The Commission of Uthman ﵁

Four Men in One Room

In the previous article, Hudayfah ibn al-Yaman ﵁ rode from the northern frontier to Medina with his cry of alarm, and we closed at the threshold of Uthman’s ﵁ house. On the strength of that alarm the suhuf of Abu Bakr ﵁ had been recovered from the chamber of Hafsa ﵂, and the men whom the Caliph had summoned had entered. What followed in 25 AH in Medina was no longer a meeting. It was the founding of a commission whose work would stand for fourteen centuries.

Four names are given in the oldest tradition, recorded by Imam al-Bukhari ﵀ in his Sahih, hadith number 4987.

Zayd ibn Thabit ﵁, the first. He was the continuity. He had recorded the revelations in the presence of the Prophet ﷺ in the final years in Medina. He had assembled the suhuf under Abu Bakr ﵁, after Yamama, when the qurra’ had fallen in ranks and the community realized for the first time that the Quran also had to be set down in writing. He had been a man in his early twenties then; now, in 25 AH, he was seasoned, experienced, and the only Ansari on the new commission. He was the bridge between what the Prophet ﷺ had dictated, what Abu Bakr ﵁ had had gathered, and what Uthman ﵁ would now have distributed.

Abdullah ibn al-Zubayr ﵁, the second. Son of Asma bint Abi Bakr ﵂ and al-Zubayr ibn al-Awwam ﵁, grandson of Abu Bakr ﵁. The first Muslim baby born in Medina after the Hijra. He had grown up in the heart of the Prophetic community; he was Quraishi by birth, Quraishi in dialect, Quraishi in lineage.

Sa’id ibn al-As ﵁, the third. A man of Banu Umayya, the same clan as the Caliph himself. Sa’id ﵁ was renowned for his eloquence. Later sources note that his Arabic stood closest to the dialect the Prophet ﷺ had spoken; one would say that his dialect was a mirror of the way Mecca had spoken in the years before the revelation.

Abd ar-Rahman ibn al-Harith ibn Hisham ﵁, the fourth. Born into the same clan as Abu Jahl. His father had converted at the fall of Mecca, and his father was a man whose speech was acknowledged as pure Quraishi. Abd ar-Rahman ﵁ inherited that same dialect, that heritage of sounds which set the Quraish apart from the ten other tribes.

Zayd ﵁ in the middle, three Quraishi linguists around him. One Ansari, three Quraish. Not from political calculation, not from clan preference, but from one deliberate reason. The Quran had not simply been revealed in Arabic. It had been revealed in a specific Arabic dialect. And about that dialect, in the definitive written edition, no doubt whatsoever was to remain.

The choice of these four was, in all the modesty of the tradition, a masterpiece of administrative wisdom. Uthman ﵁ could have chosen others. He could have appointed Ubayy ibn Ka’b ﵁, the man whom the Prophet ﷺ himself had called sayyid al-qurra’. He could have summoned Abdullah ibn Mas’ud ﵁ from Kufa, the man who said he had received more than seventy suras directly from the lips of the Messenger ﷺ. He could have appointed Ali ibn Abi Talib ﵁, his son-in-law and the man with the deepest juridical knowledge among the younger generation. Yet the Caliph chose Zayd ﵁ and the three Quraish.

Why. The answer lies hidden in the nature of the work itself. Whoever had bundled the Quran since Abu Bakr’s ﵁ time carried a knowledge no one else could match: the peculiarities of the parchment, the order in which the pieces had passed through Medina, the places where palm branch and shoulder blade and smoothed stone alternated within the treasure of suhuf. That was craft knowledge. And to test that craft knowledge against the Quraishi dialect, the Caliph chose three men who as children had already listened to how their uncles and grandfathers spoke in the hills around Mecca, before the pronunciation of the capital had been broadened by the influx of bedouin and the mingling with the Ansar and the Muhajirun in Medina.

Zayd ﵁ knew what stood written. The three Quraish knew how it had originally sounded. Together they formed an instrument on which the text could be tuned without deviating a single note from the original.

The Charge of Uthman ﵁

When Uthman ﵁ had become Caliph, after the murder of Umar ﵁ in 23 AH, he had asked for the suhuf of Hafsa ﵂ and taken them into his keeping. Not to hide them. To protect them. They were the official documents of the community, the work of Zayd ﵁ and Umar ﵁ and Abu Bakr ﵁ together, and they belonged to the collective.

When Hudhayfa ibn al-Yaman ﵁ returned from the north with the disturbing report that Muslims in Armenia and Azerbaijan were reciting the Quran in different dialects and quarreling with one another over it, the time had come to use the suhuf in full openness.

Uthman ﵁ called his four men together. He gave them a single instruction, and that instruction is handed down in the same wording by Sahih al-Bukhari and by Ibn Abi Dawud in his al-Masahif.

In one sentence lay the whole methodology. The Caliph gave no theological ruling, no exegetical preference, no interpretive directive. He gave a linguistic norm. When Zayd ﵁, the Ansari from Medina, and the three Quraish disagreed, the winner would not be the majority, not seniority, not position in the tribal hierarchy. What would win: the dialect in which the revelation had come down.

The question was never which verses would be set down in writing. There was no disagreement about that. All four men were huffaz; they knew the Quran by heart in its entirety. The question was: when a verse could be recited in several legitimate pronunciations, which pronunciation would the official written edition carry.

The answer was: the Quraishi pronunciation. Not because Quraish was better. Not because Quraish was nobler. But because the Quraishi dialect was the dialect in which the Book itself had come down.

This detail remains, in most popular accounts, undiscussed. And yet it is crucial. The commission of Uthman ﵁ did not write a new Quran. It did not gather materials that had never before been brought together. It chose, with deliberate humility before the original revelation, one of the officially permitted ways of setting that revelation down in writing.

Umar ﵁ had already anticipated that linguistic norm. When he earlier received a letter from Kufa reporting that Ibn Mas’ud ﵁ was reciting there in the dialect of Hudhayl, not in that of Quraish, Umar ﵁ wrote a rebuke back to Ibn Mas’ud ﵁ in his own hand. The message, recorded by Abu Dawud, was that the Quran had come down in the Quraishi dialect, and that teachers were to instruct the people in that, not in the dialect of Hudhayl. Not because the other dialects were invalid in the original revelation, but because unity in the written and publicly taught recitation was now becoming a necessity for an umma that stretched from Egypt to Khorasan.

What under Umar ﵁ had been a correspondence became under Uthman ﵁ an institution. The letter was now a commission. The warning was now a method. And the three Quraishi men who sat around Zayd ﵁ carried that institution personally in their voice.

The Methodology

Here the working method of the commission becomes visible in its full sobriety. Al-Azami ﵀ sets out the procedure point by point, comparable to the way Zayd ﵁ had worked ten years earlier under Abu Bakr ﵁, but now with the experience of that first compilation as a backbone.

First, verse by verse was compared with the suhuf of Abu Bakr ﵁. No verse was taken into the definitive mushaf that had not first been found again in the suhuf, and that was not then confirmed once more by the community of huffaz present in Medina. The suhuf was the principal source; the living reciters were the second witness.

The commission did not sit alone in a closed room. It worked in the open, in the midst of the community of Medina, where in 25 AH there were still hundreds of Companions ﵃ present. Ibn Sirin, one of the elder tabi’in, hands down in a second report that Uthman ﵁ had entrusted the work to a wider college of twelve men, composed of both Quraish and Ansar, to put to the test everything that was set down. Whether one now reads those twelve as a broader circle alongside the four, or as a second round of verification for added certainty, in either case the report points to the same thing: the work did not rest with one hand, but with a shared college. In that wider circle we read names such as Ubayy ibn Ka’b ﵁, Anas ibn Malik ﵁, and Abdullah ibn Abbas ﵁. Men who knew the Quran as a father knows the face of his son.

The crucial distinction to which Al-Azami emphatically draws attention is this. What Uthman ﵁ had carried out was not a second collection. Zayd ﵁ did not go around again with the call “whoever possesses verses, let him bring them”. That call had been made under Abu Bakr ﵁, in the hours after Yamama. The work of those years already lay completed in the suhuf. What the commission of Uthman ﵁ did was a re-verification. A second capstone. A check of the already verified against the living memory of a generation still living.

When Zayd ﵁ looks back afterwards, he uses in the tradition from Sahih al-Bukhari the plural “we”, where under Abu Bakr ﵁ he had used the first person singular “I”. That detail in the verb form is no coincidence. The first compilation had been the heavy personal charge of one man, assisted by a few others. The second was teamwork; a communal project in which the entire Muslim world in Medina worked together on one shared edition.

In the course of the work one familiar difficulty surfaced again. Zayd ﵁ mentions in the tradition in Sahih al-Bukhari hadith 4988 that, while he was assembling the mushaf for Uthman ﵁, he missed one verse from Sura al-Ahzab that he had formerly heard from the mouth of the Prophet ﷺ. He knew the verse by heart. He knew exactly where it belonged. But memory alone was, even for him, no criterion of admission. He set out to search, asked around, consulted the Muhajirun, consulted the Ansar, until he found the written version with Khuzayma ibn Thabit al-Ansari ﵁. Only then was the verse taken into the mushaf.

With that, Zayd ﵁ repeated in the time of Uthman ﵁ what he had also done in the time of Abu Bakr ﵁. Nothing admitted on the strength of his own memory, nothing admitted on the strength of recitation from memory by a single Companion ﵁, nothing admitted without supporting parchment that had been written in the presence of the Prophet ﷺ. A scholarly discipline to which modern textual critics would reach back centuries later.

At the same time the commission members, according to Malik ibn Abi Amir al-Asbahi ﵁ who was himself among them, worked not in one closed group but in several small teams. Some dictated, others wrote; some compared, others checked. When a spelling uncertainty arose over a particular word, such as at-tabut, or a divergence surfaced across three different suras, the problem was not solved indoors by a vote. One of the commission members carried the three shoulder blades with the three different spellings to Ubayy ibn Ka’b ﵁ outside the commission, and personally asked him to write a revision. The definitive mushaf would carry the spelling that Ubayy ﵁ had recommended.

So the commission worked. Not hierarchically, not despotically, not in secret. But according to a humble methodology in which every doubtful case was carried outward to be confirmed by a wider college of qurra’.

The Rumor of the “Altered” Verses

Centuries later a persistent story would circulate in orientalist circles: that Uthman ﵁ had struck out verses, that he had abridged the Quran, that through his commission he had had a polished political version assembled which deviated from what the Prophet ﷺ had dictated. Al-Azami treats this accusation with the patient precision of a scholar who knows the evidence.

The commission removed no verse. It chose, out of the seven legitimate ahruf in which the Quran had been revealed, one rasm, one consonantal skeleton, which at the same time left room for different pronunciations. What they could not accommodate within that skeleton was not lost; it lived on in the oral tradition and would later be codified as the qira’at, the officially recognized recitations.

The accusation breaks on one fact. Had Uthman ﵁ removed even a single verse, one of the hundreds of living Companions ﵃ in Medina would have protested. A man like Ali ﵁, a man like Ibn Mas’ud ﵁, a man like Ubayy ﵁, a woman like Aisha ﵂. None of them would have chosen silence over the mutilation of the Book. And the traditions show the opposite. When Uthman’s ﵁ work was completed, Ali ﵁ himself would say, according to the tradition in Ibn Abi Dawud:

Such unanimity of Companions ﵃ cannot be forged. It was the generation that had seen, heard, recited the Prophet ﷺ. A generation that had, more than anyone on earth, an interest in the purity of the Book they had received from his lips. That they kept silent was no indifference. It was assent.

A second safeguard lay in the nature of the content of those supposedly removed verses. Suppose, for the duration of a thought experiment, that an invented verse had stood somewhere in the Quran. What motive could have driven the commission to erase it. Al-Azami poses this question with the same sobriety with which he has answered hadith scholars across the centuries. A conspiracy of removal would only be conceivable if such a verse contained a theological change, a tribal panegyric, or a political pronouncement that a group wished to suppress. But the Quran itself, in its definitive form, is so balanced about the tribes, so restrained in naming living names, so theologically consistent with all its other passages, that no conceivable advantage can be gained from a fabricated removal. Who invents a crime from which no one profits.

When Uthman ﵁ was confronted with this accusation in later years in Egypt and elsewhere, he answered in a sermon, recorded by Ibn Asakir in his History of Damascus:

No secret room, no hidden striking-out, no Caliphal sole rule over the revelation. A public call, with an oath among the donors, that every piece of parchment brought in had truly been recorded from the mouth of the Prophet ﷺ and not from memory alone.

The Script Without Dots

Here comes a detail that at first sight confuses the modern reader and on closer reflection proves to be a masterly decision. The mushaf that the commission had written bore no diacritical marks. No dots above or below the consonants to distinguish the letters from one another. No vowel signs to fix the pronunciation. A rasm of pure consonantal skeletons, in the Hijazi script, austere and bare.

To one who knows nothing of this, that sounds like a shortcoming. How can a text without vowels be unambiguous. How can a reader know whether a letter represents a ba, a ta, or a tha when the three without dots are identical.

The answer is that the mushaf was never read alone. That was never the intention. The written copy was a chassis. The engine was the oral tradition, the living chains of qurra’ who recited the Quran as they had received it from the Prophet ﷺ, in an unbroken chain from mouth to ear, from mouth to ear. The mushaf froze the skeleton; the reciter carried the flesh.

And precisely there lay the beauty of the decision. By choosing a dotless and vowelless script, one and the same written copy could at the same time accommodate several legitimate pronunciations. A Quraishi reciter from Mecca, a Tamimi reciter from Basra, a Yemeni reciter in Sana’a, all of them could read the same consonantal skeleton aloud to their pupils according to the qira’a they had inherited from a chain of teachers. None of those qira’at was arbitrary. Every valid pronunciation required an unbroken chain up to the Prophet ﷺ himself. But the written text did not need to privilege any one of those pronunciations above the others.

The choice of the bare Hijazi script was thus no embarrassment of the material or an immaturity of Arabic orthography. It was a deliberate, intelligent invitation to plurality within authenticity. In later articles we will follow how Abu al-Aswad ad-Du’ali ﵁ in the years under Ali ﵁ would introduce the first vowel signs, and how al-Hajjaj’s contemporaries would afterwards standardize the diacritical dots. But that was work for a later generation and a wider public, not for the original commission in Medina.

A specimen from Yemen, photographed in Al-Azami’s work, shows a page from an early Hijazi mushaf. No dot, no vowel stroke, only the dry consonantal line stretching with its thin tails across the parchment. One who reads modern Arabic sees the writing, with some effort, legible but without the certainty of naming each letter unambiguously. One who already knows the text by heart, however, sees more than a text. He sees a confirmation. The written rasm confirmed what the heart already carried; it did not replace it.

That is why, alongside every mushaf sent to the provinces, a qari ﵁ had to go. A mushaf without a reciter was, in the strategic calculation of Uthman ﵁, an unfinished product. The object could be seen; the pronunciation had to be learned. We will follow those qurra’ by name in the next article, when we leave Medina together with the five copies. But already here it must be established that the Caliph never regarded the script and the oral transmission as two separate technologies. They were one instrument with two strings.

The Rasm as Anchor

Later scholars, from the second and third centuries AH onward, elevated the rasm Uthmani into a science of its own. Imam ad-Dani in his Al-Muqni’ and Imam ash-Shatibi in his 'Aqilah described word by word how each Arabic word was spelled in the Uthmanic mushaf: where an alif was written and where not, where a waw was added after a hamza, where a ya was elided. To this day, in every Mushaf printed to the Madani standard or to the Indo-Pakistani standard, that rasm remains the sacred norm. A copyist may not apply the spelling of a present-day newspaper in the Mushaf; he follows the script that was fixed in 25 AH under Zayd ﵁ and his three companions.

That is a remarkable continuity. For fourteen centuries the Muslim community has bound itself to an orthography older than any living writing system in Europe. And it has done so not out of conservatism, but out of humility: because in every irregularity of that rasm a secret of the original revelation is concealed, a deviation that accounts for a legitimate variant pronunciation which the modern reader may no longer hear, but which the hafid has indeed received from his master in an unbroken chain of teachers.

The Assembly with the Huffaz

When the commission had completed its work, there took place what al-Bukhari ﵀ and Ibn Kathir describe as one of the most crucial moments in the textual history of the Quran. Uthman ﵁ had the definitive mushaf read aloud publicly to the assembled Companions ﵃.

Picture that moment. The Mosque of the Prophet ﷺ. The sun of Medina on the palm trunks that still stand from his time. Men who have heard the revelation come down upon living lips. Aisha ﵂ in her chamber, a few meters away. Ali ﵁ present. Ibn Abbas ﵁, a young man still, but by now a scholar. Anas ibn Malik ﵁, the servant of the Prophetic household. Ubayy ibn Ka’b ﵁, sayyid al-qurra’, the master of the reciters. Talha, Zubayr, Sa’d ibn Abi Waqqas ﵁. Men whose whole lives had been shaped by this Book.

Zayd ﵁ reads. Or one of his colleagues reads. Verse by verse, sura by sura, from Al-Fatiha to An-Nas. The assembly listens. An error would be flagged. A slip of the pen would be reported. An omission would be noticed; these men knew every verse by heart.

And then, after hours, after days, the silence falls. No dissenting voice. No correction. No protest.

That silence is not an emptiness. That silence is a sealing. It is the crucial ijazah of an entire generation of eyewitnesses upon one written copy. Mus’ab ibn Sa’d would later declare that the people were content with Uthman’s ﵁ decision, and that in any case no one objected to it. Had there been even one dissenting voice, history would have noted it carefully. The Islamic tradition-culture was sharp to the point of discomfort; dissenting opinions were not suppressed, they were preserved, cited, refuted, framed.

That such a dissenting voice is absent is in itself a historical demonstration. Against the whole contestable orthodoxy of orientalist skepticism stands here one fact, simple and unchangeable: no Companion ﵁ protested against the commission of Uthman ﵁. Not one. Not over one verse. Not over one word.

It is sometimes argued that Ibn Mas’ud ﵁ in Kufa did indeed have a disagreement with Uthman ﵁ over the mushaf. That is true, but the disagreement was not about the content of the text, not about the verses, not about the suras. It was about the procedure, about the question whether his own personal copy had to be surrendered, and about the sensitive matter of who was appointed as chief compiler. On the content of Uthman’s ﵁ mushaf, Ibn Mas’ud ﵁ had no textual criticism; everything in that mushaf he knew equally well from his own experience of the revelation. How his objection ultimately resolved itself remains a subject of discussion in the sources. Part of the later traditions, including reports that Ibn Hajar discusses in Fath al-Bari, holds that Ibn Mas’ud ﵁ finally submitted to the decision of the Caliph; other readings sketch a more lasting reservation. Where the reports do agree is that the dispute concerned the procedure and not the text itself.

What the commission of Uthman ﵁ received from its generation was thus no forced submission, but an obtained consensus. Some objections were talked away, others were settled in friendship. But the end product was accepted by all the still-living eyewitnesses of the revelation, and in that acceptance lay, and still lies, the highest form of textual authentication available to humanity.

Five Copies Are Made

When the definitive mushaf had been read aloud, verified, and approved, Uthman ﵁ gave the order to make duplicates. The scribes, skilled copyists from Medina, set to work. On even parchment, in the same Hijazi script, in the same austere dotless rasm.

How many copies were made. Here the old sources differ, and Al-Azami renders the range honestly without elevating any one number to established fact. The tradition he himself favors, that of ad-Dani, counts four copies for dispatch, to Kufa, Basra, Damascus, and Mecca, with in addition one mother copy that remained in Medina; together five. Other reports reach higher. Ibn Kathir adds Bahrain and Yemen and arrives at seven; Shauqi Daif speaks of eight; al-Ya’qubi, a Shi’ite historian of later date, goes up to nine, with Egypt, Damascus, al-Jazira, and the remaining cities included. The figures thus range from five to nine, and no one can say with absolute certainty which number is the correct one.

For the narrative of these articles we hold, following ad-Dani’s preferred report and in step with the next article, to the most cited count of five. Four for the provinces: Kufa, Basra, Damascus, Mecca. One for Medina itself, as reference and as mother copy, kept under the rule of the Caliph in the Mosque of the Prophet ﷺ.

Five parchment books, written by the same skilled copyists from Medina, in one and the same school of orthography, verified by the same commission. Let one not confuse those copyists with the twelve; the twelve were the wider college of verification that Ibn Sirin mentions, the men who put the work of the original four to the test. The actual copyists were the four whom Uthman ﵁ had convened, with Zayd ﵁ in the middle. What is being prepared here is no administrative measure. It is, in no exaggeration, the origin of the first truly international textual standard in the history of human civilization. Before the printing press, before mass production, before modern philology, there already existed here a handwritten canon that would set out in identical form from Mecca to Damascus, from Basra to Kufa.

Fourteen centuries later we can still compare that text. The work of Al-Azami himself, comparison by comparison, mushaf by mushaf, shows that between the copy Uthman ﵁ had signed in Kufa, Basra, and Damascus and the copy printed today in Medina by the King Fahd Complex, there exists no single textual divergence that touches the meaning. A few isolated letters in 9000 lines, variants of a waw or an alif, with no effect whatsoever on the sense. That is no coincidence. That is the legacy of that one commission, in that one room, in that one year.

It is at this point that a brief comparison with the history of other sacred scriptures is in place, not out of polemical zeal, but out of an honest wish to do justice to the uniqueness of what happened here. The comparison is not scientific, it is historical. Of no other religious book from the first centuries of its origin do we possess a verified, signed, publicly community-approved standardization within twenty-five years of the death of the prophet who brought it. No Torah, no Gospel, no Avesta, no Bhagavad Gita knows such a moment. The commission of Uthman ﵁ stands alone in world history. Not because the Muslims were so clever, but because the Prophet ﷺ had left his community a method in which the oral and the written tradition worked together and guarded each other from the first revelation to the definitive canon.

The Silent Week After the Final Reading

We know little of what exactly happened in Medina in the week after the definitive final reading. No chronicler immortalized that moment in detail. But we can imagine what it must have been. The qurra’ who had always compared their copies with each other’s work now compared them with the same mushaf. The pilgrims who came to Medina in the season of Hajj heard for the first time recitations from a text that would be identical everywhere in the umma. The youngest pupils, children of Companions ﵃ who had never themselves seen the Prophet ﷺ, were allowed to touch a written Book that had no rival anymore.

What was sealed in that week was not only a text, but the unity of a community around that text. The Caliph had deployed no army, issued no law, imposed no sanctions; he had set a commission to work in the full sight of Medina and had it tested by a wider circle of qurra’, until the Companions ﵃ who had heard the revelation themselves had nothing more to add to it and fell silent. That is, for all the political difficulty that would follow in the later years of his rule, one of the most pure acts of leadership that history knows.

The Return of the Suhuf

When the five copies were ready, there took place what al-Bara ibn Azib ﵁ mentions so carefully in the tradition in Ibn Abi Dawud. Uthman ﵁ sent the suhuf, the original compilation of Abu Bakr ﵁ and Umar ﵁, back to Hafsa ﵂.

There is a moving symmetry there. Hafsa ﵂, daughter of Umar ﵁, widow of the Prophet ﷺ, had been the official keeper. When Umar ﵁ was stabbed to death by the hand of Abu Lu’lu’a in 23 AH, he had entrusted the suhuf to his daughter. Not because she was his daughter, but because she was the beloved widow of the Prophet ﷺ, literate, trustworthy, and herself a hafida of great authority within the community.

Uthman ﵁ had earlier asked her for the suhuf out of her keeping, with the express promise to return them undamaged. When she handed him the packet of parchments, he had sworn an oath. Now, with the commission dissolved and the copies ready, he fulfilled that oath. The suhuf were brought back to Hafsa’s ﵂ chamber. There they would remain until her death in 45 AH; afterwards, in a less happy chapter, they would be destroyed on the order of Marwan ibn al-Hakam, so that no difference of rasm would later cause discord. But that belongs to another article.

For now everything stands in its place. In one room in Medina lie five identical mushafs. In another room in Medina rest the original suhuf, returned to their official guardian. The community has its definitive written Quran. The commission has completed its charge. Four men, with the help of a larger circle of Companions ﵃, have worked in silence on a miracle: a book of which every letter is verified against the living memory of those who received the revelation, and of which five copies lie ready to set out into the world.

There is something moving in the silence of those last days of the commission’s work. No feast, no proclamation, no Caliphal ceremony. The four men lay down the pen. The five parchment books are carefully rolled up and marked. The Caliph, an old man by now, surrounded by his wife Naila and his closest confidants, knows that in this quiet act he has secured the most precious legacy of the Prophet ﷺ for the centuries still to come. It is a deed whose real magnitude would only be understood later, generations after his martyrdom. Here in 25 AH, in a room in Medina, it looked almost like routine.

In the next article we will travel with those copies. Four of them will leave Medina, each with its own qari ﵁ as companion, on their way to Kufa, Basra, Damascus, and Mecca. One will remain in Medina. And the Caliph will send along an instruction that will seal the unity of the Book for centuries: all deviating copies that do not agree with these five are to be destroyed. Not out of harshness, not out of despotism, but out of love for the unity of the umma around one Book, recited in many tongues, but sealed in one script.