6 MSS: A (31r-v) Marcabruns, C (259r-v) Bemart de Venzac (C Reg. Bernat de uenzac), I (119r) Marcabrus, K (105r) Marcabrus, a1 (295-296) Marchabras with Gothic ‘M’, d (305v-306r)

Analysis of the manuscripts

Simonelli sees the MSS dividing between AIKa1 and C, with a1 stemming from the AlK tradition but contaminated by an antecedent of C. As evidence she mistakenly adduces the versification of Ca1 which she considers to render the ‘a’ and ‘b’ rhymes of AIK identical (see ‘Versification’). She notes Aa1 convergence against IK in 17 (mistakenly, and she also gives the wrong line number), 20, 44,54 (but misreads K) and 57 (she says 56): in fact this simply points to isolated readings in the twin IK.

There may be archetype errors behind 20 and 52, The broad division is AIK versus a more loosely Enked Ca1 the latter converging in 8,15, 38, 39 and 57, and containing a common error in 25 (dentelha), a1 contains a large number of individual errors (2, 5, 17, 19, 22 and so on), and a second tornado not found in the other MSS. C shows considerable intervention: among its individual readings, 16, 43–44 and 48 are particularly noteworthy), while errors occur in 5, 6 and 7; it has preserved a good reading in 45, and in 53 where AIK are joined in error; it probably has the correct rhyme word in 20 even though the rest of the line is suspect.

IK contain joint errors in 19, 20, 23, 24, 28, 44 and 57. A contains a few isolated errors (20, 45). Some of its isolated readings (33, 38 and 57) suggest intervention, but others in 15, 43, 44 and 43 are preferable to those of the other MSS, It offers a good version of the text; C’s reworkings do not seem sufficiently coherent or interesting to warrant the space of a separate edition, and are dealt with in the notes.

Versification

Frank, Repértoire, 427.4: a7’ b7’ a7’ b7’ c7’ d8 d8 e8, with one tornado of two lines in a1; cohlas unissonans. Pace Simonelli (p. 237), Ca1’s versification does not differ from that of AIK, since the u in the ‘b’ rhyme modifies the following vowel.