Digital Analysis and Online Publication of Early Kufic Quranic Manuscripts Kept in Iranian Collections

Kufic Qur’an manuscript attributed to Ali b. Abi Talib at Mashhad, Iran. Courtesy of IQNA

By Ali Aghaei*


Over the past thirty years, research on Quranic manuscripts has made considerable progress. Fundamental works on codicology, palaeography and dating of early Quranic manuscripts have been done, although mainly on those manuscripts originally came from the Western side of the Islamic World such as al-Fusṭāṭ (Old Cairo), Damascus, and Sanaa (Yemen). To have a complete image of the Quran’s textual history, however, it is necessary to include also Quranic manuscripts kept in the collections in the Eastern side of the Islamic world. Quranic manuscripts from Iraq, Iran, Afghanistan, Pakistan and Central Asia are rarely taken into account by scholars, which is mostly due to the difficulty to have access to them. In the case of Iran, as the existing though incomplete and outdated catalogues show, the number of Quranic manuscripts and fragments kept in Iranian libraries, museums and even in private collections is considerable. Based on the available catalogues as well as the information that we were able to obtain during our exploratory trips, hundreds of Quranic manuscripts and fragments before the 4th Century CE are kept in several collections in Tehran, Qom, Mashhad, Isfahan, Tabriz, Shiraz, etc. The actual number of Quranic manuscripts from the first four Islamic centuries is most likely to be significantly higher, although it has been challenging to estimate due to lack of available or updated catalogues.

Thanks to the support of the Federal Ministry of Education and Research (BMBF) in Germany, Irankoran can approach to previously inaccessible collections. This source material offers a new perspective on the textual history of the Quran, for it is the first time that Quranic manuscripts from the Eastern part of the Islamic world are going to be systematically scrutinised.

Irankoran is using and completing the work of the project Corpus Coranicum of the Berlin-Brandenburg Academy of Sciences and Humanities (BBAW). Since 2009, Corpus Coranicum has published the first digital comprehensive online catalogue of early Quranic manuscripts, mainly based on the Western part of the Islamic world kept in European collections. It has also developed a system for digital transliteration of Quranic manuscripts.

1. Digitalisation

In the course of the project Irankoran, images of Quranic manuscripts from Iranian collections, together with their metadata, are recorded in an online digital catalogue, “Bibliotheca Coranica Iranica”, according to the guidelines of the Text Encoding Initiative (TEI). This online catalogue will present a prototype that will be used by other (previously unavailable) Iranian collections and will help develop more manuscript collections online. The publication path also ensures the accessibility of the data and results obtained in the Islamic world.

2. Digital editions

In addition to the digitalisation of the Iranian of Quranic manuscripts, which makes them accessible to other scholars all over the world, the project aims at preparing digital editions of early Kufic Quranic manuscripts which cover three aspects of the text: (a) spellings of the Quranic text (rasm), (b) variant readings (qirāʾāt), and verse numbering (ʿadd al-āy) in the manuscript. In order to perform this idea, a digital database has been developed where for each word of the Quran, all three information can be recorded.

3. Transliteration

All early Kufic Quranic manuscripts are planned to be transliterated into a typical Arabic typeface (naskh). Letters are written the same way as they appear in the manuscript, usually without, though sometimes with, diacritical dots. In the transliteration, vowel signs are not represented. These digital transliterations display different levels of readability, variant spellings (in comparison to other early manuscripts and modern prints like the Cairo edition), and modifications (corrections, over-writings, additions, and erasures) in the manuscript. To carry this out, Irankoran is following the transliteration system developed by the project Corpus Coranicum in a modified, adapted way in the presentation of modifications in the manuscript.

3. Variant readings

Irankoran also confronts the statements of the Islamic scholarly literature on the variant readings of the Quranic text with the readings as they appear in the early manuscripts. The main question is to what extent the seven readings of the Quran, canonised by the Baghdadian scholar Ibn Muǧāhid (died 936), can assert themselves in early Quranic manuscripts some of which predate his time.

4. Verse numbering

Almost all Quranic manuscripts from early period represent special signs for separating and sometimes numbering the verses. Since we have differences in numbering the Quranic verses which traditionally assumed to be originated from different locals and regions, one would expect that these differences appear in the Quranic manuscript as well. Again, the goal here is to assess the Islamic traditions on verse numbering through registration of the data in each manuscript into the database, and then a detailed survey of them.

5. Dating manuscripts

Dating of the manuscript plays a central role in understanding the history of the Quran. Unfortunately, early Quranic manuscripts usually have no colophon, most probably because the first and/or the last pages of the manuscript are usually exposed to damages. Of course, sometimes one finds pseudo-colophons which later inserted into the last page of manuscripts claiming attribution to the Shīʿī Imāms. Interestingly on the territory of the Islamic Republic of Iran, at least one codex is kept that is attributed to Caliph ʿUthmān, in the city of Negel in Kurdestan, which seems very difficult to have access.

Since the paleographical classification of script styles can only provide a relative chronology for early Quranic manuscripts, dating based on radiocarbon analysis (C-14 analysis) seems to be useful. The radiocarbon measurements are carried out in cooperation with the Laboratory of Ion Beam Physics Isotope Laboratory (ETH Zurich).

Concluding Remarks

Irankoran is an interdisciplinary research project for it incorporates the philological analysis of the Quranic manuscripts and scientific dating through C-14 analysis. The results of Irankoran will be published online. Because of its digital dimension, the main goal is to encourage its academic audience of the value of digital publication and to convince the authorities in libraries and collections who still have problems to accept the idea, as it is the case in Europe, in the United States of America, and also often in the Middle East, to put their heritage online.

*Dr. Ali Aghaei is a research fellow at the Berlin-Brandenburg Academy of Sciences and Humanities (BBAW) and the head of IranKoran project funded by the Federal Ministry of Education and Research (BMBF) in Germany.

The Discovery of an 8th-Century Qur’an Written Over a Coptic Biblical Text: An Interview with Eléonore Cellard and Catherine Louis

[This interview is also available in Arabic and French]

Edited by Ahmed Shaker and Abderrahmane Ettousy


In April 2018, Christie’s Auction House offered an early Quranic manuscript for sale as part of its Indian and Islamic World auction, which included oriental carpets and rugs. The Quranic fragment consists of nine miniature folios, the largest being 12.7 x 11.1 cm, and contains partial verses from Sūrat al-Māʼida (verses 40-58; 69-76; 83-88; 116-120) and Sūrat al-Anʻām (Verses 1-9). It is written in late Hijazi or early Kufic style on vertical parchment and dated to the 2nd-century A.H./8th-century C.E. 

Soon after the release of the auction, a French scholar named Eléonore Cellard was able to discover what appears to be a fainted Coptic text laying behind the Quranic text, which is thought to be from the Christian Old Testament. In this sense, it’s a palimpsest; that is a piece of writing material on which later writing has been superimposed on effaced earlier writing. Covering two distinct theological texts, the first layer of the palimpsest (inferior) is Coptic, while the second layer (superior) is the Arabic Quran.

The discovery is considered remarkable, not only because Quranic palimpsests are rare, but also because it is the first time we encounter an example of a Coptic fragment being washed off to make a place for the Qur’an. Moreover, on April 26th, 2018, Christie’s Copto-Quranic palimpsest has reached an outstanding result—in terms of Islamic art auction sales—when the nine fragments fetched the sum of £596,790, that is five times its estimated price (£80,000-£120,000).

Christie’s Copto-Quranic palimpsest being sold on April 26th, 2018, for £596,790. Source: ChristiesInc/Twitter

To highlight the origins, codicological peculiarities, textual readings, and overall significance of Christie’s Copto-Quranic palimpsest, we had the opportunity to speak to Eléonore Cellard and Catherin Louis, respectively. Cellard is a post-doctorate at the College de France in Paris, where she majors in the study of the textual transmission of the Qur’an in the formative period of Islam, based on the oldest and finest Quranic manuscripts available to her. She is a student of the renowned codicologist François Déroche, and has recently published a book with Brill in 2017, entitled Codex Amrensis 1—a reproduction of an 8th-century Qur’anic fragment, scattered at four different institutions in Europe and the Middle East. As for Catherine Louis, she is a research fellow at the Institute of Research and History of Texts at the Centre national de la recherche scientifique (CNRS) in Paris, where she specializes in the stud of early Egyptian Christianity, cataloguing and editing unpublished Coptic manuscript texts, and examining book’s structure techniques in Egypt around that time. Her contributions can be found at Cahiers de la Bibliothèque copte—a peer-reviewed journal, in which Louis serves as an editor along with her colleague Anne Boud’hors. 

copto-coran palimpsest
Christie’s Copto-Quranic palimpsest. In this photograph, You can fairly sight the fainted Coptic text underlying the Quranic text. Source: Christie’s

The Interview

Quranic Manuscripts Studies Blog (QMSB): How did you find out about this palimpsest?

Eleonore Cellard (EC): I discovered this palimpsest by flipping through the Christie’s Islamic Art Sale Catalogue of April 26th, 2018, which was made available online. One of the lots, entitled “nine folios from a rare Late Hijazi or early Kufic small square Qur’an” has fascinated me in particular, for the research I conduct on ancient Qur’an manuscripts dated to the 2nd-century A.H./8th-century C.E. Since the images were in a very good quality, I was able to make enlargements to make sure what was there is not simply traces of wear of the parchment but of an actual erased writing.

However, this discovery is not coincidental. For several months now, I have been working on other palimpsests, attempting first to understand in which cultural and material contexts they appeared in, and what were the techniques used to reveal these erasures. In a way, one can say that my gaze was trained to observe what it was beyond visible writing. Maybe, without this preceding work, the erased text would have escaped my eyes, just as it had escaped the eye of its former owner and the eyes of Christie’s experts. And I must say that, at the time, I thought my imagination had played me a trick.

QMSB: What’s the significance of this discovery?

EC: In my opinion, this document profoundly enriches our knowledge of the handwritten transmission of the Quran, and more generally of the texts in the book culture of the Orient of Late Antiquity. Two elements seem to be particularly important in this perspective.

Firstly, thanks to the erased Coptic script which can be easily located in Egypt, this document gives us a clue to the places of production of the ancient Qur’anic manuscripts. This is a very valuable clue because we must remember that we have no such direct information in the manuscripts themselves. They do not contain any note from before the middle of the 3rd-century AH/9th-century C.E., or even later. In addition to that, most of these manuscripts— which were discovered starting from the 17th-century towards the 18th-century—were collected in the mosques of large urban centers; that may not be related to the actual locations in which the manuscripts were copied. However, this particular palimpsest confirms that Egypt has been a place of business for the textual transmission of the Quran, at least from the mid-8th century.

Secondly, this new palimpsest completes our understanding of the scribal practices of the Quran. Today, there are countless witnesses of palimpsests preserved in the Jewish and Christian book cultures of East and West. To give just one example, more than 20% of identified New Testament Lectionaries in uppercase Greek, are palimpsests. Conversely, we have only a few palimpsests for the Quran (they represent approximately less than 0.03% of the total manuscripts). The study of these exceptional witnesses will enable us to understand the matters and contexts in which such artifacts have been assembled. In this case, the Copto-Quranic palimpsest reflects certain economical constraints that fit well with the book culture of multicultural Egypt in the 8th century.

QMSB: In addition to Christie’s Copto-Qur’anic palimpsest, do we have further examples of Quranic palimpsests in presence today?

EC: Currently, we know 4 other Quranic palimpsests, each one of them were produced in different contexts—aside from Mingana-Lewis palimpsest of Cambridge University Library [Ms. Or. 1287]. In contrast to the other three, the Mingana-Lewis palimpsest contains folios, originally Quranic that has been erased to be reused in a Christian context. The two other palimpsests were discovered in the collection of Ṣan’ā’ in Yemen[1]. Both of the latter palimpsests are entirely Qur’anic; whether it is the erased text or the one that has been rewritten over it.

The third palimpsest was identified through a replica published in the early 20th century[2]. It is a folio from the collection of the Great Mosque of Damascus, with a Greek text being erased during the 3rd-century AH/9th-century CE to transcribe a Quranic text.

However, we must remember that we do not yet have an overview of all the collections. Other palimpsests may exist in the collections found in the great mosques of Damascus or Ṣan’ā. Others, again, may have circulated outside these collections in environments where access to writing materials was more difficult.

In conclusion, the extreme rarity of this practice of parchment reuse in the Quranic context is, perhaps, due to lack of conservation’s procedures.

QMSB: What was your first impression when you examined the palimpsest?

Catherine Louis (CL): I was initially very excited by this beautiful, completely unexpected discovery. The underlying writing showed signs of seniority, and the fact that it was erased for the fragments to be reused for copying a Quran, is unique. In addition, some Coptic fragments, the first to have been identified, showed we were dealing with at least two biblical passages, which made the lot even more interesting. The difficulties of reading the underlying text made the whole thing more exciting (it was a kind of challenge).

QMSB: Is the inferior Coptic text legible? If yes, what does it say?

CL: Two of the fragments have a rather readable view, and were identified shortly after their discovery, by A. Boud’hors (CNRS-IHT, Paris) and A. Suciu (Akademie der Wissenschaften, Göttingen) as containing excerpts from Deuteronomy and Isaiah. Then two other fragments which also contain excerpts from Deuteronomy and Isaiah were identified, but they are much less clearly legible. Regarding the remaining of the five fragments, I suspect that two of them also contains biblical excerpts, but the text is not legible enough to determine with certainty which passage it is. We still have three fragments on which we, unfortunately, only decipher a few letters that do not allow for the moment to identify them.

QMSB: Is it possible yet to date the lower Coptic text?

CL: Yes, approximately. We know that these fragments came from a Coptic manuscript which was probably of a rather small size. It is written in the so-called “biblical” capital (from the name of the uncial used in many ancient Greek Bible manuscripts). This writing also corresponds to what is found in various Coptic manuscripts originating in Egypt and were datable, quite safely, to the 6th-7th centuries; as they were found during excavations alongside other dated documents. These indices allow to suppose that these fragments were copied around the seventh century—maybe even in the sixth century—but at the moment, we can not provide a precise dating due to lack of specific indications as to where these fragments were made and the circumstances surrounding its appearance in the antique markets.

QMSB: Where does the palimpsest probably come from?

CL: it is not certain that the Coptic text and the Quranic text were copied in the same place. Some palimpsests containing an underlying text in Coptic have indeed travelled, and their underlying text may have been copied elsewhere before the manuscript or part of it was acquired in Egypt and other copies of it were reproduced; they may have been written in other languages.

The Coptic manuscript with its own characteristics (size, paleography, and language) is likely to come from the Middle and Upper Egypt, where we have obtained many similar manuscripts (in the areas from Abydos to Hermopolis, at least), but it remains difficult to further clarify this issue at this time.

EC: As I said above, the material and textual characteristics of Quranic writing do not allow us to associate it with a particular regional practice. The writing used follows a model disseminated in all the different urban centers. Some textual features, which have been added later, refer to a tradition of Medinan reading; knowing that the latter has been widely disseminated in Egypt. This observation also concords with a set of manuscripts, discovered in the Mosque of ʿAmr in Fusṭāṭ, which have the same characteristics.

QMSB: What are your further studies on this palimpsest?

CL: It seems to me that the most important work would be to read and identify the Coptic fragments of this lot, which are still remained without identification. This would require good photographs (ideally multispectral photographs) that would better visualize the underlying text. Without this step, attempts to identify certain fragments are likely to fail. However, there remains a possibility that some fragments do not come from the same book. Without good quality photographs, it remains impossible to say that all the Coptic fragments used in the Quranic codex have been biblical. This is an important question that can only be solved by careful examination of the lower layer of these fragments, with the help of the modern technologies at our disposal. We can, therefore, consider that these fragments have not yet revealed all of their mysteries, and that there is still due work to get the best out of it, but this work can only be done as it become possible to access the originals again.

EC: With the reconstructing of the lower text, using multispectral imaging, we aim at several things: to precisely locate the original manuscript, and also to understand the process of transformation of the Coptic manuscript into a Quranic manuscript.

Unfortunately, this document is no longer available since the sale of Christie’s. Despite our insistence on obtaining multispectral images of the fragments, the purchaser of the object never contacted us about it. We hope that the study we are trying to conduct and publish will change the situation.

[1] Cellard is referring to two Quranic palimpsests, currently kept at the Dar al-makhṭūṭāt in Yemen; explicitly DAM 01-27.1 and DAM 18-?.a. The first is dated to the 1st/2nd-century A.H, while the second is from the 3rd/4th-century A.H.

[2] A photograph of this folio was published in 1908 by the Berlin State Library. The 1908 edition also included photographs of oriental manuscripts, written in a variety of languages, and were previously stored at the Great Mosque of Damascus. See Photographien von ausgewählten Fragmenten aus der Omayyaden-Moschee in Damaskus in verschiedenen Sprachen. Berlin, Staatsbibliothek, Mss. simulata orientalia 6, 1908. pl.3a.

We thank E. Cellard for providing us with this information.

Auctioned Leaf of Ṣan‘ā’ Palimpsest, Possibly Acquired by a Turkish Private Collector

By Ahmed Shaker

After the discovery of a missing folio from Ṣan‘ā’ Palimpsest at the Louvre Abu Dhabi last week, it seems I was lucky enough to find more interesting stuff in this respect. As you remember, the Palimpsest has now 81 folios, preserved mainly at Dar al-Makhṭūṭāt and the Eastern Library in Yemen, in addition to 4 folios, which were stolen and sold at European auction houses like Christie’s, Sotheby’s, and Bonham’s starting in 1992 through 2008. One of these 4 folios was sold at Christie’s on April 8, 2008. This auction was unprecedented, as the Quranic leaf broke the world auction record for any Islamic manuscript by fetching the sum of £2,484,500. In their Fine Books’ Annual Report on the Top Auction Sales of 2008, Ian McKay and P. Scott Brown observes the following:

“High prices have been paid in the past for illustrated and highly decorative Eastern manuscripts—for groups or even individual leaves from the great Persian epic, the Shahnama, and for purely calligraphic manuscripts of the Qur’an,”


“In very recent times, however, there has been a marked increase in prices for very early, often single leaf specimens of Islamic calligraphy.”

If you have purchased a single Quranic leaf for 3 million US dollars–that is 20 times its original price–do you really want people to know your real identity? Of course not. Moreover, auction houses withheld names. They refer to buyers as either ‘anonymous’, or ‘private’ collectors. In this regard, I speculate this folio went to a Turkish private collector. This blogpost will try to figure this out after highlighting a few introductory points.

The Lost Folios 

Ṣan‘ā’ Palimpsest and other fragments, mostly Quranic, were discovered at the false ceiling of the Great Mosque in Ṣan‘ā’ in 1972. It was Qāḍī Ismā’īl ibn al-Akwa’ (d.2008), then the president of the Yemeni Antiquities Authority, who arranged to store the newly-discovered fragments, which filled 20 potato sacks, at the National Museum. But, as he saw the number of parchments was decreasing, he decided to move it back to the Western Library (al-Maktabah al-Gharbīyah), and lastly to Dar al-Makhṭūṭāt where it remains since. Evidently, some official workers were stealing the loose leaves and selling it to foreign antique dealers. In Yemen, such prohibited activities are common. Moreover, as the manuscripts were unbound, it was easier for someone in charge to take off the leaves without anyone noticing them. After all, if you don’t watch closely and count the leaves yourself, you will never know what has been missing. This explains, partially, how some folios of the Palimpsest went missing, consequently finding its way through European auction houses in the 1990s through 2008.

Fortunately, in the winter of 1996, German art historian Hans-Caspar Graf von Bothmer had microfilmed more than 35,000 images of the manuscripts discovered in 1972. This suggests that if any Quranic folio went missing after 1996, it would likely to be found in the microfilm of Von Bothmer. But who can access this microfilm? A very limited people indeed. As I heard from David Hollenberg, director of The Yemeni Manuscript Digitization Initiative, the Germans had made an agreement with the Yemenis, that they won’t release the photographs to the public. “They did not want the microfilm having the effect of diminishing the importance of visiting and doing research of the original,” said Hollenberg. The purpose behind making this microfilm was “merely a safeguard in case anything happened to the original,” he said.

Christie’s 2008 folio

Christie’s 2008 folio, recto and verso

The folio was identified to be part of Ṣan‘ā’ Palimpsest in 2008. It’s thought to be from the mid-7th century CE, with a possible linkage to Medina. Written in Hijazi script, the folio has parts of al-Nisāʼ [4:171-176] and al-Māʼida [5:1-10]. It measures 36.3 x 28cm, with 28-30 lines to the page. Verses are separated by clusters of short sepia dashes, one medallion of red, green, and sepia. There are wear around the edges and some minor areas of holing. On April 8, 2008, the folio was sold to a ‘private’ collector for £2,484,500. Similarly, another folio of the Palimpsest, which has parts of al-Baqarah, was sold by Christie’s on May 1, 2001, for £163,250.

A Possible Turkish Owner 

Christie’s 2008 (verso), uploaded by a Turkish user on Pinterest. Probably photographed in 2017

Back to the original question of this post: How do I possibly know the Christie’s 2008 folio was acquired by a Turkish private collector? I have found a photograph on Pinterest, which bears the verso side of Christie’s 2008 folio. The user who uploaded the photograph is Turkish. He added the following description to it: “We own this Islamic Quran page. We think it is very early on vellum.”  When you navigate through his pins on Pinterest, you will realize he owes special interest to Islamic calligraphy. Although you can’t know precisely the date of the photograph, his ‘Kufic’ pin itself was updated last year. My scenario would be that the Christie’s 2008 folio was acquired by a Turkish private collector, whom he likely uploaded on Pinterest with the aim of boasting. We may never know his identity, but I have a feeling that the Quranic folio would appear in a collection or catalogue in Turkey within the forthcoming years.