7 MSS: A (28v-29r) Marcabruns, C (I77r-v) marcabru (C Reg, marc e bru), E (154) marcabru, I (117v) Marcabrus, K (103r-v) Marcabnis, R (5v) marc e bru, d (303v) Marcabruns

Analysis of the manuscripts

The absence of stanza IV in CR and their common errors (see for example lines 4, 6, 17 (wrong verb), 26, 33, 40, 43, 50, 54) show that they derive from a common source different from that of AIK. C apparently intervenes further in line 5, 32 and 38, while JR. has a further error in the omission of lines 35–37 and the presentation of lines 31–42 as one stanza. The absence of stanzas VII and IX in E, its bizarre transpositions, metrical irregularities (see for example 2, 4, and 27), errors and isolated readings (see especially lines 10–11, 15–17, 26–30, 33, 43, 45, 47) suggest that it derives from a damaged exemplar, one which may ultimately derive in parts from someone’s imperfect memory of the song. For all its errors and quirks, E preserves the most plausible reading in line 21 (see the note).

Common errors in AIK in lines 21, 32 (minor), 35, 39 (rhyme) and 46 (rhyme) show that these MSS derive from a defective common source. While JFC has an isolated minor error in 35, and I has isolated errors in lines 8, 19, 26 and 30, A has a large number of isolated reading most probably resulting from intervention: see especially lines 4, 5 (errors), 21 (parran), 23, 26 and 28 (minor retouchings) and particularly 37, 38 (error), 39, 40, 41 and 54. Kavoids the error found in ACIR in line 7.

Versification and melody

Frank, Repértoire, 223.4: 8a 8b 8a 4b 8a 4b; nine coblas unissonans. Spanke, (Marcabrustudien, pp. 23–24) suggests that Marcabru may have been inspired by the conductus, ‘De ramis cadunt folia’, of St Martial de Limoges (before 1150) and is followed in this by Fernandez de la Cuesta, who understands our poem as a conirafactum and prints its text together with the music of ‘De ramis’ (Canpns, p. 67). Gruber argues that the song is a reply to Guilhem IX, VII.