While conducting research last summer at the historic Bodleian Library at Oxford, I requested to look at a medieval Arabic manuscript volume which, according to the online catalogue, contained three texts bound together: first, apparently a treatise by Jalāl al-Dīn al-Suyūṭī (d. 1505), one of the most prolific authors in Islamic history; second, a fragmentary grammatical work of about two dozen pages; and third, the text I was actually interested in, “Lectures on ḥadīth given in Cairo in 1231” by Bahāʾ al-Dīn ibn Shaddād (d. 1234). This is the same Ibn Shaddād who was a close friend and high official of Ṣalāḥ al-Dīn Ayyūbī and would become famous for his Sīra or biography of the sultan—first printed in 1732 at Leiden as Vita et res gestæ Saladini…, an exquisite bilingual Arabic-Latin edition by the Dutch Orientalist Albert Schultens, and available today in several English versions (the often reprinted 1897 edition was a “translation from translation” based on the 1884 French-Arabic one).
But when the friendly librarian procured the manuscript for me, I temporarily forgot all about Ibn Shaddād as I became distracted by some notations on the flyleaf. It was made of clearly different paper than the manuscript itself, suggesting a relatively more recent binding.
“But when the friendly librarian procured the manuscript for me, I temporarily forgot all about Ibn Shaddād as I became distracted by some notations on the flyleaf. It was made of clearly different paper than the manuscript itself, suggesting a relatively more recent binding.”In the middle was inscribed in pencil the library’s shelfmark or call number for the volume: “MS. Marsh 108,” indicating that it was originally part of a collection of seven hundred manuscripts bequeathed to the Bodleian in 1714 by Archbishop Narcisuss Marsh, an Oxford-educated bookman who served a long church career throughout Ireland (Marsh, in turn, had acquired several hundreds of those manuscripts from a 1696 auction in Leiden of the collection of Jacob Golius, another Dutch Orientalist known for compiling a major Arabic-Latin dictionary published in 1653, but who was also officially professor of mathematics and an influential teacher of Descartes).
The Puzzle of Marsh 108
The flyleaf of MS Marsh 108 has a bit of neat Arabic handwriting at the top, which says: al-tanwīr li’l-shaykh jalāl al-dīn, thus labeling the text as “the Tanwīr” of one Jalāl al-Din. Based on the catalogue website, I took this to be a reference to Suyūṭī’s Tanwīr al-ḥalak fī imkān ruʾyat al-nabī wa-al-malak (“Illuminating the Darkness Concerning the Possibility of Visions of the Prophet and the Angels”—a topic of mystical interest causing some debate among Muslims in the fifteenth century, responding to which Suyūṭī cited ḥadīth to argue in defense of such dreams and visions). But right below it on the flyleaf was another inscription, in Latin:
So still an attribution to Jalāl al-Dīn, and Abū ‘l-Faḍl was in fact Suyūṭī’s kunya or “paedonymic,” a typical component of traditional Arabic names. Moreover, he did write a history of Egypt, titled Kitāb Ḥusn al-muḥāḍarah fī akhbār Miṣr wa-al-Qāhirah (“The Book of Good Conversation on Narratives of Cairo and Egypt”). Liber mutilus suggests a “defective volume” where parts of the original text are missing. Sure enough, when I turned the page, it was easy to tell the manuscript was in media res: there was no basmala or any conventional introductory formulae, and the first line was clearly in the middle of a sentence. It is, of course, quite common for old manuscripts to be missing their outermost i.e. first and/or last folios: since books were not always hardbound, normal wear and tear over time often meant the unfortunate loss of the crucial beginning or end of a text.
The question then was which of these two books by Suyūṭī was the one actually in Marsh 108? The above notations were both in black ink, but likely of different origins: the Arabic had thicker strokes, possibly due to a reed pen, while the Latin was written with either a quill or nib. But then someone else had circled around the four lines of Latin using a pencil, with a side note saying simply: “false.” Now this was in a familiar handwriting I recognized from shelfmark notations in pencil on other Bodleian manuscripts. In other words, it must have been one of the more recent curators or librarians at Oxford. And as researchers know very well, nowadays a pen is strictly forbidden in any rare books room or archive worldwide.
The Inheritance of Bibliographies
My curiosity piqued, I decided to look up the library’s print catalogue, guided by one other notation on the flyleaf, in yet another handwriting (a beautiful 18th century cursive) in faded ink, which said: Urii Moh. CXVII. This refers to Uri’s original catalogue in Latin, specifically the section on Codices “Mohammedani,” as opposed to that on Christian Arabic/Syriac and Judaeo-Arabic texts. Titled Bibliothecæ Bodleianæ codicum manuscriptorum orientalium and first published in 1787, this catalogue had been worked on for two decades by John Uri, a Hungarian scholar trained at Leiden, as it happens, by Schultens whom I mentioned above. Uri’s effort to document the Bodleian’s Arabo-Islamic collection was continued by Alexander Nicholl and then by E. B. Pusey. Thus a second volume, listing a further five hundred manuscripts and with corrections to numerous errors in Uri’s catalogue, came out in 1835, once again in Latin. Clearly, Latin was still the lingua franca of Western scholarship well into the nineteenth century. Even more intriguingly, until just five years ago when the remarkable digitized union catalogue of Islamic manuscripts in the UK called Fihrist was launched in 2011, scholars would have to regularly rely on these old Latin catalogues for their research at Oxford.
In fact, perhaps we still do—at least when trying to solve this puzzle of a text’s attribution. A scanned version of Uri’s catalogue is available online, but unable to make sense of the confusing pagination between its different sections, I chose to go for the physical copy standing on a nearby reference bookshelf within the reading room (this entails quite an undertaking in itself given the sheer size and weight of the catalogue, which was printed as a giant folio-size volume measuring a foot wide and a foot-and-a-half in length). I leafed through to entry number CXVII identifying Marsh 108 as containing three texts, of which the first is described as:
“Opus de Traditionum Scientia, initio et fine truncum, folia tamen 110 habens, cui recentior manus hunc præfixit titulum: Illustre Specimen Gelaleddini.”
A footnote then reprints the title inscription from the flyleaf using Arabic typography: “Titulus: التنوير للشيخ جلال الدين”. So here we have one clue: that the Arabic notation was already present in the mid-eighteenth century when John Uri was inspecting the volume, and also that he considered it the “most recent handwriting” (recentior manus) in it. His Latin rendition of the title, Illustre Specimen or “clear indication,” basically translates the word “tanwīr.”
What then to make of the other attribution, claiming the book to be Historia Ægypti? It’s not clear if that Latin bit was there already, or if it postdates Uri. Regardless, I turned to the updated catalogue of 1835, in which the “addenda et emendada” section has a small note about this particular text: “Opus Soiuthii التنوير الحلك في امكام روية النبي والملك, memorat, sed memorat H. Khal.” Thus, following up on their predecessor, Nicholl & Pusey confirmed the text as Suyūṭī’s Tanwīr, but now also providing its full title in Arabic, and for which they cite Ḥājjī Khalīfa. This, of course, is the great Ottoman scholar otherwise known as Kâtip Çelebi (d. 1657), most famous for his own monumental catalogue, a bibliographic encyclopedia titled Kashf al-ẓunūn ʿan asāmī al-kutub wa-al-funūn (“Removal of Doubts on the Names of Books and Specialties”). And so Nicholl and Pusey hoped to remove their doubts on the names of some of the books at the Bodleian! The Kashf is an alphabetically organized listing of some fifteen thousand titles along with details about their authors, and it remains a crucial reference even today for both Muslim scholars and Western academics. And indeed if we look up Kashf al-ẓunūn, we can see for ourselves a “Tanwīr al-ḥalak by Jalāl al-Dīn” listed under the letter tāʾ.
But the problem was, as I soon realized to my further intrigue, the text in the manuscript is actually not Suyūṭī’s Tanwīr after all! We are fortunate in that we do have modern editions of the Tanwīr published in Cairo and Istanbul, against which we can cross-check the text in Marsh 108. But when I tried to do so, I could find no resemblance. There was a more obvious problem even before diving into the text: the print edition of Tanwīr is barely around 75 pages—including footnotes, so the actual text is even shorter. The Kashf describes it as a risāla or short treatise/epistle, and indeed another copy of Tanwīr al-ḥalak in a digitized manuscript from the Egyptian National Library is just 20 folios. The Bodleian copy, on the other hand, is 110 folios long; a folio includes both sides, so that’s actually 220 pages of text!
Moreover, there seemed to be a mistake in genre or category. On turning to the second page of Marsh 108, my eyes were drawn to the word al-faṣl (“section”) written in red ink instead of black, along with a subheading: “First Section: On the spread of the science of ḥadīth, and the principle of its composition and compilation.” Several pages later was another rubricated faṣl, marking a section on “the citation of chains of transmission (asānīd) and texts (mutūn).” These headings suggest a book not about dreams, but rather on the subject of ḥadīth. We must wonder then if John Uri was more attentive than I initially assumed, since in the first Bodleian catalogue he did describe the text as “opus de traditionum scientia,” which seems to translate the term ʿilm al-ḥadīth or “science of traditions,” the field of specialty in traditional Islamic scholarship dealing with the compilation and critical evaluation of Prophetic ḥadīth.
The Google Era of Research
I now resorted to a strategy only possible in the internet era: google, in the verbal sense! Over the past decade or so, a vast number of classical Arabic texts have been digitized across numerous online repositories (the major resource being Shamela). This enables us to search within thousands of books in a manner simply unthinkable in the past. And luckily for me, the script in Marsh 108 is fairly readable, unlike manuscripts that can take much longer to decipher and only after honing one’s paleographic skills. Starting from the first line, I started typing out various random phrases from the text and looking for matches online. After several rounds of failed searches, or finding only matches within books that did not otherwise correspond to the manuscript, I nearly gave up and would’ve concluded that it must be unpublished material. But then, voilà—I found a positive identification. The text, it turns out, is from Ibn al-Athīr’s Jāmiʿ al-uṣūl fī aḥādīth al-rasūl (“Compendium of Sources on the Traditions of the Messenger”), a standard reference encyclopedia of ḥadīth available today in print editions that span more than a dozen volumes. My comparison revealed that the text in Marsh 108(1) corresponds to about 480 pages in vol. 1 of the Damascus edition by the late Albanian scholar Kadri Sokoli (AKA ʿAbd al-Qādir al-Arnāʾūṭ). And now I could see that the manuscript begins at just about a couple of passages into the text, confirming that exactly one folio was lost.
The moral of this story is perhaps that old adage not to judge a book by its cover, not least when it literally has no cover! But it also highlights the conditions—indeed the contingency—of our knowledge, both mundane and academic. The classic Islamic concept of the isnād or chain of transmission was meant to serve as a persistent recognition of the predecessors upon whom one “rests” (the literal meaning of the Arabic root s-n-d).
“The classic Islamic concept of the isnād or chain of transmission was meant to serve as a persistent recognition of the predecessors upon whom one “rests”…”For information on the Bodleian’s Arabic manuscripts, the latest digital catalogue of 2011 rests on the library’s card catalogue used throughout the 20th century, and which in turn had built upon and supplemented the older published catalogues. But while this genealogy helps trace the transmission of a particular error, it’s not so easy to solely attribute the misidentification of MS Marsh 108(1) either to John Uri’s initial findings in the 1780s or to Nicholl & Pusey’s further work in the 1830s. After all, confusions tend to occur naturally in part due to common names and titles in Arabic: I realized much later, for instance, that Suyūṭī himself has another Tanwīr, a commentary on Mālik’s Muwaṭṭaʾ in three volumes, titled Tanwīr al-ḥawālik. Could it be that the ḥadīth-based nature of this other book is what led in the first place to the mistake in that fateful Arabic inscription on the flyleaf?
And lest we leave with another potential confusion, I should point out that the author of the identified text, Jāmīʿ al-uṣūl was Majd al-Dīn ibn al-Athīr (d. 1210) and not his younger brother, that most famous Ibn al-Athīr (d. 1233) with the epithet ʿIzz al-Dīn, author of the classic al-Kāmil fi’l-tārīkh (“The Complete History”). Their youngest brother Ḍiyāʾ al-Dīn (d. 1239) also became well-known as a literary critic and sometime vizier. The three Ibn al-Athīrs were contemporaries of Ibn Shaddād, discussed at the beginning of this piece; in fact, the two younger brothers also served in the armies of Saladin, and the Kāmil remains a major Arabic chronicle of the Crusades. It seems as if we have come full circle on this serendipitous historiographic adventure.