Variant Readings-Study Update

the subject of the variant readings of the Qur’ān was introduced to the reader. (See the article “Variant Readings of the Qur’ān,” ) From the time of Muhammad, there have been multiple readings of the Qur’ān (including the seven, ten, and fourteen approved variant readings listed in the article), expanding by AH 300 (AD 912) to fifty variant readings. Islamic scholars, alarmed by this dilemma, decided that any reading outside the approved list should be considered “deviant”. (See the section Deviant Readings in the article “Variant Readings of the Qur’ān,”). Yet, many of these “deviant” readings “are of equal eloquence to those agreed upon”, as noted by highly respected Islamic grammarian Ibn Jinnī (d. AH 392/AD 1002).

The plethora of readings, both accepted and deviant, does have benefits, however. By comparing and analyzing these readings, researchers may discover or identify linguistic phenomena that are no longer extant. Secondly, the study of the variant readings challenges the claim that the Qur’ān is a “linguistic miracle”: a divine record of Allah’s inimitable words, a sacred book incomparable for its unique literary form. (For some evidence refuting the inimitability of the Qur’ān, see the section What Do These Readings Reveal? in the article “Variant Readings of the Qur’ān,” )

In both Volumes One and Two, the commentary sections related to the variant readings present the important linguistic, grammatical, morphological, or verbal differences among the most pertinent readings of a selected verse. However, in Volume Two these comparisons will include differences seen between the script of early Qur’ānic manuscripts and the writing of the common Arabic language.

Research Methodology

The inclusion of all the variant readings for comparison study in this book was neither feasible nor necessary. Many of these readings appear repetitious because of their strong similarities. Slight scriptural differences among other readings are so inconsequential that no meaning in the verse is changed or lost. Thus, the variant readings that are presented in this book were chosen because of their relevant, creditable differences and the importance of these differences. Criteria were carefully established in selecting these variant readings for study and analysis:

The script of the ‘Uthmānic codex (mušhaf ), or its representative ‘Uthmānic text according to the reading of Hafš, was the primary focus for this comparative study regarding the variant readings. (Please know, however, that the original ‘Uthmānic codex, compiled about thirteen years after Muhammad’s death and is no longer extant, was devoid of hamzas, dotting, and diacritical marks, such as fatha and ďamma, and other punctuation. Evidence concerning the absence of these marks and other scriptural features, sometimes called consonantal script or rasm, in the ‘Uthmānic codex is supported by the same absence in most of the extant early Qur’ānic manuscripts, which are based upon this codex.

Note: Inclusion of this script in the comparative study is unique to Volume Two.

Selected samples included variant readings that have been previously analyzed and documented by Islamic scholars who specialize in this kind of study. For Volume Two, some of these renowned and respected sources related to this research include Kitāb al-Mašāhif (al-Sijistānī), Al-Muhtasib fī Tabyīn Wujūh Shawādh al-Qirā’āt wa al-Iďāh ‘anhā and Al-Khasā’is (Ibn Jinnī), Mu‘jam al-Qirā’āt (al-Khaţīb), I‘rāb al-Qirā’āt al-Shawādh (al-‘Ukbarī), and Mu‘jam al-Qirā’āt al-Qur’ānīya (‘Umar and Makram).

Research of the verses and their variant readings encompassed the review and evaluation of the multiple meanings and interpretations provided by the principal Islamic commentators (e.g., al-Ţabarī, al-Zamakhsharī, al-Rāzī, al-Qurţubī, Abū haīyān, Ibn Kathīr, Ibn al-Jawzī, al-Bayďāwī, al-Shanqītī, and Ibn ‘Āshūr) regarding these verses with their variant readings.

Final selection of the variant readings provided in Volume Two for discussion was first culled from readings that are highlighted in the major Islamic commentaries and references. From this initial sample was gleaned the most suitable reading, when possible, according to the general context of the verse(s) and the entire sura (chapter) in order to elucidate the intended meaning.

The resultant analysis and critique of these selected variant readings will be supported with the aforementioned references, early Qur’ānic manuscripts, and some recent related research. The addition of the early manuscripts to the comparative study of the variant readings will expand and enhance the commentary for the verses examined in Volume Two, providing readers with even more information and plausible explanation behind the differences among the variant readings.

Early Qur’ānic Manuscripts

Volume Two expands the analysis and comparison study of the variant readings by including the script of the ‘Uthmānic codex and early Qur’ānic manuscripts: Paris manuscript (MS Arabe 328a), London manuscript (MS Or. 2165), Samarkand codex, San‘ā’ manuscript (MS DAM 01-27.1), Cairo manuscript, and the Topkapi manuscript.

The study of the early manuscripts allows for a closer look at the early readings of the Qur’ān and the evolution of variances in the subsequent readings. These early manuscripts are also powerful sources for improving the process of clarifying and resolving the variances among the many readings.

Readers should note that the early manuscripts were written without any diacritical marks. Since the use of dots, marks, and the hamza (glottal stop indicator) was added to the written Arabic language almost 100 years after Muhammad’s death, researchers know that readings based on these marked scripts were created later and are thus more removed from the earliest sources. Obviously, these scriptural differences between the early manuscripts without diacritical marks and the later codices that include such marks contribute to the differences in both the Qur’ān’s text and its readings.

By way of illustration, the study of the early San‘ā’ manuscript refutes the claim that the codex of Ibn Mas‘ūd and the codex of ’Ubayy Ibn Ka‘b contain deviant readings because these two codices do not fully agree with the ‘Uthmānic codex manuscript. In fact, the San‘ā’ manuscript differs frequently with the ‘Uthmānic codex manuscript. One example can be seen in Q 9.11, where these two manuscripts show differences in words, syntax, and meaning (identified with boldfacing):

‘Uthmānic codex: “...then they are your brethren in religion—we detail the signs unto a people that do know.” (Palmer English translation)

San‘ā’ manuscript: “...then they are your brethren in religion and your servants—God details the signs so that you may understand”.

Such discrepancies uncovered through the comparison study of the early manuscripts with the variant readings question the infallibility of ‘Uthmān’s committee during the compilation of the Qur’ān.

Problems within the Qur’ān’s Text

The commentary regarding the variant readings on the selected verses in Volume Two will concentrate on one or more of the following problems that exist in the ‘Uthmānic codex.

Dotted Letters

According to Islamic sources the early codices of the Qur’ān did not include any dotting, diacritical marks, or hamzas. This absence of these markings is clearly evident in early Qur’ānic manuscripts, such as the Paris and Samarkand manuscripts, and led to many variances in readings.

The following examples show the changes that can occur when dots were later added to the undotted script, based on the reader’s understanding of the verse:

(Undotted script, Q 24.63)

بينكم (Dotted letters B-Y-N-K-M): “Make not the calling of the Apostle amongst yourselves like your calling one to the other….” (Majority reading)

نبيكم (Dotted letters N-B-Y-K-M): “Make not the calling of the Apostle your prophet like your calling one to the other….” (Reading by al-Hasan al-Bašrī)

(For more information about this particular verse and its variant readings, see the comment on Q 24.63).

(Undotted script, Q 12.16)

(Dotted letters ‘-Sh-A-’): “And they came to their father at eve and weeping” (Majority reading)7

(Dotted letters ‘-Sh-A): “And they came to their father at dinner weeping” (Reading by al-Hasan al-Bašrī)

Yet if one changes the reading of al-Ģasan al-Bašrī by one dot to make it (ghishan with dotted letters Gh-Sh-A), the word becomes “deceitfully,” which is arguably a more logical rendition of this undotted script, given its context. (For more discussion, see the comment on Q 12.16, page 122.)

Diacritical Marks

The variance in readings because of the diacritical marks (e.g., fatģa ( ), damma ( ), kasra ( ), sukūn ( ), etc.) being applied differently to the script is very common. (See the table “Common Arabic Diacritics” in the article “Key to Reading Part II”) Of course, some of these differences in the use of these marks do not affect meaning. However, word differences created by differently marked script can profoundly change word choice and consequently the overall meaning of the verse, as exemplified here in Q 19.5:

“But I fear (خِفْتُ; Kh-i-F-T-u) my heirs after me, and my wife is barren; then grant me from Thee a successor” (Majority reading)

“But my heirs decreased خَفَّتِ); Kh-a-Ff-a-T-i) after me, and my wife is barren; then grant me from Thee a successor” (Readings by ‘Uthmān Ibn ‘Affān, Zayd Ibn Thābit, Ibn ‘Abbās, and others)

(For more discussion about these variant readings for this verse within its passage context, see the comment on Q 19.5).

Multiple Readings of a Single Word

Some of the words of the Qur’ān have so many variant readings that they draw attention. For example, in Q 22.40 the word salawātun, translated by Palmer as “synagogues,” has more than eighteen variant readings. (See the comment on Q 22.40) In Q 27.66, the current Arabic Qur’ān reading uses bali addaraka (“nay, but…attains” ). However, there are at least sixteen variant readings for this expression: bal adraka, bal addaraka, bal ādraka, bal aāddaraka, bala adraka, balā aaddraka, balā addāraka, bal tadāraka, bala addrak, am tadārak, balā ādraka, bali addaraka, am adraka, bal aaddaraka, balā addrak, and bal āaddarak.

These extremely numerous variants of a word or expression in these verses reveal the confusion and uncertainty by Islamic scholars in their efforts to provide the correct meaning for the ambiguous texts. Unfortunately, the resulting multiple readings make it even more difficult to determine the intended meaning.

Pronoun Gender Disagreement

The Arabic language classifies nouns, pronouns, and even adjectives as masculine or feminine. Like the English language, where pronouns and their antecedents must agree in number (singular or plural) and gender (masculine, feminine, or neuter) to be grammatically correct, the Arabic language has similar grammatical rules regarding the proper gender form of demonstrative pronouns and the nouns they modify. For instance, the masculine demonstrative pronoun hadhā (“this”) and the feminine demonstrative pronoun hadhihi (“this”) should respectively modify masculine and feminine nouns.

Yet, in some places in the Qur’ān these demonstrative pronouns are used to modify nouns that are not the same gender—even, at times, in the same sura. For example, in Q 18.98, the phrase “This [hadhā, masculine pronoun] is a mercy [raģma, feminine noun]” is used instead of “This [hadhihi, feminine pronoun] is a mercy [raģma, feminine noun]” according to the reading of Ibn Abī ‘Abla. Another example found in the Samarkand manuscript (Q 7.73) uses the phrase “This [hadhā, masculine pronoun] she-camel” instead of “This [hadhihi, feminine pronoun] she-camel.” It should be noted that this same phrase appears correctly in the same manuscript in Q 11.64 as “this [hadhihi, feminine pronoun] she-camel”.

Scriptural Differences

The script of the ‘Uthmānic codex contains many examples of words spelled differently in different places. Sometimes these different spellings occur within the same sura. Even though these different spellings generally do not change the meaning of these words, the inconsistency is puzzling. Because there is no explanation given, these anomalies appear random or suggest multiple authors for the codex or one careless author.

Alif

The first letter of the Arabic alphabet is the alif (ا), a character with two functions: (1) It represents the long /a:/ vowel, and (2) it signifies a glottal stop /أ/ when written with the hamza.

For reasons unknown, the alif is used interchangeably with another Arabic character, the wāw (و), in the spelling of certain words in the script of the codex. The following table presents a few of these samples:

Sample 1. Scriptural Differences

Arabic characters

Transliteration

English translation

Verse

Written

wāw (و )

Pronounced alif (ا)

Q 2.85

حيوة

حياة

hayāt

life

Q 24.56

زكوة

زكـاة

zakāt

required alms

Q 24.58

صلوة

salāt

prayer

Note: Word is written with a wāw (و) but pronounced with an alif (ا).

Sample 2. Scriptural Differences

Arabic characters

Transliteration

English translation

Verse

Written alif (ا)

Verse

Written wāw (و)

Pronounced alif (ا)

Q 23.33

الْمَلَأُ

Q 23.24

الْمَلَؤاْ

الْمَلَأُ

al mala’

public

Q 24.31

أبناء

Q 5.18

أبناء

abnā’

sons

Note: Word is sometimes written with an alif (ا) or wāw (و) but always pronounced with an alif.

(For more discussion on these scriptural differences, see the comment on Q 23.24)

Sometimes the same word will be spelled with an alif and sometimes with a different Arabic character, the yā’ (ي), but still always read with an alif:

Sample 3. Scriptural Differences

Arabic characters

Transliteration

English translation

Verse

Written

yā‘ (ي)

Pronounced alif (ا)

Q 2.29

فسوىهن

فسواهن

fasawāhuna

fashioned

Q 7.89

نجىنا

نجانا

najjānā

saved us

Q 11.41

مجرىها

مجراها

majrāhā

sailing

Q 11.41

مرسىها

مرساها

mursāhā

anchoring

Note: Word is sometimes written with a yā’ (ي) but always pronounced with an alif (ا).

These scriptural letter changes—the alif with the wāw or the yā’—do not exhibit a patterned variation to suggest a grammatical rule for these differences. Instead, such randomness in these spelling differences underscores one of the obvious inconsistencies in a book acclaimed by Muslims to be a linguistic miracle.

In addition to these scriptural letter exchanges, the codex script contains words where the long vowel alif is inserted in the middle of a given word and then inexplicably omitted from this word in other locations. Unlike the aforementioned scriptural letter exchanges (alif-wāw-yā), the inclusion or omission of the medial alif does create problems in word meaning.

Singularity vs. Plurality

In Q 10.94, for example, the word

can mean “book” or “books,” depending on the inclusion or omission of a medial alif and subsequent diacritical marks. According to Islamic scholars, if the word is “book,” the inference is that the “book” may mean the Bible whereas the word “books” means a collective work, e.g., Torah, Psalms, Injīl, or other books. Obviously, this difference is an important distinction in the verse’s context: And if thou art in doubt of that which we have sent down unto thee, ask those who read the Book before thee; verily, the truth is come to thee from thy Lord, be not then of those who are in doubt (Palmer’s translation uses the singular form.)

Because the early manuscripts rarely show medial alifs, the variant readings provide evidence that the singularity or plurality of this word was not definitive. However, through the centuries this ambiguity is rectified by the inclusion of a medial alif (and later, diacritical marks) that establishes

as the singular “book.” The following table shows the written progression and determination of this word:

Sample 1. Progressive Institution of Medial Alif

Example: Q 10.94

Script

English translation

Earliest manuscripts

(no diacritical marks or alif )

book/books

Current Arabic Qur’ān

(diacritical marks and embedded alif)

book

Agent vs. Action

In another example, Q 10.2, the unmarked becomes “sorcerer” (agent) or “sorcery” (action) in the codex and its variant readings. Over time, this word has been established as “sorcery”: and when the truth came to them from us they said, verily, ‘This is obvious sorcery.’ (Palmer’s translation uses the word “sorcery.”)

The following table shows the written progression and determination of this word:

Sample 2. Progressive Institution of Medial Alif

Example: Q 10.2

Script

English translation

Earliest manuscripts and Warsh

(no alif )

sorcery

Current Arabic Qur’ān (Ģafš)

(with alif )

sorcerer/sorcery

Tā’ Marbūţa

An Arabic grammatical symbol that is only written at the end of nouns and adjectives is the tā’ marbūţa (ة). Its presence there indicates that the word is feminine in gender. For example, the word zahrat (“flower”) is considered feminine and would be written as

.As its form and name suggest, the tā’ marbūţa is related to the Arabic consonant tā’ (ت), which represents the /t/ t/sound. When a possessive pronoun is added to a feminine word, the tā’ marbūţa changes in form from a “closed” tā’ (ة) to an “open” tā’ (ت) but still pronounced as a /t/. For example, طاولة (ţāwilat or “table,” a feminine noun) becomes طاولتي (ţāwilatī or “my table”).

In today’s Arabic language, these rules of grammar and use concerning the tā’ marbūţa are applied predictably and consistently. Unfortunately, these rules are not visible in the script of the codex, which contains numerous examples of the same word spelled and pronounced differently without any regard to, or acknowledgment of, these grammatical rules. For example, in Q 2.211 the feminine word ni‘mat, which means “grace or favor,” appears with a tā’ marbūţa (نعمة), which follows today’s Arabic grammatical rules. Yet, in verse 231 of the same sura, the word ni‘mat appears with a regular tā’ (نعمت) without a reason for the change.

In an attempt to remove the randomness of these inconsistencies in the Qur’ān regarding the use of the tā’ marbūţa, Islamic scholars developed the following grammatical rule:

Any feminine noun that ends with an /a/ or /h/ sound and is part of an idāfa structure, e.g., نعمة الله (“favor of God”), then the noun can be written with either a closed tā’ (ة) or an open tā’ (ت).

However, the script of the codex does not always follow this rule. In Q 35.40, the expression bayinat (“manifest sign”) was written with an open tā’ (بَيِّنَتٍ ). Yet in Q 11.17, the same expression bayinat was written with a closed tā’ (بَيِّنَة ). Even though these two cases are not part of an idāfa structure, they too do not follow this proposed rule. (See the comment on Q 28.9).

Sample 4. Scriptural Differences

Verses(s)

Written

alif (ا)

Verses(s)

Written

tā’ (ت)

Q 2.126

بلدا

Q 25.49, Q 43.11, Q 50.11

بلدة

baladā

balada

Q 19.11, Q 19.62,

Q 30.18, Q 40.46

عشيا

Q 79.46

عشية

‘ashiyā

‘ashiya

Q 40.46

غدوا

Q 79.46

غدوة

ghudwā

ghudwa

Hamza

Even though this character has no place of its own in the Arabic alphabet, the hamza (ء) is a special character, one that is a carrier of vowel sounds. It acts as a glottal stop //, similar to the sound an English-speaking person articulates between the two syllables of the expression, “uh-oh.” In the English language, this sound has no letter equivalent; in the Arabic language, this sound is represented by the hamza. (Please note that the hamza is indicated by an apostrophe in transliterated text, e.g., “Qur’ān.)

Obviously, the hamza’s position and presence in a word will directly affect the word’s pronunciation. Yet early Qur’ānic manuscripts contain not one hamza because this character was not introduced into the written Arabic until lexicographer and philologist Khalīl Ibn Aģmad al-Farāhīdī (c. AH 110/AD 169 - c. AH 173/AD 786) added the hamza and refined the diacritical marks when he compiled the first Arabic language dictionary. Tayyar Altıkulaç, Turkish scholar of Arabic paleography and calligraphy who prepared and edited the Topkapi manuscript, states that this “Mušģaf does not include a hamza (ء) anywhere”.

Despite this lack of hamzas in the earliest Qur’ānic manuscripts, the different readings (as documented in the Islamic sources) do reflect the use of hamzas. However, their use is unsystematic. In Q 15.1 the word قُرْءآنِ (“Qur’ān”) was read by ‘Abd Allah Ibn Kathīr and Ibn Muhaysin in its older form (without a hamza), or قُرَانِ (“qurān”). These readings without the use of hamzas were repeated in many of the suras, including (but not limited to) Q 16.98, Q 20.2, Q 25.30, and Q 27.1. For instance, in Q 18.98 the word dakkā’ or دَكَّاءَ(“as dust”) was read by ‘Abd Allah Ibn Kathīr without a hamza, or dakkā ( دَكَّا).

Therefore, these earliest readings serve as evidence that hamzas were not present in the early codices. On the same hand, the Arabic dialect of North Africa and even most of the colloquial dialects of the rest of the Arab world today, retain the Arabic language in its original form without hamzas. When the peoples in these areas became Muslim during Islam’s spread into Africa and the Middle East during the seventh and eighth centuries AD, they adopted Islam’s language of that time—the Arabic tongue before al-Farāhīdī added the hamza. In this manner, the original Arabic language without the hamza became widespread. When the hamza was eventually added to the Arabic language, this version became the formal but not the colloquial language of most Arabic-speaking peoples.

Comparison between Original Arabic

and Classical Arabic Language

Transliteration

Classical

Arabic

(with hamza)

Transliteration

Original

Arabic

(without hamza)

Sample English Word

‘Ā’isha

عائشة

‘Isha

عيشة

‘Ā’isha

samā’ zarqā’

سماء زرقاء

samā zarqā

سما زرقا

blue sky

‘ishā’

عشاء

‘ishā

عشا

dinner time

fā’r

فأر

fār

فار

mouse

lū’lū’

لؤلؤ

lūlū

لولو

pearl

dhi’b

ذئب

dīb

ديب

wolf

Comparison between Two Classical Arabic Readings

Ģafš (with hamzas)

Warsh (hamza-less)

Verse

Reading

Written script

Reading

Written script

يُؤمِنُونَ

يُومِنُونَ

Q 2.3

yū’minūna

yūminūna

يَّأْكُلَهُ

يَّاْكُلَهُ

Q 12.13

yã’kulahu

yãkulahu

الذِّئْبُ

الذِّيبُ

Q 12.13

al-dhi’b

al-dhīb

Note: Samples show hamza in Ģafš reading after different preceding vowels.

Samples of Words Changed over Time

H.S. 44/32

(Topkapi MS)

Readings of Word

(includes dots)

Current Arabic Qur’ān

(includes hamzas)

Arabic Transliteration

English Translation

al-ruiyā → al-ru’iyā

vision, dream

nisānā → nisā’nā

our wives

yastanbūnaka → yastanbi’ūnaka

they ask you to inform them

mustahzūna → mustahzi’ūna

mocking

Letter Additions and Omissions

The Arabic Qur’ān of today contains words with extra letters not present in the earliest manuscripts. An example is the majority reading of nunjī (“we save”) in Q 21.88 where the word ننجي is written with two n’s. This spelling is different from the script of the ‘Uthmānic codex (ىحى), which shows only one n and no dotting. Also, in the current Qur’ān the word appears with one of the n’s as a superscript above the other n:

Others read this word as najjā (نَجَّى) and also nujjīy (نُجي). It appears that this word might have originally been nuģīy (نُحِى), which means “quicken.” (See the comment on Q 21.88).

The script of the ‘Uthmānic codex also contains words where a letter has been omitted in some instances but not in others. For example, the word ‘ayyuha (أَيـُّهَ) appears without the alif only three times but with an alif (‘ayyuhā or أَيـُّها) in more than a hundred places. (See the comment on Q 24.31).

Connected Letters

Sometimes the variant readings indicate that a “non-connecting” letter (letters that are normally written detached from other letters before or after them) has been joined with the neighboring letter to create a different letter and thus a different word (which sometimes creates a different meaning as well).

Companion and commentator Ibn ‘Abbās mentions two verses, which are relevant examples to help explain this problem:

In Q 17.23 the expression waqaďā or وقضى (“has decreed”; q is connected to ď, or ) is actually wawašā (“has commanded”; q is detached). In all likelihood, the scribe confused the second wāw (و) with the letter q (ق); by connecting the wāw (و) and the š (wص), he changed the word, resulting in a different expression: . When dotting was added later, the expression became waqaďā or (وقضى), which means “has decreed.”

In Q 24.27 the expression tasta’nisū or تستأنسوا was read as tasta’dhinū or تستأذنوا by Ibn ‘Abbās, who indicated that the scribe had made a mistake in copying this expression. Ibn Mas‘ūd, ’Ubayy Ibn Ka‘b, al-A‘mash, and Sa‘īd Ibn Jubayr all read this expression the same as Ibn ‘Abbās.

However, commentator Ibn ‘Aţīyya alleges that Ibn ‘Abbās’s is a weak reading and adds that the statement concerning the scribe making an error is not appropriately attributed to Ibn ‘Abbās. He states that it is an interpretive reading. But the known readings of ’Ubayy Ibn Ka‘b and others, which corroborate the reading of Ibn ‘Abbās, rebut Ibn ‘Aţīyya’s position. Even more importantly, the expression tasta’dhinū is also found in the codex of Ibn Mas‘ūd. (See the comment on Q 17.23 and 24.27).

Word Errors

From Islamic readings and commentaries, the script of the ‘Uthmānic codex also contains erroneous words because of changes in the characters during transcription. The following examples illustrate these mistakes:

Q 13.31 includes the word yay’as or , meaning “despair.” The origin of this word is yatabayan or يَتبيَّن(“discern”), according to Ibn ‘Abbās, who comments that the scribe wrote the word incorrectly “while he was sleepy”.

Q 20.63 includes the word hadhāni or هَٰذَٰنِ, meaning “these twain.” ‘Ā’īsha states that “this [word] is an error by the scribe”. Abū ‘Amr Ibn al-‘Alā’ and ‘Īsā Ibn ‘Umar agree that “it is a mistake by the scribe, as stated by ‘Ā’īsha”. It is worth noting that all of these persons referenced here have read the verse as “‘These twain [hadhāyni or هذين] are certainly two magicians….’” While both readings of this word have the same meaning, the first one (hadhāni) is grammatically incorrect in the Arabic language. (See the comment on Q 13.31 and Q 20.63).

Synonymous and Additional Text

The variant readings also show scriptural differences from the ‘Uthmānic codex by using phrases that may be similar in meaning but do not match exactly, word for word. Also, some variant readings contain text not found in the script of the codex.

The following verses are two examples of synonymous text:

In Q 18.53 the phrase “they are going to fall therein” was read by Ibn Mas‘ūd as “they are going to face it....”

In Q 20.40 the text “And we restored thee” is written in the codex of ’Ubayy Ibn Ka‘b as “And we returned thee….”

The following verses are two examples of variant text:

In Q 11.71 the text “And his wife was standing by…” was read by Ibn Mas‘ūd as “And his wife was standing by, and he was sitting” with the phrase “he was sitting,” which is not found in ‘Uthmānic codex.

In Q 18.80 the text “his parents were believers” was read by Ibn ‘Abbās as “he was a misbeliever but his parents were believers” and read by ’Ubayy Ibn Ka‘b as “his parents were believers but he was a misbeliever” (Differences in both verses shown with added emphasis).

Word Changes and Inconsistencies

Some of the variant readings contain entirely different words not found in the majority reading. As an example, the phrase “and the stars are subjected” in Q 16.12 is written as “and the winds are subjected” in the codex of Ibn Mas‘ūd, where the word “stars” is used instead of “winds.”30

An example of vocabulary discrepancies in the Qur’ān can be found in Q 23.17. (See inset text “Sample of Qur’ānic Vocabulary Inconsistencies.”)

Sample of Qur’ānic Vocabulary Inconsistencies

Muslim exegetes state that the word tarā’iq (“roads”) means the heavens. They are called roads for several reasons:

They are layered in layers, like garments, on top of each other. They are paths for the angels to ascend, descend, and fly. They are the pathways of the planets as they travel, the source of rains for subsistence, the dwellings of the angels, the destinations of the prophets, and the origin of revelation sent down. This same word, tarā’iq, appears in another location (Q 72.11) with a different meaning: “And of us are some who are pious, and of us are some who are otherwise: we are in separate bands.” (Palmer trans.). Muslim exegetes explain the expression here “we are in separate bands” to mean “we have variant inclinations”.

According to the Arabic concordance, Lisān al-‘Arab, the word tarā’iq has multiple meanings depending on the context. In the context of time, tarā’iq refers to its fluctuation and volatility; in the context of people groups, the word means the noblemen of the people.

Given this word’s multiple and unrelated meanings, why is tarā’iq used here? Even when it is said to refer to the heavens, the exegetes offer differing opinions as to how it could refer to the heavens.

The ambiguity here appears to be a scribal error. If one is to look at the inscription without the dots, the word appears as . It is quite possible that the original word was , with a wāw (و) instead of the consonant r (ر). Then, the word must have been mistakenly read as tarā’iq. Evidence supporting this hypothesis is found in an early Qur’ānic manuscript (MS Arabe 328a), where the Arabic word وكفى in Q 4.70 is written , treating the letter و as if it was a . If this hypothesis is true, then the verse was originally “…we have created above you seven floors…,” in reference to the seven heavens in Islam.

Key Terminology

During study of the variant readings, the reader will benefit from a review of important Islamic terms and expressions to better understand their significance. The following list provides the reader with a brief description for each of these terms or directs the reader somewhere else in The Qur’ān Dilemma for a more extensive explanation.

Variant Readings

For the most comprehensive explanation of this term, see the article “Variant Readings of the Qur’ān.

Reading vs. Codex

There is a difference between the reading of ‘Uthmān Ibn ‘Affān and his codex or the reading of Ibn Mas‘ūd (or ’Ubayy or others) and his codex:

Reading of ‘Uthmān and the codex of ‘Uthmān

The “reading of ‘Uthmān” refers to his own reading from his personal codex (“codex of ‘Uthmān”). The “‘Uthmānic codex” refers to the committee compilation of the Qur’ān ordered by ‘Uthmān when he became the third caliph. (See the article “Compilation of the Qur’ān,” Qur’ān Dilemma 1: 47.) Despite having ‘Uthmān’s name associated with both his personal reading and the compiled codex, they are not identical. For example, in Q 18.79 in the reading of ‘Uthmān, the phrase “taking every good ship” is not replicated in the ‘Uthmānic codex, which states “taking every ship” (Pickthall trans.). This and other differences between ‘Uthmān’s reading and the ‘Uthmānic codex are understandable because these two versions were drafted by different people.

Reading of Ibn Mas‘ūd and the codex of Ibn Mas‘ūd

According to Islamic sources, Ibn Mas‘ūd had some readings that differed from his own codex. As an example, in Q 18.16 the phrase “except Allah” (Pickthall trans.) was read by Ibn Mas‘ūd as min dūni Allāhi (“without Allah”). Yet the codex of Ibn Mas‘ūd has min dūnina (“without us”) instead of “except Allah”.

As readers review other commentary on the variant readings presented in this book, they should notice that these “readings” attributed to Ibn Mas‘ūd are often closer than his codex to the script of the ‘Uthmānic codex. The claim that there is a difference between the reading of Ibn Mas‘ūd and his codex is absurd. Why would a person write something down one way and then read it in a different way, especially when it concerns the Qur’ān? Perhaps these readings have been attributed to Ibn Mas‘ūd in an attempt to downplay the importance of his codex and to unify the Qur’ānic text. Similar claims about reading and codex differences have been applied as well to the reading of ’Ubayy Ibn Ka‘b and his codex and others who had their own codices. (See the comment on Q 11.71 and Q 24.35).

Current Arabic Qur’ān Reading

When there is no source to prove that the reading is the reading of the majority, or the public, The Qur’ān Dilemma classifies the reading as the “current Arabic Qur’ān reading”. This term “current Arabic Qur’ān reading” refers to the reading of Hafš according to ‘Āšim, which is the most famous and widespread reading among Muslims today. As a result, it is the foremost reading among other readings, such as the readings of Warsh, al-Dūrī, Shu‘ba, Qālūn, and al-Sūsī. It is worth mentioning that most of the Qur’ānic commentaries, Islamic studies and research, and translations of the Qur’ān are based on the reading of Hafš according to ‘Āšim.

Conclusion

The study of the variant readings is important on several levels. At its core, the multiplicity of readings challenges the Islamic doctrine stating that the Qur’ān is—letter for letter, word for word—the divine revelation given by Allah. The variances in the readings and the codices dispute this claim of inerrancy.

Historically, the readings also provide a tantalizing window into the past as “oral codices” of early Qur’ānic manuscripts no longer extant.

And now, in Volume Two, these variant readings have been compared with early “unmarked” Qur’ānic manuscripts that have been discovered to trace the origin, age and development of their differences. The commentaries on selected verses presented here in this volume will reflect the enriched analysis afforded by these additional sources, providing readers with an even more in-depth understanding about how these variances may have occurred and why they matter.

References

1. al-Ashwah 61-63.

2. Ibn Jinnī, Al-Muģtasib 1: 32.

3. E. Puin, “Ein früher Koranpalimpsest aus Šan‘ā’ (DAM 01–27.1), Teil II” 552; see folio 4v,

Lines 10-11.

4. al-Dānī 2, 6, 18; Nīsābūrī 2: 251.

5. al-Khaţīb, Mu‘jam 6: 309.

6. ‘Umar and Makram 4: 270; al-Khaţīb, Mu‘jam 6: 309.

7. al-Khaţīb, Mu‘jam 4: 204.

8. Ibid.; ‘Umar and Makram 3: 156.

9. al-Khaţīb, Mu‘jam 5: 338.

10. Ibid.

11. ‘Umar and Makram 4: 365-368; al-Khaţīb, Mu‘jam 6: 545-550.

12. al-Ţabarī, Tafsīr 22: 51.

13. Nīsābūrī 2: 251.

14. Altıkulaç, [Topkapi MS] 85 (English text), 94 (Arabic Text); see point 3.

15. al-Khaţīb, Mu‘jam 3: 508, 553.

16. Ibid. 5: 312.

17. Abū haīyān 6: 311.

18. Ibid.; compare with al-Qurţubī 14: 276.

19. al-Khaţīb, Mu‘jam 5: 37 and 6: 252, respectively. See al-Suyūţī, Al-Durr al-Manthur 9: 287

(Q 17.23) and al-Ţabarī, Tafsīr 17: 240 (Q 24.27).

20. Ibn ‘Aţīya 4: 176.

21. Ibn Kathīr, Tafsīr 10: 207.

22. al-Ţabarī, Tafsīr 13: 537.

23. al-Khaţīb, Mu‘jam 5: 451.

24. Ibid.

25. al-Khaţīb, Mu‘jam 5: 451, 452.

26. Ibid. 5: 242.

27. Ibid. 5: 434.

28. Ibid. 4: 98.

29. Ibid. 5: 283-284.

30. al-Sijistānī 320 (2002).

31. al-Rāzī, Tafsīr 23: 88; compare with al-Zamakhsharī 4: 222.

32. al-Ţabarī, Tafsīr 23: 330.

33. Ibn Manžūr 2665.

34. See MS Arabe 328a: folio 14b.

35. al-Ţabarī, Tafsīr 15: 182.

36. al-Khaţīb, Mu‘jam 5: 161.