Accueil > English > K > Kitāb: The Coming of the book in the Islamic Classical Culture

Kitāb: The Coming of the book in the Islamic Classical Culture

Toutes les versions de cet article : [عربي] [English] [français]

The Arabic term for “book,” kitāb (a noun related to kataba, “to write”) encompasses every kind of graphic activity from the simplest to the most elaborate. The latter covers the whole gamut from inscription to letter, contract to administrative archive, reminder note to book. As this article aims to elucidate the relationship between the book and Arabo-Islamic culture, the first question to ask is whether the Qurʾān, as a founding text, is the first book in Islam.

 How the Qurʾān became a book

By adopting the form of a codex, the Qurʾān clearly presents itself as the first book ever made by Muslim hands. This happened through a process of selection followed by a materialisation of the revealed Qurʾānic word. Such a dual operation became possible only because the authority of the written Qurʾān superseded that of the oral Qurʾān. Only an acknowledged and respected institution could legitimise the superiority of fixed writing over mere memorisation. Even this did not happen easily. The codification of the Qurʾān exacerbated an already tense political situation: in 35/656 Caliph ʿUthmān, who crystallised all the tensions, was assassinated, and civil war broke out. Despite this conflict-ridden situation, for the first time in their history, Arabs had at their disposal their own book, transcribed in one of their own graphic systems, that of the Northern Arabs, which had appeared three centuries earlier in a Nabataean context under strong Hellenistic influence. This Hellenistic trait has remained a lasting element of Islamic culture as a culture of the book.

The codex appeared in the first century CE in Hellenised Christian communities and was in common use by around the fourth century. Compared with the volumen, it represented a major technical development in the history of writing and of written production, first, because it allowed a much larger series of texts to be gathered together than was possible on a roll, and second, because it made it possible for a text as long as the Bible to be produced as a single book. By combining all the rolls of a work, the codex also helped identify the order in which they were to be kept. In the case of the Bible, this led to a clearer delimitation of the canon. For all these reasons, the first compilers of the Qurʾān chose the codex.

Establishing the Qurʾānic corpus meant determining what was part of it and what was not, cutting it into textual units called sūras composed of sub-units called āyāt (sing., āya), usually translated as “verses” by reference to the Bible. The āyāt were gradually distinguished by graphic symbols. To validate such a textual strategy, the Qurʾān compilers had to discuss the order in which to display the different units and sub-units. The scribes who worked under them had to intervene throughout the process of graphic composition before copying the established text into manuscripts ready to be used. The oldest Qurʾānic manuscripts—going back to the last third of the first century of the Hegira, according to the findings of codicology—reveal ongoing teamwork. The act of transforming the Qurʾānic revelation into a text naturally modified the conditions of its reception, as well as the processes of codification. The system of transcription employed remained defective, inasmuch as it reproduced only the consonantal skeleton (ductus). A legitimate question is whether this points to a deficiency in writing technique or, rather, to a concession made to eminent, respected memorisers of the Qurʾān, in order to allow them to preserve some of their authority. As early as the reign of Caliph ʿUmar (13–23/634–44), the administrative officers in charge of turning the Qurʾān into a text introduced a pointing system that greatly improved its clarity. From this early stage in the history of the book in Islam, writing has exerted its power over the Qurʾānic text, foreclosing it upstream by taking part in its very production and downstream by stabilising its transmission.

As a text, the Qurʾān could now be read as often as desired and in different modes, aloud, half-voiced, or silently. The conditions required to learn it by heart were also deeply affected, because there was now a graphic support readily available to check that the memorised information was correct. With the Qurʾān laying the foundations of a graphic culture, writing enabled a greatly increased power of storage in memorisation, thanks to an informed use of the eye. Before such a type of learned auditory and visual memory could develop, however, a remarkable shift in paradigm was needed to introduce graphic reasoning into the culture of the Hijaz.

When handling writing as a tool of reasoning, the early compilers of the Qurʾān went through a series of cognitive operations that are common to all learned practices. Collecting, sorting, classifying, collating, and computing allowed them to transform odd sayings circulating orally or in writing into an articulated whole in the form of a corpus. In doing so, writing also brought out the divergent views held by various circles of Qurʾān scholars. Differences or shifts, albeit distortions, which had so far remained implicit and of little importance, were now glaringly exposed in the manuscripts. Thus fixed, the different variants become textual traditions. They were sometimes called riwāyāt (sing., riwāya) and sometimes qirāʾāt (sing., qirāʾa), depending on whether oral communication or written transmission was stressed. They owe their conservation to the existence of stable Qurʾānic communities which ensured that they would be transmitted as an inheritance received from founding masters. In these conditions they are called riwāyat fulān or qirāʾat fulān, meaning the transmission or reading of such and such a person. This attribution, which was considered to be the result of an authorial intention, was termed ikhtiyār (choice), a lesson in reading by the religious scholars of the second-third/eighth-ninth centuries, during the ʿAbbāsid period, and taʾlīf (composition) was its formalization. The latter term recalls to mind the Latin compositio. Taʾlīf is synonymous with the activity of properly composing a book, which authors and compilers had been engaged in since Greco-Roman antiquity.

The Qurʾānic copyists’ action itself belongs to the learned practices inherited from Greco-Roman antiquity. A traditional source allows us a concrete understanding of the matter, telling us how, when ʿAlqama (d. 61/680–1), one of the earliest Qurʾānic experts in Kufa, began establishing his own Qurʾānic codex, he asked his disciples to follow his example by giving them at each stage of his proceedings “one or two note-books, one or two leaves” (al-kurrāsa wa-l-kurrāsatayn wa-l-waraqa wa-l-waraqatayn; Ibn Abī Dāwūd al-Sijistānī, Kitāb al-maṣāḥif, ed. Arthur Jeffery, in Materials for the history of the text of the Qurʾān. The old codices, Cairo 1936, Leiden 1937, reissued ed. Muḥibb al-Dīn ʿAbd al-Sajjān, Beirut 1995, 59).

Making an exercise-book out of leaves written on both sides is a technique borrowed from the early period when the codex replaced the volumen. The same third/ninth-century traditional source explains how the early Arabo-Muslim copyists learnt that technique, telling how ʿAbd al-Raḥmān b. Abī Laylā (d. 82/701?), the judge of Kufa at that time, who wished to obtain a Qurʾānic manuscript of high quality, commissioned a professional Christian copyist from the neighbouring town of Ḥīra, the seat of a Nestorian bishopric and a Syriac-speaking Hellenistic intellectual centre.

The results of such cultural interpenetration in Bilād al-Shām (Syria) are obvious in the Qurʾānic and non-Qurʾānic inscriptions written on the walls of the mosque of the Dome of the Rock twenty years later by Christian professionals. The sumptuous fragments of manuscripts discovered in the attics of the mosque in Sanaʿa in the 1970s show, through their wealth of decoration and fine execution, that they portrayed more than religious expectations. As early as the caliphate of ʿAbd al-Malik (r. 65–86/685–705), such lavish manuscripts were designed to satisfy the hedonism of rich and powerful patrons intent on obtaining artifacts worthy of admiration.

  The cognitive function of writing

At the end of the first century of the Hegira and the beginning of the second, working as a copyist-bookseller (warrāq) became a respectable profession, acknowledged as such by the Muslim code of professions. Scholars also felt that each had to establish his own Qurʾānic manuscript according to philological necessities. This gave rise to several notions, including that of ṣiḥḥa (authenticity), which was applied to some favoured Qurʾānic manuscripts. For instance, a chronicle from the third/ninth century tells that Aʾmash (d.147 or 148/764–5) used to teach the Qurʾān from a manuscript. His listeners had their own copies, which they were invited to emend according to his reading. Next to him stood the disciple who had the second “most exact” (aṣaḥḥ) Qurʾānic codex, after the master’s, which was held to be the archetypal manuscript. Just like the Greek diorthōsis and the Latin emendatio, the Arabic taṣḥīḥ is a medical metaphor that considers any textual corruption—graphic, semantic, grammatical, or other—an abnormality and its correction a therapeutic operation aiming at redressing it.

Sources occasionally use the term iṣlāḥ with the same meaning. In the lexical field of this word are the verbs ṣalaḥa (to be in good order, well-kept without defect; to be honest, righteous, virtuous), and ṣallaḥa (arranging, adjusting putting something right; correcting, rectifying; acting well, behaving as a virtuous person). As ṣalāḥ (or ṣalḥ) designates what is “in good order, exempt from defect,” it is the opposite of fasād, which is the “state of what is corrupt, spoilt.”

This conception of philological work as a reflection of medicine and ethics is the result of the transfer of the intellectual and material technology of the book from the Greco-Roman to the Islamic world. It illustrates particularly well the filiation between the new learned Arabo-Islamic culture and the other learned cultures of the Middle East and Mediterranean.

ʿAlqama’s method of transmission, called ʿarḍ (lit., collation of a text with the original copy, usually in the presence of the teacher by means of recitation or reading), which was applied to the Qurʾān during his generation, later became the means of communication of books par excellence in Islamic study circles. Like the application of writing to the Qurʾān, it aroused opposition. At the end of the first century of the Hegira and the beginning of the second, not every Muslim scholar was convinced that the graphic orientation being given to high culture, especially religious, was appropriate.

Some specialists have given too much importance to this refusal, labelled “graphophobia,” and have concluded that the Umayyad court, not learned circles, was responsible for the appearance of the first books of Islam, which took the form of the literary epistle (risāla), as distinct from the scientific epistle devoid of any literary attribute.

The scientific epistle had been in use amongst scholars in the Hijaz and Iraq since the end of the first century of the Hegira but is deemed non-literary by those same experts on the grounds that it was used for private communication and not, like the book, for public communication. If this were the case it would mean that what determines the status of writing is not so much its mode of production as its mode of circulation. Such an affirmation would be absurd, because it is obvious that not all writings for public consumption are of the nature of a book. Just as ridiculous is the idea that the so-called scientific epistle belongs to the private realm. This would entail making scientific activity a private task or at least imply that it can develop in that way. If this were the case, no specialized institutions or learned communities would be needed in order to ensure that it was promoted and developed over time and could enjoy the stability without which no accumulation of information and development of science is possible.

The self-evident social character of science cannot but affect its written productions, including the so-called scientific epistles. Whether analyzed from the point of view of their aims, their language with its technical element, or their argumentative devices, it is obvious that such writings have little in common with the letters written by individuals to individuals in a literate society.

It should be evident by now that the proper method for determining the status of a literary writing is not its circulation but its production. Some literary works can be transmitted orally without losing their status as literature. On the other hand, some written specimens can be transmitted in written form without achieving literary status.

  Literary composition and editorial traditions

Some of the first so-called learned epistles constitute genuine treatises: historical, as in the case of ʿUrwa (d. 93/711–2 or 94/712–3); theological, in the case of Ḥasan b. Muḥammad b. Ḥanafiyya (d. between 86/705 and 101/720); and religious and political, as with Ḥasan al-Baṣrī (d. 110/728).

The writing techniques used in the scientific epistles mean that they cannot be placed in the category of private letters, which have neither scientific pretensions nor didactic aims. Moreover, they circulated widely in the relevant circles and sometimes beyond them. From the very beginning of the incipit of his Kitāb al-irjāʾ, Ḥasan b. Muḥammad b. al-Ḥanafiyya calls for such publicity (Josef van Ess, Das Kitāb al-Irğāʾ des Ḥasan b. Muḥammad b. al-Ḥanafiyya, Arabica 21, 1974, 56–82, and Arabica 22, 1975, 218–70). Why should the writer of a private letter request the individual or collective recipient of his text to ensure its widest possible communication? If that were done, it would stand out as a special and unusual request, and certainly atypical. As a literary device, the epistle does not belong here. It was used by the first Muslim scholars to express their opinions and ensure they had a verbatim diffusion. Political leaders understood the public impact of such writings and persecuted many authors, throwing some into jail and putting others to death. Later on, when Islamic scholars became familiar with other models of literary composition, they did not forget the epistle. On the contrary, they continued using it throughout the Middle Ages in order in the production of religious and lay works. Was this due to its status of proof? From Judge ʿĀmir al-Shaʿbī (d. after 103/721) to the Qurʾānic exegete Qatāda (d. 117/735), including theologian Ḥasan al-Baṣrī, all Iraqi legal experts from the end of the first century of the Hegira and the beginning of the second considered that, out of all the written judiciary evidence, the epistle alone was valid on its own, precisely because of its stereotyped formula.

In that period, several types of writing called ṣuḥuf (sing., ṣaḥīfa) were circulating in learned circles. Some masters owned a great many of them, which constitute the earliest form of learned archives stemming from Islamic culture. Their owners were conscious of their great value and cared enough to protect them with a binding that made them into a codex, which in turn rendered them all the more usable. A scholar from Homs, Buḥayr al-Sahūlī, was present at the siege of Constantinople. Before dying he bequeathed his manuscript to one of his pupils who must, in turn, have done the same before his death. Like this Syrian scholar, many masters of the period took care to transmit their writings to their favourite disciples. The Iraqi lawyer living in Syria, Abū Qilāba (d. 105/723?) bequeathed his to Ayyūb al-Sikhtiyānī (d. 131/748–9), his pupil from Basra.

Since the end of the nineteenth century, these ṣuḥuf have given rise to much debate concerning their place in the graphic process that led to the appearance of the book. They reveal an early stage in the use of writing for the purpose of data storage, as well as the elaboration of a teaching system as early as the end of the first century of the Hegira. They also acted as a precious mnemonic support in the hands of teachers and their students. Some, like the Ṣaḥīfa of the Prophet’s companion Samūra b. Jandal (d. 60/679–80), were endowed with a particular aura. They circulated in study circles and were continuously copied over several generations. This has led some specialists to consider them the starting point from which Islamic culture developed from rudimentary forms of writing to the production of more complex types, the highest of which was the book in proper book form, the muʾallaf.
Gregor Schoeler is correct to criticize this evolutionary scheme. He then substitutes another type of analysis in which the contrasting pair ṣaḥīfa/muʾallaf is approached by reference to the Greek literary tradition of the hypomnēmata and syngrammata. His model places the muṣannaf—so called because of its method of exposition, which is based on the systematic classifying (taṣnīf) of the exposed material—half way between the ṣaḥīfa and the muʾallaf. The reason for drawing this distinction between the muṣannaf and the muʾallaf is that the former was published orally, and was therefore destined for a restricted audience of scribes. The latter was, by contrast, intended for written diffusion, thereby ensuring a wider readership.

Although unusual, this typology amounts to a representation of a traditional conception of writing considered from the perspective of its pragmatic function (bringing near what is far and making the absent present), and its mnemonic function (writing in order not to forget). The typology leaves aside, however, writing’s cognitive function, the very one that makes graphic reasoning possible. It is also anachronistic to reduce “true literature” solely to “well-written texts intended for written publication” (Schoeler, Écrire et transmettre, 157). Numerous studies have shown that such an editorial situation prevailed only after the printing revolution. Before that time, great writing cultures, whether Greek, Roman, Indian, or Islamic, combined aural and written resources in their model of publication.

The equating of ṣuḥuf with hypomnēmata entails reducing the Greek notion to its lowest expression. Among the scholars of antiquity, hypomnēmata did not designate only aides-memoire but also lecture résumés, rough versions of books, unfinished and unpublished works. The term was thus applied to some categories of the works of Aristotle, Galen, and Plotinus. The ṣuḥuf belong to a pre-literary stage of Islamic culture, so it is impossible to consider them the starting point for literary culture, whereas hypomnēmata are an integral part of it. Trying to work out how ṣuḥuf became muʾallafāt merely leads into a cul-de-sac, because there is no possible bridge between them.

To understand what leads a community or society familiar with the handling of writing into a true culture of writing, especially in book form, we must look at some cognitive operations that have been studied by sociologists of science as modalities of graphic inscription. Several cognitive activities fall under the heading of graphic inscription. It would be impossible, or at least difficult, for them to take place without graphic tools such as lists, made in alphabetical or thematic order, measurements established according to determined metrological norms, demographic or fiscal censuses, posologies for medical or pharmaceutical uses, plus all the experimental or cartographical processes that require mathematical calculation and technical instruments.

Such intellectual operations are the very basis of graphic culture. They are employed in administrative and scientific domains. They were born in Mesopotamia, with the advent of the state and the invention of writing. They reached their greatest sophistication in the Hellenistic world, which remained the great provider of models of literary composition to the learned Mediterranean and Near Eastern cultures, whether in the form of dictionary, anthology, résumé, monograph, or compendium.

Insofar as the expansion of Islam did not end the existence of scholarly communities in the late-antique Near East, the intellectual and literary traditions of the latter remained alive well into the ʿAbbāsid period and sometimes even beyond. Islam turned to these traditions in order to establish its own scholarly culture on graphic bases, multiplying its interaction with the cultures that gave birth to its own, to the point of disputing their hegemony.

Books produced within that framework may have been published orally or in writing. This in itself neither disqualifies nor qualifies their status as literature, especially when the very same works might be edited in either manner.



La tradition manuscrite
Déroche (F.) et al., Manuel de codicologie des manuscrits en écriture arabe, Paris, 2000.
Dutton (Y.), ed., The Codicology of Islamic Manuscripts, London, 1995.
Sayyid (A.F.), al-Kitāb al-ʿArabī al-makhṭūṭ, 2 vols., Le Caire, 1997.
Hoffmann (P.), éd., Recherches de codicologie comparée. La composition du codex au Moyen-Âge en Orient et en Occident, Paris, 1998.
Gacek (A.), The Arabic Manuscript Tradition. A glossary of technical terms and bibliography, Leiden, 2001.
Humbert (G.), éd., La tradition manuscrite en écriture arabe, in Revue des mondes musulmans et de la Méditerranée 100-101, 2002.
Wouters (A.), « From papyrus roll to papyrus codex. Some technical aspects of the ancient book fabrication », Manuscripts of the Middle East 5, 1990-1, p. 9-19.
Histoire de l’écrit (à distinguer de l’histoire de l’écriture)
Abbott (A.), The Rise of the North Arabic Script and its Kurʾānic development (Chicago 1939), spéc. chap. 3.
–, Studies in Arabic Literary Papyri, vol. 2, Qurʾānic Commentaries and Traditions, Chicago, 1967.
Cook (M.), « The opponents of the writing of tradition in early Islam », in Arabica 44, 1997, p. 437-530, repris par l’auteur in Studies in the origins of early Islamic culture and tradition (Aldershot 2004), chap. 10.
Heck (P.L.), « The epistemological problems of writing in Islamic civilization: al-Ḫaṭīb al-Baġdādī’s (d. 463/1071) Taqyīd al-ʿilm », in Studia Islamica 94, 2002, p. 85-114.
Humbert (G.), « Le Kitāb de Sībawayhī et l’autonomie de l’écrit », in Arabica 44, 1997, p. 553-67.
Graham (W.A.), Beyond the Written Word. Oral Aspects of Scripture in the History of Religion, Cambridge UP, 1987, voir en particulier le chap. 3, « An Arabic reciting. Qurʾān as spoken book ».
Kennedy (H.), « From oral tradition to written record in Arabic genealogy », in Arabica 44, 1997, p. 531-44.
Madigan (D.A.), The Qurʾân’s Self-Image. Writing and Authority in Islam’s Scripture, Princeton UP, 2001.
Schoeler (G.), The Oral and the Written in Early Islam, trad. de l’allemand par Uwe Vagelpohl, éd. James E. Montgomery, London/New York (contient six articles publiés dans des revues spécialisées).
Histoire de la transmission textuelle
Abbott (N.), Studies in Arabic literary papyri, vol. 1, Historical texts, Chicago 1957; Berkey (J.), The transmission of knowledge in medieval Cairo. A social history of Islamic education, Princeton UP, 1992.
Humbert (G.), Les voies de la transmission du Kitāb de Sībawayhi, Leiden 1990.
Khoury (G.R.), « Pour une nouvelle compréhension de la transmission des textes dans les trois premiers siècles islamiques », in Arabica 34, 2, 1987, p. 181-96.
Leder (S.), « Authorship and transmission in unauthored literature », in Oriens 31, 1988, p. 67-81.
Makdisi (G.), The rise of colleges. Institutions of learning in Islam and the West, Edinburgh 1981.
–, The rise of humanism in classical Islam and the Christian West, Edinburgh, 1990.
Nasr (S.H.), « Oral transmission and the book in Islamic education. The spoken and the written word », in JIS 3, 1, 1992, p. 1-14.
Osman (G.), « Oral vs. written transmission. The case of Tabarī and Ibn Saʿd », in Arabica 48, 1, 2001, p. 66-80.
Schoeler (G.), The Oral and the Written in Early Islam, London and New York 2006.
–, « Weiteres zur Frage der schriftlichen oder mündlichen Überlieferung der Wissenschaften im Islam », in Der Islam 66, 1989, p. 38-67.
Rudolf Sellheim, « Abū ʿAlī al-Qālī. Zum Problem mündlicher und schriftlicher Überlieferung am Beispiel von Sprichwörtersammlungen », in Hans R. Roemer and Albrecht Noth (ed.), Studien zur Geschichte und Kultur des Vorderen Orients, Leiden, 1981, p. 362-374.
Vajda (G.), La transmission du savoir en islam. VIIe-XVIIIe siècles, London 1983.
Histoire du livre
Déroche (F.), Le livre manuscrit arabe. Préludes à une histoire, Paris 2004.
Pedersen (P.), Den Arabiske bog, Copenhagen, 1946, trad. du danois par G. French, ed. R. Hillenbrand, The Arabic book, Princeton UP, 1984.
Richard (F.), Le livre persan, Paris 2003; Gregor Schoeler, Écrire et transmettre dans les débuts de l’Islam, Paris, 2002.
Touati (H.), L’armoire à sagesse. Bibliothèques et collections en Islam, Paris 2003.
Histoire de la lecture
L’histoire de la lecture en contexte islamique reste à constituer :
Hitzel (F.), éd., Livres et lecture dans le monde ottoman, Aix-en-Provence, 1999 (ce volume collectif porte plus sur l’histoire du livre et l’histoire des bibliothèques que sur l’histoire de la lecture stricto sensu).
Touati (H.), « Pour une histoire de la lecture au Moyen Âge musulman. À propos des livres d’histoire », in Studia Islamica 104, 2007, p. 11-44.

Voir en ligne : Kitâb : L’essor du livre dans la culture islamique classique

How to cite:
Houari Touati, «Kitāb: the coming of the book in the Islamic classical culture », in Houari Touati (ed.), Encyclopedia of Mediterranean Humanism, Spring 2014, URL =āb