5 MSS: A (32v-33r) Marcabruns, I(120r) Matcabrus, K (106r) Marcabrus, a1 (307- 308) Marchabras, d (307r) Marcabrus

Analysis of the manuscripts

The MSS are relatively uniform. Lines 1,6, 12, 35 and possibly 45 suggest filiation between Aa1 in lines 9, 23, 24, 37 and 41 A is isolated against JKa1; finally In lines 9, 16, 22, 49 and 56 a1 is isolated, A uniform tradition Is suggested, however, by common readings at the rhyme in lines 23, 28 and 55 (all producing irregular rhymes that no scribe attempted to correct; see ‘Versification’ and the notes) and by three apparent archetype errors; in lines 24, 36 and 48, There are isolated errors: see line 35 in A; lines 10, 35 and 47 in I; line 38 in K; and lines 4, 19, 28, 29, 35, 36, 37, 38, 39, 41, 43, 44, 45, 47, and 48 in a1 Analysis of the lines which oppose Aa1 to IK, particularly lines 1, 6, 12 and 35, suggests Aa1 pay more attention to detail than IK: it is also noteworthy that A leaves a space for line 52. In some respects a1 would be an interesting base MS as it appears to be a careless copy of a superior text: it Is the only MS to preserve line 52 and several, points of detail either are or suggest better readings than those preserved in the other MSS (for example line 56). On the other hand, the frequent errors In a1 would require a large number of corrections. A’s text is more or less error-free, yet without obvious facilior readings, so A Is the obvious choice for base.

Versification

Frank, Repértoire, 157.1: a7’ a7’ b7’ b7’ c7’ a7’. Nine coblas unissonans, with a twoline tornada, unicum. However, as the poem survives in the manuscripts, its versification is not regular: in lines 23, 28 and 55 all the MSS agree on readings that make the rhyme sounds technically ‘incorrect’ (see the notes to these lines). In addition rhyme words are repeated In lines 15/40, 16/46, 26/36 and 54/56 (this last In the tomadd). The agreement of all the MSS on these points suggests Mareabra was less than rigorous about form in this poem and that the readings at the rhyme in these lines are therefore acceptable despite evidence for archetype errors in this poem’s transmission (see lines 24, 36 and 48).