Abstract
This article reports a qualitative exploratory case study of a composer’s compositional inspiration, thinking, and decision-making. The aim was to explore the dynamics and functions of intuitive and reflective modes of thinking, thus expanding the recent trend in composition research, which has given increasing attention to implicit, imaginative, and inspirational aspects of composition. The data of the study were collected in natural settings in the composer’s work studio during the compositional process. The data comprised stimulated recall interviews, as well as all of the manuscripts that the composer wrote concerning the process under scrutiny. The findings indicate the saliency of the composer’s germinal ideas into the compositional process. The composer elaborated, transcribed, and embodied his multimodal ideas, step-by-step, into musical passages of the evolving score. The process comprised two core procedures. First, compositional thinking was manifested as concrete representations, i.e., the manuscripts. Second, intuitive aspirations entered the conscious mind and subsequently opened themselves up for reflective processes. Two dilemmas were identified in defining the composer’s uncertain and complex working circumstances. The strategic and ontological dimensions of compositional decision-making were identified. The study produced a thick description of what is probably the most complicated phenomenon of musical behaviour: composing outside the confines of a laboratory environment. It explicated compositional thinking as continuous and appropriate fluctuations of intuitive and reflective ideation, monitored by metacognitive function. It also presented a detailed analysis of the dynamics of problem accumulation, i.e., the way the composer decided not to decide, which originated from the composer’s ill-defined working circumstances and his ethos of seeking coherence and employing constructionist aesthetics.
Introduction
Empirical studies of compositional processes of professional composers mostly focus on cognitive steps that underlie their reflective decisions – whether investigated through experimental settings, interviews, sketch analyses, or through writings by the composers themselves. In terms of intuitive thinking, scholars usually settle for evidence that indicates the potential for implicit thinking (Alty, 1995; Sloboda, 1985; Wiggins, 2012). Recently, increasing attention has been given to the intuitive component of compositional thinking. Various terms are used to describe this non-reflective thinking: “insight” (Collins, 2005; Collins & Dunn, 2011), “intuition” (Brown & Dillon, 2012; Holtz, 2009; McAdams, 2004), and “within-domain” (Katz & Gardner, 2012; see also next section), “automatic” (Alty, 2002), “associative” (Schubert, 2012), or “Dionysian” (Miranda, 2012) processing. However, the functions of intuition or insightful illumination, and the functions of reflective thinking in musical composition are seldom scrutinized: intuition may rather self-evidently be seen as a means of creating novel constructs; whereas reflective thinking might be considered to lead to mundane or unoriginal solutions. Furthermore, how intuitive and reflective thinking fluctuate and facilitate each other within compositional acts is seldom investigated.
Due to ambiguous conceptions about roles and dynamics of intuition and reflection in compositional thinking, composition research has been based on various theoretical foundations, most often on a combination of them: Gestalt theory and information theory (Collins, 2005; Collins & Dunn, 2011), creativity research and education (Brown & Dillon, 2012; Katz & Gardner, 2012; MacDonald, Byrne, & Carlton, 2006), problem solving theory and musicology (McAdams, 2004), cognitive anthropology and musicology (Donin & Féron, 2012; Donin & Theureau, 2007), and information theory and cognitive science or cognitive psychology (Bailes, 2009; Miranda, 2012; Reitman, 1965; Schubert, 2012; Sloboda, 1985; Wiggins, 2012). While some theories either overlook or underrate the unpredictable and intuitive component of artistic work (e.g., information theory, problem-solving research; see Goel, 1995; Öllinger & Goel, 2010), others, such as creativity theory, may be too broad or general to establish a satisfactory basis for empirical research on a singular artistic endeavour (see Katz & Gardner, 2012; Sparshott, 2009).
This paper introduces a novel approach to the empirical study of composition: it adapts dual-process theory of human information processing to composition research. With dual-process theory, both the intuitive and the reflective part of artistic thinking can be addressed without underrating either of them. A suggestion is made to bridge the gap in research on the roles of intuition and reflection, and on the dynamics between them in compositional thinking. The aim is to investigate how intuitive and reflective modes of thinking manifest themselves in a composer’s inspiration 1 and germinal ideas, as well as in compositional decision-making. Further, the interaction and dynamics between the two thinking modes is scrutinized, and the role of intuition and reflection in generating aesthetic congruence within an individual piece of music is examined.
The epistemological intention of the study is to detect the composer’s thinking in action (Schön, 1983) within natural settings. In order to respect the compositional process as a singular and unpredictable episode, a case study was conducted in which a typical professional assignment of a composer was traced: that of creating one extensive piece of music from the very first germinal ideas to the complete score. Special methodological attention was paid to reducing the sources and ramifications of misinformation that verbal accounts commonly convey, particularly when studying creative processes, and tracing the interaction between intuitive and reflective thinking (see Pohjannoro, 2012).
Intuition, reflection, and metacognition in compositional thinking
This study examines compositional thinking according to the dual-process framework of human information processing, a theory widely accepted within cognitive sciences (see an overview in Frankish & Evans, 2009). The dual-process theory attributes human reasoning and decision-making to two different modes of thinking; a notion that is also common in artistic discourse. One type of process (type 1, called “intuition” in this study) is described as fast, unintentional, automatic, effortless, and capable of dealing with large amounts of information simultaneously. The other (type 2, called “reflection” in this study) is slow, deliberate, sequential, laborious, and limited in its processing capacity (Evans, 2009, 2011; Sloman, 1996; see Allen & Thomas, 2011, and Dietrich, 2004 about dual accounts of creative thinking).
The latest development in the dual-process framework postulates that processes of monitoring, controlling, and conflict detection account for the selection between intuitive and reflective processes (De Neys & Glumicic, 2008; Evans, 2009, 2011; Goel, 2009; Stanovich, 2009). In this study, I call these intermediate processes “metacognition” (Thompson, 2009; Thompson, Turner, & Pennycook, 2011). Thus, I consider compositional thinking as a fluctuation between intuitive and reflective processing, which is controlled by metacognitive processes and can be either unintentional (type 1) or purposive (type 2).
Regardless of increasing attention to the intuitive component of compositional process, associations between these different aspects of thinking have mostly been neglected or compromised. For example, McAdams’s (2004) study examined the composer Roger Reynolds’ method of addressing compositional problems explicitly, while pushing as many of the intuitive processes as possible into the realm of rational thinking (see also Collins & Dunn, 2011). Likewise, Collins and Dunn (2011) realized that illuminative insights were recurrent and occurred in particular stages of the linear and non-linear problem-solving cycles (see also Collins, 2005, 2007; Miranda, 2012).
The studies of Donin and his colleagues (Donin, 2012; Donin & Féron, 2012; Donin & Theureau, 2007) show interesting aspects of composers’ intuitive processing: working with vague ideas, letting the material lead the way, and recognizing creativity in the use of rules or deductive reasoning. Holtz’s (2009) study based on interviews with 17 composers elaborated Bahle’s typology that recognizes inspirational type (using mainly intuition) and working type (using chiefly reflection) among composers (Bahle, 1938). Studies by Bailes (Bailes, 2009; Bailes & Bishop, 2012), Hargreaves (2012), and Mountain (2001) have focused on composers’ imaginative processes, which comprise implicit and explicit reasoning. Likewise, the study of Brown and Dillon (2012) demonstrated various meaningful compositional activities as a mixture of directive and inspirational behaviour. Wiggins (2012) suggested a theory concerning compositional inspiration, in which information is created through unconscious statistical operations according to the information-load of the respective item, and is then selected with deliberate “hedonic assessment” (see also Bailes, 2009; Mountain, 2001).
Katz and Gardner (2012) studied how different types of impetus manifest themselves in interviews with various composers. Katz and Gardner established two prototypes of inspiration, involving two distinct processes: within- and beyond-domain processes. Interestingly, the within-domain processes (that incorporate inner hearing, imagination, improvisation, and experimentation, typically inspired by musical ideas) can easily be associated with the qualities of intuitive processing. Accordingly, the beyond-domain processes (that include selecting alternatives and accommodating constraints when setting the musical parameters, and inspired by extra-musical metaphors and associations) greatly resemble the characteristics of reflective thinking (Katz & Gardner, 2012).
Research design and procedures
Method
The data were collected utilizing the stimulated recall (STR) method. The STR method requires that the interviews be conducted as soon as possible after the actual action. Stimuli are used to help the informant to recall his or her thinking. Indeed, retrospective interviews may be one of few methods of tracing the subject’s thinking, without disturbing the actual thinking process – particularly in lengthy processes in which the mind works faster than the mouth, where forgetting is an issue, and where pre-reflective thinking plays an important role (see Collins, 2007; Ericsson & Simon, 1984/1993, p. 106; Perkins, 1977, 1981, pp. 36–37). In this study, the composer’s sketches, material matrices, and score versions were used as sources to help the composer verbalize his creative process (see also Donin & Theureau, 2007; Donin & Féron, 2012).
The STR method is widely reported and commonly used to detect thinking and communication processes (see the extensive literature review by Lyle, 2003). However, the method has several pitfalls and limitations, including those concerning issues associated with verbal data in general (see Pohjannoro, 2012). Even so, the STR method clearly eliminates some of the limitations associated with some alternative methods. First, unlike protocol analysis, the STR procedure neither interferes with thinking in action nor disturbs the process under study (Collins, 2007; Ericsson & Simon, 1984/1993; van den Haak, de Jong, & Schellens, 2003). Second, STR facilitates the informant’s recollection of long processes – a recollection, which decreases over time – and allows the interviewer to steer the informant’s attention to concrete acts. The resulting minimization of reactive (explanatory, justificatory) elements in the accounts is more difficult to achieve with un-stimulated interviews. The STR method also allows the focus to be directed by the interviewer to issues relevant to the research questions, which is not possible when using protocol analysis. Further, divergent empirical data collected through the STR procedure – in this research, the interviews and the manuscripts – can be analysed and evaluated separately, and thus validated against each other (see Ericsson, 2003). In this study, it became obvious that it was indispensable to gain an overall understanding of the whole process, and especially of the role of the Identity Idea included within (see the Results section), in order to be able to categorize the verbal data into type 1 (intuitive) or type 2 (reflective) processing modes (see Pohjannoro, 2008, 2012). The verbal data elaborated the analysis of the manuscripts, which again supported the interpretation of the verbal data.
Information about the composer
The informant in this study was a professional academic composer, educated in the Finnish modernistic school. During the time of the study, he had been active as a composer in the national and international western classical music scene for more than 20 years. He rejected the use of think aloud, but agreed to be interviewed anonymously.
Data
The main empirical data were collected in the composer’s studio between 27 December 2004 and 31 January 2005, extended by orientation and supplemental interviews conducted on 5 January 2004 and 19 January 2007. The data, their main content, and their collection periods are shown in Table 1. The verbal data comprise 12 interviews (406 minutes and 29,000 words of verbatim transcripts). The transcripts include the composer’s sighs, and exceptionally soft or emphatic speech is parenthesized or capitalized, respectively. 2 The manuscripts include 4 computer-annotated material matrices, 11 hand-made sketches, and 17 score versions typeset with Finale 2002 notation software and often including handwritten markings and additions. The verbal and written data are in Finnish. All of the data, including the copies of the manuscripts, are in the possession of the author.
The data, their main content, and their collection dates according to the compositional phases and stages.
Note. aMovements of the complete piece: 1st movement: (T1–S1, T2–S2); 2nd movement (T>S, T>S, T+S [fr]); 3rd movement (m0–m1–M).
Data handling and coding scheme
The coding categories of compositional acts were constructed using both data- and theory-oriented approaches. The dual-process theory is predominantly based on experimental research, and it has been argued to employ unsatisfactorily contextualized research environments. Such research designs dissociate intuitive and reflective processing modes, utilize restricted, biased operationalizations, and test designs which may consequentially favour the very processing mode that is being studied: either intuition or reflection (Betsch, 2008; Kruglanski & Gigerenzer, 2011). Hence, the dual-process framework in its current state of theorizing did not give an adequate ground for drawing empirical differentiation of the intuitive and reflective compositional acts from the qualitative data of this study. Moreover, the data-oriented approach enabled profound consideration of the viewpoint of the composer.
The interviews, supported by sketch analysis of the manuscripts, 3 were first encoded in a data-oriented manner into categories, according to the composer’s compositional acts. The categories were then theorized and classified according to both the triarchic classification of intuitive, reflective, and metacognitive modes of thinking and the type 1–type 2 dichotomy that incorporated the metacognitive mode into type 1/type 2 thinking. At this point, the data were partly renamed and segmented into smaller units of analysis to fit the theoretical definitions of the three thinking modes. To enhance the reliability of the analysis, a second rater coded about one-fifth of the data in three stages, with the support of the code definitions. Some minor adjustments to the definitions of the codes between the cross-coding cycles were made. The confirmation procedure was conducted until satisfactory inter-coder consensus was reached.
The intuitive compositional acts comprise imagination, experimentation, incubation, and restructuring (see Figure 1). The definitions of the compositional acts are presented in the Appendix, and the coding examples are published as supplemental online material (see link at the end of the article). Reflective processing incorporates rule-based processing, contemplating different alternatives, and analytic viewing of the music. Some of the compositional thinking could be perceived as serving controlling and monitoring functions of processing. In this study, these intermediate processes are called “metacognition” (see Thompson, Turner, & Pennycook, 2011), which comprise making operative plans, evaluation, and setting musical goals. Interestingly, the last two mentioned metacognitive acts proved to be both intuitive and reflective, while making operative plans was only reflective. Therefore, the data supported the latest versatile conceptions of cognitive control and metacognition. In recent theorizing, cognitive control (Braver, Gray, & Burgess, 2007; Deroost, Vandenbossche, Zeischka, Coomans, & Soetens, 2012) and metacognition (Koriat, Ma’ayan, & Nussinson, 2006, pp. 60–62; Sun & Mathews, 2012, pp. 110–111) are more often considered to be either deliberate and endogenous (i.e., goal driven), or intuitive and exogenous (i.e., data driven), and therefore automatic phenomena (Koriat et al., 2006; see also Hommel, 2007).
Compositional acts categorized into intuitive (type 1), reflective (type 2), and metacognitive modes of processing.
All utterances about the compositional acts were also classified as conveying feelings of doubt or confidence, thus forming the experiential aspect of the compositional acts. A total of 553 utterances were classified as compositional acts, and 617 were classified as experiential aspects of compositional acts. The data were analysed both quantitatively and qualitatively. However, only the results of the qualitative analysis are presented in this article.
Description of the completed composition
During the study, the composer wrote a 15-minute percussion quartet, commissioned by a Mid-European percussion ensemble. The piece has no traditional harmonic or melodic element other than the melodic patterns formed by similar percussion instruments of various sizes.
The formal structure of the piece can be scheduled as follows (analytic symbols used in the schedule will be explained below; see Pohjannoro, 2008, for a more detailed analysis of the structure of the piece):
The completed composition comprises three movements, played attacca. The first movement introduces the core materials, called by the composer linear music (T) and space music (S), in a binary phrase construction (T1–S1, T2–S2; the hyphens, commas, and semicolons refer to formal boundaries in order of hierarchically increasing levels). T is written with a time signature: it consists of four basic rhythmic patterns, each constructed of simple rhythmic cells (dactyls, trochees, and spondees, including their inversions; see Figure 2). S is written without a time signature: it is more varied in instrumentation and tone colour than T (see Figure 3).
Score version no. 4, dated 27 December 2004, introducing linear music (T): the beginning of T1 (page 1, starting from rehearsal no. 1.1), and T2 (page 5, starting from rehearsal no. 3.1). The rest of T1 excluded (pages 2–4, rehearsal numbers 1.4–1.10). The four basic rhythmic patterns are marked with circled numbers 1, 15, 25, 35. The composer selected the rhythmic patterns from his material matrix A (dated 7 December 2004). The score version shows some of the empty bars in T1, which resulted from the composer’s reluctance to make decisions about the manner in which he would develop and transform his material in order to make a valid development for the material T. T2 only contains two written bars, beginning with the basic rhythmic pattern no. 1, and it introduces the idea of transforming the basic rhythmic patterns by augmentation. In spite of the fact that the composer did not know how he would proceed with the augmentations, he already ventured to ponder about the relationships between T1 and T2. However, he could not come to any conclusion (see Figure 4, arrow no. 1).
Score version no. 3, dated 27 December 2004: the coda of T1 (rehearsal no. 1.11 in the manuscript; basic rhythms in contrapuntal setting) and S1 (the first instance of space music; rehearsal no. 2 in the manuscript). Numbers 1–7 indicate the order of invention (1, 2, 3, 4A–4C, 5, 7A–7C; no. 6 missing), marked by the composer during the interview of 27 December. Numbers 1, 2, and 4C indicate the repetitive element (r), no. 5 the melodic element (m), and nos. 3, 4A, 4B, 4C, 7A, and 7B the tremolo element (tr) of space music. No. 7C indicates a turn to space music’s tremolo element. The capital letters A, B, A1, and B1 indicate a binary phrase construction (“parillinen lauseke”; the first handwritten text), which begins S1. By the text “joulu-tauko” the composer indicated the beginning of a Christmas break in his work. The capital text “ISO REIKÄ 20 S TYHJÄÄ” means “big gap, 20 seconds blank”.
In the second movement, the composer operates with transitions between T and S (the transitional relations marked in the formal scheme, with >). He treats the materials T and S as themes, juxtaposing them in various ways, and making diverse rhythmic modifications. The aim is to entwine T and S together, and to allow their distinctions to melt away, such that the listener ceases to differentiate the end of one from the beginning of the other. The movement and the whole composition culminate in total fragmentation of the materials.
The third movement is a coda-like contemplation of the previous musical events. In contrast to the first movement, space music dominates the third movement, with only transient suggestions of linear music. The third movement introduces the evolution of the space music’s melodic element (m) into the full “melody” (M). Aesthetically, this is the culmination of the composer’s organic and austere use of his materials, as well as their rigorous musical arrangement.
Results
Germinal ideas, fundamental dilemmas, and core procedures
The findings of the qualitative analysis indicate the saliency of the composer’s set of original conceptions into the compositional process and compositional thinking. The germinal ideas in this study were predominantly intuitive, popping up out of nowhere and without any obvious justification. However, there was a reflective side to the ideas as well: the composer intentionally decided to investigate the way musical time is perceived and comprehended, using linear music and space music as his experimental material. The ideas of space music (S) and linear music (T) represented different ways of perceiving (musical) time to the composer, rendering even philosophical aspects of temporality. These, and other germinal ideas, were both musical and extra-musical in nature: audial, visual, auditive–structural, and philosophical–conceptual, incorporating all source types of inspiration identified by Mountain’s (2001) study of composers’ writings and interviews (cf. Holtz, 2009, and Katz & Gardner, 2012, where each composer usually had an exclusive type of germinal idea).
The audial or conceptual ideas, such as S and T, together with visual ideas, such as whiteness or glacier, established the composite of the composer’s germinal ideas. Before commencing the actual compositional process – in other words starting to write down and explore his basic musical materials and to arrange passages onto the blank staves of the score – the composer constructed an organic conception out of his perceptually and conceptually divergent germinal ideas: he called this concept the Identity Idea. The Identity Idea formed a specifier of the musical and expressive identity of the work in progress. It provided the composer with a divergent musical platform for exploring contrasting metrical structures, and a mental playground for creating extra-musical meaning (see Pohjannoro, 2008). Ultimately, the Identity Idea framed all of the composer’s creative endeavours; even the entire process.
The Identity Idea proved to be the critical determinant of the whole thinking process, functioning as the driving force of the compositional process, and serving as an evaluative element for all of the composer’s compositional acts: he repeatedly considered how his actions contributed to the Identity Idea, and whether his decisions were coherent with it. The question was: Does this (musical passage, choice, expression) belong to this piece? However, the Identity Idea, though unquestionably and clearly apprehended, appeared to be too fuzzy to predict the actual and concrete process of musical modelling (see also Donin, 2012; Donin & Féron, 2012): the use of the Identity Idea – with its eidetic and multimodal, albeit cognitively diverse and fuzzy qualities – implied comprehensive perplexities in the composer’s work. These uncertain and complex working circumstances forced the composer to utilize the best of both his intuitive and reflective thinking devices.
The composer’s first dilemma concerned the temporal and procedural nature of the music and the inherent diversity in the set of conceptions that formed the Identity Idea. While the Identity Idea manifested itself as an eidetic multimodal experience, which was as evident and as persuasive as it could be, the composer could neither image it as an extended passage of sounds, nor write it immediately down as a musical structure of this lengthy piece. The next quotation – translated into English by the author – exemplifies the crucial but perplexing role of the Identity Idea in the composer’s thinking processes:
Are they [sound ideas] like fragments?
Yes, as a matter of fact they appear to be. They are fractions, which come with a conception of the whole … When you start to write it down, you have a feeling that it is quite clear, it just needs to be put down. But one usually only writes half a minute of music, two minutes at most. Usually there will be less than half a minute, until a question arises, or many questions … It wasn’t as easy as one imagined, after all … [Sigh.] (P10/186–187/I)
4
The composer’s other dilemma was the problem of how to start on a blank sheet and work with the Identity Idea’s multiple potential configurations and implications, and to take compositional initiatives and decisions that would create and maintain the aesthetic coherence of the evolving work. The composer spoke about a “jigsaw dilemma”, which accurately reflects Schön’s (1983, pp. 141–156) discussion of the ways an expert conducts his or her “practical tests”. On the one hand, all decisions have implications for future options – foreseen or unforeseen. On the other hand, the unknown future implications reflect the decisions of the present. In the composer’s own words: ANYHOW, you now and then need to look into a fairly small detail – this is only a two-minute passage, anyway [points to the passage S1] – you constantly have to compare the outcome to what has already been done, [that is] on the one hand. But on the other hand you, above all, have to compare it to that which does not yet exist. You have to compare it to what you are trying to imagine that there will be. (P6/25/S/IX)
The compositional process, delineated by these two dilemmas, proved to be a laborious endeavour of embodying, articulating, transforming, and refining the abstract and diverse conception of the Identity Idea into the musical passages of the evolving score. The process comprised two overlapping core procedures.
The first compositional procedure involved converting compositional thinking into external representations, i.e., the arrangements in the manuscripts. The process moved forward in an iterative manner: mental representations were transcribed into concrete representations (arrangements), which again imposed new mental representations, not all of which were foreseen by the composer (see Donin, 2012; Goel, 1995, pp. 193–195; Schön, 1983, pp. 158–159). In Schön’s (1983, pp. 131–133) terminology, the situation “talked back” to the composer, driving him to accept or reject the arrangement or to adjust it. When necessary, this concretizing procedure (transcribing mental representations into concrete ones) included transformation of non-musical ideas into musical ideas (see Bailes, 2009; Katz & Gardner, 2012, p. 111). However, not every notated idea had to be audible.
The second compositional procedure is about deepening and extending the awareness of the substance at hand. The pre-reflective, intuitive aspirations enter the conscious mind, and thus open themselves up for reflective processes (Perkins, 1977; Policastro, 1995, 1999; Wiggins, 2012). The procedure evolves through rapid back-and-forth movements of different intuitive and reflective compositional acts, monitored by metacognitive acts. The composer used different thinking modes according to the vagueness of the information behind his processing, which often included the constantly evolving Identity Idea. He not only altered distinct compositional acts according to his level of awareness of the qualities in a particular musical passage, but his orientation also propelled him towards extending the scope of that (very) same material. In other words, the composer both investigated the substance and inner composition of his ideas, and aspired to understand better the ideas’ references and ramifications for the other parts of the emerging piece, whether or not they were already written. The analysis of an episode of the compositional stage two in the next section shows how the intuitions moved step-by-step up and down the threshold of consciousness, and subsequently enlarged the conscious area of the idea at hand and its references to other local ideas, to the Identity Idea, to appropriate musical passages already constructed, and even to virtual music that was just emerging.
Postponing decisions, avoiding muddle, and ending up in crisis
As discussed above, the Identity Idea constituted the quintessential work-specific (local) determinant of the compositional process. 5 The composer tackled his enterprise of creating a compelling and intriguing, as well as coherent, artefact through the prolific use of intuitive and reflective thinking. However, regardless of the composer’s intuitive and reflective compositional acts, at the beginning of the process he could only attempt to probe his vague ideas and materials, whilst being unable to devise far-reaching formulations. The most important strategic practice in forming the whole compositional process was the way the composer sometimes actively deferred important decisions until later in the process. The first quotation (P10/186–187/I) of this article shows how the composer was unable to proceed smoothly after the first arrangements of his materials into the score. He could not see all of the relevant compositional options or their implications for the congruence of the whole work. Consequently, he just suspended the decisions and moved onwards through the emerging piece, leaving behind empty bars and unanswered questions, and formed only a skeleton of the section of the score that would later become the first movement of the work.
The first substantial complication occurred after the creation, exploration, and rudimentary inscriptions of T1, T2, and S1: what kinds of relationships and musical events will there be between sections T1 and T2 (see Figure 4, arrow no. 1; Figure 2)? Next, the composer faced an analogous question within passages S1 and S2 (Figure 4, arrow no. 2). Is something more or less substantial needed for S2, compared to S1 (Figures 3, 5)? Subsequently, how will the procedures T1–T2 and S1–S2 bond with each other (Figure 4, arrow no. 3)? The solution would surely have significant ramifications for the rest of the piece, especially for the second movement (Figure 4, arrow no. 4).
Deferring decisions and the evolution of the crisis in the second compositional stage. The structure of the complete composition is presented at the bottom of the figure; the accumulating problems are numbered in order of appearance with numbers followed by question marks, indicating the three arrows. The arrows point out the questions about the relations between the two Ts, and the two Ss, and between the relations T1–T2 and S1–S2.
Score version no. 5, dated 30 December 2004: S2 (the second instance of space music; rehearsal no. 4 in the manuscript). The example shows several open questions which manifest how the composer was unable to decide about the musical continuation of the material S (see Figure 4; arrow no. 2), and how that development related to the development between T1 and T2 (see Figure 4; arrows no. 1, and 3), both of the latter two being more (T2) or less (T1) unfinished – just as S2 is left incomplete. Figures 2, 3, and 5 are published with the kind permission of the composer.
Because the composer could only resolve these problems up to a point, he moved on to the second movement, and started operating with the transformations between linear music and space music. The crisis of accumulated problems was eventually resolved in a multifaceted sequence in phase XI, after the composer invented the idea of fragmenting the materials and decided the passage’s role as the dramaturgical focus of the second movement and of the whole piece. To reach this solution, the composer had to use the whole repertoire of his operational practices (see Pohjannoro, 2008).
Validity issues
Few scholars have reflected on the fallacies concerning composers’ accounts of their work, especially with regard to the intuitive aspects. These accounts tend to be not only generic, but also confusingly diverse. Whereas some composers reinforce the common understanding of the artistic process as being mystical in nature, others characterize composition as a laborious craft, sometimes even likening it to scientific investigation (Lehmann, 2005; see Harvey, 1999; McCutchan, 1999; Wiggins, 2012). While the mystical accounts of the composers’ work can be partly attributed to the desire to maintain professional secrets, these accounts, as well as the rationality-based ones, can also be attributed to composers’ inability to describe or understand the intuitive component of their work, and the ways in which it is actualized in their concrete compositional acts. Moreover, numerous sketch studies have questioned composers’ accounts of their artistic intentions and processes (see Donin, 2009; Schubert & Sallis, 2004; Sloboda, 1985, pp. 112–115). Overall, scholarly attention to the creative processes of composers has long been restricted to posthumous studies based on historical documents or analytical arguments about music (see Bahle, 1938; Cooper, 1990; Kozbelt, 2009; Simonton, 1999), although with some seminal exceptions (see Sloboda, 1985; Reitman, 1965).
Recently, scholars have increasingly committed themselves to studying composition in natural settings (Brown & Dillon, 2012; Clarke, Cook, Harrison, & Thomas, 2005, Collins, 2005; Donin, 2012; Donin & Féron, 2012; Donin & Theureau, 2007; Wiggins, 2007). The problem is that a concurrent interview or think aloud may disturb the process itself (Collins, 2007). The stimulated research method addresses the question of data capture with its use of memory stimulants to enhance the informant’s recall of her or his life episodes. However, some precautions ought to be taken. When reporting past episodes – especially in the context of creative processes – thoughts that are not easily accessible to awareness are, for various reasons, seldom verbalized. Procedural details, intuitions, and emotions that are attached to cognitive acts (see Damasio, 1996) tend to be either unintentionally or deliberately forgotten – if ever consciously recognized – or otherwise concealed in socially constructed narratives about personal life episodes (Anderson & Huddleston, 2012; Barsalou, 1988; Ericsson & Simon, 1984/1993, p. xvi; Perkins, 1981, pp. 13–18; Schwartz, 2007). Artists might also elect to conjure up a more favourable image of themselves. 6
The STR method, when executed meticulously and analysed attentively, is considered a sound and vigorous method for investigating human interactions and thinking processes whenever concurrent thinking aloud is inappropriate (Collins, 2007; Lyle, 2003). To minimize memory-related biases (see Ericsson & Simon, 1984/1993, p. xvi; Perkins, 1981, pp. 13–18; Schwarz, 2007), the interviews in this study were conducted within a fortnight of the actual events. Further, the informant’s reactive behaviour was carefully monitored (see Dennett, 2003; Ericsson & Simon, 1984/1993; Nisbett & Wilson, 1977). During the interviews, the composer’s attention to the concrete manuscripts and to their specific points of interest was steered to avoid accounts of how he typically operates. Moreover, neither the complete work nor its premiere performance was available, against which the composer could compare his account. The composer was constantly in the focus of his creative process, which clearly restricted the energy available to him to explain and justify his actions. As a result, few interpretive or justifying utterances were recorded in the data (see Schön, 1983, reflection-on-action; cf. Donin & Féron, 2012; Donin & Theureau, 2007), and all post-event explanations and justifications could be identified and excluded from the analysis (for details about the complete analytic and validation procedure, see Pohjannoro, 2008, 2012). Consequently, a substantial amount of reflection-in-action (Schön, 1983) was obtained during the STR-sessions: after reporting what he had previously done the composer quite naturally continued to develop his ideas. The accomplishment of combining the understanding of the whole compositional process and disclosing the composer’s authentic thinking in action enabled a profound scrutiny of compositional thinking in its finest ingredients, as well as their different dynamics and functions.
The limitations of the data derive from the fact that, although the whole manuscript corpus was, beyond doubt, obtained and chronologically organized, which is unusual in the traditional study of sketches (see Donin, 2009), only a fraction of the compositional thinking was conveyed. The interviews were undertaken every fortnight, while the composer’s compositional thinking presumably persisted for dozens of hours between the interviews. The critical question is whether the masked thinking differed qualitatively from the compositional thinking included in the data. The most likely qualitative gaps may have been in the verbalization of experiential matters, especially in the realm of intuitive thinking and experiences (Pohjannoro, 2012; see also Perkins, 1977).
This study cannot answer the question of how much the research procedure changed the composer’s mind-set about the process. However, the results suggest that the interviews hardly intervened in the compositional process more than the usual exogenous factors, considering the intensity of the composer’s intrinsic motivation, and the perseverance with which he used the Identity Idea to steer his ideating, planning, and decision-making. Having said that, it should be noted that the composer probably became more aware of his compositional practices, notwithstanding that metacognition already formed an important part of his typical compositional devices.
Overall, the interview data included many indications of emotional awkwardness, doubt, and anxiety, and occasionally even a chaotic atmosphere in the actual compositional situations. Previous research on the compositional process has either not addressed or has underrated the emotional difficulties of composition (see, for example, Collins, 2005; Donin & Theureau, 2007; Donin & Féron, 2012; Katz & Gardner, 2012; McAdams, 2004; see also Moisala, 2011). Arguably, the data of this study realistically demonstrate the composer’s setbacks and difficult emotions, which indicate a factor that validates the data. This type of data validation can be attributed to the choice of method, to the fact that the composer remained anonymous, and to the relaxed and familiar mood maintained during the interviews, the latter fact evidently being a result of long acquaintance between the composer and the scholar.
Finally, in the course of the research project, the interpretations and constructs were validated through member check procedure (Creswell & Miller, 2000). Besides reading the individual research articles (Pohjannoro, 2008; the article at hand) and approving their publication, this respondent validation procedure included checking the key terminology, and the translation of the interviews into English. This was to ensure that the final presentation of the data reflected the composer’s experiences and that the findings did not contradict his self-understanding. None of the composer’s handful of remarks were concerned with anything other than the accuracy of the musical or aesthetic terminology.
Discussion
The composer had to start from the void of a blank sheet, with nothing but the vague Identity Idea to steer his actions (see Alty, 1995). There was nothing to adhere to, nowhere to look, and no options to choose from. The composer used the evolving Identity Idea as a crucial means to create and sustain aesthetic congruence within his individual work. From the complexity of the innumerable possibilities on the composer’s tabula rasa, he applied the Identity Idea as an evaluative reference for his intuitive choices and as a starting point for the constraints with which he set the boundaries of his explorative work (see Policastro, 1995, 1999). Similarly, Donin and his colleagues described a “general plan” (Donin & Theureau, 2007), and “playful and whimsy” working style (Donin & Féron, 2012) as a particular cognitive devices in the compositional process. This study elaborates and scrutinizes the roles and dynamics that intuitive and reflective modes and the germinal ideas had here in the invention of the composition, and in the generation of its aesthetic coherence.
Despite the crucial impact of associative and automatic thinking, the role of reflection approximated to the importance and the amount of intuition. Reflection, as well as intuition, played a part in the creation of novel solutions, but also in originating possible crises. After starting with a systematic construction and with the investigation of his basic rhythmic patterns (linear music), both compatible with the Identity Idea, the composer was increasingly forced to adjust all of his actions to what he had done earlier, according to his aesthetic principle of coherence. Reference was made not only to the existing substance and to the still fuzzy, albeit constantly concretizing and clarifying Identity Idea, but also to his future choices (see also Donin, 2009; Donin & Féron, 2012). When the composer could not imagine the implications for the future, either intuitively or reflectively, he refused to make a decision and went on until he was ready and able to resolve the problem. Interestingly, at the end of the compositional process, the composer claimed that he had relied increasingly on intuition. Yet, from the perspective of an analytical (music) scholar, the composer’s actions seemed more like reflection, even in its stricter form, and almost like deductive reasoning. The discrepancy between the composer’s statement and the scholar’s observations of his rational-like behaviour will be explained in an article presently in preparation.
Although all of the composer’s acts entailed a decisive element, the moment or origin of such compositional decisions could not usually be detected from the data, except when the composer was able to clearly specify several options from which one had been chosen (see Goel, 1995, p. 113). However, the decision-making function was apparent in many of the compositional acts: experimenting and accepting the result, setting musical goals, and concretely writing the music down. Hence, the decision-making function could be identified as two emergent dimensions of compositional acts: the ontological dimension and the strategic dimension. These dimensions and their constituent elements are shown in Figure 6. The first dimension of decision-making is strategic: decisions can be distinguished locally according to the work-specific Identity Idea, or globally, on the basis of the composer’s generally applied aesthetic theory. The second dimension is ontological, where decisions are established on musical or extra-musical grounds.
Dimensions of the composer’s strategic and ontological decisions with their elements, and examples of intuitive and reflective utterances in a fourfold table.
It is important to note that all of the categories of decision-making serve the purpose of creating and maintaining aesthetic congruence; and they are implemented with either intuitive or reflective processing. This is in contrast to the study of Katz and Gardner (2012), which attributes within-domain processes (imagination, experimentation, improvisation, and sound imaging, which can easily be associated with intuition) to purely musical ideation. On the contrary, beyond-domain processes (intellectual operations with parameters, constraints, and metaphors, which can be identified as reflection) are associated with extra-musical inspiration. In short, Katz and Gardner (2012) imply that musical inspiration mainly suggests the dominance of intuition, and conversely, that the extra-musical impetus predominantly induces reflection. The results of this study do not support the attribution of the processing modes to the impetus of musical or extra-musical information. In contrast, the data shows how each mode of processing may be singled out, whether consciously or unconsciously, according to the task’s situational cognitive demand, i.e., not according to the ontological nature (musical or extra-musical) of the task.
Conclusion
This study explicates compositional thinking as fluctuations of intuitive and reflective ideation, monitored by metacognitive function. As such, it extends the focus of composition research – which has previously emphasized either the intuitive or the intellectual side of compositional cognition – into the area of empirical investigation of the dynamics and functions within intuitive and reflective processes. The study discusses the way in which the informant composer utilized different processing modes to create and maintain aesthetic coherence: first, when forming the germinal ideas into the diverse and abstract meta-representation of the Identity Idea; then, when proceeding to elaborate, refine, and embody the Identity Idea into musical passages and structures through step-by-step, intuitive, or reflective compositional acts.
In previous research on composition, generating “musical form” has usually been seen as a product of predominantly deliberate and rational constructive work without notable intuitive processing. Further, the aesthetic premises often assumed by the post-serial, modernist 7 composers have often been dissociated from extra-musical inspiration and impetus (see Albright, 2004; Scherzinger, 2004). On the contrary, the notion of the strategic and ontological dimensions of the composer’s compositional decisions, as identified in this study, does not support the idea of a specific source of impetus (inspired either by the musical materials themselves or by conceptual frameworks outside the discipline of music) which would imply a particular processing mode; nor does a certain processing mode suggest a particular source of impetus (cf. Katz & Gardner, 2012).
Finally, a detailed analysis has been presented of the dynamics of problem accumulation, i.e., the way the composer decided not to decide, which originated from the composer’s ill-defined working circumstances and his ethos of seeking coherence and employing constructionist aesthetics. However, the role and ramifications of the consequential phenomenon of this kind of artistic endeavour remain unclear, calling for further investigation.
As a thick description of probably the most complicated phenomenon of musical behaviour, the study provides qualitative real-world material for scientists who theorize about human cognitive architecture and the roles of thinking, reasoning, and creating within it (see Levitin & Tirovolas, 2009, p. 211). The study may also be useful for young composers interested in expanding their compositional devices and for those who educate professional composers.
Footnotes
Appendix: Definitions of the compositional acts
Acknowledgements
The author wishes to thank the composer for his involvement in this research, and for permission to release his manuscripts; and also Professor Anne Sivuoja-Kauppala and Lauri Suurpää and Adjunct Professor Auli Toom for their valuable comments on the article.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Supplemental material
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
