (the buyer in transaction 3), might actually be the vizier himself, though the absence of a full name makes this difficult to ascertain. 48 If this were so, we would witness here Ibn Taghrībirmish purchasing the shops from his own father, before selling them on to the sultan. More meaningful though is the role that Ibn Taghrībirmish himself seems to have played in al-Ghawrī's accumulation of property. In the Wizārat al-Awqāf there are, in fact, no fewer than seventeen documents all recording the sale of property by Ibn Taghrībirmish to the sultan, several occurring on the same day as the sale of these four shops: 3 Jumādā I 915/19 August 1509. Ibn Taghrībirmish seems, therefore, to have acted as an agent of sorts, accumulating property for the sultan. As Petry has pointed out, this was a role that a number of individuals played in the establishment of al-Ghawrī's waqf endowments. 49 While Petry characterises this process as exploitative on the sultan's part, to me the picture that emerges from the ownership trajectory above is more redolent of complicity by these individuals, who actively participated in this project. 50 The small-scale property history that this scroll reveals, then, allows us to situate it in a precise and well-defined context, defined by the approaches of the sultan and his close circle of associates to the accumulation and endowment of urban real estate.

Accumulating Papers in the Papermaker's Market
Having identified the somewhat convoluted ownership trajectory that these four shops underwent, as well as their probable implication in a much larger property accumulation project, we can now examine how these are reflected in archival terms. As we have seen, this entire trajectory is visible to us from just one single scroll: the istibdāl deed described and illustrated above. Thus, in the way Loiseau outlined, this one scroll can serve as an archive recording the extended transaction history of these four shops over a period of nine years. For today's historians hoping to use these documents as a lens onto late-Mamlūk history and the political strategies of its elites, this is an enormously valuable archival feature, even if uneven survival rates dictate that we interpret them with care.
Further evidence in the Wizārat al-Awqāf collection, however, alerts us that such document-archives were never intended to stand alone, and were in fact discrete components in much more substantial property archives. The collection houses at least three further documents that contain traces of the same property trajectory outlined above, revealing a more complex network of documentation in which this scroll represented only a single element. The first of these additional documents is another similar paper scroll, this time initially drawn up to record transaction 2 in Figure 1 above: that is, a sale deed recording the transfer of the four shops to Ibn al-Shiḥna's ownership on the same day as the initial istibdāl. 51 Like the istibdāl document, the initial sale deed represents the documentary matrix of this scroll, with the subsequent transactions appearing in the margins. A third scroll is also extant, this time drawn up to record transaction 3 in the table, the sale of the property to al-Sayfī Kasbāy. 52 Original scrolls recording transactions 4 and 5 in the table do not survive in the collection. The final transaction is, however, probably to be found on al-Ghawrī's massive waqfīya. 53 Owing to the length of this scroll, I was unable to locate the mention of these particular four shops. Nonetheless, the deed contains the details of each property that was endowed in favour of al-Ghawrī's funerary complex, including references to the original deeds that proved the property was legally in the sultan's private ownership, so the shops almost certainly can be found within this lengthy text. These four shops in the papermaker's market thus merited mention on numerous separate supports.
The survival of this chain of written records pertaining to the very same series of transactions alerts us to the comprehensive and highly meticulous nature of legal documentary practice within the context of late-Mamlūk Cairo. For one, it shows us the rapid proliferation of documents that accompanied regular shifts in property status. In just the first day in the chain of transactions outlined above, two full-length deeds were produced for transactions 1 and 2, 54 and another for transaction 3, 55 thus furnishing three separate metre-long scrolls over the course of around a week. Such fast-paced document production points to a real drive to furnish a full and detailed legal record of these property transactions. This is particularly intriguing if, as I suggested above, the property was always intended to end up in the hands of the sultan. Against such a contextual backdrop, we might see this documentary acquisitiveness as an integral element in this scheme, the copious legal documentation serving to mask the traces of the sultans' own ambitions.
The extant scrolls also reveal remarkably systematic ways in which this complex expanding property archive was assembled and rationalised. The series of scrolls examined here show an extensive duplication of information, with marginal updates being added to older property records in order to ensure that all surviving documentation reflected the current status of the property concerned. As well as rendering each individual document a complete and up-to-date record, these marginal additions constitute a comprehensive system of cross-referencing. Within the text of the marginal additions, explicit references are made to the original documents drawn up to record each transaction. The details of a property, its physical description and new ownership, are usually provided on the evidence of (bi-dalāla) another cited document, and in accordance with the testimony of the qadi in his isjāl (ḥasabamā yashhadu bi dhālika isjāluhu), ordinarily written on the verso of the cited deed (al-musaṭṭar bi ẓāhir al-maktūb al-madhkūr). 56 These cross-references usually also provide the dates on which the qadi verified the cited proceedings. They are, furthermore, witnessed with the testimonies and signatures of court notaries, who confirm the information on the evidence of these separate original deeds, designated "the original document" (maktūb/kitāb al-aṣl), which they had presumably consulted. The scrolls drawn up for transactions 2 and 3 above, for instance, are explicitly mentioned in the marginal texts of the earlier istibdāl deed (for transaction 1). Documents updated in this way were thus clearly intended to work in concert with a wider constellation of separate deeds. The extensive and systematic cross-referencing serves as both a way of authenticating the chain of transactions each scroll outlines, and as a finding aid for the appropriate original legal record, thus acting to clarify and rationalise the ties between the individual pieces.
The methodical nature of this system of updating and cross-referencing is even clearer when we take into account the identities of the scribes who undertook this procedure. While the same scribe did not add all the updates to the scrolls in question, we find that on each of the transaction dates outlined in Figure 1, all three of the extant scrolls regarding these four shops were updated by the same individual. The final addition to each of these scrolls, for instance, on the occasion of the incorporation of the shops into al-Ghawrī's waqf, is written by a single scribe, a certain Muḥammad al-Munāwī. 57 His handwriting is distinctive, being singularly inelegant and challenging to decipher. 58 The systematic efforts of scribes such as he reveal to us a prevailing culture of legal documentation surrounding property that seems to have been characterised by a desire or requirement for extreme comprehensiveness.
When exploring the archival practices in evidence in this corpus, the site of document preservation remains stubbornly elusive. In the collection preserved in the Wizārat al-Awqāf there are few indications of where documents of these kinds were kept. Only for some of the large waqfīyas do we have clear indications, which suggest that they were preserved within the buildings that the waqf was established to support. Al-Ghawrī's enormous waqfīya, for instance, was kept in the khānqāh in his large funerary complex. 59 Even without a confirmation of the site of archival storage of each surviving document, though, it is possible to identify meaningful archival configurations, which can feed into our understanding of the physical composition of the archive. We can, for instance, reasonably classify the scrolls discussed above as a micro-scale property archive, each one an intrinsic and indispensable documentary element in the growing record of these four shops' legal status. Like the property the documents describe, the documents themselves share a trajectory, to which their common updating by a single scribe eloquently attests. Whether the documents were physically passed between the hands of each of the properties' owners is difficult to say. Given the extreme rapidity of shifts in these properties' ownership, this seems somewhat unfeasible for these particular scrolls, and the logic uniting them is certainly the property itself, rather than its rapidly changing owners. Nonetheless, if my identification of al-Sayfī Kasbāy is correct, we might posit a temporary home for these scrolls in a household archive of the vizier Taghrībirmish and his sons. 60 The properties were in the hands of members of this family for over a year (from 6 Muḥarram 914/7 May 1508 to 3 Jumādā I 915/19 August 1509) which, though not a very long period, still necessitated a location for their safekeeping. Konrad Hirschler has noted the importance of small "private" archives for the preservation of state documents. 61 It is easy to imagine that such practices were mirrored with legal documents, especially those detailing property in the private ownership of individuals who were themselves heavily involved in the running of the state.
The preservation of these particular three scrolls in such a household archive would only have been short-lived. The final visible stage in their documentary life-spans bears witness to a meaningful shift in their archival configuration, transitioning from a small and discrete property archive to elements in a much more substantial archive: that of sultan al-Ghawrī's waqf. As I explained above, documents connected to al-Ghawrī's waqf represent a large extant corpus. More significantly still, the scrolls discussed in this article represent three of a total of 111 documents concerning property that was endowed in favour of the sultan's waqf on the very same day: the 18 Rabīʿ II 922/21 May 1516. 62 This was the last date upon which al-Ghawrī added to his large waqf endowment, which he had built up progressively over the course of his reign. The mass of extant documents that name this date offers clear evidence of the comprehensive exercise in archive formation and rationalisation that accompanied the establishment of al-Ghawrī's waqf endowment. 63 This process involved the drawing up of a waqfīya, collating the information contained in this much broader constellation of property deeds to form a real archival matrix. The scribal additions to old documents in turn show us how the huge corpus of individual property deeds was brought into the framing of the waqf by systematically updating them with details of their new endowed status. 64 This waqf archive was thus brought together, bestowing the two kinds of documents that it contained-a "new" waqfīya and older property deeds-with a meaningful internal coherence.
This process is clearly illustrated by traces of practical archival tools added to the documents and seemingly designed to enhance the functionality of this potentially unwieldy collection of scrolls. Many of the older deeds that made it into al-Ghawrī's waqf archive contain archival filing notes, written in an upper corner, either on the inside or outside of a scroll, delineating the individual property that the document concerns. Unfortunately, the position of such filing notes, on an exposed part of the scroll, has made them vulnerable to damage and loss over the intervening centuries. None of the three documents I have detailed above contain such filing notes. 65 Other comparable scrolls do, however, show that such notes usually gave concise details about the genre of deed and the property dealt with in the scroll. Two scrolls drawn up to record the stages of an istibdāl procedure in the year 858/1454, for instance, contain notes that read: deed of istibdāl; half and a quarter [three-quarters] of a building located on the Bayn al-Qaṣrayn, known as [part of] the waqf of al-Bajāsī (maktūb istibdāl; al-niṣf wa al-rubʿ min bināʾ bayna al-qaṣrayn al-maʿrūfa bi waqf al-Bajāsī). 66 The filing notes that are extant reveal the continued relevance of individual properties in structuring the waqf archive. In this particular case, even the traces of the legal and archival memory that the property once belonged to a different waqf endowment appear in the filing note. This configuration reflects the significance of individual properties for the legal record of a waqf, these separate ownership trajectories representing potential sites of legal dispute which could throw the validity of the waqf into doubt. The surviving filing notes show how small property archives, such as the three scrolls concerning our four shops, maintained their integrity within al-Ghawrī's waqf archive, though subordinated to the waqfīya, which henceforth became the ultimate record of the waqf endowment.
From the insights that emerge in this small case study, it is difficult to escape the impression of an historical milieu marked by exceptionally dynamic archival practices. Rapid shifts in the legal status of private property resulted in the continual multiplication of documents. The occasionally bewildering constellation of scrolls that resulted thus fed into the development of methods devised to rationalise the legal record. Systems of crossreferencing enhanced the functionality of growing and mutating property archives. The eventual incorporation of property into a waqf endowment saw the documents entering into new archival configurations. In the examples I have discussed here, the documents, previously a small property archive, became elements of a much more substantial waqf archive, in much the same way as the property itself became part of the waqf endowment.
The roles of particular high-profile individuals-the sultan himself, as well as his close associates amongst the late-Mamlūk legal and political elites-hint at strategic legal motivations driving the extremely comprehensive nature of this documentation. The perceived need to disguise the means by which al-Ghawrī accumulated the property to support his waqf, for instance, may have fed into the numerous and somewhat convoluted shifts in ownership that properties such as these four shops underwent. On the other hand, the phenomena of rapid document multiplication and archival cross-referencing fit into the broader range of documentary procedures that constituted contemporary legal practice. They cannot, therefore, be explained solely by reference to the agendas of al-Ghawrī and his contemporaries. Whatever role these agendas played, it is clear that the particular legal-historical milieu surrounding private property transactions as well as waqf-related procedures during this period was reflected in a complex and meticulous range of documentary and archival practices.

Conclusion
In this article I have investigated the archival practices that emerge from some of the waqf-related documentary material that is housed today in the Wizārat al-Awqāf in Cairo. This corpus, far from being exclusively concerned with waqf endowments, offers a lens onto a much wider range of property-related activities in Cairo, most importantly the exchange of private property not yet immobilised in waqf. Through examining the frequent changes in legal status that such properties underwent, we can in turn glean illuminating insights into the documentary and archival practices that accompanied such transitions. Presenting a microscale case study of the documentation connected to four shops in the Cairene papermakers' market, I have been able to frame these practices against their direct historical backdrop, identifying some of the political and legal, as well as archival, pressures that contributed to shape the material. Many of the archival practices I have highlighted, such as the phenomenon of document cross-referencing and the inclusion of filing notes, deserve more thorough treatment than I have been able to give them here. Nonetheless, this article serves to offer a whistle-stop tour of some of the archival concerns that surrounded legal documents connected to property rights in the final decades of the Mamlūk Sultanate.
The documents scrutinised here reveal a milieu in which the demands of archiving seem to have been important considerations for those requiring or producing documentation. The rapid proliferation of documents attests to a need and desire for meticulous records to detail the numerous property transactions taking place. Beyond this, though, it must also in itself have had a profound impact on the necessity for dedicated archival tools, such as practical methods of identifying and cross-referencing documents and ways of materially rationalising complex records. Identifying the multiplication of documents that accompanied the sale of private property qualifies, to some degree, the presumed archival exceptionalism of waqf that I raised in the introduction to this article. Waqf certainly fed into the development of distinctive archival configurations. We have seen this in al-Ghawrī's waqf archive, which was made up of a mass of property deeds and small property archives logically centred on a focal waqfīya. Nonetheless, the nature of private property which, unlike waqf property, was free to be sold on or subdivided for subsequent owners, also had its own unique impact. 67 The generation of new documents on each occasion of sale fed directly into the acceleration of archive formation. The link between property transactions and archival practices that this corpus so eloquently reveals should, therefore, be understood as dependant on an intricate nexus comprising the legal framing of transactions and their broader social and political milieu, as well as practical concerns.
The historical sources made use of in this article represent only a micro-sample covering a period of less than ten years. In fact, the Mamlūk-era material in the Wizārat al-Awqāf collection in its entirety is heavily biased in favour of these same early decades of the tenth/sixteenth century, containing more documents from these decades than the entire preceding century. 68 This is to be explained by the preponderance of al-Ghawrī's waqf archive. Despite the ease with which we can explain this phenomenon, however, it evidently problematises the degree to which we can make meaningful comparative generalisations with earlier or later periods. The archival dynamism I have identified in the above case study may not be unique to this period and political context, despite the close connections we can draw with the strategies of al-Ghawrī and his associates. Our view of the sultan's strategy itself is based on this sample, the likes of which we do not have for any of his predecessors. Even so, observing the apparently high investment of energy in creating and preserving a comprehensive legal documentary record that these documents show still serves to complicate the narrative of Mamlūk-to-Ottoman archival practice with which I began this article. Certainly, it is not a straightforward matter of paucity versus abundance. Such a stark distinction fails to recognise the prevalence of multiple domains in which archiving was a real and pressing concern, outside the sphere of, for instance, fiscal documentation at the chancery level. Archiving was rarely an end in itself, but the significance of the waqf institution in late-Mamlūk Cairo, and the preoccupation with property ownership that accompanied it, created fertile ground for an extreme documentary acquisitiveness, at least amongst the political elites. This case cannot, of course, substitute for an exploration of archiving in the Mamlūk chancery. It does, however, offer a different perspective, without which our picture of the archival history of the late-Mamlūk period, and our understanding of premodern Islamicate cultures of documentation, would remain incomplete.

Acknowledgements
I would like to thank the organisers and all the contributors to the 2018 conference at the Austrian Academy of Sciences, Beyond the Islamicate Chancery: Archives, Paperwork, and Textual Encounters across Eurasia (Fifteenth-Twentieth Centuries), who offered thoughtful input to an early version of this article. The research of which this article represents a part was supported by a Bloomsbury Colleges Scholarship, a Doctoral Scholarship at the Annemarie Schimmel Kolleg "History and Society during the Mamluk Era (c.1250-1517)," Universität Bonn, and by the Cluster of Excellence "Understanding Written Artefacts: Material, Interaction and Transmission in Manuscript Cultures," funded by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) at the Centre for the Study of Manuscript Cultures (CSMC) at the Universität Hamburg. Thanks to Konrad Hirschler for comments on an earlier version, and to my two anonymous reviewers for their valuable suggestions.  " 225-68. 4 It should be noted that negative comparisons are even more regularly made with medieval and early modern Europe, a reflection of the persistence of Eurocentric approaches in scholarship and an unfortunate but stubborn legacy of Orientalism. 5 Contributions to this discussion include: