Professional Documents
Culture Documents
Berio GuigueOnofre - SequenzaIV - 1st Proofs
Berio GuigueOnofre - SequenzaIV - 1st Proofs
Although Berio’s Sequenza IV for piano (1965–66) has become, like most of the
other pieces in the series, a landmark of twentieth-century literature for the
instrument, several factors set it apart from the rest. While almost all of the other
pieces were written in close collaboration with eminent players, from Severino
Gazzelloni (Sequenza I) to Rohan de Saram (Sequenza XIV), this is the only
Sequenza to have been conceived and elaborated directly at the piano by Berio
himself, although it is dedicated to Jocy de Carvalho. According to Albèra, the
piece ‘results from the composer’s own instrumental gestures’, giving it ‘this
improvised character, influenced by jazz’.1 On the other hand, David Burge reports
that Berio talked to him ‘with great excitement’ about Stockhausen’s Klavierstück
X (1961) after he had heard it for the first time, saying that ‘this “new way of
playing piano” had inspired him to write a big piece for the instrument’ that would
use clusters ‘in a different way’ and ‘require the employment of middle pedal
“most of the time”’.2 These observations put Sequenza IV into a distinctive position
within the whole cycle, as it is more likely to demonstrate affinities with Berio’s
personal pianistic background and the acknowledged inspiration of Stockhausen,
rather than with the other Sequenzas. Indeed, apart from the almost permanent use
of the sostenuto pedal, this Sequenza contains none of the typically avant-garde
extended instrumental and vocal techniques which supply the more theatrical and
dramatic elements of the Sequenzas for voice and trombone, Sequenza IV’s
adjacent pieces in the series. This is why Gale Shaub, comparing these first
Sequenzas, can assert that Sequenza IV ‘achieves its uniqueness primarily through
purely musical means’,3 that is, through a more conventional idiomatic approach
symbolized perhaps by the use of a time signature, an atypical feature in the
Sequenzas as a whole.
1
Albèra considers this particularity a crucial vector of Sequenza IV’s originality.
Philippe Albèra, ‘Introduction aux neuf sequenzas’, Contrechamps, 1 (1983), p. 102.
2
David Burge, Twentieth-Century Piano Music (New York, 1990), pp. 163–4.
3
Gale Schaub, Transformational Process, Harmonic Fields, and Pitch Hierarchy in
Luciano Berio’s Sequenza I through Sequenza X (PhD dissertation, University of Southern
California, 1989).
210 Didier Guigue and Marcílio Fagner Onofre
However, as he decided to include this ‘big’ new piano piece in the Sequenza
series, Berio applied to it the overall generative structural concept of the cycle,
namely a ‘sequence of harmonic fields’ from which most of the other musical
dimensions are derived, associated with a ‘control of the development of harmony
and melodic density’, as well as the intent to ‘melodically develop an essentially
harmonic discourse’.4
All the materials of the piece are derived from the opening vertical chords,
which are progressively horizontalized, creating a ‘continuum of figurations’ from
dense chords to thin ‘arpeggiating and eventually melodic fragments which interact
in a syntactic flux between structurally opposite and intermediate constituents’.5
The vertical objects heard in the first moments of the piece are, in fact, a
number of examples of the two main sonic categories which constitute the core of
the music’s kinesis. The first category (A) consists of bi-triadic vertical structures
built upon two superposed major, minor, augmented or diminished triads. The
objects in this category tend to feature a resonant, if not consonant sound design.
Some of them may have an added seventh and/or ninth reinforcing this resonant,
harmonic strand. Conversely, the second category of objects (B) is based on
chromatic relationships, with a large number of seconds and fourths, making them
more inharmonic, ‘anti-resonant’, almost ‘noisy’. As Albèra says, ‘this opposition
is maintained throughout the work: the second category generates, for instance, the
clusters, while the first one, taking advantage of its ‘consonant’ intervals, develops
melodic figurations’.6
This essay aims to address the way the composer controls the evolution of the
relative sonic complexity of the Sequenza’s chord objects – one of the ‘purely
musical means’ Schaub alludes to – concentrating on the immanent harmonic
structure rather than discussing any of their linear, gestural or melodic expansions
or transformations. Other authors have observed the fluidity of the time structure,
where the music gradually shifts from one sonic category to the other. Although
the return of the initial harmonic section at the end of the piece (from 1614)7 clearly
suggests an overall ABA form, it is far from easy to pinpoint exactly when each
section begins and ends.8 ‘The image of an unstable but directed “flux” or
evolution is an appropriate description of the formal design of [this piece] since the
4
Luciano Berio, Two Interviews with Rossana Dalmonte and Bálint András Varga,
trans. and ed. David Osmond Smith (New York/London, 1985), pp. 97–9.
5
John McKay, ‘Aspects of Post-serial Structuralism in Berio's Sequenza IV and VI’,
Interface, 17 (1988, p. 223. We do not intend to develop a thorough comparison with
Stockhausen’s Klavierstück X, but it is, however, interesting that it is also built upon a
structural opposition between clusters (or chords) and fast, single-voice chromatic gestures.
See Jonathan Harvey, The Music of Stockhausen (London, 1975), pp. 42–7.
6
Albèra, ‘Introduction aux neuf sequenzas’, p. 103.
7
As the score has no bar numbers, bars are identified throughout this chapter by page:
1614, therefore, refers to the fourteenth bar on page 16.
8
Stockhausen’s Klavierstück X displays a clear ABA overall layout. See Harvey, pp.
42–7.
Sequenza IV for Piano 211
Methodology
9
McKay, p. 226.
10
Our weighting system is aided by a library of computer functions we are developing
under the name of SOAL – Sonic Object Analysis Library, distributed with the IRCAM’s
OpenMusic software (http://www.ircam.fr). See: Didier Guigue, Sonic Object Analysis
Library - OpenMusic Tools for analyzing musical objects structure (Paris, 2005). As SOAL
exclusively reads MIDI files, all sonic objects selected from the Sequenza for this essay had
to be encoded in this language before being analysed by a set of functions, each of them
corresponding to one of the musical components.
212 Didier Guigue and Marcílio Fagner Onofre
In the first instance, the specific sonic structure of an object is determined by the
relative timbral complexity of each of its inner pitches. For the piano, it is possible
to condense this complexity into the general principle of a weighted decline in the
timbral complexity across the instrument’s range, such that the higher the pitch, the
less complex its sonic quality.11 This intrinsic quality is modified by the attack
speed of the piano’s hammer as it strikes the string, an action which results from
the dynamic marking in the score as interpreted by the player. This amplitude
modulation can be broadly defined as linear: the faster the attack, the stronger the
modulation, so the louder the dynamic marking, the higher the weight. The
combination of the two values produces q, a sonic quality for each note’s pitch. To
obtain the Pitch Sonic Quality (Q) value for the chord-object as a whole, it is
necessary to both conflate the individual q values and take into account the pedal
marking, whenever present. In our model, the piano pedal marking acts as a
multiplier and is based on the assumption that the sustaining pedal boosts the
whole sonic complexity, while the una corda suppresses it, in which case the
multiplying factor is less than one.12
This evaluates the object according to the dynamic marking. The complexity
weight is obtained by dividing the chord’s MIDI velocity value (which corresponds
to the dynamic marking) by the maximum velocity value (127) which corresponds
to the theoretically loudest possible sound (usually symbolically marked with fff or
more).
Range (ambitus) refers to the space defined by two boundary pitches. The relative
range vector compares the input object’s pitch range to the piano’s complete range
and obtains the complexity weight by dividing the interval between the object’s
lowest and highest pitches by the whole pitch range. The vector goes from narrow
(0.0) to wide (1.0).
11
The three main acoustic factors of this decline are (1) the decreasing number of
audible partials for a given fundamental, (2) the decreasing rank of the loudest partial(s) and
(3) the shortening of the decay rate of sound.
12
For the sources of these assertions, see previous publications by Guigue: Sonic
Object Analysis Library (Paris, 2005); ‘Sonic Object: A Model for Twentieth-Century
Music Analysis’, Journal of New Music Research, 26/4 (1997): 346–75; and Une Etude
‘pour les Sonorités Opposées’ (Villeneuve d'Ascq, 1997).
Sequenza IV for Piano 213
Register
Seven piano registers have been defined according to some global timbral
differences caused by physical or mechanical variations, such as the number of
strings per note (producing a correlated detuning or ‘chorus’ effect), the strings’
material (wound or unwound), the presence of dampers, the average stretching rate
of the tuning, etc.13 The registers are shown in Table 12.1. Among the various data
the software can return, the most relevant for this analysis were the number of
notes by register (NNR), and the each object’s lowest and highest registers.
This measures how equidistant the pitches are. The interval between the contiguous
pitches is compared to a paradigmatic interval which corresponds to the interval
that would be necessary for pitches to be exactly equidistant (chromatically
approximated). The higher the weight, the less linear the pitch distribution, and,
consequently, the more ‘complex’ the sonic object, where complexity equates with
the object’s resonant harmonic properties.
The relative harmonicity vector evaluates the closeness of the object’s vertical
pitch distribution in relation to a paradigmatic harmonic spectrum-like structure,
deduced from a hypothetical fundamental (which is either the lowest pitch in the
object or a lower transposition) and constructed inside a given spectrum
bandwidth. The fewer the number of pitches common to both paradigm and object,
the less ‘harmonic’ the object is and the higher its complexity weight. Secondly,
the average harmonic rank of these pitches (treating them as if they were the
13
Again, more detailed information is available under Guigue’s publications in
bibliography. One of our main sources is Hideo Suzuki and Isao Nakamura, ‘Acoustics of
Piano’, Applied Acoustics, 30 (1990): 147–205.
214 Didier Guigue and Marcílio Fagner Onofre
partials of the hypothetical fundamental) is calculated: again, the higher the rank,
the further the average ‘partial’ is from the fundamental, and the less ‘harmonic’
the object is. The final H value is the mean of these two complexity weightings.
Both relative linearity and harmonicity are models for evaluating the inner pitch
distributions by scanning the object’s adjacent intervals. Therefore, the spectrum-
like distribution model does not refer to some concrete acoustic structure, as the
model works with notes, not harmonics; nor does it intend to imply tonal functions
as it only evaluates adjacent intervals, although it may help to reveal a
composition’s low-level harmonic structure. Moreover, the resulting values tend to
be negatively correlated: for obvious reasons, the more equidistant the distribution
of pitches in a chord, the less harmonic, as the harmonic model implies a
logarithmic rather than equidistant distribution of intervals.
techniques, while the other hand continues to play the same chord. Various
examples of this can be found in Ex. 12.1.
Networks of similarities between chords based upon absolute density variations
can be observed, as they are ‘part of Berio’s compositional approach as
exemplified in most of the Sequenzas’.14 In fact, common procedures ‘which alter
the original pitch-class structure, such as those based on permutational or
combinatorial operations, are rarely found’.15 Conventional transposition, when a
single interval change is applied to all pitches, only occurs once in the work,
between the A1 and A2 matrices.16
Finally, it should be noted that we have taken into account the misprints of the
1967 Universal Edition score, gathered by Gale Schaub who accessed the
composer’s manuscript, as follows:
1. 110, last chord (last chord of the second system): a G5 must be added in such a
way that the chord becomes an exact repetition of A3, the sixth chord from the
beginning of the piece.
2. 61, (first chord of the page): the low C must be sharpened in such a way that this
becomes an exact repetition of B2, the third chord of the piece.
3. According to Schaub, the dynamic mark ppp on 174 of the manuscript was
overlooked by the engravers and therefore does not appear in the printed score.
We have restored it for our analysis of the pitch sonic quality of the
corresponding objects. 17
14
Schaub, p. 77.
15
Ibid.
16
This is the reason why we chose not to consider A2 as a declension of A1. In
addition, the right and left hand transpositions are in contrary motion.
17
Schaub, pp. 76, 88, 102.
218 Didier Guigue and Marcílio Fagner Onofre
lesser extent, A1) appear (164, 1614–173, 1711–1713), becoming the main element of
the closing structure of the whole work. The other A-chords do not reappear.
These observations imply some form of pseudo-tonal bipolar construction
underlying Sequenza IV’s harmony, where A-chords function as gravity centres,
while B-chords push toward an alternative, dissonant pole. As in the most
idiomatic tonal structure, sonata-form, ‘tonic region’ chords (A-chords) introduce
the work before being quickly superseded by ‘dominant region’ chords (B-chords),
which take charge of most of the harmonic syntax. During the final ‘recapitulation’
section, ‘tonic chords’ (especially the A2 chord) return with far more intensity, to
resolve and dissolve the former tensions. However, there is no unequivocal
evidence for definitively sectioning the Sequenza according to this hypothesis. The
polar oscillation between the A and B harmonic structures retains an essentially
fuzzy behaviour.
SO Q A R NNR18 L H
A1 (11) 0.05 0.19 0.32 (0 0 0 3 3 0 0) 0.3 0.6
A2 (12) 0.04 0.19 0.18 (0 0 0 5 1 0 0) 0.07 0.46
A3 (13) 0.04 0.19 0.2 (0 0 0 6 1 0 0) 0.13 0.4
A4 (15) 0.07 0.19 0.29 (0 0 1 7 0 0 0) 0.07 0.76
A5 (814) 0.06 0.39 0.33 (0 0 0 5 2 0 0) 0.16 0.83
A6 (1210) 0.05 0.09 0.45 (0 1 2 3 2 0 0) 0.1 0.73
Note: SO – Sonic Object label (pagebar); Q – Pitch Sonic Quality; A – Amplitude Rate; R –
Relative Range; NNR – number of notes per register; L – Relative Linearity; H – Relative
Harmonicity.
As shown in the NNR column of Table 12.2, all A-chords, except the last one, are
almost entirely condensed into the piano’s central region (registers -1 to +1, see
Table 12.1 above). This compositional choice creates relatively narrow ranges (R).
Generally associated with very soft dynamics (A), the overall sonic aspect of A-
chords is characterized by relatively poor timbral richness (Q). 19
These initial sonic qualities are somewhat offset by a complex vertical
distribution, due to the bi-triadic model: A-chords, despite the triadic layout that
should theoretically produce relatively simple sonorities, tend to sound quite
18
The list of integers is formatted from the lowest piano register (first value) to the
highest one (seventh value). The integers represent the absolute number of notes in the
corresponding register.
19
0.19 in the amplitude column corresponds to a ppp marking, while 0.39 is the
average amplitude weight applied to the mf-to-pp simultaneous decrescendo of the A5
chord.
Sequenza IV for Piano 219
complex (see the relatively high values in the H column). This inherent
inharmonicity is slightly greater in the last matrices. Apart from downward register
expansion, this is due to the elimination of pure triadic relations in the left-hand
chords (for example, compare A3 with A4).20 On the other hand, the highly
inharmonic quality of A5 and A6 allows them to mix more effectively with the
generally dissonant context of the B-chords in which they are submerged (see
pages eight and twelve of the score). The desire to integrate the two sonic
categories appears also to be the reason for the unusually wide range and the
striking pppp dynamic marking of A6 (see A, R and NNR values), since it
therefore merges into a very large register span and very low dynamic context (see
the score again, page 12).
Note: Matrices are placed on the horizontal axis in the order they appear in Table 12.3
below. Weight values are computed on the vertical axis. The tendency lines, produced by
two third-degree polynomial functions, show that sonic quality complexity tends to increase
progressively towards the last matrices.
However, the NNR column demonstrates that Berio continues to avoid the most
extreme registers. In fact, most of the B chords tend to reproduce the vertical span
20
It must be also noticed that the left-hand ‘perfect’ triadic quality of matrices is
weakened in the declensions.
220 Didier Guigue and Marcílio Fagner Onofre
of the A category, with their range largely restricted to the central region and only
a small number of them expanding into the +2 register (chords B4, B15 and B17).
Ex. 12.2 above shows that declensions, with few exceptions, preserve the matrices’
global range.
Of particular note are some low-register objects, namely the first three matrices,
which contain pitches in the -2 register, and, more impressively, B19, the only
chord which reaches the -3 register, having A1 as its bass note; and also the
remarkably narrow objects B5 and B7. B7 is especially noteworthy for its cluster-
like figure, an extremely high density object concentrating nine pitches into the
span of a tenth. One can also observe the linear quality of the pitch distribution of
B2, B11 and B12, built upon superimposed fourths. This intervallic structuring
counters the triadic system of A-category chords.
There must be a structural and/or acoustic reason for Berio having
systematically avoided the extreme piano registers in his chordal material,
reserving them for melodic gestures (see page eight, for example). We suggest that
the structural and sonic function of the resonant chords activated through the
sostenuto pedal led the composer to choose the most resonant piano region for his
purpose: the poor sustaining power of the highest piano register (+3) would not
allow an effective resonance effect, while the rich lowest tones would blur the
perception of harmonic structures.
Using the data above, the inner sonic qualities of the Sequenza chords can be
mapped against the chronology of the piece to evaluate how they contribute to its
harmonic structure. For this task, we put the 137 chords shown in Exx. 12.1 and
2.2 – the matrices and all their declensions – in the order they appear in the score,
including exact successive repetitions. This sequence appears in the various graphs
below on the X axis.21 It was then correlated on the Y axis against the vectors
described in the methodology section. In the following analysis, we only comment
on the most significant results.
21
It should be noted that most of the chords are in the first and two last pages of the
work: the first 29 chords appear on the first page, while all chords from n. 93 onwards are
found on pages 16 and 17.
Sequenza IV for Piano 221
SO Q A R NNR L H
B1 (11) 0.1 0.19 0.46 (0 2 2 4 1 0 0) 0.21 0.52
B2 (12) 0.11 0.19 0.36 (0 2 2 4 0 0 0) 0.09 0.6
B3 (13) 0.09 0.19 0.36 (0 1 1 5 0 0 0) 0.21 0.59
B4 (14) 0.05 0.19 0.34 (0 0 0 5 3 1 0) 0.29 0.78
B5 (15) 0.06 0.19 0.22 (0 0 0 7 0 0 0) 0.19 0.45
B6 (15) 0.07 0.19 0.44 (0 0 2 3 3 0 0) 0.36 0.66
B7 (16) 0.06 0.19 0.18 (0 0 0 9 0 0 0) 0.25 0.49
B8 (17) 0.12 0.28 0.3 (0 0 2 6 0 0 0) 0.18 0.42
B9 (25) 0.1 0.19 0.33 (0 0 0 4 2 0 0) 0.38 0.53
B10 (28) 0.15 0.73 0.33 (0 0 0 4 2 0 0) 0.28 0.6
B11 (31) 0.04 0.19 0.24 (0 0 0 5 1 0 0) 0.18 0.54
B12 (34) 0.08 0.28 0.29 (0 0 0 5 1 0 0) 0.05 0.44
B13 (42) 0.19 0.54 0.37 (0 0 1 5 1 0 0) 0.19 0.7
B14 (57) 0.14 0.54 0.3 (0 0 0 6 2 0 0) 0.19 0.64
B15 (510) 0.1 0.36 0.4 (0 0 1 6 4 1 0) 0.15 0.75
B16 (61) 0.11 0.36 0.29 (0 0 1 7 0 0 0) 0.28 0.64
B17 (61) 0.1 0.82 0.36 (0 0 0 4 1 3 0) 0.11 0.64
B18 (85) 0.04 0.19 0.21 (0 0 0 7 1 0 0) 0.23 0.55
B19 (813) 0.18 0.28 0.31 (1 0 2 2 0 0 0) 0.25 0.52
B20 (915) 0.07 0.19 0.24 (0 0 1 5 0 0 0) 0.19 0.6
B21 (96) 0.15 0.31 0.22 (0 0 2 4 0 0 0) 0.17 0.41
B22 (98) 0.15 0.71 0.26 (0 0 0 4 2 0 0) 0.27 061
B23 (101) 0.33 0.82 0.3 (0 0 2 4 0 0 0) 0.3 0.45
B24 (121) 0.14 0.39 0.23 (0 0 1 5 0 0 0) 0.47 0.54
B25 (152) 0.21 0.73 0.33 (0 0 0 5 2 0 0) 0.3 0.63
principle of cyclic form, which is clearly shown by the tendency line in Fig. 12.2.
A slow ‘fade-in’ is created by progressively increasing the number of loud objects
in the soft atmosphere, mainly between 112 and 213 (nn. 20–36 on the X axis in the
graph).
After a section where melodic gestures predominate (from 35 to 510), the climax
is abruptly reached at 511 (n. 46). This marks the beginning of the central macro-
section and corresponds to the loudest moment of the Sequenza. It is dominated by
a non-triadic chord sequence, which begins with a repetition of the two first B-
category chords of the piece (Ex. 12.3) but closes with a bi-triadic one (A32 on 67;
n. 52). This loud section then continues with mainly melodic content (611–84).
Thereafter, a quite rapidly decreasing curve is dramatically announced by the
contrasting ppp chord at 85 (n. 57), which leads to the most subdued moment of the
222 Didier Guigue and Marcílio Fagner Onofre
piece, 1210–136, played at pppp throughout (nn. 74–77). This section acts as a
resolving agent to the previous loud section.
Note: The lower graph shows the tendency line obtained by a ninth degree polynomial
function.
Following this, a new section begins in which the dynamics globally increase,
reaching its peak at 1614 (n. 111). This peak may be considered to act as a marker
for the beginning of the ‘recapitulation’. At this point, a sequence of ppp A-
category chords begins (1614–173) countering the loud B-chord sequence that
announced the previous section (511–67). This sequence uses declensions of the two
first A-chords of the piece.
A second peak occurs at 175 (n. 120). Finally, after a short pianissimo section,
the last bars show a concluding dynamic structure, clearly described by the
tendency line (nn. 129–137), the upper extreme of which is represented by three mf
objects. It hardly needs to be pointed out that a closing amplitude which reflects
that of the very beginning, as is found here, is typical of a cyclic structure.
As noted earlier, Berio largely avoids chords which explore the outlying piano
registers, so we would not expect to find strong range and register contrasts in
Sequenza IV. However, analysis of the sequence of chord registers reveals
interesting structural elements. Fig. 12.3 immediately shows a fairly systematic
alternation of relatively narrow and wide chords, and/or high–low registers.
This ‘saw-wave’ construction of harmonic sequences, a pendulum motion
between low and medium-high registered chords, directly affects the sonic surface,
as any listener can experience on hearing the very beginning (graphically described
in Fig. 12.3, nn. 1–9). This system is sustained until 813 (n. 61), disregarding the
number of melodic events that may be inserted between two successive harmonic
objects, as can be observed on pages two to three or page six. The chord sequence
of page six is reproduced in Ex. 12.4.
There are a small number of ‘zero-registered’ objects (referring back to Table
12.1’s definitions), which correspond to some specific non-triadic chords whose
pitches are all compressed into the piano’s central register (F3–F5). Two of them
have been described in the first section: B5 (which appears on 15, 19, 1111 and 175)
and B7 (16, 115, 213 and 175). The third one is a declension of the B8 family, B83,
and twice appears briefly on page 15 (1510–1511) as a complementary sonic
component of the B24 chord, which surrounds it through the sustaining action of
the pedal.
At the start of the piece, these zero-registered chords serve to make the
harmonic sequences more dynamic and varied, as B5 is inserted into wider
structures (15) or, on the other hand, when the brief appearance of B6 (16) is
absorbed into the more restricted sonority of the sustained B7 chord. The narrow
span of B51 (111) allows it to merge better into the first appearance of non-chordal
material. At 213, chord B7, played sffz, has a clear punctuating function, closing a
somewhat agitato passage and preceding a softer moment. Its unusually
compressed harmonic structure stresses this function.
224 Didier Guigue and Marcílio Fagner Onofre
There is no doubt that the gravity centre of this oscillating structure is situated
between 89 (B42, n. 60) and 813 (B19, n. 61). These two chords, shown in Ex.12.5,
present the strongest contrast of register in the piece, due to the fact that the second
one is the only chord in the piece to reach the lowest (-3) piano register. This pair
divides the whole work into two large parts of approximately equal length.22
Note: The top line of the Y axis corresponds to the highest register, and the bottom line to
the lowest one. Registers are labeled from -3 to +3 (see Table 12.1 above).
From this moment on, the pendular system becomes somewhat diluted, and
sequences of equally spanned chords appear with more and more frequency, as
seen between 911–138 (nn. 67–80 in Fig. 12.3) and 146–154 (nn. 86–90, also shown
in Ex. 12.6). This diminution of registral activity occurs in a context where the
most extended melodic gestures of the piece are presented (cf. pages 12–13) – that
is, at the point where harmonic syntax becomes less important from both structural
and perceptual points of view.
22
This moment occurs at around 5 minutes 30 seconds in an approximately 11-minute
performance of the work.
Sequenza IV for Piano 225
Example 12.5 Sequenza IV: maximum contrast of register in the two chords
of page 8
Example 12.7 Sequenza IV: the anticipation of chords in the final section
In this sense, the short section between 138 and 1311 – a repeated sequence of two
alternately registered chords (see Fig. 12.3, nn. 79–84, and also Ex. 12.7) which
interrupts the long melodic flow – seems to push the music back into a more
chordal layout, as it was in the beginning. However, it can also be interpreted as
anticipatory, a false entry of the final section, which in fact starts at 1510 with an
identical registral movement, although using a larger collection of alternating
chords.
From this point, harmonic structures converge into the final sequences of a few
alternating repeated chords during the last bars of the work. Fig. 12.3, from n. 110
(corresponding to 1614) to the end, effectively demonstrates the resolving function
of what ought to be a recapitulation: Berio almost completely eliminates the
pendular movement he used to contrast the register of the harmonic material,
dissolving it into longer sequences in the same register. His registral strategy for
the harmonic material is, therefore, directly correlated to the actualization of the
form.
paradigm, acts as a cadential rest, which seems to anticipate the very first
introduction of melodic gestures. The second sequence begins at 27 and prepares a
second cadential gesture marked by its two final chords, B7 sffz (213), and B8 ppp
(31). This acts more as semi-cadence, introducing the last and most conclusive one,
represented by the last four chords of the section. It begins with B31 (32), a
superposition of mixed intervals; proceeds through the next bars with B11 and
B12, both built upon fourths, a rare feature in Sequenza; and concludes with that
singular F major chord, A61 (313), built upon fifths (Ex. 12.8b). It seems very clear
that the music reaches a temporary conclusion here, a point of articulation, and that
this position is essentially created through the manipulation of the inner
distribution of pitches, on the complex–simple vector of sonority.
Fig 12.5 Excerpt of Fig. 12.2 (order ns. 1 to 41) with tendency line
Note: Thick horizontal lines mark ‘cadential’ sequences of decreasing sonic qualities.
While the central part, which would correspond to a development section (pages 3–
15), reveals a more complex sequence where relatively harmonic structures
alternate with more inharmonic ones (see nn. 41–89 in Fig. 12.4),26 the same
26
In this central section, chords tend to dissolve and disappear into melodic gestures.
228 Didier Guigue and Marcílio Fagner Onofre
1. from 1610 to 176 (nn. 105–121 in the graph), passing 174 (n. 118) by a complex
bi-triadic chord, A45;
2. from this point to 178 (n.126), culminating with a very dense B44 chord;
3. from here to 1711, which begins with two occurrences of B1, the opening
dissonant chord of the Sequenza (nn. 132–133).
After this the sonic tension is dissolved in the two last bars with the gentle
repetition of one of the simplest bi-triadic chords of the piece, A23. Ex. 12.9 shows,
in musical notation, the main steps of these progressions.
Sequenza IV for Piano 229
Note: H values on the Y axis are scaled from 0.4 to 0.8 rather than from 0 to 1.
Example 12.9 Sequenza IV, chord progression between 1610 and 1713
Note: Numbers between staves are the corresponding order number of the chord. Some
repeated chords are omitted. The curve in the graph beneath the example follows the relative
complexity of their inner distribution of pitches (H) weights, as also shown in Fig. 12.6.
230 Didier Guigue and Marcílio Fagner Onofre
27
The average and mode values of our list of lowest and highest registers for matrices,
according to Table 12.1 above, show that there is a slight statistical tendency for the B-
category chords to use lower registers than in the A-category.
Sequenza IV for Piano 231
Fig. 12.7 Minimum, average (mean) and maximum values for Amplitude
(A), Relative Range (R), and Harmonicity (H) for all chords and
declensions
Fig. 12.8 The global formal tri-partition according to the three sonic vectors
Note: Numbers indicate the pagebar where the points of articulation occur (italics for sub-
sections).
The polyphonic layering of these three vectors of sonority is only part of the
complex layout of the piece. The conclusion of Schaub’s analysis states that ‘the
parameters of harmony, melody, texture, rhythm, and dynamics undergo separate
transformational processes’,28 which means that all of them are implicated in the
structural formation at some level, but in independent ways. Similar polyphonic
approaches can be seen, for instance, in the Sequenza V for trombone, with its three
layers – ‘normal’ sound, sound with plunger mute, and mixed trombone and voice.
Here, however, the structure appears to be far more complex, from a ‘surface’ layer
which alternates between mainly chordal and melodic gestures, to an underground
of resonances, sustained by the middle pedal. The surface alternation undoubtedly
offers the most readily perceived articulation. Beneath this, the sequence of
28
Schaub, p. 107.
232 Didier Guigue and Marcílio Fagner Onofre
29
David Osmond-Smith, Berio (Oxford/New York, 1991), p. 39. Some of these
‘chords’ barely deserve this definition, as they have as few as one or two pitches, as on page
six.
30
See for instance the progression between 61 and 129, which starts from a single pitch
and reaches a cluster.