Folio from the "Blue Qur'an"

A single leaf, gold and silver Kufic script on indigo-dyed parchment, cut at an undocumented point from a once-bound multi-volume Qur'an now known only as the Blue Qur'an. The Metropolitan Museum's own catalogue hedges the leaf's origin as "possibly Qairawan, Tunisia", and that hedge is the honest state of the field: at least six named scholars have argued for six different combinations of date and place, from early ninth-century Baghdad to tenth-century Umayyad Spain, and none has settled the question.

What is documented is narrower and, in its way, more striking. The original manuscript, reconstructed as seven volumes of roughly 600 folios, required the parchment of at least 150 sheep and the work of several identifiable scribal hands, luxury production on a scale that every competing theory agrees points to a ruling court as patron. Most of it stayed in the library of the Great Mosque of Qairawan, recorded there in a 1294 inventory, until Tunisia's 1967 nationalisation of religious endowments sent the bulk of the surviving leaves to what is now the Raqqada National Museum of Islamic Art, still the largest single holding. This leaf was not among them: it surfaces in the documentary record only in London, with the dealer Sam Fogg, by 2002, more than eleven centuries after its probable making, and was purchased by the Metropolitan Museum in 2004. Everything in between is unknown, and this record says so.

Other leaves from the same manuscript have gone to auction repeatedly since 2010, for hundreds of thousands of pounds apiece, completing, sale by sale, a dispersal that began somewhere in the manuscript's own past, itself only partly recovered.

A single leaf from a manuscript that no longer exists as a book

This is one page, gold and silver script on indigo-dyed parchment, 30.4 by 40.2 centimetres, holding fifteen or so lines of Qur'anic text in an angular, unpointed Kufic hand: Sura 30, al-Rum, verses 28 to 32. The Metropolitan Museum classifies it as a codex leaf, which is precise and also slightly misleading, because there is no longer a codex. The leaf was cut, at some undocumented point, from a bound multi-volume Qur'an now known only by the informal name given to it by scholars and dealers: the Blue Qur'an.

Nothing about this leaf is ordinary. Indigo-dyed parchment for a religious text was an extravagant and rare choice in any medieval tradition; gold leaf lettering compounded the cost; and the sheer size of the undertaking, discussed below, points to patronage on a scale few institutions of the ninth or tenth century could sustain. The Metropolitan Museum's own catalogue entry hedges its findspot with the word "probably": probably Qairawan, Tunisia. That single word carries the weight of more than a century of unresolved argument among specialists in early Islamic manuscripts, summarised later in this record.

The manuscript this leaf belonged to

The Blue Qur'an was, before its dispersal, one continuous work: a horizontal-format, oblong Qur'an written across what Jonathan Bloom has reconstructed, using the abjad letter-numeral sequence that marks section divisions within the surviving leaves, as seven original volumes. Estimates of the original leaf count run to roughly 600 folios for the complete text. Producing parchment at this scale and this size is estimated to have required the skins of at least 150 sheep, before a single letter was written or a page dyed.

Today only a fraction is known to survive: on the order of 100 leaves and bifolia, dispersed across public and private collections on several continents. The largest single group, and the group with the clearest institutional descent from the original library, is held by the Raqqada National Museum of Islamic Art near Kairouan, Tunisia, with some 67 folios. Other leaves are held by the Chester Beatty Library in Dublin, the Museum of Fine Arts in Boston, the Aga Khan Museum in Toronto, the Nasser D. Khalili Collection of Islamic Art in London, the Los Angeles County Museum of Art, the Seattle Art Museum, the David Collection in Copenhagen, the Zayed National Museum in Abu Dhabi, Qatar's national collections, and a number of private hands, in addition to this leaf at the Metropolitan Museum. No single institution holds anything close to a complete volume; what survives is a scattered census of individual pages, most first identified and published only after they had already left North Africa.

How the page was made

The script is Kufic: angular, without the diacritical dots (i'jam) or vowel signs that would later become standard, laid out with the horizontal emphasis typical of the grandest early Qur'ans. The letters were built in gold leaf, outlined in a dark ink usually identified in technical literature on comparable manuscripts as iron-tannate based, though the adhesive used to fix the gold to the parchment surface is not established with certainty in the sources reviewed for this leaf and has been proposed variously as gum, egg white, or an animal-glue byproduct of parchment preparation. Verse endings and larger textual divisions were marked in silver: small rosettes or medallions, sometimes a roundel enclosing a letter used as a numeral, conventions that recur with local variation across the surviving corpus.

The scale of the project argues strongly against a single calligrapher. Alain George's 2009 codicological study in the Journal of Qur'anic Studies identifies several distinct scribal hands across the surviving leaves, labelled by group and sub-group (his notation includes hands such as B.I, B.II, C.I, C.II, C.III and D.IV), evidence of a workshop of several copyists working to a shared design rather than one master's unbroken hand. No source consulted for this record identifies which hand, if any, produced the Metropolitan's leaf specifically; the point stands for the manuscript as a whole and is recorded here as evidence of an organised scriptorium, not as a claim about this folio's individual scribe.

A concealed error, found by imaging: what recent research shows about production practice

In November 2024, researchers at the Zayed National Museum in Abu Dhabi, using multispectral imaging on a Blue Qur'an leaf in that museum's own collection, reported finding Qur'anic text from Sura al-Nisa' hidden beneath a later layer of gold leaf. The imaging team, credited as curators Nurul Iman Bint Rusli and Mai Al Mansouri, concluded the gold overlay concealed a scribal error, possibly a duplicated passage, rather than serving a purely decorative purpose: instead of discarding an already-dyed, already-gilded sheet of expensive indigo parchment, the workshop appears to have masked the mistake and continued.

This finding concerns a different leaf, held by a different institution, and is reported here for what it establishes about the manuscript's production generally: real, costly, imperfect labour rather than an infallible court commission executed without error. No equivalent imaging study of the Metropolitan's leaf, 2004.88, has been identified in the sources reviewed for this record, and no claim is made here that this leaf carries a similar concealed correction.

Why blue and gold: the Byzantine comparison

Dyeing parchment for a luxury book, and writing on it in metal, was not an Islamic invention. Byzantine scriptoria had for centuries produced imperial-grade manuscripts on parchment stained purple, most famously the Codex Argenteus, a Gothic-language Gospel book in gold and silver ink on purple vellum probably commissioned for the Ostrogothic king Theoderic the Great in the early sixth century, and the Vienna Genesis, a Greek Genesis text in silver on purple parchment associated with the Justinianic court. In Byzantium the colour purple, from the murex dye, was itself restricted to imperial use; a purple manuscript was a political object as much as a devotional one.

Several scholars, prominently Jonathan Bloom, have argued that the Blue Qur'an's colour scheme should be read against this Byzantine tradition: a deliberate emulation, and implicit rivalry, aimed at a comparable register of imperial luxury book production, made plausible in Bloom's Fatimid attribution by the Fatimid state's documented diplomatic contact with Byzantium. This is an argued position, resting on comparative material culture and historical plausibility rather than on any surviving text that states the patron's intention. No source consulted claims that a specific Islamic patron cited Byzantine manuscripts as a model in writing; the connection is scholarly inference from the objects themselves, and it should be read as inference.

A second, narrower argued reading comes from the art historian Yasser Tabbaa, who has proposed that the visual instability of gold script on a dark blue ground, letters that can seem to dissolve into or emerge from their background depending on light and distance, expresses something of the Mu'tazili theological position that the Qur'an, as God's speech, was created rather than eternally co-existent with God. This is a single scholar's interpretive proposal, not a documented statement of patron intent, and it is recorded here at that grade: an argued reading of visual effect, offered by one named specialist, not a consensus finding.

Where and when: the unresolved dispute

No question about the Blue Qur'an is more contested than where and when it was made, and the honest record is that the field has not settled it. At least six distinct scholarly positions have been published, spanning more than two centuries and half the medieval Mediterranean and Near East.

The earliest modern theory belongs to Fredrik Robert Martin, a Swedish diplomat and orientalist who purchased a handful of leaves in Istanbul before 1912 and publicised them through a catalogue issued by the London bookseller Bernard Quaritch, the event that first brought the Blue Qur'an to Western scholarly attention. Martin proposed that the manuscript had been commissioned by the Abbasid caliph al-Ma'mun for the tomb of his father, Harun al-Rashid, associating the dark blue ground with mourning. This theory is now largely rejected by specialists, but it set the terms of debate for a century afterward and explains why the manuscript still carries an association with Abbasid Iraq in some popular accounts.

Jonathan Bloom directly tested and rejected Martin's Abbasid attribution in an early study whose title poses it as a question, "al-Ma'mun's Blue Koran?", and later argued in "The Blue Koran Revisited" (Journal of Islamic Manuscripts, 2015) for a North African origin, most likely Qairawan under Aghlabid or early Fatimid patronage, placing production in the mid-tenth century. Bloom's case rests on two pillars: a 1294 CE inventory of the library of the Great Mosque of Qairawan, published by Ibrahim Chabbouh in 1956, which records a Qur'an written in gold on blue parchment; and his own reconstruction of the manuscript's original seven-volume structure from the abjad numeral sequence visible on surviving leaves.

Francois Deroche and, independently, Alain George have argued for a considerably earlier date, the late eighth or early ninth century, within the early Abbasid period, and for production possibly at Fustat or Baghdad rather than in North Africa at all, based on comparison with the horizontal format and letterforms of other early Abbasid-period Qur'ans. Arianna D'Ottone Rambach has proposed a third geography entirely: Umayyad Spain, pointing to close similarities in parchment colour and page ruling between Blue Qur'an leaves and the Cava Bible, a ninth-century Latin illuminated Bible produced in the kingdom of Asturias/Christian Iberia, and suggesting the manuscript may have been executed by Christian scribes, whose tradition of dyed-parchment sacred books was older and more established than any Muslim equivalent, working for a Muslim patron. Tim Stanley has separately argued for an Andalusian origin, and M. Fraser has proposed Sicily under Aghlabid or Kalbid rule.

These are not shades of the same answer. They disagree by more than a century on date and by the width of the medieval Islamic world on place, from Baghdad or Mashhad in the east to Cordoba in the west, with Qairawan and Palermo argued as intermediate positions. What every position shares is agreement that the patron, whoever it was, commanded resources on a scale associated with a ruling court: none of the competing theories proposes a private or mercantile commission.

The Qairawan library and its 1294 inventory

The single strongest documentary thread tying the manuscript to Qairawan is administrative rather than art-historical. In 1294 CE (693 AH), an inventory was made of the library of the Great Mosque of Qairawan; the Tunisian scholar Ibrahim Chabbouh published this inventory in 1956, and it records among the library's holdings a Qur'an written in gold on blue parchment. Bloom and other scholars who favour a Qairawan attribution treat this entry as the anchor of their case: a specific, dated, documentary reference to an object matching the Blue Qur'an's description, sitting in the library from which the largest surviving group of leaves demonstrably descends.

The identification is an argued inference, not a certainty stated in the source itself. The 1294 inventory does not name the manuscript as "the Blue Qur'an", carries no accession-style identifier linking it unambiguously to the specific leaves now dispersed across museums, and could in principle describe a different, similarly luxurious blue-and-gold Qur'an that has not survived at all. This record treats the identification as the best-supported documentary link available, while noting, as the discipline requires, that a plausible match is not the same thing as a proven one.

From mosque library to the market: how the manuscript came apart

Two separate histories run alongside each other after the manuscript is documented at Qairawan. One is the history of the leaves that left; the other is the history of the leaves, the majority, that stayed, at least until the twentieth century brought its own upheavals.

Martin's pre-1912 Istanbul purchase is the earliest documented instance of Blue Qur'an leaves entering Western hands and Western scholarly literature; how those particular leaves travelled from North Africa (if that is indeed their origin) to the Ottoman capital is not established in the sources reviewed here. His Quaritch catalogue gave the manuscript international visibility and, by extension, a market value it had not previously carried in the West; further leaves reached private collectors and, eventually, museums across the following decades, including the Museum of Fine Arts, Boston, which acquired its leaf in 1931 at the International Exhibition of Persian Art at Burlington House, London, for the sum of 85 US dollars, a figure worth holding onto as a benchmark for how the market valued this material before its full art-historical importance was established.

The bulk of the manuscript, meanwhile, remained under the religious-endowment (habus, or waqf) administration of the Great Mosque of Qairawan into the twentieth century. Tunisia abolished the habus system in 1967, nationalising the properties it had held, including manuscripts, which passed to the National Library in Tunis. In 1983 the Qairawan-associated material was transferred again, to the newly established Centre of Islamic Art and Civilisation at Raqqada, which opened to the public as a museum in 1986 and remains today the single largest custodian of Blue Qur'an leaves, with roughly 67 folios on record. This administrative chain, nationalisation, transfer, museum foundation, is documented for the Tunisian state collection as a whole; it is not evidence for the individual custodial history of any specific leaf now outside Tunisia, including the Metropolitan's.

How this leaf reached the Metropolitan Museum

For this specific folio, 2004.88, the documented record begins and ends abruptly. The Metropolitan Museum's own catalogue entry gives the leaf's probable origin as the Great Mosque of Qairawan, Tunisia, "from about 900", using the hedge "probably" that the Museum's cataloguers apply throughout to the Blue Qur'an's early history. The next documented fact is that the London dealer Sam Fogg held the leaf by some point between 2002 and 2004. Between roughly 900 CE and 2002 CE, more than a millennium, no owner, custodian, sale, exhibition or publication is documented for this individual leaf in the sources consulted for this record. It may have left Tunisia with the earliest wave of dispersed leaves around Martin's time, or considerably later; it may have passed through intermediate private or institutional hands never publicly recorded. None of that can be stated as fact, only as an open gap.

The Metropolitan Museum purchased the leaf in 2004 using funds from the Lila Acheson Wallace Gift, an acquisitions endowment named for the co-founder of Reader's Digest and a major benefactor of the Museum across several curatorial departments. The purchase price has not been published in any source consulted for this record, consistent with the Museum's general practice of not disclosing individual purchase prices. The transaction was a private sale through an established dealer, not an auction; no export licence application or cultural-property dispute in connection with this specific leaf has been identified in the sources reviewed.

The market for single leaves

The Metropolitan's leaf was never auctioned, but other leaves from the same dispersed manuscript have been, repeatedly, since the later twentieth century, and the pattern is itself part of this object's context. Sotheby's London sold a Blue Qur'an leaf on 4 October 2010 (lot 7) for a reported GBP 529,250. Sotheby's London offered another leaf on 22 April 2015 as lot 62 of its Arts of the Islamic World sale, estimated at GBP 300,000 to 400,000, described in the saleroom's own catalogue essay as representing a manuscript whose patron "must have been a ruler of enormous wealth and ambition"; press reporting of the result gives a price of GBP 365,000 (reported elsewhere as 547,565 US dollars at the prevailing exchange rate), though the sources reviewed do not clearly distinguish whether this figure is the hammer price or the total including buyer's premium, a distinction this record cannot resolve from the material available and flags rather than guesses at. A further leaf was reported sold at Christie's in 2018 for GBP 512,750, again above estimate.

Each of these sales completes, leaf by leaf, a process of fragmentation that began when the codex first came apart, whenever and however that happened. A manuscript that was, within human memory of its makers, a small number of large bound volumes has become, in the twenty-first-century market, a recurring category of single-leaf lots, each individually authenticated, individually priced, and individually catalogued, sold to institutions and private collectors who are necessarily acquiring a fragment rather than the work. This record does not take a position on whether that market is a legitimate mechanism for placing rare material into public and accessible private collections or a continuation, by commercial means, of an original dispersal that scholars would otherwise regret; both views are held in the field, and neither is sourced to a specific citation in the material gathered for this record.

Condition across the corpus, and what is not established for this leaf

Silver tarnishes; it is a basic and well-understood chemical property of the metal, and it is repeatedly reported across the Blue Qur'an corpus that the silver verse-markers and medallions on surviving leaves have oxidised to black or near-black over the eleven or twelve centuries since they were applied, a condition documented, for instance, in the Museum of Fine Arts, Boston's cataloguing of its own leaf. Some accounts of the wider corpus also describe letters or sections that appear to have been deliberately scraped or removed from certain leaves, in a manner going beyond ordinary scribal correction, though the specific leaves and the scholarly source for this observation are not established with precision in the tertiary material reviewed for this record and the claim is flagged here for verification against a primary conservation report rather than presented as settled.

What this record cannot state is whether the Metropolitan's own leaf shows the same silver tarnishing, the same erasures, or neither. No condition report or conservation note specific to 2004.88 was identified in the sources consulted. Given how consistently tarnishing is reported elsewhere in the corpus, it would not be surprising if this leaf's silver markers show comparable oxidation, but an expectation is not a finding, and this record does not convert one into the other.

What remains unresolved

Put plainly: this is a page whose text, materials and general historical setting are well understood, and whose specific date, specific place of production, specific patron, and specific journey from a mosque library in Tunisia to a Manhattan museum are not. The dating and attribution dispute is not a matter of popular simplification obscuring a settled scholarly answer; it is a live, published, named disagreement among specialists that has run for more than a century and shows no sign of resolving from the evidence currently available. The open questions that follow are not gaps in this record's research; they are gaps in the field's knowledge, recorded as such.