fbpixel

The Origins of the Variant Readings of the Qur'an


Published: August 23, 2019 • Updated: July 22, 2024

Authors: Ammar Khatib and Dr. Nazir Khan

بِسْمِ اللهِ الرَّحْمٰنِ الرَّحِيْمِ

In the name of God, the Most Gracious, the Most Merciful.

Note

This article has been updated on June 2nd, 2023.

Introduction

The Qur’an is the literal word of God, the sacred scripture of the Islamic faith, and God’s final revelation to humanity. This single text produced the largest and most diverse civilization ever to exist on Earth, and for one and a half millennia it has been recited, memorized, and practiced by billions of human beings across the globe. Entire libraries of books have been devoted to the study of the text’s revelation, preservation, recitation, and interpretation. Yet, one aspect of the Qur’an that continues to astound and puzzle researchers has been the fact that Qur’anic verses are recited in diverse ‘modes of recitation’ (qirāʾāt). These different modes utilize different rules (termed uṣūl) regarding the prolongation, intonation, and pronunciation of words, in addition to differences in the vowelization or letters of particular words in individual passages in the Qur’an (termed farsh). Thus, the words of the Qur’an can be divided into two categories: those words that can only be read one way (which constitute the majority of the Qur’an), and those words that can be read in multiple ways (which constitute the basis of the qirāʾāt).
The different modes of recitation are named after the most famous early reciter known for teaching that mode, and individuals who master a mode and receive ijāzah (license to teach) in it become part of an unbroken chain of transmission of that mode back to the Prophet ﷺ. While the majority of the Muslim world is accustomed to hearing the Qur’an recited in the mode of ʿĀṣim ibn Abī al-Najūd (d. 127 AH) according to his student Ḥafs ibn Sulaymān (d. 180 AH) (frequently referred to simply as Ḥafṣ ʿan ʿĀṣim), other modes continue to be recited such as that of the Medinan Nāfiʿ (d. 169 AH) (transmitted by his students Qālūn (d. 220 AH) and Warsh (d. 197 AH)), which remains the dominant mode in many regions of North Africa. Specialists in Qur’anic recitation will be familiar with seven or ten canonical modes of recitation. All of these modes of recitation adhere to the muṣḥaf (codex) of the Qur’an compiled under the supervision of the Caliph ʿUthmān (d. 35 AH) in the year 30 AH (650 CE), which was written without diacritics, thus accommodating the variations. The vast majority of these differences are quite subtle, although in certain cases they add nuances in meaning, complementing one another.
Muslims believe that the Qur’an was taught in different ways during the lifetime of the Prophet Muhammad ﷺ, known as different aḥruf (plural of ḥarf)—a concept that will be elucidated further below. The famous ten qirāʾāt studied today represent only a limited assortment of the variations that existed prior to the ʿUthmānic codex. There are a number of reported readings that differ from the ʿUthmānic codex and were recited by companions of the Prophet Muḥammad ﷺ, including ʿAbd Allāh ibn Masʿūd (d. 32 AH), Ibn ʿAbbās (d. 69 AH), ʿAlī ibn Abī Ṭālib (d. 40 AH), Ubayy ibn Kaʿb (d. 30 AH), and ʿĀʾishah (d. 58 AH), among others رضي الله عنهم. These variant readings in a select number of verses have historically been recorded in books of qirāʾāt and classical works of tafsīr (commentary on the Qur’an) and occasionally works of jurisprudence and typically relate to the presence of additional explanatory words or word substitutions. Perhaps one of the most fascinating discoveries of the past century has been the study of ancient Qur’anic manuscripts that demonstrate wordings that precisely match those wordings attributed to the companions in the classical tradition (see below).
The burning questions for both researchers and laity alike are, of course, why the Qur’an happens to be recited differently, and where these differences come from. The prevailing Muslim understanding is quite straightforward: these different readings arose from the instruction of the Prophet Muḥammad ﷺ himself, and constitute a unique feature of the Qur’an that multiplies its eloquence and aesthetic beauty. This traditional narrative shall be elaborated and analyzed in full below. Meanwhile, Western European and American scholarship has experienced considerable debate over the nature of the variant readings and the history of the Qur’anic text. Behnam Sadeghi and Mohsen Goudarzi have categorized the prevailing viewpoints as revisionist, skeptical, and neo-traditionalist. Revisionists, including John Wansbrough, Patricia Crone, and David Powers, believe the Qur’anic text itself was only standardized after ʿUthmān; however, this conclusion requires dismissing the entirety of traditional sources except for a handful of convenient statements—an approach that has been rightly critiqued as “marshaling cherry-picked, decontextualized and misinterpreted reports.” Skeptics remain in doubt concerning both traditional and revisionist narratives. Finally, there are Western scholars who believe the evidence supports key aspects of the traditionalist account, including the historian Michael Cook (a former revisionist) and Harald Motzki. Building upon manuscript evidence from Sadeghi and Goudarzi, in addition to critical analysis of arguments offered by revisionists, Nicolai Sinai has concluded that, given the absence of any compelling evidence to challenge the traditional narrative, and given the presence of considerable data in its support, the default presumption remains the ʿUthmānic recension date of 650 CE or earlier.
Contemporary Muslim scholarship (both Western and Eastern) has responded to recent manuscript evidence and the output of writings from orientalists by defending traditional narratives surrounding the variant readings or by elaborating modified narratives. Muslim responses have outlined different perspectives on the nature of pre-ʿUthmānic variant readings (those attributed to different companions), ʿUthmānic variants (differences between the regional codices ʿUthmān sent to different cities), and post-ʿUthmānic readings (differences between the qirāʾāt traditions). The present article aims to elucidate the perspective of the Islamic tradition on pre-ʿUthmānic variant readings reported from companions in light of the latest scholarship and to explore some of the questions surrounding the origins of these variant readings.
Welcome back!
Bookmark content
Download resources easily
Manage your donations
Track your spiritual growth

The Qur’an during the Prophet’s time

The revelation of the Qur’an began over fourteen hundred years ago when the Angel Gabriel (Jibrīl in Arabic) came to the Prophet Muhammad ﷺ at the cave of Hirāʾ and commanded him to recite. Angel Gabriel was the Holy Spirit who was trusted by God to deliver the words of revelation to the Prophet ﷺ:

Say (O Muhammad) Rūh-ul-Qudus [The Holy Spirit] has brought it (the Qur’an) down from your Lord with truth, that it may make firm and strengthen (the faith of) those who believe and as a guidance and glad tidings to those who have submitted (to Allah as Muslims).

 

The Prophet ﷺ learned the recitation of the Qur’an directly from Angel Gabriel. Ibn ʿAbbās narrated that Gabriel used to meet with the Prophet every night in the month of Ramadan to revise the Qur’an with him. Allah commanded the Prophet to recite the Qur’an to people (29:45, 17:106). The Prophet ﷺ recited the Qur’an in various ways as he was commanded by Allah. Among the large number of Muslims were those who had memorized the Qur’an and had learned the recitation of it directly from the Prophet ﷺ. These well-versed reciters (Qurrāʾ) were then instructed by the Prophet to teach others the recitation of the Qur’an.For instance, the Prophet Muhammad ﷺ sent seventy Medinan companions who were reciters of the Qur’an to various tribes in the incident of Biʾr Maʿūnah. The sacredness of teaching the Qur’an was captured in the Prophet’s statement, “The best of you are those who learned the Qur’an and taught it (to others).” This practice of teaching the recitation of the Qur’anic text became embedded within the culture and the Qur’an was passed on in the same manner, generation after generation. In addition to this oral transmission, the Prophet Muhammad ﷺ had over sixty-five scribes writing down the Qur’an.
We cannot understand the history of the Qur’an during the Prophetic era without an understanding of the centrality of the Qur’anic recitation and memorization in Muslim ritual practice. Perhaps the most critical fallacy of Western academics who enter the arena of Qur’anic scholarship is that they presume that the Qur’an is like the Old or New Testament and consequently rely solely on manuscripts to construct a picture of its transmission and preservation, neglecting the importance of ritual memorization and oral recitation. Indeed, what percentage of Christians have memorized the entire New Testament in Koine Greek and recite it on a daily basis? Meanwhile, practically every Muslim community in the world boasts plentiful Ḥuffāẓ (singular Ḥāfiẓ): those who have memorized the entire Qur’an by heart in Arabic. The Qur’an is recited out loud in daily congregational prayers and from cover to cover during congregational prayers in Ramadan. This unbroken practice of reciting the Qur’an publicly in daily prayers since the time of the Prophet Muhammad ﷺ is one of the reasons why revisionist Western narratives seem so fanciful to Muslim scholars familiar with the lived practice of Islam.

Did the Prophet ﷺ teach multiple readings?

The Qur’an was revealed in the Arabic language. Allah said that the Qur’an was revealed “in the clear Arabic tongue” (26:195). However, the early Muslim community contained people of all backgrounds, young and old, those proficient in Arabic and those unlettered, as well as people from different Arab tribes, with different accents and different dialects. Since learning the Qur’an was the primary means by which the Islamic message itself was learned, practiced, and transmitted in the Prophet’s time, it was essential that learning the Qur’an be facilitated for diverse peoples. Thus, the Qur’anic text was recited in different ways (termed aḥruf) during the lifetime of the Prophet ﷺ. There are several authentic reports that support this fact, transmitted through so many chains from the earliest sources that it reaches the level of mutawātir lafẓī (massively transmitted verbatim). 
Ibn ʿAbbās narrated that the Prophet ﷺ said: “Jibrīl recited the Qur’an to me in one ḥarf. Then I requested him [to read it in another ḥarf] and continued asking him to recite in other aḥruf until he ultimately recited it in seven aḥruf.”In another narration, the Prophet ﷺ says, “‘O Jibrīl! I have been sent to an illiterate nation among whom are the elderly woman, the old man, the boy and the girl, and the man who cannot read a book at all.’ He said: ‘O Muḥammad! Indeed the Qur’an was revealed in seven aḥruf (i.e., seven different ways of reciting).’”
The early Muslim scholar Ibn Qutaybah (d. 276 AH) said:

In order for the Muslims to read the Qur’an easily, the Prophet was commanded to teach the Qur’an in accordance with people’s dialects…and if everyone was to abandon their dialect and what they were accustomed to speaking as a child, as a youth and in their old age, this would have imposed great difficulty and hardship on them…Thus, Allah intended for them ease by allowing some flexibility in the language in the multiplicity of readings.

Thus, the Qur’an was inherently a multiform recitation, with multiple diverse equally valid alternate readings of many verses. This phenomenon was explained explicitly by the Prophet ﷺ himself to the companions, as we see in the famous incident where ʿUmar ibn al-Khaṭṭāb and Hishām ibn Ḥakīm disagreed in their recitation of Surah Al-Furqān. ʿUmar narrates:

I heard Hishām ibn Ḥakīm reciting Sūrah Al-Furqān during the lifetime of Allah's messenger. I listened to his recitation and noticed that he recited in several different ways which the Prophet had not taught me. I was about to jump over him during his prayer, but I was able to contain myself, and when he had completed his prayer, I put his upper garment around his neck and seized him by it and said, “Who taught you this Sūrah which I heard you reciting?” He replied, “The Prophet taught it to me.” I said, “You are wrong, for the Prophet has taught it to me in a different way from yours.”

So I took him to Allah’s Messenger and said “O Messenger of Allah, I heard this individual reciting Sūrah Al-Furqān in a way that you did not teach me, and you have taught me Sūrah Al-Furqān.”

The Prophet said, “O Hishām, recite!” So he recited in the same way as I heard him recite it before. On that Allah's Messenger ﷺ said, “It was revealed to be recited in this way.” Then Allah's Messenger ﷺ said, “Recite, O ʿUmar!” So I recited it as he had taught me. Allah's Messenger ﷺ then said, “It was revealed to be recited in this way.” Allah’s Apostle added, “This Qur’an has been revealed to be recited in seven different aḥruf, so recite it whichever way is easier for you.”

This narration provides explicit proof that the two different readings were taught by the Prophet ﷺ and that it was the Prophet himself who instructed each companion to recite it in the precise manner in which they did.

The nature of the aḥruf

Muslim scholars discussed the meaning of the seven aḥruf. However, they differed in defining the meaning of these seven aḥruf and whether or not those “seven” were intended to be interpreted metaphorically (to indicate multiplicity) or literally. The purpose of this article is not to recapitulate the discussion of the diverse opinions, which is already available in numerous works in Arabic and English.
The viewpoint espoused and elaborated in this article is one that enjoys the support of a vast majority of specialists in Qur’anic sciences, and that is that aḥruf can be explained simply as ways of varying. For example, the difference in words manifests in the following ways:
  • Singularity, duality, plurality, masculinity, and femininity.
  • Taṣrīf al-Afʿāl (Verbal Morphology)—verb tense, form, grammatical person.
  • Iʿrāb (grammatical case endings).
  • Omission, substitution, or addition of words.
  • Word order.
  • Ibdāl (alternation between two consonants or between words).
This list of different types of Qur’anic variants was mentioned by Abū al-Faḍl al-Rāzī (d. 454 AH). Among the many scholars who adopt this opinion are Ibn Qutaybah (d. 276 AH), al-Zarkashī (d. 794 AH), and Ibn al-Jazarī (d. 833 AH). Of course, these scholars differed over the precise categorization of differences and which categories to combine or split in the list of aḥruf. Therefore, it is best not to present such a list as an exclusive or exhaustive categorization but rather to use it simply as potential examples. The proposed examples of variation include phonetic, morphological, syntactic, and semantic variation. Furthermore, one must examine all the qirāʾāt (alternative readings of Qur’an) including non-canonical (shawādh)  qirāʾāt when attempting to explain the exact meaning of the seven aḥruf. Thus, one can say that the seven aḥruf are all the categories of variation to which the differences found within qirāʾāt correspond. In other words, they represent a menu of ingredients from which each qirāʾah selects its profile.
Having said that, we can now examine the beginning of Sūrah al-Furqān in which ʿUmar ibn al-Khaṭṭāb and Hishām ibn Ḥakīm disagreed on the reading of certain verses. We know that they both recited the verses differently but we don’t know exactly what those differences were. Although the canonical qirāʾāt contain no significant differences in the first verse of Sūrat al-Furqān (Qur’anic chapter 25), there are some non-canonical readings in the first verse which may provide some clues to unveil the mystery of the disagreement between ʿUmar and Hishām in Sūrat al-Furqān; these are provided in the table below. In his explanation of this very narration, the scholar Ibn ʿAbd al-Barr (d. 463 AH) also examined different readings of Sūrat al-Furqān to provide insight into the nature of the difference between these two companions’ readings. The examples listed below of different readings also help illustrate the different categories of how the aḥruf vary.
The reading of the ṣaḥābah (companions of the Prophet) was not unified during the lifetime of the Prophet ﷺ. There were differences in the way they recited the Qur’an. This was due to the license of the seven aḥruf as they were taught different readings by the Prophet himself. When the ṣaḥābah disagreed on the reading of certain verses, they sought the guidance of the Prophet ﷺ. The Prophet would listen to each party and he would give his approval of these variants as being divine. This was reported in the previously mentioned incident of ʿUmar with Hishām in addition to several other recorded incidents. The reports of the seven aḥruf clearly indicate that the ṣaḥābah read precisely as they were taught by the Prophet ﷺ.
Ibn Masʿūd narrated a similar incident when he had a disagreement with another companion in reading the Qur’an. The Prophet ﷺ was displeased and he commanded the Muslims to recite the Qur’an according to the ways they were taught. Ibn ʿAbbās reported that the Prophet ﷺ used to teach them tashahhud as he would teach them a sūrah of the Qur’an.
ʿUmar ibn al-Khaṭṭāb said that the reading of the Qur’an is Sunnah and it is transmitted by the first generation to later generations. In other words, one must have learned directly from the Prophet ﷺ or from him through his students and successors. There was nothing left to the discretion of the individual in this regard; it was necessary to recite precisely as one learned.

Codices of companions

The aforementioned diversity in reciting the Qur’an did not pose a problem during the lifetime of the Prophet ﷺ as the companions became accustomed to the concept of seven aḥruf and diverse modes of recitation. The Prophet ﷺ sent different companions to teach the Qur’an to different tribes and communities and the different readings were transmitted. However, as Islam spread to distant lands, disputes began to arise between Muslims reciting according to different modes and dialects and it was precisely this emerging confusion that led the Caliph ‘Uthmān to compile and distribute a copy of the Qur’an to eliminate such confusion.

Hudhayfah ibn al-Yamān came to ʿUthmān at the time when the people of Shām and the people of ʿIrāq were waging war to conquer Armenia and Azerbaijan. Hudhayfah was afraid of their (the people of Shām and Irāq) differences in the recitation of the Qur’an, so he said to ʿUthmān, “O chief of the Believers! Save this nation before they differ about the Book (Qur’an) as Jews and the Christians did before.” So ʿUthmān sent a message to Ḥafṣah saying, “Send us the manuscripts of the Qur’an so that we may compile the Qur’anic materials in perfect copies and return the manuscripts to you.” Ḥafṣah sent it to ʿUthmān. ʿUthmān then ordered Zayd ibn Thābit, ʿAbd Allāh ibn al-Zubayr, Saʿīd ibn al-ʿĀṣ and ʿAbd al-Raḥmān ibn Ḥārith ibn Hishām to rewrite the manuscripts in perfect copies. ʿUthmān said to the three Qurayshī men, “In case you disagree with Zayd ibn Thābit on any point in the Qur’an, then write it in the dialect of Quraysh, the Qur’an was revealed in their tongue.” They did so, and when they had written many copies, ʿUthmān returned the original manuscripts to Ḥafṣah. ʿUthmān sent to every Muslim province one copy of what they had copied, and ordered that all the other Qur’anic materials, whether written in fragmentary manuscripts or whole copies, be burnt.

During the caliphate of ʿUthmān, the codex of the Qur’an was compiled by a designated committee led by Zayd ibn Thābit (d. 45 AH) and was subsequently copied and sent throughout the Muslim world. ʿUthmān ordered all other written copies of the Qur’an to be destroyed or corrected. However, in some regions of the Muslim world, it seems the transition to the ʿUthmānic codex took time, particularly in Kūfah owing to the massive influence of ʿAbd Allāh ibn Masʿūd. The early scholar al-Aʿmash (d. 148 AH) is reported to have said, “I reached Kūfah, and the reading of Zayd was not with them except as the reading of ʿAbd Allāh is with you today: no one recited it except for one or two people.”
When reading classical works of tafsīr, it is common to encounter narrations that mention that a particular verse was recited differently by a companion (most often ʿAbd Allāh ibn Masʿūd), typically with a word substitution, addition, or omission. A number of these variant readings are readily encountered in the canonical books of Hadith and a few are mentioned in Ṣaḥīḥ al-Bukhārī. For instance, when ʿAlqamah ibn Qays (d. 62 AH) traveled to Greater Syria, he met the companion Abū al-Dardāʾ, and the latter asked ʿAlqamah about how Ibn Masʿūd recited Sūrat al-Layl; ʿAlqamah responded that Ibn Masʿūd recited verse 3 as “wa-al-dhakari wa-al-unthá” (and by the male and the female) instead of “wa mā khalaqa al-dhakara wa-al-unthá” (and by that which created the male and the female), whereupon Abū al-Dardāʾ testified that he had learned the verse in the same way from the Prophet ﷺ. This variant is said to represent one of the aḥruf in the recitation of the Qur’an that was no longer recited after the codex of ʿUthmān was established. Similarly, a narration records that Ibn ʿAbbās would recite 18:79-80 as “wa kāna amāmahum malikun yaʾkhudhu kulla safīnatin ṣāliḥatin ghaṣbā. Wa ammā al-ghulāmu fa kāna kāfiran” (translation: and there was ahead of them a king seizing every boat in good condition by force; as for the boy, then he was a disbeliever), thus containing three lexical differences from the ʿUthmānic codex. Another narration records an additional phrase mentioned by Ibn ʿAbbās after 26:214: “wa andhir ʿashīrataka al-aqrabīn, wa rahṭaka minhum al-mukhlaṣīn” (translation: and admonish your nearest kinsmen and your exclusive tribesmen). The understanding of these variants will be elucidated in detail in the following section.

How many variant readings of companions are there that differ from the ʿUthmānic codex?

Many who write on this subject simply cite a few instances of variants without giving any sense of proportion or frequency; a systematic study requires evaluating all the transmitted material. In a comprehensive study of the primary sources undertaken by Muḥammad ʿAbd al-Raḥmān at-Ṭāsān, the author notes a total of 592 instances where a companion’s reading of a verse has been narrated. Of these narrations, 52 are identical to the ʿUthmānic codex, while the remaining 540 are variant readings which differ from the ʿUthmānic codex, thus termed shādhdh. However, of those 540 instances, only 177 are found with a chain of transmission (isnād), and of those with a chain, only 20 are classified as authentic by hadith standards. Thus, the actual quantity of variants traceable to the companions is significantly smaller than may initially be supposed. 
Of particular interest is that studies of ancient manuscripts of the Qur’an have demonstrated the incredible precision and accuracy of the Muslim tradition in faithfully transmitting these variants in circulation amongst the earliest Muslim community. The Ṣanʿāʾ palimpsest is a manuscript where the original writing (termed the lower text) was erased and written over (the upper text), although contents from the undertext can be revealed using ultraviolet light, and when studied demonstrates numerous variants ascribed to companions in the Muslim tradition. Thus, the Muslim tradition erred on the side of caution in preserving and documenting a larger volume of reported variants beyond what was established exclusively through authentic chains of transmission.

Did the Ṣahābah have muṣḥafs containing variant readings?

In his famous work entitled Kitāb al-Maṣāḥif, Ibn Abī Dāwūd (d. 316 AH) has a chapter on the differences between the codices of companions (bāb ikhtilāf maṣāḥif al-ṣahābah) which he begins by explaining his terminology: “We only say the codex of so-and-so (muṣḥaf fulān) for that which differs from our codex (muṣḥaf) in writing, by addition or subtraction. This is what I have taken from my father, who followed the same practice in his work Kitāb al-Tanzīl.” He then proceeds to list narrations of variant readings from companions under chapter titles, “muṣḥaf ʿUmar,” “muṣḥaf ʿAlī,” “muṣḥaf Ibn Masʿūd,” and so on. This terminology may lead to confusion, as the narrations mention that these companions were simply heard reciting a verse in that reading, not that they possessed a physical codex that had the verse written in that reading. As al-Ṭāsān demonstrates, these readings attributed to the companions were initially simply referred to as qirā’ah Ibn Masʿūd or ḥarf Ibn Masʿūd, but later came to be referred to as muṣḥaf Ibn Masʿūd. In the case of many companions, there is no evidence that a personal codex with exclusively variant readings ever existed. Moreover, the early bibliographer Ibn al-Nadīm (d. 380 AH) noted that he personally came across many early muṣḥafs attributed to Ibn Masʿūd but was unable to find any two in agreement (laysa fīhā muṣḥafayn muttafiqayn). Muhammad Mustafa al-Aʿzami comments:

The divergent nature of the many ‘Muṣḥafs of Ibn Masʿūd’ that materialized after his death, with no two in agreement, shows that the wholesale ascription of these to him is erroneous, and the scholars who did so neglected to examine their sources well. Sadly the less scrupulous among antique dealers found it profitable, for the weight of a few silver pieces, to add fake Muṣḥafs of Ibn Masʿūd or Ubayy to their wares.

It is essential, therefore, to take the attribution of such readings to the companions and to personal codices with a grain of salt, as there is little that is verifiable in such reports.

Did these codices contain chapter differences?

Scholars have discussed a number of reports which indicate the absence of Sūrah al-Fātiḥah and al-Mūʿawwidhatayn (chapters 113 and 114) in Ibn Masʿūd’s codex and whether this implies that he did not believe them to be from the Qur’an. For instance, ʿAbd al-Raḥmān ibn Yazīd al-Nakhaʿī (d. 93 AH) states, “I saw ʿAbdullah [ibn Masʿūd] scrape (i.e. erase) al-Muʿawwidhatayn from the muṣḥaf and say, ‘They are not from the Book of Allah’.” Amongst Muslim scholars, there are differing views with respect to addressing these narrations. Some scholars rejected them as inauthentic, while other scholars considered them authentic while excusing Ibn Masʿūd for not being aware.
The first group of scholars, including al-Bāqillānī (d. 403 AH), Ibn Ḥazm (d. 456 AH), Fakhr al-Dīn al-Rāzī (d. 606 AH), al-Nawawī (d. 676 AH), and others, rejected these narrations as inauthentic. Fakhr al-Dīn Al-Rāzī pointed to the mass transmission (tawātur) of these chapters which makes it inconceivable that Ibn Masʿūd could dispute their status. Similarly, al-Nawawī emphasized the unanimous consensus of the Muslim community on the status of these chapters, and that what has been related from Ibn Masʿūd is false and not authentically established. Ibn Ḥazm declared this to be a fabrication against Ibn Masʿūd and one that contradicts the fact that all the canonical readings contain al-Fātiḥah and al-Muʿawwidhatayn, including the reading of ʿĀṣim which is partly transmitted from Zirr ibn Ḥubaysh from Ibn Masʿūd. 
In a lengthy chapter in his al-Intiṣār, al-Bāqillānī advances several considerations to render these narrations logically implausible. There is no record of any companions directly criticizing Ibn Masʿūd for such a controversial view. Moreover, numerous narrations from the Prophet indicate that al-Muʿawwidhatayn are part of the Qur’an. Al-Bāqillānī also suggests that there may have been a misunderstanding as to why Ibn Masʿūd did not write these chapters in his muṣḥaf. It is possible he wished to only write down what he witnessed the Prophet instruct the Muslims to write down. It is also plausible that Ibn Masʿūd chose not to write down al-Fātiḥa and al-Muʿawwidhatayn because they are recited daily and known to every Muslim. 
Indeed, al-Fātiḥah is the most recited chapter of the entire Qur’an, being a component of every unit of prayer and thus recited a minimum of seventeen times a day. Any dispute about its Qur’anic status would have created a massive disruption to daily worship. Ibn Masʿūd himself alluded to this in one narration. When asked why he didn’t write al-Fātiḥah in his muṣḥaf, he stated that were he to write it, he would have written it with every single chapter of the Qur’an. Concerning this statement, Abū Bakr ibn al-Anbārī (d. 328 AH) explained, “He means that the way of recitation in every unit of prayer (rakʿah) is to start with al-Fātiḥah before reciting another sūrah, so [it is as though Ibn Masʿūd] said ‘For brevity, I have sufficed with omitting it and have entrusted it to the Muslims’ preservation/memorization of it.” Ibn ʿAbd al-Kāfī (fl. 400 AH) considered the case of al-muʿawwidhatayn as analogous to al-Fātiḥah in Ibn Masʿūd’s muṣḥaf; he wrote:

“He did not place it (i.e. al-Fātiḥah) in his muṣḥaf because he was not concerned about it being lost due to its shuhra (well-known status) and it being frequently recited during prayer and otherwise. And so in the same manner, he did not place the muʿawwidhatayn in his muṣḥaf, finding no need to do so due to their shuhrah and being frequently recited.

Similarly, al-Māzirī (d. 536 AH) explained, “It can be understood from what is related regarding the omission of al-muʿawwidhatayn from the muṣḥaf of Ibn Masʿūd that he did not consider it required for him to include all the Qur’an, and he wrote that which was other than these two and left them due to their shuhrah (well-known status) with him and with the people.”
The reality is that the ṣaḥābah used their writings of the Qur’an as memory aids for personal worship and recitation, and consequently never intended them as complete official copies of the Qur’an. Thus, al-Zarqānī (d. 1367 AH) writes:

Sometimes the author of a muṣḥaf would leave out a sūrah due to its popularity (shuhrah) and its not needing to be written down due to this popularity (i.e., it was common knowledge), as has been related about Ibn Masʿūd’s muṣḥaf not containing al-Fātiḥah. And sometimes the author of the muṣḥaf would write what he saw himself needing to include from other than the Qur’an in the same muṣḥaf as has preceded regarding al-qunūt al-ḥanafiyyah which it is reported that some of the companions included in their muṣḥaf and named sūrah al-khalʿ wa al-ḥafd.

The second group of scholars accepted the aforementioned narrations concerning Ibn Masʿūd and al-muʿawwidhatayn as authentic. They considered this to be an erroneous position that was excusable on his part due to being uninformed about their status. Sufyān ibn ʿUyaynah (d. 198 AH) states:

“[Ibn Masʿūd] used to see the Prophet use these invocations for al-Ḥasan and al-Ḥusayn and he never heard him recite them in his prayers so he presumed they were simply invocations and remained on this assumption. Meanwhile, others verified that they were in fact from the Qur’an and included them within the text of the codex.

Similarly, Ibn Qutayba (d. 276 AH) states that Ibn Masʿūd thought that al-muʿawwidhatayn were a supplication that the Prophet taught akin to other invocations which were not part of the Qur’an. Ibn Ḥajar al-ʿAsqalānī (d. 852 AH) also held these narrations to be authentic; although the Qur’anic status of these chapters was mutawātir (mass transmitted) in general, this may not have been the case for Ibn Masʿūd who therefore cautiously refrained from declaring them to be chapters of the Qur’an. Ibn Taymiyyah (d. 728 AH) writes:

Some of them [i.e. the companions] used to erase al-muʿawwidhatayn while others would write [as a Qur’anic chapter] al-Qunūt. And this is a mistake that is known by unanimous consensus (ijmāʿ) and mass transmission (al-naql al-mutawātir). In spite of this, so long as this knowledge was not mass transmitted to them [i.e. these companions], they would not have committed disbelief, even though it would be disbelief for someone upon whom the evidence was established by mass transmission.

The contemporary scholar al-Ṭāsān supports this conclusion. Ibn Kathīr (d. 773 AH) states that Ibn Masʿūd perhaps did not hear them from the Prophet or they were not mass transmitted to him, and he may have corrected himself on this point and revised his position later in life. ʿAlī al-Jaʿfarī distinguishes between Ibn Masʿūd’s position towards al-Fātiḥa (which Ibn Masʿūd accepted as a chapter of the Qur’an but refrained from writing since it was widely known) and his position towards al-muʿawwidhatayn (which Ibn Masʿūd erroneously considered Divinely revealed non-Qur’anic invocations); al-Jaʿfarī also notes that Ibn Masʿūd’s own students did not adopt this position. As the ʿUthmānic codex became widely disseminated, no such disagreement remained.

How do we view the variants reported from Companions?

We now arrive at the interesting question of how variant readings from ṣaḥābah (i.e., those which differ from the ʿUthmānic codex) are to be understood theologically, from within the Muslim tradition. Did God speak all of these readings and did Jibrīl recite all of them to Muhammad ﷺ? Are they all considered ‘Qur’an’? Do they all exhibit the miraculous Qur’anic inimitability? Is it permissible to recite them in prayer?
In order to address these questions, let us first distinguish between two concepts: qirāʾah bi-al-talaqqī (recitation based on direct learning) and qirāʾah bi-l-maʿnā (recitation based on paraphrasing the meaning). We believe that the companions’ recitation of the Qur’an was based on the former concept—they took great care to recite the Qur’an exactly as they heard it from the Prophet ﷺ, and when a reciter erred in doing so, he would be corrected by others. We will return later to the opposing concept of qirāʾah bi-al-ma’nā. For now, based on qirāʾah bi-al-talaqqī, there are three possible explanations for a variant reading of a verse authentically established from a companion that differs from what we find in the ʿUthmānic codex:
  1. Abrogated ḥarf
  2. Abandoned ḥarf
  3. Non-Qur’anic tafsīr recital
We will explain these in turn.

Abrogated ḥarf

The concept of Qur’anic abrogation is established in the Qur’an itself in the verse: “We do not abrogate a verse or cause it to be forgotten except that We bring forth [one] better than it or similar to it. Do you not know that Allah is over all things competent?” (Qur’an 2:106). Abrogation may pertain to the legal ruling of a verse, whereby an instruction provided in an initial verse is known to no longer be applicable or effective following the revelation of a later abrogating verse (nāsikh). Alternatively, abrogation can pertain to the recitation of a verse, whereby it is no longer recited as part of the Qur’an, despite being initially revealed by Allah. It was narrated that ʿĀʾishah said: “One of the things that Allah revealed in the Qur’an and then abrogated was that nothing makes marriage prohibited except ten breastfeedings or five well-known (breastfeedings).”
Likewise, another companion al-Barāʾ ibn ʿĀzib mentioned about verse 2:238: “This verse was initially revealed as ‘Guard the prayers and the ʿAṣr prayer.’ We recited it thus for as long as Allah willed. Then Allah abrogated it and it was revealed: ‘Guard the prayers and the middle prayer.’”
The earliest Muslim community was thus exposed to some Qur’anic recitation that Allah in His infinite Divine Wisdom excluded from the composition of the Qur’an that would be recited until the end of time; i.e., the Qur’an was revealed with extra passages no longer found in it today. How then was the exact composition of the Qur’an determined, considering also that verses and sūrahs were revealed in a different order? Every year, the Angel Jibrīl would review the recitation of the Qur’an with the Prophet Muhammad during the month of Ramadan—“Gabriel used to meet him every night during Ramadan to revise the Qur’an with him,” and “Gabriel used to repeat the recitation of the Qur’an with the Prophet ﷺ once a year, but he repeated it twice with him in the year he died.”
According to many scholars, it was during this ‘final review’ (in Arabic called al-ʿarḍah al-akhīrah) in the last year of the Prophet’s life that the wording of the Qur’an was finalized and much of the variants in the aḥruf were abrogated and excluded from the final recitation. The companion Samurah (d. 54 AH) said, “The Qur’an was reviewed with the Prophet several times. So they say that our qirāʾah is al-ʿarḍah al-akhīrah.” The early scholar Muḥammad ibn Sīrīn (d. 110 AH) mentioned the understanding that the ʿUthmānic codex was in conformity with what was recited in al-ʿarḍah al-akhīrah. Ibn al-Jazarī (d. 833 AH) stated, “And the maṣāḥif were written according to the wording that was confirmed in the final review from the Messenger of Allah as was explicitly stated from more than one of the Imams of the salaf such as Muḥammad ibn Sīrīn, ʿUbaydah al-Salmānī, and ʿĀmir al-Shaʿbī.” 
How was it known what the Prophet ﷺ recited with Jibrīl? The Prophet would have instructed his companions accordingly after the final review with Jibrīl. Scholars speak of specific companions who ‘attended’ the final review, in the sense that the companion reviewed the entire Qur’an with the Prophet after he recited with Jibrīl, not that they were physically present while the Prophet was reciting to Jibrīl. al-Baghawī (d. 516 AH) states:

It is said that Zayd ibn Thābit attended the final review in which it was clarified what was abrogated and what remained.

Abū ʿAbd al-Raḥmān al-Sulamī said, “Zayd recited the Qur’an twice to the Prophet during the year in which he passed away, and this recitation is called the qirāʾah of Zayd because he transcribed it for the Prophet and recited it to him and witnessed al-ʿarḍah al-akhīrah, and he taught its recitation to people until he passed away. That is why Abū Bakr and ʿUmar relied upon him in its compilation and ʿUthmān appointed him in charge of writing the maṣāḥif—may God be pleased with them all.”

However, there is no definitive proof about who, if anyone, actually attended the final review with the Prophet ﷺ. Other scholars simply stated that the reading of Zayd was the one confirmed by al-ʿarḍah al-akhīrah. Ibn Taymīyah (d. 728 AH) wrote, “And the final review is the qirāʾah of Zayd ibn Thābit and others, and it is the one that the rightly-guided caliphs—Abū Bakr, ʿUmar, ʿUthmān, and ʿAlī—instructed to be written in the maṣāḥif.”
Nonetheless, the idea that the variant readings from Ibn Masʿūd and Ubayy ibn Kaʿb represent a ḥarf that was originally revealed but subsequently abrogated does not seem to square with the fact that Muslims in the subsequent generation continued to recite according to the ḥarf of Ibn Masʿūd at least in Kūfah, as reported by al-Aʿmash and others. Moreover, abrogation is generally considered a matter that must be decisively established by an injunction from God and His Messenger. Proponents of the abrogation view may counter that the knowledge of its abrogation was not widely known and may have escaped some amongst the companions and the subsequent generation.

Abandoned ḥarf

The second explanation for the existence of these diverse variants from companions also sees them as having come from a ḥarf originally revealed as part of the Qur’an. However, when ʿUthmān compiled the codex it was not necessary for him to include every single ḥarf revealed and so it was left out and eventually became extinct. Indeed, for most reported variants we have no mention that they were abrogated, but simply that they were no longer being recited except by the companion to whom they are attributed. This is seen, for instance, in the previously discussed example of Ibn Masʿūd’s reading of 92:3. 
Many scholars were of the view that the ʿUthmānic codex left out some or all of the other aḥruf because the intention was to reduce the disagreement amongst Muslims. Since the other readings were intended only as a concession, this has no bearing on the preservation of the Qur’an, which can be recited according to any ḥarf. Ibn Jarīr al-Ṭabarī, for example, explicitly argues that the ʿUthmānic codex was based on one ḥarf because the diverse readings were a concession (rukhṣah) and therefore it wasn’t an obligation for Muslims to learn and transmit them all. The opinion that the ʿUthmānic muṣḥaf selected one ḥarf was also the view of al-Naḥḥās (d. 338 AH),  Ibn ʿAbd al-Barr (d. 463 AH), al-Abyārī (d. 616 AH), Ibn al-Qayyim (d. 751 AH), and many other scholars. However, Makkī ibn Abī Ṭālib (d. 437 AH), Ibn al-Jazarī (d. 833 AH), Ibn Ḥajar (d. 852 AH) and other scholars explained that what remained after the ʿUthmānic compilation were the differences from the other aḥruf that could still be accommodated by the skeletal text of the ʿUthmānic codices; so what remained were “some of the differences of the aḥruf, not all of them”. Ibn Ḥajar cites Abū al-ʿAbbās ibn ʿAmmār al-Mahdawī (d. 430 AH) who states, “The most correct position which is upheld by the experts is that what is recited now are some of [the differences] of the seven ḥurūf which were permitted to be recited and not all of them.” The remnants of the aḥruf are thus found amongst the various qirāʾāt recited today, while those aḥruf that did not conform to the ʿUthmānic codex were abandoned.
In reality, the view that the variant readings were abandoned is practically very similar to the view that they were abrogated. It entails that the variant reading was revealed by God, yet because God did not intend for it to be included in the final Qur’an, it became abandoned as per the Divine Decree (qadar) of God. One can say that here abrogation is seen as effectively taking place by God’s Divine Will (irādah kawnīyah) rather than an explicit revealed instruction (irādah sharʿīyah). Divine Will has effectively excluded those variant readings from the muṣḥaf of this ummah, and since history is intended by God, then the Qur’an we have in our hands today is exactly the Qur’an that God wanted us to have, and the loss of variations that did not make it into the muṣḥaf was also intended by God. What the Muslim ummah would collectively agree upon (ijmāʿ), recite, and practice was included in the foreknowledge of God prior to the creation of the universe.
With the passage of time, some variant readings were effectively ruled extinct by Allah’s Divine Decree concerning the consensus of the community, just as if such readings were abrogated by legislation—and this is precisely what some scholars said. Makkī ibn Abī Ṭālib (d. 437 AH) wrote, “As for what it is in our hands of the Qur’an, it is that which conforms to the script of that (ʿUthmānic) muṣḥaf, from those qirāʾāt with which the Qur’an was revealed, and upon which the community unanimously agreed. No longer practiced are those qirāʾāt that differ from the script of the muṣḥaf. So it is as if they were abrogated by the consensus upon the script of the muṣḥaf.”

Non-Qur’anic recital

Many scholars held the view that variant readings reported by companions were often nothing more than a companion’s method of explaining the verse by simplifying the language, adding an explanatory phrase, or substituting a word for a more familiar one. Abū Bakr al-Bāqillānī (d. 404 AH) states, “From amongst them (i.e., the companions) were those who would recite interpretation (taʾwīl) alongside revelation (tanzīl).” This has been termed qirāʾāt tafsīrīyah (exegetical recitations). Abū Ḥayyān al-Andalūsī was of the view that whatever was in conflict with the script of the muṣḥaf was in reality tafsīr, not Qur’anic recitation. Imam al-Tirmidhī comments on a hadith with an added phrase after the verse by saying, “This appears to be a statement of Ibn ʿAbbās (i.e., rather than a variant reading).”
Imām al-Nawawī (d. 676 AH) writes (quoting al-Māzirī (d. 536 AH)):

As for Ibn Masʿūd then much has been narrated from him including that which is not reliably established according to the people of transmission. And that which is established which differs from what we say (i.e., recite in our muṣḥaf), then it is interpreted to mean that he wrote in his muṣḥaf some rulings and tafsīr which he believed to not be Qur’an, and he did not believe that to be impermissible as he saw it as a parchment upon which to write what he willed. While ʿUthmān and the community deemed that to be prohibited lest with the passage of time it be assumed to be Qur’an. al-Māzirī said: So the disagreement goes back to a jurisprudential matter (masʾalah fiqhīyah) and that is whether it is allowed to include commentary interspersed in the muṣḥaf.

Similarly, Ibn al-Jazarī (d. 833 AH) writes, “It was possible that they (i.e., the companions) would include tafsīr in the qirāʾāh, as clarification and elucidation (īḍāḥan wa bayānan). This is because they were well-versed in what they had learned directly from the Prophet ﷺ as Qur’an, so they were secure from confusing between them. And it was possible that some of them would write it (i.e., tafsīr) alongside it (i.e., Qur’an).” 
Many early authorities explicitly considered this to be the explanation for variant readings reported by companions. Abū Bakr ibn al-Anbārī (d. 328 AH) wrote, “As for what is transmitted from the companions or the successors that they recited in this or that manner, that is only from the angle of explanation and clarification (innamā dhālika ʿalā jihatil bayān wa at-tafsīr), not that it was Qur’an being recited.” Abū Jaʿfar al-Naḥḥās (d. 338 AH) wrote, “It is related from Ibn ʿAbbās, ‘And the middle prayer (is) the ʿaṣr prayer.’ And this recitation is a tafsīr since it is an addition to what is in the muṣḥaf.” Furthermore, it is related that ʿAmr ibn Dīnār (d. 126 AH) heard ʿAbd Allāh ibn al-Zubayr recite 3:104 with an additional phrase “and they seek Allah’s help on that which befalls them”; ʿAmr ibn Dīnār said, “And I do not know whether it was his qirāʾah or he was doing tafsīr.” Thus, sometimes those who heard companions recite on rare occasions with a variant reading may have been unsure whether they were reading a non-ʿUthmānic ḥarf or whether it was tafsīr.
Perhaps one of the strongest pieces of evidence in favor of the Qirāʾāt tafsīrīyah viewpoint is simply an empirical examination of the wording found in many of the reported variant readings from the companions. In a survey of several readings attributed to ʿAbd Allāh ibn Masʿūd, one author compared them with the vocabulary used in other relevant passages. The Qur’an uses al-mashy for human movement on the earth, while Ibn Masʿūd’s variant replaces 2:20 with maḍaw fīh. Likewise, the variant reading from Ibn Masʿūd of verse 1:6 replaces hidāyah (guidance) with its near-synonym irshād as follows: “arshidnā al-ṣirāṭ al-mustaqīm” instead of “ihdinā al-ṣirāṭ al-mustaqīm.” However, the word hidāyah and its derivatives are associated with the path in 23 other places in the Qur’an and the straight path in 18, while the word rushd is associated with none and a linguistic analysis of the meanings of the words demonstrates the greater suitability of the word hidāyah. Thus, the author writes, “So it is possible to say that the recitation arshidnā aṣ-ṣirāt al-mustaqīm is not a recitation but rather it is tafsīr, and that is based on an examination of the Qur’anic expressions in its usage of words given that it does not use irshād with the word ṣirāt in contrast to hidāyah.” Thus, Ibn Masʿūd’s variant readings almost invariably seem to substitute Qur’anic terminology with easier to understand near-synonyms or add additional explanatory words that are at variance with the Qur’an’s usual style. This seems to represent a strange coincidence but becomes readily explicable when considered to be qirāʾāt tafsīrīyah. The observation that these variant readings serve an exegetical function did not escape the earliest of scholars—one of the most powerful evidences in favor of this viewpoint is the fact that the student of Ibn ʿAbbās, Mujāhid ibn Jabr al-Makhzūmī (d. 103 AH) stated, “Had I read the qirāʾah of Ibn Masʿūd, I would not have needed to ask Ibn ʿAbbās about much of the Qur’an which I asked about.”
Although this view terms such variant readings as qirāʾāt tafsīrīyah, it should be noted that the non-Qur’anic wording added or substituted by companions in their recital need not always have been strictly for tafsīr. Rather, it would sometimes be word alterations to facilitate pronunciation or memorization. In one incident, Ibn Masʿūd was trying to teach a man to recite verse 44:44, “food of the sinful (ṭaʿāmu al-athīm)” but the man kept repeating “food of the orphan (ṭaʿāmu al-yatīm).” When the man was unable to recite the correct word after multiple attempts, Ibn Masʿūd told him to recite instead using a near-synonym “food of the disobedient (ṭaʿām al-fājir).” If others learned this same “easy version” of the verse, it would be evident that this was not what Ibn Masʿūd believed the actual wording of the Qur’anic verse to be, but rather an alternate reading for the sake of facilitation (taysīr). Al-Qurṭubī (d. 676 AH) explained that this by no means entails that one can substitute words from the Qur’an as one may ignorantly presume, but rather this was a means for ʿAbdullah ibn Masʿūd to make things easier for the student. Al-Māwardī (d. 450 AH) states that Ibn Masʿūd’s intention was to indicate the meaning of the word so that the learner would understand the correct word to recite.

Evaluation of the aforementioned explanations

The explanation which single-handedly appears to encompass all the evidence most easily is the second explanation: that non-ʿUthmānic readings were simply abandoned by the overwhelming consensus and practice of the community. However, in comparing the aforementioned three possibilities, it also becomes apparent that there is no reason why they should be considered mutually exclusive. That is, it is certainly possible that some variant readings were simply tafsīr (such as the word substitutions from Ibn Masʿūd, which clearly differ from Qur’anic style), while other variant readings were revealed Qur’anic aḥruf that subsequently were either abrogated (as al-Barā’ ibn ʿĀzib said for 2:238) or were abandoned following the ʿUthmānic recension (as Abū al-Dardāʾ suggested for 92:3). Abrogation of a reading could have occurred at any point in the Prophet’s life, as there is nothing to definitively establish the theory that al-ʿarḍah al-akhīrah itself was a means of abrogating readings. The contemporary scholar Mufti Taqī Uthmanī notes the following five possibilities with respect to shādhdh qirāʾāt:
  1. The reading was an innovation or fabrication.
  2. The reading has not been reliably transmitted.
  3. The reading represents the addition of explanatory words.
  4. The reading was abrogated in the final days of the Prophet’s life and known to be abrogated by the majority of companions, but an individual companion who was unaware may have continued to recite it as he learned it.
  5. The reading was actually a mistake made by a successor in his recitation but the listener thought it was a variant reading and transmitted it as such.
There is overlap between these possibilities and the scenarios discussed in detail above, and each of them appears to plausibly account for at least some of the reported variants. In the estimation of the authors of this paper, this appears to be the most reasonable conclusion on this matter.

What was the scope of the concession?

Scholars differed concerning the extent of flexibility granted to the early Muslim community to recite the Qur’an in different ways. Was there a requirement for every variant to be transmitted from the Prophet, or was there a concession for Muslims to recite the approximate meaning, paraphrasing to some degree if they could not recall the exact wording? The latter scenario has been termed qirāʾah bi-l-maʿnā (reciting based on meaning), in contrast to qirāʾah bi-al-talaqqī (reciting based on direct learning).
One of the scholars who thought such an expanded concession existed was Abū Jaʿfar al-Ṭaḥāwī (d. 321 AH). He wrote, “The flexibility given to them in reciting the Qur’an was to recite according to its meanings (an yatlūhu bi-maʿānīhi) even if the wording which they recited differed from the words of their Prophet with which he recited it to them.” However, al-Ṭaḥāwī specified that this was a concession (rukhṣah) that was later abrogated as it was no longer needed. Al-Ṭaḥāwī does not consider this license to have been the origin of the widely-transmitted readings (qirāʾāt); instead he believed that the qirāʾāt came from different review sessions between the Prophet and Jibrīl. Furthermore, al-Ṭaḥāwī considered there to be no remnants of this temporary license of qirāʾah bi-l-maʿnā within the codices compiled by Abū Bakr and ʿUthmān. 
A similar idea concerning a concession for reciting according to meaning has been expressed by Abū Bakr al-Bāqillānī (d. 404 AH). He wrote, “And it is possible that in the beginning of Islam it was legislated for people to exchange one ḥarf for another for instance instead of ʿalīm qadīr using ghafūr raḥīm, and then it was abrogated after that.” He was explicit on the latter point:

And the seven [canonical] reciters are to be followed in all their readings established from them… and it is not permissible and not justifiable to recite according to the meaning (qirāʾah bi-l-maʿnā) instead of following the Divinely-revealed wording.

The statements of al-Bāqillānī and al-Ṭaḥāwī are in agreement that the existence of a concession to recite according to the meaning during the Prophetic era does not entail that the qirāʾāt originated from this license. Likewise, Ibn ʿAqīlah al-Makkī (d. 1150 AH) wrote in response to the question concerning the permissibility of substituting synonyms in recitation, “It is not allowed to do that, rather the concession was for those whose tongue was previously accustomed to something other than the mutawātir wording of the Qur’an.”
There are other scholars who may have also been referring to such a restricted concession to recite according to the meaning, although their statements are less explicit in comparison to those previously mentioned. Perhaps the earliest statement that can be cited in this regard is from Ibn Shihāb al-Zuḥrī (d. 124 AH):

Abū Uways said, “I asked al-Zuḥrī about al-taqdīm wa-al-taʾkhīr (reversing the order of words or a phrase) in Hadith and he replied, ‘This is permitted with the Qur’an so how about the Hadith? If the meaning of the Hadith is captured correctly and one does not make the impermissible permissible or make the permissible impermissible, then there is no problem and that is if the meaning is correctly captured.”

Ibn Shihāb al-Zuhrī was known to have permitted transmission of hadith based on meaning (riwayah bi-al-maʿnā); however, he also commented on the famous seven aḥruf hadith by saying, “It has reached me that these seven aḥruf are basically one in meaning, they do not differ about what is permissible or prohibited.” Taken together with his previous statement, it may be claimed that he considered the seven aḥruf tradition to essentially be the Prophet’s way of describing a concession to use qirā’ah bi-al-ma’nā. The other possibility is that he was simply referring to the fact that revealed aḥruf use taqdīm wa-l-ta’khīr in Divine speech so there is no reason why human speech cannot.
Imām al-Shāfiʿī (d. 204 AH) has an interesting statement concerning the seven aḥruf. In his work on legal theory entitled al-Risālah, he discusses the fact that there are different wordings narrated from various companions for the tashahhud (the supplication recited during the seated position in the ritual prayer). As al-Shāfiʿī explains, it was allowed for the companions to recite this prayer with slight variations according to what they memorized so long as it did not affect the meaning. He then cites as evidence for the permissibility of this the very same narration about ʿUmar and Hishām encountered earlier regarding their difference in reading Sūrat al-Furqān. He then writes:

If God, out of mercy and compassion for His creatures, revealed His Book in seven [aḥruf]—knowing that memorization is subject to slippage—making it lawful for them to recite it with different wordings as long as their differences do not distort the meaning, then it is even more appropriate that differences in the wording of texts other than the Book of God be permitted as long as the meaning is not distorted in any text that does not convey a legal ruling. Differences in wording do not distort the meaning in such cases. One of the Successors said: “I met some of the Companions of God’s Emissary, and they were united in regard to the meaning but disagreed over the wording of a certain text. I asked one of them about that and he said, ‘It is not a problem as long as the meaning is not distorted.’”

In this case, al-Shāfiʿī is using the license for reciting according to the seven aḥruf to prove the permissibility of slight variation in tashahhud using argumentum a fortiori (qiyās bi-l-awlā). However, there is no need to take that as an endorsement of qirāʾah bi-l-maʿnā. The point is that if Allah provided a license for people to read with slight differences according to the aḥruf due to the fallibility of human memory, there is all the more reason to provide a license for small variations when narrating supplications. The license for the former is more restricted than the license for the latter, given that one must still recite according to the way one was taught the aḥruf while the less restricted license is only applicable to ḥadith. This statement therefore does not explicitly indicate a concession to recite differently from the revealed wording. 
Moreover, al-Shāfīʿī has commented on the same issue elsewhere with further details. In Ikhtilāf al-Ḥadīth, he discusses the different variants of the tashahhud, again draws an analogy with the aḥruf, however he specifies, “It is plausible that all these variants are authentically established, and that the Prophet would teach the congregation and individuals the tashahhud, so that one person would memorize it according to one wording and another person would memorize it according to a different wording.” He also writes, “and no one has the right to intentionally avoid reciting a letter from the Qur’an (according to the revealed wording), except out of forgetfulness.” This makes it clear that he considered the rule to be reciting according to the exact wording, and impermissible for one to intentionally depart from it.
Abū al-Layth al-Samarqandī (d. 375 AH) was of the opinion that Muslims were given permission under the license of aḥruf to recite the Qur’an according to the pronunciation of their own dialects without precedent, while only one reading originated from the Prophet ﷺ. According to Abul-Layth, this license to pronounce words differently (e.g., buyūt or bīyūt) was applicable even in farsh as long as the changes did not impact the meaning; he did not extend this to wording alterations.
Abū Shāmah al-Maqdisī (d. 665 AH) linked the seven aḥruf with qirāʾah bi-l-maʿnā though he considered them to have not been abrogated but simply no longer applicable after the ʿUthmānic recension. He writes concerning the seven aḥruf hadith:

The meaning of the hadith is that it was a concession given to them to replace words with that which gave the same or nearly the same meaning from one ḥarf to seven aḥruf. And they were not compelled to stick to one ḥarf since it was revealed upon an unlettered nation who were not accustomed to studying, repetition, or memorization of the exact wording of something, taking into consideration those who were elderly as well as those preoccupied with striving in battle and livelihood. So a concession was given to them in that. And some raised with one dialect would find it difficult to switch to another. So the qirāʾāt differed as a result of all of that.

And this is proven for us by the Hadith which explains [seven aḥruf] using the examples of halumma and taʿāl, showing the permissibility of exchanging one word for its synonym. And this is also proven for us by what has been established about substituting ghafūran raḥīman in place of ʿazīzan ḥakīman using what conveys the same essential meaning without preserving the exact wording (dūna al-muḥāfaẓati ʿalā al-lafẓ). For all of that is praise of Allah, Glorified is He.

This is all in regards to what the reciter can naturally pronounce. As for what is not possible for him because it is not from his dialect, then his situation is clear. And nothing of the qirāʾāt goes beyond this principle, which is exchanging a word for its synonym or a similar word that conveys the same essential meaning. Then, when the maṣāḥif were codified, those qirāʾāt that conflicted with the codified script were abandoned and what remained were those readings to which the script was amenable. Then some readings which conformed to the script became popular while the transmission of others became rare.

A few pages later, Abū Shāmah does quote al-Baghawī who explicitly rejects qirāʾah bi-l-maʿnā. Moreover Abū Shāmah made it clear that qirāʾah bi-l-maʿnā did not enter into the text of the Qur’an: “That which is compiled in the muṣḥaf is that which is agreed to be Divinely revealed, decisively confirmed, and it is what was written by the instruction of the Prophet ﷺ or what was established from him that he recited or taught others to recite.” 
Thus, Abū Shāmah clearly indicates that a concession for qirāʾah bi-l-maʿnā was limited to the practice of some early Muslims prior to the ʿUthmānic codex and it was not something that entered into the Qur’anic text, which was compiled exclusively according to the revealed wording as taught by the Prophet ﷺ. Ibn Ḥajar al-ʿAsqalānī (d. 852 AH) similarly states:

The aforementioned permission [to recite in different ways] was not according to desire, meaning that everyone could exchange a word with its synonym from his dialect, rather it was necessary to adhere to what was directly heard from the Prophet, and this is indicated by the fact that both ʿUmar and Hishām said, “The Prophet recited it to me this way.”

However, it is established from more than one companion that he would recite using synonyms without precedent (wa law lam yakun masmūʿan lahu). And it is for this reason that ʿUmar critiqued Ibn Masʿūd’s recitation of “attā ḥīn” to mean “ḥattā ḥīn” and he wrote to him, “Verily, the Qur’an was not revealed in the dialect of Hudhayl so teach people the recitation of Quraysh and do not teach them the recitation according to the dialect of Hudhayl.”

Ibn al-Jazarī (d. 833 AH) used the same concept to explain one possible origin of the shadhdh qirāʾāt:

The shādhdh qirāʾāt, if they happen to be authentic, originate from readings that were permitted but were not established as Divinely revealed; or people were given the choice concerning it in the beginning and then the ummah agreed upon abandoning them due to the benefit (maṣlaḥah).

He further said:

The companions agreed upon writing the Qur’an according to the Final Review (al-ʿarḍah al-akhīrah) which the Prophet ﷺ recited during the year of his passing, and according to what was revealed by Allah Most Exalted, rather than according to that for which a permission was given (dūna mā adhina fih).

Let us pause for a moment and consider what these scholars say and what they do not say. If the Prophet ﷺ gave a concession for some people to recite according to meaning, was that the rare exception while the default remained talaqqī or was the default variation? There is no indication that these scholars considered the default to be anything other than talaqqī even if a concession was made for some to recite by meaning due to hardship and difficulty. In fact, most of the aforementioned scholars reiterate the concept of recitation being a practice to be followed precisely (sunnah muttabaʿah). Moreover, did those companions who recited according to meaning ever consider their variant wording equivalent with Allah’s Divine speech? None of these scholars suggest any such notion; if anything, it would be clear that the wording was a concession in lieu of reciting the precise wording of the Divine speech.
This provides an important corrective to recent attempts to take qirāʾah bi-l-maʿnā from being a limited concession which a small handful of scholars believed existed, and interpret it instead as an explanation for the emergence of all Qur’anic variants, including the qirāʾāt. This qirāʾah bi-l-maʿnā theory and its arguments will be evaluated and critiqued in detail in the subsequent section. However, for now it is important to note that the scholars discussed did not state that the widely-recited canonical qirāʾāt were the result of qirāʾah bi-l-maʿnā. It is also interesting to note that the orientalist Ignaz Goldziher espoused a version of this view in his claim that the qirā’āt traditions in the Muslim world were invented without precedent or transmission, based on Muslims’ interpretation of the ʿUthmānic script lacking diacritics. However, this view has been convincingly challenged by a detailed study of the nature of variants.

Evaluation of arguments

Proponents of the theory of qirāʾah bi-l-maʿnā generally cite the following evidences (the specific counter-points to which are mentioned in the footnotes while the general counter-arguments are offered later in this section):
  1. The seven aḥruf hadith which conveys a degree of fluidity so long as the message is not altered, such as the hadith wherein the Prophet ﷺ states, “Whether you say ‘Samīʿan ʿAlīman’ (Most Hearing, Most Knowing) or ‘ʿAzīzan Ḥakīman’ (Most Mighty, Most Wise) then Allah is like that, so long as you do not conclude a verse of punishment with mercy or a verse of mercy with punishment.” 
  2. Reports about the reading or teaching of companions; e.g., the hadith reported from Abū al-Dardāʾ and Ibn Masʿūd about teaching a man to recite ṭaʿām al-fājir instead of ṭaʿām al-athīm. Anas ibn Mālik recited awabu qīlan instead of aqwamu qīlan and when questioned about that, he replied that the meaning was the same.
  3. Reports about the companions seemingly criticizing or denying a Qur’anic reading, which would seem to negate the supposition that they are all equally valid and all from the Prophet. This includes, for instance, an example of ʿĀʾishah denying one of the readings of 12:110, and ʿAlī denying the ʿUthmānic reading of 56:29.
  4. The existence of shādhdh readings which do not conform to the ʿUthmānic muṣḥaf.
  5. The examples of reciters performing ijtihād based on the ʿUthmānic text or reciters who selected a reading without precedent.
  6. The argument that the sheer quantity of all readings that would be traced back to the Prophet ﷺ makes it seem implausible that he could have recited them all.
The theory of qirāʾah bi-l-maʿnā may seem to account for all the variants reported and conveniently side-step the orientalist arguments about multiple “versions” of the Qur’an that do not originate from the Prophet ﷺ. However, these are not strong reasons to adopt this view given that the well-established traditional explanations easily account for variant readings and the above-mentioned evidences are, in fact, not very convincing on closer inspection.
First, with respect to the inferences drawn from aḥādīth implying a degree of flexibility in the wording of a verse so long as one does not conclude a verse of mercy with punishment, these inferences neglect the textual evidence that demonstrates an understanding on the part of the companions that they were to recite precisely as they were taught. For instance, the incident between ʿUmar and Hishām ibn Ḥakīm would never have occurred if there was an expectation that small variations were tolerable (which is the understanding of those who consider the Qur’an akin to other compositions in oral cultures). It is precisely because the companions expected everything to be recited exactly as taught by the Prophet ﷺ that they were so cautious and a dispute occurred when they suspected a variation in reading that was not taught by the Prophet. When the Prophet ﷺ spoke of the seven aḥruf, he said “the Qur’an was revealed according to seven aḥruf” or he said about a reading “This is how it was revealed,” and “Thus Jibrīl recited to me.”
Moreover, in response to these disputes, the Prophet ﷺ did not tell people to read as they pleased but rather told ʿAlī to reiterate to people that each individual was to recite according to how he had learned. ʿUmar ibn al-Khaṭṭāb said, “Qirāʾah of the Qur’an is a sunnah that one takes from one before him.” Thus, any hadith that seems to imply flexibility in wording is taken to mean that flexibility exists in selecting one of the pre-specified Divinely revealed readings.
Al-Bayhaqī (d. 458 AH) states:

As for those reports that have been transmitted concerning the permissibility of reciting Ghafūr Raḥīm in place of ʿAlīm Ḥakīm, then it is because all of that is from what has been revealed in waḥy (revelation). So if one recites that phrase in other than its correct place so long as he does not conclude a verse of punishment with a verse of mercy or a verse of mercy with a verse of punishment, then it is as if he recited the verse of one sūrah with the verse of another so he would not be sinful according to that. And the basis is that qirāʾah which was confirmed in the year the Prophet passed away after Jibrīl reviewed the Qur’an with him twice that year, and then the companions agreed upon its compilation between the covers of the muṣḥaf.

There is a hadith that is particularly explicit about the necessity of preserving the precise wording of the Qur’an. It tells of a man from Banū al-Najjār who outwardly converted to Islam and used to write revelation for the Prophet ﷺ but made alterations; he would write samīʿ ʿalīm when the Prophet dictated samīʿ baṣīr and vice versa. He later publicly renounced the faith and claimed the Prophet knew nothing other than what he had written. When he died, his tribesmen attempted to bury him several times but would awaken the next day to find his body cast out by the earth. Eventually, they gave up trying to bury him and abandoned his corpse. This would constitute the sternest warning in the Muslim tradition against making any alterations in the Qur’anic recitation or script.
Evidence suggests that the early Muslim community had the understanding of preservation of the wording of the Qur’an, rather than a notion of preservation and transmission by meaning. A story recounted by Yaḥyā ibn Aktham (d. 242 AH) tells of a man who accepted Islam in the court of al-Maʾmūn because of a test the man had invented to confirm the veracity of the faith. He had introduced subtle scribal errors into the Torah, the Gospels, and the Qur’an and disseminated their copies; the first two went undetected while the copies of the Qur’an containing errors were detected immediately. When Yahya told this story to Sufyān ibn ʿUyaynah (d. 198 AH), the latter confirmed that this is one of the unique features of the Qur’an in contradistinction to previous scriptures which were entrusted to their respective nations to preserve, citing 5:44.
As for the narrations regarding a companion reportedly criticizing a qirāʾah, these narrations are readily understood to be based on what they were familiar with from the diverse readings of the Qur’an. For instance, when Saʿd ibn Abī Waqqāṣ criticized a reading of Saʿīd ibn al-Musayyib, he did so because he did not think it came from the Prophet (ﷺ). When ʿUthmān compiled the muṣḥaf, when they wanted to record a verse they would specifically ask a person who heard it directly from the Prophet ﷺ, “How did the Prophet recite this verse to you?” Thus, the narrations of criticizing other readings need to be understood in light of the companions’ zeal in establishing with certainty the correct reading and being stringent in accepting the validity of a reading, which is precisely the opposite of the laxity one would expect if the companions adopted qirāʾah bi-l-maʿnā. One of the clearest examples of this is an incident in which ʿUmar ibn al-Khaṭṭāb heard a variant reading from a man who stated he learned it from Ubayy ibn Kaʿb. So ʿUmar took the man to Ubayy and questioned Ubayy about whether he heard this verse directly from the Prophet ﷺ in this manner, to which Ubayy replied in the affirmative, “Yes, I learned it directly (talaqqī) from the Prophet.” However, ʿUmar repeated the question multiple times, and on the third time, Ubayy became upset and said, “Yes I swear by Allah! Allah revealed it like this to Jibrīl and revealed it to Muhammad, and He did not consult al-Khaṭṭāb nor his son (i.e., ʿUmar)!” Thereupon ʿUmar left Ubayy, raising his hands in prayer saying, “Allahu Akbar, Allahu Akbar.”
In light of these narrations, it is clear that any narrations that suggest the substitution of words when a companion was teaching a Qur’anic verse are to be understood as isolated cases of providing a rukhṣah in circumstances when the student was unable to recite, memorize, or learn the original word. This is most evident in the story of Ibn Masʿūd or Abū al-Dardāʾ substituting ṭaʿām al-fājir in place of ṭaʿām al-athīm after the student continued to mistakenly recite ṭaʿām al-yatīm despite multiple attempts. These incidents do not imply that the concession wording was then considered part of Allah’s Divine speech revealed in the Qur’an.
Numerous classical scholars rejected the notion of qirāʾah bi-l-maʿnā in explicit terms, including al-Baghawī (d. 516 AH), Ibn al-Ṣalāḥ al-Shahrazūrī (d. 643 AH), Ibn al-Ḥājib al-Mālikī (d. 646 AH), Ibn Taymīyah (d. 728  AH), Ibn al-Jazarī (d. 832 AH), and al-Suyūṭī (d. 911 AH), among many others. The early scholar Abū Bakr ibn al-Anbārī (d. 328 AH) wrote concerning the narration from Anas ibn Mālik (reciting awabu qīlan):

Some of those astray may resort to such reports to cast doubts and state: whoever recites according to a reading that conforms to the meaning of a ḥarf from the Qur’an is correct so long as he does not contradict the meaning or bring forth other than what Allah intended and meant, and they use as proof this statement of Anas ibn Mālik. And this argument is an empty claim that should not be relied upon nor should the one making it be paid any attention. And that is because if one were to recite according to words that differ with the wording of the Qur’an despite approximating its meaning and general intent, then it would be permissible to recite in place of “al-ḥamdu lillāhi rabbi al-ʿālamīn” the phrase “al-shukru lil-bārī malik al-makhlūqīn.” And the flexibility in this regard would continue to increase until the entire wording of the Qur’an was nullified and replaced with that which would be a fabrication upon God and a lie upon His Messenger.

The quote from Ibn al-Anbarī raises an incredibly profound point about the notion of unrestricted qirāʾah bi-l-maʿnā. How can such qirāʾah bi-l-maʿnā even work? Why can’t someone recite “al-shukru lil-bārī malik al-makhlūqīn”? It is very difficult to understand how the process can take place without any limits or parameters—in essence, it’s a game without rules. If the Qur’an was being recited on a daily basis by thousands of people of diverse linguistic backgrounds and imperfect recollection, each with the license to alter its wording at will, there would be no consistency in public recitation and prayers, and no one would know whether to correct the imam’s recitation or not. There would be so much variability that the task of even assembling a fixed copy of the Qur’an would be nigh impossible. In order to even recite, memorize, and record something there must be a high degree of invariance and permanence. Moreover, we don’t have any statement from the Prophet ﷺ or any of the companions ever telling a learner that they had exceeded the limits or gone beyond the scope of what was acceptable variation.
We can learn as much from the places where qirāʾāt agree as we can from where they differ. The qirāʾāt traditions demonstrate an exceptionally high degree of invariance overall, despite containing multiple near-identical passages with subtle differences (mutashābihāt) that are challenging to recall with precision without careful diligent memorization of the differences (e.g., 2:58-60 compared with 7:160-2, or compare 20:10, 27:7, and 28:29, or 26:41-5 versus 7:113-117). If variant readings were based on imperfect recall and transmission based on meaning, we should expect the greatest concentration of reported variants in such near-identical passages given that they are the hardest to remember precisely.
Does the foregoing discussion mean that any conception of qirāʾah bi-l-maʿnā is entirely without merit? To the contrary, the concept as it was discussed by the classical scholars raises some interesting questions and does suggest the possibility that, on some occasions, the companions would provide a “rukhṣah reading” of sorts that was meant to assist someone unfamiliar with the words in a particular Qur’anic verse, but these readings were only an aid and did not become part of what was considered ‘God’s speech’ (kalām Allāh), nor did they become the basis for widely recited qirā’āt in the Muslim world. Taken in this manner, this view becomes no different from the basic idea behind al-qirāʾāt tafsīrīyah.

Were all variations recited by the Prophet Muhammad ﷺ?

A final question might arise here as to what did the Prophet himself ﷺ read? Did he actually read all the uṣūl from each of the canonical qirāʾāt and the farsh as well?
1. With regard to uṣūl, the Prophet ﷺ could have recited each principle (aṣl) of these uṣūl in a few sūrahs or even in one sūrah, or it could have been that it was read to him and he gave his approval. For example, a companion whose dialect is imālah would naturally read to the Prophet in what he was accustomed to. However, one must have had the approval of the Prophet ﷺ before attempting to recite the Qur’an according to their own dialect. 
2. As for farsh, the Prophet ﷺ would have recited each word in each of its different ways at least at some point in ṣalāh or in teaching sessions at least once during the course of his lifetime, as per qirāʾah bi-l-talaqqī. However, different pronunciations of the same word would fall under the same category as uṣūl as mentioned earlier by Abu al-Layth al-Samarqandī. 

Conclusion

The qirāʾāt that Muslims recite today have been transmitted through generations after generations of reciters with uninterrupted chains of transmission tracing back to the Prophet ﷺ, containing within them a mixture of the variation permitted according to the seven aḥruf. All of the accepted qirāʾāt follow three basic rules:
1. Conformity to the consonantal skeleton of the ʿUthmānic muṣḥaf.
2. Consistency with Arabic grammar.
3. Authentic chain of transmission.
Those qirāʾāt that fall short of these conditions are shādhdh (anomalous/irregular). This article has focused on elucidating how the Islamic tradition accounts for authentically established readings that do not meet the first condition above. The variant readings reported from companions that differ from the ʿUthmānic codex may represent either an abrogated or abandoned ḥarf, or they may represent a recitation containing word alterations for commentary or for facilitation for a learner. The vast and astounding detail with which the Islamic tradition discusses the diverse modes of recitation of the Qur’an is without a doubt unparalleled by any other book in the history of human civilization.
Today, Muslims continue to recite multiple canonical recitations that conform to the muṣḥaf and have been transmitted by generations upon generations with unbroken chains of authority tracing to the Prophet ﷺ. To listen to the melodious Qur’anic recitation in different readings is to experience it in the same mesmerizing manner with which it captured the hearts of its earliest audience.

Notes

1 A rough approximation can be obtained by dividing those 703 places wherein a different reading is listed in the index to Ibn Mujāhid’s Kitāb al-sab’ah fī al-qirā’āt by the total number of words in the Qur’an (surpassing 77,400), in order to arrive at 0.9% of words with an alternative reading. For these numbers see Yasin Dutton, “Orality, Literacy and the ‘Seven Aḥruf’ Hadith,” Journal of Islamic Studies 23, no. 1 (2012): 10.
2 The early scholar from Baghdad, Ibn Mujāhid (d. 324 AH) is famous for enumerating seven acceptable modes of recitation in his work Kitab al-sabʿah fi al-qirāʾāt. He selected those readings that had become widely accepted, picking one reader for every major center of knowledge in the Muslim world (Mecca—Ibn Kathīr, Damascus—Ibn ʿĀmir, Basrah—Abū ʿAmr, Madinah—Nāfiʿ) except for Kūfah, from which he chose three reciters (ʿĀṣim, Ḥamzah, al-Kisāʾī). There were many readings that were left out; for instance, Ibn al-Jazarī (d. 833 AH) stated that the early scholar Abū ʿUbayd al-Qāsim ibn Sallām (d. 224 AH) compiled twenty-five qirā’āt. Some scholars, including Ibn al-Jazarī, took the list of seven from Ibn Mujāhid and added three other reciters to form the canonical list of ten: Abū Ja’far from Madinah, Yaʿqūb from Baṣrah, and Khalaf from Kūfah.
3 Both Muhammad Mustafa Al-Aʿzami and Yasin Dutton note the inadequacy of the term “variant” for the Qur’an, given that there is not a singular fixed original, but rather the original itself is “multiformic” to use Dutton’s terminology. See Dutton, “Orality, Literacy and the ‘Seven Aḥruf’ Hadith,” 1–49; also see Muhammad Mustafa Al-Azami, The History of the Qur’anic Text from Revelation to Compilation (Leicester: UK Islamic Academy, 2003), 154–55. Acknowledging this point, the term is used in the present article to signify a reading that does not conform to the Uthmanic recension. Shady Hekmat Nasser terms such readings “anomalous” and uses the term “irregular” for those readings that are deficient in transmission or grammar, while both are referred to as shaadh in Arabic. Nasser, The Transmission of the Variant Readings of the Qurʾān: The Problem of Tawatur and the Emergence of Shawadhdh (Leiden: Brill Academic Publishers, 2003), 16.
4 Behnam Sadeghi and Mohsen Goudarzi, “San'a' 1 and the Origins of the Qurʾān,” Der Islam 87, no. 1–2 (February 2012): 1–129. For this categorization, pp. 3–4.
5 Sadeghi and Goudarzi, 3.
6 Sadeghi and Goudarzi; see also Harald Motzki, “Alternative Accounts of the Qur’an’s Formation,” in The Cambridge Companion to the Qur’an, ed. J. McAuliffe, Cambridge Companions to Religion (Cambridge: Cambridge University Press, 2006), 59–76.
7 Nicolai Sinai, “When Did the Consonantal Skeleton of the Quran Reach Closure? Part II,” Bulletin of the School of Oriental and African Studies 77, no. 3 (2014): 509–21.
8 See for instance: Sami Ameri, Hunting for the Word of God (Minneapolis: Thoughts of Light Publishing, 2013); Al-Azami, History of the Qur’anic Text; Abd al-Fattah Shalabi, Rasm al-mushaf al-Uthmani wa awham al-mustashriqin fi qira'at al-Qur'an al-Karim (Cairo: Maktabah Wahbah, 1999); Muhammad Mohar Ali, The Quran and the Orientalists (Norwich: Jamiyat Ihyaa Minhaaj al-Sunnah, 2004). The alternative narrative is discussed below in the section on qira’ah bil-ma’na.
9 It is anticipated that the present article will form the first in a trilogy with two subsequent articles examining the ʿUthmānic variants and the post-ʿUthmānic readings respectively.
10 Ṣaḥīḥ Bukhārī, no. 3, https://sunnah.com/bukhari/1/3.
11 Qur’an 16:102.
12 Ṣaḥīḥ Bukhari, no. 6, https://sunnah.com/bukhari/1/6.
13 Al-A’zami, History of the Qur’anic Text, 64–65. Al-A’zami lists no fewer than thirty-nine companions by name who memorized the Qur’an directly from the Prophet ﷺ, and this list evidently includes only the most famous who lived to teach others, as the names of those who died in Bi’r Ma’unah and Yamamah have not been preserved.
14 Ṣaḥīḥ Bukhari, no. 4090, https://sunnah.com/bukhari/64/134.
15 Ṣaḥīḥ Bukhari, no. 5027, https://sunnah.com/bukhari/66/49.
16 Al-A’zami, History of the Qur’anic Text, 68.
17 In addition to being recorded in almost all the canonical six works (Bukhārī, Muslim, Abū Dāwūd, Tirmidhī, al-Nasā’ī), the seven aḥruf narrations are found in numerous early works including the Jāmiʿ of Maʿmar ibn Rāshid (d. 153 AH), Muwaṭṭa of Imam Mālik ibn Anas (d. 179 AH), the Musnad of Abū Dāwūd al-Ṭayālisī (d. 204 AH), Musnad al-Ḥumaydī (d. 219 AH), Muṣannaf ibn Abī Shaybah (d. 235 AH), and the Musnad of Imām Aḥmad (d. 241 AH). Given the voluminous transmitted reports from the earliest era, the fact that the earliest Muslim community understood the Qur’an to be a multiform recitation cannot be logically disputed even by the most skeptical historian.
18 A comprehensive study of the narrations demonstrates that it was reported by no fewer than twenty-three companions through numerous diverse chains. See ʿAbd al-ʿAzīz al-Qāriʾ, Hadith al-aḥruf al-sabʿ: Dirāsat isnādihi wa matnihi wa ikhtilaf al-ʿulamāʾ fī maʿnāhu wa ṣilatihi bi-al-qirā’āt al-Qurʾānīyah (Beirut: Resalah Publishers, 2002).
19 Ṣaḥīḥ al-Bukhārī, no. 4991, kitāb faḍāʾil al-Qurʾān. The hadith of seven aḥruf are cited in no fewer than four books within Ṣaḥīḥ al-Bukhari.
20 Jāmiʿ al-Tirmidhī, no. 2944, https://sunnah.com/urn/731900.
21 Ibn Qutaybah, Ta’wīl mushkil al-Qurʾān (Cairo: Maktabah Dar al-Turath, 1973), 38–40; see also Abū Shāmah, al-Murshid al-wajīz ilā ʿulūm tataʿallaq bi-al-kitāb al-ʿazīz (Beirut: DKI, 2003), 90.
22 Ṣaḥīḥ al-Bukhārī, no. 5041, https://sunnah.com/bukhari/66/65.
23 al-Khaṭṭābī, Maʿālim al-sunan, (Halab: al-Maṭbaʿah al-ʿIlmiyyah, 1932), 1:293.
24 For further details, see Makkī ibn Abī Ṭālib, al-Ibānah ʿan maʿānī al-qirāʾāt (Cairo: Dar Nahdah Misr, 1977), 71–79; Abū Shāmah, al-Murshid al-wajīz, 91-111; al-Zarkashi: al-Burhān fī ʿulūm al-Qurʾān (Cairo: Dār al-Turāth, 1984), 1:213–27; Ibn al-Jazarī, al-Nashr, ed. Muḥammad Sālim Muḥaysin (Cairo: Maktabah al-Qahirah, 1978), 1:21–31; ʿAbd al-ʿAzīz Qāriʾ, Hadith al-aḥruf as-sab’ah; Ghanim Qadduri Al-Hamad, al-Ajwibah al-‘ilmiyyah ‘alā as’ilat multaqā ahl al-tafsīr.
25 Taqi Uthmani, An Approach to the Qur’anic Sciences (Karachi: Darul Ishaat, 2007), 105–65; Ahmad Ali al-Imam, Variant Readings of the Qur’an (London: IIIT, 2006); Al-Azami, History of the Qur’anic Text, 152–64; Abu Ammar Yasir Qadhi, An Introduction to the Sciences of the Qur’aan (Birmingham: Al-Hidaayah Publications, 1999), 172–202.
26 See Ibn al-Jazarī: an-Nashr, 1:27; al-Suyūṭī: al-Itqān, 1:313–14.
27 In our time, shawādhdh has been used as a designation for all qirā’āt other than the ten canonical qirā’āt. The formal classical definition is those readings that lack one of the following three criteria: (1) authentic transmission; (2) conformity to the ʿUthmānic skeletal text; and (3) concordance with conventional Arabic grammar.
28 For a list of these variants see ʿAbd al-Laṭīf al-Khaṭīb, Mu’jam al-qirāʾāt (Damascus: Dār Sa’d al-Din, 2002), 6:315. No. 1 variant 2 is read by Abū Sawwār al-Ghunawī and Abū al-Jawzāʾ, no. 2 variant 2 by ʿĀṣim al-Jaḥdarī and ʿAbd Allāh ibn al-Zubayr, no. 2 variant 3 by Muʿādh Abū Ḥulaymah and Abū Nahik, no. 3 variant 2 by Ad´ham al-Sadūsī, and no. 4 variant 2 by ʿAbd Allāh ibn al-Zubayr.
29 Ibn ʿAbd al-Barr, al-Tamhīd li mā fī al-Muwaṭṭa min al-maʿāni wa-al-asānīd (Rabat: al-Maṭba’ah al-Malakiyyah, 1980), 8:301–2. Shady Hekmat Nasser also adopts a similar approach, The Transmission, 29–30.
30 Ṣaḥīḥ Ibn Ḥibbānkitāb al-raqā’iqbāb qirā’āt al-Qur’ān, 747, https://library.islamweb.net/newlibrary/display_book.php?idfrom=749&idto=749&bk_no=314&ID=741.
31 Ṣaḥīḥ Muslim, no. 403, http://sunnah.com/muslim/4/64.
32 Ibn Mujāhid, Kitāb al-sabʿah fī al-qirāʾāt, ed. Shawqi Dayf (Cairo: Dar al-Ma’arif, 1972), 51.
33 Ṣaḥīḥ al-Bukhārī, no. 4987, https://sunnah.com/bukhari/66/9.
34 For a discussion on this point, see Al-Azami, History of the Qur’anic Text,  97.
35 ʿAbd Allāh ibn Mas’ud was praised for his recitation by the Prophet ﷺ, being one of the four from whom the Prophet told others to learn the Qur’an (alongside Sālim, Mu’ādh, and Ubayy ibn Ka’b; Ṣaḥīḥ Bukhari, no. 4999), and being one whose recitation was described “as fresh as how it was revealed” (Sunan ibn Majah, no. 143, https://sunnah.com/urn/1251380).
36 Ibn Mujāhid, Kitab al-sabʿah fī al-qirāʾāt, 67. See also Ghānim Qaddurī al-Ḥamad, Rasm al-muṣḥaf: Dirāsah lughawīyah tārīkhīyah (Baghdad: al-Lajnah al-Waṭaniyyah, 1982), 623; and Ramon Harvey, “The Legal Epistemology of Qur’anic Variants: The Readings of Ibn Masʿūd in Kufan fiqh and the Ḥanafī madhhab,” Journal of Qur’anic Studies 19, no. 1 (2017): 72–101, 20, endnote 8.
37 Ṣaḥīḥ al-Bukhārī, no. 3742, https://sunnah.com/bukhari/62/89.
38 Ṣaḥīḥ al-Bukhari, no. 4727, https://sunnah.com/urn/180620.
39 Ṣaḥīḥ al-Bukhari, no. 4971, https://sunnah.com/urn/183110; Ṣaḥīḥ Muslim, no. 208a, https://sunnah.com/muslim/1/416. al-Nawawī (d. 671 AH) and al-Qurṭubī (d. 676 AH) both stated that this was a revealed verse that was abrogated. See al-Nawawī, al-Minhāj sharḥ Ṣaḥīḥ Muslim bin Ḥajjāj (Cairo: Mu’assasat Qurtubah, 1994), 3:102; al-Qurṭubī, al-Jāmiʿ li-aḥkām al-Qurān, ed. ʿAbd Allāh al-Turkī (Beirut: Mu’assasat al-Risalah, 2006), 16:83. Al-Kirmānī (d. 796 AH) adds the possibility of it being tafsīrSharḥ al-Kirmānī ala Ṣaḥīḥ al-Bukhārī (Beirut: DKI, 1971), 9:211.
40 Muḥammad ibn ʿAbd ul-Raḥmān al-Ṭāsān, al-Masāḥif al-manṣuba lil-ṣaḥābah wa-al-radd ʿalā shubuhat al-muthārah (Riyadh: Dar al-Tadmuriyyah, 2016), 314-315.
41 Variant readings that differ from the ʿUthmānic codex must be studied by hadith standards because, unlike the qirā’āt, they have not been transmitted by the ritual practice of one generation of Muslims from the next, but rather only as isolated reports. The successively transmitted uninterrupted living oral tradition has always been the primary factor in establishing the validity of one’s recitation.
42 While the 540 reported variants likely overestimates the quantity, the 20 variants with an authentic chain of transmission likely underestimates it due to the fact that some of the other variants are demonstrated in manuscripts.
43 In their 2012 essay studying folios of the lower text, Benham Sadeghi and Mohsen Goudarzi write: “The lower writing of Ṣan‘ā’ 1 clearly falls outside the standard [Uthmanic] text type. It belongs to a different text type, which we call C-1. . . . Ṣan‘ā’ 1 constitutes direct documentary evidence for the reality of the non-‘Uthmānic text types that are usually referred to as ‘Companion codices.’ . . . C-1 confirms the reliability of much of what has been reported about the other Companion codices not only because it shares some variants with them, but also because its variants are of the same kinds as those reported for those codices” (Sadeghi and Goudarzi, 17–20). They also note that since the Uthmanic mushaf always agrees with either C-1 or Ibn Masʿūd’s text (i.e., it is always in the majority) in the case of difference, this suggests either that it was compiled from a critical examination of other textual sources or that it is the most faithful reproduction of the Prophetic prototype (21–22). Both of these scenarios are entirely consistent with what the Muslim tradition states.
44 His father is, of course, none other than the famous Abū Dāwūd al-Sijistānī (d. 275 AH) who compiled the Sunan.
45 Ibn Abī Dāwūd, Kitāb al-maṣāḥif, ed. Muhib al-Din Wa’iz (Beirut: Dar al-Basha’ir al-Islamiyyah, 2002), 2:284.
46 al-Ṭāsān, al-Maṣāḥif al-manṣuba lil-ṣaḥābah, 63–64.
48 Ibn al-Nadīm, Kitāb al-fihrist, ed. Rida Tajadud (Teheran: 1971), 29.
49 Al-Azami, History of the Qur’anic Text, 215.
50 For a detailed list of these reports and their analysis in a recent work dedicated to this subject, see ʿAlī al-Jaʿfarī, alʿAdl wa-l-Iḥsān fī taḥrīr iʿtirāḍāt Ibn Masʿūd ʿalā muṣḥaf ʿUthmān, (Kuwait: Dār al-Ẓāhiriyya, 2022), 64-57.
51 Musnad Aḥmad 21226, Muʿjam al-Kabīr li-Ṭabarānī 9150. Declared authentic by Ibn Ḥajar, al-Suyūṭī, al-Shawkānī, al-Haythamī, Aḥmad Shākir and Shuʿayb al-Arnāʾūṭ. See al-Jaʿfarī, al-ʿAdl, 65, fn 2.
52 Al-Ṭāsān, Taḥqīq mawqif al-ṣaḥābī al-jalīl ʿAbdullah ibn Masʿūd min al-jam al-ʿUthmānī, (Riyadh: King Saud University, Chair of Quranic Sciences 1435H), 80-88. Al-Ṭāsān also mentions a third group that considers Ibn Masʿūd to have only disputed writing these chapters in the muṣḥaf and not to have disputed their Qur’anic status. However, this possibility falls within the scope of the first group since it rejects the apparent import of the reports about Ibn Masʿūd as the muʿawwidhatayn, and thus we find al-Bāqillānī mentions this possibility while also disputing the authenticity and reliability of the narrations.
53 Fakhr al-Dīn al-Rāzī, The Great Exegesis = Al-Tafsīr Al-Kabīr, trans. Sohaib Saeed (Cambridge, UK: The Royal Aal Al-Bayt Institute For Islamic Thought, 2018), 343.
54 Al-Nawawī, al-Majmūʿ Sharḥ al-Muhadhdhab, (Cairo: Idarat al-Tiba'a al-Muniriyya, 1925), 3:396.
55 Ibn Ḥazm, al-Muḥallā (Beirut: DKI, 2003), 1:32. See also al-Zarkashī, al-Burhān fī ʿulum al-Qurʾān, 2:128; and Al-Azamī, History of the Qur’anic Text, 201.
56 Abū Bakr al-Bāqillānī, al-Intiṣār lil-Qur’ān (Beirut: Dar Ibn Hazm, 2001), 1:300-330.
57 Al-Bāqillānī, al-Intiṣār, 307.
58 For instance, ʿUqba ibn ʿĀmir asked the Prophet about al-Muʿawwidhatayn, and the Prophet led them in Fajr prayer and recited them; Sunan al-Nasāʾī, 952.
59 Al-Bāqillānī, al-Intiṣār, 317.
60 Al-Aʿẓamī, 199.
61 Al-Bāqillānī, al-Intiṣār, 320; see also al-Qurṭubī, al-Jāmiʿ li-aḥkām al-Qurān, ed. ʿAbd Allah al-Turkī (Beirut: Mu’assasat al-Risalah, 2006), 177.
62 al-Qurṭubī, 177. For a discussion of the subject, see al-Tāsān, al-Masāhif, 390–97.
63 Ibn ʿAbd al-Kāfī, ʿAdad suwar al-Qurʾān, ed. Khalid Ḥasan Abū al-Jūd, (Cairo: Maktaba al-Bukhārī, 2010), 86.
64 Cited in al-Nawawī, al-Minhāj sharḥ Ṣaḥīḥ Muslim ibn Ḥajjāj, (Cairo: Mu’assasat Qurtubah, 1994), 6:157–58.
65 Muḥammad ʿAbd al-ʿAẓīm al-Zarqānī, Manāhil al-ʿirfan fī ʿulūm al-Qurʾān (Beirut: DKI, 2013), 1:200. See also the discussion of Theodor Nöldeke who argues on the basis of stylistic features that this invocation clearly does not match the Qur’anic style; Theodor Nöldeke, Friedrich Schwally, Gotthelf Bergsträsser, Otto Pretzl, and Wolfgang Behn, The History of the Qur’an (Leiden: Brill, 2013), 241–42.
66 Musnad Aḥmad 21189 and declared authentic by Shuʿayb al-Arnāʿūṭ.
67 Ibn Qutayba,  Ta’wīl mushkil al-Qurʾān, 43.
68 Ibn Ḥajar, Fatḥ al-Bārī, (Beirut: Dar al-Ma’rifah 1379H), 8:743.
69 Ibn Taymiyya, Majmūʿ al-fatawa, (Mansoura: Dar El-Wafaa, 2005), 12:493.
70 Al-Ṭāsān, Taḥqīq mawqif al-ṣaḥābī al-jalīl, 86-87. See also al-Judayʿ, 112.
71 Ibn Kathīr, Tafsīr Ibn Kathīr, (Riyadh: Dar al-Tayba 1999), 8:531.
72 Al-Jaʿfarī, al-ʿAdl, 63-64.
73 Al-Jaʿfarī, al-ʿAdl, 84.
74 See Justin Parrott, “Abrogated Rulings in the Qur’an: Discerning Their Divine Wisdom,” Yaqeen, November 15, 2018, https://yaqeeninstitute.org/justin-parrott/abrogated-rulings-in-the-quran-discerning-their-divine-wisdom/.
75 Ṣaḥīḥ Muslim, no. 1452a, https://sunnah.com/muslim/17/30. This narration refers to the extent of infant breastfeeding that is sufficient to establish a familial relationship under Islamic rulings.
76 Ṣaḥīḥ Muslim, no. 630, https://sunnah.com/muslim/5/264.
77 Ṣaḥīḥ al-Bukhārī, no. 3554, https://sunnah.com/bukhari/61/63.
78 Ṣaḥīḥ al-Bukhārī, no. 4998, https://sunnah.com/bukhari/66/20.
79 For a detailed review, see Nāsir ibn Saʿūd al-Qithamī, “al-ʿArḍah al-akhīrah: Dalālatuha wa atharuhā,” Majallah Maʿhad al-Imām al-Shaṭibī li-Dirasah al-Qurʾāniyyah, 2013, no. 15.
81 Sunan Sa’īd ibn Mansūr (Riyadh: Dar al-Samee’, 1993), 1:239.
82 Ibn al-Jazarī, al-Nashr, 1:8.
83 Osama Alhaiany, “al-ʿArḍah al-akhīrah lil-Qurʾān al-Karīm wa-al-aḥadīth al-wāridah fīhā jamʿan wa dirāsah,” al-Majallah al-Ulum al-Islamiyyah, no. 10, p. 67.
84 al-Baghawī, Sharḥ al-sunnah (Beirut: Al-Maktab Al-Islami, 1983), 4:525–26. Witnessing the final review would attest to Zaid’s merit in leading the committee commissioned by Uthman. However, Ibn ʿAbbās has a statement reported from him that states that ʿAbd Allāh ibn Mas’ud attended al-ʿarḍah al-akhīrah (Musnad Aḥmad 3422). If both ʿAbd Allāh ibn Mas’ud and Zayd attended, then it doesn’t make sense why Ibn Masʿūd would have variant readings from that which was established by Zayd in the ʿUthmānic codex. And if ʿAbdullah attended and Zaid did not, then why would the sahabah unanimously adopt a codex that did not match al-ʿarḍah al-akhīrah? Abū Jaʿfar al-Naḥḥās answers that since the reading of ʿAsim comes from ʿAbd Allāh ibn Masʿūd, we know that Ibn Masʿūd did not only teach and recite in one ḥarf, rather he also had the same ḥarf as Zayd in addition to one that differed. See al-Naḥḥās, al-Nāsikh wa-al-mansūkh (Riyadh: Dar al-Asimah, 2009), 2:408. On ʿAsim’s transmission, see al-Ṭaḥāwī, Tuḥfat al-akhyār bi-tartīb mushkil al-āthār, ed. Khalid Mahmud al-Rabat (Riyadh: Dar al-Balansiyya, 1999), 8:435.
85 al-Naḥḥās, al-Nāsikh wa-al-mansūkh, 2:407.
86 Ibn Taymīyah, al-Fatāwā al-kubrā (Beirut: DKI, 1987), 4:418.
87 See section “Codices of companions” above. Ṣaḥīḥ al-Bukhārī, no. 3742, https://sunnah.com/bukhari/62/89.
88 Note that Abū Bakr al-Bāqillānī and a few other scholars differed with this position and believed that all the seven aḥruf are contained within the ʿUthmānic codices. However, the majority believed that at least some of the aḥruf were left out.
89 Ibn Jarīr al-Ṭabarī, Tafsīr al-Ṭabarī, ed. al-Turkī (Cairo: Dār Hajar, 2001), 1:59–60.
90 al-Naḥḥās, al-Nāsikh wa-al-mansūkh, 2:405. Fa-arāda ʿUthmān an yakhtār min al-sabʿah ḥarfan wāḥid wa huwa afṣaḥuhā.
91 Ibn ʿAbd al-Barr, al-Istidhkār (Damascus: Dar Qutaibah, 1993), 8:45.
92 ʿAlī ibn Ismaʿīl al-Abyārī, al-Taḥqīq wa-al-bayān fī sharḥ al-burhān fī uṣūl al-fiqh (Doha: Wizārat al-Awqāf wa al-Shuʾūn al-Islāmīyah Qatar, 2013), 2:792.
93 Ibn al-Qayyim, al-Ṭuruq al-ḥukmīyah fī al-siyāsah al-sharʿīyah, (Mecca: Dār ʿĀlam al-Fawāʾid, 1428 AH), 1:47–48; Ibn al-Qayyim, Iʿlām al-muwaqqiʿīn (Dammam: Dār ibn al-Jawzī, 2002), 5:65.
94 See also Mannāʿ al-Qaṭṭān, Mabāḥith fī ʿulūm al-Qur’ān (Cairo: Maktabah Wahbah, 1995), 158.
95 Makkī states,
There ceased to be implementation of that which differed from the script of the (ʿUthmānic ) muṣḥaf from the seven aḥruf in which the Quran was revealed, based upon the unanimous consensus on the script of the muṣḥaf. Thus, the muṣḥaf was written based on one ḥarf, and its script accommodates more than one ḥarf as it was void of dotting and vowels.
See Makkī ibn Abī Ṭālib, al-Ibānah ʿan maʿānī al-qirāʾāt (Cairo: Dār Nahdah Misr, 1977), 34. This quote is of interest because it demonstrates that while opinions on this subject are conventionally divided into three groups—those who say all aḥruf are preserved (e.g., al-Bāqillānī), those who say only one is preserved (e.g., al-Ṭabarī), and those who say the ʿUthmānic muṣḥaf was deliberately written (free of vocalization and diacritics) to accommodate more than one ḥarf (Ibn al-Jazarī)—Makkī fell into a fourth category, suggesting that although the muṣḥaf was deliberately written according to one ḥarf, the script still accommodated other readings. See also Makkī ibn Abī Ṭālib, al-Hidāyah ilā bulūgh al-nihāyah (Sharjah: University of Sharjah, 2008), 4:2911–12.
96 Ibn al-Jazarī, al-Nashr, 1:31. He writes, “As for whether ʿUthmānic codices encompass all the seven aḥruf then this is a major topic . . . the position taken by the majority of the scholars from the earlier and later generations and the Imams of the Muslims is that these codices encompass that which the text can accommodate from the seven aḥruf.”
97 Ibn Ḥajar, Fatḥ al-Bārī (Riyadh: Dār al-Ṭaybah, 2005), 11:195–96. He further explains that this was a reason for the textual variants between ʿUthmānic codices, to increase the number of readings that could be accommodated.
98 Abū al-ʿAbbās ibn ʿAmmār al-Mahdawī, Sharḥ al-hidāyah (Riyadh: Maktabah Rushd, 1995), 5.
99 As demonstrated by the view of Makkī ibn Abī Ṭālib, there is no conflict between the view that the committee of ʿUthmān compiled the Qur’an according to one ḥarf and the view that some of the other aḥruf remained because even if the committee may have primarily had one of the various readings in mind when they transcribed the codex, that does not change the fact that other readings could still be accommodated by the skeletal text. See also Ghānim Qaddurī al-Ḥamad, Rasm al-muṣḥaf: Dirāsah lughawīyah tārīkhīyah, 152.
100 Makkī ibn Abī Ṭālib, al-Ibānah ʿan maʿanī al-qirāʾāt (Cairo: Dar Nahdah Misr, 1977), 42.
101 Abū Bakr al-Bāqillānī, al-Intiṣār lil-Qur’ān (Beirut: Dar Ibn Hazm, 2001), 1:153. He mentions as an example the reading “salāt al-ʿaṣr” in 2:238.
102 Abū Ḥayyān, al-Baḥr al-muḥīṭ (Beirut: Dār al-Fikr, 2010), 1:260.
103 Sunan al-Tirmidhī, no. 3314, https://sunnah.com/urn/740400.
104 al-Nawawī, al-Minhāj sharḥ Ṣaḥīḥ Muslim bin al-Ḥajjāj (Cairo: Mu’assasat Qurtubah, 1994), 6:157.
105 Ibn al-Jazarī, al-Nashr fī al-qirā’āt al-ʿashar, 1:44.
106 Related in al-Qurṭubī, al-Jāmiʿ li-aḥkām al-Qurān, ed. ʿAbd Allāh al-Turkī (Beirut: Mu’assasat al-Risalah, 2006), 1:134. See also, Ghānim Qaddūrī al-Ḥamad, Muḥāḍarāt fī ulūm al-Qurʾān (Amman: Dar Ammar, 2003), 145.
107 al-Naḥḥās, I’rab al-Qurʾān, (Riyadh: Dār ʿĀlam al-Kutub, 1988), 1:321–2. Note however the previously mentioned narration concerning this being an abrogated reading: Ṣaḥīḥ Muslim, no. 630, https://sunnah.com/muslim/5/264.
108 Reported in Sunan Sa’īd ibn Mansūr (d. 227 AH) according to al-Suyūṭī, al-Itqān fī ʿulūm al-Qurʾān (Beirut: DKI, 2000), 1:119.
109 Abdul Jalil, “Dhahirat al-ibdāl fī qirāʾat ʿAbd Allāh ibn Masʿūd wa-qīmatuhā al-tafsīrīyah,” Journal of Qur’anic Studies 15, no. 1 (2013): 168–213, 210.
111 al-Ḥākim al-Naysābūrī, al-Mustadrak ʿalā al-ṣaḥiḥayn, ed. Yūsuf ʿAbd al-Raḥmān Marʿashī (Beirut: Dār al-Maʿrifah, 1986), 2:451. The viewpoint adopted by al-Zarqānī is that Ibn Masʿūd selected another revealed ḥarf that was easier for the man to recite. al-Zarqānī, Manāhil al-ʿirfān, 1:133.
112 Al-Qurṭubī, al-Jāmiʿ li-aḥkām al-Qurān, ed. ʿAbd Allāh al-Turkī (Beirut: Mu’assasat al-Risalah, 2006), 19:132-133.
113 Al-Māwardī, al-Ḥāwī al-kabīr, (Beirut: DKI 1994) 2:115.
114 Taqi Uthmani, An Approach to the Qur’anic Sciences (Karachi: Darul Ishaat, 2007), 249–51.
115 Al-Ṭaḥāwī, Sharḥ mushkil al-āthār, ed. Shuʿayb al-Arnaʿūṭ, 16 vols. (Beirut: Mu’assasat al-Risalah, 1994), 8:118.
116 This can be contrasted with al-Zamakhsharī (d. 538 AH) who considered it permissible without having been abrogated. After quoting the incident of Abū al-Dardāʾ teaching a man to recite ṭaʿām al-fājir, al-Zamakhsharī writes, “And this is used to prove the permissibility of exchanging one word for another so long as it gives the same meaning.” He goes on to use this as an explanation for Abū Ḥanīfah permitting recitation in Persian. al-Zamakhsharī, al-Kashshāf (Beirut: Dar al-Marefah, 2009), 1003. See also the opinion of Abū al-Faḍl al-Rāzī in Maʿānī al-Aḥruf al-Sabʿa, 385-389.
117 For a detailed analysis of al-Ṭaḥāwī’s views, refer to Ammar Khatib, Taḥrīr madhhab al-Īmām al-Ṭaḥāwī fī maʿna al-aḥruf al-sabʿa, Tafsir Center for Quranic Studies, May 22, 2021, https://tafsir.net/article/5351.pdf.
118 Al-Bāqillānī, al-Intiṣār, 1:370.
119 Al-Bāqillānī, al-Intiṣār, 1:65.
120 Ibn ʿAqīlah al-Makkī, al-Ziyada wal-iḥsān fī ʿulūm al-Qurʾān, (Sharjah: Markaz al-Buḥūth 2006) 1:491-492.
121 Reported by al-Dhahabī in Siyar aʿlām al-nubalāʾ, 1:347.
122 See for instance, Naṣīr Aḥmad Muḥibb, Ḥukm riwāyah al-ḥadīth bil-maʿnā, AJRSP 19 (2020):251-259.
123 Ṣaḥīḥ Muslim, no. 819a, https://sunnah.com/muslim/6/330
124 Muḥammad ibn Idrīs al-Shāfiʿī, The Epistle on Legal Theory: A Translation of Al-Shafii's Risalah, trans. Joseph Lowry (New York: New York University Press, 2015), 119. Cf. al-Shāfiʿī, al-Risālah, (Cairo: al-Ḥalabī 1940), 275.
125 Al-Shāfiʿī, Ikhtilāf al-ḥadīth, (Beirut: Dār al-Fikr 1990), 600.
126 Al-Shāfiʿī, Ikhtilāf al-ḥadīth, 600.
127 See Abū al-Layth al-Samarqandī, Bustān al-ʿārifīn, chap. 29, concerning revelation of the Qur’an in 7 aḥruf, manuscript, Thomas Fisher Arabic Collection, University of Toronto, https://archive.org/details/bustanalarifin00unse/page/40; see also az-Zarkashī, al-Burhān, 1:326–27.
128 Abu Shāmah, al-Murshid al-wajeez ila ulum tata’alaq bil-Kitab al-Aziz (Beirut: Dar al-Kotob al-Ilmiyah, 2003), 105. See also p. 85 where Abu Shāmah states the ʿUthmānic recension was based only on the revealed wording (lafdh al-munazzal) and not the substituted synonym words (lafdh al-murādif).
129 Abu Shāmah, 109. The quote from al-Baghawī reads, “The meaning of these aḥruf is not that every group reads according to what matches their dialect without instruction (min ghayrī tawqīf) but rather all of these ḥurūf are mentioned (mansusah) and all of them are the speech of God (kalām Allah) with which the trustworthy spirit (i.e., Jibrīl) descended upon the Prophet.” See al-Baghawī, Sharḥ al-sunnah (Beirut: Maktabah Islami, 1983), 4:509.
130 Abu Shāmah, 112. On the same page, he also states that differences between the ʿUthmānic codices were all revealed.
131 Ibn Ḥajar, Fatḥ al-Bārī (Riyadh: Dar al-Taybah, 2005), 11:191. Commenting on Ibn Hajar’s quote, ʿAbd al-Hādī Ḥamītu supports this view and states this concession was present until the unanimous consensus of the companions on the ʿUthmānic codex. Ḥamītu, Ikhtilāf al-qirāʾāt wa atharuhu fī al-tafsīr wa istinbāṭ al-aḥkām (Manama: Islamic Affairs Ministry, 2010), 40.
132 Ibn al-Jazarī, Munjid al-muqriʾīn wa murshid al-ṭālibīn, ed. ʿAli al-ʿImrān (Makkah: Dar al-Fawāʾid 1998), 99.
133 Ibn al-Jazarī, Munjid, 95.
134 In addition to the above questions, there are other matters of uncertainty in the theory of qirā’ah bi-al-maʿnā. Did the Prophet ﷺ himself recite according to meaning? Was the primary impetus for reciting according to meaning the fact that a companion had an imperfect recollection of how the verse was recited by the Prophet and, thus, if they had recourse to a written parchment, would they subsequently revert back to the Prophet’s reading?
135 For a review of such theories, see the doctoral dissertation by Ammar Khatib,  Qirāʾāt al-ṣaḥābah al-khārijah ʿan khaṭṭ al-maṣāḥif al-ʿUthmāniyyah: dirāsah taʾsīlīyah tahlīlīyah (forthcoming). In 2018, the authors of this paper engaged in discussions with some of the proponents of such views. To date, the most significant published work advocating views of this nature is by Ṣāliḥ ibn Sulaymān al-Rājiḥī, al-Masāʾil al-kubra allatī khālafa fīhā al-Qurrāʾ al-Mutaʾakhkhirīn Ijmāʿ al-Mutaqaddimīn min al-Qurrāʾ, (Riyadh: Dar al-Ṣumayʿ 2022). See also Salwa bint Aḥmad al-Ḥārithī, al-Qirāʾah ʿalā sabʿat aḥruf bayna al-ijtihād wa al-tawqīf, BFDA, 36:5 (June 2020): 97-158 and Yasin Dutton, “Orality,” 32–34. More recently, similar views have been advanced in Tareq Moqbel, The Emergence of the Qirāʾāt: The Divine Permission Hypothesis, Journal of Islamic Studies, 2022;, etac032.
136 Arabic translation of Ignaz Goldziher’s work by 'Abdul-Halim Najjar, Madhahib al-tafsīr al-Islāmī (Cairo, 1955). Shady Hekmat Nasser claims that Ibn ʿAṭīyah “openly embraces” this view that the canonical readings are not of Divine or Prophetic origin but rather “the result of the Readers’ interpretation (ijtihād) of the defective Uthmanic consonantal script” (Shady Nasser, The Transmission, 7, fn 9). However, in the provided reference, Ibn Aṭiyyah actually says nothing of the sort. What Ibn Aṭiyyah actually says is that the seven canonical readers used their ijtihād to select whatever variation from the aḥruf they found would conform to the mushaf, not that the variation itself was invented by ijtihādIjtihād was thus in selection, not invention. See Ibn ʿAṭīyah, al-Muḥarrar al-wajīz (Beirut: DKI, 2001), 1:48. The same incorrect claim using the same reference are repeated in Nasser’s second book. See Shady Hekmat Nasser, The Second Canonization of the Qur’an, (Leiden: Brill 2020), 61.
137 See, for instance, al-Aʿẓamī, History of the Qur’anic Text, 155–59.
138 Sunan Abī Dāwūd, no. 1477, https://sunnah.com/abudawud/8/62.
139 al-Naysābūrī, al-Mustadrak ʿalā al-ṣaḥīḥayn, 2:451.
140 Musnad Abī Yaʿla, no. 4022. Ibn Jinnī (d. 392 AH) discusses this and many other examples. While discussing the variant reading ascribed to Anas ibn Mālik for 9:57 and Anas’s statement that yajmahūn, yajmazūn, and yashtadūn mean the same thing, Ibn Jinnī states, “And the apparent meaning of this is that the salaf would recite one letter in place of another without a precedent simply on the basis of agreement in meaning... However, giving the benefit of the doubt to Anas would entail believing that a reading had already preceded with these three words—yajmahūn, yajmazūn, yasthadūn—so it was said ‘Recite with any that you wish.’ So all of them constitute a reading heard from the Prophet due to his saying, ‘The Qur’an has been revealed on seven aḥruf, all of them healing and sufficient.’” See Ibn Jinnī, al-Muḥtasib fī tabyīn wujūh shawadhdh al-qirāʾāt wa al-īḍāḥ ʿanhā (Dar Sazkin, 1406), 1:296.
141 For a comprehensive discussion of these reports, refer to Hamzah Awad, “Sharṭ Muwāfaqat al-Rasm al-ʿUthmānī” (PhD diss., University of Batna, Algeria, 2014), 124–38. If authentic, these reports are simply understood to relate to them expressing a preference for the reading they learned directly from the Prophet ﷺ over a reading they were unfamiliar with.
142 Ṣaḥīḥ al-Bukhārī, no. 4695, https://sunnah.com/urn/180290. Ibn Ḥajar explains that she was likely unfamiliar with the interpretation of the other reading. Ibn Ḥajar, Fatḥ al-Bārī (Cairo: Dār al-Rayan lil-Turath, 1986), 218, https://library.islamweb.net/newlibrary/display_book.php?idfrom=8477&idto=8478&bk_no=52&ID=2464.
143 Ibn Jarīr al-Ṭabarī, Tafsīr, 23:309–10. Ibn ʿAbd al-Barr (d. 463 AH) comments that Ali considered the ʿUthmānic reading to be a valid revealed reading but was expressing a preference for the reading he had learned. Ibn ʿAbd al-Barr, al-Tamhīd (Egypt, 1967), 8:297; see also Hamzah Awad, “Sharṭ Muwāfaqah al-Rasm al-ʿUthmānī,” 138.
144 In verse 17:5, it is stated that Abū al-Sammal al-Adawi recited “faḥāsu” with a ح instead of a ج and when asked about it stated that they meant the same thing. Ibn Jinnī (d. 392 AH) states, “And this indicates that some of the reciters would choose readings without precedent (yatakhayyar bi-la riwāyah)” (Ibn Jinnī, al-Muḥtasib, 2:15). However, Fakhr al-Dīn al-Rāzī (d. 606 AH) states about this variant, “It is necessary to understand that as him mentioning it as tafsīr of the words of the Qur’an, not that he actually made it part of the Qur’an, since if we were to adopt what Ibn Jinnī says it would entail no longer relying on the wording of the Qur’an and it would permit each individual to express a meaning using a word they thought was concordant with that meaning and then either being correct in that belief or erring, and this is an attack on the Qur’an, so it is evident that we must understand it according to what we have mentioned.” See al-Rāzī, Tafsīr al-Fakhr al-Rāzī (Cairo: Dār al-Fikr, 1981), 30:177.
145 Ṣaḥīḥ Ibn Ḥibbān, no. 747.
146 Ibn Mujāhid, Kitāb al-sabʿah fī al-qirāʾāt, 51. Similar statements have been related from Zaid ibn Thabit (al-Bayhaqi, al-Sunan al-kubrā, no. 3900) and Urwah ibn al-Zubayr.
147 al-Nasāʾī, al-Sunan al-kubrā, 2:539.
148 The essence of the story is in Ṣaḥīḥ al-Bukhari, no. 3617, https://sunnah.com/bukhari/61/124; however, the details of the man substituting words are found in Musnad Aḥmad (no. 13312), Ṣaḥīḥ Ibn Ḥibbān (no. 751), and Musnad Abī Dāwūd al-Ṭayālisī (no. 2119), among other sources.
149 al-Qurṭubī, al-Jāmiʿ li-aḥkām al-Qurʾān, ed. ʿAbd Allāh al-Turkī (Beirut: Mu’assasat al-Risalah, 2006), 12:180–81. al-Bayḥaqī, Dalā’il al-nubuwwah (Beirut: DKI, 1988), 7:160.
150 Sa’d said, “Verily the Qur’an was not revealed to ibn al-Musayyib nor his family.” al-Nasa’i, al-Sunan al-kubrā, no. 10996; Sunan Saʿīd ibn Manṣūr, 1:597.
151 al-Ṭaḥāwī, Mushkil al-āthār, 8:132–33.
152 Mustadrak al-Ḥākim, no. 5381. In another narration, we find that ʿUmar considered some of Ubayy’s variant readings to have been abrogated. ʿUmar said, “Ubayy is the best of us in recitation yet we leave some of what he recites. Ubayy says, ‘I have taken it from the mouth of Allah’s messenger so I will not leave it for anything.’ However, Allah says {We do not abrogate or cause a revelation to be forgotten except that we bring that which is better than it.}” (Ṣaḥīḥ al-Bukhārī, no. 5005, https://sunnah.com/bukhari/66/27). As Ibn Ḥajar explains, “ʿUmar is implying that Ubayy may not have been aware of a reading’s abrogation.” Fatḥ al-Bārī (Riyadh, 2001), 8:17.
153 al-Naysābūrī, al-Mustadrak ʿalā al-ṣaḥīḥayn, 2:451. This is analogous to the Prophet ﷺ permitting one who cannot recite any Qur’an to simply repeat subḥān Allāh (glory be to God) or alḥamdu lillāh (praise be to God) instead. See Sunan Abī Dāwūd, no. 832, https://sunnah.com/abudawud/2/442.
154 See Ibn al-Salāḥ, Fatāwā wa masāʾil Ibn al-Ṣalāḥ (Beirut: Dar al-Ma’rifah, 1986), 1:233; Ibn Taymīyah, Majmūʿ al-fatāwā (Mansoura: Dar El-Wafaa, 2005), 13:214; Ibn al-Jazarī, al-Nashr fī al-qirā’āt al-ʿashar, 1:32; al-Suyūṭī, al-Itqān (Beirut: Dar al-Kitab al-Arabi, 1999), 1:263; al-Zarkashī, al-Burhān fī ulūm al-Qur’ān, 1:332–33, who cites Ibn al-Ḥājib.
155 al-Qurṭubī, al-Jāmiʿ li-aḥkām al-Qurʾān, ed. ʿAbd Allāh al-Turkī (Beirut: Mu’assasat al-Risalah, 2006), 21:329–30. He also points out that the narration from Anas being used as proof is in fact a weak narration due to a disconnection in the chain.
156 Adrian Brockett writes, “Thus, if the Qur’an had been transmitted only orally for the first century, sizeable variations between texts such as are seen in the hadith and pre-Islamic poetry would be found and if it had been transmitted only in writing, sizeable variations such as in the different transmissions of the original document of the Constitution of Medina would be found. But neither is the case with the Qur’an.” Adrian Brockett, “The Value of the Hafs and Warsh Traditions for the Textual History of the Qur’an,” in Approaches to the History of Interpretation of the Qur’an, ed. Andrew Rippin (Oxford: Clarendon Press, 1988), 44.
157 See Ibn Taymīyah, Jāmiʿ al-masāʾil (Jeddah: Majma’ al-Fiqhī al-Islāmī, 1422), 1:133.
158 Discussed under the section “Critiquing a different paradigm: qirāʾah bi-l-maʿnā.”

Disclaimer: The views, opinions, findings, and conclusions expressed in these papers and articles are strictly those of the authors. Furthermore, Yaqeen does not endorse any of the personal views of the authors on any platform. Our team is diverse on all fronts, allowing for constant, enriching dialogue that helps us produce high-quality research.