The Fifty Writing Companions ﵃: a census of the first generation

One document was no exception

In the preceding article we followed a single document: the Sahifa that Ali ﵁ kept in the scabbard of his sword, a written text on blood money, the release of captives, and the protected status of Medina. Whoever knows only that one story might think it was a rarity, the pastime of a single lettered man within an otherwise unlettered community. The question this article answers is therefore a different one, and a larger one. Not: did a written hadith document exist in the first generation? But: how many were there, in whose hands, and how do we know?

The answer to that question was assembled by one scholar with an almost bookkeeperly persistence. Mohammad Mustafa al-Azami ﵀, in his Studies in Early Hadith Literature, opened a catalogue that had until then simply not existed in the Western study of hadith: a numbered list of Companions ﵃ from whom hadith has been transmitted in written form, whether by their own hand, or by dictation, or through letters, or through students who wrote within their immediate circle. The list runs to fifty numbered entries, from Abu Ayyub al-Ansari ﵁ to Zaid ibn Thabit ﵁, each furnished with the classical sources from which the entry has been retrieved.

Fifty names. Anyone acquainted with the sceptical literature of the nineteenth and early twentieth centuries feels at once why that number matters. The school of Goldziher and his followers built its distrust of the hadith on a single load-bearing assumption: that for well over a century the transmission had been purely oral, a stream of words without ink, in which every generation could add whatever its politics or piety desired. An oral century is a dark room; whoever wishes to hide something in it always finds a corner. But a catalogue of fifty writing or dictating Companions ﵃ turns the light on in that room. This article walks through the census, weighs its evidential force, and treats the counter-voice honestly as well: for there do indeed exist reports in which the writing down of hadith is discouraged, and a census that passed over them would forfeit its own credibility.

The method behind the list

Azami presented his material not as a loose anthology but as an ordered survey with an accountable method. He divided the early bearers of hadith literature into four classes: the Companions ﵃ themselves; the Successors who lived mainly in the first century; the scholars who were born up to around the year 65 AH and whose literary careers still began within the first century; and the scholars who were born between 70 and 110 AH. He was strikingly candid about the limits of his own division, which he himself called somewhat arbitrary, since there were, after all, people who were born early in the first century and lived on well into the second. It is a candour the reader will meet again and again in this scholar: where the material is hard, he is firm, and where his own ordering is soft, he says so.

He accounted for the dating as well. Of many early transmitters only a year of death is known, and sometimes not even that. For this Azami used a checkable rule of thumb: subtract sixty to sixty-five years from the year of death as an average lifespan, or run through the man’s teachers, find the teacher who died first, and subtract twenty years from that date, since the study of hadith, by the custom of that time, generally began around the twentieth year of life. Whoever reckons in this way reckons with margins, and Azami accordingly presented his dates consistently as approximations. In total his study treats in this manner around five hundred scholars from the first century and a half; the fifty Companions ﵃ form its first and oldest layer.

One qualification deserves emphasis, because it makes the difference between a strong claim and an overstretched one. The census does not say that fifty Companions ﵃ wrote books with their own hands. It says that of fifty Companions ﵃ hadith has been transmitted in written form from their own lifetimes or their immediate circle: sometimes by their own hand, more often by dictation to a student or a scribe, sometimes in the form of a letter containing transmissions, sometimes as a document found within the family after their death and passed on further. For the evidential question, that variety matters little. Whether a text was written by the Companion ﵁ himself or dictated under his eyes, in both cases there existed ink within the first generation, and that is precisely what the sceptical school denied.

A walk through the census

Whoever opens the list begins with Abu Ayyub al-Ansari ﵁, the host of the Prophet ﷺ upon his arrival in Medina. It is transmitted of him that he wrote down ahadith and sent them to his nephew, and within his family a collection of one hundred and twelve transmissions circulated, of which Azami carefully notes that it is unclear whether it was compiled by Abu Ayyub ﵁ himself, by his son, or by his grandson. Here too that characteristic restraint: the catalogue records what the sources will bear, and nothing more.

Then the first caliph. About Abu Bakr ﵁ a well-known report circulates that he wrote down five hundred ahadith and later burned them. The report is eagerly cited in the sceptical literature, but the classical criticism had anticipated them: al-Dhahabi ﵀ declared the transmission false, and other scholars too rejected its chain. What is documented, on the other hand, is that Abu Bakr ﵁ as caliph had letters containing ahadith go out: a letter to Amr ibn al-As ﵁ that contained transmissions, and a written instruction on the zakat for Anas ibn Malik ﵁, then governor over al-Bahrain, which appears to have been a copy of the tax document of the Prophet ﷺ himself. The pattern that becomes visible here recurs throughout the whole list: the early administration of the community ran on documents, and in those documents hadith travelled along.

Abu Bakrah ﵁ sent his son, a governor in Sijistan, a letter with sayings of the Prophet ﷺ on the administration of justice. Mu’adh ibn Jabal ﵁ possessed a Kitab of the Prophet ﷺ on the sadaqat, of which Musa ibn Talhah later obtained a copy for inspection. Amr ibn Hazm ﵁, sent by the Prophet ﷺ as governor to Najran, was given a written instruction in which the prayer times, the manner of the prayer, the wudu, the spoils, the taxes, and the blood prices were laid down; the same man later collected and edited the letters of the Prophet ﷺ, a collection that was passed on through his son and appeared in print in the modern age together with the book of Ibn Tulun. Al-Mughirah ibn Shu’bah ﵁ wrote down ahadith on request and sent them to Mu’awiya ﵁, who addressed such written requests to several Companions ﵃, among them Aisha ﵂.

Alongside the administrative line runs a personal one. Muhammad ibn Maslamah ﵁ carried, like Ali ﵁, a written text with his weapon: after his death a sahifa with ahadith was found, fastened to his sword. Sa’d ibn Ubadah ﵁, the leader of the Khazraj, had already mastered the art of writing in the days before Islam and gathered ahadith in a book from which a family member later transmitted. Salman al-Farisi ﵁ wrote down transmissions and sent them to Abu ad-Darda ﵁. Samurah ibn Jundub ﵁ left behind an extensive collection that was passed on by his son Sulaiman, and a part of that work is preserved to this day, incorporated into the Mu’jam al-Kabir of al-Tabarani ﵀. And then there is the almost incidental mention of Shamghun al-Azdi ﵁, of whom the sources say that he was the first to write on both sides of papyrus, sheets that he pressed himself and stitched together. Whoever imagines the first community as a world without a writing culture must reconcile himself with details of this kind: not only was writing being done, the technique of writing itself was being experimented with.

The great names of the transmission are of course present as well. Abdullah ibn Amr ﵁, who had mastered Hebrew and after Yarmuk read books of the People of the Book, with his as-Sahifa as-Sadiqa which was passed on through his grandson Shu’ayb and the latter’s son Amr ibn Shu’ayb to the later collectors; his story has been told at length in an earlier article in this map and is touched on here only as a list entry. Ibn Abbas ﵁, who about a single event could question as many as thirty Companions ﵃, who wrote down what he heard and sometimes deployed his servants for it, who divided his teaching over fixed days per field of knowledge and had a translator interpret for non-Arabic speakers, and whose books after his death rested with his student Kuraib and were then entrusted to Musa ibn Uqbah. Ubay ibn Ka’b ﵁, from whom Rufai’ ibn Mihran transmitted an extensive book of Qur’anic commentary. Jabir ibn Abdullah ﵁, the last Companion ﵁ to die in Medina, compiler of a small book on the pilgrimage, from whom a series of named students built up collections.

And Anas ibn Malik ﵁, who at the age of ten entered the service of the Prophet ﷺ and lived until the end of the first century. From him is transmitted the sentence that could stand as a motto over the whole census, the advice to his sons: we attach no value to the knowledge of one who has not written it down. Hubayra ibn Abd ar-Rahman describes what the teaching of Anas ﵁ looked like in practice: when the people had gathered, he brought out books, handed them to his listeners, and said that he had heard these ahadith from the mouth of the Prophet ﷺ, had written them down, and had read them back to him. Hearing, writing, reading back for verification: in that single description the whole later methodology of the science of hadith already lies in embryo.

The pen as an aid to memory

Two entries in the census deserve a moment of their own, because they connect the writing directly to the Prophet ﷺ himself. The first is that of Abu Shah, a man from Yemen, who in the eighth year of the Hijra heard the address of the Prophet ﷺ at the conquest of Mecca and asked whether that address could be set down in writing for him. The Prophet ﷺ thereupon gave someone the command: write it down for Abu Shah. It is a small scene with a large import, for the request was not refused as improper, and the command came from the mouth of the Prophet ﷺ himself, in public, at one of the most attended events of his life.

The second entry concerns a man of the Ansar whose name the sources do not preserve, but whose complaint every student recognises. He came to the Prophet ﷺ with the lament that his memory was failing him: what he heard, he could not hold on to.

Seek the help of your right hand: use the pen. Whether the man then did so, the sources do not report, and Azami records that uncertainty faithfully. But for the question this article poses, it hardly matters. The pen is here not merely tolerated by the Prophet ﷺ but recommended, as an aid to memory for one who could not carry the living word without help. Whoever sets these two scenes beside the permission granted earlier to Abdullah ibn Amr ﵁ sees three independent moments at which the writing of hadith from the mouth of the Prophet ﷺ himself was allowed or commanded.

Letters, dictations and the dictating old age

The census further contains a whole family of entries in which the written transmission takes the form of a letter. Jarir ibn Abdullah al-Bajali ﵁ wrote ahadith in a letter and sent it to Mu’awiya ﵁. Jabir ibn Samurah ﵁ wrote down transmissions and sent them to Amir ibn Sa’d. Zaid ibn Arqam ﵁ wrote ahadith in a letter to Anas ibn Malik ﵁. Abdullah ibn Abu Awfa ﵁ wrote down transmissions of the Prophet ﷺ on the law of war and sent them to Umar ibn Ubaidullah, while his own scribe Salim made copies of them and transmitted them further. Abdullah ibn az-Zubair ﵁ too received a letter in which legal questions were laid before him and in which ahadith of the Prophet ﷺ were cited. Ad-Dahhak ibn Qais ﵁ wrote a letter that contained transmissions of the Prophet ﷺ, and of ad-Dahhak ibn Sufyan al-Kilabi ﵁ a letter to Umar ﵁ is preserved in which he, against the caliph’s initial judgement, adduced a letter of the Prophet ﷺ on the law of inheritance. A letter is, in this connection, no incidental thing. Whoever puts a hadith in a letter fixes it in a dated, addressed, and verifiable document, and the recipient preserves it in exactly the wording in which it was sent.

Besides this there stands the form one might call the dictating old age. Several Companions ﵃ who lived long were, in their later years, surrounded by students who wrote down what they passed on. Of al-Bara ibn Azib ﵁ the sources report that the students in his teaching wrote down ahadith and sat in rows one behind another, an image that suggests a lecture hall rather than an informal conversation. Wathilah ibn al-Asqa ﵁ dictated ahadith to his students. Shaddad ibn Aus ﵁ dictated transmissions on the way to the pilgrimage to two of his companions. Of Sahl ibn Sa’d ﵁, Abu Hazim set down the transmissions in a collection that was passed on further by the latter’s son. And Itban ibn Malik ﵁ made such an impression on Anas ibn Malik ﵁ with his transmissions that Anas ﵁ instructed his own son to write them down. Even within the family of the Prophet ﷺ the line continues: Hasan ibn Ali ﵁ advised his sons and nephews to write down ahadith, and himself possessed a book.

In the case of Ibn Umar ﵁, the son of the second caliph, the census moreover shows how carefully it handles contradictory impressions. There existed an assumption that he was against the copying of hadith, based on Ibn Jubair’s observation that he asked questions about matters that were written in a booklet. Azami weighs that assumption and establishes that it amounts to no more than an impression of the observer, whereas the documented practice of Ibn Umar ﵁ stands directly against it: he possessed a Kitab of his father Umar ﵁, which his servant and student Nafi’ ﵀ read to him repeatedly, and a series of named persons transmitted from him in writing. When a man asked him to write down all the ahadith for him, he answered that this was too much to write, an answer that concerns the size of the request and not the permissibility of the pen.

Precisely in the case of Anas ﵁ the census also shows its critical reverse side, and it is important not to skip over it. Because he lived so long and had so many students, false books and fabricated transmissions were attributed to him as well, and Azami catalogues the forged copies by name, in the same paragraph in which he documents the genuine ones. This is no apologetics that approves everything old; it is a science that separates the genuine from the false and dares to name both categories. A tradition that registers its own forgeries deserves more trust than a tradition that claims to have none.

The counter-voice: the prohibition reports weighed

Whoever reads only the above might think that the early community was unanimously in favour of writing. It is not so simple, and the census makes no effort to conceal this. There exists a transmission in which the writing down of hadith is discouraged, and its chief transmitter is a Companion ﵁ from this very list: Abu Sa’id al-Khudri ﵁. He refused students who wanted to make written collections from him. And yet the sources note of the same man that he himself copied out ahadith, and that when a legal ruling of Ibn Abbas ﵁ came to his ears which went against a hadith of the Prophet ﷺ, he said to his informant that Ibn Abbas ﵁ should be written to, so that he would not spread these rulings further. The means by which he wished to correct an error was a letter. What he rejected, as his own practice shows, was apparently not the ink itself but a particular use of it.

That same pattern, a reported aversion alongside a documented writing practice, recurs with several names, and Azami dissects it case by case each time. Abu Musa al-Ash’ari ﵁ is said to have rejected writing and even to have erased writings of his students, but of the same Abu Musa ﵁ it is preserved that he wrote down ahadith and sent them to Ibn Abbas ﵁. Ibn Mas’ud ﵁ is said to have erased a book; from himself the statement is transmitted that during the life of the Prophet ﷺ one wrote down only the Qur’an, the tashahhud, and the istikhara prayer, and at the same time his son later brought out a book of which he swore by Allah ﷻ that it was in his father’s hand. In the case of Zaid ibn Thabit ﵁, the scribe of the revelation who besides Arabic also mastered Hebrew, the criticism goes a layer deeper still: the report in which he is supposed to have forbidden the writing of hadith turns out to rest on a transmitter, al-Muttalib ibn Hantab, who never met Zaid ﵁, whereby the chain breaks and the report is invalid. Against that broken report stands hard positive evidence: at the request of Umar ﵁, Zaid ﵁ set down in writing his legal ruling on the inheritance share of grandfathers, and he compiled the first book on the law of inheritance, the fara’id, of which the title page has been preserved to this day in the Mu’jam of al-Tabarani ﵀.

The weightiest dossier is that of Umar ﵁ himself. On the authority of Urwah ﵀ it is related that the second caliph wished to have the ahadith written down and bundled under state patronage, that he consulted the Companions ﵃ and that they unanimously agreed, but that he reconsidered the matter and in the end refrained from it; in a related report he is even said to have demanded by circular that existing collections be burned. Azami lays these three reports side by side and establishes what a chain-critic must establish: all three are mursal, the links between Umar ﵁ and the narrators are unknown, and in one of the chains the name of Ibn Umar ﵁ proves to have been demonstrably inserted in error. But he does not stop at that formal finding, and it is precisely his second argument that deserves attention. Suppose, he writes, that one nevertheless accepts the report as genuine, then it proves precisely the opposite of what the sceptics extract from it: if the Companions ﵃ unanimously agreed to a written compilation, then that whole generation therefore did not regard the writing down of hadith as something that went against the command of the Prophet ﷺ. The prohibition report explodes, the moment one takes it seriously, in the hands of whoever wishes to use it as a weapon.

The historical Umar ﵁ who rises from the documents is moreover no enemy of the pen. He introduced the registers into the administration, asked the governor of Kufa to set certain poetry in writing, cited ahadith of the Prophet ﷺ in his official letters, and collected documents of the Prophet ﷺ on financial matters which were later in the possession of his son and which Nafi’ ﵀ read to him several times. That he at the same time called three Companions ﵃ to account because in his view they spread too many ahadith fits the picture of an administrator who guarded the priority of the Qur’an; tellingly enough, he then himself sent two of these three, Ibn Mas’ud ﵁ and Abu ad-Darda ﵁, as teachers to Kufa and Damascus. Whoever truly held that the Sunnah must not be passed on does not send its bearers off to the provinces as teachers.

Thus the apparent contradiction dissolves into a nuanced but coherent picture. There was in the first generation a genuine debate about writing, borne by an understandable concern: that sheets of hadith might become confused with the Qur’an, or that attention would be drawn away from the Book so long as the community was still occupied with learning and spreading that Book itself. That debate has been treated at length earlier in this map in connection with the permission the Prophet ﷺ gave to Abdullah ibn Amr ﵁. The census adds to it the empirical test: when one runs through the actual practice of the generation, name by name, then the writing turns out to be widespread, and the prohibition reports turn out, one by one, to be weak in chain, revoked by the practice of the person concerned, or, on closer reading, in fact a proof in favour of the acceptability of writing.

What fifty names prove

Why is this sum now so devastating for the sceptical thesis? Because the force of a census lies in its spread, not in one spectacular document. A single sahifa can be explained away as an exception. But the fifty entries spread across all layers of the community: caliphs and administrators, warriors and servants, early converts and young Medinans, men and women, Mecca and Medina, and through letters also Bahrain, Sijistan, Najran, and Damascus. They spread, moreover, across all the forms that written transmission can take: the personal notebook, the administrative document, the letter containing transmissions, the dictated dictation, the family possession that comes to light after someone’s death. A forger who had wished to construct this picture after the fact would have had to sow hundreds of independent entries in dozens of independent sources, scattered across biographical dictionaries, legal literature, and hadith collections that contradict one another on countless other points, and he would in addition have had to build in the weak spots as well: the false Abu Bakr transmission that al-Dhahabi ﵀ unmasked, the broken chain beneath the Zaid report, the forged Anas copies. Forgers do not build evidence against themselves.

There is a second layer to the argument, quieter but carrying deeper. Several of these documents are not only mentioned but traceable in their later working-out, and sometimes even physically preserved. The part of the book of Samurah ﵁ in the Mu’jam al-Kabir. The title page of Zaid’s ﵁ book on the fara’id. The letters of the Prophet ﷺ that Amr ibn Hazm ﵁ edited and that appeared in print in the modern age. The content of the Sahifa as-Sadiqa, which lies embedded through the chain of Amr ibn Shu’ayb in the canonical collections. The line from the first generation to the later books is therefore no postulated bridge over a dark century; it is underpinned with documents in several places, and every following article in this map will illuminate that line further.

Finally, here too the measure must be kept, as Azami himself kept it. The census does not prove that the whole of the hadith literature stood in writing in the first generation; by far the greater part of the teaching was and remained oral, borne by memory and by repetition, and the written documents functioned as anchor and control alongside that living teaching. Nor does it prove that each of the fifty entries has the same hardness; the catalogue itself notes where an attribution is uncertain, as in the case of the family collection of Abu Ayyub ﵁. What it does prove is precisely what the sceptical thesis needed to deny: that the writing of hadith in the generation of the Companions ﵃ was an ordinary, widespread practice, permitted by the Prophet ﷺ himself. The dark room of the oral century never existed.

From the list to the life

A catalogue shows breadth, but no depth. Fifty entries tell that writing was being done; they do not yet tell how an individual transmitter lived, taught, and passed on his material to hundreds of students, and how out of that teaching there arose checkable, mutually confirming chains. For that one must lift a single name from the list and examine it up close.

In the next article we do that with the man whose name carries the most appendices in the census: Abu Hurayra ﵁, who spent net around three years in the nearness of the Prophet ﷺ and nevertheless became the most documented transmitter in history. We shall see how the classical scholars counted and recounted his corpus, which named students set his transmissions in writing, and why it was precisely around this Companion ﵁ that the attacks of the sceptics were the fiercest and the defence the best documented. The census gave us the map; now we follow one road upon it.