Northern Africa or Central Iran? An Investigation into the Production Place of a Fragmentary Kufic Qur'an at the J. Paul Getty Museum

Introduction

Islamic calligraphy thrived in two main realms in the first centuries of its flourishing. Officially, it was mostly employed by those charged with copying the Qur'an and architectural inscriptions while adhering to prescribed rules and principles of calligraphy and attending to its aesthetic aspect through rigorous, direct strokes of the pen (qalam). Unofficially, it was generally employed in bureaucratic and everyday affairs, in which scribes used cursive and mostly unprincipled strokes of the pen. In the first realm, as early writers such as Ibn al-Nadīm and today’s researchers show, Hijazi (Makkī/Madanī) scripts shaped the first stage of writing the Qur'an.¹

Some scholars say that the classification of the scripts used in the Qur'ans of the first Islamic centuries (the seventh to ninth centuries CE) originated in copies that were preserved in the Royal Danish Library by Orientalist and theologian Jacob Georg Christian Adler (1756–1834). Adler, according to firsthand sources, first used the word Kufic to refer to the script of the five copies in the library’s collection. Years later, on the basis of information gained from the introduction to Al-Fihrist by Ibn al-Nadīm, Italian historian Michele Amari (1806–89) proposed the Makkī/Madanī scripts as the first to be used in copying the Qur'an,² although later his definitive view was challenged by Qur'anic paleographer Estelle Whelan and art historian Sheila Blair.³ Using Adler’s methodology for her 1939 publication, scholar of early Arabic scripts Nabia Abbott tried to resolve the discrepancies between texts and the surviving manuscripts by proposing a classification of scripts into the double categories of Makkī-Madanī and Basrī-Kufic under the title of Hijazi scripts.⁴ Nevertheless, the problem of the diversity of styles and their nonuniformity with names of calligraphic scripts and terms from the texts remained unresolved.

On the basis of classification propounded by Muslim historians such as Nāji Zayn al-dīn in 1972 and Habiballah Faẓāʾilī in 1973,⁵ another classification was proposed by British scholar Martin Lings in 1976. In Lings’s classification, Kufic is reckoned to be the evolved form of the early scripts and, from the tenth century on, divided into two significant styles: Eastern Kufic, nearly encompassing Iran, and Western Kufic, encompassing Andalusia and North and West Africa.⁶ Despite this classification’s vast geographic and historic scope, which provided for a more rigorous recognition of scripts, it did not prevail.

Challenging the term Kufic, codicologist and paleographer François Déroche proposed a more precise classification of straight scripts used in the Qur'ans in the first Islamic centuries. Because of the indeterminacy of the place and time of the early Kufic script’s genesis, he regarded Kufic as inappropriate for this style of writing and proposed replacing it with the term Early Abbasid. Moreover, he called the Qur'anic scripts of the tenth to thirteenth centuries—identical to Eastern or Persian Kufic—New Abbasid. After rigorous study of the letters’ forms—reported first in 1983 in his book Les manuscrits du coran: Aux origines de la calligraphie coranique (The manuscripts of the Qur'an: Origins of Qur'anic calligraphy) and then in 1992 in The Abbasid Tradition—Déroche classified the scripts of the Qur'an in the seventh to tenth centuries sequentially as Hijazi scripts (comprising four groups), Early Abbasid (comprising six groups), and New Style (comprising two groups).⁷ This way, he could organize the miscellany of the styles of the Qur'anic scripts in each of the three periods.⁸

Group D is the most numerous and diverse group of Early Abbasid script. It comprises five distinct styles, with one (D.V) being further divided into three subtypes. Déroche selected the basic letterforms for the purpose of comparison. Although the letters alif, mīm, nūn, and hā are rendered in different ways, their developmental sequence cannot be placed in chronological order. Generally, all types of Group D exhibit a thick script, with vertical upstrokes that are always perpendicular to the baseline.⁹

As Déroche declares in an elaboration of his method of classification and analyses of Abbasid style, our lack of knowledge about the copying of the Qur'an in the eastern part of the Muslim world is rooted in the fact that the surviving leaves with Early Abbasid script have mostly been found in the western part. Due to the unavailability of reliable sources and copies, Déroche regards the styles of Qur'an scripts in the east as quite indefinite.¹⁰ Considering the evidence available to him at the time, such a claim seems understandable to some extent, as all the Kufic Qur'an fragments in the collections of his investigation lack information on the place of production. Because copies and manuscripts, such as the Amajur Qur'an,¹¹ were relocated or endowed to certain places, we have no solid reason to attribute their origins to their places of discovery. He regards any attempts to pinpoint origins as futile. This, along with the issue of determining the historical order of the script styles, are the two problematics in studies of the Qur'an manuscripts of the first Islamic centuries.¹² Yet, there is evidence in some Qur'an manuscripts that can help us to determine the geographic scope of certain scripts.

A few surviving leaves of a Qur'an attributed to the ninth century offer us the possibility of surmising its script style’s geographic scope. The leaves of this manuscript are dispersed in different museums and collections; the highest number are preserved at the J. Paul Getty Museum. Before dispersion, the manuscript was discovered in Kairouan, Tunisia. Two leaves are kept in the National Library of Tunisia (Rutbi 198); therefore, it has been attributed to Kairouan by historians including Lings¹³ and to North Africa by catalogers of the Getty Museum.¹⁴ Yet, resemblance of its script to the style of the copies produced in Iran suggests that the place of production of this Qur'an manuscript was Central Iran. This study investigates the characteristics of the Getty manuscript’s leaves and relates evidence for its place of production.

The Fragmentary Kufic Qur'an at the Getty Museum

Ten leaves of a Qur'an manuscript produced in the horizontal format and written in outlined gold script on parchment are preserved in the Getty Museum (Ms Ludwig X 1).¹⁵ Other leaves are listed in other collections and at auctions. In what follows, the characteristics of the Getty leaves and the other leaves’ places of preservation are expounded.

Fragmentary Qur'an from the J. Paul Getty Museum:

3rd century AH/9th century CE, probably Central Iran
Unknown calligrapher
14.4 × 20.8 centimeters, with five lines to the page
Pen and ink, gold paint, and tempera colors on parchment
Text area: 9.3 × 15 centimeters
Script: D.I

Folios 1, 2:

3:122–26
First words: minkum an tafshalā
Last words: bihī va ma al-naṣru

Folios 3, 10:

3:129–31
First words: fi al-arḍ yaghfiru liman yashāʾu
Last words: lilkāfirīna

Folios 4r–6v:

6:106–12
First words: [a]ʿriḍ ʾan al-mushrikīna
Last words: ilā baʿḍin

Folios 7r–9v:

6:116–22
First words: allāh in yattabiʿūna
Last words: kadhālika zuyyina

Other fragments:

• Paris, Bibliothèque Nationale, MS Arabe 5178 (Déroche, Abbasid Tradition, 67).
• Tunis, National Library of Tunisia, Rutbi 198 (Lings and Safadi, The Qur'ān, nos. 16, 17 [two folios each])
• Kairouan, National Museum of Islamic Arts of Raqqada (Lings and Safadi, The Qur'ān, nos. 18 [one folio], 19 [two folios], pl. 3)
• Sotheby’s, London, 15 October 1984, lot nos. 218–19; 25 June 1985, lot no. 5; 21–22 November 1985, lot nos. 290–91; 22 May 1986, lot no. 248; 1 June 1987, lot no. 78; 2 April 1988, lot no. 114; 10 October 1988, lot no. 170; 10 April 1989, lot no. 169; 26 April 1990, lot no. 140
• Christie’s, London, 9 October 1990, lot no. 45
• Sotheby’s, London, 2 October 1991, lot no. 892
• Sotheby’s, London, 18 October 2001, lot no. 4
• Christie’s, London, 23 April 2007, lot no. 3; 4 April 2012, lot nos. 1 and 2
• Toronto, Aga Khan Museum, no. AKM 480
• Khalili Collection, no. KFQ 84 (Déroche, Abassid Tradition, 67)

Each folio of text in this Qur'an manuscript includes five relatively bulky lines, which is evidence of its production as a several-volume copy. This copy is likely to have been originally produced in thirty volumes, as five-lined Qur'an manuscripts were generally divided into thirty separate parts, called juzʾ.¹⁶ If this is true, it might be surmised that each volume must have been eighty to ninety folios, considering the approximate number of words on each page.

The ten leaves of the Getty Qur'an constitute five bifolios. There are five lines on each leaf. Leaves one and two include āyahs (verses) of ālʿImrān (3:122–26). Other leaves (folios 3–10) include āyahs of al-anʿām (6:106–22); apparently, two leaves of this part of the manuscript have been lost, as these leaves (folios 3–10) do not include āyahs 112–16. Leaf three, attached to leaf ten, includes āyahs of ālʿImrān (3:129–31). A rectangular illuminated panel is on the back of leaf 10v (fig. 1). On this leaf (recto) is a three-lined frame containing the following text:

سبع مائه و تسع و خمسون و
عدد حروفه خمسه عشر
الفا و ثمان مائه و اربعون

Seven hundred fifty-nine
number of its letters
fifteen thousand eight hundred forty

This is part of a longer text, called briefly verse count, in which the numbers of the Qur'an’s chapters (suras), words, and letters are expressed. Such text has been assumed to be a feature of Iranian Qur'ans,¹⁷ the most well-known of which is the Ibn al-Bawwab Qur'an of 1001 now in the collection of the Chester Beatty Library in Dublin (Is. 1431).¹⁸ Yet, the most ancient manuscript on paper is three volumes of a fourteen-part Qur'an written in the vertical format in New Style in 327 AH/939 CE, which, according to the autograph of its scribe, was copied in Isfahan (fig. 2). This Qur'an was endowed to the Imam Riza Shrine in the early tenth century by a person named Kashvād b. Amlās.¹⁹ In the beginning of the manuscript (folios 1r, 2v), the verse count and the date and place of copying are written in golden Early Abbasid script on two leaves.²⁰

This information was also expressed in a four-volume Qur'an in the horizontal format that was endowed to ʾalā al-Dīn Mosque in Konya, Turkey. Its scribe, Ahmad b. Muhammad b. Yāsīn, copied it in 383 AH/993 CE in Isfahan (TIEM, nos. 453–56). According to its precise calligraphy and conspicuous golden decorations, it was probably produced under the patronage of the courtiers of the Buyyid dynasty (945–1055 CE).²¹ The numbers of its chapters, words, and letters are written in circles inside a horizontal rectangular panel. Scholar of Islamic art Yasser Tabbaa lists and illustrates other manuscripts containing tables with verse counts.²²

In the Getty Qur'an, the rectangular panel on the front of the leaf (recto) is divided into two squares, each containing a circle. The four sets of lines radiating into the center of the circle intersect and form a diamond. The surface decoration of the panel is totally overlaid with gold. The chain-like pattern of its margins and the ornamental dots of the background are all outlined in brown. This structure can also be seen in certain Qur'an manuscripts copied in Abbasid style, an example of which is a fragment including al-ḍuḥā (Qur'an 93) to the end of the Qur'an, written in Early Abbasid style, kept in the Chester Beatty Library.²³ The Chester Beatty manuscript includes two such illuminated leaves at its beginning and at its end. On each of these leaves, as was described for Getty folio 10 above, is a rectangular panel made up of two square frames divided by four triangles (lachak). On the other leaves of the Getty fragments, as previously described, the text is written in five lines in gold. Each word or letter is outlined in brown ink to make it distinct from the background.

The harakāt, or diacritics marking short vowels, are shown by small circular dots in red ink: for fatḥa (ـَ), a circular dot is inserted above the letter; for kasra (ـِ), a dot underneath; for ḍamma (ـُ), a dot in the left side. Nunnation (tanwīn) is shown with two dots, one over the other. These features correspond to the elements of style attributed to Arab grammarian Abul Aswad al-Duʾali (d. 69 AH/688 CE), a poet and the founder of Arabic syntax. He invented, as Egyptian encyclopedist al-Qalqashandī informs us in 821 AH/1418 CE, three main harakāt plus tanwīn.²⁴ The vocalization in this copy is mainly used in the last letter of almost all words, for it is normally the last letter that defines a word’s function in the sentence. If no vowel pointing is applied, misunderstanding and change in meaning would be inevitable. This practice of using vocalization exclusively for the last letter is attributed to al-Duʾali.

Nunnation (tanwīn) in this Qur'an is shown by two red dots. Shadda (a diacritic for doubling a consonant) is shown by a light-blue dot. There are no signs for maddah (long vowel) or the hamza (glottal stop). Because the light blue is the same one used in the illumination, the diacritics seem to have been inserted at the time the manuscript was being copied. But the dots in dark blue otherwise surrounding the letters were inserted at a later time to complete diacritics or to demonstrate other modes of recitation (qirāʾāt) (fig. 3).

In his research on colored Arabic diacritics in Qur'an manuscripts of the first Islamic centuries, which includes analyzing information in the book Al-Muḥkam fī Naqt al-Maṣāḥif (first half of the fifth century AH/eleventh century CE; The precision in the dotting of the Qur'an) by Abu ʾAmr ʿUthmān b. Saʾīd al-Dānī, historian of Islamic art and architecture Alain Fouad George has deduced new features of the regional methods of applying diacritics to Qur'an texts. George summarizes his findings in a table.²⁵ According to his table, the vocalization of early Qur'ans copied in Iraq and Mashriq (a term designating the Arab east) was accomplished through the use of red dots for fatḥa, kasra, and ḍamma, two red dots for tanwīn, a red dot for hamza, and one dot after and above the alif for hamza followed by madd. A comparison between this table and the diacritics in the Getty Qur'an confirms that the manuscript’s place of production is Iraq or the eastern part of the Muslim world.

In the Getty Qur'an, the marker used for the ends of individual verses is a rosette inscribed within a blue circle with colored dots, as seen in many of the Abbasid Qur'ans, which was assigned the code of 3.1.4 by Déroche (see fig. 3).²⁶ On folio 2v, the sign for the numeral five āyahs is a medallion inserted between the words, inside of which the word Khamsa (five) is written and encircled by pudgy painted petals. This ornament is drawn over the last word of the āyah and almost conforms to group 4.A.I in Déroche’s classification.²⁷ The sign for the numeral ten āyahs on folio 6r and 8v greatly resembles the sign for the numeral five āyahs; it differs in that, inside the medallion, the āyah’s numeral is written in Abjad letters. Plus, in the margin in front of the same line, another round medallion is drawn containing the āyah’s number in golden letters (fig. 4). In the leaves of the Getty Qur'an, the signs for the numeral ten āyahs are inserted in the margin in front of line three for āyah 3:130 (3v), line five for āyah 6:120 (8v; see fig. 4), and line one for āyah 6:110 (6r).

In the Abjad system, a number is assigned to each letter as its value. But the order of Abjad letters in the western part of the Muslim world is different from their order in the eastern part.²⁸ Therefore, one of the main criteria for attributing a Qur'an manuscript to the eastern or western part of the Muslim world is to examine the Abjad letters used for counting the āyahs, as art historian Jonathan Bloom does in attributing the early Fatamid Kufic manuscript known as the Blue Qur'an to the west of the Islamic world.²⁹ The signs used for the numeral ten āyahs on the leaves of the Getty Qur'an, however, are assigned to āyahs 110, 120, and 130, as shown by ق، ی, ق، ک and ق، ل. There is no difference between the numeric value of these letters in the eastern and western Abjad systems. One folio of the Qur'an under discussion in this article in the National Library of Tunisia (Rutbi 198)³⁰ includes al-hajj (22:43–44, 22:63–66), which has a sign for the sixtieth āyah written with a س, one of the variable numbers used in the more common, eastern system.

Throughout, Iʾjām diacritics (the dots distinguishing the consonant pointing) are respectively shown by very thin, oblique black lines. The signs to pause while reading (waqf) are indicated by dots of light and dark blue over the letters. This method has been attributed to Arabic linguist Yahya b. Yuʾmar (d. 129 AH/756 CE) and Arab grammarian Nasr b. ʿĀṣim (d. 89 AH/708 CE). Purportedly, the fifth Umayyad caliph ʿAbd al-malik Marwān (r. 65–86 AH/685–705 CE) and governor of Hejaz and Iraq Hajjāj b. Yūsuf-i Thaqafī (r. 73–95 AH/692–714 CE) ordered its creation and establishment to help distinguish between formally similar letters.³¹

The handwriting of the fragmentary Qur'an manuscript is bulky. Letters such as alif and lām are grounded on the baseline with no deviance. Déroche classifies the one leaf of this manuscript kept at the Khalili Collections, London, under the subgroup D.I and provides a description of its script.³² Its characteristics greatly resemble those of the monumental script used for endowments, sura headings, and the beginning notes of certain Qur'an manuscripts in New Style. The similarities of style in these Qur'an scripts are evidence of their production in the same geographic scope. The Qur'an produced in Isfahan in 327 AH/939 CE (see fig. 2) is an example. The text with which this copy begins is part of an endowment, and the verse count is written in Early Abbasid style. Although the quality of the script does not show clearly due to imprecise outlining of the words, the geometry of the letters can be mapped onto the Déroche scripts, and the following properties can be extracted:

• returning stroke of the lower portion of the isolated alif (الف) is longer and tip-pointed;

• medial jīm (ـجـ) is located on the baseline and the letter preceding it is located higher;

• initial ʾayn (ع) has a wide, generally circular opening;

• medial ʾayn (ـعـ) is an inverted triangle, and the final ʾayn (ـع) has a sickle-like kern;

• mīm (م) has a short horizontal kern on the baseline;

• nūn (ن), the bowl of sīn (س), and yā (ی) are relatively big wide circles; and

• single hā (ه) is a semicircle relying on a vertical line.

These characteristics conform to the general attributes of Déroche’s group D. Still, a more accurate recognition of the style of the script is difficult because, besides the lack of clarity, there is an in-between quality to the letters’ forms. Most letters conform to subgroup D.I, but the curves of letters nūn (ن), sīn (س), and yā (ی) are written in the manner of subgroup D.Va, and alif (الف) in subgroup D.III.

A stronger resemblance can be seen in other Qur'an manuscripts in New Style. An example is the beginning of the seventh volume in a ten-volume Qur'an copied by al-ʿabbās b. Muhammad b. al-ʿabbās al-Maṣāḥifī (the copyist of Qur'an manuscripts) al-Qazvīnī, containing its scribe’s autograph on two pages in Early Abbasid script. Its script mostly conforms to subgroup D.I; only the letter hā conforms to subgroup D.III, and the curves of letters nūn (ن), sīn (س), and yā (ی) are written in the manner of subgroup D.Va (fig. 5). The autograph reads as follows:

بسم الله الرحمن الرحیم
کتب هذا الجز
و اذهبه العباس
بن محمد القزوینی

In the Name of Allah, the Compassionate, the Merciful
copied this juzʾ
and illuminated it al-Abbās
b. Muhammad al-Qazvīnī

The scribe is allegedly from Qazvīn, a city in Central Iran, located in ʾIrāq-e ʾajam (Persian Iraq), or the mountainous region (Jibāl) according to the ancient geographic divisions of Iran.³³ During most of the tenth century, this region was under the rule of the Buyyid dynasty (332–447 AH/945–1055 CE). Although the manuscript’s place of production is not noted, scholars know that, on the basis of another colophon related to a Qur'an copied by this scribe, he is associated with the Buyyid dynasty’s court. He copied and illuminated the abovementioned Qur'an manuscript between 387 and 391 AH/997 and 1000 CE for “Umm al-Umarā’s treasury” in Rayy.³⁴

Umm al-Umarā or Umm al-Mulūk (mother of kings) was the title of Sayyidah Malik Khātūn or Shīrīn Dukht-i Ispahbud Rustam-i Ṭabarī (d. 1028), the first Shi'ite woman ruler in the history of Iran. Belonging to the Bāvandiān family of Tabaristān, she married Fakhr al-Dawla Daylamī. Upon her husband’s death in 387 AH/997 CE, she officially became the ruler of the mountainous region (Jibāl) of the Buyyid realm of dominion. Her two sons—Shams al-Dawla Daylamī, governor of Hamedan (r. 387–412 AH/997–1021 CE), and Majd al-Dawla Daylamī, governor of Rayy (387–420 AH/997–1029 CE)—were mere children when they became governors. The forenamed scribe worked for the library of the Buyyid’s court; the script used in copying the manuscript is in New Style and resembles the style of the aforementioned Qur'an manuscript produced in Isfahan in 939.³⁵

These characteristics are also seen in the monumental script of another copy of the Qur'an. On the basis of its resemblance to prior manuscripts, it was most probably copied in the tenth century in Central Iran (Library of Astan Quds Razavi, no. 5015), although it is missing the date of completion and the scribe’s autograph. It is the fourth volume of a ten-volume Qur'an in which the sura headings and the insertions on the pages’ margins are written in Early Abbasid style. Its writing and the outlining of the golden Early Abbasid are more exacting compared to other examples. In the sura heading of al-anʾām, the forms of letters with the vertical tooth of medial ʾayn (ـعـ), the broad shape of mīm (م), the head of wāw (و) on the baseline, and the short arms of lā (لا) conform to subgroups D.I and D.Va (fig. 6). Some explanation about the place of the sura’s revelation is written in the same style inside the medallion beside the sura heading.

Another example of Iranian manuscripts having lines in Abbasid style is the copy of the Qur'an written by Ahmad b. Muhammad b. Yāsīn, dated 383 AH/993 CE (TIEM, nos. 453–56). This copy, in four volumes, is produced in the horizontal format like the Qur'an copied by al-ʿabbās b. Muhammad al-Qazvīnī. In circles inside the panels on its beginning pages, the numbers of its suras, words, and letters are written in Early Abbasid style. But, due to the script’s small size and the imprecise outlining of words, an exact identification of its style is not possible.

Table 1 offers a visual comparison of the main letters of the Getty Qur'an; the main letters of the monumental scripts of the Qur'an manuscripts written in New Style in Central Iran; and subgroups D.I and D.Va in Déroche’s classification. The Getty Qur'an’s script entirely conforms to the style of the monumental script of the Qur'an manuscripts from Central Iran. Therefore, not only does it seem possible to attribute the production of the Getty Qur'an to Central Iran but it is also possible to propose a strong hypothesis attributing the mix of D.I and D.Va scripts to Central Iran.

	alif	jīm / ḥā / khā		ṭā / ẓā	ayn / ghayn		qāf	mīm		nūn		hā	lā
Getty fragmentary Kufic Qur'an
Isfahan Qur'an
Abbās b. Muhammad al-Qazvīnī’s Qur'an
The unknown Qur'an (Library of Astan Quds Razavi, no. 5015)
D.I classification
D.Va classification

Summary

Efforts to recognize the place of production of the Qur'an manuscripts in Early Abbasid script and determine the transformation of Qur'an script styles in the first three centuries of hijra (622–913 CE) have always yielded ambiguous results due to the lack of reliable information. Yet, close examination of the Qur'an manuscripts written in New Style helps us acquire knowledge of regional styles of Qur'anic calligraphy, at least in the eastern part of the Muslim world. Certain Qur'an manuscripts contain inscriptions and insertions written in Early Abbasid style; due to their greater quantity and the intactness of their scribes’ autographs, they provide more accurate information regarding their place of production. Thus we have seen ample evidence that both the fragmentary Getty Qur'an and the mix of D.I and D.Va styles can be traced to Central Iran. In short, the evidence is as follows:

• The insertion of a verse count at the beginning of the Getty Qur'an, containing the number of chapters (suras), words, and letters, which is also included in New Style Qur'ans produced in Central Iran, such as the Qur'an copied in 327 AH/939 CE and the one scribed by Ahmad b. Muhammad b. Yāsīn in 383 AH/993 CE, both produced in Isfahan.
• The resemblance of the Abjad system for counting the āyahs in the Getty Qur'an to the Abjad system common in the eastern part of the Muslim world.
• The considerable resemblance of the Getty Qur'an’s script to the monumental script of the Qur'an manuscripts produced in Central Iran.
• The resemblance of the diacritics used in the Getty Qur'an to the tradition of the eastern part of the Muslim world, as found in Al-Muḥkam fī Naqt al-Maṣāḥif by Abu ʾAmr Uthmān b. Saʾīd al-Dānī.

Mahdi Sahragard is an assistant professor in the Department of Architecture at Islamic Azad University, Mashhad Branch, Mashhad, Iran.

Notes

The author would like to acknowledge the anonymous reviewers for their insightful comments and express gratitude to the editorial team for their valuable feedback. Research for this essay was conducted using the Encyclopaedia of Islam transliteration system for the Arabic and Persian terms.

1. Ibn al-Nadīm, Al-Fihrist [377 AH/987 CE], ed. Muhammad Riza Tajaddud (Tehran: Zavvār, 1971), 9; and François Déroche, The Abbasid Tradition: Qur'ans of the 8th to the 10th Centuries AD (London: Nour Foundation, 1992), 12. ↩︎

2. Déroche, Abbasid Tradition, 12. ↩︎

3. Estelle Whelan, “The Phantom of Hijazi Script: A Note on Paleographic Method” (unpublished manuscript, 1997), referenced in Sheila S. Blair, Islamic Calligraphy (Edinburgh: Edinburgh University Press, 2005), 108. ↩︎

4. Nabia Abbott, The Rise of the North Arabic Script and Its Kuranic Development with a Full Description of the Kuran Manuscripts in the Oriental Institute (Chicago: University of Chicago Press, 1939), 138. ↩︎

5. Habiballah Faẓāʾilī , Aṭlas-i Khaṭṭ: Tahqiq dar Khutūt-i Islamī [Atlas of calligraphy: Research in Islamic scripts] (Isfahan: Mashʿal, 1971), 142. ↩︎

6. Martin Lings and Yasin Hamid Safadi, The Qur'ān: Catalogue of an Exhibition of Qur'ān Manuscripts at the British Library (London: World of Islam Publishing for the British Library, 1976), 29–33. ↩︎

7. François Déroche, Les manuscrits du coran. Aux origines de la calligraphie coranique. Deuxieme partie: Manuscrits musulmans, vol. 1, Catalogue des manuscrits arabes (Paris: Bibliothèque Nationale de France, 1983); and Déroche, Abbasid Tradition. ↩︎

8. Note, however, that the Early Abbasid style in Déroche’s classification originally comprised seven groups. François Déroche, Qur'ans of the Umayyads: A First Overview (Leiden: Brill, 2013), 8. ↩︎

9. Déroche, Abbasid Tradition, 36 ↩︎

10. Déroche, Abbasid Tradition, 35. ↩︎

11. The Amajur Qur'an was endowed to the mosque of Sur (Tyre), located in modern-day Lebanon, by Amajur, the Abassid governor of Syria from 870–78 CE. The surviving folios of the Qur'an have been scattered all around the world, to Istanbul, Cairo, Dublin, Cambridge, and beyond. See Alain Fouad George, “The Geometry of the Qur'an of Amajur: A Preliminary Study of Proportion in Early Arabic Calligraphy,” Muqarnas, no. 20 (2003): 3. ↩︎

12. Allain Fouad George, “Coloured Dots and the Question of Regional Origins in Early Qur'ans,” Qur’anic Studies 17, no. 1 (2015): 1. ↩︎

13. Lings and Safadi, The Qur’ān, 26, nos. 16, 17. ↩︎

14. Anton von Euw and Joachim M. Plotzek, Die Handschriften der Sammlung Ludwig, vol. 3 (Cologne: Schnütgen Museum, 1982), 15–18. See also the museum collection catalog record at https://www.getty.edu/art/collection/object/103RVE; and Tristan Bravinder, “Ninth-Century Qur’an Studied in Depth,” The Iris (blog), 27 March 2017, https://blogs.getty.edu/iris/ninth-century-quran-studied-in-depth/. ↩︎

15. Images of the Getty leaves are available on the website of the J. Paul Getty Museum, https://www.getty.edu/art/collection/object/103RVE. According to Getty provenance records, this fragmentary Qur'an was likely part of a manuscript held by the Great Mosque of Kairouan in 1022. It is believed to have been in the possession of Sultan Mahmud II (1785–1839). Following his passing in 1839, the fragments likely moved to a private collection in Istanbul, Turkey. By 1964, the fragment was acquired by H.P. Kraus (1907–88), a well-regarded antiquarian book dealer. Subsequently, Kraus sold the fragments to collectors Peter and Irene Ludwig, who held it within their collection until the Getty Museum acquired it in 1983. Thanks to Aleia McDaniel for this provenance description. ↩︎

16. For example, see an entire copy in Early Kufic style, the beginning and ending of which conform to the fourth juzʾ of the Qur'an, from 2:192 (Inna al-Laḑīna Kafarū wa mātū) to 4:22 (wa sāʾa sabīlā), Mashhad, Library of Astan Quds Razavi, no. 12220, seventy-seven leaves, 8 × 12 cm. The eastern Kufic Qur'an manuscripts, with five lines to each leaf, were in thirty volumes. The most well known of these manuscripts is a Qur'an copy scribed by ʿuthmān b. Husayn al-Warrāq-i al-Qaznavī, ca. 1070–74, kept in the same library in Mashhad. For more information on this copy, see Blair, Islamic Calligraphy, 197; and Mahdi Sahragard, Sarţi Mastūr: Tārikh va Sabk shenāsī Kūfī Sharqī [The script in veil: The history and stylistics of eastern Kufic script] (Tehran: Academy of the Arts Press, 2020), 173–85. ↩︎

17. Yasser Tabbaa, “The Transformation of Arabic Writing: Part I, Qur'ānic Calligraphy,” Ars Orientalis 21 (1991): 129. ↩︎

18. David James has discussed this Qur'an in detail. David James, Qur'ans of the Mamluks (London: Alexandria, 1988), 24. ↩︎

19. Prior to the discovery of this volume, the manuscript copied in 361 AH/971 CE by ʿAlī b. Shādhān al-Rāzī (Istanbul University Library A. 6758) had been known as the oldest. See Jonathan M. Bloom, “Silk Road or Paper Road?,” The Silk Road 3, no. 2 (2005), www.silkroad/newsletter/vol3num2/5_Bloom.php. ↩︎

20. See Sahragard, Saţr-i Mastūr, 128–33; and Mahdi Sahragard, “Revived Leaves: The Qur'an Endowed by Kashwād b. Amlās (A Manuscript on Paper from Iṣfahān, Dated Ramaḍān 327/939),” Journal of Islamic Manuscripts, no. 14 (2023): 212–34. ↩︎

21. The Qur'an was transferred from ʾAlā al-Dīn Kay Qubād’s shrine in Konya to Türk ve İslam Eserleri Müzesi (TIEM; Turkish and Islamic Arts Museum) in Istanbul in 1913. For more information, see Massumeh Farahad and Simon Rettig, The Art of the Qur’an: Treasures from the Museum of Islamic and Turkish Arts (Washington, D.C.: Smithsonian Institute, 2016), 160. ↩︎

22. See Tabbaa, “Transformation of Arabic Writing,” 126–31. ↩︎

23. Eleven folios from a Qur'an, ninth century CE, 12.8 × 20.2 cm, ten lines to each page, CBL, Is 1411, fol. 1b., https://viewer.cbl.ie/viewer/image/Is_1411/13/. See Arthur J. Arberry, The Koran Illuminated: A Handlist of the Korans in the Chester Beatty Library (Dublin: Hodges, Figgis and Co., 1967), 6, no. 10, pl. 18. Ten other leaves of this manuscript are kept in the Türk ve İslam Eserleri Müzesi (TIEM 552). ↩︎

24. Al-Qalqashandī, Subh al-ʿashā fi ṣināʿat al-īnshā, vol. 3 (Cairo: Majmaʿat al-Amīrīyya, 1924), 157. ↩︎

25. George, “Coloured Dots,” 15. ↩︎

26. Déroche, Abbasid Tradition, 23. ↩︎

27. Déroche, Abbasid Tradition, 25. ↩︎

28. G. Weil and G. S. Colin, “Abdjad,” in Encyclopaedia of Islam, 2nd ed., ed. P. Bearman, Th. Bianquis, C. E. Bosworth, E. van Donzel, and W. P. Heinrichs, http://dx.doi.org/10.1163/1573-3912_islam_SIM_0140. ↩︎

29. Jonathan Bloom, “The Blue Koran: An Early Fatimid Kufic Manuscript from the Maghrib,” in Les Manuscrits du Moyen-Orient: Essais de codicologies et de paléographie, ed. François Déroche (Istanbul: Institut Français d’Études Anatoliennes d’Istanbul, 1989), 97. ↩︎

30. De Carthage à Kairouan: 2000 ans d’art et d’histoire en Tunisie (Paris: Association Française d’Action Artistique, 1983), 262. The illustration of the folio found in the library in Tunisia was provided to the author as part of the information in the files of the Department of Manuscripts, J. Paul Getty Museum, compiled from notes by Nabil Saidi, Dagmar Riedel, Bryan Keene, Morgan Conger, and Elizabeth Morrison. ↩︎

31. ʾuthman b. Saʾīd al-Dāni, al-Muqnaʾ fi maʾrefat rasm masāhif ahl al-amṣār maʾa kitāb al-naqţ (Damascus: Dar al-Fikr, 1983), 124. ↩︎

32. Déroche, Abbasid Tradition, 43. ↩︎

33. Yāqut Al-Hamavī, Muʿjam al-Buldān, vol. 4 (Beirut: Dār Ṣadir, 1977), 342–43. ↩︎

34. Ramezanali Shakeri, Ganj-i Hizār Sāli (The thousand-year treasure) (Mashhad: Astan Quds Razavi, 1988), 65. See Mahdi Sahragard, “Sabk Shināsī Mantaqi-ī Kūfī Sharqī: Ṣifāt-i qalami Kūfī dar Qur'an-i Tāzi-yāb-i Rayy" (Regional Stylistic Features of Eastern Kufic Script: Analyzing the Characteristics in a Recently Discovered Qur'an from Rayy), Golestān-e Honar, 23 (1401/2022): 44–58. The colophon is as follows:

    "فرغ من تذهیبه العباس بن محمد بن العباس │ فی صفر من شهور سنه احدی تسعین و ثلثمائه │ کتبه العباس بن محمد بن العباس│ المصاحفی القزوینی بالری │ لخزانة السیده ام امیرالامرا اطال│ الله مدتهما فی سنه تسع و ثمانین و ثلثمائه."

    ↩︎

35. For more information on this style, see Sahragard, Saṭr-i Mastūr, 134–38. ↩︎