Towards a Comparative Study of Libraries in the Manuscript Age

Copyright 2023. De Gruyter. All rights reserved. May not be reproduced in any form without permission from the publisher, except fair uses permitted under U.S. or applicable copyright law. EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 2/6/2024 8:47 AM via AN: 3535290 ; Nuria de Castilla, Franois Droche, Michael Friedrich.; Libraries in the Manuscript Age Account: s3628809

Libraries in the Manuscript Age

Studies in Manuscript Cultures Edited by Michael Friedrich, Konrad Hirschler, Caroline Macé, Cécile Michel, Jörg B. Quenzer and Eva Wilden Volume 29

Libraries in the Manuscript Age Edited by Nuria de Castilla, François Déroche and Michael Friedrich

The research leading to these results has received funding from the European Research Council under the European Union Funding for Research and Innovation (Horizon 2020), ERC Grant Agreement number 670628, “Saadian Intellectual and Cultural Life” (SICLE, 2016–2021). ISBN 978-3-11-077210-4 e-ISBN (PDF) 978-3-11-077965-3 e-ISBN (EPUB) 978-3-11-077977-6 ISSN 2365-9696 DOI https://doi.org/10.1515/9783110779653 This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. For details go to https://creativecommons.org/licenses/by-nc-nd/4.0/. Library of Congress Control Number: 2022951399 Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the internet at http://dnb.dnb.de. © 2023 with the author(s), editing © 2023 Nuria de Castilla, François Déroche and Michael Friedrich, published by Walter de Gruyter GmbH, Berlin/Boston. This book is published with open access at www.degruyter.com Printing and binding: CPI books GmbH, Leck www.degruyter.com

Contents Nuria de Castilla, François Déroche, Michael Friedrich Towards a Comparative Study of Libraries in the Manuscript Age 1 The Islamic World John Seyller The Operations of the Imperial Mughal Library 17 Nuria de Castilla Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection at El Escorial François Déroche, Lbachir Tahali Collecting Books in Eighteenth-Century Morocco: the Bannānī Library in Fez 63 East and South Asia Jean-Pierre Drège Two Libraries of the Tang Capital 107 Ivo Smits Institutional Libraries in Japan’s Classic Court Age (Heian Period, 794–1185) 123 Gérard Colas Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century: The Sarasvatī Library in Chidambaram 135 Byzantium Luciano Bossina How Many Books Does It Take to Make an Emperor’s Library? Constantine VII Porphyrogenitus and a Chapter of History of the Manuscript Book

Contents Inmaculada Pérez Martín Byzantine Libraries: The Public and the Private 183 Western Europe Donatella Nebbiai How Private Libraries Contributed to the Transmission of Texts Jacques Verger Libraries and Teaching: Comments on Western Universities in the Middle Ages 211 225 Vanina Kopp An Ideal Library for an Ideal King? Showcasing the Collection, Organisation and Function of the Royal Louvre Library in Late Medieval Paris 243 Index of Manuscripts

Nuria de Castilla, François Déroche, Michael Friedrich Towards a Comparative Study of Libraries in the Manuscript Age In a general sense, the word ‘library’ denotes a ‘place set apart to contain books for reading, study or reference’, and, by extension, the ‘books contained in a “library”’.1 Collecting and storing books in a ‘place set apart’, whether a shelf or a niche in a wall, may have started soon after the first books were produced. However, evidence for such places only dates back to the middle of the third millennium bce, when the archaeological record shows first traces of them both in Egypt and Mesopotamia, for the latter in close relationship to scribal archives.2 Even if their relationship to archives (literary ‘books’ vs. administrative ‘documents’) is not as straightforward as today’s institutions make believe,3 there is little doubt that the history of libraries covers more than four millennia. Libraries were established and maintained by individuals, such as rulers, priests, scribes and scholars, or by institutions, such as states, temples, monasteries and families. Access was more often limited than not; the public libraries of ancient Greek or Roman cities seem to constitute the exception to the rule prior to modern times. Depending on the size and purpose of collections, various tools for using them were created, ranging from the spatial arrangement of the items to inventories and eventually catalogues listing them, often according to an order based on size, content or other factors. In many cases, these catalogues are the only remains of what once had been impressive collections. The Chinese case may serve as an example: smaller collections, both official and private, certainly existed long before 26 bce when the imperial book collection was enlarged by an official call for submitting writings to the Han court. This collection was then examined and edited, resulting in a catalogue. Eighty percent of what it lists has been lost, much of it probably already during the burning of the imperial palace in 23 ce.4 In 583, an influential friend and high official of the first Emperor of the Sui (r. 581–604) advised him to follow the Han precedent by collecting books and creating a great imperial library in his new capital Luoyang. He mentioned 1 Oxford English Dictionary online <https://www.oed.com> (accessed on 5 November 2022). 2 See the contributions in Ryholt and Barjamovic 2019; see also Clancier 2009; Haikal 2008. 3 For Greco-Roman Egypt, see Fournet 2018, 194–195. 4 Fölster 2016. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-001

Nuria de Castilla, François Déroche, Michael Friedrich four more ‘book catastrophes’ in addition to the fire of 23 ce,5 but probably did not imagine that in 622, most of the new collection would sink in the Yellow River on the ships transporting it to Chang’an, the capital of the Tang (618–907). Its catalogue was damaged and had to be complemented by earlier ones. Both catalogues, transmitted in revised versions in the official histories of the Han and the Sui, have become major sources for reconstructing literary production and the history of textual transmission, not much different to traditions in other regions. In addition, we know about large private libraries which have been dispersed or destroyed, but left traces in literary sources. Complete inventories and catalogues of such libraries in China are extant from the twelfth century on.6 Although many libraries may originally have served rather practical purposes, such as divination, medicine or the art of war, once teaching extended beyond mere acquisition of the art of reading and writing and included literary texts, schools would build small collections as well. When erudition and scholarship came to be valued as domains of knowledge in their own right, libraries became the major habitat for scholarly activities, whether official or private. The library of Alexandria is certainly the best-known and most studied among the book collections of the ancient world: established in the third century bce by the first Ptolemies as a universal library, it was meant to contain a copy of every book in the world. Its head librarians were philologists creating tools and methods for editing, cataloguing and textual criticism, which survived in Byzantium well into the early modern period.7 It is perhaps no accident that two centuries later, the examination of the Chinese imperial book collection also led to the development of tools and methods for establishing texts.8 Therefore, what has been said about the nature of libraries in the Ancient Mediterranean may also apply to those in the Far East: Ancient libraries were created by political and military power, and there is no sign that they powered the development of any kind of intellectual activity independent of it. The librarians of Alexandria and Rome alike were royal or imperial appointments. Most of those we know to have used these libraries were drawn from the ruling elites of the Empire.9 5 Dudbridge 2000. 6 See McDermott 2006, 51. 7 There exists a huge body of literature on this library; in addition to the classic by Luciano Canfora 1990, see Bagnall 2002 for a critical review of sources, and Rico and Dan 2017 for a recent attempt to complement previous scholarship. 8 For a comparison of both libraries, see Nylan 2018. 9 Woolf 2013, 6–7.

Towards a Comparative Study of Libraries in the Manuscript Age According to archaeological evidence, the palace libraries of Assurbanipal (669–c. 630 bce) at Niniveh held about 5,000 scholarly clay tablets; the library of Alexandria may have housed 40,000 scrolls at the time of Caesar,10 and the imperial book collection of the Han had a little more than 13,000 scrolls a few decades later.11 It was, thus, similar in size to the library of Celsus in Ephesus, that was built to store 12,000 scrolls in 117–120 ce.12 The latter was the private collection of a high-ranking official, held about 4,000 scrolls at the end of the second century and was probably one of the largest of its kind. Since the length of scrolls, whether papyrus or bamboo and silk, and the quantity of content may vary to a significant degree and is certainly different to the number of characters written on clay tablets, we do not have enough evidence for comparing the numbers of characters or words they contained. Concerning the physical task of storing such quantities of written artefacts, however, the challenges may have been similar. The victory of Christianity in Western Europe put an end to the large public libraries in Greek and Roman cities, founded by rich citizens, such as Celsus. For almost a millennium, churches and monasteries were the only places for storing books, in most cases not more than a few dozen of the new codex book form, which could be kept in a box. While there is nothing known about the size of Sasanian and early ʿAbbāsid ‘Houses of Wisdom’,13 we do again have figures for China: a late source reports that the Liang court collections comprised 140,000 scrolls, and those of the Sui 370,000 scrolls. The part of the Sui Imperial Library shipped on the Yellow River in 622 amounted to almost 90,000 scrolls. Under the Tang (618–907), the Academy of Assembled Worthies alone is said to have held a similar number of scrolls in 731, surpassing that of the Imperial Library by some tens of thousands.14 The biggest private library of the late eight century held 20,000 scrolls.15 Buddhist monasteries in the capital housed various versions of the Buddhist canon with up to 5,000 scrolls.16 Islamic libraries figure somewhere in between the small collections in the West and the huge libraries in the East. While legendary figures, reaching two million books, are reported for the universal libraries of the caliphs of Baghdad, Cordoba and Cairo, the first comprehensive list of Arabic book titles, the Fihrist, compiled from various collections in the tenth century, lists about 10,000 works 10 For Niniveh, see Finkel, 2019, 369, quoting Weidner 1953; for Alexandria, see Bagnall 2002, 351–356. 11 See Fölster 2016, 72. 12 See Grünbart 2018, 321. 13 See Gutas 1998, 53–60. 14 See McDermott 2006, 49–53, 213–214, n. 16. 15 See McDermott 2006, 50–51, 214, n. 25. 16 See Wang Xiang 2014; for Japan up to 1600, see Kornicki 1998, 362–412.

Nuria de Castilla, François Déroche, Michael Friedrich of about 2,000 authors.17 However, it does not provide numbers of codicological units. The famous Fatimid palace library was sold by Saladin at auctions, with a considerable part going to al-Qāḍī al-Fāḍil, head of the Fatimid chancery and later Saladin’s vizier. In 1184–1185, the bibliophile vizier endowed parts of his collection to a madrasa he had founded in Cairo, and others went to his heirs. In the early thirteenth century, 68,000 volumes are said to have been confiscated from them, in order to fill the newly established palace library. After checking them, 11,800 were returned to the heirs, probably because they were duplicates. Other high officials of the Ayyubids are reported to have owned libraries consisting of 10,000 or 20,000 volumes.18 Under Mamluk rule, Syrian and Egyptian cities experienced the development of a tight network of local libraries only in the thirteenth century. These libraries were not only part of madrasas and mosques, but also of other institutions, such as hospitals, the different types of Ṣūfī convents and mausoleums.19 The larger of these endowed and private libraries held some thousand volumes and, in exceptional cases, up to 7,000.20 Most of the libraries in the Ottoman empire had collections of fewer than 500 books up to the middle of the seventeenth century. After the accidental founding of a library independent of religious institutions in 1678, libraries grew in size but apparently never exceeded those of the Mamluks before the reforms of the nineteenth century.21 In 1594, after Moroccan troops of the Saadian ruler Aḥmad al-Manṣūr had pillaged Timbuktu and brought their booty, including scholars and books, to their capital Marrakech, the famous scholar Aḥmad Bābā (1556–1627) reportedly complained: ‘I had the smallest library of any of my kin, and they seized 1,600 volumes’.22 Contrary to the situation in other parts of the Islamic world, there were no endowed libraries in Northern Nigeria, but only private and family collections, the most famous ones being those of Timbuktu.23 In the early seventeenth century, a royal library was established in Thanjavur by Ragunātha, the last ruler of the Nāyakara (r. 1600–1623).24 It is described by a contemporary as a huge hall where the employees and accountants of the palace 17 See Behrens-Abouseif 2019, 5; see Hirschler 2020, 10–12 on different purposes of fihrists. 18 See Behrens-Abouseif 2019, 7–14. 19 Quoted from Hirschler 2016, 135; during the Ayyubid and Mamluk periods, there were more than seventy mausoleum libraries, see Hirschler 2016, 141. 20 See Behrens-Abouseif 2019, 46–50. 21 See Erünsal 2008, 45–47, 77. 22 Hunwick 1999, 315. 23 See Last 2011, 197–199; and the other contributions in Krätli and Lydon 2011 for other parts of West Africa. 24 For this library, see Wujastyk 2007.

Towards a Comparative Study of Libraries in the Manuscript Age carry out their duties. Although we do not have figures, this library is the only one from that time in South India which has survived until today and is still active, albeit as a public library. The number of its manuscripts has grown considerably in the course of the twentieth century by adding books from the families of pandits. In 2005, the Sanskrit books numbered: Palm-leaf manuscripts: 18,877 Paper manuscripts: 21,068 Prints: 14,186 We know next to nothing about libraries in the subcontinent from earlier times, let alone their size.25 This random survey of the size of libraries in the pre-industrial age is far from complete and represents mainly collections of the ruling elites, since we do not have sufficient data for other social groups. Even if some of the figures may not be wholly reliable and the relationship between codicological and textual units is not always clear,26 the results point to a threshold of thirty or forty thousand units for a large library, which is rarely exceeded, and if so, only in imperial or royal collections usually consisting of more than one repository. This number did not change substantially even after the advent of print and the spread of the hall library in Western Europe: the Herzog August Bibliothek in Wolfenbüttel housed 36,000 volumes at the end of the seventeenth century, most of them prints.27 With very few exceptions, such as the reconstructed façade of the library of Celsus in Ephesus, the buildings housing the book collections of the ancient world have perished completely, including their contents. While book collections from later centuries have sometimes survived more or less complete, the buildings containing them today are usually no older than Renaissance times in the Latin West, such as the one of the Abbey Library of St Gall. Addressing monastic libraries in the Latin West, a publication from 2011 claims: While the history of textual transmission and the history of libraries as book collections seem to have been studied to a useful degree, the state of research on the history of libraries as spaces or buildings for housing books can only be called miserable.28 25 See Delhey 2015 for manuscripts surviving from a twelfth-century Buddhist monastery and Cort 2013 on later Jain libraries. 26 For the cultural meaning of binding booklets into composite volumes see Hirschler 2020, 115–145. 27 See Eisen 2011, 284 28 Wischermann 2011, 93.

Nuria de Castilla, François Déroche, Michael Friedrich This statement about the lack of research on the spaces where books are collected probably holds true for most cultures before the industrial age.29 However, there are noteworthy exceptions including the Mamluk libraries30 and the Southern Song Imperial Library.31 The state-of-the-art of studying ‘the library’ in West European languages is, albeit unintentionally, nicely illustrated by the introduction to a thousand-page volume titled Die Bibliothek – The Library – La Bibliothèque. Thirty-nine articles deal mainly with the medieval and early modern periods in the Latin West and its afterlife, including ‘virtual libraries’. The introductory article ‘Die Bibliothek – Denkräume und Wissensordnungen’ discloses the limits of this enterprise only at its very end: That the volume has its limitations is obvious, but not surprising if one assumes an understanding of a science by human standards, given the subject matter. This is most evident in the primarily occidental view of the library. We are aware of this. However, the fact that there are not more insights into libraries of other languages and cultures is not least due to their size. You cannot enter all the rooms of a library at the same time! However, the doors are open.32 The ‘primarily occidental view of the library’ and the regret of not providing ‘more insights into libraries of other languages and cultures’ are euphemisms, especially when combined with the excuse that this neglect owes much to their extent. This remark is the last paragraph of the introduction of ten and a half pages. There is only one very technical article on an Arabic anthology and one on colonial libraries in South America in the volume, but – ‘the doors are open’. Research on libraries, thus, corresponds to that on other domains of history. In Western languages, there is a plethora of studies on ‘the Ancient World’ and Latin Western Europe,33 followed by those on Byzantium34 and, in growing numbers, on the Islamic world, but little on Orthodox Eastern Europe, Sub-Saharan Africa, the Indic world, East Asia and South-east Asia. Scholarship has focused mainly on the history of scholarship, namely, individual collections and their contents, and on the history of libraries in a certain region or during a particular period. On the other hand, research on East Asian libraries has been 29 There are exceptions, of course, even for the Latin West; see, for example, Coqueugniot 2013, 43–55; Lehmann 1996; the contributions in Nerdinger 2011; and Staikos 2017; for the Ancient Nr East, see Ryholt and Barjamovic 2019, 44–50. 30 See Hirschler 2016, 60–101; Behrens-Abouseif 2019, 52–70. 31 See Winkelmann 1974, 15–18. 32 Speer 2020, XXV. 33 See Staikos 2004–2013 for a six-volume history of the library in Western Civilization. 34 For recent scholarship on Byzantine libraries, see Grünbart 2018, 320–321.

Towards a Comparative Study of Libraries in the Manuscript Age done mostly in the region and published in the region’s languages.35 A linguistic barrier, therefore, keeps different cultural and scholarly traditions separated, complicating exchange, comparison and knowledge about ‘the library’. Comparative studies of libraries are rare. There are promising attempts for the ancient world,36 and volumes on the ‘power’ of libraries, their relationship to knowledge and political power.37 In some cases, studies devoted to only one region also include comparative aspects.38 An attempt has been made to collect terms for manuscript collections used in different cultures,39 and a systematic approach to describe such collections has been suggested.40 Topics for future cross-cultural enquiry seem obvious:41 – Typology of collections, relationship to archives; more specialised collections: tomb libraries (Egypt, China); libraries attached to mausoleums (Celsus, Islamic world), etc. – Intended purposes of collecting – Legal status of collections – Spatial and material dimensions – Organisation of collections: staff, tools and routines – Classifying and cataloguing – Acquisition, selection, censorship – Access to collections: to whom? For which purpose: reading, copying, lending? – Environment of institutions and practices – Loss and decay – Concepts of time, memory, or history that are implied Similar to the ‘medieval period’ or ‘Middle Ages’, the ‘manuscript age’ is not a well-defined term either in scope or time. It serves to distinguish the ‘Gutenberg galaxy’, with its hallmark of typographically and mass-produced printed ‘books’, from an earlier, supposedly imperfect stage of knowledge production and trans- 35 For China, see the three-volume history of libraries of Ren Jiyu 2001. 36 See König, Oikonomopoulou and Woolf 2013; Ryholt and Barjamovic 2019; also see Too 2010 for a more theoretical approach. 37 See Baratin and Jacob 1996; Lehmann 2018; also the section on libraries and databases in Jacob 2007, 605–704. 38 For China, see McDermott 2006, 118–126; Nylan 2018; Stackmann 1990. 39 See Delhey, Lorusso et al. 2015. 40 Fölster, Karolewski et al. 2015. 41 The following list is partially inspired by Friedrich 2018, 430–439.

Nuria de Castilla, François Déroche, Michael Friedrich mission by ‘manuscripts’.42 This approach is influenced by the Western European experience and still prevails in the field of book history. However, as soon as other traditions are taken into account, the picture looks rather different. Print only started playing a decisive role in the course of the nineteenth century in large parts of the Islamic world, in many cases, giving preference to lithography and offset over moveable type. Manuscript production in West Africa continued well into the twentieth century,43 and it is still active in Laos today.44 On the other hand, printing with wooden blocks may have been invented as early as 700 ce in China and had spread there increasingly since the tenth century. This did not, however, mean that manuscripts ceased to be produced, on the contrary, only up to ten percent of the holdings of the large libraries in the twelfth and thirteenth centuries mentioned above were printed. Printed books were still expensive, and scholars continued copying texts for their personal collections, until technological advancement in the sixteenth century led to greater affordability.45 In the twelfth century, manuscripts and prints were housed separately in the building complex of the Imperial Library of the Southern Song (1127–1279) because handwritten books were still held in higher esteem than printed ones.46 Although printing with moveable type had already been described in the eleventh century, it apparently did not play a major role until much later.47 Against this backdrop, the clear-cut line between a ‘manuscript age’ and a ‘print age’ becomes blurred, even if one ignores the technological differences between printing with wood blocks or stones and with moveable type. It does not make sense to define the first appearance of printed books as borderline, since this would exaggerate the role they had. It has been observed that manuscripts continued to be produced and used, even in Western Europe, long after the first printing presses had been established.48 In the present volume, therefore, the ‘manuscript age’ is used in a heuristic manner for the period when manuscripts were still the dominant medium, ranging from the tenth century for China to the twentieth century for West Africa.49 42 The concept of the ‘manuscript age’ was used by Franz Rosenthal (1947, passim) in a seminal publication to specifically define the manuscript transmission of knowledge in the Islamic world. 43 See the contributions in Reese 2022. 44 See Grabowsky 2019 for contemporary libraries. 45 McDermott 2006, 54, 25–39; see McDermott 2011 for precise figures on large private libraries in the province Jiangxi under the Song. 46 Winkelmann 1974, 34–35. 47 See Heijdra 2004, 101–102; McDermott 2006, 13. 48 See Bouza 2001. 49 For China, see Drège 1991; for West Africa see Last 2011.

Towards a Comparative Study of Libraries in the Manuscript Age The aim of this volume, Libraries in the Manuscript Age, is to present and examine, through a dozen studies, the nature and functions of various libraries in the Islamic World, East and South Asia, Byzantium and Western Europe, from the third until the eighteenth century.50 The different papers show convergences, although they deal with very different cultural areas. The association of libraries with places of power stands out clearly (de Castilla, Drège, Kopp, Pérez Martín, Seyller, Smits), more so than that which links them to religious institutions (Colas, Drège, Pérez Martín, Smits). Under these conditions, knowledge often remained confidential, since imperial, royal or princely libraries were not open to the majority of the population. Was access easier in other cases? This question is undoubtedly more difficult to study due to the lack of documents; nevertheless, Pérez Martín concludes that monastery libraries were not accessible, and the texts analysed by Déroche and Tahali show a very clear desire to restrict access to the books to members of the same family. Verger, on the other hand, proposes a broader vision of medieval Europe that made access to books a key to the emergence of a new form of knowledge. Power and knowledge are also associated in another way, that of the control that the former exercised over the content of the latter. Smits, thus, questions which works were selected and which were kept aside, while Bossina evokes an imperial project of an ideal library marked strongly by symbolic considerations. As we read on, we discover the extent to which our knowledge depends not on the books themselves but on crucial but peripheral information: catalogues (Drège, Smits, Verger), loan lists (Pérez Martín, Kopp) or lists of librarians (Seyller), inscriptions and buildings (Colas). It is, therefore, through a strategy of circumvention that we can reconstruct manuscript libraries which were established to protect the books that were deposited there, but did not fulfil the wishes of their founders, since their holdings have not been preserved. Conversely, the contributions of de Castilla and Déroche and Tahali stand out for the place reserved for 50 The idea of the conference that gave rise to this volume arose from the problems posed by the typological diversity within a comparable chronological horizon and the same society that the members of the ERC project 670628 SICLE (Saadian Intellectual and Cultural Life), devoted largely to the study of Moroccan libraries during the Saadian period (sixteenth to early seventeenth century), had identified: the Sultanian library of Mūlāy Zaydān (r. 1603–1623) in Marrakech, now preserved in San Lorenzo de El Escorial (Déroche, de Castilla, and Tahali 2022); that which the sultan’s father, Aḥmad al-Manṣūr (r. 1578–1603), had endowed the Qarawiyyin mosque in Fez (Déroche and Tahali 2022); and finally a library of scholars, the Bannānī, of slightly later creation (Déroche and Tahali in this volume). Faced with the methodological questions raised, the members of the project wanted to open up the reflection and move away from a geographical and cultural compartmentalized approach in order to reflect on the nature and challenges of libraries in different areas.

Nuria de Castilla, François Déroche, Michael Friedrich manuscripts that have been preserved to the present day, even though the libraries of which they were originally part have disappeared. The formation of the library often involves, as has been pointed out earlier, the copying of earlier works (Bossina, Déroche and Tahali) and, once constituted, they represent a source for the transmission of texts, as Nebbiai shows for Western Europe in her contribution of this volume. This case can be brought closer to the famous example of Ibn Khaldūn, who had deposited a fine copy of his Kitāb al-ʿIbār in the library of the Qarawiyyin mosque in Fez to serve as an exemplar for those who wished to make a copy.51 This aspect was, therefore, not ignored by the authors, who saw it as part of a publication process. The fact remains that libraries and books were still exposed to disappearances and displacements. The very real human experience of the fragility of books – mainly manuscripts – no doubt explains why, despite the precautions taken and in particular the creation of libraries to house them, a certain pessimism may have taken hold. In this volume, the juxtaposition of case studies that are distant from each other in time and space brings out contrasts but also suggestive similarities. Thus, even before the word became popular, encyclopaedic ambitions were present in the design of some libraries. Several of them, not by chance, were the result of a prince’s project: they were mirrors of the world through the accumulation of knowledge or the selection of works representative of universal knowledge; in all cases, the latter belonged to the prince who made it a symbolic manifestation of his power. In contrast to these prestigious libraries (or libraries of prestige?), modest collections played a decisive role in the transmission of texts and the spread of knowledge. Limited by the means available to their owners, they reflect the interests of an era or a category of readers through the choices they made. For their part, the answers given to the question of access to these different collections are very mixed, but it emerges that consulting their books was not an easy task. The approach that we had chosen has, therefore, borne fruit and the articles gathered here seem to us to show its fruitfulness to the reader. 51 Ibn Khaldūn, Kitāb al-ʿIbār, manuscript Fez, Qarawiyyin Library, 1266; Lévi-Provençal 1923,

Towards a Comparative Study of Libraries in the Manuscript Age Acknowledgements This volume is part of the results of the ERC project 670628 SICLE (Saadian Intellectual and Cultural Life, 2016–2021) under the European Union Funding for Research and Innovation (Horizon 2020). We would like to thank Kristen de Joseph and Philip Saunders for turning the text into correct English; Luca Farina for his careful treatment of tables, images and texts; and last, but not least, Caroline Macé for her attentive readings of the texts and boundless patience throughout the long editing process of this book. References Bagnall, Roger S. (2002), ‘Alexandria: Library of Dreams’, Proceedings of the American Philosophical Society, 146/4: 348–362. Baratin, Marc and Christian Jacob (eds) (1996), Le pouvoir des bibliothèques. La mémoire des livres en Occident, Paris: Albin Michel. Bausi, Alessandro, Christian Brockmann, Michael Friedrich and Sabine Kienitz (eds) (2018), Manuscripts and Archives, Berlin: De Gruyter. Behrens-Abouseif, Doris (2019), The Book in Mamluk Egypt: Scribes, Libraries and Market, Leiden: Brill. Bouza, Fernando (2001), Corre manuscrito. Una historia cultural del Siglo de Oro, Madrid: Marcial Pons. Canfora, Luciano (1990), The Vanished Library: A Wonder of the Ancient World, Berkeley/Los Angeles: University of California Press. Clancier, Philippe (2009), Les bibliothèques en Babylonie dans la deuxième moitié du Ier millénaire av. J.-C., Münster: Ugarit Verlag. Coqueugniot, Gaelle (2013), Archives et bibliothèques dans le monde grec : Édifices et organisation, Oxford: Archaeopress. Cort, John E. (2013), ‘The Jain Knowledge Warehouses: Traditional Libraries in India’, in Orsini, Francesca (ed.), The History of the Book in South Asia, Farnham: Ashgate, 55–65. Delhey, Martin (2015), ‘The Library at the East Indian Buddhist Monastery of Vikramaśīla: An Attempt to Identify its Himalayan Remains’, manuscript cultures, 8: 3–24. Delhey, Martin, Vito Lorusso, et al. (2015), ‘Wordlists for Libraries and Closely Related Phenomena in Different Manuscript Cultures from Asia, Africa and Europe’, DOI: https:// doi.org/10.25592/uhhfdm.9798. Déroche, François, Nuria de Castilla, and Lbachir Tahali (eds) (2022), Les livres du sultan. Matériaux pour une histoire du livre et de la vie intellectuelle du Maroc saadien (XVIe siècle), 2 vols, Paris: Peeters. Déroche, François and Lbachir Tahali (2022), ‘Du bon usage des livres : Aḥmad al-Manṣūr, mécène de la bibliothèque Qarawiyyīn’, Journal des savants, 2: 329–387. Drège, Jean-Pierre (1991), Les bibliothèques en Chine au temps des manuscrits (jusqu’au Xe siècle), Paris: École française d’Extrême-Orient. Dudbridge, Glen (2000), Lost Books of Medieval China, London: The British Library.

Nuria de Castilla, François Déroche, Michael Friedrich Eisen, Markus (2011), ‘Zur architektonischen Typologie von Bibliothen’, in Nerdinger (ed.) 2011, 261–306. Erünsal, İsmail E. (2008), Ottoman Libraries: A Survey of the History, Development and Organization of Ottoman Foundation Libraries, Cambridge, MA: The Department of Near Eastern Languages and Civilizations, Harvard University. Finkel, Irving (2019), ‘Assurbanipal’s Library: An Overview’, in Ryholt and Barjamovic (eds) 2019, 367–389. Fölster, Max Jakob (2016), ‘Traces in Red: Chinese Book Collector’s Seals as a Means to Track the Transmission History of a Manuscript’, in Giovanni Ciotti and Hang Lin (eds), Tracing Manuscripts in Time and Space through Paratexts, Berlin: De Gruyter, 161–183. Fölster, Max Jakob, Janina Karolewski et al. (2015), ‘Questionnaire for the Study of Manuscript Collections (Towards a Typology of Manuscript Collections)’, DOI: https://doi. org/10.25592/uhhfdm.9810. Fournet, Jean-Lucques (2018), ‘Archives and Libraries in Greco-Roman Egypt’, in Bausi et al. (eds) 2018, 171–199. Friedrich, Markus (2018), ‘Epilogue: Archives and Archiving across Cultures–Towards a Matrix of Analysis’, in Bausi et al. (eds) 2018, 421¬–445. Grabowsky, Volker (2019), ‘Thai and Lao Manuscript Cultures Revisited: Insights from Newly Discovered Monastic Collections in Luang Prabang’, Journal of the Siam Society, 107/1: 79–118. Grünbart, Michael (2018), ‘Securing and Preserving Written Documents in Byzantium’, in Bausi et al. (eds) 2018, 319–338. Gutas, Dimitri (1998), Greek Thought, Arabic Culture: The Graeco-Arabic Translation Movement in Baghdad and Early ʿAbbāsid Society (2nd–4th/8th–10th Centuries), Abingdon: Routledge. Haikal, Fayza (2008), ‘Private Collections and Temple Libraries in Ancient Egypt’, in Mostafa El-Abbadi, Omnia Fathallah and Ismail Serageldin (eds), What Happened to the Ancient Library of Alexandria?, Leiden: Brill, 39–54. Heijdra, Martin J. (2004), ‘The Development of Modern Typography in East Asia, 1850–2000’, East Asian Library Journal, 11/2: 100–168. Hirschler, Konrad (2016), Medieval Damascus. Plurality and Diversity in an Arabic Library: The Ashrafiya Library Catalogue, Edinburgh: Edinburgh University Press. Hirschler, Konrad (2020), A Monument to Medieval Syrian Book Culture: The Library of Ibn ʿAbd al-Hādī, Edinburg: Edinburgh University Press. Hunwick, John (1999), Timbuktu and the Songhay Empire: Al-Sa‘dī’s Ta’rīkh al-sūdān Down to 1613 and Other Contemporary Documents, Leiden: Brill. Jacob, Christian (ed.) (2007), Lieux de savoir. Espaces et communautés, Paris: Albin Michel. König, Jason, Katerina Oikonomopoulou and Greg Woolf (eds) (2013), Ancient Libraries, Cambridge: Cambridge University Press. Kornicki, Peter (1998), The Book in Japan: A Cultural History from the Beginnings to the Nineteenth Century, Leiden: Brill. Krätli, Graziano and Ghislaine Lydon (eds) (2011), The Trans-Saharan Book Trade: Manuscript Culture, Arabic Literacy and Intellectual History of Muslim Africa, Leiden: Brill. Last, Murray (2011). ‘The Book and the Nature of Knowledge in Muslim Northern Nigeria (1457–2007)’, in Krätli and Lydon (eds) 2011, 176–211. Lehmann, Edgar (1996), Die Bibliotheksräume der deutschen Klöster in der Zeit des Barock, 2 vols, Berlin: Deutscher Verlag für Kunstwissenschaft.

Towards a Comparative Study of Libraries in the Manuscript Age Lehmann, Yes (ed.) (2018), Savoir/Pouvoir. Les bibliothèques, de l’Antiquité à la modernité, Turnhout: Brepols. Lévi-Provençal, Évariste (1923), ‘Note sur l’exemplaire du Kitāb al-ʿIbār offert par Ibn Haldun à la bibliothèque al-Qarawiyîn à Fès’, Journal Asiatique, 203: 161–168. McDermott, Joseph P. (2006), A Social History of the Chinese Book: Books and Literati Culture in Late Imperial China, Hong Kong: Hong Kong University Press. McDermott, Joseph P. (2011), ‘Book Collecting in Jiangxi during the Song Dynasty’, in Lucille Chia and Hilde De Weerdt (eds), Knowledge and Text Production in an Age of Print: China, 900–1400, Leiden: Brill, 63-101. Nerdinger, Winfried (ed.) (2011), Die Weisheit baut sich ein Haus: Architektur und Geschichte von Bibliotheken, Munich: Prestel. Nylan, Michael (2018), ‘On Libraries and Manuscript Culture in Western Han Chang’an and Alexandria’, in Geoffrey E. R. Lloyd and Jingyi Jenny Zhao (eds), Ancient Greece and China Compared, Cambridge: Cambridge University Press, 373–409. Reese, Scott (ed.) (2022), Manuscript and Print in the Islamic Tradition, Berlin: De Gruyter. Ren Jiyu 任继愈 (ed.) (2001), Zhongguo cangshulou 中国藏书楼, 3 vols, Shenyang: Liaoning renmin chubanshe. Rico, Christophe and Anca Dan (eds) (2017), The Library of Alexandria: A Cultural Crossroads of the Ancient World, Jerusalem: Polis Institute Press. Rosenthal, Franz (1947), The Technique and Approach of Muslim Scholarship, Rome: Pontificium Institutum Biblicum. Ryholt, Kim and Gojko Barjamovic (eds) (2019), Libraries Before Alexandria: Ancient Near Eastern Traditions, Oxford: Oxford University Press. Speer, Andreas (2020), ‘Die Bibliothek – Denkräume und Wissensordnungen’, in Andreas Speer and Lars Reuke (eds) (2020), Die Bibliothek – The Library – La Bibliothèque, Berlin: De Gruyter, XV–XXV. Stackmann, Ulrich (1990), Die Geschichte der chinesischen Bibliothek Tian Yi Ge vom 16. Jahrhundert bis in die Gegenwart, Stuttgart: Franz Steiner Verlag. Staikos, Konstantino Sp. (2004–2013), The History of the Library in Western Civilization, 6 vols, New Castle, Del.: Oak Knoll Press and Houten: HES & DE GRAAF Publishers. Staikos, Konstantino Sp. (2017), The Architecture of Libraries in Western Civilization from the Minoan Era to Michelangelo, New Castle, Del.: Oak Knoll Press and Athens: ATON Publishing House. Too, Yun Lee (2010), The Idea of the Library in the Ancient World, Oxford: Oxford University Press. Wang Xiang (2014), ‘From Nālandā a to Chang’an: A Survey of Buddhist Libraries in Medieval China (618-907)’, in Tansen Sen (ed.), Buddhism Across Asia: Networks of Material, Intellectual and Cultural Exchange, vol. 1, Singapore: New Delhi ISEAS, 207–231. Winkelmann, John H. (1974), ‘The Imperial Library in Southern Sung China, 1127–1279. A Study of the Organization and Operation of the Scholarly Agencies of the Central Government’, Transactions of the American Philosophical Society, 64/8: 1–61. Wischermann, Heinfried (2011), ‘“Claustrium sine armario quasi castrum sine armamentario”. Bemerkungen zur Geschichte der Klosterbibliothek und ihrer Erforschung’, in Nerdinger (ed.) 2011, 93–130. Woolf, Greg (2013), ‘Introduction: Approaching the Ancient Library’, in König, Oikonomopoulou and Woolf (eds) 2013, 1–20. Wujastyk, Dominik (2007), ‘La bibliothèque de Thanjavur’, in Jacob (ed.) 2007, 616–636.

The Islamic World

John Seyller The Operations of the Imperial Mughal Library Abstract: To complement the extensive literary information about the imperial Mughal library documentary evidence can be gleaned from surviving manuscripts. Dozens of notes written by emperors and princes, hundreds of inspection notes and valuations recorded by their librarians, and most recently thousands of seal impressions by Mughal officials make for a fuller picture of the actual workings of the Mughal library. This includes a practically uninterrupted chain of chief librarians and the roster of librarians active during the reign of emperors from Akbar to ʿAlamgir. The imperial Mughal library was remarkable both as a physical repository of accumulated knowledge and as an institution of meticulous record-keeping. Mughal chronicles have provided considerable information about the library’s scope and composition. An inventory completed at the time of Emperor Akbar’s death in 1605 quantified its holdings as 24,000 volumes – far larger than almost any contemporary library in the world – with a cumulative valuation of 6,500,000 rupees.1 The library’s vast size naturally led Abu’l Fazl, Akbar’s court chronicler, to boast of its comprehensiveness, and by extension, the emperor’s encyclopedic acquaintance with all human knowledge. Abu’l Fazl describes the library as divided into parts, some within the harem, and others outside its quarters (Birun), with each section further subdivided according to value, language, and prestige of the discipline it encompasses, such as poetical and prose literary texts, sciences, law, and Sufism.2 Importantly, Abu’l Fazl remarks that the inspection of books followed the same order. Because the emperor was acknowledged to be illiterate – a quality extolled as evidence that his superior understanding was not acquired from men but was an innate gift from God – the chronicler also pointedly explains that Akbar listened repeatedly to books read aloud to him, and that he made marks with his own pen where the daily reading ceased.3 Abu’l Fazl then provides an extensive list of classic texts of Persian literature, the most esteemed category of 1 Smith 1917, 424 states that these numbers were copied from Mughal records by Sebastian Manrique; the amount is equivalent in 2017 values to £23,890,674 or € 21,023,793. 2 Abu’l Fazl 1977, vol. 1, 109–110. 3 Abu’l Fazl 1977, vol. 1, 110. Although I have looked for such marks for more than three decades, I have never found anything that meets this description. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-002

John Seyller knowledge, and lauds the emperor’s commissioning of a new, more immediately intelligible version of the Kalila wa Dimna, a didactic book of animal fables, entitled ʿIyar-i danish.4 He then proceeds to mention historical texts, including the newly composed Tarikh-i alfi (History of the Millenium), and numerous translations of Hindu literature written in Sanskrit, including the Haribans (Harivamsa), or genealogy of Krishna.5 He singles out for further discussion the Hindu epics of the Mahabharata and the Ramayana, translated in the 1580s. Elsewhere, we are informed that the highest-quality books were reserved for royalty and were stored separately.6 In short, this official account of the imperial Mughal library, which was written in the mid-1590s, describes the structural foundation of an institution that was to grow larger and vastly more bureaucratic over the course of the seventeenth century. The Mughals wrote extensively on their manuscripts, rivalled in this activity perhaps only by the Chinese. The Mughal practice of marking ownership with a combination of inspection notes and seal impressions on their flyleaves started slowly and modestly, with extremely rare examples of impressions of the seals of the first and second emperors of the dynasty, Babur (1483–1530, r. 1526–1530) and Humayun (r. 1530–1540 and 1555–1556). Both their seals appear without accompanying contemporary notes at the top of the flyleaf of a famous illustrated manuscript of the Shahnama made for Muhammad Juki, a descendant of the Mughals’ ancestor Timur, in Herat about 1444 (Fig. 1).7 Around the seals are terse later inscriptions that read khassa (royal) and awwal duwum (first class, second grade) along with a note of verification – sah (correct) – below that. Below them is a large rectangular field painted over with gold, evidently intended to obscure whatever notations happened to occupy that space and to furnish a pristine ground for impressions of the seals of Emperors Jahangir (r. 1605–1627, the circular seal in top centre), Shahjahan (r. 1627–1658), and ʿAlamgir (r. 1658–1707, the distinctive teardrop-shaped seals to the right and below). The round seal of Akbar (r. 1556–1605), also very rare, is conspicuously absent. The large inscription, written by Shahjahan himself, offers a formulaic notice of the manuscript’s official entry into his library on the date of his auspicious accession (fifth of the month of Bahman Ilahi, corresponding to the eighth of the month of Jumada II 1037 ah/14 February 1628 ce), along with his own patrimony back to his grandfather, Akbar, but no 4 Abu’l Fazl 1977, vol. 1, 110–112. 5 Abu’l Fazl 1977, vol. 1, 110–113. 6 Abu’l Fazl 1977, vol. 1, 110. 7 See Brend 2010, frontispiece and 191–192. The legends and designs of the seals of Babur and Humayun are discussed and transcribed in Brend 2010, 149 and 192.

The Operations of the Imperial Mughal Library Fig. 3: Detail of Fig. 2. Fig. 4: Detail of Fig. 2. further. By substance and position at the head of the manuscript, then, this is an unusually complete proclamation of Mughal ancestral ownership from as early as about 1506, asserted by a series of successive emperors.8 Surprisingly, however, the chain of Mughal ownership was interrupted from the mid sixteenth century until 1602, as attested by two of the many inspection notes on the final folio (Fig. 2).9 One note is dated to the beginning of 997 ah/1588 ce, when the book was apparently passed from the estate of one Shihab al-Din Ahmad Khan (Fig. 3), and the second on 14 Urdibihisht regnal year (ry) 47/May 1602 ce, when the valuation from the original 50 mohurs (or 500 rupees) was recorded rising dramatically to 500 mohurs (equal to 5,000 rupees), making this first-class book one of the most expensive in the whole of the Mughal 8 Brend 2010, 148–149, makes the case that Babur acquired the manuscript in Herat in 1506. 9 See the decipherment and explication of fol. 536a by A.H. Morton in Brend 2010, 163–175, including a reproduction of the folio with an annotated key.

John Seyller library at any date (Fig. 4).10 The remaining twenty-three notes are written in haphazard chronological order on the folio and are interspersed with tangentially related seals. They document the sporadic inspection of the manuscript by individual librarians on dates specified in two different calendars (Islamic and Iranian), reiterate the manuscript’s increased valuation, or record the transfer of custody from one librarian to another at irregular intervals up to 1698. Nowhere is basic descriptive information – the name of the text or the calligrapher, the number of folios, and strikingly, the presence of fine paintings and illuminations – even mentioned. This splendid manuscript introduces the topic discussed in this essay: the human dimension of the imperial Mughal library. Scattered references in various Mughal histories establish the titles and basic duties of the principal library positions under four emperors from Akbar to ʿAlamgir, whose combined reigns covered the century and a half from 1556 to 1707. They are the nazim or muʿtamad, the library’s general manager and chief librarian, a nobleman who dealt with finances and personnel, and the darogha, the deputy librarian who oversaw the purchase and cataloguing of material. Beneath him were a variety of assistants who carried out the actual inventory and physical inspection of books. These same sources name some learned men associated with the library from the reigns of Akbar to Shahjahan, and suggest that three of them – Mulla Bilal, Shaykh Fayzi, and Maktub Khan – held the position of nazim under Akbar and Jahangir, respectively. As an art historian, I have always worked primarily with actual documents rather than literary texts. More than twenty-nine years ago I began a systematic investigation into the Mughal library, with the aim of seeking a fuller understanding of the commission, circulation, and reception of illustrated manuscripts. I became engrossed in deciphering the many thousands of notes like these, which are written in a difficult clerical script with challenging orthographic conventions and abbreviations. Some of these inspection notes confirmed that portions of the library were transported about the empire by indicating that the inspection of a particular book had been carried out in such places as Agra, Lahore, Ajmer, and Srinagar.11 Others recorded monetary valuations ranging from a high of 20,000 rupees to a low of one rupee, and qualitative categories, calibrated from first to fifth class, with some further subdivisions within the top two categories.12 But 10 See Morton in Brend 2010, 167, nos 6–7. 11 See Seyller 1997, 252 for evidence that books were inspected at these Mughal capitals or prominent summering locales. 12 See Seyller 1997, 274–275 for a chart of valuations and qualitative categories.

The Operations of the Imperial Mughal Library these inspection notes most commonly yielded the names of many dozens of lowlevel librarians, sometimes designated as tahwildars (custodians), and the regnal years during which they worked. Other, more specific functions within the library were indicated for a few individuals. For example, Khwaja Dawlat is described as the inspector of the library of the royal harem (tahwildar-i kitabkhana-yi mahall-i sharif).13 ʿAmbar held the title of custodian of the external library (tahwildar-i birun), and Muʾmin and Asad Beg were distinguished as superintendents of gifts (sahibjamiʿ peshkash).14 Eighteen years ago, I began serious work on the second phase of my study of the Mughal library, this time directing my attention to the identification of the owners of thousands of seals found on the same books and occasionally on the reverse of independent paintings. Now, with the bulk of that investigation completed and the perspective gained from time spent on other kinds of projects, I wondered how these three sources might shed light on one another. In particular, I have been keen to understand the function of these seals, which appear frequently and often seem to exist in parallel to the many inspection notes I have logged. Fig. 5: London, British Library, Or. 14139, fol. 117b (lower centre); detail of Diwan of Hafiz, made at Herat or Mashhad, c. 1470; seal of Fathullah, son of Abu’l Fath dated 1006 ah/1597–1598 ce. © British Library. 13 Seyller 1997, 248, n. 15. 14 Seyller 1997, 248 and n. 17.

The Operations of the Imperial Mughal Library sonage typically placed in charge of the library. This is what makes his mention in a routine inspection note – even a single one – so much of an aberration to the pattern that developed over time in which head librarians did not get involved in routine operations (Fig. 6).15 Two different contemporaneous seals of his are impressed, one above another, on another manuscript (Fig. 7). The change in the design and date of an individual’s seal is unusual for this period. Two other points about Fayzi are noteworthy. One is that upon his death in October 1595, he left his personal library of 4,600 volumes to the imperial library. This is the earliest and most prolific example I know of an author or librarian presenting his work or collection to the Mughal library. The second is that Fayzi was also later accorded the privilege of being portrayed in the opening illustration of a manuscript of Nal wa Daman, an Indian romance he had translated from Sanskrit, shown as he presents his work to Akbar (Fig. 8). To my knowledge, Abu’l Fazl and Fayzi are the only historians and poets to be portrayed in contemporary Mughal painting. Fig. 9: Houston, TX, The Art and History Trust Collection, no accession number, fol. 79a, Gulistan of Saʿdi, completed by Sultan ʿAli al-Mashhadi at Herat in Muharram 891 ah/January 1486 ce; seal of Maktub Khan dated 1026 ah/1617 ce. © Art and History Trust Collection. The most prominent librarian of Jahangir’s reign was Maktub Khan, whom Jahangir mentions in his memoirs as the author of a famous chronogram of his accession date and as the director of both the library and picture gallery.16 His 15 The note appears in the lower centre of the folio. It reads ‘Transferred to Shaykh Fayzi from Mir Muhammad Taqi and inspected on the date of 27 Zi’l-Hijja 990 [12 January 1583]’. A note just above this indicates that the manuscript had entered the library on 15 September 1582. 16 Thackston 1999, 67, 269.

The Operations of the Imperial Mughal Library Fig. 11: London, British Library, India Office Collection Pers. 383, fol. 1a (detail of the upper portion), Layla wa Majnun, seals of ʿAbd al-Rashid Daylami (right), Iʿtimad Khan (centre), and ʿInayat Khan (left). © British Library. full name seems to be Muhammad Muʾin, for a note recording the transfer of the custody of a book in year one of Jahangir’s reign (1606) mentions that name along with the title ‘the chief of the library’ (sahibjami-yi kitabkhana).17 His seal, dated 12 ry – the inclusion of a regnal year in a seal’s legend was an overt expression of loyalty introduced under Jahangir – and 1026 ah/1617 ce appears on nine manuscripts, including a few of the highest quality and valuation (Fig. 9). The reign of Emperor Shahjahan (1627–1658) saw remarkably frequent turnover in the hierarchy of the imperial library. We can ascertain this by occasional literary references, but more definitively by the sheer number of different Shahjahan-period seals impressed on manuscripts – often, but not always the same manuscripts. Biographical sketches of four librarians will bring out some of the implications of these seals. The first librarian of this select group is ʿAbd al-Rashid Daylami, also known as Rashida, the name under which he is mentioned in The History of Shahjahan as a calligrapher and chief librarian. Predictably, he came from a line of distinguished calligraphers. When his uncle, the famous calligrapher ʿImad al-Hasani of Qazwin, was assassinated in Iran in 1615, Rashida fled to India and joined the service of Prince Shahjahan in 1033 ah/1623–1624 ce, who was embroiled in a bitter feud with his father, Emperor Jahangir. ʿAbd al-Rashid’s own accomplishment and early loyalty to Shahjahan paid off ten years after the prince’s accession, for he was apparently appointed darogha of the imperial library in 17 Diwan of Amir Hasan Dihlawi, Walters Art Museum W.650, fol. 187b (visible only under infra-red light). For a full translation of the notes, see Seyller 1997, 280–281.

The Operations of the Imperial Mughal Library 1048 ah/1638–1639 ce. This new status was established by his seal, which proclaims him as a servant of Shahjahan and is dated ry 12/1048 ah/1638–1639 ce (Figs 10–11). More important, his seal appears on thirty-nine manuscripts as well as on the reverse of three paintings – both an exceptionally high rate of occurrence. Literary sources indicate that ʿAbd al-Rashid was relieved of his duties as darogha in 1056 ah/1646–1647 ce, ostensibly for reasons of health, though he lived for another twenty-five years. ʿAbd al-Rashid Daylami was succeeded as chief librarian by Mir Salih in 1056 ah/1646 ce, the year of his newly designed seal, which appears on twenty manuscripts and the reverse of one painting (Fig. 12).18 The professional background of Mir Salih (his full name was Mir Muhammad Salih Kamboh) was equally illustrious, for he was the son of the renowned Mughal calligrapher Mir ʿAbdullah Mushkin Qalam. Writing under the pen-name Kashifi, Mir Salih became a celebrated poet and chronicler of Shahjahan, completing the Amal-i Salih or Shahjahan-nama in 1659–1660. In short, one could hardly hope for stronger political connections to the emperor. Nonetheless, six years later, that is, at the end of 1062 ah/1651–1652 ce, he was replaced as head librarian by Iʿtimad Khan, the title bestowed upon Muhammad Ashraf, eldest son of Islam Khan, who had been paymaster of the army. Thus far I have documented his seal, dated ry 26/1063 ah/1652–1653 ce, on forty-one manuscripts and five paintings.19 According to the Shahjahan-nama, a contemporary history of the reign, Iʿtimad Khan was supplanted as superintendent of the library in 1068 (December 1657) by ʿInayat Khan, the title given to Muhammad Tahir, son of Zafar Khan. Iʿtimad Khan was not demoted; rather, according to historical sources, he was appointed to the position of diwan (minister) in 1065 ah/1654–1655 ce. His replacement, ʿInayat Khan, whose full name is Mirza Muhammad Tahir Ashena, was also involved in writing the Shahjahan-nama. His tenure as chief librarian lasted less than a year because Shahjahan fell seriously ill in 1658, and was effectively deposed by his son Awrangzeb (later ʿAlamgir) after the fierce War of Succession. During his short tenure, ʿInayat Khan impressed his seal on fifty manuscripts – the most recorded by any librarian to this point – and five paintings.20 It is no surprise that a man so intimately connected with Shahjahan would be removed immediately as head librarian under the new and antagonistic regime of ʿAlamgir. He died in exile in Kashmir twelve years later. 18 The seal of Mir Salih, servant of Shahjahan, appears at the bottom centre of Fig. 14. 19 For the seal of Iʿtimad Khan, see Figs 10–11. 20 For the seal of ʿInayat Khan, see Figs 12–13.

The Operations of the Imperial Mughal Library These short biographical sketches draw attention to the personal and political aspects of those at the highest echelon of the library administration. But as I began to work up similar biographies for the individuals named in my copious files on the inspection notes and seals on Mughal manuscripts, this question crossed my mind. If these are the seals of the head librarians, what should we make of those of the other nobles that appear with similar frequency on these same manuscripts, especially during the same years? There are, for example, many seals of Ahmad Shahid dated 1054 ah/1644–1645 ce (thirty-eight manuscripts, three paintings), Ya Karim dated 1058 ah/1648–1649 ce (twenty-one examples), Muhammad Husayn dated 1061 ah/1650–1651 ce (twelve manuscripts), and Muhammad ʿAli Shahjahani dated 1062 ah/1651–1652 ce (eleven manuscripts). Who were these men? Their identities are difficult to determine, for some of the names are either very commonplace, abbreviated versions of proper historical names, or merely allusive. Ya Karim, for example, is a simple invocation to the Merciful God, but it was almost certainly used by a man whose name included the word Karim. And if they were not the chief librarian, what function did they have in the library that obliged them to assert by proxy continued Mughal ownership of these manuscripts? The answer to the latter question will require us to develop a still more nuanced picture of the imperial library, with different high-level librarians placed in charge of books with particular subject matters and qualitative levels. I also suspect that we will discover that even so bureaucratic a system as has been documented to date had its share of intrigues and messy personnel situations, much as the administrations of our universities and museums suffer from time to time. A final example demonstrates both the usefulness and limitations of the evidence of seals as well as the movement of books from Bukhara in Central Asia to Mughal India. Most of these books seem to have arrived in India around 1600, apparently as a result of unsettled political conditions in Bukhara. A copy of the Diwan of Shahi is one such Mughal manuscript of Bukharan origin (Fig. 13). Written at Bukhara by an eminent calligrapher, Mir ʿImad al-Husayni, it was illuminated and illustrated with a double-page frontispiece in the middle of the sixteenth century. The manuscript has an unusually comprehensive record of seals and notes. Some notes are distinguished, such as an accession-year inscription by Jahangir; others are less so, such as this curious one that reads ‘Nawab Khan purchased a black horse through the kindness of the Buluch living in the village of Nukhud from Muhammad ʿAli Khan on the date of 23 Ramadan 1177 (26 March 1764)’. What interests us here is the implication of the earliest seal, the seal of Akbar’s mother, Hamida Banu Begum, which is dated 968 ah/1560–1561 ce: ‘When one’s seal bears the sign of love (Hamida Banu Begum) her stamp shall become a reflection of good fortune’. The square seal is one of two types used by Hamida Banu Begum on a total

John Seyller of twelve known manuscripts (Fig. 14). The other seal is quite different in shape (twelve-lobed) and ten years earlier in date (957 ah/1550–1551 ce) (Fig. 15). Fig. 14: New Delhi, National Museum of India, 48.6/1, fol. 1a, Anthology, seal of Hamida Banu Begum 968 ah/1560–1561 ce. © photo by the author. Fig. 15: Doha, Museum of Islamic Art, Rashahat ʿayn al-hayat, fol. 1a, seal of Hamida Banu Begum, dated 957 ah/1550–1551 ce. © Museum of Islamic Art, Doha. If we follow the normal logic that dated seals are virtually contemporaneous with their use, the presence of a seal dated 968 ah/1560–1561 ce should indicate that the manuscript entered Hamida Banu Begum’s possession about that time, which would mean that this Diwan of Shahi reached Mughal India just after it was made, or almost forty years before most of the other Bukharan manuscripts. This is possible but anomalous, and invites further research on the extent of friendly contact between these two courts in the mid sixteenth century. There is, of course, another explanation: that her seal was well out of date, perhaps even some thirty-five years or so. This is not unimaginable. Her earlier seal (957 ah/1550–1551 ce), for example, is seen on a manuscript of the Diwan of Haydar Kaluj written in 967 ah/1559–1560 ce, and thus was already ten years out of date when the manuscript was produced (Fig. 16).21 Keeping in mind, then, the notion that dates on seals are not absolutely reliable historical markers, we should consider the likelihood that royal seals changed relatively infrequently, and are therefore the least reliable indicators of the dates 21 For the two annotated flyleaves of a Mughal Ramayana manuscript that once belonged to Hamida Banu Begum and is impressed with the seals of many prominent Mughal librarians, see Seyller, Sardar and Truschke 2020, 174–177. A comprehensive inventory of the queen mother’s library appears in Seyller, Sardar and Truschke 2020, 178–179 and 189.

John Seyller of actual transactions. Some non-royal individuals did obtain new seals frequently – sometimes as often as four times within a decade – but most did so only when they received a new title or wished to express their loyalty to a new master. In the best-documented situation, that is, the seventeenth-century Mughal court, seals are used most effectively as documentary evidence in conjunction with literary sources and inspection notes, the three amplifying each other so well that we can reconstruct the custody and use of a number of manuscripts practically every five years for the better part of a century. Considered together, these sources enable us to recreate quite fully the operations of the imperial Mughal library, an institution whose exceptional size and sophistication befit one of the world’s great empires. References Abu’l Fazl (1977), Aʾin-i Akbari, 3 vols, New Delhi: Oriental Books Reprint Corporation. Brend, Barbara (2010), Muhammad Juki’s Shahnamah of Firdausi, London: The Royal Asiatic Society of Great Britain and Ireland and Philip Wilson Publishers. Seyller, John (1997), ‘The Inspection and Valuation of Manuscripts in the Imperial Mughal Library’, Artibus Asiae, 57: 243–349. Seyller, John, Marika Sardar, and Audrey Truschke (2020), The Ramayana of Hamida Banu Begum, Queen Mother of Mughal India, Doha: Museum of Islamic Art. Smith, Vincent (1917), Akbar the Great Mogul, Oxford: Clarendon Press. Thackston, Wheeler (1999) (tr. and ed.), The Jahangirnama. Memoirs of Jahangir, Emperor of India, Washington, DC: Freer Gallery of Art.

The Operations of the Imperial Mughal Library Appendix: Librarians of Akbar’s Reign (1556–1605) Librarians named in literary sources ʿAbd al-Ghaffur Shaykh Abu’l Fayz, a.k.a. Fayzi Mulla Bilal Fathullah, son of Abu’l-Fath Ghiyath Allah of Shiraz Hakim ʿAli ʿInayatullah Muhammad Taqi Mulla Pir Muhammad Khan Librarians named in extant manuscripts ʿAbd al-Rahim ry 45 Mulla ʿAli ry 39, 40, 41 ʿAmbar ry 39, 40, 41 Bahadur Chela ry 42 at Srinagar Shaykh Fayzi 27 Zi’l-Hijja 990 ah [22 January 1583 ce] Hajji Fazlullah 988 ah [1581 ce], ry 46 [1601 ce] Hakim ʿAli ry 47, 49 Habibullah ry 47 ʿInayatullah ry 21, 31, 33 [998 ah/1590 ce], 39 [1002 ah/1594ce], 41, 42, 43 [1598 ce] Luqman ry 40, 47 Muhammad ʿAli ry 39 [1002 ah/1594 ce], 49 Muhammad Baqir ry 49 Muhammad Husayn 1002 ah [1594 ce] Muhammad Mansur 997 ah [1589 ce] Muhammad Nur ry 40 Muhammad Rashid ry 41 Muhammad Sadiq ry 39, 40, 50 Muhammad Taqi 27 Zi’l-Hijja 990 ah [22 January 1583 ce] Muhammad Yusuf ry 47 Salih ry 32 [996 ah/1588 ce], 33 [998 ah/1590 ce] Librarians named in seals Abu’l Fayz (a.k.a. Fayzi) Fathullah

Nuria de Castilla Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection at El Escorial Abstract: Sultan Mūlāy Zaydān’s (r. 1603–1627) library was captured at sea by the Spaniards in 1612 and his books, except for those which burned in 1671, have remained mostly untouched until now. They reflect an acquisition process which started with his father Aḥmad al-Manṣūr (r. 1578–1603) and carried on by his sons, Abū Fāris and above all Mūlāy Zaydān. Poetry ranks roughly third in the library holdings as far as the number of volumes is concerned. Most had been produced in the Middle East rather than in Morocco and mirror the tastes of the princes who read them. The comparatively high amount of religious poems reflects directly a local evolution of the period. Classical Abbasid poetry is also present, as well as a more limited number of compositions originating in al-Andalus. In this section, texts on prosody are comparatively numerous: the reason may be found in the courtly practice of poetry and the need of a technical training in this field. Contemporary authors actually transmit poems composed by Aḥmad al-Manṣūr. 1 Mūlāy Zaydān’s library The Saadian sultans’ library kept in the Royal Library of San Lorenzo de El Escorial (RBME) following its capture in 1612 under the reign of the Moroccan sultan Mūlāy Zaydān (1603–1627) is an exceptionally preserved ‘time capsule’. In contradistinction with other medieval or early modern manuscript libraries in the Islamic world, its contents were not altered to the point of making a reconstruction of the original holdings completely impossible. Although no list or catalogue of the collection in the late sixteenth or early seventeenth century has been preserved, we can still examine the books, retrace their history and sometimes identify their readers. They mirror the interests and the tastes of the Saadian rulers and of the elite at large. This situation is almost unique in the Islamic world, with the exception of the Ottoman Empire. We do not possess the complete Saadian library as it may have stood in 1612, before Mūlāy Zaydān had to relinquish Marrakech as a revolt led by Ibn Abī Mahallī was reaching the city. The sultan took with him most of his library to the coastal city of Safi, but it seems that some volumes were left behind. The latter are now scattered among various libraries. It comes as no surprise to find some of Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-003

Nuria de Castilla them in Morocco, for instance in the National Library of the Kingdom of Morocco.1 Some surfaced in other countries: a copy of the Kashshāf by al-Zamakhsharī with Aḥmad al-Manṣūr’s ownership statement penned on its first folio is part of the Cambridge University Library collections.2 Most of the books were however loaded onto a French boat, the Notre-Dame-de-la-Garde.3 The trip was to prove fateful for Mūlāy Zaydān’s library: as the sultan was unable to pay for the transportation, the captain decided to sail towards France in order to sell the cargo. On her way, the Notre-Dame-de-la-Garde was captured by a Spanish fleet and the books were taken to Spain, then to the Spanish Royal library in the monastery of San Lorenzo de El Escorial. When the Saadian library reached the Escorial, it was incorporated into the small Arabic manuscripts collection that began to be assembled under Philip II’s rule.4 They are still kept there, although many were lost during a fire in 1671.5 While in Marrakech, the library was – at least at some moment – under the custody of a librarian, Muḥammad b. al-Ḥājj al-Andalusī, who left a note on one of the manuscripts.6 However, there is no sign that he, or another librarian, ever prepared a catalogue or an index of the Saadian library. In the Escorial, the identification of the topics of each manuscript was started as early as 1617 by David Colville, well before the 1671 fire, but no inventory of the Arabic manuscripts in the Escorial during the seventeenth century seems to have survived.7 From an early date, the Arabic manuscripts kept in the San Lorenzo de El Escorial monastery played a role in the study of Arabic literature in Europe,8 but this exceptional collection did not elicit so far any study of its contents as such or of their relationship with Saadian Morocco. In the frame of the SICLE project, a research was started on the intellectual culture of that period, trying to understand what the people were reading and how it was related to their culture. 1 Hajji 1976, 224 indicates that ‘thousands of books [from the sultan’s library] are scattered in Morocco’s libraries until today’; he gives as example the copies of al-ʿAwdu Aḥmad in the Ḥasaniyya library in Rabat, no. 4911 (174, n. 75) and ʿAbd al-ʿAzīz al-Fishtalī’s Tartīb Dīwān al-Mutanabbī in the National Library of the Kingdom of Morocco 609 G (175, n. 82). 2 Cambridge, UL, Add. 815 (Browne 1900, 178). The owner’s name, Aḥmad al-Manṣūr, has not been identified. 3 Déroche 2016, 43–48. 4 On the history of the collection, see Justel Calabozo 1987, 133–169. 5 Justel Calabozo 1987, 188–190. 6 RBME 1730, fol. 1a. References to manuscripts in the El Escorial collection are directly to the shelfmark; for manuscripts in other collections, the location and name of the library are indicated. 7 Antolin 1921, 80–81; Cano 1996-2004, I, 29. 8 Casiri’s catalogue was an important source of information in this regard.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection Muhammad Hajji’s important research on the intellectual culture of Saadian Morocco is a key source on this subject and especially on the intellectual production of that time.9 The texts circulating were partially known to him through various fahrasa, the record of a scholar’s training, listing the texts he had learned from his various masters – which does not always imply books physically present in Morocco.10 Libraries were obviously also a source of knowledge that was put to use by scholars and the cultivated elite as well, but Hajji did not fully include them in his study. In fact, many libraries of Saadian times have disappeared or are only known through a few items that have come down to us. Luckily, the sultans’ library has been in good part preserved and is a primary witness of the culture of this period. A better understanding of the collection through its manuscripts provides fresh views about the history of the library itself. It developed under Aḥmad al-Manṣūr (r. 1578–1603) who rose to power after the ‘Battle of the three kings’ in 1578 and strove to establish his figure as that of the ‘caliph of the scholars and scholar of the caliphs’ according to a formula coined by a contemporary scholar, Aḥmad al-Manjūr (d. 1587 ce).11 The library survived the internal strife that broke out between Aḥmad al-Manṣūr’s sons after their father’s death. Abū Fāris who became sultan in Marrakech at that moment also collected books that were taken by Zaydān when he succeeded in conquering Marrakech. As could be expected, many volumes bear a handwritten mark stating that they belonged to Zaydān, but it comes as no surprise that many also bear his father’s ownership statements. The latter are almost as numerous as those mentioning Abū Fāris. Aḥmad al-Manṣūr was keen on establishing his figure as a scholar and the sources concerning his reign, notably the chronicle by al-Fishtalī, tell much about his efforts in that direction.12 He took pains to abide by the procedure of traditional Muslim scholarship and some of the manuscripts in his collection have actually preserved the licences (ijāza) he received from contemporary scholars. A munāwala in which Aḥmad al-Manjūr gives his permission to the sultan to transmit a treatise he had written has already been published.13 His intellectual coquetteries set apart, Aḥmad al-Manṣūr was very conscious of his position and his library reflects also his political vision. As a member of 9 Hajji 1976. 10 See for instance al-Manjūr 1976. 11 Smith 2006; García-Arenal 2009; Mouline 2009. 12 al-Fishtālī, Manāhil al-ṣafā fī maʾāthir mawālīnā al-shurafā, ed. Kurayyīm 1972, 266–294; Mouline 2009, 82–90. 13 Déroche 2012.

Nuria de Castilla a dynasty claiming descent from the Prophet Muḥammad and bearing for this reason the title of caliph and amīr al-muʾminīn, he established himself as an equal of those who, before him, claimed that title. One of the handwritten marks of Aḥmad al-Manṣūr can be compared to another one that has been attributed to an Almohad prince.14 Aḥmad al-Manṣūr was emulating his predecessors, but also contemporary rulers. Although his relationship with the Ottoman sultans was somewhat strained and he was vying with them, he was impressed by Ottoman culture and took over Ottoman manners, for instance in the organization of his army.15 His books also show that he imitated the habit, well established in the Eastern part of the Muslim world but quite unusual in the West, to have his seal stamped on one of the first folios of a manuscript. The inscription engraved on the seal takes over his (still undeciphered) official signature, but the shape and layout are probably an imitation of an Ottoman model, probably Sultan Beyazıt’s own seal (Figs 1a–1b). His son, Mūlāy Zaydān, followed suit: in addition to the handwritten marks already mentioned, a seal impression with his name in the central circle is found on a few manuscripts in the Escorial. However, its style does not conform to Eastern canons. Fig. 1a: Seal of Aḥmad al-Manṣūr. © San Lorenzo de El Escorial, RBME 48, fol. 3a. Fig. 1b: Seal of Sultan Beyazıt’s. © Private collection. 14 See Paris, BnF, Arabe 6090, fol. 2a, with the name of ʿIbrāhīm b. amīr al-muʾminīn b. amīr al-muʾminīn b. amīr al-muʾminīn b. amīr al-muʾminīn’. It has been suggested that he could be an Almohad prince, son of al-Manṣūr. 15 Mouline 2009, 151, about the army; Hajji 1976, 63 provides a sample of Turkish loan words in Moroccan Arabic.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection The sixteenth century witnesses a technical and aesthetic change in the decoration of the bindings that is contemporary with the Saadian rule. Instead of tooling the covers with small tools requiring to be stamped many times in order to produce an ornament, Moroccan binders began to use engraved plates that allowed to print directly a complete composition. The technique as well as the style of the decoration are clearly borrowed from Ottoman binders, a new element in our appreciation of the taste for Ottoman fashions in Saadian Morocco.16 As stated before, the Saadian manuscripts have been incorporated into the Spanish royal library in the Escorial monastery. In its present state, the collection has been arranged according to subject matter. Of course, the classification is not a clear cut one and texts that belong to another field are often found within another section, especially in the case of multiple-text manuscripts. Aurora Cano’s index provides some help when one wants to get a broader view of the collection contents,17 but it relies heavily on previous catalogues without always correcting their mistakes, keeping for instance the reading Aḥmad al-Munjūz instead of al-Manjūr,18 and failing to record texts that have not been described previously, like a qaṣīda found in multiple-text manuscript RBME 788.19 Anyhow, the classification by subject matters makes things easier when one wants to assess the contents of the Moroccan sultans’ library on a thematic basis in an historical-quantitative perspective. As an example of the intended approach of the collection, I shall focus here on the section of poetry manuscripts. 2 Poetry manuscripts In the Escorial library, the poetry manuscripts are gathered between numbers RBME 270 to 490, then arranged according to their size. As a first step, it is however necessary to disentangle the various components of the Arabic manuscripts collection that are nowadays present side by side on the shelves. Two copies, RBME 290 and RBME 303 were clearly not part of the Saadian library20 16 See de Castilla 2019. 17 Cano Ledesma 1996–2004. 18 The author’s name had been read incorrectly as ‘al-Moundjoûz’ by Dérenbourg (Dérenbourg 1884–1903, I, 163) and his mistake was not corrected by Cano Ledesma (Cano Ledesma 1996– 2004, II, 54, no. 75 and 57, no. 3). See also Déroche 2012, 1219. 19 See Déroche 2019, 169. The text on fol. 175 b is not described by Dérenbourg (Dérenbourg 1884–1903, II, 78). 20 Both were part of Diego Hurtado de Mendoza’s library purchased by Philip II in 1572, see Cano Ledesma 1996–2004, I, 19.

Nuria de Castilla and manuscripts RBME 400, 401, 485 and 490 are either Persian or Ottoman Turkish poetry – or both. They will not be taken into account in the following remarks. Six entries in the catalogue are indicated as missing.21 The 208 remaining manuscripts were part of the Saadian collection and constitute the basis of the present study. Among them are 52 multiple-text manuscripts (MTM), that is to say a quarter of the total. In many instances (22), these manuscripts include only two texts;22 they are conversely fifteen in RBME 470 which is quite exceptional within the group under study. As a result, the number of texts involved is larger than that of the volumes.23 The total of poetical texts present in the collection (257) is however smaller than the total of codicological units since some items are found in different copies.24 Forty-eight texts are found in two, three or even four different copies, which asks the question of the control by the librarians over the purchase of duplicate items. The accession process is actually largely escaping us as we still do not have a precise idea of the way in which the books were acquired. Some were certainly bought locally. The description of the book market in Morocco written by Nicolas Clénard, a Belgian scholar in search of manuscripts who visited Fez towards the middle of the sixteenth century, sounds somewhat gloomy: according to him, few were available and often only parts or fragments of a text, ‘so that one is often forced to buy various times the same thing if one wants to get the complete text and this can take years’.25 More possibilities were obviously available to the sultans than to a Westerner. The former were actually able to secure one way or another books that were part of local scholars’ libraries,26 a procedure that is also known in other parts of the Islamic world.27 We may surmise that most if not all the copies in a Maghribi hand were produced in the Western part of the Islamic world, al-Andalus and North Africa: this is the case for 52 of them. Although one cannot fully exclude that they 21 RBME 319, 402, 403, 447, 478 and 486. 22 In the waqf deed published in this volume by Déroche and Tahali, the miscellanies with two texts are not seen as the other multiple text manuscripts. 23 Miscellanies and composite will be considered as a whole. 24 For instance Arabe 2883, 3305 and 3961 with Ibn al-Saqqāṭ’s Muqaddima fī al-ʿarūḍ (a text on prosody, see below). It seems to happen mostly with multiple-texts manuscripts. 25 Le Tourneau 1934, 58. 26 Volumes part of a family waqf established by Aḥmad al-Manjūr are found in the collection (see RBME 1302 for instance). 27 See for instance Uluç 1999, 88–89.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection may have been transcribed outside of this area, there is a very high possibility that they were actually produced there and never left North Africa.28 However, most of the manuscripts that were part of the Saadian collection of poetry were produced in the Middle East: almost three quarters (153) are written in a script typical of the Eastern part of the Islamic world, a point that is sometimes confirmed by dated notes of Eastern owners or readers. If we only take into account the dated manuscripts (101), the distribution remains identical: 27 (26,5%) are in a Maghribi hand and 74 (73,5%) were produced in the Eastern part of the Islamic world. One cannot absolutely exclude that copyists originating from the Middle East worked in the Western Islamic world, but one may assume that this has been a marginal feature.29 Conversely, a copy in a Maghribi hand, RBME 357, was at some moment in the East as indicated by notes by various Eastern owners, the latest being dated 1564–1565. The imbalance between ‘imported’ and locally produced manuscripts is less noticeable when it comes to the multiple-text manuscripts: 23 of them are in a Maghribi hand against 29 in an Eastern script. The latter is not the only feature that could strike as the hallmark of a foreign provenance: two volumes in safina format made their way to that part of the Saadian library.30 The study of the history of the collection can throw some light on the accession process: in a very few cases only, it will be possible to reconstruct the way in which manuscripts were acquired in the East.31 The data available show that the importation process to Morocco could be quite fast. In the section under study, the most recent colophon with an Eastern provenance is dated 1015 ah/1606 ce, that is to say only six years before the library was captured.32 A general overview of the dated manuscripts indicates that many of the volumes that have been brought to Morocco from the Eastern Islamic lands were purchased under the reigns of Aḥmad al-Mansūr and of his son. When looking at the table below,33 the chronological distribution is quite meaningful: 28 See below an exception with RBME 357. 29 Hajji 1976, 82–84 states that some scholars from the Near East used to visit Morocco during the period, but he does not provide names. 30 RBME 481 and 488. 31 ʿAbdallāh Gannūn ed., 1954, 78–81 and 179–183; I owe this reference to Dr. M. Zouihal who has been able to identify an important source in Egypt. 32 RBME 446, dated 1015 ah/1606 ce. 33 Three multiple-texts manuscripts (RBME 328, 410 and 419) include texts of various periods; in the table, only the latest item has been taken into account. The copies produced between 1601 and the capture of the library are included into the sixteenth century.

Nuria de Castilla Table 1: Distribution of dated poetry manuscripts in the Saadian collection. 12th century 13th century 14th century 15th century 16th century Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Unitary 6 2 5 3 5 2 17 2 23 4 Coll. 1 0 3 2 1 6 2 0 11 8 Subtotal 7 2 8 5 6 8 19 2 34 12 Total 9 (9%) 13 (12.5%) 14 (13.5%) 21 (20%) 46 (45%) As could be expected, the sixteenth century manuscripts are more numerous and account for almost half of the total. A closer look at the corpus of dated and datable poetry manuscripts shows that a conspicuously high proportion among those produced in the East were brought to North Africa after 1550. This conclusion can be reached by combining two groups of manuscripts in an Eastern hand: the first one includes all those bearing a colophon (74 items) and the second one includes those containing a dated note in an Eastern hand later than 1550 (18 items),34 which strongly supports the hypothesis that they had not yet reached Morocco at the moment when the note was added. When taking together the copies that were completed after 1550 according to their colophon and those of the second group, it appears that 47 manuscripts representing more than half of the corpus (54% of a total of 87 manuscripts) reached Morocco after 1550. The implication is that the constitution of the Saadian library relied on an important movement from East to West over a comparatively short period of time. The proportion may actually be even higher as many manuscripts that do not bear any indication that they left the Eastern Islamic world after 1550, like a colophon or a dated note in an Eastern hand, may have also been included in this movement. Full conclusions about book trade and the history of the library will have to wait for the end of the study of the collection when it will be possible to compare the results about its various components. However, I would like to draw the attention to the high proportion of multiple-text manuscripts among the sixteenth century copies. Verses are ubiquitous in the collection since in addition to the texts we subsume under the heading ‘poetry’ we come across many didactic poems (it will be suffi- 34 Twelve manuscripts bear a note in an Eastern hand with a date later than 1550 and are therefore presumed to have reached Morocco after that date; six manuscripts belonged to one of the al-Mallāḥ brothers who sold many items to Aḥmad al-Manṣūr. Among those 18 manuscripts, some do not have a colophon; conversely, those with a colophon prior to 1550 were not included in the first group but in the second – hence the total of 87 manuscripts in our corpus.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection cient to mention here the Alfiyya by Ibn Mālik, a poem on grammar)35 that I shall not take into account here. Apart from these texts, almost half of the 257 texts found in the poetry section of the Escorial collection are either diwāns (i.e. collected works of an author) and single poems, or anthologies of various extant. The rest is made up of commentaries on specific poems or treatises related with prosody. When looking into more detail at the poems present in the collection, some features draw the attention. Religious poetry is well represented, with 61 texts, more than half of them being poems in praise of the Prophet Muḥammad or their commentaries. The importance of this group within the Saadian library can be related to what Muḥammad Hajji called the ‘religious character’ of Moroccan thought in Saadian times.36 The advent of a dynasty claiming a sharifian origin gave more weight to the devotion to the Prophet. A look at the dated manuscripts, covering 35 texts, supports Hajji’s observation since they indicate a steep rise of interest for these poems during the sixteenth century (Table 2).37 Table 2: Religious poetry. 12th century 13th century 14th century 15th century 16th century Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Unitary 0 0 0 0 0 0 3 1 10 1 Coll. 1 (1) 0 1 (2) 1 (5) 0 3 (4) 0 0 3 (4) 2 (4) 1 0 1 1 0 3 3 1 13 3 Subtotal Total 1 2 3 4 16 The significantly higher proportion of sixteenth century copies of religious poetry, in comparison with the other dated manuscripts of that group, is not paralleled at the level of the dated poetry manuscripts (Table 1). This observation will receive further support when these results will be compared with other parts of the poetry collection in the library.38 A closer look at the manuscripts containing a poem in the praise of the Prophet highlights the importance of this specific genre (Table 3): 35 See RBME 922, 135, 2488, 5474. 36 Hajji 1976, 72–73. 37 On this table and the next one, the number of texts in the multiple-text manuscripts that are related to the topic dealt with (here for instance: religious poetry) is indicated between parenthesis after the number of copies. 38 RBME 304 (1000 ah/1592 ce), 315 (1005 ah/1597 ce), 367 (962 ah/1554–1555 ce), 385 (967 ah/1559 ce), 414 (984 ah/1576 ce), 417 (994 ah/1586 ce), 420 (1004 ah/1595 ce and 1009 ah/1601

Nuria de Castilla Table 3: Praise of the Prophet. 12th century 13th century 14th century 15th century 16th century Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Unitary 0 0 0 0 0 0 2 1 7 1 Coll. 1 (1) 0 0 1 (4) 0 3 (4) 0 0 1 (2) 2 (2) 1 0 0 1 1 3 2 1 8 3 Subtotal Total 1 1 3 3 11 With a total of 24 texts in 19 dated copies, this genre is well represented. It is no surprise to find among the texts of two very popular devotional poems, the Burda by al-Būṣīrī (d. 1294 ce)39 and the Bānat Suʿād by Kaʿb b. Zuhayr40 – mostly in commented form. Actually, commentaries of both works are known to have been produced in Saadian times.41 However, only one text comes in various copies, the Minaḥ al-samīʿ bi-sharḥ tamlīḥ al-badīʿ bi-madīḥ al-shafīʿ by al-Ḥumaydī (d. 1596 ce).42 These extracts from a poem by a contemporary of Aḥmad al-Manṣūr, are found in four manuscripts: three are in an Eastern hand (Fig. 2), one of them being a copy by the author.43 They are another indication of the way in which the collection relied on the importation of manuscripts from the Near East at a late date, but also of the interest shown by the library’s owners for recent compositions. ce), 421 (1000 ah/1592 ce), 422 (994 ah/1586 ce), 423 (994 ah/1586 ce), 440 (944 ah/1537 ce), 446 (1015 ah/1606 ce) and 487 (910 ah/1504 ce) are sixteenth century copies in Eastern script; RBME 318 (983 ah/1576 ce), 361 (993 ah/1585 ce) and 384 (914 ah/1508 ce) were written in a Maghribi hand during the same period. 39 See GAL I, 264–267, particularly 265–266; S I, 467–470. The text or its commentaries are found in manuscripts Arabe 282 (s.a.), 283 (s.a.), 3042 (1000 ah/1592 ce), 316 (s.a.), 318 (983 ah/1576 ce), 3902 (s.a.), 413 (881 ah/1476 ce), 414 (984 ah/1576 ce), 4705 (s.a.) and 4871 (910 ah/1504 ce). In addition, it also appears in manuscripts that are not included in the section on poetry, for instance in Arabe 2481 (970 ah/1562 ce). 40 See GAL I, 38-9; S I, 68–69. The text or its commentaries are found in manuscripts RBME 2701 (799 ah/1397 ce), 3041 (1000 ah/1592 ce), 305 (s.a.) and 4706 (s.a.). 41 Hajji 1976, 175, n. 84 (Bānat Suʿād) and 85 (Burda). 42 See GAL II, 272, no. 7; S II, 383. 43 RBME 354 (s.a., after 993 ah/1586 ce), 421 (1000 ah/1592 ce), 422 (994 ah/1586 ce) and 423 (994 ah/1586 ce, in the author’s hand); RBME 421 is in a Maghribi hand.

Nuria de Castilla In various multiple text manuscripts, a poem in the praise of Muhammad comes first in the sequence: this disposition occurs in twelve manuscripts found in the poetry section of the Escorial collection, most of them being Maghribi copies (8).44 The distribution of the texts related to Abbasid poetry (30 in total) according to their date is notably different from the previous example (Table 4). In Hajji’s words, ‘Muḥammad al-Mahdī aš-Šayḫ était un lettré, ayant appris par cœur le dīwān d’al-Mutanabbī. De même ses fils Muḥammad al-Mutawakkil et Aḥmad al-Manṣūr, ainsi que ses petits-fils Muḥammad b. ʿAbd al-Qādir et Zaydān étaient tous des savants, des écrivains et des poètes’.45 In the library of sultans who were cultivating the image of princes fond of poetry, it is no wonder to find the works of the most famous poets of that period, like the dīwāns of al-Mutannabbī (d. 965 ce),46 Abū Tammām (d. 846 ce),47 al-Maʿarrī (d. 1057 ce)48 or Abū Nuwās (d. c. 810 ce) to name only the most famous.49 However, there is no similar gap between the sixteenth century copies and those that were produced earlier as observed in the case of religious poetry. The picture offered is that of a rather steady distribution. Table 4: Abbasid poetry. 12th century 13th century 14th century 15th century 16th century Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Diwan 1 1 3 2 2 0 0 0 3 1 in Coll. 0 0 0 0 1 (1) 0 0 0 1 (1) 1 (2) Commentary in Coll. 1 0 0 1 1 0 2 0 3 1 0 0 0 1 (1) 0 0 0 0 2 (4) 0 Subtotal 2 1 3 3 3 0 2 0 6 2 Total 3 6 3 2 8 44 RBME 304 (1000 ah/1592 ce), 399 (s.a.), 436 (s.a.) and 442 (562 ah/1167 ce) are in Eastern script; Arabe 270 (s.a.), 361 (993 ah/1585 ce), 362 (742 ah/1342 ce), 363 (771 ah/1370 ce), 384 (914 ah/1508 ce), 404 (666 ah/1268 ce and 676 ah/1278 ce), 470 (s.a.) and 487 (910 ah/1504 ce) are in Maghribi script. 45 Hajji 1976, 65. 46 GAL I, 86–89; S I, 138–141. RBME 272 (699 ah/1300 ce) and 394 (588 ah/1192 ce). 47 GAL I, 84–85; S I, 134–137. RBME 291 (1188 ce according to Dérenbourg) and 415 (556 ah/1161 ce). 48 GAL I, 254–255; S I, 449–454. RBME 273 (s.a.) and 435 (s.a.). 49 GAL I, 75–77; S I, 114–118. RBME 311 (1002 ah/1593 ce).

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection Among the thirteen dated manuscripts of Abbasid poems (Table 4),50 seven are copies from the twelfth and thirteenth centuries,51 a situation contrasting sharply with what has been observed about the religious poetry. By contrast, only four are dated to the sixteenth century.52 This distribution might indicate a preference for trustworthy manuscripts – two of them having been previously in a princely library (Fig. 3).53 Conversely, among the nine dated manuscripts with commentaries of Abbasid poetry included in this group,54 the proportion of copies of the fifteenth and sixteenth centuries, more frequently produced in the East, is higher.55 Within the manuscripts of Abbasid poetry, the various copies, dated or undated, of collected works by al-Mutanabbī or commentaries of his poems confirm Hajji’s observation about the taste for his poetry during the Saadian period.56 A commentary of his work by ʿAbd al-ʿAzīz al-Fishtālī, the Tartīb Dīwān al-Mutanabbī, was actually ‘composé sur ordre d’Aḥmad al-Manṣūr, classé d’après l’alphabet marocain, avec une belle introduction littéraire et des titres attrayants. De nombreuses gloses ont trait à la lexicographie, la rhétorique, les plagiats poétiques, etc.’57 The manuscript 609 G with this text is kept in the Bibliothèque nationale du Royaume du Maroc in Rabat and is a witness of the great value the sultan attached to this poetry: ‘c’est une copie royale ayant appartenu à Aḥmad al-Manṣūr, dont elle porte les notes séchées à l’aide de poudre d’or’.58 50 On this table and the next one, the number of manuscripts containing either diwans or commentaries is shown on the first line of each category; on the next line, the number of multiple text manuscripts and the number of texts concerned is indicated for information. 51 RBME 415 (556 ah/1161 ce), 394 (588 ah/1192 ce), 320 (640 ah/1242–1243 ce), 277 (652 ah/1254 ce), 435 (653 ah/1255 ce), 344 (694 ah/1295 ce) and 272 (699 ah/1300 ce). RBME 415, 344 and 272 were copied in a Maghribi hand. 52 Arabe 3693 (1002 ah/1594 ce) is in a Maghribi hand; Arabe 311 (1002 ah/1593 ce), 375 (984 ah/1577 ce) and 4202 (1009 ah/1601 ce) were transcribed in the East. 53 Arabe 277 (652 ah/1254 ce) and 320 (640 ah/1243 ce). 54 Arabe 4761 (619 ah/1222 ce) and 321 (970 ah/1563 ce) are in a Maghribi hand; Arabe 308 (529 ah/1134–1135 ce), 309 (736 ah/1335–1336), 322 (844 ah/1440 ce), 323 (876 ah/1471 ce), 324 (1004 ah/1595 ce), 3251 (1005 ah/1597 ce) and 361 (1004 ah/1595 ce) were transcribed in the Near East. 55 This is the case for manuscripts Arabe 308 (529 ah/1134–1135 ce), 309 (736 ah/1335–1336 ce), 322 (844 ah/1440 ce), 323 (876 ah/1471 ce), 325 (1005 ah/1597 ce) and 480 (881 ah/1476 ce), Arabe 312 (700 ah/1300–1301 ce) and 321 (1004 ah/1563 ce) being Maghribi copies. 56 In addition to the two copies of his collected works, commentaries are found in manuscripts Arabe 3061 (s.a.), 307 (s.a.), 308 (529 ah/1134–1135) and 309 (736 ah/1335–1336 ce). 57 Hajji 1976, 175, n. 82. About the author, see GAL S II, 680. 58 Hajji 1976, 175, n. 82.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection One should however note that the poems of al-Ṭughrā’ī (d. after 1121 ce),59 especially his Lāmiyat al-ʿajām, are particularly well represented in the collection: in addition to an early copy of his dīwān,60 five commentaries of his most famous poem were kept in the Saadian library,61 the abridged version of al-Ṣafadī’s by al-Damīrī (d. 1405 ce) being found in three sixteenth century copies.62 Hajji devotes a few pages to the Andalusi component in the Saadian cultural and intellectual life, but he does not mention the afterlife of the poetry of al-Andalus in Saadian Morocco – except for songs and music.63 Table 5: Poetry from al-Andalus and strophic poetry. 12th century 13th century 14th century 15th century 16th century Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Classical 0 1 1 1 0 2 0 0 0 2 in Coll. 0 0 1 (1) 1 (1) 0 1 (2) 0 0 0 2 (3) Strophic 0 0 0 0 0 0 1 0 0 2 in Coll. 0 0 0 0 0 0 0 0 0 2 (2) Subtotal 0 1 1 1 0 2 1 0 0 4 Total 1 2 2 1 4 The classical poetry from al-Andalus and the muwashshah and zajal genres are actually not represented by a large number of manuscripts, but they were nevertheless of special interest for their princely readers. According to the evidence available, the collection covers a wide spectrum of the production, with poets like Ibn Khafāja (d. 1139 ce),64 Ibn Sahl al-Ishbīlī (d. 1251 ce)65 or Ibn Khātima al-Anṣārī (d. c. 1369 ce).66 Anthologies are also found in the collection, for instance the Qalāʾid al-ʿiqyān wa maḥāsin al-aʿyān compiled by Ibn Khāqān (d. 1134 ce) and 59 GAL I, 247; S I, 439–440. 60 RBME 320 (1242–1243 ce). 61 RBME 321 (970 ah/1563 ce), 322 (844 ah/1440 ce), 323 (876 ah/1471 ce), 324 (1004 ah/1595 ce) and 325 (1005 ah/1597 ce to 1004 ah/1596 ce, with three commentaries). 62 RBME 321, 324 and 3252. For al-Damīrī, see GAL II, 138; S II, 170–171. 63 Hajji 1976, 77–80. 64 GAL I, 272; S I, 480–481; RBME 378 (s.a.). 65 GAL I, 273–274; S I, 483; RBME 379 (s.a.). 66 GAL II, 259; S II, 369; RBME 381 (738 ah/1338 ce).

Nuria de Castilla preserved in a twelfth century Maghribi copy that was at some moment in the East, at least until 1564–1565, before it was acquired by the Saadian library.67 Most of the copies are actually in Maghribi script: among the dated manuscripts, there is only one in an Eastern script.68 The Andalusian strophic poetry, muwashshaḥ and zajal, are also found in the Saadian collection, represented by four texts.69 Their distribution according to the region of copy is in itself interesting of the exchanges between the two parts of the Mediterranean. The only dated volume in an Oriental hand is a fifteenth century copy.70 The two Maghribi manuscripts were completed in the second half of the sixteenth century, one of them a few months only after the ‘Battle of the three kings’ and the accession to the throne of Aḥmad al-Manṣūr in 1578.71 Three of the manuscripts of this group witness the keen interest of the Saadian élite for these specific genres.72 RBME 455 contains the al-Siḥr wa-l-shiʿr by Lisān al-dīn Ibn al-Khaṭīb (d. 1374 ce)73 followed by a part of the al-Muqtaṭaf min azāhir al-ṭuraf by Abū al-Ḥasan ʿAlī Ibn Saʿīd al-Magistri al-Andalusī (d. 1286 ce)74 and bears notes relating it directly to Aḥmad al-Manṣūr, a clear indication of his appreciation of Ibn al-Khaṭīb (Fig. 4). The al-Siḥr wa-l-shiʿr is found again in RBME 456, prepared for the library of an unidentified Ḥasanī sharif in the second half of the sixteenth century. The third copy, Arabe 369, has been written by Idrīs b. ʿAlī b. Ibrāhīm b. ʿAlī b. Rāshid al-Ḥasanī between 998 ah/1590 ce and 1002 ah/1594 ce: a member of another branch of the Ḥasanī sharifs, he was living in Marrakech under Aḥmad al-Manṣūr’s reign but fell into disgrace and had to leave the city.75 It may be in the aftermath of this episode that his books were incorporated into the sultan’s library. The uncommon layout of the first and second texts in RBME 369 (Fig. 5) is similar to that of two examples pointing to an Andalusian fashion: the first is RBME 344, transcribed in Granada, and the other a fourteenth-century Andalusian copy of Abū Firās al-Ḥamdānī’s Dīwān (now in a private collection in Tiznit).76 The 67 GAL I, 339; S I, 579; RBME 357 (591 ah/1194 ce). 68 RBME 408 (657 ah/1259 ce). 69 RBME 3691 (999 ah/1591 ce), RBME 434 (863 ah/1459 ce), RBME 438 (s.a.) and RBME 4552 (986 ah/1578 ce). On the muwashshaḥ, see Rosen 2000, 165–189. 70 RBME 434 (863 ah/1459 ce); the other Oriental copy is not dated (RBME 438). 71 RBME 3691 (999 ah/1591 ce) and RBME 4552 (s.a., but the first text of this copy bears a date of 986 ah/1578 ce). 72 RBME 356 (990 ah/1582 ce), RBME 455 (986 ah/1578 ce) and RBME 456 (s.a.). 73 For al-Siḥr wa-l-shiʿr, see GAL II, 263, no. 14; S II, 373. 74 The text has been identified by ‘Abd al-‘Azīz al-Ahwanī, see al-Ahwanī 1948. 75 al-Maqqarī 1983, 188–189. I am indebted to Dr. M. Zouihal for this reference. 76 I am indebted to Prof. L. Tahali for this reference.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection latter was also part of Mūlāy Zaydān’s manuscripts. This uncommon layout might be indicative of a specific taste for Andalusian models in Saadian times. The Andalusian poetry certainly had a special appeal for the Saadian élite since a fourth manuscript can be added to the list: RBME 356, a princely copy with two anthologies of Andalusian poetry, the first one by al-Tujībī (d. 1202 ce)77 and the second one by al-Bulfiqī, has been transcribed for the sultan’s library (Fig. 6). The connection with Aḥmad al-Manṣūr and the Saadian élite is clear, a situation that we did not come across previously in this survey of the poetry manuscripts, except for ʿAbd al-ʿAzīz al-Fishtalī’s Tartīb Dīwān al-Mutanabbī. Actually, the Saadian library contains only a very few copies commissioned by the sultans. Many volumes bear a statement of ownership, more infrequently the seal of Aḥmad al-Manṣūr or that of his son, Mūlāy Zaydān. This quite unusual number of volumes specifically connected with the sultan or the contemporary élite are indicative of a strong interest which cannot fail to remind of a comment by Hajji:78 Madad al-ǧayš de ʿAbd al-ʿAzīz al-Faštālī, pour lequel l’auteur fait la réplique de Ǧayš at-tawšīḥ de Lisān ad-dīn Ibn al-Ḫaṭīb, rappelant à ce dernier un ensemble de muwaššaḥs andalous omis et ajoutant une grande quantité de muwaššaḥs composés par les poètes marocains sous les Saʿdides, dont ceux consacrés à l’éloge d’Aḥmad al-Manṣūr dépassant les trois cents […]. A la suite vient un muwaššaḥ d’al-Faštālī sur le vin, la nuit et l’aube, y employant allégoriquement les couleurs noire et blanche qui sont les emblèmes des deux dynasties ʿabbaside et saʿdide. Because poetry was an important part of the traditional culture, a technical knowledge of prosody was part of the curriculum and various treatises related to the technical and theoretical aspects of this topic were part of the Saadian library. Table 6: Prosody. 12th century 13th century 14th century 15th century 16th century Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Or. Maghr. Unitary 0 0 0 0 1 1 2 0 1 1 Coll. 1 (1) 0 0 0 0 1 (4) 1 (1) 0 1 (1) 1 (6) 1 0 0 0 1 2 3 0 2 2 Subtotal Total 1 0 3 77 GAL I, 273; S I, 482. 78 Hajji 1976, 178.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection Significantly enough, al-Khazrajī’s (fl. towards 1250) al-Rāmiza al-shāfiyya fī ʿilm al-ʿarūḍ is part of the miscellany RBME 248 that was transcribed in 1562 for a Saadian prince, Muḥammad b. ʿAbd al-Qādir (d. 1567 ce).79 The manuscript is of particular interest since its contents coincide with what we know about Aḥmad al-Manṣūr’s education – and he also read al-Khazrajī’s treatise which is found as well as its commentaries in various manuscripts, including outside of the section devoted to poetry.80 Al-Khazrajī is with Ibn al-Saqqāṭ among the authors on prosody most present in the Saadian library,81 a fact that can be related to the information found in sources about various scholars of this period who are reported to have written their own commentary of a work by either al-Khazrajī or Ibn al-Saqqāṭ.82 Al-Khazrajī’s work is found in various commentaries.83 In the section with poetry, eight texts are either one of their compositions, or a commentary. The proportion of Maghribi copies containing texts about prosody and related matters (6) is higher than is the case usually (Table 6);84 six copies are written in a Maghribi script against ten in a hand typical of the Eastern part of the Islamic world. Multiple-text manuscripts are also more common: seven manuscripts, containing 36 texts, against 9 single-text copies. As far as dated copies are concerned, many were written during the sixteenth century, indicating a strong and enduring interest for this topic.85 Many treatises are found in multiple-text manuscripts. Three of these are of special interest: one is a collection including material from the fourteenth and sixteenth centuries, the other two being miscellanies dated to the sixteenth century. A special mention should be made of RBME 288, probably more representative of contemporary trends. First, it bears Aḥmad al-Manṣūr’s ex libris, followed by that of his son, Mūlāy Zaydān, a sequence that is not very frequent in the collection and tends to single out works considered as important. Second, it includes three texts on dubayt, a quatrain that became popular in Saadian times – notably for devotional compositions. 79 See Déroche 2014; Déroche 2019. 80 See RBME 24814 (1562 ce), 3302 (1582 ce), 332 (1598 ce), 4893 (1598 ce), 6793 (1393–1394 ce), 78811 (1409 ce) and 96412 (1554 ce). See GAL I, 312; S I, 545. 81 See manuscripts RBME 1862 (1414 ce), 2882 and 3 (s.a.), 3305 (1582 ce), 3961 (s.a.), 4103 (1569 ce) and 78810 (1409 ce). See GAL I, 282; S I, 495. 82 Hajji 1976, 177 and n. 100. 83 RBME 3304 (text no. 2 in the volume is dated 990 ah/1582 ce), 332 (1007 ah/1598 ce), 333 (899 ah/1493–1494), 334 (908 ah/1502 ce), 4104 (977 ah/1569 ce), 416 (s.a.) and 4893 (1007 ah/1598 ce). 84 The information on this table is shown according to the same principles as on Table 2. 85 Five out of nine manuscripts: RBME 3304 (990 ah/1582 ce), 332 (1007 ah/1598 ce), 334 (908 ah/1502 ce), 4104 (977 ah/1569 ce) and 4893 (1007 ah/1598 ce).

Nuria de Castilla 3 Conclusion The present study is a first step into a broader research on the history of the Saadian library and reading in the Saadian period. It is too early to draw general conclusions – and in addition we know that we don’t possess the whole library. However, a few facts draw the attention. As could be expected, sixteenth century copies are more numerous than the older ones (see Table 1). It is however quite striking that, when looking at specific areas, the proportion varies greatly: the religious poetry stands clearly apart and shows a surge in interest during the sixteenth century. The presence in the Saadian library of various copies of the same work is puzzling. We expect to find duplicates the works of famous poets, the diwān of Mutanabbī (d. 965 ce) for instance.86 The four copies of al-Ḥumaydī’s Minaḥ al-samīʿ bi-sharḥ tamlīḥ al-badīʿ bi-madīḥ al-shafīʿ could be the result of the interest for new compositions -perhaps combined with a lack of coordination between agents buying for the sultan’s library. But how can we explain that three copies of the Muqaddima fī al-ʿarūḍ (a text on metric) by Ibn al-Saqqāṭ were part of the collection?87 Various sources confirm that it was a very popular handbook on prosody, and the same also applies to Ibn al-Qaṭṭāʿ’s al-ʿArūḍ al-bāriʿ,88 but we would not expect so many copies of elementary treatises in a sultan’s library. The multiple-text manuscripts may explain this situation, or at least partly, since their contents were more difficult to handle during a process of acquisition which, as we have seen, involved long distance. In addition, the contents of a multiple-text manuscript are often described with reference only to the first one as can be seen in documents related to books.89 As was hoped, the interests of contemporary readers begin to emerge. The para-textual data found on the manuscripts is of enormous importance in our assessment. The copies with Andalusian poetry, both classical and strophic, provide some support to the conclusion that Aḥmad al-Manṣūr was very interested in this specific tradition. In a similar way, the holdings of the collection and the copy of al-Fishtalī’s Tartīb Dīwān al-Mutanabbī show that al-Mutanabbī’s poems enjoyed a particular place in the sultan’s literary culture. On the other 86 RBME 272 (699 ah/1300 ce) and 394 (588 ah/1192 ce). 87 RBME 2883 (s.a.), 3305 (s.a., text no. 2 dated 990 ah/1582 ce) and 3961 (s.a.), all Maghribi copies. 88 RBME 3283 (s.a.), texts no. 2 dated 877 ah/1472 ce and no. 5 dated 750 ah/1349 ce, 3303 (s.a., text no. 2 dated 990 ah/1582 ce) and 331, the first being an Eastern copy. 89 See in this volume the paper by F. Déroche and L. Tahali, ‘Collecting Books in Eighteenth-Century Morocco: the Bannānī Library in Fez’.

Princes’s Readings: The Poetry in Mūlāy Zaydān’s Collection hand, there is no trace of any earlier Saadian library: the names of ʿAbd Allah al-Ghalib (r. 1557–1574) and of his son Muḥammad al-Mutawakkil (r. 1574–1576) may have been erased in a damnatio memoriæ, but those of Aḥmad al-Manṣūr’s father, Muḥammad al-Shaykh (r. 1529–1557), and brother ʿAbd Allāh (r. 1576–1578) should have been preserved. In his ownership statements the sultan quite systematically names his father and grandfather. Did such a library exist? Or was it Aḥmad al-Manṣūr’s initiative to establish a princely library which went further than a collection of some books? The research on the library is only at its beginning, but our first findings provide at least some answers about the holdings of a late Medieval or early Modern library in the Islamic world and shed a new light on the culture of Saadian Morocco. Acknowledgements The research leading to these results has received funding from the European Research Council under Horizon 2020 framework Grant Agreement no. 670628, project SICLE (Saadian Intellectual and Cultural Life). I thank heartily Lbachir Tahali for his remarks on the draft of the present paper. Abbreviations GAL = Carl Brockelmann, Geschichte der arabischen Literatur, 2 vols, Weimar: Felber, 1898–1902. S = Carl Brockelmann, Geschichte der Arabischen Literatur. Supplement-Band, 3 vols, Leiden: Brill, 1937–1942. References Primary sources al-Fashtālī, ʿAbd al-Azīz, ‫عبد العزير الفشتالي‬, Manāhil al-ṣafā fi ma’āthir mawālīnā al-shurafā, ‫مناهل‬ ‫الصفا في مىاثير موالينا الشرفا‬, ed. ʿAbd al-Kurayyīm, Rabat: Matbūʿāt Wizārat al-Awqāf, 1972. Gannūn, ʿAbdallāh (ed.), ‫عبد الله غ ّنون‬, Rasāʾil saʿdiyya, ‫رساىل سعديّة‬, Tétouan: Dār al-Ṭibāʿa al-Maġribiyya, 1954. Le Tourneau, Roger, ‘Histoire de la dynastie saʿdide. Extrait de al-Turğumān al-muʿrib ʿan duwal al-Mašriq wal Maġrib d’Abū al-Qāsim b. Aḥmad b. ʿAli b. Ibrāhīm al-Zayyānī. Texte, traduction et notes présentés par L. Mougin et H. Hamburger’, Revue de l’Occident musulman et de la Méditerranée (1977), 23/1: 7–109. al-Manjūr, Aḥmad, ‫احمد المنجور‬, Fihris ‫فهرس‬, ed. Muḥammad Hajji, Rabat: Dār al-Maġrib, 1976.

Nuria de Castilla al-Maqqarī, ‫المقري‬, Rawḍat al-ās al-ʿāṭirat al-anfās fī dhikr man laqītuhu min aʿlām al-ḥaḍratayn Murrākush wa Fās, ‫روضة اآلس العاطرة االنفاس في ذكر من لقيته من اعالم الحضرتين مرّ اكش وفاس‬, ed. ʿAbd al-Wahhāb Ben Manṣūr, 2nd edn, Rabat, Imprimerie Royale, 1983. Secondary literature al-Ahwanī, ʿAbd al-ʿAzīz M. (1948), ‘El “Kitāb al-Muqtaṭaf min azāhir al-ṭuraf” de Ibn Saʿīd’, al-Andalus, 13/1: 19–33. Antolín y Pajares, Guillermo (1921), La Real Biblioteca de El Escorial (Discursos leídos ante la Real Academia de la Historia en la recepción pública del P. Fr. Guillermo Antolín y Pajares, O.S.A., el día 5 de junio de 1921), [San Lorenzo de El Escorial]: Impr. del Real monasterio del Escorial. Browne, E. G. (1900), A Hand-List of the Muhammadan Manuscripts, Including All Those Written in the Arabic Character, Preserved in the Library of the University of Cambridge, Cambridge: Cambridge University Press. Cano Ledesma, Aurora (1996–2004), Indización de los manuscritos árabes de El Escorial, 3 vols, San Lorenzo de el Escorial: Ediciones Escurialenses. Casiri, Miguel, Bibliotheca arabico-hispana Escurialensis, sive Librorum omnium mss. quos arabice ab auctoribus magnam partem arabo-hispanis compositos bibliotheca coenobii Escurialensis complectitur, Madrid, A. Pérez de Soto imprimebat, 1760-1770 (2 vol.). de Castilla, Nuria (2019), ‘Maghrebi Bindings in Ottoman Dress. About Techniques and Changes of Taste in Saadian Morocco’, Turcica, 50: 89–113. Dérenbourg, Hartwig (1884–1903), Les manuscrits arabes de l’Escurial, 2 vols, Paris: Ernest Leroux. Déroche, François (2012), ‘Le sultan à l’école. A propos d’une licence délivrée au sultan saadien Ahmad al-Mansur’, Comptes Rendus de l’Académie des Inscriptions et Belles Lettres, 2012: 1217–1228. Déroche, François (2016), ‘Des risques et des livres’, in Institut de France. Séance solennelle de rentrée des cinq académies. Mardi 25 octobre 2016, Paris: Institut de France, 43–48. Déroche, François (2019), ‘The Prince and the Scholar. A Study of Two Multiple-text Manuscripts from Fifteenth and Sixteenth Centuries Morocco’, in Alessandro Bausi, Michael Friedrich and Marilena Maniaci (eds), The Emergence of Multiple-Text Manuscripts (Studies in Manuscript Cultures, 17), Berlin: De Gruyter, 171–197. García-Arenal, Mercedes (2009), Ahmad al-Mansur. The Beginnings of Modern Morocco, Oxford: Oneworld. Hajji, Muhammad (1976), L’activité intellectuelle au Maroc à l’époque saʿdide, 2 vols, Rabat: Dar El Maghrib. Justel Calabozo, Braulio (1987), La Biblioteca Real de El Escorial y sus manuscritos árabes. Sinopsis histórico-descriptiva, 2nd edn, Madrid: Instituto hispano-árabe de cultura. Mouline, Nabil (2009), Le califat imaginaire d’Ahmad al-Mansûr. Pouvoir et diplomatie au Maroc au XVIe siècle, Paris: Presses Universitaires de France. Rosen, Tova (2000), ‘The Muwashshah’, in María Rosa Menocal, Raymond P. Scheindlin and Michael Sells (eds), The Literature of Al-Andalus, Cambridge: Cambridge University Press: 165–189. Smith, Richard L. (2006), Ahmad al-Mansur: Islamic Visionary, New York: Pearson Longman. Uluç, Lale (1999), ‘Ottoman Book Collectors and Illustrated Sixteenth Century Shiraz Manuscripts’, Revue des mondes musulmans et de la Méditerranée, 87–88: 85–107.

François Déroche, Lbachir Tahali Collecting Books in Eighteenth-Century Morocco: The Bannānī Library in Fez Abstract: A recently found waqf deed with a list of books documents the contents of a part of Muḥammad b. ʿAbd al-Salām Bannānī’s library. This Moroccan scholar from Fez, who died in 1163 ah/1750 ce, wrote and collected books; he established a zāwiya close to the Qarawiyyin mosque where his collection of manuscripts was kept. The deed is published in this paper. It is the starting point of a research into Muḥammad Bannānī’s library: with the documentation at hand, it is possible to identify eighty-three titles that once belonged to him, but also to study the strategies of a Moroccan scholar in the eleventh ah/eighteenth ce century to collect books and to organize the survival of his library for his children. An effort to defend the prestige of the family is also apparent since the collection helped to preserve the successive contributions of various generations of scholars related by kin. The Qarawiyyin library in Fez keeps, besides its collection of manuscripts which has been the subject of a series of catalogues, a deposit of manuscripts and papers in a state that does not allow to have them kept with the rest of the collection. One of us (FD) was allowed to look at these documents and was able to identify a list of manuscripts that had been presented as a waqf: it has been the starting point for a wider research (Fig. 1). The deed had been written on a sheet of paper that has been folded in six. Its left side has been torn away, as well as the bottom of the sheet, so that the texts that were in these places have either been partially mutilated or completely lost. Its current dimensions are c. 220 × 140 to 150 mm. 1 The waqf deed The layout immediately helps identifying the different parts of the document: at the top and bottom is the text of the waqf itself; between the two, the list of the books has been arranged in six columns and ten rows, according to a familiar arrangement for this kind of information. Reference can be made here to the waqf of Shaykh Khālid al-Shahrazūrī of 1242 ah/1827 ce1 or to the catalogues of 1 See de Jong and Witkam 1987, 76. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-004

Collecting Books in Eighteenth-Century Morocco Ottoman libraries. The identification of the works concerned is facilitated by a lay out intended to make its reading easier: in each case one finds either the title, often followed by the name of the author, or the name of the latter only; relatively frequently, an indication about the nature of the manuscript (majmūʿ, for instance) or the number of volumes is also found. The whole document has been written by the same scribe who used a script Muḥammad al-Manūnī suggested to call al-musnad al-zimāmī in his classification of Maghribi scripts2 and where appear some peculiarities, like the way of writing ‘Sīdī’ or a heart-like shape of ṭāʾ when the letter is inside a word (Fig. 1, lines 2 in Part 1 of the deed and 2 in item 22 respectively). The text of the waqfs is as follows: Part 1: َ ‫ َحب ََّست‬.‫الح ْم ُد ِ َّهَّلِل‬ ْ ‫السيِّ َد ُة‬ َّ ‫ات‬ ُ ‫الز ِكي‬ ُ ‫صو َن‬ ُ ‫األ َخ َو‬ ]‫عشو[ة‬ ِ ‫ات ال َف‬ ُ ‫اض َاَل ُت ال َم‬ َّ ‫ َو‬،‫والسيِّ َد ُة َفاطِ َم ُة‬ َّ ،‫ار َحة‬ َّ ‫َّات؛‬ َ .1 ِ ‫السيِّ َد ُة َف‬ َّ ْ ْ ْ ْ ُ َّ ‫ات ْال َعالِم ْال َع‬ ُ ‫ َب َن‬.٢ ‫ِيع الكتُ ِب التِي َو ِرث َن َها م ِْن‬ ‫م‬ ‫ه‬ ‫ال‬ ‫ة‬ ‫م‬ ‫اَّل‬ ِ ُ َّ ‫اع ِة َسيِّدِي م َُح َّم ِد ب ِْن َع ْب ِد‬ َ ‫ام َشي ِْخ ال َج َم‬ َ ‫الس َاَلم َبنَّانِي َجم‬ َ َ ِ ِ َ ْ ْ ِ ‫أَ ِب‬ ]‫يه َّن‬ ِ ِ‫ور [ َوم ِْن أخ‬ ِ ‫يه َّن ال َمذ ُك‬ َ َ َ ْ ْ ْ َّ ْ َ َ َّ ْ ُ َ ‫ َوهِ َي‬،‫س َسيِّدِي م َُحمَّد‬ ِ ِ‫ َومِن أخ‬،‫ ال َع َّاَّل َم ِة اأَلفض ِل َسيِّدِي َح ْمدون تغ َّم َد ُه هللا ِب َرحْ َمتِه‬.٣ ِ ِّ‫يهن ال َعاَّل َم ِة ال ُمدَر‬ َ َ ْ ْ ُ َ َ ُ َ ]‫ور‬ َ ‫ْال َم ْذ ُك‬ ِ ِ‫ َعلى [أخ‬،‫ورة أ ْسفل‬ ِ ‫يه َّن ال َمذك‬ ْ َ َ َ ْ َّ ْ ً ً ْ ً َ ْ‫ َو َوقفا م َُخلدا إِلَى أ ْن َي ِر َث هللا اأْلر‬،ً‫ فِي ال َوثِي َق ِة أ ْس َفلَ ُه؛ َحبْسا ُم َؤبَّدا‬.٤ .‫ارثِين‬ ِ ‫ َو ُه َو َخيْرُ ال َو‬،‫ض َو َم ْن َعلَ ْي َها‬ ُ ْ ‫أَواخِ َر جُ َمادَى‬ ]‫سِ ٍّت‬/[ ‫اأْلولَى َس َن َة‬ َ .‫ َو َث َمانِينَ َومِا َئة َوأَ ْلف‬.٥ 1. Praise be to God! Have formed in good hands the virtuous, protected and generous sisters, the ladies Fāriḥa, Fāṭima and ʿAshw[a],3 2. daughters of the learned, erudite and excellent shaykh of the community, Sīdī Muḥammad b. ʿAbd al-Salām Bannānī, all the books they had inherited from their father – already mentioned – [and from their brother] 3. the wise and good Sīdī Ḥamdūn, may God cover him with His grace, and from their brother, the learned professor Sīdī Muḥammad; these are the books mentioned below, in favor of [their brother ʿAbd al-Karīm] 2 al-Manūnī 1991, 14. 3 ʿAshwa is an affectionate form of the name ʿAysha, especially in Fez. Various sources in the seventeenth and eighteenth centuries document this: see al-Fāsī al-Fihrī, al-Mawrid, ed. al-Ṣiqillī 2008, 83; al-Manālī al-Zabādī, Sulūk al-ṭarīq, ed. ʿal-Ḥayy al-Yamlāḥī 2012, 408; Muḥammad b. Jaʿfar al-Kattānī, Sulwat al-anfās, ed. al-Kāmil al-Kattānī 2004–2006, II, 211.

François Déroche, Lbachir Tahali 4. mentioned in the deed below; a perpetual legacy and an eternal waqf until God ‘inherits the earth and those who are there’ (Q. 19: 40) because he is ‘the best of those who give inheritance’ (Q. 21: 89). End of Jumādā I 5. 118 [6]/August 28, 1772.4 Part 2: َ ُ ‫ور‬ َ ْ ‫ َشهدَت‬.‫ اَ ْل َح ْم ُد هلل‬.1 َ ُ ‫اأْل َخ َو‬ ‫اج ِم أَعْ َاَل ُه الَّتِي َع َد ُدهَ ا‬ َ ‫ات ْال َم ْذ ُك‬ ِ ‫ات أعْ َاَل ُه أنَّه َُّن َحب َّْسنَ ْال ُكتُ َب ْال ُم َقيَّ َد َة فِي التَّ َر‬ ِ ‫ت ِْس َع ٌة [ َو َخم ُْسونَ َترْ َج َم ًة]؛‬ َ ْ ‫ اَل ِّنهَ اي َُة ِاِلبْن‬:‫ الَّتِي أَ َّولُ َها‬،“‫ ” َترْ َج َمة‬:‫ بالتَّا ِء ْالم َُثنَّا ِة أَ َّو َل‬.٢ َ ‫ َش ْرحُ ْال ُق‬:‫ َوآخِ رُ هَ ا‬،‫اأْلثِير فِي سِ ْف َري ِْن‬ ‫اعي ِاِلب ِْن‬ ِ ‫ض‬ ِ ِ ]ِ‫ْال َوحْ شِ ي؛ َعلَى أَخِ ي[هِ َّن ْال َع َّاَّل َمة‬ َ ْ .٣ ْ ‫ض ِل َسيِّدِي َع ْب ِد ْال َكريم‬ َ ‫اأْل ْف‬ ‫ ثُ َّم َعلَى‬،‫ َوا ْم َت َّد ْت ُفرُ وعُ ُه ْم‬،‫اسلُوا‬ ْ ‫ َو‬،‫[ال َبنَّان ِِّي] النَّ َس ِب‬ َ ‫ َما َت َن‬،‫ َوأَ ْو َاَلدِهِ ْم‬،ِ‫أو َاَل ِده‬ ِ ِ َ ‫َم ْن َي َت َع‬ ]...‫[عم ِِّه َّن‬ َ ‫اطى ْالع ِْل َم م ِْن أَ ْب َنا ِء‬ ْ ْ ُ ْ ً ُ ُ ُ ُ َ ُ َ َ ُ ْ ْ ُ ‫ فإِ ْن ُهم انق َر‬.‫ َوا ْم َت َّدت فرُ وعُ ُه ْم‬،‫اسلوا‬ ‫او َي ِة‬ َ ‫ضوا َر َج َعت الكت ُب ال َمذك‬ َ ‫ َما َت َن‬،‫ َحيْث ُه ْم‬،‫ َبنَّانِي‬.٤ ِ ‫ورة َحبْسا َعلى َز‬ ]‫ات الَّتِي‬ ِ ‫[س‬ َ ‫َوالِ ِد ْالم َُح ِّب‬ َ ‫الص‬ َ ‫ ِب َها َم َح ُّل‬.٥ ‫ور ِة‬ ُ ‫اغ ِة َو َفاس ْال َق َر ِويِّينَ ؛ لِ َي ْن َتف َِع الم َُحب‬ َّ ‫ان م ِْن َح ْو َم ِة‬ َ ‫َّس َعلَي ِْه ْم ِب ْال ُكتُ ِب ْال َم ْذ ُك‬ ِ ‫ ِبال ِّدي َو‬،ِ‫ض ِريحِ ه‬ ]‫ق َِرا َء ًة و ُم[ َقا َبلَ ًة‬ ْ َ‫ َو َم ْن َب َّد َل أ ْو‬. َ‫ َو ُه َو َخيْرُ ْال َوارثِين‬،‫ض َو َم ْن َعلَ ْي َها‬ َ َ َّ ً ً ْ ً ً َ َ َ ْ‫ َو َوقفا م َُخلدا إِلى أ ْن َي ِرث هللا اأْلر‬،‫ و َن ْسخاً؛ َحبْسا ُم َؤبَّدا‬.٦ ِ ‫َغي ََّر َفاهللُ َح[سِ يبُ ُه] َو‬ ُ ُ ‫ َوأ ْش ِه َد َشق‬. َ‫ َو َس َيعْلَ ُم الَّذِينَ َظلَمُوا أَيَّ ُم ْن َقلَ ٍب َي ْن َقلِبُون‬.‫ َسائِلُ ُه‬.٧ َ ‫ات َسيِّدِي َعبْد ْال َك ِريم ْال َم ْذ ُكور أَنَّ ُه َح‬ ]‫ا[ز‬ ِ ‫ِيق ْالم َُحب َِّس‬ ً َ َ ْ َ َ َ ‫ وصارت‬،‫ات‬ ِ ‫ فحِ يزت َعن األخ َو‬،‫ مُعا َينة ك َما َي َج ُب‬،ً‫ َو َب َس َط َي َد ُه َعلَ ْي َها؛ َح ْوزاً تامّا‬،‫ ْال ُكتُ َب ْالم َُحب ََّس َة ُكلَّ َها‬.٨ ]ً‫حبسا ُم[ َؤبَّدا‬ َ ُ ً ً َ َ ُ َ َ َّ‫ َو َعر‬،‫ِيع َما ذك َِر‬ .]...[ ‫يخ أعْ َاَل ُه‬ ِ َّ‫ َوفِي الت‬،]‫ف ِبالنِّ ْس َو ِة َتع ِْريفا [ َتامّا‬ ِ ‫ار‬ ِ ‫ ش ِه َد َعلي ِْه ْم ِب َجم‬.ُ‫ َع َرفوا ق ْد َره‬.٩ 1. Praise be to God! The sisters mentioned above attest to bequeath all the books registered in the descriptions (tarājim) above, the number of which is [5] 9, [description (tarjama)] 2. with a double pointed tāʾ at the beginning of ‘tarjama’, the first of which is al-Nihāya of Ibn al-Athīr in two volumes and the last the Sharḥ al-Quḍāʿī of Ibn al-Waḥshī to [their brother the learned], 3. the excellent Sīdī ʿAbd al-Karīm, descendant of [Bannānī], to his children and the children of the latter as long as they reproduce themselves and their [genealogical] branches spread, then to those who devote themselves to knowledge among the sons [of their paternal uncle...] 4 The date is partly missing; if the number of the units is not 6 and the sīn is the first letter of sana instead, the date is then comprised between 1183 ah, year of Muḥammad b. Muḥammad b. ʿAbd al-Salām Bannani’s death (10 Ramaḍān 1182 ah/18 January 1769 ce) and 1189 ah, i.e. between 1st October 1769 and 28 July 1775.

Collecting Books in Eighteenth-Century Morocco 4. Bannānī5 where they are, as long as they reproduce themselves and their [genealogical] branches spread. And if they disappear, the books mentioned will be bequeathed to the zāwiya6 of the father of the don[ors, where is] 5. the location of his mausoleum, in the dīwān of the jewellers district and of Fez al-Qarawiyyin so that the beneficiaries benefit from these books in terms of reading, [collation] 6. and copy, in perpetual bequest and eternal waqf until God ‘inherits the earth and those who are there’ [Q. 19: 40] because he is ‘the best of those who give inheritance’ [Q. 21: 89]. Whoever destroys or transforms [this ḥabūs], God [will chastise him] 7. and question him and ‘those who are unjust will know to what destiny they turn” [Q. 26: 227]. The brother of the donors, the aforesaid Sīdī ʿAbd al-Karīm declared that he [got] 8. all the books made waqf and that he has laid his hand on these as a whole pre-emption and that he has verified them properly. As a result, the sisters strip themselves of books that have become a legacy [in perpetuity] 9. which is recognized at its true value. All of the aforesaid has been attested by oath and the women were exhaustively identified on the date indicated above [by...].7 [...] The deed involves various members of the Bannānī family (see Appendix II):8 Muḥammad b. ʿAbd al-Salām in the first place, and his children, three daughters, Fāriḥa, Fāṭima and ʿAshwa, and their three brothers Ḥamdūn, Muḥammad and ʿAbd al-Karīm. Sources agree that Muḥammad b. ʿAbd al-Salām had actually six children. The books in the possession of the three sisters came partly from their father, partly from their brothers Ḥamdūn and Muḥammad. It is a family waqf, the beneficiary being the only surviving brother at the time of the deed, ʿAbd al-Karīm. However, the clauses provide that it will remain in the family or pass to the zāwiya the construction of which was started by Muḥammad b. ʿAbd al-Salām. 5 We suppose that a paternal uncle was named here, perhaps Abū ʿAbd Allāh Muḥammad alʿArabī b. ʿAbd al-Salām Bannānī (see below) who could be the father of the cousin Abū Ḥāmid ʿAmm al-ʿArabī (d. 1182 ah/1768 ce). Al-Manālī al-Zabādī mentions the latter as a servant of the shaykh Muḥammad b. ʿAbd al-Salām Bannānī and his children, in charge of his uncle’s zāwiya, but does not provide his own father’s name (al-Manālī al-Zabādī, Sulūk al-ṭarīq, ed. ʿal-Ḥayy al-Yamlāḥī 2012, 295; see also Muḥammad b. Jaʿfar al-Kattānī, Sulwat al-anfās, ed. al-Kāmil al-Kattānī 2004–2006, I, 158). 6 The zāwiya was close to his home and located in the al-Qaṭṭān alleyway. The construction was started when Muḥammad b. ʿAbd al-Salām Bannānī was still alive and finished after his death in order to include his tomb (see al-Manālī al-Zabādī, Sulūk al-ṭarīq, ed. ʿal-Ḥayy al-Yamlāḥī 2012, 294). 7 The names of the witnesses should have been written on the lower part of the document, now missing. 8 This family, of Jewish descent, belongs in J. Schacht’s words to the Fasi aristocracy of Islamic studies (Schacht 1960, 1019).

François Déroche, Lbachir Tahali 2 Muḥammad b. ʿAbd al-Salām Bannānī The latter was a prominent representative of the Bannānī family. He was born in 1082 ah/1671 ce to a learned scholar, ʿAbd al-Salām (d. 1136 ah/1723 ce), and Fāṭima, the daughter of a notary of Fez, Abū al-ʿAbbās Aḥmad, known as Ḥaddū, a son of Maḥammad Mayyāra al-Fāsī (d. 1072/1662), a prominent scholar of the Saadian period.9 Muḥammad b. ʿAbd al-Salām b. Ḥamdūn Bannānī has left a fahrasa that his son ʿAbd al-Karīm has integrated into his Tuḥfat al-fuḍalāʾ and where he enumerates the list of his father’s twenty-two masters among whom the figure of Ḥasan b. Masʿūd al-Yūsī (d. 1102 ah/1690 ce) stands out.10 The Law seems to have been the area where he excelled, since he later taught the Mukhtaṣar at the Qarawiyyin mosque and Ibn Abī Zayd al-Qayrawānī’s Risāla as well as Bukhārī’s Ṣaḥīḥ at the Madrasa al-Miṣbāḥiyya.11 Forced to leave Fez during the famine of 1150 ah/1737 ce, he continued to teach in Tetouan where he had taken refuge.12 He taught at al-Azhar in 1141 ah/1729 ce on his way back from the pilgrimage to Mekka and he may have returned there shortly thereafter.13 He died in 1163 ah/1750 ce. He left behind an imposing body of work that confirms the importance of this scholar from the beginning of the Alawite dynasty. However, it should be noted that his writings remain unpublished, which prompted us to publish a list drawn up by one of us (LT). 9 al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, III, 136; ʿAbd al-Karīm b. Muḥammad b. ʿAbd al-Salām Bannānī, Tuḥfat al-fuḍalā’ al-aʿlām fī al-taʿrīf bi al-shaykh abī ʿAbd Allāh Maḥammad al-Bannānī ibn ʿAbd al-salām, autograph copy, Rabat, Ḥasaniyya Library, 13900, fol. 5v. Muḥammad b. ʿAbd al-Salām Bannānī was considered the last great figure of the old Fāsi school (Schacht 1960, 1019). On Maḥammad b. Aḥmad Mayyāra, see Pellat 1990, 932. 10 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fols 8r–27r. On al-Yūsī, see for instance Hajji 1976, passim; also El-Rouahyeb 2015, 207–212, 217–229. 11 al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, IV, 81. An evaluation of his expertise in legal matters is found in Muḥammad b. Jaʿfar al-Kattānī, Sulwat al-anfās, ed. al-Kāmil al-Kattānī 2004–2006, I, 156. 12 al-Ḍuʿayyif al-Ribāṭī, Tārīkh al-Ḍuʿayyif, ed. al-Būzīdī al-Shīkhī 2007, II, 218; al-Rahūnī, ʿUmdat al-rāwīn, ed. al-Ḥājj al-Sulamī 2001–2015, IV, 130; al-Manālī al-Zabādī, Sulūk al-ṭarīq, ed. ʿalḤayy al-Yamlāḥī 2012, 294. See also Muḥammad b. Jaʿfar al-Kattānī, Sulwat al-anfās, ed. al-Kāmil al-Kattānī 2004–2006, II, 157; Muḥammad Dāwud, Tārīkh Tiṭwān, ed. Dāwud 2013, II, 183. 13 Maḥammad al-Fāsī al-Fihrī, al-Mawrid, ed. al-Ṣiqillī, 91.

Collecting Books in Eighteenth-Century Morocco Works of Muḥammad b. ʿAbd al-Salām b. Ḥamdūn Bannānī Faḍāʾil al-Ḥaramayn al-sharīfayn – Religion.14 Ifādāt wa Inshādāt wa Muqayyadāt.15 Laqṭ nadā al-ḥiyaḍ min azhār nasīm al-riyāḍ fī sharḥ Shifāʾ al-Qāḍī ʿIyāḍ – Religion.16 4. Maghānī al-wafāʾ li-sharḥ maʿānī al-Iktifāʾ – History.17 5. Mukhtaṣar risālat al-Tājūrī fī maʿrifat waḍʿ bayt al-ibra ʿalā al-jihāt al-arbaʿ – Astrology.18 6. Riḥla wa manāsik kubrā – Travel.19 7. Riḥla wa manāsik ṣuġrā – Travel.20 8. Risāla fī faḍl al-niṣf min Šaʿbān – Religion.21 9. Sharḥ ʿalā naẓm ʿAbd al-Raḥmān al-Fāsī fī al-usṭurlāb – Astronomy.22 10. Sharḥ al-ḥikam of Ibn ʿAṭāʾ Allāh – Sufism.23 11. Sharḥ al-Ḥizb al-kabīr li al-Shādhilī – Sufism.24 12. Sharḥ Ḥizb al-Naṣr – Sufism.25 13. Sharḥ Khuṭbat al-Mukhtaṣar – Law.26 14. Sharḥ Kubrā al-Sanūsī – Religion.27 15. Sharḥ al-Kulliyyāt al-Fiqhiyya – Law.28 1. 2. 3. 14 Mentioned by ʿAbd al-Ḥayy al-Kattānī, Fahris, ed. ʿAbbās 1982, I, 225. The titles are listed alphabetically, except for the fahrasa listed at the end. 15 ʿAbd al-Ḥayy al-Kattānī, Fahris, ed. ʿAbbās 1982, I, 225, these notes are the basis for Chaps 6 and 7 of ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fols 8r–27r. 16 Manuscripts Fez, Qarawiyyin Library, 840; Rabat, National Library of the Kingdom of Morocco (now: BNRM), 1456 K. 17 Manuscripts Fez, Qarawiyyin Library, 275; Rabat, BNRM, 162 K (majmūʿ), part of vol. I. It is Muḥammad b. ʿAbd al-Salām’s major work, a commentary of al-Kalāʿī’s K. al-Iktifāʾ on the expeditions of Prophet and the first three caliphs. 18 Manuscript Rabat, BNRM, 1/2188 D. 19 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 21r. 20 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 21r. 21 Manuscript Rabat, BNRM, 16 K (text no. 11), see below. 22 Manuscripts Wazzān, Library of the Great Mosque, 4/1217 and 1/1214; Rabat, BNRM, 1/2237 D, 1411 D and 1469 D. 23 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 8r–27r. 24 Manuscripts Rabat, BNRM, 470 D and 1599 D; Rabat, Ḥasaniyya Library, 29 copies. 25 Manuscript Rabat, Ḥasaniyya Library, 50 JK (autograph copy). 26 Mentioned by al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, IV, 81. 27 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, 40; al-Fihrī al-Fāsī, al-Mawrid, ed. al-Ṣiqillī 2008, 90. 28 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, 40; mentioned by al-Fihrī al-Fāsī, al-Mawrid, ed. al-Ṣiqillī 2008, 91.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. François Déroche, Lbachir Tahali Sharḥ lāmiyat al-Zaqqāq – Law.29 Sharḥ Mukhtaṣar al-Saʿd – Rhetoric.30 Sharḥ Qaṣīdat Ibn Nāṣir – Religion.31 Sharḥ al-Qaṣīda al-rāʾiyya al-lughziyya fī al-Naḥw – Grammar.32 Sharḥ al-Ṣalāt al-Mashīshiyya – Religion.33 Sharḥ al-Tanwīr fī isqāṭ al-tadbīr – Sufism.34 Sharḥ Urjūzat al-Shātibī fī iltiqāʾ al-Sākinayn – Religion.35 Takmīl Sharḥ Ḥudūd Ibn ʿArafa – Law.36 Taʾlīf fī mā lā yanṣarif min al-Asmāʾ – Grammar.37 A collection of poems – Poetry. Fahrasat Aḥmad b. al-ʿArabī b. al-Ḥājj – Bio-bibliography.38 al-Fahrasa al-kubrā – Bio-bibliography.39 al-Fahrasa al-ṣuġrā – Bio-bibliography.40 His three sons are mentioned in the deed: by the date of the waqf, two of them, Ḥamdūn and Muḥammad, had already died. Due to the disparition of the units, we only know that it was written between 11 Ramaḍān 1182 ah/19 January 1769 ce, the day following Muḥammad’s demise, and 28 July 1775. Ḥamdūn, died in Mekka on 1st Ṣafar 1170 ah/26 October 1756 ce.41 He had made the pilgrimage a first time with his father in 1141 ah/1729 ce and returned to the Holy City toward 1167 ah/1753–1754 ce and settled there definitely. Like his father, he was a specialist of the Law, both in teaching and 29 Manuscripts Rabat, ʿAllāl al-Fāsī Library, 276; Wazzān, Library of the Great Mosque, 1155 (partial copy). 30 Mentioned by al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, IV, 81. 31 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, 40; mentioned by al-Fihrī al-Fāsī, al-Mawrid, ed. al-Ṣiqillī 2008, 91. The qaṣīda is known as Sayf al-naṣr. 32 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, 40; al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, IV, 81. 33 Manuscripts Fez, al-Qarawiyyin Library, 1512; Rabat, BNRM, 245 D and 1599 D. 34 Mentioned by al-Fihrī al-Fāsī, al-Mawrid, ed. al-Ṣiqillī 2008, 91. 35 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 21r; al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, IV, 81. 36 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 21r; al-Fihrī al-Fāsī, al-Mawrid, ed. al-Ṣiqillī 2008, 90. 37 Manuscript Rabat, BNRM, 16 K (text no. 1), see below. 38 Manuscripts Rabat, BNRM, 425 D; Rabat, Ḥasaniyya Library, 6778; Casablanca, King ʿAbd al-ʿAzīz Institute, 330. 39 Mentioned by ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 8r–27r. 40 In Muḥammad b. Aḥmad al-Ḥuḍayjī, al-Kunnāsha, MS Rabat, Ḥasaniyya Library, 13003, 339–364. 41 al-Qādirī (1982-1986), IV, 201; al-Manālī al-Zabādī (2012), 295.

Collecting Books in Eighteenth-Century Morocco in practice. He was also the author of an abstract of a treatise on grammar, al-Anwār al-saniyya wa al-azhār al-badīʿa al-bahiyya fī sharḥ khuṭbat al-Alfiyya li Ibn Mālik, composed by his maternal uncle, Muḥammad b. Muḥammad b. Ḥamdūn Bannānī, known as al-Muḥawjab (d. 1140 ah/1728 ce).42 He was also a copyist, as attested by a manuscript written in his hand, the Muntakhab kanz al-ʿUmmāl fī sunan al-aqwāl wa-l-afʿāl by ʿAlī b. Ḥusām al-dīn al-Muttaqī al-Hindī.43 Muḥammad (d. 10 Ramaḍān 1182 ah/18 January 1769 ce) was noted for his interest in Sufism. His training had been provided by his father and his brother Ḥamdūn, but he seems to have limited himself to transcribing works as evidenced in particular by the manuscript Rabat, National Library of the Kingdom of Morocco, 16 K which will be discussed later.44 The third brother, ʿAbd al-Karīm, is mostly mentioned in passing within biographical entries devoted to his father or to his two other brothers;45 however, ʿAbd al-Salām b. Sūda wrote a few lines about him and preserved the date of his death on 16 Dhū al-Qaʿda 1196 ah/23 October 1782 ce.46 He acquired a sound knowledge of legal matters, but his work is limited to the Tuḥfat al-fuḍalāʾ al-aʿlām bi al-taʿrīf bi al-shayḫ Abī ʿAbd Allāh Muḥammad Bannānī ibn ʿAbd al-Salām: as indicated by the title, it referred to his father. We were hoping to find some information on the latter’s life in a copy of the text kept at the Ḥasaniyya Library, but one of us (LT) who was able to consult it found that although a chapter on this subject is announced in the introduction, this part of the text was missing in the manuscript.47 When the deed was written, ʿAbd al-Karīm was the only survivor of the three brothers. The names of the three sisters are known to us by the deed we publish: Fāriḥa, Fāṭima and ʿAshwa. Were it not for this document, we would not know them. At most do we know that one of them married ʿAbd Allāh b. Muḥammad b. Aḥmad al-ʿUnṣurī al-Gānnūnī al-Ḥasanī: he came to Fez in order to study at the madrasa al-ʿAṭṭārīn and became the son-in-law of Muḥammad b. ʿAbd al-Salām Bannānī. Three boys were born from this marriage, Muḥammad, Aḥmad and ʿAlī, who are at the origin of all the Fasi families named Gānnūn, but we do not know which of the three sisters was their mother.48 42 Manscripts Rabat, BNRM, 1780 D/1, 2048 D/1, 2217 D/6, etc.; Riyadh, King Saʿūd University Library, 7036 (autograph copy completed on 12 Ramaḍān 1145 ah/26 February 1733 ce). 43 al-Sibāʿī 2017, II, 103. 44 al-Sibāʿī 2017, II, 132. 45 al-Manālī al-Zabādī, Sulūk al-ṭarīq, ed. ʿal-Ḥayy al-Yamlāḥī 2012, 295; ʿAbd al-Ḥayy al-Kattānī, al-Tanwīh, ed. al-Majīd Khiyālī 2012, I, 226; Muḥammad b. Jaʿfar al-Kattānī, Sulwat al-anfās, ed. al-Kāmil al-Kattānī 2004–2006, I, 157; Muḥammad Makhlūf, Shajarat al-nūr, ed. al-Majīd Khayālī 2003, I, 508. 46 Abd al-Salām b. Sūda, Itḥāf al-muṭāliʿ, ed. Ḥajjī 1997, 53. 47 Manuscript Rabat, Ḥasaniyya Library, 13900. 48 al-Rahūnī, ʿUmdat al-rāwīn, ed. al-Ḥājj al-Sulamī 2001–2015, IX, 94.

François Déroche, Lbachir Tahali 3 The books Now let us turn to the list of the books that were collectively deposited by the three sisters and represented their share of the inheritance with regard to these items. In the table below with a transcription of the list, an identification number has been added in order to make cross-references easier (see Table 1). Table 1: List of the books according to the presentation of the waqf deed (we have added numbers for reference’s sake). 1. Al-Nihāya li-Ibn al-Athīr fī sifrayn 2. Al-Safāquṣī: Muʿrab al-Qurʾān, fī sifrayn 3. Ḥusn al-Muḥāḍara, fī sifrayn 4. Al-Tatāʾī ʿalā al-Mukhtaṣar, fī sifrayn 5. Shāmil Bahrām fī sifr 6. Sifr […] al-s[…] 7. Al-niṣf al-thānī min al-Madkhal li-Ibn al-Ḥājj 8. Juzʾ min Tabṣirat al-Lakhmī 9. Al-awwal min sharḥ al-Miftāḥ 10. Sifr min Ṣaḥīḥ Muslim 11. Al-Maḥallī ʿalā al-Subkī 12. Ibn Sa[yyid al-nās] fi si(frayn) 13. Al-Ujhūrī ʿalā Mukhtaṣar Ibn abī Jamra 14. Ibn Ghāzī ʿalā al-Alfiyya 15. Al-Niṣf al-awwal min Ibn ʿUmar 16. Al-Ujhūrī ʿalā al-Risāla 17. Al-Azharī, al-Muʿrib 18. Ibn ʿA[rafa] wa maʿahu al-[…] 19. Al-Hadī li-Ibn al-Qayyim al-Jawziyya 20. Al-Shaykh Yāsīn ʿalā al-Saʿd 21. Sharḥ al-Shawāhid li-l-ʿAynī 22. Ḥāshiyat Yāsīn ʿalā al-Nāẓm 23. Ibn al-Ḥājib, al-Farʿī 24. Ḥawāshī […] fī sif[r]. 25. Ḥāshiyat al-Saktānī wa-l-Fāsī ʿalā al-Ṣughrā 26. Al-Jazūlī wa mā maʿahu fī mujallad 27. Majmūʿ fīhi shurūḥ al-Burda 28. Sifr min Ibn ʿAṭiyya

Collecting Books in Eighteenth-Century Morocco 29. Al-Saʿd ʿalā al-Talkhīṣ 30. Shal[…] al-[…] 31. Al-Ṣawāʿiq al-muḥriqa li-Ibn Ḥajar 32. Ibn al-Nafīs fī al-ṭibb 33. Wathāʾiq al-Fashtālī 34. Naṣīḥat al-mughtarrīn li-l-shaykh Mayyāra 35. Al-Furūq li-l-Wansharīsī 36. Shar[ḥ…] fī […] 37. Al-Naṣīḥa al-kāfiya fī sifr wāḥid 38. Al-Muḥaṣṣal fī sifr 39. Ibn Baṭṭūṭa fī sifr 40. Majmūʿ mabdaʾuhu Taʾlīf fī mā lā yanṣarif 41. Majmūʿ fī awwalihi Qiṣṣat ahl al-kahf 42. Majm[ūʿ fī awwalihi] al-Tāj [al-Subkī] 43. Sifr mushtamil ʿalā Rawḍat al-azhār 44. Majmūʿ fī awwalihi al-Jawāhir al-ʿaḍudiyya 45. Majmūʿ fī awwalihi Fahrasa 46. Majmūʿ fī awwalihi Tuḥfat Ibn Mālik ʿalā al-Kāfiya 47. Majmūʿ fī awwalihi al-Azharī ʿalā al-Jurrūmiyya 48. Majmū[ʿ fī awwalihi] Ikhtiṣā[r Sharḥ khuṭbat] al-Alfi[yya] 49. Majmūʿ fī awwalihi Qaṣāʾid li-l-Baghdādī 50. Riḥlat wālid al-muḥabbisāt 51. Majmūʿ awwaluhu Ḥāshiyat al-Zarqānī ʿalā al-Luqānī 52. Majmūʿ fīhi Ijāzāt li-l-shaykh al-Manjūr 53. Al-Niqāya li-l-Suyūṭī fī sifr 54. al-[…] li-[…] 55. Nuskhat al-Tashīl fī sifr wāḥid 56. Al-Raṣṣāʿ wa maʿahu al-Shamāʾil fī sifr 57. Kitāb al-Jumān fī akhbār al-zamān 58. Sharḥ Shawāhid Ibn Hishām li-l-Fāsī 59. Sharḥ al-Quḍāʿī li-Ibn al-Waḥshī The list of books includes fifty-nine entries, but the number of volumes and works is slightly higher. With regard to the former, it should be noted that, on the first row, four works are in two volumes, bringing the total to 63 items. When one or two volumes contain the whole text of a work (see for instance nos 1, a two volume set, or 55 and 56, a single volume), the description sometimes notes the fact. On the other hand, several texts are only part of a work in two volumes or more: the first title of the second row (no. 7) is thus the second half of a treatise by Ibn al-Ḥājj and an isolated tome of Ibn

François Déroche, Lbachir Tahali ʿAtiyya’s tafsīr (no. 28) is mentioned in the fourth column of the fifth row. Six texts seem to be in this situation. At first glance, it is quite reminiscent of Nicolas Clénard’s description of the book trade in Fez toward the middle of the sixteenth century: the offer was then limited and often consisted of pieces of manuscripts, ‘so much so that one is often forced to buy the same thing several times if he wants to get the entire work and spend several years at that’.49 But, as we shall see below, other reasons may explain this point. The titles are given in an abbreviated form by the person who drew up the waqf deed. He may have been inspired (at least partially) by the short indications that are usually written on the edge of the manuscripts according to a widespread habit in the Muslim world. As we cannot check with the manuscripts themselves, the identification of the works is sometimes complex; the list given below is only a tentative one. Texts mentioned in the three sisters’ waqf 1. Majd al-dīn al-Mubārak b. Muḥammad al-Jazarī Ibn al-Athīr (d. 606 ah/1209 ce), al-Nihāya fī gharīb al-ḥadīth wa al-athar.50 2. Burhān al-dīn Ibrāhīm b. Muḥammad al-Safāquṣī (d. 743 ah/1342 ce), al-Mujīd fī iʿrāb al-Qurʾān al-majīd.51 3. Jalāl al-dīn ʿAbd al-Raḥmān b. Abī Bakr al-Suyūṭī (d. 911 ah/1505 ce), Ḥusn al-muḥāḍara fī tāʾrīkh Miṣr wa al-Qāhira.52 4. Shams al-dīn Abū ʿAbd Allāh Muḥammad al-Tatāʾī (d. 942 ah/1535 ce), Fatḥ al-Jalīl ʿalā Mukhtaṣar al-shayḫ Khalīl.53 5. Tāj al-dīn Bahrām b. ʿAbd Allāh al-Damīrī al-Dimyāṭī (d. 805 ah/1402 ce), al-Shāmil fī fiqh al-imām Mālik.54 7. Abū ʿAbd Allāh Muḥammad b. Muḥammad Ibn al-Ḥājj al-ʿAbdarī al-Fāsī (d. 737 ah/1336 ce), al-Madkhal.55 8. ʿAlī b. Muḥammad al-Rabaʿī al-Lakhmī (d. 478 ah/1085 ce), al-Tabṣira. 9. ʿAlī b. Muḥammad al-Sayyid al-sharīf al-Jurjānī (d. 816 ah/1413 ce), al-Miṣbāḥ fī Sharḥ al-Miftāḥ.56 49 Le Tourneau 1934, 58. 50 GAL I 387; S I 609, II. 51 GAL I 249; S II 350, no. 3. 52 GAL II 157, no. 279; S II 196. 53 Commentary of the Mukhtaṣar by Khalīl b. Iṣḥāq al-Mālikī al-Miṣrī (d. 767 ah/1365 ce). GAL II 84, 1e and 316. Also see below, manuscript Fez, Qarawiyyin Library, 1060. 54 GAL II 84, 1a. 55 S II 95, no. 1-1. 56 Commentary of the Miftāḥ al-ʿulūm by Abū Yaʿqūb Yūsuf al-Sakkākī (d. 626 ah/1229 ce).

Collecting Books in Eighteenth-Century Morocco 10. Abū al-Ḥasan Muslim b. al-Ḥajjāj b. Muslim (d. 261 ah/875 ce), Ṣaḥīḥ Muslim.57 11. Jalāl al-dīn Muḥammad b. Aḥmad al-Maḥallī (d. 864 ah/1459 ce), al-Badr al-lāmiʿ fī ḥall Jamʿ al-jawāmiʿ.58 12. Abū al-Fatḥ Muḥammad b. Muḥammad Ibn Sayyid al-nās al-Yaʿmurī al-Ishbīlī (d. 734 ah/1333 ce), ʿUyūn al-athar fī funūn al-Maghāzī wa al-shamāʾil wa al-siyar. 13. Nūr al-dīn Abū al-Irshād ʿAlī b. Muḥammad al-Ujhūrī (d. 1066 ah/1655 ce), Sharḥ Mukhtaṣar Ṣaḥīḥ al-Bukhārī.59 14. Muḥammad b. Aḥmad Ibn Ghāzī al-Miknāsī (d. 919 ah/1513 ce), Itḥāf dhawī al-istiḥqāq bi baʿḍi murād al-Murādī wa Zawaʾid Abī Isḥāq.60 15. Yūsuf Ibn ʿUmar Abū al-Ḥajjāj al-Anfāsī al-Salāsī al-Fāsī (d. 761 ah/1359 ce), Sharḥ al-Risāla.61 16. Nūr al-dīn Abū al-irshād ʿAlī b. Muḥammad al-Ujhūrī (d. 1066 ah/1655 ce), Sharḥ al-Risāla.62 17. Khālid b. ʿAbd Allāh al-Azharī (d. 905 ah/1499 ce), Tamrīn al-Ṭullāb fī ṣināʿat al-iʿrāb: Iʿrāb Alfiyyat Ibn Mālik.63 18. Muḥammad b.Muḥammad Ibn ʿArafa al-Warghamī al-Tūnusī (d. 803 ah/1400 ce), al-Ḥudūd fī al-taʿrīfāt al-fiqhiyya.64 19. Shams al-dīn Muḥammad b. Abī Bakr al-Zurʿī Ibn Qayyim al-Jawziyya (d. 751 ah/1350 ce), Zād al-maʿād fī hady khayr al-ʿibād.65 20. Yāsīn b. Zayn al-dīn b. ʿUlaym al-Ḥimṣī al-ʿUlaymī (d. 1061 ah/1651 ce), Ḥāshiya ʿalā Mukhtaṣar sharḥ Talkhīṣ al-Miftāḥ.66 21. Badr al-dīn Maḥmūd b. Aḥmad al-ʿAynī (d. 855 ah/1451 ce), al-Maqāṣid al-naḥwiyya, Sharḥ shawāhid shurūḥ al-Alfiyya.67 22. Yāsīn b. Zayn al-dīn b. ʿUlaym al-Ḥimṣī al-ʿUlaymī (d. 1061 ah/1651 ce), Ḥāshiya ʿalā Alfiyyat Ibn Mālik.68 57 GAL I 160. 58 Commentary of the Jamʿ al-jawāmiʿ by Tāj al-dīn al-Subkī (d. 771 ah/1369 ce). GAL II 89, 1c; S II 105. 59 Commentary of the Mukhtaṣar Ṣaḥīḥ al-Bukhārī by Abū Bakr Muḥammad b. Aḥmad b. Abī Jamra al-Mursī (d. 599 ah/1203 ce). Also Hajji 1976, 30, 303–305, 369. 60 Commentary of the Alfiyya by Jamāl al-dīn Muḥammad b. ʿAbd Allāh b. Muḥammad b. ʿAbd Allāh Ibn Mālik al-Ṭā’ī al-Jayyānī (d. 672/1273), see GAL I 299; S II 337; Hajji (1976), passim. 61 Commentary of the Risāla by Ibn Abī Zayd al-Qayrawānī (d. 386 ah/996 ce). 62 As above, GAL I 178, 1k. 63 Commentary of the Alfiyya by Ibn Mālik, see GAL I 299; II 27, no. 5. 64 GAL I 384, 51; S II 347, no. 6, 3. 65 GAL II 106, no. 12; S II 126, no. 12. 66 Commentary of the Mukhtaṣar sharḥ Talkhīṣ al-Miftāḥ by Saʿd al-dīn al-Taftāzānī (d. 792 ah/1389 ce), see GAL I 295. 67 Commentary of Ibn Mālik’s Alfiyya, see GAL I 299, II.11 68 Commentary of Ibn Mālik’s Alfiyya.

François Déroche, Lbachir Tahali 23. Jamāl al-dīn ʿUṯmān b. ʿUmar Ibn al-Ḥājib (d. 646 ah/1248 ce), Jāmiʿ al-Ummahāt aw al-Mukhtaṣar al-farʿī.69 25. Abū Mahdī ʿIsā b. ʿAbd al-Raḥmān al-Saktānī al-Ragrāgī (d. 1062 ah/1651 ce), Ḥāshiya ʿalā Sharḥ Umm al-barāhīn: al-ʿAqīda al-ṣughrā.70 25 bis. ʿAbd al-Raḥmān b. Muḥammad al-ʿārif al-Qaṣrī al-Fāsī (d. 1036/1626), Ḥāshiya ʿalā Sharḥ Umm al-barāhīn: al-ʿAqīda al-ṣughrā.71 26. Abū Zayd ʿAbd al-Raḥmān b. ʿIffān al-Jazūlī (d. 741 ah/1340 ce), Sharḥ al-Risāla.72 27. The reference is too imprecise to identify properly the commentaries of the Qaṣīdat al-Burda by Muḥammad b. Saʿīd al-Būṣīrī (d. 696 ah/1296 ce) described in the deed.73 28. ʿAbd al-Ḥaqq b. Ghālib Ibn ʿAṭiyya al-Andalusī (d. 542 ah/1147 ce), al-Muḥarrar al-wajīz fī tafsīr al-kitāb al-ʿazīz.74 29. Saʿd al-dīn Masʿūd b. ʿUmar al-Taftāzānī (d. 792 ah/1390 ce), al-Sharḥ al-mukhtaṣar known as Talkhīṣ al-maʿānī.75 31. Abū al-ʿAbbās Aḥmad b. Muḥammad Ibn Ḥajar al-Haytamī (d. 973 ah/1565 ce), al-Ṣawāʿiq al-muḥriqa ʿalā ahli al-rafḍ wa al-ḍalāl wa al-zandaqa.76 32. ʿAlāʾ al-dīn ʿAlī b. al-Ḥazm al-Qurashī al-Dimashqī Ibn al-Nafīs (d. 687 ah/1288 ce), al-Mūjiz fī al-ṭibb.77 33. Muḥammad b. Muḥammad al-Fishtālī (d. 777 ah/1376 ce), al-Fāʾiq fī aḥkām al-wathāʾiq. 34. Muḥammad b. Aḥmad Mayyāra al-Fāsī (d. 1072 ah/1661 ce), Naṣīḥat al-mughtarrīn wa kifāyat al-muḍṭarrīn fī al-tafrīq bayna al-muslimīn bi mā lam yunzilhu rabb al-ʿālamīn wa lā jāʾa bihi al-rasūl al-amīn wa lā thabata ʿan al-khulafāʾ al-mahdiyyīn.78 35. Abū al-ʿAbbās Aḥmad b. Yaḥyā al-Wansharīsī (d. 914 ah/1508 ce), ʿUddat al-burūq fī jamʿ mā fī al-madhhab min al-jumūʿ wa al-furūq. 69 GAL I 306, VIII. 70 Super-commentary of the Sharḥ Umm al-barāhīn, commentary by the author, Muḥammad b. Yūsuf al-Sanūsī (d. 895/1489), of his ʿAqīda al-ṣughrā (GAL II 251, IIb). 71 As above. Hajji (1976), 165, 287, 308, etc. 72 We surmise that it is a commentary of Ibn Abī Zayd al-Qayrawānī’s Risāla, see nos 15-6 above. See GAL I 178. 73 Or al-Kawākib al-durriyya fī madḥ khayr al-bariyya. 74 GAL I 412, 9; S I 732, no. 9. 75 Commentary of the Talkhīṣ al-Miftāḥ by Jamāl al-dīn Muḥammad b. ʿAbd al-Raḥmān al-Qazwīnī (d. 739 ah/1338 ce). 76 GAL II 388, no. 4. 77 GAL I 493, 37; S I 900. 78 See Hajji 1976, 209, 210 and 333.

Collecting Books in Eighteenth-Century Morocco 37. Shihāb al-dīn Aḥmad b. Aḥmad al-Burnusī al-Fāsī Zarrūq (d. 899 ah/1493 ce), al-Naṣīḥa al-kāfiya li man khaṣṣahu Allāh bi al-ʿāfiya.79 38. Fakhr al-dīn Muḥammad b. ʿUmar al-Khaṭīb al-Rāzī (d. 606 ah/1209 ce), Muḥaṣṣal afkār al-mutaqaddimīn wa al-mutaʾakhkhirīn min al-ʿulamāʾ wa al-ḥukamāʾ wa-l-mutakallimīn.80 39. Muḥammad b. ʿAbd Allāh al-Lawwātī al-Ṭanjī Ibn Baṭṭūṭa (d. 779 ah/1377 ce), Tuḥfat al-nuẓẓār fī gharāʾib al-amṣār wa ʿajāʾib al-asfār.81 40. Muḥammad b. ʿAbd al-Salām Bannānī (d. 1163 ah/1749 ce), Taʾlīf fī mā lā yanṣarif min al-asmāʾ. 41. Qiṣṣat Ahl al-kahf: it might be an extract from the al-Bidāya wa al-nihāya by Abū al-Fidāʾ Ismāʿīl b. ʿUmar Ibn Kathīr (d. 774 ah/1373 ce).82 42. We surmise that it might be the Jamʿ al-jawāmiʿ fī uṣūl al-fiqh by Tāj al-dīn ʿAbd al-Wahhāb b. ʿAlī al-Subkī (d. 771 ah/1369 ce).83 43. ʿAbd al-Raḥmān b. Muḥammad al-Jādirī (d. after 818 ah/1415 ce), Rawḍat al-azhār fī ʿilm waqt al-layl wa al-nahār.84 44. ʿAḍud al-dīn ʿAbd al-Raḥmān b. Aḥmad al-Ījī (d. 756 ah/1355 ce), Jawāhir al-kalām.85 45. Fahrasa: it might be the Fahrasa of Muḥammad b. ʿAbd al-Salām Bannānī (d. 1163 ah/1749 ce). 46. Jamāl al-dīn Muḥammad b. ʿAbd Allāh Ibn Mālik (d. 672 ah/1273 ce), al-Tuḥfa ʿalā Kāfiyat Ibn al-Ḥājib.86 47. Khālid b. ʿAbd Allāh al-Azharī (d. 905 ah/1499 ce), Sharḥ al-Ajurrūmiyya fī uṣūl ʿilm al-ʿarabiyya.87 48. Ḥamdūn b. Muḥammad b. ʿAbd al-Salām Bannānī (d. 1170 ah/1756 ce), Ikhtiṣār Sharḥ khuṭbat al-Alfiyya.88 79 GAL II 253, no. 7; S II 361; Hajji 1976, 31, 280, 501, 562. 80 GAL I 507, no. 22; S I 923, no. 22. 81 GAL II 256. 82 GAL II 49, no. 1; S II 48. 83 GAL II 89, no. 1; S II 105, 1. 84 GAL II 169; S II 217. 85 GAL II 209, no. VI; S II 291, no. VI. 86 Commentary of the Kāfiya by Ibn al-Ḥājib (d. 646 ah/1249 ce). 87 Commentary of the Kāfiya by Ibn al-Ḥājib (d. 646 ah/1249 ce). Commentary of the Ajurrūmiyya by Abū ʿAbd Allāh Muḥammad b. Muḥammad b. Dā’ūd al-Ṣanhājī b. Ajurrūm (d. 723 ah/1323 ce), see GAL II 238, 6. 88 Abbreviated version of Ibn Mālik’s Alfiyya.

François Déroche, Lbachir Tahali 49. Muḥammad b. Abī Bakr Ibn al-Rashīd al-Baghdādī (d. 662 ah/1263 ce), al-Qaṣāʾid al-witriyya fī madḥ khayr al-bariyya.89 50. Muḥammad b. ʿAbd al-Salām Bannānī (d. 1163 ah/1749 ce), al-Riḥla al-ḥijāziyya. 51. ʿAbd al-Bāqī b. Yūsuf al-Zarqānī (d. 1099 ah/1687 ce), Ḥāshiya al-Zarqānī ʿalā Sharḥ al-Luqānī ʿalā khuṭbat Mukhtaṣar Khalīl.90 52. Abū al-ʿAbbās Aḥmad b. ʿAlī al-Manjūr al-Fāsī (d. 995 ah/1586 ce), Fahrasat al-Manjūr. 53. Jalāl al-dīn ʿAbd al-Raḥmān b. Abī Bakr al-Suyūṭī (d. 911 ah/1505 ce), al-Nuqāya fī arbaʿata ʿashara ʿilman.91 55. Jamāl al-dīn Muḥammad b. ʿAbd Allāh Ibn Mālik (d. 672 ah/1273 ce), Tashīl al-fawāʾid wa takmīl al-maqāṣid.92 56. Muḥammad b. Abī al-faḍl b. Qāsim al-Raṣṣāʿ al-Tilimsānī (d. 894 ah/1488 ce), al-Hidāya al-Kāfiya al-shāfiya li-bayān ḥaqāʾiq Ibn ʿArafa al-wāfiya.93 56 bis. Muḥammad b. ʿĪsā al-Tirmidhī (d. 279 ah/892 ce), al-Shamāʾil al-Muḥammadiyya wa al-khaṣāʾil al-muṣṭafawiyya.94 57. Muḥammad b. ʿAlī al-Burjī al-Ḥājj al-Shuṭaybī (d. 963 ah/1555 ce), Kitāb al-Jumān fī mukhtaṣar akhbār al-zamān.95 58. Muḥammad b. ʿAbd al-Qādir al-Fāsī (d. 1116 ah/1704 ce), Takmīl al-marām bi Sharḥ shawāhid Ibn Hishām.96 59. ʿAbd Allāh b. Yaḥyā al-Tujībī Ibn al-Waḥshī (d. 502 ah/1108 ce), Sharḥ Musnad al-Shihāb fī al-Ḥadīth by Muḥammad b. Salāma al-Quḍāʿī (d. 454 ah/1062 ce).97 89 GAL I 250; S I 443. 90 Super-commentary by al-Luqānī (d. 958 ah/1551 ce), see GAL II 84, 1h. 91 GAL II 156, no. 268. 92 GAL I 298, I. 93 S II 737, no. 1, 7. 94 GAL I 162, no. 2; S I 268. 95 Hajji 1976, 146, 185, 189, 518, 533, 534-6, 614. 96 Commentary of the verses used as examples by Ibn Ḥishām al-Anṣārī (d. 761 ah/1360 ce) in his Awḍaḥ al-masālik, a commentary of the Alfiyya. 97 Commentary of the Musnad al-Shihāb fī al-Ḥadīth by Muḥammad b. Salāma al-Quḍāʿī (d. 454 ah/1062 ce).

Collecting Books in Eighteenth-Century Morocco Here again, the total number of titles is slightly higher than 59: they are 61 since, twice, two titles are indicated in the same entry (nos 25 and 56), which explains the presence in the above list of nos 25 bis and 56 bis. No. 18 exhibited a similar situation, but as the text is damaged in this place it is not possible to read the title (or the author’s name) of the second work. A glance at the list suggests that it does not adhere to a classification principle decided beforehand. The books were obviously not put into order according to their subject matter. On the other hand, a material principle seems to have been followed loosely: the person or persons in charge of the operation placed the works in two volumes (nos 1 to 4 and then no. 12) at the beginning, then a little further the isolated volumes from series of two or more volumes (nos 7 to 10, then 15 and 28). Multiple texts manuscripts with more than two texts are found between nos 45 and 52, with the exception of no. 50; the wording used is similar and one of these manuscripts has survived, thus confirming their typology.98 The volumes containing only two texts were seen in a different way and are scattered throughout the inventory: nos 18, 25 and 56 are in this case. No. 26 is slightly problematic: the wording used is unique in the inventory and does not allow to decide whether this manuscript contained only two texts, which would be the most likely because of its place, or if it was a miscellany. Sixteen items contain from two to more texts, and are either multiple-text manuscripts (MTM) or composite volumes. In other words, they represent together a good quarter of the books that were made waqf. As an indication, a similar proportion of manuscripts of the same type is found among the poetry manuscripts in Mūlāy Zaydān’s library.99 It is difficult to detect the intentions behind the three sisters’ move: why did they decide to favor their brother ʿAbd al-Karīm rather than their own children, in fact relinquishing to him a portion of the inheritance? The situation of the lineage at this time can explain it: ʿAbd al-Karīm was then the only son of Muḥammad b. ʿAbd al-Salām still alive and a form of family solidarity could have played. However, it is possible to go a step further on the basis of the data we have, the waqf deed itself, but also other documents. 98 See below, MS Rabat, BNRM, 16 K. 99 See de Castilla in this volume.

François Déroche, Lbachir Tahali 4 Collecting books For a scholar, a common way of getting a text for personal use was to make a copy. The Bannānī have not failed in this tradition and sources report that ʿAbd al-Salām b. Ḥamdūn had transcribed Bukhārī’s Ṣaḥīḥ in eight volumes and then made them a waqf in favour of his son Muḥammad.100 The latter had then carefully collated this manuscript with the copy which was then kept at the al-Qarawiyyin mosque, that of Ibn Saʿada al-Andalusī (d. 522 ah/1128 ce), even writing a note on the latter to commemorate his work.101 The grandson of ʿAbd al-Salām b. Ḥamdūn, ʿAbd al-Karīm, reports that his grandfather had also executed a copy of al-Tirmidhī’s Shamāʾil al-Muḥammadiyya, on which his father had subsequently added his own glosses;102 it is highly probable that it is no. 56bis on the list accompanying the waqf deed. This example highlights the importance of notes and glosses in the hand of a scholar, written in the margins of a book that one has either been copying himself or bought on the market: they increased significantly the value of the copy. In addition to the transcription of texts as a means of increasing a personal collection, the composition of works the autograph of which remains in the possession of the author within his own library can also contribute to its growth. In addition to the sources already mentioned, we have a very remarkable factual testimony of this copy activity. A multiple-text manuscript currently kept at the National Library of the Kingdom of Morocco, manuscript 16 K, contains two marks of possession of Muḥammad b. Muḥammad b. ʿAbd al-Salām Bannānī.103 It includes at the beginning a table of contents which is similar in its presentation to the inventory accompanying the waqf of the three sisters – each title introduced by a number in Fāsī numerals. Thirteen titles are indicated there, but the manuscript contains seventeen texts in its current state. However, two of the works mentioned in the table of contents are no longer part of the manuscript -a point we shall discuss later. The name of the former owner, Muḥammad b. Muḥammad b. ʿAbd al-Salām Bannānī, also appears in the colophon of three texts: he transcribed the Taʾlīf fī mā lā yanṣarif (no. 1) and the Sharḥ khuṭbat Mūkhtaṣar al-Saʿd al-Taftāzānī li Talkhīṣ al-Miftāḥ (no. 2),104 both composed by his father, 100 ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 6r. 101 ʿAbd al-Ḥayy al-Kattānī, al-Tanwīh, ed. al-Majīd Khiyālī 2012, 61. 102 ʿAbd al-Ḥayy al-Kattānī, al-Tanwīh, ed. al-Majīd Khiyālī 2012, 3. 103 Manuscript Rabat, BNRM, 16 K, unnumbered folio at the beginning of the volume and unnumbered folio before p. 426. 104 According to his own statement, he collected in these folios the notes he wrote in the margins of his own copy of Taftāzānī’s Mukhtaṣar – another example of the importance of such notes (see above).

Collecting Books in Eighteenth-Century Morocco and the Manāsik ʿAlī Baraka al-Tiṭwānī by ʿAlī b. Muḥammad Baraka al-Tiṭwānī (d. 1120 ah/1708 ce, no. 16). The copying of the texts was respectively completed on 4 and 12 Rabīʿ I 1182 ah/18 and 26 July 1768 ce (nos 1 and 2), and on Saturday 28 Shawwāl 1151 ah/7 February 1739 ce (no. 16). The interest of this volume lies not only in the fact that it was copied by Muḥammad b. Muḥammad b. ʿAbd al-Salām Bannānī, but also because it is one of the very few volumes that match an entry of the inventory of the three sisters’ contribution to the family library. No. 40 is actually described as a multiple-text manuscript that began with the Taʾlīf fī mā lā yanṣarif: this is the case of manuscript 16 K, which also confirms that manuscripts of this type were described in this way in the inventory. But it is still more important for us because it contains a series of texts related to the father of the copyist, Muḥammad b. ʿAbd al-Salām Bannānī. The latter probably transcribed himself various texts included in the manuscript: he could have completed the beginning of the Intiqāʾ Miʾat Ḥadīth fī al-targhīb fī iṣṭināʿ al-maʿrūf by Muḥammad b. Muḥammad b. Sulaymān al-Rūdānī (d. 1094 ah/1683 ce; no. 10), the end of which in an Oriental hand is presumably what remains of an original but mutilated copy. He may also have copied both the Taʾlīf al-imām Ibn ʿAbd al-Barr fī al-Basmala hal hiya āya min al-Fātiḥa am lā by Yūsuf b. ʿAbd Allāh Ibn ʿAbd al-Barr al-Numayrī (d. 463 ah/1071 ce; no. 14) and the Izālat al-labs ʿan al-masāʾil al-khams by Abū al-ʿAbbās Aḥmad b. Mubārak al-Sijilmāssī (d. 1156 ah/1743 ce; no. 15). He is the author of various texts related to the transmission of knowledge, first of all his own Fahrasa (no. 3),105 then an incomplete Fahris asānīd where he mentions the isnād he could claim in the field of ḥadīth (no. 8),106 as well as two ijāza given one to a Muḥammad b. Muḥammad b. Ḥajjāj al-Mālikī, residing in Egypt, during their meeting in Rosetta on 15 Jumādā I 1143 ah/25 November 1730 ce (no. 4),107 the other to a person whose name has disappeared (no. 5). Texts nos 7, 9 and 11 are short treatises composed by Muḥammad b. ʿAbd al-Salām Bannānī: the Sharḥ al-Ḥizb al-Kabīr li Abī al-Ḥasan al-Shādhilī,108 the Sharḥ al-Ṣalāt al-Mashīshiyya109 and the Risāla fī faḍl al-niṣf min Shaʿbān.110 We 105 It is a short version akin to the Fahrasa ṣuġrā (no. 28 in the list of his works), but with a final addition – an ijāza given to his master Ibrāhīm b. Ḥasan al-Kūrānī al-Shahrazūrī (d. 1101 ah/1690 ce). 106 The text was prepared in Mekka during the author’s pilgrimage in 1141 ah/1729 ce. 107 It concerns mainly Bukhārī’s Thulāthiyyāt and the Dalāʾil al-khayrāt by al-Jazūlī (d. 870 ah/1465 ce). 108 Manuscript Fez, Qarawiyyin Library, 1512 (no. 11 in the list of the author’s works). 109 Manuscript Fez, Qarawiyyin Library, 1512 (no. 20 in the list of the author’s works). 110 No. 8 in the list of the author’s works. The copy is partly in Maghribi script, partly in eastern script.

François Déroche, Lbachir Tahali also owe to him the notes on various subjects which constitute text no. 6 in the collection. As stated above, two texts which had been included in the table of contents (probably by Muḥammad b. Muḥammad b. ʿAbd al-Salām Bannānī) are no longer found in manuscript 16 K: the first one, no. 8, was another treatise by Muḥammad b. ʿAbd al-Salām, the Sharḥ Qaṣīdat sayyidī ʿAbd al-Raḥmān al-Fāsī fī ʿilm al-usṭurlāb which seems to be an abridged version of text no. 9 in the list given above, and the second, originally no. 9 in the manuscript’s table of contents, was the Taʿẓīm al-aimma al-arbaʿa by the owner’s grandfather, ʿAbd al-Salām b. Ḥamdūn Bannānī. The three texts that have not been described so far are a collection of excerpts (Majmūʿ fīhi aḥādīth fī makārim al-akhlāq, Uṣūl al-Ṭarīqa al-Nāṣiriyya and Mudāwāt ʿuyūb al-nafs; no. 12), an Eastern copy of the Ḥāshiya Kamāl Bāshā ʿalā al-Bayḍāwī by Aḥmad b. Sulaymān b. Kamāl Bāshā (940 ah/1534 ce; no. 13) and the Rawāʾiq al-ḥulal fī alqāb al-ziḥaf wa al-ʿilal of Muḥammad b. Ubb b. Aḥmīd al-Muzzammarī al-Tuwātī (1160 ah/1747 ce; no. 17); the last two could be a later addition. Manuscript 16 K provides us with a valuable insight into the double contribution that family members could make to the family library, as authors and as copyists. A scholar wishing to enrich his own library could also buy books on the market. The commentary by N. Clénard, cited above, highlights the difficulty encountered by potential buyers to acquire a manuscript. Five initial folios found by one of us (FD) next to the three sisters’ waqf deed in the Qarawiyyin library bear witness to this in the case of the Bannānī library. The notes recording the purchase of the manuscripts are frequently inscribed on the upper left corner of the folios: a first one concerns a copy of the Miʿyār al-muʿrib wa al-jāmiʿ al-mughrib ʿan fatāwā ahl Ifrīqiya wa al-Andalus wa al-Maghrib by Abū al-ʿAbbās Aḥmad b. Yaḥyā al-Wansharīsī (d. 914 ah/1508 ce).111 A second note records the acquisition of the Mudkhal ilā tanmiyat al-amāl bi taḥsīn al-niyyāt by Abū ʿAbd Allāh Muḥammad b. Muḥammad al-ʿAbdarī al-Fāsī, known as Ibn al-Ḥājj (d. 737/1336), on Rajab 1st, 1142 ah/19 January 1730 ce.112 Another example is found on a copy of the famous grammar treatise, the Mughnī al-labīb ʿan kutub al-aʿārīb by ʿAbd Allāh b. Yūsuf b. ʿAbd Allāh Ibn Hishām (Fig. 2), that Muḥammad b. ʿAbd al-Salām bought from his teacher, Muḥammad al-Irāqī (d. 1142 ah/1729–1730 ce).113 A copy of the Iḥkām al-dalāla ʿalā taḥrīr al-Risāla bears a list of three persons who 111 Fez, Qarawiyyin Library, Fgmt 6, see the equivalence in the bibliography (copy in Maghribi script, 28 × 20 cm). GAL II 248; S II 348. 112 Fez, Qarawiyyin Library, Fgmt 4 (copy in Maghribi script, 28,5 × 21,5 cm). GAL S I 95. 113 Fez, Qarawiyyin Library, Fgmt 7 (copy in Maghribi script, 24 × 16,5 cm). GAL II 23; S I 17.

Collecting Books in Eighteenth-Century Morocco owned the book before it became Muḥammad b. ʿAbd al-Salām’s property.114 The last example deals with the purchase of a copy of the Qawāʿid al-aḥkām fī iṣlāḥ al-anām by ʿIzz al-dīn ʿAbd al-ʿAzīz Ibn ʿAbd al-Salām Sulamī (d. 660 ah/1261 ce); it was concluded in Mekka, perhaps in 1141 ah/1729 ce.115 The latter recalls the importance of the travels to the East for Moroccan scholars looking for manuscripts.116 The waqf was a third way to enrich a library; it also had the advantage of offering a solution to stabilize it in a sustainable way (so was it hoped). The Bannānī, experts in the Law, resorted to it in two ways. The first, illustrated by the three sisters’ waqf, was implemented by Muḥammad b. ʿAbd al-Salām’s father in favor of his own son, as we have seen, for the eight volumes of Bukhārī’s Ṣaḥīḥ. With the same legal tool, Muḥammad b. ʿAbd al-Salām and later his children expanded the family library. A deed is found on eleven fragments in the Qarawiyyin Library, for instance on a multiple-text manuscript:117 Muḥammad b. ʿAbd al-Salām names as beneficiaries in the first place his own children, then the children of the latter; if their lineage were to die out, the children of his two brothers would become the beneficiaries (see Appendix II),118 then, failing these, his cousins, who would be devoting themselves to study (Fig. 3).119 As a last resort, if all disappeared, the waqf would return to his zāwiya – which has already been mentioned above. This deed was written in 1164 ah/1750–1751 ce, in the year following the death of the settler. A son of the deceased was therefore responsible for registering the will 114 Fez, Qarawiyyin Library, Fgmt 1 (copy in eastern script, 28,5 × 17,5 cm). 115 Fez, Qarawiyyin Library, Fgmt 8 (copy in eastern script, 25 × 17 cm). GAL I 430-1; S I 766. 116 M. al-Manūnī underlines Muḥammad b. ʿAbd al-Salām Bannānī’s activity in this respect (alManūnī 2005, 32). The example of the Saadian library goes in the same direction (see de Castilla in this volume). 117 Fez, Qarawiyyin Library, Fgmt 9. The manuscript contains two treatises on logic, the Sharḥ al-Jumal fī al-manṭiq li al-Khūnajī by Muḥammad b. Aḥmad Ibn Marzūq al-ḥafīd al-ʿAjīsī al-Tilimsānī (d. 842 /1439) and the Sharḥ al-Shamsiyya by al-Rāzī al-Taḥtānī, Quṭb al-dīn Muḥammad b. Muḥammad al-Rāzī (d. 766/1364). 118 One of the brothers is Muḥammad al-ʿArabī b. ʿAbd al-Salām (see Ibn al-Ṭayyib al-Sharqī, al-Riḥla al-Ḥijāziyya, ed. al-dīn Shūbad and Būtwādī 2018, II, 464 and 483) who could be the father of ʿAmm al-ʿArabī (see above). 119 We know three cousins, all sons of al-Ḥasan Bannānī: Maḥammad (sic!), ʿAlī and Muḥammad (d. 1194 ah/1780 ce). The latter is the most prominent of these three brothers: a jurisconsult, he was imam of Mawlāy Idrīs mosque where he taught Māliki law and tradition (see al-Shafshāwnī, Thamrat unsī, ed. al-Ḥaqq al-Ḥīmar 1996, 84; al-Qādirī, Nashr al-mathānī, ed. Ḥajjī and al-Tawfīq 1982–1986, IV, 214; al-Manālī al-Zabādī, Sulūk al-ṭarīq, ed. ʿal-Ḥayy al-Yamlāḥī 2012, 255).

François Déroche, Lbachir Tahali of his father in the presence of two witnesses as is clearly stated in one of those deeds: ‘this book in its entirety has been made a waqf by our father’.120 On these eleven manuscripts, the deed is repeated with small variants. On Fragment 7, the beginning of a copy of the Mughnī al-labīb ʿan kutub al-aʿārīb by ʿAbd Allāh b. Yūsuf b. ʿAbd Allāh b. Hishām (d. 761 ah/1360 ce), one reads for instance the following text (Fig. 4): َّ ‫ يُعْلَ ُم ويُ َت َح َّقق‬.‫الح ْم ُد هلل‬ ،‫الس ْفر من ُم ْغ ِني اب ِْن هِ شام والَّذي َب ْعدَه‬ ِّ ‫أن هَ ذا‬ َ .1 َ ُ ُ َّ ُ َّ َ َ ُ ْ ‫اإل َما َم ال َعاَّلمة القد َوة ال َب َركة أبُو َعبْد هللاِ َسيِّدِي م َُحمَّد‬ ِ ‫ ِممَّا َحب ََّسه الشيخ‬.٢ ‫اسلُوا‬ َّ ‫ بن ال َع َّاَّل َم ِة َسيِّدي َعبْد‬.٣ َ ‫السالم َبنَّانِي َعلَى أَ ْو َاَلدِه وأَ ْوال ِد أَ ْو َاَلدِه َما َت َن‬ َ َ ْ ُ ُ ُ ‫او َي ِت ِه‬ ِ ‫ ث َّم َعلَى َز‬،‫ ث َّم َعلَى َطلَ َب ِة العِل ِم م ِْن َبنِي َعمِّه َب ْعدَهم‬،‫ َوا ْم َت َّدت ُفرُ وعُ هُم ث َّم َعلَى أ ْوال ِد أ َخ َو ْي ِه َك َذل َِك‬.٤ َ ‫الص‬ َ ‫ َم َح ِّل‬.٥ ‫ َو َم ْن َب َّد َل أَ ْو َغي ََّر‬.‫وس ِة َفاس ال َق َر ِويِّين‬ َّ ‫ض ِريحِ ه ِبدَرْ ِب ال َق َّطان م ِْن َح ْو َم ِة‬ َ ُ‫اغ ِة م ِْن َمحْ ر‬ ْ َ ُ‫ال ْن ِت َقام ِم ْن ُه؛ َحبْساً ُم َؤبَّداً وو ْقفاً م َُخلَّداً إلى أَ ْن َير َث هللا‬ ِّ َ ُ َ‫اأَلرض‬ َ َ َ ِ ‫ َفاهللُ َحسِ يبُه و ُمت َولي ا‬.٦ ِ ِ َ َ َّ َ َ َ َ َ َ َ ْ َّ ‫ ف َم ْن َعلِ َم ذلك َو َحقق ُه قيَّ َد ِب ِه ش َها َد َت َه أ َوا َئل َر َبيع الن َب ِوي األن َور‬.‫ارثِين‬ ِ ‫ َو َم ْن َعلَ ْي َها َوه َو َخيْرُ ال َو‬.٧ َ ْ .‫ َعا َم أَرْ َب َع ٍة َوسِ تِّينَ َومِا َئ ٍة َوألف‬.٨ 1. Praise be to God! It should be known and approved that this volume containing the Mughnī of Ibn Hishām and what follows 2. is constituted in waqf by the erudite imam, model and blessed, Abū ʿAbd Allāh Sīdī Muḥammad, 3. son of the scholar Sīdī ʿAbd al-Salām Bannānī, in favour of his children and the children of his children, as long as they reproduce themselves 4. and their [genealogical] branches spread. Then, likewise, in favour of the children of his two brothers and the students among his cousins after them, then of the zāwiya 5. which contains his mausoleum in the alleyway of al-Qaṭṭān in the jewellers’ district of Fās al-Qarawiyyin the protected. Whoever destroys or transforms [it], 6. God will hold him to account and take revenge on him; a perpetual legacy until God ‘inherits the earth 7. and those who are there’ (Q. 19: 40) because he is ‘the best of those who give inheritance’ (Q. 21: 89). He acquainted himself with this and approved it by recording his testimony at the beginning of Rabīʿ I 8. in the year 1164. [Signatures of the witnesses] On the copy of the Qawāʿid al-aḥkām by Ibn ʿAbd al-Salām (d. 660 ah/1261 ce), the children of his two brothers are not mentioned and the beneficiaries are suc- 120 See above, n. 114.

Collecting Books in Eighteenth-Century Morocco cessively his own children, their descendants and, in case of extinction, those of their cousins, who dedicate themselves to science; in the last resort, the book would come back to Muḥammad b. ʿAbd al-Salām’s zāwiya.121 A different formulation, similar to that of the second document, appears on the manuscript with al-Wansharīsī’s Nawāzil, that is to say his Miʿyār.122 The beneficiaries are designated in more general terms: the settler speaks of descendants (dhurriyya) and posterity (ʿaqib). A new category, those of his relatives (aqāribuhu) who dedicate themselves to science, is mentioned: it appears before the cousins and, as in the previous cases, the zāwiya. After Muḥammad b. ʿAbd al-Salām, his children followed the same path. On the initial folio of the already mentioned copy of the Madkhal by Ibn al-Ḥājj,123 a deed has been written, stating that the daughters of Muḥammad b. ʿAbd al-Salām Bannānī, in other words the three sisters, constituted the book as a waqf in 1164 ah/1750–1751 ce; they did so in favour of those who dedicate themselves to the study among their relatives (aqāribuhunna) and among the children of their paternal uncle, or finally in favour of the zāwiya of Darb al-Qaṭṭān. As indicated on the first line of the deed, the book comes from their share in their father’s inheritance. It must the first half of the work, the second one joining it afterwards: it is actually the no. 7 of the waqf in favour of their brother ʿAbd al-Karīm.124 It thus appears that the inheritance of the three sisters was divided into two parts: the first one came directly from their father and included the first half of the Madkhal; it was partly turned into a waqf one year after the death of Muḥammad b. ʿAbd al-Salām. According to the deed in favour of their brother ʿAbd al-Karīm, there were still books that had come to them in this way and to which had been added those which they had inherited from their two brothers Ḥamdūn and Muḥammad. Among the 59 works included in the waqf published above was the second part of the Madkhal. Another modality of waqf has also been used in this strategy of book accumulation. In this case, donors foreign to the family constituted books as a waqf in favour of either the father, Muḥammad b. ʿAbd al-Salām Bannānī, or of his sons. One of us (LT) found two examples of this kind in the Qarawiyyin library collection. Both involve a certain Abū al-ʿAbbās Aḥmad b. Abī al-Ḥasan ʿAlī al-Maghribī al-Fāsī, known as al-Sharāybī al-Wuraydī al-Khawāja, a merchant born in Fez 121 See above and n. 115. 122 See above and n. 111. 123 See n. 112. 124 The beneficiaries are not the same and it would have been legally difficult to have the same book presented as a waqf for two different beneficiaries.

François Déroche, Lbachir Tahali whom Muḥammad b. ʿAbd al-Salām Bannānī met during a trip to the Near East. The first work is a fifteen-volume copy of a commentary on Khalīl’s Mukhtaṣar, the Mawāhib al-jalīla al-mustanbaṭa min kalām kull jalīl li yatayassara bihā ḥall alfāz Mukhtaṣar Khalīl (known as al-Sharḥ al-kabīr) by Muḥammad b. ʿAbd Allāh al-Kharshī (d. 1101 ah/1690 ce).125 The deed of waqf, written by Muḥammad b. ʿAbd al-Salām Bannānī, is dated to the middle of Jumādā II 1143 ah/end of December 1730 ce; it stipulates that Muḥammad and his son Ḥamdūn are the beneficiaries for the duration of their existence, after which the work will pass on to the Qarawiyyin mosque library – which seems to have actually occurred on 2 Rabīʿ II 1201 ah/21 January 1787 ce.126 The second manuscript is a six-volume autograph copy of the Fatḥ al-qarīb al-Mujīb ʿalā Sharḥ al-Targhīb wa al-tarhīb by ʿAbdallāh b. Bahāʾ al-dīn al-Shinshawrī (d. 999 ah/1590 ce).127 The same al-Wuraydī states that he makes the manuscripts a waqf in favour of Muḥammad b. ʿAbd al-Salām Bannānī and his son Ḥamdūn, then of all his descendants (fig. 5); ‘And if they disappear, the book will go to the illustrious library in front of al-Qarawiyyin Mosque’. The deed was written in an Eastern script. The date is partly illegible: we distinguish the month, Jumādā II, and the number of the units: three, which suggests a reading: Jumādā II 1143 ah/December 1730 ce, the same date as that of the other deed. We know that, at this moment, Muḥammad b. ʿAbd al-Salām Bannānī was in Egypt. The two preceding deeds call for some remarks. The list of the beneficiaries is more limited than in the previous examples, especially with regard to the commentary of Khalīl’s Mukhtaṣar for which only Muḥammad b. ʿAbd al-Salām Bannānī and his son are concerned. It should be noted that the eventual ultimate return is to the Qarawiyyin – perhaps because the construction of the zāwiya was not yet started at that time. The various documents found so far provide a few hints about the contents of a Moroccan scholar’s library in the first half of the eighteenth century. A study of the actual contents of the books can only be partial because we only know the title of the first text of the multiple-text manuscripts, except, apparently, when they contained only two, a situation observed in nos 25 and 56, and probably also in no. 18, judging by the parallel wording of no. 56 description.128 With this 125 Manuscript Fez, Qarawiyyin Library, 1060. GAL S II, 84. 126 al-Ribāṭī, Muqaddimat al-fatḥ, ed. Najīb 2008, 123. The author states that each volume bears a statement by two eyewitnesses, Muḥammad b. ʿAbd al-Majīd al-Fāsī and Aḥmad b. Muḥammad b. Sulaymān al-Gharnāṭī (see ʿAbd al-Kabīr al-Kattānī 2002, I, 152). 127 Manuscript Fez, Qarawiyyin Library, 1013, fol. 370. GAL S II, 442. 128 The first title is obviously not a basis for a thematic evaluation of the contents of a multiple text manuscript.

Collecting Books in Eighteenth-Century Morocco proviso, we observe that among the 83 texts identified so far, with the exception of Muḥammad b. ʿAbd al-Salām Bannānī’s own compositions, there is an almost equal distribution of the original compositions on the one hand and the commentaries and super-commentaries on the other.129 5 The Bannānī library When examining more closely the constitution of the library, or more exactly of the books that are known to us, one notices a relatively high number of recent works: nineteen of them date from the seventeenth and eighteenth centuries, sixteen of them being composed by Moroccan authors.130 On this last point, the trend is confirmed at a more general level: among the 83 texts identified so far, 28 (a third of the total) are authored by Moroccans.131 If we take into account together the names of the authors originating from North Africa and al-Andalus, it turns out that the Muslim West is well represented: its share amounts to a half (38 titles, against 39 for the rest of the Muslim world). This leaves aside the question of the place of copy since only a few manuscripts can be analysed on this point.132 Among the twelve fragments, those written in a Maghribi hand are six; and the two manuscripts of the Qarawiyyin as well as manuscript 16 K of the National Library are also in Maghribi script. In terms of contents and on the basis of the texts that could be identified, the books can be divided into three major groups of roughly equivalent size: the largest one comprises the books with religious content, 23 in total,133 followed by those of law (20),134 then grammar and philology (17).135 In the first category, we can note a comparatively high number of works (6) on the sīra: they may reflect Muḥammad b. ʿAbd al-Salām’s interest for this field, since his major work is a commentary of al-Kalāʿī’s K. al-Iktifāʾ on the expeditions of the Prophet and the first three caliphs. Among the remaining texts, two smaller groups can be suggested. The scientific component is reduced to two texts, one on medicine (no. 32) and the other 129 Nos 24 and 36 are included into this group, even if the full title is not known. 130 Nos 25, 25 bis, 34, 40, 45, 48, 50 and 58; Fgmt 5 and 16 K1, 3, 7, 8, 9, 11, 17. 131 Nos 14, 15, 26, 33, 35, 37, 39, 43, 44, 52 and 57, as well as Fgmt 6 in addition to the texts listed in n. 130. 132 Texts 7, 40 and perhaps also 56 bis of the list above. See below. 133 Nos 1, 2, 10, 12, 13, 19, 25, 25 bis, 27, 28, 31, 34, 38, 44, 49, 56 bis and 59. We did not include nos 7 and 37 (Sufism) or no. 41 (unidentified Qiṣṣat Ahl al-kahf). 134 Nos 4, 5, 8, 11, 15, 16, 18, 23, 26, 33, 35, 42, 51 and 56. 135 Nos 9, 14, 17, 20, 21, 22, 29, 40, 46 à 48, 55 and 58.

Collecting Books in Eighteenth-Century Morocco on astronomy (no. 43): it should be noted that the latter is included in a multiple texts manuscript without being apparently the first work: however, it has benefited from an exception to the rule prevailing for other manuscripts of this type of which only the first text is indicated.136 History and geography are better represented with four volumes137 to which could be added the two fahrasa138 and the bibliography by Suyūṭī, al-Nuqāya fī arbaʿata ʿāshara ʿilman.139 Poetry, with the exception of religious poetry (nos 27, with commentaries, and 49), belles lettres and philosophy are absent, but the texts related to these fields may have been included in multiple texts manuscripts of which only the first text was identified in the inventory. It cannot be ruled out either that works of these genres were among the books specifically inherited by their brother ʿAbd al-Karīm. A manuscript with two texts on logic, one by Quṭb al-dīn al-Rāzī al-Taḥtānī and the other by Ibn Marzūq al-ḥafīd al-Tilimsānī, was also part of this library.140 Regarding religious poetry, it may be noted in passing that the multiple text manuscript no. 49 begins with a composition of this kind, al-Baghdādī’s Qaṣāʾid al-witriyya fī madḥ khayr al-bariyya, a situation parallel to that observed in the collection of the Escorial (manuscripts RBME 270, 304, 361, 362, 363, 384, 399, 404, 436, 442, 470 and 487).141 There is little evidence available about the readers’ interest in Morocco during the period. It is however possible to make a parallel with the contents of a princely manuscript kept in the Escorial. According to a note found on fol. 385a, RBME 248 was prepared for the library of Abū ʿAbd Allāh Muḥammad b. ʿAbd al-Qādir b. Muḥammad al-Sharīf al-Ḥasanī and completed on 23 Rabīʿ II 969 ah/31 December 1561 ce, according to the colophon on fol. 281a.142 The date and the name of this Abū ʿAbd Allāh who had these miscellany prepared for himself invite us to identify him with a nephew of the Saadian Sultan ʿAbd Allāh al-Ghālib who had him killed in 975/1567.143 The list of the manuscript contents is as follows. 136 No. 52 may also be an exception. No. 27 is a different case as it contains various commentaries on the Burda. 137 Nos 3, 39, 50 and 57. The two geographical works are travel accounts, one being the famous work by Ibn Baṭṭūṭa (no. 39), the other one being by Muḥammad b. ʿAbd al-Salām Bannānī (no. 50). 138 Nos 45 and 52. 139 No. 53. 140 See n. 117. 141 See de Castilla in this volume. GAL S I, 443. 142 Déroche 2019. 143 See Fagnan 1924, 388 (= ‘Anonyme sur la dynastie saʿdienne’); Le Tourneau 1977, 26. Another date for the death of this Saadian prince is given there (Le Tourneau 1977, 25, n. 71).

François Déroche, Lbachir Tahali Contents of the Escorial manuscript, RBME 248 1. Abū ʿAbd Allāh Muḥammad b. Saʿīd al-Ṣanhājī al-Būṣīrī (d. 694 ah/1294 ce), al-Burda.144 2. Ṣafī al-dīn ʿAbd al-ʿAzīz b. Sarāyā al-Ḥillī (d. 750 ah/1349 ce), Kāfiyya al-badīʿiyya.145 3. Abū ʿAbd Allāh Muḥammad b. Yūsuf al-ḥasanī al-Sanūsī (d. 892 ah/1486 ce), ʿAqīda ahl al-tawḥīd also called al-ʿAqīda al-kubrā.146 4. al-Sanūsī, Umm al-barāhin also known as al-ʿAqīda al-ṣugrā.147 5. al-Sanūsī, treatises on the same topics.148 6. al-Sanūsī, al-Isāghugī.149 7. al-Sanūsī, theological treatises.150 8. Abū Muḥammad ʿAbd Allāh b. Abī Zayd al-Qayrawānī (d. 386 ah/996 ce), K. al-risāla.151 9. Khalīl b. Iṣḥāq b. Mūsā al-Mālikī al-Miṣrī (d. 767 ah/1365 ce), al-Mukhtaṣar.152 10. Jamāl al-dīn Muḥammad b. ʿAbdallāh b. Muḥammad b. ʿAbdallāh Ibn Mālik al-Ṭāʾī al-Jayyānī (d. 672 ah/1273 ce), al-Alfiyya (part).153 11. Abū ʿAbd Allāh Muḥammad b. Muḥammad b. Dāwūd al-Ṣanhājī b. Ajurrūm (d. 723 ah/1323 ce), Ajurrūmiyya.154 12. Ibn Mālik, Lāmiyya al-afʿāl (or al-Miftāḥ fi abniyat al-afʿāl).155 13. Abū al-ʿAbbās Aḥmad b. Abī ʿAbd Allāh Muḥammad b. ʿUthmān al-Azdī al-Maghribī, known as Ibn al-Bannā (d. 721 ah/1321 ce), Talkhīṣ aʿmāl al-ḥisāb.156 14. Abū Isḥāq Ibrāhīm b. Abī Bakr al-Tilimsānī, urjuza about rhetoric and style. 144 See GAL I, 264; S I, 467. 145 See GAL II, 160; S II, 199. 146 See GAL II, 250 (I). 147 GAL II, 250 (II). 148 This text could not be identified in the list of al-Sanūsī’s works published by C. Brockelmann (S II, 356). 149 See GAL II, 252 (XVII); also GAL I, 46521. 150 This text could not be identified in the list of al-Sanūsī’s works published by C. Brockelmann (S II, 356). 151 See GAL I, 177; S I, 301. 152 GAL II, 84; S II, 96. 153 GAL I, 298/4-ii; S I, 522. 154 See GAL II, 237; S II, 332. 155 See GAL I, 300; S I, 256. 156 GAL I, 255/1; S I, 363.

Collecting Books in Eighteenth-Century Morocco 15. Jamāl al-dīn Muḥammad b. ʿAbd al-Raḥmān al-Qazwīnī (d. 739 ah/1338 ce), Talkhīṣ al-miftāḥ.157 16. Ḍiyāʾ al-dīn abū Muḥammad ʿAbd Allāh b. ʿUthmān al-Khazrajī (fl. c. 650 ah/1252 ce), al-Rāmiza al-shāfiyya fī ʿilm al-ʿarūḍ.158 17. ʿAbd al-Raḥmān b. Muḥammad al-Jādirī (d. 818 ah/1416 ce), Rawḍat al-azhār fī ʿilm waqt al-layl wa-l-nahār.159 Eight of the seventeen texts of this miscellany, which represent in a way the basis of the knowledge of an honnête homme in the Saadian period, coincide in some way with some of the contents of the books the three sisters presented as a waqf two centuries later. Miscellany The three sisters’ waqf No. 1 27 No. 4 25, 25 bis No. 8 15, 16, 26 No. 9 4 No. 10 17, 21, 22, 48, 16 K16 No. 11 47 No. 15 20, 29 ?, 16 K2 No. 17 43 It is, of course, out of the question to put on the same footing a library in one volume, a sort of compendium of fundamental texts in the education of the elites, and the various volumes which were belonging to scholars. However, finding either the same texts or commentaries of texts present in miscellany RBME 248 tells much about the permanence of a body of knowledge, or more particularly of works that have reached the status of classics. In RBME 248, the most recent text is the ʿAqīda al-ṣugrā by al-Sanūsī; among the works presented as a waqf by the three sisters, some of those which are related to the contents of the Escorial princely miscellany date from the seventeenth century, for instance nos 16, 20, 22 or 25 and 25 bis, but they are all commen- 157 GAL I, 295; S I, 516. 158 GAL I, 312/10; S I, 545. 159 See Benchekroun 1974, 247–250.

François Déroche, Lbachir Tahali taries. Is it a choice or the consequence of the availability on the market? As noted above, commentaries represent half of all books. 6 Conclusion The three sisters’ waqf constitutes in itself an interesting document on the way books circulated in Morocco during the second half of the eighteenth century. Put in the context provided by the different documents we have found, it takes on another meaning and reflects a deliberate policy to build a library and avoid the dispersion of books accumulated over two or more generations. The stakes are very clear: the various actors are the members of a family of scholars for whom access to books was of particular importance. When did the Bannānīs start collecting books? There is little doubt that the grandfather, ʿAbd al-Salām, had already begun to work in this direction by making a copy of Bukhārī’s Ṣaḥīḥ in eight volumes a waqf in favour of his son. This example and the documents we have found make clear that the three sisters therefore did not take an initiative but acted according to what was already a family policy: after the death of their brother Muḥammad, they did again what they had done previously after their father’s death or followed a tradition established by their grandfather. To what extent Muḥammad b. ʿAbd al-Salām’s personal experience when he was preparing his Maghānī al-wafāʾ li maʿānī al-Iktifāʾ induced him to implement a more ambitious approach of acquisition then of stabilization, even of hoarding of books? We are indeed told that the composition of this work has spread over sixteen years, mainly due to the lack of sources the author finally managed to find during his trip to the Near East – and we saw that a manuscript of the library had been actually bought in Mekka.160 His intellectual stature was an asset in his collecting business, as the waqf of both Qarawiyyin Library manuscripts 1013 and 1060 suggests: the merchant al-Wuraydī had agreed to make Muḥammad b. ʿAbd al-Salām and his son Ḥamdūn the beneficiaries of two waqfs that provided them with two series of books totalling 21 volumes. It seems that only the extinction of the lineage after the death of his son ʿAbd al-Karīm allowed the ultimate beneficiary, the library of the Qarawiyyin Mosque, to come into possession of these books on 2 Rabīʿ II 1201 ah/21 January 1787 ce – and not, as was stipulated, after that of Ḥamdūn’s demise in 1170 ah/1756 ce .161 Such a procedure was a clever way to increase the 160 See above and n. 115. Also ʿAbd al-Karīm Bannānī, MS Ḥasaniyya 13900, fol. 21r; ʿAbd alḤayy al-Kattānī, Fahris, ed. ʿAbbās 1982, I, 225. 161 al-Ribāṭī, Muqaddimat al-fatḥ, ed. Najīb 2008, 123.

Collecting Books in Eighteenth-Century Morocco library size, which was also possible through purchases. The importance of the travel to the East to make acquisitions has to be stressed again here. As shown by items nos 27 and 28 in the list of his works as well as by the incomplete Fahris asānīd found in manuscript 16 K, Muḥammad b. ʿAbd al-Salām was keen on following the traditional forms in the acquisition of knowledge from masters whose names he recorded carefully. Next to this model of aural transmission, his activity in the accumulation of books leaves no doubt about the importance of the latter in his own intellectual pursuits and for his scholarly status. His children tried to maintain and increase the library they inherited from him. There was a collective and coherent effort by the Bannānīs at building up by accumulation a patrimonial library as well as at keeping it together. Despite the demises, the divisions among the heirs and the necessary adaptations that they entailed, the Bannānī library developed during the second half of the eighteenth century. The construction of the zāwiya was certainly not only intended to receive the library as a last resort, but could give it a more formal dimension. Were the three sisters encouraged by their brother ʿAbd al-Karīm to give up their share of inheritance or had their father left his recommendations? Nothing allows to decide. According to the data currently available to us, we are able to list 83 titles in this collection (see Appendix I), but the number was of course higher: the exact contents of the multiple-text manuscripts escape us, except for the manuscript 16 K, and we do not know what was in ʿAbd al-Karīm’s possession and the exact number of manuscripts involved in the 1164 ah/1750 ce operation. It was anyhow a library of a respectable size for the standards of the time. Accumulating essential books to support the intellectual activity of the Bannānīs was probably not the only objective. Obviously, as suggested by the contents of manuscript 16 K, a copy of the works of the scholars belonging to this family (including their maternal ancestor Muḥammad b. Aḥmad Mayyāra al-Fāsī and perhaps others) was part of the collection: the latter defended the prestige of the family since it helped to preserve the successive contributions of various generations. It was therefore working ad majorem familiæ gloriam. Did the waqf guarantee its durability? We can doubt it. The presence of these manuscripts in the Qarawiyyin Library shows that the volumes were partly scattered and that the zāwiya did not fulfil its function of ultimate stronghold. But this is not, after all, an exceptional situation: two manuscripts that the great Saadian scholar Aḥmad al-Manjūr had established as a family waqf in 1572–1573 passed to Mūlāy Zaydān’s library and are nowadays in El Escorial.162 It is however quite exceptional to have the possibility to reconstruct, even partially, the contents of a private Moroccan library of the eighteenth century. 162 RBME 1032, fol. 1a (980 ce/1572 ah) and 1072, fol. 3a (981 ce/1573 ah).

François Déroche, Lbachir Tahali Acknowledgements The research leading to these results has received funding from the European Research Council under the European Union Funding for Research and Innovation (Horizon 2020) ERC Grant Agreement number 670628, ‘Saadian Intellectual and Cultural Life’ (SICLE, 2016-2021). The authors are particularly grateful to the Director of the Qarawiyyin Library, A. Boukchouf, for his generosity and help. They also want to thank M. Aouad for his invitation to explore the khurūm collection at the Qarawiyyin. Abbreviations EI2 = P. Bearman, T. Bianquis, C. E. Bosworth, E. van Donzel and W. P. Heinrichs (eds), Encyclopaedia of Islam, Second Edition, Online, Leiden: Brill <referenceworks.brillonline. com/browse/encyclopaedia-of-islam-2>. GAL = Carl Brockelmann, Geschichte der Arabischen Litteratur, I–II, Weimar: Felber, 1898 and 1902. GAL S = Carl Brockelmann, Geschichte der Arabischen Litteratur. Supplementbände, I–III, Leiden: Brill, 1937, 1938 and 1942. References Primary sources Manuscripts ʿAbd al-Karīm b. Muḥammad b. ʿAbd al-Salām Bannānī ‫عبد الكريم بن محمد بن عبد السالم بناني‬, Tuḥfat al-fuḍalāʾ al-aʿlām fī al-taʿrīf bi al-shaykh abī ʿAbd Allāh Maḥammad al-Bannānī ibn ʿAbd al-salām ‫تحفة الفضالء األعالم في التعريف بالشيخ أبي عبد الله محمد البناني ابن عبد السالم‬, autograph copy, MS Rabat, Ḥasaniyya Library, 13900. Muḥammad b. Aḥmad al-Ḥuḍaygī ‫محمد بن أحمد الحضيكي‬, al-Kunnāsha ‫الكناشة‬, MS Rabat, Ḥasaniyya Library, 13003. From the khurūm collection of the Qarawiyyin Library in Fez: Fgmt 1= 8-9-1667, Zakariyyāʾ al-Anṣarī ‫( زكريا األنصاري‬d. 926 ah/1520 ce), Iḥkām al-dalāla ʿalā taḥrīr al-Risāla ‫إحكام الداللة على تحرير الرسالة‬. Fgmt 2= 2-6-367, Abū al-Suʿūd Muḥammad b. Muḥammad al-ʿImādī ‫أبو السعود محمد بن محمد العمادي‬ (d. 982 ah/1574 ce), Irshād al-ʿaql al-salīm ilā mazāyā al-Qurʾān al-karīm ‫إرشاد العقل السليم إلى‬ ‫مزايا القرآن الكريم‬. Fgmt 3= 17-5-473, Ibn Ḥajar al-ʿAsqalānī ‫( ابن حجر العسقالني‬d. 852 ah/1449 ce), al-Iṣāba fī tamyīz al-Ṣaḥāba ‫اإلصابة في تمييز الصحابة‬. Fgmt 4= 1-4-110, Ibn al-Ḥājj ‫( ابن الحاج‬d. 737 ah/1336 ce), al-Madkhal ‫المدخل‬. Fgmt 5= 16-17-1663, 16-11-2400, 16-24-170, Abū Zayd ʿAbd al-Raḥmān b. ʿAbd al-Qādir al-Fāsī‫أبو‬ ‫( زيد عبد الرحمن بن عبد القادر الفاسي‬d. 1096 ah/1685 ce), Miftāḥ al-shifā fī muḥādhāt al-Shifā ‫مفتاح‬ ‫الشفا في محاذاة الشفا‬.

Collecting Books in Eighteenth-Century Morocco Fgmt 6= 1-21-589, al-Wansharīsī ‫( الونشريسي‬d. 914 ah/1508 ce), al-Miʿyār al-muʿrib wa al-jāmiʿ al-muġrib ʿan fatāwā ahl Ifrīqiya wa al-Andalus wa al-Maġrib ‫المعيار المعرب والجامع المغرب عن‬ ‫فتاوى أهل إفريقية واألندلس والمغرب‬. Fgmt 7= 5-49-626, ʿAbd Allāh b. Yūsuf b. ʿAbd Allāh b. Hishām ‫عبد الله بن يوسف بن عبد الله بن هشام‬ (d. 761 ah/1360 ce), Mughnī al-labīb ʿan kutub al-aʿārīb ‫مغني اللبيب عن كتب األعاريب‬. Fgmt 8= 6-8-624, ʿAbd al-ʿAzīz Ibn ʿAbd al-Salām al-Sulamī ‫عبد العزيز بن عبد السالم السلمي‬ (d. 660 ah/1261 ce), Qawāʿid al-aḥkām fī iṣlāḥ al-anām ‫قواعد األحكام في إصالح األنام‬. Fgmt 9= 10-43-2, Ibn Marzūq al-ḥafīd al-Tilimsānī ‫( ابن مرزوق الحفيد التلمساني‬d. 842 ah/1438 ce), Sharḥ al-Jumal fī al-manṭiq li al-Khūnajī ‫شرح الجمل في المنطق للخونجي‬. Fgmt 10= 16-2-270–280 and 16-7-658, Muḥammad b. Yūsuf al-Ṣāliḥī al-Shāmī ‫محمد بن يوسف‬ ‫( الصالحي الشامي‬d. 942 ah/1536 ce), Subul al-hudā wa-l-rashād fī sīrat khayr al-ʿibād ‫سبل الهدى‬ ‫والرشاد في سيرة خير العباد‬. Fgmt 11= 16-10-684, Ibn Sayyid al-nās ‫( ابن سيد الناس‬d. 734 ah/1333 ce), ʿUyūn al-athar fī funūn al-Maghāzī wa al-shamāʾil wa al-siyar ‫عيون األثر في فنون المغازي والشمائل والسير‬. Fgmt 12= 17-4-685, Ibn Khallikān ‫( ابن خلكان‬d. 681 ah/1282 ce), Wafayāt al-aʿyān wa-anbāʾ abnāʾ al-zamān ‫وفيات األعيان وأنباء أبناء الزمان‬. Editions ʿAbd al-Salām b. Sūda ‫عبد السالم بن سودة‬, Itḥāf al-muṭāliʿ bi wafayāt aʿlām al-qarn al-thālith ʿashar wa al-rābiʿ ‫إتحاف المطالع بوفيات أعالم القرن الثالث عشر والرابع‬, ed. Muḥammad Ḥajjī, Beyrouth: Dār al-Gharb al-Islāmī, 1997. al-Ḍuʿayyif al-Ribāṭī, Muḥammad b. ʿAbd al-Salām ‫ محمد بن عبد السالم‬،‫الضعيف الرباطي‬, Tārīkh al-Ḍuʿayyif ‫تاريخ الضعيف‬, ed. Muḥammad al-Būzīdī al-Shīkhī, 2nd edn, 2 vols, Casablanca: Dār al-Ṯaqāfa, 2007. al-Fāsī al-Fihrī, Maḥammad [sic] b. Aḥmad ‫ محمد بن أحمد‬،‫الفاسي الفهري‬, al-Mawrid al-haniyy bi akhbār al-imām ʿAbd al-Qādir al-Ḥasani ‫المورد الهني بأخبار اإلمام عبد القادر الحسني‬, ed. Khālid al-Ṣiqillī, Rabat: Dār al-Amān, 2008. Ibn al-Ṭayyib al-Sharqī ‫ابن الطيب الشرقي‬, al-Riḥla al-Ḥijāziyya ‫الرحلة الحجازية‬, ed. Nūr al-dīn Shūbad and Ḥasnāʾ Būtwādī, 2 vols, Rabat: Ministère des Affaires Islamiques, 2018. al-Kattānī, ʿAbd al-Ḥayy ‫ عبد الحي‬،‫الكتاني‬, Fahris al-fahāris wa al-athbāt wa muʿjam al-maʿājim wa al-mashyakhāt wa al-musalsalāt ‫فهرس الفهارس واألثبات ومعجم المعاجم والمشيخات والمسلسالت‬, ed. Iḥsān ʿAbbās, 2nd edn, 3 vols, Beirut: Dār al-Gharb al-Islāmī, 1982. al-Kattānī, ʿAbd al-Ḥayy ‫ عبد الحي‬،‫الكتاني‬, al-Tanwīh wa al-ishāda bi maqām riwāyat Ibn Saʿāda li Ṣaḥīḥ al-Bukhārī ‫التنويه واإلشادة بمقام رواية ابن سعادة لصحيح البخاري‬, ed. ʿAbd al-Majīd Khiyālī, Beirut: Dār al-kutub al-ʿilmiyya, 2012. al-Kattānī, Muḥammad b. Jaʿfar ‫ محمد بن جعفر‬،‫الكتاني‬, Sulwat al-anfās wa muḥādathat al-akyās bi man uqbira min al-ʿulamāʾ wa al-ṣulaḥāʾ bi Fās ‫سلوة األنفاس ومحادثة األكياس بمن أقبر من العلماء والصلحاء‬ ‫بفاس‬, ed. ʿAbd Allāh al-Kāmil al-Kattānī et al., 4 vols, Casablanca: Dār al-thaqāfa, 2004–2006. al-Manālī al-Zabādī, Muḥammad b. ʿAlī ‫ محمد بن علي‬،‫المنالي الزبادي‬, Sulūk al-ṭarīq al-wāriya fī al-shaykh wa al-murīd wa al-zāwiya ‫سلوك الطريق الوارية في الشيخ والمريد والزاوية‬, ed. ʿAbd al-Ḥayy al-Yamlāḥī, Tetouan: Association Tiṭwān Asmir, 2012. Muḥammad Dāwud ‫محمد داود‬, Tārīkh Tiṭwān ‫تاريخ تطوان‬, ed. Ḥasnāʾ Dāwud, 12 vols, Tetouan: Institute Muḥammad Dāwud, 2013. Muḥammad Makhlūf ‫محمد مخلوف‬, Shajarat al-nūr al-zakiyya fī ṭabaqāt al-Mālikiyya ‫شجرة النور الزكية‬ ‫في طبقات المالكية‬, ed. ʿAbd al-Majīd Khayālī, 2 vols, Beirut: Dār al-kutub al-ʿilmiyya, 2003.

François Déroche, Lbachir Tahali al-Qādirī, Muḥammad b. al-Ṭayyib ‫ محمد بن الطيب‬،‫القادري‬, Nashr al-mathānī li ahl al-qarn al-ḥādī ʿashar wa al-thānī ‫نشر المثاني ألهل القرن الحادي عشر والثاني‬, ed. Muḥammad Ḥajjī and Aḥmad al-Tawfīq, 4 vols, Rabat: Maktabat al-Ṭālib, 1982–1986. al-Rahūnī, Abū al-ʿAbbās Aḥmad ‫ أبو العباس أحمد‬،‫الرهوني‬, ʿUmdat al-rāwīn fī tārīkh Tiṭṭāwīn ‫عمدة‬ ‫الراوين في تاريخ تطاوين‬, ed. Jaʿfar b. al-Ḥājj al-Sulamī, 11 vols, Tetouan: Association Tiṭwān Asmīr, 2001–2015. al-Ribāṭī, Muḥammad Būjandār ‫ محمد بوجندار‬،‫الرباطي‬, Muqaddimat al-fatḥ min tārīkh Ribāṭ al-fatḥ ‫مقدمة الفتح من تاريخ رباط الفتح‬, ed. Aḥmad Najīb, Cairo: Markaz najībawayh, 2008. al-Shafshāwnī, Sulaymān al-Ḥawwāt ‫ سليمان الحوات‬،‫الشفشاوني‬, Thamrat unsī fī al-taʿrīf bi nafsī ‫ثمرات‬ ‫أنسي في التعريف بنفسي‬, ed. ʿAbd al-Ḥaqq al-Ḥīmar, Shafshāwn: Markaz al-buḥūth wa al-dirāsāt al-Andalusiyya, 1996. Secondary literature Benchekroun, Mohamed (1974), La vie intellectuelle marocaine sous les Mérinides et les Waṭṭāsides (XIIIe XIVe XVe XVIe siècles), Rabat: [s. n.]. Déroche, François (2019), ‘The Prince and the Scholar. A Study of Two Miscellanies from Fifteenth and Sixteenth Centuries Morocco’, in Alessandro Bausi, Michael Friedrich and Marilena Maniaci (eds), The Emergence of Multiple-Text Manuscripts (Studies in Manuscript Cultures, 17), Berlin: De Gruyter, 165–191. El-Rouahyeb, Khaled (2015), Islamic Intellectual History in the Seventeenth Century. Scholarly Currents in the Ottoman Empire and the Maghreb, Cambridge: Cambridge University Press. Fagnan, Edmond (1924), Extraits inédits relatifs au Maghreb (géographie et histoire), traduits de l’arabe et annotés, Algiers: Jules Carbonel. Hajji, Mohamed (1976), L’activité intellectuelle au Maroc à l’époque saʿdide, 2 vols, Rabat: Dar el Maghrib. de Jong, Frederick and Jan Just Witkam (1987), ‘The Library of al-Šayḵ Ḵālid al-Šahrazūrī al-Naqšbandī (d. 1242/1847). A Facsimile of the Inventory of His Library (Damascus, Maktabat al-Asad, No. 259)’, Manuscripts of the Middle East, 2: 68–87. al-Kattānī, ʿAbd al-Kabīr ‫ عبد الكبير‬،‫( الكتاني‬2002), Zahr al-ās fī buyūtat ahl Fās ‫زهر اآلس في بيوتات أهل‬ ‫فاس‬, 2 vols, Casablanca: al-Najāḥ al-jadīda. Le Tourneau, Roger (1934), ‘Notes sur les lettres latines de Nicolas Clénard relatant son séjour dans le royaume de Fès (1540-1541)’, Hespéris, 19: 45–63. Le Tourneau, Roger (1977), ‘Histoire de la dynastie saʿdide. Extrait de al-Turğumān al-muʿrib ʿan duwal al-Mašriq wal Maġrib d’Abū al-Qāsim b. Aḥmad b. ʿAli b. Ibrāhīm al-Zayyānī… Texte, traduction et notes présentés par L. Mougin et H. Hamburger’, Revue de l’Occident musulman et de la Méditerranée, 23/1: 7–109. al-Manūnī, Mohamed ‫( محمد المنوني‬1991), Tārīkh al-Wirāqa al-Maghribiyya ‫تاريخ الوراقة المغربية‬, Rabat: Presses Université Muḥammad V. al-Manūnī, Mohamed ‫( محمد المنوني‬2005), Dūr al-kutub fī māḍī al-Maghrib ‫دور الكتب في ماضي المغرب‬, Rabat: Ḥasaniyya Library. Pellat, Charles (1990), ‘Mayyāra’, in EI2, VI, 932. Schacht, Joseph (1960), ‘Bannānī’, EI2, I, 1019. al-Sibāʿī, Khaled ‫ خالد‬،‫( السباعي‬2017), Tāʾrīkh al-maktaba al-Kattāniyya ‫تاريخ المكتبة الكتانية‬, 2 vols, Beirut: Dār al-Ḥadīth al-Kattāniyya.

Collecting Books in Eighteenth-Century Morocco Appendix I: The Bannānī library c. 1770 (list of holdings according to currently available sources) Authors’ names ʿAbd Allāh b. Yūsuf b. ʿAbd Allāh b. Hishām (d. 761 ah/1360 ce) ʿAbd al-ʿAzīz Ibn ʿAbd al-Salām al-Sulamī (d. 660 ah/1261 ce) ʿAbd al-Bāqī al-Zarqānī (d. 1099 ah/1687 ce) ʿAbd al-Raḥmān al-Fāsī (d. 1036 ah/1626 ce) ʿAbd al-Raḥmān al-Jādirī (d. after 818 ah/1415 ce) ʿAbd al-Raḥmān al-Jazūlī (d. 741 ah/1340 ce) ʿAbd al-Salām b. Ḥamdūn Bannānī (d. 1136 ah/1723 ce) Abū al-Fidāʾ Ibn Kathīr (d. 774 ah/1372 ce) Abū Isḥāq al-Zajjāj (d. 311 ah/923 ce) Abū al-Suʿūd Muḥammad b. Muḥammad al-ʿImādī (d. 982 ah/1574 ce) Abū Zayd ʿAbd al-Raḥmān b. ʿAbd al-Qādir al-Fāsī (d. 1096 ah/1685 ce) ʿAḍud al-dīn al-Ījī (d. 756 ah/1355 ce) Aḥmad b. Mubārak al-Sijilmāssī (d. 1156 ah/1743 ce) Aḥmad b. Sulaymān b. Kamāl Bāshā (d. 940 ah/1534 ce) Aḥmad b. Yaḥyā al-Wansharīsī (d. 914 ah/1508 ce) Aḥmad Ibn Zarrūq (d. 899 ah/1493 ce) ʿAlī b. Muḥammad Baraka al-Tiṭwānī (d. 1120 ah/1708 ce) Badr al-dīn al-ʿAynī al-Bukhārī (d. 855 ah/1451 ce) Burhān al-dīn al-Safāquṣī (d. 743 ah/1342 ce) al-Damīrī (d. 805 ah/1402 ce) Fakhr al-dīn al-Rāzī (d. 606 ah/1209 ce) al-Ḥājj al-Shuṭaybī (d. 963 ah/1555 ce) Ḥamdūn b. Muḥammad Bannānī (d. 1170 ah/1756 ce) Ibn ʿAbd al-Barr al-Numayrī (d. 463 ah/1071 ce) Ibn ʿArafa (d. 803 ah/1400 ce) Ibn al-Athīr (d. 606 ah/1209 ce) Ibn ʿAṭiyya (d. 542 ah/1147 ce) Ibn Baṭṭūṭa (d. 779 ah/1377 ce) Ibn Ghāzī (d. 919 ah/1513 ce) Ibn Ḥajar al-ʿAsqalānī (d. 852 ah/1449 ce) Ibn Ḥajar al-Haytamī (d. 973 ah/1565 ce) Ibn al-Ḥājib (d. 646 ah/1248 ce) Fgmt 7163 Fgmt 8 51 25 bis 43 26 (16 K) 41 (?) 40 Fgmt 2 Fgmt 5 44 16 K15 16 K12 (?) 35 37 16 K16 21 2 5 38 57 48 16 K14 18 1 28 39 14 Fgmt 3 31 23 Ibn al-Ḥājj (d. 737 ah/1336 ce) 7, Fgmt 4 163 The numbers alone refer to the items listed in the three sisters’ waqf; ‘Q’ designates a manuscript in the Qarawiyyin Library main collection, ‘Fgmt’ a fragment in the deposit of the same library (see the call number in the bibliography); ‘16 K’ is the manuscript in the Royal Library of the Kingdom of Morocco. In some cases, the title may be followed by two references when it is known in a document and through the manuscript itself (for example al-Madkhal).

François Déroche, Lbachir Tahali Authors’ names Ibn Khallikān (d. 681 ah/1282 ce) Ibn Mālik (d. 672 ah/1273) Fgmt 12 46, 55 Ibn Marzūq al-ḥafīd al-Tilimsānī (d. 842 ah/1438 ce) Fgmt 9 Ibn al-Nafīs (d. 687 ah/1288 ce) 32 Ibn Qayyim al-Jawziyya (d. 751 ah/1350 ce) 19 Ibn al-Rashīd al-Baghdādī (d. 662 ah/1263 ce) 49 Ibn Sayyid al-nās (d. 734 ah/1333 ce) 12, Fgmt 11 Ibn ʿUmar al-Anfāsī (d. 761 ah/1359 ce) 15 Ibn al-Waḥshī (d. 502 ah/1108 ce) 59 al-Jurjānī (d. 816 ah/1413 ce) 9 (?) Khālid al-Azharī (d. 905 ah/1499 ce) 17, 47 al-Kharshī (d. 1101 ah/1690 ce) Q 1060 al-Lakhmī (d. 478 ah/1085 ce) 8 al-Maḥallī (d. 864 ah/1459 ce) 11 al-Manjūr (d. 995 ah/1586 ce) 52 Mayyāra al-Fāsī (d. 1072 ah/1661 ce) 34 Muḥammad b. ʿAbd al-Qādir al-Fāsī (d. 1116 ah/1704) 58 Muḥammad b. ʿAbd al-Salām Bannānī (d. 1163 ah/1749 ce) 40, 45 (?), 50, 16 K1, 3, 7, 8, 9, 11 and two missing texts 33 Muḥammad al-Fishtālī (d. 777 ah/1376 ce) Muḥammad b. Muḥammad b. ʿAbd al-Salām Bannānī (d. 1182 ah/1769 ce) 16 K2 Muḥammad b. Yūsuf al-Ṣāliḥī al-Shāmī (d. 942 ah/1536 ce) Fgmt 10 al-Mundhirī (d. 656 ah/1258 ce) Q 1013 Muslim b. al-Ḥajjāj (d. 261 ah/875 ce) 10 al-Raṣṣāʿ al-Tilimsānī (d. 894 ah/1488 ce) 56 al-Rāzī al-Taḥtānī, Quṭb al-dīn (d. 766 ah/1364 ce) Fgmt 9 al-Rūdānī (d. 1094 ah/1683 ce) 16 K10 al-Saktānī (d. 1062 ah/1651 ce) 25 al-Subkī (d. 771 ah/1369 ce) 42 (?) al-Suyūṭī (d. 911 ah/1505 ce) 3, 53 al-Taftāzānī (d. 792 ah/1390 ce) 29 al-Tatāʾī (d. 942 ah/1535 ce) 4 al-Tirmidhī (d. 279 ah/892 ce) 56 bis al-Tuwātī (d. 1160 ah/1747 ce) 16 K17 al-Ujhūrī (d. 1066 ah/1655 ce) 13, 16 al-Wansharīsī (d. 914 ah/1508 ce) Fgmt 6 Yāsīn al-ʿUlaymī (d. 1061 ah/1651 ce) 20, 22 Zakariyyāʾ al-Anṣarī (d. 926 ah/1520 ce) Fgmt 1

Collecting Books in Eighteenth-Century Morocco Titles al-Badr al-lāmiʿ fī ḥall Jamʿ al-jawāmiʿ Fahrasa d’al-Manjūr Fahrasa de Muḥammad Bannānī (?) Fahris asānid al-Fāʾiq fī aḥkām al-wathāʾiq Fatḥ al-Jalīl ʿalā Mukhtaṣar al-šaykh Khalīl Fatḥ al-qarīb al-Mujīb ʿalā Sharḥ al-Targhīb wa al-tarhīb Ḥāshiya ʿalā Alfiyyat Ibn Mālik Ḥāshiya ʿalā Mukhtaṣar sharḥ Talkhīṣ al-Miftāḥ Ḥāshiya ʿalā Sharḥ Umm al-barāhīn: al-ʿAqīda al-ṣughrā Ḥāshiya Kamāl Bāshā ʿalā al-Bayḍāwī Ḥāshiya al-Zarqānī ʿalā Sharḥ al-Luqānī ʿalā khuṭbat Mukhtaṣar Khalīl al-Hidāya al-Kāfiya al-shāfiya li-bayān ḥaqāʾiq Ibn ʿArafa al-wāfiya al-Ḥudūd fī al-taʿrīfāt al-fiqhiyya Ḥusn al-muḥāḍara fī tārīkh Miṣr wa al-Qāhira Iḥkām al-dalāla ʿalā taḥrīr al-Risāla Ikhtiṣār Sharḥ khuṭbat al-Alfiyya Intiqāʾ Miʾat Ḥadīth fī al-targhīb fī iṣṭināʿ al-maʿrūf Irshād al-ʿaql al-salīm ilā mazāyā al-Qurʾān al-karīm al-Iṣāba fī tamyīz al-Ṣaḥāba Itḥāf dhawī al-istiḥqāq bi baʿḍi murād al-Murādī wa Zawaʾid Abī Isḥāq Izālat al-labs ʿan al-masāʾil al-khams Jāmiʿ al-Ummahāt aw al-Mukhtaṣar al-farʿī Jamʿ al-jawāmiʿ fī Uṣūl al-fiqh (?) Jawāhir al-kalām Kitāb al-Jumān fī mukhtaṣar akhbār al-zamān al-Madkhal Manāsik ʿAlī Baraka al-Ṭiṭwānī al-Maqāṣid al-naḥwiyya, Sharḥ shawāhid shurūḥ al-Alfiyya Mawāhib al-jalīla al-mustanbaṭa min kalām kull jalīl li yatayassara bihā ḥall alfāz Mukhtaṣar Khalīl Miftāḥ al-shifā fī muḥādhāt al-Shifā al-Miṣbāḥ fī Sharḥ al-Miftāḥ al-Miʿyār al-muʿrib wa al-jāmiʿ al-muġrib ʿan fatāwā ahl Ifrīqiya wa alAndalus wa al-Maġrib Mughnī al-labīb ʿan kutub al-aʿārīb al-Muḥarrar al-wajīz fī tafsīr al-kitāb al-ʿazīz Muḥaṣṣal afkār al-mutaqaddimīn wa al-mutaʾakhkhirīn min al-ʿulamāʾ wa al-ḥukamāʾ wa-l-mutakallimīn al-Mūjaz fī al-ṭibb al-Mujīd fī iʿrāb al-Qurʾān al-majīd al-Naṣīḥa al-kāfiya li man khaṣṣahu Allāh bi al-ʿāfiya Naṣīḥat al-mughtarrīn wa kifāyat al-muḍṭarrīn fī al-tafrīq bayna al-muslimīn bi mā lam yunzilhu rabb al-ʿālamīn wa lā jāʾa bihi al-rasūl al-amīn wa lā thabata ʿan al-khulafāʾ al-mahdiyyīn EBSCOhost - printed on 2/6/2024 8:47 AM via . All use subject to https://www.ebsco.com/terms-of-use 11 52 45, 16 K3 16 K8 33 4 Q 1013 22 20 25, 25 bis 16 K12 (?) 51 56 18 3 Fgmt 1 48 16 K10 Fgmt 2 Fgmt 3 14 16 K15 23 42 44 57 7, Fgmt 4 16 K16 21 Q 1060 Fgmt 5 9 Fgmt 6 Fgmt 7 28 38 32 2

François Déroche, Lbachir Tahali Titles al-Nihāya fī gharīb al-ḥadīth wa al-athar al-Nuqāya fī arbaʿata ʿashara ʿilman 1 53 al-Qaṣāʾid al-witriyya fī madḥ khayr al-bariyya 49 Qawāʿid al-aḥkām fī iṣlāḥ al-anām Fgmt 8 Qiṣṣat Ahl al-kahf 41 Rawāʾiq al-ḥulal fī alqāb al-ziḥaf wa al-ʿilal 16 K17 (?) Rawḍat al-azhār fī ʿilm waqt al-layl wa al-nahār 43 al-Riḥla al-ḥijāziyya 50 Risāla fī faḍl al-niṣf min Shaʿbān 16 K11 Ṣaḥīḥ al-Bukhārī Ṣaḥīḥ Muslim 10 al-Ṣawāʿiq al-muḥriqa ʿalā ahl al-rafḍ wa al-ḍalāl wa al-zandaqa 31 al-Shamāʾil al-Muḥammadiyya wa al-khaṣāʾil al-muṣṭafawiyya 56 bis al-Shāmil fī fiqh al-imām Mālik 5 Sharḥ al-Ajurrūmiyya fī uṣūl ʿilm al-ʿarabiyya 47 Sharḥ al-Ḥizb al-Kabīr li Abī al-Ḥasan al-Shādhilī 16 K7 Sharḥ al-Jumal fī al-manṭiq li al-Khūnajī Fgmt 9 Sharḥ khuṭbat Mūkhtaṣar al-Saʿd al-Taftāzānī li Talkhīṣ al-Miftāḥ 16 K2 Sharḥ Mūkhtaṣar Ṣaḥīḥ al-Bukhārī 13 Sharḥ Musnad al-Shihāb fī al-Ḥadīth 59 Sharḥ Qaṣīdat sayyidī ʿAbd al-Raḥmān al-Fāsī fī ʿilm al-usṭurlāb (16 K) Sharḥ al-Risāla 15, 16, 26 Sharḥ al-Ṣalāt al-Mashīshiyya 16 K9 Sharḥ al-Shamsiyya Fgmt 9 Sharḥ al-Targhīb wa al-tarhīb Q 1013 Subul al-hudā wa-l-irshād fī sīrat khayr al-ʿibād Fgmt 10 al-Tabṣira 8 Takmīl al-marām bi Sharḥ shawāhid Ibn Hishām 58 Taʾlīf fī mā lā yanṣarif min al-asmāʾ 40, 16 K1 Taʾlīf al-imām Ibn ʿAbd al-Barr fī al-Basmala hal hiya āya min al-Fātiḥa am lā 16 K14 Tamrīn al-Ṭullāb fī ṣināʿat al-iʿrāb: Iʿrāb Alfiyyat Ibn Mālik 17 Tashīl al-fawāʾid wa takmīl al-maqāṣid 55 Taʿẓīm al-aimma al-arbaʿa (16 K) al-Tuḥfa ʿalā Kāfiyat Ibn al-Ḥājib 46 Tuḥfat al-nuẓẓār fī gharāʾib al-amṣār wa ʿajāʾib al-asfār 39 ʿUddat al-burūq fī jamʿ mā fī al-mathhab min al-jumūʿ wa al-furūq 35 ʿUyūn al-athar fī funūn al-Maghāzī wa al-shamāʾil wa al-athar 12, Fgmt 11 Wafayāt al-aʿyān wa-anbāʾ abnāʾ al-zamān Fgmt 12 Zād al-maʿād fī hady khayr al-ʿibād

Collecting Books in Eighteenth-Century Morocco Appendix II: The Bannānī family ʿAbd Allāh Bannānī Aḥmad Būsta Maḥammad Aḥmad Mayyāra al-Fāsī ʿAbd al-Wāḥid Muḥammad (d. 1662) Ḥamdūn (d. 1681) Aḥmad said Ḥaddū (d. 1701) ʿAbd al-Salām (d. 1723) Fāṭima Muḥammad al-ʿArabī Abū Ḥāmid ʿAmm al-ʿArabī (d. 1768) Muḥammad ʿAbd al-Karīm (d. 1782) ʿAshwa Third brother Muḥammad Muḥammad al-Makkī (d. 1829) Fāriḥa Fāṭima EBSCOhost - printed on 2/6/2024 8:47 AM via . All use subject to https://www.ebsco.com/terms-of-use Ḥamdūn Muḥammad

East and South Asia

Jean-Pierre Drège Two Libraries of the Tang Capital Abstract: Little is known of medieval libraries in medieval China before the eleventh century. Their organisation has been deduced almost solely through catalogues of manuscripts and lists of officials. Somewhat better documented are two very different libraries – one generalist, the other religious – founded in Chang’an, the western capital of the Tang Empire (seventh−eighth centuries). The first is the Library of the Academy of the Hall of the Gathered Wise Men (Jixian dian shuyuan 集賢殿書院), established in the first quarter of the eighth century and located in the Chang’an Imperial Palace. The second is the library of the Monastery of Western Brightness (Ximing si 西明寺), a Buddhist monastery founded in 656, and located near the imperial palace. Several great monks, famous pilgrims, and translators of Buddhist scriptures stayed there. 1 Introduction The organisation of pre-eleventh-century medieval libraries in China is documented in the surviving catalogues of ancient manuscripts and in the lists of officials assigned to these libraries, as included in official dynastic histories. It was under the Former Han Dynasty (202 bce–8 ce) that a relatively firm concept of a library as an imperial government service first appeared. At the time, libraries belonged to the emperor’s treasury (bifu 祕府); they were used to store books as well as administrative records (memorials, legal documents, etc.), maps, and census registers. Manuscripts to be made public were moreover required to be copied there. This practice, which resulted in censorship, was probably an extension of the strict control measures of the first emperor, Qin Shi Huangdi 秦始皇帝, during his reign from 221 to 209 bce. He has been associated with the destruction of all works liable to trigger discussion or challenge his authority.1 Under the Han rulers (202 bce–220 ce), book collections increasingly grew and had to be divided across multiple locations. At the time, a distinction was made between ‘inner’ and ‘outer’ books, which, based on their contents, were stored either in the palace or outside the palace. For the period from Antiquity 1 On the history of Chinese libraries, see Drège 1991; for Han Dynasty libraries specifically, see 19–23 and 95–108. See also Fölster 2018. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-005

Jean-Pierre Drège until the third century ce, little is known about this distribution of the books. It is during this period that the Directorate of Secret Writings (Bishu jian 祕書監) – in other words, the library – became an independent body of the Palace Secretariat (Zhongshu sheng 中暑省). Books were stored separately from archives from this point on; from a theoretical point of view, however, this distinction had already existed under the reign of previous Han rulers, as the earliest surviving imperial library catalogue, which dates back to the late first century bce, contains no archives, but virtually only books. The first catalogues of religious books appeared in the fourth century. Buddhism, a foreign religion originating in India, paved the way. For about three centuries, there were enough Buddhist texts from India and Central Asia translated into Chinese to establish libraries and catalogues. Moreover, Taoism, a philosophical doctrine that transformed into a religious system, was expanded with a multitude of sutras and doctrinal texts modelled on Buddhist works that at times were simply adapted. Amid the subsequent fierce competition between the two religions, numerous Taoist monasteries and libraries were established. From the official viewpoint, even though the Han-era distinction between ‘inner’ and ‘outer’ books was no longer maintained by this time, religious books, mainly Buddhist and Taoist ones, remained clearly distinct from general works, with the exception of a few religious biographies or encyclopaedias that were included in the general collections. In the capital cities, the countless Buddhist and Taoist texts were stored not in buildings belonging to the palace, but in nearby monasteries. Religious books were generally considered complementary to non-specialist works in various book-classification systems. While explicit information on medieval libraries is scarce, two libraries created under the Tang Dynasty (618–907) provide some evidence. They differ broadly in nature – one is generalist, the other religious –, though both were founded in Chang’an, the western capital. I shall first discuss the Library of the Academy of the Hall of the Gathered Wise Men (Jixian dian shuyuan 集賢殿書院), which was created in the first quarter of the eighth century. It was located at the Imperial Palace of Chang’an and associated with the compilation of a vast catalogue listing the titles of all the books preserved at the various official libraries. I shall then turn to the library of the Monastery of Western Brightness (Ximing si 西 明寺), a Buddhist monastery founded in the previous century, in 656, and located near the imperial palace. It was home to several prominent monks, famous pilgrims, and translators of Buddhist texts.

Two Libraries of the Tang Capital 2 The library of the academy of the hall of the Gathered Wise Men Traditionally, and up until the beginning of the eighth century, the Emperor’s library was called the Bishu sheng 祕書省 (Department of Secret Writings). Of course, it was not the only official library. Other libraries were associated with educational institutions that trained senior state officials, such as the College for the Advancement of Literature, (Hongwen guan 弘文館), or with organisations, such as the Historiography Institute (Shiguan 史館), where official history was written. The latter organisation had equivalents at the level of the Crown Prince, namely the Library of the Crown Prince (Sijing ju 司經局) and the College for the Veneration of Literature (Chongwen guan 崇文館). The gynaeceum also had a library, although it was smaller. Each of these institutions had a workforce ranging from about thirty to 140 employees, without counting the scholars and higher officials acting as supervisors.2 Upon his ascent to the throne, Emperor Xuanzong 玄宗, who reigned from 712 to 756, found that the Imperial Library was inadequate. Throughout history, rulers inaugurating a dynasty and emperors leaving their mark on history have made this same observation. In the case of Xuanzong, the period before his rule had been characterized by significant unrest at the highest level, starting from the end of the reign of Gaozong 高宗 (649–683), and particularly with the usurpation of power by Empress Wu Zetian 武則天 (690–705). This was a period of rapid turnover among the senior officials responsible for the imperial library. Rules had been established in 684 stating that new books had to be presented at the beginning of every year, and that a complete inventory had to be drawn up every three years. These requirements do not appear to have been upheld over the long term. Based on a variety of works, it is possible to retrace the establishment of a new library that gradually came to compete with the Department of Secret Writings and even replace it.3 In 715, the emperor appointed two senior officials, primarily to take responsibility for organising the books – which mainly consisted of copies produced under the reigns of his predecessors – and to submit a commentary on the books to him. Following the report made by each of these officials, it was decided to create a new, detailed catalogue describing all existing works. This became the Catalogue of All Books in Four Categories (Qunshu sibu lu 群書四部 2 Drège 1991, 70–82. 3 Regarding this library, see Drège 1991, 52–63; Ikeda On 1971, 45–98; Zheng Weizhang 1983, 65–85; and Lu Yining, online.

Jean-Pierre Drège 錄), which extended over 200 scrolls and described 2,655 titles spanning 48,169 scrolls. This catalogue was soon lost, but was used to create the bibliography preserved in the Old History of the Tang (Jiu Tangshu 舊唐書), which was completed in the mid-tenth century. The main feature that distinguishes this episode from other moments in library history is that this catalogue contained descriptions of the works and information on their locations. Historical sources usually provide few details on the physical nature of libraries; for this type of information, it is necessary to refer to manuscripts found at the beginning of the twentieth century, not in any of the capital cities, but in the far reaches of Western China, in the famous library cave of the Dunhuang oasis (Gansu). There, tens of thousands of manuscript scrolls dating back to the fifth to tenth centuries were discovered.4 The project to create a new library gradually took shape. About twenty senior officials were appointed to organise the books; the works, including both new and old copies, were first stored in the eastern wing of the Palace of the Wing of the Academy of Heaven (Qianyuan dian 乾元殿), in the eastern capital of Luoyang, where the emperor was located at the time. They were divided into four categories based on a system dating back to the third century: classics, histories, masters, and literary works (jing 經, shi 史, zi 子, ji 集). The scrolls were placed on shelves. Over the winter, the emperor went to the western capital, Chang’an, and the books were then transferred to the Hall of Elegance and Rectitude (Lizheng dian 麗正殿), located in the Palace of the Crown Prince, which bordered the Imperial Palace to the East (see Fig. 1).5 In the following year, 719, to increase the pace of copying, a large number of officials from other imperial departments were brought in, and an official four-category catalogue was established. Three years later, in 722, the Emperor decided to transport the books from the library to the Palace of Great Brightness (Daming gong 大明宮). The following year, during the banquet held in the Hall of the Gathered Immortals (Jixian dian 集仙殿), the Emperor decided to commemorate the transfer of the books by bringing together the empire’s wise men and by changing the name of the hall from that of the ‘Gathered Immortals’ to that of the ‘Gathered Wise Men’. This was home to the Library of Elegance and Rectitude, which at the same time received the name of the ‘Library of the Academy of the Hall of the Gathered Wise Men’. In 736 this library was transferred to the Palace of the Stimulation of Auspiciousness (Xingqing 興慶宮), outside the Imperial City, directly against the east wall. Shortly afterwards, in 740, it was established at the Palace of Elegant Purity (Huaqing 華清宮), which was not yet the one built later near the hot springs of Mount Li (驪山). That is the abridged history of this library. 4 For an overview of the types of works discovered in Dunhuang and their materiality, cf. Drège 2014. 5 On the city of Chang’an under the Tang Dynasty, see Thilo 1997–2006; Xiong 2000.

Jean-Pierre Drège As usual for this period, books are scrolls of sheets of paper, wrapped around a central axis, to which the written scroll was attached via a blank sheet. At the beginning of each scroll, a protective cover sheet was glued to the first sheet on which the text begins. The cover sheet displayed the title of the work and the 1. Four-category library and storage room for paper, brushes, and other instruments 2. Middle courtyard 3. Interior room of the small courtyard 4. East room or room for scholars 5. North-east courtyard 6. North courtyard 7. Scholars’ kitchen courtyard 8. West courtyard 9. Copyists’ kitchen 10. Temporary teachers’ building 11. Star observation terrace (where omens were taken) Fig. 2: Layout of the courtyard of the Academy of the Hall of the Gathered Wise Men in Daming Gong, Chang’an (drawing from Ikeda 1971). scroll number, which doubled as the chapter number; a silk ribbon attached to the sheet was used to tie the scroll closed. Each scroll contained an average of fifteen sheets. The scrolls were assembled into groups of ten (on average) and wrapped in bamboo weave or brocade. These bundles were placed on shelves. Each bundle (not each scroll, as was often believed) featured a label containing the title of either the main work in the bundle, or the work that appeared first

Two Libraries of the Tang Capital in the order of the library’s book catalogue. The four categories of books were marked by the colour of the central rolling sticks and the silk ribbons, and were also distinguished by the appearance of their wrapping and the colour of their ivory labels. The rolling sticks generally consisted of an engraved or decorated wooden shaft with protruding, coloured ivory tips. The ribbons were made of woven silk of one or more colours. The thin strips of bamboo used for the wrappers were assembled with coloured silk threads, imitating the brocades. All these details are recorded, but vary between sources (see Tables 1 and 2 below). Table 1: Materials and colours of the books in the Academy of the Hall of the Gathered Wise Men, based on Jixian jizhu (Abridged History [of the Academy] of the Hall of the Gathered Wise Men). Rolling sticks Ribbons Wrappers Classics White ivory Red (chi 赤) and yellow Yellow brocade History Sky-blue ivory Blue-green Masters Purple sandalwood Purple Literature Green ivory Red (fei 緋) Red (hong 紅) Labels Green ivory and sky blue White ivory Table 2: Materials and colours of the books in the Academy of the Hall of the Gathered Wise Men, based upon the Da Tang liudian (Six Canons of Great Tang). Rolling sticks Ribbons Classics Inlaid white ivory Yellow History Inlaid blue-green ivory Pale blue Engraved purple sanPurple dalwood Green ivory Vermilion Masters Literature Wrappers Labels Red (hong 紅) ivory Green ivory Sky-blue ivory White ivory This description of the books copied for the Library of the Academy of the Hall of the Gathered Wise Men, which demonstrates a clear set of rules (even though they fluctuate), clearly does not apply to ancient manuscripts surviving from previous eras. In fact, century-old scrolls going back to the Liang Dynasty (502–557) were also found at this institution, preserved but regarded with a certain disdain by cataloguers, who had them hastily copied. The paper quality was considered

Jean-Pierre Drège mediocre, the writing clumsy and at times incorrect, and the copies were not signed by important names. Only manuscripts dating back to the Sui Dynasty (581–618) were admired. As a result of this coexistence of copies from multiple periods, the summary of the four categories of works provides figures that differ substantially from those of the catalogue presented to the emperor in 720: on the one hand, 3,060 titles across 51,852 scrolls (with a few variants depending on the sources), on the other hand, more than 80,000 scrolls.7 Unfortunately, the sources do not always list the number of titles, neither per category nor in total. We can surmise that among the more than 80,000 scrolls, some duplicates existed. These concerned the older manuscripts mentioned above, for which we have indications of the diversity of their central rolling sticks, which came in seventeen different types, as well as of their silk ribbons (eighteen types). There were also multiple copies of the same work. We can ascertain this because the catalogue recorded only a single copy of each work, as was the custom, whereas an inventory would have recorded the total number of the scrolls present. It is possible that some of these multiple copies were identical content-wise and created in the same context, as we know that under the Tang Dynasty, each work was reproduced for the emperor’s library in three copies: a main copy (yuzheng 御 正), a secondary copy (yufu 御副), and a reserve copy (zhu 貯). It is difficult to know where these three copies were stored. If three copies were stored at the Department of Secret Writings, the same was probably true for the Library of the Academy of the Hall of the Gathered Wise Men, as well as the Library of the Crown Prince and likely elsewhere. In addition to the main copies, the sources mention the secondary copies, but say nothing about the reserve copies which probably did not benefit from materials as precious as the main copies. The sources mention secondary copies of classics, the rolling sticks of which were purple. Moreover, we know that in 615, under the preceding Sui Dynasty, fifty secondary copies of 37,000 scrolls were made, to be stored at palaces, ministries, and the main administrative departments of the two capital cities. On the other hand, we know nothing about the reserve copies. The secondary copies of the Library of the Academy of the Hall of the Gathered Wise Men differ in appearance from the main copies, such as in the quality of their rolling sticks. At this library, there were therefore approximately 30,000 scrolls, which must have had duplicates. In other words, not all works would have been copied multiple times, but possibly only 7 The numbers vary, depending on the sources, due to copying errors, which were frequent when it came to numbers. Here, the total number of scrolls varies between 80,090 and 89,000.

Two Libraries of the Tang Capital the works considered most important or those belonging to the most important categories, namely classics and history.8 3 The library of the Monastery of Western Brightness The works pertaining to the two great religions, Buddhism and Taoism, were generally not counted in official bibliographies or stored alongside the generalist works in official libraries. They were stored at monasteries subsidised by the imperial authorities, including in capital cities. We know almost nothing of the organisation or operation of monastic libraries, except in the case of the oasis of Dunhuang, where loan lists, acquisition lists, and other Buddhist documents have survived, thus shedding some light on this subject in the ninth and tenth centuries.9 We have nothing of the sort for Tang Dynasty capitals. All that remains are the sometimes very detailed catalogues of the works of the Buddhist canon, as well as bibliographic treatises on the history of texts. Regarding the library’s location within the monasteries, it appears that the books were kept in a pavilion on the second floor, the sutra pavilion, which was generally built to the left of the monastery’s great hall. This was sometimes called the ‘drum tower’ and mirrored the bell tower.10 We are unaware of the arrangement of the works contained in the library, in particular between floors. While a monastic library should technically have had a more or less complete copy of the Buddhist canon, it probably also contained manuscripts offered by believers who ordered copies of the sutras for their deceased parents or on special occasions, possibly stored on one of the upper floors. To get an idea of the layout of a Buddhist monastery library, we must turn to the Yingzao fashi 營造法式 (‘Construction Methods’), an early twelfth-century architectural textbook by Li Jie 李誡 (c. 1065–1110?), presenting the features and dimensions of a model wall library (see Fig. 3) and a Buddhist revolving library.11 8 Drège 1991, 64–65. 9 Drège 1991, 216–236. 10 See Xiao 1987. 11 Li 1933, 11: 16–27 and 23: 12–21. We may also refer to the remains of the library of the Buddhist monastery of the Ornament Scripture, Huayan si, in Datong (Shanxi), which dates back to the Liao Dynasty, more precisely to 1038. See Luo 2016, 94–109.

Jean-Pierre Drège Fig. 3: Representation of a wall library; Li Jie (c. 1065–1110?), Yingzao fashi (‘Construction Methods’), ed. Wanyou wenku 1933. The wall library, which does not follow a straight line but rather appears to consist of either two lateral fore parts or a central recess, measures 19 feet high by 30 feet long, that is, approximately 5.70 m by 9 m. Neither the height nor probably the width corresponds to the space necessary to store books. It is necessary to take into account the size of the decorative elements, as shown in the illustration accompanying the text. Our information remains scarce and concerns only one specific type of library. We do not actually know what type of library it was: a conservation library or a reading and copying library containing multiple copies of the same works. Dating a few centuries earlier, a list of works kept at the Monastery of Western Brightness in Chang’an around the mid-seventh century offers us further details on Buddhist monasteries. At that time, Chinese Buddhism was flourishing; it was when the eminent pilgrim and translator Xuanzang 玄 奘 (602?–664) returned from India with 657 Sanskrit texts, which he proceeded to translate with the deferential support of Emperors Taizong (r. 626–649) and Gaozong (r. 650–683).12 The Monastery of Western Brightness was founded in 656 to heal the crown prince, who was ill;13 at the same time, a Taoist monas- 12 One of the best depictions of Xuanzang’s life remains the story dedicated to him by Waley 1952. See also Rongxi 1995. 13 Wang Xiang dedicated several studies to the Monastery of the Western Brightness: Wang Xiang 2010, 432–444; Wang Xiang 2015.

Two Libraries of the Tang Capital tery called the Monastery of Eastern Brightness (Dongming si 東明寺) was also founded for good measure. The Buddhist monastery was erected at the former residence of Princess Wanchun 萬春公主,14 later that of Li Tai 李泰 (618–652), son of Emperor Taizong and brother of the reigning emperor, Gaozong. Xuanzang was invited to stay and even take command of this temple, which he refused. However, another famous monk, Daoxuan 道宣 (596–667), did take it over. It quickly became one of the most important monasteries in the western capital.15 Daoxuan was the author of a commented bibliography of the Buddhist canon, a treatise in ten chapters, published in 664.16 While all of the surviving catalogues generally describe the state of the canon at a given point in time, this treatise gives us a list of the works present in the library of the Monastery of Western Brightness, specifying their arrangement on the shelves.17 The books consist of approximately 800 titles and slightly more than 3,360 scrolls, amounting to approximately 50,000 sheets of paper. They are arranged in the following order (see Table 3): sutras of the Greater Vehicle with a single translation; sutras of the Greater Vehicle with multiple translations; sutras of the Lesser Vehicle with a single translation; sutras of the Lesser Vehicle with multiple translations; discipline texts (vinaya) of the Lesser Vehicle; doctrinal works (śāstras) of the Greater Vehicle; doctrinal works of the Lesser Vehicle; and, last of all, compilations and biographies. Missing are discipline texts of the Greater Vehicle, which at the time consisted of approximately 20 titles across 31 scrolls. The wall library consisted of three bays, each bay containing nine shelves. One of these shelves, situated at the lower part of the central bay, was apparently empty; it was probably occupied by the Vinaya of the Great Vehicle. The sutras of the Greater Vehicle with a single translation occupied the six upper shelves, in the centre, whereas the sutras of the Greater Vehicle with multiple translations were stored on the two upper shelves of the left element, and just under the sutras of the Greater Vehicle with a single translation in the central element. The sutras of the Lesser Vehicle with a single translation occupied shelves 3 to 6 of the left bay moving downwards; shelf 6 was completed by a few sutras of the Lesser Vehicle with multiple translations. This category contained 96 titles, but only 114 scrolls; these were sometimes very short works. In this library, a bundle could contain up to 34 works, which implies that several works were likely written on the same scroll, as a bundle generally contained about ten scrolls. Discipline works of the Lesser 14 Princess Wanchun was the sixth daughter of Emperor Gaozu of the Tang (r. 618–626). 15 The remains of this monastery have been studied by Luo 2006, 76–80. The written sources have been explored by Ono 1989, 227–236. 16 Daoxuan, Da Tang neidian lu 大唐內典錄, Taishō shinshū daizōkyō, vol. 55, n° 2149. 17 Daoxuan, chap. 8, 302b–312c; Drège 1991, 212–214.

Jean-Pierre Drège Vehicle were at the bottom of the left element on three shelves, whereas compilations and biographies were placed on the lower shelves of the right element. Each shelf contained 6 to 14 bundles, with the exception of two right shelves that contained 25 and 46 bundles. These small numbers reflect the desire to isolate particularly popular sutras, such as the Flower Ornament Scripture, which at the time had only one translation, that of Bhuddhabhadra (359–429), the Dafangguang fohuayan jing 大方廣佛華嚴經. Other translations followed, in particular that of Śikṣānanda (653–710) at the end of the seventh century. In the case of the shelves with 25 and 46 bundles, we are told that each of these shelves was divided in two. We can therefore guess that the two corresponding compartments were larger than the others, and large enough to be divided into two sections. Each of these could occupy a height of 1.5 times that of the other compartments. That would fill up the supposed empty space in the right stand. Regarding the empty compartment in the middle span, it may have been filled with discipline works of the Greater Vehicle. Table 3: Distribution of the works of the Buddhist canon at the library of the Monastery of Western Brightness; taken from Daoxuan, Da Tang neidian lu, chap. 8: 302b–312c. Sūtra of the Great Vehicle Multiple translations 1 8 bundles Sūtra of the Great Vehicle Single translation 1 6 bundles Śāstra of the Great Vehicle 1 10 bundles Sūtra of the Great Vehicle Multiple translations 2 14 bundles Sūtra of the Great Vehicle Single translation 2 14 bundles Śāstra of the Great Vehicle 2 25 bundles Sūtra of the Small Vehicle Single translation 1 12 bundles Sūtra of the Great Vehicle Single translation 3 11 bundles Sūtra of the Small Vehicle Single translation 2 11 bundles Sūtra of the Great Vehicle Single translation 4 13 bundles Sūtra of the Small Vehicle Single translation 3 13 bundles Sūtra of the Great Vehicle Single translation 5 12 bundles Sūtra of the Small Vehicle Single translation 3; 4 bundles Sūtra of the Great Vehicle Single translation 6 10 bundles Sūtra of the Small Vehicle Multiple translations; 6 bundles EBSCOhost - printed on 2/6/2024 8:47 AM via . All use subject to https://www.ebsco.com/terms-of-use Śāstra of the Great Vehicle 3 16 bundles Śāstra of the Small Vehicle 1 46 bundles Śāstra of the Small Vehicle 2 12 bundles

Two Libraries of the Tang Capital Sūtra of the Great Vehicle Multiple translations 1 8 bundles Sūtra of the Great Vehicle Single translation 1 6 bundles Śāstra of the Great Vehicle 1 10 bundles Vinaya of the Small Vehicle 1 10 bundles Sūtra of the Great Vehicle Multiple translations 3 12 bundles Śāstra of the Small Vehicle 3 9 bundles Vinaya of the Small Vehicle 2 11 bundles Sūtra of the Great Vehicle Multiple translations 4 12 bundles Compilations 1 10 bundles Vinaya of the Small Vehicle 3 7 bundles [Vinaya of the Great Vehicle 10 bundles] Compilations 2 8 bundles The layout of the shelves raises some questions. Is there a hierarchical order that corresponds to a symbolic value of the texts for believers? We would have expected the three elements to correspond to the Tripiṭaka (sūtra, vinaya, and śāstra), even though the number of works in each piṭaka is different. This organisation does not appear to have been carried out for convenience, due to practical requirements. However, to be certain, it would be necessary to know the dimensions of the shelves and that of the library as a whole. The 27 shelves (including the two largest ones) contained 3,361 scrolls (according to the quantitative inventory, or 3,367 according to the breakdown by category), consisting of 326 bundles. It is impossible to know whether the shelves were almost full, just as it is impossible to know the height and width of the library. We can imagine that the bundles, which formed large rolls, were placed horizontally on top of each other. If the cases were approximately 20 cm high, the library would occupy a useful height of approximately 180 to 200 cm across a width of almost 380 cm (approximately 125cm × 3). This is therefore a relatively limited area that could easily be contained in a modest-sized tower (which may not have been the case at the Monastery of Western Brightness). This is even more plausible considering that a manuscript discovered in the famous Dunhuang cave describing the books of another, later monastery library, probably from the ninth century, tells us that the libraries spanned three sides, in other words, three walls – the west, south, and north walls – probably in a room with modest dimensions.18 We could develop various hypotheses and conjectures concerning this library or the one briefly discussed above. Through these two rare examples of libraries for which we have some information, and by relating them to the archives of some Buddhist libraries (lists of Buddhist works in monastery libraries, memoranda 18 Drège 1991, 215–216.

Jean-Pierre Drège of missing manuscripts, lists of manuscripts to be copied, lists of imperfections, etc.) of a rather small town and to a few manuscripts, all discovered in Dunhuang, we can form some idea of seventh- and eighth-century libraries. Many aspects remain largely unknown, such as the functioning of the libraries and, in particular, the circulation of books. These libraries were initially designed as storage places. At the Library of the Academy of the Hall of the Gathered Wise Men, there was a scriptorium where copyists worked, but the works that they copied likely came from elsewhere, and new manuscripts were intended to be added to the book pavilion (shuge 書閣). Books were also borrowed and sometimes lost. The purpose of these libraries was not only to collect books; their best scholars were responsible for discussing the difficulties of the classics and historical texts with the emperor every day, and for writing the emperor’s decrees and the memoirs addressed to him. Regarding the library of the Monastery of Western Brightness, it participated in two essential activities: translation and compilation. While this monastery was admittedly not as important as the Monastery of Great Benevolence (Da ci’en si 大慈恩寺), where the eminent Xuanzang translated the 600 chapters of the Great Perfection of Wisdom Sutra (Da bore boluomi jing 大般若波 羅蜜經), it was home to the translation of a certain number of sutras. In particular, it housed the team of another famous translator, Yijing 義淨 (635–713), who spent his time between the Monastery of Western Brightness and the Great Monastery of Offerings for Happiness (Da jianfu si 大薦福寺), founded in Chang’an in 684.19 We know that several monks from this monastery participated in the translation teams that met at other monasteries. Extensive compilation work was carried out at this monastery in the years after its founding, in particular under Daoxuan’s authority. References Daoxuan 道宣, Da Tang neidian lu 大唐內典錄, in Taishō shinshū daizōkyō 大正大藏經, vol. 55, n° 2149. Drège, Jean-Pierre (1991), Les bibliothèques en Chine au temps des manuscrits (Ve-Xe siècles), Paris: École française d’Extrême-Orient. Drège, Jean-Pierre (ed.), with the collaboration of C. Moretti (2014), La fabrique du lisible. La mise en texte des manuscrits de la Chine ancienne et médiévale, Paris: Collège de France, Institut des hautes études chinoises. 19 At the Monastery of Western Brightness, Yijing signed several important translations: in particular, those of four sutras, three doctrinal texts, and three discipline texts in 703. Zhisheng 智 昇, Kaiyuan shijiaolu 開元釋教錄, Taishō shinshū daizōkyō, vol. 55, n° 2154, 567a–568a.

Two Libraries of the Tang Capital Fölster, Max Jakob (2018), ‘Libraries and Archives in the Former Han Dynasty (206 bce–9 ce): Arguing for a Distinction’, in Alessandro Bausi, Christian Brockmann, Michael Friedrich and Sabine Kienitz (eds), Manuscripts and Archives: Comparative Views on RecordKeeping (Studies in Manuscript Cultures, 11), Berlin: De Gruyter, 201-230. Ikeda, On 池田溫 (1971), ‘Sei-Tō shi shūken in 盛唐之集賢院’, Hokkaidō daigaku bungakubu kiyō 北海道大學文學部紀要, 19/2: 45–98. Li Jie 李誡, Yingzao fashi 營造法式, ed. Wanyou wenku, Shanghai: Shangwu yinshuguan, 1933. Li Linfu 李林甫 (1974), Da Tang liudian 大唐六典, Taipei: Wenhai chubanshe. Liu Xu 劉昫 (1975), Jiu Tangshu 舊唐書 (Old History of the Tang), Beijing: Zhonghua shuju. Lu Yining 盧意寧 (online), ‘Zhang Yue yu Jixianyuan 張說與集賢院’, <www.ihp.sinica.edu. tw/~tangsong/papers/%20lui950731.pdf> (accessed on 30 November 2020). Luo Di (2016), A Grain of Sand. Yingzao fashi and the Miniaturization of Chinese Architecture, PhD, University of South California. Luo Xiaohong 羅小紅 (2006), ‘Tang Chang’an Ximing si kao 唐長安西明寺考’, Kaogu yu wenwu 考古與文物, 2: 76–80. Ono, Katsutoshi 消野勝年 (1989), Chūgoku Zui Tō Chōan jiin shiryō shūsei 中國隋唐長安寺院史 料集成, Kyoto: Hozokan. Ouyang Xiu 歐陽修 (1975), Xin Tangshu 新唐書, Beijing: Zhonghua shuju. Rongxi, Li (1995), A Biography of the Tripiṭaka Master of the Great Ci’en Monastery of the Great Tang Dynasty, tr. from the Chinese of Śramaṇa Huili and Shi Yancong (Taishō, 50), Berkeley: Numata Center for Buddhist Translation and Research. Steinhardt, Nancy Shatzman (1990), Chinese Imperial City Planning, Honolulu: University of Hawai‘i Press. Thilo, Thomas (1997–2006), Chang’an – Metropole Ostasiens und Weltstadt des Mittelalters 583-904. Teil 1: Die Stadtanlage, Teil 2: Gesellschaft und Kultur, Wiesbaden: Harrassowitz. Waley, Arthur (1952), The Real Tripitaka and Other Pieces, London: George Allen and Unwin. Wang Pu 王溥 (1974), Tang huiyao 唐會要, Taipei: Shijie shuju. Wang Xiang 王翔 (2010), ‘Jiqie yu jiezang. Tang Ximing si jingzangqun chutan 集篋與結藏. 唐西明寺 經藏群初探’, in Zengqin 增 勤 (ed.), Shoujie Chang’an fojiao guoji xueshu yantaohui lunwenji 首屆長安佛教國際學術研討會論文集, Xi’an: Shaanxi shifan daxue chubanshe, 432–444. Wang, Xiang (2015), Ximing Monastery. History and Imagination in Medieval Chinese Buddhism, Saarbrücken: Lambert Academic Publishing. Wang Yinglin 王應麟 (1977), Yuhai 玉海, Taipei: Dahua shuju. Wei Shu 韋述, Jixian jizhu 集賢紀注, in Sun Fengji 孫逢吉, Zhiguan fenji 職官分紀, ed. Siku quanshu. Xiao Mo 蕭默 (1987), ‘Mogaoku bihua zhong de fosi 莫高窟壁畫中的佛寺’, in Zhongguo shiku. Dunhuang Mogao ku 中國石窟. 敦煌莫高窟, Beijing: Wenwu chubanshe, vol. 4. Xiong, Victor Cunrui (2000), Sui-Tang Ch’ang’an. A Study in the Urban History of Medieval China, Ann Arbor: The University of Michigan, Center for Chinese Studies. Zheng Weizhang 鄭偉章 (1983), ‘Tang Jixianyuan kao 唐集賢院考’, Wenshi 文史, 19: 65–85. Zhisheng 智昇, Kaiyuan shijiaolu 開元釋教錄, Taishō shinshū daizōkyō, vol. 55, n°2154.

Ivo Smits Institutional Libraries in Japan’s Classic Court Age (Heian Period, 794–1185) Abstract: This paper treats three types of institutional libraries in classical Japan: royal libraries, libraries of private court academies, and monastic libraries. Two issues are explored: (1) whether we can speak of ‘institutional’ libraries or whether all libraries were in essence private libraries; and (2) which works were collected in libraries and which works do not surface in library records. The classic court age of Japan, the so-called Heian period (794–1185), was a period that knew, and then discarded, book printing techniques. It was, as it were, a manuscript age by choice. It was also an age that is primarily remembered as one that produced the first iconic works of Japanese literature, such as The tale of Genji and The pillow-book of Sei Shōnagon. A closer examination of libraries in this age will force us to rethink these perceptions; neither were printed books really absent, nor did narrative fiction in the vernacular (as opposed to Sinitic texts) feature in known book collections. Two library catalogues for the Heian court are extant today and they illustrate this point very clearly. Also, they more or less bookend this era, as one dates from 891 and the other was compiled around 1160. At the same time, they represent the two categories of libraries, the private, and the institutional, that are at the core of this paper and that may largely overlap. This paper will explore two issues: (1) whether we can speak of ‘institutional’ libraries or whether all libraries, including royal libraries, were in essence private libraries; and (2) which works were collected in libraries and which remained invisible in the records that we have. Emphasis will be on libraries in court circles, but I will also touch upon monastic institutions. Perhaps some clarification is needed with regards to the word ‘institutional’. In the context of this paper, I use the term somewhat loosely, to indicate book collections belonging to, or associated with, organisations catering to larger groups of users and created with the expectation to outlive their founders. There are three categories of institutional libraries to consider for the Heian period: [a] royal libraries; [b] libraries of private court academies; and [c] monastic libraries.1 Finally, I make a, admittedly largely artificial, distinction between libraries 1 There is not that much written about libraries and book collection in the Heian period. In EngOpen Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-006

Ivo Smits and archives. My focus is on collections of books belonging to a shared canon, not on records or archives specific to one particular organisation. The distinction is helpful in that we have ample evidence that archives were actively used, but the use made of institutional book collections remains a bit of an enigma. 1 Royal libraries When in 757 the so-called Yōrō code was promulgated in order to regulate much of Japan’s state bureaucracy, one of the many offices that it outlined was the Bureau of Books and Drawings (zushoryō var. fumi no tsukasa 図書寮), part of the Ministry of Central Affairs (nakatsukasashō 中務省).2 Its main role was to act as the state’s library and to preserve books: these were conceptualised in two categories that became dominant right away, namely Buddhist scriptures or ‘inner writings’ (naiten 内典) and non-Buddhist texts or ‘outer writings’ (geten 外典, var. gesho 外書; a category that is often somewhat misleadingly labelled ‘Confucian’). Interestingly, the Bureau also stored Buddhist statues. Its book collection served a purpose beyond mere storage: the Bureau was the designated office for the compilation of histories of Japan. However, more or less coinciding with the last national history, compiled in 901, the Bureau’s role as library became an empty shell in the Heian period and its function of storing and maintaining the court’s book collections shifted to a new office, known as the ‘royal library’ (goshodokoro 御書所). The concept of a royal palace library as initially somehow distinct from the state’s Bureau but gradually conflating the distinction between the emperor’s and the state’s book collections is probably best understood as symbolic of how the abstract idea of state and the more concrete sense of ‘the royal’ were intertwined. The umbrella notion of a royal library in practice was split up in three different palace libraries, each with their own functionaries.3 These were the ‘royal library’ proper, the ‘inner royal library’ (uchi no goshodokoro 内御書所), and the lish, there is a concise overview in Peter Kornicki’s landmark study The Book in Japan; Kornicki 1998, 364–370. In Japanese, as Kornicki rightly points out, there still is nothing that surpasses the comprehensive inventory of libraries by Ono Noriaki; Ono 1944, 23–416 (this includes the eighth century). Two of the private libraries mentioned in this paper receive scrutiny in Smits 2009. 2 For the zushoryō, see Ono 1944, 42–65; Kornicki 1998, 365. The Yōrō code, already drafted in 718, was a revision of the Taihō code of 702, which in turn was based on a Chinese code from 651. 3 For the goshodokoro, see Ono 1944, 93–97; Nagata 1988. Oddly, Ono Noriaki pays little attention to the goshodokoro; instead his focus is essentially on the zushoryō.

Institutional Libraries in Japan’s Classic Court Age ‘one-copy library’ (ippon goshodokoro 一本御書所). The inner royal library seems to have retained more explicitly the notion of being the private library of the emperor. The ‘one-copy library’ was exactly that: a palace library for rare books, of which only one copy was known to exist in the country. Frustratingly, it is not very clear how these royal libraries were used. Were they primarily repositories of texts whose primary function was to simply exist in the royal collections? Who had access to these collections and for what purposes? The fact that shadow collections existed, in the form of collections owned by for example the court academy (daigakuryō 大学寮) and its satellite institutes, suggests that royal libraries were not necessarily ‘public’ and that they were not the first port of call for scholars. However, like the Bureau of Books and Drawings, the royal libraries did seem to have a connection to the production of new state sponsored texts. For example, the main compiler of the first royal anthology of Japanese language verse, Kokin wakashū 古今和歌集 (Collection of Poems Old and New, 914), was Ki no Tsurayuki 紀貫之 (d. 945). The preface to this poetry collection explicitly mentions his function as ‘keeper of royal books’ (gosho no tokoro no azukari 御書所預).4 2 Private academies and their libraries The eight and early ninth centuries show some examples of privately founded institutions that one might call public schools. An early one, albeit pre-Heian period, is the Untei-in 芸亭院 (Camphor Pavilion Hall) established in c. 771 by Isonokami no Yakatsugu 石上宅嗣 (729–781), in the eighth-century capital Heijō-kyō (present-day Nara), and which was still in existence there in 797. Ostensibly, the Untei-in aimed to foster both Buddhist and ‘Confucian’ (or rather: Sinitic) studies, that is to study both ‘inner’ and ‘outer writings’. The governing idea was that the study of Chinese classic texts facilitated an informed reading of Buddhist texts. According to Yakatsugu’s biography in Shoku Nihongi 続日本紀 (Annals of Japan, Continued, 797), he converted his home to a temple, and there he organized a 4 Kokin wakashū, Kana no jo, 29; Nagata 1988, 363. Yet unlike the state histories for which scholarly editors extracted their information from archival records in the royal libraries, it is not at all clear that Tsurayuki and his colleagues found their material in the same libraries. This may have to do with the question of which categories of texts were collected in the royal libraries, see below.

Ivo Smits school that seems to have functioned as a privately operated public library of sorts, with some five hundred scrolls in its collection.5 One such school that did operate in the Heian period, is the Shugeishuchi-in 綜芸種智院 (Hall of Various Worldviews in an Enlightened Mind), which was in function in the years 828–845 and which was founded by the monk Kūkai 空海 (774–835) two and a half decades after his sojourn in Tang China. This school, too, trained in both ‘inner’ and ‘outer’ texts. Presumably, but this is hard to verify, Kūkai’s school also provided its students access to texts.6 What the Untei-in example shows, is that early on there existed in Japan a notion that libraries could exist outside the infrastructure of court institutions, even if such schools, at least some of which came with libraries of sorts, did not outlive their founders for long. Also, these institutions aimed at great accessibility. Yakatsugu’s library was reputedly Japan’s ‘first public library’.7 Of Kūkai’s school it is usually emphasised that it catered explicitly to ‘the general masses’ (ippan minshū 一般民衆). Nevertheless, chances of survival for libraries were greater if they did somehow connect to institutions of the court, of which the court academy is one example. The private and semi-private schools of the world of the Heian court that outlasted their founders were all tied to the court academy. The earliest examples of private tutorial schools that helped students at the academy prepare their curriculum date from the seventh century,8 but the ninth century, that is the early Heian period, saw quite a rise in them. Not surprisingly, such private academies (shijuku 私塾) were founded by scholar families with vested interests in Sinitic learning. Several presented themselves in part as cram schools for students at, or hoping to get into, the court academy. Hence, some are referred to as ‘academy boarding 5 See Yakatsugu’s biography and the regulations (jōshiki 条式) for the Untei-in in Shoku Nihongi, book 36 (sixth month of Ten’ō 1 [781]), vol. 16, 200–201. See also Hisaki 1968, 169–171; Kornicki 1998, 368. ‘Un’ in Yakatsugu’s library’s name is a term for fragrant herbs that help to keep insects out of books. Another eighth-century private school is the Nikyō-in 二教院 (Two Teachings Hall), established by the scholar of Chinese literature Kibi no Makibi 吉備真備 (695–775) in 769. This school, too, explicitly fostered the study of Buddhist as well as non-Buddhist Sinitic texts. Makibi had spent time in China, and seems to have catered especially to students wishing to study Chinese classics. 6 For Kūkai’s regulations for his school, see his Shugeishuchi-in shiki 綜芸種智院式, Shōryōshū 10, no. 102, 420–423 (esp. 422, 423). See also Ono 1944, 135–138; Hisaki 1968, 173. Kūkai’s text explicitly mentions Untei-in and Nikyō-in as two bad examples of schools that eventually folded. 7 Shoku Nihongi, vol. 16, 200, 201, 562 (note 51); Hisaki 1968, 169–171; Kornicki 1998, 368. 8 Minabuchi no Shōan 南淵請安 (dates unknown) was a scholar-monk who went to study in Sui China in 608. After his return to the Japanese court in 632, he established a private academy where he taught young nobles.

Institutional Libraries in Japan’s Classic Court Age schools’ (daigaku bessō 大学別曹), following official recognition of their status by the court academy. Several private academies became highly regarded educational institutions in their own right and would also train men (never women) who failed to get into the court academy but might hope for appointment in lower ranking positions in the court bureaucracy.9 In addition, there was a number of private schools not formally recognised as a court academy annex but certainly valued by court scholars. The best known of these was undoubtedly the private academy operated by the Sugawara family, called San’intei 山陰亭 (Mountain Shade Pavilion).10 The scholar, statesman, and poet Sugawara no Michizane 菅原道真 (845–903) described its setup in his Note on My Library, or Record of My Study (Shosai no ki 書斎記) in 893: In Sempū Ward in the eastern sector of the capital there is a house, in the southwest corner of the house is a corridor, and at the very south end of the corridor is a room. The room is hardly more than ten feet square, so that those entering or leaving have to squeeze past each other, and those occupying the room have to sit with mats pushed together side by side. And yet out of this room in the past have come close to a hundred men who passed the shūsai and shinshi examinations. […] I shifted the blinds and mats around and straightened up the room, and brought my books and stacked them away there.11 We must assume that all private academies stored books to which their instructors and students had access. In some case we know as much, since their existence is recorded. 3 Private libraries A related issue is the existence of private libraries. We know of several of these. In some instances, specifically in the case of scholars, the distinction between a private library and that of a private institution is hard to draw and arguably meaningless. The 9 For an overview of the academy boarding schools and private academies, see Momo 1947, 170– 243, 449–471; Hisaki 1968, 175–180. 10 Its more informal name was Kanke Rōka 菅家廊下 (Sugawara’s Corridor). Momo Hiroyuki has pointed out that the name Kōbaiden 紅梅殿 (Red Plum Hall), the name by which the school is often referred to, was strictly speaking the name of the entire living compound of the Sugawara family, not of the school that occupied only part of it. Momo 1947, 460; Hisaki 1968, 176–177. 11 Kanke bunsō no. 526, 535; trans. Watson 1975, 104, 105. The shūsai and shinshi examinations qualified a candidate for civil service. Momo Hiroyuki suggests that Michizane may be describing only the headmaster’s study, and that the Sugawara school and library were bigger. Momo 1947, 465.

Ivo Smits San’intei of the Sugawara family mentioned above is one such example, which was open to students and scholars associating with the Sugawara, as is probably also true for the library of the bureaucrat and scholar-poet Ōe no Masafusa 大江匡房 (1041– 1111) and his Library of the Ōe Family (Gōke Bunko 江家文庫), which at the time of its destruction by fire in 1153 contained tens of thousands of volumes.12 In fact, if we can believe legend, Masafsua himself saw his massive library in a holistic relationship with the state, and its loss was certainly widely bemoaned at the time.13 The situation is clear for the libraries of wealthy power-brokers. These were undeniably private libraries. The statesman Fujiwara no Michinaga 藤原道長 (966– 1027), for instance, was an avid book collector with an impressive library of over two thousand volumes.14 He was keen to collect printed editions of the works of the famous Chinese poet Bai Juyi and of the equally canonical Selections of Refined Literature (Wenxuan 文選), which he obtained either through Chinese merchants or through Japanese monks abroad.15 The same is true for Michinaga’s descendent Fujiwara no Yorinaga 藤原頼長 (1120–1156), who in 1145 finished the building of his Uji library (fumigura var. fugura 文庫, or fudono 文殿), located in the northeast of the capital. Until then he had stored his books in special wheeled chests (fumiguruma var. fuguruma 文車), which could easily be moved in the event that the house caught fire. It had taken him less than three months to build the storehouse for his books. It is one of the few libraries for which we have a description of its physical properties. The structure measured nearly seventy-eight square meters and was fenced off by a very thick earthen wall. The books were stored in numbered boxes which were shelved on book racks, five shelves on the east of the storehouse and six on the west side. The books were divided into four categories: study material such as the Confucian classics; Chinese historical works; miscellaneous Chinese texts; and, finally, Japanese books.16 The latter will have consisted of books in Sinitic, 12 Ono 1944, 331–335; Kornicki 1998, 369. Of course, the ‘ten thousand’ and ‘tens of thousands’ mentioned in historical sources is an extraordinary number and may simply mean ‘a lot’. For the existence of private libraries in Heian Japan, see Kornicki 1998, 369–370. 13 Zoku kojidan 2.41, 659. See there also for quotations from several journal entries from 1153 and the quasi-official history Hyakurenshō. Masafusa supposedly said that ‘as long as the country of Japan remains, these books will be preserved’. 14 Midō kanpaku ki Kankō 7 (1010).8.29, vol. 2, 74. For a translation, see Hérail 1987–1991, vol. 2, 423. 15 See for example Midō kanpaku ki Kankō 3 (1006).10.20; Kankō 7 (1010).11.28; Chōwa 2 (1013).9.14; and Chōwa 4 (1015).7.15, vol. 1, 196; vol. 2, 82, 243; vol. 3, 20. 16 Taiki, Kyūan 1 (1145).4.2 and 14, vol. 23, 149. See also Hashimoto 1964, 55–58, who provides a reconstructed plan of the library building. Yorinaga gives the storehouse’s measures as c. 7 by c. 3.3 meters, with a height of c. 3.6 meters. The building was located at the northwest corner of Yorinaga’s villa compound.

Institutional Libraries in Japan’s Classic Court Age such as histories of Japan, diaries in Sinitic, and family records, not works in Japanese. The book collection of Yorinaga’s contemporary Fujiwara no Michinori 藤原 通憲 (1106–1159) also contained only Sinitic titles. This we know from the extant catalogue of Michinori’s library that was compiled around 1160. This Tsūken nyūdō zōsho mokuroku 通憲入道蔵書目録 (Catalogue of the Library of the Lay Priest Michinori) provides fascinating insights in the reading habits of a member of the high nobility, as does Yorinaga’s journal.17 The extent to which such wealthy book collectors were willing to open up their libraries to others was probably slim, but it did happen. We do have reports of lesser nobles seeing precious books in such elite private libraries, and in the case of Yorinaga we know that he had intensive contact with scholars and academy students at his home.18 4 Visible and invisible books It is customary among book scholars to distinguish three categories of books for this period, namely ‘Chinese books’ (kanseki 漢籍; sometimes these were printed books, koshōbon 古鈔本); ‘Buddhist canonical texts’ in Chinese translation (kyōten 経典, var. butten 仏典); and ‘Japanese books’ (wahon 和本 or kokusho 国 書), including collections of poetry in Japanese (waka 和歌). The striking thing is that books in Japanese remain invisible in library infrastructures, until the twelfth century. For monastic libraries treated below this is explained by the fact that canonical texts and their commentaries were always written in Sinitic. For the more worldly libraries it must be assumed that they aimed to present an emphatically scholarly typology of knowledge, which did include Sinitic verse and prose but nothing in Japanese. The first extant Heian book catalogue is representative in that respect. Nihonkoku genzaisho mokuroku 日本国見在書目録 (Catalogue of Books Extant in Japan, 891) is in fact a list of books imported from China, rather than an attempt to also cover books produced in Japan. The few Japanese works in Sinitic in this list seem to have been included by accident.19 The categories that organise this inventory draw on Chinese dynastic bibliographies and the rhetorical effect of this valuable catalogue is to present ‘Chinese books’ as signs of state power. 17 Smits 2009. 18 Togawa 1998, 66. 19 Kornicki 1998, 422–423.

Ivo Smits Of course, there were books in Japanese around and they were collected, but one does not find them in records of book collections, but glimpses them in more personal sources. The author of The Tale of Genji, Murasaki Shikibu 紫式部 (c. 975–after 1019), for example, in her memoirs notes that she does have books in her living quarters within the royal palace. There is also a pair of larger cupboards crammed to bursting point. One is full of old poems and tales that have become the home of countless insects which scatter in such an unpleasant manner that no one cares to look at them anymore; the other is full of Chinese books that have lain unattended ever since he who carefully collected them has passed away.20 Books in Japanese indeed tended to be collections of and manuals for poetry, or fictional tales and anecdotal collections; and while the first had gained social standing since the early tenth century, the latter category did not. Such books were read, even avidly, but not given a visible place in libraries. Likewise, much if not most of the information about available book titles and their uses comes from personal journals and may also be gleaned from the many ‘classified books’ (ruisho 類書), that is, huge compendia that categorised knowledge gleaned from other sources and which one may think of as libraries in book format. 5 Monastic libraries In 1094, the monk Eichō 永超 (1014–1095) compiled a catalogue of Buddhist writings available in temples in Nara and Kyoto and their surroundings, Tōiki dentō mokuroku 東域伝灯目録 (Catalogue of [Works that] Transmit the [Dharma] Lamp in the Eastern Regions). That he could do so indicates that scholarly monastic libraries were catalogued at certain intervals and that they could be consulted by visiting monks from other monasteries.21 Eichō (1014–1095) lists 78 sutra categories for known religious treatises and sutra commentaries by a variety of East Asian monks. Often mentioned are the catalogues of Buddhist books brought over from the Chinese mainland by eight Japanese monks throughout the ninth 20 Murasaki Shikibu nikki, 497; trans. Bowring 1996, 55. See also Kamens 2007, 129. Although Murasaki was regularly reading these ‘Chinese’ books, they were not collected by herself, even though she was trained to read such books from an early age, but by her late husband. 21 Inoue 1948; see also Kornicki 1998, 420–421. For more catalogues of monastic libraries, see Ogawa 2009, 50–51. For their relationship to ideal sets of a complete Buddhist canon, see also Ochiai 1998.

Institutional Libraries in Japan’s Classic Court Age century and collectively referred to as ‘catalogues [of books] brought over [from China]’ (shōrai mokuroku 諸来目録). Together they underline the emphatically East Asian, rather than Japanese, context of a living Buddhist corpus of texts. Heian monastic libraries and their scriptoria have been little studied. Unlike the eighth and early ninth centuries, for which we have in recent years accumulated considerable information,22 the world of late ninth through twelfth centuries remains rather a closed book when it comes to the collections and their usage of the many monasteries that existed in Heian Japan. The assumption is that by and large monastic libraries functioned as they did in the eighth century, although we cannot be sure. Obviously, there were monks who wrote often extensive treatises and commentaries on religious subjects. For these they had access to large stores of Buddhist writings; they must have found these repositories in either their own or other monastic libraries. Occasionally we hear of scholarly monks lending each other books from their own collections. Japan’s early monastic libraries are credited with instituting a special protective device: they allegedly imported cats from China as early as the eighth century, to protect books from damage by mice and rats. Incidentally, it is in the monastic libraries that we find printed books manufactured in Heian Japan. The so-called ‘temple prints’ (ji’inban 寺院版) from the eleventh and twelfth centuries contained religious texts, usually sutras, printed by temples. This printing was done in commission, at the behest of wealthy patrons within the court nobility.23 These printed texts were stored in the monasteries’ sutra repositories (kyōzō 経蔵); their function was not to be read, but to result in the accumulation of merit (kudoku 功徳) that came from the reproduction of scriptures. The printing of sutras was in essence a deluxe version of the long-standing practice of sutra copying (shakyō 写経). 6 The book as object When in 757 the Bureau of Books and Drawings was established, it was stipulated that one of its duties was the manufacturing of paper, ink, and writing brushes. The reason for this was the close relationship between places of storage and places of production of books; the Bureau, then, provides one more example of the fairly 22 In English, see for example Lowe 2014a and 2014b. Peter Kornicki, on the other hand, is more sober about our knowledge of eighth-century monastic libraries. Kornicki 1998, 367. 23 Nojiri 2015, 9–10.

Ivo Smits universal overlap of bibliotheca and scriptorium. One of the Bureau’s important tasks throughout the ninth century was to copy sutras at the behest of the royal family and is close entourage.24 Increasingly, especially after the Bureau of Books and Drawings became defunct, monasteries throughout the country became the place where sutras were copied, often in a lavish execution, for karmic merit but also in the continuing endeavour to obtain complete sets of an ever-dynamic Buddhist ‘canon’ (issaikyō 一切経).25 The copying of non-religious texts was also a constant, but on a much smaller scale and the extent to which this happened in an institutionalised setting is unclear; anecdotal evidence from journals suggest that this happened very much as a personal initiative by individuals. The bulk of book production (that is, copying of books) in Heian Japan concerned religious texts, specifically sutras; books were as much, if not more, powerful objects than they were texts for reading.26 7 Conclusion To sum up, in the classic court age of Heian Japan the institutional libraries of the court seem in the first place to have been symbolic repositories of knowledge that spoke of the state’s omniscience and less hubs of readership and study. Active access to that knowledge was not their purpose. Something similar may, on occasion, have been the case with several monastic libraries. The one obvious exception, that is, repositories of books as part of a shared cultural canon that was actively read and studied, were the academic book collections with official and quasi-official status. Many individuals possessed books, sometimes large amounts of them, and read avidly. Many readers were writers, too. Yet intriguingly, the records we have of institutional libraries do not tell such readers’ histories. 24 Ono 1944, 57–60. The copying of sutras at the behest of royals had in the eighth century been outsourced to the Office of Sutra Transcription (shakyōjo 写経所), but this semi-official government bureau ceased to exist with the move to the new capital in 794 and this responsibility returned to the Bureau of Books and Drawings (zushoryō). For the shakyōjo, see Lowe 2017, 122–134. 25 Nojiri 2015, 7–9; Ochiai 1998. 26 The ritual efficacy of a book as powerful object is proven, among others, by the Heian practice of burying sutras at sacred sites. See Moerman 2007; Nojiri 2015, 9.

Institutional Libraries in Japan’s Classic Court Age References Bowring, Richard (trans.) (1996), The Diary of Lady Murasaki, London: Penguin. Hashimoto, Yoshihiko [橋本義彦] (1964), Fujiwara no Yorinaga 藤原頼長, Tokyo: Yoshikawa Kōbunkan. Hérail, Francine (trans.) (1987–1991), Notes journalières de Fujiwara no Michinaga, ministre à la cour de Heian (995–1018), Geneva: Droz. Hisaki, Yukio [久木幸男] (1968), Daigakuryō to kodai jukyō 大学寮と古代儒教. Tokyo: Saimaru Shuppankai. Inoue, Mitsusada [井上光貞] (1948), ‘Tōiki dentō mokuroku yori mitaru Nara jidai sōryō no gakumon’ 東域伝灯目録より見たる奈良時代僧侶の学門, in Shigaku zasshi 史学雑誌, 57/3:27–48 and 57/4: 23–45. Kamens, Edward (2007), ‘Terrains of Text in Mid-Heian Court Culture’, in Mikael S. Adolphson, Edward Kamens, and Stacie Matsumoto (eds), Heian Japan, Centers and Peripheries, Honolulu: University of Hawai‘i Press, 129–152. Kanke bunsō 菅家文草, ed. Kawaguchi Hisao 川口久雄 (Nihon koten bungaku taikei 日本古典文 学大系, 72), Tokyo: Iwanami Shoten, 1966. Kokin wakashū 古今和歌集, Ozawa Masao 小沢正夫 and Matsuda Shigeho 松田成穂 (eds) (Shinpen nihon koten bungaku zenshū 新編日本古典文学大系全集, 11), Tokyo: Shōgakukan, 1998. Kornicki, Peter (1998), The Book in Japan: A Cultural History from the Beginnings to the Nineteenth Century, Leiden: Brill. Lowe, Bryan D. (2014a), ‘Contingent and Contested: Preliminary Remarks on Buddhist Catalogs and Canons in Early Japan’, Japanese Journal of Religious Studies, 41/2: 221–253. Lowe, Bryan D. (2014b), ‘Buddhist Manuscript Cultures in Premodern Japan’, Religion Compass, 8/9: 287–301. Lowe, Bryan D. (2017), Ritualized Writing: Buddhist Practice and Scriptural Cultures in Ancient Japan, Honolulu: University of Hawai‘i Press. Midō kanpaku ki 御堂関白記. Dai Nihon kokiroku 大日本古記録: Midō kanpaku ki, 3 vols, Tokyo: Iwanami Shoten, 1952–1954. Moerman, D. Max (2007), ‘The Archaeology of Anxiety: An Underground History of Heian Religion’, in Mikael S. Adolphson, Edward Kamens, and Stacie Matsumoto (eds), Heian Japan, Centers and Peripheries, Honolulu: University of Hawai’i Press, 245–271. Momo, Hiroyuki [桃裕行] (1947), Jōdai gakusei no kenkyū 上代学制の研究, Tokyo: Meguro Shoten. Murasaki Shikibu nikki 紫式部日記, ed. Ikeda Kikan 池田亀鑑 and Akiyama Ken 秋山虔検雄 (Nihon koten bungaku taikei 日本古典文学大系, 19), Tokyo: Iwanami Shoten, 1958. Nagata, Kazuya [永田和也] (1988), ‘Goshodokoro to uchi no goshodokoro’ 御書所と内御書所, Kokugakuin Daigaku Daigakuin kiyō 国学院大学大学院紀要, 20: 354–379. Nojiri, Tadashi [野尻忠] (2015), ‘Gairon’ 概論, in Nara Kokuritsu Hakubutsukan [奈良国立博物館] (ed.), Maboroshi no Kunōji-kyō ni deau: Heian kokyō ten まぼろしの久能寺経に出会う：平 安古経展, Nara: Nara Kokuritsu Hakubutsukan, 6–15. Ochiai, Toshinori [落合俊典] (1998), ‘Heian jidai ni okeru nyūzōroku to shōso mokuroku ni tsuite’ 平 安時代における入蔵録と章疏目録について, in Ochiai Toshinori (ed.), Nanatsudera koitsu kyōten kenkyū sōsho 七寺古逸経典研究叢書, vol. 6, Tokyo: Daitō Shuppansha, 471–495. Ogawa, Takeo [小川剛生] (2009), Chūsei no shomotsu to gakumon 中世の書物と学問, Tokyo: Yamakawa Shuppansha.

Ivo Smits Ono, Noriaki [小野則秋] (1944), Nihon bunkoshi kenkyū (jōkan: jōdai, chūsei hen) 日本文庫史 研究 上巻 (上代中世篇), Kyoto: Rinsen Shoten (reprint 1988). Shoku Nihongi 続日本紀, Shin Nihon koten bungaku taikei 新日本古典文学大系, 12–16, ed. Aoki Kazuo 青木和夫 et al., Tokyo: Iwanami Shoten, 1989–2000. Shōryōshū 性霊集, in Nihon koten bungaku taikei 日本古典文学大系, 71, ed. Watanabe Shōkō 渡邊照宏 and Miyasaka Yūshō 宮坂宥勝. Tokyo: Iwanami Shoten, 1965. Smits, Ivo (2009), ‘China as Classic Text: Chinese Books and Twelfth Century Japanese Collectors’, in Andrew Edmund Goble, Kenneth R. Robinson, and Haruko Wakabayashi (eds), Tools of Culture: Japan’s Cultural, Intellectual, Medical and Technological Contacts in East Asia, 1000s–1500s, Ann Arbor: Association for Asian Studies, 185–210. Taiki 台記, Zōho shiryō taisei 増補資料大成, 23–25, Kyoto: Rinsen Shoten, 1982. Togawa, Tomoru [戸川点] (1998), ‘Inseiki no daigakuryō to gakumon jōkyō: Fujiwara no Yorinaga no jiseki o chūshin ni’ 院政期の大学寮と学問状況：藤原頼長の事績を中心に, in Fukutåo Sanae [服藤早苗] (ed.), Ōchō no kenryoku to hyōshō: gakugei no bunkashi 王朝の権力と 表象：学芸の文化史, Tokyo: Shinwasha, 53–86. Watson, Burton (trans.) (1975), Japanese Literature in Chinese, vol. 1, New York: Columbia University Press. Zoku kojidan 続古事談, Kawabata Yoshiaki 川端善明 and Araki Hiroshi 荒木浩 (eds) (Shin Nihon koten bungaku taikei 新日本古典文学大系, 41), Tokyo: Iwanami Shoten, 2005.

Gérard Colas Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century: The Sarasvatī Library in Chidambaram Abstract: Palm-leaf as a medium of writing occupies a major place in ancient India because of the long period of its use. Its vulnerability however required frequent copying. It seems to have been initially used to preserve and disseminate administrative and political documents; its use for transmitting literature, particularly religious literature, is accessory, late and even accidental. Libraries of manuscripts, archives, mnemonic collections (mnémothèques), and epigraphic collections (lithothèques) are four entities with various kinds of interrelationship. Two Tamil inscriptions from the thirteenth century, which attest to the activities and transformations of the Sarasvatī Library in the temple of Naṭarāja in Chidambaram, illustrate connections between these four entities. 1 Palm-leaf manuscripts, writing, and various categories of textual storing spaces This paper will mainly discuss religious libraries of palm-leaf manuscripts in Southern India. These manuscripts were made from palm leaves which, after being processed and cut into an oblong format, were a medium for engraving or writing with ink (see Fig. 1). On the Indian subcontinent, the leaves of two species of palms were used for this purpose.1 This use of palm leaves started at least as early as the century prior to the Christian era, and lasted up until the twentieth century.2 The long-standing and continuous use of palm-leaf manuscripts and their role as a historical model of the Indian book make them a major benchmark with regard to the numerous other writing and printings mediums used in India.3 When collected into 1 For more details on palm leaves, see Colas 2017. 2 To compare, paper manufacture appears to have started during the eighth century on the subcontinent, more specifically in the Kashmir region; the Chinese-Nepalese path of this technique is another possibility (see Janert 1995, 75–78). 3 Palm-leaf manuscripts were the model for the book format called pothī (a feminine term from the Northern Indian subcontinent denoting an oblong book). Regarding other manuscript mediums, see Janert 1995, 38–87. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-007

Gérard Colas a bundle, palm leaves formed a manuscript. They were pierced with one or multiple holes so that a string could be passed through to bind them. When the bundle of leaves was thick, a stick would be passed through a hole instead of string. The manuscript was protected by cover leaves or by boards that were often made of wood, sometimes of pieces of dried palm stalks, or even of ivory or other materials. Even though there were perhaps numerous manuscript libraries in pre-modern India, the historical data about them remain relatively uncertain.4 This is because the age of extant Indian manuscripts – large numbers of which still exist today – often does not date back to before the sixteenth century, though, of course, older manuscripts do exist.5 This chapter does not seek to establish a rigorous classification of manuscript libraries in India or to trace a history that could but be superficial. I will simply point out that a wide variety of manuscript libraries existed in the contexts of highly different social and intellectual practices: the libraries of Buddhist universities and monasteries, Jain libraries, ritual and technical libraries (often family ones), the sometimes important libraries of scholars,6 archives, village libraries, the libraries of Vedic colleges, of Hindu temples and monasteries, and yet probably other kinds of libraries.7 The observations that follow will be limited mainly to religious libraries in Southern India, and more specifically to that of Chidambaram temple in the Tamil country. It is nevertheless important first to understand the Indian palm-leaf manuscript tradition from its specific perspective. Whereas in other scholarly cultures in the world, the concept of the age of a manuscript relates to a material document whose conservation spans centuries, this is most often not the case for a palm-leaf document. This is because palm leaves had to be constantly re-copied, given their exposure to humidity, rodents and insects, even though in ancient India attempts were made to postpone the natural decline of palm-leaf manuscripts through different means, such as coating the leaves with insect-repellent oils. The duration of a palm-leaf manuscript could thus vary from a few years to a 4 Compared to those of the ancient Chinese world, for example; see the contribution of JeanPierre Drège in this volume. 5 See the overview article by Wujastyk 2014. 6 Such as that of the famous library of Kavīndrācārya Sarasvatī (seventeenth century) (primarily paper manuscripts), that of the Deśamaṅgalam Vāriyar family in Kerala (primarily palm-leaf manuscripts), or the large anonymous library today known as the Chandra Shum Shere Collection brought to the Bodleian Library at the beginning of the twentieth century (primarily manuscripts on paper). 7 In addition to Wujastyk 2014, see in particular Sircar 1996, 100–102; Cort 1995 (on Jain libraries); Minkowski 2010 (on the relationship between the manuscripts of technical texts and literary texts in certain libraries).

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century century, but not much longer, exceptional circumstances aside.8 It was only from the eighteenth century that people became concerned with conserving them for longer durations, influenced in particular by the Western museographic notion of the manuscript. The cultural value of palm-leaf varied, depending on the context, in a way that is paradoxical in appearance only. Palm-leaf was a medium for conserving legal and political documents (along with other materials), but erudite traditional Brahmans disdained or underrated it for religious texts in comparison with the authoritative speech; as a writing material accepted nonetheless, it was long considered superior to paper in Southern India; as a medium for texts transmitted from person to person (especially from master to disciple), it resisted the pull of printed texts which were disseminated without regard for the reader’s qualification. Economic reasons also postponed the transition from manuscripts to printing: in India, the manuscript era lasted until the twentieth century, therefore long after the introduction of the printing press by European missionaries in the mid-sixteenth century.9 Historically, writing was probably primarily a practical communication tool in Indian society. It served to publicize political and administrative decisions applicable to all individuals, regardless of their moral, social, or religious status, as illustrated by the edicts of Aśoka engraved during the third century bce and read aloud by public announcers. Writing recorded religious donations, contracts, or accounts on media with a varying lifespan, thus perpetuating legal and administrative decisions. Buildings were devoted to keeping the administrative documents of states. The political treatise Arthaśāstra, estimated to date back to between the fourth century bce and the fourth century ce, recommends that kings build an archive to hold their accounts books.10 These books listed the activities of the different government departments, wages for production workshops, the price of consumer goods, laws and customs, payments and receipts in their relationship with the friends and enemies of the king, and so on. The Arthaśāstra does not appear to expressly mention the function of an archivist, but the kāraṇika, who seems to have been the king’s accountant, partially fulfilled 8 The oldest palm-leaf manuscripts today conserved in Southern India appear to date back to the twelfth century: see Colas 2017, 119–121. 9 See Schurhammer and Cottrell 1952. 10 This building is called nibandhapustakasthāna, which R.P. Kangle translates as the ‘Records Office’ (Kangle 1963, 92); translation by Patrick Olivelle: ‘bureau of official records’ (Olivelle 2013, 111). See Arthaśāstra, ed. Kangle 1960, 2.7.1–2; tr. Kangle 1963, 92–93; tr. Olivelle 2013, 111–112.

Gérard Colas this role: he was required to keep the account books, at least over a certain period of time, in order to present them to the king.11 Up until recently, the conservation of archives on palm-leaf continued to exist on a small scale in family libraries, where ownership deeds recorded on palm leaves sat side-by-side with manuscripts of texts.12 It also developed on a large scale at the archives of Hindu temples. A more lasting and prestigious medium of recording these temple archives was the inscriptions which covered the temple walls (see Fig. 2).13 These inscriptions form veritable epigraphic collections whose contents are often comparable to that of palm-leaf archives. They can be called, in French, lithothèques to highlight the particularity of their material in contrast with palm-leaf.14 Carved by engravers who were at times illiterate, they reproduced and perpetuated, on stone, documents that had previously been written on palm-leaf15 and kept in temple archives.16 Stone inscriptions were intended to be read out loud for and in front of a sometimes illiterate audience.17 They concerned the donations made to the temple (lands from which the income allowed for the maintenance of certain acts of worship, or of a garden, of staff members and so on). When engraved on the walls of sanctuaries of a regional importance, they 11 See Arthaśāstra, ed. Kangle 1960, 2.7.22 and 2.7.34, as well as the discussion by Olivelle 2013, 518. 12 See Sarma 2012, and in particular the top photograph on Plate XVII in that same volume, here Fig. 1. I thank Mr S.R. Sarma and Brill Publishers for having allowed me to reproduce this photograph in the present volume. 13 One may also mention copper plates, another long-lasting, but transportable medium frequently used for legal deeds and kept in family as well as temple libraries: see Sircar 1996, 103–160; Salomon 1998, 113–115. Naturally, Indian epigraphy, in both Southern and Northern India, also includes genres of texts other than legal, some of which are literary; on the variety of engraved texts, see Salomon 1998, 110–126. 14 I wish to thank Jean-Pierre Drège for having pointed out the French term ‘lithothèque’ to me. The use of the English term library or French bibliothèque for epigraphic inscriptions may be considered inappropriate given that these terms specifically and etymologically mean ‘book deposit’. 15 As we can surmise with respect to the Tamil country. Other mediums may have been used as intermediaries in other parts of India and at other times. For more details on the transfer of text onto stone and on epigraph engravers, mainly in Northern India, see Sircar 1996, 85–88; on the process, see also Salomon 1998, 65–66. 16 According to V.G. Ramakrishna Ayyar, based on Epigraphical Reports (1913), 298, 299, 306 and 309, ‘Original documents, pertaining to gifts of lands made to the temple were preserved in the treasury of the temple and engraved on its walls’ (V.G.R. Ayyar 1932, 41). Here, ‘treasury’ probably translates a term denoting a place for storing archives and objects of value (Sanskrit bhāṇḍāra, Tamil paṇṭāram). 17 The audience for the oral reading of stone inscriptions was most likely similar to that of the readers and listeners of the inscriptions on copper plates: inhabitants of the village, administrative officials relaying royal orders on-site, etc. See Sircar 1996, 162–169.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century also recorded legal deeds and royal decrees of public interest, at times in relation also to the surrounding villages. The temple’s epigraphic collection, which had a close functional relationship with that of its palm-leaf archive library, therefore had an almost notarial dimension. At times, the inscriptions also contained poems, such as praises of monarchs. While palm-leaf, just like its epigraphic copies, played a leading role for keeping legal, administrative, and political documents, it had only a secondary role in transmitting religious texts, as was the case with all other physical writing media. Certain Indian textual collections long remained exclusively oral and mnemonic, to the extent that they could even be called mnémothèques in French, that is, ‘mnemonic collections’. While writing had a testimonial value in everyday affairs (including the concrete management of religion), it had none compared with the speech of an authority delivering religious and spiritual texts: in this domain, what prevailed was the principle of oral transmission from a spiritually, intellectually, or socially authorized person to a select audience. This authority of the transmitter and trust in his reliability were fundamental for the conservation, understanding, but also at times the emendation and updating of the text. The example of the ancient Vedic corpus, the core of which could date back to approximately 1500 bce, is significant. Its oral transmission has been maintained almost perfectly to date. Eleven different but complementary styles of recitation have enabled a quality of transmission which no manuscript transmission could have provided.18 Works of this ancient corpus are conserved and provided for learning in a traditional taxonomy that bears no relation to their history in the Western sense of the word. Even today, certain traditional Brahmans still possess immense mnemonic collections, to which access is in principle reserved to those who are ritually fit.19 To take another example, the early Buddhist corpus was also transmitted exclusively orally during the first five centuries of its existence – the Buddha died during the sixth or fifth century bce, according to the chronology adopted – whereas the use of writing had been known in India since at least the third century bce.20 It was not until 35–32 bce in Ceylon that the decision was 18 See Colas 2011, 495. 19 Certain comparatively late Vedic texts state that learning the Veda from a manuscript is incorrect (see Malamoud 2002, 142). To commit a text to a physical medium was to risk it falling into the hands of unqualified individuals (see also Colas 1997, 352 and 359). The Jains long restricted access to their libraries to members of their religion only (Cort 1995, 79–80). Some scholars preferred to drown their collections than let them be read by people deemed to be unqualified (Minkowski 2010, 89). 20 But Buddhism did not seek to protect the words of the Buddha from diffusion in any unau-

Gérard Colas made to methodically record the existing canonical texts on manuscripts, as war and famine had endangered the oral transmission of the words attributed to the Buddha.21 A dramatic historical event thus probably accidentally established a relationship between Buddhist texts and the specific technique which is writing, through transcription on manuscripts. The mainly oral nature of the ancient Vedic tradition is not an absolute historical model for Indian religious texts, but their recording on palm leaves took place in a constant dialectic with orality. Orality constituted a fundamental benchmark in both transmission and reading, with manuscripts being a much weaker testimony in textual traditions than they were in Europe.22 The example of the coexistence during the third century bce of the oral transmission of the Buddhist canon and the Mauryan emperor Aśoka's edicts on rocks and pillars publishing Buddhist moral and political measures confirms a dichotomy at that time between mere oral transmission and the use of writing: on the one hand there were texts supported above all by a tradition of recitation and a socio-spiritual authority, and on the other hand there were political or legal documents engraved on stone or another long-lasting material such as copper plates. In such context, palm-leaf, whether engraved or written, formed a fragile medium between the extreme and strict orality epitomized by the oral transmission of the Veda, and the mineral pseudo-eternity of epigraphy (on stone or other durable materials such as copper plates). In a palm-leaf manuscript library, texts survived only through regular re-copying, and therefore thanks to permanent or periodic scriptorium activity. Libraries themselves existed and perhaps flourished only owing to the financial efforts of donors, that is, an uncertain and all too human support. To sum up, palm-leaf manuscript libraries pertain to a broader whole that also encompasses mnemonic collections, epigraphic collections, and archives. These four types of collections involve four sorts of storing spaces (library building, mnémothèque-memory, lithothèque-temple walls, archive building). Each of these storing spaces is to a certain extent and in various ways contiguous with one or several of the other three, but has a specific purpose: for example, epigraphic collections partly perpetuate the content of documents on palm leaves; archival libraries also contain reference books (containing doctrinal texts, for instance), as we will see; and the restorers of palm-leaf manuscripts use their mnemonic collections. thorized social environments, contrary to what the Brahmans did with their Veda. The Buddha gave speeches in multiple languages, and these were disseminated in multiple languages. 21 See Lamotte 1988, 368 and 557–558. 22 See Colas 2011, 495.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century 2 The prescriptive ideal The observations that follow will be limited to a few collectively used religious libraries, mainly in Southern India at the beginning of the second millennium. The Pauṣkarasaṃhitā prescribes an ideal way of establishing such libraries. This text is a Hindu ritual manual from the Pāñcarātra denomination that may date back to before the eleventh century. It was most likely written in Southern India. Around twenty verses from this Sanskrit work recommend the ‘installation of knowledge’ or ‘of the seat of knowledge’ or even ‘of the seat of scriptures’; in other words, the establishment of a library,23 possibly in a temple.24 First, a collection of manuscripts is assembled, the core of which naturally consists of texts of the Pāñcarātra. It is specified that the texts from this religious denomination are to be edited (pariśuddha) beforehand, so that the manuscripts can then form an up-todate scriptural reference for the followers of the Pāñcarātra.25 The remainder of the collection consists of Vedic texts, socio-religious codes (smṛti), mythological stories, as well as works on logic, etc. The textual ensemble is therefore not exclusively religious. These manuscripts are placed in solid cases (saṃpuṭa). The passage then presents the prescriptions concerning the library itself. It should be built in a very stable neighbourhood with a good reputation, near (or in?) a royal city. This neighbourhood must be populated with dvija-s (‘twice-borns’)26 and maintained by a community of upstanding people. Such a social environment was probably assumed to promote the maintenance and financing of the future library. The building should be built of stone, have a metal lock (lohayantra), and door leaves with bolts (argala). It should be whitewashed (sudhādhavalita) and decorated with a painted representation of Vāgīśvarī, another name for Sarasvatī, the goddess of speech and knowledge, protectress of literature and the arts. 23 In the text, respectively: jñānapratiṣṭhāna, vidyāpīṭhapratiṣṭhāna, śāstrapīṭhapratiṣṭhāna. See Śrīpauṣkarasaṃhitā, ed. Yadugiri Yatirājasampatkumāra Rāmānujamuni 1934, 41, verses 77–98a. For a brief summary of the passage in question, see Sankaranarayanan 1993, 26–27. My interpretation of this known passage disagrees on multiple occasions with that of Florinda De Simini, who sometimes modifies the text according to her particular understanding (for example, she reads sapuṭeṣv instead of saṃpuṭeṣv in the edition [verse 80], lohayantrāsane instead of lohayantrāyane [verse 86]) (De Simini 2016, 346–347). 24 As hypothesized by Sankaranarayanan 1993, 26. 25 Another famous text from the Pāñcarātra denomination, the Jayākhyasaṃhitā, mentions the work of experts in sacred scriptures to edit the texts of the Pāñcarātra, restoring where necessary passages that have been corrupted, and the work of keepers of Scriptures who gather the texts: see Colas 1994, 118 and 131 and n. 37 and 38. 26 This term can refer either to people from the three upper classes of Hindu society, or only those of the highest class, the Brahmans.

Gérard Colas After the manuscripts have been revered and enclosed in pieces of fabric, they are to be transported to this locked building and venerated once again. Thereafter, the passage recommends daily worship, which mainly consists in carrying incense and walking clockwise around a pedestal-altar (pīṭha) located in front of the library. The text then requires that a maintenance fund (vṛtti) be procured for the conservation of the manuscripts (jñānakośānupālana) and confirmed by its inscription in a charter (śāsana). We may surmise that this fund should cater also for the salaries of employees of the library (see below on Chidambaram Sarasvatī Library). These prescriptions are followed by the recitation of two mantras in the library then designated as a divine temple (devagṛha). 3 On the importance of the librarian Regardless of whether this primarily religious library is built in a temple or elsewhere, it is thus built within a conservative social context. However, let us move beyond the recommended ideal. A few rare inscriptions from southern India record the foundation or extension of public libraries of Brahman villages, schools, monasteries, and temples, sometimes mentioning the donation of lands generating regular income to finance the employment of librarians. The compound words sarasvatī-bhaṇḍāra (or -bhāṇḍāra), literally ‘storehouse of Sarasvatī’, and pustaka-bhāṇḍāra, ‘storehouse of books’, or their Tamil forms, refer to manuscript libraries and sarasvatī-bhaṇḍārika to their librarians. In Tamil epigraphy, these establishments were distinguished from the archives of temples, called tirukkai-ōṭṭi-paṇṭāram, which kept official documents, including inter alia temple ownership deeds, rulings of the village assembly regarding tax exemptions, and sometimes sale contracts.27 The number of librarians varied, depending on the size of the manuscript library, and their position appears to have been hereditary. An inscription in the Kannaḍa language on an obelisk of the Nāgavāvi temple,28 dated to 1058 (śaka 980), mentions the creation of six librarian positions at a traditional school (called ghaṭikāsthāna) consisting of two hundred students of the Veda and fifty-two students of established disciplines (śāstra). They received land as reward for their work, as did the three commentators (vyākhyātṛ) in speculative disci- 27 V.G.R. Ayyar 1932, 32 and 41. See also below the inscriptions of Sarasvatī Library at the Chidambaram Temple. 28 Nagai today, located near Chitapur in the current state of Karnataka.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century plines and the bell-ringer. The size of the land assigned to each of the librarians was as large as that of the land given to one of the commentators and to the bell-ringer.29 This seems to indicate that the status of the librarians was comparable to that of the commentators (and the bell-ringer!), whereas the tutors (upādhyāya), totalling six, do not seem to have received land.30 The number of librarians (six) is important with respect to the total number of teachers (nine), commentators and tutors included. Last of all, librarians also benefited from a general allowance that covered the food and accommodation expenses of all of the scholars and students at the school, totalling two hundred and sixty-seven people.31 Librarians were therefore well paid at this eleventh-century Brahman teaching establishment. More important than ordinary tutors, librarians were on the same level as commentator scholars, who were masters capable of interpreting texts and revealing their meaning, and not only of teaching them. Librarian status was most likely hereditary, or at least it was the case in certain contexts. The inheritance of the lands received as a part of this position usually obliged the heir to perform the same professional service, in the same way as the donations of lands to temple priests engaged their heirs and enabled continuity of worship. Two inscriptions from the beginning of the fifteenth century in connection with the library of the Śaṅkarian monastery in Śṛṅgeri (in the modern-day state of Karnataka) confirm the hereditary nature of this position and the family accumulation of real estate assets from one generation to the next. The first (1406, śaka 1328) records a gift of land to librarian Purāṇika Kavikṛṣṇabhaṭṭa, the second (1432, śaka 1354) a gift of land to his son Kaviśaṅkarabhaṭṭa.32 The planning of new Brahman villages could include a library for collective usage. According to an inscription from the second half of the thirteenth century, the four vēli-s of land that the founding patron bought for the site of the future 29 Inscriptions of Nagai 1928, 7, 15 (lines 193–196), 16 (line 206), 23. Each librarian received 30 mattar-s of land; the commentator in Nyāya (‘logic’, the engraver most likely carved by mistake nyāsa for nyāya) and the bell-ringer as well. The commentator of the Bhāṭṭa school of the speculative discipline of Mīmāṃsā exegesis received 35 mattar-s of land, and that of the Prābhākara school of Mīmāṃsā received 45. We cannot help but wonder, on the one hand, about the importance of the ‘bell-ringer’s’ socio-ritual status (ghaṭikā-prahāri, ‘striker of hours’), which this gift of lands places on the same level as a commentator; and on the other hand, on the reasons for the difference in treatment between the commentators in the different disciplines (age, reputation?). 30 Three of these tutors taught the Veda (lines 194–195); the disciplines taught by the three others are not mentioned (see line 193). 31 In addition to the six librarians, there are the nine teachers and the two hundred and fifty-two students. The engraver most likely carved by mistake the number 257 instead of 267. 32 A.S.R. Ayyar 1939–1940, 325.

Gérard Colas village of Vikramapāṇḍyacaturvedimaṅgalam in the Tamil region were mainly intended for the temple, the houses of the one hundred and eight Brahmans, and those of the librarians and other professionals from the village.33 This document thus appears to indicate the relative importance of a library for collective usage (which a priori does not seem to have been religious) and its librarians.34 Libraries were at times expanded when donations allowed. For example, a 1269 inscription records the addition of a pavilion to the library of the temple of Raṅganātha in Śrīraṅgam. The patron ordered the images of three divinities of knowledge – Sarasvatī, Vedavyāsa, and Hayagrīva – to be ritually installed there, and provided for their regular worship through a donation.35 Another example of expansion during the thirteenth century was at the library of the temple of Naṭarāja in Chidambaram, which I will now discuss. 4 Sarasvatī Library at Naṭarāja temple in Chidambaram The manuscript library of the temple of Naṭarāja (a form of the god Śiva), in Chidambaram in Southern India, is an exceptional example, for its epigraphy and architecture record the development of its space and its manuscript copying activities as a result of the support of a thirteenth-century patron. Like most great Hindu sanctuaries in Southern India, the temple of Naṭarāja included the four kinds of textual and documentary repositories mentioned above. First there was a mnemonic collection, in this case consisting of the texts memorized and recited by priests without the help of manuscripts during rites and of other texts which 33 This would be equivalent to 9.72 hectares if the vēli measurement unit is equal to ‘six acres’ (Anglo-Saxon), as indicated by the dictionary of Johann Philipp Fabricius (Fabricius’s Dictionary 1972, s.v.). The number 108 is auspicious in ancient India. For this inscription, see Annual Reports (1911–1914), 1914, in the section ‘G.O. N° 920, 4th Aug. 1914’, 92. 34 For other mentions of libraries, see Sankaranarayan 1993, 26–32. According to a 1359 inscription (śaka 1282), the god Hastagirīśa himself, at the request of the temple administrator and scholars (bhaṭṭa) of the god, instituted a certain ‘Vaishnavite servant’ (vaiṣṇavadāsa) as director of a monastery in Kanchipuram, responsible for the worship taking place there, but also for the manuscripts (postaka) that he had collected and the accessories (upakaraṇa) associated with them: see A.S.R. Ayyar 1939–1940, 321, 324 and line 4 of the inscription. 35 The interpretation here is based on that of Sankaranarayanan (1993, 32), according to which the library probably existed already during the twelfth century. For a seemingly different interpretation, see Annual Reports (1938–1939), 1986, 95.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century certain scholars and devotees knew by heart. Second there was an epigraphic collection, consisting of the set of inscriptions which covered the walls of the temple, including those dealing with archive and manuscript libraries (see Figs 2, 3a, 3b). I will focus mainly on the other two collections: the library and the archive. Two inscriptions provide valuable information on their functioning.36 There are twenty-five or thirty years between these thirteenth-century epigraphs in Tamil language, engraved on the base of the pavilion part of the library, here referred to as inscriptions no. 1 and no. 2, in chronological order (see Figs 3a and 3b).37 Reading these two inscriptions is difficult because they are badly damaged and, besides, a staircase was built over the portion of the base where inscription no. 1 was engraved (see below).38 The archive39 was likely located in the second enclosure,40 but it seems impossible to situate its exact location today. The original building of the manuscript library was and is still today located in the third enclosure, at times called the exterior enclosure, beside the monumental west gate, slightly to the north of this gate.41 More specifically, it is found to the north (that is, to the right) of a small sanctuary that lies against the monumental west gate (see Fig. 4a) and of which it is separated by a modern building. This tiny sanctuary could be that which inscription no. 1 refers to as being a temple to the god Subrahmaṇyappiḷḷaiyār. Today the deity is identified as Śrīmuttukkumārasvāmi, in other words, a form of the god Subrahmaṇya.42 The two inscriptions name the manuscript library the 36 I will follow the transcription (here modernized), and mostly the interpretation of K.G. Krishnan (1988); see also the brief analysis in Sankaranarayanan 1993, 27–31. The two inscriptions have many gaps. 37 See Krishnan 1988, 219 (inscription no. 2 is separated from no. 1 ‘by not more than about twenty-five years’) and 223 (‘thirty years may be considered to be a reasonable interval between the two records’). In this article, I assume that the library is located in the same building where these inscriptions are engraved (see Krishnan 1988, 202: ‘the maṇḍapa [pavilion] on the tiers of which this inscription [no. 1] is engraved, situated on the north of the Subrahmaṇya temple was perhaps the place where the library was located’). It makes sense that the inscriptions are engraved on the main monument that they mention. Moreover, inscription no. 1, while significantly damaged, puts the location of the library in the northern portion of the west gallery of the third enclosure, near (to the north of?) a sanctuary of Subrahmaṇya (regarding which see n. 42 below). This location seems to correspond to that of the current building on which the two epigraphs are engraved. Inscription no. 2 mentions an extension to the original building, which seems to be the left wing of the current architectural unit (see below). 38 The extent of these missing elements makes it difficult or even impossible to translate with certainty. 39 tirukkai-ōṭṭi, for which see Krishnan 1988, 221. 40 Called vikkiramacōḻantirumāḷikai in inscription no. 1. See Krishnan 1988, 221. 41 See Krishnan 1988, 220 and footnote 37 above. 42 The date of this small, simply decorated sanctuary remains to be determined, along with

Gérard Colas ‘storehouse of Sarasvatī’, after the goddess of knowledge, hereinafter ‘Sarasvatī Library’ in this article. Sarasvatī Library may have been built during the twelfth century.43 The main room44 is placed within the gallery that runs along the entirety of the wall of the enclosure.45 In front of this room, in other words, projecting from the gallery, there is a pavilion that was originally open and which was accessed via a staircase located on the north side (to the right of this pavilion in Fig. 4b). Yet another room was later added to the south (that is, to the left of the main room within the gallery and set back from the pavilion), as mentioned in inscription no. 2 (see Fig. 5a); this extension appears to date back to the thirteenth century. At an unknown date subsequent to the thirteenth century,46 a new staircase was built against the moulded base on the east side (in other words, on the front part) of the pavilion, partially blocking inscription no. 1 (see Figs 3b and 5a), and the formerly open intercolumniations of the pavilion were walled in (except for the fifth, which is located the furthest to the left of the pavilion, behind the new staircase). As a result, the ancient north entrance was blocked off and the staircase leading to that entrance naturally fell into disuse. At the same time or later, a garden was created, formed of flowerbeds separated by strips of paving stones that of the monumental west gate. In 1932, V.G.R. Ayyar summarized the hypothetical dates for the other monumental gates, but not this one (32–33). Be it as it may, inscription no. 1 could not be alluding to the other Subrahmaṇya Temple (which is found in the north-west corner of the third enclosure, beyond the Civakāmi-ammaṇ Temple), for the construction of this chariot-shaped building was subsequent to inscription no. 1, in all likelihood. Local tradition places this construction during the reign of King Sundara Pāṇḍya I (from 1251 to 1268), and therefore after inscription no. 1 in any event (see Krishnan 1988, 219). This Subrahmaṇya Temple is clearly subsequent to Sundara Pāṇḍya I according to Fergusson 1910, 379: ‘we cannot feel sure of its date. From its position, however, and the character of its ornamentation, there seems little doubt that it belongs to the end of the 17th and first half of the 18th century. From the style, however, I would be inclined to ascribe it to the earlier date’. 43 According to Sankaranarayanan 1993, 28. 44 The length of this early main room and of the pavilion placed in front of it is five intercolumniations, a modular distance that is a typical unit of measurement in Sanskrit architectural treatises. The gallery that runs along the enclosure is itself dotted with pillars separated by the same modular distance. Regarding the concept of intercolumniation (and the similar concept of centre-to-centre distance between columns), which translates the Sanskrit term paṅkti, see Dagens 1984, 15–16. 45 For material and administrative reasons, I could not enter the Library building, the entrance to which was locked during my study, and my interpretation of the architectural layout therefore remains general and partially hypothetical. 46 In any event, a certain lapse of time following the two inscriptions.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century placed perpendicularly in front and on the right side of the library complex, with a passage paved with stones leading to the new staircase (see Figs 5a and 5b). The first of the two inscriptions is an oral order that appears first to have been transcribed onto palm-leaf before being engraved on the base of the pavilion part of the library. This order came from a committee of eight people: three senior officials responsible for managing the temple, and five representatives from different groups, appearing to include priest and accountant groups.47 The inscription relates and makes applicable an operation to rescue, restore and conserve ancient texts of the Sarasvatī Library (sarasvatipaṇ[ṭāram]). The donation from an erudite patron named Uṭaiyār Svāmitēvar allowed eight or nine people to be hired48 to unfasten and tie again49 the manuscript bundles (see Fig. 1) and to rewrite damaged leaves.50 Two or three other people were hired to place boards51 enclosing manuscript bundles and to thread their palm leaves.52 This made for a total of ten to twelve positions. The employees are referred to by their names. The inscription mentions their daily payment in cash (kācu) and in kind (fabric, unhusked rice).53 All of the expenses generated by these payments are financed by income from lands, the dimensions and location of which are provided by the inscription. In the second inscription, we will see that the same employees (or their descendants if they died?) appear to have kept their position for at least two decades. Inscription no. 1 mentions a second aspect of the rescuing operation, which concerns the archive (situated in another building, see above). As we have seen, this kind of storing space was supposed to conserve mainly the legal and administrative documentation of the temple on palm leaves.54 However, the inscription tells us 47 According to Krishnan, a person with the title of Tirumantiravōlaināyakam, in other words, ‘the head of the section which drafts or commits to writing the oral orders of the king while he is in counsel’, was among the signatories to this order (Krishnan 1988, 221). However, it appears to be a presumed name in the inscription. 48 Including two or three ascetics. The missing parts in the inscription make it impossible to estimate with any certainty the number of people hired for the different tasks. 49 aviḻttuk-kaṭṭavum. Krishnan 1988, 220: ‘unfolding, binding’. See above and the illustration on Fig. 1. Regarding the meaning of these Tamil terms, see Tamil Lexicon 1982, vol. 1, 158, col. B (meaning no. 1) and vol. 2/2, 650, col. B (meaning no. 1). 50 jīrṇṇittavai +...+ eḻutavum. 51 pala[kai] (meaning no. 1 in the Tamil Lexicon 1982, vol. 4/1, s.v.). 52 Tentative interpretation of kōkkavum. Here, the rethreading includes freshly rewritten palm leaves. See the illustration on Fig. 1. Krishnan (1988, 220) explains: ‘in placing the planks (palagai) and tagging on all the leaves (kōkka)’. 53 See lines 3 and 4 of the inscription, as well as Krishnan’s interpretation (1988, 220); Sankaranarayanan 1993, 29. 54 Sometimes also on copper plates.

Gérard Colas that it was also used to conserve manuscripts of works in Sanskrit and Tamil. One may suppose that the purpose was to keep reference and back-up copies of these works. The inscription states that these reference books kept in the archive room are to be copied and edited by comparing them to those of the Sarasvatī Library, and that after that, the originals and copies are to be stored at the two libraries. We may assume that the staff responsible for this work on the manuscripts of the archive was that which had been hired to partially or totally copy the manuscripts of Sarasvatī Library. Inscription no. 2 at Sarasvatī Library (also on the base of the pavilion part) mentions two operations with regard to that library. The responsible for the first operation apparently is Tēvar Ñānasamutratēvar, literally ‘Tēvār (named) TēvārOcean of Knowledge’. This is not sure however, for the inscription is damaged around this name. This Tēvar Ñānasamutratēvar could be the Uṭaiyār Svāmitēvar of Inscription no. 1. The first operation consisted in a new rescuing operation with the provision of manuscripts to readers. It required hiring twenty new employees, the names of which are given. Six were hired to thread in right order (?) and read (aloud)55 the manuscript of the Divyāgamam, a text or collection of texts that may have formed the temple’s ritual corpus of reference. Four others were responsible for unfastening manuscripts and tying them once again, probably to help scholarly readers handling them, as bundles could be very thick. Eight other employees were responsible for preparing new copies of old manuscripts. Another was specifically appointed to astrological manuscripts, and another to mythological texts (purāṇas).56 However, the inscription does not present the amount paid to these twenty or so new employees. As for the second operation mentioned in this inscription, it is clearly attributed to Tēvar Svāmitēvar, which tends to confirm that it was the same patron, Svāmitēvar, who directed and likely financed all of the operations mentioned by the two inscriptions. In this case, the operation in question was a physical extension of the library. This extension was constructed alongside the former building of the Sarasvatī manuscript library.57 In the layout of the current architectural complex, it is the left wing (the southern extension mentioned above) that one can see today in the gallery (see Figs 5 A and 6). Manuscripts from the former room were moved 55 ārppikkavum vācippikkavum. Krishnan (1988, 123) apparently interprets ārppikkavum as ‘arrange’. Even though Indians were familiar with silent (‘mental’) reading, reading aloud was the common practice in the editing process: see Colas 2001, 313; Colas 2011, 496 and 505. 56 Here, we are merely following the interpretation of Krishnan 1988, 222. 57 muṉpattai-sarasvati-paṇṭarām.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century to this new space, and employees that had been there for twenty-five or thirty years (or their descendants?) were to continue performing their tasks.58 The audience of this order engraved into the stone consisted, in addition to seven (or more) dignitaries, of two new library employees, most likely ranked as senior administrative officials: the librarian – literally ‘the master responsible for the books of the library’59 – and the accountant.60 5 The teachings of the Sarasvatī Library inscriptions The archive, in addition to its specific role of filing administrative and legal documents, was thus used to conserve certain manuscripts. However, it did so passively, one could say, probably as an almost administrative reference. In the absence of frequent consultation, the mere conservation of such manuscripts – which were as a matter of fact neglected, because they were rarely or never consulted by scholars – implied their rapid disappearance, considering their exposure to insects, rodents, and the climate. They did however sometimes have a chance of being replaced by fresh copies. This is what happened at Naṭarāja temple in Chidambaram when an open-minded patron overcame the institutional division between these two types of library and financed the restoration and copying of manuscripts conserved in the archives as well as in Sarasvatī Library, with the backing of the temple authorities. Sarasvatī Library’s manuscripts were intended to be read and consulted by scholars. The fragility of palm-leaf meant that a workshop of regular copying and restoration of texts had to be closely associated with conservation. Making new copies of manuscripts leaves, whether some of them or their entirety, was probably done with the assistance of people reading out loud.61 In principle, this was followed by a new reading to check the copying and make corrections.62 It was thus an editorial process in which the editor’s mnemonic library played the role of the authoritative reference, resulting in corrections and updates that were partially planned (under the influence of readings contemporary with the editor, 58 See Krishnan 1988, 222. 59 paṇṭārap-pottakam-uṭaiyār. 60 kaṇakkar. 61 As in the case of paper manuscripts. See references to Colas 2001 and 2011 in n. 55 above. 62 Colas 2011, 500 and 505.

Gérard Colas considered to be the valid ones), or even deliberate (in other words, aimed at transforming passages, terms, or syllables deemed to be faulty).63 The Sarasvatī Library also offered a service to scholars: it was most likely there and not at the archives that scholars came to consult texts – with the assistance of employees who unfastened the bunches of palm leaves and tied them once again – and copied or ordered the copying of the texts that they needed. Over a span of twenty-five or thirty years, the space and activities of Sarasvatī Library grew significantly thanks to the generosity of the same patron, Svāmi Tēvar, it seems, who was also a scholar personally involved in the collection. The number of employees whose salaries he financed went from ten or twelve to thirty or thirty-one, in addition to the hiring of a librarian and an accountant. His actions were supervised and legitimized by temple dignitaries. The separate hiring of a librarian, who seems to have partially taken over the reins from the erudite administrator Svāmi Tēvar, confirms the importance and scholarly component of this profession during the thirteenth century. The activities of Sarasvatī Library therefore depended heavily on private donations, which were most likely sporadic. Such spectacular increase in staff over twenty-five or thirty years must not be over-interpreted: the financing of salaries, despite being engraved into the temple’s stones and having the formal approval of the senior officials of the temples, did not guarantee the permanence of the jobs or that of the services provided by the employees.64 The creation of a flower bed garden around the building could be an indication that Sarasvatī Library existed for a certain time following the thirteenth century. According to the present temple’s management,65 today the building that was the Sarasvatī Library holds materials used during an annual temple celebration; the manuscripts of the temple were supposedly moved to the library of the Mahārāja of Tanjore at an unknown date. This library of manuscripts and printed texts, originally founded in the sixteenth century, and which is today lively and renowned – as Saraswati Mahal Library or Tanjore Maharaja Serfoji’s Mahal Library – developed mainly during the eighteenth century, partly based on the European model.66 As an institution of a different type to that of the library of Chidambaram temple, the Tanjore Maharaja Serfoji’s Mahal Library raises the 63 Regarding this process, see Colas 1994, 117 and 118; Colas 2001. 64 The epigraphic donations by no means had the eternal application they often sought to establish, one reason being that they could no longer easily be read after a certain period, due to changes in epigraphic engraving style and language conventions. 65 Interviewed on this topic on 31 August 2017. 66 See Wujastyk 2007.

Palm-leaf Manuscript Libraries in Southern India Around the Thirteenth Century question of the dispersal and disappearance of Indian manuscript collections, their merging with other collections, and their further transmission during different periods, which are other subjects for reflection. References Annual Reports on Indian Epigraphy (1911–1914), New Delhi: Archaeological Survey of India, 1914. Annual Reports on Indian Epigraphy (1938–1939), New Delhi: Archaeological Survey of India, 1986 [reprint]. Ayyar, V. G. Ramakrishna (1932), ‘The Chidambaram Temple’, Journal of the Annamalai University, 1/1 (April): 29–42. Ayyar, A.S. Ramanatha (1939–1940), ‘N° 34: Conjeeveram Inscription of Brahma-tantra-svatantra-jīyar’, in N. P. Chakravarti (ed.), Epigraphia Indica, 25, Delhi: Archaeological Survey of India, 318–326 [reprint: 1985]. Colas, Gérard (1994), ‘Cultes et courants du vishnouisme en Inde du sud. Quelques observations à partir des textes’, in Marie-Louise Reiniche and Henri Stern (eds), Les ruses du salut. Religion et politiques dans le monde indien (Collection Puruṣārtha, 17), Paris: Éditions de l’École des hautes études en sciences sociales, 111–138. Colas, Gérard (1997), ‘Les manuscrits envoyés de l’Inde par les jésuites français entre 1729 et 1735’, in François Déroche and Francis Richard (eds), Scribes et manuscrits du Moyen-Orient, Paris: Bibliothèque nationale de France, 345–362. Colas, Gérard (2001), ‘Critique et transmission des textes dans la littérature sanskrite’, in Luce Giard and Christian Jacob (eds), Des Alexandries I. Du livre au texte, Paris: Bibliothèque nationale de France, 309–328. Colas, Gérard (2011), ‘Relecture et techniques de correction dans les manuscrits indiens’, in Christian Jacob (ed.), Lieux de savoir, vol. 2, Paris: Albin Michel, 494–520. Colas, Gérard (2017), ‘La feuille de palmier, support d’écrit en Inde ancienne’, in Claude Laroque (ed.), Autour des papiers asiatiques, Paris: HICSA, 119–137. Cort, John E. (1995), ‘The Jain Knowledge Warehouses: Traditional Libraries in India’, Journal of the American Oriental Society, 115/1: 77–87. Dagens, Bruno (1984), Architecture in the Ajitāgama and the Rauravāgama (A Study of Two South Indian Texts), New Delhi: Sitaram Bhartia Institute of Scientific Research. De Simini, Florinda (2016), Of Gods and Books. Ritual and Knowledge Transmission in the Manuscript Cultures of Premodern India (Studies in Manuscript Cultures, 8), Berlin: De Gruyter. Fergusson, James (1910), History of Indian and Eastern Architecture, vol. 1, London: John Murray. The Inscriptions of Nagai (Hyderabad Archaeological Series, 8), Calcutta: The Baptist Mission Press, 1928. Janert, Klaus Ludwig (1995), Bibliographie mit den Berichten über die mündliche und schriftliche Textweitergabe sowie die Schreibmaterialen in Indien, Teil I: Berichtszeit bis 1995 (Sarasvati Series, 13), Bonn: VGH Wissenschaftsverlag. J. P. Fabricius’s Tamil and English Dictionary Based on Johann Philip Fabricius’s ‘Malabar-English Dictionary’ [1779], 4th edn, Tranquebar: Evangelical Lutheran Mission Publishing House, 1972. Kangle, R. (ed.) (1960), The Kautilīya Arthaśāstra, Part I: A Critical Edition with a Glossary (University of Bombay Studies: Sanskrit, Prakrit and Pali, 1) Bombay: University of Bombay [new edition].

Gérard Colas Kangle, R. (trans.) (1963), The Kautilīya Arthaśāstra, Part II: An English Translation with Critical and Explanatory Notes (University of Bombay Studies: Sanskrit, Prakrit and Pali, 2) Bombay: University of Bombay [new edition]. Krishnan, K.G. (1988), ‘N° 47 — Two Inscriptions on Sarasvatī-bhaṇḍāra from Chidambaram’, in K.G. Krishnan (ed.), Epigraphia Indica, 40/6 (July 1974), New Delhi: Archeological Survey of India, 219–224. Lamotte, Étienne (1988), History of Indian Buddhism from the Origin to the Śaka Era, Louvainla-Neuve: Université Catholique de Louvain, Institut Orientaliste [1st edn in French: 1958]. Malamoud, Charles (2002), Le jumeau solaire, Paris: Éditions du Seuil. Minkowski, Christopher (2010), ‘Sanskrit Scientific Libraries and Their Uses: Examples and Problems of the Early Modern Period’, in Florence Bretelle-Establet (ed.), Looking at It from Asia. The Processes that Shaped the Sources of History of Science, Dordrecht: Springer, 81–114. Olivelle, Patrick (2013), King, Governance, and Law in Ancient India. Kauṭilya’s Arthaśāstra. A New Annotated Translation, Oxford: Oxford University Press. Salomon, Richard (1998), Indian Epigraphy, New York: Oxford University Press. Sankaranarayanan, S. (1993), ‘Sanskrit Education in Ancient Tamil Country (An Epigraphical Survey)’, Adyar Library Bulletin, 57: 7–33. Sarma, Sreeramula Rajeswara (2012), ‘From my Grandfather’s Chest of Palm Leaf Books’, in Saraju Rath (ed.), Aspects of Manuscript Culture in South India, Leiden: Brill, 207–233. Schurhammer, Georg and G.W. Cottrell (1952), ‘The First Printing in Indic Characters’, Harvard Library Bulletin, 6/2: 147–160. Sircar, D.C. (1996), Indian Epigraphy, Delhi: Motilal Banarsidass [reprint, 1st edn: 1965]. Tamil Lexicon, Published in Six Volumes Under the Authority of the University of Madras, Madras: University of Madras, 1982. Wujastyk, Dominik (2007), ‘La bibliothèque de Thanjavur’, in Christian Jacob (ed.), Lieux de savoir, vol. 1: Espaces et communautés, Paris: Albin Michel, 616–636. Wujastyk, Dominik (2014), ‘Indian Manuscripts’, in Jörg Quenzer, Dmitry Bondarev and Jan-Ulrich Sobish (eds), Manuscript Cultures: Mapping the Field (Studies in Manuscript Cultures 1), Berlin: De Gruyter, 159–182. Yadugiri Yatirājasampatkumāra Rāmānujamuni (ed.) (1934), Śrīpauṣkarasaṃhitā, Bangalore: A. Srinivasa Aiyangar and M. C. Thirumalachariar.

Byzantium

Luciano Bossina How Many Books Does It Take to Make an Emperor’s Library? Constantine VII Porphyrogenitus and a Chapter of History of the Manuscript Book Abstract: In tenth-century Byzantium, the Emperor Constantine VII Porphyrogenitus commissioned a series of works that aimed to collect and select the knowledge accumulated over the centuries. In the Excerpta Constantiniana this enormous material was distributed in 53 treatises, divided into specific book units and meticulously numbered. A survey of other textual traditions – Epictetus’s Handbook, Evagrius’s De oratione, Nilus’s Correspondence – shows that the choice of this number is not accidental. Based on an arithmetic symbolism, the number 53 represents the aspiration to complete knowledge (earthly and divine). The emperor therefore wanted his ideal library to denote, even in the material structure, the possession of universal knowledge. 1 Constantine VII Porphyrogenitus’s Excerpta and their programmatic purpose In his famous book Le premier humanisme byzantin, Paul Lemerle describes the most remarkable features of the ‘encyclopaedism of the tenth century’, which reached the peak of its development under Emperor Constantine VII Porphyrogenitus.1 At its height, the ‘Macedonian Renaissance’ (a definition originally found in art history) infused Byzantine culture with new splendour following the crisis of iconoclasm. Yet, over the past decades, categories such as ‘encyclopaedism’ and ‘Renaissance’ have often been challenged, and a new appraisal of the cultural underpinnings of the iconoclasm – more nuanced and less negative at one and the same time – has been conceived. It is not simply a matter of minor terminological issues, rather they determine the overall interpretation of a period and characterize the nature of works that had 1 Lemerle 1971; English translation: Lemerle 1986, especially Chapter 10, ‘The Encyclopedism of the Tenth Century’, 309–346. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-008

Luciano Bossina a considerable impact. The definition of ‘Byzantine encyclopaedism’ had already been proposed decades earlier,2 but Lemerle’s authority and the adoption of this expression in leading art history and Byzantine literature textbooks ultimately endorsed its systematic spread. In recent times, however, several studies have highlighted the limits of such a definition and the risk of running into anachronism which it brings. Scholars have closely contemplated the independent historical and literary value of the Byzantine συλλογαί, insisting on the different nature of works which must not be analysed from the same perspective. As Paolo Odorico has pointed out,3 Lemerle somehow creates a ‘catalogue’ of ‘encyclopaedic works’, including the great undertakings tied to the name of Constantine (Vita Basilii, De administrando imperio, De thematibus, De cerimoniis).4 He also lists Symeon Metaphrastes’ Lives, the Basilica, and the Geoponica, works likely to be respectively regarded as religious, legal, and agricultural encyclopaedias. Furthermore, even the Souda lexicon and the collections of epigrams of the Greek Anthology were to be interpreted in the wake of this new ‘encyclopaedic spirit’. The consistency of this categorization system and above all its alleged encyclopaedic nature were nevertheless called into question. Odorico stressed the fact that the classification of these works should mirror their different purposes and the different operational criteria behind them. He repeatedly points out that so-called ‘compilation’ literature, which deconstructs and reconstructs sources in the form of a new whole, obeys a particular inner logic and implies a very specific objective.5 Hence, its interest lies not so much in the sources used but rather in the structure and function of the work. The debate was destined to continue since each of these works had to be studied in greater detail,6 without giving up 2 Already present in Büttner-Wobst’s seminal 1906 article, it was a question of historische Encyklopädie. 3 Following the 1990 study, other studies by Odorico followed up on this reflection: see, at the very least, Odorico 2011, and, more recently, Odorico 2017. 4 Vita Basilii, edition: Ševčenko 2011; De administrando imperio (which according to Lemerle 1986, 320 was ‘a sort of encyclopaedia of Byzantium’s foreign policy’), edition: Moravcsik 2008, comments: Jenkins 2012; De thematibus (description of the provinces of the empire), edition: Pertusi 1952; De cerimoniis (compilation of the ceremonial protocols of the imperial court), edition: Reiske 1829–1830, Vogt 1967. 5 Odorico 2017, 25–26. 6 For a discussion on Byzantine encyclopaedism, with specific focus on the most important works from the ninth to fourteenth centuries, see the studies gathered by Van Deun and Macé 2011 (in particular Schreiner 2011; Magdalino 2011; and Odorico 2011, which address the theoretical aspects of the different viewpoints). The conference on Paul Lemerle ‘forty years later’ (Paris, 23–26 October 2013) was also a productive occasion to return to these questions. The conference proceedings were published by Flusin and Cheynet 2017: see in particular Magdalino 2017; Odorico 2017; Markopoulos 2017; Ceulemans and Van Deun 2017.

How Many Books Does It Take to Make an Emperor’s Library? on comparing them to the encyclopaedic undertakings of other periods,7 both ancient and modern, while avoiding anachronistic parallels. In the context of this vast Byzantine literary production, the most important work is actually to be found in the enormous collection and compilation known under the name of Excerpta Constantiniana, which may now be analysed in the light of a remarkable and thoroughly documented study by András Németh.8 Németh opportunely insisted on the idea that this way of selecting and restructuring sources in view of the primary goal of creating a new whole is a ‘Byzantine appropriation’ of the past.9 Furthermore, he has had the merit of emphasizing the importance of the physical setting of the Excerpta. In this respect, the fundamental programmatic text is the prologue preceding each book. There is no need, here, to analyse in depth such a highly elaborate text.10 It will be sufficient, instead, to concentrate on the main passages, which explain the historical premises and goals behind this enormous collection:11 (a) ‘the number of events has become countless and the writings have become more complex’; (b) ‘the fabric of the history has been infinitely magnified to the point of becoming unmanageable’; (c) truly ‘useful books’ are a rarity and the writings inspire ‘fear and dread’. To resolve such a situation, Emperor Constantine is thought to have come up with the following solutions: (a) ‘to collect by means of diligent research all manner books from all over the known world’; and (b) ‘to divide and distribute their great quantity and extent […] into small sections’ in order to make this ‘fertile material’ […] ‘available unstintingly to common use’. From the operational point of view, to achieve this, it was therefore necessary (c) to establish a well-defined number of subjects, which he called the ‘principal topics’ [κεφαλαιώδεις ὑποθέσεις], and (d) to dedicate a section to each theme. 7 Comparing different periods always has its risks, but it is an essential condition for anybody who wishes to understand long-term historical phenomena. Extremely useful studies on encyclopaedism in Antiquity and the Middle Ages were published in Cahiers d’histoire mondiale 9, 1966 (cf. in particular Lemerle 1966). More recently, studies on encyclopaedism from Antiquity to the Renaissance were collected by König and Woolf 2013. 8 Németh 2018. See also Németh 2013 (with rich bibliography). 9 Németh 2018, 15: ‘Appropriation is an improved form of anthologization’. 10 Németh 2018, 54–87. In this book, the author offers a new edition of the prologue (Németh 2018, 267–268). 11 The prologue of the Excerpta de legationibus is cited according to Németh 2018 (edition: 267– 268; translation: 61–62).

Luciano Bossina Each section opens with the same prologue, containing its specific title and the place that it occupies within the entire series, according to the fixed scheme below: ὧν κεφαλαιωδῶν ὑποθέσεων ἡ προκειμένη αὕτη καὶ ἐπιγραφομένη [XYZ] [00] τυγχάνει οὖσα Of these principal topics, the present, entitled [XYZ], is number [00] As is already known, only a small portion of this enormous collection has survived: the section De legationibus and the section De virtutibus et vitiis (the first in its entirety, and only half of the second), and fragments of other two, De insidiis and De sententiis.12 Because we have the prologue of the two sections, we also know the place that they occupied: De legationibus was number 27 and De virtutibus et vitiis was number 50. The concepts on which this enormous undertaking was based were ‘wholeness’, ‘division’, and ‘order’. The emperor’s political and cultural goal was to distribute [καταμερίσαι] and organize [ἀπαρίθμησις] specific knowledge [ὑποθέσεις διάφοροι], to know ‘all the great achievements of history’ [ἅπασα ἱστορικὴ μεγαλουργία]. All other considerations aside, the interest of this passage in the overall assessment of the work is that it presents the work itself as an entire library, as a library in the form of a book, with a well-determined scope and a specific order. In this respect, one can speak of a ‘miniaturized library’,13 meaning that this immense work aspired to bring together and condense, within a single space, the most valuable content of all previous books, and thereby, as explicitly stated in the prologue, to hold all of the great lessons that history could teach the emperor in order to build and preserve a universal empire. 12 Editions: De legationibus: De Boor 1903; De virtutibus et vitiis: Büttner-Wobst and Roos 1906– 1910; De insidiis: De Boor 1906; De sententiis: Boissevain 1906. 13 I use this expression in a slightly different sense than that meant by Odorico 2017, 27. He speaks of ‘bibliothèques miniaturisées’ to define ‘des anthologies byzantines’ that bring together ‘des textes entiers, ordonnés selon une logique propre à chaque auteur, en vue de leur utilisation’. These libraries are supposedly ‘constituées par l’intégralité de la source, et non par son extrapolation’, and are motivated by the desire to gather ‘tout ce qu’on voulait conserver d’un genre littéraire précis’. These distinctions are certainly useful to avoid grouping together works of a different nature in the same category, as Odorico remarks by criticizing the excessively imprudent use of the category of encyclopaedism. However, I would like here to emphasize another aspect of the issue, namely the construction of a work that aspires to contain everything concerning a subject, gathering portions of other works, without distinguishing whether the source is incorporated in its whole or in a summarized form. In this sense, Constantine’s Excerpta is a book composed of other books, a ‘miniaturized library’.

How Many Books Does It Take to Make an Emperor’s Library? Within the context of the present volume, reflecting on libraries during the manuscript age, the goal of this paper is not to study the history of a Byzantine library from a codicological or archaeological perspective, but to shed light on this ambitious cultural project from the point of view of book history and the physical organization of the work. Nonetheless, this will require a long journey through documents that, though appearing superficially unrelated, may help to reconstruct a chapter of the history of the manuscript book in the Christian tradition.14 2 Epictetus’s Handbook and the Christian paraphrases In the spring of 1479, Angelo Poliziano concluded the last product of his suave otium: a translation of Epictetus’s Enchiridion or Handbook. In his epistle dedicated to Lorenzo the Magnificent, the author explained the specific virtues of the work: An admirable aspect across the work is its internal order: even though the text is divided into multiple chapters, every line, so to say, converges towards a single centre […] Moreover, the style – as required by the situation – is concise, clear, and devoid of ornament, just like the precepts that the Pythagoreans call diathekai.15 With these words, Poliziano is simply repeating the observations of the Neoplatonist Simplicius (sixth century), one of the Enchiridion’s most famous commentators in antiquity, who in his Preface on the Handbook says more or less the same: The speeches are pithy and gnomic, in the form the Pythagoreans called ‘precepts’. But practically all of them have a certain orderly relationship to one another and a logical sequence, as we shall see as we proceed. And, although the chapters were written sepa- 14 My interpretation here goes back to my PhD thesis: Bossina 2004, vol. 2, 331–369. The relationship between Constantine’s Excerpta and the prologue of Evagrius Ponticus’s De oratione was later independently argued by András Németh (see Németh 2013, 245–247; and Németh 2018, 71–77). I am therefore particularly pleased that my hypothesis has been confirmed by someone with a greater knowledge of Constantine. Here, I am reusing the entire demonstration, because it is based on a larger corpus of documents, as it had already been established in my 2004 study. 15 Maltese 1990, 60: Omnia vero ordinem inter se mirum habent omnibusque veluti lineis, quamvis in plura id opus capita sit distinctum, ad excitandum rationalem animum quasi ad ipsum centrum contendunt […] Stylus autem, qualem res postularet, concisus est, dilucidus quique omnem respuat ornatum, Pythagoreorumque praeceptis, quas illi diathecas vocant, quam simillimus.

Luciano Bossina rately, they all aim at one art – the art which rectifies human life. The speeches are also all directed towards one goal – rousing the rational soul to the maintenance of its proper value.16 On this point, the ancient commentator and the modern translator agreed: the Handbook was divided into chapters [κεφάλαια / capita], but this did not affect the overall unity of the project. However, Epictetus’s Handbook17 – as Pierre Hadot also notes in his translation – was ‘incredibly successful’ over the centuries18 and became one of the most stable sources for anyone wishing to find direction in his life through ancient philosophy. Yet, in view of a general analysis of this work, it might be necessary to ask a question that sounds rather odd, at first: How many chapters are there? How many ‘lines converg[ing] towards a single centre’? In the printed tradition, several divisions are witnessed.19 In Gregor Haloander’s edition (1529),20 the text is divided into 62 chapters; Hieronymus Wolf (1560)21 increased them to 79, and then, again, with John Upton (1741)22 they decreased to 52. The structure that became canonical was established by one of the eighteenth- and nineteenth-century greatest scholars, Johann Schweighäuser, who in his 1799 edition determined the number of chapters as 53.23 After fully examining the manuscript tradition of the Handbook, the utmost authority on the text 16 Simpl., In Ench., Prooem. 62–70, ed. Hadot 2001, 3–4: Κομματικοὶ δέ εἰσιν οἱ λόγοι καὶ γνωμολογικοί, κατὰ τὸ τῶν ὑποθηκῶν καλουμένων παρὰ τοῖς Πυθαγορείοις εἶδος, πλὴν καὶ τάξις τίς ἐστι πρὸς ἀλλήλους ἐν πᾶσι σχεδὸν αὐτοῖς καὶ ἀκολουθία, ὡς προϊόντες εἰσόμεθα. Κἂν τὰ κεφάλαια δὲ διωρισμένα γέγραπται, εἰς μίαν πάντα τείνει τέχνην, τὴν διορθωτικὴν τῆς ἀνθρωπίνης ζωῆς· καὶ πάντες οἱ λόγοι πρὸς ἕνα τείνουσι σκοπόν, τὸ τὴν λογικὴν ψυχὴν διεγεῖραι πρός τε τὴν φυλακὴν τοῦ οἰκείου ἀξιώματος. For an overview of the work and the author, see the introductory study by Hadot 2001, VII–CLII. English translation: Brittain and Brennan 2002, 38. 17 The text of the Handbook is here given according to the critical edition by Boter 1999. 18 Hadot 2000, 7: ‘il serait lu en Chine au XVIe siècle et, dix-huit siècles après sa rédaction, il figurerait dans les programmes scolaires’. Regarding Epictetus’s fortune, see the rich collection of data by Boter 2011, 2–10. 19 For a complete list of editions up until 1952, see Oldfather 1927 and Oldfather 1952. The data mentioned here are from Boter 1999, 146–147. 20 Epicteti Enchiridion cum interpretatione latina Angeli Politiani, ed. Haloandri, Norimbergae: Petreium, 1529 (Oldfather 1952, no. 249). 21 Epicteti Enchiridion, h.e. Pugio, sive ars humanae vitae correctrix […] Hieronymo Wolfio interprete una cum annotationibus eiusdem […], Basileae: Oporinus, 1560 (Oldfather 1952, no. 35). 22 Epicteti quae supersunt Dissertationes ab Arriano Collectae Nec non Enchiridion et Fragmenta Graece et Latine […] recensuit, notis & indice illustravit Joan. Uptonus, Londini: Woodward, 1741 (Oldfather 1952, no. 30). 23 Epicteteae philosophiae monumenta I–III, ad Codicum Manuscriptorum fidem recensuit, Latina Versione, adnotationibus, Indicibus illustravit Johannes Schweighaeuser, Lipsiae: In Libraria Weidmannia, 1799 (Oldfather 1952, no. 26).

How Many Books Does It Take to Make an Emperor’s Library? and its most recent editor, Gerard Boter in turn reconfirmed the division into 53 chapters, with a noteworthy remark: ‘It is a lucky coincidence that the tradition is more or less in accordance with the chapter division that has been current in the nineteenth and twentieth centuries’.24 At the same time, it should be recalled that the manuscript tradition of the Handbook is quite broad but also quite recent, and none of the 59 manuscripts dates back to before the fourteenth century. In the words of its last editor, this tradition is ‘more or less in accordance’ with the division into 53 chapters. In spite of this, much older evidence is preserved that contradicts this structure. The Handbook experienced enormous success during Late Antiquity and the Middle Ages, in both pagan and Christian circles, and was the subject of noteworthy commentaries and rewritings. With regard to the pagans, the superb commentary by Simplicius has already been mentioned, but there are also three strange Christian paraphrases worth citing: (a) the Paraphrase of St. Nilus; (b) the Paraphrasis Christiana; and (c) the Paraphrase of Vaticanus. gr. 2231.25 These works clearly demonstrate the enormous capacity of Christianity to appropriate masterpieces of pagan philosophy. Precisely because the chapters of the Handbook were short and sententious – as Simplicius acknowledges –, it was very easy to export and adapt them to Christian thinking.26 Moreover, the Handbook almost exclusively contains ethical teachings, all the while marginalizing the physical and ontological aspects of Stoic thought. The Christian Paraphrases rewrite the Handbook with small adjustments that transform the original version into a veritable Christian text (and specifically a monastic one). While Epictetus speaks of ‘gods’ in the plural (Handbook 31: Τῆς περὶ τοὺς θεοὺς εὐσεβείας ἴσθι ὅτι etc.), Christians obviously write ‘God’ in the singular (Paraphr. of St. Nilus 38: Τῆς περὶ θεὸν εὐσεβείας ἴσθι ὅτι etc.). While Epictetus mentions Socrates, Christians replace him with Jesus or the apostles. The adjustment is often delightful: 24 Boter 1999, 147, who then declares: ‘Therefore, I have maintained Schweighäuser’s chapter numbers’. 25 Following the studies of Wotke 1892; Piscopo 1969–1970; Piscopo 1972; Spanneut 1972; Piscopo 1978; De Nicola 1998; a critical edition of the Christian Paraphrases was published by Boter 1999. The Paraphrasis Christiana was in turn the subject of an exegetical commentary preserved in a rich manuscript tradition now available in the edition by Spanneut 2007. 26 The Christian appropriation of Epictetus is also the result of undeniable similarities in thought already recognized by Byzantine readers (for example, by Arethas of Caesarea, ninth–tenth century) and then later largely analysed by modern criticism, at least starting with Zahn 1895.

Luciano Bossina Epict. Hand. 46 Christ. Paraph. 60 Μηδαμοῦ σεαυτὸν εἴπῃς φιλόσοφον, μηδὲ λάλει τὸ πολὺ ἐν ἰδιώταις περὶ τῶν θεωρημάτων, ἀλλὰ ποίει τὰ ἀπὸ τῶν θεωρημάτων. οἷον ἐν συμποσίῳ μὴ λέγε πῶς δεῖ ἐσθίειν, ἀλλ’ ἔσθιε ὡς δεῖ. Μηδαμοῦ σεαυτὸν ἡσυχαστὴν εἴπῃς, μηδὲ λάλει ἐν ἰδιώταις περὶ τῶν δογμάτων. ἐν ἑστιάσει μὴ λέγε πῶς δεῖ ἐσθίειν, ἀλλ’ ἔσθιε ὡς δεῖ. Never call yourself a philosopher nor speak excessively, in the presence of ordinary people, of theorical principles, but practice that which is prescribed by these principles: in the same way, during a meal [symposion], do not hold conversations on the way that you must eat, but eat how you must. Never call yourself a hesychast (monk) nor speak, in the presence of ordinary people, of dogmas, in the same way, at the table, do not hold conversations on the way that you must eat, but eat how you must. So ‘philosopher’ becomes ‘monk’, ‘theoretical principles’ become ‘dogmas’, even the ‘symposium’, which is a word with an ideological connotation, is replaced with a general reference to the ‘table’. Various problems raised by these Paraphrases would deserve to be carefully studied, but one specific issue remains here as to how many chapters the text of the Handbook was supposed to contain, at the moment that it was reworked? If one looks at the four works together, namely Simplicius’s Commentary and the three Christian Paraphrases, none of them is found to divide the text into 53 chapters. All of them, instead, split it into approximately 71/73 chapters. This difference is achieved by means of progressive unification. For example, Chapter 1 of the Handbook, as one reads it today, covers numerous chapters of the Paraphrases (six chapters in Simplicius, five in the Paraphrase of St. Nilus). Chapter 33 even covers a dozen chapters. As a result, the last chapter of the Handbook, Chapter 53, corresponds to Chapter 71 in Simplicius, Chapter 72 in the Paraphrase of St. Nilus (which then adds an independent chapter, 73), Chapter 71 in the Paraphrasis Christiana, and Chapter 73 in the Paraphrase of Vaticanus gr. 2231. A specific study, which cannot be carried out here, should examine the structure of the Handbook in detail throughout the entire manuscript tradition, providing a systematic comparison with parallel texts. In any case, doubt remains concerning the original division of the Handbook into 53 chapters. I am inclined to think that Epictetus’s manuscripts, because of their recent origin, convey a subsequent, artificial arrangement (not surprisingly often regarded as inadequate and incoherent by modern scholars).27 Other textual traditions, which we will 27 See, for example, the exact evaluation of Maltese 1990, XXXI: ‘Benché in più punti inadegua-

How Many Books Does It Take to Make an Emperor’s Library? now discuss, may perhaps illuminate the reasons why this more recent chapter division was introduced. 3 Nilus of Ancyra’s Treatise in fifty-three chapters The first Paraphrase is attributed to a certain Nilus, and due to the typical tendency to group works together under the same author, this Paraphrase is also included within the proteiform corpus of Nilus of Ancyra.28 Among the works of this monk, who lived between the fourth and fifth centuries, there is an immense Correspondence, one of the largest collections from Late Antiquity, consisting of more than one thousand letters.29 Nonetheless, it is wise to question the authenticity of these letters, as they pose numerous historical and editorial problems, and in several cases, they are nothing but excerpts from other authors’ work.30 The issue, though hard to solve, must be discussed in the light of a further aspect worth mentioning. Many manuscripts transmit, under the name of Nilus, a text known as Treatise in Fifty-Three Chapters. This Treatise is as yet unpublished on its own,31 because its chapters are nothing else than extracts from Nilus’s Correspondence.32 It is therefore quite clear that this work is a later redactional product, assembled from the juxtaposition of numerous letters (and this explains why it has remained marginal so far). However, its specific interest here lies in the text distribution over exactly 53 chapters. Furthermore, the history of the text shows a striking phenomenon: in the oldest manuscript, which dates back to the tenth century (Città del Vaticano, BAV, Vat. gr. 1524), the work is presented not as a Treatise in Fifty-Three Chapters but simply as a selection of letters [Ἐκ τῶν ἐπιστολῶν τοῦ ἁγίου πατρὸς ἡμῶν Νείλου], and the numbering in the margins of that manuscript indicates that the total number was not equal to 53. That means that subsequently, the manuscripts gradually started to split the text to achieve this number. Let us look closely at two examples: ta, si è mantenuta la tradizionale articolazione in capitoli e paragrafi’. 28 For an overview of the works and the historical figure of St. Nilus, see Heussi 1917; Stählin 1924; Bardenhewer 1924, 161–178; Disdier 1931; Quasten 1963, 496-504; Guérard 1982; and the introduction to Guérard 1994. For an orientation on the pseudepigraphs of this corpus: Bossina 2011. 29 Patrologia Graeca 79, 81–581. See Gribomont 1969; Fatouros 2003. 30 Cameron 1976; Bossina 2013; Bossina 2017; Bossina 2019. 31 Although it was known to Nilus’s seventeenth-century editor: see J. M. Suares, Dissertatio de operibus sancti Nili quae e graeco latina fecit et primus edidit, in Patrologia Graeca, vol. 79, cols 1354–1356. 32 Heussi 2017, 44–45; Gribomont 1969, 248–251.

Luciano Bossina Nilus of Ancyra ep. III 283 (= 242 Poussines) 53 cap. Ἀλλ’ ἡμεῖς τοῦτο ἐπιστάμενοι, ἀντιστρατευσώμεθα τῷ ἡμετέρῳ ἐχθρῷ, καὶ ὅταν σταθῶμεν εἰς προσευχὴν ἢ καὶ γόνατα κλίνωμεν, μηδένα λογισμὸν εἰς τὴν καρδίαν ἡμῶν εἰσελθεῖν συγχωρήσωμεν, μὴ λευκὸν μὴ μέλανα, μὴ δεξιὸν μὴ ἀριστερόν, μὴ γραφικὸν μὴ ἄγραφον, πλὴν τῆς πρὸς Θεὸν ἱκεσίας καὶ ἐνατενίσεως καὶ τῆς ἐκ τοῦ οὐρανοῦ ἐγγινομένης τῷ ἡγεμονικῷ ἐλλάμψεως καὶ ἡλιοβολίας. Ῥίψαντες τοίνυν πᾶσαν ἀφορμὴν καὶ πᾶσαν ὀκνηρίαν, ἀκηδίαν καὶ εὐλογοφανίαν, τῷ μεγάλῳ ἔργῳ τῆς προσευχῆς νηφόντως καὶ θερμῶς σχολάσωμεν, ὅπερ ἐστὶ ῥίζα ἀθανασίας. 26: Ἀλλ’ ἡμεῖς τοῦτο ἐπιστάμενοι, ἀντιστρατευσώμεθα τῷ ἡμετέρῳ ἐχθρῷ, καὶ ὅταν σταθῶμεν εἰς προσευχὴν ἢ καὶ γόνατα κλίνωμεν, μηδένα λογισμὸν εἰς τὴν καρδίαν ἡμῶν εἰσελθεῖν συγχωρήσωμεν, μὴ λευκὸν μὴ μέλανα, μὴ δεξιὸν μὴ ἀριστερὸν, μὴ γραφικὸν μὴ ἄγραφον, πλὴν τῆς πρὸς Θεὸν ἱκεσίας καὶ ἐνατενίσεως καὶ τῆς ἐκ τοῦ οὐρανοῦ ἐγγινομένης τῷ ἡγεμονικῷ ἐλλάμψεως καὶ ἡλιοβολίας. 27: Ῥίψαντες πᾶσαν ἀφορμὴν καὶ πᾶσαν ὀκνηρίαν, ἀκηδίαν καὶ εὐλογοφανίαν, τῷ μεγάλῳ ἔργῳ τῆς προσευχῆς νηφόντως καὶ θερμῶς σχολάσωμεν, ὅπερ ἐστὶ ῥίζα ἀθανασίας. The original letter 242 was divided into two different chapters (26 and 27). The editorial intervention may be further detected through a revealing piece of evidence, that is the deletion of the word τοίνυν. Indeed, the person who split up the text into two sections erased the term which provided a logical link in the discourse. As a result, two texts were created from a single homogeneous text. Another example may be drawn from Chapters 37 and 38. Nilus’s original letter 192 is no other than the exact reproduction of a passage from Gregory of Nyssa’s On Virginity.33 It was later divided into two different chapters: Greg. Nyss. de virg. 22, 2 Nilus of Ancyra ep. III 268 (= 192 Poussines) 53 cap. μήτε διὰ τῆς ἀμέτρου κακοπαθείας νοσώδη καὶ λελυμένην καὶ ἄτονον πρὸς τὴν ἀναγκαίαν ὑπηρεσίαν κατασκευάσῃ. οὗτος ὁ τελεώτατος τῆς ἐγκρατείας σκοπός, οὐχὶ πρὸς τὴν τοῦ σώματος βλέπειν κακοπάθειαν, καὶ κατάλυσιν, ἀλλὰ πρὸς τὴν τῶν ψυχικῶν διακινημάτων εὐκολίαν. μήτε διὰ τῆς ἀμέτρου κακοπαθείας νοσώδη καὶ λελυμένην καὶ ἄτονον πρὸς τὴν ἀναγκαίαν ὑπηρεσίαν κατασκευάσῃ. οὗτος ὁ τελειότατος τῆς ἐγκρατείας σκοπός, οὐ τὸ βλέπειν πρὸς τὴν τοῦ σώματος κακοπάθειαν καὶ κατάλυσιν καὶ παντελῆ ἀχρείωσιν, ἀλλὰ πρὸς τὴν τῶν ψυχικῶν κινημάτων εὐκολίαν. 37: μήτε διὰ τῆς ἀμέτρου κακοπαθείας νοσώδη καὶ λελυμένην καὶ ἄτονον πρὸς τὴν ἀναγκαίαν ὑπηρεσίαν κατασκευάσας. 38: Οὗτος ὁ τελειότατος τῆς ἐγκρατείας σκοπός, οὐ πρὸς τὴν τοῦ σώματος βλέπειν κακοπάθειαν καὶ κατάλυσιν καὶ παντελῆ ἀχρείωσιν, ἀλλὰ πρὸς τὴν τῶν ψυχικῶν κινημάτων εὐκολίαν. 33 Edition and translation: Aubineau 1966.

How Many Books Does It Take to Make an Emperor’s Library? Changes in the text are self-evident: a page by Gregory of Nyssa results into a letter by Nilus, and a letter by Nilus into two chapters of the Treatise. Apparently, the text was progressively divided in order to obtain a specific number of chapters. This phenomenon becomes even clearer when looking at the modifications of the title. In the oldest manuscript, as said before, the work is presented as a selection of letters taken from Nilus’s Correspondence (Ἐκ τῶν ἐπιστολῶν). Later on, the text is presented as a Letter with 53 Chapters and eventually as Exhortations to Monks in 53 Chapters:34 Vat. gr. 1524 (10th–11th c.) Marc. gr. 131 (11th c.) Laur. Plut. IX. 18 (12th c.) Ἐκ τῶν ἐπιστολῶν τοῦ ἁγίου πατρὸς ἡμῶν Νείλου τοῦ ἁγίου Νείλου ἐπιστολὴ ἔχουσα κεφάλαια νγ´ τοῦ αὐτοῦ [Νείλου] πρὸς μονάζοντα παραινέσεις κεφαλαίοις τρισὶ καὶ πεντήκοντα The editorial history of this Treatise therefore reveals two coinciding phenomena: (a) the deletion of the epistolary and plural nature of the original text; (b) the artificial attainment of the number 53. Such a process is almost analogous to that of Epictetus’s Handbook. Is this merely a ‘lucky coincidence’? 4 Evagrius Ponticus’s De oratione and the symbolic value of the distribution of a work A work that gained enormous success in the Byzantine tradition (‘more than 120 manuscripts from the ninth to the nineteenth century’)35 might help to clarify things: Evagrius Ponticus’s De oratione. It should be noticed that in its Greek manuscript tradition this text was transmitted under Nilus’ name, since Evagrius was 34 To understand this progressive transformation of the literary genre, it should be noticed that the text in the form of a treatise was then adapted into the form of a homily and translated into Georgian by the famous translator Euthymius the Athonite: Tarchišvili 1955, 144 (no. 14), Gribomont 1969, 250. 35 Géhin 2017, 73 (index of Greek manuscripts 407–421): the text was also passed down in the Syriac (two different versions), Armenian, Arabic (three different versions), Georgian (two different versions) Ethiopian, and Slavic traditions. A complete and up-to-date overview is offered by the excellent edition of Géhin 2017, who also promised ‘une étude détaillée […] qui rassemblera tous les matériaux relatifs au traité, en grec et dans les versions orientales et slaves’ (73).

Luciano Bossina included in the condemnations of the Council of Constantinople in 553, and his oeuvre was consequently doomed to damnatio memoriae. The attribution to Nilus thus saved it from destruction. On the other hand, De oratione was also translated into Arabic and Syriac, traditions which did not recognize the ecclesiastical condemnation, thus preserving its true authorship in Oriental languages.36 In the prologue, the author proposes a highly complex numerological interpretation37 to explain why he organised his work into a specific number of chapters. As a friend asked him to write a treatise on prayer, Evagrius decided to divide it into 153 chapters, which directly references the Gospel of Saint John (John 21: 1–14). Following Jesus’s death, the apostles go fishing, without catching anything. Then, a man appears before the boat, and the apostles recognize that he is Christ resurrected. Meanwhile, Peter throws his net once again and fishes 153 fishes. Based on the evangelical story, Evagrius developed a highly elaborate numerological interpretation:38 Ἐγὼ δὲ οὐκ ἀρνηθείην, ὡς ὅλην τὴν νύκτα κοπιάσας πεπίακα οὐδέν· πλὴν ἀλλ’ ἐπὶ τῷ σῷ λόγῳ χαλάσας τὰ δίκτυα, ἤγρευσα ἰχθύων πλῆθος, οὐκ οἶμαι μὲν μεγάλων, ἑκατὸν δὲ ὅμως καὶ πεντηκοντατριῶν καὶ τούτους ἐξαπέστειλα ἐν τῇ σπυρίδι τῆς ἀγάπης, διὰ τῶν ἰσαρίθμων κεφαλαίων, τὴν πρόσταξιν πεπληρωκώς. [...] As for myself, I would not deny that having toiled all night I have caught nothing. Yet at your word I have let down the nets and caught a great quantity of fish; they are not big, I think, but there are still one hundred and fifty-three. I have sent these to you in a basket of love arranged in an equal number of chapters in fulfilment of your order. […] ἀλλ’ ἐπεὶ πάντα δισσά, ἓν κατ’ ἔναντι τοῦ ἑνὸς κατὰ τὸν σοφὸν Ἰησοῦν, δέχου πρὸς τῷ γράμματι καὶ τὸ πνεῦμα. Σύνες ὡς πάντως τοῦ γράμματος νοῦς προηγεῖται· οὐκ ὄντος γὰρ τούτου, οὐδὲ γράμμα ἔσται. Οὐκοῦν καὶ προσευχῆς διττὸς ὁ τρόπος, ὁ μέν τις πρακτικός, ὁ δὲ θεωρητικός· οὕτως καὶ ἀριθμοῦ, τὸ μὲν πρόχειρός ἐστι ποσότης, τὸ δὲ σημαινόμενον ποιότης. But since all things come in pairs, one opposite the other [Sir. 42:24], according to the wise Jesus, accept them according to the letter and according to the spirit; understand, that intelligence is prior to any writing, for if this were not so there would be no written work. The way of prayer, therefore, is also twofold: it involves the practical on the one hand and the contemplative on the other. Similarly, in the case of number, the immediate sense indicates quantity, but the meaning can refer to quality. 36 A decisive role in re-attributing this work to Evagrius was played by the studies of Hausherr 1934, 1939, 1960. Regarding Evagrius’s ecclesiastical condemnation and the censorship of his works, the reader is referred to Guillaumont 1962, 166–168. 37 The text is so complex that certain copyists passed on the work without the prologue or copied it with numerous mistakes. Even from a stemmatic point of view, the prologue constitutes an exception in the tradition of De oratione: Géhin 2017, 156–163. 38 Text by Géhin 2017, 210–214; translation by Sinkewicz 2003, 191–192.

How Many Books Does It Take to Make an Emperor’s Library? Εἰς ἑκατοστὸν πεντηκοστὸν τρίτον τὸν περὶ προσευχῆς λόγον διειληφότες, εὐαγγελικὸν ὀψώνιόν σοι πεπόμφαμεν, ἵνα εὕρῃς συμβολικοῦ ἀριθμοῦ τερπνότητα καὶ σχῆμα τρίγωνον καὶ ἑξάγωνον, ὁμοῦ μὲν εὐσεβῆ γνῶσιν τριάδος, ὁμοῦ δὲ καὶ τῆσδε τῆς διακοσμήσεως τὴν περιγραφὴν ὑπεμφαῖνον. Having divided this treatise on prayer into one hundred and fifty-three chapters, we have sent you an evangelical feast (cf. John 21:12–13), that you might discover the delightfulness of the symbolic number as well as the figure of the triangle and the hexagon: the former indicating the pious knowledge of the Trinity and the latter the description of the ordering of the present world. Ἀλλ’ ὁ ἑκατοστὸς ἀριθμὸς καθ’ ἑαυτὸν τετράγωνός ἐστιν· ὁ δὲ πεντηκοντὸς τρίτος, τρίγωνος καὶ σφαιρικός· ὁ γὰρ εἰκοστὸς ὄγδοος μὲν τρίγωνος, σφαιρικὸς δὲ ὁ εἰκοστὸς πέμπτος· πεντάκις γὰρ πέντε, εἰκοσιπέντε. The number 100 in itself is a square, while the number 53 is triangular and spherical, for 28 is triangular and 25 is spherical, for 5 × 5 = 25. Οὐκοῦν ἔχεις τὸ τετράγωνον σχῆμα, οὐ μόνον διὰ τῆς τετρακτύος τῶν ἀρετῶν, ἀλλὰ καὶ τοῦδε τοῦ αἰῶνος τὴν ἔνσοφον γνῶσιν τῷ εἰκοστῷ πέμπτῳ ἀριθμῷ ἐοικυῖαν, διὰ τὸ σφαιρικὸν τῶν χρόνων. Ἑβδομὰς γὰρ ἐπὶ ἑβδομάδα καὶ μὴν ἐπὶ μῆνα δινεῖται καὶ ἐξ ἐνιαυτοῦ εἰς ἐνιαυτὸν ὁ χρόνος κυλινδεῖται καὶ καιρὸς ἐπὶ καιρόν, ὡς ἐπὶ κινήσεως ἡλίου καὶ σελήνης, ἔαρος καὶ θέρους καὶ τῶν ἑξῆς ὁρῶμεν. You then have a square figure not only for the fourfold of the virtues but also for a wise knowledge of the present age, represented by the number 25 on account of the cyclical nature of time periods; for week moves on to week and month to month, and time rolls round from year to year; and season follows season, as we see in the movement of the sun and moon, of spring and summer, and so on. Τὸ δὲ τρίγωνον σημαίνοι ἄν σοι τὴν τῆς ἁγίας τριάδος γνῶσιν The triangle might indicate to you the knowledge of the Holy Trinity. Apart from some discrepancies, Evagrius’s entire numerological interpretation39 resorts to Nicomachus of Gerasa’s Introduction to Arithmetic (first-second century)40 and establishes the very structure of De oratione. The passage above is therefore based on the following principles: (a) first of all, the number expresses the quantity [ποσότης], but its true meaning relies on the quality [ποιότης]; (b) for this reason, the organization of a work is not only an external and material 39 Regarding this interpretation, see Muyldermans 1952, 41–46, Zigmund-Cerbu 1961, Sinkewicz 2003, 274–275, Géhin 2017, 375–381. 40 For the text by Nichomacus: Bertier 1978. Evagrius differs from Nichomacus in the concept of ‘theoretical’ number: for more details, see Géhin 2017, 378.

Luciano Bossina matter but also a structural one in the deepest sense of the term, as it conforms to a ‘symbolic number’ [ἀριθμὸς συμβολικός]; (c) the number 53 is symbolic because it is the sum of 25 and 28, in other words, a spherical number and a triangular number; (d) the spherical number is the product of a number multiplied by itself (5×5 = 25) and represents the ‘wise knowledge of the present age’; (e) the triangular number is ‘one equal to the sum of all preceding successive numbers starting with 1’41 (28 = 1+2+3+4+5+6+7), as in the following diagram: . 1 1 . . . 1+2=3 1+2 = 3 . . . . . . 1+2+3=6 1+2+3 = 6 . . . . . . . . . . 1+2+3+4=10 1+2+3+4 = 10 In Evagrius’ interpretation, the triangular number represents the ‘knowledge of the Holy Trinity’. It must be concluded that the sum of a triangular number and a spherical number represents the sum of ‘the knowledge of the Holy Trinity’ and ‘the knowledge of the present age’: 28+25 = 53. In other words, this number thus guarantees knowledge of the heavens and the earth. It seems rather inconceivable that behind these cases there be simply a lucky coincidence. On the contrary, between these three works – Epictetus’s Handbook, the Treatise in 53 Chapters, and De Oratione (all of which, whether correctly or incorrectly, were passed down under the name of the same author, Nilus) – a certain continuity should be assumed. The Christianization of Epictetus’s Handbook not only involved, now and then, adjustments of the text, according to the above-mentioned methods, but it also concerned the internal division into chapters. The architecture of the text was adapted to a symbolic number, following the same phenomenon that took place in the tradition of the Treatise in 53 Chapters, and on which the allegorical interpretation of De Oratione depended. All of these works are presented to the reader in a structure that itself clearly claims to be exhaustive. Thus, the physical organization of the text, namely its subdivision based on a symbolic number, itself conveys the meaning of the text. 41 Sinkewicz 2003, 274.

How Many Books Does It Take to Make an Emperor’s Library? 5 Constantine’s 53 volumes Now that this long trip has come to an end, we can get back to Constantine and read the prologue to his work from a new perspective. What may be specifically inferred is that the division into 53 sections (‘the material was distributed into principal topics, fifty-three in number’), as forcefully emphasized in the prologue opening each of the sections, is neither a product of chance nor is it meaningless, but rather it acquires a specific programmatic value. Here is the text of the prologue:42 Ὅσοι τῶν πάλαι ποτὲ βασιλέων τε καὶ ἰδιωτῶν μὴ τὸν νοῦν παρεσύρησαν ἡδοναῖς, ἢ κατεμαλακίσθησαν, ἀλλὰ τὸ τῆς ψυχῆς εὐγενὲς ἀκηλίδωτον ἀρετῇ συνετήρησαν, οὗτοι δὴ οὗτοι καὶ πόνοις ἐνεκαρτέρησαν καὶ λόγοις ἐνησχολήθησαν, καὶ ἄλλος ἄλλο τι τῶν ὅσοι λογικώτερον ἐπεβίωσαν παιδείας ἐρασταὶ γεγονότες σπουδαιότερόν τινα συνεγράψαντο, τοῦτο μὲν τῆς σφῶν αὐτῶν πολυμαθίας δεῖγμα ἐναργὲς τοῖς μετέπειτα καταλιπεῖν ἱμειρόμενοι, τοῦτο δὲ καὶ εὔκλειαν ἀείμνηστον ἐκ τῶν ἐντυγχανόντων καρπώσασθαι μνώμενοι. All those, both among the emperors of old and the persons of no public station, who did not allow their mind to be turned aside or weakened by pleasure, but who by dint of virtue preserved unblemished the noble quality of their soul, propelled themselves into action or gave themselves over to literary activities. Of these, the ones who undertook literary pursuits as a consequence of having been passionate about acquiring knowledge have – each in his own manner – written something remarkable. <This was> both so as to leave to posterity some brilliant proof of their vast learning, and at the same time with the intention of reaping the fruit of an imperishable fame among those who would read their work. ἐπεὶ δὲ ἐκ τῆς τῶν τοσούτων ἐτῶν περιδρομῆς ἄπλετόν τι χρῆμα καὶ πραγμάτων ἐγίγνετο καὶ λόγων ἐπλέκετο, ἐπ’ ἄπειρόν τε καὶ ἀμήχανον ἡ τῆς ἱστορίας ηὐρύνετο συμπλοκή, ἔδει δ’ ἐπιρρεπέστερον πρὸς τὰ χείρω τὴν τῶν ἀνθρώπων προαίρεσιν μετατίθεσθαι χρόνοις ὕστερον καὶ ὀλιγώρως ἔχειν πρὸς τὰ καλὰ καὶ ῥᾳθυμότερον διακεῖσθαι πρὸς τὴν τῶν φθασάντων γενέσθαι κατάληψιν, κατόπιν γινομένης τῆς ἀληθοῦς ἐπιτεύξεως, ὡς ἐντεῦθεν ἀδηλίᾳ συσκιάζεσθαι τὴν τῆς ἱστορίας ἐφεύρεσιν, πῆ μὲν σπάνει βίβλων ἐπωφελῶν, πῆ δὲ πρὸς τὴν ἐκτάδην πολυλογίαν δειμαινόντων καὶ κατορρωδούντων, With the passage of so many years, however, the number of events has become uncountable and the writings have become more complex, the fabric of the history has been infinitely magnified to the point of becoming unmanageable. Thus people’s inclinations and their choices have with time inclined ever more towards the worse, and they have become indifferent to the good and careless of understanding the events of the past. As a consequence of the truth being less accessible, the investigation of history is rendered obscure, both by lack of the useful books and because of people fearing and dreading their extreme complexity. 42 Németh 2018 (edition: 267–268; translation: 61–62).

Luciano Bossina ὁ τῆς πορφύρας ἀπόγονος Κωνσταντῖνος, ὁ ὀρθοδοξότατος καὶ χριστιανικώτατος τῶν πώποτε βεβασιλευκότων, ὀξυωπέστερον πρὸς τὴν τῶν καλῶν κατανόησιν διακείμενος καὶ δραστήριον ἐσχηκὼς νοῦν ἔκρινε βέλτιστον εἶναι καὶ κοινωφελὲς τῷ τε βίῳ ὀνησιφόρον, πρότερον μὲν ζητητικῇ διεγέρσει βίβλους ἄλλοθεν ἄλλας ἐξ ἁπάσης ἑκασταχοῦ οἰκουμένης συλλέξασθαι παντοδαπῆς καὶ πολυειδοῦς ἐπιστήμης ἐγκύμονας, ἔπειτα τὸ τῆς πλατυεπείας μέγεθος καὶ ἀκοὰς ἀποκναῖον ἄλλως τε καὶ ὀχληρὸν καὶ φορτικὸν φαινόμενον τοῖς πολλοῖς δεῖν ᾠήθη καταμερίσαι τοῦτο εἰς λεπτομέρειαν ἀνεπιφθόνως τε προθεῖναι πᾶσι κοινῇ τὴν ἐκ τούτων ἀναφυομένην ὠφέλειαν, ὡς ἐκ μὲν τῆς ἐκλογῆς προσεκτικωτέρως καὶ ἐνδελεχέστερον κατεντυγχάνειν τοὺς τροφίμους τῶν λόγων καὶ μονιμώτερον ἐντυποῦσθαι τούτοις τὴν τῶν λόγων εὐφράδειαν, μεγαλοφυῶς τε καὶ εὐεπηβόλως πρὸς ἐπὶ τούτοις καταμερίσαι εἰς ὑποθέσεις διαφόρους, τρεῖς ἐπὶ τοῖς πεντήκοντα τὸν ἀριθμὸν οὔσας, ἐν αἷς καὶ ὑφ’ αἷς ἅπασα ἱστορικὴ μεγαλουργία συγκλείεται. So it is that Constantine, born in the purple, that most orthodox and most Christian of emperors up to the present time, fitted to the task by extremely keen discernment regarding what is good and possessing an enterprising intellect, judged that the best thing, the most conducive to the common good and useful for governing conduct is – in the first place – to collect by means of diligent research all manner books from all over the known world, books teeming with every kind and variety of knowledge. Next, he thought it necessary to divide and distribute their great quantity and extent, which weigh heavily on the understanding and seem too many to be irksome and burdensome, into small sections. Hence, the profit of this fertile material could <he thought> be made available unstintingly to common use, so that, by the virtue of the selection, they might find more carefully and persistently the nourishment of texts, while the beauty of the texts could be more permanently impressed upon them. In addition, <his intention was> to distribute [the material] in an ingenious and careful manner into principal topics, fifty-three in number, in and through which all the great achievements of history might be grouped together. κοὐκ ἔστιν οὐδὲν τῶν ἐγκειμένων, ὃ διαφεύξεται τὴν τοιαύτην τῶν ὑποθέσεων ἀπαρίθμησιν, οὐδὲν τὸ παράπαν ἀφαιρουμένης τῆς τοῦ λόγου ἀκολουθίας τῇ διαιρέσει τῶν ἐννοιῶν, ἀλλὰ σύσσωμον σωζούσης καὶ ἑκάστῃ ὑποθέσει προσαρμοζομένης τῆς τηλικαύτης οὐ συνόψεως, ἀληθέστερον δ’ εἰπεῖν οἰκειώσεως. Nothing contained in the texts will escape this distribution into topics; <since> by the division of the content this procedure omits nothing of the continuous narration, but rather preserves it in a corpus and establishes the correspondence with each topic, it is not a summary but, to speak more properly, an appropriation. ὧν κεφαλαιωδῶν ὑποθέσεων ἡ προκειμένη αὕτη καὶ ἐπιγραφομένη περὶ πρέσβεων Ῥωμαίων πρὸς ἐθνικοὺς τυγχάνει οὖσα ἑβδόμη ἐπὶ τοῖς εἴκοσι. Of these principal topics, the present text, bearing the heading On Embassies of Romans to Foreigners occurs as [number] twenty-seven.

How Many Books Does It Take to Make an Emperor’s Library? It is well known that the Christian religion (but not only the Christian religion) developed many forms of numerology over the centuries. In ancient Christianity, the Alexandrine exegetical tradition was particularly keen on this type of allegorical interpretation and played a crucial role in spreading it. In the case presented here, there is more to be glimpsed, though, because it cannot be simply included among the numerous symbolic interpretations that necessarily characterized biblical exegesis (with the number three symbolizing the Trinity, the number seven as the image of the week of creation, the number forty as a reference to the Flood or Christ’s temptation in the desert, etc.). Here, the most interesting aspect is the direct relationship between the symbolism of the number and the organization of a work (in books or chapters). Such symbolism becomes a phenomenon that influences book history, and text editors, in turn, cannot help but take it into consideration. Based on a tradition that had already known previous illustrious figures, Constantine wanted to organize a collection that allowed him to be familiar with and manage universal history in the form of a ‘book-library’, a collection professing its aims through its material structure. On the other hand, the prologue to De cerimoniis explains that it is the order itself, the τάξις, that portrays the imperial power as magnificent in the eyes of subjects and foreigners.43 This is why Constantine delved into all past and present documentation so as to organize the material in a way that made them easy to understand: ‘conducted with rhythm and order, the imperial power could thus reproduce the harmony and the movement given to the universe by the Creator’.44 Therefore, the world was truly in the hands of the Emperor who, in his library, contemplated the earthly and divine order of universal knowledge. References Aubineau, Michel (ed.) (1966), Grégoire de Nysse, Traité de la virginité (Sources chrétiennes, 119), Paris: Cerf. Bardenhewer, Otto (1924), Geschichte der altkirchlichen Literatur, vol. 4, Freiburg: Herder. Bertier, Janine (ed.) (1978), Nicomaque de Gérase, Introduction arithmétique (Histoire des doctrines de l'Antiquité classique, 2), Paris: Vrin. Boissevain, Ursulus Philippus (1906), Excerpta historica iussu imperatoris Constantini Porphyrogeniti confecta, vol. 4: Excerpta de sententiis, edidit Ursulus Philippus Boissevain, Berlin: apud Weidmannos. 43 See Németh 2018, 139–141. 44 Const. Porphyr. De cer. I, praef.: ὑφ’ ὧν τοῦ βασιλείου κράτους ῥυθμῷ καὶ τάξει φερομένου, εἰκονίζοι μὲν τοῦ δημιουργοῦ τὴν περὶ τόδε τὸ πᾶν ἁρμονίαν καὶ κίνησιν.

Luciano Bossina Bossina, Luciano (2004), Nilo di Ancira. Ricerche di storia e filologia, unpublished PhD Dissertation, Università degli Studi di Torino. Bossina, Luciano (2011), ‘Falsi antichi e moderni tra le opere di Nilo di Ancira. Primi sondaggi’, in Javier Martínez (ed.), Fakes and Forgers of Classical Literature: Ergo decipiatur!, Madrid: Ediciones Clásicas, 61–78. Bossina, Luciano (2013), ‘Il carteggio di Nilo di Ancira con il generale Gainas è un falso’, in Antonio Rigo, Andrea Babuin and Michele Trizio (eds), Vie per Bisanzio. Atti del VII Congresso nazionale dell’Associazione Italiana di Studi Bizantini, Bari: Edizioni di pagina, 215–249. Bossina, Luciano (2017), ‘Nilo, Crisostomo e altre lettere false’, in Francesca P. Barone, Caroline Macé and Pablo A. Ubierna (eds), Philologie, herméneutique et histoire des textes entre Orient et Occident. Mélanges en hommage à Sever J. Voicu (Instrumenta Patristica et Mediaevalia, 73), Turnhout: Brepols, 823–849. Bossina, Luciano (2019), ‘L’eresia a Costantinopoli tra fine IV e inizio V secolo. Note su qualche fonte poco esplorata’, Studi e Materiali di Storia delle Religioni, 85: 136–164. Boter, Gerard (ed.) (1999), The Encheiridion of Epictetus and its Three Christian Adaptations (Philosophia Antiqua, 82), Leiden: Brill. Boter, Gerard (ed.) (2007), Epictetus, Encheiridion (Bibliotheca Teubneriana), Berlin: De Gruyter. Boter, Gerard (2011), ‘Epictetus’, in Catalogus Translationum et Commentariorum, vol. 9, Washington: The Catholic University of America Press, 1–54. Brittain, Charles and Tad Brennan (2002), Simplicius, On Epictetus’ Handbook, New York: Cornell University Press. Büttner-Wobst, Theodor (1906), ‘Die Anlage der historischen Encyklopädie des Konstantinos Porphyrogennetos’, Byzantinische Zeitschrift, 15: 88–120. Büttner-Wobst, Theodor and Antonius Gerardus Roos (1906–1910), Excerpta historica iussu imperatoris Constantini Porphyrogeniti confecta, vol. 2: Excerpta de virtutibus et vitiis, Berlin: apud Weidmannos. Cameron, Alan (1976), ‘The Authenticity of the Letters of St Nilus Ancyra’, Greek Roman and Byzantine Studies, 17: 181–196. Ceulemans, Reinhart and Peter Van Deun (2017), ‘Réflexions sur la littérature anthologique de Constantin V à Constantin VII’, in Flusin and Cheynet (eds) 2017, 361–388. De Boor, Carolus (1903), Excerpta historica iussu imperatoris Constantini Porphyrogeniti confecta, vol. 1: Excerpta de legationibus, Berlin: apud Weidmannos. De Boor, Carolus (1906), Excerpta historica iussu imperatoris Constantini Porphyrogeniti confecta, vol. 3: Excerpta de insidiis, Berlin: apud Weidmannos. De Nicola, Francesco (1998), ‘Osservazioni critico-esegetiche alla Parafrasi Cristiana del Manuale di Epitteto’, Bollettino dei Classici, 19: 35–61. De Nicola, Francesco (2007), ‘Per la fortuna di Dione Crisostomo e dell’Epitteto cristiano a Bisanzio’, AION. Annali dell’Università degli studi di Napoli ‘l’Orientale’, 29: 95–111. Disdier, M. Th. (1931), ‘Nil l’Ascète ou Nil d’Ancyre dit Le Sinaite’, in Dictionnaire de Théologie Catholique, vol. 11, Paris: Letouzey et Ané, 661–674. Fatouros, Georgios (2003), ‘Zu den Briefen des hl. Neilos von Ankyra’, in Wolfram Hörandner and Michel Grünbart (eds), L’Épistolographie et la poésie épigrammatique: projets actuels et questions de méthodologie. Minutes of the 16th Annual Round Table organized by Wolfram Hörandner and Michel Grünbart (XXe Congrès international des Études byzantines. Collège de France – Sorbonne, Paris, 19–25 Août 2001), Paris, 21–30. Flusin, Bernard and Jean-Claude Cheynet (eds) (2017), Autour du “Premier humanisme byzantin” et des “Cinq études sur le 11e siècle”, quarante ans après Paul Lemerle, Paris 2017.

How Many Books Does It Take to Make an Emperor’s Library? Géhin, Paul (ed.) (2017), Évagre le Pontique, Chapitres sur la prière (Sources chrétiennes, 589), Paris: Cerf. Gribomont, Jean (1969), ‘La tradition manuscrite de saint Nil. I. La correspondance’, Studia monastica, 11: 231–267. Guérard, Marie-Gabrielle (1982), ‘Nil d’Ancyre’, in Dictionnaire de Spiritualité Ascétique et Mystique, vol. 11, Paris: Beauchesne, 354–356. Guillaumont, Antoine (1962), Les ‘Kephalaia gnostica’ d’Évagre le Pontique et l’histoire de l’origénisme chez les Grecs et chez les Syriens, Paris: Seuil. Hadot, Ilsetraut (1978), ‘La tradition manuscrite du commentaire de Simplicius sur le Manuel d’Épictète’, Revue d’histoire des textes, 8: 1–108. Hadot, Ilsetraut (1987), ‘La vie et l’œuvre de Simplicius d’après des sources grecques et arabes’, in Ilsetraut Hadot (ed.), Simplicius. Sa vie, son œuvre, sa survie (Peripatoi: PhilologischHistorische Studien zum Aristotelismus, 15), Berlin: De Gruyter, 3–39. Hadot, Ilsetraut (ed.) (2001), Simplicius, Commentaire sur le Manuel d’Épictète, vol. 1: Chapitres 1-29, Paris: Les Belles Lettres. Hadot, Pierre (ed.) (2000), Arrien, Manuel d’Épictète, Paris: Librairie Générale Française. Hausherr, Irénée (1934), ‘Le traité de l’oraison d’Évagre le Pontique (Pseudo Nil)’, Revue d’Ascétique et de Mystique, 15: 34–170. Hausherr, Irénée (1939), ‘Le De oratione d’Évagre le Pontique en syriaque et en arabe’, Orientalia Christiana Periodica, 5: 7–71. Hausherr, Irénée (1960), Les leçons d’un contemplatif. Le Traité de l’Oraison d’Évagre le Pontique, Paris: Beauchesne. Heussi, Karl (1917), Untersuchungen zu Nilus dem Asketen, Leipzig: Hinrichs. Jenkins, Romilly James Heald and František Dvornik (eds) (2012), Constantine Porphyrogenitus, De Administrando Imperio: A Commentary, Washington: Dumbarton Oaks. König, Jason and Greg Woolf (eds) (2013), Encyclopaedism from Antiquity to the Renaissance, Cambridge: Cambridge University Press. Lemerle, Paul (1966), ‘L’encyclopédisme à Byzance à l’apogée de l’Empire, et particulièrement sous Constantin VII Porphyrogénète’, Cahiers d’histoire mondiale, 9: 596–616. Lemerle, Paul (1971), Le premier humanisme byzantin. Notes et remarques sur enseignement et culture à Byzance des origines au Xe siècle, Paris: Presses Universitaires de France. Lemerle, Paul (1986), Byzantine Humanism: The First Phase. Notes and remarks on education and culture in Byzantium from its origins to the 10th century, Canberra: Australian Association for Byzantine Studies. Magdalino, Paul (2011), ‘Orthodoxy and History in Tenth-Century Byzantine “Encyclopedism”’, in Van Deun and Macé (eds) 2011, 143–160. Magdalino, Paul (2013), ‘Byzantine Encyclopaedism of the Ninth and Tenth Centuries’, in König and Woolf (eds) 2013, 219–231. Magdalino, Paul (2017), ‘Humanisme et mécénat impérial aux IXe-Xe siècles’, in Flusin and Cheynet (eds) 2017, 3–21. Maltese, Enrico V. (1990), Epitteto, Manuale (con la traduzione latina di Angelo Poliziano e il volgarizzamento di Giacomo Leopardi), Milano: Garzanti. Markopoulos, Athanasios (2017), ‘L’éducation à Byzance aux IXe-Xe siècles : problèmes et questions diverses’, in Flusin and Cheynet (eds) 2017, 53–73. Moravcsik, Gyula (ed.) and Romilly James Heald Jenkins (trans.) (2008), Constantine Porphyrogenitus, De administrando Imperio, Washington: Dumbarton Oaks.

Luciano Bossina Muyldermans, Joseph (1952), Evagriana Syriaca. Textes inédits du British Museum et de la Vaticane, Louvain: Institut Orientaliste. Németh, András (2013), ‘The Imperial Systematisation of the Past in Constantinople: Constantine VII and His Historical Excerpts’, in König and Woolf (eds) 2013, 232–258. Németh, András (2018), The Excerpta Constantiniana and the Byzantine Appropriation of the Past, Cambridge: Cambridge University Press. Odorico, Paolo (1990), ‘La cultura della συλλογή: 1) Il cosiddetto enciclopedismo bizantino. 2) Le tavole del sapere di Giovanni Damasceno’, Byzantinische Zeitschrift, 83: 1–21. Odorico, Paolo (2011), ‘Cadre d’exposition/cadre de pensée : la culture du recueil’, in Van Deun and Macé (eds) 2011, 89–107. Odorico, Paolo (2017), ‘Du premier humanisme à l’encyclopedisme : une construction à revoir’, in Flusin and Cheynet (eds) 2017, 23–43. Oldfather, W. A. (1927), Contributions toward a Bibliography of Epictetus, Urbana: University of Illinois Press. Oldfather, W. A. (1952), Contributions toward a Bibliography of Epictetus. A Supplement, edited by Marian Harman, Urbana: University of Illinois Press. Pertusi, Agostino (ed.) (1952), Costantino Porfirogenito, De thematibus (Studi e testi, 160), Vatican City: Biblioteca Apostolica Vaticana. Piscopo, Mercedes (1969–1970), ‘La tradizione manoscritta della Parafrasi Cristiana del Manuale di Epitteto di S. Nilo’, Helikon, 9–10: 594–599. Piscopo, Mercedes (1972), ‘Utilizzazioni cristiane di Epitteto in alcune parafrasi del Manuale’, in Studi classici in onore di Quintino Cataudella, vol. 2, Catania: Edigraf, 601–605. Piscopo, Mercedes (1978), ‘La tradizione manoscritta della Paraphrasis Christiana del Manuale di Epitteto’, in Jurgen Dummer (ed.), Texte und Textkritik. Eine Aufsatzsammlung, Berlin: Akademie Verlag, 501–508. Quasten, Johannes (1963), Patrology, vol. 3, Utrecht: Westminster Spectrum publishers. Reiske, Iac. (ed.) (1829–1830), Constantini Porphyrogeniti imperatoris de cerimoniis aulae Byzantinae libri duo (Corpus scriptorum historiae Byzantinae, 5), 2 vols, Bonn: Weber. Schreiner, Peter (2011), ‘Die enzyklopädische Idee in Byzanz’, in Van Deun and Macé (eds) 2011, 3–28. Ševčenko, Ihor (ed.) (2011), Chronographiae quae Theophanis Continuati nomine fertur Liber quo Vita Basilii Imperatoris amplectitur, Berlin: De Gruyter. Sinkewicz, Robert E. (2003), Evagrius of Pontus, The Greek Acetic Corpus. Translation, introduction and commentary, Oxford: Oxford University Press. Spanneut, Michel (1972), ‘Épictète chez les moines’, Mélanges de Science Religieuse, 29: 49–57. Spanneut, Michel (ed.) (2007), Commentaire sur la Paraphrase chrétienne du Manuel d’Épictète (Sources chrétiennes, 503), Paris: Cerf. Spieser, Jean-Michel (2017), ‘La “Renaissance macédonienne”: de son invention à sa mise en cause’, in Flusin and Cheynet (eds) 2017, 43–52. Stählin, Otto (1924), ‘Neilos der Asket (Neilos vom Sinai)’, in Wilhelm von Christs and Otto Stählin (eds), Geschichte der griechischen Litteratur, vol. 2/2, Munich: Beck, 1470–1473. Tarchišvili, Michael (1955), Geschichte der kirchlichen georgischen Literatur (Studi e testi, 185), Vatican City: Biblioteca Apostolica Vaticana. Treadgold, Warren T. (1984), Renaissances before the Renaissance. Cultural Revivals of Late Antiquity and the Middle Ages, Stanford: Stanford University Press. Van Deun, Peter and Caroline Macé (eds) (2011), Encyclopedic Trends in Byzantium?, (Orientalia Lovaniensia Analecta, 212), Leuven: Peeters.

How Many Books Does It Take to Make an Emperor’s Library? Vogt, Albert (1967), Constantin VII Porphyrogénète, Le livre des cérémonies, 1–2, Paris: Les Belles Lettres. Wotke, Carl (1892), ‘Handschriftliche Beiträge zu Nilus’ Paraphrase von Epiktets Handbüchlein’, Wiener Studien, 14: 69–74. Zahn, Theodor (1895), Der Stoiker Epiktet und sein Verhältnis zum Christentum, Erlangen: Deichert. Zigmund-Cerbu, Anton (1961), ‘La préface du De oratione d’Évagre. Contribution à l’histoire de l’ésotérisme chrétien’, Acta Philologica (Societas Academica Dacoromana), 2: 251–257.

Inmaculada Pérez Martín Byzantine Libraries: The Public and the Private Abstract: We discuss the nature of the Byzantine libraries by examining their degree of accessibility as witnessed by the loan lists of the monasteries of Saint John Prodromos in Patmos and Saint Nicholas in Casole, and by the typikon of Boilas’ foundation near Edessa. We also draw on the place occupied by books in some monasteries in Egypt (Shenute’s White Monastery in Suhag) and in Byzantine territory, to conclude that their accessibility was non-existent. In the second part of this paper, we review the history and public nature of the Imperial Library of Constantinople: in Late Antiquity it occupied a facility near the Basilike Stoa (centrally located in the city), but later on was replaced by a ‘palace library’ accessible only to members of the imperial family and to palace officers. The use of the terms ‘public’ and ‘private’ when referring to a pre-modern culture requires a few clarifications. Both are concepts that, in the case of Rome or Byzantium, denote realities that are very different from those of today, particularly in the period starting in the nineteenth century. This is equally true for ‘private’ (or ‘family’) and for ‘public’, the meaning of which is shared by other concepts such as civic, community or collective, state, and institutional. Byzantine studies have yet to carry out a general analysis of the meaning of the terms ‘public’ and ‘private’, especially outside of the legal and religious spheres, in order to better convey the meaning of words such as τὸ δημόσιον (to dēmosion), τὸ κοινόν (to koinon), τὸ ἰδιωτικόν (to idiōtikon), or similar ones. In any case, ‘public’ should not a priori be translated as ‘that which is universally accessible’, but rather and more restrictively, as ‘that which is available to a group and which exceeds the usage of its owner and those around him’.1 In order to better determine the degree of accessibility of Byzantine libraries,2 we can look at information on book donations and/or loans, the purpose and use of books, as well as the space dedicated to storing and reading them. In the second section of this paper, we will examine the history of the Imperial Library 1 This reservation with respect to the concept of a ‘public library’ was already stated by Woolf, who considers it ‘profoundly misleading if applied to antiquity’ (Woolf 2013, 5). 2 On Byzantine libraries, see first the classic studies of Wilson 1967; Cavallo 2007, 122–128; and more recently, Grünbart 2018. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-009

Inmaculada Pérez Martín of Constantinople. Even though, contrary to the case of the Roman Empire, we have very little archaeological evidence on Byzantine libraries,3 written documentation and the study of manuscripts offer substantial evidence on their functioning and the use of their collections. 1 The lending system The information available to us to determine the degree of accessibility of Byzantine libraries is very limited, and we have specific elements on only a few monastery libraries.4 Certain typika or monastery founding charters, such as that of the Monastery of Saint John the Forerunner in the Constantinopolitan neighbourhood of Stoudiou, written by Theodore of Stoudiou (Theodore the Studite),5 suggest that Byzantine monastery libraries were reserved for the internal use of the community.6 Yet the dissemination of copies of Theodore’s works across the four corners of the Greek-speaking world, the application of the Stoudiou rule (and the principles of monastic life and spirituality established by Theodore) by a large number of monastic communities, and the letters written by Theodore himself, attesting to the intense circulation of books – they often speak of sending biblia (‘books’) and 3 For the case of Constantinople, see Van Millingen 1912, 264–271, who suggested identifying certain separated buildings with niches in the walls hosting books: these are Balaban Mesjedi, which no longer exists, and Sancaktar Hayrettin Camii (Gastria), a fourteenth-century octagonal building belonging to an architectural complex that appears to have been a mausoleum. Regarding these buildings, see also Müller-Wiener 1977, 194–195 and Kırımtayıf 2001, 41–43. The niches were however actually intended to store corpses and not books. As to the monastery of Chora, the library would hypothetically be located on the upper floor of the north annex of the church: Ousterhout 2002, 119. 4 On the dissemination of texts in monastic circles, see Cavallo 2002; Déroche 2006, 119–123; and Schreiner 2002. The case of the Chora’s library at the beginning of the fourteenth century is special, since the monastery housed the personal library of Metochites, which included many imperial books. On the accessibility of the Chora’s library, see Theodore Metochites, Poem 1, vv. 1113−1116: ‘anyone who has an ardent desire for learning and an unchanging love for the logoi can take whatever he wishes from [the monastery’s library] easily, my library being like a common treasury (tamieiou koinou)’ (Engl. transl. Polemis 2017, 86). 5 Theodore’s typikon of Stoudiou was edited by Mai and Cozza-Luzi 1844–1905, vol. 5, 111–125, repr. Patrologia Graeca, vol. 99, cols 1704–1720; it was translated and commented by Thomas and Constantinides Hero 2000, vol. 1, 83–119. On Theodore the Studite, see Leroy 2002 and Cholij 2002. 6 Eleopoulos 1967, to be completed with the studies of Julien Leroy collected and presented by Delouis 2007; cf. also Perria 1993.

Byzantine Libraries: The Public and the Private tetradia (‘notebooks’)7 –, are evidence against limitations on the use of the library of Stoudiou. They reveal, instead, that it was active and open. Two rare documents consisting of a few loan lists attached to collection inventories show the openness of other Byzantine monastic libraries. Because loan lists are by definition documents that must constantly be updated, very few of them have survived. Their temporary nature meant that people preferred to write them on loose sheets, which were rarely conserved, rather than on more lasting media. These two exceptional monastic documents have nonetheless preserved a few loan lists attached to library inventories, which serve as a ‘snapshot’ of the time when they were written, maybe because book loans were considered long-term loans. The first example is that of the Monastery of Saint John the Theologian on the island of Patmos, founded in 1091 by Christodoulos of Patmos, who along with other monks sought refuge there while fleeing from the Seljuq conquest of Asia Minor.8 There is an exceptional number of documents on the library in Patmos: six inventories of books written between 1103 and 1307 have survived; those of 1200, 1229 (or 1244), 1262 (or 1277), and 1307 were official inventories, whereas those of 1103 and 1118 were gifts from monks joining the community.9 The number and frequency of the creation of these inventories is likely evidence that monks were aware of the value of their books, particularly at this location, which had become a Christian stronghold against Islam. This is also most likely the reason why loan lists have been added to these inventories in different places.10 These lists belong to two categories: the first comprises loans to individuals living on distant islands such as Crete11 or Samos 7 See the edition of Theodore’s letters by Fatouros 1991. The contents of the letters give the impression that the term tetradia refers to written texts or anthologies written by Theodore, whereas the biblia are reference works from his office or library. 8 The monastery’s founding charter (the law, the will, and the codicil of Christodoulos of Patmos, the first of which dates back to May 1091 and the latter two to March 1093) were edited by Miklosich and Müller 1860–1890, vol. 6 (1890), 59–90; an English translation and commentary can be found in Thomas and Constantinides Hero 2000, vol. 2, 564–606. We do not yet have an in-depth study on Patmos, aside from more general works on Byzantine monasteries. The ‘Introduction’ to the edition of the monastery documents written by Vranoussis and Nystazopoulou-Pelekidou 1980 is still a useful summary; see also Morris 1995, 39, 47–50; Smyrlis 2006, 257–260 (an exhaustive portrait of the monastery domains); and Benoit-Meggenis 2017, especially 164–166. 9 Diehl 1892 and Astruc 1981. 10 Astruc 1994 and Waring 2002, 170–176. 11 In 1462, to pay a debt owed to the monks of Patmos, the scribe Georgios Chomatas (better known as Georgios Alexandros), who lived in Crete, provided two manuscripts to the monastery of St John, a Gospel and a book by John of Damascus, which is evidence of the continuity of relations between Patmos and the surrounding region; see Saint-Guillain 2009, 173–176.

Inmaculada Pérez Martín (which bears witness to the scope of Patmos’ cultural influence) as well as to members of rich families in the nearby city of Miletus (currently Balat) on the Asia Minor coast.12 The second category consists of loans that could be called ‘institutional’: Patmos loaned batches of books to its metochia (or ‘dependent monasteries’; these were therefore more or less internal loans) and to other communities on the mainland (Latros, Miletus), and on the Aegean Islands, with which the monastery maintained close relations. These books are liturgical or spiritual in nature, which suggests that they were lent to facilitate the life of monks, given the importance of prayer and religious rites in daily life. Patmos was capable of providing these volumes because its library must have had multiple copies of the most common liturgical texts. Moreover, the dissemination of these works through loans also played an essential role in maintaining Orthodox Christianity and supporting the small, isolated monastic communities in the region, located in hostile territory straddling the Islamic and Latin worlds.13 These documents demonstrate the intense cultural influence that the Monastery of Patmos had over the region. This was possible owing to imperial support in particular,14 but also to the support of the Patriarch of Constantinople, in the form of numerous donations of manuscripts, still conserved at the monastery library (it is one of the few collections of books that has remained in place since it was founded).15 In the founding documents of the Monastery of Patmos, Christodoulos also provides additional information on the origin of some of the library’s books:16 they were from an earlier monastery founded by Christodoulos on the slopes of Mt Latros (or Latmus, in Asia Minor, near the Aegean Sea); during 12 Benoit-Meggenis 2017, 22–23. 13 On the mobility of the monks of Patmos, see Gerolymatou 2019. Regarding the adaptation of Patmos to new geopolitical conditions during the thirteenth century, see Saint-Guillain 2004. During the Ottoman period, Patmos took on the ‘role of sentinel of the Christian world’ due to its strategic location that allowed it to monitor the arrival of pirate ships: cf. Saint-Guillain 2009, 168. 14 In 1088, Emperor Alexios I Komnenos donated the entirety of the island of Patmos to Christodoulos. The patronage by the Komnenoi continued, and the first century of Palaiologan rule (mid-thirteenth to mid-fourteenth century) was also a period of expansion for the monastery: cf. Benoit-Meggenis 2017, 99–100, 164–166; and Saint-Guillain 2004. 15 On the manuscripts of Patmos, see the partial catalogue of Kominis 1988, and the collection of facsimiles published by Kominis 1968–1970. Regarding the inventories of its library, see above n. 10. A few codices were sold in the past and are now a part of other collections. The most famous case is that of Plato’s Clarkianus (Oxford, Bodleian Library, Clarke 39), purchased by Edward Daniel Clarke (1769–1822): cf. Constantinides 2006, 237–238. 16 Thomas and Constantinides Hero 2000, vol. 2, 596 [B7] and 599 [C6], in which we find Christodoulos’ arrangements regarding these books, as well as the inventory that he signed and transferred to his successor, Theodosios, in order to prevent their dispersion.

Byzantine Libraries: The Public and the Private the Turkish conquest of the region (1079), the works were transferred to Constantinople and stored at Hagia Sophia, seat of the Patriarchate. When founding his new monastery, Christodoulos convinced Patriarch Nicholas III (1084–1111) to transfer some of the books to Patmos. By describing them as ‘non-decorated’, he was probably suggesting that they were books for everyday use and study, in other words, not so luxurious and expensive books, so that the patriarch would be more likely to donate them.17 The signatures of dated manuscripts conserved in Patmos attest to the influx into the monastery of books copied elsewhere, particularly in its area of influence or in Constantinople. On the other hand, there is no evidence of the copying of manuscripts at the monastery during the Byzantine period, which may suggest that initially the scriptorium was not very active. In any event, the ‘secretaries and calligraphers’ of the monastery are mentioned in a note in the twelfth-century Patmos, St John Monastery 175, a liturgical book: the copyist, the monk Neilos, states that he completed his task in 1180 and apologises for having done it at another monastery, that of Artamitou or Artamytene (located further to the north on the Asia Minor coast), because he was competing with the ‘secretaries and calligraphers’ from Patmos.18 Nonetheless, it was a captatio benevolentiae with respect to a copying task that ultimately proved to be somewhat irregular. Although Patmos needed secretaries and copyists, an organised and productive scriptorium would have left behind more evidence. Our second example comes from the other end of the Greek world, in southern Italy. It is the Monastery of San Nicola di Casole, also founded at the end of the eleventh century. For three centuries, up until its destruction by the Ottomans in 1480, it was the largest centre of Greek culture in this region of the Strait of Otranto.19 The typikon or founding charter of Casole, a late twelfth-century copy of which is conserved at the Biblioteca Nazionale in Turin,20 draws inspiration 17 One of the books still conserved today at Patmos (Patmiacus 140, a copy of the works of John Chrysostom dated to 1056) was copied at the Monastery of Latros in 1056: cf. Kominis 1968–1970, 20–21. The fact that part of the Library of Patmos is originally from Latros is also confirmed by the famous note of Paris, BnF, gr. 598, which mentions its copying in Latros in 943 by the monk Michael. The note, added at the end of the manuscript by a copyist who restored the last folio (fol. 187), explains that Christodoulos had brought books from the Monastery of Latros to Patmos: cf. Omont 1888, whose dating is inaccurate however. On the copyist, with references to prior publications, see Bianconi 2012, 32. 18 Kominis 1968–1970, 26–27. 19 On the history of the monastery, see Parlangèli and Parlangèli 1951, Arnesano 2008, and Mazzotta 2012. On the cultural reach of the monastery, Jacob 1980. 20 On the typikon of Casole, see Diehl 1886, 173–188, and Omont 1890 (partial edition), and Apostolidis 1983–1984.

Inmaculada Pérez Martín from the rules of the Monastery of Saint John of Stoudiou, whose organization as established by his abbot or hēgoumenos Theodore, served as a model for other monasteries in the Byzantine world, as noted above.21 The provisions regarding its scriptorium and library as contained in the typikon (§ 23) clearly follow the Studite model, yet the particularity of Casole resides in its role as a centre for teaching the Greek language and studying texts. This suggests the existence of a vast library that was assuredly accessible to students, where copies of poems, lexicons, and scientific texts could be obtained. However, as in the case of Patmos, external loans also existed. According to the typikon, during its first fifty years of existence, the monastery library loaned sixty-eight books to the priests and monks in neighbouring communities, to small hermitic laurae in the surrounding area,22 and to judges and notaries.23 This shows that the Casole library was open not only to other monastic communities requiring religious and liturgical works in order to operate, but also to educated people in a society in which Casole was a centre for diffusing Greek culture. In both cases – Patmos and Casole – the loans themselves are potentially the result of a deliberate wish by the monastery to disseminate the writings of one of its founders or to propagate the fundamental texts of Orthodox Christianity, a pillar of the imperial ideology, especially in peripheral or threatened regions such as the Asia Minor coast and Southern Italy during the eleventh and twelfth centuries. Other documents also show the value attributed to books in the Byzantine world and the desire to conserve them by storing them at monastic libraries: wills and donations to religious foundations contain numerous references to books, which are listed among the personal belongings. In this sense, it is worth noting that in Byzantium, most monastery collections enabled laypersons to preserve a portion of their wealth, putting it out of the reach of confiscation and other vicissitudes. However, the majority of these monasteries did not survive for more than a century, experiencing periods of growth and decline. While a few were successful in prolonging their existence through their restoration (this is the case of Stoudiou in Constantinople, mentioned above, the history of which is longer than that of the Byzantine Empire), the majority of monasteries disappeared after one or two generations, and we are now losing evidence of them in the documentation (possibly also due to changes in the monastery’s name). In the wills of members of the aristocracy or affluent groups in society who founded or restored 21 Arnesano 2014 and Chiriatti 2016. 22 Diehl 1886, 186–187. 23 Apostolidis 1983–1984, 68; Chiriatti 2016, 115–119.

Byzantine Libraries: The Public and the Private monasteries, the contradiction is evident between their concern to protect books from the damages of time and use, and their desire for these books to serve as witness to the history and the wealth of their owners. For instance, a member of the Komnenos Dynasty, Sebastokrator Isaak Komnenos (1093–1152), son of Emperor Alexis I (1081–1118), founded a monastery dedicated to the Theotokos Kosmosoteira in Bera (Pherai or Feres) in Thrace in 1152, where he was later buried.24 According to the founding charter or typikon that he wrote, Sebastokrator Isaak mentions the books that he donated to his monastery (without, however, providing an inventory), adding: ‘I bequeathed another book in addition to these [donated books], one that I composed with great effort. It [contains] heroic, iambic and political verse, as well as various letters and ekphraseis. I do not want this [book] to lie in an obscure place, but to be displayed often as [something to] read (and in memory of me) to those especially industrious men (and they [are the ones who] want to come upon books and pictures). I do not wish these books to be alienated by the monastery but to survive here forever. For what has once been dedicated to God is inalienable’.25 Because Feres was on the road between Constantinople and Thessaloniki, Isaak may have been thinking about travellers who would visit the church, to whom the monks had to show this manuscript containing his writings – a manuscript which ultimately did not survive. Another very well-known example of a collection of aristocratic books is the will of Eustathius Boilas, drafted not long before 1059, the date of manuscript Paris, BnF, Coislin 263 in which it was copied.26 Boilas was an important imperial official, hypatos and protospatharios, of Cappadocian origin, who lived in the County of Edessa at a fortress called Salem. Like many rich Byzantine citizens, Boilas took measures concerning the future of his property by creating a monastery that he called a ‘church’ (dedicated to the Theotokos or the Virgin of Salem).27 The ‘church’ belonged to Boilas’ heirs, his two daughters and their husbands, as did the library, which hosted a collection of almost sixty books, a third of which were liturgical, another third spiritual, and a final third with secular authors and entertaining literature.28 After creating the inventory, Boilas provides a few 24 The typikon from 1152 was edited by Petit 1908, and more recently by Papazoglou 1994. English translation and Commentary in Thomas and Constantinides Hero 2000, vol. 2, 782–858; see also Kaplan 2017, 321–322. 25 On this passage (ed. Petit 1908, 69, § 106; English Translation in Thomas and Constantinides Hero 2000, 844), see Grünbart 2018, 332–333. 26 Vryonis Jr. 1957, 263–277; Lemerle 1977, 13–63 (with an edition of the Greek text). 27 Lemerle 1977, 63. 28 Ed. Lemerle 1977, ll. 141–166.

Inmaculada Pérez Martín instructions concerning the use of these volumes: ‘And be these [books] dedicated to the holy church, so that my two daughters may have the use and possession of them for chanting, reading, and learning. They may not be alienated by other persons, except by the will and consent of the clergy, since these must be used as it is required and where it is required’.29 Boilas therefore planned a very limited use of his collection, referring exclusively to his daughters. For the eleventh-century Duchy of Edessa, Boilas’ book collection may have been quite exceptional, and its owner may have not had a social circle with whom to share these books; nonetheless lending between friends was frequent and necessary in a world like Byzantium, where there were no book merchants selling copies of a text, which means that lending was often the only way to obtain the copy of a work. There are a large number of letters accompanying the sending of a book or requesting its return; they also mention book loans to correct a copy of the same text. Surviving lists of loans between individuals are nonetheless very rare, which does not necessarily mean that these loans were infrequent, but is rather due to the perishable nature of the medium on which they were recorded. Regarding the loan lists added to the flyleaves of manuscripts belonging to the lender, they were most likely victims of elementary operations to erase evidence by subsequent owners of the book. We do nevertheless have a valuable account in one of the flyleaves of Vatican City, BAV, Vat. gr. 207 (fol. VII), a manuscript containing philosophical and technical works from the second half of the thirteenth century. In this case, the lending notes were crossed out instead of erased because the poor quality of the paper would not withstand erasing.30 The owner had carefully annotated the date, month, and year of loans to his family members and friends, carried out between 1268 and 1282. Among the borrowers were an abundance of patriarchal and imperial officials,31 suggesting the owner’s closeness to circles of power, which were characterised by an in-depth knowledge of ancient Greek, philosophy, and the sciences of the quadrivium. These items show that the book owner, who was partially the copyist and organiser of the volume, had a library consisting of several volumes that he loaned from time to time: Aristotle’s corpus on logic, a treatise on music (probably on harmony), arithmetic, rhetoric, civil and canonical law, Homer, John of Damascus, and Basil of Caesarea. 29 English translation Vryonis Jr. 1957, 270. 30 On the manuscript in general and its paper, of Arab-Spanish origin, see Canart 1982, 279–280 [Canart 2008, 767–768], with a new edition of the lending list. 31 Regarding the identity of the members of the circle, see Trapp 1978, 198–199; Wilson 1975, 7; and Constantinides 1982, 139.

Byzantine Libraries: The Public and the Private 2 The space dedicated to books A library such as the one reflected by the list of manuscript Vat. gr. 207, ‘average’ in size if we compare it to other contemporary lists, could be conserved in a chest,32 but if the owner was also a real book user, a reading space would become desirable.33 We know very little about domestic architecture in Byzantium and the space reserved for storing and reading books. Certain elements are only vaguely mentioned by written texts. However, we have once more abundant information on the place of books in Byzantine monasteries, where we know that there were not only collections of spiritual or liturgical manuscripts, but also a significant number of secular books, which a priori had no link to the community’s lifestyle or its spiritual goals. They had been left there by the founder(s) of the monastery or belonged to an erudite monk living there (a rather common circumstance, at least during the thirteenth and fourteenth centuries). As the furniture inventories clearly show, the books used during liturgy were kept, along with liturgical objects, in the skeuophylakion or sacristy.34 We also know that at Mt. Athos, a few monasteries kept their non-liturgical books on the top floor of towers, under the roof (in other words, in the least accessible location in the monastery) or above the narthex of the main church,35 but never on the ground floor or in a vestibule beside the church.36 It appears that in monasteries, books were only rarely found near open spaces. Monks would borrow their books and go to their cells to consult them. In Casole, the monastery’s typikon vaguely 32 The archaeological evidence on these chests or trunks is very limited, and images of furniture destined for book storage mainly come from Latin manuscripts, such as the famous seventh-century Amiatinus codex or Alfonso X of Castile’s Cantigas (Escorial T.I.1). In the illuminations decorating Greek manuscripts of the New Testament, trunks are often portrayed beside each of the evangelists, used to store writing instruments and offering a support for the model being copied. This type of representation suggests the existence in Byzantium of a piece of furniture destined to contain a small number of books. 33 Cavallo 2007, 122–128. 34 Orlandos 1958, 73. 35 Orlandos 1958, 108–110. The narthex is an elongated and at times double arched lobby that precedes the central nave of the church. The library of Great Lavra on Mount Athos is located near the narthex of the church: Litsas 2000. This has also been the case for Patmos since the sixteenth century; Bogiatzis 2012, 97–100. I must thank Zisis Melissakis for his precious information on this monastery. With regard to the library of Chora, see above, note 4. 36 According to Orlandos 1958, 109, the independent tetragonal or octagonal building, with two series of windows, which appears in a drawing at the Russian monastery of Mt Athos (St Panteleemon), made by a traveller, Vasilìj Grigorovič-Barskìj (1725–1744) and therefore prior to its complete reconstruction during the nineteenth century, could be a library.

Inmaculada Pérez Martín mentions a ‘place for books’ (o topos tōn bibliōn).37 The case of the Monastery of Saint John of Stoudiou, with an almost ‘Soviet’ organization of the scriptorium and the library,38 must have been different. It is easy to imagine the existence, at this large Constantinopolitan monastery, of a big room where monks shared the tasks of creating books (maybe not preparing the parchment, but copying, cutting quires, and binding them) and where they spent time reading. Yet in his writings, Theodore never uses the word ‘library’.39 Given that the monasteries of Byzantine origin that still exist today are hard to access and have undergone numerous transformations, it may be useful to turn to Egypt, where many monastic sites have been excavated. An example is the White Monastery of Sohag founded by the monk Shenoute during the fifth century in the Thebes region, 400 km to the south of Cairo, facing Akhmim (or Pannopolis in Greek).40 Sohag was the location of a splendid library that contained more than a thousand books, today distributed between multiple libraries, and in particular the Bibliothèque nationale de France.41 Shenoute’s works preserved by the White Monastery are evidence that its scriptorium still operated during the tenth century, so the community of monks would have been able to read these works.42 Regarding the place reserved for books in the Sohag Monastery, we have the accounts of two Coptologists who visited the monastery in the late nineteenth and early twentieth century. In 1892, Gaston Maspéro pointed out the existence of a ‘secret’ chamber where books were piled up on the floor: ‘the incomplete books, the pages torn out of Bibles in disuse, Gospels, or collections of sermons used at the monastery, were piled up pell-mell on the ground of a cell located behind the choir, in a refuge tower connected to the main body of the church only by a very narrow secret passage’.43 A little later, in a 1904 article,44 Coptologist Walter Ewing Crum, a student of Maspéro, in turn re-transcribed the inscriptions that Canon Oldfield had deciphered in a room of the monastery with five niches in its walls.45 The inscriptions, done with a reed pen or a brush, are the 37 Omont 1890, 391. 38 Featherstone and Holland 1982. 39 Typikon, § 26, 33; Thomas and Constantinides Hero 2000, vol. 1, 93, 108, 112–113. 40 Grossmann 2002, 61, 529–532; Grossmann et al. 2004; Davis 2010. 41 Regarding the monastery library, Orlandi 2002, Boud’hors 2008, and Emmel and Römer 2008. 42 A tenth-century manuscript contains a letter from Shenoute that had to be read out loud four times in a year: Emmel and Römer 2008, 13. 43 Maspéro 1892. Regarding the location of this ‘secret chamber’ in the sacristy or pastophorion, to the north of the church apse, see Davis et al. 2014, 80–81. 44 Crum 1904, 564–568. 45 On niches as rudimentary places to conserve books, see Cavallo 1991, 52.

Byzantine Libraries: The Public and the Private catalogue or inventory of the books that each niche had contained in the past, according to a thematic organization, namely Gospels, doctrinal books, and hagiography. This type of organization corresponds to that of a library, and may demonstrate a certain degree of usability. However, given the hidden location of this chamber, access to it must have been restricted. Upon closer analysis, these inscriptions instead appear to constitute a late inventory of a library containing rather damaged manuscripts that were no longer read. Today, we no longer know which was the secret room with the niches. From Shenoute’s times up until the nineteenth century, the monastery library would most likely have changed locations multiple times. Regardless, the spaces in the White Monastery detected by archaeologists as being suitable to have held a library are all hidden and difficult to access.46 The monastic library in Byzantium therefore appears to have been more of a place to store books than a place dedicated to reading. 3 From the Imperial Library to the Palace Library Any institutional or imperial effort aiming to create a library or to provide it with books had the goal of preserving a heritage considered to be a common good, or of creating a place for pleasure and knowledge acquisition among potential readers, mainly students who needed texts in order to become health, legal, or administrative professionals. Outfitting a library, but also and above all making it accessible, were actions driven by a political will. We have already shown the cases of Patmos and Casole; it is now time to turn to the Imperial Library of Constantinople. Since its establishment during the first half of the fourth century, Constantinople was designed as a new capital of the Roman Empire. This is why emperors invested all of their efforts in it in order to make the city worthy of its title: they promoted the construction of new buildings and urban spaces according to the blueprint of ancient Rome, making them more majestic by decorating them with Antique artwork. Books played an essential role in this political and architectural undertaking, as did works of art. Like the libraries created by Roman emperors during Antiquity, the founding of Constantinople involved the creation of a large 46 I must thank Professor Stephen Davis, director of the White Monastery archaeological digs, for this information. On this subject, see also Emmel and Römer 2008, 10–11. In 2011, more than one thousand fragments of Arabic and Coptic manuscripts were discovered in a room called the ‘Candle Room’ which did not have any niches: Davis et al. 2014.

Inmaculada Pérez Martín public library.47 The goal and uses of such large Roman libraries have recently been re-examined by Matthew C. Nicholls,48 who has drawn attention to their location in the city, in which they belonged to larger architectural complexes, and to their significant visibility. For example, this was the case of the Library of Celsus at Ephesus, the remains of which are still visible today. Because the Roman elite did not need the ostentatious libraries of Augustus or Trajan to read, given that its members had books in their homes and at their villae, the function of these libraries must have been very different. Nicholls has highlighted the number of activities of all sorts (literary, legal, but also political or representational) that they displayed. Access to libraries was not restricted and the public was composed not only of students but also of people who did not necessarily want to read but simply wanted an outing, to look at the sculptures or wall decorations, and listen to a speech or recital. We should therefore consider these libraries open spaces or, at the very least, buildings integrated into frequented, multi-purpose public spaces. The first record on the Imperial Library of Constantinople goes back to the year 357, the date of a speech addressed by statesman, rhetorician, and philosopher Themistius (317–389/395) to Constantius II (337–361), an emperor who, as Lemerle wrote, turned Constantinople into an intellectual capital.49 According to Themistius, Constantius organised the copying of numerous manuscripts, and not only the classics studied at school, the preservation of which was assured, but also of other specialised and rarely disseminated texts.50 The rhetor is very imprecise around the type of books that were copied at that time, but this undertaking undoubtedly suggests the formation of a collection with the aim of preserving Hellenic culture, literature, and ideas. Themistius does not speak of Latin books, but a constitution from the year 372 by Emperor Valens orders the village prefect, who was the procurator of the Imperial Library, to hire seven antiquarii, four Greek and three Latin, that is, secretaries responsible for restoring manuscripts or making new copies of them.51 The library consequently possessed books in both languages – its purpose clearly being the preservation of texts from Greco-Roman Antiquity. 47 Fedeli 1989, 48–59. 48 Nicholls 2013, 261–276. 49 Lemerle 1971, 54–55; cf. Dagron 1968. 50 Themistius, Panegyric of Constantius II (or. 14), 59d–61a, ed. Downey and Schenkl 1965–1974, vol. 1, 69–89. 51 Lemerle 1971, 55; Irigoin 1975, 19.

Byzantine Libraries: The Public and the Private While Constantius II promoted the large-scale copying of books during his rule, the construction of an imperial library building not related to teaching is attributed to Julian the Apostate (361–363) by historian Zosimus. He wrote that Julian ‘built a library to the Imperial portico [Basilike Stoa], in which he placed all the books he possessed’.52 This was probably the transformation of the emperor’s personal library into a library open to the general public, located in a building beside the Basilike Stoa. The library was likely supplied by the formidable copying undertaking commissioned by Emperor Constantius. The location chosen by Julian is significant: this ‘imperial’ colonnade, later named as such (basilikē), was to the north of the street called Mese Regia (Ῥηγία), which linked the Great Palace of Constantinople, thus cutting the city in two from east to west. It was part of a complex, the centre of which was a large square surrounded by four porticos, with two pagan temples from the times of Julian and an octagon located near the Mese.53 Procopius (Buildings, 1.11) writes that the court, which was long and filled with fountains, was a square with porticos, built not on the ground but on stone; on the other hand, the Parastaseis, a heterogeneous collection dating back to the eighth century,54 state that its roof was golden, thus emphasising the sumptuousness of the building.55 The books were probably stored in an adjacent building; this is mentioned by John Zonaras, a historian from the late eleventh century and early twelfth century, who wrote the following on the fire that destroyed the city centre in 475: ‘When that emperor (Basiliskos) ruled, an enormous fire ravaged Constantinople. It started in the Chalkoprateia quarter and spread, burning the entire neighbourhood, the porticos of the public squares and the buildings found there, and certainly also the building called Basilike, where there was a library that held 120,000 books. People tell that among these books was the intestine of a snake where the poems of Homer, The Iliad and The Odyssey, had been engraved in letters of gold, as noted by Malchos, who wrote about those emperors’.56 The fire destroyed not only this building, but also the ancient statues located in the Lausus neighbourhood, not stopping until it reached the Forum of Constantine. The building mentioned in this passage must be identified as the Imperial Library of Constantius and Julian, that which was an annex to the Basilike. According to 52 Zosimus, New History, 75; transl. Ridley 1982, 75. 53 On the Basilike, see Guilland 1954; Janin 1964, 157–160; Berger 1988, 417–422; Magdalino 2006, 17–18, 39–40, 63, and Bardill 1997. 54 Among the numerous publications on this text, see Cameron and Herrin 1984 and Odorico 2014. 55 Lemerle 1971, 61 and n. 43; Cameron 1985, 103; partial trans. by Guilland 1954, 4. 56 Zonaras, Epitome Historiarum, ed. Dindorf 1868–1875, vol. 3, 130–131.

Inmaculada Pérez Martín two epigrams of The Greek Anthology, Emperor Zeno (474–491) rebuilt the library and the entire Basilike,57 but we must question his ability to obtain a copy of all of the books in the former library. In this Basilike complex there were multiple social activities: a law from 440 prohibited the introduction of horses, the celebration of marriages, and the sale of books;58 other sources mention philosophy classes59 and legal proceedings, two activities that implied the intervention of professionals who most likely used the library’s collections.60 Today, all that remains of the Basilike is the so-called Basilike cistern (and this has probably been the case since the end of the eighth century), built by Justinian to store plenty of water.61 Today it is known by the name of Yerebatan; its monumental size (140 × 70 m) is probably indicative of the size of the Basilike,62 because the area of the current cistern could correspond with the open courtyard surrounded by colonnades up until the fire of 475. The Basilike library subsequently disappeared from the sources; it is mentioned for the last time by chronicler George Hamartolos in the mid-ninth century. He describes its destruction in 726 by the iconoclast Emperor Leo III (717–741), who supposedly burned the books, the teachers, and the building: ‘some people say – and these are very reliable men – that beside the cistern of the Basilike, which is near the Chalkoprateia, there was a splendid palace to which, according to an ancient law (kata typon archaion), was assigned an ecumenical professor, with his students in attendance, twelve pre-eminent figures who covered all of the domains of knowledge and mastered the tenets of the Church; they received imperial salaries and likewise used the books;63 emperors did not enact laws without consulting them and obtaining their agreement’.64 Emperor Leo III attempted to 57 Guilland 1954, 5–6. 58 Guilland 1954, 3. During the sixth century, Agathias, Histories, 2.29, mentions that a bookstore still existed in the portico. The presence of horses is explained because close by, the archontes awaited the emperor on horseback to accompany him in procession over the Mese. 59 Theophylact Simocatta, an author who left us a history of the rule of Maurice (582–602), introduces his tale with a dialogue between Philosophy and History, in which the first complains of having been banished ‘from the porticoes of the Emperor’ (§ 1.5), which according to the text supposedly happened after the reign of Maurice. According to Lemerle 1971, 78 and n. 20, 22, these porticoes are likely those of the Basilike. 60 As Guilland 1954, 5, believed. 61 Magdalino 1998, 226. 62 Magdalino 2006, 40. 63 The text says basilikas diaitas kai biblous echontas, suggesting that the books belonged to the Imperial Library. 64 George Hamartolos, Chronicon, ed. De Boor 1904, vol. 2, 742; for comments, see Lemerle 1971, 89–90. We do not discuss here the imperial school mentioned in this passage.

Byzantine Libraries: The Public and the Private convince them to join the iconoclast doctrine, but because they refused, he locked them up in the didaskaleion and burned the men with the buildings, the books, and their personal possessions. This destruction appears in subsequent legends, starting in the ninth century, through the proponents of restoring the worship of images in order to discredit the iconoclast emperors. There are nonetheless interesting elements in this passage retaining the memory of an imperial library and teachings, even though the title of professor, oikoumenikos didaskalos, assisted by twelve assistants, like the twelve Apostles, more likely corresponds to the position of a teacher within the Patriarchate. Despite what ‘iconodule’ propaganda claims, ninth-century iconoclast emperors promoted the training of officials and the copying of books, thus contributing to renewing intellectual and literary life. At the time, the sources no longer spoke of an imperial library but rather of the ‘Palace Library’. It must have been located in the Great Palace, the main residence of the emperors and the administrative headquarters. Accounts are too intermittent and sparse to allow us to trace the history of this library. According to Pseudo-Simeon’s (tenth century) Chronicle, Emperor Theophilos (829–842) ‘found a text in the Palace Library that he could not understand’.65 This is why he sent it to Patriarch John as well as to Leo ‘the philosopher’, teacher at the imperial school of Magnaura, so that they would explain it to him.66 The Life of Patriarch Ignatius (composed between 886 and 890) also alludes to this library when it tells of Nicetas David Paphlagon’s accusation of former patriarch Photios of having created, around 870, a fictitious genealogy of the first emperor of the Macedonian Dynasty, Basil I (867–889): ‘He then finished off his composition with countless falsehoods, which he knew would delight the ears of the emperor, and wrote it on very old sheets of papyrus in Alexandrine letters, imitating as far as possible the ancient style of writing.67 And he also put it inside very old cover plates, which he took from a very old book, before depositing it in the great library of the imperial palace’.68 Even if the accusation against Photios is false, the passage nonetheless suggests the existence, in the imperial collection, of ancient books or books with a remote origin, holding rare texts. In fact, for the Byzantines, Photios’ genealogy was not the only text in the Imperial Library of uncertain origin or shrouded in mystery. The historian John Zonaras, as well as other sources up until the six- 65 Ed. Bekker 1838, 603–760, here 644. 66 Signes Codoñer 2010, 113–114, 434–437. 67 This passage demonstrates the prestige of the Library of Alexandria. 68 Ed. and Engl. transl. Smithies 2013, 120–121 (§ 89); see also Moravcsik 1961, 67; Mango 1975, 35, n. 26, and Vinson 2003, 11.

Inmaculada Pérez Martín teenth century, mention a prophetic work known by the name of The Oracles of Leo the Wise, describing the fate of Byzantine emperors, multiple late copies of which have survived. In a passage of his Chronicle on the reign of Emperor Michael II (820–829), Zonaras writes: ‘they say that the Imperial Palace Library holds a book that contains the prophecies of emperors and that it was illuminated with human and animal images; among others, we find the coloured image of a proud lion with the letter X written behind it […]’.69 Zonaras uses the present tense (‘they say’) to refer to the library and to the fact that the book is stored there (enapokeisthai); after that, he changes to the imperfect tense (êsan, enegraphonto) to describe the contents of this book. Even though the historian lived in Constantinople and had been an imperial official, he creates a description of the Imperial Library as a place that holds books shrouded in mystery. Is this evidence that the Imperial Library of Middle Byzantine times still contained a few ancient works from the library built during Late Antiquity? Even though we should be cautious, existing accounts seem to point in this direction. The book-loving Emperor Constantine VII Porphyrogenitus (905–959), known for his undertaking to compile excerpts from the works of historians, established a library in the Kamilas Palace, a three-floors recreation pavilion. His famous Excerpta Constantiniana were organised by subject, with the aim of facilitating access to moral, political, and diplomatic exempla from ancient historiography.70 In the prologue to the Excerpta, the emperor stated that he wanted to make the work available to the public,71 but despite his good intentions, he did not finish his ambitious project, of which only very few copies have survived. A contemporary chronicle known as Theophanes Continuatus, written around 944–959 in Constantine VII’s circles, mentions the library of the Great Palace; it specifies that it was a mezzanine with a look-out located between the chapel of the Archangel Michael and the Chrysotriklinos.72 This hall, separated from the Chrysotriklinos by a marble grill, was likely a space for storing books but not reading them. Moreover, the Chrysotriklinos itself, as the palace’s ceremonial room par excellence, constituted a transitional space between the ‘public’ and ‘private’ sections of the palace. The location of the library in this place, near the Chrysotriklinos, can also be explained by the location of the central administrative offices,73 due to officials’ need to consult its collections. This was also the case of Constantine’s col- 69 Zonaras, Epitome Historiarum, ed. Dindorf 1868–1875, vol. 3, 330–331. 70 On the Excerpta Constantiniana, see recently Németh 2018. 71 See the prologue of the Excerpta, ed. Németh 2010, 185, l. 38–50. 72 Featherstone and Signes Codoñer 2015, 206 l. 80–83 [145 Bonn]. 73 Featherstone 2005.

Byzantine Libraries: The Public and the Private laborators in the undertaking to compose his Excerpta and other historical works produced in the court environment, who found the historical works necessary for their writing in this library. In sum, the texts cited are more or less consistent in their definition of the Imperial Library of the Great Palace: they describe it as being no longer a library open to the general public and/or used by students, but as a book collection accessible to officials and frequented by emperors as well as a large, disorganised collection of ancient and rare works. The Palace Library was not the only library in Constantinople placed under the patronage of the emperors. Any school founded by the ruler also benefited from his support and book donations: we have proof of this in the legal school created by Emperor Constantine IX Monomachos in 1047 in a monastic and residential complex that he re-founded, the Monastery of Saint George of Mangana.74 The school’s founding charter, composed by the scholar John Mauropous, highlights the utility of the library of Mangana, which allowed students (presented as young people impassioned with knowledge) to more easily and directly access legal texts, eliminating the need for them to memorise them by heart, and facilitating their understanding. Thanks to the new library, these writings were no longer mysterious texts or ‘abstruse oracles’ to them.75 Mauropous continues stating: ‘there was [in Mangana] a place/a room, the seat of the magistral chair, which henceforth must be called didaskaleion nomōn (‘school of laws’) and a teacher [called] nomophylax (‘guardian of the laws’). The latter would also be the guardian of the library’s legal books, which the librarian would freely loan to him so that he did not have to obtain them elsewhere’.76 This provision, while admittedly out of the ordinary, reflects the capacity of the ruler or the people who have advised him on the creation of the school to acknowledge the need to provide students with access to legal texts to ensure that they learn properly. Because the books were too valuable to allow unrestricted access to them, this procedure therefore had to be designed: the teacher owned the books from the school, that he used for his teaching, and that the students could also read and most likely copy in order to better comment on their contents. While the Great Palace appears to have held a large library up until the tenth century, the sack of Constantinople by the Crusaders in 1204 may have resulted in damage and in the dispersion of its collection. At the time, certain Greek rulers 74 See Michel Attaliates, Historia, ed. Pérez Martín 2002, 36, 254 n. 134. 75 Ed. Salač 1954, 16–37; cf. Lemerle 1977, 207–212. 76 According to Lemerle 1977, 209, n. 34, this passage suggests that, normally, teachers did not have a public library available to them, but used their own library.

Inmaculada Pérez Martín who controlled territories outside of the authority of the Latin Empire of Constantinople, such as Epirus, Trabzon, and Nicaea, undertook efforts to collect new libraries. According to the Synopsis Chronike, attributed to Theodore Skoutariotes, Nicaean rulers John III Doukas Vatatzes (1221–1254) and Theodore I Laskaris (1254–1258) gathered ‘in some cities’ libraries books from all of the arts and all of the sciences’.77 He continues: ‘(Theodore I Laskaris) gathered books on the largest variety of arts and sciences, in an even larger number than Ptolemy did – even though we glorify him for this reason; and by placing these books in cities, he ordered that they be made available to those wishing to read them and for the better understanding of the disciplines that they cover. Culture thus started to grow under his reign, and to make such progress that across the Roman regions and cities, meetings of scholars were seen […]’.78 This is the only undertaking of its type known in the history of Byzantium, and must be explained by the desire of the rulers of Nicaea to recover imperial heritage, identifying as they did the survival of the idea of the empire with that of Hellenism, in other words, of classical Greek culture. The tone is excessively encomiastic, and forces us to remain cautious with respect to the actual scope of these shipments of books to cities in Asia Minor, such as Pergamon, Smyrna, Bursa, Philadelphia, or Ephesus. Common sense leads us to believe that to be effective, the measure had to include a teacher and to organise teaching in each of the cities. However, just the desire to do so demonstrates that for these rulers, control of the territory was not exclusively dependent on the strength of their ramparts and fortresses, but also on schools and books. During the period from the reconquest of Constantinople in 1261 by the sovereigns of Nicaea up until the Ottoman conquest of 1453, the emperors continued to promote education but not the creation of libraries.79 It is true that at the time, Byzantium became a small, often desperate, state, but the dissemination of books, facilitated by the use of good-quality paper and the circulation of texts, grew exponentially from the late thirteenth century. Even though officials produced copies of great beauty, we are aware only of a single codex intended for the Imperial Library: Paris, BnF, gr. 1115, a manuscript copied, according to its colophon or copying note, by a man called Leo Cinnamus in 1276, and stored at the Imperial Library. It contains a compendium of synodal deeds, the model for 77 Ed. Sathas 1872–1894, vol. 7, 507. About this undertaking of the Nicaean Emperors, see Constantinides 1982, 12–13, 20, and in particular Förstel 2005; Angelov 2019, 85. 78 Skoutariôtès, Synopsis Chronike, ed. Sathas 1872–1894, vol. 7, 535–536; Förstel 2005, 130. 79 Wilson 1967, 58.

Byzantine Libraries: The Public and the Private which was a manuscript dated from 758–759, found in Rome.80 Because the date of the copying coincides with increasing contacts between Byzantium and Rome, as a part of the uniting of the Churches (at the Council of Lyon in 1274), the copyist may have had access to this manuscript during a diplomatic mission to Rome. The colophon mentions Emperors Michael VIII and Andronikos II Palaiologos and their wives, who must have ordered the copy because it was intended for the Imperial Library. Because the Great Palace had already been destroyed at the time, we think that the Imperial Library to which the colophon of the Paris manuscript is alluding is the Palace of Blachernae, which became the regular residence of emperors during the Palaiologan period.81 A set of bindings with the monogram of the imperial family dating back to the fourteenth and fifteenth centuries may suggest that the imperial family wished to give a consistent appearance to its book collection, which was a part of its wealth.82 These manuscripts were likely a part of the library that Spanish traveller Pero Tafur saw in 1436–1439 at the residence of Emperor John VIII.83 He writes that the palace appeared to have been magnificent, but at the time, it was in disrepair and the emperor and his family lived in poverty: ‘at the entrance to the palace, under some chambers, there was a loggia or vestibule (lonja) with marble floor, the arches of which have nicely tiled benches, and beside them long tables placed on low pillars, themselves tiled, on which there are many books, ancient writings and histories; on the other side, there are game boards, since the emperor’s residence is always crowded’.84 This pleasant open space, suitable for reading and playing at the end of the day, with benches to sit on and tables, was a peaceful place that brightened up palace life. However, instead of adopting the Roman model of a ‘public’ library, it is reminiscent of the libraries of Roman patricii and Byzantine aristocrats. The 80 Astruc 1989, 46–48, Pl. 46–47, and Alexakis 1996. 81 Magdalino 2006, 75–76. Here, we do not discuss the rather spiny topic of the library of the ‘imperial monastery’ where scholar Maximus Planudes found precious books, and which Wilson 1967, 57, clearly distinguished from the Imperial Library. On the monastery where Planudes lived, see more recently Estangüi Gómez 2013. On the accessibility of this Imperial Library, see Planudes, Ep. 67 (French transl. Schneider 2020). 82 Irigoin 1982, Hoffmann 1985, Olivier 2002, and Pérez Martín 2008, 156–158. 83 Wilson 1967, 54. 84 Andanças e viajes de Pero Tafur 1874, 180–181: ‘á la entrada del palacio debaxo de unas cámaras está una lonja sobre mármoles, abierta, de arcos con poyos en torno bien enlosados é junto con ellos como mesas puestas de cabo á cabo sobre pilares baxos, ansí mesmo cubiertos de losas, en que están muchos libros é escrituras antiguas é estorias, é á otra parte, tableros de juegos, porque siempre se falla acompañada la casa del Emperador’.

Inmaculada Pérez Martín Roman model had therefore disappeared; it endured as long as did the imperial desire to transform collections of books and works of art into an instrument of political propaganda. Like in Alexandria, fires and political turmoil ultimately swept away this last vestige of ancient culture. While Byzantine emperors undoubtedly continued to invest in books, they did so more as private collectors than in accordance with the principles of the emperors of the past. Acknowledgements The author is a member of the Instituto de Lenguas y Culturas del Mediterráneo y del Oriente Próximo, CCHS-CSIC, Madrid. This research has been funded by the project ‘The Byzantine author (III)’ [PID2019-105102GB-I00]. Thanks are due to Raúl Estangüi and Alfredo Calahorra. References Alexakis, Alexander (1996), Codex Parisinus Graecus 1115 and Its Archetype (Dumbarton Oaks Studies, 34), Washington, DC: Dumbarton Oaks Research Library and Collection. Andanças e viajes de Pero Tafur por diversas partes del mundo avidos, Madrid: Imprenta de Miguel Ginesta, 1874. Angelov, Dimiter (2019), The Byzantine Hellene. The Life of Emperor Theodore Laskaris and Byzantium in the Thirteenth Century, Cambridge: Cambridge University Press. Apostolidis, Apostolos (1983–1984), Il typikon di S. Nicola di Casole secondo il codice Taur. Gr. C III 17 (Introduzione, testo critico, indici), PhD Thesis, Roma and Bari. Arnesano, Daniele (2008), ‘San Nicola di Casole e la cultura greca in Terra d’Otranto nel Quattrocento’, in Hubert Houben (ed.), La conquista turca di Otranto (1480) tra storia e mito. Atti del Convegno internazionale di studio (Otranto – Muro Leccese, 28–31 marzo 2007), Galatina: Congedo, vol. 1, 107–140. Arnesano, Daniele (2014), ‘La penitenza dei monaci a S. Maria del Patir e a S. Nicola di Casole’, Revue des Études Byzantines, 72: 249–273. Astruc, Charles (1981), ‘L’inventaire dressé en septembre 1200 du trésor et de la bibliothèque de Patmos. Édition diplomatique’, Travaux et Mémoires du Centre de Recherches d’Histoire et Civilisation Byzantines, 8: 15–30. Astruc, Charles (1989), Les manuscrits grecs datés des XIIIe et XIVe siècles conservés dans les bibliothèques publiques de France, vol. 1: XIIIe siècle, Paris: Bibliothèque nationale. Astruc, Charles (1994), ‘Les listes de prêt figurant au verso de l’inventaire du trésor et de la bibliothèque de Patmos dressé en septembre 1200’, Travaux et Mémoires du Centre de Recherches d’Histoire et Civilisation Byzantines, 12: 495–499. Bardill, Jonathan (1997), ‘The Palace of Lausus and Nearby Monuments in Constantinople: A Topographical Study’, American Journal of Archaeology, 101/1: 67–95. Bekker, Immanuel (1838), Theophanes Continuatus, Ioannes Cameniata, Symeon Magister, Georgius Monachus, Bonn: Weber.

Byzantine Libraries: The Public and the Private Benoit-Meggenis, Rosa (2017), L’empereur et le moine: les relations du pouvoir impérial avec les monastères à Byzance (IXe–XIIIe) (Travaux de la Maison de l’Orient et de la Méditerranée, 73), Lyon: Maison de l’Orient et de la Méditerranée – Jean Pouilloux. Berger, Albrecht (1988), Untersuchungen zu den Patria Konstantinupoleos (Poikila Byzantina, 8), Bonn: R. Habelt. Bianconi, Daniele (2012), ‘Michele della Lavra di Stilo. Qualche nuova attribuzione e considerazione’, Scripta, 5: 31–41. Bogiatzis, Sotiris (2012), E oikodomike istoria tes Ieras Mones Agiou Ioannou Theologou Patmou, Thessaloniki: University Studio Press. Boud’hors, Anne (2008), ‘Copie et circulation des livres dans la région thébaine (VIIe–VIIIe siècles)’, in Alain Delattre and Paul Heilporn (eds), ‘Et maintenant ce ne sont plus que des villages’: Thèbes et sa région aux époques hellénistique, romaine et Byzantine. Actes du colloque tenu à Bruxelles les 2 et 3 décembre 2005, Brussels: Fondation égyptologique Reine Elisabeth, 149–161. Byzantine Books and Bookmen. A Dumbarton Oaks Colloquium, 1971, Washington, DC: Dumbarton Oaks, Center for Byzantine Studies, 1975. Cameron, Averil (1985), Procopius in the Sixth Century, London: Routledge. Cameron, Averil and Judith Herrin (1984), Constantinople in the Early Eighth Century. The Parastaseis syntomoi chronikai, Leiden: Brill. Canart, Paul (1982), ‘À propos du Vaticanus Graecus 207. Le recueil scientifique d’un érudit constantinopolitain du XIIIe siècle et l’emploi du papier à zig-zag dans la capitale Byzantine’, Illinois Classical Studies, 7: 271–298 [reprinted in Paul Canart (2008), Études de paléographie et de codicologie, reproduced with the collaboration of Maria Luisa Agati and Marco D’Agostino (Studi e Testi 450–451), Vatican City: Biblioteca Apostolica Vaticana, 759–786]. Cavallo, Guglielmo (ed.) (1989), Le biblioteche nel mondo antico e medievale, Bari: Laterza. Cavallo, Guglielmo (1991), ‘Modelli bibliotecari in Occidente e in Oriente nel Medioevo’, Byzantiaka, 11: 49–59. Cavallo, Guglielmo (2002), ‘Πόλις γραμμάτων. Livelli di istruzione e uso di libri negli ambienti monastici a Bisanzio’, Travaux et Mémoires du Centre de Recherches d’Histoire et Civilisation Byzantines, 14 [= Mélanges Gilbert Dagron]: 95–113. Cavallo, Guglielmo (2007), Leggere a Bisanzio, Milano: Sylvestre Bonnard. Chiriatti, Mattia C. (2016), ‘El typikón de San Nicolás de Casole: un monasterio, un scriptorium y una biblioteca’, Estudios bizantinos, 4: 107–121. Cholij, Roman (2002), Theodore the Stoudite. The Ordering of Holiness (Oxford Theological Monographs), Oxford: Oxford University Press. Constantinides, Costas N. (1982), Higher Education in Byzantium in the 13th and Early 14th Centuries (1204–ca. 1310), Nicosia: Cyprus Research Center. Constantinides, Costas N. (2006), ‘British Collectors of Manuscripts Visiting Cyprus in the Nineteenth Century’, in Julian Chrysostomides and Charalambos Dendrinos (eds), ‘Sweet Land…’. Lectures on the History and Culture of Cyprus, Camberley: Porphyrogenitus, 231–258. Crum, Walter Ewing (1904), ‘Inscriptions from Shenoute’s Monastery’, Journal of Theological Studies, 5: 552–569. Dagron, Gilbert (1968), ‘L’Empire romain d’Orient au IVe siècle et les traditions politiques de l’hellénisme: le témoignage de Thémistios’, Travaux et Mémoires du Centre de Recherches d’Histoire et Civilisation Byzantines, 3: 1–242. Davis, Stephen J. (2010), ‘Archaeology at the White Monastery, 2005–2010’, Coptica, 9: 25–58.

Inmaculada Pérez Martín Davis, Stephen J. et al. (2014), ‘Left Behind: A Recent Discovery of Manuscript Fragments in the White Monastery Church’, Journal of Coptic Studies, 16: 69–87. De Boor, Carolus (ed.) (1904), Georgii Monachi Chronicon, Leipzig: Teubner. Déroche, Vincent (2006), ‘Écriture, lecture et monachisme à la haute époque Byzantine’, in Brigitte Mondrain (ed.), Lire et écrire à Byzance, Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance, 113–123. Diehl, Charles (1886), ‘Le monastère de S. Nicolas di Casole près d’Otrante, d’après un manuscrit inédit’, Mélanges d’archéologie et d’histoire, 6: 173–188. Diehl, Charles (1892), ‘Le trésor et la bibliothèque de Patmos au commencement du 13e siècle (d’après des documents inédits)’, Byzantinische Zeitschrift, 1: 488–525. Dindorf, Ludwig (ed.) (1868–1875), Ioannis Zonarae Epitome historiarum, 3 vols, Leipzig: Teubner. Downey, Glanville and Heinrich Schenkl (eds) (1965–1974), Themistii orationes quae supersunt, 3 vols, Leipzig: Teubner. Eleopoulos, Nikephoros X. (1967), Hê bibliothêkê kai to bibliographikon ergastêrion tês monês tôn Stoudiou, Athens: Makre. Emmel, Stephen and Cornelia E. Römer (2008), ‘The Library of the White Monastery in Upper Egypt’, in Harald Froschauer and Cornelia E. Römer (eds), Spätantike Bibliotheken: Leben und Lesen in den frühen Klöstern Ägyptens, Wien: Phoibos, 5–14. Estangüi Gómez, Raúl (2013), ‘Saint-Sauveur de Chôra. Un monastère catholique à Constantinople dans le troisième quart du XIVe siècle’, Estudios Bizantinos, 1: 140–197. Fatouros, Georgios (ed.) (1991), Theodori Studitae Epistulae, Berlin and New York: Walter de Gruyter. Featherstone, Jeffrey and Meridel Holland (1982), ‘A Note on Penances Prescribed for Negligent Scribes and Librarians in the Monastery of Stoudios’, Scriptorium, 36: 258–260. Featherstone, Michael (2005), ‘The Chrysotriklinos as seen through De Cerimoniis’, in Lars M. Hoffmann (ed.), Zwischen Polis, Provinz und Peripherie. Beiträge zur byzantinischen Kulturgeschichte (Mainzer Veröffentlichungen zur Byzantinistik, 7), Wiesbaden: Harrassowitz, 845–852. Featherstone, Michael and Juan Signes Codoñer (eds) (2015), Chronographiae quae Theophanis Continuati nomine fertur, libri I–IV (Corpus Fontium Historiae Byzantinae, 53), Berlin: De Gruyter. Fedeli, Paolo (1989), ‘Biblioteche private e pubbliche a Roma e nel mondo romano’, in Cavallo (ed.) 1989, 29–64. Förstel, Christian (2005), ‘Entre propagande et réalité: la culture dans l’empire de Nicée’, Revue française d’héraldique et de sigillographie, 73–75: 129–134. Gerolymatou, Maria (2019), ‘Un monastère isolé, des moines en mouvement: le cas de Patmos entre le XIe et le XIIIe siècle’, in Olivier Delouis, Maria Mossakovska-Gaubert and Annick Peters-Custot (eds), Les mobilités monastiques en Orient et en Occident de l’Antiquité tardive au Moyen Âge (IVe-XVe siècle), Rome: Publications de l’École française de Rome [publ. open access DOI: 10.4000/books.efr.4283], 223–234. Grossmann, Peter (2002), Christliche Architektur in Ägypten, Leiden: Brill. Grossmann, Peter et al. (2004), ‘Report on the Excavation in the Monastery of Apa Shenute (Dayr Anba Shinuda) at Suhag’, Dumbarton Oaks Papers, 58: 371–382. Grünbart, Michael (2018), ‘Securing and Preserving Written Documents in Byzantium’, in Alessandro Bausi, Christian Brockmann, Michael Friedrich and Sabine Kienitz (eds), Manuscripts and Archives: Comparative Views on Record-Keeping (Studies in Manuscript Cultures, 11), Berlin: De Gruyter, 319–338.

Byzantine Libraries: The Public and the Private Guilland, Rodolphe (1954), ‘La Basilique, la Bibliothèque et I’Octogone’, in Mélanges d’histoire littéraire et de bibliographie offerts à Jean Bonnerot, Paris: Librairie Nizet, 97–107 [repr. in Rodolphe Guilland, Études de topographie de Constantinople Byzantine (Berliner byzantinistische Arbeiten, 37), 2 vols, Berlin: Akademie-Verlag, 1969, vol. 2, 3–13]. Hoffmann, Philippe (1985), ‘Une nouvelle reliure byzantine au monogramme des Paléologues (Ambros. M 46 sup. = Gr. 512)’, Scriptorium, 39: 274–281. Irigoin, Jean (1975), ‘Centres de copie et bibliothèques’, in Byzantine Books and Bookmen 1975, 17–27. Irigoin, Jean (1982), ‘Un groupe de reliures Byzantines au monogramme des Paléologues’, Revue française d’histoire du livre, n.s., 51: 273–285. Jacob, André (1980), ‘Culture grecque et manuscrits en Terre d’Otrante’, in Pier Fausto Palumbo (ed.), Atti del III Congresso internazionale di studi salentini e del I Congresso storico di Terra d’Otranto (Lecce 22–25 ott. 1976), Lecce: Centro di Studi Salentini, 53–77. Janin, Raymond (1964), Constantinople Byzantine. Développement urbain et répertoire topographique (Archives de l’Orient Chrétien, 4A), Paris: Institut français d’études Byzantines. Kaplan, Michel (2017), ‘Pouvoir des fondateurs et pouvoir des higoumènes dans les monastères byzantins, Xe–XIIe siècle’, Travaux et mémoires du Centre de Recherches d’Histoire et Civilisation Byzantines, 21/1 [= Béatrice Caseau, Vivien Prigent and Alessio Sopracasa (eds), Hou dôron eimi tas graphas blepôn noei. Mélanges Jean-Claude Cheynet]: 311–324. Kırımtayıf, Süleyman (2001), Converted Byzantine Churches in Istanbul, Their Transformation into Mosques and Masjids, Istanbul: Yayinları. Kominis, Athanasios D. (1968–1970), Facsimiles of Dated Patmian Codices, Athens: Center of Byzantine Studies. Kominis, Athanasios D. (1988), Patmiake bibliotheke…, vol. 1: Kodikes 1–101, Athens: Ekdotike Hellados. König, Jason, Katerina Oikonomopoulou and Greg Woolf (eds) (2013), Ancient Libraries, Cambridge: Cambridge University Press. Lemerle, Paul (1971), Le premier humanisme byzantin, Paris: Presses universitaires de France. Lemerle, Paul (1977), Cinq études sur le XIe siècle byzantin, Paris: Centre National de la Recherche Scientifique. Leroy, Julien (2002), ‘Le monachisme studite’, in Florence de Montleau, Théodore Stoudite, Les Grandes Catéchèses (Livre I), présentation, traduction et notes (Spiritualité orientale, Monachisme primitif, 79), Bégrolles-en-Mauges: Abbaye de Bellefontaine. Leroy, Julien and Olivier Delouis (2007), Études sur le monachisme byzantin, Bégrolles en Mauges: Abbaye de Bellefontaine. Litsas, Evthymios (2000), ‘Paleographical Researches in the Lavra Library on Mount Athos’, Hellenika, 50: 217–230. Magdalino, Paul (1998), ‘Constantinopolitana’, in Irmgard Hutter and Ihor Sevcenko (eds), Aetos: Studies in honour of Cyril Mango presented to him on April 14, 1998, Stuttgart: B.G. Teubner, 220–232 [repr. in Paul Magdalino, Studies on the History and Topography of Byzantine Constantinople (Variorum collected studies series, 855), Aldershot: Ashgate, 2006, no. VIII]. Magdalino, Paul (2006), ‘Medieval Constantinople’, in Paul Magdalino, Studies on the History and Topography of Byzantine Constantinople (Variorum collected studies series, 855), Aldershot: Ashgate, 1–101. Mai, Angelo and Giuseppe Cozza-Luzi (1844–1905), Nova patrum bibliotheca, 10 vols, Rome: Typis Sacri Consilii Propagando Christiano Nomini.

Inmaculada Pérez Martín Mango, Cyril (1975), ‘The Availability of Books in the Byzantine Empire, A.D. 750–850’, in Byzantine Books and Bookmen, 29–45 [repr. in Cyril Mango, Byzantium and Its Image. History and Culture of the Byzantine Empire and Its Heritage, London: Variorum Reprints, 1984, no. VII]. Maspéro, Gaston (1892), Fragments de la version thébaine de l’Ancien Testament, Paris: E. Leroux. Mazzotta, Oronzo (2012), ‘Monaci e libri nel Salento medievale. Il monastero di San Nicola di Casole’, in Paolo Pellegrino (ed.), Sergio Stiso tra Umanesimo e Rinascimento in Terra d’Otranto, Galatina: Congedo, 149–164. Miklosich, Franz and Josef Müller (1860–1890), Acta et diplomata graeca medii aevi sacra et profana, Vienna: C. Gerold. Moravcsik, Gyula (1961), ‘Sagen und Legenden über Kaiser Basileios I’, Dumbarton Oaks Papers, 15: 59–126. Morris, Rosemary (1995), Monks and Laymen in Byzantium, 843–1118, Cambridge: Cambridge University Press. Müller-Wiener, Wolfgang (1977), Bildlexikon zur Topographie Istanbuls, Tübingen: Ernst Wasmuth. Németh, András (2010), Imperial Systematization of the Past. Emperor Constantine VII and His Historical Excerpts, PhD Thesis, Budapest, Central European University. Németh, András (2018), The ‘Excerpta Constantiniana’ and the Byzantine Appropriation of the Past, Cambridge: Cambridge University Press. Nicholls, Matthew (2013), ‘Roman Libraries as Public Buildings in the Cities of the Empire’, in König, Oikonomopoulou and Woolf (eds) 2013, 261–276. Odorico, Paolo (2014), ‘Du recueil à l’invention du texte: le cas des Parastaseis Syntomoi Chronikai’, Byzantinische Zeitschrift, 107/2: 755–784. Olivier, Jean-Marie (2002), ‘Encore une reliure au monogramme des Paléologues’, Scriptorium, 56/2: 323–331 and Pl. 27–29. Omont, Henri (1888), ‘Note sur un manuscrit grec copié en 1050 au Mont Latros (Anatolie)’, Revue des Études Grecques, 1: 336–339. Omont, Henri (1890), ‘La bibliothèque de S. Nicolas de Casole’, Revue des Études grecques, 3: 389–391. Orlandi, Tito (2002), ‘The Library of the Monastery of St. Shenoute at Atripe’, in Arno Egberts et al. (eds), Perspectives on Panopolis: An Egyptian Town from Alexander the Great to the Arab Conquest. Acts From an International Symposium Held in Leiden on 16, 17 and 18 December 1998 (Papyrologica Lugduno-Batava, 31), Leiden: Brill, 211–232. Orlandos, Anastasios K. (1958), Monasteriake architektonike, Athens: Esemene. Ousterhout, Robert (2002), The Art of the Kariye Camii, London and Istanbul: Scala Publ. Papazoglou, Georgios K. (1994), Typikon Isaakiou Alexiou Komnênou tês monês Theotokou Kosmosôteiras (1151/52) (Thrakike bibliotheke, 3), Komotini: Demokriteio Panepistemio Thrakes. Parlangèli, Ada and Oronzo Parlangèli (1951), ‘Il monastero di San Nicola di Casole, centro di cultura bizantina in Terra d’Otranto’, Bollettino della Badia Greca di Grottaferrata, 5: 30–45. Pérez Martín, Inmaculada (2002), Miguel Ataliates, Historia (Nueva Roma, 15), Madrid: CSIC. Pérez Martín, Inmaculada (2008), ‘Encuadernaciones bizantinas en los manuscritos griegos de España’, in Niki Tsironi et al. (eds), The Book in Byzantium. Byzantine and Post-Byzantine bookbinding. Proceedings of the International Symposium Athens 13–16 October 2005, Athens: Hellenike Hetaireia Vivliodesias, 277–314. Perria, Lidia (1993), ‘Scrittura e ornamentazione nei manoscritti di origine studita’, Bollettino della Badia Greca di Grottaferrata, 47: 245–280.

Byzantine Libraries: The Public and the Private Petit, Louis (1908), ‘Typikon du monastère de la Kosmosotira près d’Aenos (1152)’, Izvestija Russago Archeologicheskago Instituta v. Konstantinople, 13: 17–75. Polemis, Ioannes (2017), Theodore Metochites, Poems (Corpus Christianorum in Translation, 26), Turnhout: Brepols. Ridley, Ronald T. (1982), Zosimus, New History. A Translation with Commentary (Byzantina Australiensia, 2), Canberra: Australian Association for Byzantine Studies. Saint-Guillain, Guillaume (2004), ‘L’Apocalypse et le sens des affaires. Les moines de Saint-Jean de Patmos, leurs activités économiques et leurs relations avec les Latins (XIIIe et XIVe siècles)’, in Damien Coulon, Catherine Otten-Froux, Paule Pagès et al. (eds), Chemins d’outre-mer. Études d’histoire sur la Méditerranée médiévale offertes à Michel Balard (Byzantina Sorbonensia, 20), Paris: Publications de la Sorbonne, vol. 2, 765–790. Saint-Guillain, Guillaume (2009), ‘Le copiste Géôrgios Chômatas et les moines de Patmos’, in Chrysa Maltezou, Angeliki Tzavara and Despina Vlassi (eds), I Greci durante la Venetocrazia: Uomini, spazio, idee (XIII–XVIII sec.). Atti del Convegno Internazionale di Studi. Venezia, 3–7 dicembre 2007, Venice: Istituto ellenico di studi bizantini e postbizantini di Venezia, 163–181. Salač, Antonín (1954), Novella Constitutio Saec. XI Medii a Ioanne Mauropode conscripta a Constantino IX Monomacho promulgata (Textus Breves Graeci et Latini, 1), Prague: Academia Scientiarum Bohemoslovenica. Sathas, Konstantinos N. (1872-94), Mesaionike bibliotheke, 7 vols, Athens: Vas. N. Gregoriades. Schneider, Jean (2020), Maxime Planoudès, Lettres (Collège de France – CNRS, Centre de recherche d’histoire et civilisation de Byzance, Monographies, 54), Leuven: Peeters. Schreiner, Peter (2002), ‘Klosterbibliotheken in Öst und West. Unterschiede und Gemeinsamkeiten’, in V. Guzelev and A. Miltenova (eds), Medieval Christian Europe: East and West, Sofia: Gutenberg, 19–29. Signes Codoñer, Juan (2010), The Emperor Theophilus and the East (830–842), Aldershot: Ashgate. Smithies, Andrew (2013), David Nicetas, The Life of Patriarch Ignatius (Dumbarton Oaks Texts, 13), Washington, DC: Dumbarton Oaks Research Library and Collection. Smyrlis, Kostas (2006), La fortune des grands monastères byzantins (fin du Xe–milieu du XIVe siècle) (Monographies du Centre de recherche d’Histoire et Civilisation de Byzance, 21), Paris: Association des Amis du Centre d’Histoire et Civilisation de Byzance. Thomas, John and Angela Constantinides Hero, with the assistance of Giles Constable (2000), Byzantine Monastic Foundation Documents, Washington, DC: Dumbarton Oaks Research Library and Collection, 5 vols. Trapp, Erich (1978), ‘Probleme der Prosopographie der Palaiologenzeit’, Jahrbuch der Österreichischen Byzantinistik, 27: 181-201. Van Millingen, Alexander (1912), Byzantine Churches in Constantinople. Their History and Architecture, London: Macmillan. Vinson, Martha (2003), ‘Rhetoric and Writing Strategies in the Ninth Century’, in Elisabeth Jeffreys (ed.), Rhetoric in Byzantium: Papers from the Thirty-fifth Spring Symposium of Byzantine Studies, Exeter College, University of Oxford, March 2001, Aldershot: Ashgate, 9–22. Vranoussis, Era and Maria G. Nystazopoulou-Pelekidou (1980), Byzantina engrapha tēs monēs Patmou, 2 vols, Athens: Ethniko Idryma Erevnon. Vryonis, Spiros Jr. (1957), ‘The Will of a Provincial Magnate, Eustathius Boilas (1059)’, Dumbarton Oaks Papers, 11: 263–277.

Inmaculada Pérez Martín Waring, Judith (2002), ‘Literacy of Lists: Reading Byzantine Monastic Inventories’, in Catherine Holmes and Judith Waring (eds), Literacy, Education and Manuscript Transmission in Byzantium and Beyond, Leiden: Brill, 170–181. Wilson, Nigel G. (1967), ‘The Libraries of the Byzantine World’, Greek, Roman and Byzantine Studies, 8: 53–80 [repr. in Dieter Harlfinger (ed.) (1980), Griechische Kodikologie und Textüberlieferung, Darmstadt: Wissenschaftliche Buchgesellschaft, 276–309]. Wilson, Nigel G. (1975), ‘Books and Readers in Byzantium’, in Byzantine Books and Bookmen 1975, 1–15. Woolf, Greg (2013), ‘Approaching the Ancient Library’, in König, Oikonomopoulou and Woolf (eds)

Western Europe

Donatella Nebbiai How Private Libraries Contributed to the Transmission of Texts Abstract: Private libraries developed in Western European cities at the end of the eleventh century, in connection with the spread of the practice of writing and autographs. In this article we study the influence that these libraries exerted as a cultural model, not only through the works they contain, but also through the ways in which books were supplied and transmitted. The first libraries appeared in Western Europe during the Carolingian period, but became more established in the eleventh century, before spreading, especially in cities, from the thirteenth to the fifteenth century. These libraries attested to the renewed relationship that their owners had with the practice of writing and with the imparting of knowledge by teachers, both orally and in writing, and its reception by pupils.1 Supporting the spread of vernacular languages and document production, the books gradually came to embody the emergence of autography and personal writing. These aspects can be studied by combining two approaches. The first approach, which is historical, examines the documents that bear witness to the concrete presence of books in an environment, and namely inventories, lists, contracts, and other practical sources. It cannot be applied exclusively, because owning a book did not necessarily imply that it was read or understood. The second approach pertains to literary history: literature is primarily made of writing and books.2 This approach takes into consideration textual models and autographs, as well as their diffusion. It cannot be exclusive either, because the manuscripts were not always copied to be circulated or reproduced, or to serve as a model. This is the case of monastic chronicles, for example, which were often written to remain at the monastery and were not duplicated.3 1 I am mainly referring to Vernet 1989; Cavallo 1988; Petrucci 1983a and 1983b. For a recent and up-to-date list of the bibliography of the history of libraries cf. Géhin 2017. 2 Bartoli Langeli 2010, 43. 3 Guenée 1973; see also Écrire son histoire 2005. Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-010

Donatella Nebbiai 1 The libraries of masters: from creation to succession People whose profession was to study – theologians, canonists, physicians – collected books for teaching and pastoral purposes. As they were affiliated to the institutions where they trained or exercised their ministry, these teachers bequeathed their books to them, thus disseminating their knowledge even after their death.4 Books were generally passed down through wills. Inherited from Roman law and established through the custom of pious bequests, the will became prevalent in the Western world in the thirteenth century, the era of the consolidation of private fortunes in towns. Drawing up a will was a unilateral, voluntary act through which an owner disposed of his assets after his death. This final act, that was nonetheless revocable as long as the person was still alive, contained provisions relating to fortunes as well as other elements. In medieval society, dying intestate was considered harmful to the salvation of the soul. Initially, the act was formally declared before an adequate number of witnesses who, at the required time, would be capable of passing on the contents of the will and ensuring that it was applied. The first version of a will was written by the owner himself, healthy in body and mind, in his own hand, or under direct supervision if his condition prevented him from doing so (nuncupative will or will in scriptis).5 It is therefore important to emphasize the strong ties between the practice of making wills and the personal expression of writing – ties that are probably not unrelated to the importance always granted to books in the provisions of wills. To leave one’s possessions to a church, a monastery, or an order was an act of charity and devotion, but it was also the expression of a cultural choice and a social strategy. First, at the end of the Middle Ages the books most frequently bequeathed to churches were prayer books or personal devotion books, which the testator recommended be firmly secured, in his memory, in the places where the community gathered, whether the chapter, cloister, or refectory. These books often stated the ex dono of their initial owner so that his name could be evoked in prayers. Second, books were associated with life choices. At the beginning of the period under consideration here, books were often a part of the luggage of masters abandoning secular life to return to the monastic world. Their intention would be to spend their last years there, with a lifestyle that they believed to be 4 Parkes 1976; Bataillon 1983; Nebbiai 2013. 5 Regarding wills as a historical source and an expression of personal writing, cf. Bartoli Langeli 1985; Bartoli Langeli 2008; Paravicini Bagliani 1980. On the custom of pious bequests in Paris during the Middle Ages, cf. Gorochov 1999; Courtemanche 1997.

How Private Libraries Contributed to the Transmission of Texts more compatible with the teachings of the Gospels than that which they had beforehand. For example, Peter Cantor (d. 1197) bequeathed his books to the Cistercian Abbey in Longpont, where he had retired.6 Adam de Curtilandonis, the dean of Laon, offered his to the Abbey of Cuissy.7 Lastly, regarding the criteria for choosing recipients, even though members of the clergy generally bequeathed their books to religious institutions, this was far from being the only option from the second half of the thirteenth century onwards. Many clerics and cannons, for example, preferred to give priority to their direct heirs (nephews, grandnephews, other family members) under the condition that they study. As for laypeople, they most often handed down their books to other laypeople and the transmission took place within the same social environment. For example, in 1334, a blacksmith in Pistoia bequeathed his blacksmith’s manual to his colleague in the same city.8 In many cases, however, institutions, whether religious or secular, were called upon to take over for private heirs if they proved to be unworthy or simply inadequate. Let us consider a few cases. In 1347, Richard de Granges, a canon from Montbéliard, sold his books to his uncle, the executor of his will, but reserved a copy of the ‘Decretals, a compilation of canon law’, for his nephew, who was studying law.9 Johannes Gasqui, Bishop of Marseille (d. 1344) bequeathed his books to the cathedral of Marseille, except for his natural science and medicine texts, which were to go to his nephews.10 In 1433, Petrus Fica, a Sicilian master of the arts and medicine, appointed his daughter Constance, wife of Johannes de la Rocha, as his universal heiress, but stated that if she proved to be unworthy, the books would go instead to the church of San Domenico de Trapani.11 This type of dual option was also found at the end of the Middle Ages in the will of scholar Giovanni Pico della Mirandola, who appointed his brother as the executor of his will and authorized him to sell his library to a religious foundation, with a limit of two years following his death to carry out the transaction.12 If this were impossible, the brother was allowed to dispose of the books as he pleased, with the exception of giving them to Giovanni’s grandson, Giovanni Francesco, should he wish to purchase them. The intent was therefore the most interesting aspect, even if it was not always explicitly stated. Allocating books was a personal choice; it allowed the owner 6 Bondéelle-Souchier 1991, 192. 7 Bondéelle-Souchier 2000, 115. 8 Florence, Archivio di Stato, Mercanzia 1054, c. 140, 1334, 22 apr.; cited by Leonelli 1985, 41 n. 2. 9 Nebbiai 2013, 202 and 258. 10 Marseille, Archives départementales des Bouches du Rhône 6 G 104, n° 696. Albanès 1899, cols 309–310. 11 Bresc 1971, 154–159, no. 67. 12 Bianca 2008, 454; Kibre 1966, 119–297 (Inventarium librorum Ioahnnis Pici Mirandulae).

Donatella Nebbiai to live on at an institution or simply through the members of his family and own circle. At the end of the Middle Ages, these motives are found in a literary account, the famous Philobiblon by Richard de Bury, Bishop of Durham (1287–1345). He wrote of books as members of his family, brothers or children that passed down a heritage, that ensured the preservation of knowledge from father to son.13 Books were perceived in a new way. As opposed to being simply a receptacle for writing, they were performative and exerted a new influence over the collection to which they belonged. Reciprocal adaptation thus took place: by becoming an integral part of institutional libraries, private libraries contributed to the evolution of the content and organization. For example, in 1257, Guillaume Ribot, Bishop of Vence, died at the infirmary of the Monastery of Saint Victor in Marseille, where he had retired a few years earlier. At the time he donated his books to the abbey, thereby enriching it with works that it had not had beforehand: a copy of Gervase of Tilbury’s Otia imperialia, one of Einhard’s chronicle, another of Marbodius of Rennes’ De lapidibus. These texts reflect the literary tastes of a clergyman active in the fight against heresy.14 The seventy or more books of Dominican Friar Jean, bequeathed to the Convent of Turin in 1287, were organized by subject according to a categorization that would be readopted by the receiving library: Bible and biblica, originalia, and sermons.15 The will of Gerard d’Abbeville, a master at the Collège de Sorbonne, refers to two types of books: everyday books, and books on theology and canonical law. This distinction may have foreshadowed that of the Collège library. Gerard d’Abbeville also ordered that his contribution be maintained in its original state, without a doubt to strengthen the inalienable nature of the bequest. He therefore also bequeathed a chest to contain the books.16 A few decades later, in 1311, the Florentine priest Amat did the same thing, bequeathing six volumes of biblical and theological works to the Cistercian Abbey of San Salvatore de Settimo, along with a cabinet in which to place them.17 13 Altamura 1954, 122–123. 14 Nebbiai 2005, 90–93. 15 Nebbiai 1992, 110–113; Ferrua 1992, 111–166. 16 Denifle and Châtelain 1889, 402, no. 430 (1271, oct. 19). 17 Nebbiai 1992, 117–119.

How Private Libraries Contributed to the Transmission of Texts 2 The model book Created in the scholarly world, the concept of the model book emerged from a change in writing production that occurred in the eleventh century. The traditional practice of copying in the scriptorium, within the monastic community, declined as it was gradually replaced by the activity of workshops located in towns and managed by scriptores. These were generally laypeople who were paid for their work. Chosen on the basis of their skills, they worked on behalf of the religious authorities, who were responsible for verifying the quality of the books produced, as well as their compliance with orthodoxy. As a result, a two-way relationship was established between the scriptor and his commissioner: a partnership similar to that of teacher and student. The goal was to produce high-quality books for clerics’ training. Every book was conceived of as a whole, whether it was isolated and contained a complete work with comments, consisted of all of the works of an author, or formed an overall collection when combined with other volumes. The fact that Bibles appear among the most well-crafted and important results of this production is no coincidence. For example, the copying of the monumental Bible in four volumes of the Dominican Convent of Lille was entrusted to William of Sens in 1264. Its ‘philological’ rereading was entrusted to the convent prior, Michel de Novirella, famous for being expertissimus in Biblia. In the colophon, he states that he followed the version ‘of the Hebrews and the ancients’.18 It was up to the secretaries of Saint Thomas to write the summary of his classes, ultimately to complete the references cited by the theologian, and always under his guidance, to carefully verify the work of the copyist.19 Around 1270, one of the theologian’s assistants, Rinaldo de Priverno, reviewed his own notes (reportationes) on the commentary on the Gospel of St. John, on behalf of another theologian, in this case a layperson, the dean of Saint-Omer cathedral chapter, Adinolfo d’Anagni, who was active in Paris and associated in particular with the Collège de Sorbonne.20 It must once again be noted that this two-way writing relationship also existed at the chancelleries of Dominican provinces, where ministers always had one or two associates trained in rhetoric and dictamen.21 18 Anno domini M°CC° sexagesimo quarto scripta fuit hac Biblia a Guillermo Senonensi et diligenter correcta secundum Hebreos et antiquos a fratre Michaele de Novirella tunc priore fratrum Praedicatorum Insulensium et cappellano domini pape, expertissimo in Biblia. Cf. Catalogue des manuscrits 1897, 11–12; Beer 1969, 37 n. 42. 19 Dondaine 1956. 20 Nebbiai 2018. 21 Bartoli Langeli and d’Acunto 1999, 399.

Donatella Nebbiai The emergence of the practice of autography is reflected in these evidences. ‘Authorial’ autography is used to describe manuscripts composed and annotated by the author himself, and ‘editorial’ autography is used when the author takes on the task of revising and annotating the work of others in his own hand.22 In the realm of religious institutions, which are the subject of this article, the most famous examples of ‘authorial’ autography are found in the monastic world. Many of the handwritten and hand-annotated manuscripts of Guibert de Nogent, Otloh of Sankt Emmeram, and Sigebert de Gembloux have survived from the tenth and eleventh centuries. Throughout their works, these authors also provide evidence of their writing practices, the ways that they had learned, and the pain it caused them to no longer be able to write, due to age and disease.23 Probably because these monks considered the practice of writing to be a means of personal redemption, related to prayer and manual work, they adopted it far more readily than did their secular confreres in educational institutions. In fact, manuscripts copied and claimed as such by secular masters are rare. These scholars tended to delegate this task to their assistants and simply to verify their copies. In the fourteenth century, at the Parisian Collège de Dormans-Beauvais, there were two escripvains [writers] who, under the direction of a master, created the manuscripts necessary for the needs of the socii and hosts, each of whom was specialized in producing a particular type of work.24 At the Collège de Sorbonne, Gérard d’Utrecht did not sign his manuscripts, no more so than did Gerard d’Abbeville or the dean of Saint-Omer, Adinolfo d’Anagni. Pierre of Limoges and John of Murs used scriptores and checked their work. These verifications were important, because the manuscripts of the Collège de Sorbonne could be used as models for other institutions. One of them has been conserved, with texts by Albert the Great, originating from the Abbey of Saint Victor (Paris, BNF lat. 14728, note on fol. 274). In bequests, the quality of a copy was emphasized by precise terminology, which is worth studying. Nicolas de Vrigny (d. 1264) bequeathed to the Sorbonne his exemplaria, reference manuscripts that were recognized by the university.25 Gerard d’Abbeville (1271) did likewise, leaving his originalia (copies containing the complete works of an author) to the institution,26 while Franciscan cardinal 22 See mainly Gasparri 1994; Lehmann 1959; Giové Marchioli 2015; Garand 1981. 23 Grégoire 1998. 24 Pellegrin 1947 (from college’s accounting registers in Paris, Archives nationales, H3 2875, 1-17 and M 94 A). 25 Douet d’Arcq 1867, 217–219. 26 Denifle and Châtelain 1889, 402.

How Private Libraries Contributed to the Transmission of Texts Matthew d’Acquasparta divided his up among the convents of Todi and Assise (1287).27 Many statements confirm that these volumes served as models. No amendments could be made to them except in authorized places, as emphasized by the will of Bishop of Lille, Wautier Marvis, who allowed the transportation of his manuscripts only to the Dominican studium of Paris, where students could use them.28 A few years earlier, in 1194, the canon Cotta gave seventeen theological and biblical volumes to the Cathedral of Vercelli, requiring that no one but the Cathedral chancellery may be authorized to reproduce them. Cotta mentioned moreover that the texts were correct, because these books had been checked against Parisian copies.29 It was also in Paris and only in Paris that the copy of the commentary on the Decretals donated in 1271 by Cardinal Henry of Susa to the Embrun Cathedral could have been corrected, in copies created exclusively in the cathedral chapter.30 One of the most meaningful terms that translates the notion of reference is without a doubt ‘corpus’, the usage of which to refer to a collection of books can be traced back to Salatiele, a native of Bologna, in 1240.31 According to the sources consulted, ‘corpus’ indicates the collection, the complete series of biblical books. The term denotes the collection’s unity, in other words, its organic nature, which remains even when the will stipulates that it will be divided up among multiple recipients. While Gerard d’Abbeville left his theology ‘corpus’ – the same one that he used to study – to a single heir, the Collège de Sorbonne, in 1287 Galien of Pisa, a former canon of Saint-Omer, protector of the Franciscans, chose by contrast to distribute his corpus between the Cordeliers of Paris (who received two thirds of it), and the Dominicans of the Convent of Saint-Jacques (one third).32 Galien’s will is also interesting because the provisions concerning his ‘corpus’ of books immediately follow those of his burial. The document thus meaningfully alludes first to the physical body of the canon, and then to his spiritual body, composed of books. 27 Menesto 1982, 104–108. 28 Lille, Archives Départementales du Nord, 127 H 30, n° 212; Lille, Archives de la province dominicaine, 27.80.9 (Littere domini Walterii quondam episcopi Tornacensis de libris ad usus fratrum Predicatorum de Insula comparatis). 29 Nebbiai 2001, 87; Frova 1989. 30 Paravicini Bagliani 1980, 133–141. 31 Edited by Orlandelli 1961. 32 Brunel 1997.

Donatella Nebbiai 3 From autography to the library project To have had the idea of passing down one’s books, and above all, to have wanted to make them available to an audience of readers, in both the present and the future, it must have been necessary for a person to consider them as ‘his own’, like a personal work; in short, to have had, with his books, the same relationship that an author has with his writings. Born during the scholastic period, when autography became an established practice, this relationship was largely shared by readers in urban society at the end of the Middle Ages. It facilitated the emergence of numerous projects to supply existing libraries or to create new ones. While these projects required favourable conditions to materialize, and especially the political authorities’ backing, the dissemination of scholars’ work through their books could also have been effectively relayed by social circles – fellow believers, family members, friends, and followers who were responsible for promoting and reproducing their writings and disseminating the content. Therefore, it was not necessary for their collections to remain intact; no more than it was necessary for them to be passed down to an institution. Consider a few examples that illustrate this. The majority of these private projects to supply libraries were born in the fourteenth and fifteenth centuries in northern and central Italy, and particularly in Tuscany, where the book market was widely diverse and newly founded mendicant orders received numerous private bequests. For example, in 1370 the physician Guido di Bagnolo donated a total of 60 books to the school he founded in Bologna;33 in the mid-fifteenth century, Angelo da Gambiglioni, a jurist from Arezzo, left his books to his male children and in turn their male descendants, specifying that otherwise, the library should go to the Brotherhood of St. Mary of Mercy;34 and Florentine merchant Filippo Pieruzzi bequeathed his library to the Cistercian Abbey of San Salvatore a Settimo, where he also chose to retire and where he hosted a cenacle on Greek and Latin studies.35 In humanist Florence, we find the practice of monastic retirement, as practised by the master theologians of the previous centuries. The dynasties in power supported some of these projects when they believed that they backed their interests. Still in Florence, at the end of the fourteenth century, it fell on the Medicis to arbitrate between two competing projects to create a public library: the first was put forth by Palla Strozzi, who planned to 33 Livi 1918. 34 Maffei and Maffei 1994. 35 Fiesoli and Somigli 2009, 7, n° 14.

How Private Libraries Contributed to the Transmission of Texts bequeath his manuscripts to Santa Trinita; the second was promoted by Niccolò de’ Niccoli, who was also the owner of a vast collection of books that he preferred to leave with the Dominicans of San Marco.36 The second option was chosen. However, these favourable conditions were not always the case, and numerous library projects stagnated, despite the prestige of the people who initiated them. For example, the creation of a study library to be used by clerics in the city of Amiens, imagined in the thirteenth century by the cleric Richard de Fournival in his Biblionomia, was never completed.37 A century later, Francesco Petrarch successively came up with two configurations to guarantee an institutional future for his book collection, and both failed. He initially planned to combine it with Boccaccio’s collection so that, according to him, they would remain indecerptae in the future, and could, following their death, be passed on to the guardianship of a religious institution. Boccaccio however subsequently chose to entrust all of his books to the Florentine Convent of the Santo Spirito. At a later date, Petrarch thought of transferring his books to a public library, planning to leave them at the Basilica of St Mark, with the support of the city authorities. However, the death of a few of his close associates resulted in the loss of all of his political support. It is consequently no longer possible to find evidence of the project to found a library in the will that he drew up in 1370. His books were divided up among different recipients. His prayer book, for example, was bequeathed to the Padua Cathedral, of which Petrarch was one of the canons. There were no provisions concerning his books in vernacular language, even though they were numerous and played an important role in establishing his cultural knowledge. Moreover, a few years earlier, in a list-curriculum of ideal reading material, the poet only mentioned Latin and Greek works. Lastly, probably for emotional reasons but also considering their philological value, Petrarch bequeathed some of the original copies of his works plus a few duplicates to his son-in-law, Francescuolo da Brossano, and the da Brossano’s children.38 The dispersion of Petrarch’s library – albeit it not absolute, as several of his manuscripts made it to the library of the lords of Padua – did not harm the cultural dissemination of his works or the reputation of excellence of his manuscripts. To cite only a few examples: the sumptuous complete copy of the Enarrationes in Psalmos by St Augustin (Paris, BnF, latin 1989, eleventh century), originally from central Italy, which had been offered to him in 1355 by Boccaccio, who had acquired it from the Franciscan convent in Siena, and which Petrarch used 36 Gargan 1988, 169–171. 37 Lucken 2017. 38 Regarding Petrarch’s library, cf. Nolhac 1892; Billanovich 1947; Petrucci 1967; Gargan 1988.

Donatella Nebbiai to proudly show off to his friends when he lived in Milan, remained associated with his reputation as a scholar, even after his death. The manuscript of Cicero’s Familiares (Florence, Biblioteca Mediceo-Laurenziana, Plut. XLIX. 9, ninth–tenth century), discovered and annotated by Petrarch in Vercelli, was acquired by Leonardo Aretino and then later by his son, finally ending up in the hands of Donato Acciaioli. The other manuscript by Cicero, discovered by Petrarch in Verona in 1345 and immediately and carefully re-copied, has not been identified, but we still have Titus Livius’ Decades (London, British Library, Harley 2493, twelfth– fourteenth century), partially written in his own hand, which were acquired during the fifteenth century by the humanist Lorenzo Valla.39 The transmission of the works of Arnaldus de Villanova, a Catalan theologian, may be one of the most interesting cases of actions to spread the writings of an author carried out by his close associates. Initiated by Arnaldus while he was still alive, with the help of his private chancellery, these actions took place following his death in 1311. We have a long document drawn up in 1318 by his followers, two years after the inquisitorial sentence of Tarragona condemning his doctrines. The document goes over the operations of his inheritance in an openly memorial way. It details the books that Arnaldus bequeathed to the Carthusian Monastery of Scala Dei, where his nephew was a cleric oblate; it makes reference to the volumes that the Inquisitor, Arnaldo Burguete, had seized at his residence to support his conviction, as well as the works that had been taken to Sicily by a group of followers guided by his assistant and secretary, Andreas Ferrandi. In an albaran (in other words, a quittance), the philosopher, preparing to head out on a trip abroad, which he then later decided against, had also left a portion of his books to another one of his associates. Other books were left to communities of Beguines, his followers in Valencia, Barcelona, Montpellier, and Marseille.40 A few concrete pieces of evidence of this dissemination have survived. Starting in 1311, Pope Clement V wrote a memorandum requesting that someone find him a copy of the book on kidney stones that Arnaldus had dedicated to him. One of Arnaldus’ nephews, John Blasi, also a doctor, who later became a banker and was known for having founded the Hospital of the Holy Spirit of Marseille, related, in 1324, in his commonplace book, that he had two of his uncle’s books at his home: a copy of the medical Speculum and a book on the crusade (Livre du passage de la terre d’outremer).41 The latter work discussed one of the epi- 39 Sambin 1958. 40 On the history of the writing of Arnaldus de Villa Nova, his entourage, and his books, cf. Ballester, McVaugh and Rubio Vela 1989; Manselli 1994; Santi 1994; Nebbiai 2004. 41 Marseille, Archives municipales I/1, 187; Hauck 1965; Villard 1934; Baratier 1968, 272–275.

How Private Libraries Contributed to the Transmission of Texts sodes in the life of the Catalan teacher mentioned above. In an inventory of goods and books at the Hospital of the Beguines of Valencia in 1354, we find another of Arnaldus de Villanova’s books, following one of the provisions of his will. It was a copy of his commentary on the Apocalypse. This scholarly work stood out in a collection of administrative and religious books in vernacular, but its presence can be explained by the fact that this hospital belonged to an environment that supported radical Franciscanism.42 In short, despite the dispersal of his books, the dissemination of this Arnaldus’ work appears to have been carried out as he had himself planned, not only for a literate public but also for the people who read and wrote in the vernacular and who had enthusiastically supported his reformist theories. A while later, in 1405, Bertran Boysset, a land surveyor from Arles, stated that he had borrowed from one of his fellow citizens, the notary Arnoldus del Puey, a land surveying treatise that he attributed to Arnaldus de Villanova. He provided a translation for it, which was according to him ‘word for word’. In Bertran’s manuscript (Carpentras, Bibliothèque L’Inguimbertine, MS 327), several sketches depicting the philosopher, which are simple but not devoid of liveliness, testify to the dissemination of his works in Provence even still at that time.43 Translations also bear witness to this. I will mention only, for his medical and scientific tests, the translations into Hebrew by Abraham Abigdor.44 References Albanès, Joseph-Hyacinthe (1899), Gallia christiana novissima, Histoire des archevêchés, évêques et abbayes de France / d’après les documents authentiques recueillis dans les registres du Vatican et les archives locales, vol. 2: Marseille, Valence: P. Hoffmann. Altamura, Antonio (1954), Riccardo da Bury Philobiblon, Naples: F. Fiorentino. Baratier, Edouard (1968), ‘Le “Journal” de Jean Blaise’, Provence historique, 18: 272–275. Bartoli Langeli, Attilio (ed.) (1985), ‘Nolens intestatus decedere’. Il testamento come fonte della storia religiosa e sociale, Archivi dell’Umbria. Inventari e ricerche, 7), Perugia: Regione dell’Umbria. Bartoli Langeli, Attilio (2008), ‘Il testamento di Enrico Scrovegni (12 marzo 1336)’, in Chiara Frugoni (ed.), L’affare migliore di Enrico. Giotto e la cappella degli Scrovegni, Turin: Einaudi, 397–540. Bartoli Langeli, Attilio (2010), ‘Autografia e paleografia’, in Di mano propria. Gli autografi dei letterati italiani, Rome: Salerno editrice, 41–60. 42 Rubio Vela and Lizondo 1984. 43 Leonelli 1985. 44 Born probably in Arles in 1350; cf. Ballester and Feliu 1993.

Donatella Nebbiai Bartoli Langeli, Attilio and Nicola d’Acunto (1999), ‘I documenti degli ordini medicanti’, in Giuseppe Avarucci, Rosa Marisa Borraccini and Giammario Borri (eds), Libro, scrittura documento nella civiltà monastica e conventuale nel basso Medioevo (sec. XIII-XV). Atti del convegno dell’Associazione Paleografi e Diplomatisti 1997, Spoleto: CISAM, 381–417. Bataillon, Louis Jacques (1983), ‘Les conditions de travail des maîtres de l’université de Paris au XIIIe siècle’, Revue des sciences philosophiques et théologiques, 67: 417–433. Beer, Ellen J. (1969), ‘Das Scriptorium des Johannes Philomena und seine Illuminatoren. Zur Buchmalerei in der Region Arras-Cambrai, 1250 bis 1274’, Scriptorium, 23: 24–38. Bianca, Concetta (2008), ‘Dal privato al pubblico. Donazioni di raccolte librarie dal XV al XVI secolo’, in Fiammetta Sabba (ed.), Le biblioteche private come paradigma bibliografico. Atti del convegno internazionale. Roma, Tempio di Adriano, 10–12 ott. 2007, Rome: Bulzoni, 453–462. Billanovich, Giuseppe (1947), Petrarca letterato, I: Lo scrittoio del Petrarca, Rome: Edizioni di storia e letteratura. Bondéelle-Souchier, Anne (1991), Bibliothèques cisterciennes dans la France médiévale. Répertoire des abbayes d’hommes, Paris: CNRS. Bondéelle-Souchier, Anne (2000), Bibliothèques de l’ordre de Prémontré dans la France d’Ancien Régime, vol. 1: Répertoire des abbayes, Paris: CNRS. Bresc, Henri (1971), Livre et société en Sicile (1299–1499), Palermo: Centro di studi filologici e linguistici siciliani. Brunel, Ghislain (1997), ‘Un Italien en France au XIIIe siècle : Galien de Pise, chanoine de Saint-Omer et fondateur du couvent des Cordelières de Lourcine à Paris, d’après son testament de 1287’, in Histoires d’archives. Recueil d’articles offert à Lucie Favier par ses collègues et amis, Paris: Société des archives de France, 249–276. Catalogue des manuscrits des bibliothèques des départements, XXVI: Bibliothèque de Lille, Paris: Ministère de l’Instruction publique et des Beaux-Arts, Librairie E. Plon, Nourrit et Cie, 1897. Cavallo, Guglielmo (ed.) (1988), Le biblioteche nel mondo antico e medievale (Biblioteca universale Laterza), Bari: Laterza. Courtemanche, Danielle (1997), Œuvrer pour la postérité. Les testaments des gens du roi au début du XVe siècle, Paris: L’Harmattan. Denifle, Henri and Emile Châtelain (1889), Chartularium Universitatis Parisiensis, vol. 1, Paris: Ex typis fratrum Delalain. Dondaine, Antoine (1956), Secrétaires de saint Thomas, Rome: Edizioni di S. Tommaso. Douet d’Arcq, Louis (1867), Inventaire de la bibliothèque du roi Charles VI, fait au Louvre en 1423, Paris: pour la Société des bibliophiles. Écrire son histoire. Les communautés régulières face à leur passé (Actes du 5e congrès international du CERCOR, Saint-Etienne, 6-8 Novembre, 2002) (Travaux et recherches, 18), Saint-Etienne: Cercor, 2005. Ferrua, Valerio (1992), ‘I frati predicatori a Torino. Dall’insediamento a tutto il secolo XIV’, Bollettino storico-bibliografico subalpino, 90: 111–166. Fiesoli, Giovanni and Elena Somigli (eds) (2009), Ricabim. Repertorio di inventari e cataloghi di biblioteche medievali. Repertory of Inventories and Catalogues of Medieval Libraries. Italia. Toscana, Florence: Sismel. Frova, Carla (1989), ‘Città e Studium a Vercelli’, in Luoghi e metodi dell’insegnamento medioevale, Lecce: Galatina, 84–99. Garand, Monique-Cécile (1981), ‘Auteurs latins et autographes’, Scrittura e civiltà, 5: 77–104.

How Private Libraries Contributed to the Transmission of Texts García-Ballester, Luis and E. Feliu, (1993), ‘Intellectual Relations Between Jewish and Christian Physicians: The Hebrew Translation of the Medicationes Parabolae de Arnaldus de Villanova by Abraham Abigdor’, Asclepio, 45/1: 55–88. García-Ballester, Luis, Michael Rogers McVaugh, and Antoni Rubio Vela (1989), Medical Licensing and Learning in 14th Century Valencia (Transactions of the American Philosophical Society, 79/6), Philadelphia: The American Philosophical Society. Gargan, Luciano (1988), ‘Gli umanisti e la biblioteca pubblica’, in Cavallo (ed.), Le biblioteche nel mondo antico e medievale (Biblioteca universale Laterza), Bari: Laterza Editori, 163–186. Gasparri, Françoise (1994), ‘Authenticité des autographes’, in Lucia Pinelli, Paolo Chiesa, and Claudio Leonardi (eds), Gli autografi medievali: Problemi paleografici e filologici. Atti del convegno di studio della Fondazione Ezio Franceschini, Erice, 25 settembre – 2 ottobre 1990 (Quaderni di cultura mediolatina, 5), Spoleto: Fondazione CISAM, 3–22. Géhin Paul (ed.) (2017), Lire le manuscrit médiéval. Observer et décrire, 2nd edn, Paris: Armand Colin [1st edn: 2005]. Giové Marchioli, Nicoletta (2015), ‘Letteratura e scrittura’, Studi medievali, 56/1: 317–331. Gorochov, Nathalie (1999), ‘La notion de pauvreté dans les statuts des collèges fondés à Paris de Louis IX à Philippe le Bel’, in Jean Dufour and Henri Platelle (eds), Fondations et œuvres charitables au Moyen Age, Paris: Éditions du CTHS, 119–129. Grégoire Réginald, (1998), ‘L’autobiografia monastica’, in L’autobiografia nel Medioevo. Atti del XXXIV convegno storico internazionale (Todi, 12-15 ottobre 1997), Spoleto: CISBAM, 81–95. Guenée, Bernard (1973), ‘Histoire, annales, chroniques. Essai sur le genre historique au Moyen Age’, Annales, 28/4: 997–1016. Hauck, Dietrich (1965), Das Kauffmannbuch des Johann Blasi 1329–1337, Sarrbrücken: s.n. Kibre, Pearl (1936), The Library of Pico della Mirandola, New York: Columbia University Press [repr. New York: AMS Press, 1966]. Lehmann, Paul (1959), ‘Autographe und Originale namhafter lateinischer Schriftsteller des Mittelalters’, in Paul Lehmann, Erforschung des Mittelalters, vol. 1, Stuttgart: A. Hiersemann, 359–381. Leonelli, Marie-Claude (1985), ‘Les manuscrits de Bertran Boysset’, in Bertran Boysset. Un arlésien au Moyen Age. Catalogue de l’exposition, Arles: Archives municipales, 22–24 and 41–42. Livi, Ridolfo (1918), ‘Guido di Bagnolo, medico del re di Cipro’, Atti e memorie della R. Deputazione di storia patria per le province modenesi, ser. 5, 11: 78–91. Lucken, Christopher (2017), ‘La Biblionomia et la bibliothèque de Richard de Fournival. Un idéal du savoir et sa réalisation’, in Claire Angotti, Gilbert Fournier and Donatella Nebbiai (eds), Les livres des maîtres de Sorbonne. Histoire et rayonnement du collège et de ses bibliothèques du XIIIe siècle à la Renaissance, Paris: Publications de la Sorbonne, 63–96. Maffei, Domenico and Paola Maffei (1994), Angelo Gambiglioni, giureconsulto aretino del Quattrocento. La vita, i libri, le opere (Biblioteca di storia del diritto italiano, 34), Rome: Fondazione Sergio Mochi Onory per la Storia del Diritto Italiano, 175–218. Manselli, Raoul (1994), ‘Arnaldo di Villanova e i papi del suo tempo tra religione e politica’, in Raoul Manselli, Scritti sul Medioevo, Rome: Bulzoni, 329–349. Menesto, Enrico (1982), ‘La biblioteca di Matteo d’Acquasparta’, in Francesco d’Assisi. Documenti e archivi. Codici e biblioteche. Miniature, Milan: Electa, 104–108. Nebbiai, Donatella (1992), I documenti per la storia delle biblioteche medievali (secoli IX-XV), Rome: Jouvence. Nebbiai, Donatella (2001), ‘Bibliothèques en Italie jusqu’au XIIIe siècle’, in Giuseppe Lombardi, Donatella Nebbiai-Dalla Guarda (eds), Libri, lettori e biblioteche dell’Italia medioevale.

Donatella Nebbiai Fonti, testi, utilizzazione del libro. Livres, lecteurs et bibliothèques de l’Italie médiévale. Sources, textes et usages, Rome: ICCU / Paris: CNRS, 7–129. Nebbiai, Donatella (2004), ‘L’école de Montpellier et les bibliothèques médicales’, in Daniel Le Blévec (ed.), L’Université de médecine de Montpellier et son rayonnement (XIIIe-XVe siècles). Actes du colloque international de Montpellier, 17-19 mai 2001, Turnhout: Brepols, 255–272. Nebbiai, Donatella (2005), La bibliothèque de l’abbaye de Saint-Victor de Marseille (XIe-XVe siècle), Paris: CNRS. Nebbiai, Donatella (2013), Le discours des livres. Manuscrits et bibliothèques en Europe IXe-XVe siècle, Rennes: Presses universitaires de Rennes. Nebbiai, Donatella (2018), ‘Les livres et les relations d’un théologien italien en France, Adinolfo d’Anagni (m. en 1290)’, in Lucien Reynhout and Benjamin Victor (eds), Librorum studiosus. Miscellanea paleographica et codicologica Alberto Derolez dicata (Bibliologia, 46), Turnhout: Brepols, 277–298. Nolhac, Pierre de (1892), Pétrarque et l’humanisme d’après un essai de restitution de sa bibliothèque, Paris: Bibliothèque de l’Ecole des Hautes Etudes. Orlandelli, Gianfranco (ed.) (1961), Salatiele. Ars notarie (Istituto per la storia dell’università di Bologna. Opere dei maestri, 2), Milan: Giuffré. Paravicini Bagliani, Agostino (1980), I testamenti dei cardinali del Duecento, Roma: Società di storia patria. Parkes, Malcom Beckwith (1976), ‘The Influence of the Concepts of Ordinatio et Compilatio on the Development of Books’, in Jonathan J. Alexander and Margaret T. Gibson (eds), Medieval Learning and Literature. Essays Presented to Richard W. Hunt, Oxford: Clarendon press, 115–141. Pellegrin, Elisabeth (1947), ‘La bibliothèque de l’ancien collège de Dormans-Beauvais à Paris’, Paris: Imprimerie nationale [extract from Bulletin philologique et historique, 1944–1945; reprinted in Elisabeth Pellegrin, Bibliothèques retrouvées, Paris: CNRS, 1988, 3–68]. Petrucci, A. (1967), La scrittura di Francesco Petrarca (Studi e testi, 248), Vatican City: Biblioteca Apostolica Vaticana. Petrucci, Armando (1983a), ‘Le biblioteche antiche’, in Alberto Asor Rosa (ed.), Letteratura italiana, vol. 2: Produzione e consumo, Turin: Einaudi, 527–554. Petrucci, Armando (1983b), ‘Il libro antico’, in Alberto Asor Rosa (ed.), Letteratura italiana, vol. 2: Produzione e consumo, Turin: Einaudi, 499–526. Rubio Vela, Agustin and Mateu Rodrigo Lizondo (1984), ‘Els beguins de Valencia en el segle XIV. La seua casa-hospital i els seus libres’, in Estudis en memoria del professor Manuel Sanchis Guarner. Estudis de llengua i literatura, Valencia: Universitat de València, vol. 1, 327–341. Sambin, Paolo (1958), ‘Libri del Petrarca presso i suoi discendenti’, Italia medioevale e umanistica, 1: 359–368. Santi, Francesco (1994), ‘Teologia per immagini e mistica. Le tre denuncie di Marsiglia di Arnaldo di Villanova e altri punti per una macchina metaforica’, in Immagini del Medioevo. Saggi di cultura mediolatina, Spoleto: Centro italiano studi sull’alto medioevo, 234–255. Vernet, André (ed.) (1989), Histoire des bibliothèques françaises, I: Les bibliothèques médiévales du VIe siècle à 1530, Paris: Promodis [reprint: 2008]. Villard, Henri. (1934), ‘Un médecin du roi Robert de Naples, négociant à Marseille au XIVe siècle’, Bulletin officiel du Vieux Marseille, 24–25: 82–90.

Jacques Verger Libraries and Teaching: Comments on Western Universities in the Middle Ages Abstract: Medieval learning always relied on written texts. Nevertheless, medieval schools and the universities themselves in the first decades of their existence were usually deprived of libraries. However, university libraries (of faculties, of colleges, of professors) progressively appeared in the fourteenth and fifteenth centuries, although generally at a limited scale. University libraries reflected up to a certain point the structures and contents of the lectures given in the classrooms. But, as far as they gave masters and students (or at least a part of them) access to a large amount of texts and knowledges otherwise inaccessible through the only means of memory or isolated books, they might also have contributed to the development and orientation of the contents of teaching and teaching methods. Eventually, from a wider point of view, the existence of university libraries probably changed the very relationship to written culture, the image of knowledge, the practical forms of intellectual work for university men and the institutional and financial frame of medieval universities, and paved the way for the emergence of the printing press. Drawing inspiration from an already abundant bibliography, the following comments are somewhat theoretical and general, despite naturally being based on specific examples, taken mainly from the French case. My intention is to inform the comparative thinking that the current volume supports, for beyond simply listing and describing medieval libraries in the East and the West, it focuses on the role, particularly the educational role, of these libraries during the age of manuscript culture. Let us start with some facts. Teaching, as practised in the Middle Ages at universities – which appeared during the thirteenth century and subsequently continuously multiplied – as well as at the urban schools that preceded them during the twelfth century, was essentially based on the study of texts. These texts were both ancient and new: ancient because most of them dated back to Greek, Roman, and Christian Antiquity or the Early Middle Ages; but also new because the majority had been rediscovered, reformatted, compiled, translated, annotated, and commented on during the twelfth century. Sometimes this took place in the very same schools where they were used, but often the work was done outside of them, at ecclesiastical institutions, chancelleries, royal courts, and so Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-011

Jacques Verger on, not to speak of Islamic countries where commentaries and scholarly treatises such as that of Averroes were at times quickly taken to the West and translated.1 These were of course written texts. None of them, not even the Bible, could simply be entrusted to memory and oral tradition.2 Conserving and disseminating them therefore required manuscripts, normally in the form of a parchment codex (as well as paper, from the end of the thirteenth century).3 The best evidence of the importance of these texts, which served as a medium for the scholarly culture passed on through teaching, and of their necessary ties to writing, is the large number of manuscripts which, despite inevitable losses, have survived up until today. For the most important texts, dozens and even hundreds of copies still exist, which is in contrast with the precariousness of the manuscript tradition of many vernacular texts, even the most famous of them. The question that then arises concerns the concrete use made of these manuscripts in school and university environments. How were they copied? How were they disseminated? How were they conserved? How were they consulted and used? What were the respective place and role of isolated books and libraries – the specific focus of this volume – considered, from the angle not only of their contents but also their establishment, organization, and use by readers? 1 Paradoxes of the twelfth century From this perspective, the situation was somewhat paradoxical during the twelfth century. As everybody knows, this century, at least at certain privileged institutions, witnessed the spectacular growth of urban, cathedral, canonical, or ‘private’ schools, as well as the discovery or rediscovery of new or forgotten texts, the creation of aids for study, and the translation of ancient or modern works from Greek or Arabic. In short, it was a particularly dynamic period of intellectual revival.4 However, at the same time, libraries during this period – as well as the highly productive scriptoria – appear still to have been confined to large ecclesi- 1 Regarding the significant textual revival that took place in twelfth-century schools, see the collection of studies in Benson, Constable and Lanham (eds) 1991, particularly Rouse and Rouse 1991 and d’Alverny 1991. 2 Riché and Lobrichon 1984. 3 Glénisson 1988, 14–37. 4 Naturally, there exists much literature on the ‘twelfth-century Renaissance’; see my short summary: Verger 1996b.

Libraries and Teaching: Comments on Western Universities in the Middle Ages astical institutions, monasteries, abbeys, or cathedrals.5 In the schools, teachers and pupils alike, or at least some of them, could get access to these libraries, but did they find exactly what they wanted? Above all, how did teachers prepare their lessons when they had little access to such resources to establish even the most modest personal library? For example, Abelard, who must have been in this case, mentioned nothing about it.6 There is even more uncertainty regarding the situation of students. It seems that teaching was primarily oral for many twelfth-century students. Their difficulty to access real books explains the appearance and success, starting during this period, of a series of improvised substitutes – with a truly promising future – including reportationes of varying degrees of accuracy, anthologies, or collections of ‘sentences’, whose somewhat haphazard creation and circulation from one school or monastery to another is evidenced by many accounts.7 This situation of a chronic deficiency of books therefore explains the success of intellectual institutions such as the Abbey of Saint-Victor in Paris, home not only to a renowned school but also to a beautiful library and a very active scriptorium which, at the initiative of Abbot Gilduin (1113–1155), systematically undertook the ‘editing’, copying, and dissemination of the works of in-house authors, in particular Hugh of Saint-Victor (d. 1141). Regardless of the specific quality of these works, this editorial organization – which is at times described as a sort of ‘proto-pecia’ – ensured that the writings of the eminent authors of the abbey would have a large echo throughout the Western educational world during the twelfth century.8 Other monasteries, abbeys, or cathedrals likely played a comparable role, albeit on a smaller scale, to the benefit of some of their teachers. 2 University libraries Things changed of course from the thirteenth century, with the appearance of universities. Even though this did not completely transform the contents or methods of teaching, it had significant institutional and quantitative effects espe- 5 For example, Vernet 2008, 31–43 and 65–91. 6 Regarding Abelard’s relationship with the books, see the comments by Clanchy 1997, 41–42, 76–79. 7 Regarding the establishment of masters’ collections of theological ‘sentences’ and their circulation during the twelfth century, see the works of Cédric Giraud, in particular Giraud 2010, 185–240 and 339–436. 8 Gilduin’s ‘editorial policy’ has been studied for example in his work on De tribus diebus of Hugh of Saint-Victor, by Dominique Poirel (Poirel 2002, 27–86).

Jacques Verger cially in the largest universities such as Bologna and Paris.9 The rapid growth of the university population and the establishment of a structural organization with statutes and privileges transformed the very conditions of the book market, among other things. Apart from a considerable increase in demand, the question is whether these new conditions modified scholars’ relationship to books, and whether the newly established universities adopted a proactive policy in this respect. Explicit statutes were adopted rather late, but it is clear that from the beginning of the university period, the need to own or at the very least to consult books containing the curricular texts, preferably with glosses or the canonical apparatus, ceased to be exclusively a problem of teachers, extending to their students as an almost sine qua non condition of success in their studies.10 Up to a certain point, universities encouraged students’ access to books. This is evidenced in certain regulations, including the prohibition of dictated classes – which was probably complied with to varying degrees11 –, and verification, during the examen secretum prior to public examinations, that the candidate did indeed own or had at least read the books of compulsory curriculum subjects.12 To make things easier, the university authorities also quickly obtained control rights and jurisdiction in the university city over the book-related professions, which were clearly on the rise. These included parchment makers, copyists, binders, booksellers, and so on.13 An effort was also made to maintain prices at reasonable levels, with the encouragement of the development of the pecia14 9 On the birth of the University of Bologna, see, for example, Greci 1988, 13–44, and on the beginnings of the University of Paris, see Gorochov 2012. 10 The first founding texts of the University of Paris, and namely the statutes issued by the papal legate Robert of Courçon in August of 1215, and the papal bull Parens scientiarum of 13 April 1231, already listed a certain number of obligatory readings for students of the arts and theology (CUP, I, nos 20 and 79). 11 CUP, III, nos 1229 and 1697, §17, and CUP, IV, 726 (no. 2690). 12 We find an account of this verification of books owned or read in multiple statutes on bachelor’s degree examinations in Paris: CUP, II, 673 and 678 (no. 1185); CUP, III, 145, nos 1319 and 1704; CUP, IV, 716, 728 and729 (no. 2690); as well as in Avignon: Fournier 1890–1891, II, no. 1245, §12, no. 1279, §2 and 6. 13 On the relations between book professions in Paris and the university, see Rouse and Rouse 1988, 41–114. The first University of Paris document regulating the activities of booksellers dates back to December 1275 (CUP, I, No. 462); a similar regulation appears in the statutes of the University of Toulouse from 1329: Fournier 1890–1891, I, no. 558, §9. 14 On the pecia system, in addition to the classical book by Destrez 1935, see Bataillon,Guyot and Rouse 19888 and in particular, the article by Shooner 1988. Originating in Bologna, the pecia was found in Paris during the second half of the thirteenth century, as demonstrated by the library regulation of 1275 cited supra note 13, and the rental fees for exemplaria as mentioned infra note 15. Other universities sought to adopt this book production system with varying degrees of success, in particular in France in Toulouse (Fournier 1890–1891, I, no. 545, §27) and Montpellier (Fournier 1890–1891, II, no. 1053).

Libraries and Teaching: Comments on Western Universities in the Middle Ages and taxed prices for the sale or rental of manuscripts,15 all in a favourable technical environment (the spread of paper and the development of a specific module of the university text book at the lowest possible price).16 While it does not appear that French universities attempted to prohibit study books from leaving the university town, as Italian universities did,17 we do find mentions of free loans of codices to poor students,18 for example. One may wonder about the efficiency of this system. As it appears, medieval universities never implemented a systematic policy regarding this subject, designed to directly encourage book production or the creation of translations, even though certain founding statutes or theoretical texts, such as the Siete Partidas in Castile (c. 1260) seem to have envisaged this at times.19 Was the cost of such a policy, which must have been high, the only reason to forgo it? This is uncertain. Moreover, we know that even during the fifteenth century, universities only very gradually opened up to the printing press. Even though the first printing press to be installed in France would be – very briefly around 1470 – at the Collège de Sorbonne, it was not until 1500 that printed books started to replace manuscripts at university libraries, especially for the founding Patristic and Scholastic texts.20 In short, purchasing or ordering books was largely left to the personal initiative – and the financial means – of the scholars who wanted them. It is therefore likely that many students and even teachers, especially young teachers of the arts, still had trouble accessing books and had to settle for a few mismatched volumes. Recurrent complaints around the lack of books and their excessive cost, as well as the theft and damage they suffered, were not simply literary topoi.21 15 CUP, I, no. 530, list dated c. 1275 and CUP, II, no. 642, list dated to 25 February 1304. 16 Cf. Nebbiai 2013, 117–148. 17 This practice has been thoroughly studied in Gargan 1994, 385–400. 18 For example, the medical university in Montpellier in 1240 (Fournier 1890–1891, II, no. 885, §8). 19 The original statutes of the University of Avignon made it the obligation of the general beadle to act as stationer (Fournier 1890–1891, II, no. 1245, §17); this provision seems to have been respected if we believe the example of beadle Antoine de Sexto studied by Léonelli 1977, 115–122. More broadly, the obligatory presence of stationers at every studium generale is mentioned in the school legislation proposed by the second Partida, title XXXI, law 11: ‘Cómo los estudios generales deben haber estacionarios que tengan tiendas de libros para enxemplarios’ (‘How the studia generalia must have stationers that have bookstores for copies’) [ (López [ed.] 1851, 358). 20 See Chartier and Martin 1989, 186–227, and n. 5, 123 and 333–347. 21 The only surviving regulations for French university libraries in the Middle Ages, those of the Sorbonne (dated to 1321) and that of the University of Angers (1431), both contain provisions to combat the loss, theft, or deterioration of books: Glorieux 1968, 214–216 and Fournier 1890–1891, I, no. 469, §14.

Jacques Verger In any event, for our purpose it is not enough to note the increased presence of books in university life, attested to in its own way by a rather stereotypical iconography in which books became a compulsory element of the décor of teaching scenes or representations of scholars at work.22 We must now ask whether and how the concept of a ‘university library’ could have been born and taken shape in this context. Of course there already existed libraries in the West, especially at ecclesiastical institutions, and mainly monasteries and abbeys, but the principles of their composition and operation did not exactly meet the needs of the university population, even though this population could use them when it had the possibility of doing so, such as in the case of the library of the Abbey of Saint-Germain-desPrés, or likely more directly useful, that of Saint-Victor for Parisian students.23 However, to be able to truly speak of ‘university libraries’, the disciplines taught at universities would have had to be well represented, if not predominant, in the libraries. The volumes in their collections would moreover have had to be directly usable – through their contents, presentation, and accessibility – for both the teachers and their scolares. In defining this concept of a ‘university library’, it is necessary to distinguish between ‘virtual’ and ‘real’ libraries. Without knowing the extent to which they corresponded to real book collections, ‘virtual’ libraries consisted of lists of works considered to be essential references for the various university disciplines; in other words, in the minds of medieval teachers, both the fundamental auctoritates and the texts of the glosses and commentaries of ‘modern doctors’ that were essential to properly understand the former, plus the work instruments (tables, directories, distinctiones) necessary for an effective use of both. Because they drew up the curricula and examination programmes of the different faculties in detail, official statutes contained such lists, especially with respect to authoritative texts and textbooks. Examples are the curriculum of the faculty of arts of Paris established in 1255,24 and that of the university of medicine of Montpellier in 1309.25 The lists of puncta taxata composed in certain faculties of law (Bologna or Toulouse, for example) to determine the exact contents of civil or canon law lessons, are somewhat similar.26 22 See Von Hülsen-Esch 2006, and Von Hülsen-Esch 2010, 297–320. 23 For Saint-Germain-des-Prés, see Verger 2015, 63–77, and for Saint-Victor, the observations of Gilbert Ouy in the introduction to Ouy 1999, 26–29. 24 CUP, I, no. 246. 25 Fournier 1890–1891, II, no. 910. 26 The first account of the taxatio punctorum goes back to Bologna in 1252 (cf. Maffei 1975, 73– 101); for Toulouse, it is a list from the end of the thirteenth century or beginning of the fourteenth

Libraries and Teaching: Comments on Western Universities in the Middle Ages In addition to these, there were the less official lists of texts given in certain accessus and other ‘student guides’, containing what we can call the fundamental bibliography of each discipline or the works that exam candidates necessarily had to be familiar with, and which had an evident practical aim, even more so than official lists.27 In Paris, the lists of exemplaria intended for the pecia put into circulation by stationers, the rental fees of which were established by the university, even though they were created somewhat empirically according to the state of the supply, inform us on the main authoritative texts in use as well as the teachers whose commentaries were the most in vogue among students and their professors. We have two copies, from circa 1275 and 1304.28 There were also ‘ideal libraries’, which were more literary: display of scholarship, commentated bibliographies, or ambitious reading curricula. The status of these lists of ancient or modern works, with their varying degrees of detail and which certain authors took pleasure in compiling, is clearly ambiguous. One of the first and most famous examples is the Biblionomia (c. 1250) of Richard de Fournival, a physician and erudite canon. This list probably drew on his fine collection of manuscripts, but given the number of works recorded and their systematic distribution into disciplines and sub-disciplines, it seems to have been a well thought-out and almost exhaustive presentation of the knowledge theoretically accessible in schools of the time.29 This literary genre extended through to the end of the Middle Ages. Another typical example among many is the list of 136 works drawn up in 1408 by Richard de Bazoques, a modest student of theology in Paris who became a schoolmaster in Normandy. It was likely less of an organized catalogue of his personal collection than a sort of encyclopaedic bibliography that he imagined, probably knowing many of the volumes that he listed by their name only.30 Shifting from these theoretical libraries to real libraries, the picture is less clear-cut. Once again, it is first necessary to agree on what can be called a ‘university library’. The answer to this question is not only quantitative: it is not enough for an individual or institution to have possessed a certain number of books for this to be called a library. As noted above, this collection had to have had some coherence, both in its contents and the use that could be made of it.31 century published in Fournier 1890–1891, I, no. 535. A similar device is mentioned in the statutes of the university of law of Montpellier from 1339 (Fournier 1890–1891, II, no. 947, §11). 27 See, for example, Lafleur and Carrier 1997. 28 Cf. supra note 15. 29 The Biblionomia was edited by Léopold Delisle in Delisle 1874, 518–535, and commented in Lucken 2013, 89–125, and Lucken 2017, 63–96. 30 Bignami-Odier and Vernet 1952, 124–153. 31 On the subject of books and university libraries, refer to the regional overview that I presented in Verger 1995, 403–420.

Jacques Verger In that case, could we say that there were student libraries during the Middle Ages? I hesitate to use this term, but we do know that some of the most well off, fortunate, or persevering medieval university students were often able to acquire a small collection of study books, possibly an average of between ten and fifteen volumes.32 These minimalist libraries were however most often limited to fundamental authoritative texts (the Bible and the Sentences for theologians; the Corpus iuris canonici or the Corpus juris civilis for jurists), and potentially a few mismatched ‘reportations’ of lectures, questions, or sermons. In the case of teachers, or at least those with fairly long careers who ended up bequeathing their books to an ecclesiastical institution or a college – which is how we know they existed –, we find the appearance of real libraries from the thirteenth century. These collections, composed with a certain degree of coherence, contained several dozen volumes, including the basic teaching texts, the most authoritative commentaries, and compendiums of ordinary or quodlibetic questions that had some impact.33 To give only one relatively early example of such an authoritative library, I can mention that of Gerard d’Abbeville (d.1272), a Parisian theologian who was well known for his hostility to mendicant religious orders, who inherited a large portion of the books owned by Richard de Fournival (mentioned above), which he in turn bequeathed to the Collège de Sorbonne.34 The Collège de Sorbonne library typified the most interesting type of library with regard to the subject of this article: school, college or university libraries, institutional libraries intended for collective use and guaranteed to last. Once again, while ‘institutional’ libraries had admittedly existed for a long time in the West, but those intended for teaching (including the libraries of study priories or convents in university cities) were a novelty during the thirteenth century. Many of them have already been the subject of studies or at least monographs. Without going into detail, it is fair to say that during the Middle Ages, non-university schools, with the exception of cathedral or monastic schools, usually had no library or had at best a modest collection of books in addition to the handful of grammar textbooks normally owned by the teacher.35 More surprisingly, the same can be said for medieval universities and university faculties, which rarely had their own library. When they acquired one, which was often at a 32 See the brief summary in the chapter by Jullien de Pomerol 2008, 93–111, especially 93–94; a few specific examples in Verger 1995, 409–410. 33 Cf. Verger 2010, 101–116. 34 See Angotti 2014, 29–70. 35 Unfortunately, this subject has not been studied in depth in France; for England, see Orme 2006, 152–155.

Libraries and Teaching: Comments on Western Universities in the Middle Ages late stage, during the fourteenth and fifteenth centuries, it was generally a small library created from a single bequest and which does not appear to have been very useful for students.36 It was not until the fifteenth century that we see certain universities make an effort, albeit one that was not always sustained, to develop their library. Cases of university libraries that were truly worthy of the name, such as the famous Amploniana library in Erfurt,37 were rare. The largest university libraries of medieval times were actually found at secular colleges, such as the Sorbonne or Navarre in Paris, or at what were somewhat incorrectly called ‘regular colleges’, that is, priories, abbeys, or mendicant, monastic, or canonical convents in university cities provided with a studium and integrated within the university. However, the general orientation underpinning the composition and functioning of these libraries depended not only on their educational purposes but also on the spiritual and liturgical requirements of everyday life. That is why I have treated them as secondary in this article.38 Yet not all of these secular or regular institutions, and especially not the most modest of them, actually had a library. Their founder or the first people responsible for them had to have had the idea and desire to have one. Moreover, college libraries were generally established by an initial donation, at times followed by subsequent bequests, as opposed to being the result of a consistent policy to purchase or order manuscripts, which would not have been considered because it probably exceeded the financial means of colleges.39 The largest and best known college library was that of the Sorbonne.40 From the late thirteenth century it stood out, not only for the number of volumes it contained – bordering 2,000 during the first half of the following century – but primarily because of its organization, which is also exceptionally well known to us and has been studied extensively. The elements of this include its precise cataloguing, detailed numbering, distribution of volumes by discipline, distinction between the ‘large’ and the ‘small’ librairie, consultation and loan libraries, multiple copies of the most sought-after works, detailed internal regulations, 36 On the medieval libraries of French universities see Jullien de Pomerol 2008, especially 96– 100, and more specifically, for the southern universities of Avignon and Montpellier-medicine, Verger 1996a, especially 100–102. 37 Speer (ed.) 1995. 38 On the libraries of mendicant orders, see the studies collected in Bériou, Morard and Nebbiai 2014. 39 On the establishment of libraries at southern colleges, see for example Verger 1995, 414–416 and Verger 1996a, 102–106. 40 The medieval library of the College de Sorbonne has been the subject of many works, both ancient and recent, the list of which would be too long to include here. The most recent, which will allow us to go through the prior bibliography, is Angotti, Fournier and Nebbiai 2017.

Jacques Verger and so on. All of this suggests an operation that was already similar to that of a modern library, and attests to a direct link between the structure of this library and the theology studies of the socii at the Sorbonne. It is nevertheless clear that, however fascinating it may be, the case of the College of Sorbonne library is exceptional – we can ask why –, and not only because of the surviving documentation on it. While certain colleges likely drew inspiration from it, the majority were most likely content to provide their fellows with a collection of a couple of dozen books varying in their heterogeneity, with no rules other than those aimed at limiting risks of the loss or excessively fast deterioration of volumes.41 The services that these libraries could provide for their users were therefore not negligible but were certainly limited: there were too few books; collections were not renewed adequately; and use by readers was disorganized. 3 Libraries and teaching This informal overview ultimately brings us back to our initial question: what did libraries, as we have defined them, contribute to teaching at medieval universities? The first element of the answer to this question is that teaching was able to emerge and develop to a certain extent, not without written texts and books, but without libraries. Libraries were almost non-existent during the twelfth century. They remained relatively rare during the thirteenth and fourteenth centuries, and were often inadequate or inconvenient. It was only at the end of the Middle Ages and especially in the modern era, with the invention of the printing press, that we witness the ‘sacralization’ of libraries which really set them at the heart of intellectual life and of the creation, conservation and transmission of knowledge – a position that they are undoubtedly losing today due to the upsurge of digital technology. How was it possible to study and teach without libraries – or virtually without them? This question raises multiple lines of thought: the considerable place granted to the oral tradition and memory; the highly active circulation of people and manuscripts (correspondence, book loans); and the capillary diffusion of occasional writings, which were light and inexpensive (reportationes, sentences, anthologies, excerpta, etc.), are all factors to take into account.42 41 Regarding these libraries, see the typewritten thesis of Rebmeister-Klein 2005, and by the same author, Klein 2017. For colleges in the Midi region, I have suggested an average of 79 volumes (containing 88 works) per library (Verger 1995, 415). 42 Even libraries held these types of productions, even though they got rid of them periodically: the regulations of the Sorbonne speak of multi libri parvi valoris, non ligati, occupantes locum,

Libraries and Teaching: Comments on Western Universities in the Middle Ages As noted above, things started to change in the university era, gradually, not radically. In non-university schools in particular, and often in universities themselves, the difficulty of accessing books continued and the solutions mentioned above remained relevant. Even when the books existed, it was never possible to be sure of receiving the required text in time, or of obtaining a complete volume in good condition. Overall, however, the greater number of books in circulation in university environments certainly modified and improved the conditions for teachers, and even more so for their audiences. This resulted in better training of students, which went hand-in-hand with the statutory formalization of curricula and the precise definition of examination procedures.43 In this respect, we may wonder whether differences in the degree of access to books constituted a discriminating factor in the success or failure of the studies of medieval scolares. All prosopographical studies have shown that medieval universities had high ‘dropout rates’.44 Naturally, there are multiple possible explanations for this, but the inability to own or consult the required books was certainly one of them.45 By contrast, one of the reasons for the success of college institutions was likely that many – but not all – of them offered to their fellows the services of a library, thus contributing to the success of their studies, which was on average greater than that of students studying outside the colleges.46 Moreover, the exceptional fortune of the College of Sorbonne, which ultimately absorbed the entire theology faculty of Paris in modern times, is certainly closely tied to its equally exceptional library. Access to this library was also a highly sicut reportationes et antiqui sermones that will be distributed to poor students (Glorieux 1968, 215); and those of the University of Angers of voluminum, librorum, textuum, lecturarum, quaternorum, sisternorumve atque peciarum collected by the library (Fournier 1890–1891, I, no. 469, §3). 43 The gradual increase in the number of manuscripts available at the end of the Middle Ages was highlighted by Bozzolo/Ornato 1983 (see in particular the first essay, ‘La production du livre manuscrit en France du Nord’, 13–121). 44 I presented the example of the universities in France’s Midi region in Verger 1986: 313–332; other studies have given different rates, but the very existence of the phenomenon is not globally challenged, in my opinion. 45 In Verger 1995, I mention, according to a document conserved in the Toulouse Departmental Archives (Archives of the Château de Pinsaguel, deposited at the Haute-Garonne Departmental Archives, 6 J 164, f. 207r-261v), the proceedings that took place between 1409 and 1423 when a former Toulouse student, Jean de Roaix, sued his guardians, which he reproached for not having provided him with the money necessary (which he evaluated as a minimum of 150 francs) to purchase the books the absence of which caused him to completely fail his legal studies, not allowing him to obtain any degree. 46 The hypothesis that college fellows would have had better grades than other students has been carefully advanced with respect to the universities of Toulouse and Cahors by Foissac 2010, 381–418.

Jacques Verger sought-after advantage among Parisian students, who could be admitted there based on a recommendation from an exercising fellow.47 Furthermore, we can justifiably assume that by offering the possibility of not only possessing isolated volumes but also of individually or collectively establishing book collections of varying sizes, but always with a minimum of internal consistency – in other words, libraries –, the development of the book market during the university era modified the very relationship between scholars and knowledge. First of all, the existence of libraries – whether real or virtual – made studies more systematic in nature by allowing to impose curricula that were themselves defined by a coherent series of books, which theoretically could be found in libraries or with stationers: the Organon and Aristotle’s philosophy, completed by the commentaries of Averroes, the two Corpus iuris and their ordinary glosses, the Articella and Avicenna’s Canon of Medicine, and so on.48 We can go even further and say that disciplines themselves were increasingly identified with corpuses of canonical books and authorized commentaries. Richard de Fournival’s Biblionomia is characteristic of this tendency to identify knowledge with libraries,49 as is the catalogue of the library of the Sorbonne, from a few decades later, because of what it reveals regarding the internal structure of this library and its numbering system, which reflect an analogous desire to embrace all of the exegetical and theological knowledge in the specific varieties of a vast collection of books.50 Libraries were collections of books that varied in their coherence and size, but were not infinite. They contained some books but excluded others, whether voluntarily or involuntarily, which were ignored, refused, excluded, destroyed, censored, or cast aside. As an instrument for defining disciplines, they also established what was orthodox and censured doctrines that were considered deviant or dangerous. The history of censorship at medieval universities has shown that it mainly concerned the actual teaching, but also affected the books associated with it. Certain texts that had been condemned or deemed suspicious – with varying degrees of effectiveness – nevertheless remained quite widespread. The 47 This system is perfectly described in the introduction to Vielliard and Jullien de Pommerol 2000, 21–50. 48 On the relationship between collection of books and university curricula, for the faculty of arts, see Weijers 1996, 9–37; the same demonstration could easily be carried out for other faculties. 49 Cf. supra note 29. 50 Fourteenth-century catalogues for the Sorbonne library were published by Delisle 1881, 8–114. In a presentation that is sometimes criticized, Palémon Glorieux attempted to reconstruct the exact contents of this library according to the first thirteenth- and fourteenth-century catalogues, in Glorieux 1968, 237–289.

Libraries and Teaching: Comments on Western Universities in the Middle Ages 1366 reform of the University of Paris established a committee of theology teachers responsible for examining – and eventually rejecting – the texts of commentaries on the Sentences liable to be copied to become exemplaria available to scribes.51 The library of the Sorbonne, in its composition and the operation of its loan service, gives some idea of the succession of doctrinal movements and the suspicion that fell on certain authors.52 So as not end on this negative note, I can furthermore posit that university libraries might well also have been factors in intellectual expansion and growth. This might have been the case by permitting, sometimes on the borders of Christian orthodoxy, a certain doctrinal eclecticism, sometimes considered characteristic of the fourteenth and fifteenth centuries. These libraries provided teachers and commentators with a variety of information that was at times contradictory and would not have been possible without their collections of books.53 This could also have been the case by enabling the dissemination of university texts outside of the academic environment per se, whether by providing non-university people with access to university libraries, which in reality has little evidence to support it, or by having enabled former students and teachers to conserve their own personal library after having left the university, thus continuing to make use of it when practising a ‘professional’ activity (preacher, judge, lawyer, doctor, etc.) or allowing the members of their social circles to use it.54 Lastly, this could also have been the case considering that, for a variety of reasons, these libraries contained – albeit in a minority and marginally – works that did not belong to university disciplines stricto sensu. While these disciplines accounted for the majority of the books contained in university libraries – those that were the most sought out, most used, and most expensive –, many libraries also had a few varied volumes which historians classify today under the category 51 CUP, III, 144, no. 1319; on the censorship of books at the University of Paris in general, see Bianchi,1999, 23–32. 52 The unpublished thesis of Fournier 2007 and subsequent articles by the same author (e.g. Fournier 2017) have shown that the acquisitions and loans of manuscripts at the library of the Sorbonne were not neutral; they reflected clear ideological choices. From another point of view, I have noted that apart from the works of local teachers, the legal and medical libraries of the colleges of Southern France were filled with writings of Italian authors, but almost completely overlooked the works of scholars from Paris and Orléans, which is once again a reflection of obvious doctrinal choices (Verger 1996a, 111–112). 53 De Libéra 1993, 420 and 471, describes the philosophy of the fourteenth century through its capacity for ‘continuous innovation’, and that of the fifteenth century through its ‘diversity’. 54 Autrand 1973, 1219–1244, has shown, for example, that the libraries of the members of the Parlement de Paris at the end of the Middle Ages were mainly those that they had established at the beginning of their legal studies.

Jacques Verger of ‘miscellaneous’, for want of a better term. These included devotional books, chronicles, astrology or magic books, encyclopaedias, books in vernacular language (secular literature, customary law), and so on.55 It was possibly through these volumes likely to spark their curiosity that students were at times able to escape the strict disciplinary framework established by statutes and become informed in intellectual domains as varied as history, popular devotion, vernacular poems and novels, occult sciences, the classics of Antiquity and of the early Italian humanists, and so on. As we know, the College of Navarre at the end of the fourteenth century and the Sorbonne fifty years later were among the first centres of French humanism. The Collège de Sorbonne appears to have housed the first French printing press in 1470. Their libraries may have played a role in the emergence of these innovative developments.56 4 Conclusion To briefly conclude, we know that medieval universities consecrated the primacy of written material in teaching, and by doing so, the need to use books. This is evidenced in the regular presence of books, despite the high price of some of them, and in a certain prestige expressed through flattering iconography and repeated literary praise. Yet the libraries of medieval universities and colleges had not yet become the intellectual working spaces par excellence, the emblematic images of knowledge, or the privileged beneficiaries of religious or secular patronage that they would become during the modern era, following the invention and spread of the printing press. It is however clear that the milestones leading to this ‘sacralization’ of libraries, by making it possible and desirable, were laid during the last centuries of the Middle Ages. Alongside the princes and prelates who collected beautiful books, the universities and colleges of the time played an essential role in this process. 55 At the Sorbonne, according to Savoye 2017, 185–201, there were only very few books in vernacular language: a few translated scientific texts (geometry, optics) and a few spiritual treatises (Somme le Roi, Evast et Blaquerne, Miserere by the Reclus de Molliens, etc.). In the libraries of colleges in Southern France, the proportion of ‘miscellaneous’ books not belonging to university disciplines was a bit larger, with up to 16% of the total (cf. Verger 1996a, 110–111 and 121). 56 See Verger 2006, 63–75.

Libraries and Teaching: Comments on Western Universities in the Middle Ages Abbreviation CUP = Heinrich Denifle and Émile Chatelain (eds), Chartularium Universitatis Parisiensis, 4 vols, Paris: Delalain, 1889–1897. References Angotti, Claire (2014), ‘Les bibliothèques des couvents mendiants, un modèle pour les séculiers ? L’exemple des deux premiers bienfaiteurs du collège de Sorbonne (Robert de Sorbon, Gérard d’Abbeville)’, in Nicole Bériou, Martin Morard and Donatella Nebbiai (eds), Entre stabilité et itinérance. Livres et cultures des ordres mendiants, XIIIe – XVe siècle (Bibliologia, 37), Turnhout: Brepols, 29–70. Angotti, Claire, Gilbert Fournier and Donatella Nebbiai (2017) (eds), Les livres des maîtres de Sorbonne. Histoire et rayonnement du collège et de ses bibliothèques du XIIIe siècle à la Renaissance, Paris: Publ. de la Sorbonne. Autrand, Françoise (1973), ‘Culture et mentalité. Les librairies des gens du Parlement au temps de Charles VI’, Annales E.S.C., 28: 1219–1244. Bataillon, Louis J., Bertrand G. Guyot and Richard H. Rouse (1988) (eds), La production du livre universitaire au Moyen Âge. Exemplar et pecia, Paris: CNRS. Benson, Robert L., Giles Constable and Caro D. Lanham (eds) (1991), Renaissance and Renewal in the Twelfth Century, Toronto: Toronto Univ. Press. Bianchi, Luca (1999), Censure et liberté intellectuelle à l’université de Paris (XIIIe–XIVe siècles), Paris: Les Belles Lettres. Bignami-Odier, Jeanne and André Vernet (1952), ‘Les livres de Richard de Bazoques’, Bibliothèque de l’École des Chartes, 110: 124–153. Bozzolo, Carla and Ezio Ornato (1983), Pour une histoire du livre manuscrit au Moyen Âge. Trois essais de codicologie quantitative, Paris: CNRS. Chartier, Roger and Henri-Jean Martin (eds) (1989), Histoire de l’édition française, I: Le livre conquérant. Du Moyen Âge au milieu du XVIIe siècle, Paris: Fayard-Promodis. Clanchy, Michael T. (1997), Abelard. A Medieval Life, Oxford and Cambridge, MA: Blackwell. d’Alverny, Marie-Thérèse (1991), ‘Translations and Translators’, in Benson, Constable and Lanham (eds) 1991, 421–462. Delisle, Léopold (1874), Le Cabinet des manuscrits de la Bibliothèque nationale, II, Paris: Imprimerie nationale. Delisle, Léopold (1881), Le cabinet des manuscrits de la Bibliothèque nationale, III, Paris: Imprimerie nationale. Destrez, Jean (1935), La pecia dans les manuscrits universitaires du XIIIe et du XIVe siècle, Paris: Vautrain. Foissac, Patrice (2010), Histoire des collèges de Cahors et Toulouse (XIVe–XVe siècles), Cahors: La Louve éditions. Fournier, Gilbert (2007), Une ‘bibliothèque vivante’. La libraria communis du collège de Sorbonne (XIIIe–XVe siècle), PhD Thesis, École Pratique des Hautes Études (Ve section), Paris. Fournier, Gilbert (2017), ‘Lire l’archive. Les commentaires sur l’Éthique à Nicomaque au collège de Sorbonne à la fin du Moyen Âge et à la Renaissance’, in Angotti, Fournier and Nebbiai (eds) 2017, 125–183.

Jacques Verger Fournier, Marcel (1890–1891), Les statuts et privilèges des universités françaises depuis leur fondation jusqu’en 1789, 2 vols, Paris: Larose et Forcel. Gargan, Luciano (1994), ‘Le note “conduxit” – Libri di maestri e studenti nelle università italiane del Tre e Quattrocento’, in Jacqueline Hamesse (ed.), Manuels, programmes de cours et techniques d’enseignement dans les universités médiévales (Textes, études, congrès, 16), Louvain-la-Neuve, Université catholique de Louvain: Publ. de l’Institut d’études médiévales, 385–400. Giraud, Cédric (2010), Per verba magistri. Anselme de Laon et son école au XIIe siècle (Bibl. d’histoire culturelle du Moyen Âge, 8), Turnhout: Brepols. Glénisson, Jean (1988), Le livre au Moyen Âge, Paris: Presses du CNRS. Glorieux, Palémon (1968), Aux origines de la Sorbonne, I: Robert de Sorbon. L’homme, le college, les documents (Études de philosophie médiévale, 53), Paris: Vrin. Gorochov, Nathalie (2012), Naissance de l’université. Les écoles de Paris d’Innocent III à Thomas d’Aquin (v. 1200–v. 1245) (Études d’histoire médiévale, 14), Paris: H. Champion. Greci, Roberto (1988), ‘L’associazionismo degli studenti dalle origini alla fine del XIV secolo’, in Gian Paolo Brizzi and Antonio Ivan Pini (eds), Studenti e università degli studenti dal XII al XIX secolo (Studi e memorie per la storia dell’università di Bologna, n.s., 7), Bologna: Istituto per la storia dell’università, 13–44. Klein, Karine (2017), ‘Les bibliothèques des collèges séculiers parisiens et la Sorbonne’, in Angotti, Fournier and Nebbiai (eds) 2017, 97–122. Jullien de Pomerol, Marie-Henriette (2008), ‘Livres d’étudiants, bibliothèques de collèges et d’universités’, in Vernet (ed.) 2008, 93–111. Lafleur, Claude and Joanne Carrier (1997), L’enseignement de la philosophie au XIIIe siècle. Autour du ‘Guide de l’étudiant’ du ms. Ripoll 109 (Studia artistarum, 5), Turnhout: Brepols. Léonelli, Marie-Claude (1977), ‘Un libraire d’Avignon à l’époque du Grand Schisme’, in Bulletin philologique et historique (jusqu’à 1610) du Comité des travaux historiques et scientifiques, Paris: Bibliothèque nationale, 115–122. Libéra, Alain de (1993), La philosophie médiévale, Paris: PUF. López, Gregorio (ed.) (1851), Las siete partidas del rey Alfonso el Sabio, II, Paris: Librería de Rosa Bouret y Cia. Lucken, Christopher (2013), ‘La Biblionomia de Richard de Fournival : un programme d’enseignement par le livre. Le cas du trivium’, in Jacques Verger and Olga Weijers (eds), Les débuts de l’enseignement universitaire à Paris (1200 – 1245 environ) (Studia artistarum, 38), Turnhout: Brepols, 89–125. Lucken, Christopher (2017), ‘La Biblionomia et la bibliothèque de Richard de Fournival. Un idéal du savoir et sa réalisation’, in Angotti, Fournier and Nebbiai (eds) 2017, 63–96. Maffei, Domenico (1975), ‘Un trattato di Bonaccorso degli Elisei e i più antichi statuti dello Studio di Bologna nel manoscritto 22 della Robbins Collection’, Bulletin of Medieval Canon Law, 5: 73–101. Nebbiai, Donatella (2013), Le discours des livres. Bibliothèques et manuscrits en Europe, IXe– XVe siècle, Rennes: Presses Universitaires de Rennes. Ouy, Gilbert (1999), Les manuscrits de l’abbaye de Saint-Victor. Catalogue établi sur la base du répertoire de Claude de Grandrue (1514) (Bibliotheca Victorina, 10), I, Turnhout: Brepols. Orme, Nicholas (2006), Medieval Schools. From Roman Britain to Renaissance England, New Haven: Yale University Press. Poirel, Dominique (2002), Livre de la nature et débat trinitaire au XIIe siècle. Le De tribus diebus de Hugues de Saint-Victor (Bibliotheca Victorina, 14), Turnhout: Brepols.

Libraries and Teaching: Comments on Western Universities in the Middle Ages Rebmeister-Klein, Karine (2005), Les livres des petits collèges à Paris aux XIVe et XVe siècles, PhD, Université Paris I–Panthéon-Sorbonne. Rexroth, Frank (ed.) (2010), Beiträge zur Kulturgeschichte der Gelehrten im späten Mittelalter (Vorträge und Forschungen, 73), Ostfildern: Jan Thorbecke Verlag. Riché Pierre and Guy Lobrichon (1984), Le Moyen Âge et la Bible (Bible de tous les temps, 4), Paris: Beauchesne. Rouse, Richard H. and Mary A. Rouse (1988), ‘The Book Trade at the University of Paris, ca. 1250 – ca. 1350’, in Bataillon, Guyot and Rouse (eds) 1988, 41–114. Rouse, Richard H. and Mary A. Rouse (1991), ‘Studium invenire: Schools, Preachers, and New Attitudes to the Page’, in Benson, Constable and Lanham (eds) 1991, 201–225. Savoye, Marie-Laure (2017), ‘Les textes vernaculaires dans la bibliothèque du collège de Sorbonne’, in Angotti, Fournier and Nebbiai (eds) 2017, 185–202. Shooner, Hugues V. (1988), ‘La production du livre par la pecia’, in Bataillon, Guyot and Rouse (eds) 1988, 17–37. Speer, Andreas (ed.) (1995), Die Bibliotheca Amploniana: ihre Bedeutung im Spannungsfeld von Aristotelismus, Nominalismus und Humanismus (Miscellanea mediaevalia, 23), Berlin: De Gruyter. Verger, Jacques (1986), ‘Prosopographie et cursus universitaires’, in Neithard Bulst and Jean-Philippe Genet (eds), Medieval Lives and the Historian. Studies in Medieval Prosopography, Kalamazoo: Medieval Institute Publications, 313–332. Verger, Jacques (1995), ‘Le livre dans les universités du Midi de la France à la fin du Moyen Âge’, in Monique Ornato and Nicole Pons (eds), Pratiques de la culture écrite en France au XVe siècle (FIDEM. Textes et études du Moyen Âge, 2), Louvain-la-Neuve: FIDEM, 403–420. Verger, Jacques (1996a), ‘Les bibliothèques des universités et collèges du Midi’, in Livres et bibliothèques (XIIIe–XVe siècle) (Cahiers de Fanjeaux, 31), Toulouse: Privat, 95–130. Verger, Jacques (1996b), La renaissance du XIIe siècle, Paris: Cerf. Verger, Jacques (2006), ‘Les collèges médiévaux et l’apparition des premiers enseignements humanistes à Paris’, in André Tuilier (ed.), Histoire du Collège de France, I: La création (1530-1560), Paris: Fayard, 63–75. Verger, Jacques (2010), ‘Les bibliothèques des professeurs comme témoignages de leur culture et de leurs méthodes de travail (France, XIIIe–XVe siècles)’, in Rexroth (ed.) 2010, 101–116. Verger, Jacques (2015), ‘L’abbaye et l’université: entre le Pré-aux-Clercs et la Montagne Sainte-Geneviève’, in Roland Recht and Michel Zink (eds), Saint-Germain-des-Prés. Mille ans d’une abbaye à Paris, Paris: Académie des Inscriptions et Belles-Lettres, 63–77. Vernet, André (2008), Histoire des bibliothèques françaises, I: Les bibliothèques médiévales du VIe siècle à 1530, Paris: Éd. du Cercle de la Librairie. Viellard, Jeanne and Marie-Henriette Jullien de Pommerol (2000), Le registre de prêt de la bibliothèque du collège de Sorbonne, 1402-1536 (Documents, études et répertoires publiés par l’IRHT, 57), Paris: CNRS. Von Hülsen-Esch, Andrea (2006), Gelehrte im Bild. Repräsentation, Darstellung und Wahrnehmung einer sozialen Gruppe im Mittelalter (Veröffentlichungen des Max-PlanckInstituts für Geschichte, 20), Göttingen: Vandenhoeck & Ruprecht. Von Hülsen-Esch, Andrea (2010), ‘Gelehrte in Miniaturen spätmittelalterlicher Handschriften. Stereotype, Differenzierungen, Deutungsschemata’, in Rexroth (ed.) 2010, 297–320. Weijers, Olga (1996), Le maniement du savoir. Pratiques intellectuelles à l’époque des premières universités (XIIIe–XIVe siècles) (Studia artistarum. Subsidia), Turnhout: Brepols.

Vanina Kopp An Ideal Library for an Ideal King? Showcasing the Collection, Organization and Function of the Royal Louvre Library in Late Medieval Paris Abstract: The librairie du Louvre existed for a short period of time, from 1368 to 1429. The Louvre library was the institutional frame in which all kinds of manuscripts were collected, inventoried by order and materiality due to a slowly professionalising procedure. Regularly, books were taken out for reading, as presents, or for loans. The inventories, and especially the lists of taken out and returned books, allow us to catch a glimpse of how translations, the circulation and reading of texts functioned at the late medieval French court. Apparently hardly followed up by the next kings and never as systematically institutionalised at the courts of the mighty contemporary dukes, this library underlines how rare and scarce this kind of institution was by then among the lay people in the medieval West. This article presents the historical sources as well as its discursive construction: The first part deals with the organisation and structure of the royal libraries, with a special focus on the Louvre library under the rule of Kings Charles V and Charles VI. The second part analyses the function of the book collection, their use and circulation, as well as reading habits. The third part argues for a discursive construction of an ‘ideal’ library in the service of royal power. The collection will be compared to the narrative choices made by medieval chroniclers in order to depict the literate king as the ideal king, and analyses how the Louvre collection helped in shaping and framing that trope. 1 Introduction In his mirror for princes, Le Songe du vieil Pelerin, the royal counsellor Philippe de Mézières advised the young French King Charles VI to make time for reading: ‘[…] you have to please yourself by reading and hearing the ancient stories for your education. These books, you should read them patiently, for example after Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-012

Vanina Kopp the holy Mass’.1 This text features an extended reading list, suggesting explicitly that the King should be learned in specific disciplines: his education started with moral and theological texts, and proceeded to an intellectual grounding in historical and scholastic knowledge. The King should gain from books not only the basics of military strategy, but also a political education; the list includes the Aristotelian books, translated under Charles V. However, the counsellor warns the King against ‘bad’ literature, including all types of courtly novels, such as the Arthurian stories, which ‘are full of errors that attract the reader to impossibilities, madness, vanitas and sins...’.2 He also alerts him against judicial astrology, necromancy, geomancy or other kinds of astrology that were not authorized by the Church, and ‘other writings that might harm the soul and the good government of [the] royal person’.3 As always with this kind of recommendations, it is very tempting to think that the young King did not follow his tutor’s advice exactly, nor adhere completely to his extensive reading list. However, we have a good idea of where the books (both the ‘good’ as well as the ‘bad’ ones) the tutor mentions came from: the royal library, known as the librairie du Louvre. This was the royal book collection, assembled in small rooms one above the other in a three-storey tower of the Louvre. Here, at its high point in 1411 under Charles VI, almost 950 manuscripts were stored.4 At the time, only the libraries of the Popes in Avignon and the library of the Sorbonne possessed more books,5 and contemporary princely collections in France, Burgundy or England held nowhere near these numbers.6 The librairie 1 ‘[…] Beau Filz, te demoustre que tu te doys delicter en lire et oyr les anciennes hystoires pour ton enseignement, esquelx livres spaciamment tu dois lire, voire apres le divin office, et l’escripture de tes deux tables et ceste present eschequier. Et aussi de quelx livres et ystoires eu te doys garder. […] c’est assavoir que tu doys ensuir les oeuvres des vaillans et des bons et fuir le contraire’ (Coopland 1969, 220). 2 ‘Et […] tu (te) doyes garder de toy trop delicter es escriptures qui sont apocrifes, et par espicial des livres et des romans qui sont rempliz de bourdes et qui attrayent le lysant souvent a impossibilite, a folie, vanite et pechie, comme les livres des bourdes de Lancelot et semblables, comme les bourdes des Veuz du Payen […]. Et combien que les dictes ystoires et bourdes attraient les lisans a vaillance de chevalerie, toutesfois elles attrayent, qui pis est, a amer par amours, qui mal se puet faire sans grant pechie en attrayant au pechie de luxure’ (Coopland 1969, 220). 3 ‘Si te doys bien garder, […] de lire ou faire lire les livres […] toute escripture qui pourroit nuire a l’ame et au bon gouvernement de ta royalle mageste’ (Coopland 1969, 220). 4 This article draws on my monograph: Kopp 2016a, here p. 52. 5 Pommerol/Monfrin 1991; Anheim 2006, 1–27; Angotti/Fournier/Nebbiai-Dalla Guarda 2017; Berndt 2018; see also the articles in this volume by Donatella Nebbiai and Jacques Verger. 6 Wijsman 2010; Winter 1985; Beauvoir 2005, 65–118; Stratford 1994, 187–197; for the French nobility, see Fourcade 2021; for female book holders in a European context, see the database https://booksofduchesses.com/ (accessed on 20 December 2021).

An Ideal Library for an Ideal King? du Louvre existed for a short period of time, from 1368 to 1429. In the latter year, John of Bedford, the English regent for France, bought and took the books with him when leaving Paris. However, the exact fate of the collection is unknown, it is likely that most books crossed the Channel and were dispersed in England, mostly after the regent’s death.7 Today, around 120 surviving manuscripts which can be identified as having belonged to the medieval collection at the Louvre, or to other royal residences outside of Paris.8 The present article does not attempt a reconstruction of the library and its manuscripts. That is a task at which librarians and archivists (especially in France) have excelled at over the last decades: finding extant manuscripts, and analysing the library’s codicological, linguistic and artistic background.9 Instead, the focus here is placed on the virtually existing medieval library as an organic entity in order to contextualise and historicise the place it held at the French court. Following a historical anthropological approach, this paper concentrates on the uses of the book collection and the functions of the library. In so doing, it demonstrates clearly that, indeed, the King’s reading habits ranged more widely than the educational list recommended in the initial didactic example taken from a mirror for princes. The first part of the article will deal with the organization and structure of the royal libraries, with a special focus on the Louvre library under the rule of Kings Charles V and Charles VI. The second part will analyse the function of the book collection, the use and circulation of books, as well as reading habits. The third part will argue for a discursive construction of an ‘ideal’ library in the service of royal power. The collection will be compared to the narrative choices made by medieval chroniclers in order to depict the literate king as the ideal king, and analyses how the Louvre collection helped in shaping and framing that trope. 2 Book collections and organizing knowledge A few sources inform us that even before the late Middle Ages, the royal household was in possession of books owned by both male and female members of the royal family.10 As the topic of this volume is libraries in the East and the West, it is 7 Stratford 1987, 329–350. 8 On the existing manuscripts, see the database europeanaregia.eu/en/historical-collections/ library-charles-v-family (accessed on 30 October 2021). 9 For the general history of the Louvre library, see Kopp 2016a; Delisle 1907; Potin 1999, 25–36; Avril/Lafaurie 1968. 10 Kopp 2016a, 38–46; Avril/Lafaurie 1968.

An Ideal Library for an Ideal King? The biographers also report that after the King’s death, in accordance with the royal will, the books were disseminated among the religious institutions he had favoured in Paris, Compiègne, and Royaumont. Due to this decision, the task of modern researchers attempting to reconstruct this collection is a difficult one.14 Moving into the late Middle Ages, we have more information on book collections. Though Louis IX kept his books in a religious setting, assimilated with relics and the treasury, some source evidence suggests that after him, book collections were housed in the immediate vicinity of the royal person and his living spaces. A ‘tour de la librairie’ might even have existed in the royal palace on the Île de la Cité around 1300.15 In 1368, the French King Charles V left this traditional royal residence on the Île de la Cité, the island in the middle of Paris, where the aforementioned Sainte-Chapelle stood next to Notre-Dame cathedral. The Île de la Cité became the centre of the royal administration, with the chancery (chancellerie), the law court (parlement), and the exchequer (chambre des comptes), whereas the Louvre was dedicated to the royal family, political representations and other royal daily duties.16 Owing to this spatial and functional separation, the books of the royal household and their associated furniture, such as benches and reading wheels, were collected together in one place and the royal library was established on two, and later three, floors in one of the towers of the Louvre castle.17 In the first inventory in 1373, 583 manuscripts were counted, and eventually during the reign of his son Charles VI, the library held over 950 books from a variety of disciplines, ranging from courtly poetry and novels, French historiography, and sciences to encyclopaedias. It also featured a lot of devotional literature; many fine books of piety were inherited from royal predecessors, especially from women. Newer titles entered the collection owing to the patronage of poets, copies presented to the King, confiscations of enemies’ belongings, but only a few due to royal commissions.18 Besides this, kings and queens owned their personal manuscripts. After the Louvre collection was taken to England in 1429, it was only in the sixteenth century, with the Renaissance King Francis I, that a new royal library emerged as an institution that can be continuously traced down to today’s French National Library.19 14 About 15 manuscripts are identified, see Tesnière 2002, 36–40 and 37; Nebbiai-Dalla Guarda 1985. 15 Guérout 1996, 219–288; Potin 2007, 119–140. 16 On the Louvre, see Whiteley 1992, 60–71; Bresc-Bautier 1995. 17 On the architectural developments, see Kopp 2016, 46–50. The Louvre as used by the late Valois Kings can be seen in the October-miniature in the Très riches heures du duc de Berry by the Limburg brothers. 18 For a reconstruction of the procedures, see Kopp 2016a, 82–103; Boucher 2005; for existing manuscripts see Avril/Lafaurie 1968 (exhibition catalogue); Potin 1999, 25–36. 19 Balayé 1988; Bloch 1989, 311–332.

Vanina Kopp 3 Book use: circulations, loans, readings From 1368 on, the library can be studied through written evidence and serial sources, mostly from the inventories of the library. They started in 1373 and were updated intermittently in 1380, 1411, 1413–1416, and finally in 1424–1425. The entries in the inventory note the materiality of the manuscripts: their size and bindings, rich decorations, sometimes the previous owner (if identified), a keyword on their content, and sometimes also mention the style of handwriting and language of the text. Most manuscripts were in the Latin and French languages, whereas some were recorded as being in Occitan, Picard, Spanish or Lombard. The four manuscripts in codex form, among them one with figures in the Hebrew language, are a noted exception.20 Inventorying after the death of Gilles Mallet, the first ‘librarian’, in 1411, the commission proceeded to identify the texts with more details,21 and also created registers of missing books and lists of new books added to the aforementioned inventories, or even lists of separate collections entering the royal collection.22 These separate registers draw a lively picture of the incoming and outgoing books: the so-called ‘guards of the library’ sometimes wrote down dates and contexts for the removal of books, repairs and new bindings, manuscripts on loan or given out as presents, and the eventual recovery of books.23 The notes in the margins of the inventory and the separate list of ‘checked out’ manuscripts allow an insight into the circulation of books from the Louvre library: within the royal household, as diplomatic gifts, or as loans to courtiers or other individuals close to the royal person. Despite a dark legend constructed by nineteenth-century scholars and librarians about the careless ‘dissipation’ of the books, leading to a ‘decadence’ of the library,24 it should be emphasized that among the more than 150 manuscripts taken out of the collection, the exact destination and use of only 18 are unknown.25 For all the other items, the manuscripts either came back to the collection or had always been intended to be given away, thus forming part of the reciprocal system of services and gifts in the courtly environment. It is noteworthy that many of the most pre- 20 See Kopp 2016a, 112–113. 21 For the procedure and the techniques used in detail, see Potin 2007, 119–140. 22 For a first insight see Delisle 1907, vol. 2; for the missing books or those in surplus, see the registers in Paris, BnF MS fr. 2700 (see Fig. 2) and BnF MS fr. 9430; Paris, Mazarine MS 2030, for later copies, see the annex 2 in Kopp 2016a. 23 The fullest register of outgoing books is in Paris, BnF MS fr. 2700, fol. 41r–49v. 24 Delisle 1907, vol. 1 125–141. 25 For detailed analyses, see Kopp 2016a, 218–255.

Vanina Kopp cious manuscripts never left the collection;26 some of them were even stored in the Vincennes castle, away from the daily life of the library, where they were unlikely to be used as loans or gifts.27 Some examples showcase the circulation and use of manuscripts: An illuminated version of the Apocalypse was signed out from the library and shown to Louis d’Anjou, the brother of Charles V, in the 1370s, to ‘make his beautiful tapestry’, as the annotation in the inventory says. It is likely that the Duke was offered several illuminated manuscripts and chose one to take with him to show to his main artist, Hennequin of Bruges, who used the illumination as an inspiration and model for the drawing of the tapestry’s motifs. This is probably the still existing and well-known ‘tapestry of Angers’, displayed today in the castle of Angers. The manuscript was then sent back to the library in Paris.28 Not only highly luxurious manuscripts were signed out. A devotional book, characterized as ‘little booklet’29 and containing French prayers, Latin monthly vigils, and the Hours of Our Lady in French, was probably the book that circulated most widely. Thanks to inventories and the complementary lists of missing and returned books, this booklet’s travels through the royal family can be tracked from before 1368 up to 1416: originally from the Anjou family, the devotional manuscript was given by Charles V to his infant daughter Mary, probably for educational purposes. On her death at the age of seven, the book was returned to the library. It was Charles VI who gave the booklet to his uncle the Duke of Berry, probably for his daughter, also called Mary.30 Other books were signed out to scholars for ‘a lifetime’ and eventually came back, as the inventory notes.31 Evrard de Trémaugon, for example, received a version of the Songe du Vergier, which re-entered the library after the counsellor’s death as noted in the later inventories. This example is interesting, because Evrard de Trémaugon is said to be the author of the Somnium Viridarii and its translation Le Songe du Vergier, a royalist, anti-papal political treaty, which also 26 There is one exception to the rule: while regent of France for the young Charles VI, Louis d’Anjou made frequent requests and received many expensive manuscripts, some of which he took away; some came back, Potin 2005, 181–212; Kopp 2016a, 66–68. 27 Avril 1996, 329–340; on the other residences see Kopp 2016a, 60–66. 28 Delisle 907, vol. 2, 19, Nr. 92: ‘L’Apocalipse, en françois, toute figurée et historiée, et en prose’; in the inventory of 1380 in Paris, BnF, MS fr. 2700, fol. 6r, with the note ‘Le roy l’a baillé à monseigneur d’Anjou, pour faire son beau tappis’; the return mention refers to this item at fol. 42r. On the tapestry, see Delwasse 2007; Yates 1975. 29 ‘petit livret’. 30 For all stations and quotes from the inventories see Kopp 2016, 233–234. 31 Kopp 2016a, 237, here a psalter for Philippe de Mézières.

An Ideal Library for an Ideal King? deals with all the other political issues during the reign of Charles V.32 We do not know if Charles V ordered the first Latin version in 1376, but it is certain that two years later he ordered Evrard to undertake the French translation and adaptation as a kind of ‘political testament’.33 Books from the library were also used as thank-you presents for services to the crown, the texts being chosen from a genre befitting the recipient. The books given as gifts to the King’s medical advisers are striking examples: Charles VI, struggling for most of his reign with mental health issues, gathered around him several physicians/astrologers, and presented several of them with medical, astrological and astronomical texts (even if their medical efforts were in vain).34 The Kings also took books with them on their travels: for example to the Mont Saint-Michel in 1393, Charles VI brought anthologies of antique stories or chansons de geste.35 In the context of knightly renaissance at the time, this type of literature must have been appropriate reading matter for the young Charles VI and his followers—read out loud, for example, in the evening in the army’s tents, the heroic stories of the king’s predecessors would have helped to create a sense of aristocratic community, united by ideals of war and heroic deeds.36 Charles V, on the other hand, seems to have preferred another genre: besides devotional literature, the high number of astrological books signed out of the library, as well as the collections assembled in the other residences of the King, testify to some personal interest. Many astrological texts, such as a horoscope of the Dauphin, astrological tables, or two judicial astrological manuals, bear his autograph signature,37 which is a sign that a book had either been written for him or that he cherished it.38 32 Schnerb-Lièvre 1982; 1993–1995. On the historical background see Cazelles 1982; for a comparative approach see Fletcher/Genet/Watts 2015. 33 Kopp 2016a, 236–237. 34 Among them, a book by Trota, a medical treaty by the eleventh-century Egyptian physician ʿAlī ibn Riḍwān, as well as several meteorological, astronomical and medicinal classics and their commentaries from the Greek tradition, see Kopp 2016a, 225; 248–249. On the place of astrology and medicine at the courts, see Boudet 2006; Boudet/Ostorero/Paravicini Bagliani 2017. On the reciprocity of gifts at court, see Hirschbiegel 2003; for the pre-modern society at large see Zemon-Davis 2000. On the mental health of Charles VI, see Guenée 2004. 35 “Des Faiz de Troye, des Romains, de Thebez, de Alexandre le Grant” are annotated with the comment that the King took them out for this purpose: ‘Le roy le prinst quant il alla au MontSaint-Michel’, Paris, BnF MS fr. 2700 fol. 7r; fol. 42v. 36 Coleman 1986. 37 See the appendix 3 for the number of manuscripts with royal autographs in Kopp 2016a. 38 See Kopp 2016a, 186–200.

Vanina Kopp Christine de Pizan offers a glimpse of how reading customarily took place. She records that the valet Gilles Malet, the so-called ‘first guard of the Louvre library’, was the usual reader to the King because ‘he was good at reading and punctuation’ (which means the modulation of the text when spoken aloud).39 Malet was not only the reader, but also the person who had the keys to the library tower and made the first inventory of the library. From 1368 to his death in 1410, he was a crucial figure for the book collection: This close royal servitor was able to locate the books in the library and signed out what might interest the King, and also read the works aloud to him. So, what did the King choose besides astrology? Again, Christine de Pizan reports on the ideal daily routine of the King:40 especially in the winter, before supper, she says, the King preferred to hear stories from the Bible, historical stories or even philosophical or other scientific topics.41 Although Charles V gathered some scholars around him, the Louvre library was not open to the public. Besides invaluable old books belonging to the King’s family, many books were plain and simple manuscripts, some described in the inventories as ‘old’ or ‘in very old script’,42 and these would never leave the rooms nor get any attention from the King. Most scientific and political works, such as Aristotle’s Ethics and Politics, entered the library for the first time in their translated versions, as well as some encyclopaedias, ‘mirrors for princes’ and royal chronicles. These collections of knowledge came to the royal library not even as commissions, but owing to patronage or gifts, and in vernacular language.43 The manuscripts the King commissioned were mostly compilations of treaties and juridical texts, such as a new ordo for the coronation ceremony,44 as well as encyclopaedias and astrological books. The selection may look odd to a modern reader, and the anecdote at the beginning of this article highlights that not all contemporary scholars shared this view, but it takes on its true significance in 39 ‘[S]ouverainement bien lisoit et bel pontoit, et entendens homs estoit, comme il y ppert […] cellui propre jour fu devant le roy, lisant longue piece par autel semblant et chiere, ne plus ne moins comme acoustumé avoit’, Solente 1936, vol. 2, 63. 40 On the construction in her biography, see Brauer 2015, 46–63. 41 ‘En yver, par especial se occupoit souvent à ouir lire de diverses belles hystoires de la Sainte Escripture, ou des Fais des Romains, ou Moralités de philosophes et d’autres sciences jusques à heure de soupper’ (Solente 1936, vol. 1, 47–48). 42 Kopp 2016a, 82. 43 See the analyses of the acquisition methods in Kopp 2016a, 280–303, and appendix 4; insisting on less royal commission and more voluntary gifts by courtiers see Boucher 2005; arguing for a coordinated translation project see Autrand 1995, 99–106; Sherman 1995. 44 On these, see Kopp 2010, 55–72. On the coronation ordo see OʼMeara 2001, Jackson 1969, 305–324.

An Ideal Library for an Ideal King? a political context, where all kinds of knowledge, even astrology, were deemed useful for counselling a king. To understand the late medieval French context, it is important to differentiate two spheres of language: the administrative, in which Latin remained predominately the language of prestige and law; and the courtly, in which French had gradually overtaken Latin under the political and social pressure of the nobility. Latin remained the language of royal ordonnances and judicial exchange with parts of the kingdom that used medieval Roman law, as well as for decisions made for eternity such as ennoblements and legitimations.45 But even on the political side, French was also used in communicating with the English, whose nobility was still French-speaking, and sometimes with the Emperor of the Holy Roman Empire Charles IV, a Luxembourger raised in Paris.46 Nicole Oresme himself, the well-known philosopher, argued in favour of translations: the classical text, he says in his introduction to the Ethics and Politics, has already been translated into many languages in the past. By translating it into French after it has been translated from Greek to Latin, the translator fulfilled another step of translatio studii. Translatio studii moved, so to speak, from the Greeks to the Romans and now to the French.47 In his eyes, translations into French made Paris a new centre of knowledge, and, therefore, translations were not only a linguistic, but also a political concern. Oresme met the expectations and general trends at court, where all writings undertaken to define the King’s rights and legitimise the Valois dynasty were written in French: the Chroniques de Saint-Denis, the traditional royal chronicle originally composed by the monks of Saint-Denis in Latin, were now being continued in French, and written by the King’s chancellor under the King’s eye.48 Political treaties or papal bulls were compiled by the King’s archivist, and brought to the library; documents on peace treaties or marriage contracts were translated into French and compiled as manuals for use in the library. On the other side, papal letters were still written in Latin, but now included a French summary and brought into the Louvre.49 These examples, very far from luxury manuscripts and expensive commissions, illustrate the contemporary bilingualism: the question was not which language to choose over all, but which language to use appropriate to a specific functional context.50 45 Lusignan 1999, 509–521; Lusignan 2004. 46 Monnet 2020; on his ties to France, see Šmahel 2006. 47 Paraphrased from Lusignan 1989, 311; for the full Excusation, see Menut 1940, 97–101. 48 On this, see Hedeman 1991; Spiegel 1978; Nora 1986, 189–214. 49 These texts have rarely been analysed, see Kopp 2016a, 292; 183-186. See also appendix 4. On the relationship between the archive and the library, see Kopp 2010, 55–72; Artonne 1955; Potin 2000. 50 See Kopp 2016a, 259–279; Lusignan 2004; Boucher 2005.

Vanina Kopp 4 Constructing an ideal This final section deals with the French King Charles V as the prototype of the wise king, how that figure was represented, and the importance of the Louvre collection for its discursive construction. It will focus on the way in which Charles V’s long lasting image as sage roy (‘wise king’) was coined, and how the royal book collection helped to build this trope. The representation of the ruler as a wise king is a trope throughout medieval history. New Solomon, rex literatus, verus philosophus are just some of the labels that were attached to kings such as the French Louis IX ‘the saint’,51 and Charles V, le sage, the Castilian Alphonso X el rey sabio,52 Robert of Anjou il saggio53 or the Staufer Kings Frederic II, known as stupor mundi, and his son Manfred.54 Books, reading and knowledge, and the figure of a reading king are all reflected in contemporary discourses. ‘An illiterate king is like a crowned ass’, claimed John of Salibury in his Polycraticus.55 In the fourteenth century, the biographer of Charles V, Christine de Pizan, used the same image in order to emphasize the importance of knowledge for a good ruler.56 In her biography of the King, she describes him as worshipping the university community and sciences. More importantly, she gives a vivid description of the Louvre library as being a treasure trove of many excellent translations made by the most renowned scholars of their time.57 The attributes and requirements for an ideal king were discussed in ‘mirrors for princes’ throughout the Middle Ages and rose to new popularity under the rule of the Valois kings.58 Wisdom is one aspect needed by an ideal king for just government, and it included art, intelligence, science, but also prudence. It is according to these topoi that the author Christine de Pizan depicts Charles V in her panegyric biography: The wise King was vray philosophe (‘a real philosopher’), astrologien (‘astrologer’), and droit artiste, which means that he was familiar with 51 Le Goff 1996. 52 Burns 1990. 53 Kelly 2003. 54 Imbach/König-Pralong 2013; Boudet 2008, 545–566. 55 ‘Rex illiteratus est asinus coronatus’, Webb 1909, 254. 56 ‘Puis conclut que roy non savant / Tout son fait n’estoit que droit vent / Et qu’autant valoit au regné / Com feist un asne couronné’ (Püschel 1887, verses 5089–5092). 57 Solente 1936, vol. 2, 42–46, the chapter XII of book 3 starts with ‘Item, comment le roy Charles amoit livres, et des belles translacions que il fist faire.’ 58 Krynen 1993, 170–208; Kopp 2021; Delogu 2008, 153–183.

An Ideal Library for an Ideal King? the university’s seven arts.59 For Christine, Charles really assumed his role as wise king by actively supporting the transfer of knowledge under his patronage. His predecessors, such as the twelfth-century Louis IX, the collector and explicator of biblical texts and patron of Robert of Sorbon, or the thirteenthand fourteenth-century Robert of Anjou, Alfonso of Castile, or the Staufer Kings Frederic II and Manfred, are further examples of wise kings, singled out for their intellectual capacities. All these kings were said to be either the authors of prayers, poems, juridical compendiums, astrological treaties, and translations of Latin, Hebraic or Arabic texts, or to have corresponded with well-known scholars or found other forms for the patronage of science and knowledge. Therefore, by the time of Charles V, not only religious knowledge, but also intellectual abilities belonged to the topic representation of wise kings, showing the gradual shifts in the main virtues for an ideal ruler. This narrative was well understood and well used, for example by Christine de Pizan, who devotes one-third of her whole panegyric biography of Charles V to show the wisdom of the ideal king. When awarding Charles V the attribute le sage (‘the wise’), she did not choose her models from the chronologically close medieval Spanish or Italian kings. Rather, she went back to antique models, such as the Egyptian Pharaoh Ptolemy II Philadelphus.60 For her, the Pharaoh had also been an astrologer and she draws a parallel with Charles V’s fondness for astrology as political advisory science, a fable Christine, herself the daughter of an astrologer, always emphasized. Another important parallel, for Christine, between the Pharaoh and King was the library and the commissioning of translations. Thus, the Pharaoh possessed over fifty thousand books in the great library of Alexandria, whereas Charles established the Louvre library. The Pharaoh ordered the translation of the Septuagint, the Torah, from Hebrew into Greek, while Charles commissioned the translation of all the main titles of philosophical and theological texts from Latin to French.61 59 Chapters III; IIII; XI.; XIII of her third book (Solente 1936, vol. 2, 12, 15, 33 and 46). 60 Nevertheless, we have to note that she mixes up the Pharaoh Ptolemy with the astrologer Claudius Ptolemy. 61 Solente 1936, vol. 2, 44–45: ‘De la grant amour, qu’il avoit en avoir grant quantité livres, et comment il s’i [sic] delictoit, et de ses translacions, me souvient d’un roy d’Egipte appellé Ptolomée Philardelphe, lequel fu homme de grant estude, et plus ama livres que autre quelconques chose, ne estre n’en pouvoit rassadié; une fois, demanda à son libraire quans livres il avoit; cellui respondi que tantost en aroit accompli le nombre de .L. mille; et comme cellui Ptholomée oist dire que les Juifs avoient la loy de Dieu escripte de son doy, ot moult grant desir que ceste loy fust translatée de ebrieu en grec, […]. Celle translacion fu moult agreable au roy. Moult fu sage cellui roy Ptholomée, et moult sceut de la science d’astronomie, et mesura la reondeur de la terre.’

An Ideal Library for an Ideal King? One last visual source may show the importance of books and learning in the topic representation of a wise king. This is one of the most emblematic images of Charles V.62 It is taken from the frontispiece of the translation by Dennis Foulechat of John of Salisbury’s Polycraticus.63 This text is not, like the earlier Aristotelian texts, from a university canon, but is a well-known ‘mirror for princes’ and belongs to counselling literature. Charles is depicted in a regal pose, wearing his crown and dressed in a cloak coat covered by the heraldic fleur de lis. He is sitting in a richly ornamented chair in front of a book-wheel. Books are piling up both on the wheel and in a chest on the ground, and on a little desk in the background two books are lying open. God, symbolized by a small hand, comes out of a little cloud above the King and blesses him. Charles points his finger at the big book, opened at a page showing the biblical saying Beatus vir, qui in sapientia morabitur, et qui, which continues [in justitia mediabitur, et in sensu cogitabit circumspectionem Dei].64 This illumination synthesized perfectly the representation of a wise king, as personified by Charles: we see erudition in books, and a learned king, who takes the best from them in order to fulfil his duties as a just ruler. By doing this as best he can through wisdom, Charles V becomes the new prototype of a wise king. These are just a few of many examples of how book prologues and biographers insist on the importance of books and book collections as the core of a royal education, intended to produce a wise, and therefore just and victorious king. From Charles V on, books and translations formed part of this enriched wisdom trope, thus bolstering the narrative of the French king’s ascendancy as the rex christianissimus.65 The intellectual requirements for a king as described in the following ‘mirrors for princes’ culminated in a reading canon,66 which effectively would have required a king to read the entire Louvre library... 62 On the representations, see Hedeman 1991; Sherman 1969. 63 Paris, BnF MS fr. 24287, fol. 2r. See Fig. 3. 64 Eccles. 14: 22. 65 On this narrative in French state building and dynastic legitimation see Krynen 1989, 88–91; Gauvard 2010. 66 Jean Gerson for example recommended the Bible commentaries by the well-known theologian Nicolaus de Lyra as well as his own sermons, Krynen 1993, 191–195.

Vanina Kopp 5 Conclusion Books being a ‘treasure’ for a king is a thought widely spread out at the late medieval court in Paris. This citation comes from Evrard de Trémaugon, jurist and author of one of the royal political compositions. He has one of his protagonists say that ‘It is clear that it is not a hateful, but a very profitable thing, even for a king, to have some books, old and new, it is a beautiful treasure for a king to have a multitude of books’. Later he adds that ‘a king without wide reading (lettrure) is like a ship without helms/rudders, and like a bird without wings’.67 Evrard, like Christine de Pizan, insists on King Charles V’s bibliophilia, and on the commission of translations as one aspect of the King’s sagesse, his wisdom.68 It might not be a coincidence that wisdom became such a prominent cardinal virtue of the king’s government at the very time the French kings were collecting manuscripts in their residences and receiving many translations, done by experts. Mostly the Louvre library was the institutional frame in which all kinds of manuscripts were collected, inventoried by order and materiality due to a slowly professionalizing procedure. Books were eventually taken out for reading, as presents, or for loans. The inventories of the Louvre library, and especially the lists of taken out and returned books, as well as other narrative sources concerning the Louvre library, allow us to catch a glimpse of how translations, and the circulation and reading of texts, functioned at the late medieval French court. This unique evidence for a royal book collection between 1368 and 1429 (apparently hardly followed up by the next kings and never as systematically institutionalized at the courts of the mighty dukes), also underlines how rare and scarce this kind of institution was in the medieval West among the lay people. It might be the uniqueness of it, both for medieval contemporaries and for today’s scholars, that explains the longevity of the library in the collective memory and the ongoing fascination around the surviving lavish manuscripts. Evidently, the narrative of the medieval king’s library went well beyond its short-lived existence: in the year 1995, the French president François Mitterrand laid the foundation stone of the new building of the Bibliothèque nationale de 67 ‘Il appiert donques clerement que ce n’est pas chose detestable, mez est profitable, mesmement a un Roy, avoir plusieurs livres, vieux et nouveaux, pour y avoir recours en temps et en lieu, selon lez divers cas qui luy avienent de jour en jour; et est biau tresor a un roy avoir grant multitude de livres’ (Schnerb-Lièvre 1982, 227). ‘Et pour ce est il ailleurs escript que un Roy sans lattreüre est conme une nef sanz avyrons et come oysel sanz elles’ (Schnerb-Lièvre 1982, 223). My own translation above. 68 See book 3, chapter XII: ‘Item, comment le roy Charles amoit livres, et des belles translacions que il fist faire’, (Solente 1936, vol. 2, 42–46).

An Ideal Library for an Ideal King? France, the national library, as a part of the monumental institutional and cultural programme, commissioned by the president himself. For this symbolic action, he used a stone taken from the recently rediscovered foundations of the medieval Louvre. This act shows the deliberate, even if historically inaccurate, narrative continuation of the Louvre library through the centuries from the Middle Ages into the digital era.69 For some, from this perspective, Charles V as founder of the national library and supporter of the emergence of the French language through translations deserves more than ever his title of a truly wise king. The Louvre library was considered as the starting point of an institution and linguistic movement that is deeply rooted in the self-perception of the French nation.70 Here we can grasp the remains of the historiographical construction of an ideal king from the Middle Ages, and how powerful it is still today. References Primary sources Solente, Suzanne (ed.) (1936), Christine de Pisan Le livre des fais et bonnes meurs du sage roi Charles V, 2 vols, Paris: H. Champion (reprint 1977). Püschel, Robert (ed.) (1887), Christine de Pisan Le livre du chemin de long estude, Geneva: Slatkine (reprint 1974). Schnerb-Lièvre, Marion (ed.) (1993-1995), Évrard de Trémaugon, Somnium Viridarii. Songe du verger, 2 vols, Paris: CNRS. Schnerb-Lièvre, Marion (ed.) (1982), Évrard de Trémaugon, Le Songe du vergier, 2 vols, Paris: CNRS. Daunou, Pierre-Claude-François and Joseph Naudet (eds) (1840), Gaufridus de Bello loco [Geoffroy de Beaulieu], Vita et sancta conversatio piae memoriae Ludovici quondam regis francorum, Recueil des Historiens des Gaules et de la France, Dom Martin Bouquet, vol. XX, Paris: Imprimerie royale. Delaborde, H.-François (ed.) (1899), Guillaume de Saint-Pathus Vie de saint Louis, Paris: Alphonse Picard et fils. Delisle, Léopold (1907), Recherches sur la librairie, 2 vols, Amsterdam: van Heusden (reprint 1967). Hiver de Beauvoir, Alfred (2005), ‘La librairie de Jean duc de Berry au château Mehun-sur-Yevre 1416’, Pecia, 2: 65–118. 69 One example: ‘Toutes ces préoccupations montrent bien que la Bibliothèque de Charles V a vraiment constitué la première ébauche, si timide fût-elle, de ce que fut, au cours des siècles, la Bibliothèque royale et de ce qu’est devenu aujourd’hui la Bibliothèque nationale’ (Dennery 1968, ix, see also figure 6). 70 On the living myth of Charles V as founder of the ‘national library’ and its implication in nineteenth-century French historiography, see Kopp 2016b, 63–86.

Vanina Kopp Jackson, Richard A. (1969), ‘The Traité du sacre of Jean Golein’, Proceedings of the American Philosophical Society, 113:305–324. Webb C., Clemens (ed.)(1909), Ioannis Saresberiensis [John of Salisbury], Policraticus sive de nugis curialium et vestigiis philosophorum, 2 vols, Frankfurt am Main: Minerva (reprint from 1965). Menut, Alfred Douglas (1940), Maistre Nicole Oresme. Le livre des Ethiques d’Aristote, New York: G.E. Stechert. Coopland, G. W. (ed.) (1969), Philippe de Mézières Le songe du vieil pelerin, 2 vols, Cambridge: Cambridge University Press. Secondary literature Angotti, Claire, Gilbert Fournier and Donatella Nebbiai-Dalla Guarda (2007), Histoire et rayonnement du collège et de ses bibliothèques du XIIIe siècle à la Renaissance, Paris: Publications de la Sorbonne. Anheim, Étienne (2006), ‘La bibliothèque personnelle de Pierre Roger/Clément VI’, in Jacqueline Hamesse (ed.), La vie culturelle, intellectuelle et scientifique à la cour des Papes dʼAvignon, Turnhout: Brepols, 1–48. Artonne, André (1955), ‘Recueil des traités de la France composés par ordre de Charles V’, in Recueil de travaux offert à M. Clovis Brunel, Paris: Mémoires et documents de l’Ecole des chartes, 53–63. Autrand, Françoise (1995), ‘La culture dʼun roi. Livres et amis de Charles V’, Perspectives médiévales, 21: 99–106. Avril, François (1996), ‘Les livres de Charles V au château de Vincennes’, in Chapelot, Jean and Elisabeth Lalou (eds), Vincennes aux origines de lʼÉtat moderne, Paris: Editions Rue d’Ulm, 329–340. Avril, François and Jean Lafaurie (eds) (1968), La librairie de Charles V, Paris: Bibliothèque nationale. Balayé, Simone (1988), La Bibliothèque nationale des origines à 1800, Geneva: Droz. Berndt, Rainer (ed.) (2018), Der Papst und das Buch im Spätmittelalter (1350–1500). Bildungsvoraussetzung, Handschriftenherstellung, Bibliotheksgebrauch, Münster: Aschendorff. Bloch, Denise (1989), ‘La formation de la bibliothèque du Roi’, in André Vernet (ed.), Histoire des bibliothèques françaises. Les bibliothèques médiévales. Du VIe siècle à 1530, Paris: Prémodis-Éditions du Cercle de la Librairie, 311–332. Boucher, Caroline (2005), La mise en scène de la vulgarisation. Les traductions d’autorités en langue vulgaire aux XIIIe et XIVe siècles, thèse Paris EPHE, 5e section. Boudet, Jean-Patrice (2008), ‘Le modèle du roi sage aux XIIIe et XIVe siècles. Salomon, Alphonse X et Charles V’, Revue historique, 647: 545–566. Boudet, Jean-Patrice, Martine Ostorero and Agostino Paravicini Bagliani (eds) (2017), De Frédéric II à Rodolphe II: astrologie, divination et magie dans les cours (XIIIe-XVIIe siècle), Firenze: Edizioni del Galluzzo. Boudet, Patrice (2006), Entre science et nigromance. Astrologie, divination et magie dans lʼoccident médiéval (XIIe–XVe siècle), Paris: Publications de la Sorbonne. Brauer, Michael (2015), ‘Politics or Leisure? A Day in the Life of King Charles V of France (1364-80)’, The Medieval History Journal, 18: 46–63. Bresc-Bautier, Geneviève (1995), Le Louvre. Histoire, architecture et décors, Paris: Assouline. Burns, Robert I. (ed.) (1990), Emperor of Culture. Alfonso X the Learned of Castile and his Thirteenth-Century Renaissance, Philadelphia: University of Pennsylvania Press.

An Ideal Library for an Ideal King? Cazelles, Raymond (1982), Société politique, noblesse et couronne sous Jean le Bon et Charles V, Geneva: Droz. Coleman, Joyce (1996), Public Reading and the Reading Public in Late Medieval England and France, Cambridge: Cambridge University Press. Delogu, Daisy (2008), Theorizing the Ideal Sovereign. The Rise of French Vernacular Royal Biography, Toronto: Toronto University Press. Delwasse, Liliane (2007), La tenture de lʼApocalypse dʼAngers, Paris: Regards. Dennery, Etienne (1968) ‘Préface’, in Avril, François and Jean Lafaurie (eds), La librairie de Charles V, Paris: Bibliothèque nationale, ix–xiv. Durand, Jannic and Marie-Pierre Laffitte (eds) (2001), Le trésor de la Sainte-Chapelle, Paris: Musées nationaux. Fletcher, Christopher, Jean-Philippe Genet and John Watts (2015), Government and Political Life in England and France, c.1300–c.1500, Cambridge: Cambridge University Press. Fourcade, Sarah (2021), La Noblesse à la conquête du livre, France v. 1300–v. 1530, Paris: Honoré Champion. Gauvard, Claude (2010), La France au Moyen Âge du Ve au XVe siècle, 2nd ed., Paris: Presses Universitaires de France (1st ed. 1996). Glending, Olson (1986), Literature as Recreation in the Later Middle Ages, Ithaca: Cornell University Press. Guenée, Bernard (2004), La folie de Charles VI. Roi Bien-Aimé, Paris: Perrin. Guérout, Jean (1996), ‘L’hôtel du roi au palais de la Cité à Paris sous Jean II et Charles V’, in Jean Chapelot and Elisabeth Lalou (eds), Vincennes aux origines de lʼÉtat moderne, Paris: Presses de l’ENS, 219–288. Hedeman, Anne D. (1991), The Royal Image. Illustrations of the Grandes Chroniques de France, 1274–1422, Berkeley: California University Press. Hirschbiegel, Jan (2003), Étrennes. Untersuchungen zum höfischen Geschenkverkehr im spätmittelalterlichen Frankreich zur Zeit König Karls VI. (1380–1422), Munich: Oldenburg. Imbach, Ruedi and Catherine König-Pralong (2013), Le défi laïque. Existe-t-il une philosophie de laïcs au Moyen Âge?, Paris: Vrin. Jullien de Pommerol, Marie-Henriette and Jacques Monfrin (1991), La bibliothèque pontificale à Avignon et à Peniscola pendant le Grand Schisme dʼoccident et sa dispersion, Rome: École française de Rome. Kelly, Samantha (2003), The New Salomon. Robert of Naples (1309–1343) and Fourteenth-Century Kingship, Leiden: Brill. Kopp, Vanina (2010), ‘Königliche Archive und Herrschaftsinformation am Beispiel des spätmittelalterlichen Frankreichs’, in Vanina Kopp and Anja Horstmann (eds), Archiv, Macht, Wissen. Organisation und Konstruktion von Wissen und Wirklichkeiten in Archiven, Frankfurt am Main: Campus, 55–72. Kopp, Vanina (2016a), Der König und die Bücher, Sammlung, Nutzung und Funktion der königlichen Bibliothek am spätmittelalterlichen Hof in Frankreich, Ostfildern: Thorbecke. Kopp, Vanina (2016b), ‘Konstruktion, Rezeption, Narration. Karl V. von Frankreich und die Louvrebibliothek im Zerrspiegel ihres Nachlebens’, Francia, 43: 63–86. Kopp, Vanina (2021), ‘Königseinflüsterer oder Regierungsberater? Innovative Impulse von Ratgebern am französischen Königshof unter Karl V. und Karl VI. ’, in Julia Burkhardt and Julia Becker (eds), Kreative Impulse. Innovations- und Transferleistungen religiöser Gemeinschaften im mittelalterlichen Europa, Regensburg: Schnell und Steiner 2021, 375–397.

Vanina Kopp Krynen, Jacques (1981), Idéal du prince et pouvoir royal en France à la fin du Moyen Age (1380–1440): étude de la littérature politique du temps, Paris: Picard. Krynen, Jacques (1993), L’empire du roi. Idées et croyances politiques en France, XIIIe-XVe siècle, Paris: Gallimard. Krynen, Jacques (1989), ‘Rex Christianissimus: A Medieval Theme at the Roots of French Absolutism’, History and Anthropology, 4: 88–91. Le Goff, Jacques (1996), Saint-Louis, Paris: Gallimard. Lusignan, Serge (1989), ‘La topique de la translatio studii et les traductions françaises de textes savants au XIVe siècle’, in Geneviève Contamine (ed.), Traduction et traducteurs au Moyen Age, Paris: Éditions du CNRS, 303–315. Lusignan, Serge (1999), ‘Lʼusage du latin et du français à la chancellerie de Philippe VI’, Bibliothèque de l’École des Chartes, 157: 509–521. Lusignan, Serge (2004), La langue des rois au Moyen Âge. Le français en France et en Angleterre, Paris: Presses universitaires de France. Monnet, Pierre (2020), Charles IV. Un empereur en Europe, Paris: Fayard. Nebbiai-Dalla Guarda, Donatella (1985), La bibliothèque de lʼabbaye de Saint-Denis-en-France du IXe au XVIIIe siècle, Paris: Éditions du CNRS. Nora, Pierre (1986), ‘Les Grandes chroniques de France, le roman aux roys 1274–1518’, in Pierre Nora (ed.), Les lieux de mémoire, vol. 2: La Nation, Paris: Gallimard, 189–214. OʼMeara, Carra Ferguson (2001), Monarchy and Consent. The Coronation Book of Charles V of France, London: Harvey Miller. Potin, Yann (1999), ‘A la recherche de la librairie du Louvre: le témoignage du manuscrit français 2700’, Gazette du livre médiéval, 34: 25–36. Potin, Yann (2000), ‘LʼÉtat et son trésor’, Actes de la recherche en sciences sociales, 133: 48–52. Potin, Yann (2003), ‘Saint Louis l’Africain. Histoire d’une mémoire inversée’, Afrique & histoire, 1: 23–74. Potin, Yann (2005), ‘Le coup dʼÉtat “révélé”? Régence et trésors du roi (septembre–novembre 1380)’, in François Foronda, Jean-Philippe Genet and José Manuel Nieto Soria (eds), Coups dʼÉtat à la fin du Moyen Âge? Aux fondements du pouvoir politique en Europe occidentale, Madrid: Casa de Velázquez, 181–212. Potin, Yann (2007), ‘Des inventaires pour catalogues? Les archives dʼune bibliothèque médiévale. La librairie du Louvre (1368–1429)’, in Hans Erich Bödecker and Anna Saada (eds), Bibliothek als Archiv, Göttingen: Vandenhoeck & Ruprecht, 119–140. Sherman, Claire R. (1969), The Portraits of Charles V of France (1338–1380), New York: New York University Press. Sherman, Claire R. (1995), Imaging Aristotle. Verbal and Visual Representation in FourteenthCentury France, Berkeley: University of California Press. Šmahel, František (2006), Cesta Karla IV do Francie 1377–1378 (The voyage of Charles IV to France, 1377–1378), Prague: Argo. Spiegel, Gabrielle M. (1978), The Chronicle Tradition of Saint-Denis: A Survey, Brookline, MA: Classical Folia. Stratford, Jenny (1987), ‘The Manuscripts of John, Duke of Bedford: Library and Chapel’, in Daniel Williams (ed.), England in the Fifteenth Century. Proceedings of the 1986 Harlaxton Symposium, Woodbridge: Boydell Press, 329–350. Stratford, Jenny (1994), ‘The Royal Library in England before the Reign of Edward IV’, in Nicholas Rogers (ed.), England in the Fifteenth Century. Proceedings of the 1992 Harlaxton Symposium, Stamford: Paul Watkins, 187–197.

An Ideal Library for an Ideal King? Tesnière, Marie-Hélène (2002), ‘Les livres royaux. Construction idéales et réalités’, in Myriam Bacha and Christian Hottin (eds), Les Bibliothèques Parisiennes. Architecture et décor, Paris: Action artistique de la ville de Paris, 36–40. Wijsman, Hanno (2010), Luxury Bound. Illustrated Manuscript Production and Noble and Princely Book Ownership in the Burgundian Netherlands (1400–1550), Turnhout: Brepols. Whiteley, Mary (1992), ‘Le Louvre de Charles V: dispositions et fonctions d’une résidence royale’, Revue de l’Art, 97: 60–71. Winter, Patrick de (1985), La Bibliothèque de Philippe le Hardi, duc de Bourgogne (1364–1404), Paris: Éditions du Centre national de la recherche scientifique. Yates, Frances A. (1975), The Valois Tapestries, 2nd ed., London: Routledge & K. Paul. Zemon-Davis, Natalie (2000), The Gift in Sixteenth-Century France, Oxford: Oxford University Press.

Index of Manuscripts Baltimore, Walters Art Museum 29 Cambridge, University Library Add. 815 40 Carpentras, Bibliothèque L’Inguimbertine 327 221 Casablanca, King ʿAbd al-ʿAzīz Institute 330 70 Doha, Museum of Islamic Art n.n. 34 Dublin, Chester Beatty Library Per 257 32 Fez, Qarawiyyin Library 275 69 1013 88, 90, 94, 100–102 1266 10 1060 74, 88, 94, 100–101 1512 70 Fez, Qarawiyyin Library, Khurūm collection 1-4-110 [= Fgmt 4] 84, 97, 99, 101 1-21-589 [= Fgmt 6] 84, 89, 97, 100–101 2-6-367 [= Fgmt 2] 96, 99, 101 5-49-626 [= Fgmt 7] 80–81, 84, 97, 99, 101 6-8-624 [= Fgmt 8] 85, 97, 99, 102 8-9-1667 [= Fgmt 1] 85, 96, 100–101 10-29 64 10-43-2 [= Fgmt 9] 81, 85, 97, 100, 102 16-2-270—280 [= Fgmt 10] 97, 100, 102 16-7-658 [= Fgmt 10] 97, 100, 102 16-10-684 [= Fgmt 11] 97, 100, 102 16-11-2400 [= Fgmt 5] 89, 97, 99, 101 16-17-1663 [= Fgmt 5] 89, 97, 99, 101 16-24-170 [= Fgmt 5] 89, 97, 99, 101 17-4-685 [= Fgmt 12] 97, 100, 102 17-5-473 [= Fgmt 3] 96, 99, 101 Florence, Archivio di Stato Mercanzia 1054, c. 140, 1334 213 Florence, Biblioteca Medicea Laurenziana Plut. IX. 18 171 Plut. XLIX. 9 220 Houston, Art and History Trust Collection n.n. 27 Lille, Archives Départementales du Nord 127 H 30, n° 212 217 Lille, Archives de la province dominicaine 27.80.9 217 London, British Library Harley 2493 220 Or. 14139 23 London, British Library, India Office Collection Pers. 383 28–29 London, Royal Asiatic Society of Great Britain and Ireland 239 19–20 Marseille, Archives départementales des Bouches du Rhône 6 G 104, n° 696 213 Oxford, Bodleian Library Clarke 39 186 Paris, Archives nationales H3 2875 216 Open Access. © 2023 the author(s), published by De Gruyter. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. https://doi.org/9783110779653-013

Index of Manuscripts 216 1780 D/1 1411 D 1456 K 1469 D 1599 D 2048 D/1 1/2188 D 2217 D/6 1/2237 D Paris, Bibliothèque nationale de France arabe 6090 42 Coislin 263 189 est. VA-225-FOL 246 fr. 2700 248–251, 262 fr. 9430 248 fr. 24287 256–257 gr. 598 187 gr. 1115 200, 202 lat. 1989 219 lat. 14728 216 Riyadh, King Saʿūd University Library 7036 71 Tiznit, Private collection n.n. 54 Paris, Bibliothèque Mazarine 2030 248 Patmos, St John Monastery 140 187 175 187 Patna, Khuda Bakhsh Oriental Public Library 451 35 470 24 Patna, Private collection Christie’s 14 October 2003, lot 25 71 69 69 69 69–70 71 69 71 69 25 Rabat, Ḥasaniyya Library 50 JK 69 4911 40 6778 70 13003 70, 96 13900 68–71, 82, 94, 96 Rabat, ʿAllāl al-Fāsī Library 276 70 Rabat, National Library of the Kingdom of Morocco 16 K 69–71, 79, 82–84, 89, 93, 95, 99–102 162 K 69 245 D 70 425 D 70 470 D 69 609 G 40, 51 San Lorenzo de El Escorial, Real Biblioteca del Monasterio de El Escorial 48 42 922 47 135 47 1862 59 248 47–48, 59, 91–93 270 43, 48, 50, 91 272 50–51, 60 273 50 277 51–52 282 48 283 48 288 44, 59–60 290 43 291 50 303 43 304 47–48, 50, 91 305 48 3061 51 307 51 308 51 309 51 311 50–51 312 51 315 47 316 48 318 48 319 44 3305 44, 59–60

Index of Manuscripts 323 324 325 328 330 331 332 333 334 344 354 356 357 361 362 363 367 369 375 378 379 381 384 385 3902 394 3961 399 400 401 402 403 404 408 410 413 414 415 416 417 419 420 421 422 423 51, 53 51, 53 51, 53 51, 53 51, 53 45, 60 44, 59–60 60 59 59 59 51, 54 48 54, 57–58 45, 54 48, 50–51, 91 50, 91 50, 91 47 51, 54, 56 51 53 53 53 48, 50, 91 47 48 50–51, 60 44, 59–60 50, 91 44 44 44 44 50, 91 54 45, 59 48 47–48 50–51 59 47 45 47, 51 48 48–49 48 434 435 436 438 440 442 446 447 455 456 470 4761 478 480 481 485 486 487 488 4893 490 5474 6793 788 96412 1032 1072 1302 1730 T.I.1 267 54 50–51 50, 91 54 48 50, 91 45, 48 44 54–55 54 44, 48, 50, 91 51 44 51 45 44 44 48, 50, 91 45 59 43–44 47 59 59 59 95 95 44 40 191 Toulouse, Archives of the Château de Pinsaguel 6 J 164 235 Vatican City, Biblioteca Apostolica Vaticana Vat. gr. 207 190–191, 203 Vat. gr. 1524 169, 171 Venice, Biblioteca nazionale Marciana Marc. gr. 131 171 Wazzān, Library of the Great Mosque 1155 70 1/1214 69