Inside Luciano Berios Serialism [PDF]

Zitiervorschau

DOI: 10.1111/j.1468-2249.2011.00275.x

christoph neidhÖfer Inside Luciano Berio’s Serialism

musa_275

301..348

Like many other composers who later distanced themselves from serialism, Luciano Berio (1925–2003) embraced the technique for a number of years in the 1950s. His ultimate rejection of serialism notwithstanding, Berio credited it as a significant source of inspiration during a period of his life in which, as he later put it, ‘I really made up for all the time I’d lost in the provinces, especially during the war, and in Milan immediately after the war’ (Berio 1985 [1981], p. 63).Without aligning himself too closely with any particular school of serial thought for too long, Berio adopted and developed the techniques he encountered in the music of his contemporaries – especially Luigi Dallapiccola, Henri Pousseur, Karel Goeyvaerts, Karlheinz Stockhausen and Bruno Maderna – before relinquishing serialism altogether by 1958. Despite Berio’s eventual rejection of the technique, the serial experience of those years continued to have a strong impact on his development as a composer into the 1960s and beyond. Ex. 1 lists the serial works from 1951 to 1958, spanning the time from his early twelve-note composition Due pezzi for violin and piano to the works just prior to the flute Sequenza. The first six works listed, up to Nones, employ largely orthodox serial procedures where the pitch rows are generally recognisable on the musical surface, with serial principles eventually extending into parameters other than pitch (in Nones). In the remaining works listed, written between 1955 and 1958, Berio subjects his serial materials to more elaborate processes of transformation which are much more difficult to decipher. Although the principles of Berio’s early serialism from 1951 to 1954 are well known, his later serial techniques from 1955 to 1958 are still little understood.1 There are three reasons for this: first, in his writings and interviews Berio provided only limited information on his serial works;2 second, it is nearly impossible to decipher the composer’s later, complex serial techniques from the published scores alone; and third, only one sketch survives for the works listed from 1955–8, for Allelujah I, making this the only one of these serial compositions whose serial structure can be determined with certainty. This article examines Berio’s compositional techniques in three serial works from his serial period: Nones (1954), Quartetto per archi (1955–6) and Allelujah I (1955–6). The aim of this study is threefold, namely to show which serial materials Berio used, how he employed them, and why he used them in the ways he did. I have chosen these pieces not only because of their chronological proximity and shared compositional aesthetic, but also because they are advanced serial works from the composer’s oeuvre for which a number of helpful, if incomplete, sources exist. These include Berio’s own comments on Music Analysis, 28/ii-iii (2009) © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA

301

302

christoph neidho¨ fer

Ex. 1 Berio’s major serial works of 1951–8 Due pezzi for violin and piano (195I, rev. 1966) Study for string quartet (1952, rev. 1985) Cinque variazioni for piano ( 1952–3, rev. 1966) Chamber Music for female voice, cello, clarinet and harp ( 1953) Variazioni for chamber orchestra (1953–4) Nones for orchestra (1954) Quartetto per archi (1955-6) Allelujah l for six instrumental groups (1955–6) Serenata l for flute and fourteen instruments (1957) Allelujah II for five instrumental groups (1956–8)

Allelujah I in his 1956 article ‘Aspetti di artigianato formale’ (‘Aspects of Formal Craft’), two pages of analytical notes on Nones, a preliminary draft score for Allelujah I, program notes for Nones and Allelujah II and various discussions of serialism in his writings and interviews. No sketches survive for the Quartetto; nevertheless, valuable information on its construction can be found in the article on Berio by Piero Santi, published in Die Reihe 4 in 1958. In view of the sparse extant primary sources, I shall demonstrate that Berio’s serialism from 1955 onwards is best understood from a historical angle which has thus far been little explored: the influence of Bruno Maderna (1920–1973), Berio’s mentor and close collaborator at the Studio di fonologia musicale in Milan at the time. Since first-hand information on why and, albeit to a much lesser extent, how Berio employed serial techniques can be obtained from his own commentaries, his thoughts on serial composition will be reviewed in Part I. Part II examines Nones and shows how the integral serialism of this work (involving pitch, rhythm, dynamics and modes of attack), while subject to specific rules, presented Berio with considerable flexibility in his compositional choices. Part III investigates the serial materials in the Quartetto in comparison with Maderna’s String Quartet, written a year earlier, whose manuscript draft score, complete with its composer’s analytical markings, Berio owned. As will be shown, Maderna paired strict serial techniques with a flexibility of application in composition, a procedure which appealed greatly to Berio. Part IV demonstrates how one can decode the serialism employed in Allelujah I through a close reading of Berio’s discussion of the work (as incomplete, from a technical point of view, as that discussion may be) in conjunction with an analysis of the surviving draft score. While prior commentators, possibly misled by Berio’s own account, have suggested that the work’s principal structure is only partially serial, it will be shown here that the structure is in fact serialised throughout. The later stages of the compositional process in Allelujah I – where Berio recombines and transforms the serial materials with considerable freedom – clearly show the influence of Maderna. I. Berio’s Views on Serialism Berio expressed his thoughts on serialism in largely critical terms. A number of excerpts from his writings and interviews illustrating what he saw as the benefits © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

303

and pitfalls of serial composition can help us better understand why and how the composer adopted serialism in the 1950s, and why serial thinking strongly influenced his entire compositional career. A close reading of Berio’s comments, in conjunction with an examination of his compositional strategies in the three selected works, will confirm that his famous attack on the ‘Twelve-Tone Horse’, from 1968, did not represent a change in his attitude towards serialism. Rather, with this attack he pointed precisely at the kinds of problems which he himself had recognised and already overcome a decade earlier. Berio saw serialism as, at its best, a powerful tool for discovering new musical territories; at its worst, however, it was too vulnerable to formalistic attitudes devoid of musical substance.The latter point lies at the core of his 1968 polemic: I would go as far as to say (as my anger comes back) that any attempt to codify musical reality into a kind of imitation grammar (I refer mainly to the efforts associated with the Twelve-Tone System) is a brand of fetishism which shares with Fascism and racism the tendency to reduce live processes to immobile, labeled objects, the tendency to deal with formalities rather than substance ... . This is why I am very much against the formalistic and escapist attitude of twelve-tone composition. In losing himself in the manipulation of a dozen notes, a composer runs the risk of forgetting that these notes are simply symbols of reality; he may, in addition, end up ignoring what sound really is. (Berio 1968, p. 169)

For Berio, at the heart of the problem lies a common misunderstanding of the relationship between analysis-turned-theory and composition: A composer can give a descriptive analysis of his own work and can bring to bear the analytical tools from past musical experience. A structural description of a piece of music cannot, however, account for the meaning of that piece unless it is placed in a historical continuity. By the same token a ‘theory’ derived from analysis can never legitimately be used as a tool for producing music. Attempts to do this betray an idea of musical language based solely on procedures for combining elements, which is, to say the least, irrelevant to any serious discussion of music. (Berio 1968, pp. 169–70)

And Berio concludes: A theory cannot substitute for meaning and idea; a discrete analytical tool can never be turned to creation by dint of polishing and perfecting it. It is poetics which guide discovery and not procedural attitudes; it is idea and not style ... . This basic fact has been missed by those who insist on trying to create a twelve-tone utopia of ‘twelve-tone coherence’ by forcing on us the dubious gift of twelve-tone melodies in which, as someone has written, ‘the twelve-tone rhythmic structuralization is totally identical (sic) with the structuralization of the twelve tones’.3 Alas, this industrialized twelve-tone horse, dull on the outside and empty inside, constantly being perfected and dragged to a new Troy in shadow of an ideological war long since fought and won by responsible minds like Schoenberg, with neither systems nor scholarship for armor! (Berio 1968, p. 171)

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

304

christoph neidho¨ fer

Although Berio briefly embraced integral serialism himself, he never believed in ‘excessive interchangeability between acoustic parameters’ (Berio 1985 [1981], p. 65). Separating out musical parameters made sense to him only insofar as he could be certain that these came out of, and would ultimately be reintegrated back into, a meaningful whole. Looking back in 1981, he explained: As everybody knows, one of the most important and symptomatic aspects of the serialist experience was the separation of musical ‘parameters’ ... . When this dividing up of ‘parameters’ was applied scholastically, for analytical purposes, to musical pieces where the solidarity between intervals, durations, instrumental timbre, intensity and register was organically implicit in the expressive and structural design of the piece, then the operation had, and still has, a meaning. It was rather like examining the separate pieces of a motor while knowing that the elementary sum of these parts didn’t constitute the motor (our perception always plays such tricks on us).The problems started when, inevitably, people began going in the opposite direction, taking unattached pieces, separate ‘parameters’, and putting them together under the indifferent and uniform light of abstract proportions, and the waiting for the unveiling of the piece (or the non-piece – which is after all the same thing because, as you know, by night all cats are grey).4 (Berio 1985 [1981], pp. 68–9)

In order for a structure to be meaningful, Berio believed, it must thus be conceived as an entity, as a concrete musical object which makes sense, rather than as an assembly of disparate components (as integrated as the compilation of the parameters may be from a serial point of view).5 At the centre of Berio’s serial practices lies the design of such concrete musical objects which in turn are subjected to various processes of transformation, serial or otherwise. In Nones, the basic materials are pitch series with characteristic rhythmic and dynamic profiles which are transformed according to a rule regulating the possible choices of durations, dynamics and articulation. The Quartetto consists of different readings of a basic sound material which freely omit (and possibly add) pitches, change rhythms and registers, and vary timbre and instrumentation.6 And in Allelujah I Berio established an initial material, fully worked out in terms of pitch, rhythm and (provisional) register, whose internal serial pitch layers are then reread by projecting them onto different rhythmic planes and then superimposing and re-orchestrating the resulting structures.The three works to be discussed here are serial in the sense that their initial sound materials (whose pitch and rhythmic dimensions, at least, are fully worked out) are built from one or several pitch series.The transformations of these materials may be guided by serial rules (as in the choices of parameter values in Nones or the rhythmic projections in Allelujah I), or may be free. Whatever the principles of transformation, however, Berio’s aim was to create coherence and musical sense that transcended the serial machinery. He had a clear vision of, and maintained full control over, how the music would ultimately sound. In Piero Santi’s words: Never during the entire creative process [in Nones] does Berio forget what is to be its end-product. Here is the basis of his artistic freedom and his excellence as a

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

305

craftsman. These are still more clearly manifest in the Quartet, since the connections tying them to the basic scheme, though less directly visible than in Nones, are clear within the musical coherence of the whole work, as that unity of all details that I have already mentioned. In his most recent work [the Quartet] Berio again shows, more clearly than before, that he relies not on the formal guarantee provided by an abstract, cerebral scheme, but on his own creative energy. Berio’s fantasy does indeed always create a plan, but this is in order to play within its limits, to vary it without invalidating it, to enrich it without obscuring it beneath a mass of dovetailings and superstructures. His fantasy loves clear form, of the kind demanded by the artistic tradition to which Berio himself belongs. (Santi 1960 [1958], pp. 101–2)

II. Nones Berio composed Nones in 1953–4, after he first attended the Darmstadt Summer Courses.7 Whereas the previous works leading up to the Variazioni were modeled on the serial counterpoint of the Second Viennese School and Luigi Dallapiccola, with whom the composer had studied at Tanglewood in 1952, Nones was Berio’s first (and possibly only) integrally serial work in which serial transformation is applied to four distinct parameters. The choice of the four parameters – pitch or pitch class, duration, dynamics and mode of attack – was influenced by developments which took place at Darmstadt in the four years prior to Berio’s arrival: Olivier Messiaen had defined these parameters, although without treating them serially, in his piano étude Mode de valeurs et d’intensités, written at the Darmstadt courses in 1949.8 Soon thereafter, Karel Goeyvaerts (in his Sonata for Two Pianos, 1950–1), Karlheinz Stockhausen (in Kreuzspiel, 1951) and Pierre Boulez (in Structure Ia, 1951–2), among others, began to subject each of these four parameters to serial permutation.9 In the years which followed, the number of parameters was expanded to include more dimensions, such as density (the number of attacks per time unit and the number of pitches per set, among others), tempo, register, and so on.10 Beginning in 1952, Stockhausen adopted what he would later come to call group composition (Gruppenkomposition), a technique dedicated to producing an agglomeration of sound.11 Berio rapidly absorbed what he encountered in Darmstadt and soon went beyond it. Characterising Nones as his ‘first reaction to Darmstadt’, he subjected the by then classical four parameters of pitch class, duration, dynamics and mode of attack to a permutational procedure that was based on a clearly defined rule, yet which at the same time gave him a welcome degree of choice.12 Ex. 2 presents a translation of an analytical note for Nones in which Berio explains how numerical values were assigned to the four parameters and then combined.13 The choice of parameters followed the rule stated at the bottom of the note: for any event, the numerical values of the four parameters must always add up to nine or more; if the sum exceeds nine, the event will have to be followed by a quaver rest. The series, shown at the top of the example, is Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

306

Ex. 2 Translation of Berio’s analytical note for Nones (Berio 1985, plate 4, second page) pitches

+ 1

2

3

4

5

6

7

8

9

10

11

12

13

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[6]

[5]

[4]

[3]

[2]

[1]

3 durations

+

[a] choice

3

[b]

dynamics

+

all values beyond 9 become a quaver rest

[a]

[b]

mode of attack +

[a] free legato [= no ind.] [b] tenuto stacc.

[a] trill [b] frull[ato] [c] tremolo

The pitches will be realised always keeping in mind that the sum of the individual elements reaches and also surpasses 9 – every unit exceeding 9 is worth a quaver rest.

RI-symmetrical and contains thirteen elements, duplicating pitch class D in the second- and penultimate-order positions. The members of the series are numbered from 1 to 13, but in the compositional process Berio used the numbering added below in square brackets.14 The durations are assigned values 1 to 4, with a choice of two durations for each of values 2 to 4. (The second choices consist of durations shorter than a quaver, the first choices of durations longer than a quaver.) The dynamics are listed with values 1 to 5, with two choices each for values 1 and 2. (The second choices present the softest dynamics.) The modes of attack are assigned values of 1, 2 or 3, with multiple choices for each of them.15 Berio’s rule stated at the bottom is modelled after the ‘synthetic number’ pioneered by Goeyvaerts in his Sonata for Two Pianos. In the central two movements of this work, pitch classes, durations, dynamics and modes of attack are assigned numerical values ranging from 0 to 4; every pitch in the score is assigned a duration, dynamic level and mode of attack such that the numerical values sum to exactly 7.16 Berio’s ‘synthetic number’ is 9, a reference to the title of the poem by W. H. Auden which inspired Nones.17 As in Goeyvaerts’s work, Berio’s preparatory materials and governing rule define a type of integral serialism which permits the composer a good deal of freedom. Not only are there multiple ways of balancing the numerical values among the four parameters, but there are frequently multiple choices for a particular value.18 This allows Berio to influence the outcome of his serial © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

307

processes more directly and to a degree unavailable in more rigid serial structures. For example, the composer may decide to use mainly the short note values (between a semiquaver and a quaver) or mainly the longer note values (between a quaver and a crotchet) within the full range of numbers 1 to 4. Or, he may choose to use only soft dynamics, balancing the corresponding numerical values in the other parameters accordingly. Ex. 3 reproduces the opening twelve bars of the work, where four serial layers (P5, P7, P10 and P11) are superimposed.The P11 layer is extracted in Ex. 4a (the harp and alto saxophone of bars 1–10). Ex. 4b summarises Berio’s choices for each of the four parameters. Ex. 4c converts the entries in Ex. 4b to the corresponding numerical values and shows the sum for each of the thirteen events.19 The sums are either 9 or 10; in addition, they form a palindrome,20 a property not imposed by any a priori stipulation. Rather, Berio chooses to stretch the succession of thirteen pitches by inserting a quaver rest following every odd-numbered event except the first and last, as shown by the vertical arrows in Ex. 4a. The option of inserting a quaver rest is provided by Berio’s rule positioned at the bottom of Ex. 2, which requires a sum of 10 or higher for the addition of such a rest. The choice of the actual numerical value (above 9) and location within the series is free, however. As is evident from Ex. 4c, the fact that all sums are 9 or 10 requires Berio to counterbalance the gradual numerical increase and decrease on the first line (given by the pitch-class series) elsewhere in the chart. He chooses to do this by gradually decreasing and increasing the numerical values for the durations and dynamics on the second and third line (although exceptions occur). The entries on the fourth line are mainly set to the smallest value, 1, with few if any attack indications given.21 The flexibility built into Berio’s rule allows him to generate textures with widely differing characteristics and to create a kind of musical coherence which lies beyond the abstract serial structure.At the beginning of the work (see again Ex. 3), all four serial layers start out with predominantly louder dynamics; they then turn to primarily softer dynamics in bars 4–7 before achieving a mixture of soft and loud in bars 7–12. This clear overall dynamic development, in which all serial layers participate, contributes – in tandem with other factors – to the passage’s sense of direction and cohesion.As is typical for much of Berio’s serial music, these bars combine pointillist attacks with melodic gestures, such as the expressive leap in the clarinet in bars 2–3 and similar leaps in the other instruments, including violin (bar 5) and contrabassoon (bar 6). The succession of these latter gestures generates direction; not only does the clarinet crescendo on D in bar 2 lead us to anticipate a consequent event – an expectation which is fulfilled by the high A in bar 3 – but the entire clarinet gesture in bars 2–3 is then echoed and carried on by the ensuing expressive leaps in the other instrumental parts.22 In other words, these gestures are not isolated events, but rather form a larger network of corresponding elements. And this is why the gestures are meaningful; they have a function beyond their individual appearance. (The compositional aesthetic here, as throughout Nones, owes much to Webern in this respect.) Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

308

Ex. 3 Nones, bars 1–12, with serial analysis = 72–76 Fl. 1

Ob. 1

Cl. 1 3

Bsn

Cbsn

3

Hn 2 3

sord .

non troppo

Tbn. 1

Timp.

Tamb. milit.

2 Tam-Tam 3

Cel.

Vibr. 3

El. guitar 3 3

Harp

3

3

sord . Vn A sord . Vn B sord . Vn C

Vla 3

3

sord . pizz.

arco

Vlc. molto dim . 3

pizz.

3

arco

Cb. P t10 : P5 : P7 : P 10 : P 11 :

F/A /E

D

B B

G B B

D

D B

F A G

E

C

A B

B

D

A

F

D F

E

D

E E A

A D D

G A G C [B?]

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

309

Ex. 3 Continued 7 Fl.

Ob.

Cl.

Sax.

[

]

I. Bsn

II.

via sord. I.

3

Tbn. 3

gliss.

Timp.

Tamb. 3

Cel. 3

El. guitar

3

Harp

Vns

Vlas

Vlc.

pizz. Cb.

(P 10 :) (P 5 :) (P 7 :) (P 10 :) (P 11 :)

F

A

E [E ] F [B ?] D B [B ?] F

D CG

G

C

G

F (etc.)

D C

B D

F

A

D

E

F (P 3 :) E

Music Analysis, 28/ii-iii (2009)

B

B

F

D

B (etc.)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

310

Ex. 4 Analysis of Nones, P11 layer, bars 1–10 (a) P11 layer (b) Parameters of P11 layer (c) Analysis of numeric values of the four parameters in the P11 layer (a) 1

2

3

4

5

6

7

3

1

Harp

3

6

3

5

4

3

3

6

2

3

1

Sax.

3

(b) pcs:

B

D

B

G

3

E 3

E

A

D

C

3

A

3

F 3

D

F

3

durations : dynamics: modes of attack: no attack indication

legato staccato

(c) pcs:

1

2

3

4

5

6

7

6

5

4

3

2

1

durations :

4b

3a

3b

2a

3b

1

1

1

3b

2a

3b

4a

4b

dynamics:

3

3

3

2a

1a

1a

1b

1a

1b

2a

3

1

2a

modes of attack: 1a

1a

1b

1a

1a

1a

1a

1a

1a

1a

1a

2a

2b

sums:

9

10

9

10

9

10

9

10

9

10

9

9

9

The rule guiding the combination of parameters illustrated in Ex. 2 requires the composer to make decisions with a clear sense of what the result is intended to sound like, since the choice of one parameter affects the choices available with respect to all of the others. In addition, the flexibility built into the rule (sums can be greater than 9 leading to added rests, multiple choices for some parameters) provides further options which again need to be considered with a clear vision of the expected sonic outcome. In the excerpt shown in Ex. 5, Berio chose a texture (via the same serial rule) which pits the solo violin’s mostly rapid and delicate gestures against a background of longer sustained dyads and single notes as well as non-pitched percussion. All dynamics are soft. Ex. 6a analyses the parameters © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

311

Ex. 5 Nones, bars 40–48, with serial analysis = 126

40

(suono d’eco )

3

Cl.

Timp.

G. C. 3

Cymbals

Tamb. T. T.

3

Vibr.

3

Guitar

sord . Vn solo 1 Solo

tutti

3

pizz.

arco

div.

uniti

Vlc.

pizz. div.

div.

uniti

uniti

Cb.

pizz.

pizz. arco I0:

E

C A C

G

A

B

E

I 5 (1–4):

F

D

G

A

P11(1–4):

B

D

B

G

B

D F

A F

I 5 (1–5):

F

D

F

A

C

p 5 (1–5):

F

A

E

C

B

assigned to the main series I0 in the solo violin. (All other serial statements are fragments.) The low values for the dynamics (1 or 2 for ppp to p) and modes of attack (1 and 2 for tenuto, legato, staccato and no attack mark) require Berio to counterbalance the gradually increasing and decreasing values for the pitch classes with overall decreasing and increasing numbers for the durations, in order to keep the sums within a narrow band (between 8 and 11).23 Where he has a choice of two note values, Berio always picks the same alternative (in every case a semiquaver for 4, a dotted quaver for 3 and a dotted semiquaver for 2), and generally prefers the shorter duration.24 The note values form a palindrome which is ultimately distorted by the rests inserted in the final version. Ex. 6b and Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

312

Ex. 6 Nones, parameters for bars 40–48 (a) Parameters for I0 (solo violin) (b) Parameters for I5/P11 (fragments) in bars 40–42 (c) Parameters for I5/P5 (fragments) in bars 43–48 (a) pcs:

1

2

3

4

5

6

7

6

5

4

3

2

1

durations :

4b

4b

4b

2b

2b

3a

1

3a

2b

2b

4b

4b

4b

dynamics:

2b

2b

2b

1b

1b

1a

1b

1b

1b

1b

1b

1b

2b

modes of attack: 2a

2a

2b

1a

1a

1b

1a

1b

1a

2a

2b

2b

2b

sums :

9

10

11

8(!)

9

11

10

11

9

9

10

9

9

pcs:

1

2

3

4

durations :

4a

3b+

1

3a

dynamics:

2b

2b

(b)

2b

1b

modes of attack: 2*

2**

2b*** 1a

sums :

9

7(!)

9

10

(* legato/staccato in percussion, ** ‘legato’ in guitar, *** pizz. = stacc.)

(c) pcs:

1

durations : dynamics:

4

5

1

3a

4b

1b

2b

1b

2a

2a

2a

2b

9

7(!)

11

12

2

3

4a

3b+

2b

2b

modes of attack: 2a sums:

9

c analyse the parameter values assigned to the remaining serial fragments in Ex. 5. Here, Berio tends to realise durations using the larger of the available note values (for example, value 4 in Ex. 6b is realised as a crotchet rather than a semiquaver in the double basses of Ex. 5) in order to create the sustained sonorities which contrast with the faster violin gestures.25 In 1981 Berio described his experience in Nones: My first reaction to Darmstadt and to Bruno’s beneficial influence, in other words my first exorcism[,] was Nones for orchestra which has nothing of Darmstadt or Maderna in it, but which develops what was for me the main focus of research and musical excitement during those years: the possibility of thinking musically in terms of process and not of form [that is, form types] or procedure.26 (Berio (1985 [1981]), p. 62)

By combining twelve-note serialism in the pitch domain with a kind of multiplechoice integral serialism involving the other parameters (see again Ex. 2), Berio provided himself with a framework which pushed his imagination towards discovering new musical avenues that would otherwise have remained unexplored. © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

313

And it is in this sense that integral serialism led him to what the composer termed ‘an objective enlargement of musical means, the chance to control a larger musical terrain’ (Berio 1985 [1981], p. 65). III. Quartetto per archi Transformation of the parameters assigned to pitch material – as in the Nones series – remained a central feature of Berio’s serial music. But given the absence of sketches for most of the works from the 1950s, determining the transformation processes and the structures to which they were applied is no easy task. For the serial Quartetto per archi, written in 1955–6, no sketches survive which would document the compositional procedures, nor has the manuscript fair copy been preserved. The only source of analytical information – which probably goes back to the composer himself – can be found in Piero Santi’s article of 1958. Santi explains, without providing score examples: In the String Quartet there is less inner dependence [than in Nones] between material and the scheme of construction, on one side, and, on the other, the way they are carried through in music.The Quartet is built up wholly on permutations of pitch-series, which recur in each sequence, and on sequence-permutations which recur in each structure, because of the use of six different durations and a particular intensity for each sequence. [In footnote:] Each structure consists of six series of six sequences each. All the durations in these six series of six sequences, i.e., 36 durations, are multiples of one of six basic values: semiquaver, demisemiquaver, triplet semiquaver, quintuplet semiquaver, triplet demisemiquaver, and quintuplet demisemiquaver.Thus for example in the first structure the durations in each of the six series of sequences are multiples of 1, 3, 5, 7, 9 or 11.This means that each duration in the first sequence-series is one of the six fundamental values, while in the second sequence-series each duration corresponds to one of the fundamental values multiplied by three; in the third series the fundamental value is multiplied by five, in the fourth by seven, etc. Sequences, sequence-series and structures follow each other exactly according to the scheme, in order then to achieve a synthesis in the free articulation of the quartet-texture. [Continued in main text:] Thus it is a matter of six different ‘readings’ of the same material.27 (Santi 1960 [1958]), p. 100)

Ex. 7a–c reproduce three excerpts from the one-movement work, each of which likely corresponds to what Santi calls a sequence. Each passage is built from the same pitch-class materials, the two chromatic hexachords A and B. In Ex. 7a–c, the solid circles mark the members of hexachord A (A–B–B–C–C–D), and the dotted circles contain its complement, hexachord B.The segmentation into these complementary hexachords is suggested by the rhythmic values used in Ex. 7a, to be discussed shortly. With one exception, each statement of hexachord A in Ex. 7a–c presents the six pitch classes in a different ordering.28 Likewise, hexachord B is reordered each time it recurs. Some statements are fragmented, such as in bars 150 (Ex. 7b) and 224–228 (Ex. 7c). The three excerpts present different readings of the same hexachords. Santi describes Berio’s rereading practice as follows: Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

314

Ex. 7a Berio, Quartetto, opening

= 96–100

arco

pizz.

VM e

sempre

arco

arco 3

pizz.

VM e

3

3

pizz.

sempre

3

3

3 3

pizz. VM e

sempre 3

3

pizz.

5

3

5

arco

pizz. pizz.

VM e

5

arco

sempre

Ex. 7b Berio, Quartetto, bars 145–150 = 112 145

3

3

pizz.

legno b.

3

arco

arco

legno b.

5

pizz.

3

pizz.

ord.

arco tast.

3

via sord.

Ex. 7c Berio, Quartetto, bars 224–228 = 72 224

5

5 5

3

3

pizz.

arco

3

pizz.

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

arco

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

315

But Berio makes the rigid skeleton of the structures produce stimuli and ideas, and also a certain coherence within his material. Here he moves with unrestricted freedom; he may leave out notes and durations or add some, he divides up durations into periodically beaten rhythms, chooses registers with complete freedom, and in all this he adheres by and large to the prescribed dynamics, within the limits of his own taste, exploiting effects of timbre and instrumentation very delicately. It would be interesting to follow from bar to bar the onward course and the melting-down of the elements, while keeping the basic scheme before one. It is typical of Berio that he lingers a short time over each of the individual elements, till these take on a figurative shape within the resulting overall picture – they do this less as pointillistic formations than as a collective agglomerate. (Santi 1960 [1958], p. 100)

The three passages in Ex. 7a–c give us a good idea of how this works. In Ex. 7a Berio creates coherence by means of two timbral strata. Percussive, irregular pizzicato attacks are pitted against sharp arco gestures of single or double attacks, most of them played downbow. The two timbres chase each other, creating forward momentum. Only the central register of the quartet is used here, making the four instruments sound alike (all four parts here could in fact be played by violins) and leaving the high and low ranges for later exploration. The distribution of timbres (pizzicato versus arco) cuts across the hexachordal structure. This also holds for Ex. 7b, where a third type of attack is added, col legno battuto. Unlike the beginning of the work, the texture here is widely spaced and the mood calm; the passage ends with a stark dynamic contrast in the last bar. In Ex. 7c different types of attack again frequently cut across the two hexachords. This passage too is quiet in character, this time contrasting short arco and pizzicato gestures with longer sustained notes, the last two played as ethereal harmonics. The semiquaver leaps which succeed each other in the first, second and fourth bars (first violin, viola and cello), together with the sustained pitches, provide gestural coherence. But what is the ‘rigid skeleton of the structures’ or ‘basic scheme’, mentioned by Santi, which is being reread and transformed? Santi’s description suggests that pitch (or pitch-class) structure, durations and possibly dynamics are part of this scheme, while other dimensions such as register, timbre and instrumentation are not prescribed by a particular plan. Since we have no documentation of the ‘basic scheme’, it is perhaps appropriate to turn to a historical source which does provide a plausible context for Berio’s serial techniques, namely Bruno Maderna’s Quartetto per archi in due tempi from 1955. Maderna dedicated his Quartet to Berio; Berio returned the favour the following year, dedicating his own Quartetto to Maderna, at a time when the two composers were in very close contact.29 Maderna’s Quartet is in two movements, with the second presenting an altered reading of the retrograde of the first, freely filtering out pitches and changing rhythms, dynamics, register and instrumentation. In 1981 Berio discussed the relationship between the first and second movements of Maderna’s Quartet: Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

316

Ex. 8a Bruno Maderna, Quartetto per archi in due tempi, end of first movement 3

8

3

184

C

(V)

NV

3

3

T

V T

3

( pizz.)

3

3

3

3

3

3

T

C

(V)

(NV)

5

188

5

T

C NV

T

V

NV

(T) arco

T pizz.

NV

V 5

5

T 5

[Maderna’s] Quartetto is in two parts. The first, in all its aspects, is the product of a strict combinatorial procedure; the second part is a retrograde reading of the first. But on the quantitative level it’s an impoverishing reading, one that filters, eliminates, introduces spaces, and thus reorganizes the time-span and the material that have just been heard on a different level, a level of the highest expressive quality. (Berio 1985 [1981], p. 68)

Ex. 8a reproduces the end of the first movement and Ex. 8b the beginning of the second movement of Maderna’s Quartet. Ex. 8b is a varied retrograde of Ex. 8a, projecting a markedly different, more aggressive character. Longer note values are often subdivided or realised as loud tremoli in Ex. 8b (as in the first and second violins of bars 1–2). I have indicated the omitted pitch classes in square brackets in Ex. 8b. In all probability, Berio must have studied not only the final version of Maderna’s Quartet, but also the latter’s short-score draft with its analytical annotations.This manuscript, a brief excerpt from which is transcribed in Ex. 9, was in Berio’s possession.30 The dotted line in the second bar indicates © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

317

Ex. 8b Maderna, Quartetto, beginning of second movement (omitted pcs shown in square brackets) = 112 circa C

metà arco

al tall.

NV C

al tall.

NV al tall.

C

NV

NV C

pizz.

5

5

5

[G, B ]

4

3

[D, F]

3

[C , E, G , E , A , B ]

P

T 3 3

3

al tall.

legno batt.

3

3

[G, A, C ]

pizz.

[D, F]

3

3

legno

[G]

[E ]

[C ]

the juncture between the two movements, from the point at which the texture runs backwards. As the sketch reveals, Maderna superimposes two different rhythmic strata. The one on the upper stave moves from triplet semiquavers to quintuplet semiquavers and crotchets, and vice versa for the lower stave.31 Berio’s Quartetto makes similar use of rhythmic layering. Ex. 10 segments the first six bars into the three distinct rhythmic layers, each of which is shown on a separate stave (demisemiquavers on stave 1, triplet demisemiquavers on stave 2 and quintuplet demisemiquavers on stave 3). One can easily recognise how each rhythmic layer articulates its own pitch-class material.The first layer presents two statements of hexachord A, as bracketed in the example, omitting D on the second occasion. The third layer contains the same hexachord, in permuted order and with C omitted. By contrast, the second layer uses mainly members of the complementary hexachord B, with the addition of one D (at *) and a single Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

318

Ex. 9 Maderna, Quartetto, excerpt from the short-score draft (Paul Sacher Foundation, Luciano Berio Collection) e

b e 3

3

5 5

d

5

Db 5

d

e

Dd

Dd

e

5

5

f 3

3

3

g

f

3

3

3

De

Ex. 10 Pitch-class material of the three distinct rhythmic layers at the opening of Berio’s Quartetto bars: 1

2

3

4

5

6

a [D] hexachord A

(1) 3

3

3

3

3

3

3

3

3

3

3

3

3 hexachord B (+D, C )

(2) *

** 5

5

5

5 hexachord A

(3) [C ]

C (at **). As will become clear, the latter two pitch classes are migrants from the first and third layers (D is missing once from the first layer and C once from the third).32 A partial rereading of the same pitch-class material occurs in the excerpt shown in Ex. 11a (from the third section of the work).The analysis of Ex. 11b illustrates how the pitch-class succession of the first layer is slightly rearranged (compared to Ex. 10), with A omitted on the second occasion and an additional fragment C–B added at the end. This layer is realised in Ex. 11a mainly with durations of a crotchet or five semiquavers, often subdivided into repeated notes or tremoli, or shortened by rests (in bars 127–128), similar to the example established in Maderna’s Quartet.The second layer in Ex. 11b remains incomplete. The segmentation into the distinct pitch-class layers shown in Exs. 10 and 11b is © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

319

Ex. 11a Berio, Quartetto, bars 120–128 = 96 120

pizz.

3

3

3

3

sord .

arco

pont .

pizz.

3

124

3

via sord.

arco

legno s.

3

pizz.

arco

Ex. 11b Pitch-class material of the two layers in bars 120–128 a

a

1

hexachord A [A]

2

from hexachord B

suggested by Berio’s rhythmic structure, which in turn is most likely modelled on the rhythmic layering found in Maderna’s Quartet. But Berio does not always realise the different pitch-class layers as rhythmically distinct units. Ex. 12a reproduces the full score of the beginning of the third section (bars 92–99). Ex. 12b presents a ‘distributional analysis’ of the pitch-class material used in this excerpt and illustrates how Berio again combines the two chromatic hexachords A and B.33 (Members of hexachord A are stemmed upwards, those of hexachord B downwards.) The pitch-class succession of the entire excerpt is shown in three large segments (bars 92–94, 95–97 and 97–99), aligned in the example to illustrate how each segment starts with the same pitch-class orderings. Bars 92–93 correspond to bars 95–96 and bar 97; other occasional correspondences occur later as Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

320

Ex. 12a Berio, Quartetto, bars 92–99 92

= 72 circa (sord.)

5

arco

pizz.

3

arco

legno b.

pizz.

5 5

arco

arco

legno b.

sord .

5 5

95

3

3

5

sord .

arco

3

sord .

arco 3

5

5

pizz.

98

via sord.

pizz.

arco

3

via sord.

legno b.

via sord.

legno b.

arco

3

3

arco

3

legno b.

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

3

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

321

Ex. 12b Analysis of pitch-class materials bars: 92

93

94

b

c

a

fragm. 95

96

fragm. of c

b

97

[G] fragm. 97 cont’d

98

99 d

b

e [C]

[G]

[D , F ] a

[F]

fragm.

well. The pitch classes are grouped by beams to show the distinct hexachordal statements, some of them fragmented. Fragments occur mostly at the end of each segment, as indicated; they are effectively interrupted by what follows.34 The lower-case letters identify specific orderings of hexachord A (ordering a is shown in Exs. 10 and 11b). The foregoing examples illustrate – with the help of information from Santi’s article and through comparison with Maderna’s Quartet – the ways in which Berio’s Quartetto is built from rereadings of a basic pitch-class material generated from ‘permutations of pitch-series’. Santi tells us, as noted earlier, that the work consists of six large sections (‘structures’), each subdivided into six subsections (‘sequences’). Each of the six large sections – with the exception of the fourth – starts with durations taken predominantly from the six basic note values, followed by subsections which introduce increasingly longer durations that are multiples of these basic values.35 Exs. 7a and 12a reproduce the beginnings of large sections (sections 1 and 3 respectively) using mostly the six basic durations, whereas Ex. 11a reproduces a third subsection (of section 3) which introduces quintuples of semiquavers and of triplet semiquavers (subdivided in bar 121) alongside shorter values.36 According to Santi, Berio groups rhythmic values into cells of various patterns. They range from single attacks and groups of two or more successive attacks to patterns containing rests. Ex. 13 shows the most prevalent cells as listed by Santi.37 Smaller cells are frequently embedded within larger ones, such as the two demisemiquavers at the beginning of (a) embedded in the second cell Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

322

Ex. 13 Rhythmic cells mentioned by Santi (1960 [1958], pp. 100–1) (a) two attacks in a row 3

;

5

;

;

etc.

(b) cells of ‘3 + rest + 1’ 3

;

3

3

;

3

;

etc.

(c) cell of ‘4 + rest + 1’

Ex. 14a Analysis of rhythmic cells assigned in bars 1–6 (compare with Ex. 10) bar:

1

2

3

4

5

6

* Hexachord A

(1) 3

3

3

3

3 Hexachord B

(2) 3

3

3 3

5

5

5

5 Hexachord A

(3)

Ex. 14b Analysis of rhythmic cells assigned in bars 92–94 bar:

92 5 5

93 3

3

94 5

5 Hexachord A

Hexachord B

of (b), and the two dotted semiquavers at the end of (a) embedded in the first cell of (b). Some of these patterns occur in the score excerpts we have seen. The analysis of Ex. 14a shows how the opening of the work is built from single attacks and cells of double attacks. Here, most rhythmic cells are assigned to pitch classes from the same hexachordal layer (an exception is the D in the first layer, marked with an asterisk). In Ex. 12a, on the other hand, the rhythmic cells cut across the hexachordal layers. Most cells consist of two successive attacks. Single attacks and patterns of 1 + rest + 2, 1 + rest + 1 and three unequally spaced © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

323

attacks also occur (Ex. 14b).38 Not all pitches assigned to these attacks are equally prominent in the texture, however, since Berio mixes together arco (sordino), col legno battuto and pizzicato timbres. Berio’s work with rhythmic cells parallels similar practices found in the music of other Darmstadt composers at the time. In particular, Pierre Boulez – under the influence of Olivier Messiaen’s rhythmic techniques and his own (and presumably Messiaen’s) analysis of The Rite of Spring – designed various procedures to synthesise a handful of basic rhythmic cells into larger patterns, as found in works such as Polyphonie X (1950–1, withdrawn) and Le marteau sans maître (1953–5, rev. 1957).39 Maderna’s and Luigi Nono’s early serial works often employed rhythmic cells as well, many of them abstracted from popular music and political songs.40 In the fourth section of the Quartetto, Berio combines rhythmic cells with another technique which at the time was frequently associated with serialism: canon. The opening of this section is reproduced in Ex. 15a, with the first three canonic entries signalled by arrows (bars 161, 168 and 175). The successive events in each canonic voice, including rests, are numbered. The order numbers for statement 2 are shown in square brackets, those for statement 3 in italics. Ex. 15b analyses the canonic theme, reduced here to its succession of pitch classes and rests (the latter indicated generically by crotchet rests).41 As the beamed groups illustrate, the pitch-class material again arises from a combination of the two complementary hexachords A and B, this time combined to form a single canonic voice. The first ordering of hexachord A corresponds to permutation ‘a’ (see again Ex. 10, bars 1–2, with D omitted, and Ex. 12b, bars 94 at ‘a’ and 97–98 at ‘a’). The other orderings of the hexachords in Ex. 15b introduce new permutations. Berio’s canon is a proportion canon: the first entry of the theme moves in dotted crotchets, followed by the second entry in minims and the third again in dotted crotchets.42 Irregularities, such as shortened or lengthened events, attest to the flexibility with which Berio handles his materials.43 This canon presents a contrapuntal technique that is not used anywhere else in the Quartetto, but that ties in nicely with Berio’s general approach to serialism: like the other sections of the Quartetto, the canon of section 4 consists of different readings of the same pitch-class material, in this case a fixed succession of 32 events, read at different speeds in contrapuntal imitation. In each reading Berio freely omits and adds pitch classes, freely alters rhythms and freely rearranges register, articulation and (probably) dynamics. In addition, the canonic theme itself is a rereading of pitch-class combinations used elsewhere in the work, constructed from permutations of the two chromatic hexachords A and B. Many of the gestures in the canon use rhythmic cells found throughout the other sections of the work in augmentation.44 This investigation into the serial materials of the Quartetto per archi necessarily remains speculative. Although the excerpts discussed here confirm ‘permutations of pitch-series’ by reorderings of the two chromatic hexachords, in the end it is Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

324

Ex. 15a Berio, Quartetto, beginning of fourth section, bars 161–178, with canonic voices marked 6 161

7

8

9

sord.

pont . 5

SV

tast.

1

via sord.

4

3

2

10

no. 1

167

11

12 MV

13 VO

14

[1] sord. ord .

15

[2] VO

VM

16

17

[4]

[3]

18

[5]

[6]

[7]

[no. 2]

173

sord .

[8]

21

22

[12] balz.

[11]

26

[9]

via sord.

19 20 sord. ord .

[10]

24

25

sord. VM 23

2

3

27

28

6

4

5

[13]

segue no. 3

24 1

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

325

Ex. 15b Theme of the canon in bars 161–214 1

2

3

4

a

5

6

7

8

9

10

11

12

13

14

15

16

17

[D]

(or D /E)

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

impossible to be sure that Berio in fact developed the permutations from these exact hexachords, as strong as the analytical evidence may be. In addition, the principle of permutation remains unclear. In light of the influence of Maderna’s serial practices at the time, one might wonder whether Berio’s ‘permutations of pitch-series’ may have followed a strict principle comparable to Maderna’s use of magic and other squares in order to generate serial permutations.45 Since no documentation of the compositional process survives, and since, as noted above, Santi states that Berio ‘move[d] with unrestricted freedom’ in realising his serial materials (‘he may leave out notes and durations or add some’), the basic serial scheme remains hidden in the final version. IV. Allelujah I As the following examination of the draft score for Allelujah I shows,46 Berio developed the basic materials for the work from strict serial procedures. As in the Quartetto, these materials were then subjected to multiple readings. Berio describes the process in his early article ‘Aspetti di artigianato formale’, which appeared in the first issue of his journal Incontri musicali in 1956. He explains that Allelujah I (then still titled Allelujah) is based on a continually recurring material, first presented in the opening 21 bars, which Berio calls the ‘matrix for the entire piece’ (Berio 1956, pp. 56–7). More specifically, he states: In Allelujah, the initial structure (first group) was conceived from the outset as a single and, in certain aspects, intuitive whole where the vertical pitch relationships were not the consequence of a horizontal pitch succession (or vice versa), where the distribution and disposition of the instruments was [sic] not a direct consequence of [predetermined] registral zones, and where the succession of durations was not analysable as a series of note values ... [b]ut where, on the contrary, all sonorous aspects were chosen and given unequivocally because they had to be chosen and given thus, and not otherwise; and where, finally, the sonorities of this first ‘formal object’ [the first 21 bars] could successively provide materials to be broken down [‘elements of analysis’] and for the formal structure, whenever taken deliberately in their ‘concrete’ sense.47

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

326 Ex. 16a Allelujah I, bars 1–4 = 132 ca. 1 Fl. 1

Picc.

Picc.

I Picc.

Ob. 1

Cl. 1

8va [sic] Harp 1 8va

III

Harp 2

Vibr.

VI

Vn

Ex. 16a reproduces the opening of the work. Each time the material established in the first 21 bars reappears – Ex. 16b and c show the beginning of the second and third sections – the pitch-class structure is preserved, while the rhythms are varied to a limited degree, and register, orchestration and mode of attack are changed more drastically.48 For instance, most pitch classes in bar 1 of Ex. 16a are reassigned new registers and completely different timbres in bars 22–23 (Ex. 16b) and 61–62 (Ex. 16c).49 In addition, Berio alters the temporal alignment. The two simultaneities from bar 1 (C–G and C–D in Ex. 16a) are pulled apart in bars 22–23 (Ex. 16b) and 61–62 (Ex. 16c).Whereas all attacks in Ex. 16a fall on a quaver beat or semiquaver offbeat, Ex. 16b and c introduce new triplet and quintuplet subdivisions of the beat, obliterating the metric pulse audible in Ex. 16a. Allelujah I is built from different readings of the first 21 bars and from different combinations of such readings, such as the superimposition of one version and © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

327

Ex. 16b Allelujah I, bars 22–26 3

22 Fl. 1

Fl. 3

I Cl. 1 3 3

sord .

3

3

Vlc. 3

3

3

3

E cl.

Alto sax.

II

3

3

Bsn 1

pizz. Cb.

III

Harp 1

the retrograde of another. Berio’s strategy of generating new textures by completely recasting the attributes of his chosen material arose, in the composer’s own words, from ‘the conviction, that to render unrecognisable, or better, to vary continuously the acoustic characteristics of the same sonorous material means equally (in relation to a formal design) to produce a new sonorous material’.50 But how did he construct the basic material of the first 21 bars in the first place? Berio’s draft housed at the Paul Sacher Foundation presents the pitch-class and rhythmic structure in short score (31 pages), to be worked out further in the final version. The draft contains only a few analytical annotations, including the listings of two series in letter notation, one twelve-note (on p. 12) and one eleven-note repeating one pitch class (on p. 27). No other series are identified, however. Berio’s discussion of the work in ‘Aspetti di artigianato formale’ does not clarify to what extent, or even whether, he used pitch-class series. Pointing out how various readings and recombinations of such readings of the first 21 bars enabled him to create widely different textures, Berio writes: Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

328

Ex. 16c Allelujah I, bars 61–65 61 E cl. 3

3

3

3

3

3

3

3

3

Alto sax.

Ten. sax.

II

3

Bsn

3

3

Cbsn 3

arco Cb.

sord. scura 1 Tpts 2 + 3

IV

Hns

5

+

2

Cymbals

3

Tamb. mil.

(pizz.)

3

3

arco

pizz.

Vns

VI

via sord.

3

5

Vlas

The interest I have put into cancelling the signs of the continuous presence of the material of the first group of pitches [that is, the first 21 bars] was not an end in itself. Nothing, indeed, could have prevented me from reconstituting the groups [that is, the different sections of the work] on the basis of a twelve-tone series, permuting and transposing its elements. What interested me was to go along with

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

329

the formal suggestions derived from the ‘destruction’ of that initial material and, inversely, to discover which material would have satisfied those suggestions, overcoming, that is, the concept of interval and pitch series.51

In order to understand what is meant here, we need to examine Berio’s commentary alongside his draft score. Transcribed at the top of Ex. 17 are the first 8 bars of the draft (at I).52 Below this, at II, appears a transcription of the corresponding bars 22–35 of the second section, aligned here with I so as to show the shared pitch-class material. The bottom of the example, at III, presents a transcription of the corresponding bars 54–65 from the third section, again aligned in order to show how this section rereads the same pitch-class materials. It soon becomes evident that section I opens with the successive entries of five different twelve-note series, as labelled in bars 1–5.53 In section II these five series are realigned temporally. Series 4 enters earlier. Series 2 starts slightly sooner and unfolds somewhat faster than in bars 1–5. In section III, the five series are slightly shifted once again.54 Of these five series, the second is the one later listed in the draft in letter notation.55 Although none of the others are identified by Berio, their identities become evident once we compare the rhythmic profiles of I with those of II and III. Series 1, 3 and 5 appear in II with the same note values and rests as in I. Series 2 and 4 retain the same note values but shorten all rests by one-third. Series 2 and 5 occur in III with the same durations as in I. Series 1 and 4 keep the same note values (quavers) but shorten the rests by one-third, while series 3 expands the note values to quintuplet dotted quavers and shortens all rests by one-fifth (with some exceptions). The layering of different series with distinct rhythmic profiles resembles what we have already seen in Maderna’s and Berio’s String Quartets. Once we recognise this general principle in the draft for Allelujah I, it is possible to determine – via a distributional analysis which takes into account Berio’s rhythmic transformations – that the entire first section of 21 bars is in fact constructed serially. The result of this analysis is shown in Ex. 18a–c. Ex. 18a demonstrates that section I (bars 1–21) is constructed from twelve different twelve-tone series, none of which relates to any of the others via canonical twelve-tone operations. Each pitch class is assigned a duration of one quaver. All rests are multiples of quavers or semiquavers.56 Ex. 18b shows that section II (bars 22–53) is built from exactly the same twelve pitch-class series. The odd-numbered series retain the same rhythmic profile as in Ex. 18a; all even-numbered series preserve the durations assigned to the pitch classes (always a quaver) but shorten the rests by one-third, including the rests which precede the first pitch class to enter.57 As a result, the temporal relationships among the odd-numbered series remain the same, while those involving the even-numbered series change. The latter unfold more quickly in section II than in section I. As Ex. 18c demonstrates, section III (bars 54–80 of the draft, bars 61–87 of the final version) is again built from the same twelve twelve-note series, twothirds of which is subjected to rhythmic diminution of the kind seen in Ex. 18b. Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

330

Ex. 17 Allelujah I, beginnings of sections I–III as they appear in Berio’s draft, superimposed (Paul Sacher Foundation, Collection Luciano Berio) 8 8

[series 2]

2

3

[ ]

4

8

[series 1]

8

I

[series 3]

[series 4]

[series 3] 22

23

24

25

26

27

28

[series 1]

II

3

3

3

3

[series 2]

3

3 3

3

3

3

3 3

[series 6]

[ ] 3

8

3

3

3

[series 4]

[series 1] 8 54

8 3

3 3

3

55

56

3

3

57

3

58

3

3

59

60

3

3

[series 2]

III

[series 3] 5

5

[] 5

5 3

3

[series 4]

3 3

3

3

5

[series 6]

8

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

331

Ex. 17 Continued [series 6]

8

5

6

7

8

[ ] 8

[8

8

]

[ ] [series 7]

[series 5]

29

30

31

32

33

34

35

[series 8] 3

3 3

3

3

3

3

3

3

3

3

3

3

3

3

[series 5]

3 3

61

5

3

62

5

63

3

64

5

5

5

5

[series 7]

65

5

3

3 5

3 5

[series 5]

3

3

3

3 3

[series 7]

3

3

3

3

3

3

[sic]

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

332

Ex. 18a Allelujah I, the twelve twelve-note series and their assigned durations in section I 1

2

3

(9 )

4

(20 )

5

(32 )

6

(37 )

7

(59 )

8

(69 )

9

(89 )

10

(101 )

11

(109 )

12

(124 )

In series 1, 4, 7 and 10 of Ex. 18c Berio retains the note values (always a quaver) and shortens the rests by one-third compared to Ex. 18a.The rests in series 3, 6, 9 and 12 of Ex. 18c are shortened by one-fifth.The durations of the pitch classes increase in series 3 and 6 to a quintuplet dotted quaver, while in series 9 and 12 the durations are changed irregularly. Since series 11 remains mostly unaltered in Ex. 18a–c and enters in approximately the same place in all three sections (after a rest of 109 or 108 semiquavers respectively), and since series 12 always ends before series 11, all three sections have approximately the same length in the draft (section II is one semiquaver shorter and section III two semiquavers shorter than section I).58 As Ex. 18a–c prove, the temporal realignment of the pitch-class material in sections I–III follows strict transformational procedures; sections II and III are not simply free rhythmic rereadings of the same pitch-class material. Berio’s comment, cited above, that in section I ‘the vertical pitch relationships were not © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

333

Ex. 18b Allelujah I, the twelve series and their assigned durations in section II 1

3

3

3

3

3

3

3

3

2

* 3

(9 ) 3

4

(20 )

5

(33 )

3

3

3

* 3

3 3

* 3

3

3

3

3

3

3

3

3

3

* 3

3

3

3

3

3

* 3 3 3 3

3

3

3

*

3 6

(37 )

7

(59 )

8

(69 )

9

(89 )

* 3 3

3 3 3

3

*

3

3

* 3

*

3

3

*

3

3 3

3 10

(101 )

11

(109 )

3

3

3

* 3

3

3

* 3

3

*

3

3

* 3 3 *

3

3

* 3 3

3

3

3

3

3

3

*

3 12

3

3

3

3

* 3

* 3

3

* 3

3

(124 )

the consequence of a horizontal pitch succession (or vice versa)’ is at first glance cryptic, because he did work with specific horizontal pitch successions, that is, the twelve twelve-note series. What he in fact meant is that the horizontal pitch successions are not locked a priori into a particular vertical alignment (because the alignment is altered in sections II and III).The analysis in Ex. 18a–c also clarifies Berio’s earlier comment that ‘[n]othing, indeed, could have prevented me from reconstituting the groups on the basis of a twelve-tone series, permuting and transposing its elements’.The composer plainly did not mean to say that he did not employ any twelve-note series at all, a claim that is false as it stands. Rather, he seems more concerned to emphasise the transformational over the additive Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

334

Ex. 18c Allelujah I, the twelve series and their assigned durations in section III 3

3

3

3

3

3

3

3

3

3

3

3

1

2 * 5

5 3

5

5

5

5

5

5

* *

5

5

5

5

* 5

5

* 5

5 5

5

5

5

5

*

5 5

3

3

(20 )

5

(32 ) 5

5

3

5 5

3

* 3

3

5

5

*

5

5

3

5

5

5 5

3

3

5

*

3

3

5

3

*

5

* 3

3

5

*

* 3 3

3

*

* 5 5

*

*

3

5

5

5

5

5 5

(37 )

5

* 3

3 7

(59 )

8

(69 ) 5

9

* 5 5

5

(9 )

4

6

5 5

3

5

3

5

3

5

3

3

5

3 3

5

3

3

5

5

3 3

3

3

5

*

3

3

* 5

5

3

3

(89 ) 3

10

(101 )

11

(108 )

12

(124 )

* 3 3

3

3

3

3

3

3

*

3

3

3

* 3 3

3

3

3

3

durations irreg. 5

3

*

5

5

5

5

5

5

5

5

5

5

5

durations irreg. 5

process of composition: in Allelujah I new textures are generated by transforming a complete larger block of material consisting of the first 21 bars, which Berio conceived as a ‘single whole’.59 Still, that block is put together exclusively, and perhaps surprisingly given the attendant claim of ‘overcoming, that is, the concept of interval and pitch series’ from a collection of a dozen twelve-note rows.60 In a letter to Luigi Nono, probably written in March 1957, Berio argued – regarding Maderna’s String Quartet and his own music – that in ‘the latest developments of serial music ... the series, as such, is dead and buried: it only serves to prepare a material over which the music is invented’.61 In Allelujah I, sections II and III – and the rest of the work – are ‘invented’ by rereading the first © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

335

21 bars, each time completely recasting the registers, dynamics, modes of attack and orchestration.62 Key to this process was the fact that Berio chose material for the first 21 bars of his draft which he considered to be broadly flexible in terms of possible compositional realisations. Thus it is possible to detect the seeds of what would become a central element in the composer’s music: namely, the notion of openness. In the article of 1956, Berio speaks of the ‘multi-polarity’ of the music of Allelujah I, with respect to the act of composition as well as the process of listening. For Berio, the basic material of Allelujah I (the first 21 bars) was ‘multi-polar’ in that it availed itself of a wide range of compositional realisations. Furthermore, Berio scored this material and its transformations in such a manner that the resulting textures, and with them the work itself, remained ambiguous in the sense that each listener would, and was expected to, hear them in his or her own way: In short, I wished to grant each aspect of the composition the possibility of being ‘equivocal’ and to provide a multiplicity of resolutions as regards not only the sonorous and structural aspects of the work, but also those strictly practical and functional that concern the habits of listening; in order also to give the listener an active part in the realisation of the work.63

For Berio, the physical location of the sounds in the concert hall plays a central role in the listening process.The six groups (‘zones’) of the orchestra are seated on stage as far apart from each other as possible.64 Each section of the work rereads the same pitch (or pitch-class) material (varying the rhythms, registers, and so on) but distributes it differently among the orchestral groups.65 Hence, in each section the pitch materials move differently in space. In addition, their paths sound somewhat different for each listener depending on where he or she is seated.The work is thus multi-polar not only in the sense that each listener will likely perceive the complex textures in a different way (focusing on different aspects of them), but also in that the sounds move differently in space depending on where the listener is positioned.66 Ultimately, however, Berio was dissatisfied with the result of the distribution of the six orchestral groups on stage and subsequently reworked the composition into an expanded version for five orchestral groups scattered through the audience. In his program note for this new version, Allelujah II, Berio addressed the function of space and its role in the listening process: In 1955, when I composed Allelujah for six orchestral groups (dedicated to Karlheinz Stockhausen and first performed in that same year [sic] in Cologne with Michael Gielen conducting), I was interested in an extremely elaborate and concentrated development of a simple initial statement. But the distribution of six orchestral groups on a conventional concert stage was not acoustically suitable. This is why, in 1957–1958, I wrote Allelujah II for five orchestral groups, where I further developed that same initial statement, in search of a deeper homogeneity and coherence between the acoustic and spatial dimension on [the] one side and the musical dimension on the other ... . The five orchestral groups of Allelujah II are no longer crowded together on the stage: they are distributed in the hall so as

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

336

to surround the audience. The purpose, given the complexity of the score, is to help the audience approach the work from different acoustic standpoints and to become more involved with the musical development, in a continuous ‘alleluiatic’ expansion.67

In his next instrumental work Berio extended the notion of openness beyond the compositional means and the listening process to include the act of performance itself. In the flute Sequenza (1958) the realisation of the rhythms is flexible in that Berio’s notation no longer fixes the exact note values. The performer makes the specific rhythmic choices according to the distribution of the pitches within the time units marked in the score.68 Although it uses some serial elements, Sequenza I is no longer serial in any strict sense.69 Berio recognised early the dangers of using serialism in dogmatic and inflexible ways. The examples from the mid-1950s examined here show clearly the ways in which Berio evaded ‘the formalistic and escapist attitude of twelve-tone composition’ in his own serial music. Looking back in 1968, he wrote: ‘To me ... it is essential that the composer be able to prove the relative nature of musical processes: their structural models, based on past experience, generate not only rules but also the transformation and the destruction of those very rules’ (Berio 1968, p. 169).Although Berio had abandoned serialism by 1958, thinking in terms of musical parameters and serial ordering processes would remain characteristic of his musical aesthetic.Traces of serial thinking can be found throughout his later oeuvre, and in this sense serialism shaped the rest of his compositional career. NOTES An earlier version of this paper was presented at the 2005 Annual Meeting of the Society for Music Theory. I wish to thank Talia Pecker Berio for sharing her extensive knowledge of Berio’s music and writings with me, and for her comments and suggestions on the draft of this article. All primary sources are quoted and reproduced here with her permission. Research visits to the Paul Sacher Foundation in Basel were supported by grants from McGill University and the Social Sciences and Humanities Research Council of Canada. My thanks go to the scholars and staff at the Paul Sacher Foundation for their assistance. My transcriptions of Berio’s and Maderna’s sketches are reprinted by permission from the Foundation. Berio’s note for Nones is translated in Ex. 2 with permission from Marion Boyars Publishers, London. Excerpts from Berio’s Nones, Allelujah I and the Quartetto per archi and from Maderna’s Quartetto per archi in due tempi are reproduced by permission of Sugarmusic S.p.A. – Edizioni Suvini Zerboni, Milan. Ex. 13 is cited by permission of Universal Edition A.G., Vienna. An excerpt from a letter from Berio to Luigi Nono is quoted by permission of the Luigi Nono Archive,Venice. Its president, Nuria Schoenberg Nono, and artistic director, ClaudiaVincis, gave much helpful advice during my time there. I am grateful to Federico Andreoni for his help with my translations. 1. The serialism of Due pezzi is analysed in Borio (1997), pp. 383–6. Seither (2000), p. 12, discusses the general features of Study.The work is analysed in Hermann (2009). Cinque variazioni and Chamber Music are briefly discussed in Allen (1974), pp. 23–4. Excerpts from these two works are also analysed in Osmond-Smith (1991), pp. 6–10. © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

337

Santi (1960 [1958]), p. 101, addresses selected features of Variazioni. The most frequently discussed work from this period is the integrally serial Nones; see Santi 1960 [1958]), pp. 99–100; Smith Brindle (1958), pp. 96–101; Allen (1974), pp. 24–30; Stoianova (1985), pp. 379–82; Hicks (1989); Osmond-Smith (1991), pp. 16–19; and Carone (2007–8), pp. 28–46. Score excerpts from Allelujah I are discussed in Berio (1956), Osmond-Smith (1991), pp. 19–21, and Fein (2001), pp. 251–63. No extensive analyses of Allelujah I, Quartetto or Serenata I have been published to this date. The earliest and most specific analytical information on the Quartetto is found in Santi (1960 [1958], pp. 100–1). Excerpts from this work are also discussed in Allen (1974), pp. 30–3; Fein (2001), pp. 263–8; and Hermann (2009). Allelujah II is examined in detail in Carone (2007–8). 2. Quartetto, Serenata I, Allelujah I and Allelujah II are mentioned (but not discussed in any detail) in Berio (1985 [1981]), pp. 63, 65, 90 and 154. Allelujah I is discussed in Berio (1956). Additional brief comments by Berio on Serenata I are reproduced in Stoianova (1985), pp. 383–5. 3. Berio must be quoting Milton Babbitt here, who in his review of René Leibowitz’s Schoenberg et son école and Qu’est-ce que la musique de douze sons? from 1950 discussed the possibility of applying twelve-note principles in both rhythmic and pitch domains. Babbitt’s exact wording is: ‘Thus there arises the reality of a rhythmic structuralization totally identical with the tonal structuralization, the two elements integrating with each other without harm to the individuality of either one’ (Babbitt 1950, p. 14). Babbitt clearly uses the term ‘tonal’ here to mean ‘pitch’ in the context of twelve-tone composition.The paragraph from Babbitt’s review which contains this sentence had been cited three years prior to Berio’s article in Peter Westergaard’s critique of Babbitt’s procedures in ‘Some Problems Raised by the Rhythmic Procedures in Milton Babbitt’s Composition for Twelve Instruments’. Westergaard’s article appeared in what was at the time the journal of the American serialists, Perspectives of New Music (Westergaard 1965). 4. Berio here paraphrases Hegel’s ‘to give out its [knowledge’s] Absolute as the night in which, as we say, all cows are black – that is the very naïveté of emptiness of knowledge’ (Hegel 1964, p. 79). In his first Norton lecture, given in 1993, Berio likewise emphasises ‘solidarity among musical elements’ (Berio 2006, p. 11). 5. As an example of what constitutes a meaningful whole, Berio recalls: ‘As I pointed out to Pousseur myself, the processes that generate melody cannot be manufactured from one day to the next – melodies are born spontaneously within collective groups or in a stylistic frame when all the “parameters” of music are at peace, and start “singing” together’ (Berio 1985 [1981], p. 79). 6. See Santi (1960 [1958]), p. 100. 7. See Berio (1985 [1981]), pp. 51 and 62.Whether Berio first attended Darmstadt in 1953 or 1954 remains uncertain, however. See Carone (2007–8), p. 29. 8. Messiaen was probably not aware of Milton Babbitt’s work at the time. Babbitt’s Three Compositions for Piano (1947), with their individual treatment of the parameters pitch, rhythm, dynamics and articulation pre-date Mode de valeurs et d’intensités by two years. See also Mead (1994), pp. 23–5. 9. Stockhausen wrote Kreuzspiel under the influence of Goeyvaerts’s Sonata for Two Pianos after the two composers first met in Darmstadt in the summer of 1951

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

338

christoph neidho¨ fer (Goeyvaerts 1994, p. 45). Also predating Kreuzspiel and Structure Ia is Michel Fano’s Sonata for Two Pianos (1951), which serialises the parameters of pitch class, rhythm and dynamics, but not the modes of attack. See Toop (1974), pp. 164–9.

10. See for example Pousseur (1959), especially pp. 67–88. 11. See Stockhausen (1963). The article was written in 1955. 12. See Berio (1985 [1981]), p. 62. 13. Ex. 2 is a translation of the second of the two pages of this note. On the first page Berio explains the intervallic properties and symmetries of the thirteennote series and mentions the use of harmonies ranging from the interval of an octave to the total chromatic. A facsimile of this note appears in Berio (1985), plate 4 (n.p.). All translations of sources in Italian, unless indicated otherwise, are mine. 14. This is mentioned in Hicks (1989), p. 255. 15. Added information which does not appear in Berio’s original note is shown in square brackets in the example. I have identified multiple choices with [a], [b] and [c] for later reference. 16. Goeyvaerts assigns the values 0, 1, 2, 3, 2, 1, 0, 1, 2, 3, 2 and 1 to the twelve pitch classes from E through to D, values 3, 2, 1, 0, 1, 2 and 3 to seven durations (ranging from a quaver to nine quavers), values 1–4 to four dynamics (pp, p, mf and f) and values 1 and 2 to four different modes of attack. See Sabbe (1981), pp. 9–10, and (1994), p. 55. 17. The title refers to the ninth canonical hour. Berio had originally planned Nones as ‘a great secular oratorio with solos, chorus and orchestra’, but the length and complexity of Auden’s poem stalled the ambitious project. The final version of Nones assembles ‘five orchestral episodes’ from the original uncompleted project (Berio 1985 [1981], pp. 62–3). 18. In addition, unlike Goeyvaerts, Berio allows his sums to exceed the ‘synthetic number’, adding even more flexibility to his choices. 19. The suffixes ‘a’ and ‘b’ denote the specific choices made where Berio would have had multiple options. 20. See also Hicks (1989), p. 267. 21. Similar tendencies in the numerical distribution are apparent in the remaining three serial layers which open the work, although the sums do not form palindromic patterns and, mistakenly, occasionally even fall below 9. As in layer P11, the values for the durations and dynamics in the remaining three layers largely decrease from either end towards the centre (again, there are exceptions): © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

339

P 5 layer pcs:

1

2

3

4

5

6

7

6

5

4

3

2

durations :

4a

4a

4a

1

3b

2b

2b

3a

2b

1

3a

3b

1 4b

dynamics:

2a

2a

2a

3

5

1a

1a

1b

2a

1a

1a

2a

3

modes of attack: 2a

2a

2a

1a

1b

2a

1b

1a

1a

2a

1a

2a

2b

sums :

9

10

11

9

14

11

11

11

10

8(!)

8(!)

9

10

pcs:

1

2

3

4

5

6

7

6

5

4

3

2

1

durations :

4a

3b

3a

2b

2b

4b

1

1

3a

2b

2b

4a

4a

dynamics:

1a

3

2a

1a

1b

1a

1a

1a

1b

1a

2a

1a

2a

P 7 layer

modes of attack: 3c

1b

1a

2a

2a

1a

2a

1a

1a

1b

1b

2a

3c

sums:

9

9

9

9

10

12

11

9

10

8(!)

8(!)

9

10

pcs:

1

2

3

4

5

6

7

6

5

4

3

2

1

durations :

3b

4a

4a

3b

3b

1

3a

3a

2b

2b

4b

3b

4b

dynamics:

4

1a

2a

1a

1b

1a

1b

1a

1a

2a

1b

2a

2a

modes of attack: 1b

2a

2a

1a

2a

2a

1a

1a

1a

1a

1a

2a

2a

sums:

9

11

9

11

10

12

11

9

9

9

9

9

P 10 layer

9

Berio does not always add a quaver rest to an event whose sum exceeds 9, as otherwise required by the rule. Another statement of P10 starts in bar 5 (violin A). Bruno Maderna analysed the first four serial layers (without calculating the numerical values) in his lecture notes for Darmstadt in 1954. The corresponding page is reproduced in Berio (1985), plate 5 (n.p.). 22. Not all of the leaps are equally prominent in the full texture, however, depending on their surroundings. The forward drive of such gestures is strongest where a crescendo and/or glissando is involved, such as in the electric guitar in bars 8–9, the saxophone in bars 9–10 and the timpani in bars 9–10 and 11–12. 23. As before, sum 8 does not satisfy Berio’s rule and must be an exception. 24. The exception is 3, which occurs only twice, realised with the larger of the two possible values. 25. Again, the sums smaller than 9 in Ex. 6b and c are inconsistent with Berio’s rule. Berio also follows only partially the stipulation that any event whose sum is larger than 9 be followed by a quaver rest. David Osmond-Smith (1991), pp. 17–18, with reference to Bergian practice, analyses the lower strings and timpani in bars 40–42 as the first half of a derived series which reads P11 from both ends to the centre (B–F–D–[D]–B–F, and so on). This reading corresponds closely to the analysis shown in Ex. 5, as I5 and P11 are literal retrogrades of each other. Osmond-Smith interprets what I have analysed as the superposition of the beginning of I5 and P5 in bars 43–48 as a partial statement of P8 (‘P9’ in his terminology), reading the pcs in the order 7–1–13–4–10–5–9–6–8. An analysis of the sums based on Osmond-Smith’s reading also leads to values occasionally smaller than 9. I have no explanation for the timpani in bar 45.

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

340

christoph neidho¨ fer

26. Berio uses the term ‘form’ here in the sense of formal scheme or form type, that is, in the sense of a preestablished, conventionalised framework. The preference for thinking in terms of ‘process’ (or ‘formation’) rather than ‘form’ is likely influenced by Edgard Varèse, among others; see Varèse (1971), pp. 28–31. For an excellent analysis of Berio’s concepts of form and formation in the broader intellectual context of the 1950s and beyond, see Carone (2007–8). 27. The authors who wrote for Die Reihe were either the composers themselves or authors close to them (Grant 2001, p. 223). In addition to the Quartetto, Santi’s article also discusses Nones and Variazioni and briefly mentions Cinque variazioni, Chamber Music and Mimusique No. 2. The two musical diagrams in Santi’s article pertaining to Nones and his description of the properties of the thirteen-note series for the work are virtually identical with what appears on the first page of Berio’s own analytical note (the second page of which was seen in Ex. 2), pointing to the composer as the source of information. 28. The exception is the ordering of hexachord A in bars 2–5 (Ex. 7a) and bars 224–6 (Ex. 7c). 29. ‘I was very close to him [Maderna] for a number of years: from 1953 to 1959 it was almost as if we were living together’ (Berio 1985 [1981], p. 52). 30. It is now held in the Collection Luciano Berio at the Paul Sacher Foundation. 31. Space does not permit me to go into the complex serial structure of Maderna’s Quartet, analysed in Fein (2001), pp. 133–83, Borio (2003), pp. 107–11 and Neidhöfer (2009). Maderna subjects the twelve-note series of the work to an elaborate and strict permutational procedure which regroups the pitch classes into successions of single pitch classes, dyads, trichords and rests. Ex. 9 shows the different permutations of the series labelled by Maderna with lowercase letters. Each distinct permutation is realised with one of the twelve basic note values used in the work (ranging from septuplet demisemiquavers to crotchets). Maderna’s sketches suggest that aside from the pitch-class structure and rhythms, no other aspects were serially determined. 32. Allen (1974), pp. 30–3, demonstrates how the ordered set C–B–B–C, canonical transformations thereof and unordered sets of set class 3–3 [0, 1, 4] from the opening of the work recur in later sections. As the present analysis shows, these and other sets are part of a larger transformational structure characterised by the use of the two complementary hexachords. 33. The term ‘distributional analysis’ was coined by David Lidov (1992), pp. 67–8.The method was first introduced, as ‘paradigmatic analysis’, by Nicolas Ruwet (1966) and later integrated into a semiological model by Jean-Jacques Nattiez (1990). 34. Berio’s work with chromatic sets such as the two complementary hexachords A and B may have been influenced by his study of the music of Anton Webern and by the discussions of Webern’s music which had taken place at Darmstadt, especially after 1953. Particularly influential at the time was Henri Pousseur’s analysis ‘Webern’s Organic Chromaticism’, which eventually appeared in the second volume of Die Reihe in 1955 (Pousseur 1958). 35. As mentioned by Santi (1960 [1958]), p. 100, in the first section Berio multiplies the six basic note values by factors of 1, 3, 5, 7, 9 and 11 respectively. © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

341

36. The six large sections of the work, according to the assignment of rhythmic values span bars 1–57, 58–91, 92–160, 161–223, 224–249 and 250–287. See also Fein (2001), pp. 264–7. 37. See Santi (1960 [1958]), pp. 100–1. 38. Other analytical interpretations would be possible too. Each of the rhythmic cells shown in Ex. 14b uses one or two of the six basic note values. It is likely that Berio thought of these small cells as forming larger ones. Santi states, for instance, that the cell shown in Ex. 13c ‘returns in different forms at the beginning of each structure [i.e. section]’ (Santi 1960 [1958], p. 100). This longer cell is a compound of two double attacks followed by a rest and a single attack. The compound could be shown at the beginning of Ex. 14b, which reduces the opening of the third section, by grouping together the first five attacks, including the rest between the fourth and fifth attack. 39. The technique is explained in Boulez (1991b), pp. 121–6. Boulez’s analysis of The Rite of Spring (completed in 1951) appears in Boulez (1991a), pp. 55–110. Messiaen’s analysis of the same work was published posthumously in Messiaen (1995), pp. 93–147. 40. For a discussion of Nono’s early serial rhythmic techniques and a comparison with Boulez’s practice, see Borio (2003). Maderna’s use of rhythmic cells is discussed in Borio (1990), pp. 32–3, Fearn (1990), p. 14 and Borio (1997), pp. 375–81. 41. This canon has been analysed in part previously by Fein (2001), pp. 266–7. My reconstruction of the theme differs from his in a few places, making it possible to account for more of the pitch material. In particular, events 12, 14 and 22–32 of the theme (shown in Ex. 15b) are not included in Fein’s reconstruction. Allen (1974), pp. 31–2A, identifies the first five events of the theme (called ‘motive’) in bars 161–163 and their restatement in bars 168–171, 175–178 and 194–197. He also shows various recurrences between bars 174 and 216 of the first four pitch classes of the motive or twelve-note transformations thereof. 42. A fourth and last thematic statement (not shown in the example) in mostly dotted crotchets starts in bar 194 and ends in bar 214. 43. As marked (underlined), events 12 and 14 in bars 167–170 double the note value (dotted minim instead of dotted crotchet). Events 6 and 10 of the second statement of the theme in bars 172 (dotted crotchet rest in the second violin) and 175 (crotchet G in the viola) shorten the regular note value (minim). Event 12 of the same statement in bars 176–178 (C in the first violin) is extended and subdivided into repeated quaver attacks. 44. See, for instance, the two-note gestures in the cello and viola, followed by a single attack in the second violin at the beginning of Ex. 15a. 45. For a discussion of Maderna’s use of such squares, see for example Rizzardi (2003). 46. The draft is housed in the Collection Luciano Berio at the Paul Sacher Foundation. The published score of Allelujah I, issued under the title Allelujah by Suvini Zerboni in 1957, was copied by Juan Hidalgo in December 1956, as indicated on the last page of the score. The work was probably composed after the Quartetto per archi,

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

342

because Santi’s article does not mention Allelujah and refers to the Quartetto as Berio’s ‘most recent work’ (Santi 1960 [1958], p. 101). 47. ‘In “Allelujah”, la struttura iniziale (primo gruppo) è stata concepita sin dall’inizio come un tutto unico e, per certi aspetti, intuitivo. Cioè: ove i rapporti verticali delle frequenze non fossero la conseguenza di uno svolgimento orizzontale delle stesse (o viceversa), ove la distribuzione e la disposizione strumentale non fosse direttamente una conseguenza delle zone di registro e ove la successione delle durate non fosse analizzabile come serie di durate ... . Ma, invece, ove tutti gli aspetti sonori fossero inequivocabilmente scelti e dati perchè così dovevano essere scelti e dati, e non altrimenti; e ove, infine, i dati sonori di questo primo “oggetto formale” potessero fornire successivamente gli elementi di analisi e di struttura formale, qualora deliberatamente presi nella loro accezione “concreta” ’ (Berio 1956, p. 64). 48. Berio (1956), p. 57. 49. The opening of Ex. 16a uses only the middle to high register, whereas Ex. 16b and c immediately include the low (but not yet the lowest) range of the orchestra. 50. ‘Questo principio generale della composizione mi è stato suggerito dalla persuasione che, anche nella musica strumentale, il rendere irriconoscibile, o meglio, il variare continuamente le caratteristiche acustiche di uno stesso materiale sonoro vuole anche dire (in rapporto a un disegno formale) produrre un nuovo materiale sonoro’ (Berio 1956, pp. 56–7; italics in the original). 51. ‘L’interesse che ho posto nell’annullare i segni della presenza continua del materiale del primo gruppo di frequenze non era fine a se stesso. Nulla, infatti, avrebbe potuto impedirmi di ricostituire i gruppi sulla base di una serie di 12 suoni, permutando e trasportando gli elementi di essa. Quello che mi interessava era di assecondare i suggerimenti formali derivati dalla “distruzione” di quel materiale iniziale e, inversamente, scoprire quale materiale avrebbe soddisfatto quei suggerimenti, superando cioè il concetto di serie di intervalli e di altezze’ (Berio 1956, p. 62). 52. My transcription omits Berio’s circle around the first three bars, labeled ‘DOPO!’ (later), and the indication above bar 1 of ‘SEMPRE DIVISI’. The opening must thus originally have been intended for strings. From bar 9 onwards, Berio reuses selected materials from bars 1–3 (and beyond). He highlights certain pitches by labeling them (fa in bar 1, sol in bar 4, and do in bar 5). I have omitted these labels in the transcription. They select pitches which project one of the twelve-note series (series 2) to be discussed. 53. In the transcription, any information added by the present author, such as the identifications of these series, is shown in square brackets. 54. The five series and their rhythmic profiles were previously identified by Fein. He states, however, that beyond these five series the rest of the first section until bar 21 is not serial (Fein 2001, pp. 254–5). My analysis will show that this is not the case. 55. See page 12 of the draft. Berio does not assign a number to this series. The other (non-twelve-note) series shown in letter notation on p. 27 summarises the pitch classes prolonged in bars 233–247 of the draft. 56. Berio adds additional pitch materials to the structure of the 12 twelve-note series from bar 9 onwards. The added pitches are notated in red in the draft. Fein (2001, © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

343

p. 255), suggests that these were added later, restating pitch-class materials selected from bars 1–9. Berio’s selection appears to have been systematic in the sense that he chose a group of from three to five pitch classes out of each of bars 1–8, adding these groups, in chronological order, to bars 9–12. 57. Irregularities are marked with an asterisk (*) in the example. Berio mentions the rhythmic transformations, in very general terms, in his 1956 article: ‘The variations in the durational relationships in the second and third group [section] take place through gradual and proportional interpolation of irrational values, yet without perceptively influencing the vertical pitch relationships already determined in the first group [section]’ (Le variazioni nei rapporti di durata nel secondo e terzo gruppo avvengono per graduale e proporzionata interpolazione di valori irrazionali, senza tuttavia influire sensibilmente sui rapporti verticali delle frequenze già determinati nel primo gruppo [Berio 1956, p. 57]). 58. The near-equal length of sections I–III is abandoned in the final version, in which Berio extends section II by seven measures of 4/8. 59. In his 1956 article Berio emphasises the holistic conception of the first 21 bars: ‘Whereas once it seemed logical – at the time of “contrapuntal” purification that by now has born its fruits in the unity of method and intuition in the compositional process – to search for a series of durations, dynamic values and timbres that could coincide “a priori” with a pitch series, via systematic procedures often “external” to the composer, it is possible today to carry out a simultaneous and unified choice of the sound properties by grasping the totality of their reciprocal formal predispositions’ (‘Mentre un tempo sembrava logico – quel tempo della purificazione “contrappuntistica” che ormai ha dato i suoi frutti nell’unità di metodo e di intuizione nel lavoro di composizione – cercare che una serie di durate, di intensità e di qualità timbriche potesse coincidere “a priori” con una serie di frequenze, attraverso procedimenti sistematici spesso “esterni” al compositore, oggi è possibile operare una scelta simultanea e unificata dei valori sonori, cogliendo la totalità delle loro reciproche predisposizioni formali’ [Berio 1956, p. 63]). 60. In the first Norton lecture (1993), Berio returned to the relationship between subtractive and additive procedure: ‘Carl Dalhaus pointed out a similar idea regarding the relationship between material and matter: “The brick is the form of the piece of clay, the house is the form of the bricks, the village is the form of the house”. I would like to bring this quotation closer to my own point of view, inverting the order of the images to fit a subtractive rather than additive perspective: “The village is the form of the house, the house is the form of the brick, the brick is the form of the piece of clay” ... . In other words, the elaboration of the cell with additive criteria can be temporarily suspended, and the path that leads to musical sense may move in an opposite direction, calling upon subtractive criteria to a heterogeneous, even chaotic whole of acoustical data. Like the sculptor who extracts the sculpture, a forza di levare (as Michelangelo said), from the block of marble. Such criteria may lead to the discovery of a specific figure: the generating cell’ (Berio 2006, pp. 19–20). 61. It is conceivable that the twelve pitch-class series reconstructed in Ex. 18a–c were generated through a single permutational procedure, as was common in Maderna’s

Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

christoph neidho¨ fer

344

and Nono’s serial music at the time using Latin and magic squares, for example. In the absence of any detailed sketches, I have so far not been able to determine whether Berio in fact used any such permutation strategies. For an introduction to Maderna’s and Nono’s procedures in what Gianmario Borio and Veniero Rizzardi have termed the ‘tecnica degli spostamenti’ (‘shifting technique’), see Rizzardi (2003), pp. 31–54. 62. ‘Penso che se tu devi parlare degli “ultimi sviluppi della musica seriale” devi stabilire il fatto che la serie, in quanto tale, è morta e sepolta: serve solo a preparare un materiale su cui viene inventata la musica’. The original of this letter from Berio to Nono is conserved in the Luigi Nono Archive, Venice. 63. ‘Insomma, disederavo dare ad ogni aspetto della composizione una possibilità di “equivoco” e una molteplicità di risoluzioni che riguardasse non solo gli aspetti sonori e strutturali del lavoro ma anche quelli strettamente pratici e funzionali che riguardano le consuetudini dell’ascolto; per dare anche all’ascoltatore una parte attiva nella realizzazione dell’opera’ (Berio 1956, p. 65). My analysis of the draft score indicates that Berio used the three readings in sections I–III to construct the rest of the work, as shown in the list below. In the left column, the bar numbers from the draft are followed, in parentheses, by those from the final version. Rereadings of sections I–III may be re-rhythmicised and may omit pitch classes and add other materials. In addition, in the final version Berio may add, omit and conflate material (not indicated here). Bars in draft (final / version)

Readings

1–21 (1–21) 22–53 (22–53) 54–80 (61–87) 80–94 (87–101) 95–115 (102–112+) 116–154

I II III IV: superposes retrograde of first half of II over second half of II V: superposes I over retrograde of III VI: bars 116– 35 superpose I over its own retrograde; bars 135–154 reread III; in addition, a prolonged version of series 2 (with selected partials from the harmonic series added to its individual pitches) is superposed in bars 116–154 I III II adds V empty bar in draft (worked out as a transition in bars 173–178 of final version) V prolonged version of series 2 (without first three pitch classes; they are contained, as part of V, in bar 190) I I; bars 233–247 add prolonged series of eleven pitch classes (G –E–F–B –C–A–D–E –B–C –D –F ) II

154–174 (138–159) 154–178 (138–162) 170–188 (154–172) 174ff 189 190–209 (179–192+ ) 191–212 (180–192+ ) 213–232 (205?–208) 227–246 (205?–222) 248–286 (224–278)

64. ‘Compatibilmente con lo spazio disponibile le 6 “zone” in cui è divisa l’orchestra devono essere distanziate il più possibile’ (preface to the score, Edizioni Suvini Zerboni 5372). 65. Generally, Berio has the 12 twelve-tone series move back and forth among different orchestral groups, thus obliterating the original serial counterpoint of his draft score. 66. The movement of sound in space was explored by many composers at the time, especially in connection with electronic composition. Stockhausen’s Gesang der Jünglinge, whose sounds move in space through five loudspeakers located throughout the audience, was premiered on 30 May 1956, at which time Berio was working © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

345

on Allelujah I. Berio (1985 [1981]), p. 65, mentions Gesang der Jünglinge as one of the ‘pieces that affected me most during those years’. Stockhausen was also working on Gruppen, for three orchestras (1955–7), at the time, and it is he to whom Allelujah I is dedicated. For a discussion of the role Berio assigns to the listener in the perception of formal processes see Carone (2007–8), especially pp. 101–2. 67. I am grateful to Talia Pecker Berio for providing me with this programme note. 68. Owing to this feature Umberto Eco cites Sequenza I as an example of an open work (Eco 1989, pp. 1–19). In his fourth Norton lecture, Berio presents a critical assessment of composition with open forms, rejecting those approaches in which (just as in the case of certain serial practices) composers were led ‘not to assume all of their perceptive responsibilities’ (Berio 2006, p. 85). 69. For an analysis of the influence of twelve-tone technique on Sequenza I and a survey of the analytical literature on this work, see Priore (2007).

REFERENCES Allen, Michael Paul, 1974: ‘The Music of Luciano Berio’ (MA thesis, University of Southampton). Babbitt, Milton, 1950: Review of René Leibowitz, Schoenberg et son école and Qu’est-ce que la musique de douze sons?, Journal of the American Musicological Society, 3/i, pp. 57–60, reprinted in The Collected Essays of Milton Babbitt, ed. Stephen Peles with Stephen Dembski, Andrew Mead and Joseph N. Straus (Princeton, NJ: Princeton University Press, 2003), pp. 10–15. Berio, Luciano, 1956: ‘Aspetti di artigianato formale’, Incontri musicali, 1, pp. 55–69. ______, 1968: ‘Meditation on a Twelve-Tone Horse’, Christian Science Monitor (15 July 1968), p. 8, reprinted in Richard Kostelanetz and Joseph Darby (eds.), Classic Essays on Twentieth-Century Music: a Continuing Symposium (New York: Schirmer Books, 1996), pp. 167–71. ______, 1985 [1981]: Two Interviews with Rossana Dalmonte and Bálint András Varga, trans. David Osmond-Smith (New York: Marion Boyars). ______, 2006: Remembering the Future (The Charles Eliot Norton Lectures) (Cambridge, MA: Harvard University Press). Borio, Gianmario, 1990: ‘La tecnica seriale in Studi per ‘il processo’ di Franz Kafka di Bruno Maderna’, Musica/Realtà, 32, pp. 27–39. ______, 1997: ‘L’influenza di Dallapiccola sui compositori italiani nel secondo dopoguerra’, in Mila De Santis (ed.), Dallapiccola: Letture e prospettive (Milan: Ricordi), pp. 357–87. ______, 2003: ‘Tempo e ritmo nelle composizioni seriali. 1952–1956’, in Gianmario Borio, Giovanni Morelli and Veniero Rizzardi (eds.), Le musiche degli anni cinquanta (Archivio Luigi Nono,Venice) (Florence: Leo S. Olschki), pp. 61–115. Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

346

christoph neidho¨ fer

Boulez, Pierre, 1991a: ‘Stravinsky Remains’, in Paule Thévenin (ed.), Stocktakings from an Apprenticeship, trans. Stephen Walsh (Oxford: Clarendon Press), pp. 55–110. ______, 1991b: ‘Possibly ... ’, in Paule Thévenin (ed.), Stocktakings from an Apprenticeship, trans. Stephen Walsh (Oxford: Clarendon Press), pp. 111–40. Carone, Angela, 2007–8: ‘Forma e formazione nella musica strumentale di Luciano Berio’ (PhD diss., Università degli Studi di Pavia). Eco, Umberto, 1989: The Open Work, trans. Anna Cancogni (Cambridge, MA: Harvard University Press). Fearn, Raymond, 1990: Bruno Maderna (Chur: Harwood). Fein, Markus, 2001: Die musikalische Poetik Bruno Madernas: zum ‘seriellen’ Komponieren zwischen 1951 und 1955 (Frankfurt am Main: Peter Lang). Goeyvaerts, Karel, 1994: ‘Paris–Darmstadt 1947–1956. Excerpt from the Autobiographical Portrait’, Revue belge de musicologie, 48, pp. 35–54. Grant, Morag Josephine, 2001: Serial Music, Serial Aesthetics: Compositional Theory in Post-War Europe (Cambridge: Cambridge University Press). Hegel, G. W. F., 1964: The Phenomenology of Mind, trans. J. B. Baillie (London: George Allen & Unwin). Hermann, Richard, 2009: ‘On Becoming Berio: Evidence from the First Three String Quartets’, in Evan Jones (ed.), Intimate Voices: Aspects of Construction and Character in the Twentieth-Century String Quartet (Rochester, NY: University of Rochester Press), pp. 99–137. Hicks, Michael, 1989: ‘Exorcism and Epiphany: Luciano Berio’s Nones’, Perspectives of New Music, 27/ii, pp. 252–68. Lidov, David, 1992: ‘The Lamento di Tristano’, in Mark Everist (ed.), Models of Musical Analysis: Music before 1600 (Oxford: Basil Blackwell), pp. 66–92. Mead, Andrew, 1994: An Introduction to the Music of Milton Babbitt (Princeton, NJ: Princeton University Press). Messiaen, Olivier, 1995: Traité de rythme, de couleur, et d’ornithologie, vol. 2 (Paris: Alphonse Leduc). Nattiez, Jean-Jacques, 1990: Music and Discourse: Towards a Semiology of Music, trans. Carolyn Abbate (Princeton, NJ: Princeton University Press). Neidhöfer, Christoph, 2009: ‘Vers un principe commun: Intégration de la Hauteur et du Rythme dans le Quartetto per archi in due tempi de Bruno Maderna (1955)’, in Geneviève Mathon, Laurent Feneyrou and Giordano Ferrari (eds.), À Bruno Maderna, vol. 2 (Paris: Basalte), pp. 323–58. Osmond-Smith, David, 1991: Berio (Oxford: Oxford University Press). Pousseur, Henri, 1958 [1955]: ‘Webern’s Organic Chromaticism’, Die Reihe, 2, pp. 51–60. ______, 1959 [1957]: ‘Outline of a Method’, Die Reihe, 3, pp. 44–88. Priore, Irna, 2007: ‘Vestiges of Twelve-Tone Practice as Compositional Process in Berio’s Sequenza I for Solo Flute’, in Janet K. Halfyard (ed.), Berio’s Sequenzas: Essays on Performance, Composition and Analysis (Aldershot: Ashgate), pp. 191–208. © 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)

Inside Luciano Berio’s Serialism

347

Rizzardi, Veniero, 2003: ‘La “Nuova Scuola Veneziana”. 1948–1951’, in Gianmario Borio, Giovanni Morelli and Veniero Rizzardi (eds.), Le musiche degli anni cinquanta (Archivio Luigi Nono,Venice) (Florence: Leo S. Olschki), pp. 1–59. Ruwet, Nicolas, 1966: ‘Méthodes d’analyse en musicologie’, Revue belge de musicologie, 20, pp. 65–90. Sabbe, Herman, 1981: ‘Die Einheit der Stockhausen-Zeit ... ’, in Heinz-Klaus Metzger and Rainer Riehn (eds.), Karlheinz Stockhausen / ... wie die Zeit verging ... , Musik-Konzepte, 19 (Munich: Edition Text + Kritik), pp. 6–96. ______, 1994: ‘Goeyvaerts and the Beginnings of ‘Punctual’ Serialism and Electronic Music’, Revue belge de musicologie, 48, pp. 55–94. Santi, Piero, 1960 [1958]: ‘Luciano Berio’, Die Reihe, 4, pp. 98–102. Seither, Charlotte, 2000: Dissoziation als Prozess: Sincronie for String Quartet von Luciano Berio (Kassel: Bärenreiter). Smith Brindle, Reginald, 1958: ‘Current Chronicle – Italy’, Musical Quarterly, 44, pp. 95–101. Stockhausen, Karlheinz, 1963: ‘Gruppenkomposition: Klavierstück I (Anleitung zum Hören)’, in Dieter Schnebel (ed.), Texte zur elektronischen und instrumentalen Musik (Cologne: DuMont), pp. 63–74. Stoianova, Ivanka, 1985: Luciano Berio: Chemins en musique (Paris: La Revue musicale). Toop, Richard, 1974: ‘Messiaen/Goeyvaerts, Fano/Stockhausen, Boulez’, Perspectives of New Music, 13/i, pp. 141–69. Varèse, Edgard, 1971: ‘The Liberation of Sound’, in Benjamin Boretz and Edward T. Cone (eds.), Perspectives on American Composers (New York: W. W. Norton), pp. 25–33. Westergaard, Peter, 1965: ‘Some Problems Raised by the Rhythmic Procedures in Milton Babbitt’s Composition for Twelve Instruments’, Perspectives of New Music, 4/i, pp. 109–18.

ABSTRACT Like many other composers who later distanced themselves from serialism, Luciano Berio (1925–2003) embraced its principles in the 1950s and beyond. While Berio’s early serial techniques from the Due pezzi of 1951 to Nones of 1954 are well known, his subsequent serial practice is still little understood for three principal reasons: in his writings and interviews Berio provided only limited information on his serial works; it is very difficult to decipher Berio’s later complex serial techniques from the published scores alone; and only one sketch survives for any of his serial works from 1951 to 1958 (for Allelujah I, 1955–6). Music Analysis, 28/ii-iii (2009)

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

348

christoph neidho¨ fer

Following a brief examination of the integral serialism in Nones (whose principles have been known for some time thanks to an analytical note by Berio), the present study investigates the serial techniques deployed in the Quartetto per archi (1955–6) and Allelujah I. Berio’s serial materials are reconstructed with the help of distributional analyses and from an historical angle that has been little explored thus far: the influence of Bruno Maderna (1920–1973), Berio’s mentor and close collaborator at the Studio di fonologia musicale in Milan.

© 2011 The Author. Music Analysis © 2011 Blackwell Publishing Ltd

Music Analysis, 28/ii-iii (2009)