Revealing the Secrets of an Early Coptic Manuscript – PML M.910

PML M.910 last folio
MS M.910, a Coptic manuscript at the Pierpont Morgan Library, NY

I recently made a model/ prototype of a Coptic manuscript from the Pierpont Morgan Library [PML M.910], containing the Acts of the Apostles, written in Sahidic Coptic some time after the fifth century [the Copts are Egyptian Christians]. The original manuscript will be scanned by digital unwrapping guru from the University of Kentucky, Brent Seales, and a team including philologist and religious scholar Paul Dilley [U. Iowa], and book conservator Maria Fredericks [Pierpont Morgan Library]. Here is the New York Times article about the project. The tripartite collaboration began early last year, out of meetings that happened in conjunction with the Mellon Sawyer Seminar I ran as a postdoc fellow, ‘Cultural and Textual Exchanges: The Manuscript Across Premodern Eurasia’, held at the University of Iowa.

Holding a model of the charred, 2,000 year old, En-Gedi scroll in my hand – a tiny thing!

Seales visited Iowa in February 2017 to present his research on digital unwrapping of the ancient Hebrew En-Gedi scroll [50-100 CE] and preliminary work on ancient Roman Herculaneum papyri which date to the mid 1st c. CE or earlier. They were both written on papyrus [a sheet material made of the reed plant] and rolled up for storage, only to be charred and fused together 2,000 years ago. These ancient charred scrolls cannot be opened without permanent damage [many have tried with the scrolls from Herculaneum’s libraries], and cannot be read with the naked eye [they are charcoal black, turd-like objects]. Seales revolutionary approach to imaging these objects is described in this 2016 NY Times article and involves scanning them using x-ray tomography, and processing the data using a complex computerized algorithm [Seales’ Volume Cartography software], then reconstructs them in a process Seales calls ‘digital unwrapping’. For a look at this process, see Seales Mellon Sawyer lecture here.


The En-Gedi scroll, digitally unwrapped []
Scanning the 2,000 year old En-Gedi scroll was an early attempt at testing this system and was fantastically successful, revealing a legible scroll section containing the oldest known copy of the Book of Leviticus! Scrolls / rolls are generally only written on one side, which allowed Seales to perfect his scanning and data processing without the complication of sorting out ink on both sides of the sheet material. The Seales team has perfected this system over the last few years, and were ready to take it to the next level – imaging a codex, which contains sheet material written on both sides – a more complicated data processing challenging.

Paul Dilley, philologist and scholar of Coptic and Manichaean manuscripts, suggested that Morgan M.910 might be a good candidate for this. Maria Fredericks, the Drue Heinz Book Conservator at the Morgan, agreed to consider the project and start the process of inquiry. The manuscript will be scanned at the Morgan in December 2017, and data processed in the following months. I will be posting more on this once the images are complete.

M.910, from the top or bottom [spine to the left] showing the wave-like distortion of the parchment quires.
Unlike the earlier charred papyrus scrolls, M.910 is written on parchment [animal skin] and contains a known text, the Acts of the Apostles [companion to the Gospel of Luke], a text authored in the late first century and copied out by the Copts. The manuscript is too fragile to be opened and is unreadable due to the water and heat damage it has suffered. The manuscript is also extremely warped creating wave-like sheets – which is common with water and heat damaged parchment. All of these factors make it a good test case for perfecting data processing for parchment codex, and working out how to digitally separate and ‘flatten’ the writing on each folio.

Before scanning the original manuscript, the Seales team needed a prototype model to make sure the manuscript would fit in the custom-fit stand and the scan would work – this is the model I made using new parchment and scribed with iron gall ink. The x-ray tomography detects the presence of iron in the iron gall ink, so I needed to source iron gall ink for my model – and learn how to write in ancient Coptic using the Sahidic Coptic alphabet!

Detail of last folio of M.910 showing Sahidic Coptic and this scribe’s particular hand
Super thin parchment skins from Jesse Meyer – a mix of goat, deer, calf.

Much had to be approximated in recreating the original manuscript, since it had been water damaged and has suffered from the heat of the Egyptian desert [I think there may still be sand in the original binding!]. The manuscript is multi-dimensional [not a perfect block], so I went with the largest dimension 12.5 x 14.5 with a 5mm spine width. Usually when I make models of historical book structures, I’m not trying to replicate exact quire structure or collation [the number of bi-folia per gathering/ quire], so this was a difficult exercise, made especially challenging because it is impossible to replicate sheet material thickness and guess how thick it will be or how it will behave once it’s gathered into quires and sewn. We ordered super thin parchment from Jesse Meyer of Pergamena, scraped and sanded to as close as possible to the thickness of .13 mm. I requested a variety of skins for the job, goat, deer and calf. Each has a different look and feel. I measured and trimmed the large parchment skins to size, trying to maximize the number of bi-folia I could get out of each skin. I needed about 13 quires [gatherings of bi-folia], so finished the book block with a thick paper when I ran out of parchment. I punched 4 sewing stations [holes for sewing] for each quire using a jig, punching cradle, and awl – and sewed each quire with a simple tacket to keep the quire together while I was scribing.

In regular scriptorium practice, the scribe would be writing on unbound sheets of parchment [which would then be collated and bound], but because of the complexity of the collation and scribal work on this project, I decided to scribe directly on to collated and tacketed quires, then resew into the binding once the scribal work was complete.

Last folio in M.910, showing the Sahidic Coptic script.

Now for scribing. To complicate an already complex task, I copied out the Acts of the Apostles, from the typescript Sahidic Coptic on the computer. These are non-calligraphic letter forms, so they do not translate to forms that can be made with quill or pen and ink. I had to ‘translate’ the typescript letterforms to a calligraphic alphabet [with my own cheat sheet], then scribe it in one line in the exact center of the manuscript. Instead of creating a full calligraphic copy of the original [a much larger task], the imaging team wanted a single line down the center, running parallel to the spine of the book, so they could test detection of the iron gall ink through one spot [the center] of the manuscript.

My cluttered scriptorium space – copying Acts of the Apostles from the typeset edition on the computer onto parchment bi-folia using iron gall ink with a variety of pens and quills.
Sahidic Coptic Alphabet []
My first attempt at scribing Sahidic Coptic using iron gall ink – approximating the height of lettering in the original. It reads: The Acts of the Apostles – The first word indeed I made, O Theophile, concerning every thing which Jesus began to do and to teach in them…

Once I got my zetas and horis, upsilons and kyimas straight, I was off and quickly got into the scribal groove. This required a zen-like balance between concentration and ease. I scribed a single line on the manuscript, with only one mis-step – which is no problem on parchment since it can be ‘erased’ by scraping away the surface of the skin in a process called palimpsesting [not possible with untreated paper!]. This was satisfying.


IMG_5032 (1)
Paul Dilley, ‘Scribe B’

Paul Dilley even tried his hand at scribing one of the 104 folios. I refer to him as ‘Scribe B’ – not as a statement about his untrained hand, but because he was the second scribe of the manuscript.

Once the scribal work was complete [took a few days], I was on to the binding. After consultation with Coptic binding expert, book conservator Julia Miller, I chose one of the sewing styles illustrated in Szirmai’s Archaeology of Medieval Bookbinding.  I had to sew the link stitch loosely in order to meet the 5mm spine width requirement [the original spine is much narrower, but the book block at its largest dimension is 5mm so I had to match that].

I sewed the book with a link stitch, a sewing style that was commonly used in Coptic book production. The link stitch has a fascinating history, since this sewing technology appears in early Christian manuscript production throughout northeast Africa [Egypt, Ethiopia], the near east [Syriac, Georgian, Byzantine, Armenian] – and all the way to the British Isles, as exemplified in the 8th-c. Stonyhurst Gospel of John. It makes sense that the technology would travel as the books traveled and were shared throughout the early Christian world, but it is also became the dominant sewing style for Islamic manuscripts, which means that craft technologies were not sequestered within religious camps – and also suggests that early book makers were drawing on pre-existing craft technologies that were applied to the production of books. Georgios Boudalis, Head of Book and Paper Conservation at the Museum of Byzantine Culture in Thessaloniki presents a case for this in his lecture and exhibit on the origins of book technologies, The Codex and Crafts in Late Antiquity [up through 8 July 2018 Bard Graduate Center Gallery, NYC]. In any case, each culture – Christian and Muslim – adapted the link stitch to their purposes.

I sewed the book like this – using one of many sewing styles for this era of Coptic manuscript. From Szirmai, The Archaeology of Medieval Bookbinding.

I sewed the book using this link stitch sewing [in two paired stations], being careful to sew loosely so that the book dimension would fit in the scanning stand the Seales team had made. Like the current state of the original, my model had no covers. I shipped the book off, and look forward to hearing about the results of the final scanning of M.910 at the Morgan. Historical bookbinding is something I do regularly to understand binding structures and book action – but it was a thrill to work on this model knowing it was part of this exciting digital imaging project!

Melissa Moreton


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s