Fragmentology in the COVID-era Classroom

The Flight into Egypt, Walters Art Museum, MS W.188, f.112r

It’s been a rough few years, friends. We have all been through so much during the pandemic, and for students and teachers the pivots and policies have been particularly difficult and frustrating. Trying to stay safe, learning to teach online, dealing with trauma and loss, getting used to social-distanced and masked pedagogy, in addition to the usual pressures of teaching and learning. It was a particular delight, then, to see my extraordinary Simmons University School of Library Science students here in Boston make their way through this semester’s course “The Medieval Manuscript from Charlemagne to Gutenberg” not only with resilience and steadfastness, but with enthusiasm, intellectual curiosity, and the joy of discovery.

Stony Brook University, “Fifty Original Leaves from Medieval Manuscripts,” no. 31

As always, the final project for my class this semester was a digital reconstruction of one of the Books of Hours dismembered by Otto Ege in the first half of the twentieth century. This year, we chose the lovely ca. 1430 Book of Hours from France whose leaves became no. 31 in the Ege portfolio, “Fifty Original Leaves from Medieval Manuscripts.” Each of my fourteen students was assigned one leaf from one of the known portfolios (such as the leaf at Stony Brook University, shown at right) to research and catalogue. They had to use online resources such as the Hypertext Book of Hours to identify the text on their leaf, and then catalogue the leaf using the Fragmentarium database. Some of the students were so enthusiastic about the project that they catalogued more than the one leaf originally assigned to them. I did some as well, so that we could work with as many leaves as possible.

One of my students took extraordinary initiative and spent hours searching the internet to try to find more leaves. She found several, including a calendar page at Dartmouth College and a miniature that was recently sold by the Manhattan Rare Book Company. Both of these are rare and important finds for an Ege manuscript. Generally, it is very difficult to definitively identify miniatures from the Ege manuscripts, since they were sold separately from the text leaves that are found in portfolios and often are framed so that the text side is not visible, making it quite difficult to determine if the miniature came from the target manuscript. In this case, however, the bookseller had reproduced the text side as well, so we could tell for sure that this miniature was from our manuscript. The miniature (below) was a depiction of King David at prayer, the opening of the Penitential Psalms section of the Book of Hours. The gold ring surrounding the vines in the lower margin is a motif that appears in other leaves, and may suggest that the book was commissioned to commemorate a marriage.

I wrote to the bookseller to ask for more information, and he informed me that the miniature had just been sold to a private collector in New York City. He contacted the owner on my behalf, and the new owner emailed me directly to let me know that he in fact owned THREE miniatures from this manuscript! In addition to David at Prayer, he had acquired miniatures of the Annunciation (Matins, Hours of the Virgin) and the Nativity (Prime, Hours of the Virgin).

In the meantime, following the trail left by Scott Gwara in his entry for this manuscript from his monograph, Otto Ege’s Manuscripts ((Cayce, SC : De Brailes Publishing, 2013), pp. 128-129), the same student tracked down a copy of Judith Oliver’s catalogue of a now-defunct collection formerly belonging to the Boston University School of Theology, where four miniatures, including the Annunciation and the Nativity, were illustrated.

Judith Oliver, Manuscripts Sacred and Secular (Boston: Endowment for Biblical Research, 1985), pp. 58-59 (no. 97, figs. 20-23).

Memphis Brooks Museum of Art, Acc. 56.27

Gwara recorded a miniature as well, this burial scene belonging to the Memphis Brooks Museum of Art that was purchased from Ege’s widow Louise in 1956. Now we had SIX miniatures to work with, two of which are untraced (the Annunciation to the Shepherds and the Adoration of the Magi, shown at right above). There are almost certainly several more miniatures that have not been located, completing the series for the Hours of the Virgin (the Visitation, the Presentation, the Flight into Egypt, and the Coronation of Virgin). There may also have been one or more miniatures illustrating the Gospel Readings that would have followed the calendar. With so much evidence, an art-historical analysis was possible. A group of students worked together to craft this stylistic description:

“Marginal rinceaux and painted line fillers, smallish acanthus leaves on miniature pages only. Margins also include gold trefoil and red, blue, and green flowers growing on the rinceaux. Rinceaux often seems to “sprout” out of the text, usually from a single gold initial or line-filler. Borders on recto and verso are mirrored for efficiency. Some leaves show a gold ring motif among the rinceaux. Miniatures with gold U-borders with flowers/ leaves in red and blue. Continental color palette: Blues, purples, jewel tone & continental design: botanical, leafy, organic. Miniature composition similar to Bedford Master Workshop (see Oliver, pp. 58-59), Dunois Master Workshop, and occasionally elements of Boucicaut Master Workshop (as suggested by Sotheby’s).”

[it is important to note that of the five other non-portfolio leaves identified by Gwara as coming from this manuscript, upon inspection only the Memphis leaf could be affiliated with Ege 31; the other four are from a different manuscript entirely]

In the end, we identified a total of thirty-seven leaves of this manuscript. After each student had catalogued their leaf/leaves in Fragmentarium, we then worked together in class to use Fragmentarium’s IIIF-sequencing functionality to digitally recreate the manuscript:

https://fragmentarium.ms/overview/F-f25b

While some students were scouring the internet for additional leaves, others were conducting codicological research. One student selected two leaves at random to catalogue – from Massey College at the University of Toronto and the Wadsworth Athenaeum in Hartford, Connecticut – that turned out to be consecutive. Several other students identified formerly-consecutive leaves, and we even found eight leaves in a row. These consecutive runs, combined with evidence such as surviving catchwords on several leaves, allowed us to partially reconstruct several quires (using an innovative and intuitive resource called VisColl). In those three quires, we identified four pairs of formerly-conjoint bifolia, highlighted in green above. Yale University’s Beinecke Library preserves a still-conjoint consecutive bifolium from the manuscript in its portfolio. Other pairs were reconstructed using textual and codicological evidence. The catchword on the verso of the leaf at Harvard University’s Houghton Library, for example (below left), indicates that this bifolium was the outermost of its quire (L9/L16 in the diagram above). In other words, each of these pairs of leaves were once attached to one another at the gutter as a single sheet of parchment, folded in half and sewn into the quire. When Ege dismembered the manuscript, he disbound the quires and split the conjoints. These pairs of leaves haven’t seen each other in nearly a century. Below, the leaf at Harvard University’s Houghton Library is virtually reunited with its former conjoint, now at the University of Saskatchewan, nearly 2,300 miles away.

We have analyzed art historical evidence and codicological evidence. What about the contents? The surviving leaves preserve portions of several sections of the manuscript: the Calendar, Hours of the Virgin, Hours of the Cross, Penitential Psalms, and the Office of the Dead. While we did not recover the critical pieces of liturgy that are usually used to determine Use (the antiphon and chapter reading for Prime and None of the Hours of the Virgin, as well as the Matins Responsories of the Office of the Dead), independent research by several students helped identify the origins of the manuscript. One student determined that the particular hymn used on the page she had been assigned seemed to indicate that the manuscript was for the Use of Paris. Another found that an atypical Psalm used at Vespers for the Office of Dead was also suggestive of Use of Paris. The few Matins responsories recovered for the Office of the Dead were consistent with Paris Use and, after a careful in-class analysis of the saints named on the calendar page, we felt we could confidently identify this manuscript as made for the Use of Paris.

That’s as far as we’ve gone so far. We spent several hours conducting provenance research in the Schoenberg Database of Manuscripts, trying to identify any pre-Ege sales of this manuscript (with help from provenance-researcher-extraordinaire Laura Cleaver and the always-helpful reference librarians at The Grolier Club Library), but we haven’t found it yet. The dimensions of the manuscript (190 x 160 (107 x 68) mm) are fairly typical, as is the number of lines (15), making it difficult to definitively identify this manuscript in an early sales record.

The semester is over, but the work continues. The linked-open data model and interoperable image sequencing reflect best digital practices. This means that if more leaves are identified, they can easily be added to the reconstruction, and if I ever do manage to find a sales record that seems to represent this manuscript when it was whole, I can update the Fragmentarium record accordingly.

This annual project accomplishes many of my goals for my students, all of whom are pursuing a Masters of Library Science: craft clean, consistent, linked data; work with digital images in a IIIF environment; analyze paleographical, art historical, and codicological evidence to determine the date and place of origin of a medieval manuscript; understand how to research and work with Books of Hours, which are among the most common genres of medieval manuscripts in North American collections; and work collaboratively. Each student’s record has its own persistent and citable URL, as does the reconstruction. The work of previous years can be found here, and for more information about other scholars who are doing reconstructions and studies of other Ege manuscripts, see this blogpost.

I hope to take on another reconstruction with Simmons students next fall, hopefully in a post-pandemic world. In the meantime, I wish you a very Merry Christmas and a Happy, and Healthy, New Year.

Manuscript Road Trip: Reverse-Engineering the Codex

The Flight into Egypt, Walters Art Museum, MS W.188, f.112r

Like all of us, I’ve been working and teaching from home for nearly a year, waiting with bated breath for the vaccine distribution, wearing my mask to keep myself and my community safe, and working exclusively from digital images of medieval manuscripts. My recent appointment as a lecturer in Latin Paleography at Yale meant that when libraries on campus opened to faculty and students I was allowed to, at long last, visit an actual library to do some research with real, not digital, medieval manuscripts. And not just any library, but the Beinecke Rare Book & Manuscript Library at Yale, where I first catalogued and studied medieval manuscripts in earnest while in graduate school there in the early 1990s.

At the Beinecke Library on January 19, 2021
(action photo by Michael Morand)

Because the pandemic is still raging in the U.S., my class meets remotely, like most university classes in the U.S. this semester. Since the Beinecke has digitized hundreds of manuscripts in recent years, I am able to teach a survey of Latin Paleography with reference to those images. And with the launch just last week of a new IIIF-compliant digital viewer at Yale, discovering, browsing, and sharing those images just got a whole lot easier. But digital images, while extremely useful for paleographers thanks to deep zoom and interoperability, aren’t sufficient if you also want to study codicology, that is, the structure of medieval manuscripts. You have to get your hands on the books. Paleography isn’t just about the letters written on the paper or parchment. It’s also about context, what Leonard Boyle called “integral paleography.” It’s about the substrate, the ink and pigments, the decoration, the format, the construction of bifolia, quires, and binding and, ultimately, the institutional context within which the manuscript was written and read, whether monastic, secular, or professional, and its journey from there-and-then to here-and-now. And so I wanted to find a way to give my students, some of whom are Zooming in from overseas, a way to engage with the three-dimensional multi-sensory experience of the medieval codex. Hence my day at the Beinecke, surveying manuscripts that will be made available to my students, on appointment, to study onsite (for my overseas students, I’ve found local collections for them to visit).

I spent time with an old friend that day, MS 699, the fifth of a five-volume set recording Augustine’s Enarrationes in Psalmos, written in Lambach, Austria, in the third-quarter of the twelfth century. Why is it an old friend? Because it played a significant role in my PhD thesis and first book, The Gottschalk Antiphonary: Music and Liturgy in Twelfth-Century Lambach. MS 699 was illustrated by the same Gottschalk who illustrated, wrote, and notated the eponymous antiphonal (a manuscript that was recycled for binding use at Lambach in the fifteenth century and that I recently reconstructed online here). Multiple scribes worked on these Augustine volumes, but the four extant volumes of the set were all illustrated by Gottschalk of Lambach:

Clockwise from top: initials by Gottschalk of Lambach in Vol. I, Vol. II, Vol. IV, and Vol. V (i.e. Beinecke MS 699) of Augustine’s Ennarationes in Psalmos (Vol. III is lost. Image from Vol. I taken from Holter 1993, p. 435, fig. 19)

MS 699 is a perfect example of a typical 12th-c. monastic manuscript in its size, layout, script, decoration, use of signatures (at the end of each quire in Roman numerals), the original incised pigskin binding, and the quality of the parchment. Don’t get me wrong, the Lambach monks were very skilled when it came to making parchment. You can easily distinguish hair-side from flesh-side, both by color and by texture. The quires are arranged according to “Gregory’s Rule,” with hair-side facing hair-side and flesh-side facing flesh-side, giving a consistent color and texture across each opening. The parchment is supple and smooth, perfectly prepared to receive the iron gall ink. But the monks were frugal. Instead of cutting around flaws in the skin to create blemish-free bifolia, they used all of the skin available to them, flaws and all. The leaves are full of holes, tears, and slits, whether from natural processes like bug bites or scarring from when the animal (presumably a sheep) was still alive, or tears inflicted by the parchmenter during scudding or stretching. These flaws are why I wanted my students to see this manuscript in person.

• 

• 

• 

• 

Parchment flaws in MS 699

One of the things I find most engaging about working with medieval manuscripts is the palpable connection they provide to the past. When I’m turning the pages of a manuscript like MS 699, I’m imagining the butcher, the parchmenter, the scribe, the artist, the binder, and hundreds of years of readers and owners who’ve handled and read the codex. I’m walking its path from 12th-century Lambach to 21st-century New Haven. While studying the manuscript in person this time, though, I was able to take it back another step, to literally start piecing together the manuscript’s first journey, from sheep to parchment to bifolia, by reverse-engineering the codex.

It’s sometimes easy to forget that parchment is an animal’s skin. It has an outside (the hair side, yellowed and rough) and an inside (the flesh side, light and smooth). The size of the codex depends on the size of the bifolia cut from the skin, which in turn depends on the size of the skin itself, which in turn depends on the size of the animal. I’ve seen giant choirbooks in which each folio is an entire skin, with the spinal column running vertically down the middle. That skin-sized single sheet might be folded in half to create one bifolium. During the 12th century, when monastic textual manuscripts tend to be around 350 x 250 mm, it was typical for a skin to be cut into four bifolia. This diagram shows how that might happen, using the bibliographic terms folio, quarto, and octavo (two sheets, four sheets, and eight sheets):

Diagram from D. S. Farnsworth,
Handmade Paper Method Cinquecento:
Renaissance Paper Textures
 (Oakland, CA: Magnolia Editions, 2018),
p. 20, fig. 2

Because there are so many flaws in these leaves, I had hoped that I might be able to find evidence of how the bifolia were cut from the skin. After carefully scanning the outer edges of each leaf, I found two whose edges fit together perfectly, showing exactly how the knife cut them from the skin. The edges of f. 110 and f. 115 fit together like puzzle pieces, which is exactly what they are.

The two bifolia (ff. 105/110 and ff. 115/116) were originally connected at their short edges, as shown below.

When together, they must have spanned the length of the skin. This means that the monks who cut the skin into bifolia were cutting four bifolia from each skin. Each leaf is 195 x 227 mm. Each bifolium is 390 x 227. The usable portion of the skin, then, measured around 780 x 454 mm, or 30 3/4 x 18 3/4 inches, a fairly typical size for a sheepskin. And now we can even partially reconstruct the very sheepskin from which these two bifolia were cut:

I haven’t yet been able to find evidence of the other two bifolia cut from this skin. Because I haven’t found the missing lower part of the hole in any other bifolium of the manuscript, I’ve oriented the two bifolia so that the flaw was close to the outer edge of the skin rather than the spine.

So now we’ve made our way backwards from 21st-century New Haven to 12th-century Lambach, all the way back to the parchmenter’s workspace at the Lambach Abbey. Now let’s see what else we can learn about book production in Lambach by moving forward in time just a bit.

The five-volume set of Augustine’s Enarrationes in Psalmos was an enormous project, requiring hundreds of bifolia in all:

Volume I (Psalms 1-50): Codex membranaceus lambacensis XVII (now Leutkirch, Fürstlich Waldburgschen Gesamtarchivs MS 5). 216 leaves = 108 bifolia = 27 skins

Volume II (Psalms 51-100): CML XVIII (sold at Christie’s in 2000, then by Les Enluminures to a private European collection). 279 leaves = 140 bifolia = 35 skins

Volume III (Psalms 101-117): CML IX (lost, number of leaves unknown)

Volume IV (Psalms 118-133): CML LXV (now Frankfurt, Stadt- und Universitätsbibliothek Lat. qu. 64). 192 leaves = 96 bifolia = 24 skins

Volume V (Psalms 134-150): CML LXIV (now Beinecke MS 699). 141 leaves = 71 bifolia = 18 skins

In all, well over 100 sheepskins. That’s a lot of sheep and a lot of labor.

These volumes are constructed like most 12th-century Germanic monastic manuscripts, in quires of four nested bifolia. The bifolia were assembled and slip-stitched together to make a quire for writing. After the manuscript was written and illustrated, the quires were stacked in order, sewn together on cords perpendicular to the spine, and secured between wooden boards covered with leather. The first step in the writing process is for a scribe to walk over to the pile of prepared parchment and pick out four bifolia to work with. You might assume that two bifolia cut from the same skin would end up together in the pile. They might, then, be grabbed by our monk as he assembled his quire and end up close together, if not consecutive, in the final codex. In this case, however, the bifolia are separated by several sheets:

Bifolium 105/110 is the second bifolium (i.e. the 2nd and 7th leaves) of the fourteenth quire of MS 699. Bifolium 115/116 is the central bifolium (i.e. the 4th and 5th leaves) of the next quire, the fifteenth. These two bifolia, cut from the same skin, ended up near one another in the codex, but in different quires. This suggests either that somewhere in the process the bifolia were separated in the pile, or that the monk who chose the parchment for his work put some judgement into selecting bifolia for each quire, rather than just grabbing the four at the top.

Reverse-engineering the codex has brought us from my hand turning the leaves in New Haven in the year 2021 past 850 years of readers to the Lambach bindery, scribe, artist, parchmenter, and butcher, all the way back to a lone sheep grazing along the River Traun near the Lambach Abbey around the year 1175.

Thanks, sheep.

Bibliography:

Holter, K. “Initialen aus einer Lambacher Handschrift des 12. Jahrhundert (Ms. 5 des Fürstlich Waldburgschen Gesamtarchivs in Schloß Zeil in Leutkirch)” Wiener Jahrbuch für Kunstgeschichte 46-47 (1993-4): 255-265, 443-436.