Machaut-De Toutes Flores

If the surviving sources are at all representative, then De toutes flours (B31) was
one of Machaut's most widely disseminated songs. Not only is it contained in

all but the earliest of the collected `Machaut manuscripts' but it is also known
from seven other sources from the fourteenth and early fifteenth centuries.
1
Although it is perhaps all too easy automatically to equate such relatively
abundant survival with popularity, it seems likely that this song was more
widely known than most of Machaut's output. One of the reasons for this is the
poetic text detailing the lover's angry plea that Fortune not destroy his `rose'
which is concerned with two extremely popular late-medieval themes: the
action of Fortune and the image of the beloved as a rose. The poetic text of B31
appears without music only in the unnotated music section of the collected
Machaut manuscript M (Paris, Bibliothe que Nationale, fonds franc ais 843).
Unlike the other Fortune poems, Il mest avis (B22) and De Fortune (B23), to
which it is linked both thematically and through shared diction and keywords,
B31 is not in Machaut's unnotated lyric collection, the Loange des dames. All
three Machaut musical balades that set texts concerning Fortune are present
outside the core Machaut sources in manuscripts of assorted songs, and all are
copied with extra or alternative voice parts.
2
All three are present in PR along
with an anonymous balade based on B23. In addition to its having an added
triplum in PR, B31's music was well-known enough to merit an instrumental
arrangement in Fa, alongside other examples of songs that have been identified
(for reasons of transmission) as a core ìnternational' repertory.
3
Using a few textually difficult moments in B31, this article shows how an
analysis oriented by (but not limited to) medieval counterpoint teaching can
assist in text-critical matters. The overall tonal planning of the song has been
considered by Jehoash Hirshberg, Peter Lefferts and Yolanda Plumley, and the
piece has recently been thoroughly analysed in this and other details by Sarah
Fuller.
4
However, two small but related moments in the song present
particular textual problems. Their discussion by previous commentators will
provide a context for subjecting B31 to further examination.
Firstly, the marking of the cantus b in bar 15 with a hexachordal fa sign ([)
has variously been treated as an error (Fuller), used to support theories of
hexachordal approaches (Hirshberg), compositional exceptionality (Hirshberg
and Brothers), and to refute counterpoint-based understanding of musica ficta
Music Analysis, 19/iii (2000) 321
Blackwell Publishers Ltd. 2000. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK
ELIZABETH EVA LEACH
INTERPRETATION AND COUNTERPOINT: THE CASE OF GUILLAUME
DE MACHAUT'S DE TOUTES FLOURS (B31)
and analytical approaches with a basis in fourteenth-century theory (Brothers).
Secondly, the appearance of an unusual dissonance between the cantus and
contratenor in the final few bars of the B section (the section before the refrain,
in bar 42) has been explained both as an early instance of word painting
(Wolfgang Do mling) and as a further index of compositional exceptionality
(Fuller). Both this dissonance and another potential contratenor-cantus (Ct-
Ca) dissonance slightly earlier in the same phrase (bar 40) will be discussed in
terms of what they might tell us about scribal and compositional priorities and
the modern interpretation of the normative-exceptional dialectic in relation to
Machaut's compositional practice.
The basic premise of this article is that analysis grounded in the tools
provided by medieval counterpoint teaching can reveal the extent of the
stability of the musical text of medieval songs. Whilst, as with B31, a medieval
song may be available in a varied number of voice parts, the added voices
depend to a great extent on a perception of, in this case, a three-part original as
being fixed in terms of relative pitch, poetic text and overall rhythmic pacing,
but flexible in terms of individual ornamental rhythmic figuration, text
underlay and its ability to support additional parts.
5
Brothers, arguing against a systematic application of ficta in line with prin-
ciples of counterpoint, and Hirshberg, arguing in favour of a hexachordally-
based modal understanding of tonality in Machaut, have both given the
example of one cadence in Machaut's De toutes flours B31 where a notated
accidental seems to contradict the inflection normally associated with a 68
sequence of sonorities. Ex. 1 outlines the source situation against a modern
transcription of the first 16 bars of the piece. Fuller summarises the situation,
outlining two possibilities:
. . . at the end of phrase two, on vergier (breve 15), a notated B flat in the cantus
in two principal sources [Vg and A] contradicts the normal inflection (major
sixth D-B natural to octave C) expected at a cadence and specified at other C
endings in this song. In other copies where B flat stands as a `signature' in the
cantus, the flat is not specified. Trained performers . . . would surely sing B
natural to provide the conventional cadence progression, to fulfil the resolution
of the B natural at breve 14, and to be consistent with the other C cadences in
the piece. Does the notated B flat on vergier signify an express desire on
someone's part to override the conventional performance mode, or is it a scribal
lapse, a premature indication of a restored `signature' B flat? Although a good
case can be made for the second position, based on specific disposition in the
two principal sources, patterns of voice-leading and Machaut's normal practice
elsewhere, the other view cannot be completely ruled out.
6
Fuller's two possibilities (shown as Ex. 1a and Ex. 1c respectively) may be
joined by a third (Ex. 1b) in which the hexachordal sign is interpreted as
implying a knock-on effect in the tenor so that the tenor voice itself makes the
Blackwell Publishers Ltd. 2000 Music Analysis, 19/iii (2000)
322 ELIZABETH EVA LEACH
T
8
G Vg A FP
G Vg A
FP Pit
Ct
8
Ca
G Vg A FP Mod Pit
9
8 na
12
voitt de
G Vg A E
FP Mod
tous fruis
E
2. En mon
G Vg A
ver
16
gier fors
8
8
8
T
Ct
Ca
4 8
( )
( )
Pit
Mod
Mod
1. De
tou tes flours
G Vg A FP Mod Pit
Ex. 1 De toutes flours (B31), bars 116
fruis 2. En mon ver gier
Ca
T
8
8
T Ca
8
Counterpoint
8
Ex. 1a Bars 1516 àt face value'
Music Analysis, 19/iii (2000) Blackwell Publishers Ltd. 2000
GUILLAUME DE MACHAUT'S DE TOUTES FLOURS 323
semitone approach instead of the cantus. These three possibilities will be
discussed below.
Ex. 1a interprets the manuscripts at `face value' as a case of a minor sixth
proceeding to an octave but avoiding the semitone movement of the directed
progression. This has been argued by both Brothers and Hirshberg, both of
whom seek to avoid marginalising or explaining away the idiosyncratic and
exceptional. Brothers quotes Fuller's summary of the situation (given above)
and concludes conversely that the `flat' is:
. . . interesting for two reasons: it `cancels' b-natural in the cantus of two
measures earlier, before this pitch has a chance to resolve to the implied c; and it
undermines the arrival on c in measure 16 by creating a whole step instead of a
half-step leading tone. It is easy to understand how the flat could have been
Ca
T
8
8
T Ca
8
Counterpoint
8
Ex. 1b Bars 1516 with semitone approach in the tenor
Ca
T
8
8
T Ca
8
Counterpoint
8
Ex. 1c Bars 1516 with semitone approach in the cantus
dropped by a skeptical scribe (and Fuller shows herself to be one when she
drops it from her edition of a source that includes it). But the transmission of
the sign is not weak. It is firmly communicated by the principal sources, and it
is dropped only by the secondary sources . . . Rather than imagining, as Fuller
does, that trained singers knew how to `correct' unconventional cadences, I
imagine them being skillful in negotiating those cadences as written.
7
Brothers additionally invokes the text-critical principle of difficilior lectio potior
as a reason for accepting the b-fa in bar 15. His view accords substantially with
that of Hirshberg, who comments that à cadence on c
0
without leading tone is
formed'
8
and that `Machaut's deliberate avoidance of the cadential progression
conflicted with the habit of modern performers of applying a cadential leading
tone'.
9
Although they ascribe the cliche respectively to the habits of medieval
theorists and modern performers, both Brothers and Hirshberg accept the b-fa
and argue that in bars 1516 Machaut deliberately manipulates a cadential
cliche . Objections to their view can be proposed under two related headings.
Firstly, taking notation at `face value' often actually means inappropriately
reading fourteenth-century notation by twentieth-century rules. Asserting, as
Brothers does, that a singer would have skilfully negotiated such cadences às
written' falsely assumes that where the signs of fourteenth-century notation
resemble twentieth-century ones they can be read as if they meant then what
they would mean to us now.
10
Secondly, treating the directed progression as a
cliche is arguably an error of category. The directed progression is not just a
cliche , a normative formula to be abandoned at will, but a fundamental part of
a notational system very different from ours. The norms which composers may
entirely circumvent belong to continuous categories, such as form, style, types
of melody lines, types of interval succession, relationships between open,
closed and final cadences, and so on. These norms are established (by us)
retrospectively on statistical grounds (and all are discussed by Hirshberg to
argue for Machaut's exceptionality as a composer in these respects).
11
By
contrast, the semitone placement required by counterpoint is a discrete
quantity and beneath the notice of this kind of exceptionality it is more
fundamental than that. It is below the level of style which I would argue is
constituted exactly by the relationship of the musical surface to an underlying
conceptual basis in simple counterpoint.
12
The basis provided by counterpoint
teaching constitutes a meeting-point for composer and singer, referring every
singer to the tenor as the locus of musical articulation, and functioning as a
conceptual element in the composer's imagination. To achieve the pitches he
wants (both vertically and horizontally) the composer must manipulate the
content of each vocal line so that, by applying an understanding of a
`contrapuntal grammar' that they both share, the singer may perform the
desired pitches.
13
If this is a fair reading of the notational conventions within which composers
worked (and notation is part of a conceptual world for a composer), then all
perfect sonorities preceded by imperfect ones with at least one voice moving by
step should be interpreted as being approached by a semitone. Since such a
progression opens phrases as often as it ends them, the degree of closure is as
immaterial as (in fourteenth-century notation) is the presence or absence of an
accidental sign (except sometimes to specify which semitone, in cases where
both voices move by step). In our notation so-called àccidentals' are in fact
obligatory and must be specified. Thus, for us, a contrapuntal analysis of the
song precedes and enables the translation of the instructions for singing, in so
far as they are specific to pitch, into our notational-conceptual terms. When
this is done, the articulated surface of the music may then undergo further
analysis of a wide variety of types.
The placing of the semitone in the approach to perfect intervals receives
theoretical justification from many fourteenth-century sources. The author of the
Berkeley treatise, for example, writes that instructions to place semitones àre
often effectively present in b-fa b-mi even though they are not always written'.
14
However, the relationship between theory and practice is questioned by those
who understandably mistrust any systematic approach to this issue.
15
As
medieval composers appear to flout many theoretical recommendations, theory's
prescriptions may be thought incompatible with composers' realities. Theorists'
own disagreements and inconsistencies reinforce this view. However, Margaret
Bent has shown that the basic agreed tenets of the theory fit the music better than
its detractors would have us believe.
16
She comments directly on the issue of so
called `signatures', and accidental signs, saying that manuscript and theoretical
evidence only appear in conflict when seen from `the premise that absence of a
notated inflection or presence of a signature were prescriptive, as in modern
notation'.
17
Instead she interprets accidentals and `signatures' as weakly
prescriptive and apt to being overruled by contrapuntal necessity.
From evidence internal to Machaut's music itself it is possible to show that,
for him at least, the directed progression (the correct approach to perfect
intervals) should always involve a semitone approach. The evidence for this is
drawn from the way in which Machaut plays with, exploits and writes around
such progressions. Where both the imperfect and perfect elements necessary to
form a directed progression are present, Machaut will frequently delay the
imperfect element in some way, reduce its actual sounding value, or present its
two notes non-simultaneously (in hocket form, for example). Conversely he
uses the expectation created by mensurally strong, durationally long imperfect
sonorities to create forward movement, even in cases where immediate
resolution of these sonorities is absent.
18
Ex. 2 is taken from Machaut's two-part balade Pour ce que tous (B12). The
mi sign in bar 14 in all manuscripts serves to prevent the cantus singer singing
b-fa.
19
For a medieval singer, b-fa and b-mi are equally available recta pitches
and several factors may affect a singer's choice. Reading from a single part he
might well assume that his b above a held tenor d
0
at the end of a poetic line
when the next cantus note is c
0
would be a directed progression whose force
overrules that of mere melodic smoothness. Therefore, even in the absence of
any hexachordal sign and despite the melodic descent from e
0
-fa, he would
probably initially sing (the composer's desired) b-mi. Then, discovering that
the linear descent from e
0
-fa is not overruled by a directed progression, since
the tenor avoids its note of resolution (the expected c), a singer too clever for
his own good might employ the weaker default of considerations of line. To
avoid a linear tritone from e
0
-fa he would then sing b-fa on a second run-
through. To forestall this and signify that a juxtaposition of e
0
-fa and b-mi is
required and that the listener and not the singer should be caught out by
avoidance of the expected resolution, the mi is marked. That all sources
nevertheless feel the need to copy an accidental that in twentieth-century terms
is totally redundant, points once more to the fundamentally different
conception of pitch in the fourteenth-century.
This instance is one of many that could be cited to show that Machaut uses
held imperfect chords to set up expectations that, because the semitone
approach is not a stylistic convention (and therefore optional) but a notational
one (and therefore embedded conceptually in the compositional framework),
he can then exploit. Using the implication of a semitone inflection to tease the
listener depends on the directed progression's notational (and conceptual)
5. Mais qui vrai e ment sa roit 6. Ce
(All) ( ) ( ) (All)
C Vg B E G A
C Vg B G A C Vg B G A
8
8
Ca
T Ca
T
8
8
T Ca
Underlying counterpoint
x
(avoided)
Ex. 2 Pour ce que tous (B12), bars 1315
obligation. The expectational force of a held imperfect sonority would surely
be diminished if elsewhere a stepwise link between an imperfect sonority and a
perfect sonority were sometimes a matter of a whole tone that is, if the
directed progression were ever optional.
In bars 1516 of De toutes flours (B31), the dyad succession 68 occurs in the
tenor-cantus (T-Ca) duet, at a point of poetic articulation the caesura of line 2
8
8
mour 3. Car il
In: C Vg B G A E
C Vg G A
8
8
NOT
< >
G C
Ex. 3a On ne porroit (B3), ouvert ending and return to opening
da mi peut de si rer. 3. Et
G Vg B
C A E G Vg B
C A E G Vg B
In: C Vg B G A E
8
8
8
8
8
< >
8
< >
8
8
NOT
Tr
Ca
Ct
T
T Ca
T Ct
Ex. 3b Se quanque (B21), ouvert ending and return to opening
which parallels the similar cadence at the caesura of line 1 (bars 78). Under what
circumstances would the cantus singer be persuaded not to inflect his part when
the elements of a directed progression are presented clearly and adjacently (and
additionally cadentially)?
20
The closed and final cadences (the end of the A
section and the end of the song, respectively) are also 68 progressions to the c
octave, approached in the cantus from b-mi. I am sceptical that an accidental (and
notationally extraneous) sign in itself would be sufficient to break the pattern set
up by this framing, let alone to bring one aspect of the notational system (a
semitone inflection caused by the use of solmisation signs) into conflict with
another (a semitone inflection necessitated by counterpoint). Other songs suggest
that in order to notate the succession of two dyads that might have formed a
directed progression but without the necessary semitone movement in either part,
Machaut prevents stepwise voice-leading by crossing the voices. This avoids a
semitone connection between the individual elements of the two dyads in any
single voice. An example drawn from On ne porroit (B3) will illustrate this. Many
of Machaut's open endings lead back by directed progression to the opening
sonority, some of which require a ficta note and some of which simply encourage
a certain recta choice.
21
However, in B3 the ouvert sonority <a-c
0
> that appears
to augur one resolution (to b-fa unison), actually proceeds to g5 without a
semitone (see Ex. 3a).
22
Similarly, in Se quanque (B21, Ex. 3b), that the Ca is
below the T for the ouvert sonority <e-g> prevents a ficta adjustment of the Ca
e, because the opening sonority d8, whilst a resolution of the sonority e3, is not
presented as such in terms of its voice-leading. Neither voice moves by step: the
cantus leaps a seventh from e to d
0
and the tenor falls a fourth from g to d.
23
The evidence gleaned from the use of expectation-building imperfect
sonorities is reinforced by the use of voice-crossing to write around the
inflection of the semitone, both implying that the directed progression is not
just a singerly practice but a notational-conceptual one that cannot simply be
cancelled by an accidental sign but has to be compositionally avoided. If
Machaut had wanted to break the pattern of semitone approaches to the c
octave in balade 31, he could (and arguably, to be entirely clear and not run the
risk of subsequent scribal normalisation, would have had to) have disposed the
relevant notes differently between the voices. I therefore reject the reading of
Ex. 1a, which may be discounted from the possibilities available, and propose
that some kind of directed progression, complete with semitone, occurs in bars
1516, and that the fa sign must therefore mean something else.
*
One possible interpretation of the b-fa sign (Ex. 1b) is that it implies a tenor d-
flat, an inflection that might be confusing were it to be signed directly. If sight-
reading, then the cantus singer might solmise the notation as in the now-
rejected Ex. 1a b-fa c but both tenor and cantus singers would be aware that
the perfect sonority the octave c had been approached incorrectly. In
general, a counterpoint inflection is more likely to be in the cantus voice, but
here the sign in the cantus voice tells the cantus singer not to inflect his part
and thus implies that the tenor singer should provide the semitone. Musically
this would be highly significant for the piece. Setting up two different
solmisations of the same imperfect sonority in approaches to the most
important cadential goal of the piece would be an èxceptional' but not
unprecedented stylistic feature. It occurs, for example, in Riches damour (B5),
connecting the end of the first poetic line to the beginning of the second. Ex. 4
shows the manuscript situation.
24
What would otherwise be a simple matter of
a directed progression involving the tenor choice of the recta note b-fa over b-
mi, is complicated by the marking of the tenor b with a hexachordal mi sign in
all sources. The meaning of the sign is clarified for twentieth-century eyes in
three of the sources: the earliest (C); the one most closely associated with
Machaut (A); and the one that is probably posthumous but that is in the earlier
two-column format and that, arguably, transmits some early texts (G). Instead
of being an avoidance of the directed progression (as the modern collected
editions would have us believe), the mi sign placed in the cantus in bar 14
reveals that the directed progression is to be effected by the very unusual ficta
note d-mi. That the cantus d
0
is not marked with a mi sign in some of the
sources, but the tenor b is marked in all of them, strongly suggests that, far
from the hexachordal sign being able to deter singers from correctly
8
8
8
men di ans da mi e 2. Po vres des poir
C Vg B E A G
C G A
C Vg B E A
15 10
Ca
T
T Ca
< >
8
Underlying counterpoint
Ex. 4 Riches damour (B5), bars 1017
approaching a perfect sonority by preventing a semitone movement, the
marking of the tenor alone is sufficient to `notate' the d
0
-mi because of the
ingrained `contrapuntal grammar' of placing semitones in a correct approach to
perfect consonances. This piece shows that the very prescription inherent in
the principle of nearest approach, which raises objections in those opposed to
its systematic application, allows composers to do strange and idiosyncratic
things. Exceptional and unusual lines and harmony occur because of, and not
despite, the necessity for semitone placement in such sonority successions. In
B5, observing the rules of counterpoint is more important more fundamental
than avoiding an augmented leap or an unusual ficta note. Having negotiated
the correct but unusual solmisation (which might take singers two attempts),
the singers are wittily applauded by the text as they pronounce the correct
solmisation syllable `mi' to their held notes in bar 14, where it forms part of the
first rhyme word of the first line of the first stanza `damie'.
25
Thus there are precedents for the argument that a tenor d-fa is implied by
the cantus sign in bar 15 of B31. To my knowledge, no-one has made this
suggestion, although it would be more consistent with an understanding of
fourteenth-century cadences and counterpoint than taking the sign at face
value as a total avoidance of semitone movement. In the context of this article,
Ex. 1b forms one of two possible readings. However, I shall argue below that it
is one that, while possible, does not seem likely, because it breaks an otherwise
regular pattern of cadences in B31. In the larger musico-poetic context of the
balade as a whole, this regularity makes possible an extensive and elaborate
game that will be examined in the next section. The other balades in which
Machaut uses hexachordal signs in one part to bring about unusual ficta in
another, lack such pattern and have a more diverse tonal logic within which the
differently solmised approach to the same perfect sonority is meaningful.
*
The final interpretation (and the second of two that preserve the semitone
approach to c8) is that the cantus ìgnores' the fa sign and approaches bar 16's c
0
fromb-mi. This means that the fa sign in bar 15 should be omitted from modern
editions, but this does not make it an error in fourteenth-century terms. If, as I
am arguing, its lack of relevance to the note immediately following would have
been contextually clear, then this implies that the sign has a more general
default meaning. This is the implication behind the omission of the sign from
both Leo Schrade's and Sarah Fuller's editions. As Fuller states, à good case
can be made [for omitting the fa sign] based on specific disposition in the two
principal sources'.
26
The `specific disposition' referred to seems to allude to the
layout of the sources. Machaut manuscript C has a two-column format that
implies that this was the earlier layout. Although a late source, G is also copied
in a two-column format; all the other manuscripts are in single columns.
27
The
Machaut sources all have an initial fa sign in the cantus that is not repeated at
the start of every line.
28
Signature or default signs occurring at the beginning of
a line or column in a two-column format are often copied at the same musical
point at least half the time now visually mid-line in later sources, to become
so-called `floating signatures'. As the original sign was a `signature', a weakly
prescriptive default setting (telling the singer that, except where counterpoint
demands otherwise, b-fa rather than b-mi should be selected), the repositioning
of the sign when subsequently copied into the one-column format is usually of
little consequence.
B31 is not present in the other two-column source, C. In manuscript G,
however, the fa `signature' of B31 occurs not just at the beginning of the first
line, but also at the top of the second column of the folio (see Ex. 5). The
column break happens to precede the note b (bar 15), which therefore follows
directly after the restatement of what the mis-en-page suggests is merely a
global (default) b-fa after the clef at the top of the column. This visual context
would aid the singer's recognition of the sign as a default `signature' rather
than as an immediately applicable local sign.
Default settings were alterable locally for directed progressions. Just as the
initial (`signature') b-fa does not affect the b-mi contrapuntally required in bar
7, neither does the hexachordal fa sign, which represents a global default,
overrule the contrapuntal necessity of b-mi in bar 15, particularly when the
musical sense of the phrases is that the resolution to c8 in bars 1516 clearly
parallels the one in bars 78. On a first reading, a somnambulant fourteenth-
century singer might perhaps have been confused, but could subsequently have
correctly identified this `fa' sign as global, rather than local, in meaning. For
him, it is not an error as such it just means something else, a `something' that
such signs can no longer mean in our usage. In the later single-column sources,
this sign is copied immediately before the same note that now occurs mid-line,
where, to the modern eye at least, its positioning bestows upon it a value
greater than any in the original format. However, even here it is no more
specific in meaning it is still a global sign, restating a general default after the
ùpset' of the b-mi (signed in bar 14 and not inferable from counterpoint
alone). In a modern score, assisted by a strict bar's length acting of non-
signatures and a required following of signatures unless explicitly cancelled by
signs, the global/local distinction is much more tightly defined.
This reading implies almost the opposite of Brothers's and Hirshberg's
claim that the fa sign in bar 15 `cancels' the mi sign in bar 14. Instead, it is the
mi sign of bar 14 that temporarily suspends a default b-fa setting. For Brothers,
the cancelling of the b-mi annuls the expectation before it has time to resolve
and thereby ùndermines the arrival on the c in measure 16 by creating a whole
step instead of a half-step leading tone'.
29
Reversing the direction of
Ex. 5 De toutes flours (B31) in MS G
Photographic reproduction by the Bibliothe que nationale de France.
cancellation, however, the b-fa default is explicitly suspended by the mi sign in
bar 14 because a deviation from the default is discouraged by the melodic line,
which does not proceed directly to the note of resolution, c
0
. After bar 14, the
default setting is restated in a few sources and, even though the note b follows
the fa sign immediately, it is sung mi because counterpoint considerations
overrule a default hexachordal solmisation, even one so adjacently stated.
30
In
B31, by coincidence, several sources happen to place the sign in this place,
which seems to lend it more authority. Such coincidence is hardly surprising
given the evidence for the interdependent copying of the Machaut
manuscripts, their possible derivation ultimately from a single authoritative
exemplar, and indeed by the sheer force of probability in such a large output.
The above interpretation (Ex. 1c) is supported by the notation of the piece
outside the collected Machaut sources. The cantus in these other manuscripts
notates the mi sign in bar 14 and does not notate b-fa in bar 15 because every
new staff line re-states a default (or signature) b-fa.
31
Invoking the philological
principle of difficilior lectio potior privileging a more unusual reading so as
not to marginalise idiosyncrasy Brothers interprets the reading of the
Machaut manuscripts (with locally notated b-fa in bar 15) as more difficult, and
therefore more likely to represent compositional exceptionality. The later non-
Machaut musical anthology sources (with no locally notated b-fa) are seen as
having been subject either to scribal hyper-correction or ignorance.
32
Although
generally a sound principle by which to distinguish authorial from scribal
initiative and older sources from more recent non-authorial circulations,
difficilior lectio potior requires that all readings considered should first be
possible. In a literary text, those readings deemed nonsense against a semantic
yardstick are discarded so as to avoid accepting errors in attempting to
privilege authorial exceptionality. I am arguing that the directed progression is
similarly a basic unit, whose positive, forward manipulation (as a creator of
expectation) relies on the impossibility of its negative manipulation, that is, its
abandonment. If desired, the effect that abandoning the semitone approach
would have had in terms of sonority succession could be achieved by the
crossing of voices, but the semitone component of a directed progression is too
fundamental to be prevented by a sign alone. Counterpoint, therefore, almost
becomes to a musical text what semantics is to a literary one, and can help us
dismiss errors from the choice of possible readings.
Brothers' use of difficilior lectio potior is also influenced by a division of the
sources for B31 into two distinct and unequal groups: the primary sources (the
`Machaut manuscripts') and secondary sources (all other sources). Whilst such
a division may seem both obvious and useful, it merits closer scrutiny. As both
scribe and poet, Machaut has a dual authority in the context of the late-
medieval culture of the book.
33
Sylvia Huot has interpreted the scribal
ordering of manuscript collections in the thirteenth and early-fourteenth
centuries as an exercise of authority and an attenuated manifestation of
authorship.
34
The choice and order of items for a single codex can create
meanings additional to those of the individual items themselves, thereby
èstablishing the affinities between scribal and poetic practices'.
35
Machaut
probably oversaw the copying of at least two of the codices which survive C
and A, the latter of which carries the famous index rubric `vezci lordenance que
G. de Machau wet quil ait en son livre' (behold the order that Guillaume de
Machaut wishes to have in his book).
36
Machaut's authority manifests itself
chiefly as the interest of a scribe in overall order. Whilst the Machaut
manuscripts are thus authoritative in terms of the additional means gained
from such overall order, copy-editing at the level of the individual piece or
poem was not part of Machaut's remit. Errors could and did enter the
transmission, and remain there, as Lawrence Earp has shown.
37
Furthermore,
in the Prologue to the later anthology MSS, Machaut personifies Musique as the
third child (and the younger daughter) of Nature, presented to him as one of
the technical means by which the content of his poems will be animated;
Machaut is identified by his younger contemporary Eustache Deschamps as a
poe te and not as a composer.
38
Thus, the prevailing aesthetic is more literary
than musical in our sense, something not true of the non-Machaut sources
which are not literary, but musical, anthologies. As John Na das has
commented about PR:
. . . even if gatherings 6 and 7 of the Reina Codex are not the ideal source for the
French repertory of the late fourteenth century, we might still appreciate the
fact that this section of the manuscript . . . may have transmitted certain features
of a composition properly and others incorrectly. The editor's quest need not be
for manuscripts that are on the whole more correct than others, but rather for
individual readings and even certain elements of readings, such as texting and
vocal scoring that are more correct.
39
The non-Machaut sources do not transmit the poetic texts fully and the pieces
are removed from their meaning-giving order. In this respect, much of
Machaut's meaning is lost. Preserved, however, is their musical substance.
Some non-Machaut manuscripts may therefore represent a specifically musical
tradition for Machaut's songs that might transmit certain elements more
reliably than the Machaut manuscripts themselves.
40
As Na das urges, only `[a]
careful codicological and paleographical study of the great late-medieval
anthologies of secular polyphony can form the proper basis for an analysis of the
readings they contain'.
41
As a general observation, those inflections required by
directed progressions are far less frequently signed in the non-Machaut sources
than in the Machaut manuscripts, whilst those which set general defaults
(`signatures') are more frequently used.
42
This could indicate that more singerly
knowledge is assumed outside the Machaut manuscript tradition than within it,
(after all, by the late fourteenth century these now-famous songs are imprinted
on singers' memories far more than at the point of the copying of the Machaut
manuscripts), and that therefore, as far as the music is concerned, the readings
of these sources are not to be dismissed lightly.
Simply inverting the importance of the two categories of manuscripts
transmitting Machaut's work runs the risk of both reinforcing the duality
between the two groups and participating in the same value-system, merely
basing àuthenticity' a concept which is inappropriately value-laden on new
criteria. Although it is of obvious benefit to the musicologist that the Machaut
manuscripts have survived, a blanket ascription of power and authority to them
does not make sense for every last detail of the music, since it privileges a
literary musical aesthetic over an arguably more musical one (and a poietic one
over an esthesic one). The non-Machaut sources are central to the reception
history of Machaut's songs. Their witness to the lack of a hexachordal fa sign in
bar 15 is an important indicator of how this famous song was understood.
It is also possible to argue for the importance of the b-mi in bar 15 from an
analysis of the musico-poetic structure of B31. The whole song plays with a
clever poetic structure in which internal, caesural rhymes are privileged so that
the listener is disorientated as to the poetic (and, by extension, the musical)
structure.
43
Fuller finds that `[t]he formal regularity of the poetic stanza is . . .
quite upset in the musical setting' because `Machaut is far less concerned with
the balance of ten-syllable lines than with syntax and nuance of language'.
44
I
would argue, however, that the stanzaic regularity is not so much ùpset' as
offset. As well as an undoubted `concern for syntax and nuance of language',
this offsetting is integral to an elaborate musico-poetic game which relies on
the exploitation of just such poetic regularity for its working.
The basic premise of the game is to suppress real rhyme words, to use
musical phrases to privilege caesural words, treating them as if they were
rhyme words, and to use the verse structure to confuse the listener as to when
the music finally makes poetic lines congruent with musical phrases. The
functioning of the game is allowed by the careful establishment of expectation
in the A section. Table 1 shows the text as sectioned by the music juxtaposed
with the formal poetic structure. The first musical phrase sets the first four
syllables of the first line up to the caesura of the first line at the word `flours'
that anticipates the real c-rhyme. The musical phrase is closed by a T-Ca
cadence to c8 in bar 8. The real a-rhyme at the end of the first line is internal to
the second musical phrase that sets the text from the caesura of line 1 to the
caesura of line 2. The a-rhyme appears on the b-mi of bar 14 and is thus
connected to and resolved by the caesural word of line 2, `vergier' in bar 16.
Significantly, `vergier' anticipates the real d-rhyme that will end line 7 and
which retrospectively will break up the succession of offset lines and
rhymes and force acknowledgement of the musico-poetic deception.
The beginning of the B section in Machaut's balades frequently a locus for
exploring novel or extreme elements of range, harmony and rhythm is where
the text-music game is most prominent. This verse form (one of Machaut's
favourites) has eight lines, seven of which are decasyllabic but one of which
the fifth line has only 7 syllables. This fifth line occurs at the beginning of the
B section and is set to a musical phrase that does not end until the T-Ca
cadence to c8 at the caesura of line 6. Although there are in fact 11 syllables in
this phrase, it is almost as if the musical phrase were coterminous with a full-
length (decasyllabic) text line as if the game of the A section has been
abandoned. The word at the end of this musical phrase (bar 37) is àmatir'. The
next musical phrase, which in reality sets the text from the caesura of line 6 to
the caesura of line 7, confirms the (false) perception of àmatir' as the end of a
line and a rhyme word (the pseudo c-rhyme), by ending with the word
`cueillir'. But, as the listener is now forced to acknowledge, something is not
right. The second T-Ca c8 cadence (bar 41) is metrically offset, it is overlaid
with a T-Ct directed progression creating an extreme dissonance (bar 42; see
below) which connects it to the remainder of the B section text. These final six
syllables of line 7 (bars 436) retrospectively reveal the true poetic structure of
the B section. These six syllables are separated from the body of the poem
thereby revealing the prior musico-poetic deception (and forewarning of the
deception that will be the cause for the narrator to think that Fortune has
plucked his rose).
The anticipation of the B section's rhymes in the A section's caesuras aptly
occurs only in the first stanza after this it would cease to be anticipation (the
real c- and d-rhymes already having happened). However, in the final stanza
the `fake' internal rhyme between the caesura of line 6 (bar 38) and that of line
7 (bar 42) involves the words `plour' and `tour' echoing firstly the c8 cadence
of the first phrase of the first stanza (`flours') which ushered in the start of the
musico-poetic deception. Because `flours' at the caesural cadence of the first
line foreshadowed the real c-rhyme, its ècho' in the caesuras of lines 6 and 7 of
the last stanza creates an identity between the fake c-rhyme and the real c-
rhyme. At a third run-through, when the deception might finally have been
untangled, this further layer of deceit and confusion is added, linking
Fortune's false `honnour' with `plour' (tears) and her honour's reality
`deshonnour' with the `faus tour' (false trick) that will cause the speaker's
rose to wither.
The musico-poetic parallel between bars 1316 and 3942 (which empha-
sises the second caesura of the A and B sections respectively) promotes more
purely musical similarities. The correspondence of their accidental signs
suggests strongly that they should be performed identically.
45
The b in bar 14
and its parallel in bar 40 are both helpfully marked with a hexachordal mi-sign,
because the recta choice for the mi is not prescribed by immediate localised
Table 1 The poetic text of B31 divided by musical phrases and cadences
Musical Text in musical units psuedo sonority Text in poetic units real sonority
Form pseudo rhymes/real rhymes rhyme real rhymes rhyme
Stanza I
A1 De toutes flours a = c c8 De toutes flours navoit et de tous fruis a d-b
navoit et de tous fruis en mon vergier b = d c8 en mon vergier fors une seule rose b d8
ouvert fors une seule rose b d8
A2 Gastes estois c8 Gastes estois li seurplus et destruis a d-b
li seurplus et destruis par fortune c8 par fortune qui durement soppose b c8
clos qui durement soppose b c8
B Contre ceste douce flour pour amatir c c8 Contre ceste douce flour c d-b
sa coulour et sodour mais se cueillir c c8 pour amatir sa coulour et sodour c g-b
le voy ou tresbuchier d d8 mais se cueillir le voy ou tresbuchier d d8
R Autre apres li jamais avoir ne quier d c8 Autre apres li jamais avoir ne quier d c8
Stanza II
A1 Mais vraiement c8 Mais vraiement ymaginer ne puis a d-b
ymaginer ne puis que la vertus c8 que la vertus ou ma rose est enclose b d8
B
l
a
c
k
w
e
l
l
P
u
b
l
i
s
h
e
r
s
L
t
d
.
2
0
0
0
M
u
s
i
c
A
n
a
l
y
s
i
s
,
1
9
/
i
i
i
(
2
0
0
0
)
3
3
8
E
L
I
Z
A
B
E
T
H
E
V
A
L
E
A
C
H
ouvert ou ma rose est enclose b d8
A2 viengne par toy c8 viengne par toy et par tes faus conduis a d-b
et par tes faus conduis eins est c8 eins est drois dons naturex si suppose b c8
drois dons
clos naturex si suppose b c8
B Que tu navras ja vigour damentir son c c8 Que tu navras ja vigour c d-b
pris et sa valour lay la moy donc (c) c8 damentir son pris et sa valour c g-b
quailleurs nen mon vergier d d8 lay la moy donc quailleurs nen mon vergier d d8
Stanza III
A1 He fortune c8 He fortune qui es gouffres et puis a d-b
qui es gouffres et puis pour engloutir c8 pour engloutir tout homme qui croire ose b d8
ouvert tout homme qui croire ose b d8
A2 ta fausse loy c8 ta fausse loy ou riens de bien ne truis a d-b
ou riens de bien ne truis ne de seur c8 ne de seur trop est decevant chose b c8
clos trop est decevant chose b c8
B Ton ris ta joie tonnour ne sont c8 Ton ris ta joie tonnour c d-b
que plour c = c !
tristesse et deshonnour Se ti faus tour c = c ! c8 ne sont que plour tristesse et deshonnour c g-b
font ma rose sechier d d8 Se ti faus tour font ma rose sechier d d8
M
u
s
i
c
A
n
a
l
y
s
i
s
,
1
9
/
i
i
i
(
2
0
0
0
)
B
l
a
c
k
w
e
l
l
P
u
b
l
i
s
h
e
r
s
L
t
d
.
2
0
0
0
G
U
I
L
L
A
U
M
E
D
E
M
A
C
H
A
U
T
'
S
D
E
T
O
U
T
E
S
F
L
O
U
R
S
3
3
9
directed progression and is, in fact, discouraged by both the global b-fa setting
and the local linear descent from a cantus figure turning around e
0
-fa (bars 13
and 39). The function of b-mi in both places is to extend the length of the
directed progression by exploiting the expectation that imperfect sonorities
create. To do this, the b-mi is needed a little sooner than is strictly inferable
from counterpoint rules which work with adjacent sonorities in simple
counterpoint.
46
The resolution sonority, c8, does not occur until the second
half of the breve (bar 41) and whilst Mod Pit and FP restate the default b-fa
immediately after it in bar 42, PR restates its fa sign immediately before the b-
mi in bar 41. Although a better musical case could be made for bar 41 than for
bar 15 (since the phrase is joined to the following one by an overlaid T-Ct
directed progression that might warrant the diffusing of the c8 arrival), to
apply difficilior lectio potior here would be somewhat perverse (and no editor
hints at such a practice, probably because it is only a case of one non-Machaut
source against others). If b-mi is to be accepted here (which no editor seriously
questions), then it is very likely that b-mi should also be sung in bar 15.
To summarise: the caesural c8 cadences of lines 1, 2, 3, 4, 6 and 7 acquire
value because c8 is the terminal section-end goal sonority. However, only
lines 4 and 8 of each stanza (made identical through `musical rhyme' the
occurrence of the same extensive closural passage to end both A2 with the
closed cadence and the refrain) have such a cadence at their end. Lines 2 and
7 have directed progressions whose resolution sonorities are the secondary
terminal goal d8. All other text lines have b-mi in the cantus and form an
imperfect sonority with the tenor (see Table 1). This is mostly d-b-mi, that
is, with the tenor note of the secondary terminal sonority (d8) but a sonority
of imperfect quality, making it both tonally òpen' and contrapuntally
unstable. If lines 1 and 3 were to have b-fa in bar 15, then the patterning of
cadences, and in particular the paralleling of the caesuras of the A section
with the caesuras of the B section (which the A section caesura words also
anticipate in the real text line ends of the B sections), would be considerably
weakened for no real gain.
47
*
One further text-critical issue may now be approached. In bars 4042 there are
two striking Ct-Ca dissonances, one of which is caused by the cantus b-mi at
the end of line 6 (bar 40). Since this is the moment the offsetting game
disintegrates, it is a particularly climactic one. The musical phrase that starts in
bar 38 defines its opening with a directed progression in both tenor duets to e-
fa5/8 (bar 39), the contratenor sings b-fa. The contratenor then restates the b-fa
at the start of bar 40 as an ornament (effectively a `re-sounded' suspension) as it
descends to its contrapuntally essential note, f-mi, which conceptually occupies
bars 4041. This means that the b-fa of the contratenor's resolution, whilst
conceptually over by the end of bar 39, is actually still sounding when the
cantus's b-mi for its progression to c
0
is first sounded (bar 40) (see Ex. 6a). The
only alternative would be for the contratenor to make a chromatic movement
from b-fa to b-mi between bars 39 and 40. Although similar chromatic
movements occur at the level of simple counterpoint in some of Machaut's
other songs, they are always carefully disguised (and thereby facilitated?) by
surface ornamentation, possibly pointing to their unacceptability or difficulty
at a surface level. Although the contratenor-cantus relationship is not bound by
the rules of counterpoint, since each has a discant relationship ìndependently'
with the tenor and not with each other, and although working within a dyadic
harmonic system, Machaut has a clear concern for overall sonority (as is
evinced by his four-part writing).
48
Were the b-fa longer and also, like the
cantus's b-mi, part of the underlying simple counterpoint with the tenor at this
point, this would offer grounds on which to suggest emendation. However, the
dissonance lasts only a minim and the contratenor b-fa is no longer a
contrapuntally essential note of the T-Ct duet when its sounding forms the
dissonance with the cantus. Thus, the brief b-fa-b-mi dissonance seems to be
plausible and stylistically possible.
49
This brief but extreme dissonance may serve an added function. It may
prepare for (and retrospectively be revealed not to be an error by) the following
and more lengthy Ct-Ca dissonance in bar 42 see Ex. 6b. Here both tenor
duets have directed progressions, which are offset so that the T-Ca resolution
37 40
Pit
Mod
G Vg A FP G Vg A FP E
G Vg B A
FP Mod
Mod
Pit
8
8
8
tir sa coulour et so dour 7. Mais se cueil lir le
8
8
T Ca
T Ct
a b
Ex. 6 De toutes flours (B31), 3842
to c8 is overlaid with a tension chord for a T-Ct progression to B-fa8. This
would work unremarkably if it were not for the cantus's falling back to g just as
the contratenor rises to a, giving a dissonance of a second for the whole breve
unit (bar 42). As this cantus note is not strictly part of the directed progression,
it would be possible to emend it to f or g. Fuller writes that:
The preparatory element (breve 40) is a third, not a sixth. The sixth arrives but
its resolution falls early, on the second part of breve 41, before the phrase ends.
The cantus subsequently drops to rest on G, seemingly oblivious of the
contratenor a. This (breve 42) is the point of utmost dissonance and tension in
the song, a moment due not to sonority and pitch relations alone, but to the
rhythmic framework within which they are cast.
50
In a footnote, Fuller considers the dissonance in bar 42 to be augured by the
brief but identical Ct-Ca sonority in bar 15 (its parallel phrase) although this is
much shorter and clearly part of the decoration of the consonances forming the
underlying simple counterpoint. That all types of sources contain this
dissonance furthers editorial reluctance to emend this dissonance. The poetic
text has also been cited in support of the dissonance. Wolfgang Do mling, for
instance, considers the dissonance to be explained by the word `cueillir' which
may be additionally signified by the downward leap of a fourth in the cantus.
He also notes that the dissonance is unusual (unprecedented?) in Machaut's
works organised at the breve level.
51
Perhaps there is some sort of localised
notional organisation at the modus level here, created by the stringing out over
almost two breves (and several tenor notes) of the cantus b-mi in bar 40 before
its resolution in the second half of bar 41 (arguably resembling the similar
instance in bars 1415).
*
Analysis led by counterpoint considerations has helped fix the musical text,
both in the sense of stabilising those of its unusual elements that are part of the
design, and in the sense of adjusting those that can be shown to merit further
interpretation or emendation. It is worth noting that, in fact, nothing has been
identified as an error in fourteenth-century terms, although to signify the
sounds implied in the fourteenth-century notation, ours requires different ones
to be added or theirs to be omitted. The analytical premises have, in their
working out, revealed much about the interpretation of signs and the
acceptability of levels of dissonance in this style. I have argued that the very
directionality central to the so-called directed progression means that it can
only be manipulated in one direction. Therefore, those instances that seem to
us to evade the necessary semitone either imply the exceptional use of an
unusual ficta note to provide the semitone or are simply misleading when
transcribed into a modern score, translated into which context they warrant
omitting (or moving) to render their original meaning. As a basic part of the
notational-conceptual framework, simple counterpoint is useful in identifying
inflections not given a separate sign in the original notation. It is against basic
counterpoint that we can describe the individuality and exceptionality of
Machaut's songs in terms of their dissonance treatment and, in particular, the
dissonances permissible between voices not in discant relationship. Although it
may not be sufficient for all our analytical needs, it is a necessary starting point
because of the role it serves in generating the text for analysis, that is, the piece
in our notational-conceptual translation. Counterpoint does not prevent
exceptionality but, rather, provides a framework within which it is made
possible. Exploiting its possibilities and playing imaginatively within its
parameters is Machaut's skill at which (to the extent that these groups are
separate) singers and listeners alike must marvel.
Appendix A
Key to Musical Examples
Abbreviations
T Tenor
Ct Contratenor
Ca Cantus
Tr Triplum
The musical examples aim to give some idea of the manuscript situation for
each passage. Accidentals marked on the staff carry above them the sigla of the
sources in which they occur. No accidentals carry the force of a modern
signature. Instead, all accidentals last the bar's length imposed by the modern
bar line. A circled siglum above an accidental indicates a start-of-line
placement for the sign in that particular source. Editorial accidentals realising
contrapuntal semitones are indicated above the staff.
The contrapuntal analysis is shown in an arhythmic format. In the initial
parsing, dissonances are shown with crossed noteheads. In both the initial
parsing and the analysis of underlying simple counterpoint, consonances are
shown with filled noteheads for imperfect consonances (thirds, sixth and their
compounds) and empty noteheads for perfect sonorities (unisons, fifths,
octaves and their compounds).
The bar numbers for B31 accord with those in Fuller. The bar numbers for
other examples accord with those in Schrade.
Editions
Ludwig: Ludwig, Friedrich (ed.), Guillaume de Machaut: Musikalische Werke, 4
vols. (Leipzig: 192654).
Schrade: Schrade, Leo (ed.), The Works of Guillaume de Machaut, 2 vols.,
Polyphonic Music of the Fourteenth Century (Les Remparts, Monaco:
L'Oiseau-Lyre, 1956).
Fuller: Fuller, Sarah, in Mark Everist (ed.), Models of Musical Analysis (Oxford:
Blackwell, 1992), pp. 5961.
NOTES
1. De toutes flours (B31) is in the Machaut manuscripts G (Paris, Bibliothe que
Nationale, fonds franc ais 225456), E (Paris, Bibliothe que Nationale, fonds
franc ais 9221), Vg (New York, Wildenstein Collection [no shelfmark]) and its
copy B (Paris, Bibliothe que Nationale, fonds franc ais 1585), A (Paris,
Bibliothe que Nationale, fonds franc ais 1584) and its copy Pm (New York,
Pierpont Morgan Library, M 396). With the exception of E, the song appears in
three parts with contratenor. In addition E transmits a triplum. The other
sources for B31 are the late fourteenth-century Florentine musical anthology
manuscript FP (Florence, Biblioteca Nazionale Centrale, Panciatichiano 26), the
early fifteenth-century Italian manuscript Pit (Paris, Bibliothe que Nationale,
fonds italien 568) and the fifteenth-century settentrional source Mod (Modena,
Biblioteca Estense, a.M.5.24). It is listed in the index which survives from the
fourteenth-century royal manuscript Tre m (Paris, Bibliothe que Nationale,
nouvelles acquisitions franc aises 23190), appeared in the German source Str
(Strasbourg, Paris, Bibliothe que Municipale, M. 222 C. 22, a MS destroyed in a
fire in 1870 and accessible only through a partial transcription made by
Coussemaker) and found in an instrumental intabulation in the Ferrarese source
Fa (Faenza, Biblioteca Communale, 117) from the second decade of the fifteenth
century. It is transmitted in the sixth fascicle of the Veneto source PR (Paris,
Bibliothe que Nationale, nouvelles acquisitions franc aises 6771) with an added
triplum which is the same in all but minor details as that in E. These sigla
concord with those in Lawrence Earp, Guillaume de Machaut: A Guide to
Research (New York: Garland, 1995), see especially Ch. 3. The balade is also
cited in an anonymous fifteenth-century German treatise, ed. Martin Staehelin,
`Beschreibungen und Beispiele musikalischer Formen in einem unbeachteten
Traktat des fru hen 15. Jahrhunderts', Archiv fu r Musikwissenschaft, 42 (1989),
pp. 220.
2. For a discussion of these and other poems by Machaut on the theme of Fortune,
see Leonard W. Johnson, Poets as Players: Theme and Variation in Late Medieval
Poetry (Stanford, CA: Stanford University Press, 1990), pp. 4158. For a
discussion of these songs and comments on Johnson's readings, see also Elizabeth
Eva Leach, `Fortune's Demesne: The Interrelation of Text and Music in
Machaut's Il mest avis (B22), De fortune (B23), and Two Related Anonymous
Balades', Early Music History, 19 (2000), forthcoming.
3. See the persuasive arguments of Reinhard Strohm in his `The Ars Nova
Fragments of Gent', Tijdschrift van der Vereniging voor Nederlanse Muziek-
geschiedenis, 34 (1984), pp. 11723.
4. Jehoash Hirshberg, `Hexachordal and Modal Structure in Machaut's Polyphonic
Chansons', in J. W. Hill (ed.), Studies in Honour of Otto E. Albrecht: A Collection
of Essays by His Colleagues and Former Students at the University of Pennsylvania
(Kassel: Ba renreiter, 1980), pp. 357; Peter Lefferts, `Signature Systems and
Tonal Types in the Fourteenth-Century French Chanson', Plainsong and
Medieval Music, 4/ii (1995), p. 133; Yolanda Plumley, The Grammar of 14th
Century Melody: Tonal Organization and Compositional Process in the Chansons of
Guillaume de Machaut and the Ars subtilior (New York: Garland, 1996), pp. 18
19, 174; and Sarah Fuller, `Guillaume de Machaut: De toutes flours', in Mark
Everist (ed.), Models of Musical Analysis: Music before 1600 (Oxford: Basil
Blackwell, 1992), pp. 4165. The triplum, seemingly added at a later stage by
someone other than Machaut, is briefly discussed in Wolfgang Do mling, Die
Mehrstimmigen Balladen, Rondeaux und Virelais von Guillaume de Machaut
(Tutzing: Hans Schneider, 1970), pp. 7980, but is not central to Fuller's analysis.
5. Even the keyboard arrangement in Fa, which may appear to be a two-part
reduction of B31 in major prolation, has been shown to use elements of the
contratenor part in its compound upper part and to preserve the basic breve
harmonic pace, despite the different figuration available in the new mensuration;
see Jane E. Flynn, `The Intabulation of De toutes flours in the Codex Faenza as
Analytical Model' (unpublished paper read at the Medieval and Renaissance
Music Conference, York, 1998).
6. Fuller, `Guillaume de Machaut: De toutes flours', p. 57.
7. Thomas Brothers, Chromatic Beauty in the Late Medieval Chanson: An
Interpretation of Manuscript Accidentals (Cambridge: Cambridge University
Press, 1997), pp. 116, 117.
8. Hirshberg, `Hexachordal and Modal Structure', p. 35.
9. Jehoash Hirshberg, `The Exceptional as an Indicator of the Norm', in U. Gu nther,
L. Finscher and J. Dean (eds.), Modality in the Music of the Fourteenth and
Fifteenth Centuries (American Institute of Musicology: Ha nssler, 1996), p. 54.
Hirshberg asserts that the `b rotundum, which cancels b quadratum, is designated in
bar 15 in all Machaut Repertory manuscripts as well as in P[it], Reina [PR] and
FP', (p. 53). However, having examined PRand seen microfilms of FPand Mod, I
concur with Brothers' representation of the situation, that is, that the sign is not
present in these `secondary' sources.
10. Margaret Bent has written persuasively against making such an anachronistic
assumption. See Margaret Bent, `Diatonic Ficta', Early Music History, 4 (1984),
pp. 148; Èditing Early Music: The Dilemma of Translation', Early Music, 22/
iii (1994), pp. 37392; `Diatonic ficta Revisited: Josquin's Ave Maria in Context',
Music Theory Online (September 1996) (mto.96.2.6.bent.tlk); and `The Grammar
of Early Music: Preconditions for Analysis', in Cristle Collins Judd (ed.), Tonal
Structure in Early Music (New York: Garland, 1998), pp. 1559.
11. Exceptionality in Machaut is often also recognised from elements that can be
retrospectively construed as progressive. See Hirshberg, `The Exceptional as an
Indicator of the Norm', especially p. 64.
12. Musical notation either represents performance actions that produce certain
sounds or represents the sounds themselves through conventional graphic-sonic
analogies. What, in twentieth-century terms, are unsigned accidental inflections
are, in fourteenth-century terms, no less explicit than those which carry a
hexachordal sign both depend on the singer's knowledge of the relationship
between representation and sound. Since fourteenth-century singers could not
have anticipated our more strongly prescriptive notation, they were trained to
realise in performance what the notation in conjunction with their training
implied. Most things that our notation implies without any separate symbolic
representation can be subsumed under the heading of performance practice and
are to do with continuous quantities, for example, tempo (unnotated rubato in
Chopin, for example). It is therefore striking to us that in the fourteenth century
pitch a discrete musical element is dependent upon `performance practice' in
our sense. Unlike the elements that lack separate graphemes in our notation, the
performer of inflections caused by counterpoint is not at liberty to choose the
degree of his response. Thus, the placement of semitones for reasons of
counterpoint is not really a `performance practice' which implies an element of
conscious choice as to the degree to which the practice is followed. Instead it is a
notational convention. This is not to argue that every single interval decision is
absolutely prescribed, since some decisions will not be connected with the correct
approach to perfect consonance. The singer may then decide how to solmise a
note (especially the note b) based on other factors, such as his understanding of
line and phrase structure, which may be open to variance from one singer to
another.
13. For an exposition of the idea of counterpoint as analogous to grammar, see Bent,
`The Grammar of Early Music'.
14. `Sed ipsa frequenter sunt in B-fa B-mi virtualiter licet semper non signetur'.
Leofranc Holford-Steven's translation is taken from Bonnie J. Blackburn,
`Review of Thomas Brothers, Chromatic Beauty in the Late Medieval Chanson:
An Interpretation of Manuscript Accidentals', Journal of the American
Musicological Society, 51 (1999), pp. 63036.
15. Most recently in Brothers, Chromatic Beauty, where he states that there is `poor
support for it [counterpoint inflections] by writings from the period' (p. 22).
Brothers' designation of counterpoint inflections as an ùnnotated convention'
reveals his twentieth-century perspective, since it is only in the light of our later,
more prescriptive approach to accidentals that fourteenth-century ones are
unnotated.
16. Bent, `The Grammar of Early Music', pp. 359.
17. Ibid., p. 36.
18. Over the course of the fourteenth century, the general trend was that the
conceptual length of the imperfect sonority was extended over more than one
tenor note leading to (fairly short-range) prolongation of the expectation of
resolution. This might well be the place at which `harmonic tonality' began,
rather than with the mirage of VI cadences (often rather more an archaic use of a
38 directed progression than anything progressive see for example the frequent
examples in Machaut's Nes com porroit (B33)) which attracted so much attention
from earlier musicologists; see Kevin N. Moll (ed.), Counterpoint and
Compositional Process in the Time of Dufay (New York: Garland, 1997), pp. 364.
19. The bar numbers are taken from Schrade's edition because it is more widely
available than the preferable edition of Ludwig. One of the more unequivocal
results of this study is to emphasise the importance of a consideration of the full
contexts of the original manuscript sources for medieval songs. Readers are thus
advised to consult microfilms and facsimiles if at all feasible, or, at least, to
consider the corrections to the modern editions detailed in Earp, Guillaume de
Machaut: A Guide to Research, Ch. 7.
20. I am using the term cadence to refer specifically to closural articulations.
Although `cadence' is a fourteenth-century term for what (following Sarah
Fuller) I have termed the directed progression, the latter term is used here to
avoid confusion. As Fuller shows, directed progressions are used to join phrases
and to open them as well as to close them; cadences (closing formulae) are not
always marked by directed progressions. See Sarah Fuller, `Tendencies and
Resolutions: The Directed Progression in Ars Nova Music', Journal of Music
Theory, 36/ii (1992), pp. 22957.
21. The balades that have a T-Ca ouvert sonority which resolves back to the opening
of the A section are Jaim miex (B7) and Donnez signeurs (B26) (cantus recta e
0
),
Se je me plaing (B15) (tenor recta b-mi), Dame comment quamez (B16), Je sui
aussi (B20), Honte paour (B25), and Pas de tor (B30) (tenor ficta f-mi) and Je puis
trop bien (B28) (tenor recta b-fa). In addition, Biaute qui (B4) (contratenor ficta
f-mi) and De fortune (B23) (triplum ficta c
0
-mi) have directed progressions
leading back to the beginning in other tenor-discant pairs.
22. Sonorities are listed from the lowest note upwards. Where the parts are
inverted from their normative relative positions (from the bottom up: T-Ct-
Ca-Tr) the sonority is shown in triangular brackets, < >, as here, where the
formula <a-c
0
> means that the tenor has c' and the cantus has the a below.
23. It is interesting to note that the T-Ct directed progression from the ouvert ending
back to opening, which has the potential for being a directed one (g-b resolving to
d5) if the contratenor sings b-fa in the ouvert, is overruled by the voice-crossing
directed progression avoidance strategy of the primary T-Ca pair. In three sources
(all arguably related GVg B) the contratenor b in the ouvert is marked with a mi
sign for clarification. That this is cautionary (because it is not present in all sources)
suggests, firstly, that the T-Ca pair has priority and, secondly, that concern for the
overall sonority (an issue when it is of a breve's duration) in the ouvert would
preclude a b-fa when the cantus does not have an e-fa. These observations depend
upon, but are additional to, those that may be gleaned from theoretical precepts.
24. The modern editions do not give an accurate impression. Ludwig has rationalised
the mi sign to apply to the c
0
at the end of bar 13 even though the sign is, in all its
sources, clearly in the d
0
space, immediately preceding the d
0
at the start of bar 14.
Although signing ficta in anticipation is common, a retrospectively-acting sign
would be a scribal oddity for all three sources. Schrade simply omits the sign.
Both Ludwig and Schrade preserve the tenor b-mi sign, thereby effectively
suppressing the directed progression in bars 1516.
25. Because bar 14 sets a rhyme word, all presentations of this sonority will have the
solmisation vowel ì' at this point. That the consonant `m' is also present in the
first line makes the pun even clearer. Within the structure of B5 the use of two
different approaches to a5 (the eventual ouvert sonority and secondary terminal
goal) is interesting and worthy of further analysis, for which there is no space
here. B5 shares the presence of two different approaches to the same resolution
with Dous amis (B6) and Samours ne fait (B1) in the latter of which the sonority in
question is also a5. In B1 (bars 910) the directed progression has the tenor b
marked mi in all sources. No source marks the cantus d
0
but I would argue that,
by analogy with B5, this instance is nevertheless clear enough. Using a
(questionable?) narrative of linear development, it might be possible to argue
on these grounds that B5 was written before B1. The hexachordal sign also in the
cantus part in its earliest source and a congratulatory textual pun in B5 (whereas
B1 has neither) may indicate the authorial over-caution of first time experiment.
27. The single surviving folio of W (Aberystwyth, National Library of Wales, 5010
C) which contains music but not B31 (f. 74v; most of the triplum of Quant en moy
(M1)) is also in a two-column format.
28. A further indication that this is not coterminous in definition or function with a
modern signature.
29. Brothers, Chromatic Beauty, p. 116.
30. There are several other cases where single Machaut sources state fa signs at the
beginning of lines immediately adjacent to notes that should be sung mi for
contrapuntal reasons: Dame comment quamez (B16), second cantus b, bar 27
directly follows a signature fa sign in C, but should be locally altered for the
caesural cadence; De triste / Quant / Certes (B29), cantus III b, bar 39 directly
follows a signature fa sign in the two-column source, G, a sign replicated in the
same position as a `floating signature' in the single column sources, Vg B and A,
but the b should be sung mi to effect the final cadence of the piece (Vg and Beven
mark the contratenor f with a mi sign); En amer (B41), contratenor b, bar 18,
follows a start-of-line fa sign in A, but should be sung mi to make a phrase-end
T-Ct directed progression during the cantus rest; and Dame de qui (B42), cantus
b, bar 41, directly follows a start-of-line fa sign in A, although it should be sung
mi to effect the caesural cadence. In De toutes flours (B31) itself, E has a start-of-
line fa marking in, and then a cautionary mi sign preceding, the cantus b in bar
29a. The mi sign is local, the fa global. If such a functional dichotomy did not
exist there would be little point stating both signs consecutively.
31. Except PR.
32. See Brothers, Chromatic Beauty, p. 118: `the unsettled quality of the piece is
gradually washed away... For the scribe of PN 6771 [PR], the piece is humdrum'.
33. See Sylvia Huot, From Song to Book: The Poetics of Writing in Old French Lyric
and Lyrical Narrative Poetry (Ithaca: Cornell University Press, 1987), pp. 242
301.
34. Ì proceed from the assumption that although some codices are disorganized
miscellanies, a great many perhaps the majority are carefully organized
literary constructs. The scribe responsible for the production of the book played a
role that combined aspects of editor and performer', Huot, ibid., p. 5.
35. Ibid., p. 5.
36. The arguments for Machaut's authority and involvement with copying are well-
rehearsed. See Lawrence Earp, `Machaut's Role in the Production of his Works',
Journal of the American Musicological Society, 42 (1989), pp. 461503; Huot,
From Song to Book; and Sarah Jane Williams, Àn Author's Role in Fourteenth
Century Book Production: Guillaume de Machaut's ``Livre ou je met toutes mes
choses''', Romania, 90 (1969), pp. 43354.
37. See Earp, `Machaut's Role in the Production of his Works', pp. 4947.
38. See Kevin Brownlee, Poetic Identity in Guillaume de Machaut (Madison:
University of Wisconsin Press, 1984), pp. 323.
39. John Na das, `The Reina Codex Revisited', in Stephen Spector (ed.), Essays in
Paper Analysis (London: Associated University Presses, 1987), p. 107; B31 is in
gathering 7.
40. This has been shown to be the case for Qui es promesses / Ha Fortune / Et non est
(M8) (see Earp, `Machaut's Role in the Production of his Works', pp. 4947), for
the poetic text of the Jugement de Roy de Behaigne and the Remede de Fortune (see
James I. Wimsatt, William W. Kibler, and Rebecca A. Baltzer (Music) (eds.),
Guillaume de Machaut: Le Jugement du Roy de Behaigne and Remede de Fortune
(Athens, GE: University of Georgia Press, 1988)), and for the music of the
Remede (see Elizabeth Eva Leach, `Counterpoint in Guillaume de Machaut's
Musical Ballades' (DPhil diss., University of Oxford, 1997, pp. 4758)).
41. Na das, `The Reina Codex', p. 107.
42. Those signs which cannot be inferred as necessary from the basic counterpoint
over adjacent tenor notes tend to appear as much in the Machaut sources as the
non-Machaut sources.
43. `The only line to be set as an integral, self-contained phrase is the refrain ... The
other ten-syllable lines are all segmented so that the initial four-syllable unit
either stands free ... or is attached to the preceding line'. Fuller, `Guillaume de
Machaut: De toutes flours', p. 43.
44. Ibid., 43.
45. This is especially obvious in the keyboard arrangement of Fa where the two
passages share identical figuration. See Dragan Plamenac (ed.), Keyboard Music
of the Late Middle Ages: Codex Faenza 117, Vol. 57, Corpus Mensurabilis Musicae
(American Institute of Musicology, 1972), No. 5.
46. Again, PR does not sign this.
47. The maintenance of b-mi from bar 14 to bar 15 also provides a context for the T-
Ca fourth at the start of bar 15. Fuller incorporates this into her contrapunctus as
an instance of where a conventional contrapunctus (from which fourths, as
dissonances, would be excluded) is ìmpossible'. She wishes the preferences of the
theorists here to yield to the practice of Machaut who `does, on occasion, prepare
a strong resolution by incorporating dissonance at the contrapunctus level' (Fuller,
`Guillaume de Machaut: De toutes flours', p. 50). However, it is not a dissonant
preparation but, if the d-b sonority is seen as established in bar 14, it is an
ornamental upper neighbour-note figure in the tenor coinciding with an
ornamental lower neighbour note in the cantus, both of which would be
eliminated in a reduction to simple counterpoint because the tenor note is entirely
dissonant in presentation and less than a breve long (if it were a breve long, this
might point to a manuscript error).
48. Overall sonority is important. This is shown in the four-part pieces by the brief
adoption of tenor function by the contratenor with respect to the triplum only,
when the contratenor is below the tenor, and preserving a T-Tr relationship
would create unacceptable overall dissonance. See Elizabeth Eva Leach,
`Machaut's Balades with Four Voices', Plainsong and Medieval Music, 10
(2001), forthcoming.
49. This moment is rhythmicised slightly differently in the contratenor of Mod
which otherwise transmits the text of Machaut's three-part original. The
contratenor has a minim rest at the start of bar 40 followed by three minims
and two semibreves (bound in a ligature). This evinces near-contemporary
discomfort with the decision that the counterpoint forces here and either gives the
contratenor time to hear the cantus b-mi and adjust his own solmisation
accordingly, or simply removes the dissonance from the strongest part of the
breve.
51. Èine fu r den Machaut-Satz auf Brevis-Ebene ungebra uchliche Dissonanz als
solche ist der Zusammenklang doch wohl zu bezeichnen: c-g-a (T.43) in der
Ballade 31 ko nnte ihre Erkla rung finden durch die betreffende Textstelle
``cueillir'' `àbbrechen'' (die Rede ist von der einzigen Rose in des Dichters
Garten); mo glichweise soll auch der betonte Quartsprung abwa rts im Cantus auf
dieses ``Brechen'' deuten.' (`The following dissonance, unusual for Machant's
compositions organised at the breve level [i.e. with no modus organisation],
should be noted: c-g-a in bar 43 [Fuller's breve 42] of balade 31. This dissonance
may be explained through its concomitant text ``cueillir'' to break off (the text
tells of a single rose in the poet's garden); ``breaking'' might possibly also explain
the downwards leap of a fourth in the cantus.') Wolfgang Do mling, Àspekte der
Sprachvertonung in den Balladen Guillaume de Machauts', Die Musikforschung,
25 (1972), p. 302. Undue reliance on word painting to èxplain away' unusual
harmony is all too tempting in dealing with a repertoire for which little about how
the music-text relationship was conceived is known. It is my impression that
there is at least a broad structural link between the poem and the music of the
type that would allow games like the caesural rhyme game played out in B31 or
the congratulatory solmisation syllable of B5 (see n. 25 above). However, despite
Fuller's comment that `the other strophes do not evidence so close a fit between
text and musical arrival points' (Fuller, `Guillaume de Machaut: De toutes flours',
p. 44), this dissonance would serve a text-related function in all stanzas since it
occurs before the revelation of the musico-poetic offsetting caused by the short
fifth line in all three: in the second stanza it sets the narrator's admonition to
Fortune to leave the rose to him (`Lay la moy donc') and in the third it sets the
mention of Fortune's subterfuge (`faus tour'). Even without specific word-
painting (the concept of which as formulated with later music may not be
appropriate for this repertoire) the location of the dissonance in bar 42 is one
which in terms of both musical and poetic structure represents the most tense
moment in the song.

Machaut-De Toutes Flores

Uploaded by

Copyright:

Available Formats

Machaut-De Toutes Flores

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Machaut-De Toutes Flores

Uploaded by

Copyright:

Available Formats

If the surviving sources are at all representative, then De toutes flours (B31) was

one of Machaut's most widely disseminated songs. Not only is it contained in

You might also like