Narrative Mapping and Narrative Sections

As explained under “Narrative Mapping” in the site, citation practice for the Roman de la Rose and most medieval texts has traditionally referenced the currently accepted critical editions. Yet this scholarly protocol inhibits the cross-manuscript comparative study that the Roman de la Rose Digital Library promotes. Since the number of lines for the work varies from one manuscript to another, depending on interpolations or excisions, the narrative mapping of the Roman de la Rose divides the text into reading segments instead of lines. In the Roman de la Rose, G and J represent the authors Guillaume de Lorris and Jean de Meun while Arabic numerals and roman letters indicate segments and shorter sub-segments within them. For example, a section such as “g3” includes two segments, “g3a” (Amans discovers the Garden of Deduiz) ranging from verses 129 to 138 in Felix Lecoy’s edition, and segment “g3b” (Ekphrastic description of the portraits of the courtly vices) ranging from 139 to 460. The second segment is further subdivided in individual descriptions: “g3b.1” (Haïne 139-151), “g3b.2” (Felonie 152-155), “g3b.3” (Vilanie 156-168), “g3b.4” (Covoitise 169-194), “g3b.5” (Avarice 195-234), “g3b.6” (Envie 235-290), “g3b.7” (Tritesce 291-338), “g3b.8” (Vielleice 339-404), “g3b.9” (Ypocrisie 405-438), “g3b.10” (Povreté 439-460).

These divisions and the range of verses means that comparable passages across different manuscript can be readily locatable while number of lines given for each section facilitate tracking variations in section length from one exemplar to another. The narrative mapping protocol (and the explanation on which I base this blog) was devised by Prof. Stephen G. Nichols borrowing from that used for classical texts, where one cites not a page number or a given edition or translation but a segment of the text. Jeun de Meun’s section was done by Prof. Tracy Adams and a full revision by Jeanette Patterson. Prof. Sylvia Huot added information on some of the main known interpolations. These interpolations have been given different numbering, such as “g20a.5*1” (Anonymous continuation in which Jealousy falls asleep, and the Lover gets to spend a night of bliss with the Rose) or “j2d*1” (‘Litany of Love’ in which Reason describes spiritual love, refers to the Passion of Christ, and outlines a mystical ascent through contemplation, ultimately advising the Lover to select the Virgin Mary as love object).

Once the Roman de la Rose was divided into the narrative segments explained above we started to match particular manuscripts to these general narrative segments. Some exemplars were matched manually, segment by segment, and in those an actual transcription of the manuscript’s first line is included. For this case see Bodmer 79, where a section like G5e (Cortoisie invites Amans to dance) starts with the actual transcription “La karole quiert si plaisant”, followed by the verse number in Lecoy’s edition (L775).

In most cases, however, the matching of a specific exemplar with the narrative segments was generated automatically through an algorithm and so it will only approximate the actual start of each reading segment. Most of the remaining narrative mapping will be generated automatically. In these automatically generated cases, the first line of text is provided but instead of matching the actual manuscript it matches the transcription of Selden Supra 57. The line will thus start with (SS57) and the narrative segment will include a question mark. For an example see Arsenal 3338, where section “g6a?” (Amors introduced & described) is transcribed as “(SS57) A lui se tint de l’autre part.”

The transcription from SS57 allows users to check themselves for the actual placement of the starting line. Although generated automatically the algorithm gives a fairly good approximation and should be accurate within one to two columns to the actual start of each reading segment.

To access narrative segments in the manuscripts that have been mapped so far select Narrative sections in the Home page and click on the desired section id. The available narrative segments can also be accessed through the Page Turner view with the drop-down menu.

If you gather more accurate information on narrative segments please feel free to let us know at

Usage statistics (through March 2009)

Since the launch of the site on September 2008 through March 2009, the Rose team has noted the following usage statistics:

• 14,112 visits from 112 different countries or territories
• The top five countries represented (in order): United States, United Kingdom, Canada, France, Spain
• 11,205 absolute unique visitors
• 21% of these visitors have returned to the site
• 314 of these visitors have used the site 9-14 times
• 192 of these visitors have used the site 15-25 times
• 605 of these visits lasted between 10-30 minutes
• 82 of these visits lasted over 30 minutes

We experienced a spike in visits in late March after someone (not a member of the Rose team) registered the Rose Digital Library on StumbleUpon. Starting in April, we have refined the analytics to gather more fine grained statistics such as visits to individual folios or pages on the site. We will report on these new statistics as we gather them.

Major site update

Below is a list of the changes. Later posts will give more detailed explanations of the new features.

Help document:

  • Content updates
  • Layout fixes

Collection spreadsheet: New feature!

  • Mined manuscript descriptions to create spreadsheet
  • Tabular view with Google visualizations
  • Downloadable in CSV format for use with local spreadsheet applications

Illustration titles: New feature!

  • Tabular view of titles, average position, and frequency.
  • Data generated from image tagging

Extant manuscripts:

  • Moved to tabular Google visualization
  • Link to view in Google docs

Narrative sections: New feature!

  • Poem divided into sections. Each section has an id, a description, and a Lecoy range.
  • Tabular view of ids, titles, and Lecoy range
  • Click on section id to search for folios containing that section

Narrative mapping: New feature!

  • A manuscript is mapped by finding the start and end of each section in that manuscript.
  • The mapping may be done manually, which takes a long time, or automatically.
  • The automatic mapping takes as input a description of the book structure and guesses the column of the section start/end.
  • Only a few manuscripts are mapped so far.
  • The sections can be searched by title and id.
  • Section in the current page can be displayed to the right of the page turner.
  • Manual mapping includes a transcription of the first line of each section.

Character names: New feature!

  • List of characters in the poem


  • Lecoy number always displayed with leading L
  • Display date when site last updated
  • Navigation on left reworked
  • Rubrics that are transcribed when describing the book structure for automatic narrative mapping are searchable
  • Normalization of illustration titles in image tagging
  • Hit enter in text box to do advanced search
  • Additional and updated French translations
  • Guessing names of printed book folios works better.
  • Finer grained tracking of site usage with Google analytics.
  • Mention contactus email when internal error occurs
  • Summary book descriptions added
  • More manuscripts cropped for page turner
  • Google visualization api used for tabular views of site. Click on column title to sort (unless sort is disabled for that table)

Known issues:

  • Webkit (Safari, Chrome) layout problem with header being offset.
  • Tabular Google visualization sometimes has columns misaligned. Click on column title to fix.