Manuscript Laboratory: Determining Provenance and Descriptive Bibliography

Provenance:  What does the page or book's script suggest about its age and country of origin?  Is there any accompanying provenance information from a book-seller or collector, and is it or does it seem accurate?  For a medieval MS, consult Drogin and/or Tillotson for MS script examples to find those which most closely resemble the script used by the scribe who wrote your leaf.  Look at sample leaves from the various Ege collections ("Ege Boxes" like USC's).  For an Early Modern or Modern MS, consult Andrew Zurcher's English Handwriting 1500-1700.

Identification and Description of Content (Author / Text / Places and Dates Mentioned / Relation to Other Versions in MS or Print): Most Medieval manuscript books come from monasteries whose "scriptoria" organized the production of books to support the Church's operation.  For this reason, most Medieval manuscripts you encounter will be religious books.  Secular manuscripts are increasingly rare in private hands, and those in institutional collections and private hands tend to be very well described so that you would not have to guess what they were in most cases (though you should never fail to investigate what else might be between the covers!).    The exceptions are legal documents, which scribes continued to write on parchment well into the nineteenth century.  If you have an Early Modern or Modern legal document, you will need to identify its scribe's characteristic letter forms and you will need to become familiar with legal terms (see the O.E.D. for English law) and local place names.  Early Modern documents' dates often are recorded in the "reignal year" of the current monarch, i.e., the number of years since s/he was crowned, so you will need a table of reignal years corresponding with years "CE" or "current era" style.

        What are the usual types of religious book you might encounter, whole or in disbound leaves?  In France, the emperor Charlemagne issued an order in 787 commanding establishment of schools and libraries in all cathedral churches in his realm (Edwards 17).  The Polyptychon, a volume from the ninth century preserved at St. Remigius, Rheims, listed the following types of books as necessary to a proper library: "A book of the Gospels [i.e., Matthew, Mark, Luke, and John], a psalter [the 130 Psalms in biblical order], an antiphonary [a collection of chants of the Divine Office for Sundays, feast days, and the Common of the Saints who had no dedicated feast days], a breviary [i.e., a collection of excerpts from the Gospels, Pslams, and other prayers for the year, sometimes in full or abbreviated], a computus [or perpetual calendar for figuring each year's church services], an order of baptism, a martyrology, a penitential, a passional, a volume of canons [celebrations of Divine Mass], [and the] forty homilies of St. Gregory" (Edwards 17).  Each of these foundation texts was organized according to one of several regional orders that help scholars familiar with their variants to localize the scriptorium of their production. 

        Medieval manuscript works often are anonymous or even pseudonymous (i.e., the real author attributed it to someone older and more famous, hence "Pseudo-Aristotle" and "Pseudo-Dionysus").  Medieval MSS often contain more than one "work."  Books were rare enough that each one tended to function like a library of its own, or a shelf in such a library.  MS contents usually are copied from previous MSS, some of which survive whole or in part, but each scribal copy differs from its exemplar so that each MS constitutes an unique instance of the works they contain.  For this reason, general content descriptions by title and/or author help to establish the kind of MS we are dealing with, and individual variants establish the specific artifact we hold.  Religious books (psalters, breviaries, etc.) can vary from era to era or from region to region, but variances in content/order will be of a limited number of types.  From the contents and their order, you may be able to infer what kind of book even a single leaf was removed from.  Later versions of religious books economically combined features of others to give users more for their goat-skin, to so speak.  Later (C14-16) psalters, breviaries and books of hours often began with a complete calendar, focusing on saints important to the local cathedral, and they might include services for saints important to the owner or region.  Books of hours, a kind of lay-person's religious library, often also included baptismal and funeral texts, as well as brief descriptions of martyred saints (a martyrology).

Physical Description: Pay close attention to the parchment, itself.  Which is the "recto" side the leaf, in which the binding sewing holes will be to the left if they exist, and which is the "verso" side, with the binding holes to the right, and sometimes pricking holes to demarcate lines of text if the fore edge has not been trimmed too radically at some point in the MS's past life?  Using a flashlight held parallel to the surface of the leaf, providing "raking light" to highlight surface irregularities, can you detect which side is the "hair" side (patterns of bumps where the follicles emerge from the epidermis) and which is the "flesh" side (noticeably smoother)?  Carefully, with a dry and clean finger, try gently rubbing an uninked portion of the page on each side to see whether touch will tell you which is rougher.  What colors are the inks, how big and what kinds of capital letters are there, and are there any non-textual, decorative markings on the pages?  Note especially passages in red ink, which are called "rubrications" (L. "rubio" = red) and often distinguish actions from things spoken or things spoken by priests from those spoken by the worshipers or sung by the choir.  Capital letters ("majiscules") are distinguished by the number of lines they take up (e.g., 2-line capital, 3-line capital, etc.).  Usually capitals indicate beginnings of verses or chapters of a larger unit.  See these resources if you want to specialize in this aspect of the lab: Arnie Sanders, "Oak Gall Ink," English 240.  Available online at:; Amy Baker, "Common Medieval Pigments," Cochineal: A Forum for Student Work at the Kilgarlin Center for Preservation of the Cultural Record (U.Texas, Austin) November 2004, Available online at:

Number of leaves/folia/pages:  Working from individual leaves, it is difficult to tell for certain how many leaves the original book contained, but some estimate can be made because the scribes must work regularly within the text block, filling each line with a reliable number of characters.  If we know how many leaves there are in a book with similar content but a different sized text block, multiplying that number of leaves times the relative size (in percent or fraction) of the text blocks in square centimeters (e.g., for a sample larger book divided by a smaller similar book: 138/43 = the larger book's text block's carrying capacity is 3.44 times that of the smaller book's text block).

Imposition of text on leaves/folia/pages:

Leaf size (h/w in cm.): Remember that parchment is inherently less regular in shape than paper.  Old manuscripts may have been trimmed several times by binders, sometimes leading to pages becoming un-square.  Also, parchment is more biologically responsive to changes in humidity, stretching and other deforming forces (see Clarkson, "Rediscovering Parchment"). Finally, if the leaf was cut carelessly from its binding, its binding holes may not be preserved along the left side of the recto page, if at all.  For these reasons, the page must be measured at several points to determine its maximum possible dimensions.  In passing, you might want to note whether the leaf's height to width proportion was close to the "Golden Section" (Φ [the Greek "Phi"]= 1.618033988749895... ), which would make it more likely a Christian liturgical book, or a larger number, indicating a "taller" page design that more commonly might contain secular text.

Text block size (h/w in cm): Scribes will attempt to rule their parchment regularly to maximize productivity per skin and to produce a text of pleasingly regular appearance.  Nevertheless, over the long duration of the copying project of a large MS of several hundred leaves, the copyists and their assistants may have changed, their tools may have changed, and their practices may have changed.  For this reason, it's a good idea to measure the text block at fifty or hundred page intervals to look for changes.

Text outside text block: endorsements on verso of single-leaf or single-gathering documents; catch-words between pages; marginalia; head and foot notes; interlinear notes.  Endorsements are intended to identify the internal contents of a document in storage so that it does not have to be unfolded/opened and read each time it is moved or searched for.  Compare the address and return address on a USPS letter's envelope, or the sender address and subject in an email in one's inbox.  Catch-words are written by scribes at the bottom of verso leaves to help binders assemble the finished book with the leaves in the proper order.  Early MSS catch-words are horizontal, parallel to the lines of text, but C15 humanist scribes developed the habit of writing catch-words vertically, usually positioning them on the side rule below the text block.  The verso leaf word would correspond to the first word of the first line on the facing recto leaf.  If leaves were copied in "threes" or "fours" etc., sometimes only the last leaf's verso contained the catch-word which would link to the recto of the first leaf of the next gathering..  Marginalia, interlinear annotations, and notes at the foot and head of the page can be distinguished from the original scribes' marginal annotations by differences in hand and ink, and they constitute important evidence about how the text was read and interpreted.

Further issues you may encounter--

Columns per page:  Single-column documents often are smaller format, though astoundingly small two-column bibles and books of hours have been produced.  Secular texts are not infrequently single column, though content tests are your best guide.

Lines per page (most/least, average):  Scribes usually will produce exactly the same number of lines per page.  Variations are a source of immediate interest, as they may indicate another scribe has begun working on the book.  Examine the hands of the leaves before the change.

Words per line (most/least, average):  This depends on the text being copied in its finest measure, but if you consistently get word counts per page in one range and suddenly get more or fewer per page, on average, that also may be a sign that a new scribe may be working.

Characteristics of the script (era, formality, abbreviations, distinctive letter forms):  See Drogin for the distinction between "script" and "hand," but basically it is the difference between the letter forms followed in general by scribes in a region and era, and the distinctive pen strokes used by particular scribes to construct the letter forms.

Other document evidence not visible or clearly interpretable in digital surrogate: parchment tears, flay holes, and sewing repairs (see Clarkson), blind markings (i.e., pressure grooves without pigment), faint markings, especially curators' pencil notes or scribal plumb ruling.



Need a hint to get started?