Published online 28 May 2022

Journal of Islamic Studies (2022) pp. 299 of 330 https://doi.org/10.1093/jis/etac032

THE EMERGENCE OF THE QIR282T: THE

DIVINE PERMISSION HYPOTHESIS

TAR E Q MO Q B E L *
Regent’s Park College, University of Oxford

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

INTRODUCTION

The topic of qir:8:t, the textual variants of the Qur8:n, is a difficult,

sensitive and contentious one. Yet, it has had enduring influence on
Muslim understanding of the Qur8:n—Bauer did not go too far when
he wrote that ‘the discipline of the Quranic readings has led to an in-
tensive reflection on the nature of the Quran’.1 This paper is an attempt
to revisit this topic. It focuses, out of various domains of 6ilm al-qir:8:t,2
on the issue of the divine origin of the variant readings. In asking
whether all the variant readings were considered by Muslim scholars
to be of divine origin, it takes on the question of what they understood
the source of the variants to be. It aims to show that, according to

* Author’s note: My thanks to the anonymous readers of this Journal for

their comments on an earlier draft of this paper. All translations herein are
mine except where stated otherwise.
1
Thomas Bauer, A Culture of Ambiguity: An Alternative History of Islam
(transl. Hinrich Biesterfeldt and Tricia Tunstall; New York: Columbia University
Press, 2021), 37; Bauer, Die Kultur der Ambiguität: Eine andere Geschichte des
Islams (Berlin: Verlag der Weltreligionen, 2011), 64.
2
For accounts of the qir:8:t and the different trajectories in the scholarship,
see Theodor Nöldeke, Friedrich Schwally, Gotthelf Bergsträsser, Otto Pretzl,
The History of the Qur8:n (ed. and transl. Wolfgang Behn; Leiden: Brill, 2013),
471–584; Ahmad 6Ali Al-Imam, Variant Readings of the Qur8an: A Critical
Study of Their Historical and Linguistic Origins (Herndon, VA: International
Institute of Islamic Thought, 2006); Shady Hekmat Nasser, The Transmission of
the Variant Readings of the Qur8:n: The Problem of Taw:tur and the
Emergence of Shaw:dhdh (Leiden: Brill, 2013). See also, more recently, the
surveys in Suheil Laher, ‘Between dogmatism and speculation: a critical assess-
ment of Qir:8:t studies’, Journal of College of Sharia and Islamic Studies, 38/1
(2020): 176–90; and Mustafa Shah, ‘The corpus of Qur’anic readings (qir:8:t):
history, synthesis, and authentication’ in Mustafa Shah and Muhammad Abdel
Haleem (eds.), The Oxford Handbook of Qur’anic Studies (Oxford: Oxford
University Press, 2020): 194–216.

ß The Author(s) (2022). Published by Oxford University Press on behalf of the Oxford Centre for Islamic Studies.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial
License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution,
and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please
contact journals.permissions@oup.com
300 t a r e q mo qb e l
eminent classical Muslim scholars,3 not all the variant readings were
regarded as divinely initiated; some of them were not necessarily initi-
ated by God. This position did not seem, in their view, to pose any threat
to their theology, nor to their understanding of the Qur8:n being divinely
protected and preserved (Q. 15:9). The question of whether or not emi-
nent Muslim scholars believed that some of the variant readings fell
short of a divine status4 leads into an exploration of how that belief
could be deployed to deal with some of the issues to which the existence

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

of the qir:8:t gave rise.
Before I begin, a number of preliminaries need attention. First of all,
the qir:8:t, qua variants, are by and large, minor, not substantive, differ-
ences: ‘Regardless of how and when variant readings emerged, there is
no denying that the differences among these variants are miniscule, to
the extent that they could often be inconsequential at the hermeneutical
level’.5
Secondly, I should stress here the importance of distinguishing be-
tween two things: the emergence of the qir:8:t as envisaged by the
tradition, the concern of this study, and the actual emergence of the
qir:8:t as it happened historically, for which a (different) study of ma-
terial evidence and manuscripts is necessary.6 What the tradition
believed happened and what ‘really’ happened are not necessarily the
same. The contention here is that eminent scholars until at least the

3
This paper mainly focuses on a sample of scholarly opinions until Ibn al-Jazar;
(d. 833/1429), though it does also present briefly the views of some modern schol-
ars. Ibn al-Jazar; is the cut-off point given that his system of qir:8:t became the
standard model, and has remained prevalent to this day. See Shady Hekmat Nasser,
The Second Canonization of the Qur8:n (324/936): Ibn Muj:hid and the Founding
of the Seven Readings (Leiden: Brill, 2021), 7–8. (Study of the discourse between
Ibn al-Jazar; and the modern period must be deferred to a future paper.)
4
To this end, I am following the general idea of Shahab Ahmed, Before
Orthodoxy: The Satanic Verses in Early Islam (Cambridge, MA: Harvard
University Press, 2017), 1–10.
5
Nasser, The Second Canonization of the Qur8:n, 184. Likewise, Benham
Sadeghi and Mohsen Goudarzi (‘4an6:8 1 and the origins of the Qur8:n’, Der
Islam, 87 [2012]: 1–129, at 19) note that whereas ‘most variants do not affect
the meaning significantly’, only a ‘small fraction of the variants do make a differ-
ence in meaning’. John Burton (The Collection of the Qur8:n [Cambridge:
Cambridge University Press, 1977], 171) makes a similar point: ‘no major differ-
ences of doctrine can be constructed on the basis of the parallel readings based on
the ‘Uthm:nic consonantal outline, yet ascribed to muBAafs other than his . . . None
of these variants is of great import’.
6
See, for example, the recent attempt by Marijn van Putten, Quranic Arabic:
From its Hijazi Origins to its Classical Reading Traditions (Leiden: Brill, 2022).
 8A
THE EMERGENCE OF THE QIRA T 301
ninth/fifteenth century appear to have accepted the proliferation of
qir:8:t based on the divine permission hypothesis; they do not appear
to have given much consideration to the difficulty of interpreting the
bare consonantal skeleton of the text (rasm) as a factor in the rise of
variant readings.7 This study is framed within what the qir:8:t trad-
ition said about the emergence of the variants. Thus one could simul-
taneously maintain that the tradition upheld what I have called ‘the
divine permission hypothesis’, while seeking and coming to a different

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

account of why, and when, the qir:8:t emerged.8 Alternatively, one
could make a more moderate claim that the hypothesis could account
for some of the qir:8:t. The point is that this paper does not necessarily
stand in opposition to the growing body of work on Qur8:nic manu-
scripts and linguistics in relation to the emergence of qir:8:t.
Thirdly, and to clarify the direction and scope of this paper, it is
worthwhile stating what this paper will not be discussing. This paper
is not about taw:tur (transmission of a report ‘by many through many’,
i.e., by multiple narrators through multiple lines of transmission).

7
Nevertheless, there are a few reports from which one might infer that some
qir:8:t were thought to have resulted from scribal errors. For instance, al-Fabar;
(d. 310/923) records a report claiming that, in Q. 13:31, Ibn 6Abb:s read yatabayyan
instead of yay8as, and that he (Ibn 6Abb:s) rejected the latter reading by saying that the
scribe wrote it while drowsing (kataba l-k:tibu l-ukhr: wa-huwa n:6is). See Ibn Jar;r
al-Fabar;, Tafs;r al-Fabar;: J:mi6 al-bay:n 6an ta8w;l :y al-Qur8:n (eds. 6Abd All:h b.
6Abd al-MuAsin al-Turk; et al.; Riyadh: D:r 62lam al-Kutub li-l-Fib:6a wa-l-Nashr
wa-l-Tawz;6, 26 vols., 2003; cited hereafter as J:mi6al-bay:n), xiii, 537. This report is
rejected by some scholars as ‘weak’; for example, MaAm<d al-2l<s;, R<A al-ma6:n; f;
tafs;r al-Qur8:n al-6aC;m wa-l-sab6 al-math:n; (Beirut: D:r IAy:8 al-Tur:th al-6Arab;,
30 vols., n.d.), xiii, 156.
8
For example, one can take the view, based on manuscript evidence, that some
variants resulted from imperfect dictation. However, this is not a view that can be
confidently attributed to the overwhelming majority of classical scholarship on
qir:8:t. Similarly, the literature advocating a post-canonization emergence of the
qir:8:t (i.e., arguing that the dialect(s) of the reading traditions differ(s) from the
original dialect of the Qur8:n) cannot be projected onto the qir:8:t tradition. For
these and related issues, see Benham Sadeghi and Uwe Bergmann, ‘The codex of a
Companion of the Prophet and the Qur8:n of the Prophet’, Arabica, 57/4 (2010):
343–436; Benham Sadeghi, ‘Criteria for emending the text of the Qur8:n’ in
Michael Cook, Najam Haider, Intisar Rabb, and Asma Sayeed (eds.), Law and
Tradition in Classical Islamic Thought: Studies in Honor of Professor Hossein
Modarressi (New York: Palgrave Macmillan, 2013): 21–41; Marijn van Putten
and Phillip W. Stokes, ‘Case in the Qur8:nic consonantal text’, Wiener
Zeitschrift für die Kunde des Morgenlandes, 108 (2018): 143–79; Pierre Larcher,
‘Une «rime cachée» dans Cor 23, 12–14? Histoire du texte et histoire de la langue’,
Arabica, 68/1 (2021): 36–50.
302 t a r e q mo qb e l
Although taw:tur is very much intertwined with qir:8:t, and discussions
on the latter normally involve the former, an exposition of taw:tur is not
necessary for our purposes: the emergence of qir:8:t is one thing, the
transmission mechanism another. Taw:tur could equally transmit divine
or non-divine texts and is, to some extent, not directly relevant here.
Furthermore, whether and what change and development the divine/
non-divine ur-qir:8:t underwent after their emergence is outside the
scope of this paper. The foremost concern here is the emergence, not a

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

diachronic analysis. This paper also does not look into the question of
ikhtiy:r, selecting and amalgamating variant readings from an existing
pool.9 An acquaintance with this topic is necessary for understanding the
qir:8:t issue as a whole10 but has no direct bearing on the emergence of
the variant readings. Only once the qir:8:t are in place does ikhtiy:r, in
its technical sense, begin. For the same reason, the paper does not enter
into the 6Uthm:nic codex debate in any detail; its scope relates to the pre-
standardization phase.
I do not pretend to be advancing a novel view in this paper. Its con-
tribution to the academic scholarship on qir:8:t is to restore and shed
new light on an explanatory model that has been present for a long time
in the Islamic scholarly tradition. The paper is set out as follows: after a
brief survey of views on the origins of the qir:8:t, I begin with the divine
permission, al-idhn al-Il:h;, which I take to be central to understanding
the emergence of qir:8:t and which enables us to solve a number of
outstanding problems about them; it is the kernel of the paper. Next, I
argue that the divine permission was not absolute; it was restricted with-
in semantic parameters. Following this, I question the separation of
Qur8:n and qir:8:t. I argue that theorizing the Qur8:n and qir:8:t as
practically one and the same thing allows us to understand how some
Muslim scholars were able to accept, even if implicitly, that here and
there the Qur8:n might include what, in however insignificant a way,
was not expressed exactly as sent down from God. The general findings
of the study are presented in the conclusion. To reiterate, my aim here is

9
I am adopting, to some extent, Bergsträsser’s interpretation of the old defin-
ition of ikhtiy:r: ‘The older technical meaning of the word refers to a reader who is
mainly following an older authority, but departs from it in some isolated instances
and follows his own way’. Nöldeke et al., The History of the Qur8:n, 486.
10
On the concept of ikhtiy:r, see al-Imam, Variant Readings of the Qur8an,
92–111; Yasin Dutton, ‘The form of the Qur’an: historical contours’ in Shah and
Abdel Haleem (eds.), The Oxford Handbook of Qur’anic Studies: 182–93, at
188–92. For a detailed exposition, see Am;n Idr;s Fall:ta, al-Ikhtiy:r 6inda
al-qurr:8 mafh<muh wa-mar:Ailuh wa-atharuh f; al-qir:8:t (Riyadh: Kurs;
al-Qur8:n al-Kar;m wa-6Ul<mih, J:mi6at al-Malik Sa6<d, 2015).
 8A
THE EMERGENCE OF THE QIRA T 303
to explore to what extent it was accepted and thinkable, especially by the
Companions and subsequently by the scholars of qir:8:t, that some
qir:8:t did not have explicit divine authority.

THE EMERGENCE OF THE QIR282T

The generally accepted Sunni view is that the qir:8:t constitute holy writ;

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

their source is God. The locus classicus of this doctrine, which states that
multiple versions of the Qur8:n were introduced by the Prophet, is the
‘seven aAruf’ Aad;th.11 To this effect, Ibn Qutayba (d. 276/889) writes:
‘And all these modes of reading are the speech of God’ (wa-kullu h:dhihi
l-Aur<fi kal:mu Ll:h),12 and al-D:n; (d. 444/1053) follows suit: ‘[the
qir:8:t] were all sent down from God’ (k:nat kulluh: min 6indi Ll:h

11
A standard explanation of this tradition and related issues is provided by Ab<
al-Khayr Ibn al-Jazar;, al-Nashr f; al-qir:8:t al-6ashr (ed. 6Al; MuAammad al-
Dabb:6; Beirut: D:r al-Kutub al-6Ilmiyya, 2 vols., n.d.), i. 23–33. Contemporary
scholarship on this Aad;th includes 6Abd al-6Az;z 6Abd al-Fatt:A al-Q:ri8, Ead;th al-
aAruf al-sab6a dir:sa li-isn:dih wa-matnih wa-khtil:f al-6ulam:8 f; ma6n:h wa-
Bilatih bi-l-qir:8:t al-Qur8:niyya (Beirut: Mu8assasat al-Ris:la, 2002). A summary
of the versions of the Aad;th, and a useful conspectus of Ibn al-Jazar;’s explanation is
provided in Yasin Dutton, ‘Orality, literacy and the ‘‘seven aAruf’’ Aad;th’, Journal
of Islamic Studies, 23/1 (2012): 1–49, at 18–30.
12
6Abd All:h Ibn Qutayba, Ta8w;l mushkil al-Qur8:n (ed. al-Sayyid AAmad
4aqr; Cairo: Maktabat D:r al-Tur:th, 2006), 94. It is by no means a straightfor-
ward task to decide how some of the Arabic terms relating to qir:8:t should be
rendered in English. The same Arabic word can bear more than one meaning, and
indeed a writer may use a word with two different meanings in the course of the
same discussion. That said, I try to offer the closest rendering in English for the
Arabic usage in its context, and to translate the same usage consistently whenever
possible. I will give two cases. The Arabic lis:n can be translated, depending on
context, as ‘tongue’, ‘language’, ‘lexicon’, or ‘dialect’. In this paper, I have trans-
lated (bi-lis:n quraysh) as ‘in the dialect of Quraysh’ although ‘in the lexicon of
Quraysh’ is also plausible. A further complication is that when the qir:8:t manuals
use, for example, bi-lis:n quraysh or lis:n Ras<l Allah, they could be intending both
lexical variations and variants in pronunciation/vocalization—so a single English
term (dialect/lexicon) might not capture the nuance of the discourse in Arabic.
Another example concerns the term Aarf (pl., Aur<f/aAruf) which can mean ‘phon-
eme’, ‘letter’, ‘word-form’, ‘style of reading’, among others. When the context is
that of the seven aAruf, I render Aur<f as ‘modes of reading’ in full awareness that a
choice has to be made and in a given instance alternative renderings are possible. In
sum, my translations of these terms should be understood as approximations.
304 t a r e q mo qb e l
munazzalatan).13 Similar statements are not uncommon in contempor-
ary qir:8:t manuals and introductory works.14
A markedly different picture is drawn in a number of Western aca-
demic studies on the Qur8:n. According to Nöldeke, the qir:8:t are at-
tributable to a number of factors: the process of oral transmission;
mistakes of copyists; differences in interpreting the consonantal forms;15
as well as free judgment and discretion.16 Influenced by Nöldeke,
Goldziher submitted that the qir:8:t resulted from attempts to make

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

13
Ab< 6Amr al-D:n;, J:mi6 al-bay:n f; al-qir:8:t al-sab6 al-mashh<ra (ed.
MuAammad 4ad<q al-Jaz:8ir;; Beirut: D:r al-Kutub al-6Ilmiyya, 2005), 29. It is
also well established in the secondary literature that this is the mainstream position
in (Sunni) Islam. For instance, we find in Nöldeke et al., The History of the Qur8:n,
487: ‘In the science of Koranic variant readings different possibilities exist a priori,
which, according to the prevalent dogma, are all equally divine’. Similarly, Nasser
(The Transmission of the Variant Readings, 10): ‘the mainstream view of Muslims
is that the variant readings of the Qur8:n are of a divine nature’, and Laher
(‘Between dogmatism and speculation’, 178): ‘Many Sunn;s (and Zayd;s) . . . assert
. . . that the variations between the readings are of divine origin’. A reiteration of the
traditional Muslim position is also presented by MuAammad MuB3af: al-A6Cam;,
The History of the Qur8:nic Text from Revelation to Compilation: A Comparative
Study with the Old and New Testaments (Leicester: UK Islamic Academy, 2003),
152–5.
14
For example, both 6Al; MuAammad al-Dabb:6(al-I@:8a f; bayan uB<l al-qir:8a
[Cairo: 6Abd al-Eam;d AAmad Eanaf;, 1938], 4) and 6Abd al-Fatt:A al-Q:@; (al-
Bud<r al-z:hira f; al-qir:8:t al-6ashr al-mutaw:tira min 3ar;qay al-sh:3ibiyya wa-l-
durra [Beirut: D:r al-Kit:b al-6Arab;, 1981], 7) state, without considering other
possibilities, that the qir:8:t can be traced back to the Prophet. And also note the
word ‘al-mutaw:tira’ in al-Q:@;’s title. In the same manner, AAmad Kh:lid Shukr;,
MuAammad AAmad MufliA al-Qu@:h and MuAammad Kh:lid ManB<r
(Muqaddim:t f; 6ilm al-qir:8:t [Amman: D:r 6Amm:r, 2001], 23) assert that all
the qir:8:t were sent down.
15
Nöldeke et al., The History of the Qur8:n, 474. I include oral transmission as a
source of qir:8:t according to Nöldeke on the basis of his statement: ‘Naturally, it is
conceivable that multiple versions of the text appeared when orally transmitted’.
That oral transmission was a source for many of the qir:8:t is also held by
Christopher Melchert, ‘Ibn Muj:hid and the establishment of seven Qur’anic read-
ings’, Studia Islamica, 91 (2000): 5–22, at 15.
16
Nöldeke et al., The History of the Qur8:n, 475. This ‘discretion’ is, however,
sometimes restricted by certain parameters such as agreement with the consonantal
text (476). It is not clear to me whether Nöldeke accepts—in addition to the above-
mentioned factors—the possibility that some variant traditions could be traced
back to the Prophet. However, Laher (‘Between dogmatism and speculation’,
179) writes that Nöldeke ‘appears to acknowledge that some variant readings
are ‘‘genuine’’ in the sense that they are transmissions of oral recitations from early
generations’.
 8A
THE EMERGENCE OF THE QIRA T 305
out the deficient skeletal text of the Qur8:n,17 and also from hermeneut-
ical and exegetical considerations.18 On the same lines, Jeffery remarked
the role of the 6Uthm:nic codex in the emergence of the qir:8:t.19 Other
suggestions include Burton’s that many forms of the qir:8:t were later
inventions fabricated by the schools of fiqh to provide proof-texts for
their legal doctrines.20 Common to many of these views is that the
qir:8:t belong to the post-standardization period.
Between these two broad approaches, there are also a number of

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

dissenting voices in the classical Muslim tradition. In his important
monograph, The Transmission of the Variant Readings, Nasser rightly
noted that al-Fabar;21 and al-Zamakhshar; (d. 538/1144)22 rejected
some canonical readings and preferred some of them over the others.
On this basis he concluded:
For both scholars, it seems that the canonical Readings were being stripped of
their divine nature, and their origin was not attributed to the Prophet himself,
but to the Qur8:n readers and transmitters, i.e. to their own selectivity in
reading and understanding (ijtih:d) of the 6Uthm:nic consonantal outline
supported by the context of the Qur8:nic verse.23

Nasser further argues that Ibn Muj:hid (d. 324/936) did not hold all the
readings to have divine authority: for Ibn Muj:hid, ‘the status of the variant
readings of the Qur8:n was similar to the status of the legal rulings
(aAk:m)’, and ‘he did not consider the seven Readings to be of divine
and absolute value’.24 The case Nasser makes here ultimately depends on
his inference that Ibn Muj:hid treated the differing qir:8:t as akin to

17
Ignác Goldziher, Die Richtungen der islamischen Koranauslegung (Leiden:
Brill, 1920), 3–4.
18
Ibid, 5.
19
Arthur Jeffery, The Qur8:n as Scripture (New York: R. F. Moore Co., Inc.,
1952), 97.
20
Burton, The Collection of the Qur8:n, 43, 165–86.
21
al-Fabar;’s approach to qir:8:t has received considerable attention, especially
as he is the author of J:mi6 al-bay:n, a major work of Sunni tafs;r. For a detailed
study, see Zayd b. 6Al; Mahd; Muh:rish, Manhaj al-im:m al-Fabar; f; al-qir:8:t wa-
@aw:bi3 ikhtiy:rih: f; tafs;rih (Riyadh: D:r al-Tadmuriyya, 2012).
22
On al-Zamakhshar;’s treatment of qir:8:t in his exegesis, see MuAammad
MaAm<d al-D<m;, ‘al-Qir:8:t al-mutaw:tira f; tafs;r al-Zamakhshar;; dir:sa naq-
diyya’ (PhD diss., Yarmouk University, 2004); Andrew J. Lane, A Traditional
Mu6tazilite Qur8:n Commentary: the Kashsh:f of J:r All:h al-Zamakhshar; (d. 538/
1144) (Leiden: Brill, 2006), 121–7; Kifayat Ullah, al-Kashsh:f: al-Zamakhshar;’s
Mu6tazilite Exegesis of the Qur8:n (Berlin: De Gruyter, 2017), 90–3.
23
Nasser, The Transmission of the Variant Readings, 7.
24
Ibid, 230.
306 t a r e q mo qb e l
disagreements in the fiqh; domain of aAk:m. As for his treatment of al-
Fabar;, some comments are in order. Nasser’s conclusion is that al-Fabar;,
among other scholars, ‘attributed the Qur8:nic variants to human origins;
either to the reader’s ijtih:d in interpreting the consonantal outline of the
Qur8:n or simply to an error in transmission’.25 This is not an accurate
account of al-Fabar;’s position if we base the account on the evidence to
hand.
First of all, al-Fabar; accepts, in principle, that the Qur8:n was

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

revealed in more than one mode of reading.26 This entails that he con-
sidered at least some of the variants to be divine. Secondly, I was unable
to find in the introduction to al-Fabar;’s J:mi6 al-bay:n any statement to
the effect that his own opinion was that the variants resulted from the
reciters’ interpretation of the 6Uthm:nic consonantal outline or from an
error in transmission. If I missed something, I stand to be corrected. The
reasoning Nasser attributes to al-Fabar; may be due to projecting onto
him some contemporary academic views.27 It would have been more

25
Ibid, 77.
26
al-Fabar;, J:mi6 al-bay:n, i. 41–59, clearly states that the Qur8:n was sent
down in different modes of reading (aAruf), which included phrasal variations
(innam: huwa khtil:fu alf:Cin).
27
Nasser, The Transmission of the Variant Readings, 40 n. 17, adds that Ibn
6A3iyya (d. 546/1151) states that the seven readings originated due to the deficien-
cies of the 6Uthm:nic muBAaf. I disagree with this reading of Ibn 6A3iyya, for al-
though he does say that the muBAaf was not vocalized (ghayru mashk<lin), nowhere
does he unequivocally say that this deficiency gave rise to the seven readings.
Perhaps the source of Nasser’s reading is Ibn 6A3iyya’s confusing statement that
‘the readers in the metropolises sought the differences that were transmitted to
them, especially those that adhered to the text of the muBAaf, and read that accord-
ing to their interpretation (thumma inna l-qurr:8 f; al-amB:r tatabba6< m: ruwiya
lahum mini khtil:f:tin l:siyyam: f;m: w:faqa kha33a l-muBAaf fa-qara8< bi-
dh:lika Aasaba jtih:d:tihim)’. It seems that Nasser took the antecedent of bi-
dh:lika to be the muBAaf; I maintain, however, that it refers back to ‘what was
transmitted to them’ (m: ruwiya lahum). In this sense, the reciters were practising
their ijtih:d in selecting from that which was transmitted to them; and this is pre-
sumably a reference to the notion of ikhtiy:r. What supports my reading is that Ibn
6A3iyya wrote in that same sentence that the readers sought what was transmitted to
them (m: ruwiya lahum), and hence it was primarily a matter of transmission.
Secondly, Ibn 6A3iyya states very clearly, on the preceding page, that the qir:8:t
were spread from the Prophet (intasharat 6an Ras<li Ll:h). See 6Abd al-Eaqq Ibn
6A3iyya, al-MuAarrar al-waj;z f; tafs;r al-Kit:b al-6az;z (ed. 6Abd al-Sal:m 6Abd al-
Sh:f; MuAammad; Beirut: D:r al-Kutub al-6Ilmiyya, 6 vols., 2001), i. 47–8. This
statement is inconsistent with Nasser’s reading which requires the qir:8:t to be a
post-MuAammadan phenomenon, that is, a consequence of (the difficulties with)
the later 6Uthm:nic codex.
 8A
THE EMERGENCE OF THE QIRA T 307
accurate to say that al-Fabar; believed that some of the variant readings
were not of divine origin, rather than to generalize. How, then, did these
non-divine, humanly originated variants emerge according to al-Fabar;?
Part of the task of this paper is to suggest a possible answer without
claiming that this is an explanation particular to the stance of al-Fabar;.
Rather, this paper offers a more general framework within which the
positions of various early Muslim scholars may be understood.
Before moving on, it is pertinent to draw attention here to a useful

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

point made by Yasin Dutton, one that seems to be in line with the gen-
eral trajectory of my argument. In an effort to hold the differences in the
variants together ‘with the theological position that there is only one
Qur8:n which has never changed and will never change’,28 Dutton
makes a case for looking at the qir:8:t through the lens of orality theory.
On the premise that the Qur8:n is an oral phenomenon, he argues that ‘a
limited amount of ‘‘variation’’ was not only accepted but also expected,
if it was even noticed; it was, in a sense, built in to the very text’. It
follows that ‘the variant readings are predominantly an expression of the
Qur8:n’s oral, and thus ‘‘multiform’’, nature’,29 indeed the Qur8:n ‘was a
typically ‘‘multiform’’ phenomenon at the beginning of its life’.30 The
upshot is that, in the reception of the Qur8:n, the Companions were
given some sort of discretion. Dutton makes this more plain where he
writes:
It also seems to me possible that, in the intervening time between the death of
the Prophet and 6Uthm:n’s decision some twenty or so years later, some
Companions were reciting parts of the revelation using the same sort of vari-
ation that they had heard and witnessed the Prophet using and allowing others
to use, and which they thus assumed they had ongoing Prophetic authority
for.31
Combining Dutton’s argument with what we established above,
namely that eminent scholars in the Muslim tradition—al-Fabar;,
al-Zamakhshar;, and (according to Nasser) Ibn Muj:hid, to name
but three—believed that some of the variants did not hold absolute
divine authority, the remainder of the paper suggests an alternative
paradigm through which we can understand the foundation under-
lying these dissenting yet authoritative voices.

28
Dutton, ‘Orality’, 30.
29
Ibid, 34.
30
Ibid, 42. The relevance of orality is also stressed in Bauer’s discussion of
qir:8:t: Die Kultur der Ambiguität, 63, 65–6, 75–7.
31
Dutton, ‘Orality’, 43.
308 t a r e q mo qb e l

AN ALTERNATIVE PARADIGM: THE DIVINE

PERMISSION (AL-IDHN AL-IL2HĪ)

Goldziher suggested that there was ‘freedom’ (Freiheit) in early Islam

with regard to the composition of the Qur8:n,32 which I partially agree
with but for reasons different from his. This ‘freedom’, I suggest, was a
response to what I shall call a ‘divine concession’, or ‘divine permission’
(al-idhn al-Il:h;). When the Qur8:n was sent down, its first recipients

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

were permitted some discretion as to its recitation with some variation.
They were informed, through the Prophet, that they were allowed to
introduce some (minor) variations in reciting the words of the Qur8:n,
perhaps to facilitate the experience of reciting it aloud and thereby en-
able a wider circulation of its message among people of varied dialectal
background.33 When verses of the Qur8:n were sent down and then
taught by the Prophet, the believers were informed, whether explicitly
there and then or through the general directives of the seven aAruf
Aad;th, that it was permissible for them to read the verses (slightly) dif-
ferently. The consequence is that the qir:8:t were primarily taught by the
Prophet and partly initiated by its first recipients. Therefore, when it
comes to the emergence of the qir:8:t we have, according to the classical
Muslim tradition, as we shall see, two sources: the Prophet first and
foremost, and, secondarily, his Companions on the basis of his indirect
authority, that is, his permission for them to recite differently from him-
self and from each other within a parameter to be discussed below. The
Muslim scholarly tradition refers to many traditions to the effect that the
qir:8:t were conveyed by the Prophet.34 As for the divine permission
hypothesis, this can be reconstructed from some classical Muslim texts.
Doing so will enable us to better appreciate the occasionally critical

32
Goldziher, Die Richtungen, 33.
33
The rationales given for the existence of the qir:8:t commonly include tays;r,
ease––the removal of hardship––of recitation and of further transmission (tays;ru
naqlih) of the Qur8:n. See Ibn al-Jazar;, al-Nashr, i. 52–3.
34
The classic reference for this traditional belief is the incident of 6Umar b. al-
Kha33:b and Hish:m b. Eak;m where they disagreed about how to recite S<rat al-
Furq:n (Q. 25). What is notable, for the topic here, is that each attributed his
recitation to the Prophet: Hish:m said that the Prophet taught him that recitation
(aqra8an;h: Ras<lu Ll:h), and likewise 6Umar (Ras<lu Ll:h qad aqra8an;h: 6al:
ghayri m: qara8ta). This is a clear indication that the Prophet, it was believed, was
the source for (at least, some of) the qir:8:t. For this version of the Aad;th, see al-
Bukh:r;, 4aA;A al-Bukh:r; (Liechtenstein: Thesaurus Islamicus Foundation, 3
vols., 2000), iii. 1049 [Aad;th 5043]. (Printed within the multi-volume set Jam6
jaw:mi6 al-aA:d;th wa-l-as:n;d wa-maknaz al-BiA:A wa-l-sunan wa-l-mas:n;d.)
 8A
THE EMERGENCE OF THE QIRA T 309
attitudes of Muslim scholars towards the qir:8:t. It will, moreover, pro-
vide supporting evidence for Dutton who hints at this notion of divine
permission when he says, referring to the reconstruction of the ur-
Qur8:n, that it is not possible to ‘reconstruct all of what was at one
time or another recited as part of the seven aAruf of the Qur8:n (or
perhaps we should say, that was allowed to be recited as part of the
seven-aAruf nature of the Qur8:n)’,35 and when he avers that the options
not included in the 6Uthm:nic codex also had either direct Prophetic

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

authority, or, and this is crucial, ‘the indirect authority of the Prophet
behind them, in the sense that he had allowed others to recite, or read,
them that way’.36
What are the possible scenarios for the divine permission? I envisage
two main possibilities. The first is that the Companion(s) would recite to
the Prophet, and the Prophet would authorize that recitation. Its author-
ity derives from the fact that it received divine assent via the Prophet.37
Even so, this reading originated, narrowly speaking, from a performance
of the Companion(s), and not from God. Another possible scenario is
that the Companion(s), on the basis of the sufficiency of the general
divine permission mediated by the Prophet, would recite with some
variations without necessarily going back to the Prophet.
35
Dutton, ‘Orality’, 35.
36
Ibid, 42–3. The possibility of the Prophet endorsing some of the qir:8:t is also
noted by Sadeghi and Bergmann (‘The codex of a Companion of the Prophet and
the Qur8:n of the Prophet’, 413): ‘If the scribes recited the text back to the
Prophet—and we do not know whether this happened—one wonders if the
Prophet tacitly endorsed some of these differences, relatively small as they generally
seem to be. If so, that would not negate the fact that one version better represented
what the Prophet himself actually recited; but which one?’
37
The reasoning is that the Companion recitation, which has no normative force
per se, is raised through the Prophet’s approval into some form of revelation.
Underpinning this proposition are two connected issues. The first is the notion
that the Prophet’s authority to pronounce the Qur8:n in additional (unrevealed)
ways amounts, in the final analysis, to divine authority. This is in line with the
mainstream Sunni view of the twofold nature of revelation, that is, the authority of
the Sunna with that of the Qur8:n. See, on this question, Aisha Y. Musa, Ead;th as
Scripture: Discussions on the Authority of Prophetic Traditions in Islam (New
York: Palgrave Macmillan, 2008). Secondly, this scenario where the Companion
would read, and the Prophet would approve (or, at least, would not disown) the
reading could be understood within the framework of iqr:r, the Prophet’s acqui-
escence in what was said or done in his absence or his presence. Iqr:r is usually
discussed in uB<l al-fiqh literature: see Ab< al-Ma6:l; al-Juw:yn;, al-Burh:n f; uB<l
al-fiqh (ed. 6Abd al-6AC;m al-D;b; Qatar: n.p., 2 vols., 1399 [1979]), i. 498–502;
Mann:6 al-Qa33:n, T:r;kh al-tashr;6 al-Isl:m;: al-tashr;6 wa-l-fiqh (Riyadh:
Maktabat al-Ma6:rif li-l-Nashr wa-l-Tawz;6, 2nd edn., 1996), 124–31.
310 t a r e q mo qb e l
In sum, there are different shades for the divine authority of the words
of the Qur8:n for its first recipients. They may be arranged within a
continuum of divine authority, with those qir:8:t taught by the
Prophet, recited by him, appearing at the upper end of the continuum,
and those originated by the Companion(s) on the basis of the general
divine permission placed at the lower end. Between these two lie the
recitations originated by the Companion(s) and explicitly approved by
the Prophet.38

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

I turn now to the principal concern of the paper, namely to explore
whether classical Muslim scholars believed in the divine permission hy-
pothesis. My argument is in two parts; first, scholarly discussions that
imply the divine permission, and, second, discussion in which scholars
all but explicitly state the divine permission hypothesis. Thus, I will be
proceeding from weaker inferences to stronger evidence.

THE DIVINE PERMISSION HYPOTHESIS

IMPLIED IN THE ISLAMIC SCHOLARLY
TRADITION

Consistent with the mainstream tradition, al-Fabar; held that the seven
aAruf Aad;th meant that the Qur8:n came down in different lugh:t (dia-
lects).39 He contrasts this position with the view of an imagined learned
interlocutor that the Aad;th indicated differences in meaning (annah<
nazala bi-sab6ati ma6:nin),40 variations in content as opposed to

38
One could also envisage a scenario where a Companion would come to the
Prophet and suggest a variation with regard to a particular expression, and the
Prophet would allow him to apply this variation to other similar expressions in the
Qur8:n, though he would not have heard the latter from the Prophet nor recited
them in his presence. For example, he may read a single word with an im:la
(fronting the /:/ vowel towards /;/) for dialectal ease, and take permission from
the Prophet to apply it, by analogy, to similar words. A similar process was sug-
gested by Ab< Bakr al-B:qill:n; (d. 403/1013), al-IntiB:r li-l-Qur8:n (ed.
MuAammad 6IB:m al-Qu@:h; Amman: D:r al-FatA li-l-Nashr wa-l-Tawz;6, 2
vols., 2001), i. 346, in the context of Gabriel’s Qur8:nic instruction to the
Prophet. For example, he suggests that it might be possible that Gabriel taught
the Prophet to read 6alayhim (without terminal vowel) and 6alayhim< (with a long
vowel, <) but did not go over with him all the instances of 6alayhim (and its qir:8:t-
cognates, ilayhim and ladayhim) in the two forms. Rather, the Prophet was
instructed to apply this rule wherever the form appeared.
39
al-Fabar;, J:mi6 al-bay:n, i. 42.
40
Ibid, 49.
 8A
THE EMERGENCE OF THE QIRA T 311
variations in word-forms. Moreover, he objects to another opinion,
spelled out by the same interlocutor, namely that the seven aAruf are
dispersed through the whole Qur8:n (annah: sab6u lugh:tin mutafarra-
qatin f; jam;6i l-Qur8:n).41 What is common between these two views, to
which al-Fabar; objects, is that they reject the possibility of having dif-
ferent expressions conveying the same meaning. Whereas he holds that a
verse may be revealed in different versions, advocates of the two other
views do not allow this. For positing that the seven aAruf are semantic

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

variations rules out their being merely linguistic variations and, likewise,
saying that the seven aAruf are dispersed throughout the Qur8:n does not
allow for one word to be expressed in different ways—the dispersed
multiplicity view excludes simultaneous multiplicity.42 The upshot of
these two opinions is that the Qur8:n was revealed, language-wise, in
one version—and al-Fabar; will not accept this.
The relevance of this discussion for the divine permission hypothesis
can be cued from an observation made by al-Fabar;’s interlocutor. This
imagined discussion partner says that the proponent of the dispersed
multiplicity view—that the seven aAruf are dispersed in the Qur8:n—
holds that the Qur8:n came down in one of two readings, for example,
either ‘BayAa’ or ‘zaqya’ (alladh; nazala bih; al-Qur8:n 6indah< iAd: al-
qir:8atayn). This is problematic, in the eyes of the interlocutor, since this
contradicts the evidence invoked for it: there is no coherence between the
opinion and its evidence.43 Detailed clarification of this view is not par-
ticularly an issue for us. What is of interest here are its implications,

41
Ibid, 51–2.
42
If one takes the statement of Ibn Qutayba (Ta8w;l mushkil al-Qur8:n, 91)
that the seven aAruf represent seven dispersed dialects (6al: sab6ati awjuhin mina
l-lugh:ti mutafarriqatin f; al-Qur8:n) at face value, it would appear that he was
endorsing the second view discussed by al-Fabar;’s imagined interlocutor.
However, the examples Ibn Qutayba gives of the seven modes of reading include
instances where one word admits more than one linguistic variation (93), and he
clearly declares that all the seven aAruf were sent down (95). Therefore, I lean
towards the view that the dispersed multiplicity described by al-Fabar;’s imagined
interlocutor is different from that of Ibn Qutayba who seems to allow for a simultan-
eous multiplicity. It is possible that al-Fabar;’s interlocutor is pointing to Ibn Sall:m
(d. 224/838) who clearly stated that the Aad;th does not mean that a single word could
be read in seven modes (wa-laysa ma6n: tilka l-sab6ati an yak<na l-Aarfu l-w:Aidu
yuqra8u 6al: sab6at awjuh), but that the Qur8:n was revealed in seven Arabic dialects
dispersed in the Qur8:n (nazala 6al: sab6i lugh:tin mutafarriqatin f; jam;6i l-Qur8:n). See
Ab< 6Ubayd al-Q:sim Ibn Sall:m, Fa@:8il al-Qur8:n wa-ma6:limuh wa-:d:buh (ed.
AAmad 6Abd al-W:Aid al-Khayy:3;; al-MuAammadiyya: al-Mamlaka al-Maghribiyya:
Wiz:rat al-Awq:f wa-l-Shu8<n al-Isl:miyya, 2 vols., 1995), ii. 168–9.
43
al-Fabar;, J:mi6 al-bay:n, i. 51–2.
312 t a r e q mo qb e l
namely that the Qur8:n was sent down in one mode. This leads to the
question: How did the advocates of this view—of whom Ibn Sall:m
might have been one—account for the other qir:8a (or qir:8:t)? If only
as a tentative inference, I suggest the divine permission hypothesis as a
possible explanation.
Another promising way to derive that hypothesis is to consider the
concept of ‘first reader’, the first person to read a particular variant, or
the first to introduce a qir:8a into the pool of qir:8:t. To explore this

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

concept, which follows from the divine permission hypothesis, let us
turn to Ibn Ab; D:w<d al-Sijist:n;’s (d. 316/929) Kit:b al-MaB:Aif. In
looking at the variant readings in Q. 1:4, ‘m:liki’ and ‘maliki’, he lists
reports to the effect that the Prophet read both.44 However, and this is
relevant to the hypothesis, he also records a tradition, on the authority of
al-Zuhr; (d. 124/742), saying that whereas the Prophet, Ab< Bakr, 6Umar
and 6Uthm:n read ‘m:liki’, Marw:n (d. 65/685) was the first to read
‘maliki’ (wa-awwalu man qara8ah: maliki Marw:n).45 The historical ver-
acity of this piece of information—this includes the reliability of the trad-
ition and whether or not al-Zuhr; was right—is not the question here.
Indeed, there are reports to the contrary—that the Prophet read ‘maliki’
too—which Ibn Ab; D:w<d also records. The pertinent issue is that Ibn
Ab; D:w<d records the concept of ‘first reader’ in the first place. The fact
that he included it in his book implies that the concept was not unthink-
able for him—that a reciter may originate a new qir:8a. We find this same
tradition in other books of qir:8:t46 as well as later tafs;rs.47 Yes, they
might dispute the reliability of the riw:ya on the basis that other reports
equally attribute ‘maliki’ to the Prophet, and therefore deem the report
factually incorrect,48 but—and this is crucial—they do not seem to be
concerned with the conceptual question of ‘first reader’—that a qir:8a

44
Ibn Ab; D:w<d al-Sijist:n;, Kit:b al-MaB:Aif (ed. MuAibb al-D;n 6Abd al-SubA:n
W:6iC; Beirut: D:r al-Bash:8ir al-Isl:miyya, 2 vols., 2002), i. 388–96.
45
Ibid, 390.
46
For example, Ab< 6Al; al-F:ris;, al-Eujja f; 6ilal al-qir:8:t al-sab6(eds. 62dil AAmad
6Abd al-Mawj<d, 6Al; MuAammad Mu6awwa@ and AAmad 6Īs: Easan al-Ma6Bar:w;;
Beirut: D:r al-Kutub al-6Ilmiyya, 4 vols., 2007), i. 108.
47
See MuAammad al-F:hir Ibn 62sh<r, Tafs;r al-TaAr;r wa-l-tanw;r (Tunis: al-D:r
al-T<nisiyya li-l-Nashr, 30 vols., 1984), i. 175.
48
Thus, they might try to reconcile al-Zuhr;’s statement by qualifying it to mean
Marw:n was the first to recite ‘maliki’ in his time or place (f; dh:lika l-6aBri awi l-balad):
see Ibn 6A3iyya, al-MuAarrar al-waj;z, i. 69. This does not seem to be challenging the
‘first reader’ concept itself; it is basically an attempt to harmonize al-Zuhr; with other
reports. What is being problematized is, essentially, the factual value of al-Zuhr;’s
report, not the concept underlying the report.
 8A
THE EMERGENCE OF THE QIRA T 313
may be ‘introduced’ by other than the Prophet.49 How can we explain this
attitude? Again, I suggest that they were, possibly but not necessarily,
operating within the divine permission paradigm and accordingly took
the concept of ‘first reader’ in al-Zuhr;’s report for granted.
A third argument for the divine permission hypothesis could be made
from the discussion on the Qur8:nic dialect: was the Qur8:n revealed in
the dialect of Quraysh, in a different dialect, or in a combination of
dialects? It is not my purpose here to answer these questions—they

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

form part of a different enquiry.50 Rather, and more simply, my purpose
is briefly to review the classical debate on Qur8:nic dialects for what it
can tell us about the divine permission hypothesis. To this end, I analyse
two traditions and underline their bearing on our understanding of the
emergence of the qir:8:t.
The first tradition is attributed to 6Umar b. al-Kha33:b. It is related in
Ibn Jinn;’s (d. 392/1002) al-MuAtasab and al-Zamakhshar;’s al-Kashsh:f
that 6Umar heard a person reading ‘6att: A;nin’—in place of ‘Aatt:
A;nin’—and inquired: ‘Who taught you [this recitation]?’ The person
answered that it was Ibn Mas6<d (d. 32/652). Thereupon, 6Umar wrote
to Ibn Mas6<d: ‘God sent down this Qur8:n and made it Arabic (fa-
ja6alah< 6arabiyyan), and He sent it down in the dialect of Quraysh
(wa-anzalah< bi-lughati quraysh), so teach people in the dialect of
Quraysh, not that of Hudhayl’.51 If one accepts both the reliability of
the tradition—that 6Umar actually said it—and the factual content—that
the Qur8:n was indeed sent down in the dialect of Quraysh, and that Ibn
Mas6<d was teaching in a different dialect—it supports the divine per-
mission hypothesis in the sense that Ibn Mas6<d, unlike 6Umar, believed
that although the Qur8:n was sent down in one dialect, it was

49
To clarify, I am not assuming that the inclusion of these reports in those
collections means that their respective authors agreed to the contents of the
reports––that they did not criticize the ‘first-reader’ reports is not evidence of ap-
proval. My point is that the concept was, at least, thinkable even if not approved by
them. (My thanks to an anonymous reader of an earlier draft of this paper for their
comment on this point.)
50
The Islamic tradition is itself not clear on this question. Various views on the
issue are advanced, some conflicting. For an overview, see Rafael Talmon, art.
‘Dialects’ in Jane Dammen McAuliffe (ed.), Encyclopaedia of the Qur8:n
(Leiden: Brill, 6 vols., 2001–2006), i. 529–31; Gh:nim Qadd<r; al-Eamad,
AbA:th f; al-6Arabiyya al-fuBA: (Amman: D:r 6Amm:r, 2005), 67–77.
51
Ab< al-FatA Ibn Jinn;, al-MuAtasab f; taby;n wuj<h shaw:dhdh al-qir:8:t wa-
l-;@:A 6anh: (eds. 6Al; al-Najd; N:Bif, 6Abd al-Eal;m al-Najj:r and 6Abd al-Fatt:A
Ism:6;l Shalab;; Istanbul: D:r Sazk;n li-l-Fib:6a wa-l-Nashr, 2 vols., 1986), i. 343;
J:r All:h al-Zamakhshar;, al-Kashsh:f 6an Aaq:8iq ghaw:mi@ al-tanz;l wa-6uy<n
al-aq:w;l f; wuj<h al-ta8w;l (Beirut: D:r al-Ma6rifa, 2009), 515.
314 t a r e q mo qb e l
permissible to read in a different dialect, one that was not as sent down
but selected for by the reciter, in this case Ibn Mas6<d. This, I think, is a
consequence of the divine permission being present in Ibn Mas6<d’s
mind.52 If one were to reject the report as unreliable or fabricated and
its matter inaccurate, the divine permission hypothesis will not necessar-
ily be ruled out. The existence of this report in the classical sources
shows that the idea was present in the Muslim scholarly memory and
not something unthinkable. Furthermore, we note that Ibn Jinn; and al-

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

Zamakhshar; include it in their works without objecting to it, which
might imply some degree of acceptance of the idea.
The same reasoning can be applied to a tradition attributed to
6Uthm:n. In the process of instructing the committee on how to compile
the Qur8:n, and in his attempt to limit the variants, he is reported to
have said: ‘If you disagree with Zayd b. Th:bit on a point related to the
Arabic of the Qur8:n (f; 6arabiyyatin min 6arabiyyati l-Qur8:n), then
write it in the dialect of Quraysh (bi-lis:ni Quraysh), for the Qur8:n
was sent down in their dialect (bi-lis:nihim)’.53 On the face of it,
6Uthm:n appears to accept that the pool of variants in circulation
includes those which do not belong to the dialect of Quraysh and, con-
sequently, includes variants that were not revealed—they were ‘gener-
ated’ by the reciters themselves.54 This perhaps points to the divine
permission hypothesis and the initial fluidity of the qir:8:t.

52
Moreover, it would not be wrong to say that 6Umar’s instruction does not
necessarily mean that he was objecting to the permissibility of reciting in a non-
revealed qir:8a. Rather, he was instructing Ibn Mas6<d to refrain from teaching in
the dialect of Hudhayl because of his (6Umar’s) preference (ikhtiy:r) for the dialect
of Quraysh, not because what Ibn Mas6<d was doing was wrong in principle. An
explanation could be drawn from Ibn 6Abd al-Barr, al-Tamh;d li-m: f; al-Muwa33a8
min al-ma6:n; wa-l-as:n;d (vol. 8 ed. MuAammad al-Fall:A; [Rabat]: Wiz:rat al-
Awq:f wa-l-Shu8<n al-Isl:miyya, 26 vols., vol. 8: 1980), 279. However, it appears
that Ibn 6Abd al-Barr’s own view was that Ibn Mas6<d was choosing from within
what was divinely revealed—thus, his position and mine are not the same.
Alternatively, one could say 6Umar believed that Ibn Mas6<d was going beyond
the limits of what was permissible, stretching the divine permission too far. Yet a
third possibility is to view 6Umar’s objection not as a point of doctrine on his part,
but as an administrative decision with an eye to stabilizing the text, akin to
6Uthm:n’s later codex. On these grounds, we can maintain that 6Umar did not, in
principle, object to the divine permission hypothesis.
53
al-Bukh:r;, 4aA;A al-Bukh:r;, iii. 1047 [Aad;th 5035].
54
I say ‘on the face of it’, given that 6Uthm:n’s statement could be interpreted in
other ways. For example, it could be understood to mean that most of the Qur8:n
was revealed in the dialect of Quraysh: see Jal:l al-D;n al-Suy<3;, al-Itq:n f; 6ul<m
 8A
THE EMERGENCE OF THE QIRA T 315

THE DIVINE PERMISSION HYPOTHESIS

BECOMING EXPLICIT IN THE ISLAMIC
SCHOLARLY TRADITION

The three arguments presented above in support of the divine permission

hypothesis are tentative inferences. In this section, I provide the stronger
evidence, in the form of almost-explicit statements, for my argument
that some scholars in the Muslim tradition believed the hypothesis to

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

be true.
I first consider Ab< Ja6far al-FaA:w; (d. 321/933). His treatise SharA
mushkil al-:th:r contains useful discussions and reports on the emer-
gence of the qir:8:t. On the divine permission, al-FaA:w; writes that,
because the Companions were required to recite the Qur8:n in ritual
prayer and to understand their religion through it, they were allowed
to read the Qur8:n according to its meanings (an yatl<hu bi-ma6:n;h)
even if their expressions differed from those recited to them by the
Prophet (wa-in kh:lafat alf:Cuhumu l-lat; yatl<nah< bi-h: alf:Ca
Nabiyyihim Balla Ll:hu 6alayhi wa-sallam al-lat; qara8ah< bi-h: 6alay-
him).55 Al-FaA:w; makes it clear that it was allowed on account of their
need to do that (li-@ar<ratihim il: dh:lik), even if what was sent down to
the Prophet was a single expression (wa-in k:na l-ladh; nazala 6al: l-
Nabiyyi Balla Ll:hu 6alayhi wa-sallam innam: nazala bi-alf:Cin
w:Aida).56 In another formulation, al-FaA:w; repeats the distinction
between what the Companions were reading under the authority of
the concession and the Prophet’s single mode of reading. Eventually,
al-FaA:w; argues, the concession was lifted and the Companions’
ways of reciting the Qur8:n reverted to the dialect of God’s Messenger
(Aatt: 6:dat lugh:tuhum il: lis:ni Ras<li Llah Balla Ll:hu 6alayhi wa-
sallam).57 Given al-FaA:w;’s clear distinction between what was sent
down to the Prophet and what the Companions were allowed to read,
it is safe to say that he implicitly recognized the divine permission
hypothesis.

al-Qur8:n (ed. Markaz al-Dir:s:t al-Qur8:niyya; Madina: Majma6 al-Malik Fahd

li-Fib:6at al-MuBAaf al-Shar;f, 7 vols., 1426 [2005]), iii. 931.
55
Ab< Ja6far al-FaA:w;, SharA mushkil al-:th:r (ed. Shu6ayb al-Arna8<3; Beirut:
Mu8assasat al-Ris:la, 16 vols., 1994), viii. 118.
56
Ibid, 124.
57
Ibid, 125. For a useful exposition of al-FaA:w;’s opinion, see 6Amm:r al-
Kha3;b, ‘TaAr;r madhhab al-Im:m al-FaA:w; f; ma6n: al-aAruf al-sab6a’, available
at https://tafsir.net/article/5351/thryr-mdhhb-al-imam-at-thawy-fy-m-na-al-ahrf-
as-sb-h (last accessed 14 November 2021). (My thanks to an anonymous reader
of an earlier draft of this paper for drawing my attention to al-FaA:w;’s opinion.)
316 t a r e q mo qb e l
After al-FaA:w;, al-B:qill:n;, the author of al-IntiB:r li-l-Qur8:n
declares a strong ‘conservative’ line on the emergence of the qir:8:t.
He writes, for instance, that all seven aAruf emerged from the Prophet
(Caharat wa-staf:@at 6ani l-Ras<l),58 and that the Prophet did not die
until he had taught all seven aAruf (fa-l: yakhruju 6alayhi l-sal:m mina l-
duny: Aatt: yuqri8a jam;6ah:).59 Nevertheless, in his response to a ques-
tion about the Prophetic tradition (reported on the authority of Ubayy)
allowing one to substitute, for example, ‘raA;m’ for ‘ghaf<r’, one of the

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

solutions al-B:qill:n; offers is to say that—if the report is authentic and
established in the first place (idh: BaAAat wa-thabatat)—it is possible
that in the beginning of Islam Muslims were allowed to substitute single
words (wa-yaj<zu an yak<na qad shurri6a f; Badri l-Isl:mi an yuj6ala
mak:na l-Aarfi l-w:Aidi khil:fuh) as long as the opposite of the intended
meaning was not obtained (fa-amm: an yuj6ala mak:nah< @idduh< . . .
fa-lam yakun dh:lika j:8izan bi-l-ijm:6). Thereafter, this was abrogated
(thumma nusikha min ba6di dh:lik).60 From this, it appears that al-
B:qill:n; held a limited version of the divine permission hypothesis.
For him it was a time-limited concession granted by divine authority
that ceased to exist by virtue of, apparently, another divine command.
A clearer statement of the divine permission hypothesis is found in
Ab< Sh:ma’s (d. 665/1267) widely quoted al-Murshid al-waj;z. He
devotes a chapter to the seven aAruf Aad;th in which the notion of per-
mission (ib:Aa) recurs with noticeable frequency. Let us consider three of
his statements. In one passage, Ab< Sh:ma quotes some scholars as
saying: ‘God revealed the Qur8:n in the dialect of Quraysh and those
eloquent Arabs in their vicinity, and then made it permissible for the
Arabs (thumma ab:Aa li-l-6arab)—who are addressed by it, and to whom
it was revealed—to read it in their dialects to which they were accus-
tomed (an yaqra8<hu bi-lugh:tihim)’.61 Ab< Sh:ma concurs with that
view (wa-h:dh: kal:mun mustaq;mun Aasan), and then adds that God
made it permissible (ab:Aa Ll:hu ta6:l:) to read in seven aAruf those
Qur8:nic phrases which accept being read in seven modes of reading,
and in less than seven those phrases which admit fewer modes of read-
ing.62 Similarly, he adds, a couple of lines later, that to read in a dialect

58
al-B:qill:n;, al-IntiB:r li-l-Qur8:n, i. 60.
59
Ibid, 336.
60
Ibid, 370.
61
Ab< Sh:ma al-Maqdis;, al-Murshid al-waj;z il: 6ul<m tata6allaqu bi-l-Kit:b
al-6az;z (ed. Fayy:r 2lt; Q<l:j; Beirut: D:r 4:dir, 1975), 95.
62
Ibid, 96. Further support for Ab< Sh:ma (and the scholars he was quoting
and approving) is found in al-Qas3all:n;, La3:8if al-ish:r:t li-fun<n al-qir:8:t
(ed. Markaz al-Dir:s:t al-Qur8:niyya; Madina: Majma6al-Malik Fahd li-Fib:6at
 8A
THE EMERGENCE OF THE QIRA T 317
other than that of Quraysh was permitted in order to make it convenient
for the Arabs (innam: ub;Aa an yuqra8a bi-ghayri lis:ni quraysh tawsi6a-
tan 6al: al-6arab).63 Third, Ab< Sh:ma comments on one version of the
seven aAruf Aad;th that it means that the recipients of the Qur8:n were
given the concession to use substitute-words (rukhkhiBa lahum f; ibd:li
alf:Cih; bi-m: yu8add; ma6n:h:) whereby their meanings are effectu-
ated.64 These statements, as well as some other views that he cites,65
seem to indicate that Ab< Sh:ma held a more accommodating concep-

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

tion of the divine permission than al-B:qill:n;. However, earlier on in the
chapter, he comments on the use of substitute-words, mentioned above,
by saying that the substitution of words was allowed before the
Companions compiled the muBAaf (wa-k:na h:dh: s:8ighan qabla
jam6i l-BaA:bati l-muBAaf), and that they understood that the dispensa-
tion was no longer required when those who had memorized the Qur8:n
proliferated.66 One may conclude from this that Ab< Sh:ma was of the
opinion that the divine permission was still operative after the Prophet’s
death, and remained so until curtailed by the muBAaf project, though he
does not specify which one. Thus, the permission eventually came to an
end. On the basis of these statements, it is safe to say that Ab< Sh:ma
endorsed the divine permission hypothesis in principle.67
But any discussion of qir:8:t remains incomplete without engaging Ibn
al-Jazar;. To him we now turn. In al-Nashr we find (at least) three
allusions to the divine permission hypothesis. Following his account of
the 6Uthm:nic compilation, and the distribution of the maB:Aif, he

al-MuBAaf al-Shar;f, 10 vols., 1434 [2013]), i. 70–1, where he observes, on the

basis of a Aad;th, that the seven aAruf concession was made after the Prophet
emigrated to Madina; in particular, it was communicated to the Prophet while he
was in Madina in a location known as A@:t Ban; Ghif:r.
63
Ab< Sh:ma, al-Murshid al-waj;z, 97.
64
Ibid, 126.
65
See, for example, ibid, 102.
66
Ibid, 89.
67
The hypothesis appears in another work of Ab< Sh:ma, his commentary on
the didactic poem, al-Sh:3ibiyya, an abridgement and versification of al-D:n;’s al-
Tays;r. In it Ab< Sh:ma notes that the authentic qir:8:t are divided into what the
Prophet read and what he allowed (annah< qara8ah< aw adhina f;h). See Ab<
Sh:ma al-Maqdis;, Ibr:z al-ma6:n; min Airz al-am:n; (Cairo: MuB3af: al-B:b; al-
Ealab;, 1349 [1931]), 4. Although one might argue that what the Prophet ‘allowed’
denotes cases in which a Companion reads and the Prophet approves, and hence
there is a divine approval via the Prophet, this does not, I think, put the Prophet’s
readings on a par with those initiated by his Companions and subsequently
approved by him. There is, as I have suggested above, a hierarchy of divine author-
ity for the different variant readings.
318 t a r e q mo qb e l
wrote: ‘and the inviolable nation agreed upon what was contained in
these codices, and on leaving out what was inconsistent with them—in
the form of addition, omission, and substituting one word for another—
from what was allowed’ (mimm: k:na ma8dh<nan f;h). He then adds
that the 6Uthm:nic codices were devoid of vocalization in order to allow
for what was soundly transmitted and what the Prophet certainly recited
(m: BaAAa naqluh< wa-thabata til:watuh< 6ani l-Nab;).68 Two observa-
tions can be made here. The first is that Ibn al-Jazar; is explicit that the

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

readings excluded from the 6Uthm:nic project belong to the permissible
category. Secondly, he seems to make a distinction between what the
Prophet recited and what was being read under the authority of the
divine permission. He does not use the concept of permissibility in a
pejorative manner. All he is saying is that 6Uthm:n decided to include
only those variants lying at the upper level of the divine authority con-
tinuum—those that the committee was certain the Prophet recited.
Moving on, Ibn al-Jazar; looks at the question of why there are seven
modes of recitation instead of some other number. One of the opinions
he includes is that the Aad;th is not meant to indicate an exact number.
Rather it indicates ease (al-mur:du l-sa6a wa-l-tays;r), and the permissi-
bility of reading the Qur8:n in any Arabic dialect on the basis—and this
is important for us—that God has given permission to do so (min Aaythu
inna Ll:h: ta6:la adhina lahum f; dh:lik).69 The third occurrence of the
divine permission hypothesis in al-Nashr appears in Ibn al-Jazar;’s dis-
cussion of whether or not all seven aAruf are contained in the muBAaf of
6Uthm:n. He gives two views. Whereas the first holds that the 6Uthm:nic
codices contain all of the aAruf, the second opinion, attributed to the
majority of scholars, maintains that the 6Uthm:nic codices contain only
the aAruf which the consonantal skeleton can support (mushtamilatun
6al: m: yaAtamiluhu rasmuh:).70 Ibn al-Jazar; then proceeds to highlight
some of the responses given by the latter group to the objections made by
those who hold the first opinion. Here, mention of the divine permission
notion appears two times. The first is when the second group defends its
position by positing that reading in all the seven modes was not obliga-
tory; ‘rather, that was something that was permitted to them as an al-
lowance [wa-innam: k:na dh:lik j:8izan lahum wa-murakhkhaBan f;h],
and they were allowed to choose whichever Aarf they wished to recite, as
is mentioned in various authentic Aad;ths’.71 A second mention is in the

68
Ibn al-Jazar;, al-Nashr, i. 7.
69
Ibid, 25–6.
70
Ibid, 31.
71
Ibid, 31. The translation of the quote is Dutton’s, ‘Orality’, 29, with the add-
ition of the Arabic words.
 8A
THE EMERGENCE OF THE QIRA T 319
following statement of Ibn al-Jazar;: ‘Some people say that the allowance
to recite according to seven aAruf [inna l-tarkh;Ba f; al-aArufi l-sab6a] was
valid at the beginning of Islam, because of the difficulty involved in
having to recite only according to one Aarf’.72
These statements, whether Ibn al-Jazar;’s own or quoted by him, seem
to give the impression that he accepted the divine permission hypothesis.
However, his position is not straightforward. The complication arises
from the fact that, a couple of lines later, he writes: ‘he who says that

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

some of the Companions, like Ibn Mas6<d, allowed reading [the Qur8:n]
according to sense, has misrepresented him’ (fa-qad kadhaba 6alayh). In
contradistinction, Ibn al-Jazar;’s own position was that Ibn Mas6<d was
inserting exegesis in his recitation, or that Ibn Mas6<d’s reading was
subsequently abrogated.73 Under the latter reasoning, if I have under-
stood him correctly, Ibn al-Jazar; is attempting to make a case that all
Ibn Mas6<d’s readings were received from the Prophet, though some of
them were later abrogated in the final review (al-6ar@a al-akh;ra)—Ibn
Mas6<d was not ‘creating’ readings.74 Admittedly, this somewhat under-
mines the divine permission hypothesis which rests on reading according
to sense (al-qir:8a bi-l-ma6n:). So, should we suppose that Ibn al-Jazar;
accepted the hypothesis, or may we assume, on the basis of his inter-
pretation of Ibn Mas6<d, that he rejected it? I have no unequivocal an-
swer to offer. But I will adduce another work of Ibn al-Jazar;, his earlier

72
Ibid, 32. Again I have relied on Dutton’s translation (‘Orality’, 29), adding
some of Ibn al-Jazar;’s wording to give a more accurate picture.
73
Ibid, 32.
74
Several reports relate that Gabriel would recite the Qur8:n with the Prophet
every Rama@:n, and that he did this twice in the year before the Prophet’s death.
This gave rise to what has been known as the final review, or final recitation, al-
6ar@a al-akh;ra. The tradition is then split on who attended the final review. Some
reports say that Zayd b. Th:bit was present, while others maintain that Ibn Mas6<d
attended. See al-FaA:w;, SharA mushkil al-:th:r, vii. 138–42; Ab< Sh:ma, al-
Murshid al-waj;z, 68–9; Ibn al-Jazar;, al-Nashr, i. 32. (An anonymous reader of
an earlier draft of this paper alerted me to the possible bearing of this debate on the
divine permission hypothesis, suggesting that the reports may imply a tension be-
tween advocates of the codex of Ibn Mas6<d and those of the codex of 6Uthm:n.
This, according to the reader, weakens the hypothesis because, if the early Muslims
held that hypothesis, they would not be committed to prioritizing their favoured
codices. However, even if we accept that a tension existed and that it gave rise to
these competing reports, it need not rule out the divine permission hypothesis. It can
be argued that the concern of each group was to place their readings at the higher
end of the divine authority continuum. This is because, as explained earlier, read-
ings that were taught directly by the Prophet or approved by him are more authori-
tative than those generated by the Companions without explicit Prophetic assent.)
320 t a r e q mo qb e l
short monograph Munjid al-muqri8;n, which refers to this hypothesis.
On the question of what was included in the Companions’ compilation,
Ibn al-Jazar; supports the view that they unanimously agreed on writing
the Qur8:n according to the final review, adding: ‘and [the Companions
compiled the muBAaf] according to what God sent down, not what He
had permitted (wa-6al: m: anzala Ll:hu ta6:l: d<na m: adhina f;hi).75 In
this statement, Ibn al-Jazar; is unequivocal that the Qur8:nic variants in
the pre-canonization phase belonged to one of two categories: what God

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

sent down and what He permitted.
Having presented evidence that the divine permission hypothesis, was
(implicitly or explicitly) believed by eminent Muslim scholars until at
least the ninth/fifteenth century, it will be useful to review contemporary
thinking on the qir:8:t in two works particularly worthy of notice. The
first is 6Abd al-4ab<r Sh:h;n’s (d. 2010) discussion in his T:r;kh al-
Qur8:n. His chapter on the seven aAruf begins by noting that the prob-
lem of qir:8:t is a function of the dispensation allowed by the Prophet
(atharun min :th:ri tilka l-rukhBa).76 Reflecting on the seven aAruf
Aad;th, he makes a number of observations, and asks: did every mode
of reading issue from the Prophet (fa-hal k:na kullu wajhin mina l-
wuj<hi l-mukhtalifati B:diran min Ras<li Ll:h)? He acknowledges that
he could not provide a definitive answer and distinguishes, based on
what he was able to gather from the context of the seven aAruf tradi-
tions, three types of qir:8:t: those issuing directly from the Prophet;
those approved by him in the sense that he might hear a Companion
reading and approve that reading; and, third, readings arising due to the
general concession (fa-inna mina l-Aur<fi m: k:na mansha8uhu l-
rukhBata l-6:mma). Sh:h;n is clearly recognizing that a form of the divine
permission hypothesis was entailed in this practice. However, the con-
cession though general was by no means unrestricted or absolute. Sh:h;n
holds that it ended with 6Uthm:n’s compilation, though the spirit of it
endures (anna r<Aa h:dhihi l-rukhBati l: taz:lu b:qiya) in the form of the
different accents in which Muslims read the Qur8:n today.77
Sh:h;n also makes a useful observation, one that I have not seen in
earlier works, about the terminology of the seven aAruf Aad;th. He notes
that the two verbs that appear in the different versions of the Aad;th,
‘unzila’ and ‘iqra8’, have different connotations. Whereas ‘unzila’,

75
Ab< al-Khayr Ibn al-Jazar;, Munjid al-muqri8;n wa-murshid al-3:lib;n (eds.
MuAammad Eab;b All:h al-Shinq;3; and Ab< al-Ashb:l AAmad MuAammad
Sh:kir; Cairo: Maktabat al-Quds;, 1350 [1932]), 22.
76
6Abd al-4ab<r Sh:h;n, T:r;kh al-Qur8:n (Cairo: Nah@at MiBr li-l-Fib:6a wa-l-
Nashr wa-l-Tawz;6, [2005] 2007), 49.
77
Ibid, 68.
 8A
THE EMERGENCE OF THE QIRA T 321
according to Sh:h;n, signifies that the aAruf were sent down by God, the
term ‘iqra8’ does not necessarily indicate this; rather, it signifies that the
seven aAruf can be obtained directly via the Prophet’s reading, or (indir-
ectly) through his allowance on the basis of permissibility and choice.78
Another important contribution on the qir:8:t is that of MuAammad
Easan Easan Jabal (d. 2015), a scholar of Arabic and the Qur8:n who
taught in al-Azhar’s Kulliyyat al-Qur8:n al-Kar;m li-l-Qir:8:t wa-6Ul<mih:
in Tanta. Toward the end of his life Jabal published on the major issues in

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

the science of qir:8:t a book that can be considered a culmination of his
decades-long study of the Qur8:n. In the course of his argument that the
Qur8:n was revealed in the dialect of Quraysh, Jabal records around ten
quotations from eminent scholars (most of them taken from Ab< Sh:ma’s
al-Murshid al-waj;z) to the effect that the Companions were allowed to
read in their own dialects.79 He concludes that the majority of those emi-
nent scholars did not qualify the permissibility of reading in one’s own
dialect with the requirement that the qir:8a be acquired directly from the
Prophet (aghfal< qayda l-talaqq;). Rather, he says, the Prophet’s concession
permitting reading in one’s own dialect is to be regarded as a general per-
mission (idhnan 6:mman).80 He argues further that those who hold that a
reading should have been acquired directly from the Prophet fall into two
errors: they renounce the making easy, which was the whole point of allow-
ing the qir:8:t in the first place, and they ignore the fact that it was prac-
tically impossible for the Prophet, given his duties and responsibilities, to
teach in various dialects.81 However, Jabal understands this general permis-
sion to apply only narrowly, only for dialectal variations, such as fronting a
letter (im:la), as opposed to changing letters and words. Thus, while Jabal
allows variations in what are commonly described as the uB<l (general
principles pertaining mainly to dialectal variations in pronunciation), he
regards changes to letters and words to be permissible only as revealed
and taught by the Prophet.82 Another difference between Jabal and previ-
ous works is his view that the dialectal general permission was operative
from the Prophetic era until the time of the ten readers (al-a8imma al-
6ashara).83 In contrast to Ab< Sh:ma, for example, this is a considerable
extension of the operative period of the divine permission.

78
Ibid, 82.
79
For some of the original quotes paraphrased here, see MuAammad Easan
Easan Jabal, Min al-qa@:y: al-kubr: f; al-qir:8:t al-Qur8:niyya (Cairo:
Maktabat al-2d:b, 2012), 62–4.
80
Ibid, 73.
81
Ibid, 60–1.
82
Ibid, 76–7.
83
Ibid, 73.
322 t a r e q mo qb e l
So, on the one hand, Jabal construed the permission for the qir:8:t
more narrowly while, on the other, he furnished it with a longer life. It is
also worth mentioning that, for Jabal, the fact that the first generation of
Qur8:n reciters read in their own dialects, without going back to the
Prophet for his endorsement, is something obvious and does not even
require to be substantiated by recourse to previous authorities. Jabal also
seems to have found it troubling that his students at Kulliyat al-Qur8:n
al-Kar;m struggled with this issue.84 In a different work, Jabal expressed

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

the view that the dialectal variations introduced by the reciters were
needed (@ar<r;).85 Thus, Dutton and Jabal seem to converge here; for
Dutton also, writes that some limited variation is ‘in a sense, built in to
the very text’.86
To recapitulate the argument thus far: it is safe to say that some
Muslim scholars held the divine permission hypothesis to be true––it is
clearly present in the Islamic scholarly tradition. We found that the per-
mission was variable in form and scope. According to one of its forms,
which could be inferred from al-B:qill:n;, the permission was granted by
the Prophet and subsequently abrogated by him during his lifetime.
According to another form of the hypothesis, which seems to be more
common, the permission was granted by the Prophet and remained in
place until the compilation of the muBAaf, whether by Ab< Bakr or
6Uthm:n. Yet another form suggests that the permission granted by the
Prophet remained in place until the canonization of the ten readings.
Clearly there is a degree of ambiguity surrounding the topic of variant
readings. First of all, it must be emphasized that the permission remains
a hypothesis. Secondly, we note the arbitrary cut-off dates set by the
scholars, which raises a number of questions: on what basis does al-
B:qill:n; advance the view that the Prophet himself abrogated the per-
missibility of substituting synonyms? Is there any evidence, for Ab<
Sh:ma and others, that 6Uthm:n annulled the permission? Could it be
said that although 6Uthm:n limited the available options, he did not rule
out the permission completely? Is it even justifiable to curtail a divine
permission on the basis of an alleged consensus, ijm:6, in favour of that
curtailment?87 And is there any sound basis for arguing that the

84
Ibid, 61.
85
MuAammad Easan Easan Jabal, Wath:qat naql al-naBB al-Qur8:n; min Ras<l
All:h Ball: Allah 6alayhi wa-sallam il: ummatih (Tanta: D:r al-4aA:ba li-l-Tur:th,
n.d. [2001]), 154.
86
Dutton, ‘Orality’, 34.
87
Cf. ibid, 42: ‘So 6Uthm:n’s decision was not only caliphal but also effectively
one of ijm:6, consensus, i.e. the consensus of the Companions, which is why, as al-
Fabar; points out, and as M:lik’s judgements and Ibn Muj:hid’s legal decisions
 8A
THE EMERGENCE OF THE QIRA T 323
permission was operative until the ten readings were canonized? It
appears that there is some room for speculation here. The questions
are difficult to answer and remain far from resolved. Such issues not-
withstanding, I hope that the paper has shown that the hypothesis was
recognized and discussed in the Muslim intellectual tradition.
I now move on to the qualitative limits of the permission: was it ab-
solute or did the Prophet place restrictions on it, and if so, what were
they?

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

READING ACCORDING TO SENSE AS A
SEMANTIC CONTROL

Reading according to sense, al-qir:8a bi-l-ma6n:, has received consider-

able attention with a range of opinions, explanations and rejoinders
available in the literature.88 There is no need to go over that ground
in this paper. My concern here is to show that the qir:8a bi-l-ma6n:
concept can (and should) be understood as closely linked to the divine
permission hypothesis, specifically as a restriction on or control over that
permission. In support of that, I reflect on two versions of a tradition
attributed to the Prophet.
The tradition appears in a number of Aad;th collections, including the
Musnad of AAmad Ibn Eanbal (d. 241/855) and the Sunan of Ab<
D:w<d (d. 275/889). AAmad’s version, which is reported on the author-
ity of Ab< Bakra, includes: ‘Read it [the Qur8:n] in accordance with
seven modes of reading (6al: sab6ati aAruf), each of which is healing
and sufficing (kulluh: sh:fin k:fin), provided that you do not end a verse
of mercy with [one of] punishment, or a verse of punishment with [one
of] mercy’.89 The version of Ab< D:w<d, on the authority of Ubayy b.
Ka6b, reports that the Angel increased the concession up to seven modes

indicate, it is an act of obedience for the Muslims to abide by it, and an opening of
the door of fitna and conflict to do anything else’.
88
See Ibn Jinn;, al-MuAtasab, i. 296; Nöldeke et al., The History of the Qur8:n,
463–7; Goldziher, Die Richtungen, 33–6; 6Abd al-Jal;l 6Abd al-RaA;m, Lughat al-
Qur8:n al-kar;m (Amman: Maktabat al-Ris:la al-Ead;tha, 1981), 155–75;
MuAammad MuAammad Ab< Shahba, al-Madkhal li-dir:sat al-Qur8:n al-kar;m
(Riyadh: D:r al-Liw:8 li-l-Nashr wa-l-Tawz;6, 1987), 201–12; Sh:h;n, T:r;kh al-
Qur8:n, 119–35; 4ubh; al-4:liA, Mab:Aith f; 6ul<m al-Qur8:n (Beirut: D:r al-6Ilm li-
l-Mal:y;n, 1977), 107–8.
89
AAmad Ibn Eanbal, Musnad al-Im:m AAmad b. Eanbal (vol. 34 eds. Shu6ayb
al-Arna8<3, 62dil Murshid and Haytham 6Abd al-Ghaf<r; Beirut: Mu8assasat al-
Ris:la, 50 vols., 1993–; vol. 34 1999), 70–1 [Aad;th 20425].
324 t a r e q mo qb e l
of reading and, directly after that, reads: ‘there is not in these modes of
reading anything except what is healing and sufficing (laysa minh: ill:
sh:fin k:fin); if you say [God is] Most Hearing, Most Knowing [in place
of/or if you say] Most Mighty, Most Wise (in qulta: sam;6an 6al;man
6az;zan Aak;man), [to do so is permissible/equally healing and sufficing]
so long as you do not end a verse of punishment with [a word of] mercy,
or a verse of mercy with [a word of] punishment’.90
Al-Zuhr;, in this context, allowed reading according to sense. It is

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

reported that when he was asked about reversing the order of phrases
in Aad;th, he said that, given this (reversal) was permitted with the
Qur8:n, it should be so with Aad;th (inna h:dh: yaj<zu f; al-Qur8:n
fa-kayfa bih; f; al-Aad;th), if the meaning is rendered correctly (idh:
uB;ba ma6n: al-Aad;th). Likewise, al-Sh:fi6; (d. 204/820), in the course
of his discussion of why there are various formulas for the tashahhud in
the ritual prayer, refers to the qir:8a bi-l-ma6n: issue. After citing the
incident of 6Umar b. al-Kha33:b and Hish:m b. Eak;m (referred to
above), he says that God revealed the Qur8:n in seven modes of reading,
out of His mercy, knowing that human memory can fail (ma6rifatan
minhu bi-anna l-AifCa qad yazillu), and made it lawful for people to
read the Qur8:n in different expressions as long as the meaning is not
affected (li-yuAilla lahum qir:8atahu wa-ini khtalafa l-lafCu f;hi m: lam
yakun f; khtil:fihim iA:latu ma6nan). Al-Zamakhshar; also makes this
same point.91
We see from both versions of the Prophetic tradition that the author-
ization of the seven aAruf appears in conjunction with the concept of al-
qir:8a bi-l-ma6n:. The juxtaposition of the two seems to indicate that
they are interrelated and intended to be taken together. While not

90
Ab< D:w<d al-Sijist:n;, Sunan Ab; D:w<d (vol. 2 eds. Shu6ayb al-Arna8<3 and
MuAammad K:mil Qara Balal;; Beirut: D:r al-Ris:la al-62lamiyya, 7 vols., 2009),
602 [Aad;th 1477]. Another way to render the Aad;th would be something like:
‘each one of them is healing and sufficing whether you say: [God is] Most Hearing,
Most Knowing [or] Most Mighty, Most Wise’. My clarificatory insertions partly
draw on Ab< al-Fayyib MuAammad Shams al-Eaqq al-6AC;m:b:d;, 6Awn al-
ma6b<d: sharA Sunan Ab; D:w<d (ed. 6Abd al-RaAm:n MuAammad 6Uthm:n;
Madina: al-Maktaba al-Salafiyya, 2nd edn., 14 vols., 1968–69), iv. 350–1.
91
Shams al-D;n al-Dhahab;, Siyar a6l:m al-nubal:8 (vol. 5 ed. Shu6ayb al-
Arna8<3; Beirut: Mu8assasat al-Ris:la, 25 vols., 1996), 347; MuAammad b. Idr;s
al-Sh:fi6;, al-Ris:la (ed. AAmad MuAammad Sh:kir; Cairo: MuB3af: al-B:b; al-
Ealab;, 1938–40), 273–4; al-Zamakhshar;, al-Kashsh:f, 1003. I learned about
the views of al-Zuhr;, al-Sh:fi6;, and al-Zamakhshar; from the useful paper by
Ammar Khatib and Nazir Khan, ‘The origins of the variant readings of the
Qur’an’, available at https://yaqeeninstitute.org/read/paper/the-origins-of-the-
variant-readings-of-the-Qur8:n#ftnt100 (last accessed 14 November 2021).
 8A
THE EMERGENCE OF THE QIRA T 325
altogether discounting the plausibility of some scholars’ efforts to ex-
plain away the authorization of the qir:8a bi-l-ma6n: in the Aad;th,92 I
venture to say that qir:8a bi-l-ma6n: is a function of the seven aAruf
concession––it is there because the concession is. Indeed, the rationale
of the Aad;th suggests that they were both authorized together. It follows
that the ending of the seven aAruf concession entails the ending of the
concession to read in accordance with sense.
Further, I suggest that the qir:8a bi-l-ma6n: mentioned in the second

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

part of the Aad;th, serves as a control over the divine permission given in
the first part, namely for the seven aAruf. Whereas the first part contains
the concession in respect of the seven aAruf, the second part sets the
general limits to this concession by subjecting it to a semantic control,
namely the qir:8a bi-l-ma6n:. I therefore suggest seeing this control as a
mechanism enabling the functionality of the seven aAruf concession. It is
a necessary control accompanying the divine permission; otherwise, we
would have, so to speak, a door without hinges––one liable to fall out
and leave an unregulated opening, or liable, in order to prevent that, to
be forcibly sealed up. The qir:8a bi-l-ma6n: parameter was, therefore,
not a problem; it was a solution in anticipation of the proliferation of
qir:8:t. The regulatory principle established for the first audience of the
Qur8:n was that, although some degree of fluidity here and there was
permissible, the meaning should not be compromised. This is a wide
control—simply saying that the meaning should be maintained does
not offer a close control. Where do the believers draw the line?93
While I concede that the Aad;th sets only a general semantic control,

92
For example, some have suggested that the Aad;th only refers to God’s names
at the conclusion of verses, and hence the qir:8a bi-l-ma6n: is restricted in two ways:
it only applies to substituting God’s names and only when they appear in verse
endings. It has also been suggested, in another attempt to circumvent the problem,
that the substitution is restricted to the individual addressed in the Aad;th, namely
the Prophet. See al-6AC;m:b:d;, 6Awn al-ma6b<d, iv. 351, and for expanded discus-
sions, see the literature cited in n. 88. The latter explanation becomes difficult when
the practice of the Companions is taken into account: it is mentioned in Ibn Eajar
al-6Asqal:n;, FatA al-b:r; bi-sharA al-Bukh:r; (eds. 6Abd al-6Az;z Ibn B:z,
MuAammad Fu8:d 6Abd al-B:q;, and MuAibb al-D;n al-Kha3;b; Cairo: al-
Maktaba al-Salafiyya, 3rd edn., 13 vols., 1987), ix. 27, that a number of
Companions were reading with synonyms even though they had not heard those
readings from the Prophet (thabata 6an ghayri w:Aidin mina l-BaA:ba annah< k:na
yaqra8u bi-l-mur:dif wa-law lam yakun masm<6an lah).
93
This point was raised by Ibn al-Anb:r; (d. 328/940). See MuAammad al-
Qur3ub;, al-J:mi6 li-aAk:m al-Qur8:n (vol. 21 eds. 6Abd All:h b. 6Abd al-MuAsin
al-Turk; and MuAammad Ri@w:n 6Irqs<s;; Beirut: Mu8assasat al-Ris:la, 24 vols.,
2006), 329–30.
326 t a r e q mo qb e l
in the larger scheme of things, the historical record of the qir:8:t stands
as evidence that the first and later generations understood and knew how
to apply that control. The minor variations in recitation were permitted,
accepted, and they were, or because they were, faithful to the meanings
of the Qur8:n—the qir:8a bi-l-ma6n: parameter achieved its ‘delimiting’
task.

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

THE QUR82N AND THE QIR282T—DOES THE
DISTINCTION HOLD?

In this last section I consider the distinction between the Qur8:n and the
qir:8:t. In what appears to be an effort to distance the Qur8:n from the
problems associated with the qir:8:t, some scholars have sought to
distinguish them as separable. One of the sources in which this differen-
tiation is explicit is al-Zarkash;’s (d. 794/1392) al-Burh:n f; 6ul<m al-
Qur8:n. Al-Zarkash; explains that the Qur8:n and the qir:8:t are two
different realities (wa-6lam anna l-Qur8:n wa-l-qir:8:t Aaq;qat:n muta-
gh:yirat:n) by arguing that whereas the Qur8:n is the revelation sent
down upon the Prophet, the domain of the qir:8:t concerns the expres-
sions of the Qur8:n.94 This distinction was picked up by al-Suy<3; (d.
911/1505),95 among others, and found its way into modern works on the
qir:8:t.96 However, there is no consensus on this matter; some scholars
dispute this distinction.97 An alternative take on this issue is to say that
Qur8:n and qir:8:t are the same thing.

94
MuAammad al-Zarkash;, al-Burh:n f; 6ul<m al-Qur8:n (ed. MuAammad Ab<
al-Fa@l Ibr:h;m; Cairo: Maktabat D:r al-Tur:th, 4 vols., n.d.), i. 318.
95
al-Suy<3;, al-Itq:n f; 6ul<m al-Qur8:n, ii. 523.
96
See, for example, Sh:h;n, T:r;kh al-Qur8:n, 40; Jabal, Min al-qa@:y: al-
kubr:, 80–1.
97
On this view, see MuAammad S:lim MuAaysin, al-Qir:8:t wa-atharuh: f;
6ul<m al-6Arabiyya (Cairo: Maktabat al-Kulliyy:t al-Azhariyya, 2 vols., 1984), i.
10. Shukr;, al-Qu@:h, and ManB<r, Muqaddim:t, 49, also dispute this distinction
although they try to find a way out for al-Zarkash;. They quote al-Zarkash;’s al-
Burh:n to the effect that he did not deny the interrelatedness of the Qur8:n and
qir:8:t. However, this statement does not appear in al-Burh:n. Additionally, its
language is far from what an eighth-/fourteenth-century scholar might have writ-
ten. This statement, which recurs in other modern works too, appears to have been
misattributed to al-Zarkash;—at least, it is not found in the printed editions of al-
Burh:n. Another scholar who seems to object to this distinction is Al-Imam,
Variant Readings of the Qur8an, 88, as he writes, after quoting al-Zarkash;: ‘In
 8A
THE EMERGENCE OF THE QIRA T 327
But before I explain that alternative, it is worth noting that an explicit
separation of the qir:8:t from the Qur8:n marks a break with, at least the
majority of, the scholarship on qir:8:t before al-Zarkash;. In the fourth
century AH Ibn Ab; D:w<d gave to the chapter on Prophetic readings in
his Kit:b al-MaB:Aif the following title: ‘The readings reported from the
Prophet are to be treated as his codex’ (m: ruwiya 6ani l-Nabiyyi mina l-
qir:8:ti fa-huwa ka-muBAafih).98 Here Ibn Ab; D:w<d equates the
qir:8:t with a hypothetical Prophetic muBAaf. This muBAaf and the

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

qir:8:t are not seen as separate; rather, the qir:8:t are seen as constitutive
of the muBAaf. The inseparability of Qur8:n and qir:8:t can also be
inferred from al-R:z;’s (d. 606/1210) comments on the variants in Q.
20:63. He writes that the notable scholars ruled that the ‘irregular var-
iants’ in Q. 20:63—meaning those not transmitted by taw:tur—should
not be accepted, for if we allow any addition to the Qur8:n by way of
isolated mono-transmitted reports, we cannot definitively say that what
we have to hand is the complete Qur8:n. That is because, he adds, if
these ‘irregular readings’ are accepted as part of the Qur8:n (ma6a kaw-
nih: mina l-Qur8:n), even though they were not transmitted to us by
taw:tur, then the same claim could be made for yet more ‘irregular
readings’ too,99 and open the door to the destabilization of the
Qur8:n. What is most relevant here is that al-R:z; treats the qir:8:t as
mina l-Qur8:n, ‘in and of’ the Qur8:n. Indeed, it is precisely because
qir:8:t are mina l-Qur8:n that controls such as ruling out ‘irregular
readings’ are applied by al-R:z; and the scholars he is quoting. This
lends more weight to my contention that the line between the Qur8:n
and qir:8:t is blurred.
We can now look more closely into the distinction between Qur8:n
and qir:8:t. For those who adopt the view that all the qir:8:t are divinely
inspired/authorized, the distinction cannot stand inasmuch as every
qir:8a is then a building block of the Word of God. The position of those
who hold that the qir:8:t were later inventions is consistent with a dis-
tinction between the Qur8:n and the qir:8:t on the ground that their
authorship is distinct––and this position is coherent with the view of
those who assume the existence of an ur-Qur8:n.100 However, the

fact, no major difference exists between the authentic recitations and the Qur’an,
and the relation between them is that of the parts to the whole’.
98
Ibn Ab; D:w<d, Kit:b al-MaB:Aif, i. 388.
99
Fakhr al-D;n al-R:z;, Tafs;r al-Fakhr al-R:z;: al-mushtahir bi-l-Tafs;r al-
kab;r wa-Maf:t;A al-ghayb (Beirut: D:r al-Fikr, 32 vols., 1981–85), xxii. 75.
100
For a discussion on the ur-Qur8:n, see Fred Donner, ‘The Qur8:n in recent
scholarship: challenges and desiderata’ in Gabriel Said Reynolds (ed.), The Qur8:n
in Its Historical Context (London: Routledge, 2008): 29–50, at 31–43, and the
rejoinder in Dutton, ‘Orality’, 44–9.
328 t a r e q mo qb e l
either/or option is not the only one. The view that some of the qir:8:t are
not divinely inspired entails that there is some distance between qir:8:t
and Qur8:n, a partial distinction. If the divine permission hypothesis is
accepted, how should we regard, on the one hand, the qir:8:t initiated by
the Companions and approved by the Prophet and, on the other, those
qir:8:t initiated by the Companions under the general permission but
without reverting to the Prophet for explicit divine assent? Although
these are moot points, it could be said that the qir:8:t initiated by the

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

Companions which received Prophetic approval are ‘in and of’ the
Qur8:n. As for the qir:8:t that emerged under the general permission
without specific Prophetic authorization, I am less confident. However, I
would conjecture that they were deemed to be ‘in and of’ the Qur8:n in
the eyes of the Companions inasmuch as they presumably recited those
qir:8:t in their ritual prayer. At least for the first generation, I would
think, there was in practice no distinction between the Qur8:n and the
qir:8:t.

CONCLUSION

Various theories for the emergence of the qir:8:t have been suggested.
This paper has proposed an alternative, the divine permission hypoth-
esis. Its core argument is that eminent scholars until at least the ninth/
fifteenth century believed that from its inception the Qur8:nic revelation
was accompanied by a divine permission allowing the first recipients of
the Qur8:n some flexibility in reading it. What followed from this per-
mission is the circulation of variant readings that did not have one and
the same divine authority. They ranged from those directly taught by the
Prophet, to those approved by him, to those generated by the
Companions under the general divine permission but without any refer-
ral to the Prophet. It was also believed, by some of the scholars who
endorsed the divine permission hypothesis, that this flexibility, con-
trolled within strict semantic boundaries, was lifted during the
Prophet’s lifetime, by others of them that it lasted up to the time of
the 6Uthm:nic codex, even though that codex is said to have included
only the qir:8:t at the upper end of the divine authority continuum—
those explicitly taught by the Prophet and confirmed to him during the
final review (al-6ar@a al-akh;ra).
While the divine permission hypothesis may not be the only tenable
account of the emergence of the qir:8:t, it does have the merit of satis-
factorily accounting for many of the problems connected with the qir:8:t
 8A
THE EMERGENCE OF THE QIRA T 329
and, more widely, the history of the Qur8:n. Its explanatory efficacy is
evident in a number of areas:
it allows us to take much more of what is said in the Islamic tradition
at face value;
it is able to accommodate the statements of 6Umar and 6Uthm:n that
the Qur8:n was revealed in the dialect of Quraysh;
it is consistent with the qir:8a bi-l-ma6n: traditions;

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025

it streamlines our understanding of the so-called ‘Companion codices’
problem;
it offers a useful perspective on the discomfort of some Companions
with some qir:8:t;
it accounts for the existence of an unlimited number of variants in
early Islam (al-Badr al-awwal) as Ibn al-Jazar; points out;101 and,
finally,
it enables a more charitable interpretation of the line of qir:8:t
critique associated with such towering figures in the history of
Qur8:nic scholarship as al-Fabar; and al-Zamakhshar;102—for, as
I have suggested, the permission hypothesis is a way to understand
their general argument and their particular criticisms.
That said, I do not claim that the divine permission hypothesis
resolves all the questions associated with the emergence of the
qir:8:t. Rather, it leaves a number of questions unanswered that offer
directions for future research. Among these are––how to distinguish
between the qir:8:t that the Prophet actually recited and those that he
tacitly endorsed;––how and by what criteria to identify, if one both
accepts the permission hypothesis and attributes qir:8:t to difficulties
with the rasm, the variants resulting from these different prompts; ––
can it be convincingly demonstrated that the divine permission lasted
beyond the stabilization of the text as a muBAaf, beyond the gener-
ation of the Companions. Finally, it would be of the highest interest to

101
Ibn al-Jazar; comments on the view espousing that the widely transmitted
variants are unlimited (l: Aadda lah:) by saying that if it is meant by that view
the first generation of Islam, then it is possible. See Ibn al-Jazar;, Munjid al-muq-
ri8;n, 16. Clearly, then, Ibn al-Jazar; does not see a problem with saying that there
were numerous, in fact unlimited, variants in the formative period of the qir:8:t.
102
This is to say that their criticisms of the qir:8:t, understood in light of the
divine permission hypothesis, could be taken as a criticism of the variants which
they thought were initiated by the Companions, not the divinely inspired variants.
330 t a r e q mo qb e l
be able to understand how and when, and for what reasons, the dom-
inant emphasis among believers came to be that the Qur8:n is and
always was entirely and exclusively God’s speech dictated to God’s
Messenger. Many of the issues around the formative history of the
Qur8:n remain, for the present, unresolved, and may well be unre-
solvable, but they undoubtedly deserve further reflection and
research.

Downloaded from https://academic.oup.com/jis/article/33/3/299/6594398 by guest on 10 January 2025