PART 2: Tellur

[This part of the exposition contains video examples. You can click the still images to play them]

The issue of stacked identities, with which I concluded the first part of this exposition, is strikingly pertinent to Tellur. Unlike the opening section of Gondwana, this work is not composed by means of instrumental synthesis. However, a new instrument of process is certainly superimposed on the guitar-of-old. Murail writes, ‘Tellur starts out as a kind of wager. How can one produce the long sound continua necessary for my work on procedures, transitions and evolutions, on an instrument that produces brief, plucked sounds?'^[1] Thus, Murail must superimpose on the traditional guitar—with its many ‘brief’ sounds—an instrument that can not only generate continuous sound masses, but can also accommodate his brand of gradual transformations between these sound masses. 

As in ensemble works such as Gondwana, the creation of this instrument of process is an analytical decomposition of instrumental sounds into their ‘elementary components’, followed by a re-composition of suitable sonic aggregates.^[2] However, the nature of the superposition is arguably different. The instrumental synthesis in the opening section of Gondwana is an impressive act of instrumental unification. In typical spectralist fashion, the manifold of the ensemble is replaced with a singularity derived from an acoustic model. Instrumental synthesis directs the listener’s expectation away from the traditional ‘cellular’ approach, and toward an approach that progresses ‘from the global level to the level of detail’ (Murail 2005a: 155).

In Tellur the redirection of expectation is less achieved by an act of unification, and more by a shift of focus toward elements of sound production typically ignored in the classical tradition. To this end, Tellur makes ample use of well-researched and innovative extended techniques. 

The work opens with a rasqueado or strumming technique derived from flamenco practice. Traditionally this is used for strumming chords, which drown out the extraneous sounds created by the attack of the hard nail surface on the string. Here it is exactly those percussive attack sounds that take on a central role in the structure of the musical discourse. 

[Video 1: first line, attack sounds.]

The emphasis on traditionally ignored elements of sound production Tellur has in common with the contemporary work Territoires de l’Oublie for solo piano. As pointed out by Marilyn Nonken, it proved impractical to entirely translate such an approach into a notational system inextricably linked with the conventions Murail tried to upend (2014: 83–85). As a result, the notation of Territoires de l’Oublie, and similarly that of Tellur, allows the performer a certain amount of freedom in realization.

From Murail’s writing, we can surmise this freedom is to direct the performer away from conventional minutiae and toward the totality of the sounding result. Notational excess would actually ‘erode the message it is trying to convey’ (Murail 2005a: 160). But, given the context of Murail’s ideas and the rest of his oeuvre at the time, we can safely assume freedom of realization is not to be confused with freedom of outcome. The goal was very unlikely the creation of an aleatoric musical structure dependent on performer agency.^[3] If one were to be faithful to Murail’s ideas of authorship, access to the ‘ideal score’ would be critical to traverse the landscape of choices an open notation creates. 

This is especially true for Tellur, which often requires the player to stretch their technique far beyond familiar reference.^[4] As a result, the score is for a considerable part intabulation, meaning that it refers more to physical action than sonic result. The instructions the score provides for these tablatures—though impressively researched for a non-guitarist—leave a wide margin for realization. The transitions between different playing techniques, which are mostly left to the discretion of the performer, are but one illustration.

[Video 2: transitions between playing techniques, page 2, systems 4-5.]

Tellur’s performance instructions prescribe that sound aggregates evolve slowly and methods of playing are substituted

one for another indiscernibly. The notation highlights some (but not all) of these transitions with arrows. 

All of this is set in a musical language that embraces Schaeffer’s turn toward sound itself and to sonic exploration far beyond the limits of traditional instrumental practice. But at the same time, it replaces the eidetic solidity of sonic objects with relational perception in the flux of process. 

In such a language, ‘any variation in scale, duration, frequency, density, etc. will instantly alter the overall equilibrium’ (Murail 2005a: 154). In other words, every element of sonic perception is tied up in a sonic totality, and that totality is tied up in the totality of the process. ‘Everything is connected’ (2005a: 152). Every element of a performer’s sound production potentially bears on the most fundamental level of Tellur’s compositional structure. With so much left to their discretion, performers who wish to be ‘faithful’ to this work face a tall order. Their first step to meet that order is to find clues to the ‘ideal score’ in the models that inform its gradual processes.

Bird's Eye View

In Question de Cible, Murail writes: ‘We have certainly sought, after rather intuitive phases, “objective” phases to renew composition’ (2005a: 150). Thierry Alla (2008: 46) marks the period from 1967 to 1979 in Murail’s career as an intuitive phase, a period of experimentation that precedes a period of ‘emblematic works’ inaugured by Gondwana (1980). Tellur was written in 1977, right after works like Sables (1974) and Mémoires/Erosion (1975–76) established Murail’s personal style.

While the separation between these periods is not razor sharp,^[5] classifying Tellur as more intuitive is certainly defendable. The guitar cannot accommodate the intricate spectral techniques Murail used in the ‘emblematic’ ensemble works of the early 80s. Still, Tellur clearly exhibits the preoccupation with processes and continuous change that is congenital to Murail’s spectral writing (Murail 2000). 

As in all of Murail’s other works from the late seventies, this preoccupation translates into a succession of slow and gradual transformations. In the case of Tellur, The composer refers to the transformations as ‘a series of “ascents” and of “descents”’ (Murail 1978). There is little published on the design of these ascents and descents. Rafael Andia (1984) discusses in broad terms a sinusoidal outline of the sound/noise relation as well the role of entropy; and in an undergraduate paper, Jeffrey Bowen (2010) discusses a few isolated constituent aspects of Tellur’s processes.^[6]

Jérôme Baillet (2002) offers a larger foothold for analyzing Tellur’s ascents and descents. He suggests that all works written between 1974 and 1980 move along slow rises and falls in tension, which are generated by the density of the sonic material, frequential ambitus, intensity, and the speed with which phenomena succeed one another.^[7] However, Baillet’s model still turns up a considerable surplus of unexplained information. For example, it leaves undiscussed rudimentary spectral experiments that grew into the more elaborate structures of Murail’s later works; and does not shed much light on how these factors combine in musical discourse.^[8]

Contrary to theorists, performers must figuratively eat all parts of the score. Further headway then requires a descent into the particular. What follows is a practitioner’s analysis. It identifies factors that play a key role in shaping Murail’s processes,^[9] but does so starting from the particularity of Tellur’s material and its place in the context of the composer’s catalog. 

I will make liberal use of my own recordings. These do not serve as illustrations of verbal discourse, quite the opposite: keeping with Murail’s focus on sound proper, the words are meant to highlight a few structural factors in my rendition. They work in tandem with my recordings to shed light on some of the analytical approach involved in my performances. 

Descent into the Particular

The series of ascents and descents that make up Tellur roughly correspond to the sections that are marked in the score with capital letters. They form a chain of gradual transformations that effect increases and decreases in entropy.^[10] In this exposition, I will present an analysis of only sections A and B (an ascent or increase in tension/entropy, followed by a descent). A practice-based analysis of the whole score would be too lengthy, and its opening sections alone provide sufficient material to support the current discussion. 

As is common in Murail’s work from the period, sonic transformations carry clear fingerprints of the electronic studio (Murail: 2005b). In the opening sections of Tellur, the treatment of musical material invites analogies with moving faders, varying tape speeds, and the application of filters. The detailed discussion below highlights how these and other operations play a structural role in the work’s transformational processes.

Section A

Sounds ignored by tradition dominate the beginning of the work. The rasgueado technique, traditionally used to strum chords, is limited to a tremolo on the lowest string, which the guitarist muffles with the left hand. Without cover of the string’s resonance, sounds of nail faces colliding with the string step into the limelight. Above noisy ‘thumps’ with a trace of the muted A₂, high pitches catch the ear, generated by the stretch of strings between the nail and the bridge.

In the minute or so that follows, it is as if Murail installs a system of filters and faders that will help realize his gradual transformations. In fact, given a quarter-tone resolution, all the pitch material specified in the first six lines of the score could be seen as a gradual fading in or out of parts of the natural overtone series on F₀. In the first two systems, an asymmetric cross-fade of rasgueado sounds and string resonance transforms a noisy timbre-harmony^[11] into a single resonant pitch. This is followed by a cross-fade between A₂ and G₃; and then the explicit articulation of an E4 already introduced as an overtone of the intensifying A₂ attacks.

[Video 3: faders in systems 1-6. Numbers in red indicate partials of the F₀ spectrum]

Increases in amplitude and spectral ambitus carry through this play of faders, and the latter continues to the end of section A (amplitude takes a detour but catches up eventually). They are boosted on the way by the introduction of string-crossed tremolos that open the door to expanding arpeggios and eventually strummed chords. Only the fifth system disturbs this gradual expansion with a spike of an accented dyad, followed by a resurgence of attack sounds—a brusque movement of the faders.

[Video 4: evolution spectral ambitus, section A]

The rising intensity is enhanced by a move toward inharmonicity that marks the pitch material, from the A₂ tremolo in the second time bracket up to the next-to-last chord of section A. The example below illustrates this by showing the descending virtual fundamental that comes with such an evolution.^[12] Note that this increase in inharmonicity works in tandem with a densification of pitch aggregates (the final chord excluded).

[Video 5: evolution towards inharmonicity, p.1, system 3 to p.2, system 2.]

When pitch aggregate density and inharmonicity slightly slacken right before section B, rasgueado kicks in and the right hand moves to the bridge. This helps overall tension stay an upward course in two ways:

• Rasgueado noise and attack pitches are reintroduced. The latter move to an extremely high register and almost become a high-frequency noise band.
• High partials of the resonating strings gain prominence. 

The increasingly inharmonic pitch aggregates have finally led up to a twilight state between dense harmony and noise. After the left hand then gradually mutes resonating strings, all that remains are noise and unintelligible attack pitches. This is the noisiest, and therefore most inharmonic point in sections A and B.

[Video 6: Rasgueado at the end of section A]

While harmonicity, aggregate density, and noise help define the ascent of section A on the vertical plane, the periodicity of figurations helps shape it horizontally. For example, the salient periodicity of the attack pitches helps to give the beginning of the piece less tension than the tremolo on A₂ it morphs into—even though the former is a more inharmonic pitch aggregate.^[13]

[Video 7: differing periodicity between the beginning of the first and second time brackets]

Or the perceptual proximity of disorder and stasis are leveraged for a smooth transition from periodic arpeggiations to rasgueado.

[Video 8: transition from periodic arpeggio to rasgueado in section A]

Loose ends and Section B

Before we continue to section B, there are some loose ends to tie, namely those resulting from that ‘brusque movement of the faders’ on line 5 of page one. While a series of shifting cross-string tremolos and arpeggios figures have partially prepared this event, the sudden spike in amplitude and change in texture come as a surprise in an otherwise predictable process. But this interruption is not of the capricious kind.^[14] It instigates two additional processes of which different steps are dispersed along the first two sections of the work. 

The first loose end is the sudden resurgence of rasqueado attack pitches. This draws attention to a dispersed process in three steps:

• Step 1: Recognizable attack pitches at the beginning of the work.

• Step 2: Attack pitches ascend into unrecognizability and turn into noise.

• Step 3 (section B): The dispersed process merges with the processual evolution of the (virtually) uninterrupted sound mass that runs through section A. The merger marks a change in direction. Attack pitches descend and decelerate back into complete recognizability, much like a gradual decrease of speed in the playback of a tape recording.^[15] This results in a rarefication of the spectral ambitus to the high register and a texture that gradually thins out both vertically and horizontally.
  Throughout their descent, the attack pitches first make a stop in the natural overtone series on G₂ and then continue their microtonal slip to a dyad on the thirteenth harmonic of A₂ and the eleventh harmonic of G₂ (both fundamentals appear at the beginning of section C, at the top of page 3). At the beginning of section C, they give way to a soft harmonic on F₅.

[Video 9: first dispersed process]

The second loose end, another dispersed process, consists of accented attacks: 

• Step 1: drastic three-stage reduction of spectral ambitus and spectral density.

• Step 2: A₄ harmonics accelerate and are absorbed into the underlying sound mass.

• Step 3 (Section B): A₄ harmonics suddenly re-emerge. Contrary to step 2, they now decelerate and gain prominence. The deceleration carries over to the ordinario A₄ notes in section C.

[Video 10: second dispersed process]

Bird's Eye View Revisited

The detailed description above shows that three transformational processes run through the first two sections of Tellur, one continuous and two dispersed. I highlighted a few important factors^[16] that shape these processes: some of the ‘elementary components’ into which the sound is decomposed:

• the ratio between the noise factor of the attack and the strength of pitch resonance;
• spectral ambitus;
• the degree of harmonicity; 
• the density of sonic material;
• the periodicity of figurations;
• amplitude.

Each of these factors is not just an open concept, but a space instilled with hierarchy, magnetized by two extremes:

• pure pitch resonance versus pure noise;
• a sine tone versus the whole range of perceptible frequencies;
• a sine tone (extremely harmonic) vs noise (extremely inharmonic);
• from a sine tone to a saturized spectral region;
• regular, salient periodicity vs complete stasis;
• and with hierarchies of amplitude, we are of course all familiar. 

Hierarchies like these enable the ‘vectorization’ of Murail’s musical discourse, the sense that processes have a trajectory and directionality (Murail 2005a)—though it bears repeating they do not exhaustively describe these directional trajectories. 

The ‘ascent’ of section A and the ‘descent’ of section B is a multiplication of the three processes these factors help shape.^[17] Multiplication implies these processes do not just run their independent course. They interact and modulate each other. For example, the ‘brusque movement of the faders’ modulates the evolution of the continuous sound mass that had previously only evolved gradually. Only after a few ‘aftershocks’ (textural shifts and accents of decreasing intensity) can the trajectory return to gradual change.

Entangling processes thus creates a periapt against mechanical predictability, while maintaining a clear overall trajectory. That trajectory is spiral: between the start of the ascent and the bottom of the descent, some characteristics return and others have changed. Attack pitches give way to string resonance twice over, and both times pitch classes A and G take center stage. However, the second time around that resonance enters suddenly instead of fading in; is discontinuous (both a souvenir from multiplication); and the G (and it’s overtones) appears an octave lower.