1. Dancing to Architecture?

Framing all the great music out there only drags down its immediacy. […] Writing about music is like dancing about architecture—it’s a really stupid thing to want to do.

Elvis Costello (b. 1954), singer-songwriter1

Starting this book with such a quote is not just a flippant rhetorical device. It flags my very strongly felt unease regarding the subject matter of the undertaking and my research in general. It is not that I agree with music theorist Heinrich Schenker (1868-1935), who famously began his thesis The Art of Performance by stating that “a composition does not require a performance in order to exist. […] The reading of the score is sufficient,”2 thus similarly negating the importance of his topic. No, I believe all perception of music is performative whether it is reading a score, hearing with one’s inner ears, imagining, playing and singing, or listening to someone else performing. I even contend that when we say “music,” when Elvis Costello speaks of “great music,” we think of performances, sounds that “live” in our bodies, in our memories. We do not think of inscriptions on pages (scores) even if we are speaking of western art music with its long tradition of notated, authored compositions. And this is exactly why it is so difficult—if not stupid—to talk or write about it. When we do, we are trying to express in words that is in effect, a subjective, physical-affective experience. So a better question might be, “Why is it that we cannot readily recover for our ordinary speech what is so tantalizingly offered by practice?”3 If this question seems sensible then we have identified the reason why we need a musicology of performance.

Musicology—a discipline invented in nineteenth-century Austria and Germany along the analogies of philology, historical and literary studies—has traditionally been concerned with the written text of music. To a great extent it still is. However, over the past few decades there has been an exponential growth in scholarship that focuses on music performance. The contention of this book is that we need a better theoretical framework for such studies; a framework that enables engagement with this richly complex phenomenon so that “talking about music” may be regarded less like “dancing to architecture,” less of “a stupid thing to want to do.”

The theoretical framework and analytical approaches I propose in this book are for studying musical performance. They do not shift the thinking about music to the different paradigm advocated by Nicholas Cook: music as performance.4 I am interested in a musicology that might assist us to deconstruct the complex that music performance entails: the act and its perception, the aesthetic and the technical, the cultural and the historical, the personal and the common. Therefore I propose a model that engages not only with the various elements and aspects but, importantly, with the interactions of these. I argue that music performance shows overwhelming similarities to the characteristics of complex dynamical systems. We may gain a better understanding of its layers and the functioning of its contributing elements if we approach it with an adequately complex method. I will demonstrate this by studying forty recordings of Johann Sebastian Bach’s Six Sonatas and Partitas for Solo Violin (BWV1001-BWV1006, dated 1720) made during the past thirty years or so. In this introductory chapter I will first outline the problems we face when studying classical music performance as well as some of the specific questions and debates that relate to playing music composed almost 300 years ago. In the second part of the chapter I will introduce my material and outline how I proceed in the rest of the book.

1.1. Problems in Researching and Writing about Music Performance

Music perception is multi-modal, whence it lays the crux of our problem. The role of visual, spatial and kinaesthetic inputs and sensations has been repeatedly demonstrated. What we hear depends on the context, on what we see, on our disposition and health, and our prior experiences and knowledge; on our mental and muscle memory, and the function of mirror neurons, not to mention “species-saving” mechanisms of evolutionary significance.5 Edmund Husserl (1859-1938) considered perception an extraordinarily creative mental act while Maurice Merleau-Ponty (1908-1961) called for a more “holistic theory incorporating embodiment and action.”6 The music theorists David Lewin argued that “musical perception was a type of skill, built over time, which can manifest in an infinite number of creative responses.”7 He claimed that it was erroneous and leading to false dichotomies to “suppose that we are discussing one phenomenon at one location of phenomenological space-time when in fact we are discussing many phenomena at many distinct such locations.”8

It is indeed difficult to accept the view that our perceptual experiences may be understood on cognitive terms alone. However, it is equally difficult to be convinced by models that discount the role of cognitive processes and endeavour to explain everything in neurological or evolutionary terms. As Merleau-Ponty noted, “the distinction between subject and object is blurred in my body […].”9 Or, as Günther Stern [Anders] (1929-1930) opined, “When listening to music we are out of the world and in music.”10 This sounds rather romantic, does it not?—even though it goes beyond the romantic notion of music communicating the inexpressible. Anders’ formulation, as I see it, implies that listening to music is inwardly in direction. It is internalized (unlike sight that canvasses the outside world); it has no objective subject, and in this regard John Lennon’s quip might ring quite true as it evokes another very internalized and personal experience: “Listen, writing about music is like talking about fucking. Who wants to talk about it? But you know, maybe some people do want to talk about it.”11

The “talking about” music performance has long been the domain of music critics on the one hand, and music psychologist on the other hand, both with characteristics that duly raises the question of “why bother.” The former deals in metaphorical description and tends to reflect normative thinking. Given the space limits of most magazines and dailies, reviews tend to be very short, having no room for detailed observation or definition of subjective terms. More recently they often include phrases to the effect, “you really need to hear the disk, it is not possible to describe it properly.” The latter, that is music psychology, is driven by empiricism and laboratory testing. As such it is limited to what it can measure and test. Although recent developments in technology make their investigations increasingly sophisticated and influenced by neuro-science, the tendency to look for universals through what is measurable yields results of moderate interest to musicians and lovers of music. These studies are not really concerned with either the phenomenological experience or the aesthetic-affective impact of technical and stylistic differences. And when they do, the results do not necessarily provide particularly penetrating new insights. Instead, they simply confirm what practicing musicians have known for long through practice.12 For example, they have rightly noted the importance of visual cues in music performance and drew attention to it when most investigations focused on audio-only formats—both in terms of performer-to performer communication and, more importantly, perhaps, for listeners’ enjoyment and aesthetic judgement.13 By now, however, psychological research into the multi-modal perception of music performance seems to be focusing on anything but the aural sensation and perception, as witnessed at a highly successful conference on the topic held at the University of Sheffield in March 2015. Yet musical gestures—aurally perceived gestures as well as gestures made visible to the mind through aural stimulus—are crucial in the affective communication between performer and listener, between “the music” and the perceiver.

I mention gesture because it is a “hot topic” nowadays. Three edited books have been published on music and gesture in the past ten years and they provide important propositions that would benefit from systematic psychological, cognitive or neurological investigations.14 Several of the contributing writers address the multi-modal and experiential perception of music. Rolf Godøy, for instance notes that “music perception is embodied in the sense that it is closely linked with bodily experience” and it “is multi-modal in the sense that we perceive music with the help of both visual/kinematic images and effort/dynamics sensations, in addition to the ‘pure’ sound.”15 Tapping into the mentioned debate between the Husserlian emphasis on “cognitive processes” and the Merleau-Pontian emphasis on the embodied sensation, that “thought and sensation as such occur only against a background of perceptual activity that we always already understand in bodily terms,”16 Godøy states that “ecological knowledge in listening, [means] knowledge acquired through massive experience of sound-sources in general and musical performances in particular.” He adds: “[…] the main point is […] not so much the kinematics (the gesture trajectory shapes that we see) as it is the dynamics of movement (the sensation of effort that we feel through our embodied capacity for mental simulation of the action of others).”17

In a similar vein, Lawrence Zbikowski proposes the possibility of “no correlation between the gesture and the sound that produced it” and argues that “gesture [and music] give access to a dynamics, imagistic mode of thought that is inaccessible to language.”18 As such it may function like a metaphor by reflecting “a conceptual mapping” of knowledge in one domain to the experience in another domain. The multi-modal perception of music and the subjectivity of meaning are underlined by such assertions, and highlight the difficulties commentators on music performance face. Yet our fascination is such that we do not give up easily. The dance to architecture goes on.

Focusing on actual rather than metaphorical gestures but similarly building on James Gibson’s ecological theory of hearing and listening,19 Luke Windsor argues: “Gestures are actions that musicians make, and the supreme virtue of music in this respect is that it can make audible gestures that are near invisible.”20 I propose that deciphering these gestures that are made audible by music is a key to a better understanding of aural modes of communication, of our capacity for “imagistic thought” that are visible only to our minds but triggered by sound. We want to talk about music because we are fascinated by our experience and want to understand why and how these strongly felt reactions come about.

A comprehensive approach to the study of music performance is therefore important. It paves the road towards such insights. How, in what manner does sound specify the actions of performers and how do these aural cues give meaning to the musical experience? My analyses will aim to answer these questions step by step. First looking at the separate performance elements and then contemplating their contribution to the overall effect. Given our multi-modal, cognitive as well as embodied and affective perception of music, analysis must attempt to consider cues not in isolation but in their complex, non-linear and dynamic interactions with each-other as well as with both the performer’s and the listener’s historical-cultural disposition. This complexity hints at the biggest problem experimental investigations of music performance face: for such an understanding “we must imagine a hearing that involves no lapse of attention and a perfect operation of all the faculties—sensation, memory, understanding, imagination, and so on. In an actual, empirical hearing it is difficult to imagine such perfection.”21 In contrast, in an analytical, contemplative framework that relies on rich, empirically derived data and a transdisciplinary approach the chances seem better to achieve this goal. It is this kind of musicology of performance that I propose.

Problems with Historical Investigations of Music Performance

It is not only the multi-modal nature of our engagement with music that causes problems in investigations of musical performance. There are also the historical, cultural and aesthetic dimensions to be accounted for. How and why does a performance evolve? How and why do styles of interpretation change? How do performers interact with the past and the present; how do they influence each-other? Are stylistic practices developing through communities sharing aesthetic sensibilities, geographical location, cultural or educational history? What is the current scene of performing the music of Bach like compared to earlier times? Are there trends and if so, who are the trend setters? The availability of more than one hundred years of recordings makes such investigations possible. In fact the study of sound recordings as evidence of changing performing styles has been a growing field of musicological investigation since the mid-1990s.

The explosion of digital reissues of old recordings at the end of the twentieth century suddenly put the history of music performance on centre stage as hundreds of items from early catalogues have had again become easily accessible. These provided fascinating and undeniable evidence for considerable changes in the interpretations of canonical compositions within the European concert tradition. Normative thinking regarding how Beethoven, or Bach, or any other composer’s music “should go” was challenged, eventually leading to Nicholas Cook’s call for a re-evaluation of the framing of musicological investigations to be not music as text (scores, compositions), but music as performance. Or how I prefer to think about it: music as sound.

Anything that develops so rapidly brings with it the danger of “running ahead of itself.” Understandably, historians have focused on the earlier recordings and on the playing of musicians of bygone eras. Many of them (e.g. Jascha Heifetz, Fritz Kreisler, Beniamino Gigli, Tito Gobbi, Adelina Patti, Pablo Casals, Ignacy Jan Paderewsky, Alfred Cortot, the Lehner Quartet, Willam Mengelberg, Bruno Walter, to name a few) are legendary and becoming familiar with their artistry is not only informative regarding earlier styles of playing or singing but also satisfies curiosity. But a narrative of the history of twentieth-century performance styles that is based primarily on studying earlier musicians while relying on impressionistic information regarding more recent or living artists, leads to false conclusions. In this book I aim to supplement this near singular focus of detailed research on early recordings and pre-war artists by similarly engaged, systematic work on current musicians. This is important to do for an accurate picture to emerge and to prevent potentially unwarranted conclusions regarding a “golden age” (that is, pre-1930s) from taking hold.22

A lack of sufficient balance in scholarly attention may also foster premature notions about the roles various stakeholders and cultural-historical-social forces play in the development of performing styles and interpretative approaches. I will explore these at length in the next chapter. So here I only introduce some of the key issues that are problematic and need further investigation.

A commonly expressed view is that the recording industry has fostered a de-personalisation of musical expression through its demand for technical perfection and repeatability, that performances have become much less individual than they used to be during the proverbial “golden age” prior to and at the beginning of sound recordings.23 Theorists of music performance in the second half of the twentieth century also seem to be of the opinion that performances today are less communicative because they are less detailed.24

The individual liberties observed in early recordings, including piano rolls, are of course stunning for a modern listener. Performances have an air of spontaneity because they do not seem to strive for uniformity of tone or phrasing, steadiness of tempo, or ensemble and (in case of pianists) hands co-ordination, and because they freely arrest the musical movement to highlight a melodic pitch, or glide and “scoop” to notes of affective significance. Importantly, the flexibilities heard in these recordings are of a “fluid” nature; they sound very engaged because they seemingly follow unhinged the texture and assumed dramatic impetus of the music. However, closer study reveals that these performance characteristics were quite common during the turn of the nineteenth and twentieth centuries; they represent the general trend, the convention.25 They seem individual and idiosyncratic to us only because our current conventions are very different. Systematic and comprehensive analyses are important—whether dealing with old or recent performances—if we wish to avoid the problems of premature generalizations and misrepresentation of historical-cultural developments. Performances are constantly changing and evolving. The impressions about conventions in the 1980s and 1990s reported in narratives of the recorded history of performing western classical music may not hold water when considering more recent performances, or specific repertoires.

Nevertheless, broad and unsystematic listening to classical music seems to confirm, at a basic level, that performance has been “tidied up.”26 There is ample evidence of current players’ dazzling technical proficiency, and this precision of intonation and ensemble, steadiness of tempo, accuracy of rhythm, bow and vibrato control, dexterity, virtuosity and so on have become minimum standards. The often less polished and more wayward playing on early recordings, on the other hand, may make such performances sound more personal and musically engaged because of their seeming vulnerability and the ad hoc, “in the moment” solutions musicians are able to pull off.27 But what are the assumed characteristics of the modern style?

Since in this book I study performances of Bach’s Solos for violin, it seems appropriate to report how the “modern” style is described in literature specific to violin playing. Eric Wen refers to a “codification of violin technique,” namely a focus on the left hand and “overemphasis on vibrato at the expense of shadings in the bow.” This, Wen claims, “led to an increase in digital facility [and] set a standard in sound production which lacked variety.”28 Jaap Schröder explains the situation by highlighting that the modern legato stroke, which is based on the whole bow being under a certain pressure, “tend[s] to produce a straight and uninteresting tone that must be made more attractive through vibrato.”29 Mark Katz agrees but also posits a direct link between the increased prominence of vibrato and the advent of sound recording. In his view, during the 1910s-1920s vibrato may have been considered an aid in adding personality in the absence of visual contact and also helped project the tone into the recording equipment—while also hiding imprecision in intonation.30 In short, a more homogeneous tone has developed aided by standardized and continuous vibrato and seamlessly even up-and-down bow strokes. The aesthetic preference for unity and smoothness has also fostered fingering that enables the use of single strings for the entire length of phrases. As each string of the violin has a slightly different timbre and because one cannot use vibrato on open strings, violinists have tended to shift hand position to avoid open strings and create a unified tone by remaining on the same string for a given melodic passage. Later I will explain the differences between this way of playing and what historical sources tell us about early eighteenth-century practice. However, it remains to be seen if currently recording violinists subscribe to this modern aesthetic when performing J. S. Bach’s Solos.

As I have already mentioned earlier, in spite of these observations, their general arguments regarding the current state of affairs are not necessarily entirely accurate. Although the grand narrative of increasingly less “interesting,” more homogeneous performance practice dominates public and scholarly opinion, these are based on broad impressions, not systematic examinations. Yet by now several data-rich studies exist that demand a more refined account of how performance styles are developing and what characteristics are typical today (cf. fn. 25). On the one hand, they point to the fact that trends and fashions existed in earlier times also, while on the other hand, they reveal considerable variety among recent versions of given repertoires, thus throwing the theory of growing homogeneity into doubt.31 Furthermore, it is recognized that every generation expresses nostalgia towards a “golden age” which invariably seems to refer to the period during which that generation gained its formative musical experiences.32 Whether idealizing the performances of certain teachers and older artists or gradually canonizing particular aspects of their playing or musical approach, when a performance style becomes a convention it also becomes normative, fostering a level of uniformity: one orthodoxy gradually replaces another.33 The problems with currently available historical studies of music performance are becoming increasingly transparent.

Is HIP a Modern Invention/Aesthetic or Does it Have Historical Grounding?

Another problem, perhaps more pertinent for the case study in this book, is the reception and critical theory of what we call historically informed performance, or HIP. Without wanting to pre-empt the discussion of this complex area in chapter two I should explain here some of the stakes involved.

At the beginning of sound recordings the most often performed repertoire was from the nineteenth century and many works from periods prior to Mozart were hardly known at all. The revival of early music became an important preoccupation of musicians throughout the twentieth century but in particular during the post WWII decades of the 1950s to the 1980s. The history and aesthetics of the “early music” or “authenticity movement,” as it was originally called, have been mapped from a range of viewpoints. Harry Haskell wrote primarily about its history; I investigated its stylistic development; Richard Taruskin and John Butt theorized its ideologies and cultural connections; while Bruce Haynes focused on articulating stylistic and aesthetic differences among various twentieth-century approaches to repertoire, to name but only a few larger studies.34

During the course of the movement not only forgotten repertoire but also old instruments and past performance conventions have been revived. Or so people specializing in HIP claim. They assert that by studying period instrumental treatises, by playing on historical instruments or copies thereof and by implementing ornamentation, articulation, bowing, tonguing and fingering techniques recommended in historical sources, they are able to recreate performing conventions that were typical at the time of composition. Critics of the movement argue, on the other hand, that HIP is not much else than a selective cherry-picking of elements our modern sensibilities like; their sum total has little historical credence, in fact they are essentially similar to the mainstream performance (MSP) style.

Although currently less discussed, the extent to which HIP is an invented, modern aesthetic tradition is still a matter for debate. As I will show in the next chapter, and throughout the book, a systematic and comprehensive analysis of recordings and repertoire does not support a generalized view of HIP. The analogies between the styles of MSP and HIP are not water tight when it comes to performances of Bach’s Solos for violin. “It depends” – as a lawyer would say and as I will show in chapters three to five. More importantly, performers have moved on, and believing that what might have been true of the 1970s and 1980s still holds today is naive. One constantly has to re-evaluate findings about “current practice,” because practice never stands still. When the theoretical framework and analytical approaches proposed in this book are implemented a more nuanced picture develops that allows for the complexities to shine through and a fairer account of what performers do to be formulated.

Data versus Narrative–Letting Go of Dancing or Returning to the Dance Floor?

Having outlined problems encountered in various approaches to investigating music performance it is time to address the dilemma the “systematic and comprehensive” analyst faces: How to “talk” (write) about recordings meaningfully and interestingly? When interpretations are played one after the other, the differences between them seem all too obvious. Yet, when enlisting scientific apparatus to account for these differences, the resulting descriptive words, numerical tables and graphs often seem opaque and thus hopelessly inadequate; really, like overkill. As Taruskin put it, “data can’t solve aesthetic issues. To think it can is utopian, and all utopian thinking ends up being authoritarian.”35 I might retort that without data all we have is opinion. But what knowledge do we gain? How much and what kind of detail do we need? On what grounds and through what kinds of expression can we verbalise sensed aural and temporal experiences? What do we learn that we did not know already, even if only intuitively or perceptually? This, of course, may be asked of many academic endeavours, especially in the arts. In this book I argue that we need much more detailed empirical data before an evidence-based narrative of the historical-cultural evolution of performance styles can be written because the devil is indeed in the detail. We need not only more data but we need to approach our subject, music performance, differently because it is like a complex system and complex systems are not “constituted merely by the sum of [their] components but also by the intricate relationships between these components. In ‘cutting up’ a system, the analytical method destroys what it seeks to understand.”36

I argue that individual performances are so varied and unique that making generalizations on the basis of a select few samples is simply not adequate from a scholarly point of view. It is impossible to know whether an observed element is typical until we have widespread evidence for it. Furthermore, personal traits cannot be distinguished from conventions until a representative sample has been thoroughly investigated. As José Bowen noted, “It is altogether too easy to mistake a performance characteristic for a unique interpretive feature when it is, in fact, a general style trait of an unfamiliar style.”37 For instance, a-synchrony between parts (e.g. melody versus accompaniment) could be regarded as a unique feature of, say, Paderewsky’s playing, until one realises that most pianists trained during the nineteenth century performed that way. The seemingly idiosyncratic style turns out to be a historical convention.

Groups of performers belonging to the same generation or “school” of performing practice and playing within broadly similar stylistic conventions nevertheless exhibit personal characteristics otherwise they would not likely become soloists with established careers. To account for these individual artistic signatures one has to study differences in degree, not just in kind. The problem is that once such evidence is carefully amassed and examined with a fine-toothed comb, the richness and variety of observations are staggering. The resulting picture tends to be so complex that drawing neat conclusions appears to be farcical, if not outright untenable, for anybody intimately familiar with the data. The modernist scientific project of categorizing, of putting products in the either/or baskets quickly fails. It is all too clear that for any particular finding there could be several counter examples. Take for instance Richard Tognetti’s recording of Bach’s Six Sonatas and Partitas for Solo Violin. If one listens to the B minor Partita’s Sarabande movement, one may come to the conclusion that Tognetti adds ornaments and embellishments where appropriate, at least in slow movements. However, the only other movements in which a few additional embellishments are heard are the E Major Gavotte and the Andante of the A minor Sonata.

To preview some of the results of my analyses of tempo in chapter four, here is a more general example of how varied some findings can be. In conversations people often claim that HIP performers play faster than MSP instrumentalists. However, my systematic study of Bach recordings reveals three important characteristics that correct this generalization. Firstly, it is not so much the tempo that is faster in HIP versions of slow movements, but the perception of the tempo. The overall duration is frequently similar but because of the different approach to articulation and rhythm, the HIP versions tend to move along at an apparently brisker pace. Secondly, HIP players tend to play slow movements faster while MSP players often have a more virtuosic speed in fast movements, especially in finales. The exceptions among HIP violinists include Rachel Podger, whose tempos are usually fast all round, and Monica Huggett, whose tempos are among the slowest regardless of movement type. Thirdly, and as a consequence of the second observation, HIP performers tend to use less extreme tempos than MSP performers (i.e. the differences between the speeds of fast and slow movements is less pronounced). It is clear then, that generalizing on the basis of select examples (e.g. an impression gained from listening to fast or slow movements) can lead to incorrect claims. But of course preparation of comprehensive data is time consuming and the results are likely to be valid only for the particular repertoire examined. Moving along at a snail pace is not conducive for an academic career and many people rather read provocative assertions written in an engaging and persuasive style then hair-splittingly detailed accounts with no neat or debatable conclusions, however accurate they might be. After all, most of us prefer to dance (especially when it comes to music) than to labour in a lab.

Finally, we circle back to the issue of subjectivity of perception and the problems with quantification. Aural analysis is often more instructive in identifying nuanced but clearly audible individual differences that may “disappear” once measured data are averaged, percentages calculated and other means of numerical-statistical or graphic representations are enlisted to show “scientifically” potential trends and regularities. Surely, one has to decide if the aim is to find generalizable trends or to account for individual differences. Yet in either case quantification or visualization are essential because without them the aurally based observations may be tainted by the researcher-listener’s own biases, musical attention and memory. To recall Hasty cited earlier, in empirical investigations of listening it is difficult to maintain the “perfect operation of all the participating faculties—attention, sensation, memory, understanding, imagination, and so on.” What one can hope for is to constantly move back and forth (or in and out) of all these faculties and engage with them one by one while also paying attention to their interactions and combined effect. My proposed approach entails exactly such a conduct and my analytical comments throughout chapters three to five are peppered with cross references looking forward and back across the various observations made according to a particular “faculty” and questioning or confirming previous or later conclusions.

Overall the aim of my analyses is to emphasize the importance of individual differences and to advocate for a more nuanced, more circumspect picture of “general trends”; to be less lofty and more honest about “how things are.” I posit that for a while—until we have accumulated significant quantities of systematically gathered, specific information—the main goal of historical studies of performance and sound recordings needs to be the provision rather than the explication of data. And the data must be generated by a combination of quantitative and qualitative-descriptive analyses and be supported by ample Audio examples.38 If we are serious about putting the performer centre stage and claiming for her or him a pivotal role in the identity and reception of western classical compositions (or “pieces,” to make their identity more fluid than the words “work” or “composition” may imply),39 we must undertake the painstaking data gathering and analytical tasks that were typically applied to written texts / scores until the 1980s. We need to deconstruct and name the specifics constituting players’ artistry, because style history is the history of choices performers make. If we wish to understand how these choices fit with or reflect broader cultural movements we first need to know their exact nature as far as it is possible to know, and how they vary from one generation to the next.

1.2. Summary: Recordings, Aims and Methods

Being a historical musicologist, this book is primarily a historical-analytical investigation of music performance. Given the diverse approaches noted throughout the introduction and problems encountered in all of them, my contention is that we need a different approach and a different theoretical framework. I propose to consider music performance as a complex dynamical system and to use comprehensive analytical methods that are commensurate with the object of study. My aim is to model such an approach through the case study of Bach recordings and meanwhile demonstrate the complex, non-linear and dynamic nature of music performance.

As I have noted, there seems to be a lack of systematic work on contemporary styles of music performance; the depth and detail of discussion rarely going beyond that of a record review.40 Therefore I intend to start correcting the balance by putting the contemporary performer in the spotlight. Because the earliest recordings featured singers and violinists, much has been written about legendary violinists of the early twentieth century.41 It seems appropriate then to also focus on violinists, but (mostly) living ones. As for repertoire, J. S. Bach’s Six Sonatas and Partitas for Solo Violin (BWV1001-BWV1006) are ideal because of their long and varied performance history that enables engagement with HIP / MSP debate.42 Importantly, they are unaccompanied pieces, making the examination of performance choices less complicated.

The forty recordings selected for study were all made between 1980 and 2010 except for Sergiu Luca’s. His was issued in 1977 and is included here because it was the first to use a baroque bow and to attempt a historically informed interpretation of the Solos.43 I do not at all claim that my selection is fully inclusive of all recordings made during the designated period of 1980 to 2010, but I certainly believe they form a substantial collection and a representative cross section of such recordings (Table 1.1).44 I also believe that recordings can be studied as if they were performances because the listener experiences them as such, regardless of how they might have come about.

In a study like this, it is not enough to note the date of a recording and its release on the market. It is also important to be aware of the age of the performers. It has been shown that performers develop their personal style early on and then tend to stick with it by and large throughout their careers.45 The majority of violinists studied here were born in or after 1945, except for Oscar Shumsky (b.1917), Ruggiero Ricci (b.1918), Jaap Schröder (b.1925), Gérard Poulet (b.1938), Rudolf Gähler (b.1941), Sergiu Luca (b.1943), Ugo Ughi (b.1944) and Sigiswald Kuijken (b.1944). The younger players, born since 1970, are represented by Lara St John (b.1971), Isabelle Faust (b. 1973), Rachel Barton Pine (b.1974), James Ehnes (b.1976), Hilary Hahn (b.1979), Ilya Gringolts (b.1982), Julia Fischer (b. 1983), Sergey Khachatryan (b.1985) and Alina Ibragimova (b.1985). Table 1.1 and the Discography provide the complete list and full details.

Table 1.1. Studied Recordings Listed by Chronology of Violinists’ Year of Birth

In this selection approximately twenty-five46 recordings represent non-specialist or “mainstream” violinists, with multiple recordings by Gidon Kremer, Viktoria Mullova, Christian Tetzlaff and Rachel Barton Pine. There are eleven violinists representing the historically informed performance movement, with two recordings by Sigiswald Kuijken. Initially (and in this summary) I categorized violinists who play on modern violins and perform the broad gamut of the violin repertoire as representing “mainstream performance” (MSP), while those playing with eighteenth-century violins and bows (whether originals, copies, or reconstructions) and specializing in performing largely pre-1800 repertoire are regarded as “period” (HIP) violinists. However, it has been pointed out that the distinction between HIP and MSP is increasingly difficult to make as time passes. For instance, Colin Lawson and Robin Stowell noted the application of “period principles […] to mainstream situations” whereas in 2006 Eitan Ornoy demonstrated “clear similarities” between the styles of playing.47 This trend began when Nikolaus Harnoncourt and other HIP specialists started conducting large symphonic orchestras in the early 1980s, demonstrating that the HIP style was transferable to modern instruments. However, as noted in the previous section, there are also commentators who believe HIP was never really different from MSP. This book explores where things are now in this regard and one of its main findings is the revision of this initial categorization.

As a practical note, it needs to be said, that throughout the book I generally refer to Bach’s Sei Solo a Violino senza Basso accompagnato as the Solos and I abbreviate the identification of movements by stating the key of the sonata or partita of which they are a part, followed by the movement title. This ignores the fact that certain movements are in a different key from the work’s main tonality (e.g. the Andante of the A minor Sonata is actually in C Major) but should cause no confusion as to the identity of the excerpt. The spelling of the movement titles follows Bach’s own indication in the autograph manuscript.

To foreshadow some elements of this analysis, it can be noted that many of the younger players in the sample claim to have been influenced by HIP ideology (e.g. Zehetmair, Tetzlaff) and quite a few of them undertook periods of specialist training (e.g. Ibragimova) or regularly play with period ensembles (e.g. Barton Pine). According to the liner notes for their respective recordings, three MSP violinists use a baroque bow and / or gut strings: Rachel Barton Pine, Richard Tognetti, and Viktoria Mullova in 2008. On the other hand, Rudolf Gähler plays with the so-called curved-bow, also referred to as “Bach-bow” or “Vega-bow.” Although the arguments for using such a bow bespeak of a search for period practices, the playing style enabled by such bows is closer to the aesthetic principles upheld by the MSP tradition. The historical existence of such a bow has been discredited by researchers since the 1960s.48

This book is not the place to discuss at length what constitutes the performance characteristics of MSP or HIP in principle; there is plenty of literature on that as mentioned earlier. I will refer to them in detail later on, particularly in chapters two and four, as they relate to performance features I analyse. Here it should suffice to recall that, as we have seen in earlier comments on twentieth-century violin playing, in MSP the aesthetic ideal is essentially a purity and evenness of tone (achieved through vibrato and seamless bowing) in service of power and projection of the melody. Inflections are kept to the minimum to foster long-spun phrasing shaped through graded dynamics. In contrast, HIP takes its cue from the bass line and meter of a piece and inflects melodic-rhythmic-harmonic groups accordingly. The use of shorter, uneven bow strokes, bouncing rhythm and locally nuanced articulation of smaller musical units are among the main characteristics of such interpretations. These and other differences will of course be explored and expanded upon since they are key matters to consider when teasing out the qualitative differences between performances. Identifying them and explaining their characteristics and interactions provide the basis for a more textured description of performances and a more nuanced “categorization” of violinists.

Importantly, although versions on period instruments are contrasted with mainstream ones, my goal is not to assess historical verisimilitude or whether an interpretation reflects Bach’s possible intentions. The futility of such exercises has been well argued by others many times over, as discussed in the next chapter. The aim is neither a search for the ultimate “authentic,” “definitive,” or “normative” performance nor the matching of interpretations to particular editions, analytical observations, or performance suggestions, even though certain specific differences between HIP literature and execution are noted.49 My goal is, rather, to conduct a thorough examination of these recordings as a model for analysing performances as complex dynamic systems. Such an approach is likely to unveil the actual level of uniformity versus individuality in playing Bach’s Solos at the turn of the twentieth and twenty-first centuries and thus contribute to a more circumspect narrative of baroque performance styles on record. Throughout I aim to problematize methods to show their limitations and how misleading the results of any one approach might be when used in isolation. Yet when used cumulatively, the various analytical approaches unpack the complex, dynamic network of interactions at play and allow us to see what performers do to achieve particular effects and how differences in execution lead to different readings of the music. At this point the “talking about music” ceases to be “like dancing to architecture”; it starts tapping into the holistic experience.

The different interactions and transformations such an approach unveils can be elucidated by reference to Deleuzian terminology and thinking. I will introduce these in chapter two as part of my argument for tackling music performance as a complex dynamical system. Once I modelled such an approach through chapters four and five but also chapter three, I return to this crucial matter in the final chapter to make the parallels more explicit.

To achieve my goals, both descriptive and quantitative methods need to be employed based on repeated close listening (aural analyses) and software assisted measurements of audio signals. When working with variants of a musical “text” the human brain is often more effective than current computer programs in discerning and sorting individual resemblances or nuances. The use of software is more advantageous in identifying basic or categorical differences and trends across the larger body of data. Moreover, since music performance is ephemeral—“evanescence is its essence” even if recorded—what is available to the listener-analyst is “only what is present in the moment, what is remembered of the past, and expectations for the future.”50 The data a computer might provide may have very little ecological validity. Contemplating the subjectively experienced sound is therefore essential, both globally as a whole and also in its details through close listening. While I focus more on the measurable in chapter four, in chapter five I aim to engage with the holistic and affective experience. However, both chapters are cross-referenced to highlight discrepancies between points of view and underscore the importance of the comprehensive, non-linear approach.

Violin technique is examined as far as vibrato is concerned while bowing is discussed in relation to articulation and phrasing as well as multiple stops and timbre. Expressive timing (rhythmic flexibility) is studied in detail as this contributes significantly to the performance style and plays a decisive role in aesthetic perception and in generating homogeneity or diversity among performances.51 Nevertheless, rhythmic flexibility and phrasing remain somewhat resistant to quantification and categorization. The discussion of them is therefore necessarily limited to a few detailed, striking examples and generalizations. Other elements, such as tempo choices, ornamenting, the shaping of dotted rhythms, the performance of multiple stops, and the extent of reliance on pulse and / or harmony to project musical character and structure lend themselves better to tabulation, transcription or various other forms of visualization. Together, I hope, they provide ample complementary evidence in support of my main points and arguments.

I would like to stress my conviction that, at the time of writing, the methods of studying musical performance have important limitations. Treading the fine line between “impartial” measurements and “subjective” description raises many problems throughout that I do not claim to be able to solve. But in my account I attempt to retain that semi-conscious moment of perception before I as a critic become “modern”; when I “still possess [a] pristine and unspoiled exoticism” and a sense of wonder about my subject; the moment before one tries to “desecrate” what one believes to be the phenomenon under investigation.52 I am interested as much in the sonic means that violinists use as the expressive effect that the sound creates. I want to avoid the obvious danger in attempting to use science to account for human experience: the danger that what is quantifiable may become equated with the phenomenon one originally set out to study. More often than not I am frustrated by the recognition that the “objectively” measurable in a performance does not seem to fully explain my impression of it and that words fail to grasp the internalized perception.

In the Epilogue of chapter six I contemplate the implications and significance of the crucial difference between holistic-affective perception and abstract analytical dissecting with reference to Ian McGilchrist’s proposition put forward in his 2009 book on brain laterality, The Master and his Emissary.53 The discrepancy between “scientific” data and “sensed experience” underlines the importance of multi-modal approaches; of cross-checking acoustically measured and subjectively perceived data and if in doubt, giving priority to the listening experience. Music performance is a complex aural system rooted in oral cultures, as discussed in chapters two and six. This book is an attempt to explain aspects of it and map interactions in the rather opposite domains of text (language) and sight (the visual) even though I also provide numerous Audio examples, and advocate for the “and / as well as” view instead of the “either / or” dichotomy.

In the next chapter I explore some of these theoretical-aesthetic and methodological concerns in much more detail and present my argument for an approach that embraces the complexity of music performance. Such an approach highlights the interactions among a performance’s varied and layered elements, allowing for a more balanced view that accounts for differences in degree, not just in kind. In the subsequent, central chapters I discuss the recordings that provide the empirical data, the case study, for my argument: first a more general overview (chapter three), then considerably detailed analyses (chapters four and five). At the end of the book (chapter six) I return to cultural issues and propose an alternative view of “how things [may] stand now,” at the beginning of the twenty-first century.