Using Mirador to compare editions of Cicero's letters
A collaboration between the Bodleian Libraries and the Biblioteca Apostolica Vaticana
Although the Bodleian and Vatican Libraries are primarily digitizing works that they do not hold in common, the cases of overlap between their selections offer a unique opportunity to compare edition-specific or copy-specific features. Once more incunabula have been digitized, we will be highlighting instances of overlap, such as our several editions and copies of Ars moriendi, a very popular medieval text that was printed in several editions during the 15th century. Today we will be using the most famous example at our disposal, the Gutenberg Bible (Arch. B b.10-11 at the Bodleian and Stamp.Barb.AAA.IV.16-17 at the Vatican), to explore what a particular printing error tells us about each copy's history.
Some differences between the Bodleian's and Vatican's copies of the Gutenberg Bible are obvious. The Bodleian's is printed on paper, while the Vatican's is vellum; the Vatican's is more lavishly decorated, with multicolor initials and gold leaf; and while both copies are in excellent condition, the Vatican's has some water damage along the top edge. But there are also more subtle differences between the two copies. One of these has to do with the ways in which the Bibles' rubricators and early owners dealt with a well-known printing error, which is discussed in Paul Needham's article "Copy-Specifics in the Printing Shop" (pp. 9-20 of Bettina Wagner's Early Printed Books as Material Objects, vol. 149, Walter de Gruyter 2010).
The Gutenberg Bible's typesetters did not provide chapter or book headings. These were filled in later by the rubricators, who could make their additions plain or elaborate depending on how much the customer was willing to pay. To facilitate this process, the printers left a blank space where each chapter heading, titulus, or initial letter was meant to go, and they also printed a rubric guide to be sold with each copy of the Bible, which specified the location and text for each of these additions. However, in the book of Matthew, the printers missed the line break between chapters 21 and 22. This error was not caught and corrected before the books were sold, and neither was it noted in the original rubric guide, which gave instructions for 27 chapter headings in Matthew, rather than the correct 28.
Plenty of copies of the Gutenberg Bible include the full 28 chapter headings, however, and as Needham points out, this is thanks more to the printers than to the rubricators. The copy of the rubric guide held at Munich contains a manuscript annotation underneath the instructions for Matthew 21. This addition, presumably made before the guide left the printing shop, identifies the Matthew 22 error and instructs the rubricator to fix it by squeezing a paragraph mark into the space between sentences and writing the chapter header in the margin. (See Needham, p. 16, for a photograph and translation of this annotation.) The other surviving copy of the rubric guide (held at Vienna) does not include this annotation. Needham posits that it was added only to later copies of the guide, noting that some extant copies of the Bible lack the correction, despite having otherwise followed the guide very closely (p. 17).
So how do the Bodleian's and Vatican's copies deal with this error? Both copies are rubricated, and both include running book titles. They do not, of course, have tables of contents, so in order to find a particular section it is necessary to make use of the metadata supplied by the libraries. The Vatican's page-viewing site includes section listings and navigation options that make it possible to click directly through to a particular book of the Bible (in this case, Matthew). The metadata on the Bodleian's page-viewing site is a little sparser, but it does link to the relevant entry in the Bodleian's online catalogue of incunabula, which lists the folio numbers for each book of the Bible.
Here are the Vatican and Bodleian copies open at Matthew 22, displayed side-by-side using Mirador (which we can use thanks to both libraries' compliance with the standards of the International Image Interoperability Framework):
In the Bodleian's copy, chapter headings are given in-line in red, and two- or three-line chapter initials are given in red or blue, in the space allotted to them by the printer. For the most part, the rubricator appears to have followed the guide, although there are some small variations in wording (for example, at the beginning of Leviticus, where Needham quotes the guide as calling for an inscription of "Explicit hellesmoth id est exodus. Incipit vagecra id est leuiticus" (p. 15), the rubricator has dispensed with the hellesmoth id est in order to make space for a Capitulum primum at the end of the line (see Arch. B b.10 f. 49r)).
Here is the chapter heading for Matthew 24:
At chapter 22, the difference is obvious. There is, of course, no line break, and just as the annotated rubric guide specified, the rubricator has inserted a paragraph marker in blue and written the chapter number in red in the margin. The rubricator or decorator has also included pen-flourishing—which was not indicated in the rubric guide, and which otherwise occurs in this copy only at principal initials, never at chapter initials—presumably to draw attention to the chapter break.
The Vatican's copy tells a different story. While its chapter initials are generally more ornate than the Bodleian copy's (most include some pen-flourishing in blue or green), the rubricator missed the chapter 22 correction completely. At Matthew 22.1, the initial is unmarked, and the following chapters are misnumbered (23 as 22 and so on). However, these errors have been corrected by a later hand in brown ink. "Capitulum xxii" has been written in the margin near the beginning of what should be chapter 22:
The subsequent chapter headings have also been modified to show the correct number, and a clarifying arabic numeral has been added in the margin in each case:
These additions appear to have been made in a roughly contemporary hand, but it's unclear who made them, or what authority they were working from. They do not appear to have been working from the rubric guide annotation, as the "Capitulum xxii" label has been erroneously placed in line with verse 2 of chapter 22, rather than verse 1. (It seems unlikely that the rubricator worked from the guide either, as the tituli vary widely from the guide's specifications.) It's likely that the correction was simply made by a diligent early reader, although the Vatican librarians may be able to provide more information on this subject.
Comparing two copies in this way makes it easier to imagine the complex chain of influence involved in the history of these artifacts, from the printers to the rubricators and decorators to the owners of the books themselves. And the Vatican and Bodleian Libraries aren't alone in digitizing their copies of the Gutenberg Bible. The British Library has digitized its two copies and made it possible to compare the images side-by-side, and the Bibliothèque nationale de France has not only digitized its two copies but made them available via IIIF, which means we can expand our Matthew 22 Mirador viewer even more:
In fact, if you don't confine yourself to the final volume of the Gutenberg Bible, which contains the Book of Matthew, you can gather together even more digitized copies of the first volume. Here are six copies (the Bodleian's, the Vatican's, the BnF's vellum and paper copies, one from the Bavarian State Library and one from the Library of Congress, courtesy of the World Digital Library):
To learn more about how to create comparative workspaces like this, see our post on IIIF and the Polonsky Project.
Using Mirador to compare editions of Cicero's letters
An update on scanning technology at the Vatican Library
A comparison of Bodleian and Vatican manuscripts using Mirador