An edition of trouvère songs, with decades-long
work on their lyrics and melodies, has given this writer much insight into,
and admiration for, the artistry - and at times lack of it - of both the
poets and the composers involved. Although all trouvère melodies
are short, each invented to serve only a single, usually short stanza of
a poem or, in lais, one of several sections, they are often intricately
formed. The lyrics offer much variety in versification, ranging from
stanzas of eight regular iambic or trochaic tetrameters to schemes with
more verses of varying lengths, involving two or three different meters.
Similarly, the rhyme schemes may employ as few as two or three rhymes,
connecting verses of equal length, or a great variety of sophisticated
Analyses of trouvère poems have almost exclusively dealt with counting syllables and with probing caesuras in long verses; in printing, short verses are variously indented to give poems pleasing contours on the page. In such “analyses,” the metric intent of the versification - whether a verse starts or ends on a strong or unstressed syllable, the structure of the stanza, and the supporting rhyme scheme - often remains unobserved. Similarly, musical analyses have often obscured rather than clarified the intended plan by going into too great melodic detail while neglecting the rhythm.
The purpose of this paper is to present three randomly selected trouvère chansons and to show the artistry involved in their poetic and musical formulation. In the standard catalog of the trouvère repertoire, Hans Spanke’s revision of Gaston Raynaud’s Bibliographie des chansonniers français,1 they are the three consecutive songs Nos. 1547, 1548a and 1551, which are separated only by poems extant without melodies; in my Complete Edition 2 they are Nos. 889, 890 and 891.3 All three are anonymous, the second being a crusader’s song, the other two Mary songs, and all three occur with their music in a single manuscript each.
The stanza of the first song is unusual in that it has eleven verses but only two rhyme syllables. All verses have eight syllables. It is evident that the rhymes must fall on stressed syllables and that the preceding seven syllables are intended to follow a regular pattern of stresses; the problem is to discover this pattern. It is further evident that all the stanzas that follow are intended to be sung to the melody of the first stanza. To establish the intended stress pattern or meter of the stanza, it is best to see how the usual speech or prose stresses best fit into such a pattern. A survey produces the following result:
According to evidence furnished by the contemporary motets, this is
a very acceptable proportion of so-called wrongly accented syllables and
a very normal feature.4
By expanding the survey to the remaining stanzas, it appears that there the poet frequently falls back into the more usual iambic tetrameters, which do not serve stanza 1 too well. Both readings are well served by the melody, which, however, is notated without indicating the rhythm, i. e. in premensural symbols - symbols that do not indicate relative lengths.5
To do justice to the metric poetry, the melody presumably supports it with the rhythms current in the music of the time other than chants, namely, those of the rhythmic modes. In general, these rhythms can satisfy the several simple poetic meters involved in trouvère poetry in various ways:
The choice of mode seems to depend on the most singable,
most comfortable distribution of multi-note figures. (Nota bene:
It is the succession of syllables which reflects the rhythmic pattern,
not that of the notes; many syllables carry ornamental multi-note figures.)
In this song the iambic rendition probably favors mode 1, but the triplet figures in several phrases introduce an unevenness which may render other versions preferable. Mode 5 seems too stodgy, though quite possible. The trochaic-dactylic or trochaic-tribrachic interpretation suggests modes 2+3 or 2+6. This approach hardly admits a 5th-mode rendition, because it would lead to uneven phrases. In all renditions, the multi-note ornaments comfortably fall on long parts of the pattern or belong to chains of graceful figurations. All these versions are shown in Example 1d.
Whichever approach is taken, the formulation of this melody is very interesting and differs from the rhyme scheme. The poetic and musical structures parallel only with regard to the overall A-A-B formulation, while the syntactic formulation remains independent from both the musical formulation and the rhyme scheme. The sentences close in vv. 6 and 8:
The second song is rather clearly cast in regular
iambic tetrameters throughout, but there are again several “misplaced”
stresses, here seven of 48 or one in seven (the neglected stresses are
marked by underscoring).
Of the twelve verses of the stanza, the last four form a refrain which is repeated at the end of all stanzas. The rhyme structure is very simple, alternating a- and b-rhymes. The rhyme syllables remain the same throughout all stanzas except in the last one, though not always strictly the same. This rather amateurish approach is also revealed in the somewhat humdrum melody, which includes many nonchalant variants. The structure of the melody nevertheless remains quite sophisticated (Example 2a, and Example 2b).
A(a-b) A'(a-b') A"(a'-b') B(c-d) B'(c-e) B
Only the first of the three songs has the number of stanzas most usual for trouvère songs namely, five. The second song, just discussed, seems to have seven stanzas, but stanzas 5 and 6 are preserved incomplete. Our third song has eight stanzas. Moreover, its first stanza is unusual in that it has, in addition to the three-line refrain, nine verses, whereas the other seven stanzas contain only seven lines. The refrain (rf) is sung both at the beginning of stanza 1 and at the end of all stanzas. The meter is trochaic throughout, employing both strong and weak rhymes, but the verses have various lengths of 5, 7, and 8 syllables, and an interesting rhyme order (Example 3a).
The formulation of the melody is particularly intricate.
The trochees admit approaches in modes 1, 2, or 5. A performance
in mode 5 would seem quite acceptable and in keeping with the religious
text; the rich and varied ornamentation provides much variety. When
one tries to apply the livelier modes 1 or 2, however, it appears that
several phrases tend towards one mode, while others to the other, because
otherwise the long figures would fall awkwardly on short beats. Such
mixtures of modes, however, are not unusual in the trouvère repertoire.
In Example 3b, both versions are shown.
Of great interest here are the intricate melodic interrelationships among the phrases, which are indicated above the staves. The most obvious relationships, despite important rhythmic variants, are listed in Example 3c.
Because of the varied repeats, which integrate the tune, the melodic form contradicts that of the poem’s text. The latter, apart from the refrain, has a rather conventional A A B structure, in which two lines are added to stanza 1. Most of the melodic phrases derive from the “motives” presented in the refrain. They are arranged in only two sections, both of three phrases, supporting, respectively, vv. 4-6 and 7-9, derived from variations of the melody of verse 3 of the refrain. Another verse is added to the second group of three verses, anticipating the melody of the repeated refrain. Together with the two phrases unique to stanza 1, which also refer to the melody of verse 3, this phrase forms a third group of three verses. Thus, the unusual form of the melody unfolds throughout the song in groups of three verses, whereas the lyrics divide into 2-2-5 lines framed by a three-line refrain (Example 3d).