8 MSS: A (32v) Marcabruns, C (174r-v) marcabru (C Reg. marc e bru), E (152) Marcabru, I (119v–120r) Marcabrus, K (105v–106r) Marcabrus, R (5r) marc e bru, a1 (306) Marchabms, d (306v-307v) Marcabrus

Analysis of the manuscripts

There is probably an archetype error behind 15–17 (see the notes), and possibly another in 14 (donx) which ACR corrected independently.

The manuscripts divide broadly between AEIKa1 and CR. The versification in R consists of regular coblas altemadas with alternating ‘b’ and ‘c’ rhymes. C has the same framework but gaps in stanzas I and II (which probably resulted from the physical deterioration of Cs exemplar), and VI and IX-X absent. In C VI should probably be regarded as missing since otherwise the rhyme-scheme is disrupted. AEIKa1 break off after 39; 40-42 are in R, but are presented as a new three-line stanza. The simplest explanation for this is that R obtained the lines from a source separate from the one used by any of the other MSS, IX–X in R might be an addition, but there seems no strong reason not to accept that R has preserved what is closest to an original text. If R’s transmission is complete and the order of stanzas correct, it folows that AEIKa1 are joined in error with stanzas II and III interverted, and likewise lines 12–13, besides missing all lines after 39.

Dejeanne thought the stanzas should rhyme as coblas doblas, and that the transmission of AEIKa1, which start with two pairs of such stanzas, is ‘fort incomplet’. Spanke (Marcahrustudien, pp. 104–5) considered that of the stanzas not found in AEIKa1, only VI and VII are by Marcabru, VIII definitely not being his since it refers to a dead Guiscart/Richart, and IX being ‘ein schwachlicher Aufguss von schon Gesagtem’. Neither of these comments is decisive. The first presumes arbitrarily that Marcabru could not have known such a person. The second is both subjective and untrue. Van Vleck (Memory, p. 99) describes AEIKa1’s version as coblas doblas with one unmatched stanza, pointing out that coblas doblas and coblas altemadas are easily confused in transmission. Peragi (Saggi, pp. 59–72) accepts the stanza order of AEIKa1 and considers CR’s entire version to represent a reworking, visible primarily in the ‘inversion’ of stanzas III–II and the extension of the ending.