Wikipedia:Documenting Rare Book Images

Preamble: a declaration of love

Historiated letter S (for suscepimus deus) from the Sherbrooke Missal: Aberystwyth, NLW, MS 15536E, fol. 246r, c. 1310–20, with the presentation of the baby Jesus in the temple, and musical notation for Candlemas.

Rare books have a fascination and a beauty that can capture our imagination. They stand as witnesses to the wit and wisdom - and the foolishness- of ages gone by. And many are works of art in their own right. Medieval manuscripts or early modern incunabula are among the most precious, and it is a true privilege if you get the chance to work with them directly. Simply the smell of an old binding, or the almost numinous experience of holding in your hands a page on which a scribe wrote with pen and ink a thousand years ago, is enough to leave a lasting impression. It is an almost sacred duty to do them justice.

But old books are also among our most important sources for much of our cultural history. How do we actually know what happened at the Battle of Stirling Bridge? Or how people imagined an elephant in centuries that had never seen one? Or what melodies were sung at Candlemas in 13th-century Wales? Or what was the original story behind the legends of King Arthur? We can look in academic sources, but they are borrowing from older sources that are analyzing older ones still, and so often the trail leads back to a rare-books reading room. If we really want to get to the bottom of the matter, we will want a look at the real thing.

Manuscript images in Wikipedia

"Labours of the Months: May", by Simon Bening, a full-page miniature in the calendar cycle of a Flemish Book of Hours. Munich, BSB, Clm 23638, fol. 6v, early 16th century.

So it is not surprising that Wikipedia contains many thousands of photographs showing pages from rare books. They appear where you might expect them, in articles on medieval history, art or literature, but also where you might not: our article on beer has a sixteenth-century print depicting an early-modern brewery; our article on Lithuania contains an image of the page in the Annals of Quedlinburg where the name "Lithuania" first appears in writing; our article on democracy reproduces both the Cotton copy of the Magna Carta and a 17th-century engraving of a session of parliament; and our article on algebra sports a page from al-Khwarizmi's Arabic treatise Al-Jabr.

These illustrations enrich the articles, adding colour to the page and giving an impression of antiquity that the text alone wouldn't transmit. Even as a purely decorative element, they enhance the encyclopedic content. However, their usefulness is greatly enhanced if they are accompanied by the information that will allow the reader to make the most of them. With documentation, they become sources of deeper understanding and gateways to further research. But for this, they need captions that tell us exactly what we are looking at.

Correct manuscript citation

Genealogical roll recording the Norman ancestry of William the Conqueror. London, BL, Royal 14 B vi, ca. 1300-1307. This scroll is around five meters in length, allowing an extended vertical family tree.

A manuscript is any handwritten source, usually a codex (book) or a scroll, though it can also be a single sheet. As with all sources cited or used, the absolute minimum of documentation is a reference that allows the reader to find their way directly back to the original to verify it themselves. It is not enough to give a link to an on-line collection of open source images that we plundered the picture from, especially if this resource itself lacks proper documentation. And all-too-often this includes Wikipedia Commons, which usually gives some information, but seldom enough. If we want to add a manuscript photo, we may need to invest a little effort to confirm what it actually is. Fortunately, many of the most important manuscripts are now digitalized on-line, which allows us to verify exactly where the image is taken from.

In Medieval Studies, there is an established norm for manuscript citation, and it should be followed strictly. If we wish to reference a whole manuscript, we state the name of the town in which it is kept, then the name of the library, and then the call number, in that order. The library name can be shortened, especially if it repeats the name of the town. Often there is a standard abbreviation (BAV = Bibliotheca Apostolica Vaticana, i.e. the Vatican Library; BnF = Bibliothèque nationale de France), which can be used if linked to Wikipedia's article on the library. There are also more general abbreviations that can be used if space is tight in a caption, such as UL = University Library.

• Oxford, Bodleian, Dodsworth 116. (Contains the chronicle of Kirkstall Abbey.)
• Munich, BSB, Clm 4660. (The Carmina Burana.)
• Aberdeen, UL, Ms. 21. (The Middle-English Polychronicon of John Trevisa.)
• Brussels, KBR, Ms. 14697. (Gottfried's Tristan.)
• London, BL, Harley 1808. (A metrical history of the kings of England.)

How medieval England imagined an elephant. Illustration from Matthew Paris, Chronica Majora vol II. Cambridge, Parker, MS 16, fol. 151v, mid-13th century.

If we are presenting an image from within a manuscript, we also need to identify the page. Some manuscripts are paginated using our modern system, so there will be a page number, but most use the medieval system of counting folios, i.e. leaves of the book. A folio has two sides, called "recto" (the front) and "verso" (the back). So what in a modern book would be p.1 is fol.1r, and p.2 is fol.1v.

• New York, Pierpont Morgan Library, MS M.1044, fol. 45r. (A page from a book on hunting with a picture showing net-weaving and dog-training.)
• Dublin, Trinity College Library, MS A. I., fol. 7v. (A full-page illustration of the Madonna and child from the Book of Kells.)
• Leiden, University Library, Or. 1551, fol. 27b-28a. (A double page with townscapes from the Arabic travelogue by Qabudli al-Haggi Mustafa Wasfi Efendi.)

One word of caution is needed here: data in circulation may not be current. It can happen that a manuscript is transferred from one institution to another, as when holdings from Berlin libraries were taken to Kraków in 1945, or that a collection is reorganized, as when the rarest items at the British Museum Reading Room became part of the new British Library in 1973. In the 20th century, many major libraries produced modern catalogues in which manuscripts were given new call numbers. Precise citation conventions also change: a German manuscript in the Palatine collection would in the past have had a call number prefixed Cod. Pal. ger., but today this is cpg. Unfortunately, information from 19th-century scholarship is still being copied unchecked into contemporary books. So when we are citing a manuscript, it is best to go to the library's on-line catalogue and take the information from there.

Incunabula and other early prints

William Caxton's edition of Chaucer's Canterbury Tales, Westminster, 1476/77, ISTC ic00431000, fol. 108v. This copy Washington, D.C., Folger Shakespeare Library, STC 5082.

Incunabula are books, broadsheets etc. that were printed in Europe before 1500. Together with printed material from the 15th to 17th centuries, these are precious artifacts typically found alongside manuscripts in the rare books reading rooms of major libraries. Like modern printed works, they are cited by author - title - publisher - date - place of publication. However, if they are rare, they are likely to appear in scholarly catalogues, such as the Incunabula Short Title Catalogue (ISTC), and it is helpful to give a catalogue number, much as we would give the ISBN of a modern book. If only a single copy of a printed book survives, or if idiosyncrasies of a particular copy are visible in the image, then as with a unique manuscript we would cite by town - library - call number.

Inscriptions

Incidentally, although it is not the subject of this essay, much the same considerations apply to photographs of inscriptions carved on stone, engraved on a tool or painted on walls, wood panels, etc., or indeed to any unique work of art. If it is in a museum, citation form is town - museum - catalogue number. If it is on the wall of a building, we should be told which building.

Fuller documentation

In addition to the obligatory identification of the manuscript or print and the page or folio number, it can be helpful to contextualize the image with more information. How much to include is a matter of taste and judgment, but there's a lot to be said for being generous.

To make sense of a manuscript image, the reader certainly wants to know the approximate date of the manuscript, which frequently is different from the date of authorship of the text contained in it. Codicologists like a mention of the physical material: paper, parchment, vellum. But the average reader will be more grateful for help to explain what is to be seen in the image: what person, event or story is represented by a picture, what does that highly-stylized writing actually say, is there anything visible that we might easily miss? We don't want to detail the obvious, but we do need to remember that what might have been obvious to a medieval reader is not necessarily so today.

Here, it's important to remember why the image is in the article. We want to lead with the information that most obviously supports the text in the article's main text. The caption should establish the picture's relevance for the article, and draw the reader into the material. But if the image is worth including, it is worth understanding it, so within limits, it is legitimate to describe the image for its own sake.

The principle of succinctness

Miniature of Pope Joan from a 15th-century German manuscript of Jans's Weltchronik: Heidelberg, UL, cpg 336, folio 203r, paper, c. 1420.

We also need to remember the principle of succinctness. There is a rule of thumb that it is good to restrict captions to three or four lines, as a shorter caption has more of an impact. However, image captions in published books are often much longer, so exceptions can be justified. It is usually possible to shorten a first draft simply by thoughtful rephrasing, and judicious use of abbreviations can help a lot. However, deleting helpful information just to reduce a caption length is seldom a good approach.

In principle, relevant image documentation for rare books should be in the article somewhere, but it does not necessarily have to be in the caption. If the image is to be discussed in the article, much of the documentation can go into the article text, but if a picture is only illustrative, it is better for the information to be directly linked to it. If it can't be in the caption itself, a good option is to put it into a footnote built into the caption.

The image of Pope Joan at the top of this section is an example of one where the original caption length of around twelve lines was seen as problematic. General tightening up reduced it to nine lines, but shorter would have involved cutting material. It would of course be possible simply to delete the last two sentences, but in the context of the article where it appears (Jans der Enikel), it is all relevant and of interest to readers curious about the article's subject. The solution used in that article is to put the second half of the caption into a footnote, which can be seen by hovering with the cursor over the blue footnote letter. The resulting balance is commensurate with its Good Article status.

A good method of entering explanatory footnotes that are kept separate from citations is to use the sfn coding system. The extension of the caption is placed within the template {{efn|xxx}}, which is kept inside the caption itself. Then at the bottom of the article, we enter {{notelist}} under the header == Notes ==.

Image description page

The incipit of Matthew in the Lindisfarne Gospels, with the decorated letters LIB (for Liber generationis Iesu Christi). London, BL Cotton MS Nero D.IV, fol 27r, 8th century.

Clicking on any image on a Wikipedia page leads the reader to the image description page. This is the place for details of the history of the digital file, including copyright status or credits for digitalization, which don't belong in the article. If the image was lifted from a book or copied from a website, that information belongs on the description page.

However, the image description page also has space for fuller information, and should duplicate some of what is in the caption. Remember that the same image may be used in multiple articles, and the description page is an anchor for all of them. So where a caption should show why the image is relevant to the article subject, and draw the reader back into the main text, the description page does not have those tasks. But the task of identifying what exactly is in the image is the same in both the caption and the description page, so that for example a correct manuscript reference with library and call number belongs in both places.

The image description page can also include a more detailed description of what can be seen in the image, and it is not bound by quite the same need to be succinct. Some featured pictures have image descriptions which are several hundred words long (e.g. the image from the Lindisfarne Gospels above). So if there is a conflict about how much information belongs in a caption, a possible solution is to move some of it there.

Gallery

• 
  The funeral of Diocrès from the Très Riches Heures du Duc de Berry. Chantilly, Musée Condé, MS 65, fol. 86v.
• 
  One of the four surviving exemplified copies of the Magna Carta, London, BL, Cotton MS Augustus II.106, 13th century, vellum.
• 
  The earliest English poem, Cædmon's Hymn. Oxford, Bodleian, MS Hatton 43, fol. 129r.
• 
  Portrait of Konrad von Würzburg from the Codex Manesse. Heidelberg, UL, cpg 848, fol. 383r.