Pluralising the Musical Text: Improvising on Canonic Repertoire

JONATHAN AYERST

Zürcher Hochschule der Künste

This chapter grew out of a long-term autoethnographic study in which I learnt to improvise on various Baroque models over a period of six years (Ayerst 2021). During this time, I noticed significant qualitative changes in my musical knowledge and perspectives as I changed from an interpretive musician to an improvising one—changes that were particularly noticeable as I began to use canonic works as models to guide and inspire my own creativity. Initially even the idea of improvising on a work of Johann Sebastian Bach seemed impossible: How could I change the notes and still stay in touch with the original? Which notes should I change? How could I make creative decisions like Bach? What were the criteria for my choices? What worked well and what did not? Tackling these questions (which functioned as barriers to my engagement in improvising) pushed me to critically examine my habitual beliefs about musical practice, to more objectively understand the culture of classical music within which improvisation has typically been marginalised as the rare preserve of a few individuals, and to confront the seeming incompatibility of my own cognitive processing with the task. The answers I discovered are addressed throughout this chapter as I describe the development of an “improvisational knowledge” that is based on the perception and abstraction of musical features and underlying functions, used conceptually to guide one’s improvisation within the confines of a particular historical style.

Understanding the Werktreue approach to the musical text

Clearly, through long training as an interpretive performer, I had acquired a knowledge of music that made even the idea of improvisation very difficult. I first encountered the expression Werktreue in Lydia Goehr’s seminal The Imaginary Museum of Musical Works (2007). In this work I seemed to read the story of my own thinking as a musician, particularly in the detailed historical analysis of the aesthetic and sociocultural changes affecting musical practice through the late eighteenth and nineteenth centuries, changes that resulted in attitudes of reverence towards musical scores as perfected works and in a division of labour between a composer/creator and a performer/ interpreter:

An important aspect of Werktreue reverence and subservience to the score was the belief that the score itself contained a transcendent spiritual message—“meaning-like qualities that survive substantially intact, over time, inherent in the relations between their notes” (Leech-Wilkinson 2012, paragraph 1.5). Thus, in a Werktreue doctrine, the role of the interpretive performer becomes that of communicating the encoded message through an exact replication of the notes of the score—the transmutation of “silent, symbolically coded prescriptions” into sounding matter, as the introduction to this volume puts it. Naturally this caused a huge problem for my improvising because, according to this belief, if I changed any of the notes by improvising a new version, the meaning of the music would be lost or corrupted in some way. Worse still was the sense of wrong-doing that accompanied my initial attempts—the idea that I was doing something embarrassing or perverse, outside what was accepted by my musical colleagues and friends.

The culture of Werktreue in classical music

Although I could situate my own musical outlook in the Werktreue approach described above, could I also describe other performers in the same way? Is it viable to talk of a typical “interpreter’s attitude” or a universal “improviser’s knowledge”? To do so would mean looking beyond the considerable variation at an individual level to the cultural practices that create shared perspectives, values, and beliefs—sociocultural knowledge that “allows us to meaningfully act, interact and communicate with other members of the same culture” (Dijk 2008, 222). This knowledge, obtained through immersion in a particular culture and social group, clearly results in ontological beliefs about what music is—that is, the text or score—and consequently what one “does” with it (the text or score) as a musician. Which is to say that much of what we learn about music is culturally situated: what we learn and how we learn it regulates and shapes our behaviour as a musician. ¹ Nor, according to the anthropologist Clifford Geertz, should we underestimate the role of cultural knowledge in shaping individual perceptions of the world:

Therefore, if we want to understand the construction of general attitudes towards improvisation among classical musicians—for example, attitudes of non-identification, suspicion, and marginalisation (Sarath 2015; Biasutti 2017); attitudes of prejudice, with the assumption that improvised music is of lower quality than composed music (Hill 2017); attitudes of fear, embarrassment, and anxiety when asked to improvise (Dolan 2005; Woosley 2012)—then we have to critically examine Western classical music as a set of cultural practices identifiable through a highly distinctive set of norms, values, and beliefs. As Rabinow suggests:

Thus, for musicians like myself, born as it were into a pre-existing cultural milieu in which one struggles to find one’s creative voice and the mental freedom to improvise, I think it necessary to distance oneself from the practice until the moment it appears distinct and idiosyncratic: as one way of making music among a multitude of possibilities presented by global cultures. At this moment one begins to notice how, in Western classical music, highly specialised mental models and representations of music are continuously recreated and reinforced, revealed particularly through the words people use to describe themselves as musicians and their attitude and approach towards music—the “discourse strategies that typically influence socially shared beliefs” (Dijk 2008, 223).

With space limited, one example will suffice. When the pianist András Schiff, in a 2013 interview with Jeff Spurgeon on WQXR, New York’s classical music radio station, was asked about his experience in performing the music of J. S. Bach, he offered the following comments:

To begin with, Schiff’s words seem quite innocuous: few would question that it is, after all, good to be responsible in one’s work, to be humble in one’s practice, or to respect another’s work. Yet it is worth examining the unspoken implications: first, that modern musicians are, by default, interpreters, separate and distinct from a historical practice of creativity; second, that in this day and age, an interpreter is a lesser mortal, a servant of the composer. Why? Because composers possessed innate gifts and spiritual insights presumably denied to modern musicians—“we have to know our place.” Following this is the emphasis on every note in the score—“you never play a note without this divine connection” (my italics)—a concept that is closely associated with integrity and responsibility—“if you don’t feel that, you don’t play. You cannot betray that.” Thus Schiff asserts that the performer has a sacred obligation to honour the written score; and that, by adopting an attitude of extreme humility, even religious fervour, one seeks spiritual communion with the composer—“I must be on his spiritual wavelength.” In this way the behaviour and attitude of the (ideal) performer is reinforced within extremely narrow limits by an authoritative source within the culture. ² What if a different attitude were adopted? One, for example, in which the individual does not feel that every note is divinely ordained? Such an attitude would constitute a “betrayal,” an act habitually associated with the basest human motives and one invoking the severest punishments. In this way Schiff (and others) reproduce a “noble struggle” marked by “guilt and insufficiency” (Botstein 2001, 591) between the contemporary performer and the historical text: composers and their music represent “other” beings, removed from normal everyday humanity by virtue of their gifts, a race of Gods and heroes to be approached only with excessive humility, caution, and a correct reverence.

Needless to say, such an attitude makes improvisation an impossibility. Nor, I believe, is it possible for improvisation to occur within the institutions and socio-cultural circles founded on Werktreue beliefs. To improvise, one must first fundamentally challenge such views, by understanding them as modern constructions that are grounded in implicit ideologies.

In the following section, I describe a different approach: one in which a historical figure such as J. S. Bach, however “gigantic” the cultural status and musical achievements, is recreated as a more human, “everyday” figure. This reconstruction, which bypasses the aesthetics of intervening eras, seeks to restore Baroque practices of modelling and copying from others’ works, treating the composition as an exemplar for one’s own music. But it also draws on a different philosophical approach from that of Werktreue: one that places the contemporary practising musician (be it performer, composer, musicologist, etc.) at the centre, as a creative agent able to take possession of the music as a language to manipulate at will. This is a far step from “knowing one’s place”; yet it is also, I would propose, far from the arrogant or presumptive position in which one places oneself on an equal footing with historical figures of genius. Perhaps it is a change of emphasis better explained through the ideas of the social philosopher Mikhail Bakhtin, who was anxious to portray individuals engaged in everyday creative acts, fully accepting themselves as positioned at a unique moment in time, rather than acting through the timeless abstraction of roles—genius, performer, improviser, and so on. As Courtney Bender reflects: “Bakhtin finds both aesthetic and theoretical thinking problematic precisely because they abstract what they imagine to be ‘important’ from actual events located within real time and space” (1998, 187). In this way the emphasis of musical creativity is shifted from theoretical abstractions, such as the imaginary legacy of great composers and their canonic works, to a more everyday dialogue between individuals and their contemporary social situations. The improvised music comes from within the individual—in this case, the improviser—constructed through participation with real (or imaginary) listeners in lived time. Thus, as an improviser, one learns to overcome the sense of moral fallibility raised through Werktreue’s enthralment to history by focusing more on the reality of the moment:

In this way, through constructing our music in the moment-to-moment act of creatively improvising I believe it is possible to reconstruct an approach that is historically situated, rather than historically dominated, and thereby to glimpse something of the original composer’s perspective of the task.

Describing an improviser’s knowledge and mental representation of musical elements

When I first started learning to improvise, my friend the composer and improviser Wolfgang Mitterer advised me simply to “not turn the page”; and it is true that taking a sonata and not turning the first page forces one to confront exactly the limits of one’s knowledge about a particular composer. However, at the same time, Mitterer’s advice is too radical to be of practical use, for, by not turning the page, one confronts all aspects of musical creativity and construction in one go! Discussions of improvisation treatises from the eighteenth century (Berkowitz 2010; Callahan 2010; Sanguinetti 2007)—an era of widespread improvisational practice—demonstrate that a great deal of stepwise, preparatory work in the form of patterns, formulae, and rules were inculcated in the novice improviser before whole forms were attempted. It is these formulae and patterns that underlie the improviser’s skill and give a clue to the cognitive processes involved. Because each pattern serves as an abstract, skeletal structure that can be realised in many different ways (for example in different keys, with more or fewer embellishments, diminutions, etc.), it is possible to understand the improviser’s knowledge as essentially conceptual in nature—that is, as a mental structure that, while clearly represented and defined within the improviser’s imagination, is not specific to a particular arrangement of notes but instead can be realised in an infinite number of different ways. An example of this might be a cadence, but mental structures can also be applied to concepts such as chord sequences, “opening moves” of a particular genre or form, or even the idea of musical “form” itself. Thus, practising the acts and exercises of improvisation (especially repetition, transposition, and elaboration of basic formulae and patterns within a particular style) results not only in physical skills but also in a specialised kind of knowledge—a way of perceiving and interpreting musical structure—which is suitable for improvisation because it is fundamentally conceptual in quality. ³

Acquiring an improviser’s knowledge and using Sei gegrüßet Jesu Gütig as a model for improvisation

My own acquisition of basic conceptual formulae and patterns was acquired as I tried to improvise (in a rather generic way) Baroque-style preludes. Over the course of a year, I pieced together four-part textures in diatonic tonalities, avoiding parallel motions in fifths and octaves, gradually gaining fluency in performing chains of sequences, opening phrases, and cadences. A considerable breakthrough was achieved through the discovery of Johann Joseph Fux’s rules of motion and categorisation of consonance-dissonance in the introduction to his 1725 Gradus ad Parnassum (Fux 1965, 20–22):

Figure 3.1. The classification of consonant and dissonant intervals and their permitted motions as proposed in Fux’s Gradus ad Parnassum.

Fux categorised the intervals as “perfect” (unison, octave, and fifth), “imperfect” (sixth and third), and “dissonant” (second, fourth, tritone, and seventh), explaining their use in musical textures with respect to three different kinds of motion: parallel, oblique, and contrary. Perfect intervals, for example, cannot appear successively in parallel and must be approached obliquely or through contrary motion. Thirds and sixths, however, can appear in parallel motion, and it is a characteristic of the Baroque style to have passagework in parallel sixths and thirds. These rules facilitated the speedy categorisation of intervals and so reduced the amount of time spent in calculating forwards. They also reduced retrospective assessment (i.e., was that a good or bad move?) as I now felt guided by clear rules towards what was effective, permitted, and expressive. Having acquired these basic conceptual formulae and patterns and assimilated them as skills, I was able to focus on the more specific conceptual language and features of specific models such as the chorale partita.

So far, my analyses have been quite theoretical: I have described a cognitive model of improvisation—in which the basic patterns and elements of musical structure come to be conceptually represented—in comparison with that used by the trained interpreter of scores who represents one version of musical events more literally. I now discuss how improvisers come to manipulate the communicative functions and events of music as a language as constructed in the moment. Progressing beyond the construction of basic patterns and formulae towards concert performance, specific genres, and complete musical forms, improvisers now “enter upon the stream of verbal communication; indeed, only in this stream does their consciousness first begin to operate” (Vološinov 1973, 81). This new consciousness or knowledge is also shaped by behavioural and social influences on musical decision-making that occur within “the field,” that is, the social contexts in which improvisation occurs. For example, it should be remembered that the act of improvising variations on a chorale theme commonly occurs within a church setting where the chorale is sung immediately following the organist’s Vorspiel. This means the improvisation serves to introduce the melody to the listeners, who benefit from many repetitions of the theme appearing in different guises. As a result, the improviser is constrained to choose clarity over complexity, especially at the beginning and end of the form, for the social purpose of the partita is, as my teacher Jürgen Essl remarked to me during a lesson, “to encourage people to sing,” i.e. to teach them the melody and inspire them towards active participation once the improvisation is finished.

Additionally, it is known that, within the Protestant community since the eighteenth century, the practice of improvising on chorale melodies has been widespread among professional organists and continually specified as a requirement in auditions for church positions until this day. ⁴ Thus, it was quite natural for organists to be aware of and draw on a shared library of resources, techniques, “scripts”—that is, customs and ways of treating the theme that both informed and extended the common language of the partita as an improvised event. Likewise, listeners in the Protestant faith could be expected to be aware of the conventions of a chorale partita through regular church attendance and immersion in the music of their culture. We can therefore view the improvisation as an ongoing dialogue between organist and congregation, with musical decision-making being led not only by theory and form but also through an exchange of meaning with the listeners—that is, ascertaining which decisions listeners most value and respond to, which in turn motivates the improviser to attempt new tasks, avoid others, and to construct the music in step with the listeners’ reactions. As Shotter and Billig (2008, 24) describe: “The kind of understanding indicated here is not of a cognitive, representational-referential kind, but is a practical, dialogical kind of understanding, a kind of understanding that is ‘carried’ in our ongoing languaged-activity, and is continually updated, utterance by utterance, as it unfolds.”

Improvising the chorale partita

The theme chosen for this demonstration is:

Figure 3.2. Discant of the chorale Aus tiefer Not schrei’ ich zu dir.

The improvisation begins with an initial presentation of the theme. It is harmonised in a simple fashion (MF1), leaving many possibilities of modulation and chromaticism to be explored later on. During the theme one establishes a basic harmonic plan (involving memorised movements for voice-leading principles, cadences, etc.) that serves as a conceptual structure for the subsequent variations. This means not only that the harmonic plan can be realised in a multitude of different ways through the different diminutions, rhythms, and textures characterising each variation but also that certain elements of the plan can be altered (for example using a different cadence) without changing the essential mental representation.

This procedure can be seen immediately in the first variation, which presents the following distinctive features: a series of running sixteenth notes in the left hand (MF2); an embellished version of the chorale theme in the right hand (MF3); and imitative sequences forming episodes between each right-hand phrase (MF4).

Of course, the performance of these selected concepts in the context of the chorale prelude rests on the foundation of the generic, basic, voice-leading movements and principles described above. In this stylistic improvisation one’s basic formulae and patterns (which realise these voice-leading movements almost without thinking) are drawn on automatically as a resource to be steered and shaped towards the distinctive features of the variation. As one gains in experience, the procedural knowledge used can be continually refined to bring one closer to the model. In effect, this process of refining multiplies the number of concepts used but also facilitates their selection, as is noted in skill-learning literature: Anderson (1982), for example, describes a hierarchical organisation of routines and subroutines all serving a common goal state and also explains how new solutions grow in “specificity and strength” (373) by making them more readily available in the moment of performance. To return to our example of the first variation, particular ways of embellishing the theme or of executing the left-hand sixteenths to make a smooth transition between the right-hand phrases or to give a pleasing rise and fall to the phrase are discovered and rehearsed in multiple ways until a new mental representation emerges. With many conceptual resources at one’s disposal, the improviser can respond to any suitable chorale theme with a variation of this character.

I have already mentioned that the opening harmonisation leaves many possibilities unexplored. My experience as an improviser suggests that these possibilities now unfold through variations that increase in complexity as both improviser and audience become more attuned to the theme and its harmonic possibilities. Certainly, in Sei gegrüßet all the subsequent variations gradually gain in complexity and chromaticism, beginning with several variations using only two or three parts, in which the theme generally appears in the upper voice before sounding in the bass (MF5), alto (MF6), and tenor (MF7) lines, and becoming increasingly adventuresome in the use of chromaticism and contrapuntal devices (MF8). After analysing, selecting, and abstracting features from many different chorale partitas, I would further propose that all the possibilities can be reduced to three rules or guidelines that will serve to construct a variation:

• texture is restricted to two, three, or four parts (more voices than this represents an unusually rich texture—see, for example, the final variation of Sei gegrüßet Jesu Gütig), with the chorale theme appearing in any of these parts;
• Metre is in duple or triple time (note that 2 can also yield 4 and 8; 3 can yield 6 and 9);
• Constructive devices are those commonly used, including distinctive melodic and rhythmic motifs, characteristic organ registrations, contrapuntal imitation, and so on.

Conclusions

A thread that runs throughout this volume is how to overcome the tyranny of the historical “text” in classical music by instead creating new relationships, new ways of perceiving canonic works. Improvisation emerges as one of several methods to establish a new musical practice, one in which historic texts, despite their undeniable stature and the achievements they represent, serve as models for an unlimited amount of new music. In discussing the practice of improvising new versions of a canonic work I have touched on several areas including musical analysis, cognition and mental representation, sociocultural context and knowledge (also sociocultural criticism), and music as a language constructed dialogically in the moment. Throughout I have compared the cognitive perspective of the interpretive performer with that of the improviser, and many of these insights arise from my own experience in both roles. Clearly, as a convert to improvisation, my preference is for the improviser’s perspective, which, in the plurality of its conception, seems to me a more natural, creative, and truthful way of making music. But what do I mean by “truthful”?

Certainly, through improvising we recreate a truer historical practice, documented by sources such as treatises and eyewitness reports. As mentioned above, Baroque organists not only improvised chorale partitas but also notated versions of their improvisations, the latter serving as models for new improvisations and new scores; thus, plurality of text was intrinsic to a practice that seemed more preoccupied with day-to-day craftsmanship than encoding a “definitive” version.

There is also a more truthful musical knowledge to be created through improvisation: by inserting itself into the living language of the musical style, by trying to speak in this style, by manipulating and solving the problems and constraints of the language in a similar way to the original composer, the improviser recreates a similar cognitive experience. As Baily remarks: “the structure of the music comes to be apprehended operationally, in terms of what you do, and by implication, what you have to know” (2008, 122). This also means that, as an improviser, one can perform and interpret the original text with “inside” knowledge.

Lastly, I propose that the Werktreue belief that meaning—a meaning, any meaning—can be woven into the text as interpreted and/or communicated by the performer is fundamentally false, ideological in nature, and thus untruthful: “a pure dream,” “the imaginary relationship of individuals to their real conditions of existence” (Althusser 1971, 160, 162). And I offer this proposition despite my admiration for the undeniable achievements of interpreters who focus so intensively on single texts. The improviser’s attitude of plurality emphasises the potential of music as a shared and living language, manipulating its forms and expressions in order to create a unique dialogue with present conditions and contexts: acoustic surroundings, audience expectations, instrumental characteristics, and so on. It is this potential that recreates the text (even one from the classical canon) as a more relevant, everyday, moment-to-moment aesthetic experience. Whatever relationship exists between an improvised version and the score-as-model, the former is always in search of something new, as Bakhtin states: “an utterance is never just a reflection or an expression of something already existing and outside it that is given and final. It always creates something that never existed before, something absolutely new and unrepeatable” (1986, 119–20).