This is a draft stemmatic analysis of a fourteenth century metrical, rhymed poem about the siege and sack of Jerusalem that I’ve done some work on in the past (click the image to see a larger version). The rectangular boxes along the bottom row represent the three printed editions of the work (as of a couple of years ago, though I’m fairly sure there are no new editions). “Fischer” was published in a journal in 1903 and 1904; “Herbert” in a Roxburgh Club edition in 1905 and “Wilson” is an unpublished 1967 dissertation available through Proquest. Wilson drew on both printed editions for historical information but the bulk of his disseration was editing a particular manuscript that neither of the other two had looked at. Herbert does not mention Fischer.
The impetus for this diagram was that I was trying to figure out, for a given printed edition of the text or article about it, which manuscript traditions it was drawing on. One of the big textual questions for this poem is that there are two clearly differerent traditions that are indicated by “short” and “long” on the diagram, different by about 1200 lines.
The scarcity of scholarship on this poem also means that authors who have written about don’t always use the same designations for a given manuscript, so comparing commentary from one to another can be tricky. For example, one manuscript, designated “R6” on this chart, is referred to as M by Kurvinen, R by Millar, and as either Pepys 37 or Magdalene College 2014 in bibliographies.
Nineteenth and early twentieth century scholarship took the “long” version as the original and catalogued and referred to “short” versions as “defective.” The “long” versions of the poem include sections on the passion of Christ (800 lines) and a biography of Judas (400 lines) which are omitted in the short versions. Phyllis Moe in 1963 called the originality of the long version into question in part because the extra material is not integral to the “plot” of the rest of the poem.
While Fischer and Wilson were producing diplomatic editions of particular manuscripts, Herbert’s intention was to produce an authoratative text. He selected one particular manuscript, R1 (indicated with the single line), as the best text and made corrections based on four other manuscripts (dotted lines). Two of those manuscripts actually seem to be older than Herbert’s base text.
This chart shows just how messy tracing the “descent” of manuscripts can be. R12, a relatively late manuscript, appears to draw from both R4 and R5 which are from two different branches of the manuscript tree (see Greetham’s reference to Maas comparing contamination to including female descent in a human family tree (Textual Scholarship 324)). Someone in the late 15th century was also combining two manuscripts when creating an edition of the text.
Another oddity is labelled “Harlech Ms Prose” which is a fifteenth century prose summary of the poem. It seems to draw primarily from a parent manuscript to R2, R6 and R10 (they have specific features not found in the prose version but also have specific features not found elsewhere that are found in the prose version). It shows some influence from R1.
Although it’s fun when there are many extant manuscripts of a poem to work from, it also adds potential complexity not found in unique manuscripts when we can more easily say that we just don’t know.
Link to Bibliography