A codex is produced by binding together a number of quires (pl. karārīs, sg kurrās). Quires are small stacks of folded sheets of paper or parchment, and each individual leaf of paper/parchment in a quire stack is called a bifolium (pl. bifolia). In turn, each bifolium is divisible into two halves, each called a folium. The side of the folium on which the scribe begins writing is known as a 'recto' (or rectum folium), and the other side of the page is called a 'verso' (or versum folium).
In the Arabic-Islamic tradition, the most commonly-encountered quire is what is known as a 'quinion': i.e. a stack that consists of five bifolia. A scribe would write on each bifolium when it was loose, and when he had finished copying the book, he would bring the whole stack of bifolia to a bookbinder (called mujallid in the Mashriq, musaffir in the Maghrib), or a bookseller (warrāq) for binding the bifolia into a codex. Given that all the bifolia remained loose until the entire book was bound, individual bifolia could easily become disordered, or, for example, if the copyist tripped on his way to the bookbinder, the whole book could easily become disordered...
To prevent folia and quires from getting mixed up, quires were numbered (individual folia were numbered only since the 10th/16th century). These “quire signatures” are most often found in the upper margin of the recto side of the first folium of a quire. Until the 5th/12th century, they were noted in letters with numerical values, whereas most manuscripts from the second half of the 5th/12th century onwards the quire signatures were usually spelled out. Authors and booksellers sometimes mentioned the total number of quires on the fly-leaf or first folio of a codex, to give an indication of its size.
In order to ensure that bifolia were inserted in the right order, scribes also wrote catchwords (taʿqībāt) on every verso side of a folio. The practice of adding catchwords may have begun as early as the 3rd/9th century. It became common practice by the 7th/13th century and became such an integral part of the book-making tradition that it was long preserved even in printed books with sewn (or glued) and numbered pages.
In Sudanic Muslim Africa, manuscript folios were never sewn together. Valuable texts were kept between wooden or carton covers in a leather envelope which was tied with a strap of goat leather.
Book-case made in Chad, late 20th century.
About this manuscript
Title: Al-Musnad al-jāmiʿ (also Sunan al-Dārimī)
Classmark: Or. 364
Subject: Hadiths
Author: (Compiler) ʿAbd Allāh b. ʿAbd al-Raḥmān al-Dārimī (181/797-255/869)
Copyist: ʿAbd Allāh b. Muḥammad Yūsuf
Format: Codex
Extent: 285 folios, leaf height260 mm, width 175 mm
Date of creation: 634 AH, 1236/37 CE.
Language(s): Arabic
Script: Naskh
Al-Dārimī's compendium of sayings of the prophet Muhammad (hadiths) is a collection of some 3,500 hadith arranged by subject matter. Although it is not considered one of the six most reliable hadith collections by Sunni scholars, the book is considered to narrate very authentic hadiths and is often cited in specialist hadith scholarship.
The copy in this lesson is almost eight hundred years old. According to the colophon on f 283v the copy was made in the Mustanṣiriyya Madrasa (in Baghdad) in the month of Sha'ban, 634 Hijri, i.e. April 1237 CE. It is so well preserved that its age hardly shows. The codex came to Leiden in 1668, as part of the legacy of the diplomatic representative for the Dutch Republic Levinus Warner, who had purchased it in Istanbul (see lesson 21 for more on Warner). The cover is not original.
Most quires of this codex consist of five bifolios (i.e. quinions). Even without investigating their binding, this can be deducted from the quire signatures, which are still visible in the upper left quires on f10r (thāni, second), f78r (ʿāshar, tenth), f 88r (ḥādī ʿashar, eleventh) for instance. Note this manuscript does not contain catchwords in the lower margins: catchwords were only beginning to become commonplace around the time of this manuscript's copying.
Questions
Somewhere in the process of making this codex, two bifolia were discarded. New quires then started on folia with numbers such as 78, 88, 99 etc. Note the quire number on f78r, عاشر. The quire number on f 68r, in the same place, is difficult to read in this image, but knowing what to expect, you will be able to decipher it. Then go to folium 58r. How do you explain the absence of a quire signature altogether there? For a clue, see f 198r.
The quire number may have been written quite close to the edge of the folium and been lost when the folia were trimmed. After years of use, the edges of folia would fray and the owner, perhaps a new owner, would ask a warrāq or mujallid to straighten the bookblock by trimming the folia. Marginal notes, catchwords and quire-numbers did not always survive the operation intact.
In the digital images, it is impossible to see where the seventh and eight quire do start. Somewhere between the sixth quire that started with f50r and the ninth, starting with f78r, a number of bifolios were lost or discarded because of mistakes.
