An electroacoustic society: schismogenetical trends in Trevor Wishart’s Globalalia

Author(s): 
Kevin Dahan
Date of publication: 
2014-08
ISSN: 
2052-7888
DOI: 
10.3943/001.2014.08.0201
 

1.0 Introduction

There is a strong tendency in music analysis to consider works as an assemblage of fixed musical structures. While this is perfectly understandable in the context of written music – after all, the composer operates a reduction in translating his musical thoughts and feelings into two-dimensional graphemes drawn on paper – it is less the case in non-written music, and notably in electroacoustic music. There is no alternative space for music representation and there is no established set of graphical elements to use in order to construct an “interpretation” or from which to derive an analysis. Consequently, a number of analytical tools and methods for electroacoustic music have been developed, some of which can be found on the OREMA website[1]. In this respect, this article won’t try to propose another method for electroacoustic music analysis; it is rather an experiment into shifting some conceptions we might have on electroacoustic music analysis, describing the different processes at work. Worded differently, the approach described below is neither a set of tools nor a method; however some aspects of it have been used and transposed in altogether different contexts.

What is analysis anyway? Going to back to the root of words (something we will do frequently in this article...) often sheds new light on our current usage and conception; in the case of analysis, the Greek verb αναλυειν has the primary sense of “unloose”- which can be thought of as finding ways of making things clearer, and less complicated. In the praxis of Aristotle, it involves deconstructive mechanisms, “breaking up” a problem into smaller, more contained, parts. Most of the time, indeed, analytical works present a thorough deconstruction of a musical work into individual “atomic” elements pertaining to the particular methodological bias chosen by the analyst: for example set theory based analysis would go to greater length in order to determine musical phrases “fit” the theory[2], while some analytical theories are strongly tied to a specific musical language, effectively being a superset grammar- making them worthless in other contexts (Schenkerian analysis is a prime example). This particular deconstructive/reconstructive approach is sometimes mandatory, however it may lead to effectively derive from analysis to interpretation.

In 1942, Susanne Langer came up with the concepts of presentational and discursive symbols, in her landmark book Philosophy in a New Key. Roughly, discursive symbols are commonly found in language and in mathematical knowledge: created from a vocabulary, with conventional meaning, they can be mutually defined and translated into different systems (language). Presentational symbols, on the other hand, cannot be described by statements of logical truth, and simply “are”. If one can permute discursive symbols and not change the meaning of a statement, altering presentational symbols would lead to another proposition. Interestingly, in chapter 8, Langer posits that Music is a kind of presentational symbol, and consequently should not be described as “a language”. While this approach can be thought of as naïve at best, Langer goes further by explaining that Music does offer a “logical” expression of emotions. The interesting aspect of her research lies more in the attempt to weave connections between vastly different fields of knowledge (from logic to psychology, and from art to algebra) than in the resulting propositions[3].

One interesting aspect of Trevor Wishart’s music is that his works lie at the intersection of many fields, from social commentary to symbolist approach, and from acousmatic “tradition” to digital manipulation; there is usually a profound connection to man-made sounds, of which voice is prime. In this respect, Globalalia is a particularly interesting piece to work on, as it exemplifies some aspects of Langer’s thoughts on Music – however naïve and old-fashioned they may seem at first glance. The key here is to understand Globalalia not only as an electroacoustic work, but as an experiment in “musical sociology”: due to the profusion of sounds, manipulations, transformations, and to the length of the piece itself, we are confronted to a complex ecosystem in which musical entities – which we’ll term from now on musical beings – interact.

An analysis of Globalalia will be proposed, structured around the central idea of considering musical elements not as much as structures, but rather as musical beings, with characteristic morphologies, but that are also exhibiting distinctive behaviours, and notably when in contact with each other. This will necessitate introducing an interesting concept to music analysis, borrowed from Gregory Bateson’s cultural anthropology, schismogenesis, which we will present first. We will then proceed to a linear analysis, focusing on specific elements while following the discourse of the piece, before some concluding remarks on this work, on electroacoustic music analysis as praxis, and on the idea of evolving away from musical structures to the more organic concept of musical beings and musical society in the difficult exercise of music analysis.

2.0 Schismogenesis

The concept of schismogenesis is first mentioned in the works of Gregory Bateson, in an article dating from 1935, called “Culture Contact and Schismogenesis”[4]. It is termed as an attempt to describe and analyse situations in which two different cultural groups are forced into contact. The word schismogenesis was coined by Bateson to describe processes that lead from a stable social system, defined by two social groups, to a new organization of this system, or its destruction. The signification of this term is evident, if we consider, once again, the Greek roots: the verb σχιξειν means “to cut, to divide” and γενεσις, meaning “origin, birth”. According to Bateson, schismogesis

[...] is probably the most instructive of the possible end results in contact, since the factors active in the dynamic equilibrium are likely to be identical or analogous with those which, in disequilibrium, are active in cultural change. Our first task is to study the relationships obtaining between groups of individuals with different behaviour patterns, and later to consider what light these relationships throw upon what are more usually called “contacts.

Therefore, in order to analyse such a system through schismogenesis, one has to follow a three-step process: (1) identify factors resulting in a stasis or a breakdown of a system, (2) study relationships between elements of a system, and (3) describe and reveal how theses relations have an impact on the system as a whole.

Schismogenesis can be described and understood thanks to “differentiation mechanisms”. In his particular case, Bateson defines two cases of “differentiations” in order to characterize the relationships between two social groups:

The possibilities of differentiation of groups are by no means infinite, but fall clearly into two categories: (a) cases in which the relationship is clearly symmetrical [...]; and (b) cases in which the relationship is complementary [...]. [5]

Symmetrical Differentiation

This differentiation happens when social groups, having “[...] the same aspirations and the same behaviour patterns”[6], use an identical behaviour pattern in relations between members of different groups:

Thus members of group A exhibit behaviour patterns A,B,C in their dealings with each other, but adopt the patterns X,Y,Z in dealing with members of group B. Similarly, group B adopt the patterns A,B,C among themselves, but exhibit X,Y,Z in dealing with group A. Thus a position is set up in which the behaviour X,Y,Z is the standard reply to X,Y,Z. [7]

Complementary Differentiation

The second case is equally simple. It appears when social groups have ”behaviours and aspirations [that are] fundamentally different”[8], and use a complementary behaviour pattern in relations between members of different groups:

Thus members of group A treat each other with patterns L,M,N and exhibit the patterns O,P,Q in dealings with group B. In reply to O,P,Q, the members of group B exhibit the patterns U,V,W, but among themselves they adopt pattern R,S,T. Thus it comes about that O,P,Q is the reply to U,V,W, and vice versa. [9]

Reciprocal Differentation

Bateson added a third case of differentiation, in order to accommodate for the concept of homeostasis – with which a system stays stable through regulating mechanisms – called reciprocal differentiation.

In this type of behavior patterns X and Y are adopted by members of each group in their dealings with the other group, but instead of the symmetrical system whereby X is the reply to X and Y is the reply to Y, we find here that X is the reply to Y. Thus in every single instance the behavior is asymmetrical, but symmetry is regained over a large instances since sometimes group A exhibit Y and group B reply with X. [...] The reciprocal pattern, it may be noted, is compensated and balanced within itself and therefore does not tend toward schismogenesis. [10]

Musical Use of Schismogenesis

How pertinent is this concept of schismogenesis in a musical setting, and in particular in non-score based music? The technical and esthetical problems raised by electroacoustic music – heterogeneity of composing tools and musical styles, complexity of compositional strategies, instability and precariousness of technical solutions – are frequently questioned in the literature. All these factors have in common the same root: the impossibility of a well-defined representational system, due to the polymorphic, multi-level and multi-scale nature of electroacoustic music works. As a matter of fact, these works exist in a plurality of forms, from sound files to graphical sketches[11], and from processing techniques down to diffusion – this echoes some of Langer’s remarks about art in general and music in particular.

This is where schismogenesis can be of help: since its modus operandi does not take into account the nature of the constituting elements of a system, it can be used whatever the nature of the system being studied; only the global behaviour is important – and there are no preconceptions on what a musical “system” is. In a musical context, schismogenesis proves to be very flexible: for example, applied to a local level, it helps model the evolution of different sonic elements put in contact; in the same manner, from an analysis of a global movement, one can better understand the local behaviours of constituents. It proves also useful to put into contact symbols of different natures, viz. presentational or discursive symbols, if we are to borrow the terms from Langer.

Conditions of use

It is difficult to determine what criteria are necessary for a schismogenetic analysis to work. Firstly, schismogenetic analysis is not intended as an analytical method, but rather as a conceptual tool. Rather than proposing a “recipe” to follow in order to analyse a musical work (producing a reduction or a model – but we already talked about this!), schismogenetic analysis tries to relate the complexity, by showing how the different organizational levels interact in a complex system.

Nevertheless, the analyst should take care in respecting several points before undertaking a schismogenetic analysis:

(a) Existence of a dynamical evolution – given the temporal nature of music, it is almost unavoidable[12];
(b) Evidence of contrasting elements (corollary of (a) – how do you track changes?);
(c) Existence of several analytical levels (independent of their nature: temporal, formal, conceptual, technical).

If these conditions are fulfilled, the analyst should be able to designate the following elements:

(1) Elements or groups of elements involved in schismogenesis and communication channels between them;
(2)  Contextual characteristics of the schismogenesis[13];
(3)  Mechanisms or factors controlling the schismogenesis, being (at least) of three different types:

a. Loops of degenerative causality can be superimposed on schismogenesis, implying that when a certain intensity level is reached, some form of constraint is applied;
b. Other cumulative interactions, besides schismogenesis, may exist[14], and may act in an opposite direction;
c. Growth of the schismogenesis may be limited by environmental factors (internal or external), different from groups in contact in the schismogenetic loop.

Once these aspects have been determined, the analyst may use any technical method he wishes to prove the schismogenesis.

Methodology, from musical structures to musical beings

There is no “ready-made” method for schismogenetic analysis. Instead, we should keep in mind that such an analysis is merely of schismogenetic orientation, in that the principle of schismogenesis is simply used to show the differentiation processes at work in a given musical system. Therefore, a “classical” approach to electroacoustic music must primarily be used: for example, research of principal elements, then spectrographic, spectromorphological, or UST characterisation of these elements. Once these perceptual elements have been found, identified, and the structure analysed, it is then possible to proceed to the schismogenetic analysis proper, leading to a description of the global behaviour of the elements, and of the global evolution of the musical system.

However, such an approach forces us to consider the sound elements under investigation differently: since we are describing elements that exhibit behaviours, we are brought to consider that we are not dealing with immutable, structural and self- contained objects (which is the Western traditional concept of a “note”), but rather with (musical) beings[15]. That effectively changes some perspectives on the way analysis should be conducted.

These are the aspects on which we will be working on with Globalalia.

3.0 Analysing Globalalia

Trevor Wishart’s Globalalia lies at the centre of the issues raised by Langer’s writings on language and music: while this work uses human language, it focuses on syllables and not on words; while it extracts elements from narrative contexts, it seems to refuse to create a discursive story from these; while it explores biologically-produced sounds, it does so by transmigrating them to complex compounds. And it does so very carefully, so as to lie strictly at the boundary between these different categories; using Simon Emmerson’s language grid categories, it uses over its complete duration all the available cells, from abstract to abstracted syntax, and from aural to mimetic discourses.

Consequently, using well-known methods for electroacoustic music analysis may prove a thorough challenge and would require focusing on one particular aspect, ignoring the multiple resonances and echoes that are found (and used) at the different levels of the musical discourse. The sheer amount of sound elements used makes complicated, if not impossible, resorting to genetic or procedural approaches, that attempt to reconstruct the origins of each individual sound (tracing back to the 4000+ sound samples used by the composer[16]); on the other end of the spectrum, going for a structural or syntactic methodology, focusing on the higher-level gestures or forms with little or no interest for the underlying components would fail to take into account the surgical precision with which the numerous sonic complexes are organised (and possibly ignoring nuggets of semantic information in the process).

This is precisely where schismogenesis provides an interesting approach. A work of this duration provides the listener with a sense of evolution, and also with clues relating to the modifications and transformations of its constituting elements. Globalalia is a complex musical “society”, made up of many tiny musical beings that engage in communication and exchanges; it does however have a strong architectural backbone, a place musical beings inhabit, which we will be analysing first.

Architectural elements

Figure 1. Complete piece, Hanning (8192), overlap 25%.

For a work of this length, it is not uncommon to feel the need to “cut” at given points and (sometimes arbitrarily) define parts and sections. Fortunately enough, Trevor Wishart uses a very distinctive sound of a man screaming throughout the piece – the most salient element from a series. It is heard the first time at the very beginning of the work, and will act from this moment as a kind of signal, generally triggering changes in the musical discourse, as we will discuss in detail below. Using this element, among others, it is possible to divide Globalalia into seven sections, each with distinctive organizational characteristics. Each of these sections may then be subdivided into smaller parts, since they seem to be articulated around different sounds, or clearly make use of different processing treatments. By using a spectrogram[17] (see Figure 1), it is possible to obtain a graphical representation – to act as a visual presentational symbol – on which to rely, possibly helping with the definition of the different segments; likewise, for elements resembling pitched melodies, an approximate Western notation score can be provided.

Section 1 – from 0’00'' to 5’44''

Figure 2. Section 1, FFT Kaiser 8192, overlap 50%.

“Theme” and introductive exposition (0’00'' to 0’35'')
This section starts by an exposition of approximately 35 seconds, during which several highly distinctive elements are put into front; they (for the most part) will be developed throughout the piece.

  • The first element presented is the “scream” (male voice), along with two other sounds;
  • Another distinct element is the use of the German word[18] “Traum” (0’06'');
  • Yet another is the use of the interjection “wow” (0’24'');
  • Finally a syllable “mow” is distinctly heard and repeated (0’30''-0’33'').

Interspersed between those fragments are several bridges created by accumulation of specific syllables or deconstructing syllables into consonants and vowels. This “exposition” permits to set the stage for several evolutions, differentiations and/or oppositions that will be explored during the piece.

An important aspect of the very beginning of the piece (the scream itself, plus the two following elements) is its pitch structuring. When analysing this small segment, it is clear there is an importance given to pitch structure, even if it seems secondary to the more prominent features of the voice. In fact, over these mere two seconds, we are evolving from clearly pitched voice to a more noisy “duck-like” sound. The pitch profiles are given in Figure 3, showing an approximation in standard notation that includes the most important partials heard. These characteristics are particularly meaningful: the scream (an inarticulate vocalisation) related to pitch, whereas consonant (an extracted element of an articulated language) pertains to noise: when put in contact, they produce a third element, combining the two aspects – effectively exhibiting a symmetrical differentiation.

Figure 3. Approximate notation of the initial motive.

After having exposed some kind of pitch structuration, rhythmical aspects are featured prominently in the next sections of the introduction. At first, the duration chosen for sonic elements from 0’04'' to 0’11'' is enough for some words to be understood – or thought to be understood, with elements around 0.20s in size. This allows Trevor Wishart to hint at a “dream”, using the German word Traum distinctively. The next elements (0’11'' to 0’18'') begins with a modification of the initial scream – this time centred around a distinctive B to Bb pitch profile, that directly leads to a rhythmical section using sonic grains of 0.06s to 0.07s in duration. This hints at a complementary differentiation: when putting the scream into contact with rhythmical (quasi- ) loops, it triggers a shortening of the constituting elements of loop group, as well as a pitch-modification of the scream.

The next subsection (0’18'' to 0’24'') confirms this particular aspect of the schismogenesis, by introducing new aspects of dealing with time: repetition and accelerando, creating a more continuous sound, but made up by accumulation of very short discrete elements at first (0.02s) to longer, identifiable syllables (0.13s), creating a chiasmus between length of individual elements and perception of speed. This is reinforced by the use of a distinctive (female) “wow” sound, clearly pitched, echoing the (male) scream – again one can hear clear pitches in this element, attacking from C then followed by a longer, prominent C#, finally descending down to G# or A (since there’s a lot of breath noise at the end). The final part of this small exposition (0’27'' to 0’34'') consists of a juxtaposition of different layers: (a) an initial background layer made of small repeated elements generating a rhythmical texture, (b) transpositions of vowel and diphthongs sounds, creating a melody, and (c) conclusive “meow” sounds that use descending two-tone pitch profile found both in the “scream” and the “wow” sounds, with breathy (noise) features. These aspects are summarized in Table 1.

0’00''

0’04''

0’11''

0’18''

0’27''

Scream

“Traum”

Rhythms

Repetition

Juxtaposition

Pitch à noise

Short, but “understandable” elements

Shorter, percussive grains

Chiasmus:
Short to long grains, accelerating

(a) Repetitive rhythmical background (b) Melody
(c) Noise

Table 1. Salient aspects of initial exposition.

Already there are schismogenetical trends operating at different levels: symmetrical differentiation between pitched and noisy (unpitched) elements, complimentary differentiation between rhythmical (looped) and continuous sounds[19], and there are reciprocal differentiation operating between sounds of “male” or “female” origin.

Looking at the spectrogram, it is clear that this section is afterwards structured around different episodes, all confirmed by listening[20]:

  • From 0’35'' to 1’58'': accumulation of syllables centred on /t/ consonants;
  • From 1’58'' to 3’30'': change of focus to /m/ consonants, slowly transforming into an harmonic texture;
  • From 3’30'' to 5’44'': rhythmical loops using /n/, /k/ and /kl/consonants creating harmonic layers, then filtered.

Subsection 1a (0’35'' to 1’58'')

There is a pattern taking place that echoes the structuration found in the introduction: from 0’35'' to 0’46'', an accumulation of discernible syllables creates a synthetic Sprechgesang-like “melody”, cut by the scream, which triggers a loop made-up of shorter sonic elements (0’47'' to 1’03''), confirming the diagnosis of a complimentary differentiation taking place. This device is then used at different speeds, creating contrasting moments between pitched and noisy sections, different levels of time, and sonic qualities (using effects to enhance reverberation quality, time, or filter resonances) throughout this first part.

Subsection 1b (1’58'' to 3’30'')

Time-stretching is introduced and used to heavily process /m/ consonants sounds, creating an exploratory texture, from 1’58'' to 2’37''. A short episode, using granular structuration then follows (2’37'' to 2’49''), and finally a combination of the two approaches is made (2’49'' to 3’30''), proposing a rich harmonic canvas onto which granular structures are juxtaposed. Here we are confronted by a new aspect of the schismogenesis at play in Globalalia, exhibiting symmetrical differentiation between granular and harmonic elements – in fact by evolving together in a common “place”, musical beings of these morphologies exhibit changes in the same manner, reinforced by filtering to enhance their harmonic qualities.

Subsection 1c (3’30'' to 5’44'')

Returning to more clearly identifiable recorded syllables triggers the use of rhythmical loops, with two different episodes (from 3’30'' to 4’00'', then to 4’35''). Finally a long crescendo is built (from 4’35'' to 5’44''), using accumulation of rhythmical loops and processing (principally filtering and time-stretching) – these accumulations are typical of a system that is not regulated, building up on cumulative behaviours or complementary differentiation. In this case, these accumulations are made of mid- sized sonic elements, that are repeated and repitched; note that in order to prevent the breakdown of the system due to this scishmogenesis, filtering effects are used.

Section 2 – from 5’44'' to 10’28''

Figure 4. Section 2, FFT Kaiser 8192, overlap 50%.

Likewise, it is possible to find several subsections:

  • From 5’44'' to 6’16'', there is a sort of reexposition, using the same elements found at the beginning of the piece (scream, “Traum”, “wow”, “mow” syllable, notably);
  • From 6’16'' to 8’30'', there is a development centred around a pitched motive, and consonant /r/ sounds;
  • From 8’32'' to 9’18'', there is a contrasting section, using primarily consonant /s/ and /z/ sounds;
  • From 9’18'' to 10’28'', a very rhythmical moment takes place, using sounds extracted from /v/, /t/, /l/, and /m/ consonant and syllables.

Reexposition (5’44'' to 6’16'')

This section starts by using the scream, using time-stretching in order to smooth the sound and stress the differences with the next elements introduced at 6’16'' (based around /r/ sounds and syllables such as /bree/ /bra/ /bro/). Complete syllables (consonant and vowel), put into contact with the scream, seem to make them evolve towards repetitive behaviour, and more clearly towards pitch, as the next subsection exemplifies. This particular interaction is typical of a symmetrical differentiation: while the scream is being modified by time-stretching – hence losing its well-defined pitch – the syllables, which are mostly noisy (prevalence of consonant sounds) are made shorter (incidentally, even more when put into contact a second time with a scream at 5’56'') and more clearly pitched.

Subsection 2a (6’16'' to 8’32'')

At 6’53'' a melodic motive is clearly put into front (however, it can be heard in the background since 6’20''); progressively this arpeggio-like motive is transformed and strongly smoothed, to make it sound like a synthetic (quasi-brass) instrument. Pitches are clearly defined (see Figure 5) and this introduces another layer in the musical discourse: rhythmical melodic loop, an effect of the symmetrical differentiation taking place – repetitive elements inducing smoothing in the timbre qualities, making the sound evolves from speech to quasi-instrument.

Figure 5. Approximate transcription of the arpeggio-like motive at 6'53''.

Gradual transformation using effects on this motive over the course of subsection are then cut short by a sudden reexposition of the raw motive using the initial voice (at 7’58''): this is a brutal “return to reality” that leads the musical discourse to another development. Effectively, this schismogenesis results in a collapse of the musical system at this point: speech-like sounds have evolved too far away from their initial “musical social group”.

Subsection 2b (8’32'' to 9’18'')

This is a short development around short elements (around 0.13s in duration) using sibilant consonants that are successively shortened to half duration (0.05s) then elongated to double duration (0.3s) and finally arranged in combination. This subsection concludes by a quasi-melodic and rhythmical descending motive, a result of the contact between short elements, consonants, and repetition: this is clearly the effect of a complementary differentiation.

Subsection 2c (9’18'' to 10’28'')

Rhythmical loops are then put into front, putting once again repetition as an essential procedure. Gradually adding other loops at different speed (there are four different superposed loops at 9’32''), much like procedural pieces, there is a phase moment at around 9’48'', possibly generating a word. The loops then gradually dissolves into the breathing aspects of consonants, effectively degenerating into noise. Again we are facing another aspect of the schismogenesis: a symmetrical differentiation that implies loops exhibit phase and antiphase patterns – seemingly leading to understanding, but ultimately degenerating to noise.

Section 3 – from 10’28'' to 15’08''

Figure 6. Section 3, FFT Kaiser 8192, overlap 50%.

As usual, the scream signals the beginning of a new section. At this point, the use of the scream is symptomatic of a reciprocal differentiation at work: each time the musical system is about to collapse (in loudness, for section 1, and in noise, for section 2), this signal is used, preventing a complete breakdown; for these reasons, some may term it “primal”... Once again, there is a segmentation in subsections that follows the use of various syllables and consonants:

  • From 10’28'' to 10’36'', there is a shortened reexposition;
  • From 10’36'' to 11’50'', discourse is based on /p/ consonants;
  • From 11’50'' to 13’25'', /st/ and /f/ consonants are used;
  • From 13’25'' to 15’08'', /dr/ and /tr/ consonants are the primary sounds. In this context, the quasi-word “Traum” is exposed again (15’).

Reexposition

This time, elements constitutive of the initial exposition are used but shortened and compressed in time – obviously the “scream” is exposed, but also some of the rhythmical loops, differently arranged. The “wow” sound is then subjected to an interesting transformation, by using only the extracted noise (breath) content (10’33''), effectively inverting the usual perspective on the initial scream/syllable schismogenesis. In this case, elements extracted from languages are used for their pitch qualities, while more incidental sounds such as screams or onomatopoeias are retained for their noisy aspects.

Subsection 3a (10’36'' to 11’50'')

The percussive aspects of /p/-based syllables induces several juxtaposed rhythmical loops at different time scales; from 10’36'' to 11’02'', the sounds are between 0.1 to 0.3s in length, implying an “easy” recognition of the syllables used, allowing pitch recognition as well. A small bridge, from 11’02'' to 11’05'', that incidentally exposes in element by element the different syllable that structures this section (albeit in a very fast, collage way), is used to introduce another development using much shorter sound grains (around 0.04s duration); using shorter grains implies more difficulties in identifying pitches, while retaining the qualities of the “p” consonant. Two (11’12'') then three (11’16'') layers of this loop are used, and the whole development used twice (11’19'' to 11’28'') before leading into a section that uses the same principle of rhythmical loops, but with sonic elements of a variety of durations and heavily processed – most notably by resonance filtering and pitch-shifting; this ultimately creates pitched percussive sounds, some reminiscent of gamelan. Evidently, the inherent percussive qualities of the /p/ consonants are triggering a rhythmical development, however since the basic materials used by Wishart in this section are syllables, they somehow retain some of their pitched aspects (formants) – another complimentary differentiation. Filtering, repetition and superposition are essentially external and artificial processes: in this case, they may act metaphorically as a symbol of changes in communication in the digital age: even without content (in our case, without pitch) and reduced almost to noise (the noisy components of the percussive “p” sound), technology can produce an illusion of meaning.

Subsection 3b (11’50'' to 13’25'')

Derived from the last sounds of the preceding section, the discourse now focuses on /st/ consonant syllables, from the start exposed with filtering effects applied. At 12’03'' one can infer the word “stop” being exposed twice. At 12’29'', there are two unmistakable bass notes produced by heavy processed and lowered sounds (see Figure 7), followed by a quasi-word evoking “skip”. Working as a signal, this leads into a drone-like moment, created by adding layers of processed sounds: (a) a very fast repetition of a percussive sound, creating a muted pluck-like line (around the frequency of the second note of the bass motive), (b) hissing sounds extracted from the noise features of consonants, (c) an harmonic texture created by time-stretching and filtering, using harmonic series from the two bass notes. The “skip” word is then clearly heard (13’23'') and closes this sequence. This continues and amplifies the schismogenesis initiated in subsection 3a, leading to a collapse of the system: interestingly this is the symptom of a complimentary differentiation: at extreme points, syllables creating semantic meaning (words), while consonants (subsyllables, in a sense) generates notes, and even distinctive intervals; by intensively relying on technology for communication, we are gradually moving away from language, creating “something else”.

Figure 7. Bass notes at 11'29''.

Subsection 3c (13’25'' to 15’08'')

The principle of accumulation is used again, creating a very dense texture that slowly evolves – the fundamental frequencies are those of the bass motive. It fades out after a clear exposition of “Traum” (14’58'') and a derivation of the scream. Evidently “Traum” is featured at the centre of the piece, and together with the scream, could beg for a psychoanalytical understanding – which we’ll skip for now[21]. This median subsection subtly makes use of the different differentiation processes we described until now: repetition and single events, accumulation to textures, pitched and noisy sound aspects, brief and long elements – evidently a manifestation of a musical society, demonstrating its diversity through the use of technology. In this case, however, technology is not presented negatively, but rather as a mean to make a “dream” emerge.

Section 4 – from 15’08'' to 19’22''

Figure 8. Section 4, FFT Kaiser 8192, overlap 50%.

Section 4 begins (again!) by a reexposition; this triggers the usual change of focus in the syllables and consonants:

  • From 15’08'' to 15’30'': reexposition;
  • From 15’30'' to 17’30'': /tsch/ consonants and /me/ /mi/ syllables are prominently used;
  • From 17’30'' to 18’44'': there is a climatic change, using heavily processed sounds, and shifting focus more on vowel sounds rather than consonants;
  • From 18’24'' to 19’22'': gradual transformation into synthetized organ-like impacts.

Reexposition

This time, the reexposition seems to be at the “normal” pace, with an initial interest on the “Traum” word, then gradually focusing at the end on the “mow” and “wow” sounds.

Subsection 4a (15’30'' to 17’30'')

During the two minutes of this subsection, a texture is created in a similar way as the one in subsection 3c; the repetitive layer consists of grains of 0.07s put 0.03s apart (a sort of synchronous granular synthesis), with variation in pitch – it evolves from an initial /tsch/ sound exposition, following a clear symmetrical differentiation (one aspect contaminating the other). 

Other elements are introduced by a brief exposition, then a time-stretching/pitch-shifting is applied and grain fragments of varying length (very short (0.09s) at the beginning (16’00''), much longer (0.275s) towards the end of this passage (17’00'')) are repeated. This whole subsection is a clear example of a reciprocal differentiation: there are antagonistic propositions (fragments of different lengths, conflicting pitches) but the complete system is homeostatic: there is no collapse in the musical discourse. Interestingly contrasting elements (such as instrument-like sounds – 16’56'') may be regarded as environmental, and this is what triggers changes in the musical structures.

Subsection 4b (17’30'' to 18’44'')

This subsection is a constructed soundscape akin to being immersed in a place with many conversations going on: effectively, this is another example of accumulation, using longer, distinctive syllables, but with a lot of layers. Towards 18’20'', a low frequency voice can be heard in the background, and afterwards (18’30'') the arpeggio motive is heard once again. This hints at a more complex schismogenesis at an higher structural level: between this subsection and the preceding one, there is a complementary differentiation taking place: temporal elements are opposed (short/long), pitch characteristics as well (moving/steady).

Subsection 4c (18’44'' to 19’22'')

Unsurprisingly, this short moment develops around the idea of the chiasmus section of the introduction, using the same loop, but gradually inducing resonances by filtering (19’20''). This clearly shows how much the musical beings have evolved since their initial exposition, focusing on repetition and pitched aspects, all filtered trough the veil of a “dream” – completing the schismogenesis described in the previous two subsections.

Section 5 – from 19’22'' to 22’00''

Figure 9. Section 5, FFT Kaiser 8192, overlap 50%.

This section starts with an exposition using shortened sounds (notably the scream). It then develops into these subsections:

  • From 19’22'' to 19’38'': reexposition, ending with an percussive sound;
  • From 19’38'' to 20’53'': development using sounds from the exposition, notably the “wow” sound, and the (heavily- modified and shortened) scream;
  • From 20’53'' to 22’00'': consonants sounds are treated to make percussive, electric organ-like background textures, on which various elements are superimposed (scream, filtered sounds).

Reexposition

This reexposition is particular, in that the scream sound and “wow” sounds are only heard late and processed this time.

Subsection 5a (19’38'' to 20’53'')

The signal of this section is a distinctive “Allons-y !” (“Let’s go!”) French expression suggesting the reexposition is over and another episode is opening. This expression is repeated several times in this subsection, along with transformations of the “wow” and scream sounds, and gives the basis to the crescendo (from 19’45'' to 20’15'') that is exposed, along with a cyclic bass motive; this procedure is repeated a second time, until 20’53''. The use of semantic elements generally triggers an accumulation process, symptomatic of a complementary differentiation, which is the case here.

Subsection 5b (20’53'' to 22’00'')

This subsection follows another crescendo allure (complementary differentiation): using heavily processed consonant sounds; it generates a texture very close to sounds created by a percussive electric organ. Over this layer the scream is prominently featured, approximately every ten seconds (21’05'', 21’19'', 21’29'', 21’37'', 21’39'', 21’49''). The predominance of pitch- gliding sound structures suggests that only non-language elements (or heavily processed sounds – through the use of technology) carry musical information[22].

Section 6 – from 22’00'' to 26’45'' 

Figure 10. Section 6, FFT Kaiser 8192, overlap 50%.

  • From 22’00'' to 22’24'': Reexposition, “Traum” is distinctly heard;
  • From 22’24'' to 23’35'': focus on /tsch/ then /z/ consonants;
  • From 23’35'' to 25’20'': plays on /k/ consonants and /kr/ derived syllables, leading to a very melodic exposition;
  • From 25’20'' to 26’45'': after a brief introduction on /tr/ consonants, filtered sounds create a rich harmonic textured background.

Reexposition

Starting in a classical manner, this reexposition gradually heads somewhere else. Just after “Traum” is heard (22’04''), all sounds are heavily reverberated, creating a texture made of the different resonances. The scream is heavily processed through pitch-shifting at 22’20'', and this signals the main development of this part.

Subsection 6a (22’24'' to 23’35'')

This moment is a pretty “classical” (at this time in the piece) development on /tsch/ consonants, by repetition of different motives. This is clearly a symmetrical development, in that the system stays stable, and there are similar behaviours from the consonants and motives.

Subsection 6b (23’35'' to 25’20'')

Using the interesting characteristics of /k/ and derived consonants, there is a strong development taking place, by slow accumulation of different layers of distinctive sonic elements. A melodic motive (see Figure 11) emerges (24’14''), after having been exposed without accentuating the pitched elements (24’00''): used as a structural pivot in this development section, it provides a base note (G#), used as a bourdon. The other layers consist of a brilliant female voice sounds (using syllables such as /ke/, /ko/, /ka/) put in loops and repeated, and a friction sounds (that fast alternates between left and right channels).

Figure 11. Approximation of the melodic motive at 24'14''.

This very interesting moment is superposing numerous distinctive differentiations: female/male sounds, pitched/unpitched syllables and consonants, short grains and long textures/long grain and short motives...

Subsection 6c (25’20'' to 26’45'')

This final section of part 6 consists of large compound of sounds, treated as grains (around 2.2s in lengths) that alternate between channels, hence following the latest development of the previous section, but extending it. At around 25’50'', filter banks are used to process this stream of sounds – concluding the development by adding a final touch on the important schismogenesis between “natural” and “cultural” (modified) sounds.

Section 7 – from 26’45'' to 29’ (coda)

Figure 12. Section 7, FFT Kaiser 8192, overlap 75%.

This last section forms in a way a recapitulation of the different episodes developed during the work: in this respect, it can be thought of as a coda, or the result of the different contacts between these musical beings. 

  • From 26’45'' to 27’00'', the original sounds are reexposed once again, but processed through heavy reverberation.
  • From 27’00'' to 27’18'', elements are exposed, one by one, without apparent connections – an organ-like melodic motive is heard in the background.
  • From 27’18'' to 27’30'', elements are exposed similarly, but effects are added (resonance, repetition).
  • From 27’30'' to 27’50'', elements are organised in loops, and the organ motive is more present.
  • From 27’50'' to 28’16'', a development is made around repetition of “meow” sounds with resonances.
  • From 28’16'' to 28’46'', heavily processed sounds are used (through time-stretching and pitch-shifting).
  • Finally, the last moment of the piece 28’46'' to 29’00'' exposes one last time the initial scream motive (at low volume) and several echoes, creating a fade out.

Global form and statistical aspects

Using a chart, it is possible to visualise the general structure of the piece, hence allowing us to directly infer the relative lengths of each section (see Figure 13).

Figure 13. Globalalia structure.

Another interesting statistical element one can determine is the relative “weight” of the different sounds used in the work. Evidently, as statistical data, this kind of approach leaves behind a number of detailed aspects that are pertinent for analysis. By labelling the subsections with the most prominent sounds used, we obtain the graphical pie shown in Figure 14, which clearly shows the importance of exposition elements and /r/ consonants. Interestingly, vowel sounds, “understandable” words and expressions, and other “incidental” sounds (like “wow” and “mow” developments) are featured in a median proportion.

Figure 14. Statistical repartition of sounds.

4. Globalalia: a musical society

The obvious pun – Globalalia that alludes clearly to “glossolalia” – captures in a single word several of the schismogenetic mechanisms that are at play in this work. “Glossolalia” comes from the Greek words γλωσσα – tongue, language – and λαλειν – to speak, and as a strong religious meaning, referring to a mystical crisis during which unknown languages (or foreign languages apparently unknown to the speaker[23]) are spoken; medically, it corresponds to a symptom of mental illness. Interestingly enough, the notion of language is strongly present in this case – however with little reference (obviously) to a particular structuration, not to mention grammar; it seems that glossolalia is clearly located in between the presentational and discursive symbols Langer described. In Globalalia, however, there is no language – or is there? The methodology used by Wishart in this case, gathering digital audio samples from broadcasts all over the world and dissecting them into syllables, suggests an approach that is very similar in classic linguistic methodology: where a linguist seeks lexemes and morphemes – respectively basic meaning and grammatical units, the composer isolates what I suggest calling sonemes – the smallest possible unit of sound to convey a sense of direction, that is, a unique aural identification. The key difference, however, is that in our case these sonemes can evolve and change, not unlike a living organism: hence the application with which the composer tries to draw connections between similar sonemes – the subsections in Globalalia, focusing almost exclusively on one soneme, are a prime example. To a certain extent, Wishart’s work on Globalalia is an attempt to provide a conceptual playground in which sonic entities are put into contact and evolve freely – a musical sandbox for exploring the qualities of human vocal sounds and languages. 

In the case of Globalalia, these differences lead to tensions and to schismogenetical trends, the most important conceptual ones being:

(a) Difference and repetition: much like Deleuze, Wishart’s efforts to going back to the soneme (an element that precedes lexemes or morphemes) demonstrates that difference is essential and not the opposite of similarity – and repetition is presented as an organic quality, and not a process;
(b) Pitch and noise: there is clearly a continuum between these aspects – however Wishart blurs the line by inverting the symbols usually associated to these characteristics, associating culture with noise (the breathy components of the consonants) and nature with pitch (the pitched quality of vowels, screams, and interjections);
(c) Technology and organicity[24]: the different effects applied to the sonemes are not trying to impose constraints or external modifications – rather they are a logical evolution of their organic qualities, made possible by the digital age.

By creating a global language, diverting the very basic elements – syllables, consonants, vowels, and incidental sounds, all in effect sonemes – from the noise generated by our global “culture” and “technology use”, Globalalia and its sonic Esperanto reach deeply to our understanding of human nature and cultures, through its careful exposition and examination of a musical society: as such, it is an interesting point of departure in the study of music analysis through the use of musical beings.

How could music analysis benefit from considering works as society and music elements as musical beings? We posit that many aspects of the musical discourse may be best viewed in this way: classical devices of musical developments (augmentation, diminution) may be understood as either an effect of a given ecosystem or by contacts between musical beings (which we can analyse – up to a certain extent – using tools such as schismogenesis); mutations of aural elements (electroacoustic, but also including modes de jeux in performance for example) are understandable as pertaining to the musical beings’ “DNA” (hence allowing a more intimate connection and explanation of musical intentions for example in pedagogical settings). This take on analysis creates a bridge over the usual gap between the technical, symbolic and structural aspects of music analysis[25]: and by doing so, reinvests time and dynamic evolution in its framework – two essential components of any music.

[1] http://www.orema.dmu.ac.uk
[2] As can be seen particularly in the famous analysis of Debussy’s Feux d’Artifice in Lewin, 1993.
[3] Which are hardly functional, unlike Wittgenstein’s closely related take on language.
[4] Which can be found in the collection of essays Steps to an Ecology of Mind.
[5] [Bateson, 1972 ; p.67]
[6] [Bateson, 1972 ; p.68]
[7] [Bateson, 1972 ; p.68]
[8] [Bateson, 1972 ; p.68]
[9] [Bateson, 1972 ; p.68]
[10] [Bateson, 1972 ; p.69]
[11] In the case of Globalalia, the archives made available on the York University website, found at https://dlib.york.ac.uk/yodl/app/collection/browse? parent=york:809081#tree=%7CTrevor%20Wishart%20archive demonstrate these different possible forms, and the amount of data it represents.
[12] At one level (e.g. structural) or another (e.g. symbolic), evolution exists...
[13] i.e. topical conditions for the observation of schismogenesis.
[14] Another schismogenesis may exist, on another level, but cause local disturbances on the schismogenesis under investigation.
[15] In certain contexts, with the right sources, this conception permits to include external elements (such as program notes, code, “genealogic trees” of transformations – including “nonviable” branches – and so on) into the analysis – recalling aspects of Battier’s application of the Constructivist concept of faktura to electroacoustic music analysis.
[16] Some details on these samples are found on the archive mentioned above.
[17] All images were generated by the Raven Pro software, initially developed for bioacoustic signal analysis: http://www.birds.cornell.edu/brp/raven/RavenOverview.html
[18] Or maybe a quasi-word...
[19] Opposition between smooth time (“temps lisse”) et striated time (“temps strié”), per Boulez definitions.
[20] Timestamps are approximate, given the continuous nature of the discourse.
[21] “Are musical beings dreaming of electroacoustic effects?”
[22] Going forward, it may also suggest that it is through culture (defined as the use of technological tools) that we are able to create meaning – and it seems to be a bold statement... [23] The correct term in this case would be xenoglossia.
[24] One could also see in this one the dilemma of the electroacoustic composer and its sound objects – to alter or not to alter?
[25] The classical semiotic tripartition : esthesic, poietic and neutral levels. 

 

Bibliography: 

BATESON, G. (1972) Steps to an Ecology of Mind. Chicago: University of Chicago Press.
BATTIER, M. (2003) A Constructivist approach to the analysis of electronic music and audio art – between instruments and faktura. Organised Sound, 8 (3), pp. 249-255.
BOULEZ, P. (1963) Penser la musique aujourd’hui. Paris: Gauthier.
DELEUZE, G. (1968) Différence et Répétition. Paris: Presses Universitaires de France.
EMMERSON, S. (1986) The Language of Electroacoustic Music. London: MacMillan.
LANGER, S. (1996) Philosophy in a New Key: A Study in the Symbolism of Reason, Rite and Art. 3rd ed. Cambridge: Harvard University Press.
LEWIN, D. (1993) Musical Form and Transformation: Four Analytical Essays. New Haven: Yale University Press.
WISHART, T. (1994) Audible Design. London: Orpheus the Pantomime.
WISHART, T. (1996) On Sonic Art. London: Routledge.
WITTGENSTEIN, L. (1922) Tractatus logico-philosophicus. New York: Harcourt, Brass & Company.