George ColouredDotsQestion 2015

Centre for Islamic Studies at SOAS
Coloured Dots and the Qestion of Regional Origins in Early Qur'ans (Part I) / ‫‬ةنولملا‫ ‬طاقنلا(‫‬
‫‬لوألا‫ ‬ءزجلا(‫ ‬ةركبملا‫ ‬فحاصملل‫ ‬ةيميلقإلا‫ ‬لوصألا‫ ‬ةلأسمو‫‬
Author(s): Alain George and ‫جروج نيلأ‬
Source: Journal of Qur'anic Studies, Vol. 17, No. 1 (2015), pp. 1-44
Published by: Edinburgh University Press on behalf of the Centre for Islamic Studies at
SOAS
Stable URL: https://www.jstor.org/stable/24280713
Accessed: 09-06-2023 17:58 +00:00
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Edinburgh University Press, Centre for Islamic Studies at SOAS are collaborating with
JSTOR to digitize, preserve and extend access to Journal of Qur'anic Studies
This content downloaded from 103.25.54.11 on Fri, 09 Jun 2023 17:58:51 +00:00
All use subject to https://about.jstor.org/terms
Coloured Dots and the Question of Regional
Origins in Early Qur'ans (Part I)
Alain George
University of Edinburgh
During the first centuries of Islam, the written notation of the Qur'an underwent a
gradual evolution.1 After an early stage represented by the 'Hijazi' tradition, red dots
were introduced into 'Kufic' Qur'ans to mark short vowels.2 This system was soon
expanded by assigning more functions to the red dots, sometimes supplemented by
yellow, green, and blue dots; and by creating new orthographic signs. These devices
were used in different ways by different vocalisers. Few textual sources dealing with
this subject survive. By far the most consequential is al-Muhkam fi naqt al-masähif
(lit. 'The Precise on the Vocalisation of Qur'ans') by Abü cAmr cUthmän b. SacId
al-Dänl (371-444/982-1053). Following a line of work initiated by Yasin Dutton,
the present study will confront the assertions of the Muhkam with a sample of key
manuscripts in an attempt to gain insights into the regional origins of early Qur'ans.
Al-DänI's shorter treatise on the subject, the Kitäb al-naqt ('Book of Vocalisation'),
will also be considered where relevant.3
Since early Qur'ans first came under the scrutiny of modem historical science some
two centuries ago, their study has been hampered by two major obstacles: chronology
and provenance. As the manuscripts were repeatedly unbound and dispersed over
the centuries, their opening and closing pages have been lost, and with them
the colophons and legal deeds (waqfiyyät) that may have contained contextual
information. Only a handful of waqfiyyät survive, and no colophons at all, among the
hundreds of thousands of early Qur'anic folios from the first three centuries of Islam.
In recent years, advances have been made in our understanding of the chronological
development of early Qur'anic calligraphy.4 Coloured dots, as will soon become
apparent, might provide some insights into the issue of provenance.
Journal of Qur'anic Studies 17.1 (2015): 1-44

Edinburgh University Press
DOI: 10.3366/jqs.2015.0178
© Centre of Islamic Studies, SOAS
www.euppublishing.com/joumal/jqs
2 Journal of Qur'anic Studies
The notation systems studied here reflect more or less complex rules of grammar and
recitation. In order not to overburden the text, readers are referred to the general
introduction to this subject published in English by Muhammad Surty, and to more
specialised texts for points of detail.5 One particular term, naqt (lit. 'dotting'), was
used by al-Däni and his contemporaries to refer to Qur'anic vocalisation and its
trappings. The cognate terms naqqata and näqit (pi. nuqqät) will therefore be
translated respectively as 'to vocalise' and 'vocaliser', while harakät will be given as
'vowels'. When needed, hamza will be noted more fully than in common
transliteration, and tanwln will be indicated through superscript. Arabic plurals
have generally been used, e.g. alifit, hamzät, silät, and so on.
The following abbreviations have been employed:
• BNF: Bibliotheque Nationale de France (Paris)
• Cambridge: Cambridge University Library (Cambridge, England)
• CBL: Chester Beatty Library (Dublin)
• Freer: Freer Gallery of Art, Smithsonian Institution (Washington, DC)
• Khalili: Nasser D. Khalili Collection of Islamic Art (London and Geneva)
• Met: Metropolitan Museum of Art (New York)
• NLR: National Library of Russia (Saint Petersburg)
• TIEM: Türk ve islam Eserleri Müzesi (Istanbul)
The main reference work consulted for Qur'anic readings is cAbd al-Latif al-Khattb's
Mucjam al-qirä'ät.6
Al-Däni on the Regional Habits of Vocalisers
Al-Däni was an Andalusi scholar of the Qur'an and religious sciences. A brief
autobiographical account of his life and travels was recorded by his student Abü
Däwüd Sulaymän b. Najäh (Valencian, 413-496/1023-1103) and repeated by several
later writers.7 Bom and educated at Cordoba (Qurtuba), al-Däni set out in 397/1007
for a journey to the central Muslim lands, spending four months in al-Qayrawän and a
year in Cairo before heading to Mecca for the Hajj. On his way back he stopped in
Cairo and al-Qayrawän again before reaching Spain in 399/1009. In each of these
cities, he learned aboutfiqh and Qur'anic readings from local authorities. About Cairo
(Misr), he notably says: 'There I read the Qur'an, wrote hadith, fiqh, qiräDät, and
other things under a number of Egyptians, Baghdadis, Syrians, and others.'8 He thus
appears to have acquired first-hand knowledge of reading systems used in different
regions of the Islamic world. Several of his biographers, including Ibn Bashkuwäl
(Cordoban, wr. 534/1139), also state that he was a proficient calligrapher.9 This seems
Coloured Dots and the Question of Regional Origins 3
confirmed by passing remarks made by al-Dänl himself in his writings, as when he

notes about läm-alif in the Kitäb al-naqt:10
Whoever has mastered the art of calligraphy (sinäcat al-khatt) among

scribes past or present (min al-kuttäb al-qudamäJ wa-ghayrihim) will
begin by tracing the left side, before the right; only those who ignore
the art of tracing (sind'at al-rasm) will proceed differently. This is the
same principle (manzila) as when one begins by tracing the alif before
the mim in mä and similar forms involving two letters.
Due to political turmoil in Cordoba, four years after his return, al-Däni moved to
Zaragoza (Saraqusta), where he remained for seven years. He then stayed in Mallorca
(Mayurqa) for eight years before settling down in Denia (al-Dänya), the town on the
eastern coast of al-Andalus after which he was eventually named. Al-Däni's renown,
especially in the field of qirä°ät, endured long after his death, both in his region and in
the central Islamic lands.11 He was the author of numerous works on the Qur'anic
sciences, of which a few are extant, including the Kitäb al-naqt and al-Muhkam fi naqt
al-masähif, respectively published in editions by Otto Pretzl and cIzzat Hasan.
Just over a decade ago, Yasin Dutton highlighted some key passages from the
Muhkam and confronted them with a sample of 21 early Qur'an fragments from the
Bodleian Library in Oxford.12 His observations showed that distinct but internally
coherent systems were applied in different manuscripts, and that some of these could
be matched with al-Däni's observations. In the present study, I will take this line of
investigation one step further by enlarging the sample to key manuscripts from other
collections and by focusing on the question of regional origins.13 The latter is not
specifically treated by al-Däni, who is primarily concerned with issues of grammar
and recitation. But he does provide scattered indications that can allow us to build a
picture of the regional habits of vocalisers, albeit incomplete.
1. The Umayyad Period
The oldest vocalisation system, writes al-Däni, involved red dots placed above, below,
or on the line to mark fatha, kasra, or damma respectively - a convention which
remained at the basis of later Kufic vocalisation. He cites traditions that ascribe its
invention to different authors of the late first to early second/late seventh to early
eighth century: Abü'1-Aswad al-Du3ali (d. 69/689), Yahyä b. Yacmur (or Yacmar,
d. before 90/710 or in 129/747) and Nasr b. cAsim al-Laythi (d. 90/710).14 The same
men were often credited with the establishment of Arabic grammatical science (nahw)
by writers of the third/ninth century onwards.15 The historicity of both sets of

assertions remains elusive as far as individual names are concerned; but with regard to
Qur'anic notation, it is possible to derive some related evidence from extant

manuscripts.
The earliest Qur'ans were written in the family of scripts called 'Hijäzl' in modern
scholarship. Recent studies have shown that the bulk of these manuscripts must date
to the first/seventh to early eighth century, and that they were probably written far
beyond the Hijäz.16 Vocalisation is absent from Hijäzl Qur'ans, with rare exceptions,
such as BNF Arabe 6140a and Cambridge Add. 1125.17 These two fragments, which
are probably from the same manuscript, have red dots for the vowels, tanwin,
and hamza\ yet it is difficult to ascertain whether this layer of notation is original.
The theoretical possibility of later additions is highlighted by instances of re-inking,
additions, and corrections observed in other Hijäzl manuscripts, which show that they
were used well beyond the time of their production.18
19
Writing about a slightly later period, al-Danl notes:
I have seen (wasala ilayya) an old mosque Qur'an (mushafjämic catlq)

written at the beginning of the caliphate of Hishäm b. cAbd al-Malik
in the year 110-the date was written at the end where it said:
'This was written by Mughlra b. MInä in Rajab of the year 110
[October-November 728].' The vowels, hamzät, tanwin, and tashdld
were all marked by red dots, as we have related was the practice of
early vocalisers of the people of the Mashriq.
The mention of an 'old mosque Qur'an' brings to mind a corpus of monumental

manuscripts, most of them in style C.Ia, reflecting official Umayyad patronage and
probably intended for use in major mosques.20 Al-Dänl appears to be citing a final
colophon, which is worthy of notice since the earliest extant Qur'an colophons date to
the fourth/tenth century. This makes Mughlra b. MTnä the earliest calligrapher - and
the only Umayyad calligrapher - whose name is known (albeit indirectly) from an
actual manuscript; he remains otherwise unknown. As regards the vocalisation
however, it is impossible to assert whether the red dots that al-Dänl saw were original.
A Qur'an fragment belonging to the same period might bring us into more certain
grounds: the so-called 'Umayyad codex of Fustat', a name coined by Francois
Deroche on the basis of its city of discovery. This manuscript was probably produced
during the reign of cAbd al-Malik (65-86/685-705); its leaves are now scattered
between Saint Petersburg (NLR Marcel 11, 13, and 15, collectively referred to as
'Marcel 13' hereafter) and Paris (BNF Arabe 330c).21 Here abjad letter numerals
written in gold and outlined in black mark every fifth verse: their content, shape, and
layout suggest that they are original22 I was able to study the vocalisation of Arabe
330c at close quarters. In two areas of overlap (ff. 13r, 16v), the abjad letter covers the
red dot: note, on f. 16v (line 15), the way its red hue reappears underneath the gold
that has flaked off (fig. I).23 This implies that the vocalisation was executed after the
text and before the illuminated letters, probably as part of the original manuscript. In
other words, red dots appear to have been used to mark vowels, tanwin, and hamza in
Coloured Dots and the Question of Regional Origins
\M
,*a\
Fig. 1. Detail of BNF Arabe 330c, f. 16v, line 15.
this Qur'an of the late first/late seventh to early second/eighth century: the basis of the
system might therefore have existed by this early date in the Umayyad period.
In Arabe 330c and Marcel 13, sila is occasionally indicated by red horizontal strokes
placed just below the middle of the shaft of alif(e.g. Arabe 330c, f. 12r, line 7): these
appear to be in the same ink as the red vocalisation, with which they may be
contemporaneous.24 Most diacritics, as well the occasional small black alif signalling
a harfzä°id in Marcel 13, are later additions made in a darker ink than the text.25 The
original diacritics are relatively sparse, and tend towards the form of a thickened dash,
sometimes almost circular. Qäf was noted by placing a dot below the letter, a
convention different from those of later Qur'ans, both Mashriql and MaghribI, and
also known from other Qur'ans of the same period.26
One might also adduce, with regard to the beginnings of vocalisation, two textual
reports from the Musannaf of cAbd al-Razzäq al-SancänI (126-211/744—827). This
early compilation of religious traditions (akhbär) contains opinions for or against the
vocalisation of Qur'ans ascribed to Ibrahim (al-NakhacI, c. 50-96/670-717), al-Hasan
(al-Basri, c. 21-110/642-728), and Ibn Sirin (c. 34-110/654-728).27 The chains of
transmission (asänld) lead back three generations from the time of the writer. Their
very inclusion implies that the introduction of Qur'anic vocalisation no longer
belonged to recent memory by the days of cAbd al-Razzäq and his teachers, Sufyän
al-Thawri (97-161/715-77), and cAbd Allah b. Kathlr, a transmitter of Shucba
(c. 82-160/702-76): had this been the case, they could not have diffused this
information credibly in the eyes of their contemporaries. One must thus look beyond
their lifespan for the source of this notation, which again leads back to the Umayyad
era. The dates of Ibrahim, al-Hasan, and Ibn Slrin would point to a time earlier
than 110/728, and probably not after the reign of al-Walld (86-96/705-15). Thus this
early textual source corroborates the evidence of Arabe 330c with regard to
chronology, although the content of the traditions need not be accepted word for
word: two slightly contradicting opinions are reported, for example, on the authority
of Ibn Slrin. These observations, being drawn from one textual source and one
early Qur'an fragment (Arabe 330c), would deserve to be expanded into a more
comprehensive study, notably of Umayyad manuscripts. As they stand, they suffice to
raise the hypothesis that Qur'anic vocalisation had been established by the late first/
early eighth century.
Turning to the question of regional origins, it is difficult to assert where this

codification first took place. As shown by previous studies, the verse count, codex
variants, and decoration of Marcel 13 and Arabe 330c point to a production in Greater
Syria (biläd al-Shäm).2H By contrast, the persons listed by al-Dänl as possible

inventors of the red dots were all active in Iraq. This attribution becomes explicit
when al-Danl marks his approval of the following statement by Abü Hätim Sahl b.
Muhammad (al-Sijistänl, Basran, d. 255/869):29
Vocalisation is [a creation] of the people of Basra; all others took it

from them, even the people of Medina. The latter used to have a
different vocalisation which they abandoned for the vocalisation of the
people of Basra.
Given the lack of Umayyad manuscripts attributable to regions other than Greater
Syria, it is difficult to evaluate the merits of this claim. Coin issues from Wäsit, the
Umayyad capital of Iraq from the late first/early eighth century onwards, do provide
related evidence: they show that, in the numismatic realm, this city was at the
forefront of the reform of Arabic script between 85/704 and 90/709, under the
governorship of al-Hajjäj (75-95/694-714). Textual sources, including an early
testimony from Mälik b. Anas (d. 179/796) recorded by Ibn Zabäla (wr. c. 199/814),
also indicate that the same al-Hajjäj sent large Qur'ans, presumably commissioned in
Iraq, to major cities of the empire.30
Other writers portray al-Hajjäj as the driving force behind the orthographic
improvements devised by the above Basrans.31 This assertion is in itself not
implausible given the historical context, yet its value is undermined by the number of
authors credited with the new system.32 What is more, these texts assert that the
diacritical signs were invented as part of the same process, erroneously since dated
inscriptions and papyri, together with the Hijäzl corpus, make it clear that these
already existed at least half a century earlier.33 Finally, the sources for this tradition
are late. The earliest known writer to attribute the foundation of Arabic grammar
to Abü'1-Aswad is Ibn Salläm al-Jumahi (c. 139-231/756-845), but his statement
is about grammatical theory and syntax, not vocalisation.34 Al-Jähiz (c. 160—255/777—
869) is cited by several later authors as having written in his Kitäb al-amsär wa- 'ajib
al-buldän that Nasr b. cÄsim was the first to vocalise Qur'ans, and that he was called
'Nasr al-Hurüf ('Nasr of the letters/variants').35 If authentic (which remains to be
confirmed), this would be the earliest such assertion to have emerged so far, dating to
a time when literature about 'the first to ...' (awä°il) was beginning to develop in
earnest.36 It would also represent an early stage in a process of amplification and
harmonisation of this historical narrative that continued into the fourth/tenth century,
and beyond. This process of 'growing backward', as Rafael Talmon called it, has been
documented for the birth of Arabic grammatical science, where part of its rationale
was to establish the pre-eminence of the Iraqi school.37 A similar bias towards Iraq
may also have been at play with regard to the origins of Qur'anic vocalisation.
In sum, while the names traditionally cited with regard to early vocalisation cannot be
completely discarded, the sources are not reliable enough to accept them as historical
information. And while the vocalisation system using red dots may conceivably have
first emerged in Iraq, the extant material and textual evidence is too limited to confirm or
reject this hypothesis; the manuscript record suggests Syria as another possibility. The
presumed link with al-Hajjäj is even more uncertain. All that can presently be stated
with reasonable confidence is that red dots were introduced into Qur'ans in the
Umayyad period, possibly between the reigns of cAbd al-Malik and al-Walld, and that
al-Hajjäj was involved in a process of calligraphic reform instigated during those reigns.
Beyond this starting point, al-Dänl reveals precious little of his views on the historical
development of notations systems; these, he dispenses in fragments scattered

throughout a work focussed on grammatical and notational matters. In order to retrace
his logic, we will proceed with a summary of his main assertions about Medina and
the Maghrib, then Iraq and the Mashriq, before confronting them with surviving
manuscripts.
2. Medina and the Maghrib
Al-DänT saw the Medinan system as having acquired a distinctive character at an

early date:38
The vocalisers of the people of Medina, both in early times and now,
solely use red and yellow for vocalising their Qur'ans (fi naqt
masähifihim). Red is used for the vowels, sukün, tashdld, and takhfif,
and yellow is used specifically for hamzät.
The contrasting functions assigned to red seem to imply the existence of several signs
in this colour. This is confirmed by the following citation ascribed to Qälün (Medinan,
d. 220/835):39
In the Qur'ans of the people of Medina, mukhaffaf letters carry a red
circle (dära), as do musakkan letters ... Letters dotted in yellow are

hamüza.
Elsewhere, al-Dänl notes that red circles were also used for letters omitted in
pronunciation (al-harf al-säqit min al-lafz, generally known as harf zädd)40 He
explains that the people of Medina chose to note hamza as a yellow dot 'in order to
distinguish it from the harakäf, since unlike them, it is a letter; Iraqi vocalisers, by
contrast, simply used red for both purposes.41 He mentions the reproved practice of
marking vocalisation in black, and more generally in the same ink as the text.42 The
underlying logic was to keep the core of the written text, the rasm as recorded in the
earliest Qur'ans, distinct from later orthographic layers. Al-Dänl also cites a passage
in the Kitäb al-naqt of Ibn Mujähid (Iraqi, d. 324/936) stating that readers
comprehend shapes more quickly than colours, which implies a cognitive rationale for
introducing different shapes in notation.43
The people of Medina, al-Dänl notes, mark consecutive hamzät as two yellow dots,
even though they were pronounced as a single hamza in their recitation.44 He quotes
Qälün again as saying:45
In the Qur'ans of the people of Medina, mushaddad letters carry a däl,
and this däl opens upwards (wa-fathat al-däl fawq)... For kasr, [it was
placed] below the letter.
This case was noted with däl, explains al-Dänl, as the last letter of the word shadld",
following a comparable convention, the first letter of the same word (shin) marked
shadda in the Mashriq. He adds that the notation of shadda through däl was used by
Medinan vocalisers 'old and new' (min salafihim wa-khalafihim).46
The Medinan notation system, writes al-Dänl, was adopted by 'the majority (cämma)
of the people of the Maghrib, be they Andalusi or other'. He himself had observed this
in Qur'ans from the days of al-Ghäz! b. Qays (Cordoban, d. 199/812), 'the companion
of Näfic b. Abl Nucaym [Medinan, d. 169/785] and transmitter of Mälik b. Anas
[Medinan, d. 179/796]'.47 Since al-Ghazi trained under these two Medinans, it is
plausible that he also learned the notation system of their city, although al-Dänl does
not explicitly credit him with its introduction to the West.
In other parts of the Muhkam, he describes various habits of 'the people of our land'
(ahl baladina), which could designate either al-Andalus or the broader Maghrib.
While the former may have been thought to be al-Dänl's natural horizon, both terms
seem to be used as equivalents in the text, as for example in the inclusion of Andalusis
amongst the people of the Maghrib in the above citation. Thus it seems that his
comments might be applicable to the Maghrib as a whole, including Spain, North
Africa, and presumably Sicily, without ruling out further differentiations. Iraq is,
likewise, linked to the Mashriq in al-Dänt's descriptions, but the area encompassed is
more difficult to circumscribe: judging from general usage in texts of the period, it
certainly also included Iran and the eastern Islamic lands, possibly along with Syria,
and perhaps Egypt.
48
In order to illustrate the early vocalisation system of his region, al-Danl observes:
I have seen a Qur'an written and vocalised by Hakim [Hukaym?] b.

cImrän al-Näqit, the vocaliser of the people of al-Andalus, which he
had written in the year 227 [842 A.D.], The vowels were indicated by
red dots, the hamzät by yellow [dots], and initial alifät al-wasl
[mubtada ', i.e. after a pause in recitation] by green [dots]. Silät, sukün,
and tashdld were marked in a thin red pen (bi-qalam daqlq
bi'l-humra), in the way that we have related about the vocalisers
of our land. The sila was above the alif if preceded by a fat ha, below
it if preceded by a kasra, and along its middle if preceded by a
damma. Alifi omitted in the rasm (al-alifät al-mahdhüjut min al-rasm)
were included in an abbreviated form (ikhtisdr) in red. There was a
small circle in red for unpronounced letters (hurüf zawd'id) and
mukhajfaf letters, as in anä la°awdacü [Q. 9:47], a-fa'ln mitta
[Q. 21:34],°ul\aYika and a-man huwa q[ä]nif" [Q. 39:9], as we have
shown about the people of Medina, and as became the custom of the
people of our land.
This description brings together an array of evidence not otherwise extant in the
record: a set of notational features combined with a date, scribe name, and indication
of regional origin. The words that introduce the historical information (katabahu wa
naqqatahu ...) suggest that al-Dänl might again be paraphrasing a colophon. The
calligraphy and vocalisation of the (lost) Qur'an of 227/842 thus appear to have been
executed by the same Hakim (or Hukaym) b. Hmrän, although one cannot rule out that
he was overseeing a collaborative enterprise. In other manuscripts, these two tasks -
calligraphy and vocalisation - may conceivably have been carried out by one or
several persons, depending on local usages and skills.
Al-DänI's numerous references to nuqqdt and ah I al-naqt ('vocalisers') does bring to

mind a specialised task performed after the text had been written by the scribe. Indeed,
a khabar cited by Ibn Abi Däwüd (230/844-316/929) attributes the following opinion
about Qur'anic manuscripts to al-Hasan al-Basri: Ί see no problem with their sale and
purchase, and with their vocalisation for pay.'49 Regardless of the historicity of this
assertion with regard to the late first/early eighth century, it implies that vocalisation
could be carried out as a separate task at the time of composition, in the late third/ninth
to early fourth/tenth century.
One might logically expect this task to have been primarily carried out by religious
scholars, given the advanced knowledge of the Qur'an, grammar, and recitation it
required. Al-DänT himself was a Qur'an scholar and calligrapher with evident practical
and theoretical knowledge of vocalisation. His forebear Hakim b. Hmrän is also

identified elsewhere in the Muhkam as a student (sdhib, lit. 'companion') of al-Ghäzi
b. Qays, which again implies a religious training.50 Given the limited evidence at our
disposal, one should not exclude the possibility that people from other social
backgrounds were sometimes involved in this activity.
The mention of al-Ghäz! as a teacher of Hakim also reinforces the presumption that
the former played a part in the transmission of the Medinan vocalisation system to
al-Andalus. Al-Däni's observation of the notation used in this particular Qur'an does
confirm the presence of several features ascribed to Medinan origins, namely the red
and yellow dots, the notation of sukün and shadda in a thin red pen, and the red circles
for khafif and hurüf zawäJid. But there are also differences that could reflect a distinct
Maghribi evolution: the green dots for alif al-wasl·, the unspecified signs made with a
thin red pen for sila; and the notation of alif mahdhüfa in red.
Other passages provide further elaboration on these points. Shadda would have
been noted as a däl in the Maghrib, both in al-Dänf s days and earlier, as in Medina.51
Al-Dänl also writes: 'Sukün is marked by the majority of the people of our land,
old and new, as a stroke (jarra) above the musakkan letter, be it a hamza or another
letter'; he draws a contrast with the Medinan convention of marking this case and
takhfif 'as a small circle above the letter'.52 The latter Medinan convention appears to
have also existed in the Maghrib, at least initially:53
Early vocalisers of the people of Medina and the people of our land
used a small red circle for letters noted in writing [but] omitted in
pronunciation (hurüf zawä°id)·, and for mukhaffaf letters, whether
accepted or not, indicating when this was so and giving the correct
recitation.
This is indeed the convention observed by al-Dänl in the Qur'an of 227/842. In the
Kitäh al-naqt, he also states that it was used for these two functions by vocalisers
(ahl al-naqt) 'old and new', following the precedent of the people of Medina - which
seems to imply its continued relevance in his time.54 As a side note, he relates its form
to zero as a placeholder in mathematics:55
This same circle is the small zero (sifr) which arithmeticians (ahl
al-hisäb) use for the absent digit in calculus (al-cadad al-macdüm fi
hisäb al-ghubär) to indicate its absence, which is like the absence of
the hurüf zawä 'id from pronunciation, the absence of tashdld from
the hurüf mukhajfafa, and the absence of haraka from the hurüf
musakkana.
The stroke (jarra) cited above for sukün might also have been used for takhfif, since
he remarks in the Muhkam that the latter was indicated by a stroke 'like a horizontal
alif (alif mabtüha). This form, he explains, stood for the initial khä' of 'khafif,
abbreviated to its lower part for practical purposes; elsewhere, the fatha of modern
vocalisation is also described as an alif mabtüha, which gives an idea of its form.56
The convention would have originated with Slbawayhi (d. c. 180/796) and his pupils,
who noted khafif with the full letter khä3.57
Having defined sila as the sign used to mark alif al-wasl, al-Däni remarks:58
The vocalisers of the people of our land, old and new (qadlma"
wa-hadith""), mark it with a stroke like the one for sukün ... But should
it be marked through a small circle, this is also correct, since the circle
stands for sukün and for unpronounced letters (al-harf al-säqit min al
lafi) among the people and vocalisers of Medina.
He notes the Maghrib! usage of adding a dot in green or dark blue (bi'1-khadrä0 aw
al-läzaward) to alif al-wasl, so that if a reader pauses after the previous word, they
may know how to pronounce the glottal stop.59 This clarifies why, in the Qur'an of
227/842, alif al-wasl was noted by both a green dot and a sila sign in a thin red pen. In
sum, two conventions, using either a small red circle or a horizontal red stroke, appear
to have existed for sukün, sila, takhfif, and hurüf zawä°id. In the Kit&b al-Naqt,
al-Dän! also notes:60
The habit of the people of our land, old and new, is to add medial alif
omitted from the rasm (al-alijdt al-mutawassität al-mahdhüfät min
al-rasm) in red, as in al-'[ä]lamin, al-f[ä]siqin, al-s[ä]lihät, s[ä]m[ä]
wät, h[ä] 'ulä ', y[ä]Dädam and suchlike.
In the case of consecutive hamzät, al-Dän! observes in the Muhkam that the Medinan
convention of marking two yellow dots was followed in 'the old Qur'ans of the people
of our land'.61 However in his day, their notation had acquired greater complexity,
and al-Dän! devotes two entire chapters to its different cases, marked through various
combinations of yellow and red dots.62 In addition:63
The vocalisers of our land, old and new, have the habit of marking the
hurüf al-madd wa'l-Ιϊη al-thalätha, alif, yd', and wäw, with an
elongated stroke in red (matta bi'1-hamrä').
He is referring here to the alif wäw, and yä° with sukün preceded by a haraka of the
same sound, and to which elongation (madd) may thus be applied in recitation. In the
Kit&b al-naqt, he also associates this red elongated stroke with madd and notes that if
the letter in question is omitted in the text (mahdhüfa), it should be drawn in red with a
madd sign added above it.64 Finally in the Muhkam, al-Dän! notes that the Medinans,
and after them the Maghribis, added the ending -ü to plural endings with mlm (dammü
mimät al-jamc, e.g. calayhumu) - something he says he had observed, amongst other
features, in Maghrib! Qur'ans from the time of al-Ghäz!.65
To recapitulate, al-Dän! saw the Medinan and Maghrib! systems as closely related, but
the latter acquired distinguishing features at an early stage, and it eventually grew in
complexity. The essential features shared by both notation systems were the use of red
dots for vowels and tanwin, of yellow dots for hamza, and of other signs written in a
thin red pen for further orthographic functions (see the summary in Table 1).
3. Iraq and the Mashriq
By contrast, al-Däni portrays the conventions that prevailed in Iraq as essentially a

continuation of the Umayyad period:66
The vocalisers of the people of Iraq only use red for vowels and other
things, and for hamzät, and in this way their Qur'ans can be recognised
and distinguished from others.
Thus al-Däni would have inferred from seeing vocalisation and orthography
done solely in red in a given Qur'an that it was Iraqi, or at least as following the
conventions of 'the people of Iraq'. Another statement corroborates the same idea,
while adding an important allusion to chronology: 'Most nuqqät of the people of Iraq,
old and new, do not note sukün, tashdld, or madd in Qur'ans.'67 This implies that the
same system would have remained dominant unto his day. Vocalisers in Iraq and the
Mashriq would thus have placed less emphasis on written notation - and presumably
more on teaching and memorisation.
This strand in vocalisation, while fundamentally based on red dots, appears to have
sometimes featured additional characteristics, which are not presented as systematic.
The people of the Mashriq, he writes, note the sila of alif al-wasl as an inverted däl
placed above the alif regardless of the inflection, instead of the Maghrib! red strokes
and green or blue dots.68 He mentions that the same sign was used for zawWid
'in books', which presumably means secular books, without specifying whether this
feature was specific to the Mashriq and whether it appeared in Qur'ans there.69
Al-Däni notes a single feature of the diacritics:70
The people of the Mashriq mark β3 with one [diacritical sign] above it,
and qäf with two above. The people of the Maghrib mark fa' with one
[diacritical sign] below it, and qäf with one above.
Citing Ibn al-Munäd! (Baghdadi, 256-336/869-947), he also notes some cases in

which the hamza maftüha is followed by alif (i.e.. madd may be applied to the latter):71
The people of Basra note two dots, one for the hamza and the other for
the fatha, following the school (madhhab) of al-Khalll [ibn Ahmad]

and others; whereas the people of Kufa note a single dot on the crown
(yäßkh) of the alif, to its left.
The Kufan usage seems to have been more widespread since al-Dän! describes it
elsewhere as a convention of the Iraqis:72
Most vocalisers of Iraq differ from the people of Medina and others in
that they place hamza maftüha at the beginning of a word and followed
by an alif in pronunciation after this alif, as in 'ämana, 'Adam and
'Äzar.
For our purposes, this case may simply be described as hamza followed by madd. One
might infer that hamza preceded by madd was noted in the same manner, since the
remark is about the lack of distinction between these two cases. The more logical
convention, asserts al-Danl, was that of his own region: to place a (yellow) dot before
or after the alif to reflect the position of hamza.73
The picture of Iraqi/Mashriqi conventions provided by al-Dänl is far from monolithic:
while it does highlight a dominant usage, it also indicates variations initiated by

different authorities, as well as some complete departures from the norm. The number
of (now lost) books on naqt by Iraqi authors of the second/eighth to fourth/tenth

centuries quoted in the Muhkam points to the absence of an overarching system. One
disputed usage of some schools (tawäDif) in Kufa and Basra would have been to note
variants (hurüf shawädh) with green dots.74 Some even recorded the accepted reading
in green and the shädh in red, which is 'confusion and deviation' (takhllt wa-taghyir).
Mostculamä" were against this usage, asserts al-Dänl, citing the reproval of Ahmad b.
Jubayr al-Antäk! (also 'al-Küf!', d. 258/872). The citation is based on an isnäd rather
than a written source; if authentic, it would indicate that the latter practice was in
existence by the second half of the third/ninth century. Another usage that plunged
al-Dänl (and presumably others) into particular dismay was the notation of several
readings in one manuscript:75
More reproved and dreadful than this [i.e. the notation of shawädh] is
the habit of some readers to gather different readings and variants

(jam' qirä 'ät shattä wa-hurüf mukhtalifa) in the same Qur'an, and to
use for each reading or variant a colour other than black, such as red,
green, yellow, dark blue (al-läzaward), while signalling this at the

beginning of the manuscript.
Nonetheless, this does appear to have been accepted by some scholars. Thus the Kitäb
al-naqt of Ibn al-Munädl is cited as saying:76
If what is read is vocalised in two layers (calä wijhayn) or more,

lay down on a leaf not affixed to the Qur'an (ruqca ghayr mulsaqa
bi'l-mushaf) the names of the colours and of the readers, so that
whoever reads it may know about this. Let the hues be clear and bright.
While the practice must therefore have been known in Baghdad, where this writer
lived, it is not associated with a specific region in the text.
On the authority of Abü'l-0Abbas Muhammad b. Yazid al-Mubarrad (Basran, 210-85/

826-98) and Abü'l-Hasan b. Kaysän (Baghdadi, d. c. 299/912), both of whom are
cited by Ibn al-Nadim as authors of treatises on orthographic notation,77 al-Däni
mentions a different orthographic system attributed to al-Khalll b. Ahmad (Basran, d.
between 160 and 175/776 and 791). Based on miniature letters, notably alif, wäw, and
yä ' for vowels, shin for shadld, and khä: for khafif, it would have been applied
to 'books' (al-kutub), i.e. works other than the Qur'an.78 This assertion is largely
accurate, since these signs are solely attested in secular documents for the third/ninth
century onwards,79 although the case studies below will show that they had begun to
be introduced in Qur'ans by al-Dänl's lifetime.
When considered in its entirety, al-Dänl's text stands out by its consistency.
The author is conveying through his occasional remarks glimpses of a coherent
conception that is essentially devoid of internal contradictions, and clearly backed by
practical experience. One rare exception occurs at the end of a passage describing the
conventions of Medina and the Maghrib, including their yellow dots, where al-Däni
writes: 'And I saw this in the rest of Iraqi and Syrian Qur'ans (wa-ka-dhälika ra3 ay tu
dhalika fi sä°ir al-masähif at-'Irüiqiyya wa'l-Shämiyya)'.80 This sentence mns against

the assertion repeated throughout the book that red dots alone were used in Iraq. It
cannot be satisfactorily accounted for, except either as an error or as a reference to an
uncommon usage of these regions: it will therefore be left aside for our present
purposes.
From the foregoing discussion, it is possible to draw a comparative table of regional

conventions as seen by al-Dänl (table 1). The level of detail in his discussion of
al-Andalus and the Maghrib makes it possible, for several features, to distinguish
an earlier phase around the days of al-Ghäzi (first half of the third/ninth century) from
a later phase nearer al-Dänl's lifetime (fourth-fifth/tenth-eleventh century). In cases

where the difference is unspecified, we will assume, as a working hypothesis, that the
system remained unchanged between these two periods.
Other Regional Habits
As a result of al-Dänl's focus on the Maghrib, Medina, and Iraq, virtually

no information is provided about other parts of the Islamic world. He cites
Ibn Ashta (al-Isfahänl, the author of a lost Kitäb al-masähif, d. 360/971) to the
effect that, in li-yasü'ü (Q. 17:7), the people of Sancä° place the dot in front of
(iquddäm, presumably meaning, here, 'after') the (first) wäw, which is written in black
(i.e. as part of the rasm, rather than as a damma), based on an analogy with
the grammatical form li-yasücü, in which the 'ayη takes the place of the hamza.8I
He also notes:82
Ibn Ashta said: I saw in the mushaf of Ismäcil al-Qust, the imäm of the
Meccans [c. 100-70 or 190/719-87 or 806], that the damma was above
the letter and the fatha in front of the letter, in opposition to what is
usually done.
In other words, in the second/eighth century, some Meccans would have departed
from the basic convention of placing damma on the line and fatha above it.
Iraq and the Medina Al-Andaius and Al-Andalus and
Mashriq the Maghrib (early) the Maghrib (later)
Fatha, damma, Red dot Red dot Red dot Red dot
kasra
Tanwin Two red dots Two red Two red dots Two red dots
dots
Hamza Red dot Yellow dot Yellow dot Yellow dot
Two consecutive Two yellow Two yellow dots Combinations of
hamzät
hamzat dots yellow and red dots
Hamza followed One dot after and One dot
by madd above the alif (KufaJ before the
Iraq) / One dot for alif
hamza and another
for fatha (Basra)
Hurüf al-madd
Huruf Not marked Elongated stroke in Elongated stroke
red in red
Alif mahdhufa
mahdhüfa Alif or its abbreviated Alif or its
form in thin red line abbreviated form in
thin red line
Atif al-wasl / sila Small inverted dal

däl Small red Red stroke and green Red stroke and
above alif circle dot / Small red circle green or blue dot /
Small red circle
Variants / " " "

Not marked / green
shawädh
shawadh dots in some schools
Shadda Not marked / small Small dal

däl Small dal
däl Small dal
däl
shin in red
~ Small red Small red circle Small red circle

Hurüf zawa'id
Huruf zawä'id
circle
Takhfif
Takhflf Not marked / Small red Small red circle Small red circle /
Marked as a small circle Thin horizontal
khä'
kha' stroke
Sukün
Sukun Not marked Small red Stroke above the Stroke above the
circle above letter letter
the letter
Fa'
Fä' Diacritic above letter Diacritic below
letter
Qäf
Qaf Two diacritics above Diacritic above
letter letter
Table 1. Regional conventions as seen by al-Dani.
This echoes al-Dänf s mention (noted above), on the authority of Abü Hätim, of an
unspecified earlier Medinan convention that was superseded by the red dots ascribed
to Basra. These assertions suggest that different usages from the standard one known
to us may have existed in the early period.
From Theory to Praxis: The Manuscripts
When turning to the manuscript evidence, one is faced with a methodological

difficulty highlighted many years ago by Gotthelf Bergsträsser and Otto Pretzl:83
coloured dots may sometimes have been added to 'modernise' an earlier manuscript, a
possibility already raised above for Arabe 6140a. In some Qur'ans, this is brought to
mind by the contrast of execution between script and vocalisation (although one
cannot rule out the possibility that a less skilful vocaliser was sometimes associated
with a proficient calligrapher). An additive process may notionally have occurred even
in manuscripts with no apparent visual discrepancy. However, this need not be the end
of the matter.
On one level, the knowledge that a given manuscript was probably vocalised in a
certain region is in itself valuable, even if the place of production remains uncertain.
Certain combinations of script and vocalisation may eventually add up to a significant
pattern across manuscripts. What is more, a recent scientific study of seventeen

Qur'ans of the second/eighth to ninth/fifteenth centuries has shown that all but two
used the same pigments for the orthography and illumination.84 Since the sample is
small in size yet broad in scope, this result cannot be taken as conclusive, but it does
suggest that vocalisation may often have been applied as the manuscripts were
produced.
It is possible to assess this feature in individual cases through close observation

with the naked eye of overlaps between the vocalisation and illumination. If
the former is covered by the latter (and was thus applied before it), and provided
the illumination is original, then one might infer that the vocalisation is also
original (a method already exemplified in the above discussion of Arabe 330c). In
some manuscripts, the appearance of the same hues - sometimes even the same
dots - in the illumination and vocalisation can provide further evidence about
contemporaneous stages of production. The reverse case, where the vocalisation goes
over the illumination, can provide an indication of a later date, but only if the
illumination is not original: otherwise, one cannot rule out that the red dots
were added shortly after it, as part of the same process. Nor is it possible to
make a pronouncement about the numerous early Qur'ans that are devoid of
illumination.
In what follows, several of the rare Qur'ans carrying evidence of date and provenance
will be analysed; wherever possible, this will be preceded by an assessment of
the relationship between their vocalisation and illumination. While these

manuscripts represent an essential starting point when seeking to establish basic
parameters, others will also be considered, especially insofar as they can be placed
within larger series.
1. Iraq, Iran and Greater Syria
The Qur'an of Amäjür (Greater Syria, in or shortly before 262/876)
The two extant waqfiyyät of this manuscript were drawn up a month apart in 262/876
in Sür (Tyre) at the request of Amäjür, the cAbbäsid governor of Greater Syria
(r. 256-64/870-8), who ruled from Damascus.85 This implies that new volumes were
endowed as they were being completed, hence that the manuscript was produced in or
shortly before 262/876, most probably in Greater Syria. The manuscript is written in
style D.I, the classical phase of the Kufic tradition. It does not have single verse
markers, though tenth-verse marker have been included; no sura headings have
been published, which presently precludes a study of potential overlaps with the
vocalisation. At any rate, since the manuscript was endowed in Sur and rediscovered
at the turn of the twentieth century at the Great Mosque of Damascus, it is likely to
have remained in Greater Syria during the whole period of its use: the vocalisation,
whether it dates from the third/ninth or fourth/tenth century, almost certainly reflects
conventions from this region.
Extant parts of the manuscript are currently scattered between Istanbul (TIEM),
Cambridge University Library, the Ashmolean Museum in Oxford and other
collections.86 In addition to published images, the 74 pages from Cambridge Add.
1116 have recently been made available online as part of the Cambridge Digital
Library project.87 Throughout the manuscript one can observe:
• Red dots alone for vocalisation, hamza and lanwin\
• A red dot to the right of alifior hamza, and to its left for hamza followed by madd
(fig. 2, line 1, 'ätaytukum) or preceded by madd (fig. 2, line 3, jä 'akum)·,
• The absence of vocalisation for alif al-wasl, which serves to distinguish it from
initial hamza (fig. 3, line 2, unzilat al-tawrät, recited unzilati-1-tawrät);
• A diacritical dash above the letter for fäD (I was unable to observe a qäf with
diacritics, as these signs are very sparsely included);88
• The use of green for the occasional notation of variants: green dots for vowels,
green dashes for diacritics and in at least one place, a green vertical stroke for alif
mahdhüfa (see below).
The Qur'an of Amäjür, in sum, follows the pattern ascribed by al-Dänl to Iraq and the
Mashriq in almost every detail, with one nuance: alif al-wasl is indicated simply by
omitting the red dot, rather than by a small inverted dal. The sporadic addition of
;i *^4. ~*LL~—
*^gssL> u y,
«•
I I
Fig. 2. Folio from the Qur'an of Amäjür (Q. 3:81). Cambridge, Cambridge University
Library, MS. Add. 1116, f. 27r. Page dimensions c. 12.5 χ 19.5 cm. Reproduced
courtesy of the Syndics of Cambridge University Library.
Μj
a ϊ—SSL. Μ S&üL
mm
ηφηετ A -L· U
_
Fig. 3. Folio from the Qur'an of Amäjür (Q. 3:65). Cambridge, Cambridge University
Library, MS. Add. 1116, f. 5r. Page dimensions c. 12.5 χ 19.5 cm. Reproduced
courtesy of the Syndics of Cambridge University Library.
green signs for variant readings echoes his observations about the habits of some
Basran and Kufan schools; al-Dänf s citation of al-Antaki had precisely suggested that
this practise was at least as old as the second half of the third/ninth century. The
following variants, in green, can be noted in the Cambridge folios (the list may not be
exhaustive; names of readers associated with a given variant in the qirä°ät literature
are provided for reference):
• f. 5r (Cambridge Digital Library, image 11): Q. 3:65, al-'injll (red, associated with
all readers except al-Hasan) / al-'anjil (green, reading ascribed to al-Hasan
al-Basri);89
• f. 14r (image 29): Q. 3:73, 3an yu'tä (red, most readers) / än yu°tä (green,
Ibn Kathir, Mujähid);90
• f. 21v (image 44): Q. 3:78, li-tahsibüh (red, most readers) / li-tahsabüh (green,
Ibn cÄmir, cÄsim, Hamza, Abü Jacfar, al-Hasan, al-MutawwicI);91
• f. 24v (image 50): Q. 3:79, tucallimün (green, cÄsim, Hamza, al-KisäT, Ibn cÄmir,
Khalaf, al-Acmash); no red vocalisation has been included for the reading that was
considered standard (ta'lamün and tacallamün are recorded as alternatives in the
literature);92
• f. 25v (image 52): Q. 3:80, lä ya'murukum (red, Ibn Kathir, Näfic, Abü cAmr, Abü
Jacfar, al-Kisä°I, cÄsim) / lä ya'murakum (green, Ibn cÄmir, Hafs, Hamza,
Hammäd, Yahyä, °Äsim);93
• f. 27r (image 55, fig. 2): Q. 3:81, lamä (red, most readers, including Hafs can
cÄsim), limä (green, al-Hasan, Hamza, al-Acshä, Yahyä b. Wathäb, Hubayra can
Hafs can cÄsim);
• f. 27r (image 55, fig. 2): Q. 3:81, °ätaytukum (red, most readers) / 'ätaynäkum
(green, Näfic, Abü Jacfar, al-Acraj, al-Hasan); a green diacritical dash and a green
vertical stroke have been added for 0ätaynäkum;
• f. 27r (image 55, fig. 2): Q. 3:8 XJä 'akum (red with madd on the alif, most readers)
/ jl'akum (green, not recorded in the literature, where only imäla is mentioned for
this word);
• f. 28r (image 57): Q. 3:81, °ä qarartum (?) (red) / 'a 'aqrartum (?) (green); it is
difficult to determine exactly which readings are intended here; several variants are
recorded in the literature, involving the softening or elision of the second hamza
and/or the inclusion of an alif between the two hamzät;94
• f. 30r (image 61): Q. 3:83, yabghün (red diacritical dashes, Abü °Amr, Hafs,
cÄsim, cAbbäs, Yacqüb, Sahl, al-YazIdl, al-Hasan) / tabghün (green diacritical
dashes, all other readers);
• f. 31r (image 63): Q. 3:83, wa-karha" (red, most readers) / wa-kurhan (green,
al-A°mash);
• f. 31r (image 63): Q. 3:83, wa-'ilayhu (red, not recorded in the literature)/ wa
'ilayhi (green, standard reading);
• f. 31r (image 63): Q. 3:83, turjacün (red dots and dashes, most readers) / yarjicun
(green dots and dashes, Yacqüb);95
• f. 35v (image 72): Q. 3:86, jä'ahum (red with madd on the a I if, most readers) /
jf'ahum (green, not mentioned in the literature, where only imäla is recorded for
this word);96
• f. 36v (image 74): Q. 3:87, calayhim (green, standard reading) / no red dot;97
Most of these variants thus feature in the qirä 'ät literature, which started developing
in the third/ninth century, if not earlier, and associated different variants with readers
of the second/eighth to early third/ninth century and their pupils.98 In the Qur'an of
Amäjür, or at least the Cambridge fragment, the readings noted in either red or green
do not fit consistently into the categories established in the qirä'ät literature: they do
not lead back to the name of the same one or even two readers. Indeed a few of
the variants observed here (ff. 27r, 31r, 35v) do not appear at all in the literature,
yet they must have been read in this period. This could reflect attitudes exemplified
by Ibn Qutayba (213-76/828-89) and, several decades later, Ibn Miqsam (d. 354/
965), who endorsed any reading as long as it was based on the cUthmänic rasm
and grammatically sound. Ibn Miqsam and another prominent Qur'an reader, Ibn
Shannabüdh, were tried and forced to recant in 322/934 and 323/935 respectively, at
the instigation of Ibn Mujähid." Until that period however, a broad spectrum of
readings and approaches had existed, just as a multifaceted movement towards the
systematisation of variants was gaining traction.100
The Qur'an of cAbd al-Muncim (possibly Greater Syria, before 298/911)
This Qur'an written in style D.I was made a waqf at the Great Mosque of Damascus in
Dhü'l-Qacda 298/July 911 by cAbd al-Muncim b. Ahmad (fig. 4).101 Its production
must have occurred in the preceding months, years, or decades. The waqfiyyät make
Greater Syria a plausible region of origin, but others remain equally conceivable: the
manuscript therefore cannot stand as a primary piece of evidence about provenance,
but will be analysed for supplementary information. The illumination clearly runs over
the red dot on CBL Is. 1421, f. 2a, which suggests that the latter are original, although
further observations along the same lines would be desirable.102 In the manuscript,
one can notice:
• Red dots for vowels, hamza, and tanwitr,
• A red dot to the right of alif for hamza, and to its left for hamza preceded
or followed by madd (Jämanü, fig. 4, line 1; 'älü, f. 2a); this notation of hamza
is not solely used for initial alif, but can also occur at the end of a word
(shuhadäf. lb);
Fig. 4. Folio from the Qur'an of cAbd al-Muncim (Q. 22:77-8). Dublin, Chester
Beatty Library, Is. 1421, f. la. Page dimensions 21 χ 32 cm. © The Trustees of the
Chester Beatty Library, Dublin
• A red dot to the left of alif can also indicate madd without hamza (fig. 4, line 8,
°Ibrählm)\
• Alif al-wasl is not vocalised, which serves to distinguish it from initial hamza
(fig. 4, line 2, wa-"asjudü wa-°acbudüh, recited wa'sjudü wa'cbudüh)·,
• The diacritical sign for fäD is above the letter; qäf is marked by two diacritics
above the letter (fig. 4, line 5,fi, haqqf, these signs appear to be original, although
their hue fluctuates differently from that of the adjoining letter strokes, which
suggests their insertion may have been a discrete task completed after the
calligraphy itself.
The manuscript is thus entirely consistent with al-Dänl's observations about Iraq and
the Mashriq, with the same minor nuances as in the Qur'an of Amäjür. Variants do not
appear in the three leaves that make up Is. 1421, though this pattern remains to be
confirmed against a larger sample of leaves from this manuscript.
The Khayqänl Qur'an
This Qur'an written in an early version of the New Style (CBL Is. 1417) contains
a text in its closing page stating in Persian and Arabic that one Abü'1-Qäsim
al-Khayqänl corrected the text in 292/905; two further texts also in his name appear
at the end of other volumes from the same manuscript.103 These statements have
been used for decades as a terminus ante quem for its production and as a basis to
suggest an origin in the eastern Islamic world.104 However, close observation of the
original casts some doubt on their authenticity: their ink is different from that of the
Qur'anic text, and appears closer to that of its re-inkings; they contain obvious Arabic
grammatical errors, which are all the more suspicious since they are meant to emanate
from the corrector of the manuscript; and some features of their script, notably the
curved ending of rä°, only find parallels in later periods. These texts may have been
added to empty pages at the end of their respective volumes long after the manuscript
was written. Until this matter has been settled, it seems preferable not to rely on this
manuscript for indications of chronology or provenance. Its Qur'anic text, which can
still be ascribed to the fourth/tenth century, displays the above uses of red dots,
including for hamza and hamza with madd, as well as additional orthographic signs.
The Qur'an of Ibn Shädhän (361/972)
This manuscript was copied by CA1I b. Shädhän al-Bayyic in 361/972. It is a fine

example of the New Style, an angular aesthetic of the Arabic script that gradually
superseded Küfic in the fourth/tenth century, and remained in use well into the
sixth/twelfth century. Its largest preserved section is CBL Is. 1434 (170 folios); about
16 additional folios, including the colophon, are at Istanbul University Library
(Ms. A6758).105 The same Ibn Shädhän was also the copyist of an extant copy
of al-Slräfi's Akhbär al-nahwiyyin al-Basriyyin completed in Jumädä I 376/
August-September 987 (Istanbul, Süleymaniyye Library, Shähid CA1T Pasha No.
1842). In its colophon, his nisba is given as al-RäzI, which implies that he or his
family was from Rayy, in Iran.106
CA1I was identified by Fritz Krenkow with a traditionist of the same name listed by Ibn
Hajar al-°AsqalänI (773-852/1372-1449), but mistakenly since the latter had Abü
Badr al-Sakünl (d. 204/820) as his teacher, and even his pupils had passed away by
the early fourth/tenth century.107 On the other hand, CA1I b. Shädhän may have been
related to Abü Bakr b. Shädhän al-Räzi (d. 376/987), who lived in NIshäpür and
frequented Süfi circles.108 But the names provided for CA1I in the colophons do not
provide enough genealogical information to substantiate this possibility. Krenkow
noted Arabic grammatical mistakes in cAlI's copy of the Akhbär al-nahwiyym
al-Basriyyin, which suggests that this was not his native language.109 Thus an origin
in the eastern Islamic world appears as likely, even though the evidence is not
sufficient to make a firm attribution.
The opening page of CBL Is. 1434 (f. la; fig. 5) carries a small sheet attached by a
modem paper frame to the recto of the opening illumination (f. lb). Although its
borders are concealed by this frame, it appears to be a cropped fly leaf, which may
previously have been pasted to this page or to the binding. It reads:
Bi'smi'lläh. H[ä]dhäΊ-mushaf manqüt bi-qirä'at cAbd Alläh b. Kathir

calä mä rawähu Ibn Abi Bazza canhu wa-bi-qirä3 at Abi cAmr b.
al-cAlä' calä mä rawähu al-Yazldl. Fa-mä käna fihi mimmä ikhtalafä
fihi min madd aw hamz aw ziyädat harf aw isqätihi aw tashdld aw
ikhtiläf fi'l-raf aw al-nasb aw al-jarr aw ghayr dhälik, mucallaman
calayhi {bi'fl-sufra, fa-li-Ibn Kathir khässaf". Wa-mä käna
mucallaman calayhi bi'l-fustuql fa-li-Ibn cAmr khässaf". Wa-mä
käna mimmä ittafaqä calayhi manqüt bi'l-humra. Wa-mä käna min
madd aw hamz aw tashdld {mucallaman}no calayhi bi'l-zinjär.
Abu cAmr idhä khatama al-süra wa-akhadha ft qirä'at ukhrä lä

yaqra3 bi'smi'lläh al-rahmän al-rahlm ka-qawlahu <al-dällln; alif
läm mim dhälika> <fa-unsurnä calä'l-qawm al-käfirln; alif läm mim
Alläh> <wa'ttaqü Alläh lacallakum tuflihün; yä ayyuhä al-näs> ,
kadhälika ilä äkhir al-Qur°än. Wa-Ibn Kathir yafsil bayn al-suwar
bi-bi'smi'lläh al-rahmän al-rahlm.
In the name of God. This Qur'an is vocalised according to the reading

of cAbd Alläh b. Kathir in the transmission of Ibn Ab! Bazza; and
according to the reading of Abü cAmr b. al-cAlä3 in the transmission

of al-YazIdi. As to what they differ about in terms of madd, hamz,
the addition of a letter or its omission, tashdld, rap, nasb, and jarr, or
anything else, what is indicated {in} yellow is specific to Ibn Kathir,

and what is indicated in pistachio green (al-fustuql) is specific to Abü
cAmr. What they both agree on is vocalised in red, with madd, hamza,
tashdld {indicated} in verdigris (al-zinjär).lu
Abü cAmr, when he reached the end of a sura and went on to

read another, would not recite bi'smi'lläh al-rahmän al-rahlm, as in
al-dällln - alif läm mim dhälika [Q. 1:1-2:1]; fa-ansurnä calä'l-qawm
al-käfirln - alif läm mim Alläh [Q. 2:286-3:1-2]; wa'ttaqü Alläh
lacallakum tuflihün - yä ayyuhä al-näs [Q. 3:200-4:1]; and so on until
the end of the Qur'an. Ibn Kathir used to mark the division between
suras with bi'smi'lläh al-rahmän al-rahlm.
To summarise, this note suggests that the general vocalisation should appear in
red and in verdigris (i.e. light green) for madd, hamza, and tashdld. Readings specific
to Ibn Kathir (c. 45-120/645-738), in the transmission of his pupil Ibn Abl Bazza
(sometimes given as BazzI, d. c. 124/742), should appear in yellow, and readings
specific to Abü cAmr (d. 154/770), in the transmission of al-Yazidi (d. 202/817), in a
distinct 'pistachio green'. There follows the interesting mention that the basmala was
not recited when reading consecutive suras in the reading of Abü cAmr.
Journal of Qur'anic Studies
vi··:.. ^
fi
■■
■
. *6^Jp*jJ)\xa
cm* L -iy'-Li^ 1^,1
4ilL· 1
**&'<■ .ϊ·. ^•SJgsty.lCUJt I
ι xSk^y^-,,
η ^ jliîjs,
$( « •• f -- - —
- nn f
I , . .. ·- - '•^r-^^'^N0-.>*l^»JU, II
I
;. , '-«
I·
V|
mn
: IB!Ui
.
Fig. 5. Opening sheet attached to the Qur'an of Ihn Shädhän. Dublin, Chester Beatty
Library, Is. 1434, f. la. Total page dimensions 26 χ 17.8 cm © The Trustees of the
This text about vocalisation was written in black ink in a hand that occurs again in
the margins of the same volume, this time in green and occasionally yellow, to repeat
phrases from the adjoining Qur'anic calligraphy (e.g. ff. 34b, 43a, 96b); as well as in
the last page of cAH's copy of the Akhbär al-nahwlyyin al-Basrlyym, in a note just
above the monumental calligraphy stating that the text was 'checked, corrected and
/♦
*
jii>\ -·. \
^41±4•»».! Ua«c
1 «· ».1 ij,ίjI*!***
kiic-
*?.
>\ "~^'UaJc.
4ALa1I
/ajAlfSÄ $4
· / /-- «'^ i<|
-7
✓
si
•>2»Uj>» ji-ai
Wy
>Ss
/
*>Uu>4^ a*
UJ3
^ ..., _„^
_ âft *T
. Jrt
4. ^ 3 ^β5 -*αΑ ·.
Fig. 6. Closing page of the Ibn Shädhän's copy of the Akhbär al-nahwlyym
al-Basriyyln. Istanbul, Siileymaniyye Library, Shähid CA1I Pasha No. 1842.
Dimensions unknown.
collated with the help of God' (fig. 6).112 This increases the likelihood that both
interventions happened at the time of copy, and that cAli b. Shädhän teamed with the
same corrector for both manuscripts. The alternative, namely that the same person
came to own two manuscripts by this calligrapher at a later point in time, then decided
to vocalise one and correct and collate the other, seems improbable. Indeed, the
corrector's bookhand bears a close affinity with other specimens dated to the late
fourth/tenth century.113 The fact that the leaf seems to have been originally detached
from the manuscript echoes Ibn al-Munädl's prescription, cited above, that indications
about qirä'ät should be laid down 'on a leaf not affixed to the Qur'an'.
In the Qur'an of Ibn Shädhän, a further layer of notation in yet another hand can be
discerned: tiny words giving indications about grammar, such as khabar, masdar,
shart, and istifliäm, added above relevant words from the main text in minute yellow
script; the hand appears to be the same as in some marginal juzD markers written in
red. Judging by the shape of its initial jlm and its ra \ this hand may be later than those
of CA1I and the anonymous vocaliser.
Returning to the Qur'anic vocalisation, one can observe that:
• Red dots mark vocalisation, tanwin, and most cases of hamza;
• A red dot is placed to the right of initial alif for hamza;
• Alif al-wasl is not vocalised, which serves to distinguish it from initial hamza
(fig. 7, lines 1-2, humu al-khäsirün, recited humu'l-khäsirün);
• A horizontal green stroke above medial or final alif marks madd (fig. 7, line 1,
3uld'ika; 1. 6, al-samäJ; 1. 8, al-malä'ika; 1. 10, al-dimä');
• A vertical green stroke marks madd for initial alif, it is sometimes combined with a
red dot to indicate the position of hamza (fig. 8, line 1, li-Ädam);
• The same vertical green stroke above letters other than alif denotes alif mahdhüfa
(fig. 7, line 6, sam[ä]wäty,
• Hamza is marked, for cases other than initial alif by the modern sign based on
cayn, in green (fig. 7, line 1, 'ulä'ika; 1. 7, shay0; 1. 8, al-malä'ika);
• Shadda is noted by a green shin;
• A small blue circle at the end of a word appears to indicate cases of waqf(the sign
recalls a modern sukün; in these cases the letter should indeed be pronounced with
sukün)·,
• A small green inflection indicates sukün;
• The word huwa ('He') regularly carries a yellow dot and an inflection in light green
(fig. 7, beginning of line 7); the former indicates the standard reading huwa, and
the latter the reading hwa (with a sukün on the ha') associated, among others, with
Abu cAmr and al-Yazid!;114
• Two oblique strokes placed respectively above and below the letter, one in yellow
and the other in light green correspond to cases of imäla, which was widely applied
by Abü cAmr (fig. 8, line 2, al-käfirin, recited al-kefirin);] 15 the green stroke
presumably indicates his reading and the yellow stroke that of Ibn Kathir, without
imäla;
• A vertical yellow stroke often appears below the final häJ of words ending with
y-h, e.g. fihi, nabihi, yadayhi, 0 Hay hi (fig. 7, line 4); these reflect a reading
specific to Ibn Kathlr in which a yä' is added at the end of the word, e.g. fihi;116
• In some phrases, light green and yellow dots indicate variant readings, e.g. in
fa-talaqqä 'Ädamu min rabbihi kalimät'" (Q. 2:37; fig. 8, line 10), the standard
reading is indicated in green and the reading of Ihn Kathir {fa-talaqqä 'Ädama min
rabbihi kalimäf) in yellow;117
• A yellow dot marks cases of mim säkina (fig. 7, lines 3-4, yumitukum thumma
yuhyikum); these may have been instances of waqf emphasised by Ibn Kathlr,
though I was unable to confirm this;
• In al-sufahä0 °a-lä (Q. 2:13; f. 4a, line 7), there is an oblique green stroke above the
initial alif of 'a-lä\ its function is probably related to the two consecutive hamzät in
this phrase, since several readings are recorded for them, with all but one having
transmissions on the authority of Abü cAmr;118
• Fa' is marked by one diacritic above the letter, and qäf by two diacritics above the
letter.
The vocalisation, in sum, corresponds to the explanations provided in the opening

sheet, notably with regard to the readings of Abü cAmr and Ibn Kathlr, as well as to
the two green hues used for variants and general vocalisation respectively. The blue
circles are the only elements omitted in this text, either because they were deemed
unimportant or because they are a later addition. At a basic level, this notation also
matches al-Däni's assertions about the Mashriq, with its red dots for the vowels,
hamza and tanwln, and its diacritics; but a range of additional signs have been
introduced for madd, alif mahdhüfa, shadda, sukün, certain cases of hamza, and
variants.
The Isfahan Qur'an (Isfahan, 383/993)
This manuscript was completed in Isfahan in Ramadan 383/October-November 993

by Muhammad b. Ahmad b. Yasln, also in the New Style.119 Having been unable to
study it in person, I have based my observations on published reproductions of the
following fragments: TIEM 453—456; Khalili KFQ50; Met, Rogers Fund, 40.164.5
(unnumbered folios, labelled Metl, Met2 and Met3 below); Freer F1937.34;
Minneapolis Institute of Arts, 51.37.7.120 The sura markers consist of rectangular
illumination bands in gold with red and green. They are probably original: a dedicated
space has been left for them, most evidently in several folios where the last words of
the preceding sura were centered on the line so that the illumination rectangle could be
articulated around these words. Furthermore, the illuminated titles feature a distinctive
medial cayn/ghayn in the form of a knot with two loops that also occurs in the
calligraphy of the main text (as in fig. 9, maghlübun)}2X
$*>==>
4lk
M)
-° -" 111y ill
^^T^ts^^c===w^ It^\aâ—
i=^43JU4^J
l*_—
9fl
i=&k aA& jjLili <4
_
sjjg1 '3ξ
... ■
*' "'Ί»'
Fig. 7. Folio from the Qur'an of Ibn Shädhän (Q. 2:27-30). Dublin, Chester Beatty
Library, Is. 1434, f. 6b. Page dimensions 26 χ 17.8 cm. © The Trustees of the
.4^^1^4
JiUi*
•^4^^ £it
.1" *
«» jj^Vj o-. fgm

*ά
φφ
4 it
■#ί·^~ 1^==4 4
if JSllÄ)
I
UP
m
ife&aMi
Fig. 8. Folio from the Qur'an of Ihn Shädhän (Q. 2:34-37). Dublin, Chester Beatty
Library, Is. 1434, f. 7b. Page dimensions 26 χ 17.8 cm. © The Trustees of the
/4
Fig. 9. Detail of folio from the Isfahan Qur'an (Q. 54:10-11). New York,
Metropolitan Museum of Art, Rogers Fund, 40.164.5.
The verse separators were executed before these sura markers, since the latter cover
the former in cases of overlap.122 At least one red dot is partially covered with gold
specks from a verse separator (fig. 9); in the Freer folio, the blue frame of the sura
illumination again runs over one red dot and is interrupted to avoid covering another.
This suggests that the red layer of vocalisation preceded the illumination, and is
therefore original. In places, the blue vocalisation is overrun by the red vocalisation
and by specks of gold from verse markers, which suggests that it is also original.123
I was unable, from the available sample, to observe comparable overlaps for the green
layer of notation. However the signs that appear in green, red and/or blue have exactly
the same shapes, whilst each colour fulfils a distinct function (see below).124 Thus all
three layers of notation are probably contemporaneous, hence original - a conclusion
also supported by the appearance of the same colours in the verse markers, marginal
ornaments and sura markers. The vocalisation can be outlined as follows:
• Red dots are used for vowels, tanwTn and most hamzät;
• A red dot is placed to the upper right of alif for hamza with fat ha (e.g. fig. 10,
line 2, 3absäruhum; same page, line 3, ka-'anna hum; fig. 11, line 3, 'ahwä'ahum,
first hamza·, same line, °amfn); below the alif for hamza with kasra (fig. 10,
lines 2, 3, °ila); to its left for hamza followed by madd (fig. 11,1. 3, 'äyät'"; fig. 12,
end of line, °äla Lüt), or preceded by madd (fig. 11, line 3, 3ahwä'ahum, second
hamza; same image, line 4, jä3ahum, min al-'anhä');
• Alif al-wasl is not vocalised, which serves to distinguish it from initial hamza;
• Sila is marked by a green horizontal stroke that cuts across the initial alif and läm
(e.g. fig. 10, line 2, min al-'ajdäth; fig. 11, line 4, min al-3anbä3);
• Hamza is sometimes marked, for cases other than initial alif, with the modern
sign based on the phonetically related letter cayn, in blue (fig. 10, line 2,
shay3);
Fig. 10. Folio from the Isfahan Qur'an (Q. 54:5-9). New York, Metropolitan
Museum of Art, Rogers Fund, 40.164.5. Page dimensions 24 χ 35.1 cm.
<*- y ûy .: I ] // s, },i . , '. ; ft 'j
Fig. 11. Folio from the Isfahan Qur'an (Q. 54:1-5). London and Geneva, Nasser D.
Khalili Collection, KFQ90 (verso). Page dimensions 23.9 χ 33.8 cm. Reproduced
courtesy of the Khalili Family Trust.
• Shadda is marked by the modern sign derived from shin, in blue, alongside the red
dot that indicates the vowel;
• A green sign in the shape of a shin marks idghäm; one example ([al-hadithi]
ta'jahün, Q. 53:59, Khalili KFQ90 recto) involves idghäm between thä ' and tä°,
which is not standard but was applied to this phrase in the readings of Abü 'Arm
and Ya'qöb;125
• Green is otherwise used to indicate variants (see below);
• A small khä0 in blue denotes khafif (fig. 9, fa-fatahnä; same page, line 4,
wa-hamalnähu);126
• A small inflection in red indicates sukün (fig. 11, line 4, laqad);
• A small blue circle denotes waqf(fig. 9,fa-antasir, fig. 10, line 1, canhunr, fig. 11,
line 3, °ahwäJahum); this includes instances in which sukün should replace the final
vowel or tanwin of a word if the reader marks a pause (fig. 10, line 4, 'asir1—;
fig. 11, line 3, mustaqirr1— );
• Fä° is marked by one diacritic above the letter, and qäf by two diacritics above the
letter.
Variants:
• At the beginning of Q. 54:7 (fig. 10, line 2), red and green signs are used to
note two different readings of the same word: khushshacan (red vowels, sila and
shadda; reading of Näfic, cÄsim, Ibn cÄmir, Ibn Kathrr, Ibn Muhaysin and others)
and khäshican (green vowels, medial alif, and khä0 for khafif above the shin; Abü
cAmr, Hamza, al-Kisäri, Ya'qüb, Khalaf and others);127 note that in order to apply
this convention, the shadda in this word has been noted in red, as opposed to blue
in the rest of the manuscript;
• The word al-Qur°än is written with a green dot and red däl above the rä°; the red
däl corresponds to the standard reading with hamza säkina, and the green dot
the reading al-Qurän (Ibn Kathir, Ibn Muhaysin), where the hamza is replaced
by alif;128
• The word al-mu°minin is written with a red dot and a green däl above the wäw; the
former reflects the standard reading with hamza, the latter the reading al-müminin,
with a wäw säkina (Hamza, Abü Ja'far, Abü °Amr, al-Azraq, Warsh, and
al-Isfahäm);129
• The hi2° of the word huwa carries both a red dot for the damma of the standard
reading huwa and a green sukün inflection for the variant reading hwa (only one
relevant occurrence of this word could be observed);130
• The phrase °inna'lläha huwa (Q. 51:58) has a green sukün inflection on the
hä° of the first word and a green shadda on the hä° of the second word;
ϋΐα.,ν,τΐ Ay
Fig. 12. Detail of folio from the Isfahan Qur'an (Q. 54:34). New York, Metropolitan
Museum of Art, Rogers Fund, 40.164.5.
this corresponds to the reading °inna'llah-huwa (with idghäm; Abu cAmr,

Yacqub);131
• Likewise in 'illä °äla Lüt (Q. 54:34; fig. 12), there is a red dot on the first läm to
indicate the standard reading with fatha; as well as a green sukün inflection on the
same letter, which together with the green shadda on the second läm indicates
idghäm ('illä :äl-lüt; reading of Abü cAmr, Yacqub).132
All these variants, with the exception of al-Qurän, reflect the reading of Abü cAmr,
which may have been consistently used in this manuscript, although the study of a
larger portion of the text would be necessary to draw a firm conclusion. Once again,
the vocalisation follows the pattern attributed by al-Dänl to Iraq and the Mashriq in its
most basic features: the red hamzät, diacritics, madd (with the nuances noted above),
and the khä' for takhfif (which al-Dänl associated with al-Khalll b. Ahmad and
Slbawayhi, rather than Iraq in general). Other layers of notation have been added for
shadda, sukün, idghäm, and silät, together with occasional variant readings in green.
Some of these features reflect the same conventions as in the Qur'an of Ibn Shädhän,
although they are noted in different colours (shadda, waqf, modem hamza, sukün);
others are distinct (sila, idghäm, takhfif).
Summary
The manuscript evidence at our disposal, in sum, suggests that the notation system
probably in place by the late first/early eighth century in Greater Syria (Arabe 330c)
was still used in the second half of the third/late ninth to early tenth century in that
region (the Qur'an of Amäjür, and possibly of cAbd al-Muncim); its amplified version
is attested in the eastern Islamic world for the fourth/tenth century (the Isfahan Qur'an,
and possibly the Qur'an of Ibn Shädhän). Although no early Qur'ans with explicit ties
to Iraq survive, this geographical spread, together with the assertions of al-Dänl,
makes it reasonable to assume that the same system was also used there. Green dots
were sometimes added for variants, as in the Qur'an of Amäjür and the Isfahan
Qur'an. The growing complexity apparent in notation of the fourth/tenth century
suggests that al-Dänl is not presenting the most up-to-date information on the
Mashriq, presumably because his acquaintance with that region was less thorough
than with the Maghrib.
(Note: Part II of this article will appear in the next issue of Journal of Qur'anic
Studies, 17:2)
NOTES
1 This research was made possible by the generous support of the Leverhulme Trust. I also
wish to thank Elaine Wright (Chester Beatty Library), Annie Vernay-Nouri (Bibliotheque
Nationale de France) and Nahla Nassar (Nasser D. Khalili Collection) for greatly facilitating my
access to their respective collections; and Shaykh Ziad Taktak (Taalbaya, Lebanon), for the
time he offered to try and solve some issues about Qur'anic readings. Any shortcomings
naturally remain my own.
2 The names 'Hijäzi' and 'Kufic' are both misnomers insofar as they suggest a link with the
Hijäz and Kufa respectively, whereas the scripts they designate had a much broader
geographical spread. They will nevertheless be used here because they are widely accepted
terms, and for want of more suitable alternatives.
3 al-Dänl, al-Muhkam, and al-Däni, al-Muqnicβ rasm masähif al-amsär, mac Kitäb al-naqt
(with the relevant section, at pp. 132-53, henceforth referred to as Kitäb al-naqt). The above
edition of the Muhkam was reissued in Damascus in 1986; a further edition was published by
Dar al-Kutub al-cllmlyya (Beirut) in 2004 (not consulted). The main other source on
vocalisation is a section of Ihn Abi Däwüd's Kitäb al-masähif (Ihn Abi Däwüd, Materials,
pp. 141-50). Regional habits are not discussed in this work.
4 E.g. Deroche, 'New Evidence About Umayyad Book Hands'; Deroche, 'Colonnes, vases
et rinceaux'; George, 'Calligraphy, Colour and Light in the Blue Qur'an', esp. pp. 81-9;
George, The Rise of Islamic Calligraphy, Deroche, Qur'ans of the Umayyads.
5 Surty, A Course in the Science of Reciting the Qur'an. Cf. also Dutton, 'Red Dots, Green
Dots ... Part I', pp. 121-2. For a detailed discussion, see al-Suyütl, al-Itqän fl culüm al-Qur'än
(many editions), esp. ch. 28-33.
6 I thank Frederik Leemhuis for this reference. An earlier work widely used in previous studies
is cUmar and Makram, Mucjam al-qirä'ät.
7 The fullest version appears in Yäqüt al-Rümi, Irshäd, vol. 5, pp. 36-7. It is also given in
more or less shortened form by different authors, e.g. Ihn al-Jazari, Ghäyat al-nihäya, vol. 1,
p. 447 (no. 2091); Ihn Bashkuwäl, al-Sila, vol. 2, p. 386 (no. 876). About Sulaymän b. Najäh,
see Ibn al-Jazari, Ghäyat al-nihäya, vol. 1, p. 287 (no. 1392).
8 Yäqüt al-Rümi, Irshäd, vol. 5, p. 37. In this account, al-Däni provides the names of two of his
teachers at Mecca: Abü'l-cAbbäs Ahmad al-Bukhäri and Abü'l-Hasan b. Firäs. Ibn Bashkuwäl
lists additional names of his teachers in Cairo and Qayräwän, but without citing a source; Ibn
Bashkuwäl, al-Sila, vol. 2, p. 385 (no. 876). On his biography, see also the editor's introduction
to al-Däni, al-Muhkam, pp. 5-10; Ben Cheneb, art. 'al-Däni'; Fesharaki and Saeedpoor, art.
'Abü cAmr al-Däni'.
9 Ibn Bashkuwäl, al-Sila, vol. 2, p. 386 (no. 876); Ibn al-Jazari, Ghäyat al-nihäya, vol. 1,
p. 448 (no. 2091); al-Däni, al-Muhkam, p. 11 (editor's introduction).
10 al-Däni, Kitäb al-naqt, p. 152.
11 For example, his work was widely drawn upon by Ibn al-Jazari (d. 833/1429), a prominent
later authority on the subject who lived between Syria, Anatolia, and Iran, and who in
his biographical notice called him the 'teacher of teachers and master of masters' fustädh
al-ustädhm wa-shaykh al-mashäyikh') (Ibn al-Jazari, Ghäyat al-nihäya, vol. 1, p. 447).
12 Dutton, 'Red Dots, Green Dots ... Part 1'; Dutton, 'Red Dots, Green Dots ... Part II'.
13 The potential of vocalisation as an indicator of regional origins has been highlighted by

Blair, Islamic Calligraphy, pp. 120-4.
14 al-Dänl, al-Muhkam, pp. 4—7. Cf. also al-Däni, Kitäh al-naqt, pp. 132-3; Bergsträsser and
Pretzl, Geschichte des Qoräns, pp. 261-2; Abbott, The Rise of the North Arabic Script, p. 39;
Hamdan, 'The Second Masahif Project', pp. 808-9 (where some confusion appears to arise
between the vocalisation and diacritics). The earliest sources cited in these studies, both by
Abbott, are Ibn Durayd (d. c. 321/933) and Ibn Qutayba (d. 276/889); but their respective
passages about Yahyä b. Yacmur and al-Hasan al-Basri are general biographical notices,
without a mention of vocalisation (Ibn Durayd, Kitäb al-ishtiqäq, p. 163; Ibn Qutayba, Kitäb
al-macärif, p. 12).
15 Talmon, 'Who Was the First Arab Grammarian?', pp. 128—45; Talmon, 'Schacht's Theory',
pp. 10-22.
16 Deroche, La transmission, pp. 152-9; George, The Rise of Islamic Calligraphy, ch. 1;
Sadeghi and Bergmann, 'The Codex of a Companion'; George, 'Le palimpseste'; Sadeghi and
Goudarzi, 'San'ä' I and the Origins of the QurTin'; Deroche, Qur'ans of the Umayyads, ch. 1-2.
17 For BNF Arabe 6140a, see Deroche, Les manuscrits du Coran, 61 (Cat. 6); for Add. 1125,
http://cudl.lib.cam.ac.uk/view/MS-ADD-01125 (accessed 29/09/2012). Two mutually related
fragments in the style classified by Deroche as Hijäzi IV (BNF Arabe 334c; Khalili KFQ59,
KFQ61) are also vocalised with red dots, but these manuscripts probably date to the second/
eighth century; see Deroche, The Abbasid Tradition, pp. 32-3 (no. 3).
18 This process is best documented in the so-called 'Codex Parisino-Petropolitanus'; see

Deroche, La transmission ecrite du Coran, pp. 45-50.
19 al-Däni, al-Muhkam, p. 87. Translation after Dutton, 'Red Dots, Green Dots ... Part I',
pp. 119-20.
20 George, The Rise of Islamic Calligraphy, pp. 74—89; Deroche, Qur'ans of the Umayyads,
ch. 4.
21 Deroche, 'Colonnes, vases et rinceaux'; George, The Rise of Islamic Calligraphy, pp. 75-8.
Other Qur'ans featuring the same script type, and recently labelled O.I by Deroche, are
discussed in Deroche, Qur'ans of the Umayyads, ch. 3, where the name 'Umayyad codex of
Fustat' was first coined. Cf. also Washington D.C., Library of Congress, Ms. AL-17 (Digital ID
ascs 295), under http://hdl.loc.gov/loc.amed/ascs.295 (accessed 17/07/2012); and the fragments
from the National Museum in Damascus published by Radiciotti and D'Ottone, Ί frammenti
della Qubbat al-khazna di Damasco', figs 1-2.
22 Deroche, 'Colonnes, vases et rinceaux', pp. 240-2; George, 'Calligraphy, Colour and
Light', pp. 92-3. The whole of Arabe 330 can be consulted on Gallica, gallica.bnf.fr (accessed
29/09/2012).
23 See also f. 13r, line 7 (available on Gallica).
24 In his catalogue of this collection, Deroche remarked that these strokes were 'added',
without further elaboration; Deroche, Les manuscrits du Coran, p. 144 (Cat. 268).
25 Deroche, 'Colonnes, vases et rinceaux', p. 238. In Arabe 330c, the later addition of
diacritics, of tails for final mlm and the re-inking of faded parts of the text are particularly clear
in places, e.g. f. 14v. Cf. Deroche, Les manuscrits du Coran, p. 145 (Cat. 268).
26 Deroche, 'Colonnes, vases et rinceaux', p. 238, n. 36; Deroche, Islamic Codicology,

pp. 220-1.
27 cAbd al-Razzäq, al-Musannaf, vol. 4, p. 322 (no. 7941), pp. 323—4 (no. 7948). The
expression al-naqt bi'l-cArabiyya used in these akhbär specifically refers to vocalisation.
28 Deroche, 'Colonnes, vases etrinceaux', p. 260; George, The Rise of Islamic Calligraphy, p. 78.
29 al-Dänl, al-Muhkam, p. 7.
30 al-Samhüdl, Wafä' al-wafä, vol. 2, p. 457; George, The Rise of Islamic Calligraphy,
pp. 71^1, p. 86, p. 91; George, 'Calligraphy, Colour and Light', p. 98, p. 100.
31 Hamdan, 'The Second Masahif Project', pp. 796-809; Bergsträsser and Pretzl, Geschichte,
p. 262; Abbott, The Rise of the North Arabic Script, p. 39.
32 Omar Hamdan ('The Second Masahif Project') has argued for the historicity of this
account by adducing the idea of a committee formed by al-Hajjäj. Ibn Abi Däwüd does cite a
tradition according to which al-Hajjäj gathered huff a ζ and qurrä' in order to have them count
the number of letters in the Qur'an and create divisions of the text (Ibn Abi Däwüd, Materials,
pp. 119-20). But the extension of this idea to the vocalisation remains speculative, even if one
chose to take the sources at face value. For a source-critical approach to these texts and the
process of 'growing backward' in historical writing, see Talmon, 'Schacht's Theory', pp. 40-6.
33 Hamdan, 'The Second Masahif Project', p. 800, pp. 807-8. On diacritics in earlier
documents and Qur'ans, see George, The Rise of Islamic Calligraphy, p. 29, p. 51; Deroche,
La transmission ecrite du Coran, pp. 43-5, p. 120.
34 al-Jumahl, Tabaqät fuhül al-shu'ara', vol. 1, p. 12; Talmon, 'Who Was the First Arab
Grammarian?', p. 131.
35 Ibn cAtiyya, al-Muharrar al-wajlz, vol. 1, p. 50; al-ZarkashT, al-Burhän, vol. 1, p. 251.
Charles Pellat mentions a manuscript of al-Jähiz's work, British Library Or. 1129, which I was
unable to consult (Pellat, The Life and Works of Jähiz, p. 22, n. 39).
36 On awä 'il, cf. Noth, The Early Arabic Historical Tradition, pp. 104-8; McCants, Founding
Gods, pp. 70-83.
37 See Talmon, 'Schacht's Theory', where the phrase is coined on p. 45.
38 al-Dänl, al-Muhkam, p. 19; Dutton, 'Red Dots, Green Dots ... Part 1', p. 117. Cf. also
al-Dänl, Kitäb al-naqt, p. 134; al-Dänl, al-Muhkam, p. 8, p. 148. In one of the above instances
(al-Dänl, al-Muhkam, p. 8), the term nabrät is used as a synonym for hamza; on this term,
cf. Lane, Arabic-English Lexicon, vol. 8, p. 2,757.
39 al-Dänl, al-Muhkam, pp. 19-20. Cf. also p. 195. Qälün is cited in the Muhkam through
the isnäd cAbd Alläh b. cIsä al-Madani > Muhammad b. Ahmad b. Munlr > Ahmad b. cUmar
al-Jlzl (in one case, the second transmitter is replaced by Muhammad b. al-Asbagh). In this
passage, Dutton interpreted the expression harf musakkan as 'consonants that are not to be
pronounced at all' (Dutton, 'Red Dots, Green Dots ... Part Γ, p. 118). However the usage
of al-Dänl, here and in other passages, as well as that of classical dictionaries shows that this
adjective designates a letter with sukün; for examples of the latter, see the definitions of the
words bakh and hinh, respectively in al-FTrüzäbädl, al-Qämüs al-muhii, vol. 1, p. 265; Ibn
Manzür, Lisän al-cArab, vol. 2, p. 432.
41 al-Dänl, al-Muhkam, ρ 147.
42 al-Dänl, al-Muhkam, ρ 19. Cf. also al-Dänl, Kitäb al-naqt, p. 132.

44 al-Dänl, al-Muhkam, ρ 8, pp. 117-18.
47 al-Dänl, al-Muhkam, ρ 8. Cf. also Ibn al-Jazari, Ghäyat al-nihäya, vol. 2, p. 3 (no. 2534);
Bergsträsser and Pretzl, Geschichte, p. 175.
48 al-Däni, al-Muhkam, p. 87. Translation after Dutton, 'Red Dots, Green Dots ... Part Γ,
p. 119, with minor modifications. The name of the scribe is erroneously printed as Hakam in the
Arabic text, see the corrigenda in al-Dänl, al-Muhkam, last page (unnumbered). For an
explanation of the above cases of hurüf z.awä'id, see Dutton, 'Red Dots, Green Dots ... Part Γ,
p. 136 (n. 45). A-man huwa qänit"", with takhfif on the mim, is the reading of Ibn Kathir, Nafi",
Hamza and others; the standard reading today, that of Hafs 'an 'Äsim, is amman huwa qänit""
(see al-Khatib, Mucjam, vol. 8, pp. 141-2). As a side note, a variant reading of a fa 'in mitta is a
fa'Tn muttu (al-Khatib, Mucjam, vol. 6, p. 17).
49 Ibn Abi Däwüd, Materials, p. 143.
50 al-Dani, al-Muhkam, p. 9.
52 al-Dänl, al-Muhkam, p. 51. Cf. Also p. 86.
55 al-Däni, Kitäb al-naqt, p. 150. Cf. also Kunitzsch, 'The Transmission of Hindu-Arabic
Numerals Reconsidered', pp. 4—10.
56 al-Däni, al-Muhkam, p. 7.
57 al-Däni, al-Muhkam, pp. 51-2.
59 al-Däni, al-Muhkam, p. 86; al-Däni, Kitäb al-naqt, p. 145.
62 al-Däni, al-Muhkam, pp. 93-118.
64 al-Däni, Kitäb al-naqt, p. 138. On the hurüf al-madd wa'l-lin, see Fleisch, art. 'Hurüf
al-hidjä3'.
65 al-Däni, al-Muhkam, p. 8. Qälün is cited with regard to the Medinans.

66 al-Däni, al-Muhkam, p. 20; translation after Dutton, 'Red Dots, Green Dots ... Part I',
p. 118. Cf. also al-Däni, al-Muhkam, p. 147.
67 al-Däni, al-Muhkam, p. 56. Cf. also al-Däni, Kitäb al-naqt, p. 137.

70 al-Däni, al-Muhkam, p. 37. Cf. Deroche, Le livre manuscrit arabe, p. 81.
71 al-Däni, al-Muhkam, p. 222. Elsewhere, Ibn al-Munädi is cited as the author of a book on
vocalisation, which could be the source of this quotation; see below, note 76.
72 al-Däni, al-Muhkam, p. 128 (my emphasis).
73 al-Däni, al-Muhkam, p. 128. Variations on this case and their notation in the Maghrib are
discussed in the previous pages.
75 al-Däni, al-Muhkam, p. 20. Cf. also Dutton, 'Red Dots, Green Dots ... Part I', p. 118;
al-Däni, Kitäb al-naqt, p. 134.
76 al-Däni, al-Muhkam, pp. 21-2. On Ibn al-Munädi, see Sezgin, Geschichte des Arabischen
Schrifttums, vol. 1, p. 44; Claude Gilliot, 'Kontinuität und Wandel in der „klassischen"
islamischen Koranauslegung', p. 28 (n. 164).
77 Ibn ai-Nadim, Kitab al-fihrist, p. 65 (on al-Mubarrad, Kitab al-khatt wa'l hijaj, p. 89 (on
Ibn Kaysän, Kitäb al-hijäj.
78 al-Dänl, al-Muhkam, p. 7, p. 22.
79 E.g. Blair, Islamic Calligraphy, pp. 145-7; George, The Rise of Islamic Calligraphy, fig. 68
(both 252/866); al-Munajjid, al-Kitäb al-'Arabi al-makhtüt, Pi. 15 (279/892), PI. 16 (280/893);
Moritz, Arabic Palaeography, PI. 119, PI. 120 (both 311/923), PI. 121 (351/962). The khä ' for
khafif must have been rarely encountered in secular manuscripts, as it is mainly of interest for
formal recitation.
81 al-Dänl, al-Muhkam, p. 235. For the different readings of this phrase, see al-Khatlb,
Mu'jam, vol. 5, pp. 16-19.
82 al-Dänl, al-Muhkam, pp. 8-9. Cf. Shah, 'Exploring the Genesis', p. 13. On Isma'll al-Qust,
see Ibn al-Jazarl, Ghäyat al-nihäya, vol. 1, pp. 150-1 (no. 771); Shah, 'Exploring the Genesis',
p. 20, p. 21.
83 Bergsträsser and Pretzl, Geschichte, p. 272.
84 Report by Bernard Guineau, in Deroche, Islamic Codicology, p. 125. The two exceptions
are not named in the text, making it difficult to assess the pattern among the four manuscripts of
the second/eighth to third/ninth centuries included in the sample.
85 Deroche, 'The Qur'än of Amägür', pp. 59-66.
86 Deroche, 'The Qur'än of Amägür', p. 65 (n. 7); George, 'The Geometry of the Qur'än of
Amäjür', p. 3; Sahin, The 1400th Anniversary, pp. 152-5 (Cat. 5-7).
87 See http://cudl.lib.cam.ac.uk/view/MS-ADD-01116 (accessed 14/09/2012).
88 For fa", see George, 'The Geometry of the Qur'än of Amäjür', fig. 1, left, line 2; Cambridge
Digital Library, Image 70 (f. 34v), line 3.
89 al-Khatlb, Mu'jam, vol. 1, p. 513.

90 al-Khatlb, Mu'jam, vol. 1, pp. 519-20.
92 al-Khatlb, Mu'jam, vol. 1, pp. 529-30.
93 al-Khatlb, Mu'jam, vol. 1, p. 531. Both readings are attributed to cÄsim by different
sources. A third reading, la ya'murkum, is attributed by some sources to Abü cAmr (who is also
cited as having read lä ya 'murukum) and to Abü Shu°ayb al-SOsl.
94 For all references to Q. 3:81, see al-Khatlb, Mu'jam, vol. 1, pp. 534—7.
95 For all references to Q. 3:83, see al-Khatlb, Mu'jam, vol. 1, pp. 538-9.
98 Cf. Leemhuis, art. 'Readings of the Qur'än'; Shah, 'The Early Arabic
Grammarians' Contributions'; Melchert, 'The Relation of the Ten Readings to One
Another'.
99 Shah, 'The Early Arabic Grammarians' Contributions', pp. 78-85; Melchert, 'Ibn Mujähid',
p. 20.
100 Shah, 'The Early Arabic Grammarians' Contributions', pp. 88-93.
101 James, Qur'äns and Bindings, p. 20 (no. 7). Cf. also Deroche, 'Collections de manuscrits
anciens', pp. 147-9.
102 This is the only overlap between vocalisation and illumination observable in Is. 1421, as
this fragment consists of only three folios. The space left in the layout for the illumination bands
shows that this layer of the manuscript is original. The gold of the verse markers appears to be
the same as for the larger illuminations, and it is outlined in the same dark brown that again
appears to match the ink of the text and larger illuminations, which makes it likely that these
verse markers are original.
103 Is. 1417d, f. 46b and Is. 1417a, f. 47b; Is. 1417b, f. 46a.
104 Arberry, The Koran Illuminated, p. xviii, p. 10 (nos. 23-6); James, Qur 'ans and Bindings,
p. 26 (no. 12); Deroche, 'Les manuscrits arabes dates', p. 349 (no. 36); Deroche, The Abbasid
Tradition, p. 134; Blair, Islamic Calligraphy, pp. 147-8; George, The Rise of Islamic
Calligraphy, p. 119.
105 James, QuCäns and Bindings, pp. 27-8 (nos. 13-14); Whelan, 'Writing the Word of God',
pp. 134—5 (n. 97); Wright, Islam, Faith, Art, Culture, p. 105 (fig. 68).
106 al-Slräfi, Biographies des grammairiens, pp. 8-9, p. 109 and Pis; al-SIräfT, Akhbär
al-nahwiyyln al-Basrlyyln, p. 81; al-Munajjid, al-Kitäb al-cArabl al-makhtüt, PI. 22;
Sayyid, al-Kitäb al-cArabi al-makhtüt, vol. 2, p. 571 and PI. 66; Edwards, Ά Study of
'Eastern Kufic' Calligraphy', pp. 58-9.1 thank Christiane Gruber for helping me to consult the
latter study.
107 For Krenkow's identification, see al-Sirafi, Biographies, p. IV (French), p. 8 (Arabic). For
the pupils of this Ibn Shädhän, see Whelan, 'Writing the Word', p. 135 (n. 97); for ai-Saküni,
see Ibn Sacd, Kitäb al-tabaqät al-kablr, vol. 9, p. 335. Cf. also Stern, '[Review of] Istanbul
Üniversitesi Kiitiiphanesi', p. 398.
108 Silvers, A Soaring Minaret, p. 36.
109 See Krenkow's remarks in al-Sirafi, Biographies, pp. 8-9 (Arabic), and his annotated
corrections in the edited text.
110 This word is damaged in the original document.
111 The word zinjär is of Persian derivation but was used in other parts of the Islamic world,
for instance Ibn Bädis's early fifth/eleventh-century treatise on inks and bookmaking, composed
in North Africa (Levey, Mediaeval Arabic Bookmaking, p. 21, n. 150).
112 Qübila wa-suhhiha wa-cürida bi-cawn Allah. Annotations by the same hand probably
appear in other parts of the manuscript, but only two other pages, including the title page, have
been published.
113 Vajda, Album de paleographie, PI. 20 (Istanbul, Köprülü Library, No. 948, dated
370/980).
114 al-Khatib, Mu'jam, vol. 1, p. 72; vol. 2, p. 381. The reading hwa is also ascribed to NälT,
al-KisäT, Qälün, Abü Jacfar, and al-Hasan.
115 For examples of imäla in the reading of Abu cAmr, see al-Khatib, Mucjam, vol. 1, p. 56,
p. 442, p. 451.
116 al-Khatib, Mu'jam, vol. 1, p. 28.
118 al-Khatib, Mucjam, vol. 1, pp. 45-7.
119 Deroche, 'Les origines de la calligraphic islamique', p. 24, p. 28; Sahin, The 1400th
Anniversary, p. 197.
120 For TIEM and Khalili, see respectively Sahin, The 1400th Anniversary, pp. 196-7
(Cat. 35), and Deroche, The Abbasid Tradition, pp. 154—5 (no. 83). For the Met
fragments, http://www.metmuseum.org/Collections/search-the-collections/140006978; for the
Freer fragment, http://www.asia. si.edu/collectk>ns/singleObject.cfm?ObjectNumber=Fl937.34
(both accessed 10/10/2012); for the Minneapolis fragment, http://collections.artsmia.org/index.
php?page=detail&id=1212 (accessed 04/07/2014).
121 Deroche and von Gladiss, Der Prachtkoran, p. 110 (text: left, lines 3,4; illumination, right,
line 2); Sahin, The 1400th Anniversary, p. 196 (Cat. 35; text: line 2; illumination: line 4).
122 As in the Freer folio (see note 120); Sahin, The 1400th Anniversary, 197 (Cat. 35);
Deroche and Gladiss, Der Prachtkoran, p. 110 (right).
123 See Deroche, The Ahbasid Tradition, p. 154 (no. 83, recto, beginning of line 2); http://
www.metmuseum.org/collections/search-the-collections/140006978?img=5 (end of line 1,
yassarnä, Q. 54:32; beginning of line 4, najjaynähum, Q. 54:34).
124 This contradicts Deroche's assumption that the vocalisation was noted in two stages
because some signs appear in two colours (Deroche, The Abbasid Tradition, p. 155).
125 Image: Deroche, The Abbasid Tradition, p. 154 (no. 83, recto, beginning of line 1).
Variant: al-Khatib, Mucjam, vol. 9, p. 208.
126 Not visible on the above image; see http://www.metmuseum.org/collections/search-the

collections/140006978?img=l (accessed 10/10/2012).
127 al-Khatib, Mu'jam, vol. 9, pp. 218-9.
128 Images: Freer folio (see note 120), line 3; http://www.metmuseum.org/collections/search

the-collections/140006978?img=5, line 1 (accessed 10/10/2012). Variant: al-Khatib, Mu'jam,
vol. 9, p. 234.
129 T1EM, unknown folio number. Image: Sahin, The 1400th Anniversary, p. 197 (lower
image), line 2. Variant: al-Khatib, Mu'jam, vol. 1, p. 310; vol. 9, p. 141.
130 T1EM 453, f. 259b, line 3. Image: Sahin, The 1400th Anniversary, p. 197. Variant: see
note 114. The same word also appears, amongst published pages, in 'inna'lläh-huwa (note 131
below), but only with a shadda for idghäm.
131 TIEM, unknown folio number. Image: Sahin, The 1400th Anniversary, p. 197 (lower
image), line 4. Variant: al-Khatib, Mu'jam, vol. 9, p. 143.
Bibliography
Abbott, Nabia, The Rise of the North Arabic Script and its Kur'änic Development
(Chicago: University of Chicago Press, 1939).
cAbd al-Razzäq al-SancänI, al-Musannaf ed. Hablb al-Rahmän al-AczamI (11 vols.
Beirut: al-Majlis al-cllml, 1970-2).
Arberry, Arthur J., The Koran Illuminated: A Handlist of the Korans in the Chester
Beatty Library (Dublin: Hodges, Figgis, 1967).
Ben Cheneb, M., art. 'al-Dänt' in Encyclopaedia of Islam, 1st edn.
Bergsträsser, Gotthelf and Otto Pretzl, Geschichte des Qoräns von Theodor
Nöldeke. III: Die Geschichte des Korantexts, 2nd edn (Leipzig: Dieterichische
Buchhandlung, 1938).
Blair, Sheila, Islamic Calligraphy (Edinburgh: Edinburgh University Press, 2006).
al-Dänl, Abü cAmr cUthmän b. SacId, al-Muqnic β rasm masähif al-amsär, maca
Kitäb al-naqt, ed. Otto Pretzl (Beirut: Orient-Institut, 2009).
, al-Muhkam fl naqt al-masähif ed. cIzzat Hasan, 1st edn (Damascus: Wizärat
al-Thaqäfa wa'l-Irshäd al-Qawmi, 1960).
Deroche, Francois, Qur'ans of the Umayyads: A First Overview (Leiden: Brill, 2014).
, La transmission ecrite du Coran dans les debuts de Γ Islam: Le codex Parisino

petropolitanus (Leiden: Brill, 2009).
, Islamic Codicology: An Introduction to the Study of Manuscripts in Arabic
Script, tr. Deke Dusinberre and David Radzinowicz (London: Al-Furqan Islamic
Heritage Foundation, 2005).
—, 'Colonnes, vases et rinceaux sur quelques enluminures d'epoque omeyyade',
Comptes Rendus des Seances de I 'Academic des Inscriptions et Belles Lettres
(2004), pp. 227-64.
—, Le livre manuscrit arabe: Preludes ä une histoire (Paris: Bibliotheque
Nationale de France, 2004).
—, 'New Evidence About Umayyad Book Hands', in Essays in Honour of Saläh
al-Din al-Munajjid (London: Al-Furqän Islamic Heritage Foundation, 2002):
pp. 611—42.
, The Abbasid Tradition: Qur'äns of the 8th to 10th Centuries A.D. (London:
The Nour Foundation, 1992).
—, 'The Qur'än of Amägür', Manuscripts of the Middle East 5 (1990-1), pp. 59-66.
—, 'Les manuscrits arabes dates du Ille/IXe siecle', Revue des Etudes Islamiques
55-7 (1987-9), pp. 343-79.
—, 'Les origines de la calligraphic islamique/the Origins of Islamic Calligraphy' in
Textes sacres et profanes/Sacred and Secular Texts: Catalogue of an exhibition at
the Musee d'Art et d'Histoire ([Geneva], 1988).
, Les manuscrits du Coran: Aux origines de la calligraphic coranique,
Catalogue des manuscrits arabes. Deuxieme partie: manuscrits musulmans.
Tome I, I (Paris: Bibliotheque Nationale de France, 1983).
—, 'Collections de manuscrits anciens du Coran ä Istanbul: Rapport preliminaire'
in Janine Sourdel-Thomine (ed.) Etudes nwdiivales et patrimoine turc (Paris:
Editions du Centre National de la Recherche Scientifique, 1983), pp. 145-65.
Deroche, Frangois, and Altmut von Gladiss, Der Prachtkoran im Museum für
islamische Kunst: Buchkunst zur Ehre Allahs (Berlin: Museum für Islamische
Kunst, 1999).
Dutton, Yasin, 'Red Dots, Green Dots, Yellow Dots and Blue: Some Reflections on
the Vocalisation of Early Qur'anic Manuscripts. Part Π', Journal of Qur'anic
Studies 2:1 (2000), pp. 1-24.
, 'Red Dots, Green Dots, Yellow Dots and Blue: Some Reflections on the
Vocalisation of Early Qur'anic Manuscripts. Part I', Journal of Qur'anic Studies
1:1 (1999), pp. 115—40.
Edwards, Frances C„ Ά Study of "Eastern Kufic" calligraphy' (Unpublished MA
Thesis, University of Michigan, 1981).
Fesharaki, M.A., and S. Saeedpoor, art. 'Abu °Amr al-Dänl' in Wilferd Madelung and
Farhad Daftary (eds), Encyclopaedia Islamica. Volume 1 (Leiden: Brill/Institute of
Ismaili Studies, 2008).
al-FTruzabadi, al-Qamus al-muhlt (4 vols. Cairo: Mustafa al-Babl al-Halabl, 1371
[1952]).
Fleisch, Η., art. 'Hurüf al-hidjä" in Encyclopaedia of Islam, 2nd edn.
George, Alain, 'Le palimpseste Lewis-Mingana de Cambridge, temoin ancien de
l'histoire du Coran', Comptes-Rendus des Seances de l'Academie des Inscriptions
et Belles Lettres (2011), pp. 377-429.
, The Rise of Islamic Calligraphy (London: Saqi, 2010).
, 'Calligraphy, Colour and Light in the Blue Qur'an', Journal of Qur'anic
Studies 11:1 (2009), pp. 75-125.
, 'The Geometry of the Qur 'an of Amäjür: A Preliminary Study of Proportion in
Early Islamic Calligraphy', Muqarnas 20 (2003), pp. 1-15.

Gilliot, Claude, 'Kontinuität und Wandel in der „klassischen" islamischen
Koranauslegung (II./VII.-XIl./XIX. Jh.)', Der Islam 85:1 (2009), pp. 1-155.
Hamdan, Omar, 'The Second Masahif Project: a Step Towards the Canonization of
the Qur'anic Text' in Angelika Neuwirth, Nicolai Sinai, and Michael Marx (eds),
The Qur 'an in Context: Historical and Literary Investigations into the Qur'anic
Milieu (Leiden: Brill, 2010), pp. 795-835.
Ibn Abi Däwüd, Materials for the History of the Text of the Qur'an. The Old Codices:
the Kitäb al-masähif of Ibn AbT Däwüd, ed. Arthur Jeffery (Leiden: Brill, 1937).
Ibn cAtiyya, cAbd al-Haqq b. Ghälib, al-Muharrar al-wajlzfi tafslr al-kitäb al-caziz,
ed. A. Muhammad (5 vols. Beirut: Dar al-Kutub al-cllmlyya, 1993).
Ibn Bashkuwäl, al-Sila fi tärlkh a'immat al-Andalus, ed. cIzzat Al-cAttär al-Husayni
(2 vols. Cairo: Maktabat al-Khanji, 1955).
Ibn Durayd, Kitäb al-ishtiqäq = Abu Bekr Muhammed ben el-Hasan ibn Doreid's
Genealogisch-etymologisches Handbuch, ed. Ferdinand Wüstenfeld (Göttingen:
Dietrichsche Buchhandlung, 1854).
Ibn al-Jazaril, Muhammad b. Muhammad, Ghäyat al-nihäya fi tabaqät al-qurrä', ed.
Gotthelf Bergsträsser (2 vols. Beirut: Dar al-Kutub al-cllmlyya, 2006).
Ibn Manzür, Lisän al-cArab (15 vols. Beirut: Dar Sädir, 1955-6).
Ibn al-Nadlm, Kitäb al-fihrist, ed. Ridä Tajaddud (Tehran: Matbacat Dänishgäh,
1971).
Ibn Qutayba, Kitäb al-macärif = Ibn Coteiba's Handbuch der Geschichte, ed.
Ferdinand Wüstenfeld (Göttingen: Vandenhoeck und Ruprecht, 1850).
Ibn Sacd, Kitäb al-tabaqät al-kabir, ed. °Al! Muhammad cUmar (11 vols. Cairo:
Maktabat al-Khänji, 2001).
James, David, Qur'äns and Bindings from the Chester Beatty Library: A Facsimile
Exhibition (London: World of Islam Festival Trust, 1980).
al-Jumahl, Muhammad b. Salläm, Tabaqät fuhül al-shuLarä', ed. Mahmüd
Muhammad Shäkir (2 vols. Cairo: Matbacat al-Madanl, 1974).
al-Khatib, cAbd al-Latlf Muhammad, Mu'jam al-qirä'ät (11 vols. Damascus: Dar
Sacd al-Dln, 2002).
Kunitzsch, Paul, 'The Transmission of Hindu-Arabic Numerals Reconsidered' in

J.P. Hogendijk and Abdelhamid Sabra (eds), The Enterprise of Science in Islam:
New Perspectives (Cambridge, MA: MIT Press, 2003), pp. 3-21.
Lane, Edward William, Arabic-English Lexicon (8 vols. London: Williams and
Norgate, 1863-93).
Leemhuis, Frederik, art. 'Readings of the Qur'än' in Encyclopaedia of the Qur'än.
Levey, Martin, Mediaeval Arabic Bookmaking and its Relation to Early Chemistry
and Pharmacology (Philadelphia: American Philosophical Society, 1962).
McCants, William Faizi, Founding Gods, Inventing Nations: Conquest and Culture
Myths from Antiquity to Islam (Princeton, NJ: Princeton University Press, 2012).
Melchert, Christopher, 'The Relation of the Ten Readings to One Another', Journal of
Qur'anic Studies 10:2 (2008), pp. 73-87.

, 'Ibn Mujähid and the Establishment of the Seven Qur'anic Readings', Studia
Islamica 91 (2000), pp. 5-22.
Moritz, Bernhard, Arabic Palaeography: A Collection of Arabic Texts from the First
Century of the Hidjra till the Year 1000 (Cairo: Publications of the Khedivial
Library, 1905).
al-Munajjid, Saläh al-DTn, al-Kitäb al-cArabl al-makhtüt ilä'l-qarn al-cäshir al-hijrl

Vol. 1 (al-namädhij) (Cairo: Machad al-Makhtütät al-cArabIyya, 1960).
Noth, Albrecht, The Early Arabic Historical Tradition: A Source-Critical Study, in
Collaboration with Lawrence Conrad (Princeton, NJ: The Darwin Press, 1994).
Pellat, Charles, The Life and Works of Jah.ii (Berkeley, CA: University of California
Press, 1969).
Radiciotti, Paolo and Ariana D'Ottone, Ί frammenti della Qubbat al-khazna di
Damasco: aproposito di una scoperta sottovalutata', Nea Rhome 5 (2008), pp. 45-74.
Sadeghi, Behnam and Uwe Bergmann, 'The Codex of a Companion of the Prophet
and the Qur'än of the Prophet', Arabica 57:4 (2010), pp. 343-436.
Sadeghi, Behnam and Mohsen Goudarzi, 'Sancä' I and the Origins of the Qur'än',
Der Islam 87:1-2 (2012), pp. 1-129.
Sahin, Seracettin (ed.), The 1400th Anniversary of the Qur'an ([Istanbul?]: Antik
A.S. Cultural Publications, 2010).
al-Samhüdl, Wafä' al-wafä bi-akhbär Dar al-Mustafä, ed. Qäsim al-Sämarrä'T
(5 vols. London: Al-Furqän Islamic Heritage Foundation, 2001).
Sayyid, Ayman Fu'äd, Al-kitäb al-cArabl al-makhtüt wa-cilm al-makhtütät = Le
manuscrit arabe et la codicologie (2 vols. Cairo: al-Där al-Misriyya al-Lubnäniyya,
1997).
Sezgin, Fuat, Geschichte des Arabischen Schrifttums (13 vols. Leiden: Brill,
1967-2000).
Shah, Mustafa, 'The Early Arabic Grammarians' Contributions to the Collection
and Authentication of Qur'anic Readings: The Prelude to Ibn Mujähid's Kitäb
al-Sabca', Journal of Qur'anic Studies 6:1 (2004), pp. 72-102.
, 'Exploring the Genesis of Early Arabic Linguistic Thought: Qur'anic Readers
and Grammarians of the Basran Tradition (Part Π)' Journal of Qur'anic Studies 5:2
(2003), pp. 1-47.
Silvers, Laury, A Soaring Minaret: Abu Bakr al-Wasiti and the Rise of Baghdadi
Sufism (Albany, NY: State University of New York Press, 2010).
al-Slräfi, Abu SacId al-Hasan b. cAbdalläh, Akhbär al-nahwiyyin al-Basriyyin, ed.
Muhammad cAbd al-Muncim Khafajl and Muhammad Tähä al-Zayni (Cairo:
Mustafa al-Bäbl al-Halabi wa-Awläduhu, 1955).
, Biographies des grammairiens de l'ecole de Basra = Akhbär al-nahwlyyin al
Basriyyln, ed. Fritz Krenkow (Paris: Geuthner/Beirut: Imprimerie Catholique,
1936).
Stem, S.M., '[Review of] Istanbul Üniversitesi Kütüphanesi Arapâ Yazmalar

Katalogu, V. 1, 1951', Bulletin of the School of Oriental and African Studies 16: 2
(1954), pp. 398-9.
Surty, Muhammad Ibrahim, A Course in the Science of Reciting the Qur'äin, 2nd edn
(Leicester: The Islamic Foundation, 2000).
Talmon, Rafael, 'Schacht's Theory in the Light of Recent Discoveries Concerning and
the Origins of Arabic Grammar', Studia Islamica 65 (1987), pp. 31-50.
, 'Who Was the First Arab Grammarian?', Zeitschrift für Arabische Linguistik
15 (1985), pp. 128-45.
cUmar, Ahmad Mukhtär, and cAbd al-cÄl Sälim Makram, Mufam abqirä'ät
al-Qur 'änlyya, maca muqaddima fi'l-qirä 'ät wa-ashhar al-qurrä ' (8 vols. Kuwait:
Jämicat al-Kuwayt, 1982-5).
Vajda, Georges, Album de paleographie arabe (Paris: Maisonneuve, 1958).
Whelan, Esthelle, 'Writing the Word of God: Some Early Qur'än Manuscripts and
their Milieux, Part I', Ars Orientalis 20 (1990), pp. 113^17.
Wright, Elaine, Islam: Faith, Art, Culture, Manuscripts of the Chester Beatty Library
(London: Scala, 2009).
Yäqüt al-Rüml, Irshäd al-arlb ilä macrifat al-adlb = Dictionary of Learned Men,
ed. D.S. Margoliouth (7 vols. Leiden: Brill, 1907-27).
al-Zarkashl, Muhammad b. Bahadur, al-Burhän fi culüm al-Qur'än (4 vols. Cairo:
Dar al-Turäth, n.d.).

George ColouredDotsQestion 2015

Uploaded by

Copyright:

Available Formats

George ColouredDotsQestion 2015

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

George ColouredDotsQestion 2015

Uploaded by

Copyright:

Available Formats

Centre for Islamic Studies at SOAS

Journal of Qur'anic Studies 17.1 (2015): 1-44

The following abbreviations have been employed:

• BNF: Bibliotheque Nationale de France (Paris)

• Cambridge: Cambridge University Library (Cambridge, England)

• CBL: Chester Beatty Library (Dublin)

• Freer: Freer Gallery of Art, Smithsonian Institution (Washington, DC)

• Khalili: Nasser D. Khalili Collection of Islamic Art (London and Geneva)

• Met: Metropolitan Museum of Art (New York)

• NLR: National Library of Russia (Saint Petersburg)

• TIEM: Türk ve islam Eserleri Müzesi (Istanbul)

Al-Däni on the Regional Habits of Vocalisers

confirmed by passing remarks made by al-Dänl himself in his writings, as when he

Whoever has mastered the art of calligraphy (sinäcat al-khatt) among

al-masähif, respectively published in editions by Otto Pretzl and cIzzat Hasan.

1. The Umayyad Period

by writers of the third/ninth century onwards.15 The historicity of both sets of

Qur'anic notation, it is possible to derive some related evidence from extant

were used well beyond the time of their production.18

I have seen (wasala ilayya) an old mosque Qur'an (mushafjämic catlq)

The mention of an 'old mosque Qur'an' brings to mind a corpus of monumental

Turning to the question of regional origins, it is difficult to assert where this

Syria (biläd al-Shäm).2H By contrast, the persons listed by al-Dänl as possible

Vocalisation is [a creation] of the people of Basra; all others took it

development of notations systems; these, he dispenses in fragments scattered

2. Medina and the Maghrib

Al-DänT saw the Medinan system as having acquired a distinctive character at an

and yellow is used specifically for hamzät.

In the Qur'ans of the people of Medina, mukhaffaf letters carry a red

circle (dära), as do musakkan letters ... Letters dotted in yellow are

In the Qur'ans of the people of Medina, mushaddad letters carry a däl,

I have seen a Qur'an written and vocalised by Hakim [Hukaym?] b.

he was overseeing a collaborative enterprise. In other manuscripts, these two tasks -

Al-DänI's numerous references to nuqqdt and ah I al-naqt ('vocalisers') does bring to

and theoretical knowledge of vocalisation. His forebear Hakim b. Hmrän is also

have also existed in the Maghrib, at least initially:53

lafi) among the people and vocalisers of Medina.

combinations of yellow and red dots.62 In addition:63

By contrast, al-Däni portrays the conventions that prevailed in Iraq as essentially a

and distinguished from others.

Citing Ibn al-Munäd! (Baghdadi, 256-336/869-947), he also notes some cases in

the fatha, following the school (madhhab) of al-Khalll [ibn Ahmad]

or after the alif to reflect the position of hamza.73

The picture of Iraqi/Mashriqi conventions provided by al-Dänl is far from monolithic:

while it does highlight a dominant usage, it also indicates variations initiated by

of (now lost) books on naqt by Iraqi authors of the second/eighth to fourth/tenth

the habit of some readers to gather different readings and variants

green, yellow, dark blue (al-läzaward), while signalling this at the

If what is read is vocalised in two layers (calä wijhayn) or more,

On the authority of Abü'l-0Abbas Muhammad b. Yazid al-Mubarrad (Basran, 210-85/

be introduced in Qur'ans by al-Dänl's lifetime.

dhalika fi sä°ir al-masähif at-'Irüiqiyya wa'l-Shämiyya)'.80 This sentence mns against

From the foregoing discussion, it is possible to draw a comparative table of regional

a later phase nearer al-Dänl's lifetime (fourth-fifth/tenth-eleventh century). In cases

Other Regional Habits

As a result of al-Dänl's focus on the Maghrib, Medina, and Iraq, virtually

Iraq and the Medina Al-Andaius and Al-Andalus and

Mashriq the Maghrib (early) the Maghrib (later)

Hamza Red dot Yellow dot Yellow dot Yellow dot