Tahammul al-'ilm: how hadith was taught and received

From the compiler to the classroom

In the previous article we saw how al-Zuhri ﵀, under the caliphal command, systematically gathered together the hadith of Medina and Damascus, and how, around his person, dozens of students assembled their own written collections. But a compiler does not fall from the sky. Behind every collector stands a teaching system that shaped him: a manner in which knowledge was offered, received, checked, and passed on. The science of hadith has a name of its own for the whole of this, tahammul al-'ilm, the bearing of knowledge, and Mohammad Mustafa al-Azami ﵀ devoted an entire chapter of his Studies in Early Hadith Literature to it, built upon hundreds of scattered reports from the earliest sources.

Why does that system deserve a place of its own on this map? Because the skeptical reading of hadith is, in essence, a reading of the teaching. Whoever claims that traditions could be freely invented and projected backward in the second and third centuries is thereby claiming that the teaching in which those traditions were passed on was formless: without administration, without control, without any memory of who had received what from whom. Azami’s chapter shows the opposite, not by way of one spectacular document, but through the accumulated ordinariness of hundreds of classroom scenes. There existed a pedagogy of hadith, with working methods we can name, with tests we can describe, and with a culture of admission whose rules have been preserved. This article walks through that system, from the circle around the Prophet ﷺ to the dictation classes of the second century.

The foundation of it all lies, as so often on this map, in the Qur’an itself. Azami opens his chapter with the verse in which the acquiring of religious knowledge is made a duty of the community, even in times of war:

From that command, Azami writes, an eruption of teaching activity arose throughout the Islamic world, and the word for knowledge itself, 'ilm, was for centuries used in the first place for the knowledge of the ahadith and what belonged to them. Whoever said in the second century that a man was seeking knowledge generally meant: he sits with the transmitters.

One warning from the author himself the reader must carry along, because it colors the whole chapter and because it is typical of his scholarly manner. In those early centuries there existed no organized universities with fixed curricula and fixed methods of instruction; every teacher and every student could choose the form that suited him, and the practice differed from time to time and from place to place. Every generalization, Azami writes, every restriction to a single fixed practice, is dangerous and misleading; what he offers is a sketch. That is no weakness of his argument but a strength: the working methods that follow are not an invented system laid retroactively upon the sources, but patterns that rise out of the sources themselves, with all their variation.

The first circle: the Prophet ﷺ as teacher

The teaching of hadith begins not with a scholar, but with the Prophet ﷺ himself, who called himself a teacher and sat with his Companions ﵃ in a circle to instruct them. His words and deeds were carefully followed by the Companions ﵃ and retold among themselves until they had mastered them. And because not everyone could be present on every occasion, something arose that one may fairly call an attendance arrangement: Companions ﵃ agreed among themselves to attend the circle in shifts, whereby whoever was present afterward informed the absent of what the Prophet ﷺ had said and done. This was not only their own arrangement; it was also the command of the Prophet ﷺ himself, and the sources preserve, as an example, the agreement of Umar ﵁ with a neighbor of the Ansar, who took turns replacing one another.

How heavily that attendance weighed appears from the small, almost moving report about the Companion Sulait ﵁. He had received from the Prophet ﷺ a piece of land that lay far from Medina, and the absence that its management imposed on him distressed him so much that he asked the Prophet ﷺ to take the land back. A gift of land did not outweigh missed lessons. Thus the first generation regarded its own presence in the circle: as something it could not allow to be replaced by wealth.

Some Companions ﵃ made their notes already during the lifetime of the Prophet ﷺ, others assembled their collections later, and it was a widespread custom to gather and repeat the ahadith individually or jointly, a practice that simply continued in the time of the Successors. Out of those lessons the first books then grew. In the middle of the first century of the Hijra, Azami notes, books of ahadith began to appear that had been composed from the lessons of teachers, and among the earliest that are mentioned are the books of Bashir ibn Nahik ﵀ and Hammam ibn Munabbih ﵀, both students of Abu Hurayra ﵁; the books from the circles of Ibn Abbas ﵁ and Jabir ﵁ belong to the same period. The reader of this map knows those names: the sahifa of Hammam ﵀ is treated further on as a physical manuscript. What becomes visible here is the delivery room in which such documents came into being: not an archive and not a chancery, but a classroom.

In the third quarter of that same century a technique then appears that betrays how written the teaching had by now come to think: the atraf, literally the ends. A student did not write out the full hadith, but only its beginning or a distinctive part, as a memory-key for a text he possessed or had mastered in full elsewhere. The earliest mention of this method that Azami could find concerns Ibn Sirin ﵀ in the lesson of his teacher Abidah al-Salmani ﵀, who died in the year seventy-two; later, complete reference works were built upon this principle, which the early scholars used as a kind of concordance. A memory-aid of this sort makes sense only in a culture where the full texts already lay fixed. Whoever writes atraf, possesses books.

By subject or by teacher

How did one order the growing mass? Azami found the earliest trace of a thematic arrangement in two men who met in Medina: Ibn Abbas ﵁ and Urwah ibn az-Zubair ﵀ revised ahadith between them by subject, heading by heading. But precisely at those sessions the sources note a telling restriction: because in such a thematic discussion one did not go through a tradition in full and in its precise wording, but only touched upon its essential points, both scholars forbade their students to learn traditions by this method. One may pause a moment over what this implies. In the teaching of the first century there already existed a worked-out distinction between discussing a tradition and transmitting it, and only the second form, the full and word-faithful one, counted as a valid channel. A community that formulates such a rule is not naive about textual corruption; it has a policy on it.

The arrangement that won the day was not the one by subject but the one by teacher. In the second century it was the common practice to teach ahadith according to the shuyukh from whom one had them: one gave the hadith of this shaykh, and then the hadith of that one. The sources preserve the conversation in which al-Fallas, asked where he came from, answered that he came from the lesson of Mu’adh, and to the follow-up question whose hadith was transmitted there, answered: the hadith of 'Awn. The lesson had a name, and that name was a source. Teachers kept the ahadith of each shaykh separately, and there are numerous references confirming this as general practice.

That this was no abstract administration but a world of concrete, retrievable books is shown by the anecdotes Azami assembled. Mujahid ﵀ went to Hammad ibn Amr and asked him to bring the book of Khusaif; they brought him the book of Husain, the wrong one, then, and the story is preserved for precisely that reason. Al-Thauri ﵀ announced that he would bring the book of the best man of Kufa; the students expected the book of Mansur and he came with that of Muhammad ibn Suqah. And when Ahmad ibn Hanbal ﵀ discussed a tradition transmitted by way of Hushaim, he sent his student al-Athram to the house of Abdullah ibn Musa to look it up there in his books of Hushaim; al-Athram describes how the books of Hushaim were laid before him and how he found the ahadith of Ya’la gathered in one place, without the tradition he was seeking. Thus sounds a science that archives its sources by teacher and consults the archive when in doubt. The negative result of al-Athram’s search is the finest detail here: one looked things up not only to confirm, one looked things up to test, and sometimes the answer was no.

The four working methods

For the actual teaching, Azami distinguishes four systems that were in general use: the oral recitation of the teacher, reading from books, question and answer, and dictation. Whoever has ever seen a modern lecture hall, a seminar, and a practical class will recognize the repertoire; the difference lies in the role that memory and control played in them. It is worth looking at these four forms one by one, for in their details the same pattern lurks each time: oral and written were, in this teaching, not opposites but partners, whereby the spoken word brought the text to life and the written text guarded the spoken word. It is precisely this interweaving that the skeptical division between an oral early period and a written late period cannot accommodate.

The first form, the free recitation of the teacher from memory, was the oldest and the most prestigious, but from the second half of the second century it began to give way, though on a modest scale it long held its ground. Characteristic of this form was the very long bond between student and teacher: one stayed years with a single shaykh, until one was regarded as an authority on his ahadith and bore the honorific title rawi or sahib of that teacher, the transmitter or companion of that man. And the pace was, by modern standards, astonishingly low: in one lesson only a few ahadith were taught, say three or four. The sources that Azami gathers in his notes make it still more concrete: Shu’bah ﵀ related that from Amr ibn Dinar ﵀ he learned in five hundred meetings only a hundred ahadith, one hadith per five lessons, then, and al-Zuhri ﵀ advised his students to learn one or two ahadith a day, with the explanation that whoever learned much at once forgot much. Slowness here was no lack of zeal; it was the pedagogy itself. A text that was to attain the status of a tradition was not consumed but ground in.

The second form was reading from books, and Azami distinguishes three variants of it. The teacher could read aloud from his own book, and that counted as the most preferable way: the text then came directly from the controlled copy of the authority himself. The teacher could also read aloud from the student’s book, a copy or a selection from his own work, and precisely this variant held great pitfalls for teachers who did not fully master their material from memory. For here appears one of the most remarkable practices of the entire system: students and scholars deliberately played tricks, inserting strange ahadith here and there into the teacher’s copy, and laid the book before him to read in order to test the soundness of his knowledge and memory. Whoever did not recognize the inserted material was exposed and declared unreliable. The science of hadith knew, in other words, the random sample and the trap-test avant la lettre, and it applied them to its own teachers. Further on in this map, at the science of Jarh wa-Ta’dil, it will emerge how this culture of testing was raised to a system; here we see it in its natural habitat, the classroom.

The third variant of book-teaching became, from the beginning of the second century, the most common practice of all: the reading aloud to the teacher. A student, or a designated reader whom one called the qari’, read the text aloud while the teacher listened and corrected, and the other students compared what they heard with their own copies or listened attentively in order to copy it later. For the copies themselves an entire infrastructure existed: many teachers made copies available and kept their own scribes for the purpose, a katib or warraq, and students who already owned a copy brought it along. The fourth working method in Azami’s list, question and answer, was the smallest but perhaps the cleverest: the student read a part of the tradition, and the teacher completed it in full. Whoever has ever been quizzed on a text recognizes the method at once.

Dictation and the discipline of memory

The fourth great form, dictation, has a history of its own that is telling for the direction in which the teaching was moving. Apart from the dictations of the Prophet ﷺ himself and the occasional dictations of his early Companions ﵃, Wathilah ibn al-Asqa ﵁, who died in the year eighty-three, was perhaps the first to hold classes for it. In the early days the method was discouraged, and the reason the sources give is at first sight surprising: by dictation a student could gather much knowledge in very little time without much effort. Knowledge that came too fast was distrusted; the grinding-in of the slow lesson counted as the royal road. It was al-Zuhri ﵀, the protagonist of the previous article, who was the first to deviate systematically from that attitude: toward the end of the first century we find him dictating, and he followed that method for the rest of his life. From the second century onward dictation became, alongside reading aloud, the most usual form, with regular classes convened for the purpose.

Around dictation a profession of its own and a technique of its own grew up. In large audiences the office of the mustamli arose, the man who repeated the words of the shaykh in a loud voice for the assembly, row by row, so that even the scribes at the back could catch them word for word. Because not every student wrote quickly enough, one sometimes chose a fast writer who took down the dictation while the others watched to see that he made no mistakes, in order to borrow his text afterward or to copy it in the presence of the owner. In the literary circles a class of professional copyists thus arose, the warraqun, and from their work a trade in books; Hammam ibn Munabbih ﵀ bought books for his brother Wahb ﵀, and the hiring of people to write or sell a hadith book goes back to the late days of the Umayyads.

Dictation was done from a book or from memory, and with the latter variant Azami shows himself at his sharpest, for he does not conceal its weakness. Leaning on memory was in that time in part a matter of prestige and reputation, he writes, and this practice led to many errors, through the inherent defects of the human memory. But then follows the other half of the sentence, and it is decisive for the question of reliability: the scholars therefore had to go through their books to refresh their memory, and in many cases, when they were uncertain, they did not dictate. The system knew its own source of error and had two remedies for it: the book as touchstone, and silence in doubt. There were also extremes, scholars who refused to dictate when students wrote on erasable wooden tablets, and others who did write down and then erased again after memorization; Azami weighs them and calls them rare and unusual practices. The mainstream was a written culture with memory-discipline, not a memory-culture with occasional ink.

After the writing came the correction, and for that too a fixed custom existed: the copies that had arisen through dictation or transcription were corrected, by the students among themselves or under the supervision of their masters, and of Nafi’ ﵀ it is preserved that he asked his students to bring their books along for correction. For the rough draft one used wooden tablets, from which fair copies were later produced, and to save time and space one sometimes wrote with abbreviated words. Draft and fair copy, dictation and collation, reading aloud and reading against: whoever sets these words in a row describes not a romantic storytelling culture, but a scriptorium.

Access to the circle

The chapter also has something to say about the students themselves, and the picture is that of an ordered course of study. First one learned the Qur’an, largely by heart, and many scholars tested new students in the Qur’an before they were admitted; alongside this one learned subjects such as law, religious practice, and grammar, and generally one joined the circles of the muhaddithin around the twentieth year of life. Hadith was in this curriculum not a beginners’ subject but an advanced study, which one began only with the Book in the heart and the language in the fingers. The examples in Azami’s notes show the range: al-Zuhri ﵀ named Ibn Uyaynah ﵀, at fifteen, the youngest student he had ever seen, Ahmad ibn Hanbal ﵀ began when he was sixteen, the Syrians reportedly began only at thirty and the Basrans as early as ten. Only in later centuries did the age drop dramatically, when, for the sake of the shortest possible chain, the isnad 'ali, it became a fashion to seat children with famous witnesses of great age; and it is characteristic of the self-criticism of the discipline that transmitters who had written down their material in their youth were later assessed for it in their reliability.

Two social traits of this teaching deserve mention at the last, because they mark its character. The teaching of hadith was free; only a few asked money for it, and they were publicly reproached for that practice. The relation between student and teacher rested on reverence, some students served their teachers, other teachers precisely refused every service, so that it might not count as payment for the teaching, and in many cases teachers helped their students financially and it was quite ordinary to offer them meals. And the most curious phenomenon of this world of teaching, Azami writes, was the unbroken traveling of students and scholars to collect ahadith; the traveling was perhaps an essential part of studentship, al-Khatib al-Baghdadi ﵀ wrote a separate book about it, and the biographies are full of travel accounts. That rihla already has an article of its own on this map; here it suffices to establish that it was no romantic exception, but the logistical consequence of the shaykh-system: whoever wanted the hadith of a teacher had to go to that teacher.

Where all this took place is then almost a footnote: from the days of the Prophet ﷺ himself the mosques served as schools, a usage that lives on to this day in the Muslim world; alongside this one finds references to the kuttab, the writing-school, and to houses that were used as schools, while separate school buildings do not seem to have been built in the early period; when the throng was very great one withdrew to a public place.

Haddathana does not mean: oral only

There remains the question that binds this whole chapter to the skeptical literature, and with which Azami closes his first appendix. The transmission formulas of the isnad, haddathana, he told us, akhbarana, he informed us, 'an, on the authority of, sound oral to the modern reader’s ear, and upon that sound many a theory of a purely oral transmission has been built. Azami tests the sound against the practice, and the practice turns out to be otherwise. One of the four regular scribes in the lessons of Shu’bah ﵀ in Baghdad, Adam ﵀, described his own time of study thus: I wrote and people copied from me; I heard two thousand traditions. He says heard, while he received them by dictation and wrote them down; the word haddathana, Azami concludes dryly, gives half the picture. Zuhair ibn Mu’awiyah ﵀, when he heard a tradition for the second time, put a mark beside it: done. One listened with the book in hand.

The decisive example is the Muwatta’ of Malik ﵀, the book that has an article of its own on this map. Azami sets side by side eight passages from the classical collections in which the same material from the Muwatta’ is cited: one says haddathani Yahya, another akhbarana Qutaibah, the third contents itself with 'an. In all these cases, he establishes, the tradition was not passed on orally but by way of the book, while only the author was referred to. And with the Maghazi of Ibn Ishaq ﵀, likewise treated elsewhere on this map, he compares two independently transmitted versions of the same work and finds only the minute differences one expects between two manuscripts of a single book; it is unthinkable, he writes, that so large a book would have been passed on orally for centuries and that students would have had to learn the entire book by heart instead of writing it down.

His conclusion enumerates the forms of documentation that lie behind the formulas: copying from a written document; writing from a written source by dictation; the reading aloud of a written document by the teacher; the reading aloud of it by a student; the oral transmission of a document while students take it down; and the oral transmission that is orally received. One key point all these ways have in common: the permission of the shaykh to the student to use the material. Whoever used material without that permission earned in the discipline a name of disgrace of his own: sariq al-hadith, the thief of the tradition. Transmission in this system was no matter of hearsay; it was a granted authorization, bound to a verifiable course of study.

The later science of hadith terminology then codified this practice. In the mustalah literature of later centuries, with the Muqaddimah of Ibn as-Salah ﵀ as the best-known milestone, the modes of reception were ordered into the classical eightfold division: the hearing from the mouth of the shaykh, the reading aloud to him, the granted authorization to transmit, the handing over of the book, the written transmission, the notification, the bequest, and the finding. That division falls outside the corpus of Azami’s study and belongs to a later phase of the history of the discipline; we name it here because it shows what became of the practice of this article. What grew in the first two centuries as living classroom practice was later not forgotten or romanticized, but analyzed, named, and furnished with rules, as a mature science deals with its own history.

From the classroom to the piece of evidence

Thus tahammul al-'ilm proves to be the missing middle piece between two worlds that the skeptic likes to keep separate: the world of the early notes and the world of the later books. Between the two lay no silent century, but a school, with working methods, tests, rounds of correction, admission requirements, and a professional ethic of its own, and through that school the material passed from hand to hand without ever being unattended. Whoever can point out how a text was taught need not guess how it was preserved. And it is that same school which explains why the number of transmission channels grew so tempestuously in the second and third centuries: hundreds of teachers, each with dozens of writing students, multiply a modest core of texts into a forest of chains. What that means for the dizzying figures in circulation about hadith is taken up at length further on in this map.

But to describe a school is one thing; to hold its product in one’s hands is another. In the next article the map returns to Medina, to a document that arose precisely in this world of teaching and illustrates it to the very page: the nuskhah of Suhail ibn Abi Salih ﵀, the small written corpus that goes back through his father to Abu Hurayra ﵁. There we shall see what a concrete text from these classes looked like, how it was passed on, and why precisely its modest size makes it one of the strongest pieces of evidence of the early written transmission.