|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
CHAPTER THREE
MUSICAL FORMS IN THE TROUBADOUR REPERTOIRE
Contrary to the usual estimates, less than 20
percent of the troubadour songs with music may be
classified as true oda
continua forms, in the sense that no musical
phrase is repeated. The proportion is doubled if one
includes forms that have one repeat, but even so, the
total remains less than 40 percent, not the two‑thirds
that is often cited. Among
composers who favoured this form, Folquet de Marseille
and Peire Vidal stand out for their avoidance of
strongly repetitive forms; neither composer used the
ABAB form at all in their surviving notated songs.
Thirteen of Folquet de Marseille's poems are preserved
with music, and at least one version of each song has
either no repeats or only one; in three versions, the
repetition of two or more phrases shows some symmetry.
Eight of Peire Vidal's 12 musical forms have either no
repeats or one; two show some symmetry, and two have
irregular repetition patterns.[1]
Apart from the tendency for abba rhyme schemes
to be matched with oda continua musical forms, one may note the fact that this form was
used with the rare songs whose rhyme schemes fall at
either end of the spectrum of possible rhyming
patterns. At one end, there is a song by Guillem de
Saint Leidier, "Pus tan mi fors' amors que mi fay
entremetre" (P‑C 234,16), which has only a
single rhyme for its entire six‑line stanzas; at
the other are two songs that repeat no rhyme within
their stanzas, which are also of six lines: Arnaut
Daniel's sextina, "Lo ferm voler q'el cor
m'intra" (P‑C 29,14), and Peire Vidal's
"S'ieu fos en cort que hom tengues drechura"
(P‑C 364,42).
Of the songs in this group whose music is
transmitted in more than one manuscript, there are ten
in which one or more versions diverge in form.[2]
In seven of these, the difference is between no repeats and
one repeat, but in two songs one of the other versions
shows an ABAB form, and in one the other version has a
symmetrical repetition pattern. Given the more
significant divergence of these three cases, they will
be looked at individually. The two versions of the
first song, by Peire d'Alvergne (P‑C 323,15),
are given below, along with their formal graphs.[3]
Example 3. Peire d'Alvergne, P‑C 323,15.
In the
version of X, lines 3 and 4 repeat the music of lines
1 and 2 exactly. In the version of R, lines 3 and 4
are related to 1 and 2 through transposition, which
might suggest a scribal error. Line 4 is consistently
a third higher than line 2, while line 3 alternates
between a second and a third in relation to line 1.
However, these two lines also have marked divergences
in neume formations, apart from transposition, and R
differs significantly from X in the last three lines
of the song; transposition would seem to be only one
factor in the divergence between versions. The
differences of form in "No m'alegra chan ni
critz" (P‑C 167,43) by Gaucelm Faidit also
affect only the first four lines, and these are
reproduced in the example, along with the graphs of
form. It should be noted, though, that although no
other phrases are repeated in the rest of the song in
any of the three versions, that of R is so different
from those of G and W from line four onwards that one
would have great difficulty identifying it as the same
music. Example 4.
Gaucelm Faidit, P‑C 167,43.
Here
it is the version of R that produces a strict ABAB
form in the melody, yet it is highly unlikely to be
correct, since none of the other 15 text manuscripts
duplicate R's addition of three extra syllables to
line 4 with the words "mos ditz me," and
without equivalence in line‑length between
corresponding phrases, the ABAB form is not possible.
The only other difference in form occurs in line 3,
where W repeats the music of line 1, but with a new
ending, while G has an entirely new phrase.
The differences in form between the two complete
versions[4]
of Folquet de Marseille's "Tant m'abellis
l'amoros pessamens" (P‑C 155,22) arise in
the last two lines of the song; it is shown in Example
5 below. Example
5. Folquet de Marseille, P‑C 155,22.
In G, a
symmetrical form is produced by repeating the music of
lines 3 and 4 in lines 7 and 8, which results in a
symmetrical balance between the two halves of the
song. The last two lines of R's version also recall
elements of lines 3 and 4, but the similarity is much
more attenuated. It is closest in line 7, which might
have been graphed as a variant of line 3, as in G;
this is a matter of interpretation, and the phrase was
judged to be different enough to be considered a new
phrase.
It will be noticed from the graphs of both
versions' musical form in this song, that other
phrases besides the last two repeat elements of
earlier phrases, namely initia
and cadential material. This kind of repetition at the
sub‑phrase level is more the rule rather than
the exception in this category of form as in the
others, and is an important structural device in a
form sometimes deprecated for its supposed looseness
or formlessness. Certainly, the freedom of a
continuous unfolding must be understood as a principle
of the non‑repeat form, as the name oda
continua suggests. But there are very few songs
that do not balance this freedom with the kind of
structural linking just mentioned, and even in these,
of course, a sense of unity is achieved through
similarity of phrase shape, modal orientation, and
motivic recall. P‑C 155,5, P‑C 364,42, and
P‑C 406,28 are three songs whose formal graphs
indicate a minimum of cadential or initia
linking, or phrase recall, in comparison to the
average. As an example, "S'ieu fos en cort que
tengues drechura" (P‑C 364,42), is given
below and discussed, but the reader will find
analogous circumstances in the other songs as well.
Example
6. Peire
Vidal, P‑C 364,42.
As in Arnaut Daniel's sestina, the poetic and musical
forms mirror one another in that both would have been
considered examples of the oda
continua by Dante. There is no repetition of rhyme
sounds within the stanza, and no repetition of musical
phrases; at the sub‑phrase level, no initia
or cadences are repeated. Yet all the rhymes are
feminine in this isometric canso,
ending in the same vowel sound 'a'; in addition, the
b, c and d rhymes are linked in their accented
syllables through assonance, as are the e and f
rhymes. The cadences all differ, but those of lines 2
to 5 all consist of a descending three‑note
neume followed by a single pitch; the first line ends
with only two single neumes, while the last line has
the most extended cadence, with four‑plus‑two
pitches (the rhyme of the first line is also the only
one not linked to the others in the sound of its
accented syllable). Furthermore, though each phrase is
clearly different from all the others and not a
variant, each is nevertheless linked to at least one
other phrase through motivic recall. The brief recto tono at the beginning of line 1 is recalled at 3,3‑5;[5]
the stepwise ascent followed by two descending thirds at 1,6‑10'
is heard again transposed at 3,1‑7; the
relatively wide leap of a fourth at 2,7‑8 is
paralleled by an ascending fifth at the same location
in line 3; the two descending thirds of 3,5‑7
are answered by two ascending thirds at 4,3‑5;
and the motive at 5,5‑9 is repeated at 6,5‑9,
and in turn recalls 4,4‑8.
The presence of these links, connecting both
rhyme sounds and musical phrases, results in a unified
structural design in a song that appears on the
surface to be one of the more extreme examples of the
most differentiated of forms. This concern for
structure in the oda
continua category of troubadour songs is an aspect
that deserves some emphasis, since it is so easily
overlooked, and will therefore dominate the remainder
of this section. In considering the poetic structure
of the stanza, the main factors are its division into
parts, the pattern of rhymes, the number of lines, and
the meters. It is common for the stanza to be
divided textually into two parts, and these may be
equal in length, or one or the other may be longer,
the relative length being a function not only of the
number of lines, but the number of syllables in each
line. Normally, the rhyme scheme will indicate the
stanzaic division where this is clear, and it will be
supported by meter as well as syntax and content. For
example, the following poetic forms all have a clear
two‑part division, indicated by the slash:
a
b
b
a
/
c
c
d
d
10'
10
10 10'
10
10
10'
10'
a
b
a
b
/
b
a
b
a
7'
4
7'
4
7
5'
7
5'
a
b
b
a
a
/
c
d
d
c
c 10
10'
10'
10
10
10
10'
10'
10
10
However, there are plenty of examples where no
clear division is evident, or where ambiguity results
in various possible divisions, or where the stanza
divides into more than two parts. In musical forms
with repeated phrases, the division can be evident
simply from the pattern of repetitions, although it
need not always coincide with the textual division. In
oda continua
forms, this option is not available, but it is not
quite true that the form can have no division, as
Dante claimed.[6]
To
some extent, each phrase shows some degree of melodic
articulation, whether it be through characteristic initia and cadences or contrasting material, and the presence of a
stronger articulation dividing the stanza is often a
matter of degree. In some cases, though, the melodic
division is readily apparent, and may be produced by
various means. For example, if the interval between
the end of one phrase and the beginning of the next is
very small or at the unison throughout the song, a
large leap at one phrase juncture will effectively
articulate the stanza into two at that point. Examples
that divide the stanza melodically in this way include
P‑C 323,4, P‑C 323,15, P‑C 30,3, P‑C
364,39, and P‑C 155,11.
Following is the formal graph for P‑C
323,4, "Amis Bernart de Ventadorn," by Peire
d'Alvergne:
The
interval between lines 4 and 5 is a seventh, while
that between every other phrase is the unison, except
for lines 3 and 4, which are separated by a third. The
leap in this case clearly underlines the division of
the stanza indicated by the rhyme scheme. P‑C
323,15, P‑C 364,39, and P‑C 155,11 function
in a very similar way, with their largest interval
between phrases falling at the point of division one
would expect from the rhyme scheme and metrical
structure. The rhyme scheme of P‑C 30,3 contains
some ambiguity, which the metrical form does not
resolve, since the stanzas are isometric; its pattern
of rhymes is abbaacc. The song with its formal graph
is shown in Example 7.
Example
7. Arnaut de Mareuil, P‑C 30,3.
One could
divide the stanza after the fourth line, on the
strength of the abba pattern, which is a standardized
type, and this would be supported by the tornadas,
which are of three lines. It would also be legitimate
to see two points of articulation in the stanza, one
after the third line, and one after the fifth, with
the latter perhaps stronger on account of the new c
rhyme in the last two lines, thus:
abb/aa//cc. In terms of the intervals at phrase
junctures, the division between the fifth and sixth
lines receives a marked emphasis by the upward leap of
a seventh at this point. All other intervals between
phrases are either unisons or seconds, except for the
third that comes between lines 4 and 5; since the
cadence of line 4 repeats that of lines 1 and 3, with
the difference that the final pitch is b instead of d,
giving an ouvert/clos
effect, the possible division after the fourth
line is respected in this way. Likewise, the secondary
articulation after the third line is supported by the
exact repetition of the cadence of the first line at
the end of the third, in effect linking the first
three lines together.
In other songs different techniques may be used
to underline the stanzaic division. In P‑C
406,39, by Raimon de Miraval, which has a rhyme scheme
that divides the eight‑line stanza into two
equal parts, the last four lines all utilize the same
cadence figure, binding these lines together as a unit
and distinguishing the second part from the first. P‑C
366,22, by Peirol, also has eight‑line stanzas,
and in this case the same sequence of final pitches in
the first four lines is repeated in the last four,
suggesting an equal division into two parts here as
well.
As we have seen, from the point of view of the
poetic structure, the stanzaic division is largely a
product of the specific sequence of rhymes and meters
that make up this structure. The pattern itself might
be considered to be relatively independent and
secondary compared to the poetic text as a whole, but
this technical aspect of poetic craft was an absorbing
concern for the troubadour poets, who strove for
originality and ingenuity in versification. Likewise,
the formal pattern resulting from the sequence of
repeated and non‑repeated musical phrases is to
some degree an abstraction from the living whole that
is the melody in its sounding reality. One of the
virtues of this level of abstraction, however, is that
it provides the basis for a dynamic interrelation
between the poetic and musical forms, dynamic because
it arises not from a simple identity between the two,
but from an often complex mixture of agreement and
tension that enriches both sides of this dual art. A
prime focal point for this interaction of forms is the
pattern resulting from the combined sequences of
rhymes and cadences. It is normal to compare the rhyme
sequence with the musical form determined by
repetition of entire phrases, but as cadential figures
may be repeated or varied independently of phrase
repetition, they often set up their own pattern, which
of course interacts with the rhyme sequence much more
directly than does the entire phrase. In songs without
repetition of phrases, such cadential patterns are an
important structural resource; to a lesser extent,
finals alone and initia may also set up formal patterns.
In many ways, each song presents a unique
configuration of interrelationships that differentiate
it from all the others; there are no standardized
types at this formal level. Nevertheless, a few
selected examples may give some idea of the formal
interest possible even in the oda continua category of musical forms. "Quant hom honratz
torna en gran paubreira" (P‑C 364,40) is a canso
by Peire Vidal; the music and text for the first
stanza are shown in Example 8. Example
8. Peire Vidal, P‑C 364,40.
cadence
pitches
c
G
C
C
F
C
C
cadence
pitches as
a
b
a
a
c
a
a rhyme
scheme The text repeats many of the standard commonplaces of
the courtly love song, without achieving stylistic
originality.[7]
The poem is based on an extended military metaphor in which
the lady plays the role of a superior foe. The poet
almost seems to have been more interested in the
technical challenges of the versification; the poem is
set in coblas
unissonans, one of the most difficult rhyming
patterns, in that the same rhyme sounds in the same
sequence must be used in every stanza. The stanzas are
isometric, consisting of seven lines of ten syllables,
all with feminine rhymes except the two c rhymes in
lines 5‑6. Since the a and b rhymes both have
the same vowel sound in their final, unaccented
syllables, and since the stanzas are unissonans,
one may imagine the ear becoming fatigued after one or
two stanzas. It would therefore seem fitting to use
the through‑composed design for the music, which
would provide the most variety in the sound pattern,
with a new phrase for every line. The music does do
this, but it also juxtaposes a different pattern which
interacts dynamically with the poetic form.
One feature of the rhyme scheme is the
dovetailing of rims
encadenatz (abab) and rims
crozatz (abba),[8]
so that the diesis or division
between lines 4 and 5 is obscured. The last four lines
form a unit like the first four, with line 4 as a
pivot common to both. Like the text, the music too has
no clear division in the middle of the stanza.
Instead, line 5 begins at the same pitch as the final
pitch of line 4. There are only three different final
pitches, as there are only three rhymes, and the
sequence of finals also produces two overlapping
patterns of four lines each, with a pivot at the
fourth line. The first, third and fourth lines in each
section end in a c or C, while the final pitch of the
second line in each section (that is, line 2 and line
5) lies a fourth or fifth from C (G in line 2, F in
line 5). The sequence can be graphed as a set of
rhymes (as shown above in Example 8) and its effect is
to set up a counterpoint to the actual rhymes,
exhibiting both tension and parallelism.
The cadences in turn set up their own pattern
which further modifies and interacts with those found
in the finals and textual rhymes. Thus lines 1 and 2
share the same final vowel sound 'a', though they have
different rhymes, and their cadences have the same
figures, but shifted to a different pitch level. A
similar use of transposition occurs with the two
masculine rhymes at the ends of lines 5 and 6, and
another example of this interplay between end‑line
patterns can be seen between lines 3 and 6 and lines 4
and 7, which are linked through their cadences; in
this case the cadences are at the same pitch level,
but are varied by the redistribution of pitches over
syllables.
Other songs from this category that are notable
for structural interplay between rhyme scheme and
cadential material include: P‑C 70,19, P‑C
155,1, P‑C 155,8, P‑C 155,11, P‑C
155,22, P‑C 29,14, P‑C 30,3, P‑C
167,52, P‑C 364,4, P‑C 370,13, P‑C
406,39, P‑C 375,19, P‑C 248,53, and P‑C
248,61. Two of these, P‑C 375,19 and P‑C
155,8, have been chosen for discussion because they
illustrate the contrasting results possible from
similar means. "S'eu fi ni dis nulla sazon"
(P‑C 375,19), by Pons de Capdoill has a very
straightforward, simple stanzaic form shared by a
large number of poems; its eight‑line stanzas
are isometric, with a meter of eight syllables, varied
slightly by the two feminine c rhymes. The song and
its graph are shown in the following example. Example
9. Pons
de Capdoill, P‑C 375,19.
What is
interesting in this song is the subtle, discreet
manner in which the cadences underscore the rhyme
structure and stanzaic articulations, without
juxtaposing a different pattern of their own. The
stanza divides clearly into two equal parts, and the
cadences help to define this division and the
secondary articulations in a number of ways. The three‑note
descending figure b‑a‑G is common to the
cadences of the first four lines, joining them
together as a unit; transposed upward and downward one
step, the figure further articulates the sub‑grouping
of the last four lines into two pairs, at the same
time connecting lines 6 and 8 with cadences of the
first half, and differentiating them. The cadence of
line 5 stands out for its shift in register compared
to all the others, and this too has the effect of
underlining the stanzaic division at this point;
slightly varied and transposed down a fifth, the same
cadence links the two c rhymes. The chiastic rhyme
structure of the first half is strictly paralleled in
the cadences by the finals; in addition, the cadences
for the b rhymes are identical, while those with the a
rhymes are slightly varied. Since the a and b rhyme
sounds are themselves linked in their consonantal
terminations, it seems fitting that all four cadences
accompanying them should also employ similar figures. "In
cantan m'aven a membrar" (P‑C 155,8), by
Folquet de Marseille, has a more complex, tangled
poetic structure, with a correspondingly more involved
set of cadential interrelationships. The music is
given below along with the graph of its form. Example
10. Folquet
de Marseille, P‑C 155,8.
The first
point to notice about the poetic form is the degree of
non‑congruence between rhyme pattern and
metrical form. The rhyme scheme would suggest
articulations at the changes of rhyme, since the
rhymes occur in a strictly sequential order. Thus, one
possible division would be:
a
a
/ b b //
c
c c
c /
d d
The use of
three different meters, however, and the order in
which they appear, would produce the following design,
which agrees with that of the rhymes at the beginning
and end, but contradicts it in the middle: a
a
b b
c
c
c
c d
d
8 8 /
10 4
8 //
10 4 8
/ 10
10
The music adds its own measure of complexity to the
structure, confirming some of the prominent points of
articulation suggested by the rhyme scheme and
metrical design, while creating other links beyond
these. The parallelism between lines 3, 4, and 5, and
6, 7, and 8, suggested by the metrical arrangement, is
present in the cadential formations of these lines.
The cadence of line 3 is repeated in line 6, while the
entire melodic material of line 7 repeats that of line
4 at a lower pitch level, with slight variation; the
parallelism is not carried through to lines 5 and 8.
The rhyme scheme articulations after lines 2 and 4 are
also echoed in the cadences; the two a rhymes share
the same pitches in their cadences, varied only by
redistribution over syllables, and a similar matching
can be seen in the cadences for the b rhymes (indeed,
all of line 4 is a minor variant of the music for the
last four syllables of line 3).
An alternate articulation for the latter half
of the song is produced musically through effects of
motivic recall and melodic enjambment. Line 8 opens
with the same initium
as line 1, suggesting a new beginning here, and thus a
point of articulation. Then, the ascending melodic
line, emphasized at the cadence by the interval of the
third, produces an effect of melodic enjambment with
line 9 which immediately descends, so that the two
together form an arch. The beginning of line 10 also
recalls an earlier phrase, this time line 3, in its
five ascending pitches on syllables 2 to 6, and the
cadence of this final line repeats the cadence of the
fourth line. By way of summary, these various links
and articulations are diagrammed below with the rhyme
and metrical schemes.
This concludes the examination of oda
continua forms; a list of all the songs classified
under this heading for at least one of their versions
can be found in Appendix
I.
[1]
Appendix I contains a summary of troubadour song
forms by composer and by category; Appendix II
contains the analytical graphs for each song.
[2]
The songs are: P‑C 323,15; P‑C 70,43; P‑C
155,16; P‑C 155,22; P‑C 155,23; P‑C
167,15; P‑C 167,43; P‑C 167,59; P‑C
364,4, and P‑C 364,39.
[3]
In the musical examples, letters above the beginning
of the
first staff of the song identify the manuscript the
version is found in; where more than one version
exists, the vertical ordering adopted is that of
number of songs contained, merely for the sake of
consistency; that is, the manuscript with most songs
is on the top stave, that with the fewest on the
bottom. The list of manuscripts can be found in
Appendix II with the catalogue of forms. Music and
text for the first stanza follow the edition of H.
van der Werf and G. Bond, eds., The
Extant Troubadour Melodies: Transcriptions and
Essays for Performers and Scholars (Rochester,
1984), unless otherwise noted, except that I retain
the original pitch level of each version, while van
der Werf transposes songs to bring all versions to
the same level. In his edition, square brackets
enclose editorial emendations, and angular brackets
missing passages in the manuscript. In a few
instances, the editor has proposed corrections where
these appeared warranted, and I have adopted these
without comment. The reader is referred to van der
Werf's edition for more detailed discussion of the
manuscript sources and editorial notes for each
song. Pitch designations follow the
medieval convention, with uppercase for the octave
from A to A below middle c, lowercase for the octave
above, and primes to indicate pitches extending in
either direction beyond this range. The hyphen joins
notes in a ligature.
[4]
The third version, found in W, has music for the
first three lines and part of the fourth, due to
damage to the manuscript.
[5]
This system of reference is borrowed from van der
Werf's edition, The
Extant Troubadour Melodies; the number before
the comma refers to the line of verse, the number(s)
after the comma to the syllable(s) in the line (and
where applicable, the music with those syllables).
[6]
"Quia quedam sunt sub una oda continua usque ad
ultimum progressive, hoc est sine iteratione
modulationis cuiusquam et sine diesi." De
vulgari II, x, 2.
[7]
In his edition, Peire
Vidal. Poesie, 390, Avalle describes the poem as
"una delle solite canzoni di maniera ... ricca
di luoghi comuni e di frasi fatte, res
nullius di tanta rimeria occitanica."
[8]
The terms are from the Leys
d'Amors. |
