Acerca del autor

Brinkley Messick is Assistant Professor of Anthropology at the University of Michigan.

De la contraportada

De la solapa interior

Fragmento. © Reproducción autorizada. Todos los derechos reservados.

The Calligraphic State: Textual Domination and History in a Muslim Society

University of California Press

The theoretical descent of Roman jurisprudence from a Code, the theoretical ascription of English law to an immemorial unwritten tradition, were the chief reasons why the development of their system differed from the development of ours. Neither theory corresponded exactly with the facts, but each produced consequences of the utmost importance.
Sir Henry Maine

Viewed in detail, the development of the shari'a across the Muslim lands was a phenomenon involving specific men and specific texts. We know something about the particulars of this history because of the existence of accounts arranged in the form of biographical entries. Created by early Muslim historians, these biographical works played a crucial role in tracking the transmission of Islamic knowledge across regions and through successive generations. One such work, composed in twelfth-century Yemen, covers the early generations of Yemeni shari'a scholars. It also traces the arrival in the highlands of authoritative shari'a texts of the era, describing the specific teacher-to-student connections by which they were introduced and then diffused. A local node in one reported line of transmission was a man by the name of Faqih al-Nahi, until his death in 1171 a resident of the town of Ibb.1 The text in question was a famous one, authored by an eminent shari'a jurist of the preceding century named al-Shirazi who lived and taught in Baghdad, an international center of Islamic learning.2

"Between me and the author are two men," al-Nahi of Ibb used to say to his colleagues. In this shorthand manner al-Nahi expressed the particular legitimacy of his stature as a scholar and teacher. It was a remark worthy of citation in an intellectual world in which the texts of knowledge were literally embodied, their conveyance reckoned in terms of known relayers. Authority of this sort relied upon the specification of human links between intellectual generations. Al-Nahi could explain that he had learned the text from his teacher, who had received it from an individual who had acquired it directly from al-Shirazi, in the author's Baghdad lesson circle. Then, as now, Yemeni scholars were

travelers, circulating in search of teachers and knowledge not only throughout their native highlands but also northward to Mecca and beyond to Cairo, Damascus, and Baghdad. Completing his studies with al-Shirazi, the traveling scholar in question returned to reside in Yemen, where he taught the text to his own students, including a man who would later teach it to a third generation, which included al-Nahi of Ibb. Each of the two intervening scholars identified in al-Nahi's personal line of reception is the subject of a separate entry in the same biographical history.

Authoritative texts are as fundamental to the history of shari'a scholarship as they are to the history of the other intellectual disciplines. Such a text was "relied upon" in a place and time:3 the knowledgeable consulted it, specialists based findings upon it, scholars elaborated its points in commentaries, teachers clarified its subtleties, students committed its passages to heart. Authority in a text depended on a combination of attributes both ascribed and achieved: there were the built-in features of textual ancestry and authorship as well as an acquired reputation and record of dissemination. The fates of such texts were diverse, ranging from an enduring general prominence or more limited respect among the cognoscenti to a purely ephemeral authority and the all-but-forgotten status of the superseded. Since this authority would change radically in the course of the late nineteenth and the early twentieth century, its modalities must be examined in some detail.

A primary sort of textual authority was derived:4 it followed from the existence and importance of an original, genuine, and ultimately reliable text, which refers back to the position and identity of the Quran. "In the genealogy of texts," Edward Said writes (1983:46), "there is a first text, a sacred prototype, a scripture, which readers always approach through the text before them." A genealogy of authoritative texts in Islam must begin with a consideration of the Quran as the authoritative original. The paradigmatic, Urtext qualities of the Quran concern both content and textual form. Substantively, the Quran and the Sunna, the practice of the Prophet, constituted the two fundamental "sources" (usul , sing. asl ) for the elaboration of shari'a jurisprudence. Discursively, the Quran represents both the end and the beginning of the kitab (text, scripture, writing, book). Just as Muhammad was the last, the "seal," of the Prophets, and also the first Muslim, the Quran was the definitive and final kitab , whose particular authority would initiate and delimit a discursive tradition.

A central problem in Muslim thought concerns the difficult transi-

tion from the unity and authenticity of the Text of God to the multiplicity and inherently disputed quality of the texts of men. A concurrent underlying tension was generated in shari'a scholarship where an unresolvable gulf opened between divinely constituted truth and humanly constituted versions of that truth. Purists of all eras, including many contemporary "fundamentalists," have made a distinction between the divine shari'a, defined as God's comprehensive and perfect design for His community, and a humanly produced shari'a, or, more precisely, the corpus of knowledge known as fiqh (usually translated "jurisprudence"), a necessarily flawed attempt to understand and implement that design. In this gap between divine plan and human understanding lay the perennially fertile space of critique, the locus of an entire politics articulated in the idiom of the shari'a.

If the transition from the divine plan to its human versions was difficult, it was also necessary, for the truth of Revelation could only be implemented through the medium of human understanding. With the death of the Prophet Muhammad in 632, the Muslim community found itself cut off both from further Quranic Revelations and from the Prophet's own practice, that is, from further elaborations of the two "sources" of secure knowledge. Thereafter, the community confronted the problems of developing a more detailed corpus of rules and procedures while continually adjusting to new social realities.

The growth of the fiqh as a body of knowledge, resulting from the work of jurists in the early centuries of Islam, brought with it the inevitable disagreements associated with purely human creativity. Authoritative fiqh works endeavored to further define the already definitive. Their derivitiveness was both necessary to their nature and the crux of their problematic: they were texts in the world of the Text. Participating with self-assurance in an authentic tradition, such texts had their recognized antecedent sources and their means of establishing legitimacy. At the same time, they pertained to history rather than to the eternal; to societies of rifts and hierarchies rather than to the ideal communitas of the umma ; to difference rather than to certainty.

Condensed and practical fiqh manuals, known as the mutun (sing. matn , lit. "text"), were a distinct category of authoritative text. At first glance, such manuals seem to have little to recommend their consideration. Since they contain neither the elaborations of material nor the refinements of method that characterize the field as a whole, the mutun might be thought the least impressive works of the jurisprudence literature, their bare-bones presentations being notable mainly for feats of

compression. But these handbooks took on a significance that went far beyond their modest literary qualities, and they eventually became the most widely disseminated of all shari'a texts. As instructional standards, they were influential gateways to the central academic discipline, representing for many the sum total of advanced (that is, post-Quranic-school) instruction. Particular manuals came to be symbols of the separate "schools" of shari'a thought, which in the Sunni tradition were one and the same as its principal subdivisions. Containing simple but authoritative statements of position on the range of substantive issues, manuals were studied by all educated adherents of a school at the outset of advanced instruction, and they were later referred to by scholars in their various interpretive activities. In addition, these compact works were among the typical source texts upon which the expansive commentary literatures of the fiqh were based.

Manuals appeared on the intellectual scene at a relatively late date, however. These were not the earliest works, handed down from the formative centuries of Islam, but were instead the products of a more mature, developed, and institutionally established later thought.5 In contrast to the discourse-setting sort of originality of the foundational fiqh works, by jurists such as Malik, Abu Hanifa, al-Shafi'i, and Ibn Hanbal, for whom the four standard "schools" are named, the manuals were determinedly derivative. They offered creative synopses, in which a body of doctrinal views pertaining to a particular school was sifted, selected, and summarized. While most of the earlier classics were in the category of exhaustive (literally "long") works, manuals were examples of an "abridgment" genre that became common across the intellectual disciplines.

Manuals were similar to the long jurisprudence works in overall topic coverage but differed radically in depth of presentation. They included sections on ritual matters, such as ablutions, prayer, fasting, tithe and alms-giving, and the pilgrimage; the numerous types of contracts, transactions, and dispositions covering agrarian production, commerce, and the family; rules concerning evidence, court procedure, and punishments; and a variety of miscellaneous matters, such as hunting and dietary rules. Principles connected with the leadership of the Muslim community were covered in sections concerning the conduct of war, certain types of crimes, taxation, and the administration of justice.

The great Ibn Khaldun, who died in 1406, viewed with disapproval the routine use of such abridgments in his era.6 They have a "corrupting influence on the process of instruction," he wrote, because they

"confuse the beginner by presenting the final results of a discipline to him before he is prepared." Ibn Khaldun's general pedagogic concern was with the acquisition of a properly developed intellectual habitus .7 While he believed that the "crowded" meanings of the abridgments were an obstacle to this acquisition, he deemed contact with the "repetition and lengthiness" of the long works "useful." Regardless of Ibn Khaldun's opposition, however, some manuals went on to become classics in their own right, remaining basic to the intellectual landscape for centuries, until the moment of modern changes in the nature of schooling, knowledge, and the law.

Shafi'i Texts

One of four standard schools of Sunni legal interpretation, the Shafi'is take their name from the early jurist Muhammad b. Idris al-Shafi'i (died 820).8 The word madhhab , conventionally translated as "school," literally means a "path," and in technical scholarly contexts the reference is to those jurists who claim intellectual descent from an eponymous jurist, in this case al-Shafi'i. In a broader, nontechnical sense, madhhab identities were a means of expressing geopolitical loyalties. In Yemen the Shafi'i school is associated with Lower Yemen, where Ibb is located, and with regions further South and East—the territory, until the unification of 1990, of the People's Democratic Republic of Yemen (South Yemen).

Prior to the tenth century, the Maliki and Hanafi schools, both developed earlier than the Shafi'i, represented the main currents of Yemeni legal thought. Although al-Shafi'i held a brief appointment (and is said to have engaged in political intrigue) in Yemen, the main centers of his teaching and later influence were Cairo and Baghdad. Ibn Samura, the historian who reported on the biography of Faqih al-Nahi, documented the advent and initial spread of Shafi'i thought in the highlands. In succeeding centuries, al-Shafi'i's intellectual descendants rose to predominance in instruction and judgeships throughout the central Middle East. With the sixteenth-century spread of the Ottoman Empire, however, the Shafi'i school was officially (but not always popularly) supplanted in most of these areas by the Hanafi school, which was favored by the Ottomans. Nevertheless, the Shafi'i school today probably has more adherents than any other in the Muslim world, its main centers other than southern Arabia being Egypt, East Africa, and populous Southeast Asia.

For many centuries now, Ibb has supported an active community of Shafi'i scholars. Despite their seemingly remote mountain valley location, town jurists were far from parochial. In terms of texts studied, theirs was not an unconventional local version of the shari'a. Beginning with Ibn Samura and continuing to the present century, biographical histories provide views of the changing scholarly community in Ibb. A recently published work (Zabara 1979) devoted to noted individuals of the just-completed (fourteenth) Hegira century contains an entry on a distinguished Ibb scholar and prominent political figure who lived from 1876 to 1922, some seven and a half centuries after Faqih al-Nahi. Like al-Nahi, 'Abd al-Rahman al-Haddad was an adherent of the Shafi'i school of shari'a jurisprudence. Both men were connected to the school through their relations with particular teachers and specific texts. In al-Haddad's case the teacher was his father, and the key text was a celebrated old manual known as Al-Minhaj . His biographical entry opens as follows:

The learned scholar and man of letters, the bright and sagacious 'Abd al-Rahman, son of 'Ali, son of Naji, al-Haddad, the Shafi'i, the Yemeni, the Ibbi, was born in the town of Ibb in the year 1293 [1876] and received instruction from his father, in Shafi'i jurisprudence [beginning with] Al-Minhaj .9

Ibb scholars such as 'Abd al-Rahman al-Haddad and, in the next generation, men such as his nephew and son-in-law, Ahmad b. Muhammad al-Haddad, whom I knew in the 1970s as an old man and a practicing court judge, commenced their higher studies with two standard Shafi'i texts, the just-mentioned Al-Minhaj by Muhyi al-Din al-Nawawi, a Syrian who died in A.D. 1277,10 and a still more radically concise manual, known as Al-Mukhtasar ("the abridgment") or simply as the matn , the "text," of Abu Shuja', a resident of Basra active in the twelfth century.11

Both were classics of the Shafi'i school widely studied beyond the boundaries of Yemen. Al-Nawawi's and Abu Shuja"s manuals are to the Shafi'i school what Khalil's well-known Abridgment , for example, is to the Maliki school predominant in North and West Africa. In nineteenth-century Cairo, students of Shafi'i jurisprudence at the famous al-Azhar mosque-university began with the shorter work by Abu Shuja', while Al-Minhaj was a principal intermediate text in the typical sequence of study.12 At the same time, at the eastern end of the Muslim world, in Malaya and Java where the Shafi'i school was pre-

dominant, the same two texts were fundamental in both instruction and Islamic court rulings. Both were translated from Arabic into French, Al-Mukhtasar just after mid-century and Al-Minhaj two decades later. For use in British-administered colonies, including the nearby Indian Ocean port of Aden, an English translation of the French edition of Al-Minhaj was available by 1914.13

In the following two sections, I examine the recitational and open identities of the fiqh-manual genre as a means of introducing some of the main features of the core discursive tradition. I elaborate upon the paradigmatic qualities of the Quran and make comparisons with the texts of the collateral science of hadith.

Recitation

Writing is the outlining and shaping of letters to indicate audible words which, in turn, indicate what is in the soul. It comes second after oral expression.
Ibn Khalbun

Concise fiqh manuals were explicitly designed to be memorized. The extreme brevity of Abu Shuja"s text was intended, as the author notes, to "facilitate its study for the educated and simplify its memorization for the beginning student." Referring to a larger work summarized in his manual, al-Nawawi says, "Due to its great size, many contemporaries, except for the most dedicated, are unable to memorize it, and I therefore undertook to shorten it to about half that size to simplify its memorization."

Recitation and memorization were at the foundation of Muslim pedagogy, in both literal and general methodological senses. In neighborhood Quranic schools in Ibb, as elsewhere in the Muslim world for many centuries, young children began their instructional lives by acquiring the sacred text by rote. The goal was recitation from memory, and the basic method was oral repetition, which was supported by the technology of the pupil's individual lesson-board. The passage to be learned was written on the board, and once it was memorized the writing was washed off, to be replaced by the next passage to be acquired. A similar pattern was repeated at the higher academic level, in training in the shari'a, the principal subject matter of advanced instruction. At this level Ibb students worked to memorize basic manuals, such as Abu Shuja"s Al-Mukhtasar and al-Nawawi's Al-Minhaj .

Pedagogical activity, in primary instruction as in advanced work, proceeded ideally from an initial oral recitation (or dictation) by the

teacher to the listening student. The student later repeated the text segment on his own, often until memorization was achieved. Finally, returning to the teacher, the student endeavored to accurately reproduce the original recitation. Writing intervened in these procedures in facilitating the work of repetition; its role was decisive but understated. Manuscript copies were often made by advanced students in the course of study, but these were considered by-products of the learning process. Students eventually received licenses to teach, which entailed permission for riwaya , the "oral transmission" of a particular text or texts through recitation.

The paradigm for this was the Quran. Received orally by a Prophet who, according to doctrine, could neither read nor write,14 the revelations contained in the Quran are considered the spoken Word of God. The textual character of the Quran is quite different from that of the Bible, or at least the Gospels, which are considered humanly authored and which constitute a "book" in a sense closer to the contemporary Western meaning. The Quran, by contrast, is a recitation-text. The Prophet was instructed by the Archangel Gabriel to "recite," and the Quran, an extended "recitation," was received by him and then orally reconveyed in this way to his companions. As the Quran circulated in the world, recitations were repeated and memorized, the text was preserved in human hearts, and, in the event, a discursive style was set in place. The Quran's written form, the physical text located "between the two covers," would always be backgrounded in relation to its emphasized recitational identity. A century ago, Snouck Hurgronje urged Western scholars "to give up the erroneous translations of Quran by 'reading,' and [the root verb] qara'a by 'to read."'15 By attending to the revealed Word and recitational (versus read) qualities of the Quran, recent scholarship has begun to revise earlier assumptions about the identity of this sacred text as a book or scripture in the conventional Western sense.16

If the links of recitational reproduction were crucial to the authoritative character of both teacher and transmission, they were essential as well beyond the lesson circle. Recitation (qira'a )17 is the mode in which the Quran figures in ritual life, and a recitational style also structures the articulation of basic shari'a principles in legal practice. Trained jurists such as shari'a court judges did not have to refer to written versions of authoritative legal manuals any more than professional Quran reciters (or ordinary individuals in prayer) had to "read" the Quran: both recalled the text directly from memory. Memorization of

at least a small portion of the Quran is essential to the performing of daily prayers, a hallmark of membership in the Muslim community. Although memorization of the whole Quran was highly valued, the minimum necessary for prayer is the brief first section known as the fatiha , or "opening." Unlike the restricted scholarly lineages associated with the authoritative texts of the intellectual disciplines, the "descent group" of the paradigmatic authoritative text comprised all those who identified themselves as Muslims. In contrast to this community-defining status, shari'a scholarship passed down narrow channels, through such men as Faqih al-Nahi of Ibb.

Elements of the recitational complex were set in place early on, in dialogue with a reciprocally constituted form of "writing." A parallel to the patterned receipt and transmission of shari'a scholarship existed in the separate but closely associated scholarly discipline of hadith studies. The most significant hadiths report on the words or actions of the Prophet Muhammad and are, therefore, crucial in ascertaining his sunna , the statements and acts that represent his authoritative practice. Like the memorized and recited Quran, the once only orally transmitted hadith reports were also set down in written recensions, but this occurred despite specific Prophetic orders (also in the form of hadiths) to the contrary. Those who would undertake the recording of such reports had to contend with contradictory dictums (e.g., "Abu Sa'id al-Khudri said: 'I asked the Prophet permission to write the hadith down but he refused it'"; "Do not write down anything from me except the Quran. He who has noted down anything from me apart from the Quran must erase it"; but, on the contrary, "'Abd Allah b. 'Amr asked the Prophet permission to write the report down. It was granted"; and, generally, "Commit knowledge to writing").18 Some early scholars held it was forbidden to place hadith reports in writing; others advocated memorization and cautioned about the unreliability of writing. One warned: "Strive eagerly to obtain hadiths and get them from the men themselves, not from written records, lest they be affected by the disease of corruption of the text."19 The recording process went forward, but only after a century during which such reports were transmitted by exclusively oral means. As had occurred when the Quran was set down in a definitive written text,20 the human interventions to preserve such reports in writing were fraught with intracommunity conflict. Collections of reports of established authenticity became the foundational works in the field of hadith studies.

In Ibb such studies were integral to scholarly activity. For example,

another of the town's twelfth-century cohort of scholars was a renowned specialist in hadith, nicknamed Sayf al-Sunna ("Sword of the Sunna").21 In the year 1184, he traveled from Ibb to Mecca to hear the recitation of one of the most authoritative of all hadith collections, a work known as the Sahih , by Muslim b. Hajjaj (d. 875).22 Returning to Ibb, Sayf al-Sunna taught this text in the same recitational way he had received it, licensing in his turn a number of regional scholars to transmit it to their own students.

As the field of hadith scholarship developed, the essential critical activity was to sift authentic from fabricated reports. Methodological emphasis in this discipline was placed on the close scrutiny of the individual transmitters. Since reports were initially handed down orally, attention was focused on the character and circumstances of the human links in the "chain" (isnad ) of transmission. It was said that "knowing hadith reports means knowing the men." The reliability of a handed-down oral report of what the Prophet had said or done depended on the existence of an unbroken and unimpeachable series of reputable and reliable word-of-mouth transmitters. A "science of men" ('ilm al-rijal ), as it was called, grew up at the center of hadith studies, influencing also the early development of biographical histories. The parallel orientation in the discipline of shari'a scholarship is indicated by Faqih al-Nahi's previously cited statement that "between me and the author are two men." In later centuries, the great length of intellectual genealogies as well as breaks in their links would contribute to local crises of textual authority.

Both a whole fiqh manual and the one or more sentences of the typical hadith report were referred to as a matn , "basic text." Manuals and individual reports shared a kind of textuality in which writing, or the text in its written form, was considered secondary and supplementary. The privileging of the recited word over the written text and an associated concern with the specific connections of oral-aural transmission marked both genres of authoritative text. Paradoxically, these distinctive attitudes grew up within the context of an encompassing literate tradition and a thoroughgoing social reliance on writing. Walter Ong (1982:26) has written that Muslim cultures "never fully interiorized" writing. Both Ong and Jack Goody (1968:14) refer to such privileging of recitational forms as "oral residues," features indicative of a "still" partly oral society. But this recitational emphasis is perhaps better understood as a complex motif of a fully realized type of civilizational literacy. Muslim societies elaborated diverse, historically

specific textual worlds, central elements of which were their particular understandings, and relative valuings, of the recited and the written. It is only with the application of Western-modeled yardsticks for complex forms of literacy and for universal (evolutionary) routes of oral-to-written shifts that cases such as the Muslim one can be made to appear incomplete, marked by residuals, or stalled in development.

One version of the divine-to-human transition was in the specific relation, and movement, of recited word to written text. Although there are major differences in structure and history, the associated Muslim attitude toward writing may be initially compared to one Derrida (1974) has detected in the Western tradition, beginning with Plato and continuing down to such figures as Rousseau and Saussure. According to Derrida, concrete writing has been consistently denigrated while a metaphorical writing, a kind of natural or divine inscription associated with the spirit, the voice, and speech has been extolled. In relation to this primacy of the spoken word, concrete writing was considered exterior instead of interior, human instead of divine, and artificial rather than natural. Consonant with the central Western metaphysical concern for "presence," speech was deemed the valued locus of truth, the means of direct voicing of the spirit, and, for Saussure, the proper focus of scientific linguistics. Writing—secondary, representational, supplementary—was frequently condemned as evil or contaminating. As such philosophical, theological, and theoretical positions were taken, however, it was concrete or "fallen writing" that was the consistent medium of intellectual discourse.

In Muslim societies, a culturally specific logocentrism, as Derrida terms this privileging of the spoken word, has had widespread institutional implications, many of which are connected to the recitational cast of the basic texts. In its varied forms, recitation purported to convey an authoritative genuineness of expression by replicating an originally voiced presence. Recitational logocentrism went hand in hand with a concern about the problem of authorial absence in certain written texts. While recitation was thought to maintain a reliable constancy of meaning, the secondary medium of writing was seen as harboring a prospect of misinterpretation. "Once a thing is put into writing," as Plato has Socrates say in the Phaedrus (1952: 158), "the composition, whatever it may be, drifts all over the place." Extending the authority-giving presence of an original author to other places and times would be the work not only of recitational devices but also of writing. In written form, however, the general misreadability of the

medium was dangerously extended by the open potentiality of the texts themselves. An author's voice, and thus his presence and his truth, could be securely recovered only through the technique of recitation.

All this was especially crucial in connection with the sacred text and similarly treated authoritative texts, including fiqh manuals and hadith reports, which partook of the privileged quality of the recited word. When textual authority, involving the restoration of an original voiced presence, was at stake, recourse was had to the faithful recitational reproduction, directly from memory or, secondarily, by reading the matn in question aloud. According to al-Shafi'i, an authoritative relayer of hadith should be "capable of transmitting the hadith letter for letter [bi-hurufihi ] as he heard it." Relaying a report in the form of its gist or paraphrased meaning was unacceptable and dangerous: "If he transmits only the meaning and is unaware of what might alter its sense, he might unknowingly transmute the lawful into the unlawful and vice-versa." Al-Shafi'i concludes that if a relayer "transmits letter for letter there remains no ground for fearing a change of the meaning."23

Recitational mechanics began at the alphabetic level. Much has been made of the supposed consequences of alphabet differences, especially the contrast between the "absence" of vowels in the Semitic languages and their presence, in the form of alphabetic letters, in languages such as the Greek. A tendency to focus on comparative civilizational advantage has obscured contextual consideration of how particular non-Western alphabets related to the systems of which they were a part.24 As it usually appears on the page, Arabic script consists of strings of unvoweled consonants. The act of voweling, whether by marking in the vowel signs over and under the consonantal string or by voicing them in recitation, is an interpretive act, lending the script a particular significance in the process. This is important because written texts often allow alternative vowelings.25 While script preserves a string of consonants, recitation units consonants and vowels, enabling the production and reproduction of a whole. Given the nature of its script conventions, there is an identifiable physical loss in "reducing" something in Arabic to writing. In comparison with a fully vocalized "word," a written text can be considered an incomplete consonantal fragment. Preserved in its voweled-consonant recitational form, by contrast, a memorized text is one that has been embodied complete.

Versification was often mobilized to enhance the mnemonic accessibility of basic texts. Further accentuating an already predominant recitational design, repetitions of what Jakobson called a "figure of

sound" facilitated the task of memorization while evoking the much admired oral declamations of the poets. The versification of scholarly texts presented the compositional challenge of compactly presenting complex matters, complicated by adherence to a set rhyme scheme. Many authoritative texts in jurisprudence and in other fields were composed and studied in verse form. Two of the basic texts used in turn-of-the-century Ibb, for example, one in hadith and the other in grammar, were compositions of a thousand verses each, and were thus known, with their respective author's names, as "al-Suyuti's Thousand" and "Ibn Malik's Thousand."26 Earlier on, Shafi'i scholars in Ibb had relied on a verse-text jurisprudential manual by Ibn al-Wardi.27 In several formal genre categories,28 verse has long represented a principal channel for the assertion of scholarly views and rebuttal, as well as for the display of erudition and ingenuity.29 Since they intimately and succinctly convey character and personal attitudes, verse excerpts also abound in biographical dictionary entries.

Composition could also be recitational, although this was typical only for the taught subset of the literary corpus. Al-Shafi'i's own works, including his famous Risala , are an example. Such works were constituted in the teaching circle. As al-Shafi'i dictated his leading pupils took down his spoken words in writing. He would later correct the text when it was read aloud in his presence.30 An origin in recitation confirmed an authorial presence and thereby enhanced the authority of the work. Additionally, the Risala takes the form of a discussion, with questions by an anonymous interlocutor and answers given by al-Shafi'i. In this respect, the Risala resembles nothing so much as the Platonic dialogues. This dialogic, question-and-answer format occurs in many treatises constructed either in the form of a scholarly disputation or as queries posed to an authoritative interpreter.

A Yemeni example of recitational composition is a work by a former ruling imam who was deposed and in prison. This work, the basic jurisprudence manual of the shortly to be discussed Zaidi school, was, like Al-Mukhtasar and Al-Minhaj of the Shafi'is, designed for memorization. The imam taught it orally to a fellow prisoner who, upon release, had a written text prepared.31 In terms of its framing, however, the imam's manual is different from al-Shafi'i's Risala . Like other manual texts, it was explicitly intended for oral instruction and memorization, but rather than presenting a quasi-modeling of direct speech, it is condensed and formulaic. In the Zaidi manual, the author's intention is preserved, not by a compositional format recalling an original inter-

change, but by a determined concision that facilitates the memorized acquisition of the fullness of his words.

This authoring in the lesson medium seems comparable to the situation in medieval England where "dictating was the usual form of literary composition" and where the individual activity of composition could be experienced as dictating to oneself (Clanchy 1979:219). But the comparison is deceptive, for the English case is one taken from the cusp of a particular historical transition to what Ong (1982:95) calls "high literacy," in which an author "puts his or her words together on paper." The Middle Eastern examples, however, pertain to another history, governed by specific cultural assumptions about the inscribing of authorial presence in recitational media. They also concern only one segment of a literate system that, in other sectors, already exhibited the compositional features of "high literacy."

A prominent feature of the intellectual and institutional landscape of Islam, the recitational complex was, at the same time, far from comprehensive. Recitational emphases mainly concerned certain core subjects, such as shari'a jurisprudence and supporting fields, which were also the instructional mainstays of the old curriculum; within these subjects, recitational structuring was especially relevant at the level of the fundamental texts. Beyond these core subjects, there were important recitational influences, but there were also intellectual fields (medicine, history, philosophy, etc.) and other domains of activity where reading and writing predominated relatively unfettered by recitational concerns.32 Qualification must also extend, however, to the place of recitation within its principal strongholds, the core subjects and the authoritative texts. There, a foregrounded recitational construction of the transmission processes, an ideological emphasis on recitation, audition, and memorization, must be weighed against the evidently important, but consistently underplayed roles of the techniques of reading and writing.33

A sectoral supervaluation of recited texts also did not rule out a comfortable general relation with written texts. The coexistence of recitational forms and their written versions was taken for granted. An early exemplification of this is found in further comments by al-Shafi'i concerning his inclusion of numerous hadiths in his famous Risala : "I did not want to cite reports that I had not fully memorized, and I have lost some of my books, but I have verified the accuracy of what I memorized by checking it with the knowledge of scholars."34 In another passage he establishes basic principles regarding memorization

and written sources. An authoritative relayer "should have learned the report by heart, if he relates it from memory, and should have memorized the written text if he relates it from its written form."35 As the importance of memorized retention is repeatedly stressed, an interdependence with writings is consistently recognized.

In Yemen, as elsewhere in the Muslim world, there was a flourishing culture of the book, expressed in such activities as library keeping, the calligraphic arts, bookbinding and selling, and manuscript copying.36 Some scholars made a personal practice or profession of copying books. One such Ibb copyist, who "copied in his own hand numerous books and . . . wrote lines of his own poetry on each of them,"37 had studied calligraphy (khatt ), among other subjects, with the town's noted early hadith specialist Sayf al-Sunna. Another is said to have wanted to make a copy of a well-known book by al-Ghazzali, but lacked gallnuts, the standard ingredient in one type of ink. He used instead the wood of a locally abundant tree and composed lines of poetry to the effect that in the Ibb region, because of the availability of that particular species, one's work need never be interrupted by the lack of conventional ink (Ibn Samura 1957:193).

While authoring was a phenomenon of the qawl , an individual's "word" or "voiced opinion," copying was one of the yad , the "hand" or "handwritten script." Manuscript copies conclude with appended statements, "in the hand of so-and-so," which identify the copyist. The relation of qawl to yad , of voiced word to handwritten script, is another variant of the basic relation of original to supplement. Within the sphere of the written supplement, however, there were conventions to establish the relative authority of the versions. A familiar patterning was reproduced: copies had their own sort of genealogy. Most authoritative, as might be expected, was the autograph, a manuscript copied out in the hand of the author. In such cases qawl and yad came as close to convergence as possible. Otherwise, in the ordinary activity of manuscript copying, there was an important distinction between the written "original" (asl ) and the written "copy" (nuskha ). This relation of original and copy, internal to the written realm of manuscripts, suggests a replication there of the wider opposition of basic text to supplement. When a copy was completed it could be vested with authority through careful comparison with the original (a collation procedure referred to as muqabala bil-asl , "comparison/encounter with the original"). This was the work of two men, present to each other, involving an activity exactly parallel to the recitation (dictation) and audition of the teacher-

student lesson interchange.38 In this manner the copy could take on the authority of the original and become, in turn, an asl (a source text) with respect to copies made from it. That authoritative copy making closely followed the structural pattern of authoritative knowledge transmission illustrates the reach of recitational models beyond the sphere of the core pedagogy.

Open Texts

There can be no commentary unless, below the language one is reading and deciphering, there runs the sovereignty of an original Text. And it is this text which, by providing the foundation for the commentary, offers its ultimate revelation as the promised reward of commentary. The necessary proliferation of the exegesis is therefore measured, ideally limited, and yet ceaselessly animated, by this silent dominion.
M. Foucault, The Order of Things

An additional set of attitudes surrounding authoritative texts may also be introduced with reference to instructional methods. As a second and integral part of the standard lesson, an initial recitation by the teacher of a segment of a matn , or basic text, was followed and complemented by his elucidating commentary, his sharh . This matn/sharh , text/commentary relationship was a fundamental one. In the pedagogical format, the relation of basic text to expansive commentary was also one of recitation meant for memorization to supplementary interpretation meant for nonmemorized understanding. While students would endeavor to acquire the matn by heart, this technique would not be applied to the sharh . The sharh served the subordinate role of informing a student's comprehension of the principal focus of instruction, which was the matn . The text/commentary relationship was crucial also to the format of books. The written literature of shari'a jurisprudence, for example, developed largely by means of interpretive elaborations on basic texts.

Radically compact expression in matns both facilitated the memorization of essential formulae and absolutely required interpretive commentary to clarify the "crowded" meanings. Standing alone, abbreviated matns are often only barely comprehensible, since they are composed in a kind of stripped-down, subconventional prose in which many of the connecting words and phrases of ordinary discourse are elided. A minimal accomplishment of a sharh , in the oral lesson as in written composition, was the filling in of such rudimentary connections.

In written works, the resulting expanded text, that is, the matn plus sharh , typically represented a much closer approximation of normally comprehensible prose. This transformation was made possible by a very important and immediately obvious fact about the text/commentary relationship: its physical aspect. While the matn is a textual "body," sharh means to "open up" (another referent is to surgery). The commentary, ranging from trivial linking words to lengthy and important doctrinal elaborations, is inserted in spaces opened up in the original text. Although they remain distinct, the two are not physically isolated from each other, as are either footnotes located on another part of the page or entirely separate volumes. With the commentary embedded directly in the original text, grafted into spaces opened up in its body, the two alternate in a jointly constituted, matn -and-sharh text: a segment of matn is followed by a segment of sharh , which is followed by another segment of the matn , and so on. A range of markers, from different ink colors (see fig. 3) and a type of overlining to conventional wording shifts ("he said"/"I say") were used to signify the transitions back and forth between the two.39 In such a situation, reading involved traversing already familiar or even previously memorized segments of matn (as well as citations of Quran or hadiths) embedded in the larger commentary text.

In a work of sharh , interpretations literally become part of the text interpreted. If it consisted only of its insertions, a work of commentary would be merely a collection of disconnected and unviable fragments. Instead, a sharh work is actually taken to be the sum of the matn and the added comments, involving a complex notion of full quotation. Encompassed by the larger sharh , the matn appears in continually interrupted bits and pieces, but the quotation is faithful, word for word, and eventually complete. Interpretive in intent, the commentary genre joins two insufficiencies. It is the destiny of a matn to be interpretively expanded by sharh and that of a sharh to depart from a global invocation of a matn . A related sort of embedded, in extenso quotation of matn is also basic to hadith works. In synopses, the quotational process worked in reverse: as interpretive reductions rather than expansions, the distillations depended on preexisting full texts. At issue in both expansive and retractive genres are distinctive conceptions of originality and authorship.

Fundamental to both the pedagogy and the organization of books, the text/commentary relationship had further structural resonances. On a microlevel, the relationship is comparable to that of consonants and vowels in Arabic. A relation of insertion similar to that between

Figure 3.
Commentary by al-Ghazzi, in black ink, on text by Abu Shuja', in red ink,
shown here in gray. Late-seventeenth-century copy, Arabic ms. 4350. By
permission of the Houghton Library, Harvard University.

consonants and vowels obtained between matn and sharh . Like an unvoweled consonantal string, a matn can and frequently does stand alone; by contrast, neither vowels nor commentary insertions are independently viable. But both consonant strings and matns are open to, and ultimately dependent upon, the interpretive interventions of voweling and textual commentary.

A text and commentary type of relation is also part of a much wider pattern, one that goes to the heart of a general understanding of the growth of the shari'a. An equivalent relationship exists between the Quran, as the paradigmatic text, and the Sunna of the Prophet, with the latter standing as a kind of expansive elaboration, through Muhammad's actions and statements, upon the Revelations contained in the former. Taken together, the Quran and the Sunna, the two "sources," stand in turn in the position of basic "texts" with respect to the body of interpretive elaboration that became shari'a jurisprudence, or fiqh . Up to the establishment of the four standard schools of law, this jurisprudence developed through the efforts of the early Muslim jurists, who employed a method of analogically based interpretation (known as ijtihad ), the refining of which was one of the major contributions of al-Shafi'i. Thereafter, within each of the established schools, the text/commentary relationship was replicated still further, as the authoritative works were themselves subjected to successive commentaries and summaries, some of which became authoritative in their own right. Muslim jurisprudence developed like a branching tree, expanding outward from the single, original, and paradigmatic Quran, complemented by the Sunna, through the foundational early works of the several schools, to the concise manuals and then the outer reaches of proliferating commentaries and glosses.

New contributions stood squarely on the shoulders of earlier works. Both Al-Minhaj and Al-Mukhtasar are original texts upon which numerous subsequent written commentaries and glosses were based. Each constituted a point of textual departure for a separately branching literary tradition in Shafi'i scholarship.40 Explicit lines of textual ancestry are traceable. Among the descendants of al-Nawawi's thirteenth-century Al-Minhaj , for example, are several sixteenth-century commentaries and synopses, while the same text's ascendant line goes back through a work by al-Rafi'i (d. 1266), which it synopsizes. Al-Rafi'i's text, in turn, was based on two short works by al-Ghazzali (d. 1111), both of which summarized one of the great scholar's own larger works, which was itself based on a work by his teacher al-Juwayni (d. 1086).41

The highly implicit and generally "unfinished" qualities of authori-

tative manuals necessitated and actively invited interventions, from recitational vowelings and the simplest reader emendations to full interpretive commentaries. Such texts were endowed with a dual and contradictory character: their marked stability was anchored in the assumed security of person-to-person instructional linkages and the authenticity of recitational transmission, while their equally fundamental instability derived from an absolute need for elucidation. While the intellectual world that this duality implied was one oriented toward a tension between a settled conservatism and a required, but criticizable, interpretive dynamism, the social world implied was one based on interpretive relations of hierarchy and power (see chapter 8).

Openness in authoritative texts was not only a consequence of concision, however, but equally a matter of internal discursive construction. Although by no means devoid of clear principle, shari'a scholarship was very much an ongoing argument, one characterized as much by recognitions of divergent views as by points of univocal consensus. The open character of such texts was crosscut by a basic positivism, a conviction that truthful positions existed, although particular truths might be known only to God. A polyvocality of competing opinion stood opposed to the notion that authorial presence was the medium of univocal truth. It was acknowledged that individual scholars could differ in their interpretations in the absence of a clear Quranic text or an authentic hadith from the Prophet.42 Decisive texts, however, especially in the paradigmatic form of individual Quranic passages, were understood to convey truth. A "text" (nass ) of this type, that is, an individual source-passage within an authoritative text, was defined by scholars as "carrying only one meaning" (al-Juwayni n.d.: 46). When, as was often the case, positions were taken but the truth of a matter remained uncertain, this could be represented in manuals and other works by the appended formula "and God knows best."

When divergent views exist on an issue, al-Nawawi is very careful to report this fact. At the beginning of Al-Minhaj he establishes terminology to be used throughout the work to identify the status of alternative opinions. He exercises his interpretive preference, but in so doing he does not cover up other voices. "In all instances" he offers an "elucidation and evaluation of opposed opinions, of opposed ways a question can be addressed, and of opposed methods."43 His terminology allows him both to specify that a particular doctrine appears preferable to him and to indicate whether the doctrine or doctrines rejected are widely accepted or not. Other terms similarly enable him to choose a manner

of addressing an issue while also noticing the strength, or weakness, of different approaches. An example of this precision is the usage of the comparative degree—"clearer" and "sounder"—to represent his choices in instances where, as he says, "disagreement is strong." When al-Nawawi simply states his own view, he utilizes the opening expression qultu , literally "I said," together with the cautious concluding phrase "and God knows best." The effect of such careful usage is to highlight the essentially disputed quality of many positions advocated. Other terms employed by al-Nawawi include madhhab , to indicate the commonly accepted positions of the Shafi'i school, and nass ("text"), to indicate the source-text positions of al-Shafi'i himself. Shafi'i's personal doctrine, which sometimes involves "a weak view or a divergent opinion," is further broken down, however, into the often opposed positions he held early and late, that is, during the Iraqi versus Egyptian periods of his career. Al-Nawawi uses the conventional terms "old" and "new" to mark off these different positions held by the school founder.

The overall impression of shari'a jurisprudence as presented in Al-Minhaj is of an unstable mix of the settled and the contested. In the text itself al-Nawawi remarks that he "has already begun work on a volume which will take the form of a commentary on the details of this abridgment." Despite the definitiveness Al-Minhaj came to represent in the Shafi'i school, it was nevertheless viewed, beginning with its author, as an open site, as a text in need of further clarification. To the extent that an extremely concise manual might be relatively univocal because of the dictates of space, commentaries inevitably opened up points of contention. This is the case with the very brief matn by Abu Shuja', the other manual relied upon in Ibb. Nothing beyond a single-view outline was possible in the scope of such a matn , but in the hands of al-Ghazzi, his principal commentator, many of Abu Shuja"s points are immediately modified, opened up to alternatives, or even bluntly reversed. As opposed to the Quran (12:1), which identifies itself as the "clear text" (al-kitab al-mubin ),44 the discourse of the shari'a literature built up from it appears markedly and openly discordant. Divergences represented by the mere existence of separate schools of shari'a interpretation were further developed within each of the schools. Fiqh manuals served to summarize but could not finally contain the essentially contentious terrain of authoritative shari'a thought.

In many historical settings, schools competed for the intellectual and, the subject of the next chapter, the political terrain. In chapter 2, late-nineteenth- and early-twentieth-century highland history is the back-

drop for a discussion of differing shari'a conceptions of the state. With the Revolution of 1962, however, the Yemeni nation-state required and eventually adopted an entirely new form of law. Chapter 3 examines the transformation of the recitational and open character of authoritative shari'a texts through processes of codification and legislation.

Excerpted from The Calligraphic State: Textual Domination and History in a Muslim Societyby Brinkley Messick Copyright © 1992 by Brinkley Messick. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.