The Binding of Codex Sinaiticus

After seeing Sinaiticus at the British Library this weekend, I immediately realized that Cockerell was not only responsible for its conservation, but had re-bound it in precisely the same way. Sewn on meeting guards to cords laced into English oak boards wrapped with pale alum-tawed pigskin. He even used the same "thorn" rolling stamp as was used on Bezae.

But after digging a little deeper, I now wonder which Cockerell was responsible for the current binding of Sinaiticus. Douglas Cockerell, author of Scribes and Correctors of Codex Sinaiticus, is credited with the conservation of Sinaiticus in 1935. It is also recorded that his son, Sydney Cockerell, aided him in this restoration before taking over for his retiring father at the Central School of Arts and Crafts in London. And then, as I can only assume, Sydney Cockerell bound Codex Bezae precisely the same way almost 30 years later. (Anyone interested in the genealogy of English bookbinders, click here.) Fortunately, I recently met someone who worked under Cockerell the Younger, perhaps he can solve this mystery. It is intriguing to think that Cockerell the Younger would have used the very same "thorn" rolling stamp on the binding of Sinaiticus that was used on Bezae. (Or else this wins as the most irrelevant detail in blogging on New Testament MSS for 2007.)

On the history of the binding of Sinaiticus, there is this interesting tidbit from one of Skeat's last articles. The Cockerell he refers to here is Douglas Cockerell, citing the relevant section of Scribes and Correctors. The second binding of the manuscript, previous to Cockerell's, was done by monks at St. Catherine's zealously following Tischendorf's instructions to carefully preserve anything that looked like the 43 leaves he was permitted to take with him:

"The monks got as far as sewing the leaves into quires, and then sewing the quires together. They then attached to the back two broad bands which were evidently intended to be attached to the binding boards. By this stage, however, the volume had become very out of shape. As Cockerell describes it, 'While the fore-edge is roughly square, the spine is badly out of shape. When the spine is straightened up, as in the new binding, the fore-edge becomes irregular. It is quite possible that this later binding was never actually completed. The sewing threads were deliberately cut from the bands, perhaps with a view to a fresh start.' However, by this time the monks seemed to have realised that their primary objective, of securing the leaves against future loss, had been obtained, and they took no further action."

(T.C. Skeat "The Last Chapter in the History of Codex Sinaiticus" NT 42.4 (2000): 314)


The Binding of Codex Bezae

“It is indeed defective and not copied correctly enough right from the beginning, nor is it in good enough condition as it should be, as may be seen from several parts in a different hand which have been inserted, and the barbarous notes of some ignorant old Greek monk added everywhere.”

- T. Bezae (in a letter to Cambridge University Library about the codex).

I was recently afforded a day of consultation with Codex Bezae at the Cambridge University Library. The Manuscripts Room of this library is one of the friendliest and most accommodating I have yet encountered. Much of my understanding of the most recent re-binding of Bezae is indebted to the head of the conservation department there who walked me through its construction and compared it to more recent binding procedures that are distinctly different.

The Current Binding.

Part of the catalog item is a two page typewritten treatment report by Stan Cockerell, who rebound Bezae in the 60’s. The treatment report chronicles the status of the manuscript at that time and its conservation process. Cockerell removed the codex from its 19th century binding and found that most of the outside fold of each folio had been severely damaged, as well as many of the interior folio leaves. Additional to this were seven saw cuts in the spine done at the time of its 19th century rebind. Such damage is consistent with a rebind in which the spine has been scraped of any original glue, and then each signature (gathering, quire) is re-sewn at five original points along with two rows of kettle stitching on the top and bottom of the spine. The conservation treatment here was similar to that of Siniaticus in that every leaf was flattened on clips. But unlike Siniaticus, “the number of repairs runs into several thousand.” Page tears and splits were either sewn or repaired with PVA adhesive along with a few different types of toned vellum and paper. Damaged folio backs were guarded with linen. (Bezae is a rather damaged text, lots of corrosive ink, ox gall stains, and tearing.) After all of these repairs were completed, the repaired signatures were sewn to vellum meeting guards with linen thread spaced every ½ inch. A meeting guard is simply a folded piece of material onto which a signature is sewn, they “meet” each other along their respective folded edges: m.g. > < sig. These meeting guards are then bound just like signatures along the spine, in this case on five large cords. The volumes were split at folios 175 and 176 to match the facsimile edition, and then laced onto English oak boards wrapped to about four inches or so in a pale alum-tawed pigskin. Cockerell said he wanted the codices to have a “mellow” feel.

These days the binding process would have been different in that rather than a flat-back binding with meeting guards, the entire codex would have been reconstructed and guarded folio by folio, re-sewn, and then bound on a hollow-back binding. Both are secure, lay flat, virtually adhesive free bindings. But the latter method tends to preserve the original codex construction and provide better visual access to the spine.

The Original Binding.

My mission in this consultation was to use Bezae as a test case to see if it were possible to find traces of the original binding structure of the earliest major rebound biblical codices. I was looking for hints of its original structure that had survived at least two re-bindings, one far more professional than its predecessor. These would be any of the following: Original alpha-numerical binding aids along the tailpiece of the codex on each signature that would have helped in the original binding process. Any tooling marks or signs of extant original adhesive. Folios on which it was possible to see a definitive and consistent sewing pattern that would most probably have been original.

1. Unfortunately, I didn’t find too much. In his monograph on Bezae, Parker enumerates 18 or 19 different hands in the manuscript, one of which is responsible for the Greek alpha-numeric numbering of each signature on the lower inner margin of the last page of each. Additional to this are three different sets of numbering and markings from binders or curators of the manuscript. One of these is in ink Roman numerals at the top fore-edge of each verso and recto. The other two are sets of ordinal numerals written in pencil and ink. I couldn’t find any catalog data on the codex that would link any of these much later hands to a particular time other than a possible match between the hand on one of the sets of folio numbering and a few sentences self-dated to 1898 on a set of folio maps in the box of scraps and whatnots that are part of the Bezae catalog item (see below). Such data would be somewhat irrelevant, however, the original folio numbering is consistent with Roman practice, albeit in Greek script (as they fall on the Greek side of the manuscript).

2. As far as tooling marks are concerned, Bezae is a very distinct example of the ruling of text-blocks. All the original adhesive had been removed excepting the possibility that one of the vellum scraps in an envelop with the codex had some adhesive (definitely not from Cockerell’s binding) that could either be from its 19th century or original binding. Other than that, the current codex has been so drastically cleaned and guarded that it would be difficult to find any tooling marks on the spine or endsheets.

3. Due to the binding previous to Cockerell’s, the sewing pattern is completely indeterminate. I got as close as Cockerell, who surmised that the original codex was bound on five cords, probably proportionally spaced as they are now (as the original codex was a bit larger than it currently is). And such a sewing pattern would be consistent with early European binding. The saw cuts on each folio have left a set of elongated holes that obscure any original manufactural marks.

As a result, there isn’t too much that one can say about the binding of Bezae. This consultation did, however, sharpen my attention to details that could be indication of original binding structure should all the right circumstances be in place. If Bezae had been rebound according to contemporary re-binding procedures, and had not been mangled by a 19th century binder, then it would be possible to surmise how the original codex had been constructed. Even if each signature had been fairly badly damaged, it would really only take three or four solid leads to detail its original sewing pattern with some certainty. Furthermore, extant endsheets and spine lining material would only enhance such conjecture, giving us further clues as to their method of attachment to actual covers. Bezae does not afford such data, but others may.

The “Bezae Box.”

There was a box of material that can be seen along with Bezae. Its contents are as following:

1. A few exhibition labels, on one of which E.A. Lowe is cited as sourcing the codex in “a near-East centre” (Egypt or Palestine) in the early fifth-century. A different label claims “Sicily” as its provenance.
2. Cockerell’s treatment reports along with a remarkable set of photos of the manuscript in various states of repair. A few of these were published in Parker’s monograph.
3. Two envelops of vellum scraps presumably leftover from the re-binding.
4. The Corpus Christianorum edition of the Vulgate text of the supplemental pages of Bezae by J. Mizzi.
5. Several sets of folio maps (a handwritten diagram of each folio) that had been checked and signed by four different librarians (1898, 1949, 1952, 1962) claiming “all here.”
6. Blank reconstructions of each folio, probably for re-binding practice.


Winchester Gospels

I can't seem to find much background on this 11th century gospel codex (The Winchester Gospels) on display in the Wren Library at Trinity College, but it is very well preserved right next to a 9th century copy of Paul's epistles and directly across from one of Wittgenstein's actual notebooks and the first, handwritten copy of Milne's Winnie the Pooh:



Also next to it in the case is this lovely gem from a manuscript of the writings of St. Jerome. I have forgotton the date, but 10-11th? Regardless, this man teaching a bear to speak is a memorable image of medieval literacy. Though it is far later than St. Jerome, it jogged my memory as to his peculiar desire to make texts more readily understandable and readable through colemetric arrangement ("Per cola et commata," in basic stichometric sense-units rather than scripta continua). There is an interesting brief discussion of this as it relates to NT textual criticism a few pages into an old Kirsopp and Lake article on the text of Acts. (Kirsopp and Silva Lake, "The Acts of the Apostles" JBL 53/1 (1934): 34-45.)


Rebinding Codex Claromontanus

While at Tyndale House I have been given the opportunity to do some repair work in the library. Fortunately, Tyndale House is one of the most well preserved libraries I have seen, mostly due to the fact that none of these books circulate. But I have been able to pluck a number of needy volumes from the shelves and give them what attention I can with so few materials and tools on hand. The above cover and spine is from F.F. Bruce's copy of Codex Claromontanus. As is characteristic of German books from this period, the text-block itself is in excellent condition. But the covers and spine were completely removed from the text.

Ordinarily, I would have recased the entire book, giving it a new cover, spine, and endsheets. But as I don't have any presses or cover material here, I opted to rebuild the book from the inside out. This above picture is of the spine interior. Note the spine card material, which was just some scrap pulp-based paper that had been lying around the bindery in 1852 (when the book was published). The first step here was to seperate the covers from the spine, leaving as much original material intact as possible. I left the spine itself attached to one of the boards, as this gutter had not completely disintegrated through use. I then scraped all the old padding from the spine while being careful not to damage the first few layers of each signature.

The second step was to retrace the steps of the original binding, and create enough space to reproduce it with new materials. This book was very sturdily bound on five flax threads that were then laced into the boards. You can see the remnant of one of these threads in the above photo. (Click for more detail.) My plan was to adhere a piece of mull to the spine with overlaps that would correspond to the distance I peeled back the covers on each board. This would replicate the threads that had originally been holding the covers on the book, but much more permanently.

Step three was the process of reattaching the covers to the book. I simply glued the one and a half inch flap of mull to the back edge of each board, and then glued the flap of cover material I had peeled back over this strip of mull. In the above picture, you can see this finished step on the first cover. The spine is rebuilt, and the mull flows neatly onto the back edge of the cover board.

Here the other cover is attached the same way. This was much trickier, as the spine material was still attached to this board. The final steps will involve affixing the left edge of the spine to the left cover, adding some archival grade paper to the end-sheet gutter margins, and then putting the book back on the shelf. I apologize for this contemporary diversion from the concerns of ancient books, but the methods you see here aren't that different than the sorts of repair that would have occured from the earliest days of the book.


Coptic Bookbinding - Between Book Historians and Biblical Studies

Images of Coptic Bookbinding

Contemporary bookbinders often refer to any non-adhesive bookbinding in which unsupported stitching across the signatures is laced directly into the covers as a "Coptic Binding." ("Unsupported" simply means that the thread passes through each signature and is linked to the identical stitch on the previous signature, rather than being sewn over a cord or a thong. See below for images.) There are a number of tutorials on the web that demonstrate this technique, which can become complicated when sewing on more than three tapes or cords, with more than one needle, or with actual ancient materials (sewing up papyrus is far more difficult than sewing on modern paper). This is a good basic tutorial on the sewing method, but this tutorial takes one much closer the sorts of limp bindings we often see in Nag Hammadi and similar finds.

Many images of contemporary "Coptic" binding can be found online, but one of the more interesting collections of such bindings can be seen in the Special Collections section of the Princeton University Library website. Though most of these images are of relatively late books, the section on Early Codex and Coptic Sewing and Early European Sewing and Board Attachment are illustrative of the influence the earliest Coptic bindings had on book technology far into the modern era. The techniques in the "Early European Sewing" section are what characterize "modern" hand-bookbinding, the key difference from their Coptic predecessor being that European bindings introduced a cord or tape onto which each signature was sewn. (If you click on each image, it will take you to an enlarged version with an extremely handy magnifying tool. If only all papyrology sites had the same coding!) You can clearly see the differences in stitching between this 17th century Ethiopic text (though it is late, it is a very clear example of the Coptic stitch) and this 15th century English text. The former "Coptic" stitch would be expected in very early Christian manuscripts, and through a scrutiny of these images one can imaginatively retrace the steps of some of our earliest Christian artifacts.

Articles on Coptic Bookbinding

Such materials provide an interesting means of reflection on early Christian manuscripts that runs parallel to, and often complete independent of, the related discussion in Biblical Studies. But additional to archives of images, there are also a number of articles on these trajectories in the technology of books written by historians and practioners of book-bindings that highlight an interest direction of scholarship that Biblical manuscript studies has rarely addressed. A report from a 2003 Guild of Bookworkers seminar summarizes a paper given by Dr. John Sharpe on the History of the Early Codex. This sort of research provides some material depth to a number of questions that have puzzled papryologists and historians of early Christian literature for quite some time:

"The model for the early codex was a wooden tablet. Single-quire text blocks of papyrus bound in simple leather covers with a fore edge flap and a strap for fastening resemble the proportions and shape of many wooden tablets that survived as archeological evidence of the Early Christian Era. The further development of the book structure resulted in the necessity of multi-quire text blocks. Papyrus failed when used for the multi-section structures. Parchment became the material of choice... Covers of wooden boards and leather spines were made separately and attached as “case bindings”. Best examples of early Coptic bindings are books from the Nag Hammadi find (1945, Egypt). A similar single-quire Coptic codex is the Berlin 8502. Leather on its cover is similar to the one used in the Dead Sea Scrolls. It neither resembles tanned leather, nor does it resemble parchment or alum tawed skin. Single-quire text block was not the only codex form during the third and fourth centuries AD. Out of surviving codices from before 400 AD the majority is of the multi-quire type. Early bindings were quite elaborate. They used fringed leather strips that were laced through numerous holes in the board and used as wrapping pieces finished with bone pegs as fasteners. Those early multi-quire text blocks were sewn link stitch and bound in wooden boards with leather spine structures that were assembled separately from the text block and attached in a manner of case bindings. Bindings found under the ruins of the Monastery of Apa Jeremiah near Saqqara (1920) were dismantled and partially described by Lamacraft in 1939."

The lack of any consistent method in our earliest extant Coptic bindings suggests that this was still a period of experimentation with this format that expresses itself in the use of different stitching patterns and finishes. It was only with the eventual hegemony of parchment that any sort of standardization set in, as it made the fairly regular stitching of multi-quire text blocks possible. I always assumed on the basis of my very limited first hand knowledge of early Christian papyri that our earliest bindings were "case-bindings," which simply means that the top and bottom leaf of the text block were attached directly to the interior of a leather or stiff paper cover. Sharpe's undocumented point here both confirms my suspicion and raises the question of how related contemporary "Coptic bookinding" really is to its ancestor. Many contemporay definitions of Coptic bookbinding will tell you that the stitching is to be laced directly into the cover (as in the Ethiopic cover above), whereas many early Coptic bindings actually skipped this step and just had the covers glued directly to the endsheets of a textblock. I have always assumed that this latter method would have been the original method practiced by early Christians as it is the least technical and least time consuming process.

Another interesting article from the Book and Paper Group Annual covers the Adoption of the Codex. Here the "African" model of the early codex is described:

"The model takes two construction types with regard to assembly. The first type is a single quire papyrus codex which can be compared with the single quire parchment notebook from Roman examples. The single quire papyrus codex is associated with a portfolio or wallet like cover made of leather. The text was restitched directly though the cover with interior leather stays positioned in the inner fold to cushion the papyrus from the cinch of the sewing. The cover was frequently reinforced inside with a cartonnage of papyrus. Then, with its protective cover flaps closed and tied with thong, the text was well protected for travel.

The second type was used for binding multiple quires. Each quire was stitched from quire to quire forming chains of stitches across the back of the text. A stitch passing through the inner fold of the gathering would pass to the outer fold connecting the separate folios together. The stitch would then drop down to pick up a previous exterior stitch and climb to enter the next gathering. Quire by quire the book would be constructed. Cover boards, of wood or skin, were also sewn to the text. This "sewn board" book would then be covered with pasted leather and perhaps provided with a second, outer leather case for travel. The result was a secure text block with a docile, flat opening provided by the pliant stitch chains. For the following discussion the African bookbinding model combines together the single and multiple quire type. This obscures the possibility that the sewn single gathering codex may be more associated with the genre of letters folded for travel. The sewn multiple gathering would then be used to accommodate assemblies of letters or larger compilations or, eventually whole Gospels."

He next makes a number of interesting points made about the possible technological backgrounds to the codex. While I haven't ever encountered these in Biblical Studies oriented literature on early Christian manuscripts, such points are frequent in literature on this period by historians of bookbinding. Among many others, these two paragraphs are especially intriguing:

The resources of many crafts must have been assimilated into early codex bookbinding. The most apparent parallel to the sewn boards binding technique is found in boat building of antiquity where the shell-first construction method created a hull from sewn boards. The V tunnel lacings connecting the planks and seam battens evoke the whole range of early wooden board book cover attachments. A boat craft connection is also suggested by the role of sea faring trade in materials such as papyrus and by the role of Mediterranean seaports in the production and distribution of manuscripts. Crafts of sewing leather tents and containers would also be relevant. The fourth century Gnostic texts found at Nag Hammadi were generally stitched with leather thongs into leather portfolios.

From the perspective of technology, many comparisons of the scroll and codex format focus on text management features of the two formats. However, during this early period all books, both scroll and codex, lacked text management devises such as word spacing, pagination or punctuation. Only the punctuation of the codex page itself could have played a part in the first century selection of the format. The influence of page format on illustration, as opposed to text, would recommend the codex since iconography could be set off into distinct fields. The African model is relevant here since the tradition of illuminating Christian books was advanced, not by Greek convention, but by the heritage of Coptic art. In pharonic times prayers and liturgies were illustrated with figures of deities and protective symbols in bright colors with boarder designs at the top and bottom. The texts were traced in black outline with catchwords written in red.

The connection here between bookbinding and shipbuilding is as intriguing as it is provisional. At the very least, it points out that we can't simply think of the origin of the codex as something that began in a material void, linked only to early Christian theological or missiological particularities. In the course of bookbinding, I often turn to the construction and engineering of other things when I am stumped on a particular sewing or restoration problem. Likewise, early Christians must have sought better ways to put together books, publish their literature, and circulate these documents about the Roman empire. And they most probably turned to other trades and crafts for material solutions to problems in these processes. As the above article points out, this may not just be the case with the codex format, but also with questions regarding ideal page size, ideal text placement and size, and the best means of illumination and punctuation. In this way, the mystery of the origin of the codex is not simply theoretical or social in scope, but is technological and historical.

Coptic Bookbinding and Early Christian Origins

Reading an article about early Christian manuscripts by someone from beyond the guild of papyrology and early Christian studies is like having someone else tell you that your shirt is untucked in the back or you have toothpaste on your lip. From the perspective of book technology there are a number of unnoticed connections, unutilized points of access, and simply unknown practical contexts to early Christian fragments that characterize the practice of NT textual criticism. While the historical particularities of this last article could certainly be debated and more sharply defined, it helpfully demonstrates the point I have made consistently elsewhere that there is a lot of space in the material date to begin talking about early Christians as book-binders rather than just book-readers or book-users. There may be enough data out there from early Christian manuscript holdings to enact a convergence between all this data and scholarship from historians of book-binders and the discussion of textual criticism and early Christian origins currently taking place in Biblical Studies.

Update (1/09): Please click the "Coptic Bindings" for additional posts on this subject.


The Bog Psalter

Excellent news today on the 8th century illumintated Book of Psalms recently found in a Tipperary bog. The Times reports that the book "is still in its original binding." As this book is probably a contemporary to the Book of Kells, which hasn't had its original covers for a long time, this 8th century bookbinding artifact is all the more exciting.

The provenance of the Book of Kells is murky, but we know it was in Kells by the 11th century because we have records of its being stolen from there during that period. Fortunately for us, said robbers then tore the text-block from its richly gilded and jewelled cover and ditched the actual pages, which were quickly recovered. They must not have been big readers. Ever since then it has been rebound several times, most of these far less than professional. In one notorious case some illustration was actually cropped from a few pages. In 1953, it was very finely rebound in four volumes of pigskin by Sir Robert Powell (who I have learned is the only person to be knighted for bookbinding). So while the actual pages of the Book of Kells may be in better condition than the Bog Psalter, the extant original binding of the latter makes it nearly as priceless.

I can't tell though from one of the last paragraphs in the article whether the binding or the text is of "a very high standard."

Coptic Book Covers at Al-Gourna

All the way back in February 2005, Al-Ahram announced the discovery of two Coptic payprus codices, one "set of parchments between two wooden labels," and an assortment of ostraca beneath a sixth-century monastery in Al-Gourna near Luxor.

They described the contents as following: "The first book has a hard plain cover embellished with Roman text from the inside while the second includes no less than 50 papers coated with a partly deteriorated leather cover bearing geometrical drawings. In the middle, a squared cross 32cm long and 26cm wide is found. As for the set of parchments, Gorecki said it included 60 papers with a damaged leather cover and an embellished wooden locker." (Egypt Today also picked up the story, explaining that "Theologists cannot wait for the restoration processes to begin..." I know, it is a bit rude to poke fun of foreign news agencies helpful enough to publish info in English. But I can't help but ask: Are there any theologists out there standing by for these results?)

Science and Scholarship in Poland later described the actual contents of these manuscripts: 1. "One of the books" is the only complete text of the "Canons of Pseudo-Basil" in Coptic, which previously has only been extant in Arabic. 2. "The other" contains the "Life of St. Pistentios." 3. The stack of "richly decorated" parchment turned out to be the only complete translation of Isaiah in Coptic. In the bindings of the two codices were found scraps of "The Suffering of St. Peter," "another religious text," and some tax receipts. Someone has dated the Isaiah manuscript to 9/10th century, the two codices to 7/8th.

(Roger Pearse also has a comprehensive page on the find that I will be checking for updates periodically.)

Other descriptions of the find have either been too hard to track down, or they don't exist. So far many descriptions of the covers, bindings, and manuscripts are a bit ambiguous, but they sound like remarkable witnesses to Coptic bookbinding, especially in light of their decorative nature. Typically, if a spine and covers are intact enough that their padding and stiffening materials (such as the apocryphal materials recovered from these codices) are both sizable and legible, then that is a good indication that a decent technical description of the binding process can be made. To be fair, I can only state this with certainty on the basis of images of Coptic bindings from which other legible papyrus or parchment manuscripts have been extracted. Unfortunately, I have no first-hand experience with seperating materials of this age from ancient Coptic bindings. (If anyone ever needs a hand in this capacity, let me know.) But I have lifted things like handwritten guard duty records on folded rag-based paper from the spines of Revolutionary War-era American bindings (if I recall correctly, it was a bound book by Thomas Paine). In such a modern context it is the case that the sizability and legibility of spine padding materials is directly related to the overall condition of the covers. I think it would be safe to say the same thing of books from the 7/8th century. All this is to say that even though reported descriptions of the covers claim they are deteriorated, the fact that legible manuscripts have been lifted from the bindings indicates that they may be fairly intact.

After requesting images and/or more detailed descriptions of the find, I just recieved an email response from the very helpful PCMA, stating that "Tomasz Górecki, the head of the mission working at Sheikh Abd el-Gurna" will be in the field for a few more months. When he returns, I may be able to get some photographs and/or more detailed descriptions of these covers. If so, I will be more than happy to share them here.