I was discussing this piece with some of my students last term and made a mental note to blog an analysis of it. Most of the two-part Machaut pieces are rather neglected: you can’t make a dance out of them like you can with the monophonic virelais, and by the time you’ve assembled singers to perform Machaut, you might as well do the pieces closer to a modern four-part texture. As two-part songs are neglected by performers, they’re also neglected by musicologists, who tend to prefer the three- and four-part pieces as if these show evidence of greater artistry in some teleological narrative of contrapuntal progress. *sigh* Ok, off the early music soap box now and down to business!
First you will need a score of this wonderful piece, so here’s a version that I prepared earlier.
If you want to look at the manuscript version of it, check out the links to the Machaut manuscripts online I blogged about earlier; it’s on f.202r-202v of C, f.317v of Vg, 316v of B, f.477r of A, f.150v of G and f.136r of E. The text and a translation of the poem is given below: The rondeau is structured so that the first three lines of text are in the music’s A section, the second three lines are in its B section. Often, a rondeau will have different rhymes in its two sections, but here the rhyme scheme of each is identical aab aab, with the a-rhyme or ‘-art’ and the b-rhyme of ‘-our’. This means that despite the musical repeating structure of the rondeau being lopsided (the music repeats ABaAabAB, with capital letters denoting the use of the refrain text, shown in italics above), the rhyme scheme makes the poem poetically into four six-line ‘stanzas’ of an identical verse structure, it’s just that the second ‘stanza’ repeats the music’s A section for its second half rather than using the B section.
Thus the music of the B section only comes three times, while that of the A section comes five times, including three times in succession starting with its first repetition. Because the letters a and b are being used both for the rhyme structure and for the musical structure — two structures that are related but not commensurate — this can cause confusion. The diagram here seeks to clarify this by representing the A section of the music in red, the B section in blue. Italic script denotes (as before) the refrain text, and the rhyme scheme should be clear. All in all it’s a fascinating poem, with the lover’s burning from love aptly pictured in the two rhymes, ‘art’ (he burns) and ‘our’ from ‘amour’ (love).
Rondeau 5: a musical analysis
The A section
The A section of the rondeau starts with the sonority a/e which swiftly moves to a directed progression to F/f as the tenor falls to G, giving the sixth G/e that will produce the octave F/f. This directed progression is immediately restated into m.2 where the F/f resolution lasts effectively a perfect long (that is, the whole measure of the transcription). Measure 3 returns to G/e, which resolves again to F/f for the whole of the perfect long of m.4. Towards the end of m.4, though, the upper voice ornamentation produces expectation that the F tenor note is itself being imperfected (by repeated cantus ds) for a resolution to E/e. The resolution follows at the start of m.5, the first resolution to a different tonal goal in the piece and the start of the second line of the poetic text. Given the emphasis on clearly articulated cadences to F/f in the first four measures—articulating the first line of the poem—the new E/e cadence in m.5 is striking, particularly as it is subject to an accented upper auxiliary note—a striking dissonance in the cantus on the note f, the very note that has, until this point, been the main tonal goal of the cantus. The f thus inflects the E/e resolution with an f dissonance at the start of the second line of poetry, weakening it by displacing it from the start of the new perfect long group; it is additionally weakened by the cantus voice not pausing on the e, which lasts only a minim. The voices, which had been moving apart from their initial fifth into spacing of an octave (for cadences), thereafter collapse together, as if running out of steam at the effort of a new cadence tone, resolving to a/a unison within m.5.
The idea of collapse is emphasized by the cantus rest, during which the tenor sings G and the cantus gratefully joins it with an e, to permit a resolution once more to the familiar tonal goal of F/f at the start of m.6. It is as if the initial weak and failed attempt at the new E/e goal has been abandoned for now and the singers have re-grouped on solid and familiar ground to decide their next move. The cantus proposes a second attempt, repeating the descent that it had tried at the end of m.4 from the f down to a d over the tenor’s F, giving the right sonority for another E/e resolution. But the tenor is having none of it and almost immediately branches out to an entirely new pitch—b-fa. This enables first a directed progression to a/aa and then, using a short cantus linear intervallic pattern, a concatenation of progressions from a/aa to G/g back to F/f and finally, from that secure familiarity, in the last measure to the very same E/e resolution, tried with limited success in mm.4-5. The whole of the text of line 3, ‘Dame d’onnour’ (honoured lady), aids the cantus singer in performing the successful progression from F/d to E/e for the end of the A section.
The F goals in the initial part of the A section of the rondeau are strongly associated with both the subjectivity of the first person je performing the poem and the Look to which he is subjected and which causes him pain. The lady whose Look it is seems is associated more directly with the E goal. Even if we set aside the fact that none of Machaut’s music has a final on E, the non-final nature of the E sonority is confirmed by its presence at the end of the A section of the rondeau, a position that occupies an ‘ouvert’/secondary tonal status: the lady, ultimately, is not going to be attainable and the weak and quickly abandoned resolution to E at mm.4-5 suggests that attempting to gain her might in itself lead to peril, collapse.
E sonorities usually serve as secondary sonorities in this repertoire in pieces whose primary tonal goals are to D sonorities. But not only are there no resolutions to D in the A section of the rondeau, there is no instance of the pitch D at all in the tenor until the B section. It is an unusual feature for the pitch of the final to be absent from the opening part of a rondeau tenor, occurring in Machaut only elsewhere in R1 and R12; it is more normal, indeed normative, for the initial pitch of the tenor A section to be identical with the final pitch of the B section.
The B section
Despite the absence of the pitch D from the A section’s tenor, the fact of the A section ending on E/e might make an eventual final cadence on D/d relatively unsurprising. But the B section is full of other surprises. Poetically, it is unique among Machaut’s rondeaux in having similar versification (aab) and identical rhymes (‘-art’ and ‘-our’) to those in the A section. This means that both rhyme sounds occur in both sections, and that the ‘-our’ rhyme ends both sections.
Musically, the B section surprises by opening with an imperfect sonority, a third a-c, with a marked c#. This clearly expects for a G/d resolution, but the tenor rises while the cantus rests as if taking stock, and then also rises so that the b-fa/d sonority of the second half of m.10 resolves at the opening of m.11 to a/e. Personally I would recommend that any singer preserve the c# as an ornamental note in the latter part of m.10, creating an abundance of hard sweet sharpness in sonority to match the hard sweet sharpness of the dart being described in the text.
The dart itself brings about the first D in the tenor, although it is not reached by a directed progression and is ornamented, including with two dissonant es, before sounding at the end of m.11 as an octave D/d. The section’s next directed progression is into m.13, with the voices at their widest spacing and the tenor reaching D for the second time, this time as the goal of a directed progression to D/aa. The text is ‘in me’ but there’s nothing ‘in’ the lover, he is instead ‘hollowed out’ by the dart of the lady’s look, which points right to the lack at the centre of his subjectivity, subjecting him to her gaze and objectifying him as her Other by looking back.
After this point, there are no more directed progressions until the repeat of the closing section of counterpoint of the A section towards the end of m.15, which runs nearly all the way to the end, setting the words ‘all sweetness’. The tenor F is extended, as is the linear intervallic pattern in the cantus.
The resolution to E/e is similar to that in m.5 with the initial dissonant upper neighbour note at the start of the perfection group; and once more the cantus continues to move in quavers. The tenor thus descends to a D as the cantus moves in similar parallel perfect motion to d and this is the end of the piece, without a directed progression to D, simply a parallel falling off from an ornamented and briefly resolved directed progression to E/e. If this is correct, the counterpoint of the ending is, strictly speaking, defective, presenting parallel octaves.
The final tenor notes F, E, D are in a ligature that in all sources is rhythmically three breves. It is possible that this is simply an error—a solution could be proposed as follows, which would give a more normative counterpoint for the end, similar perhaps to that in B6 and similar here to the cantus and tenor at the first tenor D in m.11.
But this emendation might remove something that is an integral part of the meaning of the piece. Preserving the manuscripts’ suggestion that parallel octaves to D/d form the end of this piece is an attractive move given the relative lack of preparation for the D final. Here, the song has no real end: the rondeau form goes round and round and the lady’s look and its paradoxical ‘doucour’/ ‘dolour’—the two words set at the end of the B section—flies back and forth with the simultaneous subjectivity and objectification of the je as the lady looks back at the lover. This is why the dart is sharp and hurts—it draws attention to the existence of an Other—a subject who is not the subject and for whom the subject is an object. This surprise (D) is unwelcome (parallel octaves).
The security of the F resolutions in the A section are dissolved by the sparse resolutions in the B section with only one F/f resolution that is subject to the extension of the linear intervallic pattern in m.16-17 as if trying to reclaim the subject as self at the last before the final descent to D. The vain enterprise of the B section becomes perhaps more frenetic and fraught because the structure of the rondeau ensures that the relatively stable F-E tonal axis of the opening A section is heard five times in total and is front loaded by being heard three times in succession. So after the very first presentation of the B section, with its catastrophic dissolution of the initial tonal ‘subject’ (F) and its contrapuntal failure (parallel octaves), the subject reasserts itself in a triple repetition of the A section. But the B section inevitably follows and then the whole AB refrain with its refrain text signalling an end known to be inevitable (and, paradoxically, no end at all) from the outset.