66th IFLA Council and General
Jerusalem, Israel, 13-18 August
Code Number: 087-141-E
Division Number: II
Professional Group: Rare Books and Manuscripts
Joint Meeting with: Art Libraries
Meeting Number: 141
Simultaneous Interpretation: No
Towards a database of the Arabic manuscripts in The British Library: a case history
Colin F. Baker
The British Library
London, United Kingdom
"We are all aware of libraries which still wait to announce their accessions of the last hundred years or more."
This comment, made some thirty years ago by the late Professor J. D. Pearson in the preface to his Oriental Manuscripts in Europe and North America1 was certainly relevant to the British Library's collection of Arabic manuscripts, and no doubt to other major academic libraries. In this paper, I wish to discuss a major development in collection access; how the British Library has recently made information about all its Arabic manuscripts effectively accessible and as quickly as possible to its users; how the project evolved; and how the nature of its information forms the basis of a proposed on-line union handlist of Arabic manuscripts in the UK.
Given the size of the British Library's collection of Arabic manuscripts, it is not unsurprising that Pearson's comment has held true for so long. With some 14,000 codices, containing more than 20,000 works, the British Library's collection represents more than half of the total number of some 25,000 Arabic manuscript codices within UK institutions2. It can claim with some justification to be one of the most important of such collections in Europe or North America, both numerically and in terms of quality. In fact, the British Library's collection is actually two historic collections in one, combining the Arabic manuscripts of the old British Museum Library with the holdings of the India Office Library, once part of the British Foreign & Commonwealth Office. Its riches are renowned. Among its many illuminated and illustrated treasures, for example, is what is possibly one of the earliest Qur'an manuscripts dating from the late eighth century AD; the unique seven-volume Mamluk Qur'an, written completely in gold for Sultan Baybars II in Cairo; an early fourteenth century copy of al-Qazwini's Wonders of Creation ('Aja'ib almakhluqat); and, among others, a fourteenth century manual of horsemanship (Kitab nihayat al-su'l wal-umniyah fi'ilm al-furusiyah).
The formation of what is now the British Library's collection of Arabic manuscripts over the past two centuries or more illustrates the process by which a number of originally private collections came together to form a huge public collection and international resource. This reflects the political, trading and cultural interests of administrators, missionaries, scholars and traders in the Middle East and the Indian subcontinent.
Concerning the old British Museum Library, the most important early acquisitions were the manuscripts collected by Claudius James Rich, a typical administrator-scholar figure of the period, while British resident in Baghdad. His collection of 390 manuscripts was purchased by Act of Parliament in 1825 for just £7,000! Other large collections added to the British Museum Library during the nineteenth century included 328 manuscripts bought from the Austrian traveller Dr Eduard Glaser, consisting mostly of Zaidi literature; 246 manuscripts from the estate of Colonel Robert Taylor in the 1850s, who was successor to Rich as Political Resident in Baghdad; a collection of 198 manuscripts, assembled in Damascus and Cairo between 1849-80 by Alfred Freiherr von Kremer; and 104 manuscripts from E. W. Lane's collection, many of which formed the basis of his celebrated Arabic-English Lexicon.
In contrast to the old British Museum Library, more than half of the India Office Library's collection comes from a single source. These are the 1,950 Arabic manuscripts from the so-called 'Delhi collection', representing what remained in the imperial Library of the Mughal dynasty at Delhi in 1858. Other important collections in the India Office Library include 438 manuscripts from the Bijapur collection, the remnant of the Adil Shahi rulers; 94 manuscripts from the Library of Tipu Sultan; 141 manuscripts purchased from the nabob Richard Johnson in 1807; and 72 manuscripts bought from Warren Hastings, the first Governor-General of India, in 1809.
Due to the to the size and diversity of the British Library's holdings of Arabic manuscripts, we must recognise the fact that published catalogues to these collections and to on-going acquisitions have been variously sporadic and uneven in their levels of information, ranging from the fully descriptive, scholarly catalogue to the basic handlist. For example: on the old British Museum side, since Ellis & Edwards's A Descriptive list of the Arabic Manuscripts3 of 1912 there had not been a published British Library handlist of recent acquisitions until Roderic Vassie's classified handlists of Qur'anic sciences, Hadith and Islamic law, published in 19954. To take an example from the India Office Library side: the catalogues dealing with Qur'anic literature, Sufism and Ethics, Fiqh, and Kalam were last published as fascicles between 1930 and 19405 - and only as supplements to Otto Loth's 1877 publication entitled A Catalogue of Arabic Manuscripts in the Library of the India Office.6
It must also be noted that although manuscripts dealing with Qur'anic sciences, Islamic law, Hadith, Kalam and Sufism represent obviously the major part of the British Library's collection - in fact 45% - they do not constitute the whole of the subject make-up and range. 16.5% of the collection is taken up by philosophy, history, geography, ethics and polity. Accounting for 22% of the collection is Arabic language and literature. Included in this broad subject area are Arabic grammar, lexicography, and philology, rhetoric and prosody, poetry and prose, anthologies and literary miscellanies, encyclopaedias of arts and sciences, music and other arts. While the sciences represent 9% of the collection - mathematics, proto-sciences, astronomy and medicine - 7.5% account for literature of other religions and miscellaneous documents. Among these documents are specimens of calligraphy, letters, firmans, and religious endowments (awqaf). The literature of other religions includes Arabic texts relating to the Druze, Bahais, Christians and Jews. We must also include in this category manuscripts in the Arabic language, though not in Arabic script, such as Arabic in Hebrew script (Judaeo-Arabic), Arabic in Syriac script (Karshuni) and Arabic in Samaritan script.
While the British Library's current published catalogues of Arabic manuscripts provided published information on some 7,620 manuscripts, representing just over 50 per cent of the total collection, the obvious need was to provide ready information and easy access to the whole of the collection. This situation has improved considerably within the last five years. We are now in a position to offer information on all Arabic manuscripts in the collection through a single point of access, shortly-to-be-published, namely, the Subject-Guide to the Arabic Manuscripts in the British Library.
The Microfiching Project
The catalyst for this major development was a massive project, partnered and funded by the King Faisal Center for Research and Islamic Studies in Riyadh, its purpose being to make the entire collection of the British Library's Arabic manuscripts available on microfiche. This provided the unique opportunity to carry out an extensive house-keeping operation, involving a complete volume-by-volume stocktaking and checking, with equal treatment to each manuscript in the collection - and with complete curatorial objectivity; everything, including fragments on papyri, vellum and paper, as well as documents, scrolls and decorative calligraphic panels. (Incidentally, the timing happened to be fortuitous, as this house-keeping operation coincided with preparations for the major move of the Oriental Collections from Orbit House to the new British Library building in St. Pancras.)
The question arises: why did we microfiche the collection instead of digitising it- The main reason was the date. Negotiations began in mid-1991 and we were not ready for this process at that time; moreover, microfiche was what the King Faisal Center wanted. With the benefit of hindsight, if we were starting now we would, I believe, opt for digitisation because we could then add an expanded database. Other considerations which had to be taken into account were the necessity - for security and preservation reasons - to film on site; the setting up of a dedicated filming operation to ensure that deadlines were met, and that the work proceeded steadily rather than piecemeal; and the need for linguistic assistance dedicated to the preparation of the material for filming. The problem of filming a large collection is further exacerbated when that collection is only partially and inadequately catalogued. All of which brings me to the main theme of this paper; namely, the compilation of the Subject-Guide.
The Subject-Guide developed out of the microfiche filming project. Originally, only a manuscript-microfiche number concordance was envisaged. The compilation of the indexes for this project was naturally dictated by the structure imposed by the filming. The manuscripts were filmed according to eighteen broad subject categories based on those in the British Museum catalogues of Arabic manuscripts, beginning with Qur'an, Hadith, Kalam and Fiqh, and continuing with Mysticism and Pietics, Prayers and Sermons, Philosophy, etc. This involved searching the catalogues, handlists, catalogue slips and departmental registers of both the Oriental and India Office collections to ensure that the relevant manuscripts were brought within each category.
Concerning the choice of categories, since previous catalogues were arranged according to subject, or at least provided subject indexes, compilation originally appeared to be a simple matter of amalgamation. But as previous compilers differed in their assignment of analogous manuscripts, it was decided to alter even some of their justifiable preferences for the sake of uniformity and ease of access.
Within each category are listed the authors, the titles of their works, the date of copying, the British Library manuscript number, the British Library catalogue reference and the microfiche number. Of course this is basic as it provides summary information extracted from the published catalogues and departmental registers. But this basic work had to be done. All information was checked against the original manuscript, and where information was deficient or non-existent, some preliminary research was undertaken. It must be underlined that this index is only a schematic guide and in no way a scholarly analysis. It is purely the key to the collections, as one would say in Arabic, Miftah al-kunuz, 'the key to the treasures'.
Before discussing the structure of the Subject-Guide further, I should explain why we still did not opt at this stage for a computerised database for compiling the indexes. With our large and diverse collection, it became clear to us that it would be far easier and more efficient to develop a fully descriptive database once the subject matter has been manually categorised and identified. Therefore, what this Subject-Guide offers are the results of extensive preliminary work, which would have been necessary before embarking on any descriptive Arabic cataloguing database project involving detailed codicological and palaeographical analysis of the manuscripts. Moreover, we have a tangible result with immediate benefit for all. The intellectual activity involved in sorting out the material and information manually proved exceptionally productive in this exercise.Time spent developing a suitable database for access to both on-site and remote users would have eaten considerably into a project for which the time-table was very tight, as this would have involved having to solve the problems of stand alone or on-line databases, as well as the issue of vernacular and romanised author/title searching. As the aim was to finish filming within three years, providing ready access at this stage through the traditional format of a British Library publication was, and still is, thought to be the best approach. It cannot be over emphasised that this Subject-Guide is no substitute for fully descriptive catalogues of its Arabic manuscripts which the British Library is committed to producing.
To return to the Subject-Guide, a number of basic questions had to be resolved. How, for example, to present the names of classical Arabic authors? Our objective was simplicity and convenience with appropriate cross-references. Given we are dealing not only with Muslim authors from the Near and Middle East and from the Indian subcontinent, but also with Christian and Jewish authors having both Arabic and non-Arabic elements within their nomenclature, the need for uniformity necessitated a compromise. We therefore considered how the names of Arabic authors are expressed according to common usage within the British Library; the form of the name found in the colophon, title page or the text of the introduction; and how they compared with Brockelmann, Sezgin, Kahhalah and al-Zerekly. In a number of cases we followed the Library of Congress name authority, though such authority does not of course exist for many Arabic names found in manuscripts. Our manuscripts contain not a few hitherto unrecorded authors and titles.
Taking all things into consideration, we therefore preferred the nisbah as the leading element of name, followed by the ism and nasab. Wherever possible this is the form under which authors are listed. Where authors are better known under another element, such as the nasab or kunyah, we have used that form. We ignored titles or ranks except in rare cases where the author is well known by such an appellation. Ecclesiastical ranks have been helpful in specifying Christian Arabic authors and these have been included in contrast to the rest of the Subject-Guide: The same applies to honorific laqabs for distinguishing authors, particularly those of Yamani Imams. Translators, compilers and editors of works are also listed, their contribution indicated in brackets.
As for titles of works, we listed these under each author in chronological order of copying, dated items preceding those assigned to an approximate century. To impose uniformity to the titles of manuscripts as found in the texts themselves and as listed in the standard sources, it was necessary to make amendments during the compilation of the Subject-Guide, which accounts for any variation from the catalogue entry.
Another consideration: how to establish the true identity of a work? In some cases, it has not been possible to decide whether works with titles, often differing in varying degrees, are in fact the same works, recensions, or entirely different works. We have therefore listed such works under the individual titles found. Some correlation between texts remains to be made. At this stage this task must be left for further research which more detailed cataloguing projects will hopefully solve.
How does this Subject-Guide benefit us? First, it offers summary information in a single point of access, which enables us for the first time in the history of the British Library to offer a complete listing of its collections of Arabic manuscripts. Secondly, it provides at a glance the date range of manuscripts of individual Arabic works within the history of their textual transmission. This is still an important and relevant research issue as Arabic is a field in which so many texts are even today unavailable in the form of printed critical editions. Thirdly, it is adding a fascinating 'Indo-Arabic' dimension to the provenance and transmission of the texts in our collections, by enabling us to see how the Arabic manuscripts from the India Office collections complement the Oriental collections. Finally, and obviously, we must add its usefulness to the large number of users who come to the British Library with reference enquiries, as well as internally for the selection of new acquisitions. Indeed, we are well on the way to invalidating the late Professor Pearson's apt and just comment on the endemic situation of manuscript cataloguing in Oriental collections.
Proposed On-line Union Handlist of Arabic Manuscripts in the UK
We are now not only able to make a major advance in the development of an on-line handlist, but are also at an exciting stage in the documentation of Arabic manuscripts throughout the UK. Our Subject-Guide is serving not only as a model but as a catalyst for other UK libraries. Based on what was done in the British Library, the Bodleian Library at Oxford has proposed using and developing our Guide as a subject-structure for an on-line UK union handlist. This would provide a standardised public access from a web-mounted database, and offer for the first time a national listing of the holdings of Arabic manuscript collections. The World Survey of Islamic Manuscripts7 provides important preliminary work for such a project, as it has already identified those UK libraries and college institutions with holdings of Arabic manuscripts. The advantages of this project to librarians and to specialist users are obvious. With the possibility of more subject descriptions for each manuscript, its usefulness would also be extended to non-Arabists and generalists. For example, collected information on foliation, illumination, illustration and original bindings could be added to the list.
Having sorted out our own collections, the British Library is now in the happy position to develop and set up a database for fully descriptive cataloguing and other codicological projects. Such initiatives and pilot projects are already underway in various institutions world-wide. One is mindful of the projects for computerised cataloguing of Arabic manuscripts in the Dar al-Kutub in Cairo; in the Bibliothèque nationale de France8; in the King Faisal Center in Riyadh; also the Iranian project to produce a world union catalogue of Islamic manuscripts, undertaken by the Nashr-e Hadith-e Ahl al-Bayt Institute in Tehran; as well as the project by ACSAM to catalogue and eventually digitise all Indic and Islamic manuscripts in North American collections. Certainly, there is a lot going on, and no lack of scope for co-operation and co-ordination not only nationally but at the international level.
In short, by way of summary, what has the British Library achieved? And, what next? First, a complete classified handlist of all the Arabic manuscripts in the British Library. Secondly, immediate and ready access to all the collection. Thirdly, a structure on which to set up a database for fully descriptive cataloguing, including codicological and palaeographical aspects. Fourthly, a structure for digitisation of illuminated and illustrated manuscripts. Finally, a model for an on-line national union listing.
- J.D. Pearson, Oriental Manuscripts in Europe and North America, Bibliotheca Asiatica 7 (Switzerland, 1971), p.111.
- This figure is based on information found in G. Roper (ed.), World Survey of Islamic Manuscripts, 4 vols. (London, AI-Furqan Islamic Heritage Foundation, 1992-94).
- A Descriptive list of the Arabic Manuscripts acquired by the trustees of the British Museum since 1894 (London, 1912).
- R. Vassie, A Classified Handlist of Arabic Manuscripts acquired since 1912 (British Library, 1995). Vol.1, Islamic Law, vol.2 Qur'anic Sciences & Hadith.
- Fasc.1: C. A. Storey, Qur'anic literature (London, 1930); Fasc.2: A. J. Arberry, Sufism and ethics (London, 1936); Fasc.3: R. Levy, Fiqh (London, 1937); Fasc.4: R. Levy, Kalam (London, 1940).
- O, Loth, A Catalogue of Arabic Manuscripts in the Library of the India Office (London, 1877).
- G. Roper (ed.), World Survey of Islamic Manuscripts, 4 vols. (London, Al-Furqan Islamic Heritage Foundation, 1992-94).
- See M.-G. Guesdon, 'The Bibliothèque nationale de France and the UNESCO's MEDLIB Project: A Seminar on Arabic Manuscripts Computerised cataloguing' in Manuscripta Orientalia, 5 iii (1999), pp.59-61.