Recovering Lost Manuscript Evidence

Close study of physical evidence and provenance can lead to significant insights into the history of medieval manuscripts, early printed books, and other rare or unique special collections materials. We can see this in connection with Le Roman de la rose, which was one of the most widely read, admired, and influential Old French works of literature, in some 21,000 lines of allegorical verse. It was begun by Guillaume de Lorris around 1230 and completed by Jean de Meun (d. 1305?) four decades later. So popular was Le Roman de la rose that about 300 complete manuscripts are extant, including many illuminated copies. The Manuscripts Division has two complete manuscripts of the text (Garrett MS. 126; Princeton MS. 227), as well as a fragment of a manuscript leaf and a fifteenth-century selection of handwritten extracts. The abundance of manuscript copies has allowed the work to be carefully edited with attention to the inter-relationship of textual witnesses and images. Beyond using manuscripts to establish and edit the text, researchers are interested in tracing evidence of production, provenance, and readership. Such evidence has been noted when available for about 250 manuscripts surveyed by Ernest Langlois in his book Les manuscripts du Roman de la rose (1910); and for more than 130 manuscripts included in the Roman de la rose Digital Library.

Yet evidence of this sort elusive when unreadable or lost, especially when new owners erased old inscriptions and volumes were rebound or otherwise physically modified. Fortunately, recovery of lost manuscript evidence is still possible, as we can see with Princeton MS. 227. It is a relatively recent addition to the Manuscripts Division, which understandably has received far less attention than Garrett MS. 126, the gift of Robert Garrett (1875-1961), Class of 1897. The latter is an illuminated copy dating from the mid-fourteenth century and can be viewed among Treasures of the Manuscripts Division in DPUL (Digital Princeton University Library). Princeton MS. 227 contains Le Roman de la Rose and Le testament, an Old French moralistic text by Jean de Meun, also found in Garrett MS. 126. Some of Princeton MS. 227’s provenance is well-documented. It was in the library of the French scholar Dominique Méon (1748-1829), who used it as one of his authoritative manuscripts in his edition of Le Roman de la rose (1814); and it was later manuscript no. 4363 in the library of Sir Thomas Phillipps (1792-1872), the great English private collector of medieval manuscripts.

At some point in the manuscript’s early history, the scribal colophon at the end of Le testament (fol. 209r), below the final “Amen,” was partially erased by washing with a wet finger or rag and scraping with a knife. Whoever did this was trying to remove the scribe’s name, but erased other parts of the colophon as well. Too faint to read was the partial date, essentially a “note to self” written by the same scribe at the end of Le Roman de la rose (fol. 175v). Erased and seemingly unreadable inscriptions can be recovered and deciphered by doing digital photography under ultraviolet light (UV), a standard technique for most of the twentieth century, based on the fluorescence of light and erased iron-gall ink under a “black-light” source. This temporarily makes the erased brown ink appear darker so that it can be read or photographed. Digital photography has made it possible to enhance the UV digital images by image-processing and manipulation in Adobe PhotoShop. At one’s desk, it is now possible to increase contrast, alter colors, and reverse the written text so that it displays as white or light-colored writing against a black background. One must also extend medieval abbreviations, add apostrophes where needed, and do some conjectural reading. This involves paying close attention to parts of letters (especially ascenders and descenders), counting obliterated characters, and making educated guesses about words of equal length that work in context or follow established scribal formulas. The imaging results for the two inscriptions in Princeton MS. 227 can be seen at the end of this blog post: fol. 209r (above); fol. 175v (below).

The scribal colophon can now be transcribed in full, as follows: “Ce livre est par fini guillaume charpentier et l’escrist de sa main et fut parfait de la second jour de l’octobre l’an de grace mil ccclxxv.” The faint scribal note on fol. 175v appears to read “En xv juillet.” Reading the two scribal notes, we can see that Guillaume Charpentier completed writing and correcting the entire volume on October 2, 1375, having finished work on Le Roman de la rose about ten weeks earlier, on July 15. 1375. More can be learned using this information. With the help of a perpetual calendar for 1375, readily available online, we determine that July 15 was a Saturday and October 2 a Monday. That means that the scribe, if he had worked Mondays through Saturdays each week, only resting on the Sabbath, would have had 67 days in which to copy Le testament on 69 pages (fols. 176v-209r. However, we know that medieval scribes working on a non-deluxe manuscript like Princeton MS. 227 would have been able to copy about three pages each day. For this reason, we can be sure that Guillaume Charpentier was only working part-time on this manuscript.

Who then was Guillaume Charpentier? Ernest Langlois mentions him but only in connection with the present manuscript. Charpentier’s name does not appear among scribes identified in the six-volume Colophons de manuscrits occidentaux des origines au XVIe siècle (1963-82). However, there was a royal clerk named Guillaume Charpentier, who was active in the 1360s and 1370s in the Dauphiné of Viennois (a royal province in southeastern France), then held by the son of King Charles V (r. 1364-80). Concerning this Guillaume Charpentier, see Guvtave Dupont-Ferrier, Gallia regia ou État des officiers royaux des baillages et des sénéchausses de 1328 à 1525 (1942), vol. 2, p. 382; Anne Lemonde, Le temps des libertés en Dauphiné: L’integration d’une principauté à la Couronne de France (1349-1408) (2002), p. 138. Higher-level document clerks, public notaries, and working administrators occasionally “moonlighted” by copying vernacular literary manuscripts on a part-time basis for local patrons. There is ample evidence of this phenomenon in the late fourteenth and fifteenth centuries in England, Italy, and other places. Someone like this royal clerk probably had sufficient writing skills to copy an entire manuscript. Still, his name is so common in France that an exact identification of the scribe is impossible at this time. Certainly, it is a matter meriting further investigation.

For more information about this or other manuscripts, contact Don C. Skemer, Curator of Manuscripts, at

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.